Data Visualization in Enlightenment Literature and Culture 3030549127, 9783030549121

Table of contents :
Acknowledgments
Contents
Notes on Contributors
List of Figures
List of Tables
Chapter 1: Introduction: “Speaking to the Eyes”—Reassessing the Enlightenment in the Digital Age
Eighteenth-Century Precursors to Data Visualization
Data Visualization in Eighteenth-Century Studies: A Brief Survey of the Field
Bibliography
Part I: Digital Enlightenment: Representing Big Data
Chapter 2: In Search of Enlightenment: From Mapping Books to Cultural History
Bibliography
Chapter 3: Examining the Early Modern Canon: The English Short Title Catalogue and Large-Scale Patterns of Cultural Production
Canon-Making Versus Large-Scale Patterns of Cultural Production
Defining the Canon: A Brief History
Defining the Canon: Available Data
Actors
Works
Subject-Topics
Defining the Canon: Methods
The Canon: Works, Time, and Subject-Topics
Religion and Literature
Donaldson v. Becket and the Importance of 1774
People: Authors
People: Printers and Publishers
Gender and the Book Trade
Places
Materiality
Conclusion
Bibliography
Chapter 4: Europe and Its “Others”: Visualizing Lexical Relations Between Western and Non-Western Locations of the Enlightenment in The Eighteenth-Century Collections Online
Methodology
Part One: Continental Drift
Part Two: Europe and Its “Others” in Visible Relation
Conclusion
Bibliography
Chapter 5: Text Mining and Data Visualization: Exploring Cultural Formations and Structural Changes in Fifty Years of Eighteenth-Century Poetry Criticism (1967–2018)
Introduction
Eighteenth-Century Poetry Criticism: A Model
Methodology
K-Means Clustering and Topic Modeling
Author Names over Time and Gender Inequality in the Poetry Criticism Corpus
Conclusion
Bibliography
Part II: Data Visualization and the Eighteenth-Century Corpus: Case Studies
Chapter 6: The Grid and the Visualization of Abstract Information: Three Eighteenth-Century Models
Grid as Visual Theory
Barbeu-Dubourg and the Grid as Time
Quesnay and the Grid as Relation
Linnaeus and the Grid as Organization
Classical Grid as Philosophy
Conclusion
Bibliography
Chapter 7: Exploring Data Visualization: Time, Emotion, and Epistolarity in Frances Brooke’s The History of Emily Montague
Textual Analysis: Voyant
Social Network Analysis and Mapping: Palladio
Interactive Timeline: Tableau
Conclusion: Textual Analysis and Digital Pedagogy
Bibliography
Chapter 8: Outliers, Connectors, and Textual Periphery: John Dennis’s Social Network in The Dunciad in Four Books
Methods and Goals
John Dennis in the Poem’s Plot Network
John Dennis’s Social Network
Conclusion
Bibliography
Chapter 9: Publishing Music by Subscription in Eighteenth-Century Britain: The Concertos of Charles Avison
Publication by Subscription: A Data-Driven Approach
Avison’s Subscription Lists and His Popularity
Individual Subscribers: Categories and Gender
Conclusion
Bibliography
Chapter 10: Afterword: Novel Knowledge, or Cleansing Dirty Data: Toward Open-Source Histories of the Novel
Introduction
Dreaming of Ideal Data
Waking Up
Case Studies: Where We Are Now
Conclusion: Toward Best Practices
Bibliography
Index

Citation preview

Data Visualization in Enlightenment Literature and Culture

Edited by Ileana Baird

Data Visualization in Enlightenment Literature and Culture

Ileana Baird Editor

Data Visualization in Enlightenment Literature and Culture

Editor Ileana Baird Zayed University Abu Dhabi, United Arab Emirates

ISBN 978-3-030-54912-1    ISBN 978-3-030-54913-8 (eBook) https://doi.org/10.1007/978-3-030-54913-8 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2021 Chapters 1, 3, 8 and 10 are licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). For further details see licence information in the chapters. This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the ­publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and ­institutional affiliations. This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Acknowledgments

Any book that involves the collective effort of a team of editors, contributors, and readers is predicated on trust, intellectual respect, and collegiality. Therefore, I am truly indebted, first and foremost, to the readers of our chapters, whose expert knowledge, insightful suggestions, and constructive feedback have had a significant impact on our work. I would like to thank Dr. Brad Pasanek, from University of Virginia, Dr. Michael Gavin, from University of South Carolina, Dr. Collin Jennings, from University of Miami, Dr. Shawn W. Moore, from Florida SouthWestern State College, and Dr. Rafael Alvarado, Program Director at the School of Data Science, University of Virginia. Special thanks are owed as well to the anonymous reader commissioned by Palgrave Macmillan to evaluate our book for the well-informed and enlightening suggestions made. Their expertise and generous support for this project are truly appreciated. My gratitude goes to my institution as well for their continuous support for my work. I would like to thank the Office of Research at Zayed University for a timely Short-Term Grant that allowed me to acquire the resources needed for completing this project and cover all the permission-­ related costs. I would also like to thank the wonderful staff from Zayed University Library, especially Hanin Abueida, whose prompt assistance allowed me to make good progress on my work. Finally, special thanks are due to our Palgrave Macmillan team: Allie Troyanos, our Series Editor, for her genuine interest in our book, Rachel Jacobe, Editorial Assistant, for her expert advice on all permission-related

v

vi 

ACKNOWLEDGMENTS

issues, Brian Halm, Production Editor, for his smooth monitoring of the production of this book, Mani Vipinkumar, Project Manager, for his attentive work on the proofs, and Arun Prasath, Project Coordinator, for his patience and prompt support throughout the process of preparing the manuscript for publication and beyond.

Contents

1 Introduction: “Speaking to the Eyes”—Reassessing the Enlightenment in the Digital Age  1 Ileana Baird Part I Digital Enlightenment: Representing Big Data  29 2 In Search of Enlightenment: From Mapping Books to Cultural History 31 Simon Burrows 3 Examining the Early Modern Canon: The English Short Title Catalogue and Large-Scale Patterns of Cultural Production 63 Mikko Tolonen, Mark J. Hill, Ali Zeeshan Ijaz, Ville Vaara, and Leo Lahti 4 Europe and Its “Others”: Visualizing Lexical Relations Between Western and Non-Western Locations of the Enlightenment in The Eighteenth-Century Collections Online121 John Regan

vii

viii 

Contents

5 Text Mining and Data Visualization: Exploring Cultural Formations and Structural Changes in Fifty Years of Eighteenth-Century Poetry Criticism (1967–2018)153 Billy Hall Part II Data Visualization and the Eighteenth-Century Corpus: Case Studies 197 6 The Grid and the Visualization of Abstract Information: Three Eighteenth-Century Models199 Jakub Zdebik 7 Exploring Data Visualization: Time, Emotion, and Epistolarity in Frances Brooke’s The History of Emily Montague231 Courtney A. Hoffman 8 Outliers, Connectors, and Textual Periphery: John Dennis’s Social Network in The Dunciad in Four Books265 Ileana Baird 9 Publishing Music by Subscription in Eighteenth-Century Britain: The Concertos of Charles Avison309 Simon D. I. Fleming 10 Afterword: Novel Knowledge, or Cleansing Dirty Data: Toward Open-Source Histories of the Novel351 Emily C. Friedman Index371

Notes on Contributors

Ileana  Baird is an Assistant Professor of English at Zayed University, UAE.  Her research areas include eighteenth-century British literature, visual and material culture, Orientalism, and digital humanities. She is the editor of Eighteenth-Century Social Networks: Clubs, Literary Salons, Textual Coteries (2014) and the co-editor of Eighteenth-Century Thing Theory in a Global Context: From Consumerism to Celebrity Culture (2014; 2018), and All Things Arabia: Arabian Identity and Material Culture (2020). As a SHANTI fellow at the University of Virginia (2009–2011), she worked on a social network analysis project focused on Alexander Pope and his dunces. Some of her findings are highlighted in the article, “The Strength of Weak Ties: Eliza Haywood’s Social Network in The Dunciad in Four Books (1743),” recently published in ABO: Interactive Journal for Women in the Arts, 1640–1830. Simon  Burrows is a Professor of History and Digital Humanities at Western Sydney University, Australia, where he is the leader of the Digital Humanities Research Group. He is the principal investigator of the French Book Trade in Enlightenment Europe (FBTEE) database, a project which was awarded the British Society for Eighteenth-Century Studies Digital Resource Prize in 2017. He is the author of French Exile Journalism and European Politics, 1792–1814 (2000), Blackmail, Scandal and Revolution: London’s French Libellistes, 1758–1792 (2006), A King’s Ransom: The Life of Charles Théveneau de Morande, Blackmailer, Scandalmonger, and Master-Spy (2010), The French Book Trade in Enlightenment Europe II:

ix

x 

NOTES ON CONTRIBUTORS

Enlightenment Bestsellers (2018), and co-editor with Glenn Roe of Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies (2020). Simon  D.  I.  Fleming is Head of Music at the Queen Elizabeth Sixth Form College in Darlington and part-time faculty in the Music Department at Durham University, UK. His main area of expertise is music produced in the British provinces during the long eighteenth century. He has produced numerous articles on this topic, including several on the Newcastlebased composer, Charles Avison, as well as articles on William Howgill, John Pixell, James Nares, the musicians of Carlisle Cathedral, the Georgian town waits, the Spalding Gentlemen’s Society, and the music produced in eighteenth-century Stamford. He is currently working on a book on Avison’s music and is engaged in a digital project that involves indexing the subscribers to every music-related publication issued in Britain before 1820. As part of this project, he has written on subscription and gender, and on patterns of subscription at British and Irish cathedrals. Emily  C.  Friedman is an Associate Professor of English at Auburn University, USA, and Director of 18thConnect.org, an aggregation site that peer-reviews and makes more discoverable digital projects in the field of eighteenth-century studies. She is the creator of Manuscript Fiction in the Age of Print, a small-scale digital project that describes, transcribes, and encodes fiction that survives in manuscript from 1750 to 1900. She is at work on the first monograph to emerge from the dataset, Unprinted: A Literary History of Alternative Circulation. She is also the author of Reading Smell in Eighteenth-Century Fiction (2016) and of numerous articles on eighteenth-century literary culture. Billy  Hall is an Assistant Professor of English at Brigham Young University, USA.  His professional interests include eighteenth-century poetry and poetics, Enlightenment studies, philosophy of technology, media studies, and the history of aesthetics. He has published an article on Christopher Smart in 1650–1850: Ideas, Aesthetics, and Inquiries in the Early Modern Era and his reviews have appeared in the Journal of British Studies and Eighteenth-Century Studies. He is the technical editor of the Romantic Circles’ digital edition of William Wordsworth’s Guide to the Lakes and is completing a monograph that takes a quantitative approach to eighteenth-century poetry written by women.

  NOTES ON CONTRIBUTORS 

xi

Mark J. Hill  holds a PhD from the University of Oxford and is a postdoctoral researcher at the University of Helsinki, Finland. He is interested in topics both historical (the emergence of the public sphere in eighteenth-­ century Europe, Jean-Jacques Rousseau’s political thought) and methodological (the intersection between intellectual history and digital humanities). A historian by training, Hill employs quantitative approaches to historical research ranging from quantitative text analysis to statistical approaches and social network analysis. Before Helsinki, he was Fellow in the Government Department at the London School of Economics. Courtney  A.  Hoffman  is Marion L.  Brittain Postdoctoral Fellow and Assistant Director of the Writing and Communication Program at the Georgia Institute of Technology, USA. Her research focuses on eighteenth-­ century correspondence, both fictional and real, particularly between women, and considers representations of affective relationships between letter-writers over time. Her recent essays have appeared in the edited collections The Cinematic Eighteenth Century (2017) and Global Frankenstein (2018). She is working on a book project that explores interactions between temporality and affect in the exchange of letters in eighteenth-­ century epistolary fiction. Hoffman is also collaborating as domain expert on a data visualization research project which explores machine learning and artificial intelligence tools that can allow non-expert users to determine the types of visualizations that will be most helpful for textual analysis. Ali  Zeeshan  Ijaz is a postdoctoral researcher in the Department of Digital Humanities at the University of Helsinki, Finland, and a member of the Helsinki Computational History Group. He holds a PhD in Bioinformatics from Western Sydney University. Ijaz is working on a project involving bibliographical data analysis. Leo Lahti  is a senior researcher at the University of Turku, Finland, and one of the founding members of the Helsinki Computational History Group. His academic background is in the theory, methods, and applications of statistical machine learning and high-throughput data analysis. His research interests focus on complex dynamic systems and computational history. John Regan  is a Lecturer in Literature and the Digital at Royal Holloway, University of London, UK, and a research associate at the Concept Lab, Cambridge Centre for Digital Knowledge. He is the author of Poetry and

xii 

NOTES ON CONTRIBUTORS

the Idea of Progress, 1760–1790 (2018) and the co-editor of Rethinking British Romantic History, 1770–1845 (2015). He is interested in eighteenth-century poetry, aesthetics, and ideas of human development, and is investigating how knowledge structures emerged and changed shape across the eighteenth-century corpus. Mikko  Tolonen is an Assistant Professor of Digital Humanities at the Faculty of Arts, the University of Helsinki, Finland, and Principal Investigator of the Helsinki Computational History Group at Helsinki Centre for Digital Humanities (HELDIG). Between 2015 and 2017, he worked as a Professor of Research on Digital Resources in the National Library of Finland on a project involving newspaper digitization. Since 2015, he has played a central role in building the Digital Humanities infrastructure at the Faculty of Arts at the University of Helsinki. He is the head of the Digital Humanities program at UH and the chair of the Digital Humanities in the Nordic Countries (DHN). His research focuses on the study of public discourse and knowledge production by combining metadata from library catalogs and early modern books, newspapers, and periodicals. In 2016, he was awarded the Open Science and Research Award by the Finnish Ministry of Education. Ville Vaara  is a doctoral student in Digital Humanities and History at the University of Helsinki, Finland. He is a member of the Helsinki Computational History Group, where he works on a project focused on publication networks and the evolution of the book trade in Britain in the long eighteenth century. His research interests involve the history of the book and the use of digital methods in the study of history. Jakub Zdebik  is an Associate Professor of Art History and Theory and Chair of the Visual Arts Department at the University of Ottawa, Canada. He is the author of Deleuze and the Diagram: Aesthetic Threads in Visual Organization (2012) and Deleuze and the Map-Image: Aesthetics, Information, Code and Digital Art (2019). He has written for RACAR, The Semiotic Review, ESC: English Studies in Canada, The Brock Review, and Esse. He has curated exhibitions at the Kennedy Museum of Art, Ohio, Galerie R3, and Gallery Karsh-Masson.

List of Figures

Fig. 1.1

Fig. 1.2

Fig. 1.3

Fig. 2.1 Fig. 2.2 Fig. 2.3 Fig. 2.4 Fig. 2.5 Fig. 2.6 Fig. 2.7

“Chart Representing the Extent, Population & Revenue of the Principal Nations in Europe in 1804.” William Playfair, An Inquiry into the Permanent Cause of the Decline and Fall of Wealthy and Powerful Nations (London, 1805). (Courtesy of the Thomas Fisher Rare Book Library, University of Toronto) “Exports and Imports of Scotland to and from different parts for One Year from Christmas 1780 to Christmas 1781.” William Playfair, The Commercial and Political Atlas (London, 1786) “Linear Chronology, Exhibiting the Revenues, Expenditure, Debt, Price of Stocks & Bread, from 1770 to 1824.” William Playfair, Chronology of Public Events and Remarkable Occurrences within the Last Fifty Years; or from 1774 to 1824 (London, 1824) Core aspects of the FBTEE database structure The STN sales graph (group stacked bar chart, 1770–1794) Visualizing the Parisian taxonomic system and the relative importance of different keywords Sources and gaps: a calendar of the daily volume of transactions drawn from various STN sources (1773–1778) Sales of French illegal bestsellers between 1770 and 1789, with and without foreign wholesale clients included (note how Loudun and Amsterdam disappear) Places of residence of men of letters corresponding with the STN Leading STN authors of the 1780s by volume of works traded

6

7

8 37 38 41 44 45 49 50

xiii

xiv 

List of Figures

Fig. 2.8 Fig. 2.9 Fig. 2.10 Fig. 3.1 Fig. 3.2

Fig. 3.3 Fig. 3.4

Fig. 3.5 Fig. 3.6

Fig. 3.7

Fig. 3.8 Fig. 3.9 Fig. 3.10 Fig. 3.11

Sales destinations for works originally published in German or English and traded by the STN 51 Annualized trade in religious and irreligious works as a proportion of total STN sales 52 The changing pattern of the STN’s trade in illegal pornographic works listed in Darnton’s Corpus of Clandestine Literature, 1769–178755 Total documents, works, and canon items in the ESTC per year (1500–1800) 76 The Full Canon (1500–1800). These canonical works have been sorted by the first publication year. Individual dots indicate the publishing year for the initial publication and all subsequent reprints 82 Most frequently printed works (1500–1800). The point size indicates the number of reprints for each work (rows) during a given decade (columns) 83 Temporal variation in the relative frequency of the most common subject-­topics in the canon (1500–1800). The relative variability in publishing frequency is higher in earlier time periods due to the lower number of total published works 84 The most popular subject-topics for the ten most printed works in each decade from 1500 to 1800 90 Top authors (1500–1800). The point size indicates the number of publications for each author, including reprints (rows), per year (columns). The color indicates publication before (red) and after (blue) death, respectively. The authors have been sorted by their death year 92 Posthumous publication frequency. The percentage of authors in the ESTC with posthumous publications during the first 50 years after death, grouped by decade (1470–1800). This analysis includes the 1544 authors whose lifetime data is available for the investigated period with one or more posthumous publications 93 Timeline of Shakespeare’s publications included in canon. The point size indicates the share of the publisher with most prints of the indicated work (rows) per decade (columns) 94 Titles with missing book trade actor data. This figure charts out the coverage of the book trade actor data, as found in the catalog records (1500–1800) 96 Share of publications by the largest publishers (top 1%) 97 Share of the canon by largest publishers (top 1%) for unique works, as derived from the work-­field dataset 98

  List of Figures 

Fig. 3.12

Fig. 3.13

Fig. 3.14 Fig. 3.15 Fig. 3.16 Fig. 3.17 Fig. 3.18 Fig. 3.19 Fig. 3.20 Fig. 3.21 Fig. 4.1

xv

Publishing and reprint patterns by publisher role in the printing sequence. This figure indicates changes in the reprint publishing patterns by different publishers over time and charts out the “fluidity” of the book trade. Each work had its publishers explored in chronological sequence to find out how often the publications changed hands. New publications (“New work”) and new editions of the same works by the same publishers (“Stable publisher”) were traced. Publications changing hands were traced both in the cases where the previous publisher disappeared from the book trade (“New publisher, old inactive”) and where the publication changed hands, but the previous publisher stayed active (“New publisher, old active”). Cases of the publication returning to the hands of a previous owner are relatively rare (“Return of earlier publisher”) 99 Publisher subject-topic specialization and canon share. This figure illustrates the differences and similarities in the publishing landscapes of the identified subject-topics (individual scatterplots). Each dot represents an individual publisher, the horizontal axis indicates the publisher’s topic specialization, the vertical axis indicates the portion of all publications by a publisher included in the canon, and the size of the dot indicates the publication volume of a publishers in a particular field. The large dot in the center right in “Information, general works” represents the Stationers’ Company100 Works by female authors in the data-driven canon per decade 101 Fraction of publications by place for the top publication places excluding London (1500–1800) 103 Movement of canonical editions from original print location 106 Number of prints per capita (1700–1800) 107 The fraction of canonical editions compared to all editions per city (1700–1800) 108 Dominant book formats for the most frequent subject-topics 110 The distribution of common book formats for selected canon-works over time 111 Estimated paper consumption for different formats over time for books included in the canon 112 Semantic network visualizing top 20 lexical co-associations with “Africa” (red), “Asia” (green), “America” (yellow), and “Europe” (blue), and top 20 co-associations of these bound terms between 1710 and 1720 135

xvi 

List of Figures

Fig. 4.2

Fig. 4.3

Fig. 4.4 Fig. 5.1 Fig. 5.2

Fig. 5.3

Fig. 5.4 Fig. 5.5 Fig. 5.6 Fig. 5.7

Fig. 5.8 Fig. 5.9 Fig. 5.10 Fig. 5.11

Semantic network produced based on top 20 lexical co-associations with “Africa,” “Asia,” “America,” and “Europe,” and top 20 co-associations of these four bound terms between 1750 and 1760 138 Semantic network produced based on the top 20 lexical co-­associations with “Africa,” “Asia,” “America,” and “Europe,” and top 20 co-­associations of these four bound terms between 1790 and 1800 141 Semantic network produced based on the lexis that co-associated with “Africans” between 1790 and 1800 143 Percent of poetry articles published each year in EighteenthCentury Studies (ECS) and The Eighteenth Century: Theory and Interpretation (TEC) (1967–2018) 155 (a–b) Contrasting words in the poetry criticism corpus (PCC) with and without authors’ names. These visualizations indicate how both the authors considered and the language used to discuss poetry have changed from the first twenty-five years to the second twenty-five years. The most frequent and exclusive terms to each period appear at the top and the bottom of the two charts 159 Network graph derived from a 50-topic model of all 1500+ essays. I have highlighted here topics that correlate strongly with traditional literary genres. This figure illustrates how algorithmic transformations like topic modeling identify genre—“poetry,” “novel,” and “theater”—as a structural feature within the model 166 K-means clustering of the genre subset: novel (cluster 1, red), theater (cluster 2, green), and poetry (cluster 3, blue) 168 K-means clustering with the poets’ names included 170 K-means clustering with the poets’ names removed 171 Visual representation of topic distribution between the two journals. Topics close to the center line are those more likely to be shared by both journals, while topics on the margins represent discourse more closely aligned to one journal or another175 Output from LDAvis for a twelve-topic model of PCC 177 (a–d) Topics over time in the PCC corpus (1967–2018) 181 Topics 4 (sexuality) and 8 (gender) (1967–2018) 184 Graph charting the relative frequency of a poet’s name for each year, as it appears in the PCC. The green line is the average of the relative frequency of female poets and the orange line the relative frequency of male poets 185

  List of Figures 

Fig. 5.12

Fig. 5.13 Fig. 6.1 Fig. 6.2 Fig. 6.3 Fig. 6.4 Fig. 6.5 Fig. 6.6 Fig. 6.7 Fig. 6.8 Fig. 6.9

Fig. 6.10 Fig. 7.1

Fig. 7.2 Fig. 7.3

xvii

(a–b) Anthologized male and female authors, 1967–2018. The top graph highlights Anna Seward (red), Lady Mary Wortley Montagu (blue), Ann Yearsley (orange), and Anne Finch (green); all of them appear at least once in the two journals prior to 1990. The bottom highlights the male poets prevalent prior to 1990: Alexander Pope (green), John Dryden (red), Jonathan Swift (purple), and Samuel Johnson (blue)188 (a–b) Scatterplot of the relative frequency of a poet’s name for each year (1967–2018) 190 Jacques Barbeu-Dubourg’s Machine chronologique, 1753. (Courtesy of Princeton University Library) 208 Jacques Barbeu-Dubourg’s Machine chronologique (detail), 1753. (Courtesy of Princeton University Library) 210 François Quesnay, Tableau économique, 1766. (Courtesy of Bibliothèque Nationale de France) 212 Georg Dionysius Ehret, Methodus, 1736. (Courtesy of The Linnean Society of London) 215 Carl Linnaeus, Regnum Animale, in Systema Naturae, 1735. (Courtesy of The Linnean Society of London) 217 Carl Linnaeus, Regnum Lapideum, 1768. (Courtesy of The Linnean Society of London) 220 Linnaeus, Regnum Lapideum [folded mineral shapes], 1768. (Courtesy of The Linnean Society of London) 220 Francine Savard, Tableau chronologique, 2008–2009. Inkjet printing on rag paper. (Courtesy of the artist) 224 Hans Haacke, Detail of Shapolsky et al. Manhattan Real Estate Holdings, a Real-Time Social System, as of May 1, 1971 © Hans Haacke / Artists Rights Society (ARS), New York. (Courtesy of the artist and Paula Cooper Gallery, New York) 225 Luce Meunier, Figure semblable #1, 2012. Etching on paper, edition of 5. (Courtesy of the artist) 226 The Voyant dashboard consists of several tools, all of which are interactive and can be altered based on what the researcher wishes to analyze. This image displays the visualizations produced before completing any filtering for common words 237 Frequency of words related to emotion and time in The History of Emily Montague as determined by Voyant 239 (a–c) Trend charts for specific word usage in Brooke’s novel: (a, left) illustrates trends for “moment” with regard to emotion; (b, top right) shows “happy” with the more immediate temporal terms “day” and “present”; (c, bottom right) includes terms that consider longer periods of time 240

xviii 

List of Figures

Fig. 7.4

Fig. 7.5 Fig. 7.6

Fig. 7.7

Fig. 7.8 Fig. 7.9

(a–d) Voyant produces Cirrus visualizations (word clouds) to illustrate word frequency: (a, top left) illustrates frequency prior to eliminating the novel’s title as stop words, while (b, top right) shows the new frequencies; (c, middle) demonstrates Voyant’s Contexts tool; (d, bottom) shows a mimeographed page from the text, complete with the nonsensical catchword “tremely.” The Mapping the Republic of Letters project explores data visualizations relating to the Enlightenment correspondence, including letters by Franklin, Locke, Voltaire, and others (a–d) A social network visualization for a text with a wide cast of characters can be complex, such as this one, analyzing George Eliot’s Middlemarch, posted on Twitter by Katherine Ognyanova (Katherine Ognyanova, Twitter post, September 22, 2015, 4:44 p.m., https://twitter.com/ognyanova) (a, top left), whereas a similar network visualization for The History of Emily Montague’s letter writers is less revealing (b, top right). Using location of writing is somewhat more interesting in a network visualization (d, bottom right), but not particularly helpful. It also requires a key (c, bottom left) to translate between location name and GPS coordinates, which Palladio requires for geospatial plotting (a–c) Palladio’s mapping feature connects places of writing and reading and indicates directionality by the concavity of the line illustrating travel: (a, top) shows Canada and England, while the closer views of the Canadian (b, middle) and English (c, bottom) locations allow for a magnified image of where Brooke’s characters are placed in the novel Palladio’s timeline highlights datapoints in blue according to a designated filter: in this case, the letter writer Visualization based on chronological dating of the letters in The History of Emily Montague created with Tableau. The letter that appears as number 35 in the text was written on 23 July 1766, so it appears between letters 6 and 7 on this visualization, allowing for an analysis of how its contents can shift the letter readers’ attention in the narrative from their own time to the letter writers’ past (here, John Temple’s). Letter writers are color-coded and appear in a key to the right of the image. A pop-up box identifies the writer’s name, the letter number, and the date of writing

243 247

248

250 252

255

  List of Figures 

Fig. 7.10

Fig. 7.11 Fig. 7.12

Fig. 7.13 Fig. 8.1

Fig. 8.2

Fig. 8.3

Fig. 8.4

xix

(a–b) Highlighting a single letter writer indicates the placement of one’s letters within a larger timeline. Here, Lucy Rivers does not write often, but her letters have a great deal of emotional impact (a, top). Temple’s writing is similarly infrequent, but by adjusting the height of the points on the timeline it becomes easier to see a single writer’s contributions 256 (a–b) Visualizations of the letters contained in section 2 (a, top) and section 7 (b, bottom) organized by number and color-coded by writer 258 Once date of writing is added to the visualization for section 7, where the letters are organized in the order in which they appear in the text, Lucy’s and Temple’s exchanges clearly stand apart from the others. Color-coding the addressee of the letters and using text to label each letter by writer illustrates which character might be contributing to word frequency in this particular section of the novel 259 This close-up view of section 4 indicates that several groups of letters were written on the same day within this span of the text, including an exchange between Emily and Bell 261 The owl frontispiece to The Dunciad: An Heroic Poem, In Three Books (1728) featuring Dennis’s Works. (Courtesy of Albert and Shirley Small Special Collections Library, University of Virginia) 277 Cover page of The Dunciad, Variorum. With the Prolegomena of Scriblerus (1729), containing the famous image of the ass carrying the dunces’ productions. Dennis’s Works are showcased here alongside those of Leonard Welsted, Ned Ward, Lewis Theobald, John Oldmixon, and Eliza Haywood. (Courtesy of Professor David Vander Meulen, University of Virginia)278 First page of The Dunciad, Variorum. With the Prolegomena of Scriblerus (1729), containing a frontispiece showing an owl wearing a fool’s cap, a possible impersonation of critic John Dennis, aka “Furius.” (Courtesy of Professor David Vander Meulen, University of Virginia) 279 Graph describing John Dennis’s relations in poem and apparatus as NEATO (spring-model layout). The relations described indicate similarity (green), characters attacked (red), and characters defended (dotted green). Dennis’s relations with Edmund Curll, Giles Jacob, and Leonard Welsted are bidirectional, which indicates strong ties 282

xx 

List of Figures

Fig. 8.5

Fig. 8.6

Fig. 8.7

Fig. 8.8

Fig. 8.9

Fig. 8.10

Fig. 9.1 Fig. 9.2a Fig. 9.2b Fig. 9.2c Fig. 9.3 Fig. 9.4

All relations for John Dennis as DOT (hierarchical layout). The graph highlights Dennis’s strong connection with Colley Cibber, Edmund Curll, and Giles Jacob (the edges are bidirectional)283 All relations for Giles Jacob as NEATO (spring-model layout). This graph highlights Jacob’s close relationship with critic John Dennis, with whom he is in a relation of similarity, and with Edmund Curll, whom Jacob supported in his campaign against Pope 286 Graph showing the social networks of all the six dunces considered as NEATO (spring-model layout). This graph highlights the three connectors of the poem: John Dennis, Edmund Curll, and Colley Cibber 288 Shiva Graph. This graph gives the viewer a keen sense of the relatedness of all the characters in the poem and makes visible the poems’ three connectors: Colley Cibber, Edmund Curll, and John Dennis 289 The Inner Circle. Full view of all networks of the six dunces considered as CIRCO (circular layout). This graph highlights the central and the peripheral characters of the poem and apparatus292 The Inner Circle. Detail view of all networks of the six dunces considered as CIRCO. This detail view highlights characters who appear in more than one network, or the main protagonists of the poem and the apparatus. The hall of fame/“good writers,” includes four authors: Alexander Pope, John Gay, Joseph Addison, and John Dryden. The hall of infamy/“bad writers” includes seventeen authors: Colley Cibber, Edmund Curll, Eliza Haywood, John Henley, John Ozell, John Oldmixon, Lewis Theobald, Giles Jacob, Laurence Eusden, Elkanah Settle, Thomas Cooke, John Dennis, Bernard Lintot, Charles Gildon, George Duckett, Leonard Welsted, and Richard Blackmore 293 (a–c) Subscription list to Charles Avison’s Eight Concertos, op. 4, 1755. Print from the author’s collection held at Durham University’s Palace Green Library 314 Location of subscribers to Avison’s op. 2 (1740 - List A) 320 Location of subscribers to Avison’s op. 4 (1755 - List E) 321 Location of subscribers to Avison’s op. 9, set 2 (1767 - List H) 322 The proportion of subscribers to all of Avison’s concerti grossi by gender 341 Types of subscribers to Avison’s op. 4 concertos (1755—List E) 344

List of Tables

Table 2.1 Table 2.2 Table 2.3 Table 2.4

Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 4.1

Placing Neuchâtel geo-politically through the STN database 39 Books traded by the STN, by edition type, 1769–1794 (all known “sales”) 47 Distribution of keywords by edition type. Commissioned editions and editions whose place of publication is unknown are excluded 50 Keyword sales of “illegal” works as a percentage of all “illegal” sales, by country, highlighting categories where French took at least 10 percent more than other regions (in bold) or those where France took 50 percent more (STN editions excluded) (bold plus italic). NB: “Illegal” is defined here as works appearing in Darnton’s Corpus of Clandestine Literature in France, 1769–178953 Top 20 canonical works (1500–1800) 80 Top 10 canonical literary works (1500–1800) 83 Top authors and works, canon editions, and total works recorded in the ESTC (1500–1800) 89 Distribution of subject-topics among works by top authors (1500–1800)91 Top 10 printing locations in the whole ESTC and in the canon (1500–1800) 104 Locations for the first printed editions and for subsequent editions of canonical works (1500–1800) 105 Words likely to bind to “knowledge” at distance 10 in the ECCO corpus for the period 1770–1780 126

xxi

xxii 

List of Tables

Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 5.1 Table 5.2 Table 5.3 Table 8.1

Table 9.1

Table 9.2 Table 9.3 Table 9.4 Table 9.5

Words likely to occur within sentences with “Africa,” with those most likely at the top of each list 128 Words likely to occur within sentences with “Asia,” with those most likely at the top of each list 129 Words likely to occur within sentences with “America,” with those most likely at the top of each list 130 Words likely to occur within sentences with “Europe,” with those most likely at the top of each list 131 Words likely to occur within sentences with “Africans,” with those most likely at the top of each list 144 Words likely to occur within sentences with “Americans,” with those most likely at the top of each list 146 Words likely to occur within sentences with “Europeans,” with those most likely at the top of each list 147 Top ten words for a six-cluster model of the matrix set at 61% sparsity172 Top twenty most probable terms in each topic generated by the twelve-topic model using the stm package 174 Topic prevalence in descending order of expected proportion and top ten terms for each topic in the twelve-topic model of the PCC 179 Spreadsheet that captures the salient information about John Dennis in the poem and apparatus: the character’s address in the text, the reference to the character (in bold when in the poem, in regular font when in the apparatus), the name of the person he is related to, the description of their relationship (similarity, dissimilarity, character attacked, or character defended), alternative references to Dennis, and authorial id 295 Subscription lists to Avison’s publications (1740–1767). When the year printed on the title page was not, in fact, the year when the work was issued, the actual year of publication has been given in square brackets 313 Musical societies which subscribed to Avison’s publications. The figure in brackets refers to the number of copies the society purchased 323 Musicians who subscribed to Avison’s publications 328 Members of the aristocracy and gentry who subscribed to Avison’s publications 337 Publishers who subscribed to Avison’s publications 340

CHAPTER 1

Introduction: “Speaking to the Eyes”— Reassessing the Enlightenment in the Digital Age Ileana Baird

On inspecting any one of these Charts attentively, a sufficiently distinct impression will be made, to remain unimpaired for a considerable time, and the idea which does remain will be simple and complete. —William Playfair, The Commercial and Political Atlas (1786) The greatest value of a picture is when it forces us to notice what we never expected to see. —John W. Tukey, Exploratory Data Analysis (1977)

I start my argument about the necessity of this book by imagining a dialogue of sources, almost two hundred years apart. Writing in 1786 on the expressive power of his revolutionary charts, William Playfair argued for I would like to thank Michael Gavin and Shawn W.  Moore for their attentive reading of this introduction and insightful suggestions. Their collegiality and support are truly appreciated. I. Baird (*) Zayed University, Abu Dhabi, United Arab Emirates e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_1

1

2 

I. BAIRD

their usefulness in a manner typical of Enlightenment rationalism, which is predicated on the belief that empirical knowledge is grounded in our senses. His language is notably reminiscent of this belief: the visual “impression” his charts make is “distinct,” “unimpaired,” “simple,” “complete,” and lasting.1 They convey the message vividly, clarify beyond doubt, and stand the test of time. To this description of scientific certainty John W.  Tukey adds the element of surprise: a good picture is that in which what we “see” is forced upon us almost against our will. Playfair’s focus on the attentive “inspection” of a picture is shifted in Tukey’s assessment to an unexpected epiphany, or sudden “notice,” of something that we could not see before.2 The seeing is no longer believing but serendipitously discovering something new. Simply and effectively defined by Andy Kirk as “the representation and presentation of data to facilitate understanding,”3 data visualization allows the researchers to perceive, comprehend, and interpret data in new and innovative ways, based on their preexisting knowledge. Its aim is not only to convey information in a visually memorable way, as Playfair suggested, but also to make visible information that otherwise would be buried in large amounts of data. Patterns, trends, relations, absences even—are all brought to the fore, suddenly noticed, and, as a result, explored and their possible implications unveiled. The advent of computational methods has changed, in other words, the very way in which we think about the object of our study. For scholars of the Enlightenment, the opportunities created by conducting their work in a Golden Age of data graphics have been momentous. Institutional projects, digital humanities centers, and individual initiatives in the field of eighteenth-century studies, some of which are attentively discussed in the afterword to this book, have flourished in recent years, many of them placing data visualization at the center of their endeavors. In the past decade, all major national and international conferences in the field have hosted panels, workshops, or roundtable discussions that focused on the use of digital methods to shed new light on the “big data” made available by important digital archives, such as Early 1  William Playfair, The Commercial and Political Atlas: Representing, by Means of Stained Copper-Plate Charts, the Exports, Imports, and General Trade of England, at a Single View (London: Printed for J. Debrett et al., 1786), 4. 2  John W. Tukey, Exploratory Data Analysis (Reading, MA: Addison-Wesley, 1977), vi. 3  Andy Kirk, Data Visualization: A Handbook for Data Driven Design (London: Sage, 2016), 19.

1  INTRODUCTION: “SPEAKING TO THE EYES” 

3

English Books Online (EEBO) or Eighteenth-Century Collections Online (ECCO). Such academic venues have been hotspots of significant digital work: the Digitizing Enlightenment symposium series at the University of Australia, for instance, sprang from a conversation that took place at the annual meeting of the Society for French Historical Studies in 2014, and materialized recently in the publication of an edited collection with the same title.4 More recently, the 15th Congress of the Enlightenment, organized by the International Society for Eighteenth-Century Studies (ISECS) in Edinburgh in 2019, hosted an impressive number of panels that discussed the new affordances of digital technologies and the novel ways in which eighteenth-century researchers can communicate their findings by using such methods. This interest culminated in a day-long special event organized by The Voltaire Foundation on the use of digital methods in our discipline, an event that gathered a significant number of presenters, many of whom supported their arguments with suggestive visualizations. Our own collection stemmed from a roundtable discussion on Data Visualization in Eighteenth-Century Studies organized at the 49th Annual Meeting of the American Association of Eighteenth-Century Studies (ASECS) in 2018 and from ongoing conversations that ensued in two ISECS panels: The Digital Eighteenth-Century: Directions and Opportunities, and Digital Approaches to Eighteenth-Century Studies. This is not to say that the strides made by data visualization in our discipline and beyond have come without challenges. Chart junk, distortions in presenting data, illegibility when too much data is translated in visual form, biases or access limitations in collecting data, a lack of tools to represent data in a meaningful way by humanist measures, and even loss of work due to digital tools or programming languages being constantly upgraded are all real. Moreover, as Johanna Drucker cautions, “graphical tools are a kind of intellectual Trojan horse”: we should be constantly reminded to put techniques of graphical display on a foundation that is humanistic in nature and able to reflect not only existing information but also interpretive complexity. As Drucker explains, this means that, “[a]t the very least, humanists beginning to play at the intersection of statistics and graphics ought to take a detour through the substantial discussions of the sociology of knowledge and its developed critique of realist models of 4  Simon Burrows and Glenn Roe, eds., Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies (Liverpool: Voltaire Foundation in association with Liverpool University Press, 2020).

4 

I. BAIRD

data gathering. At best, we need to take on the challenge of developing graphical expressions rooted in and appropriate to interpretative activity.”5 Or, as practitioners like Franco Moretti point out, what we do with these visualizations is sometimes lacking: “so much can be generated visually in such captivating ways, in such enormously information-rich ways, that too little time is spent analyzing what can be produced.”6 Placed at the intersection of digital humanities and Enlightenment studies, this collection explores the new interpretive possibilities created by using data visualizations to communicate large or complex information. Such visualizations include tabulations, charts, k-means clustering, topic modeling, social network graphs, data mapping, and/or other illustrations of patterns of social or intellectual exchange. As compellingly demonstrated by our contributors, these are invaluable methods that help researchers interpret large amounts of data, or data that is structurally embedded in texts but not immediately legible when using traditional methods of investigation. They are also methods, we argue, that allow eighteenth-century scholars to start thinking of the Enlightenment in ways that are not only more encompassing but also more solidly based on quantitative data extracted from existing repositories or texts. While superb recent work in our field has increasingly drawn attention to the need to think outside Eurocentric views and recognize the “constellations of global synchronicity”7 that the Enlightenment is, or to acknowledge the existence of “plural Enlightenments”8 and their global interconnectivity, digitization has opened up new avenues of inquiry that are equally exciting. To claims about the need to broaden the temporal or spatial breadth of the Enlightenment it adds the opportunity to increase the depth of our study of the Enlightenment corpus and reassess existing master narratives with the new tools afforded by the digital age. 5   Joanna Drucker, “Humanities Approaches to Graphical Display,” DHQ:  Digital Humanities Quarterly 5, no. 1 (2011): 1, http://www.digitalhumanities.org/dhq/ vol/5/1/000091/000091.html 6  Ruben Hackler and Guido Kirsten, “Distant Reading, Computational Criticism, and Social Critique: An Interview with Franco Moretti,” Le Foucaldien 2, no. 1 (2016): 13, https://doi.org/10.16995/lefou.22 7  Sebastian Conrad, “Enlightenment in Global History: A Historiographical Critique,” American Historical Review 117, no. 4 (2012): 1016, https://doi.org/10.1093/ ahr/117.4.999 8  David T.  Gies and Cynthia Wall, eds., The Eighteenth Centuries: Global Networks of Enlightenment (Charlottesville: University of Virginia Press, 2018).

1  INTRODUCTION: “SPEAKING TO THE EYES” 

5

Eighteenth-Century Precursors to Data Visualization Data visualization arguably started in the eighteenth century with the work of the Scottish engineer and political scientist William Playfair (1759–1823), the creator of the first statistical graphs.9 Playfair invented visualization techniques that are in standard use today: pie and bar charts, line graphs, and circle graphs. His Commercial and Political Atlas (1786) was the first book to contain statistical charts10 and his Statistical Breviary (1801) the first work ever to contain a pie chart.11 As Playfair memorably stated, “the best way to capture the imagination is to speak to the eyes.”12 His visualizations are, indeed, quite telling: they are high-quality copper-­ plate engravings colored by hand in three or four colors that convey meaning and chart multivariate data.13 In a chart showing the population and revenue of the main countries in Europe in 1804, for instance, Playfair represented each country as a circle (using green and red to suggest maritime or land powers), the country’s population in millions as a line (left), the taxes collected in millions of pounds sterling as a line (right), and he connected population and taxes by using a sloping line to show which one is higher (Fig. 1.1). In another example, Playfair used a bar chart to show Scotland’s imports and exports from and to seventeen countries in 1781 by using coordinates of revenue and location that allow for clear comparison (Fig. 1.2). Finally, in a line graph published in Chronology of Public Events and Remarkable Occurrences (1824),14 Playfair used color-coded lines to visualize fluctuations in revenue (brown), expenditure (green), debt (red), price of stocks (blue) and bread (yellow), imports (dark blue) and exports (purple) over time (Fig. 1.3). These examples show Playfair’s 9  For details, see Jurgen Symanzik, William Fischetti, and Ian Spence, “Commemorating William Playfair’s 250th Birthday,” Computational Statistics 24, no. 551 (2009): 551–66, https://doi.org/10.1007/s00180-009-0170-z 10  This book was published in a revised edition as Lineal Arithmetic (London: Printed for the Author, 1798) which contained thirty-three charts. 11  Cf. Edward R.  Tufte, The Visual Display of Quantitative Information (Cheshire, CT: Graphics Press, 2001), 44. 12  William Playfair, Elemens de statistique (Paris: s.n., 1802), xx. 13  See Jefferson Bailey and Lily Pregill, “‘Speak to the Eyes’: The History and Practice of Information Visualization,” Art Documentation: Journal of the Art Libraries Society of North America 33, no. 2 (2014): 168–91, https://doi.org/10.1086/678525 14  William Playfair, Chronology of Public Events and Remarkable Occurrences within the Last Fifty Years; or from 1774 to 1824 (London: Printed for G. and W. B. Whittaker, 1824).

6 

I. BAIRD

Fig. 1.1  “Chart Representing the Extent, Population & Revenue of the Principal Nations in Europe in 1804.” William Playfair, An Inquiry into the Permanent Cause of the Decline and Fall of Wealthy and Powerful Nations (London, 1805). (Courtesy of the Thomas Fisher Rare Book Library, University of Toronto)

ability “to select perceptual attributes that would best support efficient cognitive operations,”15 such as color-coded categories, slopes to compare trends, length and circles to code quantities, and so on. Such work was not singular during the “Age of Reason”: throughout the century, data visualization was variously employed by other Enlightenment intellectuals as well. In 1701, Edmond Halley created the first contour map to show curves of equal magnetic declination.16 In Systema Naturae (1735), Carl Linnaeus relied on tabular notations to classify plants, animals, and minerals in a grid-like manner. Jacques Barbeu-­ Dubourg’s Machine chronologique (1753) used a gridded graphic method to capture historic events, and François Quesnay’s Tableau économique (1758) contained an economic chart interweaving three economic 15  Ian Spence, “William Playfair and the Psychology of Graphs,” in JSM Proceedings (Alexandria, VA: American Statistical Association, 2006), 2431. 16  Edmond Halley, The Description and Uses of a New, and Correct Sea-Chart of the Whole World, Shewing Variations of the Compass (London: Printed for the Author, 1701).

1  INTRODUCTION: “SPEAKING TO THE EYES” 

7

Fig. 1.2  “Exports and Imports of Scotland to and from different parts for One Year from Christmas 1780 to Christmas 1781.” William Playfair, The Commercial and Political Atlas (London, 1786)

stakeholders to highlight their relationship of exchange.17 Directly influenced by Ephraim Chambers’  Cyclopaedia (1728), Denis Diderot and Jean-Baptiste Le Rond d’Alembert’s monumental Encyclopédie18 (1751–1772) contained 2,569 plates, including some of the first “do-it-­ yourself” diagrams ever to be created. As John Bender and Michael Marrinan  aptly put it, they are the expression of a new “culture of diagram” that used visual correlation as a form of knowledge: “[t]he plates materialize the sensuous feel of engraved printing while mapping correlations among people, places, and things to produce a fulsome, extra-optical

17  These three examples of data visualization are discussed in detail by Jakub Zdebik in the chapter “The Grid and the Visualization of Abstract Information: Three Eighteenth-Century Models” included in this collection. 18  Denis Diderot and M. D’Alembert, Encyclopédie, ou dictionnaire raisonné des sciences, des arts et des métiers, 17 vols. (Paris: Briasson et al., 1751–1765).

8 

I. BAIRD

Fig. 1.3  “Linear Chronology, Exhibiting the Revenues, Expenditure, Debt, Price of Stocks & Bread, from 1770 to 1824.” William Playfair, Chronology of Public Events and Remarkable Occurrences within the Last Fifty Years; or from 1774 to 1824 (London, 1824)

understanding of the practices, métiers, and products of contemporary life.”19 Later in the century and beyond, developments in representing data visually followed in rapid succession. Playfair himself may have been influenced by Joseph Priestley’s timeline charts (1765)20 and James Watt’s indicator diagrams (1796).21 In 1770, the French Academy’s Mémoires de mathématique et de physique included a timeline bar graph by Philippe Buache that suggestively plotted data showing month-by-month Seine water levels for the period 1760–1766.22 In Pyrometrie (1779), Johann Heinrich Lambert, “the only scientist in the eighteenth century to use  John Bender and Michael Marrinan, The Culture of Diagram (Stanford: Stanford University Press, 2010), 17. 20  Between 1765 and 1769, Joseph Priestley created the first timeline charts to illustrate eminent persons’ lifetime spans in A Chart of Biography (London: Published for J. Johnson, 1765), and the world’s most important historical moments and their distribution in space in A New Chart of History (London: Published for J. Johnson, 1769). For details, see James R. Beniger and Dorothy L. Robyn, “Quantitative Graphics in Statistics: A Brief History,” The American Statistician 32, no. 1 (1978): 1–11,  https://www.jstor.org/ stable/2683467 21  See H. Gray Funkhouser, “Historical Development of the Graphical Representation of Statistical Data,” Osiris 3 (1937): 289, https://www.jstor.org/stable/301591; and Thomas L. Hankins, “Blood, Dirt, and Nomograms: A Particular History of Graphs,” Isis 90, no. 1 (1999): 50–80, www.jstor.org/stable/237474 22  “Extrait des Profils qui representent la Crue et la Diminution des Eaux de la Seine.” For details, see Ronald K.  Smeltzer, “One for the History Books: An Early Time-Line Bar Graph,” Chance 23, no. 2 (2010): 54–56, https://doi.org/10.1007/s00144-010-0024-z 19

1  INTRODUCTION: “SPEAKING TO THE EYES” 

9

graphs extensively,”23 created the first graphical representations of variations in soil temperature and solar warming over time.24 In 1782, Charles de Fourcroy used proportional figures to compare demographic data in superposition,25 and August Friedrich Wilhelm Crome created the first statistical map, showing the geographic distribution of fifty-six European commodities.26At the turn of the century, The Philosophical Transactions of the Royal Society of London had already started to show meteorological data in a tabular and graph form; one such example is Dr. Buxton’s 1794 graph of barometric variations.27 In 1795, Louis-Ézéchiel Pouchet designed the first graph table to represent three variables at once by using a family of curves.28 Later on, in Carte figurative (1844), the French engineer Charles Joseph Minard used casualty data to illustrate the fate of Napoleon’s army in Russia during the 1812 campaign in a graph displaying a complex, multivariate time-space story considered by Edward R. Tufte as “the best statistical graph ever drawn.”29 As this brief overview suggests, during the eighteenth century data visualization becomes an important expression of visual thinking that will lay the foundations for the radical developments in the next century and beyond. From tabular recording of information to timeline charts to statistical graphics to diagrams and graphical tables, eighteenth-century visualizations indicate the growing importance of gathering data “in large and periodic series … and the usefulness of these bodies of data for planning, for governmental response, and as a subject worthy of study in its own right.”30  Hankins, “Blood, Dirt, and Nomograms,” 56.  Johann Heinrich Lambert, Pyrometrie oder vom Maaße des Feuers und der Wärme mit acht Kupfertafeln (Berlin: Bey Haude & Spenser, 1779). 25  Charles de Fourcroy, Essai d’une table poléométrique, ou amusement d’un amateur de plans sur les grandeurs de quelques villes (Paris: M. Dupain-Triel pere, 1782). 26   August Friedrich Wilhelm Crome, Europens Produkte. Zum Gebrauch der neuen Produkten-Karte von Europa (Dessau: Verfasser, 1782). 27  The Philosophical Magazine (London: Printed by Davis, Taylor, and Wilks, 1800), 7:357. Cf. Funkhouser, “Historical Development of the Graphical Representation of Statistical Data,” 289. In 1794, Buxton also invented the printed graph paper with a rectangular coordinate grid. 28  Louis-Ézéchiel Pouchet, Échelles graphiques des nouveaux poids, mesures et monnaies de France, comparées avec celles des pays les plus commercants de L’Europe (Rouen: Guedra, 1795). 29  Tufte, The Visual Display of Quantitative Information, 40. See also Bailey and Pregill, “‘Speak to the Eyes’,” 171–73. 30  Michael Friendly, “Milestones in the History of Thematic Cartography, Statistical Graphics, and Data Visualization,” (2008): 1, http://www.math.usu.edu/~symanzik/ teaching/2009_stat6560/Downloads/Friendly_milestone.pdf. For more details, see also 23 24

10 

I. BAIRD

Data Visualization in Eighteenth-Century Studies: A Brief Survey of the Field The modern origins of the field go back to the 1950s, when computer graphics were first generated. Interactive statistical applications developed by E. B. Fowlkes and M. A. Fishkeller, among others, led to the spread of high-resolution graphics in the 1970s.31 The publication of the National Science Foundation’s report on “Visualization in Scientific Computing” in 1987 marked the emergence of data visualization as a new specialization in the field of computer graphics. From then on, the term “data visualization” started to apply to visualizations in both the scientific and information visualization fields,32 branching into statistical graphics and thematic cartography. Information graphics, a closely related field whose emphasis is on representing data in visually striking ways, emerged during the same time. Nigel Holmes’ creation of the first infographics or, as he called them, “explanation graphics,”33 was an important milestone in representing data visually. In the humanities, the growing interest in information analysis and the increasing accessibility of large datasets have led to what Jefferson Bailey and Lily Pregill called “a visualization renaissance … with information analysis and visualization literacy recognized as fundamental skills in the academy.”34 This renaissance built upon the seminal work of Jacques Bertin, whose Semiologie graphique (1967) was the first book to organize Michael Friendly, “A Brief History of Data Visualization,” in Handbook of Data Visualizations, ed. Anthony Unwin, Chun-houh Chen, and Wolfgang K. Härdle (Berlin: Springer-Verlag, 2008), 15–56. 31  Cf. Friendly, “A Brief History of Data Visualization,” 40. 32  Scientific visualizations refer to 3D visualization of architectural, biological, or other type of scientific data, with an emphasis on realistic renderings of volumes, surfaces, sources, etc. In its broadest sense, information visualization refers to any display that organizes the information in a way that allows for finding relations among heterogenous elements; examples include tables, graphs, maps, and even texts. As a new research field, information visualization was launched in the early 1990s and describes the “visual representation of large-scale collections of non-numerical information, such as files and lines of code in software systems, library and bibliographic databases, networks of relations on the Internet, and so forth.” Cf. Friendly, “Milestones in the History of Thematic Cartography, Statistical Graphics, and Data Visualization,” 2. 33  See, for instance, Nigel Holmes, Designer’s Guide to Creating Charts and Diagrams (New York: Watson-Guptill, 1984) and Pictorial Maps: History, Design, Ideas, Sources (New York: Watson-Guptill, 1991), to give just a couple of examples. 34  Bailey and Pregill, “‘Speak to the Eyes’,” 169.

1  INTRODUCTION: “SPEAKING TO THE EYES” 

11

graphic visual elements according to features and relations in data,35 and Edward R.  Tufte, whose influential volume, The Visual Display of Quantitative Information (1983), provided a working theory and rules of good practice in data visualization. In Tufte’s words, “graphical excellence consists of complex ideas communicated with clarity, precision, and efficiency.”36 This involves using “well-designed presentation of interesting data,” giving the viewer “the greatest number of ideas in the shortest time with the least ink in the smallest space,” considering multivariate data, and telling the truth about the data.37 Data visualization designers like David McCandless, on the other hand, have called attention to the aesthetic dimension of such visualizations and the importance of “making information approachable and beautiful”38 [emphasis added] as an essential way to convey meaning. Moreover, as Johanna Drucker correctly pointed out, a humanistic approach to graphical display of information also involves constructive moves that should account for ambiguity and uncertainty, as well as a “recognition of the interpretive nature of knowledge”39 [emphasis in original]: The humanistic aspect of this approach should be obvious: that knowledge created with the acknowledgement of the constructed nature of its premises is not commensurate with principles of certainty guiding empirical or realist methods. Humanistic methods are counter to the idea of reliably repeatable experiments or standard metrics that assume observer-independent phenomena. By definition, a humanistic approach is centered in the e­ xperiential, subjective conditions of interpretation. Phenomena and their observers are co-dependent, not necessarily in equal measure.40

With the advent of new digital tools in recent years, such analytical possibilities have increased significantly, allowing researchers of the Enlightenment to revisit, among other things, the history of the novel, 35  Bertin considers seven visual variables: position, form (shape), orientation, color (hue), texture, value (lightness or darkness of color), and size and develops a visual semantics for linking data attributes to visual elements. See Jaques Bertin, Semiology of Graphics: Diagrams, Networks, Maps, trans. William J. Berg (Madison: University of Wisconsin Press, 1983). 36  Tufte, The Visual Display of Quantitative Information, 51–53. 37  Ibid. 38  David McCandless, Information Is Beautiful (London: Collins, 2009), n.p. 39  Johanna Drucker, Graphesis: Visual Forms of Knowledge Production (Cambridge, MA: Harvard University Press, 2014), 128–29. 40  Ibid., 130.

12 

I. BAIRD

networks of intellectual exchange, or the notion of canonicity itself. Period-related studies have made use of tabulated data, scatterplots, maps, trees, dendograms, network diagrams, and histograms for their explanatory power and ability to represent a whole corpus at one glance. Stylometry,41 computational stylistics,42 topic modeling,43 agent-based modeling,44 social network analysis,45 geospatial mapping,46 and studies in authorship attribution47 have all employed visualizations that aid in the 41  John Burrows, Computation into Criticism: A Study of Jane Austen’s Novels and an Experiment in Method (Oxford: Clarendon Press, 1987). 42  Anne Bandry-Scubbi, “Chawton Novels Online, Women’s Writing 1751–1834, and Computer-Aided Textual Analysis,” ABO: Interactive Journal for Women in the Arts, 1640–1830 5, no. 2 (2015): 1–54, https://doi.org/10.5038/2157-7129.5.2.1 43  See Jeffrey M. Binder and Collin Jennings, “Visibility and Meaning in Topic Models and 18th-Century Subject Indexes,” Literary and Linguistic Computing 29, no. 3 (2014): 405–11, https://doi.org/10.1093/llc/fqu017, and “‘A Scientifical View of the Whole’: Adam Smith, Indexing, and Technologies of Abstraction,” ELH 83, no. 1 (2016): 157–80, https://doi.org/10.1353/ELH.2016.0001; Glenn Roe, Clovis Gladstone, and Robert Morrissey, “Discourses and Disciplines in the Enlightenment: Topic Modeling the French Encyclopédie,” Frontiers in Digital Humanities 2 (2016): 1–8, https://doi.org/10.3389/ fdigh.2015.00008; and David J.  Newman and Sharon Block, “Probabilistic Topic Decomposition of an Eighteenth-Century American Newspaper,” Journal of the American Society for Information Science & Technology 57, no. 6 (2006): 753–67, https://doi. org/10.1002/asi.20342 44  Michael Gavin, “Agent-Based Modeling and Historical Simulation,” DHQ:  Digital Humanities Quarterly 8, no. 4 (2014): 1–18, http://www.digitalhumanities.org/dhq/ vol/8/4/000195/000195.html 45  Two chapters in Chloe Edmondson and Dan Edelstein, eds., Networks of Enlightenment: Digital Approaches to the Republic of Letters (Liverpool: Voltaire Foundation in association with Liverpool University Press, 2019) provide suggestive social network visualizations: Chloe Edmondson’s “Julie de Lespinasse and the ‘Philosophical’ Salon,” 139–72, and Charlotta Wolff’s “‘Un admirateur des philosophes modernes’: The Networks of Swedish Ambassador Gustav Philip Creutz in Paris, 1766–1783,” 173–200. See also, Ileana Baird, “The Strength of Weak Ties: Eliza Haywood’s Social Network in The Dunciad in Four Books (1743),” ABO: Interactive Journal of Women in the Arts, 1640–1830 9, no. 2 (2019): 1–36, https://doi.org/10.5038/2157-7129.9.2.1202 46  See, for instance, The Grub Street Project, a digital edition of eighteenth-century London that maps its print culture, literature, and trades, at http://grubstreetproject.net/. Some of these maps are discussed in Allison Muri, “Graphs, Maps, and Digital Topographies: Visualizing The Dunciad as Heterotopia,” Lumen 30 (2011): 79–98, https://doi. org/10.7202/1007717ar. An important study is also James Raven’s Bookscape: Geographies of Printing and Publishing in London before 1800 (London: The British Library, 2014). 47  See, for instance, John Burrows, “Who Wrote Shamela? Verifying the Authorship of a Parodic Text,” Literary and Linguistic Computing 20, no. 4 (2005): 437–50, https://doi. org/10.1093/llc/fqi049; Lisa Pearl, Kristine Lu, and Anousheh Haghighi, “The Character in the Letter: Epistolary Attribution in Samuel Richardson’s Clarissa,” Digital Scholarship in the Humanities 32, no. 2 (2017): 355–76, https://doi.org/10.1093/llc/fqw007; and

1  INTRODUCTION: “SPEAKING TO THE EYES” 

13

organization and interpretation of data, or that bring  to light relations between units of information otherwise hidden in the larger corpus. Many of the studies applying data visualization methods to the field of eighteenth-century studies are the product of major institutional projects. Stanford University’s Mapping the Republic of Letters, the most ambitious enterprise to date, is a collaborative project started in 2008 that aims at unveiling the correspondence networks created by scientists, salonnières, or travelers during the Enlightenment. With case studies including charting the movement of intellectuals and socialites between salons, mapping the social spaces of the Grand Tourists who visited Italy during the eighteenth century, tracing the  epistolary and publication networks of the Enlightenment, and outlining the production and circulation of the scientific knowledge during pre-modern era, this initiative uses sophisticated visualization tools to show how such networks facilitated a transnational circulation of ideas, people, and things.48 University of Virginia’s Sciences, Humanities, and Art Network of Technological Initiatives (SHANTI) provides a suite of tools that make it easy to create interactive, web-based visualizations, some of which have been employed in period-specific projects.49 Under the directorship of Laura Mandell, The Center of Digital Humanities Research (CoDHR) at Texas A&M University hosts a Humanities Visualization Space that supports the liberal arts and humanities research community by providing visualization technologies and workshops, many of them relevant to our field. Similarly, 18thConnect: Eighteenth-Century Scholarship Online offers access to The Eighteenth-­ Century Poetry Archive (ECPA), a digital archive which provides visualization and modeling tools to augment close readings of poems, and will soon offer access to Voyant, a web-based reading and analysis environment for digital texts whose benefits are attentively discussed in her chapter by one of the contributors of this book.50 Other important initiatives include University of Oxford’s Cultures of Knowledge project, which focuses on reassembling and interpreting the A.  Abdul-Rahman et al., “Constructive Visual Analytics for Text Similarity Detection,” Computer Graphics Forum 36, no. 1 (2017): 237–48, https://doi.org/10.1111/cgf.12798 48  For more details about participants, goals, and case studies, see http://republicofletters. stanford.edu/index.html 49  Shanti Interactive, University of Virginia, accessed April 1, 2020, http://www.viseyes.org/ 50  See Courtney A.  Hoffman’s chapter, “Exploring Data Visualization: Time, Emotion, and Epistolarity in Frances Brooke’s The History of Emily Montague.”

14 

I. BAIRD

correspondence networks of the early modern period (1550–1750),51 providing suggestive visualizations of letter frequency and distribution in space and time in their collection of Early Modern Letters Online. The French Book Trade in Enlightenment Europe (FBTEE) database, an award-­ winning project currently hosted at Western Sydney University, maps the production, marketing, dissemination, and reception of books and ideas during the late eighteenth century, illustrating the trade of the Swiss publishing house Société Typographique de Neuchâtel in suggestive visualizations.52 HELDIG, a Finnish research network approaching problems in the humanities and social sciences with computational methods, has been engaged in several projects involving the global Enlightenment that employ data visualization.53 Another award-winning online collection hosted by the Bodleian Library, Electronic Enlightenment (EE), attempts to recast our understanding of the period by making available over 77,000 letters by over 10,000 correspondents, linking them to their source location, and providing images of historical maps at the country level.54 The ARTFL Project, a collaboration between the French government and the University of Chicago, is a consortium-based service that provides its members with access to North America’s largest collection of digitized French resources; besides ongoing digitization projects, it includes a variety of projects that employ data visualization tools to report their findings.55 The Comédie-Française Registers Project (CFRP), the result of a partnership between the Bibliothèque-Musée de la Comédie-Française and five leading French and North American universities (Université Paris Nanterre, Université Paris-Sorbonne, MIT, Harvard University, and University of Victoria), has developed several data exploration and visualization tools, creating register statistics, visual displays of performance frequency, office box data for authors and plays, or ticket sales during 51  Cultures of Knowledge: Networking the Republic of Letter, 1550–1750, University of Oxford, accessed April 1, 2020, http://www.culturesofknowledge.org 52  FBTEE: The French Book Trade in Enlightenment Europe. Mapping the Trade of the Société Typographique de Neuchâtel, 1769–1794, Western Sydney University, 2014, http://fbtee. uws.edu.au/main/ 53  Helsinki Center for Digital Humanities, University of Helsinki, 2020, https://www. helsinki.fi/en/helsinki-centre-for-digital-humanities 54  Electronic Enlightenment Project, Bodleian Libraries, University of Oxford, 2008–2019, https://www.e-enlightenment.com/. Currently, this is the most wide-ranging online collection of edited correspondence of the early modern period. 55  The Project for American and French Research on the Treasury of the French Language (ARTFL), The University of Chicago, accessed February 20, 2020, https://artfl-project. uchicago.edu/content/about-artfl

1  INTRODUCTION: “SPEAKING TO THE EYES” 

15

particular theater seasons, to name just a few.56 Similar initiatives that use quantitative methods applied to eighteenth-century datasets or texts have bourgeoned in the past years globally, resulting in a growing number of publications that employ data visualization as a methodological tool. Among them, several recent edited collections have made a significant contribution to expanding our field. Chloe Edmondson and Dan Edelstein’s Networks of Enlightenment: Digital Approaches to the Republic of Letters assembles data-driven scholarship related to the study of the correspondence, social, and knowledge networks that made the Enlightenment possible in an effort to create a sense of a “European” identity. The case studies included here range from Catherine the Great’s epistolary networks and Voltaire’s correspondence to Casanova’s French coteries and Samuel Johnson’s lexicographic networks in his Dictionary. Five of its chapters contain suggestive epistolary, social, and knowledge network visualizations that  illustrate promotional strategies, salon demographics, trajectories of intellectual transmission, hubs of French academic networks, science super-connectors, citation frequency, or semiotic clusters.57 In Reassembling the Republic of Letters in the Digital Age: Standards, Systems, Scholarship, Howard Hotson and Thomas Wallnig gather work by scholars involved in creating a digital framework for multilateral studies on early modern letter-writing (1500–1800).58 The outcome of a COSTfunded project (European Cooperation in Science and Technology), this volume considers the challenges of planning a state-of-the art digital system that can collect highly granular data on the Republic of Letters, and the need for emerging technologies to build a workable, pan-­European institutional infrastructure. Chapters in this collection employ data visualizations to illustrate preliminary results, such as event distribution, event-­ based letter models, geographical and social network visualizations of library metadata, letter and place name distribution, epistolary itineraries, isopleth maps, and Minard diagrams of best postal routes between regions. 56  The Comédie-Française Registers Project (CFRP), accessed February 20, 2020, https:// www.cfregisters.org/en/ 57  Cheryl Smeall, “‘He Belonged to Europe’: Francesco Alagarotti (1712–1764) and His European Networks,” 75–106; Edmondson, “Julie de Lespinasse and the ‘Philosophical’ Salon”; Wolff, “Un admirateur des philosophes modernes’”; Melanie Conroy, “The Eighteenth-Century French Academic Network,” 225–50; and Mark Algee-Hewitt, “The Principles of Meaning: Networks of Knowledge in Johnson’s Dictionary,” 251–78. 58  Howard Hotson and Thomas Wallnig, eds., Reassembling the Republic of Letters in the Digital Age: Standards, Systems, Scholarship (Gottingen: Gottingen University Press, 2019).

16 

I. BAIRD

Finally, in Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies, Simon Burrows and Glenn Roe collect  contributions that chart the development of several inter-­ related digital projects that have played a transformative role in the way the history and culture of the eighteenth century is viewed. The studies included here describe major digital humanities research initiatives in eighteenth-century studies, such as the ARTFL Encyclopédie, Electronic Enlightenment, and Mapping the Republic of Letters. Especially in its second section, “Digital Methods and Innovations,” this collection discusses methods and technologies developed during these projects, digital tools and resources used by its contributors, and the role played by digital models and visualizations in uncovering evidence and building arguments. Including visualizations ranging from correspondence networks, book sale destinations, and geospatial mapping to topic distribution, histograms, and social networks of eighteenth-century French salonnières, Digitizing Enlightenment provides a suggestive view on how such methods can  contribute to the “reinvention and transformation of scholarly practices in the humanities at the dawn of the digital age.”59 Although works of exemplary scholarship that bring together contributions by authors involved in important digital projects, these studies do not make their goal to highlight the role played by data visualization per se in reassessing the Enlightenment. Moreover, not all chapters included in these collections employ visualizations, or a significant sample of them, in building their argument. They also focus mainly on network analysis, addressing to a lesser extent other types of visualizations, their limits, or their potential for pedagogical use. Reporting mainly on large ongoing projects, they rarely include samples of smaller projects developed by individual researchers at institutional level, or, with only one exception, case studies focused on particular texts that employ data visualization as a methodological tool.60 Our collection aims to fill in these gaps by providing a representative sample of both large and small projects that illustrate the variety of the work done in eighteenth-century studies employing quantitative methods. The chapters included here present groundbreaking research by leading and emerging scholars in the field that involves the use of visualizations in 59   Simon Burrows and Glen Roe, introduction to Burrows and Roe, Digitizing Enlightenment, 24. 60  I refer here to Algee-Hewitt’s chapter, “The Principles of Meaning,” cited above.

1  INTRODUCTION: “SPEAKING TO THE EYES” 

17

analyzing texts and/or cultural trends emerging from data mined from existing databases or other aggregates of sources. This approach does not only provide, as Moretti suggested, a new “model of the narrative universe” by highlighting “hidden patterns” of contact and exchange61 but also unveils connections between various pieces of data—about an individual, about individuals in relation to others, about circulation of knowledge, and about spatial or temporal dynamics. By including digital projects undertaken by eighteenth-century scholars from various disciplines (visual arts, history, musicology, literary studies), this collection addresses both the benefits and challenges of the  digital humanities scholarship in our field. In organizing this collection, we have focused on the use of quantitative methods both for distant reading of large datasets or digitized collections and for close reading of particular texts. Challenging existing opinions that “literature is not data” and, even more, that “literature is the opposite of data,”62 we have set as our goal to demonstrate the huge potential that exploring literature as data can have in advancing our field. As evidenced by our contributors, distant readings of large datasets can reveal trends in literature and criticism, disciplinary directions, dissemination patterns, canon formation processes, genre- or gender-specific ebbs and flows over time, or othering strategies. On the other hand, close readings of texts or images can highlight plot trajectories, artistic influences, defamatory or prestige-­building strategies, social, sentimental, and chrono-spatial networks, or less obvious textual meanings. Finally, this distant/close dichotomy has been a conceptual focus as well: although this collection is necessarily Eurocentric given its engagement with the Enlightenment as an eighteenth-­century European intellectual movement, it has been our goal to broaden its spatial expanse by including contributions that address the Enlightenment authors’ interest in non-European cultures, or that discuss texts falling outside a strict Western-European context. In doing so, we have aimed to acknowledge the universalizing impulse that is a hallmark of the Enlightenment, and its global circulation of ideas and texts.

61  Franco Moretti, Graphs, Maps, Trees: Abstract Models for a Literary History (New York: Verso, 2005), 54. 62  Stephen Marche, “Literature Is Not Data: Against Digital Humanities,” Los Angeles Review of Books, October 28, 2012, https://lareviewofbooks.org/article/ literature-is-not-data-against-digital-humanities/

18 

I. BAIRD

Our first section, Digital Enlightenment: Representing Big Data, includes four studies resulting from major, ongoing institutional projects that focus on large corpora of texts and engage in various ways with the idea of canonicity. The chapters included in this section address, in turn, the tensions between the “high Enlightenment” canon and popular bestsellers, construct a new, data-driven canon based on quantitative approaches to large-scale datasets, engage with Eurocentric views of “the other” to highlight perceived differences between canonical and peripheral cultures, and reassess the place of poetry in the Enlightenment canon in general, and in contemporary critical discourse in particular. As highlighted by our contributors, computational methods and data visualization can offer exciting tools for analysis that can significantly revise and even dismantle existing ideas about the period. Simon Burrows’ chapter, “In Search of Enlightenment: From Mapping Books to Cultural History,” discusses the philosophy, methodology, conceptual dilemmas, and findings of the AHRC-funded French Book Trade in Enlightenment Europe (FBTEE) project. Currently based at Western Sydney University, this project aims at creating an interpersonal-relational database that maps the pan-European book trade of the celebrated Société typographique de Neuchâtel (STN), a large-scale Swiss wholesale publisher that operated between 1769–1794 and that is, to date, the richest source on the international book trade of the later Enlightenment. In this chapter, the author showcases how careful database design, customized digital tools, and an intuitive online interface allow users to compensate for complexities, biases, and gaps in the STN archive and argues that, by using these tools, one can create datasets that can stand proxy for the wider European book trade of the late eighteenth century. By using suggestive visualizations of sales destinations, transaction volumes, leading STN authors, and circulation of illegal bestsellers during the ancien régime, to name just a few, this chapter creates a vivid sense of the reception, transmission, and transformation of the Enlightenment ideas across space and time. In “Examining the Early Modern Canon: The English Short Title Catalogue and Large-Scale Patterns of Cultural Production,” Mikko Tolonen, Ali Zeeshan Ijaz, Ville Vaara, Mark J. Hill, and Leo Lahti present the findings of an ongoing digital project focused on the history of eighteenth-­century book publication funded by the Academy of Finland. As part of the Helsinki Computational History Group at Helsinki Centre for Digital Humanities (HELDIG), the authors have created a

1  INTRODUCTION: “SPEAKING TO THE EYES” 

19

historical-biographical database based on The English Short-Title Catalogue (ESTC), a standard source for analytical bibliographic research holding close to half a million titles, to construct a data-driven canon that considers changes over time, subject-topics, top-works, authors, publishers, publication place, and materiality. This chapter provides methodological and historical insights into the development of the Enlightenment print culture and demonstrates the huge analytical potential of harmonized metadata catalogs. While quantitative analyses of the history of the book trade have been done before, this is the first attempt to engage with the complex process of canon formation at such a large scale. In building their argument, the authors use a significant number of visualizations, ranging from tabular records of top canonical works and graphic representations of frequency variations of the most common subject-topics in the canon to reprint patterns, book distribution, publication places, and gendered  trends. The authors’ work highlights the decisive role played by publishers in the process of canon formation, and the epistemological shift started at the end of the seventeenth century, when religious works lost their dominant position within the canon, being increasingly replaced by literary works. As the authors compellingly argue, this shift in the production and consumption of print allowed for a reinvention of the canon during the eighteenth century. In “Europe and Its ‘Others’: Visualizing Lexical Relations between Western and Non-Western Locations of the Enlightenment in The Eighteenth-Century Collections Online,” John Regan complicates the discussion of Enlightenment Eurocentrism by looking at the ideological structures emerging from the Enlightenment literature itself when placed within a broader, global context. The author investigates how the names of continents interacted across three historical periods of the ECCO corpus in order to assess how British authors wrote about Europe versus its colonies, the Orient, the Indies, or the Americas. Using a custom-designed measure of word co-association developed by the members of the Concept Lab at the Cambridge Centre for Digital Knowledge, Regan makes visible associations between geographical designators in the early, middle, and late eighteenth century, creating semantic network diagrams that visualize the often surprising ways in which global place names and attributes kept company within sentences in the historical corpus. By using semantic networks and tabular records of lexical bindings, this chapter makes visible how knowledge about “Asia,” “Africa,” “America,” and “Europe” was structured and how these place names associated with each other in the

20 

I. BAIRD

vast, impersonal, aggregated repository of the corpus. Thus, this chapter reconstructs the most common ways in which the printed anglophone century understood relations between “Europe” and its others, suggesting a complementary story about Enlightenment othering that would have remained closed off without the affordances of the digital. In the last chapter of this section, “Text Mining and Data Visualization: Exploring Cultural Formations and Structural Changes in Fifty Years of Eighteenth-Century Poetry Criticism (1967–2018),” Billy Hall uses quantitative methods to identify thematic shifts in the critical reception of eighteenth-century poetry. The author employs computational tools to examine trends in current criticism, as reflected by essays published in two flagship journals in the field, Eighteenth-Century Studies and The Eighteenth Century: Theory and Interpretation. By using algorithmic manipulation, k-means clustering, LDA topic modeling, network graphs of dominant genres, and topic distribution over time, Hall examines patterns in attention to various texts and/or poets, identifies trending topics, and makes inferences about disciplinary focus, direction of disciplinary practice, and the impact of gender on the poetic canon. As the first ever attempted quantitative approach to contemporary poetry criticism in eighteenth-­ century studies, this study demonstrates the expanding interpretive possibilities afforded by digitization. The second section of the book, Data Visualization and the Eighteenth-­ Century Corpus: Case Studies, gathers chapters that are the result of smaller-size projects that employ visualizations for close reading of particular works. They look at eighteenth-century visualizations that anticipate postmodern forms of graphic design, employ  computational tools to bring to the fore less obvious information within texts, or analyze in a data-driven fashion paratextual materials, such as subscription lists, footnotes, and other apparatus, to draw conclusions about processes of diffusion, association, and prestige building. The visualizations used here shed light, on the one hand, on the importance of the grid as the underlying structure operative in data visualization during the eighteenth century and beyond and, on the other hand, on the benefits of such visualizations in providing a view of the data that “highlights potentially interesting patterns”63 within a text. These case studies show the payoffs of close 63  S.  Janicke, G.  Franzini, M.  F. Cheema, and G.  Scheuerman, “Visual Text Analysis in Digital Humanities,” Computer Graphics Forum 36, no. 6 (2017): 228, https://doi. org/10.1111/cgf.12873

1  INTRODUCTION: “SPEAKING TO THE EYES” 

21

reading primary sources with digital tools which, when coupled with the reader’s expert knowledge of their context of production, can provide invaluable insights into such texts. In “The Grid and the Visualization of Abstract Information: Three Eighteenth-Century Models,” Jakub Zdebik looks at the grid as an organizational device that helps convey information for scientific inquiry. The author discusses three eighteenth-century cases: Jacques Barbeu-­ Dubourg’s Machine chronologique (1752), where the grid is used to organize time, François Quesnay’s Tableau économique (1758), where the grid is used to represent economic exchange among various stakeholders, and Carl Linnaeus’s Regnum Lapideum (1768), where the grid is used as a display method for comparative purposes. These eighteenth-century instances of data visualization are analyzed in light of contemporary art history theories to demonstrate their lasting impact on contemporary artists like Francine Savard, Hans Haacke, and Luce Meunier, whose work mirrors the patterns of these early modern graphic designs. As the author convincingly  argues, the eighteenth-century method of mingling the visual and the textual in visualizations of abstract knowledge anticipates postmodern methods of artistic expression. Courtney A. Hoffman’s chapter, “Exploring Data Visualization: Time, Emotion, and Epistolarity in Frances Brooke’s The History of Emily Montague,” elaborates on the author’s use of data visualization as a tool for textual analysis and considers the affordances of three software programs: Voyant, Tableau, and Palladio. The author examines the textual analysis capabilities of these programs by using data related to the epistolary exchanges of the characters  in Frances Brooke’s 1769 novel, The History of Emily Montague. Voyant has helped the author identify specific word-use frequency throughout the text, Palladio has created maps that visualize the progress of the letters over space and the characters’ geospatial location, and Tableau has generated displays that order letters by date, rather than by appearance in the text. Hoffman illustrates the potential for graphic production of these programs, their usefulness when examining the temporal aspects of the novel’s epistolary structure, and their pitfalls in both research design structure and text selection. Thus, this chapter narrates the process by which such visualizations can lead to new interpretations of the source text that are significant in themselves or can point to new directions of analysis  and highlights their potential for use in the classroom.

22 

I. BAIRD

In “Outliers, Connectors, and Textual Periphery: Visualizing John Dennis’s Social Network in The Dunciad in Four Books,” Ileana Baird uses social network analysis to visualize the fields of relations involving John Dennis, the most important critic of the first half of the eighteenth century, with Alexander Pope’s “dunces,” and the nature of these affiliations. Using visualizations generated by GraphViz, a program that creates topological graphs from sets of dyadic relations, and ShivaGraph, a tool that helps visualize large networks and navigate through them as through a map, the author brings to light data that is structurally embedded in Pope’s poem but not immediately legible given the amount and complexity of the information. In Dennis’s case, they reveal the competing stories told by the poem and the apparatus and the critic’s prominent role as the uncrowned king of The Dunciad’s textual periphery. The graphs also highlight Dennis’s essential position as a network connector, his camp affiliations, the role played by peripheral characters in the plot network of the poem, and the identity of the main dunces targeted by Pope, or the poem’s “hall of infamy.” Simon D. I. Fleming’s chapter, “Publishing Music by Subscription in Eighteenth-Century Britain: The Case of Charles Avison,” considers one of the most valuable resources available to researchers of eighteenth-­ century social history: the subscription lists attached to various publications. In the musical field, these lists shed considerable light on the connections individual composers forged with other like-minded individuals, the support they received from members of the middle and upper classes, and gendered trends in purchasing musical prints. In this data-­ driven study, the author employs visualizations ranging  from tabulated data to scatter maps of subscriber locations to pie charts showing the types and the gender of Avison’s subscribers, highlighting Avison’s growth into one of England’s lead musicians and the importance of the connections he forged with particular groups or individuals. This chapter underscores the role played by  Avison’s extensive networking through subscription in building prestige, as well as the benefits of a data-driven approach to subscription lists in unveiling cultural trends and gendered practices during the Georgian era. Finally, in the afterword, “Novel Knowledge, or Cleaning Dirty Data: Toward Open-Source Histories of the Novel,” Emily C. Friedman discusses “the most important, most under-rewarded, and most unsexy aspect of data visualization: the production and/or usage of reliable underlying data.” Starting from the premise that visualizations are only as

1  INTRODUCTION: “SPEAKING TO THE EYES” 

23

good as their underlying evidentiary base, the author discusses the contributions of digital projects in eighteenth-century fiction that have laid the foundation for such practices, including massive multi-institution projects like Orlando, mid-size projects like The Early Novels Database (END), or small-scale projects like her own Manuscript Fiction in the Age of Print (MFAP), which creates meaningful metadata about unpublished manuscript fiction during the period. Cautioning against the potentially disastrous effects of creating visualizations from partial or “dirty data,” Friedman proposes a set of guidelines for best practices in working with or creating large datasets so that amendable, transformable visualizations can be produced, built on collective knowledge. As suggested by this outline, this collection is a truly interdisciplinary effort that showcases the significant digital humanities work done in the field of eighteenth-century studies and its potential to transform our disciplinary practices. By addressing fundamental  period-related themes— from issues of canonicity, intellectual history, and book trade practices to novel ways of exploring canonical authors and texts, gender roles, and public sphere dynamics—, this collection also makes a broader argument about the necessity of expanding the very notion of “Enlightenment” not only spatially but also conceptually, by revisiting its very tenets in light of new data. By translating these new findings in suggestive visualizations, the contributors to this collection unveil unforeseen patterns, trends, connections, or networks of influence that could potentially revise existing master narratives about the period and the ideological structures at the core of the Enlightenment.

Bibliography Abdul-Rahman, A. et al. 2017. Constructive Visual Analytics for Text Similarity Detection. Computer Graphics Forum 36 (1): 237–248. https://doi. org/10.1111/cgf.12798. Algee-Hewitt, Mark. 2019. The Principles of Meaning: Networks of Knowledge in Johnson’s Dictionary. In Edmondson and Edelstein, Networks of Enlightenment, 251–278. Bailey, Jefferson, and Lily Pregill. 2014. ‘Speak to the Eyes’: The History and Practice of Information Visualization. Art Documentation: Journal of the Art Libraries Society of North America 33 (2): 168–191. https://doi. org/10.1086/678525.

24 

I. BAIRD

Baird, Ileana. 2019. The Strength of Weak Ties: Eliza Haywood’s Social Network in The Dunciad in Four Books (1743). ABO: Interactive Journal of Women in the Arts, 1640–1830 9 (2): 1–36. https://doi.org/10.5038/2157-7129.9.2.1202. Bandry-Scubbi, Anne. 2015. Chawton Novels Online, Women’s Writing 1751–1834, and Computer-Aided Textual Analysis. ABO: Interactive Journal for Women in the Arts, 1640–1830 5 (2): 1–54. https://doi. org/10.5038/2157-7129.5.2.1. Bender, John, and Michael Marrinan. 2010. The Culture of Diagram. Stanford: Stanford University Press. Beniger, James R., and Dorothy L.  Robyn. 1978. Quantitative Graphics in Statistics: A Brief History. The American Statistician 32 (1): 1–11.  https:// www.jstor.org/stable/2683467. Bertin, Jaques. 1983. Semiology of Graphics: Diagrams, Networks, Maps. Trans. William J. Berg. Madison: University of Wisconsin Press. Binder, Jeffrey M., and Collin Jennings. 2014. Visibility and Meaning in Topic Models and 18th-Century Subject Indexes. Literary and Linguistic Computing 29 (3): 405–411. https://doi.org/10.1093/llc/fqu017. ———. 2016. ‘A Scientifical View of the Whole’: Adam Smith, Indexing, and Technologies of Abstraction. ELH 83 (1): 157–180. https://doi.org/10.1353/ ELH.2016.0001. Burrows, John. 1987. Computation into Criticism: A Study of Jane Austen’s Novels and an Experiment in Method. Oxford: Clarendon Press. ———. 2005. Who Wrote Shamela? Verifying the Authorship of a Parodic Text. Literary and Linguistic Computing 20 (4): 437–450. https://doi. org/10.1093/llc/fqi049. Burrows, Simon, and Glenn Roe, eds. 2020. Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies. Liverpool: Voltaire Foundation in Association with Liverpool University Press. ———. 2020. Introduction to Burrows and Roe, Digitizing Enlightenment, 1–24. Conrad, Sebastian. 2012. Enlightenment in Global History: A Historiographical Critique. American Historical Review 117 (4): 999–1027. https://doi. org/10.1093/ahr/117.4.999. Conroy, Melanie. 2020. The Eighteenth-Century French Academic Network. In Edmondson and Edelstein, Networks of Enlightenment, 225–250. Crome, August Friedrich Wilhelm. 1782. Europens Produkte. Zum Gebrauch der neuen Produkten-Karte von Europa. Dessau: Verfasser. de Fourcroy, Charles. 1782. Essai d’une table poléométrique, ou amusement d’un amateur de plans sur les grandeurs de quelques villes. Paris: M. Dupain-Triel pere. Diderot, Denis, and M.  D’Alembert. 1751–1765.  Encyclopédie, ou dictionnaire raisonné des sciences, des arts et des métiers. 17 vols. Paris: Briasson et al.

1  INTRODUCTION: “SPEAKING TO THE EYES” 

25

Drucker, Joanna. 2011. Humanities Approaches to Graphical Display. DHQ: Digital Humanities Quarterly 5 (1): 1–20. http://www.digitalhumanities.org/dhq/vol/5/1/000091/000091.html. ———. 2014. Graphesis: Visual Forms of Knowledge Production. Cambridge, MA: Harvard University Press. Edmondson, Chloe. 2019. Julie de Lespinasse and the ‘Philosophical’ Salon. In Edmondson and Edelstein, Networks of Enlightenment, 139–172. Edmondson, Chloe, and Dan Edelstein, eds. 2019. Networks of Enlightenment: Digital Approaches to the Republic of Letters. Liverpool: Voltaire Foundation in association with Liverpool University Press. Friendly, Michael. 2008a. A Brief History of Data Visualization. In Handbook of Data Visualizations, ed. Anthony Unwin, Chun-houh Chen, and Wolfgang K. Härdle, 15–56. Berlin: Springer-Verlag. ———. 2008b. Milestones in the History of Thematic Cartography, Statistical Graphics, and Data Visualization. 1–79. http://www.math.usu. edu/~symanzik/teaching/2009_stat6560/Downloads/Friendly_ milestone.pdf Funkhouser, H.  Gray. 1937. Historical Development of the Graphical Representation of Statistical Data. Osiris 3: 269–404. https://www.jstor.org/ stable/301591. Gavin, Michael. 2014. Agent-Based Modeling and Historical Simulation. DHQ: Digital Humanities Quarterly 8 (4): 1–18. http://www.digitalhumanities.org/dhq/vol/8/4/000195/000195.html. Gies, David T., and Cynthia Wall, eds. 2018. The Eighteenth Centuries: Global Networks of Enlightenment. Charlottesville: University of Virginia Press. Hackler, Ruben, and Guido Kirsten. 2016. Distant Reading, Computational Criticism, and Social Critique: An Interview with Franco Moretti. Le Foucaldien 2 (1): 1–17. https://doi.org/10.16995/lefou.22. Halley, Edmond. 1701. The Description and Uses of a New, and Correct Sea-Chart of the Whole World, Shewing Variations of the Compass. London: Printed for the Author. Hankins, Thomas L. 1999. Blood, Dirt, and Nomograms: A Particular History of Graphs. Isis 90 (1): 50–80. www.jstor.org/stable/237474 Holmes, Nigel. 1984. Designer’s Guide to Creating Charts and Diagrams. New York: Watson-Guptill. ———. 1991. Pictorial Maps: History, Design, Ideas, Sources. New  York: Watson-Guptill. Hotson, Howard, and Thomas Wallnig, eds. 2019. Reassembling the Republic of Letters in the Digital Age: Standards, Systems, Scholarship. Gottingen: Gottingen University Press.

26 

I. BAIRD

Janicke, S., G.  Franzini, M.F.  Cheema, and G.  Scheuerman. 2017. Visual Text Analysis in Digital Humanities. Computer Graphics Forum 36 (6): 226–250. https://doi.org/10.1111/cgf.12873. Kirk, Andy. 2016. Data Visualization: A Handbook for Data Driven Design. London: Sage. Lambert, Johann Heinrich. 1779. Pyrometrie oder vom Maaße des Feuers und der Wärme mit acht Kupfertafeln. Berlin: Bey Haude & Spenser. Marche, Stephen. 2012. Literature Is Not Data: Against Digital Humanities. Los Angeles Review of Books, October 28. https://lareviewofbooks.org/article/ literature-is-not-data-against-digital-humanities/ McCandless, David. 2009. Information Is Beautiful. London: Collins. Moretti, Franco. 2005. Graphs, Maps, Trees: Abstract Models for a Literary History. New York: Verso. Muri, Allison. 2011. Graphs, Maps, and Digital Topographies: Visualizing The Dunciad as Heterotopia. Lumen 30: 79–98. https://doi. org/10.7202/1007717ar. Newman, David J., and Sharon Block. 2006. Probabilistic Topic Decomposition of an Eighteenth-Century American Newspaper. Journal of the American Society for Information Science & Technology 57 (6): 753–767. https://doi. org/10.1002/asi.20342. Pearl, Lisa, Kristine Lu, and Anousheh Haghighi. 2017. The Character in the Letter: Epistolary Attribution in Samuel Richardson’s Clarissa. Digital Scholarship in the Humanities 32 (2): 355–376. https://doi.org/10.1093/ llc/fqw007. Playfair, William. 1786. The Commercial and Political Atlas: Representing, by Means of Stained Copper-Plate Charts, the Exports, Imports, and General Trade of England, at a Single View. London: Printed for J. Debrett et al. ———. 1798. Lineal Arithmetic. London: Printed for the Author. ———. 1802. Elemens de statistique. Paris: s.n. ———. 1805. An Inquiry into the Permanent Cause of the Decline and Fall of Wealthy and Powerful Nations. London: Printed for Greenland and Norris. ———. 1824. Chronology of Public Events and Remarkable Occurrences within the Last Fifty Years; or from 1774 to 1824. London: Printed for G. and W. B. Whittaker. Pouchet, Louis-Ézéchiel. 1795. Échelles graphiques des nouveaux poids, mesures et monnaies de France, comparées avec celles des pays les plus commercants de L’Europe. Rouen: Guedra. Priestley, Joseph. 1765. A Chart of Biography. London: Published for J. Johnson. ———. 1769. A New Chart of History. London: Published for J. Johnson. Raven, James. 2014. Bookscape: Geographies of Printing and Publishing in London Before 1800. London: The British Library.

1  INTRODUCTION: “SPEAKING TO THE EYES” 

27

Roe, Glenn, Clovis Gladstone, and Robert Morrissey. 2016. Discourses and Disciplines in the Enlightenment: Topic Modeling the French Encyclopédie. Frontiers in Digital Humanities  2: 1–8. https://doi.org/10.3389/ fdigh.2015.00008. Smeall, Cheryl. 2019. ‘He Belonged to Europe’: Francesco Alagarotti (1712–1764) and His European Networks. In Edmondson and Edelstein, Networks of Enlightenment, 75–106. Smeltzer, Ronald K. 2010. One for the History Books: An Early Time-Line Bar Graph. Chance 23 (2): 54–56. https://doi.org/10.1007/s00144-010-0024-z. Spence, Ian. 2006. William Playfair and the Psychology of Graphs. In JSM Proceedings, Statistical Graphics Section, 2426–2436. Alexandria, VA: American Statistical Association. Symanzik, Jurgen, William Fischetti, and Ian Spence. 2009. Commemorating William Playfair’s 250th Birthday. Computational Statistics 24 (551): 551–566. https://doi.org/10.1007/s00180-009-0170-z. The Philosophical Magazine. 1800. Vol. 7. London: Printed by Davis, Taylor, and Wilks. Tufte, Edward R. 2001. The Visual Display of Quantitative Information. Cheshire: Graphics Press. Tukey, John W. 1977. Exploratory Data Analysis. Reading, MA: Addison-Wesley. Wolff, Charlotta. 2019. ‘Un admirateur des philosophes modernes’: The Networks of Swedish Ambassador Gustav Philip Creutz in Paris, 1766–1783. In Edmondson and Edelstein, Networks of Enlightenment, 173–200.

Open Access  This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

PART I

Digital Enlightenment: Representing Big Data

CHAPTER 2

In Search of Enlightenment: From Mapping Books to Cultural History Simon Burrows

This chapter appeared originally in the special issue, “The Digital Turn,” of the Journal for Early Modern Cultural Studies 13, no. 4 (2013): 3–28, doi: https:// doi.org/10.1353/jem.2013.0042. It is reprinted here with permission of the University of Pennsylvania Press. The work it describes is an output from the British Arts and Humanities Research Council (AHRC) grant project The French Book Trade in Enlightenment Europe, 1769–1794 (FBTEE). References to major related publications since 2013 have been added and the text has been expanded, clarified, and updated with reference to subsequent developments in the project it describes. Earlier versions were delivered as a valedictory lecture of the same title at the University of Leeds in November 2012 and as a contribution to a joint keynote presentation to the New Zealand e-Symposium in July 2012. My co-presenter on the latter occasion was Catherine Nicole Coleman of Stanford University’s Mapping the Republic of Letters project. I thank the organizers of these events, particularly Sydney Shep, Jonathan Topham, and Raphael Hallett for giving me the opportunity to discuss my ideas. This chapter also owes a considerable debt to the work and discussions with members of the French Book Trade in Enlightenment Europe team—Sarah Kattau, Amyas (Henry) Merivale, and Vincent Hiribarren, and particularly to Mark Curran, who was my research assistant between 2006 and 2011. His work and contribution are discussed in this chapter.

S. Burrows (*) Western Sydney University, Sydney, NSW, Australia e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_2

31

32 

S. BURROWS

For some commentators, the very notion of digital humanities is almost an oxymoron.1 Computers trade in certainty, in absolutes, in yes/no, on/off dichotomies, in solid data. The humanities, on the other hand, depend on subjective interpretation. History is doubly problematic, requiring imaginative reconstruction of an unknown and unknowable past from the incomplete residues it has left behind. There is no finite, self-contained, and self-consistent corpus of texts, there is no single agreed-upon methodology. There are problems here, then, on both physical and conceptual levels. What material survives? How are surviving archives presented? And how does the written record relate to “actual events”? What events are not recorded, and how far is our impression distorted by what does survive or the ways we approach this material? Moreover, the very phenomena we are treating remain fluid, unsettled. We discuss the past in terms of elusive concepts that evade agreed upon definitions and are open to endless discussion, debate, and re-­examination. This is particularly true in the realm of cultural and intellectual history. To take just one example, what was the Enlightenment? When did it begin? When did it end? Who was involved? What did it change? These are fundamental but unresolved questions. And I could ask the same of the Renaissance, the Scientific Revolution, the Age of Romanticism, and so on. Yet, when designed and used imaginatively, digital methods can provide tools for handling gaps and distortions, while offering multiple or open-­ ended possibilities for analysis and interpretation from a variety of methodological and disciplinary perspectives. This brings me to the French Book Trade in Enlightenment Europe, 1769–1794 (FBTEE) project. Along with a handful of other leading digital humanities projects, most notably Stanford University’s hugely impressive Mapping the Republic of Letters, the FBTEE project, now based at Western Sydney University, is both literally and metaphorically remapping the Enlightenment.2 In the process, it 1  See David Greetham, “The Resistance to Digital Humanities,” in Debates in the Digital Humanities, ed. Matthew K.  Gold (Minneapolis: University of Minnesota Press, 2012), 438–51, https://dhdebates.gc.cuny.edu/projects/debates-in-the-digital-humanities 2  Since late 2013, the FBTEE project has been hosted at Western Sydney University and has been funded by the Western Sydney University (2013–2019) and the Australian Research Council (ARC) grant (2016-present), Mapping Print, Charting Enlightenment. I thank the University of Leeds and AHRC for agreeing to the transfer of the database and website to Western Sydney, where the original FBTEE database can now be consulted at http://fbtee. uws.edu.au/stn/interface/. The database is also available at the same site in downloadable form.

2  IN SEARCH OF ENLIGHTENMENT 

33

can help to transform the way we approach, think about, and do cultural and literary history. To be sure, it takes the existence of a historical-­cultural phenomenon called “the Enlightenment” as a given, locating it chronologically in the long eighteenth century and sociologically in the burgeoning networks of intellectual and cultural exchange of the period. And since these networks can be located in physical space, computer-aided geospatial mapping can help us to analyze them in new and exciting ways. Compared to Mapping the Republic of Letters, the FBTEE project began with quite limited goals and focus. The Stanford project is a long-­ term, open-ended, large team effort that has developed across two decades. It examines both correspondence networks and the topography of international travel across the long eighteenth century, while developing cutting-­edge and inspirational visualizations and analytical tools. In contrast, in its original incarnation and current online public-facing database resource, the FBTEE project looks at the pan-European trade of a single, important Swiss publishing house, the Société typographique de Neuchâtel (STN) and tracks the origins and destinations of almost half a million books that passed through its hands. It draws on a finite, closed dataset and is driven by a tightly defined set of research questions. This has been its great strength—although, as we shall see, it is presently being expanded exponentially and in multiple directions.3 But for now, having introduced the project from several angles, this chapter will attempt to answer four ancillary questions: why did I opt for digital humanities methods? What are the project’s research questions? How has it used the technology to answer them and where does it fall short? What additional tools have we developed, and what can we learn from them? In the process, I wish to show how the FBTEE project has demonstrated both the utility of digital humanities methods, and their potential to answer both established and novel historical questions. I also wish to reveal how it has developed innovative ways to research and handle complex, incomplete, or inconsistent datasets, uncertainty, and ambiguity. Describing the unique challenge and objective of designing digital humanities tools for historical projects, Catherine Nicole Coleman of the Mapping the Republic of Letters project observes that: 3  Further database resources from the later stages of the FBTEE project are expected to be released in 2020 and 2021. By early 2020, the FBTEE project had employed the energies of more than twenty researchers and developers and had become, like the Mapping the Republic of Letters project, a “long-term, open-ended, large team effort.”

34 

S. BURROWS

When you are working with incomplete and imprecise information, the goal is not a reproducible proof but expert interpretation of the evidence … It is not about applying existing models to history. It is about thinking through data at multiple scales, from multiple perspectives, to see historical evidence in a larger context than we could before but also with the ability to navigate the information: pulling out to see the big picture, zooming into the details, working with the given properties of evidence to give it shape and tell a story.4

This is precisely what the FBTEE project set out to do, as the rest of this chapter will show. But, of course, as with any endeavor, I also had more personal reasons for adopting my chosen project and approach. I first dreamed of creating a research database in a very specific career and research context. I wished to study the dissemination and reception of books because I had already done extensive work on print journalism and pamphleteering, so looking at books was a natural progression. Also, because my previous studies had been rather narrow in focus, I wanted to do something more expansive. In the STN, I found inspiration. The STN archive is universally estimated to be the richest source on the international book trade during the later Enlightenment.5 The STN traded in cheap popular editions of books written in French—the elite language of the period—to a market that spanned Europe from Lisbon to Moscow,

4  Catherine Nicole Coleman, “Seeking the Eye of History: The Design of Digital Tools for Enlightenment Studies,” in Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies, ed. Simon Burrows and Glenn Roe (Liverpool: Voltaire Foundation in association with Liverpool University Press, 2020), 245. 5  The most important studies of the STN and its trade are Michael Schlup, ed., L’Edition neuchâteloise au siècle des Lumières: la Société typographique de Neuchâtel (1769–1789) (Neuchâtel: Bibliothèque Publique et Universitaire, 2002); Robert Darnton and Michel Schlup, eds., Le Rayonnement d’une maison d’édition dans l’Europe des Lumières: la Société typographique de Neuchâtel 1769–1789 (Neuchâtel: Bibliothèque Publique et Universitaire, 2005; Hauterive: Editions Gilles Attinger, 2005); and Jeffrey Freedman, Books Without Borders in Enlightenment Europe: French Cosmopolitanism and German Literary Markets (Philadelphia: University of Pennsylvania Press, 2012). Other important studies include Robert Darnton, A Literary Tour de France: The World of Books on the Eve of the French Revolution (Oxford: Oxford University Press, 2018), and the twin volumes resulting from the FBTEE project: Mark Curran, The French Book Trade in Enlightenment Europe I: Selling Enlightenment (London: Bloomsbury, 2018), and Simon Burrows, The French Book Trade in Enlightenment Europe II: Enlightenment Bestsellers (London: Bloomsbury, 2018). The primacy afforded to the STN archive is open to challenge, particularly by the much less wellknown Luchtman’s archive, which is discussed in note 43.

2  IN SEARCH OF ENLIGHTENMENT 

35

Stockholm to Naples.6 Based in the Prussian-ruled Swiss principality of Neuchâtel, it was also able to deal in works that were banned elsewhere on the Continent. I had worked in the STN archives half a dozen times between 1990 and 2004, and I was well aware that they had inspired Robert Darnton to write his masterful Forbidden Best-Sellers of Pre-­ Revolutionary France.7 That book offered a new empirical underpinning to Darnton’s celebrated but controversial argument that the French monarchy was discredited on the eve of the revolution due to the “drip, drip” desacralizing effect of banned free-thinking and pornographic literature.8 I admired Darnton’s attempts to use booksellers’ correspondence to draw up a representative list of underground bestsellers. It was headed by Mathieu-François Pidansat de Mairobert’s Anecdotes sur Madame la Comtesse du Barri (1775), a scandalous biography of the king’s mistress, Louis-Sébastien Mercier’s L’An 2440 (1771), which is a futuristic thought-­ experiment about a sleeper who awakens in the far future to find a morally perfect society, and Thérèse philosophe (1748), a piece of materialist-atheist pornography loosely based on a genuine clerical sex scandal and generally attributed to Jean Baptiste de Boyer, Marquis d’Argens.9 This was decidedly not the literature we were taught in school. Darnton, a superb stylist, knew it. In his hands, this became one of the most seductive, iconoclastic historical materials ever written. This, and his earlier work, inspired a generation of scholars to look at literature well beyond the traditional canon.10 Yet, the narrowness of Darnton’s project frustrated me. While the STN had traded across Europe, Darnton had dissected only a small section of their trade—booksellers’ orders for highly illegal works from just one country. There was no internationally comparative material. Moreover, though he published a whole book of appendices, The Corpus of Clandestine 6  It should be noted here that the dominance of French was contested, particularly in the east of Europe, where German was more widely spoken. 7  Robert Darnton, The Forbidden Best-Sellers of Pre-Revolutionary France (New York: Norton, 1996). 8  Darnton, Forbidden Best-Sellers, 216. 9  According to Angus Martin, Vivian Mylne, and Richard Frautschi, Bibliographie du genre romanesque français, 1751–1800 (London: Expansion, 1977), Thérèse philosophe was first published in an undated edition in 1748 and then again in 1750 and 1760; by 1800, it went through sixteen editions, not including several further non-dated editions. 10  Robert Darnton, “The High Enlightenment and the Low-Life of Literature in PreRevolutionary France,” Past and Present 51 (1971): 81–115, https://www.jstor.org/ stable/650404

36 

S. BURROWS

Literature in France, 1769–1789,11 it was impossible to use his work to trace trends over time. Twenty-five years of literary history were presented as a flat tableau, making it hard to validate his claims for a progressive political desacralization of the French monarchy. Finally, I had methodological doubts. Although Darnton argued that, because there was no system of “returns” for unsold stock, booksellers made sure to match supply and demand, it was clear that the STN did not and could not provide every work requested from them. This mattered greatly to me because, although Darnton recorded a handful orders for political pornographic works against Marie-Antoinette, I had become convinced that, while several such works had been printed before 1789, they did not actually circulate among the public before the French Revolution. Thus, they could not, as many historians asserted, have helped to undermine the ancien régime. My first forays into the STN’s accounts seemed to bear me out.12 A database of the STN’s trade—based on the STN’s accounting records rather than the booksellers’ correspondence used by Darnton—seemed an ideal way to address these issues, while bringing Darnton’s approach into the Digital Age. It would allow me to examine the STN’s whole trade, pinpoint trends in time and space, address my concerns about the limitations of Darnton’s demand-side based methodology, and explore my hypothesis concerning the circulation of scandalous pamphlet attacks on Marie-Antoinette. The basic aim of the FBTEE team has been to map the dissemination of books by the STN in time and space. We wanted to record from where the STN sourced books and to where they sent them. This, in turn, would help us to answer a tight set of research questions, including: how far did demand for French books differ over time and space? What was the volume of the STN’s trade with different regions? Did demand patterns differ between different types of urban centers? Which authors, titles, and sorts 11  Robert Darnton, The Corpus of Clandestine Literature in France, 1769–1789 (New York: Norton, 1995). 12  See Simon Burrows, “The Scandalous History of Marie-Antoinette,” in Blackmail, Scandal and Revolution: London’s French Libellistes, 1758–1792 (Manchester: Manchester University Press, 2006), esp. 147–52 and, for the STN evidence, 150–51. See also Darnton’s case study of the bookseller Mauvelain in Robert Darnton, “Trade in the Taboo: The Life of a Clandestine Book Dealer in Prerevolutionary France,” in The Widening Circle: Essays on the Circulation of Literature in Eighteenth-Century Europe, ed. Paul J.  Korshin (Philadelphia: University of Pennsylvania Press, 1976), 11–83.

2  IN SEARCH OF ENLIGHTENMENT 

37

of books were bestsellers and where? What proportion of book sales was associated with the Enlightenment in each place? How extensive was the illegal book trade? And how far would the database be able to assist us in resolving bibliographic mysteries? What I had dreamed up was, in fact, a complex relational and interpretational database (Fig.  2.1). It sought to relate datable events (mostly sales, acquisitions, and stock takes) to the books traded and the clients who traded them. Thereafter, data on books would be enriched by information on the precise editions traded (where that could be identified), subject taxonomies, and the legal status of books. Client information would be enriched by information on places of residence, professional affiliations, and gender. The level of subjective interpretation involved in our data editing and enrichment processes might appear anathema to database purists; yet, it is precisely this data enrichment necessarily accompanied by explanatory notes, interpretative tools, and query options that make the database such a powerful instrument of historical, literary, and bibliographical analysis. To help users unravel the multiple layers of interpretation and hence uncertainty in the database, we have developed original means of addressing the problems of representing complex, incomplete, and sometimes treacherous historical data in digital form. I have therefore coined the term “interpretational-relational database” to describe the FBTEE, in the hope that this concept may prove useful to others. The central spine of this “interpretational-relational database” is the day-to-day record of book trade transactions from the STN accounts. As every record is dated, we can pinpoint transactions in time down to a particular day. Of course, there is something approximate about those dates: they record the days under which the STN recorded the receipt of the dispatch of books. Quite how those dates reflect the physical movement of

Client data Profession Gender

Named Persons Place

Events data Clients

Archival Correspondence

Datable Events (transactions)

Data Source

Fig. 2.1  Core aspects of the FBTEE database structure

Book data Superbooks

Illegality markers Editions Taxonomy

38 

S. BURROWS

stock which took time to pack and ship and could spend many weeks in transit is not entirely clear. However, in essence, we have developed a self-­ consistent way of dating the transactions that interest us that allows us to pinpoint events very precisely. This means that we can graph trends over time. However, the results are problematic if we do this by raw figures, for the STN sales graph was not flat (Fig. 2.2).13 So, to compensate for this shortcoming, we provide a way to measure sales as a proportion of all sales in a given year. Locating events in space proved to be a bit more problematic. In theory, each transaction is tied to a single client, and each client has a single place of residence. But, again, there is some approximation in this. The dissemination patterns we record cannot reveal whether a book remained in the place to which it was dispatched or not. Nor can we always be sure whether books were sent to clients at their place of residence: in some cases, we know this was not the case because occasionally some of our sources record this information. Other sources, however, do not. There is an assumption, then, in our figures that most books stayed in or close to the place where they were received, and that, unless specified otherwise, those books were delivered to a client’s hometown.

Fig. 2.2  The STN sales graph (group stacked bar chart, 1770–1794)

13  The sales graph in Fig. 2.2 and other visualizations used in this chapter (save the simplified database map, which is my own work) were developed for the project by Vincent Hiribarren.

2  IN SEARCH OF ENLIGHTENMENT 

39

Defining space in geo-political terms was also challenging, even though political borders remained stable for most of our time period and major changes in 1772, and between 1790 and 1794, did not, fortunately, impact on our dataset. Even so, locating ancien régime geo-political space is notoriously difficult. The complex and overlapping jurisdictions that existed across a Europe of fragmented states are a cartographer’s nightmare. I concluded that the best way to resolve them was a “two-tier plus” system. Every town in the database was ascribed to a single sovereign territory, say France, or the Papal States, and then to an administrative subunit, such as Burgundy or Avignon. The precise nature of these provincial units necessarily varied from state to state, but, where possible, the subunits chosen were related to the administration of the book trade. The database also records whether a town belonged to any of a series of further units—the Holy Roman Empire, ecclesiastical territories, imperial-­free cities, university towns, and so on. Large empires of non-contiguous states were also broken up for mapping and statistical purposes. For example, Prussia is divided into three parts: Prussia proper, Prussia: Cleves, and Prussia: Neuchâtel (for the home of the STN was technically a Prussian principality, though geographically speaking in Switzerland; see Table 2.1). We have also included mapping options centered on modern upper and lower territories, so that users can link the eighteenth-century book trade to the shape and histories of contemporary states or provincial units, or to large “geographic zones” comprising national or super-national regions as aids to macro-level analysis. So, in all, besides mapping towns, users can record their results on five different base maps. Each map offers a different, but legitimate, representation of a complex geo-historical reality. Giving users multiple options in how they represent the data is one of our key strategies in designing the database. Aware of the complexities, biases, Table 2.1  Placing Neuchâtel geo-politically through the STN database Territorial type

Unit incorporating Town of Neuchâtel

Eighteenth-century lower territory Eighteenth-century sovereign territory Twenty-first-century administrative unit Twenty-first-century country Geographic zone Ecclesiastical land Holy Roman Empire

Neuchâtel Prussia: Neuchâtel Neuchâtel [i.e., Swiss canton] Switzerland Helvetic No No

40 

S. BURROWS

gaps, and ambiguities of our data, we have chosen to alert users to problematic issues and allow them to choose from a range of solutions. Thus, they can shape the data or its representation in ways appropriate to their purposes. For example, when it came to categorize the content of the books, we had to choose among a range of systems and philosophic approaches. Should we choose a “tree system,” where every work had a single place and where statistics could be presented as percentages of the whole? Or should we use a keyword system, with multiple descriptors for each book? Should we adopt an “off the peg” system, like the Dewey or Library of Congress systems? Or should our chosen system be adapted to fit the material? If so, should we design our own system or use an eighteenth-­ century taxonomy? The solution eventually adopted considers all these issues. Every book handled by the STN is categorized twice: once by the eighteenth-­century tree system of the Parisian booksellers (see Fig. 2.3) and once by a purpose-built project keyword system.14 We did something similar when trying to define which books were illegal. This is an impossible task under old regime jurisdictions, particularly in France, due to conflicting systems of legal and quasi-legal, tolerated books. Illegality covered a spectrum. At one end were books so innocuous that their publishers did not bother to apply for a license, or pirated editions of legally licensed works. At the other end were books that the police actively rooted out: pornography, atheistic works, and books deemed harmful to public morals. In between was a vast and uncertain middle ground. Even the authorities were often not sure what was permitted and what was not, particularly as the State, Church, and University of Paris all had powers to censor or condemn books. Our solution was to provide users with a series of “markers of illegality” and let them choose among them. These markers include books to be found in Darnton’s Corpus of Clandestine Literature, works found on the Versailles bookseller Claude Poinçot’s 1790 inventory of the secret dépôt in the Bastille, or works on the Austrian emperor Joseph II’s index of banned books.15  I would like to thank the staff of the Brotherton Library at the University of Leeds, particularly Jane Saunders and Chris Sheppard, for discussing cataloguing systems with me and pointing me to key literature, particularly Edward Edwards, Memoirs of Libraries: Including a Handbook of Library Economy, 2 vols. (London: Trübner, 1859). 15  Poinçot’s list of the books in the Bastille dépôt is reproduced in Robert L.  Dawson, Confiscations at Customs: Banned Books and the French Book Trade During the Last Years of 14

2  IN SEARCH OF ENLIGHTENMENT 

Unknown

Phys hys ysique Uncategorisable

Théologie hétérodoxe verses oeuvres théo ologiques Métaphysique Métap

Jurisprudence

Mathématique hé

Médecine, Chirurgie

Rhétorique

41

Grammaires & dictionnaires

Arts Philosophie

Catechétiques Polygraphes anciens & modernes

Histoire naturelle

Science et Arts

2

1

Belles-Lettres

La Sainte Bible & commentaires

Poétique

Atlases Histoire ecclésiastique

Droit canonique

Droit français

Histoire

Géographie

Voyages

Histoire de France

1 = Théologie 2 = Undetermined

Bubbles of Parisian Categories

Romans Droit Etranger

Histoire universelle & particulière

Histoire européenne divers

Treemap of Parisian Categories

Fig. 2.3  Visualizing the Parisian taxonomic system and the relative importance of different keywords

We have similar approaches to authorship. As literary scholars are well aware, defining authorship is so problematic that some postmodernists do away with the concept altogether. Certainly, who exactly participates in the creation of a single text is open to debate. Multiple roles and processes are involved before a text is written down; many more intervene before the text finds its way into print. Writers, secretaries, and transcribers, critics and reviewers, proofreaders, editors, typesetters, publishers, translators, and commentators can all be involved in the production process. Therefore, we felt compelled to capture and address some of that complexity in the database. In database terms, we have treated as an author anyone who was, according to our bibliographic sources, significantly involved in the creation of a piece of text that is included in a work. This “involvement” might include many possible roles, but we have reduced them to four. A primary author is the original, sole, or main writer of a piece. A secondary author is one who, while not the primary author, played a significant role in creating the text. That person might be a contributor or a collaborator, a person who adds some extra material, illustrations, or a gloss, or the Ancien Régime (Oxford: Studies on Voltaire and the Eighteenth Century, 2006), 7. For Joseph, see Joseph II, Ordonnance ultérieure de l’Empereur et Roi: du 20 mai 1788, concernant les libelles ou ecrits satyriques, diffamatoires, scandaleux ou séditieux; de même les estampes, ou caricatures & les mauvais propos de cette espèce (Bruxelles: Pauwels, 1788).

42 

S. BURROWS

someone whose work is included in an anthology or collection of essays. Then, we have editors of texts, and, finally, there are the translators. Even so, we still had a large number of conceptual problems to resolve. For example, defining and describing books and then grouping their many editions into unitary “superbooks” (or “works”) brought up multiple and complex problems.16 For example, should a text that was radically rewritten and expanded by a new author be treated as a separate work from the original? Or what combination of title, editorship, place of publication, format, and serial number, all of which could be subject to change or interruption, gave it an essential unity? And how should we treat texts which constituted parts of greater, unitary multi-author works? This last question was particularly problematic, since it applied to the most iconic and bestselling text of Judeo-Christian civilization, the Bible. This was often sold in separate sections (e.g., The New Testament or just the Gospels), in individual books (most notably the Psalms, which could also be published with or without music), and in multiple versions containing variant texts (e.g., some Catholic Bibles contained the Apocrypha, whereas Protestant and Hebrew scriptures invariably did not). These issues of “superbook” definition were so varied that we resolved them on a case-by-­ case basis. We treated the Bible itself as a unique instance, grouping all parts and versions in common not under a single work but under a single unique “author” which, after some debate, to avoid controversy and from respect for the manes of Voltaire, Diderot, and d’Holbach, we decided to name “[Bible]” rather than “Holy Spirit.” We do not attribute specific human authors to Biblical texts, even where, as in the case of Paul’s letters, their authorship appears beyond dispute. We also needed to establish what the STN treated as “books” in their accounts and determine whether we would do the same. Here, the sources dictated our approach: since we were trying to count the “books” the STN bought and sold, our definition of a book was any printed item the STN’s general accounts treated as a commercial product, including serial publications, almanacs, editions commissioned by authors or other publishers, as well as all manner of books and pamphlets.17 Then, we hit several really difficult questions. How should we deal with and count part 16  Jonathan Topham from the Science in the Victorian Periodical [SciPer] project introduced us to the “superbook” concept. I am grateful to him for this key suggestion. 17  For a longer discussion of what constituted a book for our purposes, see Burrows, Enlightenment Bestsellers, 17.

2  IN SEARCH OF ENLIGHTENMENT 

43

sales of works sold in instalments or as multi-volume sets? Were translations part of the same “superbook” as editions produced in the original language?18 Moreover, how should we deal with title changes, especially where content was changed, extended, or revised? And what about instances where works were (re)printed as part of larger works? Also, when did a radically revised edition become a separate work? These last three questions often needed to be answered on a case-by-case basis, following precedent. Beyond them lay further questions requiring us to establish protocols to govern further data fields: how should we deal with known and suspected false imprints? How and with what degree of certainty could we identify editions from incomplete or ambiguous data? Furthermore, as we had to reconstruct the STN’s trade in different periods from a variety of different sources, which had been created for different purposes and contained information of differing quality and form, we had to find ways to compensate for uneven sources and occasional gaps in the data. We drew our data primarily from the STN’s daybooks, rolling stock books, and order books, and we supplemented this data with other sources, including, in periods where no other evidence survived, booksellers’ letters. These problems raised further issues about the comprehensiveness and representative value of our data, particularly over time. Users are not always comparing like with like, and for one year, unfortunately, the great year of the 1789 Revolution, we had no information at all (Fig. 2.4). The implications of the final three of these issues are worth reflecting on here. As we had uneven sources and gaps in our data, we had to find ways to compensate for data distortion, particularly if we wanted to make claims for the representative value of our outputs. But how were we to do this? Early in the project, I realized that, for reasons of data integrity and consistency, we would need a way of isolating information gleaned from different sorts of source material. Therefore, data interface designer Amyas (Henry) Merivale and I developed an “option” menu which allowed users to exclude or include data on the basis of its original source and explain why they might wish to do so. This would allow them to use customized filters on our data. Later, we realized that this idea could have wider applicability. We could also limit our dataset by markers of illegality, client type or client gender,

 Our answer was yes, as long as it was a genuine and approximate translation.

18

44 

S. BURROWS

Fig. 2.4  Sources and gaps: a calendar of the daily volume of transactions drawn from various STN sources (1773–1778)

types of editions, or the original language in which a work was written.19 So, that is what we did. As a result, users are spared the need to know how to run complex database queries in our chosen software, MySQL. They also do not need to develop a detailed knowledge of our data structures and consult an extensive user manual. Instead, they can conduct their queries through our interface, which has a user-friendly and largely intuitive presentation.20 Used intelligently, the options we have incorporated are powerful tools for specialized queries or for ironing out biases and distortions. For 19  As I have noted elsewhere, Mark Curran deserves credit for the idea of extending the options to deal with illegality. The power of the idea struck us both immediately, and within minutes we had conceived the remaining options menus. For details, see http://fbtee.uws. edu.au/main/project-history/ 20  See Robert Darnton’s review of The French Book Trade in Enlightenment Europe, 1769–1794 for the Institute of Historical Research’s Reviews in History (2012) at http:// www.history.ac.uk/reviews/review/1355, and Jeremy D.  Popkin’s review, “French Book Trade Database,” for the British Society for Eighteenth-Century Studies (2012) at https:// www.bsecs.org.uk/criticks-reviews/french-book-trade-database/

2  IN SEARCH OF ENLIGHTENMENT 

45

example, we can filter out clients whose trade patterns are anomalous. These might include booksellers like Malherbe of Loudun in western France. In the late 1770s, Malherbe bought up cheaply a huge lot of mostly illegal books that the STN had been unable to shift and sold them across northwestern France via a gang of traveling colporteurs (“peddlers”).21 Loudun, therefore, looms large in a simple map of France’s illegal trade with the STN, but this is quite misleading. In the eighteenth century, Loudun was a small, sleepy, provincial town, so there is no reason to consider it particularly bookish. It was a mere historical accident that tied Loudun to the sole mass “dumping” of books in the STN’s entire history. We might, therefore, consider Loudun atypical and so choose to exclude Malherbe from our queries, along with a handful of other “foreign wholesale clients,” that is, entrepôt dealers who forwarded on to distant places a large proportion of the books they received. This use of our “client type” option quite literally wipes such anomalies from our maps (Fig. 2.5). How far can our dataset, based on the archives of just one publisher, stand proxy for the entire European trade in French books? Darnton argues—and others take him at his word—that, because the major publisher-­wholesalers of books swapped works among themselves, the STN could source anything from anywhere. Also, because there was no system of “returns” for unsold stock, booksellers were “cautious” about what they ordered: demand very closely matched supply.22 This appears to

Fig. 2.5  Sales of French illegal bestsellers between 1770 and 1789, with and without foreign wholesale clients included (note how Loudun and Amsterdam disappear)  Malherbe’s use of peddlers was first remarked in Darnton, Forbidden Best-Sellers, 16.  Darnton, Forbidden Best-Sellers, 56–57.

21 22

46 

S. BURROWS

suggest that the STN’s trade was very little different from that of a Dutch publisher trading in French books: Darnton insists that all publisher-­ wholesalers operating along the French border from the Low Countries to Switzerland could draw on “a kind of invisible floating stock.”23 Our geospatial analysis challenges this assumption, and not only because we discovered significant numbers of “returned books” in our data (though, in all probability, few of these were mere unsold stock). A moment’s thought about the time constraints, distance, and costs involved in sourcing books in the eighteenth century would surely have raised questions about the idea that Dutch and Swiss markets were largely interchangeable. Indeed, in his review of the FBTEE database, Darnton clarified that he did not mean to go so far as to suggest this.24 Our data offers some support for the separate markets hypothesis: it suggests that the STN dealt primarily in Swiss books.25 The reason for this is that they were cheaper and faster to stock.26 In all, Swiss books seem to have accounted for over 90 percent of all books sold by the STN (Table 2.2).27 Nevertheless, the argument can easily be overstated. Once the STN’s own editions, which comprised around two-thirds of their trade, are removed from the equation, some 30 percent of their unit sales were books that appear to have originated outside Switzerland.28 Among the  Ibid., 55.  Darnton, review of the FBTEE. 25  This total includes the so-called Swiss “international” editions (i.e., joint editions published by more than one publisher in more than one country), where one partner was based in Neuchâtel. STN joint editions are not included here. 26  For more about the implications of Swiss books, see Simon Burrows and Mark Curran, “How Swiss Was the Société Typographique de Neuchâtel? A Digital Case Study of French Book Trade Networks,” Journal of Digital Humanities 1, no. 3 (2012): 56–65, http:// journalofdigitalhumanities.org/1-3/how-swiss-was-the-stn-by-simon-burrows-and-markcurran/, and especially Mark Curran, “Beyond the Forbidden Best-Sellers of PreRevolutionary France,” Historical Journal 56, no. 1 (2013): 89–112, https://doi. org/10.1017/S0018246X12000556. The contention that Swiss books made the bulk of the STN’s trade is powerfully made by Curran’s article, but the analysis offered here suggests that significant caveats to his argument must remain. The dialogue around this issue continues across Curran, The French Book Trade in Enlightenment Europe I, and Burrows, The French Book Trade in Enlightenment Europe II. 27  Table 2.2 indicates that around 370,000 out of 410,074 books traded by the STN were “Swiss editions.” 28  The figures are slightly different for books “bought in” by the STN where, if we assume similar proportions of foreign books among the works whose place of publication is unknown to those whose place is known, it appears that around 27.8 percent of the works the STN 23 24

2  IN SEARCH OF ENLIGHTENMENT 

Table 2.2 Books traded by the STN, by edition type, 1769–1794 (all known “sales”)

Edition type STN editions, all types Other Neuchâtel editions Other Swiss editions French editions Belgian and Dutch editions British editions German editions Place of publication unknown All other editions Total (all sales, rounded)

47

Total copies traded 273,116 6364 42,037 9237 4814 2491 1756 68,116 1292 410,074

Several clarifications are needed for this table. The “Swiss edition” total includes the so-called Swiss “international” editions (i.e., joint editions published by more than one publisher in more than one country), where one partner was based in Neuchâtel. STN joint editions are not included here. British figures are particularly suspect given the high numbers of works using London as a false imprint. Some of these may have gone undetected by the bibliographical and catalog sources we have used. Of the 66,329 books disseminated by the STN whose place of publication is unknown but whose supply origin has been traced, 46,082 were sourced from Switzerland (the Helvetic zone), 13,907 from France, and the rest from across Europe. None were sourced directly from Britain, Russia, or Poland. The “All other editions” grouping comprises Avignon-sourced, Russian, Scandinavian, and international editions

STN’s outgoing trade, the figures are more striking still. Over 60 percent of the books they dispatched went directly to clients abroad. Of this foreign trade, which involved almost a quarter of a million units, only one in eight (12.6 percent) went to destinations within a two hundred-kilometer radius of Neuchâtel.29 In contrast, over 51,000 units were sent to markets as distant as the Iberian Peninsula, Russia, Poland, Scandinavia, Britain, Belgium, or the Netherlands. Clearly, then, a substantial proportion of the “bought in” were foreign editions (in all, this amounted to about 27,400 foreign-sourced and 16,350 Swiss-sourced books). Several thousand supposedly STN editions acquired from, or printed by, STN clients are excluded from these figures. 29  Measurements taken in Google Earth from Place Pury, Neuchâtel. Out of 246,519 units, 31,072 were known to have been sent abroad less than 200 kilometers away. A further 15,000 units altogether went to Lyon, 211 kilometers away, and Turin, 221 kilometers away.

48 

S. BURROWS

books flowing through the arterial networks of Europe’s French wholesale book trade comprised long-distance international trade. Nevertheless, the STN tended to trade individual foreign editions in low volumes. This no doubt reflects that, as they traveled further from their point of origin, wholesalers tended to break up stock consignments into smaller bundles before selling them on. If the STN’s trade is typical, the international “floating stock” theory appears to have a certain residual validity, despite the overwhelming significance of Swiss-produced books to the STN’s total trade statistics. A first glance at the author networks suggests that the STN was also peddling a distinctly Swiss-Parisian Enlightenment (Fig. 2.6). This is clearly borne out in some domains. In science, the main authors STN sold were Swiss luminaries: Samuel Tissot, Charles Bonnet, and Albrecht Haller. A vogue for Linnaeus passed the STN by: hardly a single work of Linnaeus or Linnaean science passed through their magazines.30 In religion, the vast majority of books sold were Protestant works, mostly by Swiss or French Protestant divines. From this data, we might conclude that impermeable intellectual borders existed across much of Europe. Yet, in other domains, the STN’s trade shows remarkable levels of cross-­ cultural transfer. For example, almost 10 percent of the STN’s unit sales were translated British or German works, mainly novels and travel writings/geography books.31 Then, if we turn to what sorts of books were sourced from where, we notice a remarkable homogeneity. 30  I am grateful to Mark Curran for first drawing my attention to the absence of Linnaeus and, as far as we have yet discerned, of Linnaean science, from the database. 31  The STN’s three top-selling translated works were Patrick Brydone, A Tour through Sicily and Malta: In a Series of Letters to William Beckford, Esq. of Somerly in Suffolk (London, 1773) with 1837 sales; Fanny Burney, Cecilia (London, 1782) with 1625 sales; and The Triumph of Benevolence; or, The History of Francis Wills (London, 1772) with 1568 sales (this book has been attributed to both Oliver Goldsmith and Arthur Murphy). These were published by the STN in editions entitled Voyage en Sicile et à Malthe (2 vols., 1776); Cecilia, ou Mémoires d’une héritière par l’auteur d’Evelina; traduits de l’anglois (5 vols., 1783); and Histoire de François Wills, ou le Triomphe de la bienfaisance (2 vols., 1774). The top-selling works translated from German included Anton-Friedrich Büsching’s Vorbereitung zur grundlichen und nützlichen Kenntniss der geographischen Beschaffenheit, published by the STN in 1779 as Introduction à la connaissance géographique et politique des Etats de l’Europe with 1543 sales; Salomon Gessner’s Oeuvres with 933 sales (including a STN edition published in 1776); and Friedrich Nicolai’s Das Leben und die Meinungen des Herrn Magister Sebaldus Nothanker, sold by the STN in various editions including their own (4 vols., 1779) under the titles L’Intolérance ecclésiastique, ou Les malheurs d’un hétérodoxe and La vie et les opinions de maître Sébaltus Nothanker, with 899 sales.

2  IN SEARCH OF ENLIGHTENMENT 

49

Fig. 2.6  Places of residence of men of letters corresponding with the STN

Analysis of the sourcing of works by subject keyword suggests that the same major themes and subjects were being produced everywhere in broadly similar volumes (Table  2.3). The anomalies over religion (the STN traded mainly Protestant works) and philosophie (which was more dangerous to produce in some states than others) are understandable, while the STN’s apparent lack of interest in scientific publishing seems to have been down to local editorial policy. This, in turn, suggests a broader representative value to the STN archive. If mined judiciously, it can be used to measure and compare reading tastes. Hence, I would suggest that the STN database offers tools for geospatial and temporal analysis that can be used to take the pulse of the later Enlightenment. Dorinda Outram has conceived of the Enlightenment as a set of “capsules of debate” or “flashpoints, where intellectual projects changed society and government.”32 The STN database allows us to track the lifeblood of many, if not all, of those debates, as well as some wider intellectual and reading trends. It permits us, for example, to track the rise 32  See Dorinda Outram, The Enlightenment, 3rd ed. (Cambridge: Cambridge University Press, 2013), 3.

50 

S. BURROWS

Table 2.3  Distribution of keywords by edition type. Commissioned editions and editions whose place of publication is unknown are excluded Keyword Literature Prose fiction Politics Philosophie Science Religion France Current affairs Work of religiosity Christianity

Swiss non-STN editions STN & allied editions Foreign editions All 30.4 21.8 20.2 17.8 16.7 16.5 15.2 13.1 13.0 12.7

28.8 17.8 17.5 9.1 4.9 19.9 18.1 14.9 15.3 16.4

32.7 24.2 18.9 10.2 8.57 12.3 16.2 13.3 9.2 8.6

29.5 19.3 18.2 11.0 7.7 18.5 17.3 14.4 14.3 14.9

Fig. 2.7  Leading STN authors of the 1780s by volume of works traded

of a new wave of writers in the 1780s and measure their shifting concerns (Fig.  2.7). Or, it allows us to trace the dissemination of English and German novels in French translation (Fig. 2.8). Such works offer a fascinating opportunity to map cultural transfers in late eighteenth-century Europe. It also allows us to explore sales trends for Christian writers and

2  IN SEARCH OF ENLIGHTENMENT 

51

Fig. 2.8  Sales destinations for works originally published in German or English and traded by the STN

their atheist and irreligious foes (Fig. 2.9).33 Most valuably, the STN database allows us to add geospatial nuance to our findings. We can, for example, sense the heat and excitement generated by the French finance minister Jacques Necker’s sensational publication of government accounts in France and his native Switzerland. These discussions dominate our bestseller tables in 1781, supplying our top six bestsellers for the year.34 33  The volume of early exchanges highlighted here reflects, perhaps, the heated debate surrounding the publication of d’Holbach’s Système de la nature (1770), but it may also exaggerate the significance of the years 1770–1774, since the STN published the first edition of d’Holbach’s work and several refutations of it. For more information on these exchanges, see Mark Curran, “Mettons toujours Londres: Enlightened Christianity and the Public Sphere in Pre-Revolutionary Francophone Europe,” French History 24, no. 1 (2010): 40–59, https:// doi.org/10.1093/fh/crp073 34  The top six works on the STN’s bestsellers list for 1781 were either penned by Necker or related to the financial debates he sparked off. They were, in descending order of sales, Mémoire donné au roi par M. Necker en 1778 (STN edition, 1781); Charles-Alexandre de Calonne, Les Comments (STN edition, [1781]); Philippe-Henri Grimoard, Lettre du marquis de Caraccioli à M. d’Alembert (STN joint edition with Société typographique de Berne, 1781); Observations modestes d’un citoyen sur les opérations de finances de M. Necker, attributed to Pierre-Augustin Robert de Saint-Vincent (STN edition, 1781); Louis-Sébstien

52 

S. BURROWS

Fig. 2.9  Annualized trade in religious and irreligious works as a proportion of total STN sales

However, their sales hardly register in Europe beyond the Bourbon realm and Necker’s own patrie.35 The STN database also enables us to reconstruct the popular market for illegal books in France and compare it to demand elsewhere (Table 2.4). This exercise reveals an apparent predilection among French readers for more and harder pornography than among French-speaking readers

Mercier, Le Philosophe du Port-au-Bled (STN edition, 1781); and Jacques Necker, Compterendu au Roi (STN edition, 1781). For most of these pamphlets, all or most of the copies sold by the STN seem to have been STN editions. 35  For example, the top ten sales destinations for Necker’s Compte-rendu, accounting for over 80 percent of sales, were all in France or Switzerland, despite the existence of many other editions produced inside France. Equally, at least 97.55 percent of copies of Calonne’s Les Comments were sold to destinations in France or Switzerland. In all, out of 14,506 book sales assigned the keywords “Necker” and “Financial Administration” in the database, 7392 went to France (including Avignon) and 6071 to Switzerland (including Neuchâtel, Geneva, and other perpetual allies of the Swiss Confederation).

2  IN SEARCH OF ENLIGHTENMENT 

53

Table 2.4  Keyword sales of “illegal” works as a percentage of all “illegal” sales, by country, highlighting categories where French took at least 10 percent more than other regions (in bold) or those where France took 50 percent more (STN editions excluded) (bold plus italic). NB: “Illegal” is defined here as works appearing in Darnton’s Corpus of Clandestine Literature in France, 1769–1789 French rank

Keyword

1 2 3 4 5 6 7 8

Literature Philosophie Erotic works Prose fiction Politics France Religion Amorous adventures Christianity Anti-clerical works Current affairs Skeptical works Pornographic works Libertine texts Satirical works Lives and letters Despotism Biography History Scandal

9 10 11 12 13 14 15 16 17 18 19 20

Total copies in dataset

% of French “Illegal” sales

% of Swiss “Illegal” sales

% of other “Illegal” sales

Total sales of “Illegal” works assigned keyword

46.49 39.95 35.45 33.96 29.48 29.38 28.45 25.35

40.93 44.17 21.3 32.53 36.57 42.8 18.61 20.92

48.74 44.43 24.75 39.31 29.86 31.97 21.69 20.36

16,672.3 15,058.7 10,652.9 12,779.9 10,907.8 11,440.1 8854.9 8218.6

23.68 23.20

13.72 10.41

16.13 13.91

7041.5 6517.9

22.68 22.64 22.39

29.33 13.34 5.97

20.84 14.25 13.51

8165.8 6588.9 6119

21.7 17.87 16.17

10.12 9.69 23.23

18.30 13.81 13.76

6800 5498.5 5781.8

12.84 12.61 12.20 11.90

16.66 15.55 10.87 13.21

8.63 8.44 9.86 10.64

4212.5 4095.5 3998.6 4149.6

18,227

4438

13,047

35,712

54 

S. BURROWS

elsewhere in Europe.36 Part of the libertine, free-thinking literature that Darnton and others associate with the origins of the French Revolution, such works do seem to have been more widely read in France than elsewhere. Whether they played a role in revolutionary causation is another question: it is possible that their deeper penetration into the French culture merely reflects a preference among readers of erotica for texts in their own language.37 Moreover, these figures also show that other sorts of subject matter were also popular. Above all, the high-minded, skeptical, rational, reformist branch of Enlightenment thought that we have labeled “philosophie.” These were heavyweight intellectual works. Darnton’s emphasis on pornography and scandal has tended to marginalize these more serious works, as has his claim that ordinary readers were unable to “assimilate” Rousseau.38 In contrast, our figures suggest that ordinary readers were demanding cheap editions of Rousseau, Voltaire, d’Holbach, and others in great numbers. So, the “high Enlightenment” canon is worth another look. Spatial-temporal analysis of sales has uncovered a much bigger truth about the most toxic of illegal book sales to France. These books were squeezed out of the French book trade after a 12 June 1783 decree, which ordered all book imports entering the country from abroad to be sent to Paris for inspection by the Parisian book guild (Fig. 2.10). This greatly added to the costs of importation and intensified the Parisian guild’s stranglehold on the French trade at the expense of its 36  Here and throughout this chapter the terms “hard pornography” or “harder pornography” are used to indicate points on a spectrum of sexually explicit writings. In contrast to the common and current usage of these terms, they do denote extreme, exploitative, and abusive visual and film materials, although at the harder end of the spectrum eighteenth-century libertine literature frequently depicted sexual exploitation of women, children, and weaker men by dominant males. Here and in the database, “pornography” is defined, following Lynn Hunt, as a sub-class of “erotica” that depicts genitalia or sexual acts for the purpose of arousal. Previous authorities have not always been so careful in their definitions around this topic, leading to sensationalism and misconceptions about the nature of pre-Sadian erotica and the prevalence of pornography before the revolutionary era. 37  For a more detailed analysis of these issues of the data in Table 2.4, see Simon Burrows, “French Banned Books in International Perspective, 1770–1789,” in Experiencing the French Revolution, ed. David Andress (Oxford: Studies on Voltaire and the Eighteenth Century, 2013), 19–45, and Simon Burrows, “The Anatomy of the Illegal Sector,” in Burrows, French Book Trade in Enlightenment Europe II, 119–36. 38  Darnton, “High Enlightenment,” 110.

2  IN SEARCH OF ENLIGHTENMENT 

55

Fig. 2.10  The changing pattern of the STN’s trade in illegal pornographic works listed in Darnton’s Corpus of Clandestine Literature, 1769–1787

regional competitors. Previous treatments of French censorship and policing of the book trade in the period had suggested they were generally ineffective and that, at most, the 12 June decree was temporarily disruptive. Our database, however, showed something rather different. The decree permanently crippled the STN’s illegal trade with France and forced it to seek markets elsewhere.39 This opened interesting possibilities about revolutionary causation. Many historians have argued in recent years that the revolution was, in large part, the product of a rising public sphere that had slipped beyond control. Our research suggested that, in the run up to the French Revolution, the government’s control was all too tight and growing tighter. The same may well have been true of its influence over the newspaper press, including the popular and well-informed extra-territorial newspapers that composed one third of the French market in the 1780s.40 Was the elite rebellion that preceded the general revolt in 39  This point and its implications have been discussed by members of the FBTEE project team at several conferences, and variously in print by Burrows, “French Banned Books,” Curran, “Beyond the Forbidden Best-Sellers,” and Louise Seaward, “Censorship through Cooperation: The Société typographique de Neuchâtel (STN) and the French Government, 1769–1789,” French History 28, no. 1 (2014): 23–42, https://doi.org/10.1093/fh/crt086 40  There is an extensive literature on these papers. For an introduction, see Simon Burrows, “The Cosmopolitan Press, 1759–1815,” in Press, Politics and the Public Sphere in Europe and

56 

S. BURROWS

part, at least, a reaction to this increasing control? This is a highly revisionist reflection and merits further investigation. Some will doubtless argue that comparative sales statistics may be a useful tool for understanding the economics of the book trade, but they remain a crude and reductive means for studying the influence of texts. I would counter that they remain a useful indicator nonetheless, and sometimes they are the best tool at our disposal. When combined with computer-­aided geospatial temporal mapping, as in the FBTEE project, they have revealed new patterns and provided useful insights into the cultural life and politics of pre-revolutionary France. As I have shown here, this approach allows us to rewrite our understanding of the eighteenth-­ century book trade, re-evaluate the greatest of European cultural movements, and, just possibly, thrown new light on the cultural origins of the most paradigmatic of all revolutions. But this is not the end of the story, it is just the beginning. The FBTEE project has so far yielded up the secrets of only one Swiss bookseller’s archive. How far this can ever be representative of the wider trade remains a big problem haunting the whole project. With a data sample representing perhaps only 0.5–1 percent of the entire French book trade, there is plenty of scope to add more data, both from other booksellers’ archives and other types of sources representing the reception and dissemination of books. First, the FBTEE database could be expanded by incorporating other comparable publishing sources, such as the Leiden-based Luchtman publishing company archives, which with surviving records spanning the entire eighteenth century is arguably a richer source than the STN archives. Likewise, the daybooks of the Paris-based Veuve Desaint survive for much of the period from the 1760s to the 1780s.41

North America, 1760–1820, ed. Hannah Barker and Simon Burrows (Cambridge: Cambridge University Press, 2002), 23–47. 41  The Luchtman company archives, held in the University Library in Amsterdam, were published in microform by MMF Publications between 1992 and 1998 under the title Luchtmans archief. Thanks to a Metamorphose grant and supplementary funding from Western Sydney University, they are now available online at https://luchtmansarchive. com/. The Desaint Archive is held in the Bibliothèque historique de la ville de Paris and has been studied by a team under Sabine Juratic. I thank Erik Jacobs and Wallace Kirsop for directing me to these archives. I am currently working with colleagues in the Netherlands toward opening up both archives through bibliometric digital resources.

2  IN SEARCH OF ENLIGHTENMENT 

57

On a more ambitious level, the FBTEE project always aspired to develop data structures and data entry tools capable of facilitating other bibliometric projects charting the dissemination and reception of books. The sources for such projects are abundant. They include publishers and booksellers’ records, book dealers’ catalogs, catalogs and acquisition and borrowing records of public and private libraries, customs, confiscations, publishing permission, and censorship registers, book reviews and advertisements, reading journals and commonplace books, remarks in private correspondence, and indices of banned books. Material from several such sources already exists in the FBTEE database and is associated with individual books in several database fields.42 However, to integrate this data into a unitary bibliometric survey, we will need to restructure these “events” on the model of our book trade “transactions,” giving each a date, a place, and a value.43 A first token step in this direction was to relabel the “transactions” field in the database as “events” shortly prior to its publication. An events-centered approach would enable users to chart the reception histories of individual titles, authors, genres, or themes in a single statistical query. However, the data presented might also incorporate qualitative data, including personal responses to reading, as found in the Reading Experience Database (RED).44 As has been noted elsewhere, this amounts to an expanded, historical variant of the service provided by online bookstores, such as Amazon, by gathering sales statistics and hosting reviews of their merchandise.45 Taken in the round, such data might give a much clearer idea of the historical reception, communication, and transformation of ideas across space and time than currently exists. This might be

42  For example, we have recorded inclusions in Darnton’s Corpus of Clandestine Literature, the Papal Index, or Poinçot’s list of books found in the Bastille in 1789 under illegality markers; in contrast, inclusion in STN catalogs is recorded in our general notes field. 43  My AHRC application documentation flagged an intention to develop tools and a data structure useful to further bibliometric projects. 44  The Reading Experience Database (RED), 1450–1945, The Open University, accessed 19 December 2019, http://www.open.ac.uk/Arts/RED/. Currently, RED databases, each with a different focus, exist for Britain, Canada, New Zealand, Australia, and the Netherlands. The British RED, for example, records examples over the long durée, whereas its “kiwi” cousin concentrates on the reading experience in World War I, a moment considered important to the formation of New Zealand’s national identity. 45  See Burrows and Curran, “How Swiss Was the Société Typographique de Neuchâtel?,” 63.

58 

S. BURROWS

seen as more than the culmination of a book history project: it is surely one of the holy grails of cultural history, and hitherto a most elusive one.46 The realization of this goal depends on whether our methods, tools, and approaches can be integrated into other projects, in such a way that their data structures become interoperable. This has been an aspiration of FBTEE since its initial grant application. The idea has been promoted in some of the earliest Internet commentaries on our project, notably in a blog by Eleanor Shevlin, organizer of the 2011 SHARP conference in Washington, DC, and a passionate promoter of FBTEE’s virtues. She opined that our data structures would be widely useful, and so we were passing them to other projects even before the database was published.47 Nonetheless, from a practical point of view, given the extensive use of time and energy that the FBTEE database entails, its methods and approaches are best fit for the study of societies where access to print and print technologies was limited. In Europe, this limits us to the early modern period, broadly defined. After the so-called Second Printing Revolution of the early to mid-nineteenth century, the volume of printed material in Europe and North America becomes just too great to manage, though this may not be the case elsewhere. FBTEE collaborator and map-­ visualization designer Vincent Hiribarren, for example, contemplated a linked project on apartheid-period South Africa; likewise, I have worked with scholars interested in mapping the literature of nineteenth-century Australia. The aspiration, then, is the creation of what we have called elsewhere “a unitary but flexible digital system for researching, accessing, and calibrating the dissemination and influence of ideas, knowledge, and culture across the entire print era.”48 The FBTEE database is an important first step toward realizing that dream.  For further reflections on these issues, see Curran, “Beyond the Forbidden Best-Sellers” and Simon Burrows, Jason Ensor, Per Henningsgaard, and Vincent Hiribarren, “Mapping Print, Connecting Cultures,” Library and Information History 32, no. 4 (2016): 259–71, https://doi.org/10.1080/17583489.2016.1220781. For a blue skies vision of what might be done in the subfield of library history, see Simon Burrows, “Locating the Minister’s Looted Books: From Provenance and Library Histories to the Digital Reconstruction of Print Culture,” Library and Information History 31, no. 1 (2015): 1–17, https://doi.org/ 10.1179/1758348914Z.00000000071 47  Eleanor Shevlin, “The French Book Trade in Enlightenment Europe, 1769–1794,” Early Modern Online Bibliography. The Humanities and Digital Scholarship, 5 September 2010, accessed 19 December 2019, http://earlymodernonlinebib.wordpress. com/2010/09/05/the-french-book-trade-in-enlightenment-europe-1769-1794/ 48  Burrows and Curran, “How Swiss Was the Société Typographique de Neuchâtel?,” 64. 46

2  IN SEARCH OF ENLIGHTENMENT 

59

Bibliography Brydone, Patrick. 1773. A Tour Through Sicily and Malta: In a Series of Letters to William Beckford, Esq. of Somerly in Suffolk. London. Burney, Fanny. 1782. Cecilia. London. Burrows, Simon. 2002. The Cosmopolitan Press, 1759–1815. In Press, Politics and the Public Sphere in Europe and North America, 1760–1820, ed. Hannah Barker and Simon Burrows, 23–47. Cambridge: Cambridge University Press. ———. 2006. Blackmail, Scandal and Revolution: London’s French Libellistes, 1758–1792. Manchester: Manchester University Press. ———. 2013a. French Banned Books in International Perspective, 1770–1789. In Experiencing the French Revolution, ed. David Andress, 19–45. Oxford: Studies on Voltaire and the Eighteenth Century. ———. 2013b. In Search of Enlightenment: From Mapping Books to Cultural History. Journal for Early Modern Cultural Studies 13 (4): 3–28. https://doi. org/10.1353/jem.2013.0042. ———. 2015. Locating the Minister’s Looted Books: From Provenance and Library Histories to the Digital Reconstruction of Print Culture. Library and Information History 31 (1): 1–17. https://doi.org/10.117 9/1758348914Z.00000000071. ———. 2018. The French Book Trade in Enlightenment Europe II: Enlightenment Bestsellers. London: Bloomsbury. Burrows, Simon, and Mark Curran. 2012. How Swiss Was the Société Typographique de Neuchâtel? A Digital Case Study of French Book Trade Networks. Journal of Digital Humanities 1 (3): 56–65. http://journalofdigitalhumanities.org/1-3/how-swiss-was-the-stn-by-simon-bur r owsand-mark-curran/. Burrows, Simon, Jason Ensor, Per Henningsgaard, and Vincent Hiribarren. 2016. Mapping Print, Connecting Cultures. Library and Information History 32 (4): 259–271. https://doi.org/10.1080/17583489.2016.1220781. Büsching, Anton-Friedrich. 1779. Introduction à la connaissance géographique et politique des Etats de l’Europe. Strasbourg. Burney, Fanny. 1783. Cecilia, ou Mémoires d’une héritière par l’auteur d’Evelina; traduits de l’anglois. 5 vols. Neuchâtel. Calonne, Charles-Alexandre de. 1781. Les Comments. Paris. Coleman, Catherine Nicole. 2020. Seeking the Eye of History: The Design of Digital Tools for Enlightenment Studies. In Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies, ed. Simon Burrows and Glenn Roe, 221–246. Liverpool: Voltaire Foundation in Association with Liverpool University Press. Curran, Mark. 2010. Mettons Toujours Londres: Enlightened Christianity and the Public Sphere in Pre-Revolutionary Francophone Europe. French History 24 (1): 40–59. https://doi.org/10.1093/fh/crp073.

60 

S. BURROWS

———. 2013. Beyond the Forbidden Best-Sellers of Pre-Revolutionary France. The Historical Journal 56 (1): 89–112. https://doi.org/10.1017/ S0018246X12000556. ———. 2018. The French Book Trade in Enlightenment Europe I: Selling Enlightenment. London: Bloomsbury. Darnton, Robert. 1971. The High Enlightenment and the Low-Life of Literature in Pre-Revolutionary France. Past and Present 51: 81–115. https://www.jstor. org/stable/650404. ———. 1976. Trade in the Taboo: The Life of a Clandestine Book Dealer in Prerevolutionary France. In The Widening Circle: Essays on the Circulation of Literature in Eighteenth-Century Europe, ed. Paul J.  Korshin, 11–83. Philadelphia: University of Pennsylvania Press. ———. 1995. The Corpus of Clandestine Literature in France, 1769–1789. New York: Norton. ———. 1996. The Forbidden Best-Sellers of Pre-Revolutionary France. New York: Norton. ———. 2012. Review of The French Book Trade in Enlightenment Europe, 1769–1794. Reviews in History: n.p. https://reviews.history.ac.uk/ review/1355 ———. 2018. A Literary Tour de France: The World of Books on the Eve of the French Revolution. Oxford: Oxford University Press. Darnton, Robert, and Michel Schlup, eds. 2005. Le Rayonnement d’une maison d’édition dans l’Europe des Lumières: la Société typographique de Neuchâtel 1769–1789. Neuchâtel: Bibliothèque Publique et Universitaire; Hauterive: Editions Gilles Attinger, 2005. Dawson, Robert L. 2006. Confiscations at Customs: Banned Books and the French Book Trade During the Last Years of the Ancien Régime. Oxford: Studies on Voltaire and the Eighteenth Century. Edwards, Edward. 1859. Memoirs of Libraries: Including a Handbook of Library Economy. 2 vols. London: Trübner. Freedman, Jeffrey. 2012. Books Without Borders in Enlightenment Europe: French Cosmopolitanism and German Literary Markets. Philadelphia: University of Pennsylvania Press. Gessner, Salomon. 1776. Collection complètes des oeuvres. Neuchâtel. Greetham, David. 2012. The Resistance to Digital Humanities. In Debates in the Digital Humanities, ed. Matthew K. Gold, 438–451. Minneapolis: University of Minnesota Press. Grimoard, Philippe-Henri. 1781. Lettre du marquis de Caraccioli à M. d’Alembert. S.l. Histoire de François Wills, ou le Triomphe de la bienfaisance. 2 vols. 1774. Neuchâtel. Joseph II. 1788. Ordonnance ultérieure de l’Empereur et Roi: du 20 mai 1788, concernant les libelles ou ecrits satyriques, diffamatoires, scandaleux ou séditieux;

2  IN SEARCH OF ENLIGHTENMENT 

61

de même les estampes, ou caricatures & les mauvais propos de cette espèce. Bruxelles: Pauwels. Martin, Angus, Vivian Mylne, and Richard Frautschi. 1977. Bibliographie du genre romanesque français, 1751–1800. London: Expansion. Mercier, Louis-Sébastien. 1771. L’An 2440. London. Mercier, Louis-Sébstien. 1781. Le Philosophe du Port-au-Bled. London. Necker, Jacques. 1781. Compte-rendu au Roi. Paris. Necker, Jacques. 1781a. Mémoire donné au roi par M. Necker en 1778. London [Paris]. Nicolai, Friedrich. 1774–1777. La vie et les opinions de maître Sébaltus Nothanker. 3 vols. London [Berne]. ———. 1779. L’Intolérance ecclésiastique, ou Les malheurs d’un hétérodoxe. 4 vols. Neuchâtel. Observations modestes d’un citoyen sur les opérations de finances de M.  Necker. 1781. S.l. Outram, Dorinda. 2013. The Enlightenment. 3rd ed. Cambridge: Cambridge University Press. Pidansat de Mairobert, Mathieu-François. 1775. Anecdotes sur Madame la Comtesse du Barri. London. Popkin, Jeremy D. 2012. French Book Trade Database. British Society for Eighteenth-Century Studies. https://www.bsecs.org.uk/criticks-reviews/ french-book-trade-database/ Schlup, Michael, ed. 2002. L’Edition neuchâteloise au siècle des Lumières: la Société typographique de Neuchâtel (1769–1789). Neuchâtel: Bibliothèque Publique et Universitaire. Seaward, Louise. 2014. Censorship Through Cooperation: The Société typographique de Neuchâtel (STN) and the French Government, 1769–1789. French History 28 (1): 23–42. https://doi.org/10.1093/fh/crt086. Shevlin, Eleanor. 2010. The French Book Trade in Enlightenment Europe, 1769–1794. Early Modern Online Bibliography. The Humanities and Digital Scholarship. http://earlymodernonlinebib.wordpress.com/2010/09/05/thefrench-book-trade-in-enlightenment-europe-1769-1794/. Accessed 19 Dec 2019. The Triumph of Benevolence; or, The History of Francis Wills. 1772. London. Thérèse philosophe, ou Mémoires pour servir à l’histoire de D. Dirag et de Mademoiselle Eurydice. 1782–1783. 2 vols. London. Voyage en Sicile et à Malthe. 1776. 2 vols. Neuchâtel.

CHAPTER 3

Examining the Early Modern Canon: The English Short Title Catalogue and Large-Scale Patterns of Cultural Production Mikko Tolonen, Mark J. Hill, Ali Zeeshan Ijaz, Ville Vaara, and Leo Lahti

But where are such critics to be found? By what marks are they to be known? How distinguish them from pretenders? —David Hume, “Of the Standard of Taste” (1757)

Canon-Making Versus Large-Scale Patterns of Cultural Production This chapter offers a data-driven approach to constructing and examining the English canon (ca. 1500–1800). This is part of a long-term project started in 2013, which involves extracting data from The English Short Title Catalogue (ESTC), a database that contains records of works published between 1473 and 1800  in Britain and its former colonies. The

M. Tolonen (*) • M. J. Hill • A. Z. Ijaz • V. Vaara University of Helsinki, Helsinki, Finland e-mail: [email protected]; [email protected]; [email protected]; [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_3

63

64 

M. TOLONEN ET AL.

project has been undertaken by the Helsinki Computational History Group (COMHIS) to which the authors of this chapter belong. COMHIS is an integrated multidisciplinary team that combines big data approaches with expert subject knowledge in intellectual and book history to study the early modern period.1 For this chapter, we have quantitatively constructed a canon of works that were published most often, most frequently, and for the longest period of time by making use of a processed version of the ESTC and analyzing it in terms of time, people, places, and materiality.2 Importantly, our aim has not been to curate a canon but to extract one based on a systematic quantitative investigation of publishing patterns. Therefore, this chapter provides both methodological insights into such a task and historical insights into the history of (mainly English) printed works. One crucial aim of this study is to demonstrate the enormous analytical potential of harmonized metadata catalogs. For us, this study functions first and foremost as a proof of concept3 and lays out groundwork for a series of case studies developed with the ESTC. Defining the Canon: A Brief History When speaking of written works, the canon commonly refers to “lists of approved authors” of literature.4 In his important monograph on the 1  For more information about the group and the authors’ backgrounds, see https://www. helsinki.fi/en/researchgroups/computational-history 2  For a technical description of our canon index, see footnote 45. The code for reproducing the figures and tables in the chapter is to be found at https://doi.org/10.5281/ zenodo.4003898 3  We are fully aware that some of our statistics, especially with respect to particulars, are bound to change as we harmonize and further enrich the ESTC data, as well as improve the overall data quality. We do not see this as a weakness, however, but as a possibility to generate transparent and reproducible inferences that are based on the most reliable and comprehensive data sets currently available. These inferences can be further refined over time, leading to more precise evidence. At the same time, we have put forward research questions that can be answered with our current data and that are focused on large-scale qualitative trends. We expect that the conclusions we have drawn will not be significantly revised by future improvements in data quality. 4  Wendell V.  Harris, “Canonicity,” PMLA 106, no. 1 (1991): 116, doi: https://doi. org/10.2307/462827. For the etymology of the canon and its different uses, see also Trevor

L. Lahti University of Turku, Turku, Finland e-mail: [email protected]

3  EXAMINING THE EARLY MODERN CANON 

65

making of the English canon in the eighteenth century, Jonathan Kramnick calls it “a pantheon of high-cultural works from the past.”5 What works are to be included in such a pantheon, however, is far from unambiguous. Samuel Johnson believed that canonical works of literature are able to communicate timeless human experiences due to the uniformity of human nature; these are works which transcend the particular.6 Similarly, David Hume believed in a “universal” form of art which is superior to lesser forms, such as those rooted in the mundane and the particular.7 It is from this perspective that the literary critic emerges as a moderator of the canon, or, as Ernst Robert Curtius put it, “the artificial safeguard of a tradition.”8 Conversely, there are those who dismiss the idea of objective canonicity, and argue, instead, that the canon is a reflection of contemporary or past values.9 While our age, 250 years after Johnson’s time, is “as much post-­ canonical as post-modern,”10 this is not a recent view. In late eighteenth century, Isaac Disraeli wrote that “different times … are regulated by different tastes. What makes a strong impression on the public at one time, ceases to interest it at another … and every age of modern literature might, perhaps, admit of a new classification, by dividing it into its periods of fashionable literature.”11 Other thinkers, such as Jean Jacques Rousseau, went even further, arguing that the very idea of culturally celebrated works Thornton Ross, The Making of the English Literary Canon: From the Middle Ages to the Late Eighteenth Century (Montreal: McGill-Queens University Press, 1998), 23–25. A canon need not be literary only; a cultural canon can also include, for example, objects, such as the Notre Dame Cathedral, UNESCO sites of world heritage, or even cultural trends, such as tulipomania in Holland. For details, see Frits van Oostrom, A Key to Dutch History: The Cultural Canon of The Netherlands (Amsterdam: Amsterdam University Press, 2007). 5  Jonathan Brody Kramnick, Making the English Canon: Print-Capitalism and the Cultural Past, 1700–1770 (Cambridge: Cambridge University Press, 1999), 1. 6  Ross, The Making of the English Literary Canon, 267–92. About the nature of literature, see also Terry Eagleton, “The Subject of Literature,” Cultural Critique no. 2 (1985): 95–104, https://doi.org/10.2307/1354202 7  David Hume, “Of the Standard of Taste,” in Essays Moral, Political, Literary, ed. Eugene F. Miller (Indianapolis: Liberty Fund, 1987), 226–50. For an insightful interpretation of this text, see Noel Carroll, “Hume’s Standard of Taste,” The Journal of Aesthetics and Art Criticism 43, no. 2 (1984): 181–94, https://doi.org/10.2307/429992 8  Ernst Robert Curtius, European Literature and the Latin Middle Ages (Princeton, NJ: Princeton University Press, 2013), 253. 9  Charles Altieri, “An Idea and Ideal of a Literary Canon,” Critical Inquiry 10, no. 1 (1983): 37–60, https://www.jstor.org/stable/1343405 10  Peter Robinson, introduction to Textual Scholarship and the Canon, ed. Hans Walter Gabler, Peter Robinson, and Paulius Subačius (Leiden: Brill, 2008), 1. 11  Isaac Disraeli, “Literary Fashions,” in Curiosities of Literature, 7th ed. (London: John Murray, 1823), 4:171–72.

66 

M. TOLONEN ET AL.

being universally good is a dangerous fiction.12 More recently, Richard Proudfoot has argued that “‘canons’ were abhorred as restrictive practices by critics and theorists [of Shakespeare’s time] who wanted to impose other kinds of restriction, of their own choosing, on the study of literature.”13 It is unsurprising, then, that defenses of particular values, such as those found in Harold Bloom’s The Western Canon, continue to raise questions of inclusion and exclusion.14 While, for Johnson and Hume, it was the task of the literary critic to identify canon-worthy authors and works, for critical theorists and post-colonial advocates building canons upon individual judgments is open to, and perhaps deserving of, criticism.15 It is this ambiguity around the very idea of canonicity that this chapter aims to overcome through a quantitative analysis of patterns of large-­scale cultural production. By moving away from debates around style, subjectivity, objectivity, and universality, and turning, instead, to availability, this chapter hopes to sidestep this debate. While most scholarship on early modern and eighteenth-century British canon formation approaches this subject from the perspective of canon-makers like Samuel Johnson or 12  Jean Jacques Rousseau, Politics and the Arts: Letter to M.  D’Alembert on the Theatre, trans. Allan Bloom (Ithaca, NY: Cornell University Press, 1968). 13  Richard Proudfoot, Shakespeare: Text, Stage & Canon (London: Arden Shakespeare, 2001), 38. 14  Harold Bloom, The Western Canon: The Books and School of the Ages (Harcourt: Brace, 1994). See also Lillian Robinson, In the Canon’s Mouth: Dispatches from the Culture Wars (Bloomington: Indiana University Press, 1997). On different perspectives of inclusion and exclusion from a post-colonial perspective see Leslie A. Fielder and Houston A. Baker, eds., English Literature: Opening Up the Canon (Baltimore, MA: Johns Hopkins University Press, 1979); Gail Low and Marion Wynne-Davies, eds., A Black British Canon? (Hampshire: Palgrave Macmillan, 2006); and Ankhi Mukherjee, What Is a Classic? Postcolonial Rewriting and Invention of the Canon (Stanford, CA: Stanford University Press, 2013). For aspects of gender, see Paul Salzman, ed., Expanding the Canon of Early Modern Women’s Writing (Newcastle upon Tyne: Cambridge Scholars Publishing, 2010). 15  Importantly, however, both of these perspectives are also visible in the more pragmatic aspect of canon formation: the syllabus. Once a work enters the teaching curriculum, canonicity is naturally enforced until a revisionist wave of canon formation emerges. In this way, even the critics of the canon find themselves acting as judges of values represented by these works. For details, see Jan Gorak, ed., Canon vs. Culture: Reflections on the Current Debate (New York: Garland, 2001); John Guillory, Cultural Capital: The Problem of Literary Canon Formation (Chicago: University of Chicago Press, 1993); Harry Levin, “Core, Canon, Curriculum,” College English 43, no. 4 (1981): 352–62, https://doi.org/10.2307/377120; Richard Bradford, Is Shakespeare Any Good? And Other Questions on How to Evaluate Literature (Malden: Wiley-Blackwell, 2015), 166–93; and Blair Hoxby, What Was Tragedy? Theory and the Early Modern Canon (Oxford: Oxford University Press, 2015).

3  EXAMINING THE EARLY MODERN CANON 

67

John Dryden, a key aim of this chapter is to introduce a systematic method for studying the canon and the process of canon formation which does not focus on particular authors or literary critics.16 To achieve this goal and move beyond ontological debates on what the canon actually is, we have also expanded the temporal limits of canon formation. Much of the debate around the early modern English canon focuses on whether literature, in its modern sense, was born in the eighteenth century or earlier.17 As Kramnick argues, “the decisive reception of the English literary past was settled during the mid-eighteenth century. Years of critical discussion coalesced then into a durable model of literary history and aesthetic value.”18 However, in a similar vein to the individual-centered 16  On the one hand, this emphasis on individual canon-makers can be seen in the debate about “when literature was invented,” culminating in Richard Terry’s study, “Literature, Aesthetics, and Canonicity in the Eighteenth Century,” Eighteenth-Century Life 21, no. 1 (1997): 80–101, https://www.muse.jhu.edu/article/10407. Even when scholars of the early modern canon are aware of theories of cultural production and pay attention to wider social processes, they still emphasize the role of individual canon-makers. Jeremy Lopez, for example, in Constructing the Canon of Early Modern Drama (Cambridge: Cambridge University Press, 2014) examines the history of canon-making through anthologies, thus focusing on selections by individuals. On the other hand, both Kramnick and Ross base their take on the canon on Pierre Bourdieu’s distinction between canon as a set of values reflected in literature, and canon as a result of large-scale cultural production. Cf. Pierre Bourdieu and Jean-Claude Passeron, Reproduction in Education, Society and Culture, trans. Richard Nice (London: Sage Publications, 1977); Pierre Bourdieu, “The Market of Symbolic Goods,” Poetics 14, nos. 1–2 (1985): 13–44, https://doi.org/10.1016/0304-422X(85)90003-8; and Pierre Bourdieu, “The Field of Cultural Production, or the Economic World Reversed,” Poetics 12, nos. 4–5 (1983): 331–56, https://doi.org/10.1016/0304-422X(83)90012-8. For a traditionally constructed historical case of pamphlets and the marketplace, see Alexandra Halasz, The Marketplace of Print: Pamphlets and the Public Sphere in Early Modern England (Cambridge: Cambridge University Press, 1997). 17  In his analysis, Douglas Lane Patey emphasizes the importance of the eighteenth century in the process of canon formation, given its focus on aesthetics; see “The Eighteenth Century Invents the Canon,” Modern Language Studies 18, no. 1 (1988): 17–37, https://doi. org/10.2307/3194698. Following the debate started by Thomas P. Miller, Clifford Siskin, Howard Weinbrot, Barbara M. Benedict, Robert Crawford, J. Paul Hunter, Thomas Bonnell, and Jonathan Brody Kramnick in the same issue of Modern Language Studies, Richard Terry emphasized the importance of earlier times in the process of canon formation; for details, see “Literature, Aesthetics, and Canonicity in the Eighteenth Century.” For a more recent discussion and a similar emphasis, see David Fairer, “Historical Criticism and the English Canon: A Spenserian Dispute in the 1750s,” Eighteenth-Century Life 24, no. 2 (2000): 43–64, https://doi.org/10.1215/00982601-24-2-43 18  Kramnick, Making the English Canon, 1. While there are several works that challenge the formative importance of the eighteenth century, all of them focus on the role of individual agents in the process of canon formation. See Ross’s Making of the English Literary Canon, for instance, which considers cultural attitudes toward literature, and Jane Spencer’s Literary

68 

M. TOLONEN ET AL.

approach, this assessment starts from the position that, before one can define the canon, one must define its contents. To avoid this question, we take the preliminary position that printing itself was the starting point for a work to be potentially included in the canon. While we do not deny that the eighteenth century, and particular types of written works, are key to the canon, we think that defining the canon by these features is to put the cart before the horse. Since Elizabeth Eisenstein’s seminal study, The Printing Press as an Agent of Change, it has been clear that the history of the book is an integral part of the formation of early modern civil society.19 On the whole, however, the role of the printing industry in the process of canon formation during the early modern period has remained understudied, especially from a quantitative perspective, despite scholars from outside of book history highlighting the benefits of such an approach. Benedict Anderson, for example, coined the term “print capitalism” to describe the complex dialectics of modernization through literacy and commerce tied to the book trade.20 Similarly, Jürgen Habermas identified the print and its growing mass consumption as key to the emergence of the public sphere.21 In the same tradition, this chapter acknowledges that large-scale cultural production is related to the formation of common values, but it starts from the premise that to engage with the canon at this level requires a wider view of what the canon could be. To come to understand what the canon could be, we turn to Alastair Fowler. Fowler’s taxonomy identifies six kinds of canon: the potential, the accessible, the selective, the official, the personal, and the critical.22 While Relations: Kinship and the Canon 1660–1830 (Oxford: Oxford University Press, 2005), which focuses on author networks. 19  Elizabeth Eisenstein, The Printing Press as an Agent of Change: Communications and Cultural Transformations in Early Modern Europe (Cambridge: Cambridge University Press, 1979); see also Lucien Febvre and Henri-Jean Martin, The Coming of the Book: The Impact of Printing, 1450–1800 (New York: Verso, 1976). For more recent literature, see Thomas Munch, Conflict and Enlightenment: Print and Political Culture in Europe, 1635–1795 (Cambridge: Cambridge University Press, 2019). 20  Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism (New York: Verso, 1983). 21  Jürgen Habermas, The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society (Cambridge: Polity Press, 2003). See also Halasz, The Marketplace of Print, 162–204. 22  Alastair Fowler, “Genre and the Literary Canon,” New Literary History 11, no. 1 (1979): 97–119, https://doi.org/10.2307/468873. Wendell Harris has suggested that

3  EXAMINING THE EARLY MODERN CANON 

69

the selective, official, personal, and critical canons are not directly within the scope of our study of large-scale cultural production, the other two get closer to what we aim to engage with here. The potential canon “comprises the entire written corpus, together with all surviving oral literature”23; that is, an ideal canon is made up of the totality of the written corpus, including what has not survived, as well as oral literature which is potentially lost and often scattered when not. Therefore, for our purposes, we turn to the “accessible” canon, or the portion of the potential canon which is available at a given time. An important upshot of this approach to the canon is that “creative writings [are] not dissociated from referential ones such as history, oratory, letter writing, and preaching.”24 To be printed is to be potentially canonical (we could call this the commercial canon) and, importantly, we have a robust record of these works in library catalogs, such as the ESTC. This study takes the ESTC and treats it as a record of cultural production, and thus as the “accessible” canon. Its contents (which we will return to in more detail shortly) are taken as data points which can be quantified and tracked over time. Based on these records, some works emerge as having been printed more frequently than others; some subjects and topics come into and fall out of fashion; and some authors gain or lose audiences at different times. When approached from this perspective, decisions about who should and can be canonical are no longer made by us, but by the cultural environment they emerge from. We disconnect ourselves from individual canon-makers and focus, instead, on large-scale cultural production. Thus, Shakespeare and cookbooks are treated as equals in their potentiality to be part of the canon. Our underlying principle is that these are works which particular people decided that should be printed at particular moments in time. By focusing on products that were printed over an extended period of time, our approach to the canon could run the risk of being exploratory, mundane, intrinsic,  and perhaps even positivist. This should not be regarded as a weakness, however. Instead, it is, perhaps ironically, in line with the tradition initiated by Quentin Skinner and the Cambridge School: it allows us to create a canon which is born out of its own historical context, similar to the one that would have been accessible to those living at a other kinds of canon can also be considered, such as the pedagogical canon. See Harris, “Canonicity,” 113. 23  Fowler, “Genre and the Literary Canon,” 98. 24  Terry, “Literature, Aesthetics, and Canonicity in the Eighteenth Century,” 98.

70 

M. TOLONEN ET AL.

given time.25 Moreover, by looking at all works in the ESTC, and comparing them across subjects and genres, a new way of seeing how cultural capital is formed emerges. That is, by comparing the popularity of aesthetically valuable literature with that of more mundane works of domestic economy, we may be able to recognize things about literature that definitions of aesthetic value miss. As Alex Thomson suggestively put it, “the very idea of literature might be a function of the way that we look at the past. What has been seen as literary in the past has often been treated dismissively by subsequent generations, so it seems perfectly reasonable to say that a book can be literary at one time and not at another.”26 Defining the Canon: Available Data The ESTC is a comprehensive union catalog listing early modern books, serials, newspapers, pamphlets, broadsides, and other ephemera printed between 1475 and 1801.27 Covering over 480,000 documents held by 25  See Quentin Skinner, Visions of Politics. Volume I: Regarding Method (Cambridge: Cambridge University Press, 2002), and Mark J. Hill, “Invisible Interpretations: Reflections on the Digital Humanities and Intellectual History,” Global Intellectual History 1, no. 2 (2016): 130–50, https://doi.org/10.1080/23801883.2017.1304162. It is worth noting that the concept of “canon” had particular meanings at specific historical moments. While we acknowledge that these meanings informed decisions regarding which works deserved reprints or critical editions, we do not aim to uncover them here, but the body of works which were available to readers at particular moments in time. For more on the historical meaning of “canon,” see Jan Gorak, The Making of the Modern Canon: Genesis and Crisis of a Literary Idea (London: Bloomsbury, 2013), 47. 26  Alex Thomson, “What Is Literature?” in The Edinburgh Introduction to Studying English Literature, ed. Dermot Cavanagh, Alan Gillis, Michelle Keown, James Loxley, and Randall Stevenson, 2nd ed. (Edinburgh: Edinburgh University Press, 2014), 3–16. 27  For a discussion of the ESTC and its possibilities and limitations, see Michael F. Suarez, “Towards a Bibliometric Analysis of the Surviving Record, 1701–1800,” in The Cambridge History of the Book in Britain. Volume V: 1695–1830, ed. Michael F. Suarez and M. L. Turner (Cambridge: Cambridge University Press, 2009), 37–65; Michael F. Suarez, “Book History from Descriptive Bibliographies,” in The Cambridge Companion to the History of the Book, ed. Leslie Howsam (Cambridge: Cambridge University Press, 2015), 199–219; Stephen Karian, “The Limitations and Possibilities of the ESTC,” in The Age of Johnson: A Scholarly Annual. Volume 21, ed. Jack Lynch (New York: AMS Press, 2011), 283–97; Peter Stallybrass, “The Library and Material Texts,” PMLA 119, no. 5 (2004):1347–52, https://doi. org/10.1632/003081204X17914; Alex Weedon, “The Uses of Quantification,” in A Companion to the History of the Book, ed. Simon Eliot and Jonathan Rose (Malden: Wiley Blackwell, 2009), 33–49; Leo Lahti, Niko Ilomäki, and Mikko Tolonen, “A Quantitative Study of History in the English Short-Title Catalogue (ESTC), 1470–1800,” Liber Quarterly

3  EXAMINING THE EARLY MODERN CANON 

71

more than 2000 libraries, it is an essential record of early English print culture. However, like other library catalogs, it is also “a greatly underestimated source of knowledge.”28 It is important to note that the original purpose of an analytical bibliography is not to support quantitative research, but to preserve as much original information regarding the printed material as possible. However, at the same time, each ESTC record represents a unique printed document, edition, reprint, or variant. Theoretically, every known variant of a work should have its own distinct record in the ESTC. When we combine this fact with the realities of the hand-press printing period, a technology which remained remarkably stable until the nineteenth century, it becomes possible, through careful harmonization,29 to treat these records as comparable units. For example, estimating the popularity of a particular work based on the number of its editions and reprints makes sense due to the relative stability of print run counts (something which, following technological innovations in the print industry during the nineteenth century, becomes more difficult due to the larger variations between print runs). Thus, the ESTC allows us to examine the canon from a data-driven perspective.30 25, no. 2 (2015): 87–116, https://doi.org/10.18352/lq.10112; and Maureen Bell and John Barnard, “Provisional Count of STC Titles, 1475–1640,” Publishing History 31 (1992): 47–66. 28  Mikko Tolonen, Leo Lahti, Hege Roivainen, and Jani Marjanen, “A Quantitative Approach to Book-Printing in Sweden and Finland, 1640–1828,” Historical Methods: A Journal of Quantitative and Interdisciplinary History 52, no. 1 (2019): 57, https://doi. org/10.1080/01615440.2018.1526657 29  The original, unprocessed bibliographic metadata is typically not directly comparable across the catalog due to varying naming conventions, spelling errors, missing entries, and other technicalities. The data quality and comparability can be substantially enhanced by automated harmonization procedures, as described in more detail in Leo Lahti, Jani Marjanen, Hege Roivainen, and Mikko Tolonen, “Bibliographic Data Science and the History of the Book (c. 1500–1800),” Cataloging & Classification Quarterly 57, no. 1 (2018): 5–23, https://doi.org/10.1080/01639374.2018.1543747 30  Our methodological approach was described in Lahti, Marjanen, Roivainen, and Tolonen, “Bibliographic Data Science and the History of the Book (c. 1500–1800),” and it was later used in Tolonen, Lahti, Roivainen, and Marjanen, “A Quantitative Approach to Book-Printing in Sweden and Finland, 1640–1828.” See also Lahti, Ilomäki, and Tolonen, “A Quantitative Study of History in the English Short-Title Catalogue (ESTC), 1470–1800.” Philip Gaskell’s idea of “London average” (750–1500) as a reliable estimate of early modern British print runs was supported, for example, by Richard Sher, in The Enlightenment and the Book. Scottish Authors and Their Publishers in Eighteenth-Century Britain, Ireland, and

72 

M. TOLONEN ET AL.

It is also important to understand that, even though an immense amount of work has gone into processing the ESTC, it remains imperfect both as a catalog and a dataset.31 Like other early modern catalogs, it cannot be considered as comprehensive because it does not contain information about every existing early modern British publication. Some works have been lost while others have been collected with concern for posterity.32 Thus, while a quantitative analysis of the early modern period is exciting, we must recognize the limits of the data we have access to. Although there are statistical approaches to estimating lost works, this chapter will focus, instead, on one aspect of the dataset which we can more confidently explore: the most popular works recorded in the ESTC.33 These are works which are both historically more likely to survive and quantitatively more representative of the overall publication output.34 With these reservations in mind, one of the strengths of the ESTC’s records lies in their robustness: they contain as many as 420 fields, each with its own attributes, ranging from a work’s physical features to information on the libraries holding copies of the work. Their weakness, however, lies in the significant effort needed to extract this information at scale. While the ESTC records use the Machine-Readable Cataloging (MARC) 21 standard, making use of

America (Chicago: University of Chicago Press, 2007), 86–87. See also Sarah Werner, Studying Early Printed Books, 1450–1800: A Practical Guide (Malden: Wiley Blackwell, 2019), 23–25. 31  Cf. Stephen Karian, “The Limitations and Possibilities of the ESTC.” 32  See, for example, the impact of the Civil War on the number of pamphlets published or the damage caused by the Great Fire of London in 1666 in terms of lost works (Fig. 3.1). There are several other issues that complicate these historical anomalies, such as Thomason’s tracts causing a peak in the number of variants during the Civil War, duplicated titles when Wing and STC catalogs are combined, approximated publication years causing a peak in publication numbers at five-year intervals starting from 1500, and other issues specific to the data collection and cataloging process. 33  For analyses and estimates of the surviving record and the ESTC, see Flavia Bruni and Andrew Pettegree, eds., Lost Books: Reconstructing the Print World of Pre-Industrial Europe (Leiden: Brill, 2016). 34  It is evident that the survival rate of frequently published canonical works is higher than that of single-sheet publications and other ephemera. On broadsheets, see Andrew Pettegree, ed., Broadsheets: Single-Sheet Publishing in the First Age of Print (Leiden: Brill, 2017).

3  EXAMINING THE EARLY MODERN CANON 

73

this data requires extensive processing.35 A brief overview of the process clearly shows this challenge.36 Actors The data for actors connected to entries in the ESTC are contained in a wide variety of MARC fields.37 The extraction process varies depending on the type of actor (author, printer, publisher, editor, etc.) and the amount of information about that actor documented in the metadata. In an ideal case, we are provided with the actors’ names, years of birth and death, and their role in a particular printed work; in less ideal cases, we may only be provided with a set of initials or a verbatim repeat of a work’s imprint.38 As 35  Lahti, Marjanen, Roivainen, and Tolonen, “Bibliographic Data Science and the History of the Book,” 5–10. The Library of Congress provides an extensive overview of the MARC 21 standard at https://www.loc.gov/marc/ 36  The end result of processing the MARC catalog is a dataset comparable to linked data. Distinct entities, such as actors connected to the titles documented in the original MARC catalog, are separated out and assigned unique identifiers; these identifiers are, in turn, used to link various entities. A conscious decision was made to keep the infrastructure around the dataset as light and as simple as possible. The whole processed dataset is stored in a collection of CSV tables of entities; these entities include works, titles, actors, and linking tables that document these connections, when needed. These tables, in turn, are stored in a thoroughly documented Git repository. From a modelling point of view, this resembles an SQL database and can be easily exported to other formats. For details, see Leo Lahti, Ville Vaara, Jani Marjanen, and Mikko Tolonen, “Best Practices in Bibliographic Data Science,” in Proceedings of the Research Data in the Humanities 2019 Conference: Data, Methods, and Tools, ed. Jarmo Harri Jantunen, Sisko Bruni, Niina, Kunnas, Santeri Palviainen, and Katja Västi (Oulu: Faculty of Humanities Linnanmaa, 2019), 57–65. 37  The process of unifying actors, and some initial research making use of this data, are detailed in Mark J.  Hill, Ville Vaara, Tanja Säily, Leo Lahti, and Mikko Tolonen, “Reconstructing Intellectual Networks: From the ESTC’s Bibliographic Metadata to Historical Material,” in Proceedings of the Digital Humanities in the Nordic Countries 4th Conference: Copenhagen, Denmark, March 5–8, 2019, ed. Constanza Navarretta, Manex Agirrezabal, and Bente Maegaard (Aachen: CEUR-WS.org, 2019), 201–19. 38  Information on book trade actors was to a large part distilled from a field (260b) described in the MARC standard as “Name of publisher, distributor, etc.” In reality, however, it most often repeated faithfully the publisher’s statement of a particular title. Examples vary widely in length, level of detail, and style. For instance, they can vary from brief statements, such as “Sold by J. Newton and R. Bland,” to detailed descriptions, such as “Printed by T. Bensley, Bolt-Court, Fleet-Street. Sold at Providence chapel on Monday and Wednesday Evenings, and at Monkwell Street Meeting on Tuesday Evenings; by W.  Baynes, No. 54, Paternoster-Row; T. Green, No. 93, and J. Baker, No. 226, Oxford-Street; J. Cobbin, No. 14, Hertford-Street, Fitzroy-Square; at the Chapel in the Cliff, Lewes, Sussex; T. Barston, Castlegate, Grantham, Lincolnshire; and by A. Batten, Sen. Wellwyn, Herts.”

74 

M. TOLONEN ET AL.

each record is distinct and there is no explicit information linking the actors or the works recorded, a large amount of cleaning and unification is needed. Strings representing data points must be extracted, cleaned, corrected, compared (internally and externally, by making use of other databases), unified (when appropriate), and finally harmonized. Incomplete or varying spellings make the task even more difficult as identifying and grouping mentions of the same historical actors requires extensive post-­ processing. We have currently compiled a database of 144,399 unique actors (of which 56,693 are primarily book trade actors and 67,924 primarily authors), with 1,107,777 links to titles documented in the ESTC.39 Works An additional issue with the ESTC is the lack of unifying links between multiple editions of the same work. If one searches for Romeo and Juliet, for instance, there may be 80 results based on exact string matching, but no explicit links among these works. Moreover, when one accounts for title variations, repetitions, or commentaries, connecting the correct items becomes a very difficult task. To address this issue, we have used a work-­ field dataset as the foundation of our process. Since this is an integral part of the workflow that enabled us to start discussing the canon based on ESTC records, it is necessary to explain this process in more detail. Our aim was to draw out relations between discrete records and link them as single works; in other words, to create a relational model. The fields we chose for matching different records included, first, the edition statement, which provides a specific edition number for each record. Currently, only a small subset of the whole ESTC contains information in this field (around 44,000 records). Second, we used the title and title remainder fields to identify the complete title of a record, while the title uniform field provided a representative title for the work. Lastly, we used the publication year of the record to provide a chronological ordering of the various editions. Distinguishing information, such as publication date or edition number (if available), was then used to determine more precisely what a particular edition is and organize all the editions of a work in chronological order. The dataset was created through a multi-stage harmonization process, which began with an initial cleaning of the various fields needed for this task. Any unwanted characters and stop words were removed, and the text  Other roles include translator, dedicatee, and unknown.

39

3  EXAMINING THE EARLY MODERN CANON 

75

was converted to lowercase. An initial dataset was then created by combining the actors responsible for works with harmonized titles; these actors were used to differentiate among different works with the same title. This provided a unique work-field identifier for each record. The dataset was then segmented and harmonized on a per-actor basis.40 This was performed by specialized algorithms, which were able to determine the representative work-fields of these editions. This allowed for an effective grouping of editions despite differences in title lengths and spelling. Finally, as some works have multiple actors attached to them (often at different times), another harmonization step was necessary to properly collate these duplicated works into single, unified work-fields. Figure 3.1 shows a comparison among the original record counts from the ESTC (including first editions and reprints), the processed unique works as derived from the work-field dataset (where the reprints have been removed), and the first prints of the works included in the canon that we extracted. The original records include prints without any actor information, which raises the number of records to over 480,000. However, by processing unique works information, we effectively normalized the whole ESTC dataset, which resulted in a significantly lower number of works for each decade. As the canon dataset is based on the work-field dataset, it follows a similar trend. It is important to note that the work-field harmonization and dataset creation is an ongoing process, which is iteratively and continuously being improved for more accurate grouping of editions. At the moment, however, the work-field dataset consists of 200,378 works covering a total of 361,245 records. Subject-Topics It was important for this study to include information about the subject-­ topics (such as “Religion,” “Literature,” and “History and Geography”) of the works examined.41 While other approaches may see this as a necessary step for determining what to include in the canon (i.e., which 40  We have chosen only those records where the actor attached to them had a specific role. The list of accepted actor roles included author, corporate author, translator, and attributed name. All the other records either did not have any actors attached to them or had actors of other roles than the ones selected, so they were not used in the creation of our dataset. 41  Fowler has underscored the relevance of the information related to genre, when thinking about the canon, in “Genre and the Literary Canon,” 100.

76 

M. TOLONEN ET AL.

Fig. 3.1  Total documents, works, and canon items in the ESTC per year (1500–1800)

subject-topics are more aesthetically or culturally valuable and, therefore, worthy of being included), for our purposes it allowed for further qualitative reflections, such as temporal changes within the canon. This information, however, has not been recorded in the ESTC comprehensively (roughly only half, or 266,207 documents in the ESTC, have subject headings) or systematically (there are 12,553 unique subjects).42 It was, therefore, necessary for us to modify and enrich this data. There are numerous models proposed for categorizing subject-topics, and they range from classification systems developed by ancient authors, to early modern attempts to revise such systems, to current efforts to 42  To complicate things even more, many subjects recorded in the ESTC are questionable. For example, John Arbuthnot’s political pamphlet on John Bull is tagged in the ESTC under the subject heading “Bulls.”

3  EXAMINING THE EARLY MODERN CANON 

77

create appropriate models for quantitative analysis.43 Because the approach in this study is computational, a hierarchical classification system by subject was aimed for. To this end, we chose the Dewey Decimal Classification (DDC) system.44 As classifying every single work in the ESTC was not an option, we chose, instead, to hand-classify a selection of works and, then, used this information with the existing subject headings in the ESTC to create a conversion table which was used to further extrapolate the data. That is, each entry with subject-topics in the ESTC was compared with the manually entered data, and this training dataset was used to find the most common equivalent ESTC subject-topic for each DDC category. The resulting DDC-to-ESTC translation table was then used to assign Dewey-style subject-topics to non-hand-categorized ESTC entries. In total, we hand-classified 1153 works, which represent a total of 47,041 individual documents. From these, we were able to classify another 62,342 documents with the conversion table. We should note that the catalog has many unique and rare topics: 7957 topics were used in only one instance and they range from individual psalms to specific years to “Granby (Race horse).” In the end, we identified 53,683 works with a subject-topic in the original ESTC but with no equivalent in the DDC. This is a typical example of the diminishing returns of manual work in digital humanities: the remaining untranslated subject-­ topics are increasingly rare, so the payoff for each additional manual entry decreases. Defining the Canon: Methods As stated, our aim in this study is to construct a canon using a data-driven approach and analyze the works contained within it. As this is a data-­ driven approach, there is no qualitative judgment made at the curation stage. Instead, we aim to construct a list of canonical works that is born out of historic publication records. To do so, it is necessary to define a set of features which could be considered representative of canonical works 43  For a quantitative example, see David. L. Gants, “A Quantitative Analysis of the London Book Trade, 1614–1618,” Studies in Bibliography 55, no. 1 (2002): 185–213, https://doi. org/10.1353/sib.2005.0004 44  We have used a two-level system. The general categories are: “Information & General Works,” “Religion,” “Social Sciences,” “The Arts,” “Literature,” “Philosophy,” “Natural Science & Mathematics,” “Applied Science,” “History & Geography,” and “Language.” These further break down to a total of 94 subcategories.

78 

M. TOLONEN ET AL.

and examine the entirety of the ESTC for these features. The features we have chosen include the total publication count and publication frequency, which have been further normalized by the overall publishing activity during the same period. Based on this canon index,45 we identified the top 1000 works that were printed in relatively high numbers, relatively frequently, and over a relatively long period of time.46 This method has resulted in the inclusion of works that qualitatively would never be considered as canonical but, nevertheless, have been included in the data-driven index as they have identical data profiles to works that we would expect to be included. For example, while almanacs may be considered less historically important to many, their frequent and 45  We have defined the canon index as Cw = Tw x (Nw/N), where Cw is the canon index for work W, Tw is the publication frequency (total number of distinct publication years) for work W, Nw is the total publication count for work W, and N is the total number of all works published between the first edition of work W and the year 1800. The ratio (Nw/N) indicates the overall share of the given work in all publications within the time period that starts from the first publication of work W and ends in 1800. The normalization of the covered time span improves the comparability between earlier and later publications. The canon index increases with the total publication count and publication frequency. We have tested multiple metrics and methods to create a reliable index of “canonical works,” making use of historical expectations and ensuring a qualitative balance among features. We have chosen the above-mentioned features for several reasons. First, while the number of unique versions (or editions) of a given work is certainly an indicator of a work’s importance, this feature alone is not an indication of canonical importance. Works could be printed in very high numbers over very short periods of time and then quickly forgotten (see, for example, political pamphlets during the English Civil War). While these works are interesting in their own right, they are more likely indicative of topicality than canonicity. Thus, the longevity of a work, i.e., the frequency and total period during which it was reprinted, was deemed a key feature. However, this metric also needs to be treated as a relative feature, allowing for works to be measured against their temporal peers, as works printed earlier had a much higher longevity potential. Thus, each work was also measured as a proportion of its potential contribution to the publication output for the period in which it was published. That is, a work published in 1750 has been measured as a proportion of all works published between 1750 and 1800 rather than against all works in the ESTC. It is important to note that works first printed at the very end of the years covered by the ESTC cannot have their future impact measured as there is no data post-1800. There are, therefore, fewer new works marked as canonical in the last decade of the eighteenth century. The identification of the top canonical works was relatively robust to the choice of index; we have chosen to use an index that is intuitive and easy to calculate. 46  Although 1000 may appear to be an arbitrary number, it meets two useful criteria: first, it is a number that a human is still able to engage with when qualitatively examining results and assigning subject-topic classifications; second, it is a number after which the total number of reprints for a work begins to drop below 20.

3  EXAMINING THE EARLY MODERN CANON 

79

stable reprinting patterns  mean  that they feature in our list. Their relevance, of course, depends on the context of publication. While an almanac lacks literary significance, from a commercial standpoint it is very significant, a point which can also be made about commercial catalogs that emerged in the latter part of the eighteenth century. Similarly, while grammar books played a key cultural role at the time of their publication, some may argue for their exclusion from lists of literary bestsellers.47 Indeed, for the most part, these works do not fit the category of printed cultural material we are interested in. For this reason, printed versions of laws and political reports, liturgies and local church documents, catalogs, almanacs, curricula, periodicals not published as collected works, and annual reports from various associations have been excluded from our canon,48 leaving us with 856 works.49

The Canon: Works, Time, and Subject-Topics When treated as one continuous historical canon, the top 20 works we have extracted based on the current data are displayed in Table 3.1.50

 Sigfrid H. Steinberg, Five Hundred Years of Printing (London: Penguin, 1955), 99–105.  Of the 154 works considered not relevant to the canon, the top ten were Church of England liturgies, general public acts (one set authored by “Britain” and one by the Parliament), Irish proclamations, Connecticut laws, Catholic church liturgies, Rider’s British Merlin, Quaker yearly epistles, miscellaneous official documents by King Charles I, and reports from the Court of Chancery. 49  It should be noted that the earlier part of the period covered by the ESTC is slightly overrepresented, as we can see in the long tail. There are at least two reasons for this. First, there is a back catalog of written material which had existed for much longer, yet it is only recorded from the start of the ESTC. This catalog includes authors who lived prior to the sixteenth century, in particular ancient authors, who can only be recorded at the beginning of the ESTC records. Second, printing itself was a much more specialized industry early on. This means that what was chosen to be printed may already have met some subjective criteria established by printers, which made it more likely to be a work with longevity. At the time, the differences between early printing and traditional manuscript production were minimal. 50  It should be noted that the publication years of these works have been extracted from the ESTC following the method laid out previously. We do not claim that this data includes every edition of a work; the many changes in the way in which a work has been recorded in the ESTC makes the detection of all editions difficult. As harmonization of public documents, such as general acts, continues, they will be mapped together in the future based on this additional information. Information about the full canon can be found in the online code and data supplement at https://doi.org/10.5281/zenodo.4003898 47 48

80 

M. TOLONEN ET AL.

Table 3.1  Top 20 canonical works (1500–1800) No. Author

Eds. No. Author

Work

Eds.

Thomas Book of Psalms Sternhold and (1553–1762) John Hopkins The Book of Common Prayer (1549–1795) William Lily Short Introduction to Grammar (1543–1800) Aesop Fables (1484–1800) John Bunyan Pilgrim’s Progress (1678–1800) Richard The Whole Duty of Allestree Man (1658–1800) John Milton Paradise Lost (1667–1800)

592 11

François Fenelon

Telemachus (1699–1800)

171

373 12

Mathurin Cordier

Colloquies (1651–1800)

140

241 13

Alexander Pope

Essay on Man (1733–1800)

172

226 14

Desiderius Erasmus Homer

8

Ovid

153 18

9

Daniel Defoe

Metamorphoses (1552–1800) Robinson Crusoe (1719–1800)

10

John Wesley

Hymns (1740–1798)

203 20

1

2

3

4 5 6

7

Work

237 15 181 16

211 17

213 19

Colloquies (1519–1800) The Iliad (1581–1796) Lewis Bayly The Practice of Piety (1612–1792) Robert The Oeconomy of Dodsley Human Life (1750–1800) Isaac Watts Hymns (1707–1800) William A Companion to Vicker the Altar (1707–1800) Edward Night Thoughts Young (1742–1800)

124 142 128

185

152 131

168

One can immediately notice the diversity of these works. Included are the expected works of poetry and fiction, but also devotional literature and language grammars—works which would normally be excluded from a literary canon. However, as we have stated, our aim is not to curate a canon but to extract one. It is possible to purge works post hoc, but before such a step is taken, it is worth noting and investigating what does make the list. Questioning the works and authors who make up the extracted canon will be a recurring theme throughout this chapter, but to give one example, we will turn now to William Vicker’s A Companion to the Altar. A relatively obscure work today, this short text was written to spiritually prepare readers for holy communion. During the eighteenth century and

3  EXAMINING THE EARLY MODERN CANON 

81

beyond, the work was often offered as a gift to those preparing for confirmation and was frequently printed along with The Book of Common Prayer. This practice can be considered as an explanation for its 131 records, and perhaps be seen as a good reason to exclude this work from our canon. However, the strength of our approach lies in confronting and reflecting on works such as these. To have been published in such large numbers means it reached a large audience, and it is, therefore, worth reflecting on the number of people who would have been familiar with its contents. We know, for example, that it was one of the 20 volumes owned by Jane Austen and that “she made constant use of the devotions contained in it.”51 Thus, while the work itself may be seen as distinctly uncanonical by other definitions, its relationship to both the canon and the historical-­ cultural moment of its publication should encourage further reflection. Of course, raw statistics offer only one view of many, especially if we are looking at the canon atemporally. For example, when one looks at the frequency of publication for works included in the ESTC, the emphasis will be on the latter part of the eighteenth century, when most printing activity took place. However, by constructing a data-driven canon which takes into consideration the relative longevity and publication frequency of these works, we can also provide a temporally representative selection. While this temporality can be seen in the overall distribution of these works (Fig. 3.2), this approach also allows us to make more specific analyses. For example, in Fig. 3.3, we can also see the works which were most frequently printed per decade. As previously noted, we have also assigned subject categories to many of the works in the ESTC, as well as to all the works which make up our canon. By categorizing these works, we can also examine the canon by subject. For example, the top-ten literary works in each category can be seen in Table 3.2.52 This data also allows us to examine the changes in the distribution of subject-topics in the canon over time and recognize that subject-topics emerge and subside at particular historical moments. These shifts are not entirely surprising. As Fowler noted, “the complete range of genres is by no means equally, let alone fully, available in any one period … Moreover, 51  Irene Collins, “The Rev. Henry Tilney, Rector of Woodston,” Persuasions no. 20 (1998): 156. 52  Although in the top-ten, the collected works of Swift, Pope, Virgil, Horace, Milton, and Shakespeare are not included.

82 

M. TOLONEN ET AL.

Fig. 3.2  The Full Canon (1500–1800). These canonical works have been sorted by the first publication year. Individual dots indicate the publishing year for the initial publication and all subsequent reprints

each age makes new deletions from the potential repertoire.”53 We can see this in Fig. 3.4: by including more than strictly literary works in the canon, we can see the emergence of new types of printed documents from the mid-seventeenth century onward. Religion and Literature By combining data covering  top-works, temporality, and subject-topics, we can begin to construct more complicated versions of the English canon. One can see, for example, the importance of grammar books during early printing era (under the category “language”), and then  Fowler, “Genre and the Literary Canon,” 110.

53

3  EXAMINING THE EARLY MODERN CANON 

83

Table 3.2  Top 10 canonical literary works (1500–1800) Fiction

Poetry

Drama

Aesop’s Fables Bunyan, Pilgrim’s Progress Defoe, Robinson Crusoe Homer, The Iliad

Milton, Paradise Lost Ovid, Metamorphoses Pope, Essay on Man Young, Night-Thoughts Thomson, The Seasons

Addison, Cato, a Tragedy Gay, The Beggar’s Opera Ramsay, The Gentle Shepherd Shakespeare, Hamlet

Gay, Fables Juvenal, Satires

Otway, The Orphan Shakespeare, Romeo and Juliet Rowe, The Fair Penitent Otway, Venice Preserv’d Rowe, Jane Shore

Goldsmith, The Vicar of Wakefield Lesage, Giles Blas Cervantes, Don Quixote Sterne, A Sentimental Journey The Fables of Phaedrus Sterne, Tristram Shandy

Butler, Hudibras Blair, The Grave Gay, Acis and Galatea

Lillo, The London Merchant

Fig. 3.3  Most frequently printed works (1500–1800). The point size indicates the number of reprints for each work (rows) during a given decade (columns)

84 

M. TOLONEN ET AL.

Fig. 3.4  Temporal variation in the relative frequency of the most common subject-­topics in the canon (1500–1800). The relative variability in publishing frequency is higher in earlier time periods due to the lower number of total published works

recognize the dominance of religious works up to the late seventeenth century, followed by the rise of literary genres (drama, fiction, and poetry) at the beginning of the eighteenth century.54 On the whole, it becomes possible to recognize epistemological shifts between the Renaissance and the eighteenth century.55 In particular, frequently printed late Renaissance works are largely religious tracts, classical works, and grammar books. While the classics, especially Aesop’s Fables and Cicero’s On Duties, continue to be printed until the end of the eighteenth century, religious works, such as The Book of Hours, The Book of Common Prayer, and The 54  For early religious publishing, see Patrick Collinson, Arnold Hunt, and Alexandra Walsham, “Religious Publishing in England 1557–1640,” and Ian Green and Kate Peters, “Religious Publishing in England c.1640–1695,” The Cambridge History of the Book in Britain. Volume IV: 1557–1695, ed. John Barnard, D. F. McKenzie, and Maureen Bell (Cambridge: Cambridge University Press, 2002), 29–66, and 67–94, respectively. 55  It is, of course, true that detecting an epistemological shift cannot be based solely on reprint records. Publication records do, however, offer one perspective.

3  EXAMINING THE EARLY MODERN CANON 

85

Book of Psalms, become relatively less apparent. When looking at the eighteenth century, however, works that are more traditionally considered canonical do emerge—particularly with respect to literature. Thus, while literature clearly had a central place in the canon, its volume in terms of printed works compared to religious works is cemented by 1700. When analyzing the subject-topic distribution in the canon before and after the eighteenth century, the decline of religious and grammar books, and the rise of literary genres, especially drama, becomes apparent. This pattern holds true for all the works in the ESTC that include subject-topic information (although, overall, religion holds its place better than history and geography, for example). Thus, confirming some scholars’ historical expectations, the eighteenth century emerges as a key moment in the history of the canon. We would hesitate to state that this makes the eighteenth century representative of the canon, however. We would, instead, note that there is an epistemological shift regarding what is considered canonical at this point in time. This does indicate that, depending on one’s analytical aims, the year 1700 is a potentially useful marker.56 Other potential markers do exist, however, as we will see below. Donaldson v. Becket and the Importance of 1774 It has been claimed that, “[o]n 22 February 1774, literature in its modern sense began.”57 With Donaldson v. Becket, the House of Lords ended the perpetual copyright and, consequently, London’s monopoly over print in Britain, allowing for a back catalog of cultural goods to enter the public domain. The impact of this event on the print industry was profound. As we have noted in previous quantitative research into the ESTC, the relationship between the London printers and publishers in the 1770s changed radically.58 Our concern in this chapter, however, is to identify what was offered to the public at a large scale after this act. While there is clear

56  Depending on one’s aims, other years may be more appropriate. For example, one may also want to look at the post-1666 print industry as it was rebuilt following the Great Fire of London. See Hill, Vaara, Säily, Lahti, and Tolonen, “Reconstructing Intellectual Networks,” 206–208. 57  Ross, Making of the English Literary Canon, 297. 58   See Hill, Vaara, Säily, Lahti, and Tolonen, “Reconstructing Intellectual Networks,” 201–19.

86 

M. TOLONEN ET AL.

evidence that Ross’s thesis of radical change should be visible in our data, the question is whether our canon diverges from this expectation or not. To answer this question, we created a post-1774 canon which was compared with the complete data-driven model. We extracted the 1000 most printed works originally published before 1746 (thus making them potentially public domain), reprinted after 1774, and printed in Great Britain (where the law applied).59 We then compared these works with those included in the larger canon, looking for substantial differences. It must be noted, however, that the limits placed on the post-1774 canon mean that it is a substantially smaller subset. While the entire processed ESTC has currently 361,245 harmonized recorded documents (representing 200,378 works), only 8925 of those documents (1997 works) meet the post-1774 criteria. Therefore, to compare the two directly is not entirely meaningful as a substantial number of works in the data-driven canon are missing. However, we can still examine how the top-works in the post-1774 canon are distributed in our data-driven canon. The results show that almost a quarter (23.2%) of all post-1774 works can still be found in the larger data-driven canon. Moreover, the overall distribution of the highest ranking post-1774 works (i.e., most printed) overwhelmingly falls at the top end of the larger canon. This is not entirely surprising, though, as printing frequency is one of the defining features of the data-driven canon. This comparison does, however, verify that this data-driven approach recovers works similar to those in the post-1774 canon with substantial coverage and accuracy. What is more important to this study, however, is what is not captured in the post-1774 canon. Amongst the top 500 post-1774 works, only 41 are not in the data-­ driven canon. Of these, 16 are works of literature. The types of works not included in these 41 titles are spread across subject-topics, but the most common type of work is drama (eight titles). On the other hand, amongst the top-500 data-driven canonical works which were in the public domain, 59 are not in the post-1774 list. Of these, 21 are works of literature. Importantly, the works which did not make the larger data-driven canon but are in the post-1774 list are generally printed less frequently. When 59  While there is evidence that printers in Scotland had accepted the end of perpetual copyright as early as the 1740s, that is, before the publishers in London, it is not clear that this was universally the case. As Scotland is important culturally and as a center for reprinting works, it was included in our analysis. See John Feather, Publishing, Piracy and Politics: An Historical Study of Copyright in Britain (London: MansFigell, 1995), 81.

3  EXAMINING THE EARLY MODERN CANON 

87

including all prints (not just those in Britain) the mean number of prints in the data-driven canon is 19, while the number of works in the post-1774 canon is 7. If we only look at prints in Britain, 90% of post-1774 works have 10 or fewer reprints. Interestingly, a total of 145 literary works found in the larger  data-­ driven canon are missing from the post-1774 canon. A significant number of these works (50) are works of fiction. These include works which were still covered by copyright in 1774, such as Henry Fielding’s Tom Jones (1749), Samuel Johnson’s Rasselas (1759), and Oliver Goldsmith’s The Vicar of Wakefield (1766), as well as older works which, while still published in the eighteenth century, did not reach the same edition count as some newer works, such as Boccaccio’s Decameron (first recorded in the ESTC in 1562), Robert Greene’s Pandosto (1588), and Philip Sidney’s Arcadia (first recorded in the ESTC in 1590).60 Additionally, 34 works of poetry, 34 works of drama, and 27 miscellaneous literary works are not included in the post-1774 canon. Authors of works that did not make the post-1774 cut include Jonathan Swift, Hannah More, Jean-Jacques Rousseau, Voltaire, Michel de Montaigne, Alexander Pope, Ned Ward, John Dryden, Benjamin Franklin, Thomas Paine, and Richard Steele.61 Overall, while there is clear evidence that the 1774 legal changes did have a substantial impact on publishing, especially of works that one would consider canonical, this is an event which takes place so late in our dataset that its impact is likely to be seen more in the following periods not covered by our dataset. This includes works which, for various reasons, fell out of favor toward the end of the eighteenth century, as well as works which were still under copyright in 1774. As we are interested in the canon as it developed and was available over the entirety of the period covered by the ESTC, the actions of a temporally specific group of people should not be over-represented. However, it is worth noting one important upshot of the comparison between the two datasets: the general overlap of coverage and the opportunity to understand the causes of the missing works provides further verification of the methods used to construct our canon. 60  For more on the shift in the works which make up the canon, see Ross, The Making of the English Literary Canon, 210–52. 61  It should be noted that many of these works may have found their way into the emerging “collected works” genre. However, new collections of works by an author are not treated as continuations of previous works in our data, but as new editions. While this is methodologically correct for our purposes, it does mean that there is potential for under-representation at the author level.

88 

M. TOLONEN ET AL.

People: Authors When discussing the people responsible for the canon (with particular emphasis on authors), we must acknowledge that the importance we place on them today is radically different from the importance their contemporaries had placed on these authors in their time. As Adam Rounce has noted with regard to Samuel Johnson, “[t]he desire to be recognized as an author and to profit from it coexists with an awareness that much work in the burgeoning world of print and literary journalism was not especially intended to be handed down to posterity; the sparsity of works with Johnson’s name on the title page in his lifetime indicate his pragmatism.”62 In fact, it was only toward the end of the eighteenth century that authorship began to take the form that we recognize as typical today and, therefore, for the majority of the era covered by the ESTC the notion of authorship was quite different. One of the clearest examples of this is anonymity. Up until the eighteenth century it was quite common for authors not to be credited, by choice or by practice, for their work. For instance, between 1679 and 1800, the ESTC has 239 records with authorship attributed to a “Lady.” Many similar examples can be found in our canon: multiple works by Defoe were initially published without any attributed authorship, Philip Francis’ criticisms of George III’s government were penned under the name “Junius” for obvious reasons, and Richard Steele used the nom de plume Isaac Bikerstaff in The Tatler (which was itself  borrowed from Jonathan Swift).63 Bickerstaff is indicative of another aspect related to authorship: the practice of collaborative writing. While Steele can be credited with most issues of The Tatler, he was not the sole author: Joseph Addison and Jonathan Swift also contributed pieces to the periodical, a practice which would be taken further with The Spectator. For many, in fact, authorship was not something that was held in any regard. The “hack” author, paid by the page or word, willing to put his or her skills to use for any topic or patron, was a common presence during the eighteenth century and was successfully immortalized in Pope’s Dunciad (1728, 1729, 1743). In other words, when thinking about the canon, it is 62  Adam Rounce, “Authorship in the Eighteenth Century,” Oxford Handbooks Online, 2015, https://doi.org/10.1093/oxfordhb/9780199935338.013.38 63  Other pseudonyms included in the canon are: “Countess of Huntingdon’s Connexion,” “Gentleman in the Country,” “Gentleman of Oxford,” “Lover of their Precious Souls,” “Person of Quality,” and “Protestant.”

3  EXAMINING THE EARLY MODERN CANON 

89

important, on the one hand, not to put undue weight on authorship, and, on the other, to recognize the exceptionality of those who were able to step into the authorial spotlight. Based on our records, 833 works in the canon have a person or organization as author and 556 of these are unique. There are different ways of speaking about who the top authors were or may have been: one can count the authors who published the most canonical works, the authors who published the most editions of these works, or the authors who have the most records in the ESTC. In each case, a different picture emerges, although we see repetition (Table 3.3). Table 3.3  Top authors and works, canon editions, and total works recorded in the ESTC (1500–1800) Author

Works Author

Canon eds.

Author

ESTC recds.

Shakespeare, William (1564–1616) Watts, Isaac (1674–1748)

20

Shakespeare, William (1564–1616) Watts, Isaac (1674–1748)

779

Wesley, John (1703–1791)

1217

570

1085

Bunyan, John (1628–1688) Defoe, Daniel (c. 1660–1731) Wesley, John (1703–1791) Lily, William (c. 1468–1522) Goldsmith, Oliver (1728–1774) Milton, John (1608–1674) Gay, John (1685–1732)

480

Pope, Alexander (1688–1744)

343

Shakespeare, William (1564–1616) Defoe, Daniel (c. 1660–1731) Watts, Isaac (1674–1748) Swift, Jonathan (1667–1745) Pope, Alexander (1688–1744) Voltaire (1694–1778) Bunyan, John (1628–1688) Christie, Mr. (James) 1730–1803a Dryden, John (1631–1700)

9

Bunyan, John (1628–1688) Goldsmith, Oliver (1728–1774) Wesley, John (1703–1791) Defoe, Daniel (c. 1660–1731) Dryden, John (1631–1700) Rowe, Nicholas (1674–1718) Virgil (70 BCE–21 BCE)

8

Voltaire (1694–1778)

7

8 8 7 7 7 7

416 377 375 366 356 347

1082 1059 770 664 648 641 616

548

The inclusion of James Christie is worth mentioning: founder of the auction house Christie’s, his attributed works are catalogs of sale. While certainly not the type of work one would generally consider canonical, his inclusion is a point of historical interest as it reflects both changes in the print industry and the rise of commerce. Additionally, it draws attention, again, to the relationship between the canon and large-scale cultural production a

90 

M. TOLONEN ET AL.

Fig. 3.5  The most popular subject-topics for the ten most printed works in each decade from 1500 to 1800

When we look at the most published works per decade (Fig. 3.3), we can generate a complementary list that attempts to estimate which works were “bestsellers.” In this case, no author who makes the top-ten-per-­ decade list has more than four works included. Additionally, the list emphasizes early authors, such as John Stanbridge (1463–1510), Robert Whittington (approximately 1560), Richard Allestree (1619–1681), John Brinsley (active 1581–1624), and Edward Coke (1552–1634), none of whom make the general list. This is likely due both to the lack of competition when printing first emerged and to the high number of reprints of grammars (see also Fig.  3.5). In contrast, authors like Isaac Watts and Daniel Defoe have only two works that make the top-ten-per-decade list. This is an important insight: it shows that there are various ways to integrate the data that allow for both contextual and longue durée insights. When examining the subject-topics by top-ten canon authors (Table 3.4) literary works dominate the list, although there are a few outliers, such as Defoe’s familial advice, The Family Instructor (1715), the conduct piece, Religious Courtship (1722), and Wesley’s medicinal textbook, The Primitive Physick (1747). While temporality plays an obvious role in these results, overall there is a decent spread of authors, covering various time periods and genres (Fig. 3.7).

3  EXAMINING THE EARLY MODERN CANON 

91

Table 3.4  Distribution of subject-topics among works by top authors (1500–1800) Category

Count

Drama Poetry Religion Fiction History & geography Philosophy Sermons Domestic economy Education & manners Medicine & health Miscellaneous literature

34 14 13 10 6 4 4 1 1 1 1

In addition to identifying canonical authors, we have also generated some general statistics regarding the lives of the authors working during the ESTC era. By looking at the authors whose birth and death years are available, we notice that the average age of authors when publishing a first work is quite high (over 40). Additionally, many authors, and especially canonical authors, continue to be published after their death, and, with the advent of the public domain, more and more frequently (Fig. 3.6). Analysis of posthumous publication frequency indicates that being published after death is more common for authors during the early modern period, ancient authors excluded (Fig. 3.6). According to our data, the median number of years to the first publication after death is one; however, for 2.7% of the 1455 authors who were included in this analysis, the first posthumous publication appears over 100 years after their death. The frequency of posthumous publications within the first 50 years after death shows a steadily declining pattern over time.64 For the first half of the sample (until the 1650s) the percentage is higher, but we should remember that at the time it was more difficult to be printed due to limitations in the  print industry. Thus, existing resources were most likely 64  We have used the 50-year window after an author’s death because it removes the bias associated with the fact that later authors have fewer years for republishing (the bias remains for the last 50 years but this is a side issue as most data is directly comparable and the declining trend is clear). In principle, the declining trend could also be explained by increasing intervals of republishing but, in our data, the average publishing time after death is getting systematically smaller, not larger, so this is not a likely explanation.

92 

M. TOLONEN ET AL.

Fig. 3.6  Top authors (1500–1800). The point size indicates the number of publications for each author, including reprints (rows), per year (columns). The color indicates publication before (red) and after (blue) death, respectively. The authors have been sorted by their death year

directed toward printing canonical works, or works by established authors (Fig. 3.7). As noted previously, while most authors we have  extracted from the ESTC can uncontroversially be labeled as canonical, there are some whose inclusion could be contentious. However, as was the case with William Vicker, we may want to reflect on their inclusion before deciding to purge them from the canon. William Lily, for example, is worth noting. Strictly in terms of publication counts, it is understandable why he ranks so high: his Latin Grammar (although written by many hands, including Erasmus) was granted a royal monopoly as the only Latin textbook to be used in schools from 1540 onward. However, while Lily was mainly known as a schoolteacher and grammarian, his contribution to humanist education should not be overlooked.65 Instead, we should acknowledge that his 65  Mary Beth Stewart, “William Lily’s Contribution to Classical Study,” The Classical Journal 33, no. 4 (1938): 217–25, https://www.jstor.org/stable/3291195

3  EXAMINING THE EARLY MODERN CANON 

93

Fig. 3.7  Posthumous publication frequency. The percentage of authors in the ESTC with posthumous publications during the first 50 years after death, grouped by decade (1470–1800). This analysis includes the 1544 authors whose lifetime data is available for the investigated period with one or more posthumous publications

work was a cultural constant amongst all “upper-class British males” for over two centuries, and that its contents, including hundreds of quotes from Roman writers, were known by most educated readers by heart.66 This fact was disapprovingly recorded by John Locke in Some Thoughts Concerning Education (1692): “Custom serves for reason, and has, to those who take it for reason, so consecrated this method, that it is almost religiously observed by them, and they stick to it, as if their children had scarce an orthodox education unless they learned Lilly’s grammar.”67 If one of the aims of a data-driven approach to canon formation is to highlight works that were essential to the cultural space of the time, Lily’s inclusion is an important one—the impact of his grammar was profound, influencing a host of canonical authors, including John Lyly, Ben Jonson, 66  Nancy A.  Mace, “The History of the Grammar Patent from 1620 to 1800 and the Forms of Lily’s Latin Grammar,” The Papers of the Bibliographical Society of America 100, no. 2 (2006): 177, https://www.jstor.org/stable/24293669 67  John Locke, “Some Thoughts Concerning Education” in The Works of John Locke, Esq.; in Three Volumes (London: Printed for John Churchill and Sam. Manship, 1714), 3:73.

94 

M. TOLONEN ET AL.

Fig. 3.8  Timeline of Shakespeare’s publications included in canon. The point size indicates the share of the publisher with most prints of the indicated work (rows) per decade (columns)

Thomas Fuller, George Borrow, Charles Lamb, Edgar Allan Poe, and, of course, William Shakespeare (Fig. 3.8).68 Finally, it is worth turning to, perhaps, a more conservative canonical author who emerges from the above analysis. One benefit of a data-driven approach to the canon is that it enables us to shed new light on the printing and publishing history of specific authors. If we focus on Shakespeare’s publications, for instance, a few points of interest emerge. For instance, while popular during his lifetime and continuously published throughout the seventeenth century, Shakespeare was printed less in the mid-­ seventeenth century. This was partially caused by censorship during the English Civil War and Interregnum years (1642–1659), which was not the most propitious time for printing and performing plays.69 However, the 68  R.  Smith, “Lily, William (1468?–1522/3), Grammarian and Schoolmaster,” Oxford Dictionary of National Biography, 2008, https://doi.org/10.1093/ref:odnb/16665 69  Emma Depledge, “Shakespeare in the Civil War and Interregnum Years, 1642–1659,” in Shakespeare’s Rise to Cultural Prominence: Politics, Print and Alteration, 1642–1700 (Cambridge: Cambridge University Press, 2018), 13–38.

3  EXAMINING THE EARLY MODERN CANON 

95

data also shows that Shakespeare reemerged most strongly in the eighteenth century. This is directly related to the impact of several publishers who, by printing individual plays by Shakespeare, encouraged a growth in the popularity of his works and initiated a newly invented canon-making business (in which Pope, Dryden, Johnson, and others were an important part). Thus, we can see in our data the effect that the publishing efforts of Robert Walker, Jacob Tonson, John Bell, Edward Harding, and others, had on Shakespeare’s canonization. It is, therefore, important that we also discuss these actors. People: Printers and Publishers Developments in the literary canon went hand in hand with developments in the book trade itself. This trade was transformed and driven by the economic expansion that occurred during the seventeenth and eighteenth centuries, when the scale and volume of all kinds of printings greatly expanded. By the end of the eighteenth century, this led to significant changes in the structure of the book trade, one of which was the increased specialization of the people involved.70 Within the existing data regarding book trade actors (printers, publishers, and booksellers), the data on booksellers is the most sporadic. Canonical works, however, provide better information on publishers than the dataset as a whole (Fig. 3.9). Of all the titles cataloged in the ESTC, 27% do not mention any book trade actors; in the subset of titles included in the data-driven canon, this number is 13%. Out of the three categories of book trade actors, publishers are the best represented in the data while booksellers are the least represented, with 85% of titles not mentioning them. The printer information is also missing from 64% of titles. Overall, the number of specialized roles linked to canonical publications increases toward the end of the period, as expected. An observation relating to data quality should be made here. The ESTC joins two major catalogs, STC and Wing, with the dividing line between the original catalogs running at the year 1640. The pre-1640 period seems to have more carefully documented metadata but, at the same time, the number of entries in the database immediately shoots up after the divide. Part of this can be attributed to an increase in the publication activity, but 70  See, for example, James Raven, Publishing Business in Eighteenth-Century England (Woodbridge: Boydell Press, 2014), 40.

96 

M. TOLONEN ET AL.

Fig. 3.9  Titles with missing book trade actor data. This figure charts out the coverage of the book trade actor data, as found in the catalog records (1500–1800)

this is also partially due to the better coverage of the published material in the latter half of the catalog. The number of unique book trade actors mentioned in the catalog follows a more consistent curve compared to the absolute number of titles, which can be taken as an indication that the individual actors involved have been detected reasonably well. Looking at the publications linked to individual book trade actors also reveals a wide variety of profiles. A publisher’s output can vary from a few to hundreds of titles, so it makes sense to create a rough categorization of the publishers based on this variable. To explore this aspect of the book trade from a data-driven perspective, we divided the publishers into percentiles according to their publication output. The publishers were ranked yearly by their output, and as expected, the highest quantiles of the book trade dominate the data. What is immediately apparent is that the top 1–5% publishers account for over half of all publications (where the publishers are known), with no major variance over time (see Fig. 3.10 for a closer look at the first percentile’s share). This matches historical expectations: the discussion around the 1774 Donaldson v. Beckett decision (and the propositions made to the Parliament by both established London booksellers and the “little low Stall Booksellers in Middle Row Holbourn”71) emphasizes the division 71  Thomas F. Bonnell, The Most Disreputable Trade: Publishing the Classics of English Poetry 1765–1810 (Oxford: Oxford University Press, 2008), 35–36.

3  EXAMINING THE EARLY MODERN CANON 

97

Fig. 3.10  Share of publications by the largest publishers (top 1%)

between the economic leaders of the book trade and the numerous, but less established, latter actors. The dividing line runs between the top London publishers, who owned many of the more lucrative and valuable copyrights, and members of the London trade who were not part of this select circle, as well as the less central publishers in Scotland and provincial England.72 The copyright battles of the eighteenth century were not over the right to print in general, but rather over the possession of intellectual property rights that had greater financial potential, such as canonical works. Here, too, the leading publishers had higher percentiles. Their share of the data-­ driven canon is proportionally higher, as illustrated in Fig. 3.11. It has been claimed that the end of the eighteenth century was pivotal in changing the nature of the publishing business, with publishers starting to rely less on profits from “safe” reprints and taking on a more modern, entrepreneurial character.73 The increase in mentions of booksellers in the publishers’ statements could be taken as a confirmation of this claim, as it reflects the commercialization of the print industry in more than one way (see the overview of the actor data in Fig. 3.9). With increasing specialization within the trade, and a growth in the market for books, an increased need for advertising followed. Indeed, publisher statements often include advertising-like language, and provide practical details, such as bookshop  Feather, Publishing, Piracy and Politics, 94.  Ibid.

72 73

98 

M. TOLONEN ET AL.

Fig. 3.11  Share of the canon by largest publishers (top 1%) for unique works, as derived from the work-­ field dataset

locations, lists of other works available, and so on. On the other hand, however, the distribution of canonical and non-canonical works does not significantly change during the eighteenth century. The canon stayed in relatively few hands, even after 1774. While there was a gradual increase in the “fluidity” of the publication business during the eighteenth century, that is, works tended to change hands more often, this trend is relatively temperate, with no sudden hike in the 1770s. Additionally, while the relative number of new works and reprints by new actors compared to reprints by established actors does increase (Fig. 3.12), the changes are less dramatic than it must have appeared to the worried copyright-­owning printing elite of the time.74 In fact, consumers’ demand for books increased dramatically toward the end of the eighteenth century, which can be seen in the number of printings documented in the ESTC for that period.75 Thus, even if the established elite was challenged by those seeking to make inroads into an expanding market, they were well positioned to defend their hegemony and exploit the new opportunities created by the public’s increased demand for books.  Bonnell, The Most Disreputable Trade, 169.  See, for example, Raven, Publishing Business in Eighteenth-Century England, 214–18, and Karen O’Brien, “The History Market in Eighteenth-Century England,” in Books and Their Readers in Eighteenth-Century England: New Essays, ed. Isabel Rivers (London: Continuum, 2001), esp. 123–30. 74 75

3  EXAMINING THE EARLY MODERN CANON 

99

Fig. 3.12  Publishing and reprint patterns by publisher role in the printing sequence. This figure indicates changes in the reprint publishing patterns by different publishers over time and charts out the “fluidity” of the book trade. Each work had its publishers explored in chronological sequence to find out how often the publications changed hands. New publications (“New work”) and new editions of the same works by the same publishers (“Stable publisher”) were traced. Publications changing hands were traced both in the cases where the previous publisher disappeared from the book trade (“New publisher, old inactive”) and where the publication changed hands, but the previous publisher stayed active (“New publisher, old active”). Cases of the publication returning to the hands of a previous owner are relatively rare (“Return of earlier publisher”)

It is also known that, after 1774, other publishing monopolies, such as those covering almanacs, grammars, and law books, were under attack.76 This highlights the importance of specialization within the trade and the significance of printed materials outside the subject-topics generally seen as canonical. When looking at the division between these categories, we can see that many publishers specialized in relatively limited subject-topics (Fig.  3.13). At the same time, new subject-topics were clearly part of a broader portfolio of publications. This means that we have different types  Feather, Publishing, Piracy and Politics, 94–95.

76

Fig. 3.13  Publisher subject-topic specialization and canon share. This figure illustrates the differences and similarities in the publishing landscapes of the identified subject-topics (individual scatterplots). Each dot represents an individual publisher, the horizontal axis indicates the publisher’s topic specialization, the vertical axis indicates the portion of all publications by a publisher included in the canon, and the size of the dot indicates the publication volume of a publishers in a particular field. The large dot in the center right in “Information, general works” represents the Stationers’ Company

100  M. TOLONEN ET AL.

3  EXAMINING THE EARLY MODERN CANON 

101

of publishers throughout the early modern era: those who specialized and those who published in response to  demand. Literature, religion, social sciences, and, to some extent, information and general works (almanacs and the like) appear to be fields where the market was large enough to accommodate specialization, and a limited group of well-established publishers were occasionally able to dominate these markets based on monopoly rights. A good example of this phenomenon is the single most voluminous publisher of the time, the Stationers’ Company of London, which dominated the information and general works publishing market of eighteenth-century London. Gender and the Book Trade While it would be exciting to claim that our data-driven approach has allowed for a reassessment of the gender imbalance (amongst other imbalances) in the history of the canon, this is not the case. As seen in Fig. 3.14, most canonical authors were male. In fact, only 1 in every 32 authors is female, and men have on average 38 reprints per work compared with only 31 by women. Within the book trade as a whole the gender imbalance is not as great, although it is still significant: 1  in roughly 14 book trade actors in our data is female.

Fig. 3.14  Works by female authors in the data-driven canon per decade

102 

M. TOLONEN ET AL.

In total, 21 authors in the canon are female, and only three of them have multiple works included: Susanna Centlivre (1667?–1723) with three, and Elizabeth Singer Rowe (1674–1737) and Hannah More (1745–1833) with two each. The remaining authors have only one work each: Hannah Glasse (1708–1770), Lady Mary Wortley Montagu (1689–1762), Anne Fisher (1719?-1778), Mary Collyer (?−1763), Madame Marie-Catherine d’Aulnoy (1650 or 1651–1705), Elizabeth Raffald (1733–1781), Madame Jeanne-Marie Leprince de Beaumont (1711–1780), Hester Chapone (1727–1801), Catherine Talbot (1721–1770), Frances Brooke (1724?–1789), Eliza Smith (died cca.1732), Frances Burney (1752–1840), Madame de Gomez (1684–1770), Madame de Graffigny (1695–1758), Elizabeth Moxon (life years not available), Mary Brook (cca. 1726–1782), Anna Letitia Barbauld (1743–1825), and Elizabeth Grey, Countess of Kent (1581–1651). There is, however, a clear growth in the number of female authors and in the number of works by female authors that make up the canon over time. In fact, if we only look at the eighteenth century—which is relevant in this case as there is only one female in the canon prior to this period, Elizabeth Grey, Countess of Kent—the split is less severe, albeit still unequal: the disparity between reprints drops by nearly two works, with one female author for every 18 male authors. The subjects covered by female authors include domestic economy, drama, education and manners, fiction, language, miscellaneous literature, and religion. Since the number of female-authored works is quite small, it is somewhat difficult to compare their subject coverage with that of the larger number of male-authored works. However, we do see that, in general, more women write fiction, and fewer women write educational and religious works.77 Places The centrality of London (and of the Stationer’s Company) to the early modern English book trade is well documented.78 This reality is visible in 77  These findings confirm and supplement the findings of William Underwood, David Bamman, and Sabrina Lee related to the nineteenth-century authorship. For details, see “The Transformation of Gender in English-Language Fiction,” Journal of Cultural Analytics (2018): 1–25, https://doi.org/10.22148/16.019 78  For more on London’s place in the history of the book trade, see James Raven, The Business of Books: Booksellers and the English Book Trade 1450–1850 (New Haven, CT: Yale

3  EXAMINING THE EARLY MODERN CANON 

103

Fig. 3.15  Fraction of publications by place for the top publication places excluding London (1500–1800)

the publication records extracted from the ESTC as well, with London publishing more works by a factor of ten to its closest rival, Edinburgh. What is more, the non-London-based publication industry only begins to mature toward the end of the seventeenth century (and even then, works printed in London continue to dominate local markets).79 As we see in Fig. 3.15, the early prints outside London are coming almost entirely from Paris, with Edinburgh, Cambridge, and Oxford dominating the seventeenth century, before Dublin and North American publishers begin to emerge in the eighteenth century.  When it comes to the canon, however, a different geographical picture emerges, as seen in Table 3.5. There are at least two findings revealed by this data: first, the importance of particular areas as producers of canonical works, and, second, the importance of particular areas as producers of reprints of popular works. Regarding the former, London unsurprisingly dominates the print market. In fact, of the 30 cities which are recorded as sources of first editions University Press, 2007). See also the works referenced in Hill, Vaara, Säily, Lahti, and Tolonen, “Reconstructing Intellectual Networks.” 79  James Green, “The British Book in North America,” in Suarez and Turner, The Cambridge History of the Book in Britain. Volume V, 544–59.

104 

M. TOLONEN ET AL.

Table 3.5  Top 10 printing locations in the whole ESTC and in the canon (1500–1800) ESTC Records

Canon Total Prints

Location

Count

Location

Count

London Edinburgh Dublin Boston, MA Philadelphia, PA Oxford Glasgow New York, NY Cambridge Newcastle

318,708 32,146 26,273 10,665 10,282 7418 5358 4716 4371 2726

London Dublin Edinburgh Glasgow Boston, MA Philadelphia, PA Oxford Paris Cambridge New York, NY

21,273 2668 1778 1071 665 646 331 289 251 232

of canonical works, London is responsible for 606, or roughly 84% of all titles. With regard to the movement of individual prints of canonical works, this means that more than ten times as many works originally printed in London were subsequently printed elsewhere than the other way around. Of course, London was the political, financial, and cultural capital of Britain, so these findings may not be surprising. However, what is surprising is the magnitude of this imbalance, with London dwarfing the rest of the canon. Edinburgh, the center of the Scottish Enlightenment, comes a distant second as the source for first editions that will become canonical, with 38 titles (Table 3.6).80 On the other hand, there are cities that are centers for reprinting editions of canonical works which were not originally printed there. Edinburgh fits this case, with the second highest number of reprints (288), followed by its Scottish compatriots Glasgow (252) and Aberdeen (57). The key player in reprints of canonical works, however, was Dublin. While Dublin did have a number of first editions (18), these are disproportionate to the number of subsequent editions it printed (427, the most of any city). This is almost certainly due to Dublin’s privileged legal and cultural position: 80  It should be noted, however, that the place of publication for 125 first editions is unknown due to the lack of detail in the ESTC about first editions, or due to multiple cities publishing editions of the same work in the same year. Also, our counts are by work and do not include reprints or editions beyond the first. Therefore, the highest number any one city can achieve is the total number of works counted in our canon, that is, 847.

3  EXAMINING THE EARLY MODERN CANON 

105

Table 3.6  Locations for the first printed editions and for subsequent editions of canonical works (1500–1800) Canon: First editions

Canon: Non-local subsequent editions

Location

Count

Location

Canon

London Edinburgh Dublin Oxford Cambridge Pairs Boston, MA Philadelphia, PA Amsterdam Manchester

599 37 18 12 8 5 4 3 2 2

Dublin Edinburgh Glasgow Philadelphia, PA Boston, MA London New York, NY Belfast Oxford Aberdeen

427 288 252 148 139 95 74 71 62 57

being beyond the reach of the British law and the Stationers’ grasp, and having a large English-speaking population, meant that Dublin’s printers were in a privileged position. And while there was certainly a local market for these works, it is also clear that Dublin was not the only intended market for these works, and many were exported to Britain and beyond. Thus, while we know that Dublin was an importer of canonical works first printed elsewhere, it was also an important exporter of these works. A similar, yet smaller, pattern is also seen in the colonies, with Philadelphia, Boston, and New  York reprinting numerous non-local works. Overall, the movement of works between Europe and the colonies remains largely unidirectional, according to the ESTC records (Fig. 3.16).81 On the whole, however, there is a remarkable movement of various editions of these works among various locations. While books obviously traveled as any other material object would, the works themselves as less tangible things were recreated in various locations throughout the world, being reprinted by local book trade actors for commercial and intellectual profit.

81  Base Mercator projection map used for visualization created by Geordie Bosanko and shared under the CC BY-SA 3.0 license at https://commons.wikimedia.org/wiki/ File:Mercator_Projection.svg. This base map was further edited by Adrienne Hawkes.

106 

M. TOLONEN ET AL.

Fig. 3.16  Movement of canonical editions from original print location

Population must also be taken into consideration when looking at publication counts. When one considers the size of London compared to that of other cities at the time, its dominance may not be surprising. However, size does not appear to be the key contributing factor to the number of prints, relative to population size, that emerge from a city. As seen in Fig.  3.17, smaller cities were capable of producing many more works than larger cities in both absolute and relative terms. The reasons for this vary. University cities like Oxford and Cambridge printed for specific niche markets, while colonial cities developed their own markets, which were less reliant on imports.82 Dublin and Glasgow were also important reprint centers, although we can see a decline in their print output compared to London following the liberalization of the market in 1774. While we have already touched upon Dublin’s relationship to

82  Bristol’s inclusion in Fig.  3.17 can be attributed almost exclusively to John Wesley, a prodigious writer who had an important impact on both British and colonial print markets in the early modern era. Interestingly, the publication place of over 300 of his works (more than 20%) was Bristol (comparatively, only Swift had more works published outside of London, i.e., in his hometown, Dublin). Moreover, when looking at Bristol’s publishing record, Wesley accounts for almost 20% of all publications from Bristol in the ESTC. When John’s brother, Charles, is included, this number increases to over 25% of all works from Bristol in the ESTC.

3  EXAMINING THE EARLY MODERN CANON 

107

Fig. 3.17  Number of prints per capita (1700–1800)

reprints, Glasgow is also worth noting given that its print history is arguably tied to the production of editions of “classics.”83 Glasgow’s dominance in canonical printing (Figs. 3.17 and 3.18) can be tied to a specific printing house run by the brothers Robert and Andrew Foulis.84 Initially elected as printers to the University of Glasgow in 1743 83  Thomas F. Bonnell, “The Elzevirs of Glasgow: Robert and Andrew Foulis,” in The Most Disreputable Trade, 39–67. 84  For more on the impact of the Foulis brothers on the eighteenth-century Scottish intellectual life, see Steinberg, Five Hundred Years of Printing, 139–41.

108 

M. TOLONEN ET AL.

Fig. 3.18  The fraction of canonical editions compared to all editions per city (1700–1800)

and charged with printing works by ancient authors, by the 1750s the Foulis brothers had turned to printing Elzeviresque editions of English classics by Milton and Gray, before engaging in mass market reprints of dubious legality. With the deaths of Andrew in 1775 and Robert in 1776, the period of Glaswegian dominance ends, although Robert’s son, also Andrew, was able to revive the business to some extent from the 1780s on, with a renewed focus on exporting books to the American market. Importantly, their genius was found not only in whom they printed but also in how they printed. While their editions of the ancients were often expensive folios or quartos, their reprints of contemporary English poets were duodecimos: they were cheaper to print and thus more affordable to a broader audience. As Thomas F.  Bonnell notes, “they had crossed a divide from an old world of monumental scholarly and typographical ventures devoted to ancient Greek and Latin texts into a new world of selling multi-volume collections of modern vernacular classics to a larger and

3  EXAMINING THE EARLY MODERN CANON 

109

more diverse readership.”85 It is this transformation of the print trade which we turn to next. Materiality There is one further aspect of print that is worth taking into consideration when thinking about the canon: materiality, or the composition of a printed work, such as page size, format, number of pages, and print area.86 This is, again, an aspect of print which is often overlooked when one engages with historic works quantitatively.87 However, the material attributes of a printed work have meaning to a reader, as made explicit by Joseph Addison in issue 529 of The Spectator: “I have observed that the Author of a Folio, in all Companies and Conversations, sets himself above the Author of a Quarto; the Author of a Quarto above the Author of an Octavo; and so on, by a gradual Descent and Subordination, to an Author in Twenty Fours.”88 In other words, within the materiality of a work, there were ingrained literary and social signifiers that contemporaries would have been familiar with. When evaluating the most dominant book formats attached to subject-­ topics over time, the significance of the smaller book format emerges. In particular, we notice the growth of the octavo and duodecimo formats (Fig. 3.19), which were more portable, easily fitting in one’s pocket, and thus more easily perused or read at different occasions. While the trend can be seen across all subject-topics, it is particularly visible in the case of  Bonnell, The Most Disreputable Trade, 62.  Print area quantifies the paper consumption in sheets for a unique copy of a document; the combined print area across different documents in a given time period can be used to quantify the breadth of the printing activity. 87  For more work on materiality, see Tolonen, Lahti, Roivainen, and Marjanen, “A Quantitative Approach to Book-Printing in Sweden and Finland, 1640–1828,” 57–78; Lahti, Ilomäki, and Tolonen, “A Quantitative Study of History in the English Short-Title Catalogue (ESTC), 1470–1800,” 87–116; Lahti, Marjanen, Roivainen, and Tolonen, “Bibliographic Data Science and the History of the Book (c. 1500–1800),” 10–23; and Eetu Mäkelä, Mikko Tolonen, Jani Marjanen, Antti Kanner, Ville Vaara, and Leo Lahti, “Exploring the Material Development of Newspapers,” TwinTalks at DHN 2018 Understanding Collaboration in Digital Humanities, https://cst.dk/DHN2019Pro/TwinTalks WorkshopProceedings.pdf 88  The Spectator No. 529 Thursday, November 6, 1712. Oxford Scholarly Editions Online, 2014, https://www.oxfordscholarlyeditions.com/view/10.1093/actrade/9780198186137. book.1/actrade-9780198186137-div1-104 85 86

110 

M. TOLONEN ET AL.

Fig. 3.19  Dominant book formats for the most frequent subject-topics

3  EXAMINING THE EARLY MODERN CANON 

111

Fig. 3.20  The distribution of common book formats for selected canon-works over time

literary and philosophical works. Legal and public administration books, on the other hand, were slower to change, remaining in folio for much longer. Interestingly, we have also noticed a regional variation with respect to preferred formats, with duodecimo being the most popular format in the New World. When looking at the changing materiality of the canon with respect to some of the most popular works published between 1500 and 1800, we notice that early publishing mixed folio, quarto, and octavo formats. This trend continues, in some cases, until the end of the eighteenth century (see, for instance, Paradise Lost, in Fig. 3.20), but with time the octavo and duodecimo formats become dominant (see Aesop’s Fables and Short Introduction to Grammar in Fig. 3.20). The choice of format, therefore, is tied to complex relations among economic viability, perceived importance, and pragmatics. A folio edition of Milton, for example, was a worthy and desirable endeavor in the late eighteenth century, while a grammar was much less likely to be imbued with the same subjective value and was, therefore, more convenient to own in a smaller format. Thus, this type of analysis allows us not only to touch upon the realities of printing practices but also to gain insight into the changing preferences of the public with respect to materiality and early modern reading habits.89 At the same time, when looking at the overall paper consumption for different formats of books included in the canon, we can see that octavo and duodecimo  See, for example, Reinhard Wittmann, “Was There a Reading Revolution at the End of the Eighteenth Century?” in A History of Reading in the West, ed. Guglielmo Cavallo and Roger Chartier (Amherst, MA: University of Massachusetts Press, 1999), 284–312. 89

112 

M. TOLONEN ET AL.

Fig. 3.21  Estimated paper consumption for different formats over time for books included in the canon

formats started to dominate the book printing market only in the second half of the eighteenth century (Fig. 3.21).

Conclusion The goal of this study has been to extract from the ESTC a data-driven canon which could be used to demonstrate that quantitative investigations of this type are valuable for historical research. While quantitative analyses of the history of the book trade exist, there has been no attempt so far to engage with the complex process of canon formation at such a large scale.90 To this effect, we have constructed a method for extracting a list of “canonical” works from the ESTC based on three publication features: count, frequency, and longevity. We have thus generated a data-driven list of canonical works that considers subject-topics, top-works, authors and publishers, publication place, and materiality from a historical perspective. 90  It should be noted that this is an ongoing process; we continue working on further harmonizing the data.

3  EXAMINING THE EARLY MODERN CANON 

113

While we believe this quantitative approach is in itself a methodological contribution worth reporting, it also allows us to make a number of historical claims worth studying further. At the same time, it is important to recognize the limitations of this type of analysis. Data reliability, representativeness, and completeness will improve over time, and this will influence all quantitative estimates derived from the data. Algorithmic questions, such as the exact definition of the canon index or genre classification, and choices made in model parameters, such as investigated time window, will also affect this analysis. From our perspective, however, making such interpretations explicit allows one to evaluate these choices and propose alternative solutions. The reproducibility of the analysis will then allow us (and others) to explore how sensitive the qualitative conclusions are to different analytical choices. Here, however, we have primarily focused on broad historical patterns and trends that are expected to remain stable to variations in data and algorithmic details. When examining the early modern English canon from this data-driven perspective, it becomes obvious that an epistemological shift takes place during the late seventeenth century-early eighteenth century, when religious works lose their dominant position within the canon and are increasingly replaced by literary works (Figs. 3.4, 3.5, 3.7, and 3.8). Although literature in all its forms was historically an important part of the canon, changes in its production and consumption allowed for its growth in the eighteenth century.91 Additionally, this analysis allows us to highlight the essential role played by the publisher in the process of canon formation, besides that of the literary critic. In particular, the role of the elite, London-based publishers (Fig. 3.11) and that of the arguably more dubious printers operating outside of London (Table 3.5, Figs. 3.17 and 3.18) become evident. Overall, we can now visualize the lengthy and arduous process of a work becoming canonical. While the total number of publications grew exponentially during the latter half of the eighteenth century, the distribution of canonical works remained relatively stable in comparison (Fig. 3.1), which indicates that the works that are most often reprinted over long stretches of time are comparatively few. This is perhaps a finding worth reflecting on further: 91  Other historical moments, Donaldson v. Becket in particular, are also visible in our data (Figs.  3.17 and 3.18). However, as shown in the canon comparisons, the impact of such moments on the ESTC from a data-driven canon perspective should not be overestimated.

114 

M. TOLONEN ET AL.

this description of large-scale cultural production and competition within the literary market, which directs the canonization process, may allow scholars of the period to extrapolate further and use these statistics to develop prediction models for an author’s or a work’s likelihood of becoming canonical. The broader claim of this chapter is that the development of the print market as a cultural producer has driven the changes we are able to witness in the ESTC when studied in a data-driven manner. This builds on previous work by Bourdieu, Anderson, and Habermas, who tied print capitalism to historical, cultural, political, and social changes.92 Our contribution is to apply quantitative methods to demonstrate the accuracy of these qualitatively-grounded studies in a manner which has not been attempted before. As this analysis demonstrates, what Wendell V. Harris wrote more than thirty years ago is truer today than ever: “The ‘canon question’ … proves much more complex than contemporary ideological criticism admits.”93 While large-scale cultural production is certainly a key factor in the canon-­ making process, were it to be taken as the only factor, we would dismiss numerous individual voices and would offer yet another version of the revisionist approach to canon formation. There is no such thing as an “absolute” canon, only different takes on it. While it is inevitable that different works matter in different ways, our main concern in this study is with the impact of print culture on canon formation. By considering these recorded works and their historical availability over extended periods of time, we hope to offer a more nuanced understanding not only of the history of the book trade but also of the cultural context from which it emerged.

92  According to Bourdieu, there are two important factors that make a difference in the print market: the restricted production of literature for like-minded audiences, and the largescale cultural production. We have limited our study to the latter category because of gaps in the information related to particular works. For example, in Shakespeare’s case, we cannot judge whether his late eighteenth-century success is due to earlier restricted production. We may only note that Shakespeare enters large scale cultural production after a gap in publishing his works in the seventeenth century, as seen in Fig. 3.8. 93  Harris, “Canonicity,” 115.

3  EXAMINING THE EARLY MODERN CANON 

115

Bibliography Altieri, Charles. 1983. An Idea and Ideal of a Literary Canon. Critical Inquiry 10 (1): 37–60. https://www.jstor.org/stable/1343405. Anderson, Benedict. 1983. Imagined Communities: Reflections on the Origin and Spread of Nationalism. New York: Verso. Barnard, John, D.F.  McKenzie, and Maureen Bell, eds. 2002. The Cambridge History of the Book in Britain. Volume IV: 1557–1695. Cambridge: Cambridge University Press. Bell, Maureen, and John Barnard. 1992. Provisional Count of STC Titles, 1475–1640. Publishing History 31: 47–66. Bloom, Harold. 1994. The Western Canon: The Books and School of the Ages. Harcourt: Brace. Bonnell, Thomas F. 2008. The Most Disreputable Trade: Publishing the Classics of English Poetry 1765–1810. Oxford: Oxford University Press. Bourdieu, Pierre. 1983. The Field of Cultural Production, or the Economic World Reversed. Poetics 12 (4–5): 331–356. https://doi. org/10.1016/0304-422X(83)90012-8. ———. 1985. The Market of Symbolic Goods. Poetics 14 (1–2): 13–44. https:// doi.org/10.1016/0304-422X(85)90003-8. Bourdieu, Pierre, and Jean-Claude Passeron. 1977. Reproduction in Education, Society and Culture. Trans. Richard Nice. London: Sage. Bradford, Richard. 2015. Is Shakespeare Any Good? And Other Questions on How to Evaluate Literature. Malden: Wiley-Blackwell. Bruni, Flavia, and Andrew Pettegree, eds. 2016. Lost Books: Reconstructing the Print World of Pre-Industrial Europe. Leiden: Brill. Carroll, Noel. 1984. Hume’s Standard of Taste. The Journal of Aesthetics and Art Criticism 43 (2): 181–194. https://doi.org/10.2307/429992. Collins, Irene. 1998. The Rev. Henry Tilney, Rector of Woodston. Persuasions 20: 154–164. Collinson, Patrick, Arnold Hunt, and Alexandra Walsham. Religious Publishing in England 1557–1640. In The Cambridge History of the Book in Britain. Volume IV, ed. Barnard, McKenzie, and Bell, 29–66. Curtius, Ernst Robert. 2013. European Literature and the Latin Middle Ages. Princeton: Princeton University Press. Depledge, Emma. 2018. Shakespeare’s Rise to Cultural Prominence: Politics, Print and Alteration, 1642–1700. Cambridge: Cambridge University Press. Disraeli, Isaac. 1823. Curiosities of Literature. Vol. 4. 7th ed. London: John Murray. Eagleton, Terry. 1985. The Subject of Literature. Cultural Critique 2: 95–104. https://doi.org/10.2307/1354202. Eisenstein, Elizabeth. 1979. The Printing Press as an Agent of Change: Communications and Cultural Transformations in Early Modern Europe. Cambridge: Cambridge University Press.

116 

M. TOLONEN ET AL.

Fairer, David. 2000. Historical Criticism and the English Canon: A Spenserian Dispute in the 1750s. Eighteenth-Century Life 24 (2): 43–64. https://doi. org/10.1215/00982601-24-2-43. Feather, John. 1995. Publishing, Piracy and Politics: An Historical Study of Copyright in Britain. London: MansFigell. Febvre, Lucien, and Henri-Jean Martin. 1976. The Coming of the Book: The Impact of Printing, 1450–1800. New York: Verso. Fielder, Leslie A., and Houston A. Baker. 1979. English Literature: Opening Up the Canon. Baltimore: Johns Hopkins University Press. Fowler, Alastair. 1979. Genre and the Literary Canon. New Literary History 11 (1): 97–119. https://doi.org/10.2307/468873. Gants, David L. 2002. A Quantitative Analysis of the London Book Trade, 1614–1618. Studies in Bibliography 55 (1): 185–213. https://doi. org/10.1353/sib.2005.0004. Gorak, Jan, ed. 2001. Canon vs. Culture: Reflections on the Current Debate. New York: Garland. ———. 2013. The Making of the Modern Canon: Genesis and Crisis of a Literary Idea. London: Bloomsbury. Green, Ian, and Kate Peters. Religious Publishing in England c. 1640–1695. In The Cambridge History of the Book in Britain. Volume IV, ed. Barnard, McKenzie, and Bell, 67–94. Green, James. The British Book in North America. In The Cambridge History of the Book in Britain. Volume V, ed. Suarez and Turner, 544–559. Guillory, John. 1993. Cultural Capital: The Problem of Literary Canon Formation. Chicago: University of Chicago Press. Habermas, Jürgen. 2003. The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society. Cambridge: Polity Press. Halasz, Alexandra. 1997. The Marketplace of Print: Pamphlets and the Public Sphere in Early Modern England. Cambridge: Cambridge University Press. Harris, Wendell V. 1991. Canonicity. PMLA 106 (1): 110–121. https://doi. org/10.2307/462827. Hill, Mark J. 2016. Invisible Interpretations: Reflections on the Digital Humanities and Intellectual History. Global Intellectual History 1 (2): 130–150. https:// doi.org/10.1080/23801883.2017.1304162. Hill, Mark J., Ville Vaara, Tanja Säily, Leo Lahti, and Mikko Tolonen. 2019. Reconstructing Intellectual Networks: From the ESTC’s Bibliographic Metadata to Historical Material. In Proceedings of the Digital Humanities in the Nordic Countries 4th Conference: Copenhagen, Denmark, March 5–8, 2019, ed. Constanza Navarretta, Manex Agirrezabal, and Bente Maegaard, 201–219. Aachen: CEUR-WS.org. Hoxby, Blair. 2015. What Was Tragedy? Theory and the Early Modern Canon. Oxford: Oxford University Press. Hume, David. 1987. Of the Standard of Taste. In Essays Moral, Political, Literary, ed. Eugene F. Miller, 226–250. Indianapolis: Liberty Fund.

3  EXAMINING THE EARLY MODERN CANON 

117

Karian, Stephen. 2011. The Limitations and Possibilities of the ESTC. In The Age of Johnson: A Scholarly Annual, ed. Jack Lynch, vol. 21, 283–297. New York: AMS Press. Kramnick, Jonathan Brody. 1999. Making the English Canon: Print-Capitalism and the Cultural Past, 1700–1770. Cambridge: Cambridge University Press. Lahti, Leo, Niko Ilomäki, and Mikko Tolonen. 2015. A Quantitative Study of History in the English Short-Title Catalogue (ESTC), 1470–1800. LIBER Quarterly 25 (2): 87–116. https://doi.org/10.18352/lq.10112. Lahti, Leo, Jani Marjanen, Hege Roivainen, and Mikko Tolonen. 2018. Bibliographic Data Science and the History of the Book (c. 1500–1800). Cataloging & Classification Quarterly 57 (1): 5–23. https://doi.org/10.108 0/01639374.2018.1543747. Lahti, Leo, Ville Vaara, Jani Marjanen, and Mikko Tolonen. 2019. Best Practices in Bibliographic Data Science. In Proceedings of the Research Data in the Humanities 2019 Conference: Data, Methods, and Tools, ed. Jarmo Harri Jantunen, Sisko Bruni, Kunnas Niina, Santeri Palviainen, and Katja Västi, 57–65. Oulu: Faculty of Humanities Linnanmaa. Levin, Harry. 1981. Core, Canon, Curriculum. College English 43 (4): 352–362. https://doi.org/10.2307/377120. Lopez, Jeremy. 2014. Constructing the Canon of Early Modern Drama. Cambridge: Cambridge University Press. Low, Gail, and Marion Wynne-Davies, eds. 2006. A Black British Canon? Hampshire: Palgrave Macmillan. Mace, Nancy A. 2006. The History of the Grammar Patent from 1620 to 1800 and the Forms of Lily’s Latin Grammar. The Papers of the Bibliographical Society of America 100 (2): 177–225. https://www.jstor.org/stable/24293669. Mäkelä, Eetu, Mikko Tolonen, Jani Marjanen, Antti Kanner, Ville Vaara, and Leo Lahti. Exploring the Material Development of Newspapers. TwinTalks at DHN 2018 -Understanding Collaboration in Digital Humanities. https://cst.dk/ DHN2019Pro/TwinTalksWorkshopProceedings.pdf Mukherjee, Ankhi. 2013. What Is a Classic? Postcolonial Rewriting and Invention of the Canon. Stanford: Stanford University Press. Munch, Thomas. 2019. Conflict and Enlightenment: Print and Political Culture in Europe, 1635–1795. Cambridge: Cambridge University Press. O’Brien, Karen. 2001. The History Market in Eighteenth-Century England. In Books and Their Readers in Eighteenth-Century England: New Essays, ed. Isabel Rivers, 105–134. London: Continuum. Patey, Douglas Lane. 1988. The Eighteenth Century Invents the Canon. Modern Language Studies 18 (1): 17–37. https://doi.org/10.2307/3194698. Pettegree, Andrew, ed. 2017. Broadsheets: Single-Sheet Publishing in the First Age of Print. Leiden: Brill. Proudfoot, Richard. 2001. Shakespeare: Text, Stage & Canon. London: Arden Shakespeare.

118 

M. TOLONEN ET AL.

Raven, James. 2007. The Business of Books: Booksellers and the English Book Trade 1450–1850. New Haven: Yale University Press. ———. 2014. Publishing Business in Eighteenth-Century England. Woodbridge: Boydell Press. Robinson, Lillian. 1997. The Canon’s Mouth: Dispatches from the Culture Wars. Bloomington: Indiana University Press. Robinson, Peter. 2008. Introduction to Textual Scholarship and the Canon, ed. Hans Gabler, Peter Robinson Walter, and Paulius Subačius, 1–8. Leiden: Brill. Ross, Trevor Thornton. 1998. The Making of the English Literary Canon: From the Middle Ages to the Late Eighteenth Century. Montreal: McGill-Queens University Press. Rounce, Adam. 2015. Authorship in the Eighteenth Century. Oxford Handbooks Online. https://doi.org/10.1093/oxfordhb/9780199935338.013.38. Rousseau, Jean Jacques. Politics and the Arts: Letter to M.  D’Alembert on the Theatre. Trans. Allan Bloom. Ithaca: Cornell University Press, 1968. Salzman, Paul, ed. 2010. Expanding the Canon of Early Modern Women’s Writing. Newcastle upon Tyne: Cambridge Scholars Publishing. Sher, Richard. 2007. The Enlightenment and the Book. Scottish Authors and Their Publishers in Eighteenth-Century Britain, Ireland, and America. Chicago: University of Chicago Press. Skinner, Quentin. 2002. Visions of Politics. Volume I: Regarding Method. Cambridge: Cambridge University Press. Smith, R. 2008. Lily, William (1468?–1522/3), Grammarian and Schoolmaster. Oxford Dictionary of National Biography. https://doi.org/10.1093/ ref:odnb/16665. Spencer, Jane. 2005. Literary Relations: Kinship and the Canon 1660–1830. Oxford: Oxford University Press. Stallybrass, Peter. 2004. The Library and Material Texts. PMLA 119 (5): 1347–1352. https://doi.org/10.1632/003081204X17914. Steinberg, Sigfrid H. 1955. Five Hundred Years of Printing. London: Penguin. Stewart, Mary Beth. 1938. William Lily’s Contribution to Classical Study. The Classical Journal 33 (4): 217–225. https://www.jstor.org/stable/3291195. Suarez, Michael F. 2015. Book History from Descriptive Bibliographies. In The Cambridge Companion to the History of the Book, ed. Leslie Howsam, 199–219. Cambridge: Cambridge University Press. Suarez, Michael F.  Towards a Bibliometric Analysis of the Surviving Record, 1701–1800. In Suarez and Turner, The Cambridge History of the Book in Britain. Volume V, 37–65. Suarez, Michael F., and M.L. Turner. 2009. The Cambridge History of the Book in Britain. Volume V: 1695–1830. Cambridge: Cambridge University Press. Terry, Richard. 1997. Literature, Aesthetics, and Canonicity in the Eighteenth Century. Eighteenth-Century Life 21 (1): 80–101. https://www.muse.jhu. edu/article/10407.

3  EXAMINING THE EARLY MODERN CANON 

119

The Works of John Locke, Esq.; in Three volumes. 1714. London: Printed for John Churchill and Sam. Manship. Thomson, Alex. 2014. What Is Literature? In The Edinburgh Introduction to Studying English Literature, ed. Dermot Cavanagh, Alan Gillis, Michelle Keown, James Loxley, and Randall Stevenson, 3–16. Edinburgh: Edinburgh University Press. Tolonen, Mikko, Leo Lahti, Hege Roivainen, and Jani Marjanen. 2019. A Quantitative Approach to Book-Printing in Sweden and Finland, 1640–1828. Historical Methods: A Journal of Quantitative and Interdisciplinary History 52 (1): 57–78. https://doi.org/10.1080/01615440.2018.1526657. Underwood, William, David Bamman, and Sabrina Lee. 2018. The Transformation of Gender in English-Language Fiction. Journal of Cultural Analytics: 1–25. https://doi.org/10.22148/16.019. van Oostrom, Frits. 2007. A Key to Dutch History: The Cultural Canon of the Netherlands. Amsterdam: Amsterdam University Press. Weedon, Alex. 2009. The Uses of Quantification. In A Companion to the History of the Book, ed. Simon Eliot and Jonathan Rose, 33–49. Malden: Wiley Blackwell. Werner, Sarah. 2019. Studying Early Printed Books, 1450–1800: A Practical Guide. Malden: Wiley Blackwell. Wittmann, Reinhard. 1999. Was There a Reading Revolution at the End of the Eighteenth Century? In A History of Reading in the West, ed. Guglielmo Cavallo and Roger Chartier, 284–312. Amherst: University of Massachusetts Press.

Open Access  This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

CHAPTER 4

Europe and Its “Others”: Visualizing Lexical Relations Between Western and Non-Western Locations of the Enlightenment in The Eighteenth-Century Collections Online John Regan

Whenever we open our mouths, it is to speak of “enlightenment.” —Japanese reformer Tsuda Mamichi (1829–1903)

Digital inquiries are well placed to further the ongoing project of challenging Eurocentric views of the Enlightenment. With the creation of historical corpora such as The Electronic Text Corpus of Sumerian Literature Qtd. in Albert M.  Craig, Civilization and Enlightenment: The Early Thought of Fukuzawa Yukichi (Cambridge, MA: Harvard University Press, 2009), 147. The implications of Mamichi’s statement are discussed in Sebastian Conrad, “Enlightenment in Global History: A Historiographical Critique,” The American Historical Review 117, no. 4 (2012): 999–1027, https://doi.org/10.1093/ ahr/117.4.999. J. Regan (*) Royal Holloway, University of London, Egham, UK e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_4

121

122 

J. REGAN

or The Japanese History Corpus, new apertures are opening onto how knowledge was structured and the Enlightenment fashioned and refashioned outside the European Republic of Letters.1 Approaches that can broadly be gathered under the umbrella of distant reading allow qualitatively new and novel views onto the fault lines and main currents of thought in various regional Enlightenments. As a result, we learn more about the non-western Enlightenments all the time. But there is still merit in inquiring into how European authors constructed knowledge about these “other” environments, although such an effort does not come without its challenges. Bradley Naranch has described the hindrances to examining the diffusion of western Enlightenment thought and values through the non-European world in the following terms: “The primary challenges to assessing the Enlightenment’s significance for global history have been interpretive in nature. The most significant of these come from postcolonial critics, who argue that all Enlightenment thinkers, even the most radical among them, were complicit in perpetuating discourses that confirmed notions of racial superiority, supported Eurocentric views of civilizational exceptionalism, and enabled ‘enlightened’ versions of European colonial power to emerge.”2 The postcolonial critique of European Enlightenment may, indeed, have become stifling. But the picture is constantly shifting, as one example demonstrates. In Orientalism, Edward Said wrote scathingly about late seventeenth-­ century court scholar Barthélemy d’Herbelot’s work with Jean-Baptiste Colbert, arguing that it propagated the idea of the inferiority of Islam among European readers. But, as Alexander Bevilacqua argued in his book, The Republic of Arabic Letters: Islam and the European Enlightenment, d’Herbelot’s work was, in fact, more nuanced than Said acknowledged.3 Its assertions about Islam were on the whole balanced and judicious. The history of the Enlightenment studies is littered with disagreements such as 1  The Electronic Text Corpus of Sumerian Literature. Revised Edition, https://ota.bodleian. ox.ac.uk/repository/xmlui/handle/20.500.12024/2518?show=full, accessed March 10, 2020. The Japanese History Corpus, https://pj.ninjal.ac.jp/corpus_center/chj/, accessed March 10, 2020. 2  Bradley Naranch, “The Enlightenment’s Global Histories. Review of Lüsebrink, HansJürgen, ed., Das Europa der Aufklärung und die außereuropäische koloniale Welt.” H-German, H-Net Reviews, May 2009, https://www.h-net.org/reviews/showrev.php?id=24139 3  Alexander Bevilacqua, The Republic of Arabic Letters: Islam and the European Enlightenment (Cambridge, MA: Belknap Press of Harvard University Press, 2018), 114.

4  EUROPE AND ITS “OTHERS” 

123

this: the pronouncement that a given author’s critique is indicative of a wider prejudice, and the recovery of a more complicated truth. This problematization of intellectual and cultural history at the level of individual authors is, of course, valuable, and Enlightenment studies gain a great deal from such praxis and the dialectics which follow from it. But what happens when we turn our attention away from individual texts and authors to survey a historical corpus in its entirety?4 In other words, in what way does the digital profitably complicate the Enlightenment? These are the questions that this chapter begins to answer, taking a diachronic view across the whole of eighteenth-century publications in Britain to reconstruct how British authors wrote about Europe versus the “other”—be it its colonies, the Orient, the Indies, or the Americas.5 This chapter explores knowledge of these “other” regions as it was produced, in an aggregated, shared, common form, by the polity of authors whose texts constitute the Eighteenth-Century Collections Online corpus (ECCO).6 In other words, it begins to make visible how authors in the British corpus most commonly wrote about non-western nation-states and regions, and their perceived national or racial characteristics. To investigate relations between geographical designators in the early, mid, and late eighteenth-century ECCO corpus, I have used a custom-­ designed measure of lexical co-association. My objective has been to draw a historically sensitive analysis of visualized semantic networks that makes visible how the eighteenth-century corpus defined the British publishing   This study draws considerably from Sebastian Conrad’s What Is Global History? (Princeton, NJ: Princeton University Press, 2017). Conrad’s book argues for a more capacious, more global, historical approach to the Enlightenment, one that uses cutting-edge tools and is nicely adumbrated by the blurb: “Until very recently, historians have looked at the past with the tools of the nineteenth century. But globalization has fundamentally altered our ways of knowing, and it is no longer possible to study nations in isolation or to understand world history as emanating from the West.” Another collection taking this view is David T.  Gies and Cynthia Wall, eds, The Eighteenth Centuries: Global Networks of Enlightenment (Charlottesville: University of Virginia Press, 2018). 5  Quotation marks around “other” indicate the well-known concept of “othering” rather than referring to a specific usage. It should also be noted that, while there is a non-negligible number of translations and foreign language texts in ECCO, this is not so large as to invalidate the phrase “British works” used to describe the corpus, or the reference to “British authors” in this chapter. 6  This chapter draws on Prasenjit Duara, Viren Murthy, and Andrew Sartori, eds., A Companion to Global Historical Thought (Hoboken, NJ: John Wiley & Sons, 2014). Particularly useful were the assertions of “comparative and connective histories” in Jennifer Pitts’ chapter, “The Global in Enlightenment Historical Thought,” 184–96. 4

124 

J. REGAN

center in relation to what it saw as its periphery. Because it takes as its field of inquiry the whole of ECCO, what is made visible below is how knowledge about the “Orient,” the “Americas,” “Africa,” and “Europe,” was structured in a common and non-individual sense. In brief, this chapter presents a view onto the vast, impersonal, aggregated repository of the historical corpus, literally a common sense of geographical naming and framing. To do so, this analysis starts with presenting in table form the words which bind to the names of four continents: Africa, Asia, America, and Europe. It then visualizes relations between these names in semantic network diagrams. Finally, it makes a morphological switch to focus upon three plural nouns: “Africans,” “Americans,” and “Europeans”7 to investigate if this lexical approach brings out facets of British writing on non-­ western locations of the Enlightenment that are not present in geographical nouns. As shown here, morphologically diverse approaches to such inquiries can yield rich results: in this case, a more visceral sense of British “othering” than had hitherto been made visible.8 The wide, zoomed-out views that these inquiries present mitigate against the kind of thinking that led William McNeill to assert in 1963 that “[w]e, and all the world of the twentieth century, are peculiarly the creatures and heirs of a handful of geniuses of early modern Europe.”9 Because ECCO is constituted of many thousands of written voices, no single voice, nor a relatively small group, can predominate. This chapter makes a virtue of this composite nature of the ECCO database to reconstruct the most common ways in which the printed anglophone corpus represented, and indeed knew, the non-European locations of the Enlightenment. 7  Unfortunately, “Asians” did not get above the threshold of frequency-in-corpus to have been uploaded into the tool that calculates the distributional probability factor (dpf). “Asiatic” did, but the author felt that there was insufficient difference, both semantically and syntactically, and ruled it out of being investigated along with the plural nouns “Africans,” “Americans,” and “Europeans.” 8  See also John Regan, “‘Beauty’ and the ‘Beautiful’: A Computational Enquiry into the Eighteenth-Century Concept of Aesthetics,” The Journal of Aesthetics and Art Criticism (forthcoming 2020). 9  William H. McNeill, The Rise of the West: A History of the Human Community (Chicago: University of Chicago Press, 1963), 599. An alternative to received narratives of Western or European dominance from epoch to epoch is presented in J.  C. Sharman, Empires of the Weak: The Real Story of European Expansion and the Creation of the New World Order (Princeton, NJ: Princeton University Press, 2019).

4  EUROPE AND ITS “OTHERS” 

125

It should be noted, though, that this exploratory work is only a first foray into a new field of inquiry. It is intended to open up new avenues of inquiry, with the hope that one day we might work more assuredly with what the shared, common authorial knowledge of western and non-­ western Enlightenments was in the British eighteenth century.

Methodology In all forms of linguistic communication, certain words are more likely than others to occur in proximity to one another. One may readily assume that “America,” for instance, would be more likely to appear in proximity to the word “Philadelphia” than it would to “Shanghai.” We can now account for this likelihood and quantify the strength of binding between the words “America” and “Philadelphia.” And if we can quantify binding between these words we may also, by computation, ascertain the strength of binding between “America” and all the other words with which it is likely to co-associate in a given corpus of historical texts. Calculations of binding in this chapter are all the result of a custom-­ designed measure called distributional probability factor (hereafter dpf ).10 Dpf allows researchers to scan through the over two hundred thousand volumes in ECCO to identify particular words and their lexical surroundings.11 In our case, this process has produced over 594 million data points. Using this as our data source, we then ascertained the likelihood of one term to co-associate with another over a notional baseline of a random distribution of terms within the dataset. This allowed us to construct a metric that gave us an indication of the strength of the co-associations between every term. To illustrate, see Table 4.1. Table 4.1 shows the words most likely to occur at a distance of ten words, both before and after the word “knowledge,” in the period 1770–1780.12 We then compared what we found in the corpus, or the 10  Dpf is a is a computational method developed by the members of the Concept Lab, Cambridge Centre for Digital Knowledge, to measure the likelihood of lexical co-associations in a corpus of texts. Dpf corrects for (i.e., limits the influence of) word frequency in a corpus, so that very rare words do not shoot to the top of a co-association list and very common words are not unduly penalized. 11  For practicality, we have excluded from our analysis terms with a corpus frequency of less than five thousand across the one hundred years of data. 12  Distance ten before and after the search term has been selected because, as shown through a great deal of testing, distance ten gives a reliable view of what occurs within sen-

126 

J. REGAN

Table 4.1  Words likely to bind to “knowledge” at distance 10  in the ECCO corpus for the period 1770–1780 Knowledge, distance 10, years 1770–1780, corpus ECCO, frequency 211,136 Knowledge Disagreement Ignorance Science Understanding Perception Godliness Sciences Ideas Attain Learning Study Mathematics Infight Experimental Attained Acquired Ignorant Passeth Attainments

3567 2766 2031 1922 1861 1832 1721 1707 1704 1675 1673 1572 1418 1411 1399 1390 1355 1352 1342 1292

“observed” results, with an “expected” baseline, which is what one should expect if all the words were randomly distributed in ECCO. Doing this allowed us to allot a dpf score to relations between words. As we can see above, the term “disagreement” is very likely to occur close to “knowledge,” having a dpf score of 2766. Comparing this binding to that of the term “attainments,” we notice that this latter word is approximately half as likely to occur in the corpus ten words away from “knowledge” as it has a dpf score of only 1292. A binding list like the one above, however, is only the beginning. Building on this first element we may then take each set of terms on this list, produce binding lists for all these words, and visualize the binding tences containing these words. The average length of sentences in the ECCO corpus is seventeen; this means that using a distance of ten is an effective means of viewing sentential space around search terms.

4  EUROPE AND ITS “OTHERS” 

127

connections between these terms.13 By doing so, we can recover the structure of the discursive environments which dominate the ECCO corpus in a range of historical periods; thus, we can reconstruct how words tend to cluster or repel one another at various moments in time. Cross-referencing binding of lists for different terms indicates structured relationships between words in the lexical record. What overlap is there between what binds to “Armenia” and “Enlightenment,” for instance? Or, how much is shared between what binds to these words and the binding lists for “Arabic” and “mathematics”? How much of what binds to “power” also binds to “executive,” “plenitude,” “legislative,” “delegate,” “magistrate,” and “endued”? Finding the answers to these questions can help to identify patterns in word use and, more pertinently, structures of knowledge which have hitherto remained invisible.14

Part One: Continental Drift Using place names as one’s field of inquiry is an approach whose complexities are determined by a range of historical and geopolitical contingencies. In this first section, I will use names of continents as all continents were, to varying degrees, subject to the establishment and disestablishment of constituent states throughout the eighteenth century. This will allow me to explore the language that is likely to occur within sentences containing terms like “Africa,” “Asia,” “America,” and “Europe.” Tables 4.2, 4.3, 4.4, and 4.5 show what that language was in the British corpus in three thirty-three-year historical tranches: 1700–1733, 1734–1766, and 1767–1800. In the last column, it also shows the language that is common to all three historical periods, that is, the language that binds to these continent names irrespective of the time they appear in the corpus. While these historical divisions are undeniably crude, they do begin to tell us something about the lexical environments within which each continent name was suspended, in different historical periods, in the British 13   The digital methods employed here have been developed by the members of The Cambridge Concept Lab: John Regan, Peter de Bolla, Ewan Jones, Gabriel Reccia, and Paul Nulty. For details, see http://www.crassh.cam.ac.uk/programmes/ the-concept-lab-cambridge-centre-for-digital-knowledge 14  This method is explained in detail in Peter de Bolla, Ewan Jones, Paul Nulty, Gabriel Recchia, and John Regan, “Distributional Concept Analysis: A Computational Model for History of Concepts,” Contributions to the History of Concepts 14, no. 1 (2019): 66–92, https://doi.org/10.17863/CAM.35748

128 

J. REGAN

Table 4.2  Words likely to occur within sentences with “Africa,” with those most likely at the top of each list 1700–1733 Africa Vandals Carthage Tunis America Mediterranean African Libya Asia Carthaginian Barbary Isthmus Tripoli Africans Scipio Sicily East Indies Arabia Cape Ethiopia Morocco

1734–1766 12,876 9639 9493 7913 6911 6418 5957 5542 5498 5437 5321 5108 4963 4580 4489 4456 4325 4145 3965 3884 3883

Atlantic Africa Vandals Carthage African Tunis Ethiopia Asia Morocco Carthaginian Barbary Scipio Africans Sicily Cape Mediterranean America Verd Algiers Ocean Tripoli

1767–1800 15,015 12,776 11,275 10,600 8926 8590 7749 7587 7116 6843 6421 6358 6290 6274 6120 6116 5816 5788 5490 5455 5174

Asia Africa Barbary Atlantic Tunis Vandals Carthage Europe Niger Ethiopia Cape Mediterranean Indies Carthaginian Ocean Verd African America Libya Algiers Gibraltar

On all three lists 15,913 13,056 11,464 11,398 10,676 10,604 10,313 8620 8618 8461 8382 8131 8060 8023 7844 7478 7293 7268 7203 6615 6520

Africa African America Asia Barbary Cape Carthage Carthaginian Ethiopia Mediterranean Tunis Vandals

corpus. Unsurprisingly, these words all bind to a great many place names. In the case of “Asia,” for instance, it is notable the emphasis on locations within what would now be called Central or Eastern Europe. One sign of the shifting nature of what was considered to be “Asia” in the British corpus is the prominence and high ranking of the term “Euxine,” an adjective meaning “of the Black Sea.” This descriptor relates to a body of water which sits between Eastern Europe, the Caucasus, and Western Asia. The term “Cappadocia” names an arid region in central Turkey, and “Ephesus” designates a region which has changed hands between Turkey and Greece several times. These co-associations indicate that, in this corpus, “Asia” means “Eurasia.”15 Only the inclusion of “Ganges” in the first two historical tranches, and “Tartary” highly on all three, appear as what we may 15  Jürgen Osterhammel, Die Entzauberung Asiens: Europa und die Asiatischen Reiche im 18. Jahrhundert (Munich: C. H. Beck, 1998).

4  EUROPE AND ITS “OTHERS” 

129

Table 4.3  Words likely to occur within sentences with “Asia,” with those most likely at the top of each list 1700–1733 Ephesus Euxine Asia Libya Tartary Cappadocia Pontus Seleucus Sardis Macedonia Cilicia Phrygia Thrace Africa Greece Smyrna Ganges Mithridates

1734–1766 10,117 9767 8944 8730 7481 6825 6435 6217 6134 6107 5805 5669 5547 5498 4846 4793 4566 4494

Euxine Asia Libya Europe Asiatic Ephesus Tartary Amer Archipelago Ganges Cappadocia Caspian Africa Pontus Sardis Seleucus Scythians Turky

1767–1800 13,284 11,298 10,633 10,216 9543 9227 9105 8818 8657 8306 8041 7638 7587 7429 6898 6209 5811 5810

Asia Europe Eur Africa Tartary Euxine Amer Ephesus Libya Caspian Mithridates Turkey Pacific Archipelago Cappadocia Scythians Cilicia Macedonia

On all three lists 29,277 20,283 16,372 15,913 12,734 11,822 11,073 10,813 10,009 9817 9545 9474 8768 8353 8290 8277 8271 8101

Africa Asia Cappadocia Ephesus Euxine Libya Tartary

designate as Asian rather than Caucasian in modern geographies. It is also interesting to note that, when the word “Asia” is used by authors across the eighteenth century, the lexical domains in which it is situated are virtually empty of references to any states in the Far East. In its most common usages, “Asia” was surrounded in sentences by place names of Western Asia and Eastern Europe rather than by references to China, Japan, Korea, or India.16 Lexical binding across the British corpus for “America” provides a stark contrast. All three historical tranches are heavily populated with the names of the states that took part in the highly riven process that would culminate in the formation of the modern United States of America at the end of the eighteenth century. The lists also feature some names of South and Central American states, the sea along America’s eastern seaboard, and, recurrently, the term “Columbus.” These “containing” concepts, to use  On the influence of Islamic commerce on Eurasia and the Caucasus, see Marshall G.  S. Hodgson, Rethinking World History: Essays on Europe, Islam and World History (Cambridge: Cambridge University Press, 1993). 16

130 

J. REGAN

Table 4.4  Words likely to occur within sentences with “America,” with those most likely at the top of each list 1700–1733 Columbus Mexico Canada America Peru Florida Carolina Atlantic Brazil New York Africa Newfoundland Virginia Isthmus Chili Domingo Colonies Maryland Jamaica Islands New England

1734–1766 14,550 13,897 13,372 10,558 9848 9273 7895 7450 7361 7232 6911 6741 6386 6289 5906 5730 5530 5408 5110 5094 5052

Atlantic America Florida Mexico Canada Colonies Brazil Pacific Columbus Chili Quebec Panama Europe Peru Africa Maryland Isthmus Virginia Carolina Breton Cartagena

1767–1800 11,739 9885 9080 8892 7839 6907 6843 6792 6772 6490 6461 6205 6071 6003 5816 5743 5598 5456 5218 5163 5079

America Europe Mexico Asia Africa Colonies Atlantic Canada Panama Florida Peru East-Indies Pacific Chili Indies Continent Settlements Columbus American Britain Exported

On all three lists 10,776 8901 8340 7633 7268 6657 6467 6457 6213 6167 6000 5930 5839 5787 5616 5436 5313 5304 5228 5201 5191

Africa America Atlantic Canada Chili colonies Columbus Florida Mexico Peru

Peter de Bolla’s term, seem to indicate that the discourse surrounding this area during the eighteenth century worked somewhat like an inventory: it was a laying out of the contents of this other place.17 It is a list of things which constitute modern America. And the interesting question for “America” is: why does this concept work as a repository while “Europe” does not? We can also notice the wide differences between what might be called the cartographical or geopolitical reality of “America” and “Asia” and the semantic or epistemic contents of these lists. What we are reading are not 17  The idea of “containing” concepts is developed brilliantly in Peter de Bolla, The Architecture of Concepts: The Historical Formation of Human Rights (New York: Fordham University Press, 2014). De Bolla’s theory is that some concepts act as receptacles for a range of other concepts. For example, the concept of “media” contains related concepts, such as “film” or “television,” the concept of “writing,” but also the concept of “thinking.”

4  EUROPE AND ITS “OTHERS” 

131

Table 4.5  Words likely to occur within sentences with “Europe,” with those most likely at the top of each list 1700–1733 America Asia European Alliances Madras Europe Africa Tartary Spain France Bourbon Monarchy Weft-Indies Commercial Potentates Indies Mediterranean Enhance Exorbitant Europa Exchanges

1734–1766 4184 3610 3484 3063 2889 2706 2699 2624 2577 2473 2456 2446 2391 2217 2203 2117 2104 2086 2067 2055 2049

Europe Amer Asia Netherlands America Africa Germany Turkey Switzerland Russia Tartary Spain Continental France Aria Muscovy Asiatic East-India Austria Perennial Poland

1767–1800 24,632 19,279 10,198 6988 6009 5134 4873 4816 4706 4099 3617 3502 3184 3148 2994 2950 2908 2836 2824 2746 2718

Europe Asia Amer America Africa Turkey Pacific Netherlands Atlantic Ocean Germany Russia France Mediterranean Sweden Poland Spain India Euxine Switzerland Muscovy

On all three lists 20,650 20,049 13,316 8889 8542 5500 5290 5104 4849 4553 4522 4265 4141 3864 3665 3662 3579 3542 3467 3409 3397

Africa America Asia Europe France Spain

This is a new, albeit minor, aperture onto how “cosmopolitan” histories were in eighteenth-century Britain. All the texts discussed in Karen O’Brien’s Narratives of Enlightenment: Cosmopolitan History from Voltaire to Gibbon (Oxford: Oxford University Press, 2004) would have been borne on the above data. But, as clarified in O’Brien’s introduction, none of those voices (including David Ramsay’s, William Robertson’s, and David Hume’s) would, or could, have supervened over the entire ECCO

objective descriptions or definitions of locations: they are what the continents were taken to mean, and associate with, in three historical periods by a vast number of eighteenth-century authors. The difference between these geographical locations as things in the world and their realities as epistemes produced by many written voices will become more telling and more visible in the following section. How things were and how the British corpus presented them are notions in fascinating, if opaque, relation. Interestingly, “Britain” does not feature in any of the lists for “America,” and nor do any of its constituent nation names. “America” does bind to

132 

J. REGAN

the word “colonies,” which is a minor indication of an emerging structuring of transatlantic relations. The inclusion of “Africa” in all three lists also hints at the sad history of slavery, a form of commercial traffic between these continents in full swing at the time. Considerably more will be written about this in the following section. “Africa” has a somewhat similar profile to “America” in the sense that lexical binding lists largely names of established nations and regions within that continent. But here, we might turn our focus away from the content of each historical list to the length of the list in the far-right column. The size of this list suggests the changeability of each word’s lexical co-­associations across the century: how susceptible each term was to altering lexical suspension in each historical tranche of ECCO.18 We can, therefore, make a tentative inference that the lexical environment around “Africa” in the one hundred years of the British corpus was relatively stable. If we compare its historically unchanging list with that of “Europe,” the brevity of this latter list is notable. The only chrono-invariant words for “Europe” are the other continents in this inquiry, and then “France” and “Spain.”19 “Germany” and “Sweden” associate with this continent name in two of the three historical periods surveyed, whereas none of the nations of the recently united United Kingdom feature in any of these lists. Based on this first evidence, it appears that British writers as a whole worked with a considerably more diverse range of lexical co-associations for “Europe” than they did for “Africa.” We also notice that lists for “America” and “Africa” are populated to a fairly large extent with the names of places within these two continents. This is true when we look both at each historical period in 18  Each word is surrounded by a given lexis in each historical period: if that lexis differs greatly from one historical period to another, the common-lexis column on the right will contain a short list of words; if there is little change, the list will be longer. Therefore, the shorter the list of terms, the more changeable and unstable the search term’s lexical surroundings are; the longer the list, the more stable and historically unchanging the term’s associations are. 19  In line with the account presented in Anthony Pagden, Lords of All the World: Ideologies of Empire in Spain, Britain and France, 1500–1800 (New Haven, CT: Yale University Press, 1995), it is possible that, irrespective of their political or ethical differences, Britain, Spain, and France would have been yoked together lexically by the simple fact of their imperial interests.

4  EUROPE AND ITS “OTHERS” 

133

isolation and at the lexis common to all three lists. “Asia,” on the other hand, has intriguing lists in ECCO both in each historical tranche and across the century. This denomination is suspended in a vocabulary of Eastern and Central Europe, the Caucasus, and the Euxine. No countries of the Near or Far East feature, and the “Ganges” appears only on one list. This Eurasian preponderance may be due to the fact that the eastern reaches of Asia remained closed to European travelers (and by extension commentators) until around 1780, while the last twenty years of the century saw a generous cosmopolitanism in European exploration and writing on China, Japan, “Arabia,” or India, which hardened into a Eurocentric, commercial, and cynical Orientalism by the turn of the nineteenth century. Apparently, for most of the eighteenth century, “Asia” meant primarily Eurasia for British authors. Indeed, as Gagan Sood has shown, in commercial terms, the Islamic determination of Eurasia was extensive in early modern history.20 It is likely, therefore, that these lexical fields constitute the abiding sense of that determination: North African, Arabic regions shaped the eighteenth-century British sense of what “Asia” meant. One potential pitfall of using geographical designators as a way to unveil the knowledge structures that authors in eighteenth-century Britain constructed to make sense of the world is that this will give preference to a certain type of text. One argument might be that these lists and visualizations are the result of the names of continents being listed in the many atlases published and republished during the eighteenth century. However, as the presence of so much unpredictable co-associating lexis indicates, these lists have not been produced by such programmatic means. Further, none of the major atlas publications that would shape ECCO data, such as Atlas Nouveau (Amsterdam, 1742), Britannia Depicta (London, 1720), and Cary’s New and Correct English Atlas (London, 1787), features lists of continents and their constituent nations to such an extent that they would determine the contents of these lists.

20  Gagan D. S. Sood, “Circulation and Exchange in Islamic Eurasia: A Regional Approach to the Early Modern World,” Past & Present 212, no. 1 (2011): 113–62, https://doi. org/10.1093/pastj/gtr001

134 

J. REGAN

Part Two: Europe and Its “Others” in Visible Relation I will now visualize relations between Europe, Asia, Africa, and America and their associated lexes in these lists. By doing so, I will make visible relations not only among these continents but also among the vocabularies surrounding each. The three decades of data that will be presented are 1710–1720, 1750–1760, and 1790–1800 (Tables 4.2 to 4.5). Obviously, they do not correspond exactly with the three thirty-three-year periods in the tables above; instead, they are intended to provide a more granular and historically specific sense of how the locations of European and nonEuropean Enlightenment were framed lexically across the eighteenth century in British print. The first semantic network diagram, for 1710–1720, gives an immediate indication of two main ways in which British knowledge of western and non-western locations was structured in the early part of the century (Fig.  4.1).21 This first visualization immediately reveals that the lexis around “Africa” in the 1710–1720 tranche of ECCO is in several cases the same as that surrounding “America.”22 This is illustrated by the number of nodes that the two continent names connect to. The words at these nodes make for grim, and perhaps predictable, reading. When British authors wrote about “Africa” and “America” between 1710 and 1720, they were very likely to use the words “traders,” “trading,” “colony,” “negroes,” and “plantation.” In other words, the shared lexical space for these two continent names in British authorial space is predicated on slavery and colonialism. “Africa” also shares lexical space with “Asia” through nodes like “Persia,” “Libya,” “Arabia,” and several second-order relations linking these two names. This space, however, is of a less overtly commercial kind. So, “Africa” is suspended in a shared lexis with these two 21  We should remember that what is visualized here is the lexical binding for all nodes, that is, all words in the binding lists for these continents in the 1710–1720 period in ECCO, and not just for the continent names themselves. 22  Some notes on what is visualized: the thickness of edges in these semantic networks indicates the strength of binding between word-nodes. A thicker line between nodes indicates a stronger likelihood of their occurring together within sentences. Secondly, a community-detection algorithm has noted when a given set of words is likely to co-associate and has colored these network visualizations accordingly. Please be aware that the colors are different in each graph: so, a given community around “Asia” may be green in one visualization, and purple in another. What is significant is the fact of their forming communities, not the color used to identify them.

4  EUROPE AND ITS “OTHERS” 

135

Fig. 4.1  Semantic network visualizing top 20 lexical co-associations with “Africa” (red), “Asia” (green), “America” (yellow), and “Europe” (blue), and top 20 co-associations of these bound terms between 1710 and 1720

continent names whose common denominators are commerce and slavery, but, interestingly, there is no general pooling of shared lexis for all three. This suggests that eighteenth-century British authors framed relations between “Africa” and “America” in dark commercial terms, while they wrote about relations between “Africa” and “Asia” in more geographical, less nakedly commercial terms, mainly relating North African sites with the Middle East. The second feature of note here is the exceptional nature of “Europe” in these lexical co-associations. In plain terms, “Europe” is, lexically

136 

J. REGAN

speaking, unusually unrelated to the other continents23: as seen in Fig. 4.1, it shares a very limited frame of reference with the other continent names. Across this decade of printed material, “Europe” is likely to bind only to “Asia,” “Africa,” “America,” and “fleets,” which is a very small shared lexical space for this word in the semantic network for this decade. Inasmuch as the whole polity of British authors can be considered European (not a resolved question by any means), this finding clearly indicates that British authors wrote about “Europe” in quite separate terms than about its “others.” There is also the question of whether, en masse, the British authors themselves considered themselves European: nothing here suggests that this was the case. Edmund Burke’s support for British affinity with the continent was a notable exception: he wrote about a “Commonwealth of Europe” before the British commonwealth was a prospect and, more controversially in our time, about Europe as “virtually one great state.”24 These positive statements by Burke are indicative of a line of argument that Britain should align with the continent and assume membership.25 According to Burke, the factors uniting Britain with Europe were their shared Christianity, their preference for monarchy, their “common custom,” their “legal heritage,” and the enshrinement of that most eighteenth-­ century scion in our education systems: manners.26 Finally, the last proof of Britain’s belonging to Europe was its propensity toward free trade and economic independence. This argument, though, does not seem to have had much traction during the time if we look at the tabulated language above and the semantic diagram in Fig.  4.1. Such a disparity does not invalidate Burke’s view, however. Rather, when zooming out from individual authors to how word 23  In Harry Liebersohn, The Travelers’ World: Europe to the Pacific (Cambridge, CT: Harvard University Press, 2006), the focus is upon the exchanges between European explorers and Pacific islanders. Libersohn’s book is a wonderful exposition of how Europeans interacted with their colonial “others”; however, as the above evidence suggests, there is a dearth of occurrences of this continent name in lexical relation to its “others” in ECCO, so this may contradict Libersohn’s conclusions. 24  Edmund Burke, Two Letters Addressed to a Member of the Present Parliament, on the Proposals for Peace with the Regicide Directory of France (London: Printed for F. and C. Rivington, 1796), 110–16. 25  A fine, succinct analysis of Burke’s position is given in Jennifer M. Welsh, Edmund Burke and International Relations: The Commonwealth of Europe and the Crusade against the French Revolution (Basingstoke: Palgrave Macmillan, 1995), 78–80. 26  Burke, Two Letters, 114.

4  EUROPE AND ITS “OTHERS” 

137

relations and, by extension, knowledge, were structured in this time, a complementary account of relations between Britain and Europe emerges. Offsetting Burke’s pro-European stance are the virtually constant wars with Holland and France at the time. It is unthinkable that ECCO would not reflect what was, essentially, a martial century for Britain. During this decade, the Whigs called for greater trade relations with Europe and the Tories rebutted this with even greater calls for martial strengthening, seeing defense as an alternative to trade. On a related note, a caveat is warranted about how much of an “other” “America” would have been considered, particularly in this period of colonial history. As shown by this analysis, there is little evidence of a shared lexical space between these two continent names, which is a simple, bottom-up proof of “othering” in the British corpus. The two continents are simply presented in starkly different terms: there is no perceived Atlantic axis of the Enlightenment in commerce or culture in the general flow of printed material. The observation that Europe does not share lexical space with other continents is quite different than that relating to the chrono-invariant bindings for “Europe.” As indicated above, lexical binding in sentences around “Europe” is susceptible to considerable variance over our corpus-­ time. This indicates that, from the British authorial perspective, “Europe” as the geographical center of the Enlightenment was a notion with widely shifting frames of reference. Moreover, this first visualization allows for another inference: it seems that “Europe” was not considered by early eighteenth-century authors in ECCO as a topic to be discussed in the same terms (literally) as the non-European locations of the Enlightenment. In other words, with regard to continent names at least, British authors wrote about Europe in different terms than they used when writing about the other continents. Moving forward in our corpus-time to the semantic network for the mid-century decade, 1750–1760, we suddenly find an extensive overlap between the lexical spaces around “Europe” and “Asia,” something that the tabulated data above did not suggest. However, we must again recall that these networks visualize lexical binding not only between the three continent names but also for the words to which these continent names are likely to bind. Thus, as noted in the relations in blue and yellow at the top of the network in Fig. 4.2, there is a remarkable tendency in the general traffic of British print to discuss “Europe” and “Asia” in

138 

J. REGAN

Fig. 4.2  Semantic network produced based on top 20 lexical co-associations with “Africa,” “Asia,” “America,” and “Europe,” and top 20 co-associations of these four bound terms between 1750 and 1760

the same, or similar, terms. What this network is showing is that, on several occasions, the two continent names bound directly to the same words; however, in at least as many cases, there are second-order binding relations, such as continent names at only two lexical removes from one another. We can observe this by following the lexical associations from “Europe” to “Asia” through “Asiatic” and “Turky,” or from “Asia” to “Europe” through “Archipelago” and “Turkey.” As noted

4  EUROPE AND ITS “OTHERS” 

139

above, what we find is a semantic field around “Asia” which is oriented, so to speak, around the Caucasus, Turkey, and Greece. Nonetheless, this British authorial and historical version of “Asia” is something lexically, semantically, and, by extension, epistemologically related to “Europe” in the mid-century ECCO data. To modern eyes, it appears to be more of a Eurasian vista, and this historical difference is a fascinating reminder of the historical situatedness of words and their special, historically determined frames of reference. Moreover, what “Europe” does not share lexical space with is also noteworthy. While this word does bind to the truncated form “amer,”27 there are no lexical connections whatsoever between “Europe” and “America” in the historical tranche of 1750–1760 publications. Again, we must remind ourselves of the usual caveats: these are lists of the top twenty most likely lexical associations between continent names. Then, we also produced binding lists for the top twenty words which bind to each continent. So, all four continent names used here, plus the lists for what binds to them, are cross-referenced to show which words in this particular part of eighteenth-century published discourse were in shared lexical space. However good this view is, it is only partial, as it arguably must always be, given the nature of digital inquiry. Yet this partiality, this thresholding of lexical co-associations, does not seem an adequate explanation for the lack of shared lexical space between “America” and “Europe.” “European” binds to “America,” but given the colonial relation of Britain to the American states at the time, it should also account for the cleavage between colony and colonizer that was about to occur, so it is remarkable that “Europe” would not share at least some lexical space with “America.” If nothing else, this feature of the network is a further suggestion of the separateness of “Britain” and “Europe” in the authorial British mid-­ eighteenth century.

27  It is to be noted that corpora are liberally sprinkled with lexical truncations and a wide range of other oddities, some owing to bad character recognition, and some to the unfamiliar nature of the texts at the time. A term like “amer,” for instance, may have been used in its truncated form in atlases or may have been a catchword (as per eighteenth-century printing practices, the first word of the following page was placed below the last line of text on the previous page).

140 

J. REGAN

Anthony Pagden has provided two feasible reasons why British authors might have lexically sequestrated Britain and Europe from their “others.”28 The first is the different colonial behaviors of the British and their mainland European counterparts: as Pagden insightfully suggests, British colonialism used markedly different strategies of colonization than those employed by the French, Dutch, and Portuguese throughout the seventeenth and eighteenth centuries. As a result, it is likely that the lack of a general lexical pooling is due to the fact that British authors described British and European forms of colonial expansion differently: pragmatic differences in forms of subjugation meant that different vocabularies were used to describe different colonial endeavors. The second reason Pagden offers is that a general denial of communication between the colonizer and the colonized was a prerequisite for colonization to actually happen. Pagden condenses this theory as follows: “If we could not accept the consequence of these melancholy facts [that conquest and annihilation were necessary for a successful colonial expansion] and clearly it was unacceptable to any enlightened person then the only alternative was to leave ‘them,’ physically and conceptually, alone.”29 The relevant point for my analysis is, of course, the latter part of this “leaving-alone”: a conceptual turning away from American colonies that extends to a common authorial discourse which registers virtually no sense of a will for peaceful integration. It is also worth noting that, in general, the 1750–1760 semantic network is busier and more populated than the one for 1710–1720. But such an increase in nodes and connections between the four continent names is by no means a necessary function of more materials being printed during this time. As we will see, although an increase in network connections does, indeed, happen for geographical designators like “Europe” and “America,” the diversity of a given word’s lexical co-associations and the strength of these associations is not necessarily linked to an increase in the output of printed material.30 As we turn our focus to the final decade of ECCO data, from 1790 to 1800, the semantic network showing what associates with continent

28  Anthony Pagden, European Encounters with the New World from Renaissance to Romanticism (New Haven, CT: Yale University Press, 1993). 29  Ibid, 187. 30  See John Regan, “Progress as Impersonal Knowledge in the Eighteenth Century: Two Computational Case Studies,” Digital Scholarship in the Humanities (forthcoming 2020). As shown in this study, the semantic networks surrounding “nature” contract sharply in the later part of the eighteenth century, just as the print proliferation is gathering speed.

4  EUROPE AND ITS “OTHERS” 

141

names and what associates with that associated lexis is, indeed, busier and more interconnected (Fig. 4.3). Perhaps the most striking facet of this view onto the final decade of the corpus time is the new centrality of “Europe” among what were previously markedly its “others.”31 The thickness of the edges in these visualized

Fig. 4.3  Semantic network produced based on the top 20 lexical co-­associations with “Africa,” “Asia,” “America,” and “Europe,” and top 20 co-­associations of these four bound terms between 1790 and 1800 31  Two publications by C. A. Bayly have formed my sense of the diffusion of Britain and British thinking in other places during the late eighteenth century. They are Imperial Meridian: The British Empire and the World 1780–1830 (New York: Longman, 1989), and

142 

J. REGAN

semantic networks indicates the strength of binding between word-nodes. In light of this fact, we can see that not only is “Europe” forming hitherto unseen associations with the other continent names in this part of the corpus but also that the likelihood of co-associations in general between nodes in this network is much increased. This is particularly true for “Europe” and “Asia.” However, the dramatic change that we notice as corpus-time has elapsed is that Europe is now very strongly bound to “America.” This is not a generalized sharing of lexical space as the two continent names do not share many first-order co-associating words. Through words like “Pacific,” though, we can clearly see a new order of lexical exchange between these terms and their related lexis. Furthermore, now that “America” is emerging as the “other” in a most formalized sense, “Britain” emerges as one of its bound terms. Thus, the tumult on the other side of the Atlantic seems to have produced in the British corpus a lexical coming together that a century of colonial relations and general discourse did not. In turn, “Asia” binds to “Japan,” “Siam,” and “Mogul,” which indicates a shift away from what is now called Eurasia or the Caucasus as its central region in the British authorial knowledge. As already mentioned, a considerable upswing in Romantic Orientalism had occurred in the British literature at the time, which brought with it a new fascination with the places, cultures, and ideas of the East. This decade predates the vast explosion in saleable literature of this kind, with authors like Lord Byron, Robert Southey, and Walter Savage Landor, to give just a few examples, capitalizing on this commercial precinct of publishing. This “Asia” network indicates a widening and deepening of the interest in the “Orient” during the last decades of the century. Finally, the new association of “Africa” with “Atlantic” and via this word with “Pacific” is an abiding reminder of a terrible form of seafaring commerce.32 Nonetheless, Fig. 4.3 does indicate if not a move away from, then a reduction in the prominence of the language related to slaves and slavery around this continent name. Here, however, we should pause and consider that the data is, perhaps, disguising the abiding presence of slavery as part of the British authorial epistemology of “Africa” and ­ The Birth of the Modern World, 1780–1919: Global Connections and Comparisons (Malden: Blackwell, 2004). 32  Felicity A.  Nussbaum, ed., The Global Eighteenth Century (Baltimore: Johns Hopkins University Press, 2005).

4  EUROPE AND ITS “OTHERS” 

143

“Africans.” Another part of the morphology allows us to dig a little deeper into this “other.” Using “Africans” as the lead term in the inquiry, we notice that the plural form of the noun is suspended in a domain of brutal commerciality as it binds with terms like “planter,” “negro,” and “slavery” (Fig. 4.4). This finding reflects a dehumanizing way of writing about the “other” as it notably shifts attention from individual to a category. Returning to the data this chapter began with—lexical binding in three historical tranches of the ECCO corpus—, we can drill down into what the plural nouns for people from the four continents surveyed tell us about Europe and its “others.” This will be my final focus, in part because data for plural nouns in the earlier part of the century is, in general, too sparse to allow computational inquiry. We must keep in mind that the historical periods in the following tables are thirds of the century rather than decades, which accounts for the fairly wide difference in lexical binding between the words in tables and their

Fig. 4.4  Semantic network produced based on the lexis that co-associated with “Africans” between 1790 and 1800

144 

J. REGAN

semantic network diagrams. In both the historical tranches and the lists of words that bind to these plural word forms across the century in ECCO, there are some clear frames of reference. Perhaps most striking in the British authorial epistemology of “Africans” is the preponderance of historical proper nouns: “Antiochus,” “Scipio,” “Hannibal,” and “Carthage,” to name just a few (Table 4.6). That vocabulary is abidingly bound to the word “Africans,” which gives a strong indication of the importance of the word to histories of the continent and its peoples. The fact that historical texts form a significant part of ECCO seems, thus, to shape the eighteenth-­ century authors’ view of “Africans” in a way that was not evident when the search term was the continent name. As seen in Fig. 4.4, this classically oriented historical nomenclature forms one clear community, while the language of slavery, including the word “abolition,” forms another. In fact, looking back to Table  4.2 and the visualizations created throughout this chapter, the language of African classical antiquity is Table 4.6  Words likely to occur within sentences with “Africans,” with those most likely at the top of each list 1700–1733 Africans Scipio African Carthage Independence Herodotus Aloe Africa Flaccid Cornelius Hannibal Polybius Antiochus Arabs Lavender Carthaginian Paulus Pauli Olympiad Arabians Eclipse

1734–1766 24,768 16,990 7335 6851 6539 4821 4687 4580 3889 3850 3795 3680 3382 3327 3206 3182 3179 3048 2980 2975 2920

Africans Scipio African Polybius Carthaginian Carthage Flos Hanniba Africa Eusebius Paulus Americans Flaccid Asparagus Europeans Antiochus Adopted Cornelius Scorching Epitome Laudes

1767–1800 37,954 16,571 12,891 9454 9312 7834 7521 7267 6366 6024 5896 5560 5549 4517 4464 4299 3752 3590 3449 3390 3211

Africans Scipio Paulus Carthage Europeans Hannibal African Carthaginian Caius Antiochus Africa Polybius Eusebius Gauls Cornelius Tiberius Extirpated Fcribit Scorching Pofita Consul

On all three lists 34,955 22,752 11,347 9606 9306 9299 8128 7747 6307 6126 5895 5774 5593 4686 4599 4505 4464 4374 4374 3743 3738

Africa African Africans Antiochus Carthage Carthaginian Cornelius Hannibal Paulus Polybius Scipio

4  EUROPE AND ITS “OTHERS” 

145

present in both the historically sensitive and century-wide view of “Africa.” This, then, seems to be a principle foundation of the British authorial epistemological construction of this “other” across morphologies: Africa was a continent both distant in time and in terms of shared lexical space. It is only when we use the plural noun that we uncover that there is a closeness between “Europe” and “Africa” in the corpus: a closeness of planters, negroes, and slavery. What else can be uncovered by using a plural noun form as the morphological way of viewing Europe’s “others” in the eighteenth-century corpus? Turning to “Americans,” we immediately note the strength of the binding between this word and “European”’ in the three thirty-three-year tranches of printed data. This appears to go hand in hand with what we might have predicated to be the most familiar story involving “Americans,” if not “America,” in the British corpus: a burgeoning and warming debate about taxation, sovereignty, colonialism, and human rights. By the end of the historical tranche studied, the independence controversy had supervened over the mass of writing using the word “Americans” (Table 4.7). Entering the data through the aperture of the plural noun “Americans,” we find that “Great Britain,” “British,” and “Britain,” are all part of this word’s lexical space. Martial language, such as “evacuated,” “retreated,” “lieutenant-colonel,” and “invaders,” is now a prominent part of the picture. Looking at the column on the right, we notice just how unstable this plural noun’s lexical associations have been in the British corpus. Of the four words which are common to all three historical lists, three are part of the “America” and “Europeans” morphology. A cursory glimpse back across the lists suggests that the word “Americans” is suspended in a shifting environment, susceptible to the vicissitudes of political tensions across the Atlantic. In other words, the plural noun “Americans” is bearing the lexical scars of a kinetic, shifting, and politically hot situation. It is also immediately clear that the chrono-invariant list for “Europeans” is longer and arguably more interesting than the one for “Europe” (Table  4.8). Scanning over the list of unchanging lexical binding, and indeed across the historical columns, we note that “Europeans” is a word that binds overwhelmingly to its “others.” Thus, Table 4.8 features terms binding to “Europeans” which are colonially inflected, featuring nouns and adjectives that are clearly used in contradistinction with the search term. Table 4.8 demonstrates the extent to which the British corpus situated the idea of “Europeans” above its “others.” This distasteful language of othering is no historical quirk or accident: “savages,” “natives,”

146 

J. REGAN

Table 4.7  Words likely to occur within sentences with “Americans,” with those most likely at the top of each list 1700–1733

1734–1766

1767–1800

Americans Europeans Asiatic American Agriculture Gothic Commoners Iceland Coalition Latest Domestic European Savages Alonzo Ceylon Prostitute Hudson Invaders Warmed

36,900 20,334 12,793 9433 9191 8984 8047 7639 7387 6914 6638 6481 6227 5846 5768 5133 4891 4761 4428

Americans Europeans Taxation Kennedy Africans Colonists Burke Invaders American Finery Legislation Planned America Engineers Fraser Scythians Commoners Lucan Spaniards

24,579 7546 7262 6390 5560 4753 4494 4407 4269 4077 4058 4055 3976 3934 3901 3862 3666 3595 3526

Echoes America

4359 4324

Despotism Taxing

3431 3415

Americans British Taxation American Cornwallis Washington Great-Britain Clinton America Independence Europeans Boston Greene Rhode Canada Arnold Evacuated Colonies Lieutenant-­ colonel Britain Retreated

On all three lists 11,980 7197 6649 5508 5324 5178 4938 4477 4063 3903 3737 3477 3363 3326 3285 3258 3169 3152 3039

Americans Europeans American America

2857 2738

“unhealthy,” “habitable,” and “fire-arms” all bind to the plural noun “Europeans” across all three historical tranches. Only “savages” does not make the historically unchanging list in the right-hand column. The term “Europeans” is very starkly suspended in a lexical space constituted of those peoples and things, places and characteristics, that are different from Europeans: “blacks,” “Americans,” “Chinese,” “Japan[ese],” and “Indians.”

Conclusion As stated at the outset of this chapter, this work is exploratory: it is only intended to open up future areas for investigation. Therefore, enumerating findings which invite further investigation is a useful final step.

4  EUROPE AND ITS “OTHERS” 

147

Table 4.8  Words likely to occur within sentences with “Europeans,” with those most likely at the top of each list 1700–1733 Americans Blacks Chinese Natives Brazil Armenians Japan Portuguese Mercantile Indians Islanders Fire-arms European Java Brasil Bengal Inland Unhealthy Tawny Columbus Canoes

1734–1766 20,334 18,928 8068 7662 7518 6541 6316 6047 5851 5677 5613 5403 5042 5022 5008 4427 4248 4153 3870 3863 3816

Blacks Europeans Americans Nabob Natives Chinese Negroes Bengal Portuguese Africans Japan European Indians Bombay Fire-arms Armenians Mogul Clive Arabians Java Unhealthy

1767–1800 9555 8688 7546 7053 6575 6530 5863 5678 4529 4464 4458 4399 4315 4167 4048 3924 3555 3334 3328 3299 3226

Europeans Africans Natives Chinese Negroes Armenians Indians Arabians Fire-arms Batavia Portuguese Blacks Habitable Africa Savages Unhealthy Japan China Americans Malabar Artillery

On all three lists 11,217 9306 7855 7461 6811 6419 5645 4997 4920 4829 4498 4494 4293 4163 4110 3901 3896 3871 3737 3651 3493

Americans Armenians blacks Chinese fire-­arms Indians Japan natives Portuguese unhealthy

In the first section of this chapter, the tabulated data for “Africa” across three thirty-three-year spans of the eighteenth century gave a stark indication of is embeddedness in a certain kind of historical discourse: essentially, the Roman history of North Africa. The same data for “Asia” indicated this continent name’s suspension in the place names of Eurasia and the Caucasus, especially from 1700 to 1733, which likely reflects the preponderance of the Arab trade in Eurasia: the Arab trade, assumed to be Asian, so dominated the region that British authors felt comfortable discussing the two as if they were one. Further inquiry will uncover the extent to which this was the case, and if British authors compartmentalized a near “Asia” of commerce in the first part of the century and a much farther east “Asia,” susceptible to a commercial Romanticism, later in the century. A consequent question would be, then: when did British authors recognize various forms of the Enlightenment in Asia, and roughly where were they located?

148 

J. REGAN

A second form of inquiry was the length of the lists of historically unchanging associated lexis in the right-hand columns of the tabulated data. The historically unchanging lexical binding around “Africa” was relatively copious, indicating the stability of at least one frame of reference in British writing about this continent: its history of slavery. The unchanging lexis around “Europe,” on the other hand, was minimal: it included references to France and Spain only. So, once again, we can surmise that authors writing in Britain produced a fairly fixed, commercial knowledge of “Africa” and a more diverse range of references for “Europe.” Visualizations of the semantic networks for the continent names from 1710 to 1720 show that “Europe” formed very few lexical connections with other continent names in the earliest data. By contrast, “Africa” was lexically well-connected to both “Asia” and “America,” although these latter continent names did not share any lexical space themselves. Moving on to the mid-century decade of data, “Europe” starts forming extensive lexical connections with “Asia.” This is certainly something which warrants further investigation: why did British writers begin discussing these two continent names to such extent around mid-eighteenth century? How and why did the printed whole of ECCO situate these two continents in a similar lexical space at this particular time? We might recall that, for the last decade of data, the presence of “Japan,” “Siam,” “India,” and “Mogul” in lists for “Asia” indicates a shift farther East in terms of reference during the latter part of the eighteenth century. So, British authors were in general building up to a richer sense of east Asia, albeit this place may, in time, have become a mysterious, saleable “other.” This is a quite separate lexical phenomenon to that whereby “Asiatic,” “Tartary,” and “archipelago” united these continent names in common usage in the mid-century ECCO corpus. In other words, “Europe” forms stronger lexical connections with Eurasia mid-century and then widens its focus, which is reflective of an evolution in the way British authors view “Asia”: from a Caucasian space of Middle Eastern and Arab trade, to a new, “Oriental,” mysteriousness. While data for “Africa” has given tentative suggestions of a turn away from language related to slavery in the latter part of the century, a change in morphological focus to “Africans” has returned a different report. The plural noun “Africans” is bound to “negro,” “planter,” and “slavery,” which indicates a clear abidance of slaving vocabulary into the late century data. Investigating plural nouns further, it has become evident that the same language of classical Roman history in Africa found in the first half

4  EUROPE AND ITS “OTHERS” 

149

of the century texts surrounds the term “Africans” later on. This clearly indicates a common, historically durable way for British authors to frame this continent name and plural noun as a repository of one, fixed set of historical references. The data for “Americans” has shown very little chrono-invariant language, which indicates that this term is clearly suspended in very mutable lexical environments from period to period in ECCO.  The list for 1767–1800 showed strong signs of the controversy related to the American independence as a key and growing frame of reference. Significantly, a preponderance of martial language is notable in all the lists throughout the century. In comparison, the term “Europeans” is situated in a much more stable lexical environment from period to period than the term “European” is. Its lists are full of names and adjectives for the “other,” and the martial and colonial pejoratives in the list indicate a rigid sense of European superiority. In discussing relatively recent attitudes of colonialism, Pagden captures an important idea that resonates with some of the findings discussed in this chapter: We all, it seems, need to salvage some notion of ourselves as potentially benevolent agents, to persuade ourselves that European civilization is not quite so rapacious and destructive of those who do not serve its ends as it so obviously appears to be. To achieve that objective, the critics of colonialism have tended to construct “others” quite as false as those invented by their opponents. Every image of the “other” reflects intently upon itself.33

The term “Europe” is strongly bound to its “others” in the last decade of the eighteenth century, sitting centrally in the semantic network. Its new lexical centrality, however, is decidedly not an indication of a new sympathy or sense of affinity with other places, unless one counts the closeness produced by the traffic in slaves. Not only does the lexis around “Europe” suggest that it was separate and, indeed, above the other continents in the British authorial imaginary but also that the lexis around all the other continent names works to distinguish Europe as superior to its “others.” British authors looked at Europe versus Africa, Asia, and America and produced a strenuous, in most cases pejorative, sense of difference both when writing about Europe itself and when writing about the other  Pagden, European Encounters with the New World, 187.

33

150 

J. REGAN

continents and their people. Thus, although these results may not countermand prevailing assumptions about the attitude of the British toward Europe and other locations during the Age of Enlightenment, they do open up new possibilities for further enrichment of a complementary story about Enlightenment othering that would have remained closed off without the affordances of the digital.

Bibliography Bayly, C.A. 1989. Imperial Meridian: The British Empire and the World 1780–1830. New York: Longman. ———. 2004. The Birth of the Modern World, 1780–1919: Global Connections and Comparisons. Malden: Blackwell. Bevilacqua, Alexander. 2018. The Republic of Arabic Letters: Islam and the European Enlightenment. Cambridge, MA: Belknap Press of Harvard University Press. Burke, Edmund. 1796. Two Letters Addressed to a Member of the Present Parliament, on the Proposals for Peace with the Regicide Directory of France. London: Printed for F. and C. Rivington. Conrad, Sebastian. 2017. What Is Global History? Princeton: Princeton University Press. Craig, Albert M. 2009. Civilization and Enlightenment: The Early Thought of Fukuzawa Yukichi. Cambridge, MA: Harvard University Press. de Bolla, Peter. 2014. The Architecture of Concepts: The Historical Formation of Human Rights. New York: Fordham University Press. de Bolla, Peter, Ewan Jones, Paul Nulty, Gabriel Recchia, and John Regan. 2019. Distributional Concept Analysis: A Computational Model for History of Concepts. Contributions to the History of Concepts 14 (1): 66–92. https://doi. org/10.17863/CAM.35748. Gies, David T., and Cynthia Wall, eds. 2018. The Eighteenth Centuries: Global Networks of Enlightenment. Charlottesville: University of Virginia Press. Hodgson, Marshall G.S. 1993. Rethinking World History: Essays on Europe, Islam and World History. Cambridge: Cambridge University Press. Liebersohn, Harry. 2006. The Travelers’ World: Europe to the Pacific. Cambridge, CT: Harvard University Press. McNeill, William H. 1963. The Rise of the West: A History of the Human Community. Chicago: University of Chicago Press. Naranch, Bradley. 2009. The Enlightenment’s Global Histories. Review of Lüsebrink, Hans-Jürgen, ed., Das Europa der Aufklärung und die außereuropäische koloniale Welt. H-German, H-Net Reviews: n.p. https://www.h-net. org/reviews/showrev.php?id=24139

4  EUROPE AND ITS “OTHERS” 

151

Nussbaum, Felicity A., ed. 2005. The Global Eighteenth Century. Baltimore: Johns Hopkins University Press. O’Brien, Karen. 2004. Narratives of Enlightenment: Cosmopolitan History from Voltaire to Gibbon. Oxford: Oxford University Press. Osterhammel, Jürgen. 1998. Die Entzauberung Asiens: Europa und die Asiatischen Reiche im 18. Jahrhundert. Munich: C. H. Beck. Pagden, Anthony. 1993. European Encounters with the New World from Renaissance to Romanticism. New Haven: Yale University Press. ———. 1995. Lords of All the World: Ideologies of Empire in Spain, Britain and France, 1500–1800. New Haven: Yale University Press. Pitts, Jennifer. 2014. The Global in Enlightenment Historical Thought. In Pransejit, Murthy, and Sartory, A Companion to Global Historical Thought, 184–196. Prasenjit, Duara, Viren Murthy, and Andrew Sartori, eds. 2014. A Companion to Global Historical Thought. Hoboken: Wiley. Sharman, J.C. 2019. Empires of the Weak: The Real Story of European Expansion and the Creation of the New World Order. Princeton: Princeton University Press. Sood, Gagan D.S. 2011. Circulation and Exchange in Islamic Eurasia: A Regional Approach to the Early Modern World. Past & Present 212 (1): 113–162. https://doi.org/10.1093/pastj/gtr001. Welsh, Jennifer M. 1995. Edmund Burke and International Relations: The Commonwealth of Europe and the Crusade against the French Revolution. Basingstoke: Palgrave Macmillan.

CHAPTER 5

Text Mining and Data Visualization: Exploring Cultural Formations and Structural Changes in Fifty Years of Eighteenth-Century Poetry Criticism (1967–2018) Billy Hall

Models return us to the process—the tools, techniques and practices— through which we construct our knowledge of phenomena that exceed our direct observation. —Andrew Piper, “Think Small: On Literary Modeling” (2017)

Introduction Originally, this chapter was concerned with a relatively modest question: what is the place of poetry in the field of eighteenth-century studies today? If the language of crises in the humanities has become commonplace, the narrative of a declining interest in poetry borders on outright cliché. Despite, or perhaps because of this perception, those few apologies for the value of poetry range from the tentative Can Poetry Matter? to the

B. Hall (*) Brigham Young University, Provo, UT, USA e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_5

153

154 

B. HALL

hyperbolic Can Poetry Save the Earth? to the eschatological Poetry’s Afterlife. In eighteenth-century studies, Jennifer Keith invokes this notion explicitly, if only to resist its disciplinary logic for eighteenth-century scholars. She opens her 2007 review essay “Why Poetry?” by claiming that “in a cultural climate and scholarly marketplace that increasingly marginalize poetry, especially eighteenth-century poetry, relatively few critics are willing to assert its value.”1 Keith goes on, of course, to defend the historical and cultural value of poetry for literary scholars of the eighteenth century and invites them to explore “how eighteenth-century poetry matters, especially how poetic forms articulated new relations among political, material, and inner worlds.”2 Like Keith, John Sitter has also mounted an apology for poetry; however, he tracks along formal and aesthetic lines as much as an historical one. As he points out, the “insistent materiality of eighteenth-­century poetry” is best understood in terms of “seeing the poems as poems and not as something else, quotable pieces of a ‘discourse’ interchangeable with any other form which ideas are to be abstracted.”3 While they may articulate different facets of poetry’s value, Keith and Sitter do not hesitate to assert that eighteenth-century poetry has professional and personal value, but that that value may be lost in the haze of disciplinary and economic forces indifferent to poetry. Consider Fig. 5.1, for instance, which plots the percent of poetry articles published in Eighteenth-Century Studies (ECS) from 1967 to 2018, and in The Eighteenth Century: Theory and Interpretation (TEC) from 1979 to 2018. While the line graph offers a visual representation of a downward trend (from a high of almost 75% to around 10%), it also invites inquiry into Keith’s argument that interest in eighteenth-century poetry may be waning, or at least that poetry has to compete with a broader range of interests within eighteenth-century studies. The question for Keith and Sitter is not if poetry matters, but how poetry matters. That more narrow question about how poetry matters in eighteenth-century studies shifted my original question from a methodologically and ideologically oriented inquiry to a cultural and historical question about how eighteenth-century scholars conceive the way that poetry matters. Rather than exploring the place of eighteenth-century poetry now, I was struck by the idea of exploring how critics have argued  Jennifer Keith, “Why Poetry?,” The Eighteenth Century 48, no. 1 (2007): 87, https:// doi.org/10.1353/ecy.2007.0002 2  Ibid., 91. 3  John Sitter, The Cambridge Introduction to Eighteenth-Century Poetry (Cambridge: Cambridge University Press, 2011), 2. 1

155

5  TEXT MINING AND DATA VISUALIZATION 

100%

Percent of Poetry Articles Published Per Year

Journal ECS TEC

Percent of Total

75%

50%

25%

2015

2010

2005

2000

1995

1990

1985

1980

1975

0%

1970

10%

Fig. 5.1  Percent of poetry articles published each year in Eighteenth-Century Studies (ECS) and The Eighteenth Century: Theory and Interpretation (TEC) (1967–2018)

that eighteenth-century poetry matters for the last fifty years. In particular, I was interested in how essays from two major journals in eighteenth-­ century studies dealt with eighteenth-century poetry, primarily British, but also, to a lesser extent, German and French as well. To do so, I algorithmically modeled the essays from those two journals to foreground the “tools, techniques and practices”4 that have shaped fifty years of knowledge production in eighteenth-century poetry criticism. This is not to suggest that ECS and TEC tell the only story. Both journals have historically published essays by British and American authors and therefore do not account for alternative narratives of eighteenth-century anglophone poetry from outside this tradition. Additionally, I do not give  Andrew Piper, “Think Small: On Literary Modelling,” PMLA 132, no. 3 (2017): 651, https://doi.org/10.1632/pmla.2017.132.3.651 4

156 

B. HALL

in this chapter a specific rationale for why poetry criticism may be in decline in eighteenth-century studies. “Computational models,” Willard McCarty writes, “however finely perfected, are better understood as temporary states in a process of coming to know rather than fixed structures of knowledge.”5 The method I use here acknowledges, instead, the temporary and tentative nature of each computational manipulation—namely, that each new model of the dataset is only one state in a process that produces many alternative states. Therefore, my goal of modeling fifty years of poetry criticism is neither to create a new master narrative nor a master model; rather, I aim to illuminate structures and patterns in the sociolect of eighteenth-century poetry criticism, as well as identify how and why we approached it the way we did in the past.

Eighteenth-Century Poetry Criticism: A Model To create a model of eighteenth-century poetry criticism, I have followed several steps. The first was to apply constraints to the mountain of writing on eighteenth-century poetry over the last fifty years. I decided to only include essays from two key journals that I felt would provide a coherent representation of the field and that included at least fifty years of criticism.6 I selected Eighteenth-Century Studies (ECS) because it is the flagship journal of the American Society for Eighteenth-Century Studies, and The Eighteenth Century: Theory and Interpretation (TEC) because it is the disciplinary vanguard of eighteenth-century studies. As such, they are uniquely positioned to model broad trends in the field. The journal contents were downloaded from Project Muse and JSTOR in PDF format and then converted to plain text files for pre-processing.7 Because I wanted a corpus that consisted primarily of article-length writing, I did an initial filtering to remove all financial reports, “Editor’s Notes,” short “Correspondence” essays, introductory essays, and most footnotes and endnotes. The initial dataset included a total of 1536 articles (941 from ECS and 596 from TEC) and covered the years from 1967 to 2018.8  Willard McCarty, Humanities Computing (New York: Palgrave Macmillan, 2005), 27.  I have not included in this study monographs or essays in collections partly for practical reasons and partly because I wanted to map poetry criticism as broadly as possible. 7  All computational work was done with R, a statistical programing language commonly used in data science. 8  This is a bit uneven in part because TEC, which was originally Studies in Burke, did not begin its run until 1978, while ECS began its run in 1967. 5 6

5  TEXT MINING AND DATA VISUALIZATION 

157

I then imported plain text documents for “cleaning.” This included removal of front and back matter, headers, footers, page numbers, and any formatting errors resulting from the import process.9 At this point, all characters were converted to lowercase and all punctuation and stopwords (i.e., high frequency function words such as “like,” “as,” “the,” “and,” “but,” etc.) were removed. Next, I proceeded to create an initial model of eighteenth-century poetry criticism from this larger set of essays. Because I was interested in articles that dealt primarily with poetry, I created a list of key terms that I felt would best typify an essay on poetry (“poet,” “poets,” “poetry,” “poetics,” “poem,” and “poems”) and filtered this initial model by using term frequency-inverse document frequency (tf-idf), as a statistical measure of semantic significance.10 This narrowed down the original text collection from 1536 to 310 articles. This 310-article model of eighteenth-century poetry is the poetry criticism corpus (PCC) that I have used for algorithmic transformation and modeling alongside the entire corpus of 1536 articles.

Methodology As in many other disciplinary fields, tactical use of specialized vocabularies constructs knowledge in the literary history and criticism of eighteenth-­ century poetry as well. Attending to the sociolects of eighteenth-century poetry criticism offers valuable insights into how poetry has mattered— the development and transformation of the field’s discursive conventions, theoretical commitments, and methodological assumptions. This type of broad sweeping survey often moves at the institutional level of big ideas. 9  Such errors include non-ascii characters, collapsed words, and encoding problems when images were read by the OCR software and converted to plain text from a PDF. 10  To rank terms, I used term frequency-inverse document frequency (tf-idf) statistics. Term frequency (tf) is simply how many times a term appears in a document. Common terms will have a higher tf while more specialized words will have a lower tf. Inverse-document frequency (idf) is computed by dividing the total number of documents in a corpus by the number of documents that contain a specific term. “Poet,” for example, might have a relatively high term frequency in an essay about poetry. However, that high frequency is offset by the idf, which adjusts to account for how many documents there are in an entire corpus with the word “poet” in them. Consequently, terms in a document with a high tf-idf score are likely indicators of a document’s content, while terms in a document with a low td-idf score are not likely indicators of a document’s content. In this case, I combined the tf-idf score for each term into one aggregate td-idf score and then selected only those essays with an aggregate td-idf score of 0.002 or higher.

158 

B. HALL

One thinks of Gerald Graff’s “institutional history,” or Richard Ohmann’s “politics of letters,” which negotiate change at relatively high levels of abstraction.11 Conversely, quantitative methods, such as text-mining and computer-assisted text analysis, shift attention to the atomic level of the word. Figure 5.2a, b illustrate how this kind of word-level attention highlights structural elements within the corpus over time. Figure 5.2a includes author names which act as the highest organizational level with respect to words with significance in text collection. If proper names are removed, Fig. 5.2b suggests that the next level of significance includes the discursive terms used (i.e., terms associated with categories like gender, genre, and subject). More important, this type of word frequency analysis is highly suggestive of patterns of usage that change over time and, with further analysis, potentially illuminate cultural changes in eighteenth-century poetry criticism. Tracking the top tf-idf terms across five decades of poetry criticism foregrounds how computational modeling can provide the “tools, techniques, and practices” that Andrew Piper argues construct our knowledge.12 The computation of word frequencies, document clustering using k-means algorithms, and LDA topic modeling offer ways to carefully read an otherwise impractical number of essays at, perhaps, the closest level possible. When these algorithmic manipulations are coupled to data visualizations that translate the numerical underpinnings of quantitative modeling into legible representations, they help us in identifying latent discourse patterns and the structures that produce those patterns. Perhaps equally important, quantitative analysis brings to light “patterns of representation that may not be visible to the individuals participating in them.”13 A quantitative approach to a historical survey of contemporary poetry criticism has not been attempted yet in eighteenth-century studies.14 11  Gerald Graff, Professing Literature: An Institutional History (Chicago: University of Chicago Press, 1987); Richard Ohmann, Politics of Letters (Middletown: Wesleyan University Press, 1987). 12  Piper, “Think Small,” 651. 13  Andrew Goldstone and William E. Underwood, “The Quiet Transformations of Literary Studies: What Thirteen Thousand Scholars Could Tell Us,” New Literary History 45, no. 3 (2014): 379, https://doi.org/10.1353/nlh.2014.0025 14  Michael Gavin’s “Historical Text Networks: The Sociology of Early English Criticism” takes a quantitative approach to what he calls “historical text networks” derived from metadata of works contained in the Early English Books Online (EEBO) database. For details, see Michael Gavin, “Historical Text Networks: The Sociology of Early English Criticism,”

5  TEXT MINING AND DATA VISUALIZATION 

159

Fig. 5.2  (a–b) Contrasting words in the poetry criticism corpus (PCC) with and without authors’ names. These visualizations indicate how both the authors considered and the language used to discuss poetry have changed from the first twenty-five years to the second twenty-five years. The most frequent and exclusive terms to each period appear at the top and the bottom of the two charts

160 

B. HALL

Fig. 5.2  (continued)

5  TEXT MINING AND DATA VISUALIZATION 

161

Unlike an introductory essay or “companion” volume which reflects broadly on the field of eighteenth-century studies by selecting a small sample of critical work, a quantitative approach does not provide a survey of historical and cultural contexts in which works are situated, nor does it engage in an analysis of the literary genres, themes, forms, and practices of eighteenth-century poetry. Rather, it uses computational methods to assess a specific domain in the broader discipline of literary studies. Specifically, this approach parallels the work of Stephen Ramsey, Andrew Goldstone, Ted Underwood, Dan Edelstein, and other scholars who have used quantitative methods to explore historical patterns in scholarship and challenge the concepts that underwrite the work of criticism.15 It also connects to the work of Matthew Jockers, Andrew Piper, and Willard McCarty, who offer new concepts and vocabulary for considering the specific questions and problems that emerge when the availability of large digitized datasets meets the affordances of text mining and data visualization.16 “Modeling of something,” McCarty argues, “readily turns into modelling for a better or more detailed knowledge of it; similarly, the knowledge gained from realizing a model for something feeds or can feed into an improved version.”17 Therefore, this exploratory modeling of a corpus of two influential journals in the field of eighteenth-century studies is useful both for clarifying what we already know and for suggesting ways to improve or develop the field in new and, perhaps, unexpected ways. It is a necessary practice of literary historians and critics to develop, adapt, and use models of the field because those models offer useful representations of an ever-expanding mass of writing that has largely exceeded our capacity to digest in its entirety. Two conceptual models, for instance, seem to hold sway over how eighteenth-century scholars conceive Eighteenth-Century Studies 50, no. 1 (2016): 53–80, https://doi.org/10.1353/ ecs.2016.0041 15   See Stephen Ramsey, “Special Section: Reconceiving Text Analysis: Toward an Algorithmic Criticism,” Literary and Linguistic Computing 18, no. 2 (2003): 167–74, https://doi.org/10.1093/llc/18.2.167; Goldstone and Underwood, “The Quiet Transformations of Literary Studies,” 359–84; and Dan Edelstein, “Enlightenment Scholarship by the Numbers: dfr.jstor.org, Dirty Quantification, and the Future of the Lit Review,” Republics of Letters 4, no. 1 (2014): 1–26, https://arcade.stanford.edu/sites/ default/files/article_pdfs/ROFL_v5_Edelstein_final.pdf 16  Matthew Jockers, Macroanalysis: Digital Methods & Literary History (Urbana: University of Illinois Press, 2013); Andrew Piper, Enumerations: Data and Literary Study (Chicago: University of Chicago Press, 2018); and McCarty, Humanities Computing. 17  McCarty, Humanities Computing, 27.

162 

B. HALL

eighteenth-century poetry. On the one hand, there are evolutionary models of eighteenth-century British poetry, which assume change and track the transformation of poetry over time. While the details vary, perhaps the most common feature of these models includes a movement through time from a Neoclassical/Augustan phase to a post-Augustan/pre-Romantic phase that eventually gives way to full-bodied Romanticism. On the other hand, there are revolutionary models, which are driven by an impulse to seek out the new. Such models might eschew the historical constraints of change-over-time as they range widely over the literary field in search of new voices and poetics subjects. They often break up old canons, seeking alternative threads to follow through a rapidly expanding field of new poets and texts. Whether evolutionary or revolutionary in nature, the models we use to interpret, categorize, unpack, and manipulate eighteenth-­ century cultural artifacts are often intricate and sophisticated. The problem, and paradoxically the virtue, of these complex historical and hermeneutic models is their transparency as organizational and interpretive tools. It seems that, in the application of interpretive frames, the assumptions of a model too often recede behind the argument of its significance, and the model stands in for the reality it exemplifies. My model of eighteenth-century poetry criticism is intended to highlight the decisions I have made in the construction and algorithmic manipulation of the model, as well as the interpretive value of those decisions. Interdisciplinary definitions of models and debates about the practice of modeling have a long history and range widely between the practical and conceptual work of building models and the manipulation of models in the process of building knowledge. In the digital humanities, computational power and access to large digitized text collections have increasingly brought the conceptual and practical issues of data modeling to the fore.18 Therefore, I begin from the accepted notion that a model is only representational and not the actual data in the world, that is, it attempts to formalize aspects of a dataset. As Julia Flanders and Fotis Jannidis correctly point out, “the data model is not identical with the data, but rather a description of it that includes the semantics of the data and from which one can derive 18  See McCarty, Humanities Computing; the essays collected in Julia Flanders and Fotis Jannidis, eds., The Shape of Data in Digital Humanities: Modeling Texts and Text-Based Resources (London: Routledge, 2018); Piper, “Think Small;” Arianna Ciula and Oyvind Eide, “Modelling in the Digital Humanities: Signs in Context,” Digital Scholarship in the Humanities 32, no. 1 (2017): i33-i46, https://doi.org/10.1093/llc/fqw045; and Graeme Simsion, Data Modeling: Theory and Practice (New Jersey: Technics Publications, 2007).

5  TEXT MINING AND DATA VISUALIZATION 

163

more formal aspects like the structural properties of the data or consistency constraints that apply to it.”19 Moreover, models need not be mimetic in a strict sense, but rather operate “informationally, as in the metaphor of a map.”20 Attending to the structural properties of data formalized in a model has significant consequences. As Piper notes, models “shift the focus toward the signifiers of research and away from the signifieds.”21 In this way, models redirect attention away from potential conclusions about the meaning of a dataset because they focus on the structures and processes that formalize knowledge—the “tools, techniques, and practices” that inform one’s choices. Thus, while far from comprehensive, the modified collection of eighteenth-century poetry criticism extracted from the two eighteenth-century journals I have selected serves as a useful model for algorithmic manipulation, k-means clustering, LDA topic modeling, and data visualization. It is, hopefully, a valuable model for drawing inferences that open new topics of inquiry.

K-Means Clustering and Topic Modeling In its either supervised or unsupervised flavors, text mining is organized around two main concepts. The first entails reducing the amount of information, and in particular “noise,” that obscures structures in large text collections. In addition to information reduction, the second guiding concept is identifying patterns within a corpus of documents. Unsupervised forms of text mining are particularly suited to exploratory work because the algorithms are designed to find patterns in unstructured data and, in the case of clustering algorithms, structural features whereby documents can be organized into groups.22 “Topic modeling,” David M. Blei explains, 19  Julia Flanders and Fotis Jannidis, “Data Modeling,” in A New Companion to Digital Humanities, ed. Susan Schreibman, Raymond G.  Siemens, and John Unsworth (Malden: Wiley/Blackwell, 2016), 230. 20  Piper, “Think Small,” 652. 21  Ibid. 22  For a concise introduction to text mining, see Matthew L. Jockers and Ted Underwood, “Text-Mining the Humanities,” in Schreibman, Siemens, and Unsworth, A New Companion to Digital Humanities, 291–306. The field-specific literature on text mining is vast and often complicated for humanists not trained in statistics and computer science. Ashok N. Srivastava and Mehran Sahami, eds., Text Mining: Classification, Clustering, and Applications (Boca Raton, FL: CRC Press, 2009) and Michael W. Berry, ed., Survey of Text Mining: Clustering, Classification, and Retrieval (New York: Springer-Verlag, 2004) are both excellent introductions to methods for text mining and knowledge discovery. More practical and hands on

164 

B. HALL

“provides a suite of algorithms to discover hidden thematic structure in large collections of texts. The results of topic modeling algorithms can be used to summarize, visualize, explore, and theorize about a corpus.”23 In the so-called bag-of-words (BoW) approach, k-means and topic modeling algorithms use statistical methods that do not rely on syntactic or semantic structures the way human readers depend on meaning, grammar, and semantics.24 The BoW approach is not concerned with word order or type, and documents are usually converted to lowercase and stripped of normal approaches to text mining include Julia Silge and David Robinson, Text Mining with R: A Tidy Approach (Sebastopol, CA: O’Reilly, 2017); Matthew L. Jockers, Text Analysis with R for Students of Literature (New York: Springer, 2014); Kasper Welbers, Wouter Van Atteveldt, and Kenneth Benoit, “Text Analysis in R,” Communication Methods and Measures 11, no. 4 (2017): 245–65, https://doi.org/10.1080/19312458.2017.1387238; Ted Kwartler, Text Mining in Practice with R (Hoboken: John Wiley & Sons, 2017); and Hadley Wickham and Garrett Grolemund, R for Data Science: Import, Tidy, Transform, Visualize, and Model Data (Sebastopol, CA: O’Reily, 2017). 23  David M. Blei, “Topic Modeling and Digital Humanities,” Journal of Digital Humanities 2, no. 1 (2012): n.p., http://journalofdigitalhumanities.org/2-1/topic-modeling-and-digital-humanities-by-david-m-blei/. The literature on topic modeling is large and ranges from technical essays on statistics and algorithm performance to application in humanities to tutorials. For useful overviews of topic modeling in the humanities, see Scott Weingart’s post, “Topic Modeling for Humanists: A Guided Tour,” The Scottbot Irregular (blog), July 25, 2012, http://scottbot.net/2012/07/; Ted Underwood’s various posts on The Stone and the Shell; and Andrew Pipers’ posts, “Topic Modelling Literary Studies: Topic Stability, Part 1,” Txtlab (blog), May 28, 2010, https://txtlab.org/2018/05/topic-modelling-literarystudies-part-1-topic-stability/, and “Topic Stability, Part 2,” Txtlab (blog), June 7, 2018, https://txtlab.org/2018/06/topic-stability-part-2/. The essays in The Journal of the Digital Humanities 1, no. 1 (2012) also provide an excellent introduction to topic modeling. David M. Blei has also written several essays both technical and introductory; among them, “Probabilistic Topic Models,” Communications of the ACM 55, no. 4 (2012): 77–84, https://doi.org/10.1145/2133806.2133826 is one of the clearest introductions. For reading and interpreting topic models, see Jonathan Chang, Sean Gerrish, Chong Wang, Jordan L. Boyd-Graber, and David M. Blei, “Reading Tea Leaves: How Humans Interpret Topic Models,” in Advances in Neural Information Processing Systems 22, ed. Y.  Bengio, D.  Schuurmans, J.  D. Lafferty, C.K.I.  Williams, and A.  Cullota (Vancouver: Curran Associates, 2009), 288–96. 24  The “bag-of-words” (BoW) approach does not include word order or other semantic markers. Instead, a BoW algorithm generates a unique vocabulary for a text or text collection and some unit of measurement—simple word frequency or tf-idf are common measures. Because the BoW method is common in natural language processing and text mining, there are a number of tutorials available. A good introduction can be found in Yin Zhang, Rong Jin, and Zhi-Hua Zhou, “Understanding Bag-of-Words model: A Statistical Framework,” International Journal of Machine Learning and Cybernetics 1, no. 1–4 (2010): 43–52, https://doi.org/10.1007/s13042-010-0001-0

5  TEXT MINING AND DATA VISUALIZATION 

165

reading markers like punctuation and stopwords. Once converted into this form, text collections are prepared for algorithmic transformation, the output of which is often matrices where document features are paired to some numerical measurement—frequency, tf-idf, or topic distribution. “The goal of algorithmic modeling,” Ted Underwood explains, “is to translate that uninformative representation, provisionally, into something more meaningful.”25 In order to work back from that cumbersome and difficult form, data visualization techniques help to transform columns of numbers and words into visual representations that are more easily legible and provisionally meaningful. To do so, I began mapping the discursive landscape with a topic model and a k-means cluster analysis26 of all 1563 essays from the text collection of the two journals.27 Because I was ultimately interested in a poetry corpus of essays drawn from a larger text collection, this initial modeling was an effort to determine if literary genre would be identifiable by machine reading as a key structure in the larger text collection. Topic modeling algorithms treat each essay in the corpus as an unstructured collection of words. In this BoW approach, we can consider each document as consisting of parts from every topic. Additionally, each topic is made up of all the documents in the corpus used to generate the model. Similar documents share a high percentage of a particular topic, and similar topics most likely share a high percentage of the same documents. The results were then mapped into a network graph to provide an overview of the underlying model, with “genre” nodes and their associated documents colored according to literary genre (“novel”/red, “poetry”/blue, “theater”/ green). As seen in Fig.  5.3, the size of the topic node depends on its 25   Ted Underwood, “Algorithmic Modeling: Or, Modeling Data We Do Not Yet Understand,” in Flanders and Jannidis, The Shape of Data in Digital Humanities, 261. I discuss topic modeling in detail below. 26  The k-means algorithm is normally used for document clustering when browsing a collection for similar documents or other search engine implementations. For details, see E.  Laxmi Lydia, P.  Govindasamy, S.  K. Lakshmanaprabu, and D.  Ramya, “Document Clustering Based on Text Mining K-Means Algorithm Using Euclidean Distance Similarity,” Journal of Advanced Research in Dynamical and Control Systems 10, no. 2 (2018): 208–14; and Mehdi Allahyari, Seyedamin Pouriyeh, Mehdi Assefi, Saied Safaei, Elizabeth D. Trippe, Juan B.  Gutierrez, and Krys Kochut, “A Brief Survey of Text Mining: Classification, Clustering and Extraction Techniques,” arXiv e-print (2017), arXiv:1707.02919. 27  I used here the standard Gibbs version of LDA in the “topicsmodels” package for R, with minimal tweaking of the default setting. However, I arrived at the number fifty after cross-validating with the algorithms from the “ldatuning” package.

166 

B. HALL

Fig. 5.3  Network graph derived from a 50-topic model of all 1500+ essays. I have highlighted here topics that correlate strongly with traditional literary genres. This figure illustrates how algorithmic transformations like topic modeling identify genre—“poetry,” “novel,” and “theater”— as a structural feature within the model

in-­degree, or, in this case, how many essays link back to that node. The larger the node in the network, the more documents are linked to that topic. As highlighted here, the three main genre topics are “novel,” “theater,” and “poetry.” Among these, the prominence of the “poetry” and “novel” nodes indicates their relative importance within the text collection as a whole. While the network graph does not provide a great deal of detail about what is in those topics, it does suggest that genre is a strong discursive marker and that the algorithm clusters essays that share that genre-based lexicon. One of the disadvantages of the k-means algorithm is that it does not perform well on high-dimensional matrices, such as text collections, when there are many variables in the form of individual words.28 Since the topic model provided a good general overview of the importance of genre, I turned to the k-means clustering algorithm to test its performance at clustering documents by genre. For the initial test to evaluate how well k-means would cluster articles, I pre-selected essays according to their association to a literary genre and called this the Genre Group Corpus (GGC). From the 1536 essay in the entire collection, I selected three 28  Underwood points out this problem in his article, “Algorithmic Modeling.” As of now, I have only begun to experiment with spherical k-means and other versions of the k-means algorithm that try to overcome this limitation.

5  TEXT MINING AND DATA VISUALIZATION 

167

groups of roughly thirty essays from each major genre—novel, poetry, theater—extracted from the larger corpus by using tf-idf scores.29 For this initial test, I performed k-means clustering on six different versions of the GGC, where the number of terms, or dimensions, ranged from 6 to 8769.30 I also set k = 3, so that the algorithm would, presumably, generate three clusters—one for each genre group. The algorithm produced different clustering results for each of these iterations depending on how many words they included. The best clustering result came at 1961 terms, which clustered these articles into groups of twenty-eight (cluster 1, novel), twenty-seven (cluster 2, theater), and twenty-four (cluster 3, poetry) articles per cluster, respectively (Fig.  5.4). At this level of granularity, the algorithm clearly clustered essays based on genre terms. Theater articles and poetry articles (clusters 2 and 3) have a relatively clear separation. However, the novel cluster (cluster 1) overlaps both the poetry and theater clusters, with its centroid straddling the dividing line between the other two clusters: this suggests that the vocabulary of novel criticism shares more terms with poetry and theater lexicons than the vocabulary of poetry and theater share with each other. Despite the failure of the k-means algorithm to plot the GGC into three distinct clusters, the fact that the algorithm picked up genre terms and grouped together essays that shared a genre marker suggested that, with some tweaking of the dials, k-means might be useful for clustering essays in the poetry criticism model. To move beyond the algorithmically modeled map of eighteenth-­ century studies as a whole and explore the details particular to the sociolects of poetry criticism, I needed to explore the textual features that the machine algorithms would identify as statistically relevant in the more limited model that included only poetry criticism. The poetry criticism 29  The essays were selected based on the highest tf-idf score for each genre term. I then handchecked the three genre lists to ensure that the included essays did, in fact, deal primarily with genre-based material. 30  To minimize the impact of dimensionality, I varied the sparsity of the genre group document matrix to limit the number of words used by the algorithm. This reduction is, of course, not trivial because it means that document clustering will be based on a set of words that appear within a range of very common to very uncommon. A very sparse matrix, for instance, will contain all the words, many of which will only appear in one or a handful of documents. On the other hand, a matrix with very low sparsity will only contain those words that appear in all or nearly all documents. The idea of multiple models was to get closer to a set of words that will be representative of an essay by eliminating idiosyncratic words and, at the same time, not becoming so generic that the words fail to distinguish between documents and become, therefore, useless for clustering.

168 

B. HALL

Cluster Plot: Genre 1961 words 40

cluster 1 2 3

20

0

−20

−10

0

10

20

Fig. 5.4  K-means clustering of the genre subset: novel (cluster 1, red), theater (cluster 2, green), and poetry (cluster 3, blue)

corpus (PCC) is a subset of 310 articles discussing aspects of eighteenth-­ century poetry filtered down from the original 1563 essays.31 If unsupervised methods worked to isolate topics and clusters of essays by genre terms, I hoped those methods might also be used to map the discursive terrain of a more field-specific collection of poetry-oriented texts. My working assumption was that the algorithms would identify three categories of terms: terms that typified a disciplinary approach (i.e., formalist, historicist, feminist, poststructuralist), poetic genres (i.e., satire, ballads, 31  I used the tf-idf scores to separate essays that had a high tf-idf score on a list of poetry terms (i.e., poet, poets, poem, poems, poetry, poetics), and to obtain an aggregate score for all poetry terms. Using those numbers, I removed articles with an aggregate tf-idf poetry terms score below 0.002, which resulted in a PCC of 310 essays.

5  TEXT MINING AND DATA VISUALIZATION 

169

odes, pastoral), and author names. Once identified, I assumed that tracking changes to those terms over the course of fifty years would show how governing narratives of the field have shifted over time. My question was if the clustering algorithms would underwrite k-means clustering and topic modeling group essays in such a way that those clusters would echo the emergence of the so-called “new eighteenth century.”32 Beyond this, though, I hoped that the clusters would reveal important structural elements of the discourse and offer new insights into the disciplinary culture informing eighteenth-century poetry criticism of the last fifty years. Starting with k-means, the initial assessment of the PCC’s clusterability was low. Additionally, unlike the genre test, which had a clear number of clusters to look for in the GGC, the PCC did not have a clear way to determine a value for k and assign a number of clusters for the algorithm to generate.33 In an effort to get a range of results, I ran nine iterations of k-means algorithm on the PCC, both with the names of poets included and with the names of poets removed (Figs. 5.5 and 5.6).34 Unfortunately, the algorithm did not perform well at separating the poetry-related essays into distinct and informative clusters. The nine plots based on this model in Fig. 5.6 show in nearly every case that the k-means algorithm produced models with one or two very large clusters and then smaller groups nested within those larger clusters. Rather than clear clusters of essays that separate along an expected structural feature (i.e., disciplinary method, poetic genre, or author), k-means seemed to fail at distinguishing document clusters within the PCC. It appears that, even when the dimensions (i.e., the number of terms) of the poetry model were reduced, the algorithm was not capable of clustering the poetry criticism essays. However, when examining the top terms for each cluster, I found that at or near the top of the list there are poet names (Table 5.1). This happens over and over with k set to various values and the matrix to various levels of sparsity. It seems, then, that either the discourse of poetry criticism is so homogenous as to 32  The term was coined in Felicity Nussbaum and Laura Brown, eds., The New Eighteenth Century: Theory, Politics, English Literature (North Yorkshire: Methuen, 1987). 33  As an unsupervised method, k-means, like topic modeling, requires the modeler to determine a value for k that determines the number of clusters the algorithm will generate. For the k-means analysis of the poetry corpus, I used the elbow method and the Hopkins factor for estimating clusterability and the number of clusters. 34  Each iteration set k to 3, 6, and 9 and included terms ranging from 25,521 to 1318 to 318 (sparsity of 93%, 51%, and 25%, respectively), with the key feature being the tf-idf score for each word in the matrix.

170 

B. HALL

Fig. 5.5  K-means clustering with the poets’ names included

5  TEXT MINING AND DATA VISUALIZATION 

Fig. 5.6  K-means clustering with the poets’ names removed

171

Johnson Pope Painting Songs Swift Dryden

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Cluster 6

Pastoral Women Thomson Howard Jonathan Milton

Top ten terms

Cluster Pope Dunciad Music Song Beauty Music

Shakespeare Sexual Landscape Lyric Woman Johnson

Renaissance Trade Visual Popular Enlightenment Biblical

Homer Female Artist Music Aesthetic King

Epic Horace Sublime Air Narrator Virgil

Table 5.1  Top ten words for a six-cluster model of the matrix set at 61% sparsity

Paris Wordsworth Melancholy National Locke Narrator

Latin Rape Portrait Collections Women Wit

Golden Lady Pictures Scholars Aesthetics Epic

172  B. HALL

5  TEXT MINING AND DATA VISUALIZATION 

173

be unclusterable or that the algorithm was pressing against its limits in a large and diverse text collection with high dimensionality. My suspicion was that the tf-idf feature, which is particularly good at ranking terms by significance to a particular document, was pushing the names of poets to the top and obscuring other structural components in the data. Clearly, tf-idf scoring is good at identifying who an essay is about, but that strong authorial presence was potentially obscuring other discursive terms. Because the author signal was strong, I assumed that k-means would cluster essays together by author, showing a Pope cluster, or a Finch cluster. However, that failed to happen. Rather, the clustering algorithm did not seem to differentiate articles on Pope, Swift, or Johnson from articles on Montague, Finch, and Barbauld. Again, my sense was that the prominence of a poet’s name was obscuring discursive structures that lay deeper down the statistical chain—terms that would indicate differences within several essays on Pope and that might cluster essays other than by the poet’s name. However, when the poets’ names were removed, k-means produced similar results (Fig. 5.6): in each case, one large generic cluster that encompassed all the smaller clusters.35 It would seem, therefore, that the k-means approach to document clustering reinforced the structural importance of the author names; however, in this case, the authorial signal was not strong enough to produce informative clusters. Unfortunately, k-means was not as helpful as expected in clustering poetry criticism essays and visualizing the discursive terrain within eighteenth-­ century poetry criticism. However, the other unsupervised machine clustering, LDA topic modeling,36 offered a powerful set of algorithms to work recursively though various models of poetry criticism articles and look for structures that organize them in ways that help illuminate the past fifty years of eighteenth-century poetry criticism. In this type of analysis, the algorithm traverses a corpus looking for terms that tend to occur together and then groups these word clusters in “topic” categories (see Table  5.2 for top terms per topic in a twelve-topic structural topic 35  I also ran k-means on two more versions of the corpus (one with author names and one with author names removed), using raw term frequency as the main feature. In each of the eighteen iterations, the algorithm performed poorly at clustering these essays. 36  Latent Dirchlet Allocation (LDA) is a statistical way of organizing data according to similarity. Topic modeling that uses LDA allows for creating a nuanced model of a text corpus because it moves beyond mere word frequency to generate lists of words likely to occur together in a document. The assumption is that every document is a collection of topics, so LDA assigns a probability for each topic to occur in each individual document.

174 

B. HALL

Table 5.2  Top twenty most probable terms in each topic generated by the twelve-topic model using the stm package Top twenty probable terms: highest probability in a topic Topic 1: World, god, nature, lines, pastoral, thy, sense, life, poetic, line, form, time, passage, dullness, language, art, light, mind, verse, song Topic 2: Life, literary, satire, nature, literature, criticism, human, essay, world, sense, reader, moral, history, critics, critical, mind, time, wit, age, theory Topic 3: Literary, public, social, authors, books, society, political, genius, community, critical, author, writing, cultural, writers, modern, text, copyright, literature, time, history Topic 4: Life, sexual, death, desire, love, body, readers, women, child, text, social, reading, children, die, pleasure, physical, woman, sense, time, world Topic 5: War, trade, political, American, slave, social, slavery, slaves, economic, nature, georgic, land, society, country, history, labor, life, America, power, natural Topic 6: Political, king, fables, power, music, war, time, lines, politics, James, sense, god, biblical, body, Jacobite, tory, peace, kings, love, Williams Topic 7: Language, literary, writing, poetic, readers, knowledge, reading, critical, social, discourse, theory, form, system, words, meaning, text, modern, criticism, critics, texts Topic 8: Women, female, love, woman, male, sexual, desire, social, body, human, women’s, feminine, world, sex, friendship, life, public, gender, power, physical Topic 9: Ballads, ballad, songs, song, history, literary, cultural, collection, language, Chinese, culture, popular, time, Scottish, oral, texts, literature, world, historical, tradition Topic 10: Art, painting, history, nature, arts, language, artist, essay, imitation, theory, translation, time, music, modern, form, fall, classical, age, style, baroque Topic 11: Nature, world, human, imagination, natural, science, mind, time, god, sense, reason, experience, power, philosophy, form, language, matter, scientific, vision, history Topic 12: Literary, time, life, letters, political, wrote, letter, public, author, George, death, printed, lord, history, garden, writing, written, figure, opera, monument

model of the PCC). An example of this kind of exploratory pattern recognition is how the topic model distributes topics between the two journals: topics 1, 2, 3, 10, and 12 are more likely to appear in ECS, while topics 4, 5, 6, 7, 8, 9, and 11 tend to appear more in TEC (Fig. 5.7). Meaningful questions about this distribution of topics by journal require drilling into

5  TEXT MINING AND DATA VISUALIZATION 

175

Fig. 5.7  Visual representation of topic distribution between the two journals. Topics close to the center line are those more likely to be shared by both journals, while topics on the margins represent discourse more closely aligned to one journal or another

176 

B. HALL

the constitution of the topics themselves, as seen below. However, before doing that, it is worth giving one example of the suggestiveness of this division of topics at journal level. Topic 7, which includes the language of criticism and theory, tends to TEC. On the other hand, topic 1, which signals an older discourse about poetry connected to the poetic genres of pastoral and hymn, leans strongly to ECS. The initial structural topic modeling (stm)37 seemed to point toward the kind of features I hoped the modeling would make salient: (1) sociolects tied to poetic genres, subjects/themes, or methodological orientation; and (2) changes in these sociolects and structural features over time. After experimenting with a number of flavors of LDA topic modeling, I settled on the stm that builds on previous work in topic modeling and that allows users “to incorporate arbitrary metadata, defined as information about each document, into the topic model.”38 The twelve-topic structural topic model built with stm allows for the inclusion of document-­ level metadata in the estimation of models. Here, the metadata features of interest in the PCC include publication year and journal of origin. The other tweak I made to the PCC prior to modeling with stm was the removal of poets’ names. I return to this important structural feature in the next section of this chapter; however, to foreground discursive features and sociolects in the PCC, I removed these names from this modeling. In terms of visualization, one of the challenges of interpreting topic models stems from the difficulty of visualizing the entire model at once. Traditionally, network graphs (like the one in Fig. 5.6) have served as the go-to method to visualize the complexity underlying such models. They are visually appealing and often dynamically interesting visualizations of topic models. However, the practice of visualizing topic models as networks is not without its own history and controversies. Primarily, the complication pertains to the way that network graphs tend to purchase their legibility at the expense of data.39 Therefore, instead of using a network 37  An stm is a version of the LDA algorithm that allows for document-level metadata to be included in the modeling process. For technical information and an overview of the stm features, see Margaret E. Roberts, Brandon M. Stewart, and Dusting Tingley, “stm: An R Package for Structural Topic Models,” Journal of Statistical Software 91, no. 2 (2019): 1–40, https://doi.org/10.18637/jss.v091.i02 38  Ibid., 1–2. 39  The main issue with network graphs representing a model in its entirety is that, in practice, every topic is connected to every node, which does not translate into very readable and meaningful visualizations. Using multidimensional scaling allows for presenting the entire model, without eliminating data, in a legible way. Ted Underwood has an excellent illustration of these issues in “Visualizing Topic Models,” The Stone and the Shell (blog), November

5  TEXT MINING AND DATA VISUALIZATION 

177

graph to visualize my model of the poetry corpus, I used the topic model visualization package LDAvis,40 which allowed me to explore the underlying model as a whole, without having to trim data. Part of the LDAvis output is the Intertopic Distance Map in Fig.  5.8. This map has two important visual features for unpacking features in the structural topic model’s clustering of essays. First, the size of each topic bubble indicates the expected prevalence of that topic in the corpus. Second, the map organizes topics in a two-dimensional space as bubbles based on the

Fig. 5.8  Output from LDAvis for a twelve-topic model of PCC 11, 2012, https://tedunderwood.com/2012/11/11/visualizing-topic-models/. See also the conversation in comments. 40  LDAvis offers a two-dimensional alternative to the network graph for capturing and interpreting the model when zoomed out to visualize the entire model. It is browser-based and is best viewed in its dynamic form.

178 

B. HALL

distribution of topics over all the words of the corpus.41 This means that overlapping topic-bubbles or topic-bubbles that are proximate in the two-­ dimensional space of the graph have a higher degree of similarity than topics that appear further away from each other. With respect to expected topic proportion, I anticipated that modeling a text collection of essays on a specific literary genre would produce a very general topic that appears in a high number of documents. I assumed that this expected generic term topic would include the terms of art that constitute poetry criticism and analysis—figurative language terms, rhetorical tropes, formal conventions, and metrical language. However, the uniformity of the bubble sizes in the Intertopic Distance Map of the twelve-­ topic model suggests otherwise. In fact, the distribution of expected topic proportion is very narrow, between 10% and 7% (Table 5.3). This even distribution of topic proportions suggests, unexpectedly, that there is no one large topic that captures a common lexicon shared by essays discussing eighteenth-century poetry. Despite the absence of a single controlling discourse topic, there are three clusters of topics in Fig. 5.8 that invite attention: topics 2, 7, and 10 in the upper left-hand quadrant, topics 3, 9, and 12 in the upper right-­ hand quadrant, and topics 4 and 8 in the lower central portion. Collectively, these three clusters account for 65% of the documents containing these topic terms and strongly suggest the main structural features and shape of the corpus. Topics 2, 7, and 10, for instance, account for roughly 26% of documents built from the model, and most closely resemble what I had expected to appear in a single large topic of general discourse terms (see Table 5.3 for top ten terms associated with these topics). In addition to the general discourse cluster, the stm model generated two additional clusters of topics that clarify more significant structural features of the PCC. In the lower-central position of the Intertopic Distance Map, topics 4 and 8 show a significant overlap. Both topics are marked by a high prevalence of discursive language associated with women, sex, and, 41  As far as I can tell, the multidimensional scaling (mds) done in LDAvis is a distance method of Jensen-Shannon and the multidimensional scaling handled by the cmdscale function, which is, in turn, plotted in a two-dimensional space. See Carson Sievert and Kenneth E.  Shirley, “LDAvis: A Method for Visualizing and Interpreting Topics,” in Proceedings of the Workshop on Interactive Language Learning, Visualization, and Interfaces, ed. Jason Chang, Spence Green, Marti Hearst, Jeffrey Heer, and Philipp Koehn (Baltimore, MA: Association for Computational Linguistics, 2014), 63–70, https://nlp.stanford.edu/ events/illvi2014/papers/sievert-illvi2014.pdf

5  TEXT MINING AND DATA VISUALIZATION 

179

Table 5.3  Topic prevalence in descending order of expected proportion and top ten terms for each topic in the twelve-topic model of the PCC Topic

Expected topic proportion

Topic 1

9.90%

Topic 2

9.70%

Topic 11

9.30%

Topic 8

9.20%

Topic 7

8.50%

Topic 5

8.00%

Topic 9

8.00%

Topic 10

7.70%

Topic 4

7.60%

Topic 12

7.60%

Topic 6

7.50%

Topic 3

6.90%

Top ten terms World, god, nature, lines, pastoral, thy, sense, life, poetic, line Life, literary, satire, nature, literature, criticism, human, essay, world, sense Nature, world, human, imagination, natural, science, mind, time, god, sense Women, female, love, woman, male, sexual, desire, social, body, human Language, literary, writing, poetic, readers, knowledge, reading, critical, social, discourse War, trade, political, American, slave, social, slavery, slaves, economic, nature Ballads, ballad, songs, song, history, literary, cultural, collection, language, Chinese Art, painting, history, nature, arts, language, artist, essay, imitation, theory Life, sexual, death, desire, love, body, readers, women, child, text Literary, time, life, letters, political, wrote, letter, public, author, George Political, king, fables, power, music, war, time, lines, politics, James Literary, public, social, authors, books, society, political, genius, community, critical

more broadly, desire and violence. The fact that the two topics cluster together highlights how these two major and interrelated strands of discourse related to women and sexuality are key methodological and thematic components in the PCC.  In the upper right-hand coordinates of Fig. 5.8, topics 3, 9, and 12 form another important cluster. Topics 9 and 12 are particularly connected to genres like epistle and ballad, but all three topics together suggest a strong relation to print culture and the social life of writing. In contrast to these clusters of topics, the individual topic bubbles in Fig. 5.8, and the top terms associated with those topics in Table 5.2, suggest a more specialized vocabulary, one concerned with poetic subjects and themes in poetry criticism rather than with broad subjects like print

180 

B. HALL

culture or methodological orientations. Topic 11, for instance, on the extreme left-hand side of Fig. 5.8 highlights a more insular group of essays concerned with the intersection of poetry and natural philosophy. Topic 1 turns out to be strongly linked to hymns, pastoral and nature poetry, and religious themes. Topic 6 is strongly correlated with fable and political poetry (it is no surprise that the uppermost essays here refer to John Dryden’s political poetry), while topic 5 relates to slavery. Taken together, the topic clusters and the individual topics mapped into this two-­ dimensional space provide a robust image of the stm model as a whole and offer a glimpse into the overall structural makeup of the PCC. In addition to mapping the methodological and thematic language clusters of the model, I wanted to explore how topics tracked over time. The stm package is particularly useful in this respect as its algorithms incorporate covariates, such as time, in the modeling process. Drilling into topics and visualizing movement though time with line graphs provides an important view on the structural shape of the corpus in terms of genre and discourse changes (Fig. 5.9a–d).42 An interesting example of such a change over time is how the two journals connect genres and/or themes to various time periods. Just like the topic clusters, the topics in the line graphs below align with each journal in interesting ways. All three topics that were popular in the 1970s and early 1980s (topics 1, 2, 10) tend to appear in ECS. In terms of genre, these topics include pastoral verse and satire, as well as poetry and painting; thematically, they seem to be focused on nature, criticism, and religion (Fig. 5.9a).43 Nevertheless, these ECS topics decline throughout the 1980s and never recover prominence in the model. Conversely, topics popular in the late 1980s through the early 2000s (topics 7, 8, 9) are strongly associated with TEC. Here, the topics refer to gender and sexuality, as well as social and cultural discourse. In the case of topic 9, there is a marked increase of interest in ballads, songs, and history during this period (Fig. 5.9b). While this chronological distribution of topics suggests that poetry criticism differed by journal, by 2010 there is a relatively even representation of topics trending upward and topics making a return between the two journals. Trending upward, topic 3 indicates a focus on essays related 42  See Benjamin M.  Schmidt, “Modeling Time,” in Flanders and Jannidis, The Shape of Data in Digital Humanities, 150–66. 43  The early prevalence of these topics in ECS is partly explained by the fact that TEC does not begin its print run until 1979.

5  TEXT MINING AND DATA VISUALIZATION 

Fig. 5.9  (a–d) Topics over time in the PCC corpus (1967–2018)

181

182 

B. HALL

to authors, readers, and books; topic 4 to sex, death, and the body; and topic 5 to essays concerned with slavery, war, and politics (Fig.  5.9c). Topics 11 (natural science) and 12 (epistolary writing) also trend upward after 2010, but curiously these are topics with a lexicon that was also prevalent in the early years of the two journals (Fig. 5.9d). If at one point the two journals had methodological or thematic differences expressed by their topic distribution, for the last decade the two journals appear to be in concert about what is happening in eighteenth-century poetry criticism. Stepping back from journal-topic-time distributions allows for a broader view of changes to the discursive structures in eighteenth-century poetry criticism. The topics divide into four main time-based groups. Perhaps not surprisingly, these trends reflect in broad strokes a narrative quite familiar to eighteenth-century scholars. Beginning in the top left-hand side of Fig. 5.9 and moving clockwise, we see an early phase of poetry criticism in the late 1960s to mid-1980s that focuses on pastoral and satire, as well as poetry and painting, nature, criticism, and religious themes. This gives way to the so-called “new eighteenth century”; during this twenty-year phase, there are three dominant sociolects linked to literary language and criticism, erotic themes, and history. This phase is followed by a “new,” new eighteenth century, with topics linked to print culture, slavery, war and sex, empire, and gender trending up after 2010. Connected to these trending topics are topics 11 and 12 (science/nature, and the “letter” genre); while only accounting for a relatively small proportion of the overall number of essays, these two topics once popular in the 1980s are increasing in frequency in recent years. Thus, the topic model and related visualizations present us with a view of the corpus as a whole. Much of its structure is familiar both in terms of the general discourse of poetry criticism and the more specialized language connected to subjects and themes. Topic clusters and topic frequency over time offer a narrative of change and development. In particular, they suggest that eighteenth-century poetry criticism undergoes a period of transformation. Trending topics and returning ideas seem to enter a new phase in the late 2000s.

5  TEXT MINING AND DATA VISUALIZATION 

183

Author Names over Time and Gender Inequality in the Poetry Criticism Corpus Part of the transformation in eighteenth-century poetry criticism is the expansion of the literary field to include poets that have either been excluded before or underrepresented—in particular, female poets. While it may be true that the author is dead, at every turn in the quantitative analysis of poetry criticism—tf-idf, k-means clustering, topic modeling—the author’s name emerges as a key structural element. This is particularly relevant when thinking about the gender distribution of the poets in the PCC and the shifting landscape of those poets over time. For scholars working on eighteenth-century poetry, it is clear that there is a significant increase in the number of female poets represented in anthologies and in the number of edited editions dedicated to a single female author. Anthologies by Roger Lonsdale,44 Paula Backscheider and Catherine Ingrassia,45 and works by literary historians like Janet Todd, Margaret Ezell, Carol Barash, Isobel Armstrong, Virginia Blain, and many more46 have demonstrated how the critical recovery of female poets and the opening of the canon have both challenged and reinforced traditional narratives of eighteenth-century poetry. Celebrating the effects of this critical and historical work on the poetic field, David Shuttleton writes: “critical engagement with eighteenth-­ century women’s poetry has undergone a seismic shift since the last decade of the twentieth century through a project of radical recovery and reassessment largely enacted by feminist literary historians, editors, critics, and biographers.”47 The consequence of this tectonic movement, he writes, is the unmasking of a twentieth-century “chauvinism” that depicted the 44  Roger Lonsdale, Eighteenth-Century Women Poets (Oxford: Oxford University Press, 1989). 45  Paula Backscheider and Catherine Ingrassia, eds. British Women Poets of the Long Eighteenth Century (Baltimore: Johns Hopkins University Press, 2009). 46  Janet M.  Todd, Feminist Literary History (Cambridge: Polity Press, 1988); Margaret J. M. Ezell, Writing Women’s Literary History (Baltimore: Johns Hopkins University Press, 1993); Carol Barash, English Women’s Poetry, 1649–1714 (Oxford: Clarendon Press, 1996); Isobel Armstrong, Victorian Poetry: Poetry, Poetics, and Politics (London: Routledge, 1993); Isobel Armstrong and Virginia Blain, eds., Women’s Poetry in the Enlightenment: The Making of a Canon, 1730–1820 (New York: St. Martin’s Press, 1999). 47  David Shuttleton, “Poetry,” in The Cambridge Companion to Women’s Writing in Britain, 1660–1789, ed. Catherine Ingrassia (Cambridge: Cambridge University Press, 2015), 103.

184 

B. HALL

Fig. 5.10  Topics 4 (sexuality) and 8 (gender) (1967–2018)

canon of eighteenth-century poetry as “entirely male: patrician, rigidly neoclassical, dominated by satire until the onset of a proto-romantic searching for sublimity.”48 However, this narrative of seismic shift is complicated by the trends shown by topics 4 and 8 in Fig. 5.10—both topics strongly marked by their association with discursive language connected to eroticism, sexuality, and gender studies. Topic 8 (gender) seems to rise in frequency during the twenty years from 1990 to 2010, before dropping off significantly as a topic associated with current eighteenth-century poetry criticism. Topic 4 (sexuality) rises in the late 2000s as if to replace the earlier discourse of topic 4, but it too falls off in frequency in recent years. So, what happened to the seismic shift? How does the picture change when we add authors back into the mix? As a baseline metric to assess the distribution of poets by gender in the PCC, I calculated the yearly relative frequency for every male poet and  Ibid.

48

5  TEXT MINING AND DATA VISUALIZATION 

185

Poet's Gender Per Year Poetry Bigrams '69-'79

'80-'90

9'0-9'9

'00-'09

'10-'18

Mean Relative Frequencey per Year Grouped by Gender

0.0004 gender female male

0.0003

0.0002

0.0001

0.0000 1970

1980

1990

2000

2010

2020

Year

Fig. 5.11  Graph charting the relative frequency of a poet’s name for each year, as it appears in the PCC. The green line is the average of the relative frequency of female poets and the orange line the relative frequency of male poets

every female poet name and then took the median relative frequency for each gender group. The orange line in Fig.  5.11 is the average relative frequency of male poets’ names, and the green line is the average relative frequency of female poets’ names. Rather than a seismic shift in the frequency of female poets included in the corpus, there appear to be two significant spikes in the late 1990s. However, the trend up does not continue and, in fact, wanes again until the early 2010s, at which point it proceeds in fits and starts rather than permanently changing the landscape of eighteenth-century poetry criticism. More problematically, for all fifty-­ one years of poetry criticism in these journals, the relative frequency of male poets is higher than that of women poets. Thus, rather than confirm

186 

B. HALL

or challenge Shuttelton’s narrative of change,49 this graph illustrates how computational methods and quantitative analysis complicate the modeler’s relationship to the data and the knowledge-base that is built up for and in that data. Given the importance of the author’s name, both conceptually and historically, the way in which this textual structure organizes poetry criticism shifts attention to a practical question: when do specific poets matter to poetry critics? Identifying patterns in attention to different poems and/or poets might allow for legitimate inferences about disciplinary focus and direction of disciplinary practice, as well as unpacking some of the tension between the structural topic model’s representation of a downward trend in the intersection of poetry and feminist discourse and the perception of significant changes in eighteenth-century poetry criticism. The first step to exploring what those poets’ critics have written about was to decide on a list of poets to look for in these journals. As a starting point to delimiting the field of potential eighteenth-century poets to include in this list, I turned to the table of contents for the third edition of David Fairer and Christine Gerard’s Eighteenth-Century Poetry: An Annotated Anthology.50 While acknowledging the editorial balancing act of representativeness, familiarity, and introduction of new poems and poets, as well as of a variety of genres and themes, I used all forty-four poets (twenty-nine male and fifteen female) included in this anthology as my base list, adding to it only John Dryden. Working with this list of poets and the poetry articles subset extracted from ECS and TEC, I mapped who matters to eighteenth-­ century poetry criticism and when. I hypothesized that exploring this list of anthologized authors in the two journals should provide a way for thinking about trends in the field and highlight the significance of these poets in the landscape of eighteenth-century poetry criticism.  Ibid.  David Fairer and Christine Gerard, eds., Eighteenth-Century Poetry: An Annotated Anthology, 3rd ed. (West Sussex: Wiley Blackwell, 2015). The Blackwell anthology is currently the only major non-gender-specific volume dedicated to eighteenth-century poetry, which makes it an important locus for a general picture of what poets and poems matter at the moment. Moreover, it cut a middle ground between anthologies dedicated to eighteenth-century poetry written by women and earlier anthologies that were by default dedicated to male eighteenth-century poets. See, for instance, Louis I. Bredvold, Robert K. Root, and George Sherburn, eds., Eighteenth-Century Poetry and Prose (New York: Ronald Press, 1932) and Geoffrey Tillotson, Paul Fussell, and Marshall Waingrow, eds., Eighteenth-Century English Literature (New York: Harcourt, Brace, & World, 1969), which combined have only three female poets versus over one hundred male poets. 49 50

5  TEXT MINING AND DATA VISUALIZATION 

187

The problem with consolidating the relative frequency of these poets into a single line by gender, as seen in Fig. 5.11, is that this picture obscures what individual poets are referenced and with what frequency. Moreover, such a visualization does not show when individual poets come into and fall out of critical orbit. While the hairball of lines in Fig. 5.12b are difficult to disentangle, their messiness actually belies a general pattern of appearance and frequency that is in concert with the narrative provided by the poetry criticism seen in the gender distribution graph (Fig. 5.11). Figure 5.12a illustrates the increasing presence and number of female poets in the two journals after 1990. Prior to 1990, only four poets appear in the poetry subset of essays: Anna Seward (red), Lady Mary Wortley Montagu (blue), Ann Yearsley (orange), and Anne Finch (green). After 1990, the number of female poets in these journals increases to include all the fifteen poets in Fairer and Gerard’s anthology. Without attempting to assign causality to such a significant shift in the number and frequency of female poets, these two key publications do seem to capture the critical zeitgeist of eighteenth-century studies in the late 1980s. Felicity Nussbaum and Laura Brown’s The New Eighteenth Century: Theory, Politics, English Literature, published in 1987, certainly staked a claim to a new ground for eighteenth-century studies where new historical and theoretical approaches to the period took root. In addition to Nussbaum and Brown’s volume of essays, Roger Lonsdale’s 1989 edition, Eighteenth-Century Women Poets, dilated the newness of the “new” eighteenth century by intensifying the recovery efforts of lost female voices. In contrast, Fig. 5.12b visualizing male poets discussed in poetry criticism is much less legible. The early decades of both journals (1967–1989) are dominated by male poets, in particular Alexander Pope (green) and Samuel Johnson (blue); this interest peaks again in the late 1980s. However, after 1990, the picture gets muddled as the frequency of male poets drops and more male poets jostle near the bottom of the graph. The line graph of poets is useful because it indicates an increase in the number of female poets discussed by criticism near the end of the 1980s and a general dampening of the frequency of male poets. Charting the poets that critics have written about seems to confirm what we already know about the canonical transformations made possible by the recovery efforts of feminist literary historiographers. However, the colored lines in both the female and male poets line graphs also suggest the perdurability of a smaller set of poets, in particular Anna Seward, Lady Mary Wortley Montagu, Alexander Pope, Samuel Johnson, and Jonathan Swift. To get a more detailed representation of

188 

B. HALL

Anthologized Female Authors '67-'79

'80-'89

'90-'99

'00-'09

'10-'18

Relative Frequency per Year

0.0015

0.0010

0.0005

0.0000 Poetry Articles: 1967-2018 Anthologized Male Authors: 30 Poets '67-'79

'80-'89

'90-'99

'00-'09

'10-'18

Relative Frequency per Year

0.0015

0.0010

0.0005

0.0000 Poetry Articles: 1967-2018

Fig. 5.12  (a–b) Anthologized male and female authors, 1967–2018. The top graph highlights Anna Seward (red), Lady Mary Wortley Montagu (blue), Ann Yearsley (orange), and Anne Finch (green); all of them appear at least once in the two journals prior to 1990. The bottom highlights the male poets prevalent prior to 1990: Alexander Pope (green), John Dryden (red), Jonathan Swift (purple), and Samuel Johnson (blue)

5  TEXT MINING AND DATA VISUALIZATION 

189

individual references to a poet and explore the shape of this smaller canon of poets, I used a scatterplot of names plotted over the same time frame (1967–2018). Figure  5.13 gives us the list of poets’ names where each panel presents a reference to a poet’s first and last name as a bigram.51 The picture here is much starker than that of the line graphs. Figure 5.13a shows how few references there are to female poets. Except for Lady Mary Wortley Montagu, the sparseness of references is somewhat unsettling: several poets have only one or a handful of citations. Equally troubling is that in the decades after Lonsdale’s anthology, there is no significant uptick in either the number of references to women poets or the relative frequency with which they are cited. Prior to 1990, there are only twelve references to the women in my list of poets, half of which are to Montagu. Of the fifteen anthologized female authors, only Anna Seward, Anne Finch, and Lady Mary Wortley Montagu appear in articles before 1990, while Mary Robinson does not appear in journals until 2009 and 2010, respectively. Martha Fowke, Mary Leapor, and Anna Seward have at least one reference, so a higher relative frequency, but otherwise these poets do not seem to have piqued a significant critical interest. Thus, what appeared to be a surge of women poets turned out to be a relatively few scattered references. This scarcity of women poets is made all the starker by Fig. 5.13b, which plots references to male poets: they almost universally have more prominence in poetry criticism in terms of both frequency of citations and frequency of references over the time span covered by these journals. Most surprising of all, however, is that, despite narratives of decentering the canon, a small canon of poets continue to dominate the critical discussions. The poets with the most references in these two journals and with the highest relative frequency are Lady Mary Wortley Montagu, Alexander Pope, Samuel Johnson, and Jonathan Swift. Moreover, the place of these poets in criticism remains relatively stable over the publication years of these journals. Essays referring to Montague and Pope have a spike in 1989 and 2004 and remain staples in poetry criticism ever after. Samuel Johnson’s case is more interesting in that he has the most frequent 51  A bigram is a sequence of two adjacent elements from a string of tokens; in our case, the authors’ first and second names. Because poets are often referred to only by their last name, I also tested the last names of poets as a unigram, that is, an n-gram consisting of a single item from a sequence. While the relative frequencies went up, the overall map of the poets across time did not.

190 

B. HALL Anthologized Female Authors: 15 ann yearsley

anna seward

anne finch

anne ingram

0.0020 '67-'79

'80-'89

'67-'79

'80-'89

'67-'79

'80-'89

'90-'99

'00-'09

'10-'18

'67-'79

'80-'89

'10-'18

'67-'79

'80-'89

'10-'18

'67-'79

'80-'89

'90-'99

'00-'09

'10-'18

'67-'79

'80-'89

'10-'18

'67-'79

'80-'89

'10-'18

'67-'79

'80-'89

'90-'99

'00-'09

'10-'18

'67-'79

'80-'89

'10-'18

'67-'79

'80-'89

'90-'99

'00-'09

'10-'18

0.0015 0.0010

Relative Frequency per Year

0.0005 0.0000 laetitia barbauld

martha fowke

mary jones

mary leapor

0.0020 '90-'99

'00-'09

'90-'99

'00-'09

'90-'99

'00-'09

'90-'99

'00-'09

'10-'18

0.0015 0.0010 0.0005 0.0000 mary robinson

sarah dixon

wortley montagu

0.0020 '90-'99

'00-'09

'90-'99

'00-'09

'90-'99

'00-'09

'10-'18

0.0015 0.0010 0.0005 0.0000

Poetry Articles: 1967-2018

Anthologized Male Authors

Relative Frequency per Year

0.0020 0.0015 0.0010 0.0005 0.0000 0.0020 0.0015 0.0010 0.0005 0.0000 0.0020 0.0015 0.0010 0.0005 0.0000 0.0020 0.0015 0.0010 0.0005 0.0000 0.0020 0.0015 0.0010 0.0005 0.0000

aaron hill '67-'79 '80-'89'90-'99'00-'09'10-'18

isaac watts '67-'79 '80-'89'90-'99'00-'09'10-'18

john philips '67-'79 '80-'89'90-'99'00-'09'10-'18

oliver goldsmith '67-'79 '80-'89'90-'99'00-'09'10-'18

thomas chatterton '67-'79 '80-'89'90-'99'00-'09'10-'18

alexander pope '67-'79 '80-'89'90-'99'00-'09'10-'18

james macpherson '67-'79 '80-'89'90-'99'00-'09'10-'18

john pomfret '67-'79 '80-'89'90-'99'00-'09'10-'18

richard savage '67-'79 '80-'89'90-'99'00-'09'10-'18

thomas gray '67-'79 '80-'89'90-'99'00-'09'10-'18

ambrose philips '67-'79 '80-'89'90-'99'00-'09'10-'18

james thomson '67-'79 '80-'89'90-'99'00-'09'10-'18

jonathan swift '67-'79 '80-'89'90-'99'00-'09'10-'18

robert burns '67-'79 '80-'89'90-'99'00-'09'10-'18

thomas parnell '67-'79 '80-'89'90-'99'00-'09'10-'18

charles churchill '67-'79 '80-'89'90-'99'00-'09'10-'18

john dryden '67-'79 '80-'89'90-'99'00-'09'10-'18

joseph warton '67-'79 '80-'89'90-'99'00-'09'10-'18

robert lloyd '67-'79 '80-'89'90-'99'00-'09'10-'18

thomas warton '67-'79 '80-'89'90-'99'00-'09'10-'18

christopher smart '67-'79 '80-'89'90-'99'00-'09'10-'18

john dyer '67-'79 '80-'89'90-'99'00-'09'10-'18

mark akenside '67-'79 '80-'89'90-'99'00-'09'10-'18

samuel johnson '67-'79 '80-'89'90-'99'00-'09'10-'18

william collins '67-'79 '80-'89'90-'99'00-'09'10-'18

george crabbe '67-'79 '80-'89'90-'99'00-'09'10-'18

john gay '67-'79 '80-'89'90-'99'00-'09'10-'18

matthew prior '67-'79 '80-'89'90-'99'00-'09'10-'18

stephen duck '67-'79 '80-'89'90-'99'00-'09'10-'18

william cowper '67-'79 '80-'89'90-'99'00-'09'10-'18

Poetry Articles: 1967-2018

Fig. 5.13  (a–b) Scatterplot of the relative frequency of a poet’s name for each year (1967–2018)

5  TEXT MINING AND DATA VISUALIZATION 

191

and significant presence in the years prior to the printing of TEC and in the first decade of that publication’s existence, before falling out of favor. He does, however, return to critical attention in the late 2000s and the second decade of the twenty-first century. In all, the picture here is of a small, largely unchanged canon of poets between the 1970s and today. In general, women poets seem to have appeared in these journals with the same or less frequency than their “minor” male counterparts and, with the exception of Lady Mary Wortley Montagu, there are considerably fewer instances of female poet names than male, which seems to suggest that far fewer articles were either focused on a female poet or refer to a female poet anthologized by Fairer and Gerard. Far from announcing the end of the male canon, the graph shows a striking perdurability of male poets. Not only do male poets appear in more articles and with greater frequency than their female counterparts in the early years of the two journals, but they continue to do so up to 2018. In other words, despite the rhetoric of “seismic” change, the downward turn of topics 4 and 8, coupled with the frustrating disparity in the representation of male and female poets, indicates a persistent interest of eighteenth-century poetry criticism in exploring poetry written by men rather than women.

Conclusion In concluding this chapter, I would like to highlight some of the complications and opportunities for scholars who work in the traditional lanes of literary studies or seek to branch out into the quantitative methods of text mining and data visualization. This cultural analysis of eighteenth-century poetry criticism over the past fifty years poses as many questions as it offers arguments and interpretations. Why are the number of articles dedicated to poetry declining while the number of articles published in both ECS and TEC is increasing? The visualizations of the model’s topics over time convincingly show that eighteenth-century poetry criticism is entering a new phase. The question is, what will we hold on to from the “new eighteenth century” and in what ways will we leave it behind? How can the field address an apparent disjunction between the perception of what is happening in the criticism and literary history of women poets of the eighteenth century and what is happening in these journals themselves? Authors and methodology seem to take precedence over poetic form in the discursive landscape, which points back to Keith and Sitter’s point

192 

B. HALL

about the need to address poetry on its own terms and poetic form as an important part of poetry criticism. Yet, it seems that such an effort has yet to be addressed and is currently overlooked as a site for important work. This project is also meant to offer eighteenth-century scholars an exploratory model that deploys machine clustering methods and data visualizations as tools for both explication and analysis. This exploration has highlighted how the effort to excavate the “tools, techniques and practices” that shape eighteenth-century poetry criticism often entails a recursive process of organizing a corpus, algorithmic analysis, and data visualization. As exploratory and descriptive methods, text mining and data visualization might resist the modeler’s impulse to draw wide-ranging conclusions. However, they also offer an opportunity to reflect not only on the object of our study but also on the presuppositions the modeler brings to the iterative processes involved in building and using models. In McCarty’s words, computational modeling “effects a sea-change by forcing us to confront the radical difference between what we know and what we can specify computationally, leading to the epistemological question of how we know what we know.”52 Computational modeling and data visualization fuel this recursive process of metacognition, of reflecting on how we know what we know, at an important moment in eighteenth-century studies. Such epistemological inquiry means widening the methodological territory we have grown accustomed to in ways commiserate with the expansion of the field of inquiry made possible by digitization. Forging a “new” eighteenth century for poetry criticism and beyond will require revisions to the models we have fashioned and used in the past. Quantitative analytics and data visualization offer exciting sets of tools for pressing once again on the boundaries of what can be done in the field of eighteenth-­ century studies.

Bibliography Allahyari, Mehdi, Seyedamin Pouriyeh, Mehdi Assefi, Saied Safaei, Elizabeth D. Trippe, Juan B. Gutierrez, and Krys Kochut. 2017. A Brief Survey of Text Mining: Classification, Clustering and Extraction Techniques. arXiv e-print. arXiv:1707.02919. Armstrong, Isobel. 1993. Victorian Poetry: Poetry, Poetics, and Politics. London: Routledge.  McCarty, Humanities Computing, 25.

52

5  TEXT MINING AND DATA VISUALIZATION 

193

Armstrong, Isobel, and Virginia Blain, eds. 1999. Women’s Poetry in the Enlightenment: The Making of a Canon, 1730–1820. New  York: St. Martin’s Press. Backscheider, Paula, and Catherine Ingrassia, eds. 2009. British Women Poets of the Long Eighteenth Century. Baltimore: Johns Hopkins University Press. Barash, Carol. 1996. English Women’s Poetry, 1649–1714. Oxford: Clarendon Press. Berry, Michael W. 2004. Survey of Text Mining: Clustering, Classification, and Retrieval. New York: Springer. Blei, David M. 2012a. Probabilistic Topic Models. Communications of the ACM 55 (4): 77–84. https://doi.org/10.1145/2133806.2133826. ———. 2012b. Topic Modeling and Digital Humanities. Journal of Digital Humanities 2 (1): n.p. http://journalofdigitalhumanities.org/2-1/ topic-modeling-and-digital-humanities-by-david-m-blei/ Bredvold, Louis I., Robert K. Root, and George Sherburn, eds. 1932. EighteenthCentury Poetry and Prose. New York: Ronald Press. Chang, Jonathan, Sean Gerrish, Chong Wang, Jordan L. Boyd-Graber, and David M. Blei. 2009. Reading Tea Leaves: How Humans Interpret Topic Models. In Advances in Neural Information Processing Systems, ed. Y.  Bengio, D.  Schuurmans, J.D.  Lafferty, C.K.I.  Williams, and A.  Cullota, vol. 22, 288–296. Vancouver: Curran Associates. Ciula, Arianna, and Oyvind Eide. 2017. Modelling in the Digital Humanities: Signs in Context. Digital Scholarship in the Humanities 32 (1): i33–i46. https://doi.org/10.1093/llc/fqw045. Edelstein, Dan. 2014. Enlightenment Scholarship by the Numbers: dfr.jstor.org, Dirty Quantification, and the Future of the Lit Review. Republics of Letters 4 (1): 1–26. https://arcade.stanford.edu/sites/default/files/article_pdfs/ ROFL_v5_Edelstein_final.pdf Ezell, Margaret J.M. 1993. Writing Women’s Literary History. Baltimore: Johns Hopkins University Press. Fairer, David, and Christine Gerard, eds. 2015. Eighteenth-Century Poetry: An Annotated Anthology. 3rd ed. West Sussex: Wiley Blackwell. Flanders, Julia, and Fotis Jannidis, eds. 2018. The Shape of Data in Digital Humanities: Modeling Texts and Text-Based Resources. London: Routledge. ———. Data Modeling. In Schreibman, Siemens, and Unsworth, A New Companion to Digital Humanities, 229–237. Gavin, Michael. 2016. Historical Text Networks: The Sociology of Early English Criticism. Eighteenth-Century Studies 50 (1): 53–80. https://doi. org/10.1353/ecs.2016.0041. Goldstone, Andrew, and William E. Underwood. 2014. The Quiet Transformations of Literary Studies: What Thirteen Thousand Scholars Could Tell Us. New Literary History 45 (3): 359–384. https://doi.org/10.1353/nlh.2014.0025.

194 

B. HALL

Graff, Gerald. 1987. Professing Literature: An Institutional History. Chicago: University of Chicago Press. Jockers, Matthew L. 2013. Macroanalysis: Digital Methods & Literary History. Urbana: University of Illinois Press. ———. 2014. Text Analysis with R for Students of Literature. New York: Springer. Jockers, Matthew L., and Ted Underwood. Text-Mining the Humanities. In Schreibman, Siemens, and Unsworth, A New Companion to Digital Humanities, 291–306. Keith, Jennifer. 2007. Why Poetry? The Eighteenth Century 48 (1): 87–91. https://doi.org/10.1353/ecy.2007.0002. Kwartler, Ted. 2017. Text Mining in Practice with R. Hoboken: Wiley. Lonsdale, Roger. 1989. Eighteenth-Century Women Poets. Oxford: Oxford University Press. Lydia, E. Laxmi, P. Govindasamy, S.K. Lakshmanaprabu, and D. Ramya. 2018. Document Clustering Based on Text Mining K-Means Algorithm Using Euclidean Distance Similarity. Journal of Advanced Research in Dynamical and Control Systems 10 (2): 208–214. McCarty, Willard. 2005. Humanities Computing. New York: Palgrave Macmillan. Nussbaum, Felicity, and Laura Brown, eds. 1987. The New Eighteenth Century: Theory, Politics, English Literature. North Yorkshire: Methuen. Ohmann, Richard. 1987. Politics of Letters. Middletown: Wesleyan University Press. Piper, Andrew. 2017. Think Small: On Literary Modelling. PMLA 132 (3): 651–658. https://doi.org/10.1632/pmla.2017.132.3.651. ———. 2018. Enumerations: Data and Literary Study. Chicago: University of Chicago Press. Ramsey, Stephen. 2003. Special Section: Reconceiving Text Analysis: Toward an Algorithmic Criticism. Literary and Linguistic Computing 18 (2): 167–174. https://doi.org/10.1093/llc/18.2.167. Roberts, Margaret E., Brandon M. Stewart, and Dusting Tingley. 2019. stm: An R Package for Structural Topic Models. Journal of Statistical Software 91 (2): 1–40. https://doi.org/10.18637/jss.v091.i02. Schmidt, Benjamin M.  Modeling Time. In Flanders and Jannidis, The Shape of Data in Digital Humanities, 150–166. Schreibman, Susan, Raymond G. Siemens, and John Unsworth, eds. 2016. A New Companion to Digital Humanities. Malden: Wiley/Blackwell. Shuttleton, David. 2015. Poetry. In The Cambridge Companion to Women’s Writing in Britain, 1660–1789, ed. Catherine Ingrassia, 103–117. Cambridge: Cambridge University Press. Sievert, Carson, and Kenneth E. Shirley. 2014. LDAvis: A Method for Visualizing and Interpreting Topics. In Proceedings of the Workshop on Interactive Language Learning, Visualization, and Interfaces, ed. Jason Chang, Spence Green, Marti Hearst, Jeffrey Heer, and Philipp Koehn, 63–70. Baltimore: Association for

5  TEXT MINING AND DATA VISUALIZATION 

195

Computational Linguistics. https://nlp.stanford.edu/events/illvi2014/ papers/sievert-illvi2014.pdf. Silge, Julia, and David Robinson. 2017. Text Mining with R: A Tidy Approach. Sebastopol: O’Reilly. Simsion, Graeme. 2007. Data Modeling: Theory and Practice. Bradley Beach: Technics Publications. Sitter, John. 2011. The Cambridge Introduction to Eighteenth-Century Poetry. Cambridge: Cambridge University Press. Srivastava, Ashok N., and Mehran Sahami, eds. 2009. Text Mining: Classification, Clustering, and Applications. Boca Raton: CRC Press. Tillotson, Geoffrey, Paul Fussell, and Marshall Waingrow, eds. 1969. EighteenthCentury English Literature. New York: Harcourt, Brace, & World. Todd, Janet M. 1988. Feminist Literary History. Cambridge: Polity Press. Underwood, Ted. Algorithmic Modeling: Or, Modeling Data We Do Not Yet Understand. In Flanders and Jannidis, The Shape of Data in Digital Humanities, 250–263. Welbers, Kasper, Wouter Van Atteveldt, and Kenneth Benoit. 2017. Text Analysis in R. Communication Methods and Measures 11 (4): 245–265. https://doi. org/10.1080/19312458.2017.1387238. Wickham, Hadley, and Garrett Grolemund. 2017. R for Data Science: Import, Tidy, Transform, Visualize, and Model Data. Sebastopol: O’Reily. Zhang, Yin, Rong Jin, and Zhi-Hua Zhou. 2010. Understanding Bag-of-Words Model: A Statistical Framework. International Journal of Machine Learning and Cybernetics 1 (1–4): 43–52. https://doi.org/10.1007/ s13042-010-0001-0.

PART II

Data Visualization and the Eighteenth-Century Corpus: Case Studies

CHAPTER 6

The Grid and the Visualization of Abstract Information: Three Eighteenth-Century Models Jakub Zdebik

The grid organizes abstract, imperceptible forces visually. Despite its rigid geometry, it has an incontrovertible link to abstract thought. It is no wonder, then, that the grid is the underlying structure operative in data visualization: it helps visualize abstract information and offers it to scientific scrutiny. In what follows, I will survey the use of the grid in three eighteenth-­century visualizations of information and analyze them from philosophical and aesthetic perspectives. I will start by exploring how, in Jacques Barbeu-Dubourg’s Machine chronologique (1753), the grid is harnessed to organize time. Then, I will argue that the geometric, repetitive pattern coordinating economic exchange in François Quesnay’s Tableau économique (1758) stands in for the notion of relationality. Finally, through Linnaeus’s illustrations in Systema Naturae (1768), which display objects arranged according to a grid pattern on a comparative table, I will consider the notion of organization associated with the grid. The grid’s

J. Zdebik (*) University of Ottawa, Ottawa, ON, Canada e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_6

199

200 

J. ZDEBIK

relationship to time, organization, and visibility is then framed according to Gilles Deleuze’s concept of the diagram; a concept that negotiates among knowledge, abstraction, and display. Finally, in order to bridge philosophy and graphic design, I will rely on texts that elaborate on the notion of the grid in visualizations from a graphic perspective. I will analyze the relationship between scientific annotation and aesthetics through Michel Foucault’s description of the Linnaean organizational grid as a table, tableau, and portrait. I will then determine what happens to the grid apparatus in contemporary data visualization and how it is used in visual art closer to us by transposing the graphic patterns surveyed in the first part of this chapter onto three examples of contemporary art that mirror the patterns of the eighteenth-century graphic works. Thus, this chapter will apply an art historical and philosophical methodology to scientific notations to demonstrate the lasting impact of eighteenth-century data visualization on contemporary aesthetics.

Grid as Visual Theory The grid is an important philosophical concept for art historians who have extensively studied its role in conveying knowledge. Erwin Panofsky, for example, in Perspective as Symbolic Form, states that the grid pattern of a tiled floor in Renaissance art precedes abstract mathematic thought by displaying the rigor of a coordinate system.1 In doing so, he cements the relationship between scientific discovery and visual representation in the art field. Likewise, art historians Rosalind Krauss, Hubert Damisch, and Christine Buci-Glucksmann focus on the time and space the grid occupies in the production of art. Krauss demonstrates the importance of the grid as an “optical unconscious” in the structure of abstract painting, Damisch shows how the grid is part of the timeline of art’s picture plane coming into being, and Buci-Glucksmann takes a diagrammatic view of the grid as the basic condition of the image as a communication device.2 All these thinkers theorize the grid as a visualization of thought.

 Erwin Panofsky, Perspective as Symbolic Form (New York: Zone Books, 1991), 58.  See Walter Benjamin, “Little History of Photography,” in Walter Benjamin: Selected Writings. Volume 2: 1927–1934, trans. Rodney Livingstone, ed. Michael W.  Jennings, Howard Eiland, and Gary Smith (Cambridge: Belknap Press of Harvard University Press, 1999), 521. 1 2

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

201

Krauss emphasizes the importance of the grid pattern in avant-garde abstract art. In her seminal essay, “Grids,” published in 1979, she marvels at how the grid is a rich source of creation despite its seeming rigidity and muteness.3 Krauss highlights the significance of this visual device by demonstrating how its source is grounded in the past. However, the grid is not traced back, as expected, to the patterns that covered the canvas of the Renaissance art and its illusionistic, three-dimensional painterly spaces. Rather, the precursors of the grid that becomes an emblem of the abstract, minimalist, and conceptual art of the twentieth century are the organizational grids that served as explanatory graphics of scientific data in the treatises on optics of the nineteenth century. Krauss also reads into the grid’s temporal and spatial components. On the one hand, the grid is a temporal marker: it announces a twentieth-century aesthetic in a nineteenth-­century source. On the other hand, the grid is spatially self-­ reflexive: it organizes the surface of the canvas and its spatial limits. The temporal and the spatial facets of this visual device are summarized by Krauss in the two opposite movements of the grid: a centripetal force that folds the grid onto itself in a self-reflexive trajectory, and a centrifugal force that pushes it outside the boundaries of the grid’s delimited space toward a greater, totalizing connectivity. Thus, for Krauss, the grid is a diagrammatic vehicle that shuttles the function of data organization from nineteenth-century scientific treatises, especially in the field of optics, to the avant-garde paintings of the twentieth century. This is possible because, through its liminal tautness, the grid reveals the optical self-reflexivity of visual expression. In her exploration of various modes of visualization, from scientific annotation to artistic expression, Krauss makes an unexpected discovery: the seemingly rigid grid possesses a spiritual and symbolic layer that makes it relevant in different historical and cultural contexts. Significantly, the structure that Krauss demonstrates is informed by Claude Levi-Strauss, who argues that the substrata of the grid is, surprisingly, spiritual (a seemingly problematic concept for contemporary art).4 If Krauss locates the layout for an artistic grid in the scientific tracts of the nineteenth century, Damisch goes further back in time. Damisch sees

3   Rosalind Krauss, stable/778321 4  Ibid., 55.

“Grids,”

October

9

(1979):

50–64,

http://www.jstor.org/

202 

J. ZDEBIK

in the grid a grounding structure for both art and history.5 The grid also lays out the rules for society (e.g., it dictates the rules of engagement in war by circumscribing spatial limits) and geography (e.g., it is a useful tool for colonialist expansion). From a philosophical perspective, Damisch follows Ludwig Wittgenstein’s pronouncement that “[t]he form is the possibility of the structure”6 and similarly poses that “a grid … is not a structure, but the possibility of it.”7 In order to find the link between form and structure, Damisch attempts to establish a genealogy of the grid. This proposed genealogy has roots in the ancient divination, which established rules for organizing a space in a grid-patterned way.8 According to the art historian, it is through divination that we have an initial understanding of time as a projection toward the future. The grid pattern, then, is a realization of time’s sequentiality. Moreover, for Damisch, the grid carries a sense of history: the chess board, for example, originated in India 3000 years prior to the invention of chess in the eight century, traveled to Europe via the Middle East and, when it landed there, its organizing function was reconfigured to fit into the scale of society itself. Thus, the initial small scale of the grid pattern changes context. Its pure functionality is abstracted from its original object. And, abstract in this way, the grid pattern is applied to larger systems. This is how, according to Damisch, we end up with the delimited space of the grid as both the adversarial terrain of battles and wars and the foundational model of empire building.9 The grid model for war or colonial expansion represents time and creates a “sense of history” within a “delimited space.”10

5  Hubert Damisch, “Genealogy of the Grid,” in The Archive of Development, ed. Henk Slager and Annette W. Balkema (Amsterdam: Rodopi, 1998), 49–53. 6  Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. C.  G. Ogden (Sweden: Chiron Academic Press, 2016), 28. 7  Damisch, “Genealogy of the Grid,” 51. 8  Ibid. According to Damisch, “a genealogy … starts with the procedures of divination: the way the ancient peoples and among them the Greeks played with pebbles and cards, and how they distributed them in a given space turned into a more or less regular scheme that led to the use of grid patterns.” 9  Ibid., 54. As Damisch explains, “[t]he chessboard existed in India for 3,000 years, long before the same Indians invented the game of chess in the eighth century A.D. It was known in Europe early in the Middle Ages, when the board was first introduced in the West. It took a certain time before Europeans were taught the rules of chess by the Arabs, but the game soon provided a model for the functioning of society and history.” 10  Ibid.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

203

From an aesthetic perspective, this statement offers a way of seeing the grid’s functions diagrammatically. According to Christine Buci-­ Glucksmann and Josh Wise, the diagram is an organizing grid: “To be certain, the diagram as ‘relationship of forces’ implies an abstract machine that grids the social and engenders an ‘intersocial in the making.’”11 The term “abstract machine” is borrowed from Gilles Deleuze and Félix Guattari and is used to conceptualize the material consequences of artistic functions. The authors equate “diagrammatic activity” with the spatial organization by the grid of visual materials, in this case, painting.12 Deleuze also identifies the diagrammatic function operative in painting, where atomized traits are able to lift off from a visual configuration—say a bird that a painter like Francis Bacon intended to paint—and land into a wholly heterogeneous visual expression, such as an umbrella sheltering a grinning Mussolini flanked by sides of beef. But where does the specific relationship between the diagrammatic function and the grid come from? Deleuze’s broader understanding of the diagram as a redistribution of pure function from one system to another is seemingly inspired by Michel Foucault’s notions of function and system. Indeed, Deleuze’s view of the diagram comes from Foucault’s description of Jeremy Bentham’s Panopticon, a prison structure with a regular grid of cells surrounding a central tower. The guard in the tower, unseen by the prisoners, can observe each cell. He is aided by the fact that the cells are traversed by light. Since they cannot see the guard, the prisoners assume that they are being watched at all times. What is important here, according to Foucault, is not the concrete material that makes up the architecture of the prison but the network of seeing enabled by Bentham’s design, or the visual function of the prison. This visual function is what makes the Panopticon system possible.13 And this function, abstracted 11  Christine Buci-Glucksmann and Josh Wise, “Of the Diagram in Art,” ANY: Architecture in New York, no. 23 (1998): 34–36, https://www.jstor.org/stable/i40087754 12  Ibid., 35. As Buci-Glucksmann and Wise further point out, “we find a diagrammatic activity in the work of a painter such as Vermeer, whose grids were made by crossing strings attached to fixed points on a horizon line.” 13  Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York: Vintage Books, 1995), 205. Foucault cautions, however, that “the Panopticon must not be understood as a dream building: it is the diagram of a mechanism of power reduced to its ideal form; its functioning, abstracted from any obstacle, resistance or friction, must be represented as a pure architectural and optical system: it is in fact a figure of political technology that may and must be detached from any specific use.”

204 

J. ZDEBIK

from its material scaffolding, can be transposed onto other systems, such as a school setting where the teachers see all the students from the front of the class, or a military barracks that the drill sergeant can assess at a glance. The example of the military system is interesting here because it brings the diagram back to the historical grid Damisch is discussing. Foucault explains that “the old, traditional square plan was considerably refined in innumerable new projects,”14 which sounds similar to the diagrammatic transfer of pure function from one system to another. After describing the strict measurements that the standard military camp adheres to, Foucault concludes that the function of the camp is one of visibility: “[t]he camp is the diagram of a power that acts by means of general visibility. For a long time this model of the camp or at least its underlying principle was found in urban development, in the construction of working-class housing estates, hospitals, asylums, prisons, schools: the spatial ‘nesting’ of hierarchized surveillance.”15 In this passage, we clearly see how he moves from the camp to various institutions whose primary function is not one of physical or mental healing, public safety, or learning, but rather of surveillance. The same grid structure is present in each system, having the same function in different contexts, whether it appears in optical treatises or in twentieth-­century art. In this sense, Damisch’s grid moves from a checkerboard with no chess to the chess game in India to the Arab world to Europe to finally become the matrix for conflict in the system of war and colonial expansion through its function of surveillance and construction. In Buci-Glucksmann’s argument, the grid as the baseline of the diagram is but a visual manifestation of the limits of thought. Her view of the relationship between the grid and the diagram is reflected in John Bender and Michael Marrinan’s reading of the perspectival grid as “a preferred model for presenting visual information.”16 The grid as perspective is a diagram. Visual catalogs are organized as grids and tableaus because a gridded informational pattern is there to “promote a humanist subject” through the visual organization of information.17 What I see here is the grid as the visual manifestation of a diagrammatic distribution of abstract  Ibid., 171.  Ibid., 171–72. 16  John Bender and Michael Marrinan, The Culture of Diagram (Stanford, CA: Stanford University Press, 2010), 34. 17  Ibid. 14 15

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

205

functions. In effect, the grid facilitates the transfer of thought into different contexts. Since the grid is both abstract and representational, it displays its own reflexivity as to its spatial and temporal transferability. Similar to art historians, Howard Wainer plots the grid’s position in the evolution of graphic design. In his opinion, “mechanical pieces necessary for data-based graphics” started to emerge in the eighteenth century, being structured around permutations of the grid pattern.18 The grid’s historical trajectory starts with vertical and horizontal coordinate systems. These systems were used in the past in geographical surveys, astronomy, and urban planning.19 From urban planning to chess board to medieval musical notation, the grid changes from utilitarian to more abstract forms. Such links between practical and intellectual can be seen in the seventeenth century as well, when the grid became standardized in graph paper and René Descartes developed his “coordinate system.”20 The relationship between Descartes’ coordinate grid and visual graphics is rendered clear in the way the art world adopts a geometrical spatial organization in its aesthetics.21 Thus, Wainer pinpoints the material and intellectual articulation of the grid. In addition to the historical adoption of the grid, Wainer provides two salient examples illustrating how this device renders data collection into coherent and readable graphics. He shows how early circular patterns plotting astronomical data were simplified and made readable by turning the circle into a grid,22 and how weather data is currently filtered 18  Howard Wainer, Graphic Discovery: A Trout in the Milk and Other Visual Adventures (Princeton, NJ: Princeton University Press, 2006), 10. 19  Wainer explains that a grid system was used by the surveyors of the Nile’s flood basin in 1400  BCE and that an intersection of horizontal and vertical coordinates was used by Hipparchus in astronomy around 400 BCE. He also charts the grid’s use in Roman urban planning, in chessboard in India, and in the Medieval musical notation. Ibid., 10. 20  Ibid., 11. In 1637, Descartes developed the coordinate system, which, as Wainer points out, was “an important intellectual milestone in the path toward statistical graphics.” In 1680, gridded graph paper was standardized. 21  See Matthew Wickman, Literature after Euclid: The Geometric Imagination in the Long Scottish Enlightenment (Philadelphia: University of Pennsylvania Press, 2015), 5. According to Wickman, “for Descartes and his followers, this meant creating a coordinate system that affixed numerical or algebraic values to, for example, a given point along a plane, making it easier to calculate values; for avant-garde artists, it meant incorporating postclassical theories about the shape of the universe and the nature of reality into the visual and narrative arts, changing ideas about the nature of truth and the realm of the possible.” 22  Wainer, Graphic Discovery, 11–12. Wainer explains that Pliny plotted astronomical data in a circular form corresponding to the locations of the celestial bodies. However, circular graphs were hard to use since one had to always go around the circle to read the data. The

206 

J. ZDEBIK

out through the grid into analyzable information, turning the invisible into visible. In Wainer’s account, the grid progresses geographically through history, reconfiguring its organizing function toward more and more abstract systems.23 The grid, then, tempers the complexity of information by abstracting data into visualizations. Not all historians of graphic display are enthusiastic about the grid, however. Edward Tufte, for example, considers the effect of the grid on graphics as visual noise.24 In fact, he believes that, in graphic design, the grid should be muted to let the evidence shine forth. Nevertheless, the usefulness of the grid in the display of evidence is undeniable. Hannah B. Higgins explains that, in the fifteenth century, the grid was a physical and material device used in the office of the exchequer to organize the space into a game board of sorts, where taxes could be tabulated visually.25 What is important in these opposite views is the preeminence of the grid as an organizing principle of information in various visual registers. The grid is not always immediately apparent, but its organizational features can be made visible even if the line that guides the information is not always evident. In what follows, I will discuss three types of grid with varying degrees of visibility in different disciplines of the eighteenth century. I believe that these types of grid can demonstrate the flexibility of the device when applied to various disciplines and its capacity to represent in visual form specific types of knowledge.

grid, on the other hand, provided a beginning and an end to the circular pattern, thus making data readable. 23  The grid was responsible for translating weather-related data into forecasting information, as Robert Plot, the originator of this visual device wrote: “We shall then be enabled with some grounds to examine, not only the coastings, breadth, and bounds of the winds themselves, but of the weather they bring with them; and probably in time thereby learn, to be forewarned certainly, of divers emergencies (such as heats, colds, dearths, plague, and other epidemical distempers) which are not unaccountable to us.” In other words, the grid helps make something intangible visible. Quoted in Wainer, Graphic Discovery, 15. 24  Edward R. Tufte, Visual Explanations: Images and Quantities, Evidence and Narrative (Cheshire: Graphics Press, 1997), 105. 25  Hannah B. Higgins, The Grid Book (Cambridge: MIT Press, 2009), 128.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

207

Barbeu-Dubourg and the Grid as Time Jacques Barbeu-Dubourg (1709–1779) was a Parisian physician who belonged to several medical societies in France, Great Britain, and the United States and was also a friend of Benjamin Franklin, Jean-Jacques Rousseau, and Denis Diderot.26 But it was not his medical background that was at the root of his invention of the gridded graphic method of capturing history that he called machine chronologique (1753). Most likely, it was his network of friends that was instrumental in his graphic endeavor: Rousseau’s view of history as a system of paradoxes,27 Diderot’s description of the “chronological machine” in his Encyclopédie,28 and Franklin’s role as a mediator between Barbeu-Dubourg and Joseph Priestley, the author of the Chart of Biography, which contains a graphic display of great historical names.29 Barbeu-Dubourg used the grid to capture the progress of history. The idea behind his temporal graphic display was to create a geographic visualization which he thought to be more stimulating than a non-visual historical narrative. Barbeu-Dubourg’s machine chronologique is a scroll of paper encased in a device that, with the help of two handles, allows the user to peruse the paper once the casing is opened (Fig. 6.1). The paper roll offers an overview of history: it is divided by longitudinal and horizontal timelines, providing information that allows for a comparison of events at particular moments in time. In his study of this device, Stephen Ferguson 26  See Stephen Ferguson, “The 1753 Carte chronographique of Jacques Barbeu-Dubourg,” The Princeton University Library Chronicle 52, no. 2 (1991): 190, https://doi. org/10.2307/26404421; and Alfred Owen Aldridge, “Jacques Barbeu-Dubourg, a French Disciple of Benjamin Franklin,” Proceedings of the American Philosophical Society 95, no. 4 (1951): 331, http://www.jstor.org/stable/3143280 27  Stephen G. Salkever, “Interpreting Rousseau’s Paradoxes,” Eighteenth-Century Studies 11, no. 2 (1977–1978): 204–26, https://doi.org/10.2307/2738384 28  Ferguson, “The 1753 Carte chronographique,” 190. 29  Boyd Davis explains how Franklin shuttled the notion of graphic history between two authors: “The two inventors shared a connection: Benjamin Franklin was a mutual friend and sent Barbeu-Dubourg a copy of Priestley’s timeline, to which Barbeu-Dubourg responded with the gracious remark, ‘I have received with gratitude and viewed with pleasure the biographical chart of Mr. Priestley which is in truth make on almost the same principle as my own, without plagiarism on either side, as I in no way claim on account of the date.’” Barbeu-Dubourg may have known from Franklin that Priestley had been making charts for some time to assist in his teaching at the Warrington Dissenting Academy before he began to publish them. See Stephen Boyd Davis, “History on the Line: Time as Dimension,” Design Issues 28, no. 4 (2012): 4, note 6, http://doi.org//10.116/DESI_A_00171

208 

J. ZDEBIK

Fig. 6.1  Jacques Barbeu-Dubourg’s Machine chronologique, 1753. (Courtesy of Princeton University Library)

explains that Barbeu-Dubourg links space and time to history in visual terms: in other words, geography and chronology are meant to be in his study the “eyes” of history.30 Aesthetically speaking, studying history through geography renders it “lively, convenient, attractive,” whereas the chronological aspect of history “is dry, [and] unprofitable” because it overburdens memory with “repulsive dates” and “a prodigious multitude of numbers.”31 Ferguson surmises that Barbeu-Dubourg’s intention was to put chronology on par with geography, to render it more accessible, and that this prompted him to reach for the visual device of the grid: “Setting out human events on the time scale is identical in his mind to mapping the earth against a rational, formal, and impersonal grid, as a

 Ferguson, “The 1753 Carte chronographique,” 197.  Jacques Barbeu-Dubourg, Chronographie; ou, Description des temps, contenant toute la suite des souverains de l’Univers, & des principaux événemens de chaque siècle, depuis la Création du Monde jusqu’a présent (Paris, 1753), 5. 30 31

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

209

geographer would do.”32 Here, the grid mediates between two different disciplines as a formal method of visual display: history, which is arranged in a temporal grid, and geography, which is expressed through a cartographic grid. For Barbeu-Dubourg, the analogy between time and space recalls mapping: his device follows the abstract patterns of a grid in showing the timeline of events, which are rendered orderly by the design. His chronological machine shows the equidistance between the vertical divisions of the horizontal line, each increment equating the span of a year. This creates a lopsided distribution of data along the timeline: while at one side of the timeline there are no events on record, their number increases as history becomes more populated with events (Fig. 6.2). The density of inscribed events in particular time periods creates a dynamic visual effect: the interlacing of lines and dotted lines at the top of the page, for instance, is in striking contrast with the precise writing at the bottom of the page, which demonstrates the stability of the grid in relation to the fluctuation of information.33 The concentration of streaks at the bottom of the timeline creates a cascading effect, suggesting the organic vitality of the entries that runs counter to the staid rigidity of the elegant grid. This is the innovation of Barbeu-Dubourg’s chronological machine: the spacing of the grid does not change. As Ferguson correctly points out, “[s]uch placing of human events against an absolute scale was unlike time-­ lines before it.”34 In this invention, the grid retains its integrity and is associated with rationality and, by extension, morality: the formal mapping of the grid supports the moral intent of the project.35 The grid appears to yield a non-objective stratum with each implementation of its model, similar to Krauss’s spiritual dimension and Damisch’s divination origins. The grid allows to place past events against the backdrop of a rigid pattern and, therefore, illuminates the past’s “absolute place in all time.”36  Ferguson, “The 1753 Carte chronographique,” 196.  As Davis explains, “time runs from left to right. The other axis is used for states and kingdoms, with two bands near the bottom edge for notable events and individuals. This detail shows the lowest band for the period from about 1693 A.D. to 1728 A.D.” See Davis, “History on the Line,” 5. 34  Ferguson, “The 1753 Carte chronographique,” 212. 35  Ibid. 36  Ibid. Ferguson explains the endurance of Mercador’s grid in cartography in the following terms: “In 1569, Mercador created a cylindrical map projection that became the standard map projection for nautical navigation because of its ability to represent lines of constant course (rhumb lines or loxodromes), as straight segments that conserve the angles with the meridians. Nowadays, many online street mapping services, such as Bing Maps or Google 32 33

210 

J. ZDEBIK

Fig. 6.2  Jacques Barbeu-Dubourg’s Machine chronologique (detail), 1753. (Courtesy of Princeton University Library)

The imposition of the grid onto the map was also a turning point for cartography. Earlier on, Gerardus Mercator had improved upon Ptolemy’s project by showing “the continents and seas laid out over a grid made by parallel vertical lines (meridians) and parallel horizontal lines (latitudes).”37 In Barbeu-Dubourg’s project, the grid allows for a visualization of history through geography. This analogy with cartography was applied in the eighteenth century to economy as well.

Maps, use a variant of the Mercator projection for their map images called Web Mercator or Google Web Mercator.” 37  Ibid., 196, note 15.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

211

Quesnay and the Grid as Relation Just like Barbeu-Dubourg, François Quesnay (1694–1774) was a physician who in the latter part of his life turned to the study of economics. From humble beginnings as a farmer’s son, he became a surgeon and then the personal physician to Louis XV and Madame de Pompadour.38 Quesnay and Barbeu-Dubourg were also similar in their reliance on cartography. In her essay, “Envisioning Capital,” Susan Buck-Morss highlights the link between cartography and economic charts by mentioning that mapping the economy was an outgrowth of the technique used in navigational maps.39 Buck-Morss goes on to explain that Quesnay’s map interweaves three different economic stakeholders—farmers, landowners, and artisans—and highlights their relationship of exchange. These stakeholders were represented as separate spaces and their exchange of goods and labor through time was depicted through a zigzag pattern entwining through the space of a grid (Fig. 6.3). The chart essentially demonstrates the economic flow between farmers, landowners and artisans. This relationship between hard to picture or even abstract notions and concrete ones was referred to by Quesnay as a graphic representation of the work of thought: “the zigzag, if properly understood, cuts out a whole number of details, and brings before your eyes certain closely interwoven ideas which the intellect alone would have a great deal of difficulty in grasping, unravelling and reconciling by the method of discourse.”40 The zigzag upon the grid represents abstract relationships between various domains and forces in a visual display of information. The non-orthogonal, diagonal grid creates a zigzag of relationships through the inherent dynamism conveyed by the crossways pattern. Another analogical element in Quesnay’s charting of economic abstraction through a path-making on a gridded space is the biological image of circulation. Here, the grid is a repository of data within which the zigzag 38  Jacqueline Hecht, “Le bicentenaire du Tableau économique: François Quesnay et la physiocratie,” Population 13, no. 2 (1958): 287–92, https://doi.org/10.2307/1524823 39  Susan Buck-Morss, “Envisioning Capital: Political Economy on Display,” Critical Inquiry 21, no. 2 (1995): 434–67, https://doi.org/10.1086/448759. To situate BuckMorss’s visualization of capitalist networks, see the geographically based links established between economy and history in David Harvey, Spaces of Capital: Towards a Critical Geography (New York: Zone Books, 2001), 109–12. 40  François Quesnay, “Letter from Quesnay to Mirabeau,” in David McNally, Political Economy and the Rise of Capitalism: A Reinterpretation (Berkeley: University of California Press, 1988), 110. Quoted in Buck-Morss, “Envisioning Capital,” 440.

212 

J. ZDEBIK

Fig. 6.3  François Quesnay, Tableau économique, 1766. (Courtesy of Bibliothèque Nationale de France)

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

213

is meant to illustrate a dynamic exchange. What is significant in this visualization, according to Buck-Morss, is that, as a physician, Quesnay is interested in charting economic circulation between states in medical terms: thus, “[t]he circulation of wealth was to him the lifeblood of society… Even before William Harvey’s seventeenth-century physiological theories, the description was common of money ‘circulating’ through the ‘body politic.’”41 The circulation as a relational exchange is the pure function isolated on the overall graph. What is important here, and what will lead us to our final example of the grid pattern in the eighteenth-century graphic design, is what Buck-Morss reads as a biological impulse in the charting of different economic forces: “At the same time, the ‘picture’ Quesnay provided was one in which these two schemes, circulation (circular flow) and production (the fertility schema), folded into each other in the same social body.”42 In fact, Quesnay’s design captures the very idea of exchange. The dynamism coiled in the sloping lines activates the grid and accounts, visually, for a way of considering the grid as the very condition that makes it possible for abstract functions to shuttle from system to system. In this case, the expenditure undergoes changes in significance as it circulates between farmer, landowner, and artisan. Furthermore, the dynamic flow of the design is reinforced by the oblique lines, which give the impression of a complex irrigation system. The grid pattern trickles down from system to system like a persistent leak opportunistically piggybacking different materials to make its way through a dominant structure. From history to geography to economy to biology, the grid’s organizational and relational functions are consistently integral to the context, or system, in which they find themselves at particular moments in time.

Linnaeus and the Grid as Organization Carl Linnaeus (1707–1778), the Swedish naturalist credited with the creation of modern taxonomy, relied on tabular notation to classify plants, animals, and minerals.43 Linnaeus had the most significant impact on  Buck-Morss, “Envisioning Capital,” 440.  Ibid., 443. 43  For details, see Isabelle Charmantier, “Carl Linnaeus and the Visual Representation of Nature,” Historical Studies in the Natural Sciences 41, no. 4 (2011): 365–404, https://doi. org/10.1525/hsns.2011.41.4.365 41 42

214 

J. ZDEBIK

eighteenth-century thought of the three graphic grid innovators discussed here. Linnaeus did not only use the grid as a way to organize elements of life visually; as Foucault emphasized, his very rhetoric, insofar as it is an organizing endeavor, has something of the grid in its discursive methodology. In “Articulate Images: Bringing the Pictures of Science and Natural History into the Art Curriculum,” Julia Marshall links the graphic organization of information in natural history illustrations with the graphic surface function of an eighteenth-century painting. This relationship can be explained by following the theories of art historian Leo Steinberg, who viewed the surface of a painting as an information repository, and of James Elkins, who described informational images as needing an extra dimension of scientific interpretation to alter the readability of the visual data immediately displayed on their surface.44 Marshall discusses the reciprocal influence of natural history illustration and fine art painting that focuses on nature. The informational illustration takes from painting its display of its subject of study: “Many display an entire organism, often in profile, with certain part-specific structures more fully displayed in peripheral drawing. Some show many animals or plants presented in rows or grids.”45 It is this very display in rows or grids that is of interest to us here, especially as employed in the work of Linnaeus. Marshall explains how, in his Classification of Plants by Sexual Characteristics, Linnaeus demonstrates the function of the grid display. She makes a distinction between the technique of the artist who actually made the drawings of flowering plants, Georg Dionysius Ehret, and Linnaeus’s idea of illustrating these images. This idea is not enwrapped in the drawing style but in the organization of the elements on the page. Thus, the plants are “delineated according to the number and arrangements of stamens. The individual specimens are numbered 1 to 24, suggesting that there is a key to the image that identifies each one”46 (Fig. 6.4). The key is important since it is what makes these illustrations informational images as opposed to fine art paintings, due to the combination of 44  Leo Steinberg, “Other Criteria,” in Other Criteria: Confrontation with TwentiethCentury Art (London: Oxford University Press, 1972), 55–92; James Elkins, “Art History and Images That Are Not Art,” in Images: Critical and Primary Sources: Understanding Images, ed. Sunil Manghani (London: Bloomsbury, 2013), 554–71. 45  Julia Marshall, “Articulate Images: Bringing the Pictures of Science and Natural History into the Art Curriculum,” Studies in Art Education 45, no. 2 (2004): 140, https://doi. org/10.1080/00393541.2004.11651762 46  Ibid., 143.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

215

Fig. 6.4  Georg Dionysius Ehret, Methodus, 1736. (Courtesy of The Linnean Society of London)

image and text. Each element of the drawing is numbered and has an accompanying textual explanation. However, the association of the image with text is only one of the two ways in which this image is informational. The other one has to do with the gridded space it occupies. The arrangement of the plant samples can be compared to Linnaeus’ Regnum Animale (1735) table, which organizes the information graphically by entering the place of each species in relation to others in a grid. The illustration is, in turn, arranged rectilinearly, which suggests the comparative function of the arrangement. The picture is not art but document, and this is due to the grid, which serves a function (i.e., comparison) but also attributes a higher, more objective authority to the image. The grid thus plays a role in the very principle at work in Linnaeus’s scientific endeavors, which called for “clear, precise, explicit and comprehensive drawings, often composed in flat, spare, and grid-like patterns devoid of any reference to

216 

J. ZDEBIK

environment.”47 In other words, the grid offers the object up for study by taking it out of its actual space and placing it into a space of scrutiny. The grid is the condition of relationality, allowing for a comparison between abstract pieces of information. Simultaneously, the grid organizes the surface that captures information visually, as in the case of the displayed flowers, or textually, as in the case of the table of categories. It is both function and display of function.

Classical Grid as Philosophy In The Order of Things, Foucault focuses on classifications as elements of knowledge built in eighteenth-century botanical illustrations. In order to get to the substance of their subject of study, these illustrations had to focus on part, number, and position so as to be able to capture information. Thus, the grid becomes a method which allows for objective scrutiny. Nowhere is this function of the grid more prominent than in Linnaeus’s Regnum Animale from Systema Naturae (Fig.  6.5). The grid pattern organizes animal life into six categories, each of which is then hierarchically arranged and boxed into a rectilinear lattice. Although less dynamic than Barbeu-Dubourg’s grid of time, which suggests a lateral movement, or Quesnay’s Tableau économique, whose oblique directionality suggests dynamic relations among different stakeholders, Linnaeus’s grid allows the eye to see and create connections. The viewer’s participation animates the knowledge displayed on the surface of the page.48 Here, Foucault describes Linnaeus’s method of using the grid as a repository for the gaze, where black squares capture invisible relations and open space is meant to contain words.49 The use of black and white spaces in visual interface with things is a way of providing clarity to information. Visibility and ­expression  Ibid.  For the active participation of the viewer in the graphic display of knowledge, see Bender and Marrinan’s discussion of the “active correlation of information” triggered by the visual diagrams in The Culture of Diagram, 8. They also discuss the “flat planar surface” arrangement of information in diagrams that “incite a correlation of sensory data with the mental schema of lived experience” (21). The image of surveying knowledge from great heights as if looking at a map has also been used by David S. Ferris in “Post-Modern Interdisciplinarity: Kant, Diderot and the Encyclopedic Project,” Modern Language Notes 18, no. 5 (2003): 1261, https://doi.org/10.1353/mln.2004.0007 49  Michel Foucault, The Order of Things: An Archaeology of the Human Sciences (London: Routledge, 2005), 148. According to Foucault, “Buffon and Linnaeus employ the same grid; their gaze occupies the same surface of contact upon things; there are the same black 47 48

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

217

Fig. 6.5  Carl Linnaeus, Regnum Animale, in Systema Naturae, 1735. (Courtesy of The Linnean Society of London)

find their medium in illustration through the aesthetic strategy of graphic clarity. This is why Foucault compares botany to anatomy: because of the heft of data to be analyzed. The body makes the graphics heavier and darker with inscription, whereas, for the purpose of classifying data in order to refine the organizational interface, plants provide distinction on their surface and lend themselves to a clearer graphic description: “Hence the epistemological precedence enjoyed by botany: the area common to words and things constituted a much more accommodating, a much less ‘black’ grid for plants than for animals; in so far as there are a great many constituent organs visible in a plant that are not so in animals, taxonomic knowledge based upon immediately perceptible variables was richer and more coherent in the botanical order than in the zoological.”50 Thus, the squares left to accommodate the invisible; the same open and distinct spaces to accommodate words.” 50  Ibid. As Foucault points out, “in the seventeenth and eighteenth centuries, anatomy lost the leading role that it had played during the Renaissance and that it was to resume in Cuvier’s day; it was not that curiosity had diminished in the meantime, or that knowledge

218 

J. ZDEBIK

intertwining of knowledge and aesthetics has never been clearer and more consequential. What is important for Foucault when analyzing the dynamics of graphic information is the murkiness of its visual impact. The “black” grid is to be avoided in favor of a multi-purpose, flexible, clear, and encompassing grid that displays objects and describes them in writing. The grid as an actual and methodological device can be applied across disciplines and animal or vegetable kingdoms.51 The grid, then, takes over some of the discursive duties of the text and is able to convey information silently: it groups and organizes individuals in categories and this makes superfluous information redundant and, therefore, not necessary to be represented.52 Much as in the case of Quesnay’s zigzag, Linnaeus’s grid offers, figuratively speaking, the possibility of creating a network through which information about relationships flows.53 The grid becomes, thus, an abstract way of conveying information. As Foucault explains, nature itself becomes visible when filtered through the grids of graphic design and intellectual methodology: “nature is posited only through the grid of denominations … it glimmers far off beyond them, continuously present on the far side of this grid, which nevertheless presents it to our knowledge and renders it visible only when wholly spanned by language.”54 Thus, Foucault’s treatment of Linnaeus stands in for a general view of the information visualization in the eighteenth century and can encompass the intellectual endeavors of both Quesnay and Barbeu-Dubourg. Interestingly, Foucault also offers some insights into the relationship between informational images and fine arts when he discusses the relationship between table, picture, and grid. These three elements express the impetus of sciences to create a table-language that would replicate the had regressed, but rather that the fundamental arrangement of the visible and the expressible no longer passed through the thickness of the body.” 51  Ibid., 154. 52  Ibid. Foucault details how the grid’s display of information functions as a network with a clear intersection of verbal and visual elements: “Each group can be given a name, with the result that any species, without having to be described, can be designated with the greatest accuracy by means of the names of the different groups in which it is included. Its complete name will cross the entire network of characters that one has established, right up to the largest classifications of all. But for convenience, as Linnaeus points out, part of this name should remain ‘silent’ (one does not name the class and order), while the other part should be ‘sounded’ (one must name the genus, the species, and the variety.).” 53  Ibid. 54  Ibid., 174–75.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

219

nature exactly in all its details, and where information about things could be inserted without any mediation.55 In the eighteenth century, he explains, the table was not there to mirror nature but rather to tabulate similarities and differences based on language. As Foucault suggests: “it was a matter of dividing nature up by means of a constant table of identities and differences for which language provided a primary, approximative, and rectifiable grid [emphasis mine].”56 In Foucault’s view, the table is a tableau, a mix of picture and language that is more akin to art than science and draws a portrait of nature instead of trying to mirror it.57 Language generates a grid of differentiation which, in turn, is visually emphasized by the graphic layout of the table. For Foucault, the grid, or its dismantling, also marks the passage between a relational network of different registers of thought and a shuttering of each modality in its own self-reflexive limit, where each discipline ceases communicating with another.58 This unfolding in the mind of the observer is exemplified by Linnaeus’s illustration of the multiple facets of minerals in Regnum Lapideum. Placed on a grid, the movement of shapes is almost dynamic, allowing for comparison (Fig. 6.6). In the second illustration, Linnaeus delves deeper into the structure of minerals by showing the folding function present in the things themselves (Fig. 6.7). The relationship between the grid and the eighteenth-century thought is so pronounced that, once the grid is dismantled, the classical era is no longer accessible to us in all its complexity: “The last ‘bastion’ to fall—and the 55  Ibid. Again, in the following Foucault shows the intermingling of verbal and visual categories in the display of information: “This is the positivist dream of a language keeping strictly to the level of what is known: a table-language, like the one Cuvier was probably dreaming of when he attributed to science the project of forming a ‘copy’ of nature; scientific discourse was to be the ‘table’ of things.” 56  Ibid., 323. 57  Ibid. As Foucault explains, “language is not so much a table as a picture, in the sense that, freed from the intricacy that gives it its immediately classifying role, it stands a certain distance apart from nature in order to draw some of it into itself by means of its own passivity, and finally to become nature’s faithful portrait.” 58  Ibid. When the function is stripped away, Foucault argues, representation remains: “And how, after all (if not by a slow and laborious technique), are we to discover the complex relation of representations, identities, orders, words, natural beings, desires, and interests, once that vast grid has been dismantled, once needs have organized their production for themselves, once living beings have turned in toward the essential functions of life, once words have become weighed down with their own material history in short, once the identities of representation have ceased to express the order of beings completely and openly?”

220 

J. ZDEBIK

Fig. 6.6  Carl Linnaeus, Regnum Lapideum, 1768. (Courtesy of The Linnean Society of London)

Fig. 6.7  Linnaeus, Regnum Lapideum [folded mineral shapes], 1768. (Courtesy of The Linnean Society of London)

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

221

one whose disappearance cut us off from classical thought forever—was precisely the first of all those grids: discourse, which ensured the initial, spontaneous, unconsidered deployment of representation in a table.”59 A chasm opens up here between the eighteenth century’s use of the grid and our own. In other words, an aesthetic function discarded from the classical era can be mined for visual expression in contemporary art. We can look, for instance, at Barbara Maria Stafford’s monograph, where she proposes a visual analogy between eighteenth-century and contemporary views of information in an effort to bridge the span from “the eighteenth-century fascination with analogy and metaphor” to “the creation of a forward-looking hermeneutic tool.”60 For Stafford, analogy is a methodology ripe for an era of information which functions as a shuttle, “allowing the viewer to go back and forth between the past and the present.”61 It is a combinatory apparatus that can connect materials and thoughts of the past and present in order to allow for a speculative path toward the future through a non-discursive, alogical approach to visual thinking, devoid of any methodological constraints: “Analogy … can even be seen as part of a global ecological mission since it does not add to disposable commodities, but reuses and reconfigures them. Considering patterns or, better, societies and their divergent artifacts as potentially congruent might allow us to enter the twenty-first century less incoherently, more open to new combinations”.62 In this case, analogy in all its inventive potential functions as a diagram in which pure functions move from one system to another. In other words, Stafford points that diagram functions in a similar way to analogy: “The transfer from a familiar to an unfamiliar referent lands us in an equivocal region.”63 It is the non-linear, approximate nature of analogy that gives it force as a potentially revelatory methodology: Analogia, then, should not be confused with establishing identity or isomorphism. When combined together, the prefix ana and the noun logos etymologically mean “according to duration,” “according to the same kind of way.” As a method, it establishes a proportional relationship either among  Ibid.  Barbara Maria Stafford, Good Looking: Essays on the Virtue of Images (Cambridge, MA: MIT Press, 1996), 120. 61  Ibid., 209. 62  Ibid., 209–10. 63  Ibid., 211. 59 60

222 

J. ZDEBIK

single items or among a group of embodied properties existing under diverse conditions and adapted to multiple purposes. Vision makes an imaginative leap, embarking on a dialectical process of seeing affinities not easy to convey in words.64

The link between Stafford’s concept of analogy, which can make connections beyond simple similarities, and diagram, which shows how a function belonging to a particular system “travels” to another system and operates similarly in a new context, is illustrated by the notion of “image-schema.” This concept links the diagram to Immanuel Kant’s schema, which translates empirical data into conceptual notions through an image-like operation. Stafford herself makes links between analogical translations, schemas, and mapping. The concept of “image-schema” is sampled from neurological mapping in consciousness studies, where analogy works by mapping relevant cognitive regions and then identifies those that are similar but heterogeneous. In this context, the map allows for a crossover from one region to another.65 This movement from one graph to the next can be telescoped to the level of methodology; thus, eighteenth-century functional aesthetics can be shifted onto twentieth-century avant-garde objects. If, for Foucault, classical thought ceases to be directly accessible to us, it is through an analogical transmission that we can make the link between the eighteenth-century grids and twentieth-century graphic design. Eve Meltzer explains how discourse does not function as a grid, as it did in the classical age, but as a grid-pastiche: “Jameson’s binaries, his data, devices, modes of decoding, and deciphering, as well as Barthes’s stoking

 Ibid.  Ibid., 211. Stafford demonstrates how, from one spatial context to another, data is mapped out and diagrammatically transferred to form a network of information based on similarities: “Devised originally to handle concepts not at home in any category, analogy is productively employed in recent studies of consciousness. According to Lakoff, the brain develops ‘image-schema’ to perform intuiting, sorting, and associating functions required in those ambiguous cases where the phenomenal world is not already classified. What is exciting about analogy is that, after many stimulus inputs, particular neuronal groups will be selected in patterns. Similar signals will not only activate the previously selected neuronal groups in one map, but whey will cross over to another map or even to a whole group of maps because their operation is linked together by reentry channels. What is described as a ‘map’ here is a sheet of neurons in which the points on the sheet are systematically related to the points on a sheet of receptor cells.” 64 65

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

223

reference … all evoke in striking similarity the image of the grid.”66 Here, again, it is not the grid’s function that comes out of the rhetoric, but rather an aestheticization of structuralist methodologies based on the “scientifist” stance the grid lends credence to. Instead of making connections between disparate elements in order to allow for knowledge to emerge, the grid is a constraining device that is inward looking and self-reflexive: “the grid functions as ‘an arbitrary framework on which to build an entity, a self-restrictive device by which to facilitate choice.’”67 In this context, the grid projects the machinations of pure function as the aesthetic manifestation of an organizing force before data was even entered: “Like a net suspended over a void, the grid permits us to picture the absence that functions as the structuring principle and to grasp the idea that all of its terms are fundamentally negative or, as Saussure says, ‘differences without positive terms’.”68 In effect, the grid makes visible the very structure of the function, imperceptibly articulating data into readable information. What Meltzer seems to suggest is that, at some point, the art world appropriates the aesthetics of the grid to capture the zeitgeist of the information era: The dream of the information world is a fantasy about being in and of the grid. Self-restriction; arbitrariness; that disciplined, autonomous, device-like quality of being both ‘run’ and ‘followed’ at once; the proposition of an absolute visibility that defies the very conditions of the phenomenal world; the very unquestionability of the laws that govern the system; and the proposition that ‘if law is anywhere, it is everywhere’—these are the conditions of the grid. There is also its scientifist claim to reason, and its peculiar notion of visibility as all-seeing, immune to opacity, and powered by, to repeat Descartes’s formulation, the ‘inward vision of [the] mind.’69

This “scientifist” impetus in art can be seen in the work of several artists who use grid patterns, resembling the graphic patterns of eighteenth-­ century thinkers. Francine Savard, for instance, is a painter who works on the edge of conceptual art. Her work, Tableau chronologique (2008–2009), 66  Eve Meltzer, Systems We Have Loved: Conceptual Art, Affect, and the Antihumanist Turn (Chicago: University of Chicago Press, 2013), 60. 67  Lucy Lippard, “Top to Bottom, Left to Right,” in Grids, Grids, Grids, Grids, Grids, Grids, Grids, Grids (Philadelphia: Institute of Contemporary Art, 1972). Quoted in Meltzer, Systems We Have Loved, 60. 68  Meltzer, Systems We Have Loved, 60. 69  Ibid., 65.

224 

J. ZDEBIK

Fig. 6.8  Francine Savard, Tableau chronologique, 2008–2009. Inkjet printing on rag paper. (Courtesy of the artist)

consists of a table on which a long paper is placed as if it were an enormous map (Fig.  6.8). The white surface is crisscrossed with gentle lines and words that indicate the names of authors whose writings were sampled in epigrams of other authors. What is interesting here is the density of connections on one side of the timeline and the large gap between the left and the right sides. This indicates that, very much like Barbeu-Dubourg, Savard is operating on an absolute demarcation of time. Savard’s names are more interconnected on one side of the chart, just like Barbeu-­ Dubourg’s timeline, which gets denser as it gets closer to the contemporary time of its production. Likewise, Hans Haacke’s Shapolsky et  al. (1971) can be compared to Quesnay’s Tableaus (Fig.  6.8). Haacke creates a chart displaying the investment paths involved in the management of slums that looks uncannily similar to Quesnay’s visualization: it consists of two columns at the opposite sides of the page, interlaced with lines. However, whereas Quesnay makes visible economic relations that were present but unseen in order to formalize a discipline, Haacke makes visible clandestine predatory practices that define a market economy. Haacke designs a graphic that depicts the origins of capitalism in order to denounce the capitalism’s excess. Finally, Luce Meunier’s work seems to link Linnaeus’s patterns to contemporary art: we can find a visual analogy between her Figure semblable #1 and Linnaeus’s visualization of the multiple facets of minerals in

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

225

Fig. 6.9  Hans Haacke, Detail of Shapolsky et al. Manhattan Real Estate Holdings, a Real-Time Social System, as of May 1, 1971 © Hans Haacke / Artists Rights Society (ARS), New York. (Courtesy of the artist and Paula Cooper Gallery, New York)

Regnum Lapideum, where they are arranged according to an invisible grid (Fig. 6.9). This gives them a relational organization for comparative purposes but also allows the viewer to see each mineral in a distinct, decontextualized, and objective depiction. Despite its contested contribution to the

226 

J. ZDEBIK

Fig. 6.10  Luce Meunier, Figure semblable #1, 2012. Etching on paper, edition of 5. (Courtesy of the artist)

history of mineralogy,70 Linnaeus’s Regnum Lapideum can thus be recuperated aesthetically due to its similarity to the folding and unfolding patterns found in Meunier’s abstract creations. Her Figure semblable #1, for example, resembles an indexical notation of a technical process that can be deciphered by using a scientific code (Fig. 6.10). In fact, just like Linnaeus’s mineral tables, Meunier’s variations of the folds are only surface-­level. Because of the lost interpretive key in an era of informational saturation, we are forced to come to terms with the fact that the actual function of the grid is aesthetic.

Conclusion The grid organizes abstract information by making it visually coherent. As demonstrated here, the rift between the abstract organization of knowledge in the classical era and our present-day data-driven society is not absolute. Foucault sees a rift in relation to the treatments and uses of the grid, Meltzer explains how the diagrammatic function of the grid stays on the surface of art, and Stafford sees a possibility of bridging this gap through analogy. Aesthetic theory helps us see the trajectory of the grid from the checker board to the battlefield, and from scientific models to 70  According to Peter J. Heaney, “the first is a misconception that modern mineral classification eschews a Linnaean structure, when, for a century and a half, mineralogists actually have employed a Linnaean tree to organize the mineral kingdom.” See Peter J.  Heaney, “Time’s Arrow in the Tree of Life and Minerals,” American Mineralogist, 101 (2016): 1027, https://doi.org/10.2138/am-2016-5419

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

227

avant-garde minimalist paintings. Similarly, the concept of the diagram helps us account for the grid’s function in a different context, establishing the importance of the grid for an epistemological-aesthetic theory. This disciplinary heterogeneity is obvious in the way the grid was used by different innovators in the eighteenth century. The grid was employed to organize time and history, to present economy as an organic system, and to convey knowledge in botany, biology, and mineralogy. In each of these three manifestations, the grid functions as an organizing principle, bringing together different visual elements. Barbeu-Dubourg’s temporal line takes its inspiration from geography, Quesnay’s diagonal grid of economic exchanges is based on the anatomical model of the circulatory system, and Linnaeus’s classification method relies on rhetorical visualizations, being a hybrid combination of text and image. Each of these three pioneers of graphic organization were, indeed, working in a multidisciplinary way, very much like the conceptual artists of today. The consequences of the eighteenth-century multidisciplinary mingling of the visual and the textual and of the visualization of abstract knowledge can be felt in the contemporary art in relation to its defining traits of conceptualism in all forms.71 But it is specifically in the artistic representation of abstract information that the innovations of the eighteenth-century thinkers are the most evident. The divide between the present and the past use of visualization has never been more keenly felt. The effects of these classical models are manifest in the aestheticization of the anxiety about the quantification of information in art.

Bibliography Aldridge, Alfred Owen. 1951. Jacques Barbeu-Dubourg, a French Disciple of Benjamin Franklin. Proceedings of the American Philosophical Society 95 (4): 331–392. http://www.jstor.org/stable/3143280 Barbeu-Dubourg, Jacques. 1753. Chronographie; ou, Description des temps, contenant toute la suite des souverains de l’Univers, & des principaux événemens de chaque siècle, depuis la Création du Monde jusqu’a présent. Paris. Bender, John, and Michael Marrinan. 2010. The Culture of Diagram. Stanford: Stanford University Press.

71  David Joselit, “On Aggregators,” October 146 (2013): 3–18, https://doi.org/10.1162/ OCTO_a_00154. Joselit explains how conceptualism is a fundamental building block of contemporary art and traces its relation to documentation, data, and information.

228 

J. ZDEBIK

Benjamin, Walter. Walter Benjamin: Selected Writings. Volume 2: 1927–1934. 1999. Trans. Rodney Livingstone. Edited by Michael W. Jennings, Howard Eiland, and Gary Smith. Cambridge: Belknap Press of Harvard University Press. Buci-Glucksmann, Christine, and Josh Wise. 1998. Of the Diagram in Art. ANY: Architecture in New York 23: 34–36. https://www.jstor.org/stable/i40087754 Buck-Morss, Susan. 1995. Envisioning Capital: Political Economy on Display. Critical Inquiry 21 (2): 434–467. https://doi.org/10.1086/448759. Charmantier, Isabelle. 2011. Carl Linnaeus and the Visual Representation of Nature. Historical Studies in the Natural Sciences 41 (4): 365–404. https:// doi.org/10.1525/hsns.2011.41.4.365. Damisch, Hubert. 1998. Genealogy of the Grid. In The Archive of Development, ed. Henk Slager and Annette W. Balkema, 49–53. Amsterdam: Rodopi. Davis, Stephen Boyd. 2012. History on the Line: Time as Dimension. Design Issues 28 (4): 4–17. https://doi.org/10.1162/DESI_a_00171. Elkins, James. 2013. Art History and Images That Are Not Art. In Images: Critical and Primary Sources: Understanding Images, ed. Sunil Manghani, 554–571. London: Bloomsbury. Ferguson, Stephen. 1991. The 1753 Carte chronographique of Jacques Barbeu-­ Dubourg. The Princeton University Library Chronicle 52 (2): 190–230. https://doi.org/10.2307/26404421. Ferris, David S. 2003. Post-Modern Interdisciplinarity: Kant, Diderot and the Encyclopedic Project. Modern Language Notes 18 (5): 1251–1277. https:// doi.org/10.1353/mln.2004.0007. Foucault, Michel. 1995. Discipline and Punish: The Birth of the Prison. Trans. Alan Sheridan. New York: Vintage Books. ———. 2005. The Order of Things: An Archaeology of the Human Sciences. London: Routledge. Harvey, David. 2001. Spaces of Capital: Towards a Critical Geography. New York: Zone Books. Heaney, Peter J. 2016. Time’s Arrow in the Tree of Life and Minerals. American Mineralogist 101: 1027–1035. https://doi.org/10.2138/am-2016-5419. Hecht, Jacqueline. 1958. Le bicentenaire du Tableau économique: François Quesnay et la physiocratie. Population 13 (2): 287–292. https://doi. org/10.2307/1524823. Higgins, Hannah B. 2009. The Grid Book. Cambridge: MIT Press. Joselit, David. 2013. On Aggregators. October 146: 3–18. https://doi. org/10.1162/OCTO_a_00154. Krauss, Rosalind. 1979. Grids. October 9: 50–64. http://www.jstor.org/ stable/778321 Lippard, Lucy. 1972. Top to Bottom, Left to Right. In Grids, Grids, Grids, Grids, Grids, Grids, Grids, Grids. Philadelphia: Institute of Contemporary Art.

6  THE GRID AND THE VISUALIZATION OF ABSTRACT INFORMATION 

229

Marshall, Julia. 2004. Articulate Images: Bringing the Pictures of Science and Natural History into the Art Curriculum. Studies in Art Education 45 (2): 135–152. https://doi.org/10.1080/00393541.2004.11651762. McNally, David. 1988. Political Economy and the Rise of Capitalism: A Reinterpretation. Berkeley: University of California Press. Meltzer, Eve. 2013. Systems We Have Loved: Conceptual Art, Affect, and the Antihumanist Turn. Chicago: University of Chicago Press. Panofsky, Erwin. 1991. Perspective as Symbolic Form. New York: Zone Books. Salkever, Stephen G. 1977/1978. Interpreting Rousseau’s Paradoxes. Eighteenth-­ Century Studies 11 (2): 204–226. https://doi.org/10.2307/2738384. Stafford, Barbara Maria. 1996. Good Looking: Essays on the Virtue of Images. Cambridge, MA: MIT Press. Steinberg, Leo. 1972. Other Criteria: Confrontation with Twentieth-Century Art. London: Oxford University Press. Tufte, Edward R. 1997. Visual Explanations: Images and Quantities, Evidence and Narrative. Cheshire: Graphics Press. Wainer, Howard. 2006. Graphic Discovery: A Trout in the Milk and Other Visual Adventures. Princeton: Princeton University Press. Wickman, Matthew. 2015. Literature after Euclid: The Geometric Imagination in the Long Scottish Enlightenment. Philadelphia: University of Pennsylvania Press. Wittgenstein, Ludwig. 2016. Tractatus Logico-Philosophicus. Trans. C.G. Ogden. Sweden: Chiron Academic Press.

CHAPTER 7

Exploring Data Visualization: Time, Emotion, and Epistolarity in Frances Brooke’s The History of Emily Montague Courtney A. Hoffman

This chapter is essentially a narrative about learning how data visualization works, how it can be used in textual analysis, and what programs can be helpful or not. As a novice in digital humanities in general, and in data visualization in particular, my approach may seem simplistic to those who are well versed in data scraping, in coding in R, or in using Gephi, or to those who know exactly what “edges” and “nodes” mean and especially what they are meant to do. This chapter is not intended to enlighten those who create programs from which the statistics used in data visualization are derived, or who produce the beautiful graphs, charts, and clouds that allow us to communicate our findings pictorially. Rather, it is meant for those, like me, who are newbies in the realms of data visualization, and who may want to learn the steps they can take to conquer what may feel like a daunting task: dipping their toe into the world of data visualization.

C. A. Hoffman (*) Georgia Institute of Technology, Atlanta, GA, USA e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_7

231

232 

C. A. HOFFMAN

Outside of the field of literary studies, the use of infographics is nearly universal. While literary scholars are limited to writing analyses that rely on the occasional use of paintings or engravings to complement their verbal eloquence, the use of tables, graphs, and charts pervades the media of communication in other disciplines. In an ever-changing, fast-paced, global environment, the need to reconsider how to apply new research methods to our academic endeavors is becoming more apparent. The availability of texts is broader than ever, and digitization provides easy access to material that previously required travel, often an expensive undertaking, especially when funding for such trips is decreasing. While we were once limited to analyzing a small number of texts in isolation, the current access to digital books allows us to broaden the scope of our projects and consider questions on the scale of hundreds, if not thousands of texts. As evidenced by other chapters in this book, big data has made inroads into the field of eighteenth-century studies. This is not to say that data visualization has no place in projects that consider only one text. Indeed, my work here is based on an analysis of the intersections of time and emotion in Frances Brooke’s epistolary novel The History of Emily Montague (1769).1 As a text, this novel is somewhat unique among epistolary novels of the eighteenth century as it focuses on not one, but several protagonists: it tells the story of ten different characters who compose 228 letters, most of which must traverse the Atlantic to reach their recipient. In short, the novel’s plot revolves around the courtship of Colonel Edward Rivers, who travels to Canada to repair the family fortunes, and Emily Montague, a young woman of somewhat mysterious origins. Rivers’ letters home to his sister, Lucy, and his friend, John Temple, intertwine with the correspondence of Lucy’s friend, Arabella (Bell) Fermor, who is the daughter of one of Rivers’ military acquaintances. The epistles comprising the text travel mainly from their writers in Canada to their recipients in England, though Lucy and Temple send a few replies from England to the colony as well. The final quarter of the novel is set in England, where Rivers completes his courtship of Emily, his fortunes restored, Lucy weds Temple, and Bell marries her suitor. Although Brooke’s work follows many genre conventions, The History of Emily Montague contrasts sharply with more famous examples of epistolary novels. Samuel Richardson’s Pamela (1740), for example, devolves 1  Frances Brooke, The History of Emily Montague, ed. Mary Jane Edwards (Don Mills, ON: Carleton University Press, 1985). All the references are to this edition.

7  EXPLORING DATA VISUALIZATION 

233

from a correspondence between the titular character and her parents to a one-sided approximation of a journal, where Pamela addresses her parents at the start of each missive, but never receives a written answer. Also, although Richardson’s second novel, Clarissa (1748), features two pairs of correspondents engaging in letter-writing, Richardson never attempted to sustain as complex a rhetorical situation as Brooke later would do, with nearly three times as many individual writing personae contained in a single novel. Published almost a decade after Brooke’s novel, Evelina (1776), Frances Burney’s most popular work, similarly commences with an exchange of letters among Evelina, her guardian Mr. Villiers, and a few other characters, but quickly shifts to letters written mainly by Evelina, with only few examples of replies to her.2 The fact that Brooke’s text displays such a broad network of epistolary exchange, combined with the novel’s setting in Canada for the bulk of the action, further complicates my rhetorical situation as I set out to analyze the emotional aspects of the correspondence among the characters over, in, and through time. The months passing before a letter’s recipient in London could read what her friend wrote from Montreal, then even more time passing before the original writer could expect a reply, not only replicates the realities of eighteenth-century transatlantic correspondence but also emphasizes the impact that the letter’s rhetorical appeals would have on the reader. As Janet Gurkin Altman argues in her influential work on the structures of genre in eighteenth-century epistolary novels, “first-­ person narrative lends itself to the writer’s reflexive portrayal of the difficulties and mysteries surrounding the act of writing. Therefore, the epistolary form is unique among other first-person narratives in its aptitude for portraying the experience of reading. In letter narratives we not only see correspondents struggle with pen, ink, and paper but also their messages being read and interpreted by their intended or unintended recipients.”3 In this case, the reader is as important as the writer of letters, and, therefore, the contents of the letters sent within the structure of the epistolary novel matter to plot—as does the time when they are read. 2  Frances Burney, Evelina, or, the History of a Young Lady’s Entrance into the World, ed. Stewart J.  Cooke (New York: W.  W. Norton, 1998); Samuel Richardson, Clarissa, or the History of a Young Lady, ed. Angus Ross (London: Penguin, 2004); Samuel Richardson, Pamela; or, Virtue Rewarded, ed. Thomas Keymer and Alice Wakely (Oxford: Oxford University Press, 2008). 3   Janet Gurkin Altman, Epistolarity: Approaches to a Form (Columbus: Ohio State University Press, 1982), 88.

234 

C. A. HOFFMAN

Altman terms this the “pivotal present,” which relates the “respective times that the letter is dispatched, received, read, and reread,”4 so that the time of reading a letter corresponds to the time of writing that same letter, thus bringing the writer and reader into a shared moment, despite the reality of time—minutes, days, weeks, months, or even years—passing between the actual events of writing and reading. A visualization that demonstrates where the letters written at earlier moments fit within a timeline while simultaneously illustrating their place within the novel’s overarching narrative would highlight, therefore, that sense of occupying a moment that has passed but still connects the letter reader to the letter writer. Stephen Carl Arch has suggested that the distance between the event of writing and the arrival of a letter in its reader’s hands blunts the emotional effect of what has been written: So, for example, in the letters that Rivers exchanges with his sister, Lucy, and with his best friend, Temple, the reader experiences a disconcerting temporal effect: letters written in heightened emotion and confidential whispers and hurried closings are received two or three or four months after their “spontaneous” compositions … The recipient and the novel reader are supposed to experience these emotions vicariously, sympathetically identifying across the distance imposed by circumstances. However, by the time Lucy had received her breathless missive, Arabella has long since moved on to other concerns, a fact that the reader actually knows and that Lucy must have sensed, even as she wrote. Those emotions are no longer alive and relevant when the letter is received.5

What Arch does not consider is that this is the nature of reading letters that Altman highlights: while time will have passed for the writer and she will have moved on to a new activity or idea, particularly when correspondents are separated by an ocean, the reader does realize that a letter was written months ago, as the date is clearly stated at the top of the page. The imaginative power of reading allows Lucy to be in the moment with Bell, to be present with her as she describes her experiences and her feelings, and to understand the emotions expressed in Bell’s writing as if they were affecting her at the moment Lucy is reading her missive. This  Ibid., 118.  Stephen Carl Arch, “Frances Brooke’s ‘Circle of Friends’: The Limits of Epistolarity in The History of Emily Montague,” Early American Literature 39, no. 3 (2004): 471, https:// www.jstor.org/stable/25057366. 4 5

7  EXPLORING DATA VISUALIZATION 

235

phenomenon is what Altman terms “temporal polyvalence”: the reader is mentally present with the writer, even while physically absent.6 Brooke very carefully uses measures of time (e.g., calendar date, day of the week, time of day) to give her characters’ correspondence an air of authenticity, so she orders her 228 letters so that a reply that has traversed an ocean will not appear juxtaposed with the original missive. Instead, letters coming from England in response to those from Canada interject their contents into the larger narrative, interrupting and influencing events described by Brooke’s colonially based characters, and shifting the plot to accommodate information that would seem to be woefully late or out-of-­ date. Thus, a letter sent from London dated 1 December 1766 appears between letters dated 10 and 11 February 1767 in Canada,7 reminding the Canadian reader of the emotions felt during the writing about events experienced and discussed months after the fact. When considering how the emotional impact of receiving letters dated so long in the past and responding to missives originally composed even farther in the past would affect the readers, I found myself somewhat bogged down. Being able to visualize the time the letters were written—rather than when they were read, as the overall temporal scheme of the novel reflected the Canadian characters’ position and put the London writers at a distance from both the characters and the extradiegetic reader—might allow me to grasp more firmly the characters’ emotions at the time of writing. Moreover, if I could also incorporate into these visualizations the letter writers’ locations, knowing that time and place are interconnected, I would potentially be able to infer how the shifts in the place of writing affected the overall emotional reactions of the characters. With these ideas in mind, I began looking for data visualization programs capable of conveying this information. There were several points I considered in my choice of which programs to try and which I ended up using. First, my own limited knowledge informed my initial search; the software programs I discuss in the rest of this essay are easy to use and relatively inexpensive to obtain. Second, I see teaching as one of my primary duties as a scholar. Research that can inform and further my ability to instruct my students is a priority for me. Therefore, a program valuable to me would not only further my research capabilities but also have potential for use in the classroom. The data visualization tools that I describe here, 6 7

 Altman, Epistolarity, 127–28.  Brooke, The History of Emily Montague, 131–35.

236 

C. A. HOFFMAN

Voyant, Palladio, and Tableau, are accessible, easy to use, and pedagogically efficient. They are also interactive, so students and researchers can play with their affordances and the data visualizations that they construct, which makes them useful tools for teaching analysis and thinking about data visualization as a whole. What I have learned from using these programs can also be transferred to more complex software; students can perform the same analysis I do and can apply the skills they gain to tasks they may be asked to complete in their future careers, outside the study of the humanities. In each of the following sections, I consider how the program in question accepted data and in what form, what visualizations the program created from that data, and what these visualizations allowed me to discover about Frances Brooke’s The History of Emily Montagu and the intersection of emotion and time in the characters’ exchange of letters.

Textual Analysis: Voyant Although I initially believed that Voyant had limited potential for the argument I intended to make about The History of Emily Montague, the type of analysis that this program allows is varied.8 The program’s visualization tools provoked ideas that I had not considered before, allowing me to make connections between specific word use in Brooke’s text in ways that could reveal significance. Parts of these ideas come from the interactivity of the site itself. As Geoffrey Rockwell and Stéfan Sinclair suggest in their introduction to Hermeneutica: Computer-Assisted Interpretation in the Humanities, Voyant is designed to be a set of web-based tools for data mining and textual analysis that allow for variation, alteration, and determination: “Voyant mixes tools as panels much like those in a comic book, creating a medley, or commedia, that encourages ‘serious play’.”9 By naming the available tools (Cirrus and TermsBerry makers, trend tracking, and context panels, for example) “embeddable toys,” Rockwell and Sinclair emphasize how versatile Voyant can be for collaborative research, presentations, and pedagogy.10 This adaptability is a beneficial feature of the program, as is its ease of use, particularly after a bit of playing with the options available to 8  Stéfan Sinclair and Geoffrey Rockwell, “Voyant Tools,” accessed August 1, 2018, https://voyant-tools.org. 9  Geoffrey Rockwell and Stéfan Sinclair, Hermeneutica: Computer-Assisted Interpretation in the Humanities (Cambridge, MA: MIT Press, 2016), 11. 10  Rockwell and Sinclair, Hermeneutica, 11.

7  EXPLORING DATA VISUALIZATION 

237

eliminate certain words or pick out the few choice phrases that could highlight the concepts I wanted to understand within the novel (Fig. 7.1). Since I am analyzing the intersections between time and emotions, rather than sorting through all of the terms Voyant identified in the text (5202 unique word forms), I plugged in a series of terms that represent temporality, such as “moment,” “hour,” “day,” “week,” “year,” “past,” and “present,” to name just a few. Like most search engines, Voyant’s algorithm allows for variations in spelling and usage: the search term “day*” would return results for both “day” and “days,” for example. I also entered terms that reflected several words relating to emotion: “happ[iy]*”11 for variations in form of “happy”; “sad”; “melancholy”; “pleas*” for variations relating to both “please” (e.g., “pleasing”) and “pleasure”; “ang*” for “angry” and “anger”; and “love.” Interestingly,

Fig. 7.1  The Voyant dashboard consists of several tools, all of which are interactive and can be altered based on what the researcher wishes to analyze. This image displays the visualizations produced before completing any filtering for common words 11  Voyant’s algorithm will search for derivations of the root term; the options indicated in brackets here limit the search to the terms “happy” and “happi-” without returning derivations of “happen.”

238 

C. A. HOFFMAN

and perhaps not too surprisingly, given the propensity for expressions of feeling to pervade eighteenth-century texts in general, and novels in particular, three of the top five words which appeared most frequently were terms with positive emotional connotation: “pleas*” (136 occurrences), “happ[iy]*” (89), and “love” (85);12 they were followed by “day*” (53) and “time*” (39). The emotional words with negative connotations for which I searched actually appeared less frequently than any of the temporal terms: only “angry” (6) returned a value higher than that of the temporal terms “past” (4), “second” (4), “minute” (3), and “future” (2), with “disgust” (2), “melancholy” (2), “anger” (1), “sad*” (1), and “shame*” (1) all coming in at the bottom of the list (Fig. 7.2). Though frequency counts do not necessarily tell me much about the text, at least in terms of my overarching argument, the fact that “day*” appears more often than any of the other terms I searched, leads me to believe that the characters are writing about relatively short spans of time. Also, the fact that “moment” appears 31 times and is counted as the sixth most frequent term, suggests that there is an element of immediacy to Brooke’s letter writers’ sense of their narrative: they are considering their present (“present” ranks seventh, with 24 occurrences, well above “past” and “future”) when writing. This concern with the “now” is much in line with the general thought in the eighteenth century about epistolarity as a form of “writing to the moment,” which makes it both immediate and somewhat implausible. A deeper engagement with the text and the contexts in which the terms are used would result in a more nuanced analysis, but the information afforded by Voyant certainly provides a place to begin. Voyant’s Trends tool allows the user to compare the places in the text where terms appear. Breaking the text into ten sections, the bar and line graph display compares the frequencies of specific terms (either the top four most frequent terms as determined by the algorithm, or a group of terms chosen by the user). Visualizing the frequency of emotion-related terms throughout the novel, particularly juxtaposed with those of other words, has the potential to illuminate any possible correlations between the terms, either positive or negative. Specifically, such comparison indicates if the increase or decrease in a term’s number of occurrences is 12  A search for variations of “love*” would also have included derivations of “lovely,” which would have shifted the focus of the search away from specifically emotion-related terms. I therefore searched the word “love” only; an additional 12 instances of “loved” occur in the text.

7  EXPLORING DATA VISUALIZATION 

239

Fig. 7.2  Frequency of words related to emotion and time in The History of Emily Montague as determined by Voyant

reflected or not in the number of occurrences of the opposite term. For example, when comparing the trends of the term “desire,” variations of “happy,” “love,” “moment(s),” and variations of “please,” and “present,” Voyant shows that in section 4 “love,” “happy,” and “please,” which are already used more frequently than the other terms, all increase from the previous section, with “happy” showing the largest jump (Fig.  7.3a). “Please” is rated most frequent as it appears through all but section 7, where “love” is used most often, but not as frequently as in section 2. “Moment” and “present” converge in frequency in section 4, as well as in the final section of the text. Neither “moment” nor “desire” appear at all,

240 

C. A. HOFFMAN

Fig. 7.3  (a–c) Trend charts for specific word usage in Brooke’s novel: (a, left) illustrates trends for “moment” with regard to emotion; (b, top right) shows “happy” with the more immediate temporal terms “day” and “present”; (c, bottom right) includes terms that consider longer periods of time

“happy” is at its lowest, and “present” is at its highest in section 3. Combing through the text will demonstrate the places in which these terms appear and provide context for an argument about the plot that may elucidate why these fluctuations occur. Such an examination of the novel in conjunction with word frequencies would allow me, for instance, to identify the letter writers in these sections and parse whether particular writers are consumed with writing about their feelings at any of these specific points in the narrative. A similar comparison performed on “day*,” “happy,” and “present” shows that “happy” dips to its lowest usage in section 9, but increases in the final section of the novel, “day*” jumps to its highest rate at the end of the novel, and “present” remains relatively steady in its middle range compared to the rest of the novel. Though the increased use of “day*” would suggest the writers are potentially concerned with their “now” or

7  EXPLORING DATA VISUALIZATION 

241

their immediate future, the low rate of “present” indicates that they are either counting days passing or referring to a specific day or days. Adding the term “week” to the mix reveals that, although this particular time frame is mentioned at its highest rate in section 9, in section 10 it does not appear at all. The trend for “year” is similar, though reversed for sections 9 and 10 (Figs. 7.3b, c). Despite the text’s denouement and happy ending, the characters appear to speak more about days than years as such, which suggests that their concerns are immersed more in the quotidian than in longevity. There is little more the visualizations can tell us without looking more closely at the text; contexts and narrative still play a role in analyzing the text as a whole, though tools like the ones Voyant offer can provide direction for further investigation, particularly in a long and complex text like The History of Emily Montague. Voyant is not without its drawbacks, however, especially regarding the type of text that concerns me. This eighteenth-century novel is not available in a modern scholarly edition since Mary Ann Edwards’s 1985 text, so the version of the text I uploaded into Voyant comes from a digitized PDF edition of the original, which can often be the only way to access older texts we wish to analyze, particularly if we are engaged with recovery work. Given the type of file I used here, I needed to consider standards of eighteenth-century publication, such as the placement of the first word of the following page below the last line of text on each page, which is termed a catchword. Such a mark would have aided the printer and typesetter in assuring that pages were printed in the correct order, while type was being set on the press; however, Voyant does not recognize this variation in the text. And why should it? As a tool that can deal with a volume of data ranging from a single short poem to a much larger corpus of multiple texts, Voyant would have no need for filtering a single word on an indeterminate number of pages based on an obsolete printing practice. As Rockwell and Sinclair explain, text analysis programs like Voyant recognize terms by “breaking a long string into simpler parts for manipulation and recombination.”13 That is, the computer takes chunks of text and finds similarities between individual words and characters. While a trained scholar recognizes that certain markers in an archaic text are not part of the narrative, the computer does not. The algorithm reads coded tags that identify spaces between words, font changes, punctuation, indications for

 Rockwell and Sinclair, Hermeneutica, 33.

13

242 

C. A. HOFFMAN

dialogue, or paragraph breaks, but has trouble with redundant typographic markers, hyphenations, misprinted text, and contractions.14 The inability of Voyant to determine that the catchword at the bottom of the page of an eighteenth-century text is extraneous is a factor for scholars to consider when using such a program, which relies on word frequency to construct its visualizations. Given the propensity for hyphenated words to be split across pages, e.g. “extremely” appearing as “ex-” and “tremely” at the turn between pages 68 and 69, so that “tremely” is printed as the catchword on the bottom of page 68 (see Fig. 7.4d), Voyant will count the hyphenated pieces as individual words. What results is the identification of terms such as “tremely” and the suffix “-tion” as datapoints. Similarly, the text’s use of the long -s, an archaic typographical practice well-known to scholars of the eighteenth century, renders the word “passed” as “paʃʃed,” which Voyant considers as a separate term. The printing of the novel’s title on the heading of its pages (“THE HISTORY OF” across the even pages and “EMILY MONTAGUE” across the odd) increases the frequency of these words in the text (as seen in Fig. 7.4a). Voyant does not consider the phrase as extra-textual for the purposes of textual analysis. In fact, without manually removing the words “history,” “Emily,” and “Montague,” “Emily” and “Montague” register as the words appearing most often in the text.15 If not for the repetition of the title, this result might have been potentially significant. Instead, each of these examples indicates the care scholars need to take in interpreting the visualizations Voyant creates: if we do not realize where the data we input is flawed, we may obtain results that are based on inaccurate information. It is up to the scholar, then, to determine how to sort through the text on a word by word basis. Should I eliminate “letter” as a word to be considered, since it appears at the beginning of each of the 228 missives that constitute the text? Voyant counts 71 instances of the word in the text, however, so something must account for the discrepancy. A closer inspection of the source text reveals that the typeface includes a space between each character of the term spelled as “L E T T E R” in the heading. So, Voyant’s tabulation of the frequency of this word reflects only the contents of the letters themselves and not the headings “LETTER X” or  Ibid., 33–34.  Such words are termed “stop words,” i.e., words that the algorithm will not consider because they appear too frequently. Other examples are “of,” “to,” “the,” “and,” and variants of the verb “to be.” 14 15

7  EXPLORING DATA VISUALIZATION 

243

Fig. 7.4  (a–d) Voyant produces Cirrus visualizations (word clouds) to illustrate word frequency: (a, top left) illustrates frequency prior to eliminating the novel’s title as stop words, while (b, top right) shows the new frequencies; (c, middle) demonstrates Voyant’s Contexts tool; (d, bottom) shows a mimeographed page from the text, complete with the nonsensical catchword “tremely.”

“LETTER LVI.”16 Scholars routinely make choices like these when analyzing texts, recognizing that, as Rockwell and Sinclair note, “orthography can shift over time and with variations in usage.”17 Such awareness of the potential pitfalls of Voyant as a tool to analyze eighteenth-century texts, particularly works that can only be accessed in digitized PDF editions, does not mean, though, that we should limit our use of the program. Rather, we must simply be judicious in how we interpret the results of Voyant’s analysis or be prepared to weed out specific terms ourselves. Modern PDF readers, such as Adobe Acrobat, allow the user to edit the digitized text. A question to consider might then be: would it save us time 16  Even here, for the sake of convenience, I have transcribed the text as I read it, rather than as it appears on the page; the program would not make such a distinction. 17  Rockwell and Sinclair, Hermeneutica, 34.

244 

C. A. HOFFMAN

Fig. 7.4  (continued)

7  EXPLORING DATA VISUALIZATION 

245

to sort through a text and erase the page headings and printer’s markings in the footers, so that fewer discrepancies in orthography will appear in the text Voyant will analyze?18 Or, would the insertion of the researcher into the text lead to the creation of a separate text, one that can no longer be argued to be the original, and would that affect the way in which scholarship might consider the text as text? Certainly, the search application in Acrobat affords the opportunity to determine word frequency, but it does not produce visualizations that can highlight specific areas of the text that need to be examined. Having a digitized copy of the source text, preferably one of a modern edition, in which an editor has already made decisions regarding text, orthography, and any cruxes apparent between earlier editions could be ideal: the language would be more recognizable to a modern reader and thus the algorithms would work to sort the text into smaller pieces. For now, however, scholars are relegated to utilizing what PDF texts are available, particularly when studying more obscure texts, so they must remain alert to the ambiguities that may arise when a program like Voyant reviews an orthographically archaic text like the one I used for this study of The History of Emily Montague.

Social Network Analysis and Mapping: Palladio Since the textual analysis program did not allow me to consider The History of Emily Montague in a way that rendered temporality visible as a series of chronological events, and I was searching for data visualizations that would represent time passing as part of the epistolary exchange, I considered other available options. The Mapping the Republic of Letters website19 highlights recent scholarship regarding eighteenth-century letter writing practices and social networks and presents data with considerations for timelines and geography. My overarching argument about the correspondence in Brooke’s text maintains that temporality and location are

18  I would like to note that, as this chapter went through the publication process, a very generous anonymous reviewer provided suggestions on how to overcome a problem like this one with a solution I had no idea was possible: save each letter as a separate text and then enter them all as a corpus, rather than enter the novel as a single text. I mention this not only because I wish to express my gratitude to this reviewer, but also to highlight how the learning process continues long after the original work seems over. 19  Mapping the Republic of Letters, Stanford University, accessed August 2, 2019, http:// republicofletters.stanford.edu/.

246 

C. A. HOFFMAN

connected, so I was attracted by the ability of Palladio20 to both visualize the place of writing and to construct a timeline that I could use to correlate the time of a letter’s composition to the emotions expressed by the character writing the letter. One of the biggest differences between using Voyant and Palladio is that Palladio requires the user to input a specific dataset of specific concepts (numbering of letters, letter writer, letter receiver, date written, and GPS coordinates), rather than simply inputting the text itself. This means more initial work for the researcher, who must collect data and compile it in the proper format for uploading to Palladio’s program. As I had already compiled a spreadsheet with the pertinent data for my own use to keep track of the letters in The History of Emily Montague, I could open a new sheet where I copied and pasted the information I needed from my own notes. Any researchers who are approaching such a project fresh, as I would with any future study, would have to consider the specific parameters needed for their data to be used in Palladio. The work can be tedious, but once done, it could easily be transferred when a slightly different organization is demanded by Tableau, the program I will discuss in the next section. Given that the Mapping the Republic of Letters project produced an intricate visualization of a social network, I began my exploration of Palladio’s capabilities with the goal of creating a similar network visualization (Fig. 7.5). The possibility of social network analysis to demonstrate where some of the more complex relationships between characters lay is one that has been pursued by scholars of eighteenth- and nineteenth-century novels in recent years. The value of computer-generated social network analysis stems from the calculations the program makes about the frequency of interactions, the types of interactions, and a variety of other related concepts that are not always apparent to a reader when engaged with the text. By creating so-called “nodes,” a social network visualization will display “multiple fields of all types, with various connectivity attributes that could be spatial, temporal, or dependent on other attributes, such as belonging

20  “Palladio,” Stanford University, accessed March 14, 2019, https://hdlab.stanford.edu/ palladio/. This program was designed at Stanford University for the Mapping the Republic of Letters project and it is offered as a free online resource. The project design team includes Dan Edelstein (Lead Investigator), Nicole Coleman (Project Director), Ethan Jewett (Lead Developer), Eliza Wells (Research Assistant/Documentation Author), Giorgio Caviglia, and Mark Braude.

7  EXPLORING DATA VISUALIZATION 

247

Fig. 7.5  The Mapping the Republic of Letters project explores data visualizations relating to the Enlightenment correspondence, including letters by Franklin, Locke, Voltaire, and others

to the same group or having some common computed values.”21 Visualizations for social networks can highlight the direction of interactions, the weight of those interactions (such as the number of letters one person or character writes, which is represented by the size of the node), and multiplicities of interactions (such as what characters’ paths may link or cross in ways or places that are not always obvious to the text’s reader). Novels like George Eliot’s sprawling Middlemarch (1871),22 where the cast of characters is large and varied, provide fertile ground for the creation of a social network visualization (Fig. 7.6a). I imagined that a similar analysis might prove enlightening when applied to the letter writers in The History of Emily Montague, since Brooke’s number of characters is much larger than that in other epistolary novels. Although the visualization Palladio produced does, in fact, demonstrate the heavier weight Brooke places on her two most prolific writers (Edward Rivers, with 75 letters, and Bell Fermor, with 78, which in total represent 67.1% of the novel’s letters), the network analysis provides little elucidation for the temporal or 21  Matthew O. Ward, Georges Grinstein, and Daniel Keim, Interactive Data Visualization: Foundations, Techniques, and Applications (Natick, MA: AK Peters, 2015), 55–56. 22  George Eliot, Middlemarch, ed. David Carroll (Oxford: Oxford University Press, 2008).

248 

C. A. HOFFMAN

Fig. 7.6  (a–d) A social network visualization for a text with a wide cast of characters can be complex, such as this one, analyzing George Eliot’s Middlemarch, posted on Twitter by Katherine Ognyanova (Katherine Ognyanova, Twitter post, September 22, 2015, 4:44  p.m., https://twitter.com/ognyanova) (a, top left), whereas a similar network visualization for The History of Emily Montague’s letter writers is less revealing (b, top right). Using location of writing is somewhat more interesting in a network visualization (d, bottom right), but not particularly helpful. It also requires a key (c, bottom left) to translate between location name and GPS coordinates, which Palladio requires for geospatial plotting

emotional components of the argument I am attempting to make (Fig.  7.6b). Even when the data ranges shift to reflect dates of writing instead of characters composing the novel’s epistles, the resulting visualization is too scattered to suggest any new interpretative path for me to travel (Fig. 7.6d). It seems that, although its high number of letter writers and letters is unique for the genre, The History of Emily Montague does not provide a large enough sample size to demonstrate any unforeseen connections among characters. The fact that I did not find any new answers related to time and location in the resulting social network visualization does not mean that

7  EXPLORING DATA VISUALIZATION 

249

Palladio is not useful for my argument. Given its ability to use geographical information to create a map visualizing the places from which Brooke’s characters write their missives, Palladio allows the user to visualize the progress of the letters over space. For The History of Emily Montague, this means that I can see the journey made by the correspondence exchanged by Brooke’s characters across the Atlantic, a voyage that, in the eighteenth century, could take anywhere from three weeks to three months depending on the weather. Indeed, part of my argument about the effects of reading a letter written in a remote past once it has finally arrived at its destination revolves around the length of that very journey. Palladio provides a visualization showing the distance between England and Canada, which gives form to the argument I am making. The map in Fig. 7.7 illustrates Ward, Grinstein, and Keim’s description of interactive application: “visualization and data go hand in hand with the goal of building a model that represents or approximates the data.”23 The blue swath between shores composed of shades of green and brown gives the time of movement for the transatlantic correspondence of Brooke’s characters a physical approximation (Fig. 7.7a). The computer-generated ocean represents the time it would take for letters to travel between writers, and for the feelings expressed in them to be written, perhaps forgotten, and then remembered when the reply reminds the writer of what had previously been felt. Representing the gap between characters and what Melissa Gregg and Greg Seigworth term the “in-between-ness” where affect occurs, the expanse of the Atlantic is a rather large space for emotions to fulminate as time passes.24 Though the specific data collected does not include “ocean” as a concrete idea, it is implicit in the intention of the letters exchanged between England and Canada and the notation of the letters’ recipient. The length of the lines between the place of sending and the place of receiving highlights the distance over which missives must move, making the transatlantic lines more prevalent when compared to the much shorter lines between locations such as Montreal and Quebec, or the correspondence between Emily and Bell in adjoining bedrooms in Silleri. Without details about the order of individual letters within the text, I cannot conclude if the brevity of the lines also indicates the pace of the  Ward, Grinstein, and Keim, Interactive Data Visualization, 31.  Melissa Gregg and Greg Seigworth, “An Inventory of Shimmers,” in The Affect Theory Reader, ed. Melissa Gregg and Greg Seigworth (Durham, NC: Duke University Press, 2010), 1. 23 24

250 

C. A. HOFFMAN

Fig. 7.7  (a–c) Palladio’s mapping feature connects places of writing and reading and indicates directionality by the concavity of the line illustrating travel: (a, top) shows Canada and England, while the closer views of the Canadian (b, middle) and English (c, bottom) locations allow for a magnified image of where Brooke’s characters are placed in the novel

7  EXPLORING DATA VISUALIZATION 

251

exchange; knowing that letters 73–79, dated 24 and 25 February 1767, constitute the women’s exchange allows me to argue that its rapidity could simulate the immediacy of in-person conversation, or, potentially, could indicate important content requiring expeditious replies.25 Palladio accounts for the location of the sender and receiver with GPS coordinates and displays the aggregate for each place via weighted circles. It also allows for the use of different colors to represent the writer and reader, in addition to curving lines that indicate the letters’ direction of travel. This feature is highly useful for visualizing place in relation to correspondence, but when several writers are positioned close to each other (as Rivers and Bell are), the weighted indicators will blend together, with the larger obscuring the less prolific writer (Fig.  7.7b). The overlap of circles makes distinguishing between specific writers difficult, though it allows any outliers, particularly those who are located at a distance from the main protagonists, to stand out. One example of an outlier is Bell, who on her journey to England writes the only letter in the novel composed on board a ship; such a letter must travel longer distances to reach its addressee, and thus may affect the reader differently than if it were written from a neighboring city, estate, or bedroom. The program also allows the user to zoom into various locations, so making distinctions between writers is a bit easier. Thus, Palladio’s visualization allows me to see that, given the larger circles in Canada and the convex lines heading east to England, the letters sent from North America take up a larger portion of the overall text. The circles in England seem nearly the same size, and appear to be darker in color, suggesting a similarity in weight; however, once I zoom in, I can see that many letters are sent from outside of London and between two locations very close together (indicating Rivers’s and Temple’s neighboring country estates). As I know that in the final section of the novel, which represents over 23% of the whole text, all the correspondents reside in England, this allows me to argue that, up to letter 175, Canada forms the temporal grounding for the text’s narrative.26 Although Palladio’s Map feature is helpful for visualizing the space between correspondents and the time that would pass between letters written by Rivers in Quebec 25  Only Letter 73 is dated on 24 February; the others are written on 25 February. However, a closer look at the text reveals that Emily notes the time of her writing at “Eleven at night” on Letter 73 and Bell’s response is timed as “Eight o’clock, just up” in Letter 74, which indicates that there is a very brief pause in their exchange, occasioned by the need to sleep. 26  This final section encompasses letters 175–228, that is, a total of 53 letters.

252 

C. A. HOFFMAN

to his sister Lucy in London and her replies to him, the map does not account for the ordering of the letters in the text. Palladio includes a timeline feature which, once the dates of the letters are organized in the fashion the program requires to read them, can produce a chronological visualization to aid in considering how the letters track over calendrical time (Fig. 7.8). This type of organization indicates when letters are dated, rather than their order of appearance within the overall narrative structure. My argument suggests that receiving a letter written several weeks or months prior to the date of reading would pull the reader into an imagined past described by the writer, as well as into the reader’s own remembered past, recalling the events about which the sender had originally written and to which this reply refers. For example, Rivers receives a letter from his sister Lucy and one from his friend Temple, both of whom announce their engagement. The missives are dated 3 January 1767, but they arrive 21 March. Ironically, Lucy’s previous letter, dated 1 December 1766, indicated her disinterest in Temple as a romantic partner. In it, she writes that “my heart is in no danger from a man of his present character” in response to Rivers’s note from 15 October desiring

Fig. 7.8  Palladio’s timeline highlights datapoints in blue according to a designated filter: in this case, the letter writer

7  EXPLORING DATA VISUALIZATION 

253

her to be cautious in her feelings toward his friend.27 However, Rivers does not receive the 1 December letter until mid-February, six weeks after Lucy and Temple have already wed. He must surely have been reminded of Lucy’s disparaging remark in his 20 February reply, where he wrote that he is “very happy in finding you think of Temple as I wish you should,” when reading their felicitous news in March, and probably somewhat abashed to know he had reciprocated in a letter they might both read, and which was written well after their nuptials.28 Palladio produces a timeline that details when the novel’s letters are dated, which is close to what I hoped a visualization might show. However, though the writers’ identities have some color-coding to differentiate them, and the datapoints are weighted as to the number of letters written by each character over a span of four months, Palladio does not indicate the specific date of each letter, nor does it show the letter number. Mousing over the lines on the graph produces a tag with the writer’s names. Thus, while Palladio shows when characters wrote their letters, it does not help me to draw conclusions regarding which letters interject into the broader narrative in ways that provide spots for emotion to swell in response to the contents of these letters, becoming the thrust for change in the characters who read them. Palladio’s geographical and temporal visualization tools provide information for me to make some of the interpretative moves I aim for the project, including the connection between the weight of letters composed in Canada and Brooke’s ordering of these letters, which indicate that the Canadian writers provide the temporal frame in the novel’s structure. But for my analysis to consider both the order of letters in the text and the date on which they were written, I envisioned a different chronological form than Palladio creates. I began to think that, perhaps due to my uncertainty about what exactly data visualizations could show, I might have been asking an impossible question.

Interactive Timeline: Tableau In my search to find a data visualization tool that would allow me to view when the letters in Brooke’s novel were written, I next experimented with the data analytics platform Tableau. This program uses an Excel spreadsheet containing the same data regarding the letters in The History of Emily  Brooke, The History of Emily Montague, 133.  Ibid., 139.

27 28

254 

C. A. HOFFMAN

Montague needed to create a visualization in Palladio, including the letter number in the text, the writer, the reader, the date of writing, and the GPS coordinates for both the place of writing and the destination place. Through trial and error, I determined a configuration of data within the program that would produce a timeline of letter writing where the date of writing was the primary organizational system, rather than the letter writer or the letter number. Tableau constructed a linear temporal visualization that color-coded each datapoint (one missive per point) according to its writer, with a tag for the letter number as it appeared in the text. Tableau allows for visualizing multiple types of data: in addition to the date, place, number, and text tags I mentioned above, the platform can display data regarding counts or specific places, such as cities, airports, or regions. While I found that the Gantt bar-style graph was most informative for my purposes, Tableau’s Workbook screen contains an infographic that highlights distinct types of graphs, charts, or tables based on the type of data provided. Each example not only gives a thumbnail image of a finished visualization but also indicates which and how many measures or dimensions are needed to create that particular graphic. Users can drag and drop the specific data type into the row or column line above the image box and Tableau will create the right visualization to match the data it was given. A bar graph requires a single measure, so dragging the solitary measure provided in my datasheet (letter number) into the row line produced a single vertical bar, while dragging it to the column line resulted in a horizontal line. Tableau assumed the data should be summed until I altered the measure format, indicating the letter number as a count of discrete digits. That the individual categories of data might be manipulated is a potential drawback of Tableau, and it is certainly a feature of the software that should be utilized with caution. However, shifting the format of the date when a letter was written from year alone to month-day-­ year is not only appropriate but crucial to constructing a legible, informative visualization based on the calendar, particularly considering the importance of the actual time of writing in my overarching argument. To add details to the graphic, Tableau allows the user to drag the dimension or measure over to icons named “Color,” “Detail,” “Size,” “Tooltip,” and “Label.” For my purposes, using the letter number as a label for each individual letter and using the color feature to indicate which character wrote each missive prompted the software to pinpoint the “location” along the timeline according to the date a letter was written. Tableau then showed the letter writer via a separate shade for each character. This

7  EXPLORING DATA VISUALIZATION 

255

resulted in a linear graphic where, once a mark on the line was moused over, a pop-up box revealed the name of the writer and the place of the letter within the narrative structure (Fig. 7.9). Adjusting the size of the points on the timeline affords an emphasis for each point and makes them easier to mouse over on multiple letters written on the same date, as does the ability to drag the bottom edge of the graphic down to vertically elongate the visualization. This manipulation of the graphic for ease of viewing is particularly helpful in terms of identifying the outliers, that is, those characters who do not write often, but whose missives contain vital information. Though neither Lucy Rivers nor John Temple write many letters to their correspondents in Canada, the integration of their epistles’ dates of arrival with the fact that weeks or months have passed since their writing is precisely indicated by the function that brings a specific writer’s datapoints forward while greying out the others (Fig.  7.10). The information in Lucy’s letters, for instance, drastically shifts the emotional landscape for both Rivers and Emily, the two

Fig. 7.9  Visualization based on chronological dating of the letters in The History of Emily Montague created with Tableau. The letter that appears as number 35 in the text was written on 23 July 1766, so it appears between letters 6 and 7 on this visualization, allowing for an analysis of how its contents can shift the letter readers’ attention in the narrative from their own time to the letter writers’ past (here, John Temple’s). Letter writers are color-coded and appear in a key to the right of the image. A pop-up box identifies the writer’s name, the letter number, and the date of writing

256 

C. A. HOFFMAN

Fig. 7.10  (a–b) Highlighting a single letter writer indicates the placement of one’s letters within a larger timeline. Here, Lucy Rivers does not write often, but her letters have a great deal of emotional impact (a, top). Temple’s writing is similarly infrequent, but by adjusting the height of the points on the timeline it becomes easier to see a single writer’s contributions

characters whose courtship underpins the narrative, and thus their arrival should be noted within the scope of my broader analysis.29 Tableau cannot analyze the text contained within individual letters; that information is not part of the dataset compiled into the original Excel sheet I uploaded into the program. However, playing with both the 29  Lucy writes of her mother’s hopes for Rivers’ future financial prospects (he has come to Canada because of his father’s profligacy), and Emily decides to sacrifice her possibility of happiness as his wife and leaves him to find a wife with an income, as she has none. So, Emily travels to England but Rivers follows her; this shifts the narrative’s geospatial frame from Canada to England.

7  EXPLORING DATA VISUALIZATION 

257

interactive textual analysis capabilities of Voyant’s word frequency charts and Tableau’s ability to highlight specific writers’ frequency of writing during particular time frames, I could determine which characters have the potential for creating shifts in word usage. For example, if I wished to investigate which writers contribute to the increased frequency of word variants of “please” in the second section of the text, I could do so (Fig. 7.2a). After creating a new workbook in Tableau by using the letter number along the X-axis to display the order in which it appeared in the text, I could determine which writers are active in any given section by highlighting those specific letters and eliminating the rest.30 The visualization I produced for individual sections afforded color-coding for either writer or receiver. I could also label with text each individual letter datapoint to indicate the tag not shown by color. Out of the 23 missives in the second section, Bell and Rivers provide nine letters each, Emily writes three, her ostensible guardian, Mrs. Melmoth, two, and Temple completes the set with the final contribution (Fig. 7.11a). It is likely, therefore, that Bell and Rivers are the characters most concerned with pleasing or being pleased.31 In addition, because Tableau is interactive and provides opportunities to change details of the visualizations it creates, when I switched out “writer” for “receiver” as a detail for the datapoints, Lucy became most prominently highlighted as the addressee of 14 of the 23 letters in this section of the text. Because the very nature of epistolarity necessitates audience awareness in composition, one conclusion to draw from an analysis of these data visualizations is that, in this particular section, Brooke constructs her letter writers as expressing ideas of what is pleasing to Lucy, rather than telling her of the less idyllic aspects of their lives. A deeper look at the text itself, perhaps even a

30  Each section of the text represents 1/10 of the whole; with 228 total letters in the text, one section would be comprised of approximately 23 letters. With this in mind, I broke the text into sections to identify which letters to highlight in Tableau according to the word frequency trend graphs generated by Voyant (section 1: 1–22; section 2: 23–46; section 3: 47–69; section 4: 70–92; section 5: 93–115; section 6: 116–138; section 7: 139–161; section 8: 162–184; section 9: 185–207; and section 10: 206–228). 31  My overarching argument about the text concerns the way that a letter sent from England to Canada brings information that affects its reader emotionally; Temple’s letter, traveling westward across the Atlantic, could instigate a deep emotional response in this paradigm. Although the odds of Temple’s single letter being so immersed in a sense of “please” that it skews the data are extremely low, they should not be ignored.

258 

C. A. HOFFMAN

Fig. 7.11  (a–b) Visualizations of the letters contained in section 2 (a, top) and section 7 (b, bottom) organized by number and color-coded by writer

repetition of the Voyant analysis with a smaller section of the text, could reveal more about how Brooke uses variants of “please” here. Similarly, both “happy” and “moment” have their highest frequency of use in the seventh section of the novel, between letter numbers 139 and 161 (Fig.  7.11b).32 When this span is examined closely, Bell writes and Lucy receives 12 of the 22 letters each. At first, it seems possible that Lucy is the recipient for all of Bell’s letters; this is not the case, however, though nearly so. Emily’s only letter is addressed to Lucy, and Bell’s final missive in this section is written to Temple. Selecting color to indicate the receiver and label to show the writer, along with adding the date of composition to the row option so that it is displayed on the Y-axis of the screen, I created a graph of this chunk of text with enough information to postulate that, in her letters to Lucy between 20 May and 13 July 1767, Bell expresses a sense of exultation regarding her immediate circumstances (Fig. 7.12).33 The text of those particular missives is much concerned with Bell’s decision to marry her suitor and her announcement of the event to her friend, which would, in fact, correspond with the increased use of “happy”  Brooke, The History of Emily Montague, 251–88.  Ibid.

32 33

7  EXPLORING DATA VISUALIZATION 

259

Fig. 7.12  Once date of writing is added to the visualization for section 7, where the letters are organized in the order in which they appear in the text, Lucy’s and Temple’s exchanges clearly stand apart from the others. Color-coding the addressee of the letters and using text to label each letter by writer illustrates which character might be contributing to word frequency in this particular section of the novel

derivations. On the other hand, in Emily’s two letters several mentions of the word “moment” appear in conjunction with the process of composing her missives and her decision to sail from Canada to England during this period in the novel. So, despite Bell’s position as the most prolific letter writer in this section, she is not necessarily the correspondent responsible for the above word frequency results; while Voyant’s and Tableau’s visualizations can indicate possible correlations, determining causation requires returning to the source text for a more accurate analysis. One potential concern when comparing Voyant’s Trends tool with Tableau in the way I have is that, while Tableau organizes the text by letter number, as stipulated in the spreadsheet I entered into the program, Voyant separates the text into sections based on page number within the text. This difference in organizational practices could lead to a misinterpretation of which characters are writing in which sections of the text. However, given the overall frequency with which Brooke assigns letter writing to Rivers and Bell, any section that weighs heavily in either of their

260 

C. A. HOFFMAN

favor is presumably close to accurate in approximating word frequency trends. As such, while Tableau’s graphs could help me make predictions about patterns in Brooke’s construction of her characters’ word choice, Voyant’s word frequency visualizations still necessitate digging deeply into the text to determine the accuracy of those predictions. Although the program itself can be a bit finicky to use for a beginner and is not available as a free online resource, as Palladio and Voyant are, the free two-week trial allowed me to explore the program and produce a visualization similar to what I had imagined, as well as several other configurations that proved informative. A subscription to Tableau may be worthwhile for a university or individual researcher, given its versatility and, with a little trial and error, its relative ease of use.

Conclusion: Textual Analysis and Digital Pedagogy The need to go back to the text to check the results does not suggest that visualizations like the ones I have produced here are unnecessary. Without the graphics provided by these programs, I would not have known, for example, that in section 4 of the text (letter numbers 70–92, Fig. 7.13), derivations of “present,” “desire,” and “moment” appear equally, or that “please,” “happy,” and “love” are all nearly equally present and much more frequent than the former three, with “love” at its highest point in the text. These findings suggest that, in this section, characters are highly concerned with affections and interpersonal relationships (Fig.  7.2a).34 Nor would I have been able to ascertain at a glance that there are five sets of letters written on the same day in this span; one of which is the spate of short notes I mentioned above exchanged by Bell and Emily while they resided in adjoining bedrooms. However, without returning to The History of Emily Montague, the context of any and all of those missives cannot be determined with only these visualizations. Nuance and a sense of connection in the plot on a broader scale than what is essentially a small combination of factors can be lost, even as other connections may be revealed. When, for instance, I removed Captain Fermor and Mrs. Melmoth from the visualization, since they only receive 3 of the 22 letters shown here, I could see that Lucy receives the majority of the correspondence, which is the norm. In contrast, Emily and Rivers wrote an equal number of ­epistles,  Brooke, The History of Emily Montague, 137–69.

34

7  EXPLORING DATA VISUALIZATION 

261

Fig. 7.13  This close-up view of section 4 indicates that several groups of letters were written on the same day within this span of the text, including an exchange between Emily and Bell

which is an unusual occurrence given their overall contributions to the novel’s total number of letters; this indicates that here the narrative is focused on their courtship, rather than on one of the other storylines that are intertwined throughout the text.35 However, unless I delve into the text itself, I cannot draw any valid conclusions about why these connections might be significant. The visualizations I produced with these interactive programs are incredibly useful in pointing toward directions for me to go within the text, but, by themselves, they do not provide adequate interpretations of the text. As with a microscope, I can examine the beauty in the narrower focus these visualizations afford, but the larger picture is much more difficult to find without additional information.

35  Robin Howells argues that the novel is a complex interweaving of citation, interlocution, and inscription when read through dialogism and Brooke uses various discourses to model the fluid and open nature of the epistolary genre in general. For details, see “Dialogism in Canada’s First Novel: The History of Emily Montague,” Canadian Review of Comparative Literature 20 (1993): 437–50.

262 

C. A. HOFFMAN

Voyant, Palladio, and Tableau are all valuable potential teaching tools beyond the applications I have demonstrated here for this epistolary novel. Voyant produces visualizations in under a minute and requires only pasting text or uploading a PDF file into the website, a task that can be accomplished during the first few minutes of a class period. Students can easily see which words or phrases repeat in a long poem, such as Alexander Pope’s Essay on Man or The Rape of the Lock, or James Thomson’s The Seasons, revealing trends across books and cantos, or across an author’s entire oeuvre. The program could prove a useful accessory for an analytical essay, supporting or undermining an initial interpretation, forcing students to look more deeply into the meaning of a text. Analyses of Frances Burney’s massive novels could be enhanced by a collective class project, where students contribute lists of character interactions in designated sections of the novel and enter data into a spreadsheet. Using Palladio, the class can take that spreadsheet’s information and produce a social network visualization which, in conjunction with a Voyant visualization, could elicit predictions regarding character interactions and word usage or tone in various sections of the unwieldy text. At the very least, students could see how a wide cast of characters can intersect in unexpected places. When examining a collection of correspondence such as Lady Mary Wortley Montagu’s, students might construct a map of her travels to and from Turkey in 1717–1718, or throughout Europe in her later years. They might also construct her social network, given how prolific a correspondent Wortley Montagu was, by identifying the addressees of her letters. A timeline such as the one Tableau can produce might illustrate changes or correlations in subject matter when she corresponded with certain people. With multiple students contributing to the construction of the dataset, the workload is lessened for everyone, and if the specific parameters required are determined before the work begins, or are easily revised during the data collection process, the result will be a visualization the entire class can admire and interpret. Given that Voyant and Palladio are open access resources and relatively easy to use, they allow the students to play with the various features of these programs. A prospect such as this means that data visualizations can aid scholars in both their research and their teaching, bringing digital humanities into the classroom in conjunction with—rather than abandoning or replacing—traditional methods of textual analysis.

7  EXPLORING DATA VISUALIZATION 

263

Bibliography Altman, Janet Gurkin. 1982. Epistolarity: Approaches to a Form. Columbus: Ohio State University Press. Arch, Stephen Carl. 2004. Frances Brooke’s ‘Circle of Friends’: The Limits of Epistolarity in The History of Emily Montague. Early American Literature 39 (3): 465–485. https://www.jstor.org/stable/25057366. Brooke, Frances. 1985. The History of Emily Montague, ed. Mary Jane Edwards. Don Mills: Carleton University Press. Burney, Frances. 1998. Evelina, or, the History of a Young Lady’s Entrance into the World, ed. Stewart J. Cooke. New York: W.W. Norton. Eliot, George. Middlemarch, ed. David Carroll. Oxford: Oxford University Press. Gregg, Melissa, and Greg Seigworth. 2010. An Inventory of Shimmers. In The Affect Theory Reader, ed. Melissa Gregg and Greg Seigworth, 1–25. Durham: Duke University Press. Howells, Robin. 1993. Dialogism in Canada’s First Novel: The History of Emily Montague. Canadian Review of Comparative Literature 20 (3/4): 437–450. Mapping the Republic of Letters. Stanford University. http://republicofletters. stanford.edu/. Accessed 2 Aug 2018. “Palladio.” Stanford University. https://hdlab.stanford.edu/palladio/. Accessed 14 Mar 2018. Richardson, Samuel. 2004. Clarissa, or the History of a Young Lady, ed. Angus Ross. London: Penguin. ———. 2008. Pamela; or, Virtue Rewarded, ed. Thomas Keymer and Alice Wakely. Oxford: Oxford University Press. Rockwell, Geoffrey, and Stéfan Sinclair. 2016. Hermeneutica: Computer-Assisted Interpretation in the Humanities. Cambridge, MA: MIT Press. Sinclar, Stéfan, and Geoffrey Rockwell. 2018. Voyant Tools. https://voyant-tools. org. Accessed 1 Aug 2018. Ward, Matthew O., Georges Grinstein, and Daniel Keim. 2015. Interactive Data Visualization: Foundations, Techniques, and Applications. Natick: AK Peters.

CHAPTER 8

Outliers, Connectors, and Textual Periphery: John Dennis’s Social Network in The Dunciad in Four Books Ileana Baird

While reports on large ongoing projects involving the use of data visualizations in eighteenth-century studies have started to emerge in recent years,1 mainly due to primary texts becoming accessible through digitization processes and data-sharing initiatives, less focus has been put so far on the potential for data visualization to unveil new information about particular texts, literary or not. The reasons are quite obvious: the texts in question should be structurally or stylistically complex enough to render such an analysis valuable. In other words, looking at a text’s  Important book-length publications include Chloe Edmondson and Dan Edelstein, eds., Networks of Enlightenment: Digital Approaches to the Republic of Letters (Liverpool: Voltaire Foundation in association with Liverpool University Press, 2019); and Simon Burrows and Glenn Roe, eds., Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies (Liverpool: Voltaire Foundation in association with Liverpool University Press, 2020). 1

I. Baird (*) Zayed University, Abu Dhabi, United Arab Emirates e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_8

265

266 

I. BAIRD

argument, organization, or style from a quantitative perspective—whether to identify hidden patterns of relationships among characters, map narrative trajectories, decipher obscure allusions, spot stylistic differences, or attribute authorship—requires a text that is complex or contentious enough to justify the use of forensic linguistics, stylometry, or other forms of data analysis involving quantitative methods. Unsurprisingly, such texts are rare: one example is Samuel Johnson’s Dictionary, which has been recently analyzed in light of the knowledge networks it creates through the use of quotations and its role in the process of canon formation.2 Another example is Samuel Richardson’s Clarissa, where the author’s ability to create distinctive character writeprints has been tested through the use of machine learning techniques.3 Another text extraordinarily rich in potential is, I argue, Alexander Pope’s mock-epic poem The Dunciad in Four Book (1743), whose comprehensive attack on a large number of reallife characters, sophisticated use of annotations, and multiple revisions over a period of seventeen years make it a perfect case study for quantitative analyses. The Dunciad remains, to date, the poem with the largest number of identifiable characters in British literature: dozens of dunces inhabit its spaces creating, through the mere frequency of the names provided, the strong impression that the individuals mentioned in the poem are not important as real characters, but as pieces in an intricate mechanism of cultural reassessment. To complicate things even more, many of Pope’s dunces change from one edition to another, are obscure individuals, or are difficult to identify due to the author’s use of sobriquets. Moreover, because of his epic’s allusive mood and playful competition between the poem and the apparatus, his dunces’ affiliations and the motivations of Pope’s attacks are many times unclear. Jonathan Swift’s concern with the Dunciad’s indecipherability4 echoes, therefore, the problems faced by current readers, who need to approach the text with a key to its social and plot networks to clarify its meanings. 2  Mark Algee-Hewitt, “The Principles of Meaning: Networks of Knowledge in Johnson’s Dictionary,” in Edmondson and Edelstein, Networks of Enlightenment, 251–77. 3  Lisa Pearl, Kristine Lu, and Anousheh Haghighi, “The Character in the Letter: Epistolary Attribution in Samuel Richardon’s Clarissa,” Digital Scholarship in the Humanities 32, no. 2 (2017): 355–76, https://doi.org/10.1093/llc/fqw007 4  “I have long observ’d that twenty miles from London no body understands hints, initial letter, or town-facts and passages; and in a few years not even those who live in London.” See George Sherburn, ed., The Correspondence of Alexander Pope (Oxford: Oxford Clarendon Press, 1956), 2: 504–505.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

267

One solution to this dilemma is, of course, punctilious footnoting. Edmund Curll was the first one to annotate the poem in his Key to the Dunciad, published in June 1728. The Scriblerians followed suit, experimenting playfully with ersatz and pedantic footnoting in The Dunciad Variorum (1729). Inspired by the 1716 Geneva edition of Boileau’s Works, Pope split his footnotes in Remarks, containing commentaries by “Modern” critics, and Imitations, containing quotes from ancient or contemporary authors whose work he admired. In the 1751 edition of Pope’s Works, William Warburton introduced a third category of notes, Variations, which recorded Pope’s successive revisions of The Dunciad.5 As pointed out by current criticism, Pope’s use of notes served not only satirical but also historical and explanatory functions.6 They were meant to ridicule authors and critics who fell short of talent and wit, to set better examples in the “Ancients”’ texts, and to elucidate the context of his war with the dunces. As such, these footnotes do provide the necessary “keys” to contextualize his defamatory assaults, with only one major problem: they are utterly unreliable. The mini-narratives they provide are contradictory, incomplete, and obviously biased; they shed only the right amount of light on events to raise questions about his dunces’ competence, but rarely provide the full context of these events. They involve authorial sympathies or antipathies, heated debates, and scandalous abuse; in short, they function as a lively replica of the disputes occurring in eighteenth-century coffeehouses and city streets. The motivations of The Dunciad’s attacks on particular characters become, therefore, even more difficult to untangle under the weight of these “clarifications” of Pope’s authorial intentions. To address this issue,

5  The poem went through three major revisions, thirty-three separate editions, and about sixty impressions by 1751, the year of Warburton’s posthumous edition of Pope’s Works. The changes it encountered ranged from paratextual additions (footnotes, illustrations, front and back matter) to more significant revisions, such as the replacement of the initial hero of the poem, playwright Lewis Theobald, with actor and poet laureate Colley Cibber, in The Dunciad in Four Books. For details, see David L. Vander Meulen, Pope’s Dunciad of 1728. A History and Facsimile (Charlottesville: University of Virginia Press, 1991), 3–72. 6  Harold Weber, “The ‘Garbage Heap’ of Memory: At Play in Pope’s Archives of Dulness,” Eighteenth-Century Studies 33, no. 1 (1999): 15, https://doi.org/10.1353/ecs.1999.0060. For insightful analyses of Pope’s footnotes, see also Anthony Grafton, The Footnote: A Curious History (Cambridge, MA: Harvard University Press, 1997), 23, and Chuck Zerby, The Devil’s Details. A History of Footnotes (New York: Simon & Schuster, 2002), 45–58.

268 

I. BAIRD

modern editions of the 1729 and 1743 Dunciads contain additional layers of notes to the text and to the Scriblerian footnotes themselves.7 I would like to call attention, therefore, to another way of approaching this text that could help in navigating its complicated argument: reading Pope’s satire by using the new quantitative methods afforded by digital technologies. This method involves the use of social network analysis to represent the text as a collection of asserted links between various character names. Rather than focusing on clarifying information gaps and/or allusions to real life events, as the critical notes to Pope’s satirical footnotes do, this approach X-rays the poem, focusing, instead, on character relationships as they emerge from both text and apparatus (front and back matter, footnotes, headpieces, illustrations). This type of analysis has different goals than untangling the poem’s annotative maze: it uncovers more obscure relationships of hostility or alliance, makes evident the main targets of Pope’s satire, identifies the protagonists who act as “connectors” or “outliers” in the social network of the poem, helps elucidate authorial intentions not clearly spelled out by the text or the apparatus, and suggestively illustrates the magnitude of Pope’s war with his dunces. As such, this approach can augment discussions about  political leanings, gender bias, promotion practices, and canon ­formation—all issues of paramount importance for the construction of publicness as a moral and political category at the beginning of the eighteenth century. After a brief description of this digital project, I will focus here on the networks of relations that involved John Dennis, the most important critic of the first half of the eighteenth century and one of the most vituperative attackers of Pope’s work. The analysis of Dennis’s social network will shed new light on the importance Pope assigned to the critic in his satire, the competing stories told by the poem and the apparatus, and Dennis’s role as a connector within the highly networked public of early eighteenthcentury London.

 I refer here to Valerie Rumbold’s exemplary editions: Alexander Pope, The Dunciad in Four Books (New York: Pearson, 1999), and The Poems of Alexander Pope. Volume III: The Dunciad (1728) & The Dunciad Variorum (1729) (New York: Pearson, 2007). All the quotes used in this chapter come from the former edition. 7

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

269

Methods and Goals As a Social Sciences, Humanities and Performing Arts Network of Technological Initiatives (SHANTI) fellow at the University of Virginia, I had the opportunity of working with Rafael Alvarado, at the time Associate Director of the program, on a project involving the visualization of the fields of relations of six of Pope’s dunces: Colley Cibber, the last King of Dunces, actor, theater manager, and Poet Laureate to King George II; Edmund Curll, notorious bookseller and publisher epitomizing the unscrupulous rush for profit of the print market of the time; Eliza Haywood, remarkable actress, novelist, playwright, and periodical publisher; John “Orator” Henley, a famous cleric and one of the most controversial public figures of Pope’s time; John Dennis, the leading critic of the period; and Giles Jacob, legal author with literary ambitions who consistently supported Dennis in his attacks against Pope. These dunces were selected based on their representative value: they all inhabit key areas of the political, cultural, and religious life of early eighteenthcentury London. The text I used was the 1743 edition of the poem, The Dunciad in Four Books, which contains Pope’s latest changes in characters and textual revisions. I started by capturing all the information relevant to each of these six characters in a spreadsheet indicating the dunce’s name, his or her location within the text (i.e., the poem or the apparatus), the textual reference to the dunce (e.g., “and all the mighty Mad in Dennis rage …”), the individuals the character is linked to, the nature of their affiliation (similarity, dissimilarity, character attacked, or character defended), alternative names or references used by the poet to identify the character (i.e., Dennis is also referred to as “Furius” or “a dry old gentleman”), and a brief description of the character’s role within the poem (critic, poet, journalist, etc.) (Table 8.1). When the protagonist is mentioned in the main text, the exact quote is recorded in bold to account for the number of instances he or she appears in the poem versus the apparatus (e.g., in Dennis’s case, the ratio is 3:92). By using a simple script (Perl), we then converted this spreadsheet into a series of graph data structures that could be interpreted by a graph visualizer.

270 

I. BAIRD

Next, we generated images of the social networks in which these characters were involved by using two tools: GraphViz and ShivaGraph. GraphViz is an open source visualization software that creates topological graphs from sets of dyadic relations. Each graph represents a view of the social network data expressed directly and indirectly in Pope’s Dunciad, as a neato, circo, or dot algorithm.8 In the GraphViz visualizations, the relationships among characters are color-coded: green edges indicate similarity (i.e., Eliza Haywood is described as similar to classicist author and translator Anne Tanneguy-Le Fèvre Dacier9), dotted red edges indicate dissimilarity (i.e., Besaleel Morris, William Bond, and John Durant Breval are less skilled than William Congreve, Joseph Addison, and Matthew Prior, whom they try to emulate), red edges indicate character attacked (i.e., Dennis attacked Pope), and dotted green edges indicate character defended (i.e., Giles Jacob defended Dennis during his quarrel with Pope).10 Shiva stands for SHANTI Interactive Visualization Application. This application can be used to create charts, graphs, tables, maps, and other data representations that help visualize large sets of networks and navigate through them as through a map. The Shiva graph shows 8  Neato are spring-model layouts of undirected graphs (i.e., graphs in which all nodes are connected, and the edges are bidirectional); this is the default tool to use if a graph is not too large (about 100 nodes) and when trying to identify high or low energy configurations (in our case, main nodes/connectors vs. peripheral characters). Circo are circular layouts of undirected graphs; they are very useful in highlighting relationships between agents/objects or positions within a network (in our case, the main targets of Pope’s attacks, or the poem’s “hall of infamy,” and the main targets of his dunces’ attacks, or the poem’s “hall of fame”). Dot graphs are hierarchical, or layered drawings of directed graphs (i.e., graphs in which all the edges are directed from one node to another); this is the default tool to use if edges have directionality, such as in dependency trees (in our case, they highlight Dennis’s relationships with characters who are not connected with each other in his social network). 9  For an explanation of this surprising relation of similarity and an assessment of Haywood’s role in the poem, see Ileana Baird, “The Strength of Weak Ties: Eliza Haywood’s Social Network in The Dunciad in Four Books (1743),” ABO: Interactive Journal of Women in the Arts, 1640–1830 9, no. 2 (2019): 1–36, https://doi.org/10.5038/2157-7129.9.2.1202 10  These four categories describe in a more nuanced way than in prior critical assessments the nature of the relationships in which Pope’s characters are involved. While the text clearly indicates relations of antagonism or support (character attacked vs. character defended), it also depicts more subtle relations of similarity or dissimilarity that do not necessarily involve a direct attack against, or support for, a particular individual.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

271

character relationships as well (green/similarity, orange/dissimilarity, red/character attacked, blue/character defended), but it is interactive, indicating, for instance, the number of connections of a particular character when clicking on his or her name. We have created, thus, social network graphs that make visible what Alex Woloch describes as the “space of the protagonist”:11 one’s network of relations with the other characters in the poem, and the nature of these affiliations. Although it is important to recognize that, just like the maps of large geographical areas, some of these graphs are difficult to read on a computer screen due to the amount of data they represent, they still provide invaluable insights into the argument of the poem. As demonstrated below, this model of social network analysis brings to light data that is structurally embedded in the poem but not obvious or immediately legible given the amount and complexity of information. Our initial goals were (a) to clarify the relationships of these particular dunces with each other and with Pope and his friends, allies, or defenders; (b) to make visible networks of relations that could bring to the surface, in Franco Moretti’s words, “hidden patterns”12 of contacts and exchange, and (c) to assess the magnitude and social ramifications of Pope’s attacks. Our final results led, excitingly, to much more than that.

John Dennis in the Poem’s Plot Network Pope’s quarrel with John Dennis, the most important critic of the time, illustrates in an exemplary way the ramifications of the literary feud started by The Dunciad with key personalities of London’s cultural life. England’s leading critic of the first decade of the eighteenth century, political pamphleteer with strong Whig sympathies, author of moral and religious tracts, and playwright with a constant presence on the English stage, Dennis is one of the most prominent public personalities drawn into the whirlpool of Pope’s satire. One of the most vituperative attackers of Pope’s 11  Alex Woloch, The One vs. the Many: Minor Characters and the Space of the Protagonist in the Novel (Princeton, NJ: Princeton University Press, 2003). 12  Franco Moretti, Graphs, Maps, Trees: Abstract Models for a Literary History (New York: Verso, 2005), 54.

272 

I. BAIRD

work, becoming, as Samuel Johnson put it, “the perpetual persecutor of all his studies,”13 Dennis is also the perfect victim of Pope’s elaborated machinations, which ended up in moral disrepute and severe literary effacement.14 None of Pope’s dunces—other than Colley Cibber—were subject to a more complicated strategy of ridicule and disparagement in his satire than John Dennis, which is a clear indication of Dennis’s prominence in London’s cultural life. A brief account of Dennis’s conflict with Pope will help clarify some of the references to the critic in the poem. Their quarrel started in 1711 with Pope’s biting (and allegedly unprovoked) remarks in An Essay on Criticism, which ostensibly baffled Dennis: There well, might Criticks still this Freedom take; But Appius reddens at each Word you speak, And stares, Tremendous! With a threatening Eye, Like some fierce Tyrant in Old Tapestry!15 (l. 584–587)

The reference to Dennis as Appius is inflammatory: it alludes to his failed play, Appius and Virginia (1709), remembered today only because of Dennis’s invention of a device that imitated the sound of thunder for use as a stage effect. The causes of Pope’s attack are uncertain; as Dennis mentioned in his Remarks upon the Dunciad, he had only met Pope three times before this attack and they never had any disagreements, although the fact that Dennis ignored Pope’s Pastorals (1709) might have precipitated the young poet’s rancor.16 In response, Dennis rushed to publish Reflections, Critical and Satyrical, upon a Late Rhapsody, Call’d, An Essay

13  Samuel Johnson, “Pope,” in The Works of Samuel Johnson, LL.D. (New York: John Dearborn, 1832), 2:236. 14  Avon Jack Murphy, John Dennis (Boston: Twayne Publishers, 1984), 62. 15  Alexander Pope, An Essay on Criticism (London: Printed for W. Lewis, 1711), 34. 16  In Epistle to Dr. Arbuthnot, Being the Prologue to the Satires (1735), Pope seems to point to Dennis’s disapproval of his Pastorals: “Soft were my numbers; who could take offence, / While pure description held the place of sense? / Like gentle Fanny’s was my flowery theme, / A painted mistress, or a purling stream. / Yet then did Gildon draw his venal quill; / I wished the man a dinner, and sat still. / Yet then did Dennis rave in furious fret; / I never answered I was not in debt” (l. 147–154). Another reason for Pope’s attack may be Dennis’s insults aimed at William Walsh, Pope’s poetic mentor. See Johnson, “Pope,” 1: 227.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

273

upon Criticism (1711),17 in which he taunted Pope for his deformity, stating, among other things, that he is “as stupid and as venomous as a hunchback’d Toad,”18 and accusing him of Jacobitism. Pope hit back with The Critical Specimen (1711), where he dubbed Dennis “Rinaldo Furioso, Critick of the Woeful Countenance” (another allusion to one of Dennis’s dramatic failures, Rinaldo and Armida), and with The Narrative of Dr. Robert Norris (1713), a veiled defense of Joseph Addison’s Cato, which had been virulently attacked by Dennis in the previous year.19 In 1715, the Preface to John Gay’s What D’Ye Call It satirized “classical” critics and Dennis’s doctrine of poetic justice, a critique in which Dennis thought he discerned Pope’s hand. After this attack, Dennis joined Curll in his disparaging campaign against Pope, and published (anonymously) A True Character of Mr. Pope, and His Writings (1716), a venomous pamphlet that described Pope as a “little, but very comprehensive Creature, in whom all Contradictions meet”: Pope was “a Beast and a Man,” “a Whig and a Tory,” “a Rhimester without Judgement or Reason,” “a Critick without Common Sense,” “a Jesuistical Professor of Truth,” “a lurking way-laying Coward, and a Stabber in the Dark,” and a “Traytor-Friend,” among other qualifications.20 Pope struck back in the collectively authored Three Hours after Marriage (1717), where Dennis appeared briefly as “Sir Tremendous Longinus.”21 Following this bout, Dennis’s anti-Pope attacks became more comprehensive: in 1717, he published Remarks upon Mr. Pope’s Translation of Homer, in which he criticized Pope’s translation of the Iliad for its many “blunders” and “errors,” Windsor Forest for “want

17  John Dennis, Reflections, Critical and Satyrical, upon a Late Rhapsody, Call’d, An Essay upon Criticism (London: Printed for Bernard Lintott, 1711). 18  Ibid., 26. 19  Interestingly, Addison denied any involvement in Pope’s attack against Dennis, and Dennis confessed that he wrote the attack against Addison at publisher Bernard Lintot’s request, who, in turn, had been persuaded by Pope to invite Dennis’s contribution. 20  [John Dennis], A True Character of Mr. Pope, and His Writings. In a Letter to a Friend (London: Printed for S. Popping, 1716). For Dennis’s authorship of this piece, see Edward N. Hooker’s compelling argument in “Pope and Dennis,” English Literary History 7, no. 3 (1940):188–98, https://doi.org/10.2307/2871490 21  John Gay, Three Hours after Marriage: A Comedy (London printed; reprinted in Dublin by S. Powell, 1717).

274 

I. BAIRD

of plan,” and The Temple of Fame for being a burlesque imitation of Chaucer’s House of Fame.22 Their conflict apparently cooled down during the next decade, when Pope subscribed to Dennis’s Select Works and his volumes of Letters. In 1727, though, Pope published Peri Bathous, or the Art of Sinking in Poetry, an upside-down Longinian treatise in which Dennis was, again, one of his favorite targets: here, he included the critic among “porpoises,” which “put all their Numbers into a great Turmoil and Tempest, but whenever they appear in plain Light … they are only shapeless and ugly Monsters.”23 The 1728 Dunciad contained, again, derogatory references to Dennis, but Pope’s most damaging attack against the critic followed the next year, in the much expanded Dunciad Variorum, where Dennis and Giles Jacob were turned into the main protagonists of the poem’s sub-textual debate. Dennis retaliated with Remarks on Mr. Pope’s Rape of the Lock. With a Preface Occasion’d by the Late Treatise on the Profound, and the Dunciad (1728), a series of seven letters in which he critiqued systematically Pope’s composition, characters, machines, sentiments, and style, and whose Preface contained a severe indictment of “Mr. A. P__E”’s literary posterity.24 The following year he elaborated on his previous attack in Remarks upon Several Passages in the Preliminaries to the Dunciad … and upon Several Passages in Pope’s Preface to His Translation of Homer’s Iliad, a pamphlet dedicated to Lewis Theobald, The Dunciad’s first King of Dunces. Pope was described here as a “scandalous Author,” showing a “monstrous and impudent Vanity,” and as “an empty … impudent Scribler” whose “Pericranium is … much out of Order.”25 As regards The Dunciad itself, Dennis developed here his famous argument about the poem’s lack of action and passivity of its hero, provided a harsh

22  John Dennis, Remarks upon Mr. Pope’s Translation of Homer, with Two Letters Concerning Windsor Forest, and the Temple of Fame (London: Printed for E. Curll, 1717). 23  Peri Bathous, or The Art of Sinking in Poetry, in Miscellanies. The Last Volume (London: Printed for B. Motte, 1727), 27. 24  John Dennis, Remarks on Mr. Pope’s Rape of the Lock. In Several Letters to a Friend. With a Preface Occasion’d by the Late Treatise on the Profound, and the Dunciad (London: Printed for J. Roberts, 1728). Dennis’s prediction of Pope’s posterity is, indeed, disparaging: “For I will venture to affirm, that Mr. A. P__E has no Admirers among those who have Capacity to discern, to distinguish, and judge; and I will venture to foretell, that Time will make this Affirmation good” (Preface, vi). 25  Ibid., 5–10.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

275

comparison between Pope’s and Boileau’s satirical methods, and elaborated extensively on the motives of his own inclusion in the poem. What in The Dunciad caused Dennis’s vehement response to Pope’s satire? It is obvious for anybody looking closely at the poem’s plot network—that is, the story told by a character’s associations rather than the story told by the poem itself—that much of Dennis’s conflict with Pope is included, albeit in a distorted form, in the poem and apparatus. Interestingly, the poem itself is scarce in references to Dennis, who is mentioned in only three brief instances. First, in Book I, l. 106, Dennis is described as one of the deranged specimens of the “Grub-street race”: “She [Dulness] saw slow Philips creep like Tate’s poor page, / And all the mighty Mad in Dennis rage.” Then, in Book II, Dennis is one of the participants in Dulness’s fourth game, the noise competition, which is a hint at his opinionated and highly reactive nature: ’Twas chatt’ring, grinning, mouthing, jabb’ring all, And Noise and Norton, Brangling and Breval, Dennis and Dissonance, and captious Art, And Snip-snap short, and Interruption smart, And Demonstration thin, and Theses thick, And Major, Minor, and Conclusion quick. Hold (cry’d the Queen) a Cat-call each shall win; Equal your merits! Equal is your din! (The Dunciad in Four Books, II, l. 237–244)

The lines describe in a brilliant way Dennis’s critical method, so often displayed in his attacks against Pope: while Dennis’s “demonstration” of The Dunciad’s libelous nature, for instance, contains attentive analyses of some of its prefatory material (such as William Cleland’s Letter to the Publisher), it constantly fails to provide a serious close reading of the poem itself, thereby undermining its own purpose. Instead, Dennis’s prose abounds in “interruptions”—ranging from analyses of classical authors’ superiority over Pope to detailed explanations of Pope’s false accusations against Dennis—, or in nasty personal attacks against Pope formulated in an often suburban lingo. Dennis’s conclusions come, indeed, too “quick” to be accepted at face value, and much of his argument is seriously undermined by the offensive language that permeates throughout.

276 

I. BAIRD

Finally, the last reference to Dennis in the main text occurs in Book III, where he is paired as “fool with fool” (l. 176) with Charles Gildon, a minor playwright and critic whom Pope suspected to have written with Dennis the infamous True Character of Mr. Pope and His Writings. Dennis’s reaction to this association is, indeed, foolish: in his Remarks on the Dunciad, he dismissed any collaboration with Gildon, admitting implicitly the sole authorship of the libel.26 Dennis is thus presented in the poem as an unreliable critic, a lunatic, a libeler, and a fool, in short, as an inconsequential figure with a brief but thunderous presence on the poem’s stage. The details of these qualifications are significantly expanded in footnotes and apparatus, where the critic becomes, surprisingly, the most vocal of Pope’s dunces. The story told by the paratext is, indeed, quite different from the story told by the poem: with 92 references in the apparatus (without counting the illustrations, Index of Persons, and Index of Matters), Dennis becomes the uncrowned king of The Dunciad’s textual periphery, being placed at the center of a dense network of relations that disclose much of the political and moral motivations of his war with Pope. Dennis is also showcased in three of The Dunciad illustrations, more than any other dunce of the poem (Figs. 8.1, 8.2, and 8.3).27 Again and again, his associations indicate Dennis’s pivotal role in some of the most heated debates of the time: Curll’s defamatory campaign against Pope, Jeremy Collier’s pamphlet war against the “profaneness” of the stage, the arguments about the institution of laureateship, the “Ancients” versus “Moderns” debate, the defense of the classical rules and the sublime in art, the influence of political and social conditions upon the production of letters, the role of religion as a social unifier, and so on. Dennis’s name appears eight times in the List of Abusers alone, which is a clear indication of his central role in dunces’ campaign against Pope. Both lists of works attacking Pope published before and after The Dunciad, 26  “As to my writing in concert with Mr. Gildon, I declare upon the honour and word of a gentleman, that I never wrote so much as one line in concert with any one man whatsoever.” Dennis, Remarks upon the Dunciad, 50; qtd. by Pope in Testimonies. 27  For more details on the satirical role of The Dunciad illustrations, see Ileana Baird, “Visual Paratexts: The Dunciad Illustrations and the Thistles of Satire,” in Book Illustration in the Long Eighteenth Century: Reconfiguring the Visual Periphery of the Text, ed. Christina Ionescu (Newcastle upon Tyne: Cambridge Scholars Publishing, 2011), 329–66.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

277

Fig. 8.1  The owl frontispiece to The Dunciad: An Heroic Poem, In Three Books (1728) featuring Dennis’s Works. (Courtesy of Albert and Shirley Small Special Collections Library, University of Virginia)

278 

I. BAIRD

Fig. 8.2  Cover page of The Dunciad, Variorum. With the Prolegomena of Scriblerus (1729), containing the famous image of the ass carrying the dunces’ productions. Dennis’s Works are showcased here alongside those of Leonard Welsted, Ned Ward, Lewis Theobald, John Oldmixon, and Eliza Haywood. (Courtesy of Professor David Vander Meulen, University of Virginia)

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

279

Fig. 8.3  First page of The Dunciad, Variorum. With the Prolegomena of Scriblerus (1729), containing a frontispiece showing an owl wearing a fool’s cap, a possible impersonation of critic John Dennis, aka “Furius.” (Courtesy of Professor David Vander Meulen, University of Virginia)

280 

I. BAIRD

for instance, open with pamphlets by Dennis (Reflections on Essay on Criticism and An Essay on the Dunciad, respectively). This highlights his severe reaction to writings containing personal offense, an attitude considered by Pope unsuitable for a critic and used to his own advantage every time he launched a new offensive against his dunce. Theobald’s branding of Dennis as “Furius” in the Censor of 5 January 1717 describes the critic perfectly: in Appendix VIII alone (A Parallel of the Characters of Mr. Dryden and Mr. Pope), Dennis’s artillery aims at Pope no less than seventeen times, accusing him of being “a mortal enemy to his country,” “a popish rhymester,” “an incompetent translator,” an ape, an ass, a frog, a coward, a knave, a fool, and “a little abject thing.” In addition, the notes to the poem abound in Dennis’s remarks on Pope’s physical deformity, (lack of) education, his “depravity of genius and taste,”28 his want of genius or admirers, and his substandard knowledge of English and Greek. Thus, Dennis’s omnipresence in the footnotes to The Dunciad and in the apparatus construct the poem’s paratext as a space ruled by the furious madness of incompetent critics.

John Dennis’s Social Network Despite general agreement on Dennis’s role in “establishing the profession of criticism in England,”29 recent assessments of his legacy have focused disproportionately on his place in The Dunciad and the motivations of Pope’s attack against him.30 This led critics like John Morillo, for instance, to conclude that “[l]ike all of the other hapless writers entombed  Pope, “Testimonies of Authors,” Dunciad in Four Books, 65.  Paul D. Cannan, “John Dennis,” The Encyclopedia of British Literature 1660–1789, ed. Gary Day and Jack Lynch (Oxford: Oxford University Press, 2015), 1: 407. 30  See, for instance, John Morillo, “John Dennis: Enthusiastic Passions, Cultural Memory, and Literary Theory,” Eighteenth-Century Studies 34, no. 1 (2000): 21–41, https://doi. org/10.1353/ecs.2000.0063; Kathrine Cuccuru, “That ‘Tremendous’ Mr. Dennis: The Sublime, Common Sense, and Criticism,” in Passions, Sympathy, and Print Culture: Public Opinion and Emotional Authenticity in Eighteenth-Century Britain, ed. Heather Kerr, David Lemmings, and Robert Phiddian (London: Palgrave Macmillan, 2016), 105–21; and, more recently, Philip Smallwood, “Petty Caviller or ‘Formidable Assaillant’? Johnson Reads Dennis,” The Cambridge Quarterly 46, no. 4 (2017): 305–24, https://doi.org/10.1093/ camqtly/bfx025 28 29

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

281

in the Dunciad, Dennis has survived primarily as the butt of Pope and his fellow Scriblerians’ jokes or, at best, as a minor figure requiring the stronger ally of a canonized author to gain entrance into modern criticism.”31 As I will try to suggest, however, this view goes against Pope’s own assessment of the critic’s role in the cultural landscape of his time. While acknowledging that Dennis figures “a great deal” in Pope’s mock epic, “especially in its sarcastic footnotes,”32 current criticism has failed to recognize his position as a main protagonist of Pope’s satire, nor has it addressed the significance of his unequal presence in the poem and the apparatus. Therefore, in the following analysis I will change focus from the history of his conflict with Pope to the way in which the poet constructs Dennis in the poem through his character associations. This change in perspective will allow for a more attentive investigation of the critic’s role in the poem and, implicitly, of Pope’s perception of Dennis’s sphere of influence outside the poem during the genesis of his mock-epic. Dennis’s social network, as extracted from Pope’s satire, includes many of the characters involved in the anti-Pope campaign mentioned above. As such, the visualizations we have created provide, first and foremost, camp visibility: they immediately clarify who the characters that Dennis supported or attacked are, which allows for interesting inferences regarding the motivations of these associations. Dennis supports, or is described as being similar to, characters like Eliza Haywood, Thomas Cooke, Leonard Welsted, Bernard Lintot, Edmund Curll, or Elkanah Settle, to name just a few, and attacks characters like John Dryden, Jeremy Collier, Pope, Gay, William Law, Abel Boyer, and Arthur Bedford (Figs. 8.4 and 8.5). These camps indicate obvious hierarchies of value, but they also open up interesting interpretive avenues when looked at in close detail. While the motivations of Dennis’s animosity against Pope are well-known, his reasons for attacking authors like Law, Boyer, or Bedford are not immediately clear. They may have clashed, as the notes to the poem suggest, in the debate

 Morillo, “John Dennis,” 21.  “John Dennis,” Encyclopedia Britannica, 2 January 2020, https://www.britannica. com/biography/John-Dennis 31 32

282 

I. BAIRD

Fig. 8.4  Graph describing John Dennis’s relations in poem and apparatus as NEATO (spring-model layout). The relations described indicate similarity (green), characters attacked (red), and characters defended (dotted green). Dennis’s relations with Edmund Curll, Giles Jacob, and Leonard Welsted are bidirectional, which indicates strong ties

over the morality of the stage or, given their job description, over religious or political issues.33 Two of Dennis’s connections are particularly interesting as they describe the critic as being at the center of literary coteries involving notorious  William Law is author of The Absolute Unlawfulness of the Stage-Entertainments Fully Demonstrated (London: Printed for W. and J. Innys, 1726). Dennis attacked him because he thought, in disregard of the actual publication date of his work, that his pamphlet was precisely timed to coincide with the Jacobite attempts at a restoration of the Old Pretender (James Francis Edward Stuart). Arthur Bedford is a vicar who wrote pamphlets against the stage; he was confused by Dennis with Hilkiah Bedford, the alleged author of The Hereditary Right of the Crown of England Asserted (London: Printed for Richard Smith, 1713), and therefore attacked for his support for the Pretender. 33

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

283

Fig. 8.5  All relations for John Dennis as DOT (hierarchical layout). The graph highlights Dennis’s strong connection with Colley Cibber, Edmund Curll, and Giles Jacob (the edges are bidirectional)

authors or journalists of the time. Dennis’s connection with Lewis Theobald, James Moore Smythe, Matthew Concanen, and Thomas Cooke, for instance, a relationship that is not immediately obvious just by perusing the poem, is made visible by these graphs. At further examination, we find that they all belonged to a “Club” of authors who, much like the Scriblerians, used to hold weekly meetings during which they produced offensive pamphlets against their rivals (Appendix II, “List of Abusers”). One of these attacks, a letter against Pope signed by W. A. (probably William Arnall) and published in the Mist’s Weekly Journal on June 8,

284 

I. BAIRD

1728, is directly referenced in The Dunciad. Given that the Mist’s Journal published several anonymous attacks against Pope, we may assume that some of them had been contributed by members of Dennis’s “Club.”34 The second reference to Dennis’s cultural leadership appears in Appendix VI, “Of the Poet Laureate,” where he is described as “the worthy president” of the Grub Street Journal “society” of authors. This affiliation seems to point to a category of authors belonging to a specific locale rather than to authors who had their works published in the journal.35 Indeed, in A Tale of a Tub (1704), Swift makes numerous references to the “Society” of Grub Street, or “the Grub-street Brotherhood,” which is described as a “spatious Commonwealth of writers” strategically located in the immediate vicinity of the “Bedlam” mental hospital.36 Thus, describing Dennis as presiding over the large category of “hack” writers and as leading the “Club” of anti-Pope authors publishing in The Mist’s Weekly Journal indicates Pope’s acknowledgement of his central cultural role in the real space of the city. This explains Pope’s deliberate choice to give Dennis a main role in his mock-epic, a role not immediately obvious when reading the poem without paying close attention to its notes. Another important connection revealed by these graphs is that between Dennis and Giles Jacob, an obscure character mentioned in only one instance in the poem (III, l. 149–150), which raises legitimate questions about the reasons for this association. Mainly known as a legal writer, Jacob is also the author of The Poetical Register, a literary history of 34  Given that the majority of these attacks were published under pseudonym, their authorship is difficult to establish. It is true, however, that a concerted attack against Pope was hosted in the pages of the Mist’s Weekly Journal, which published several malicious pieces against the poet. See, for instance, Letter XXIII, Homer’s Character Attempted in Blank Verse, Letter XXVII, BS’s Scurrilous Reflections upon Mr. Pope, and Letter LVII, The Great Mischief Accrued to Church and State from the Assaults of Illiterate Pamphleteers, republished in A Collection of Miscellany Letters, Selected out of Mist’s Weekly Journal, vol. 2 (London: Printed by N.  Mist, 1732). The Mist’s Journal also appears on the cover of the Dunciad Variorum (1729), together with other pro-governmental publications, such as The London Journal, The Daily Journal, The British Journal, Pasquin, and The Flying Post. 35  The Grub Street Journal (1730–1737) was a publication that satirized popular journalism and hack writing that was believed to have been started by Pope himself. However, although he did sporadically contribute to the journal, Pope was not the initiator of this venture but the clergyman Richard Russel and botanist John Martyn. 36  Jonathan Swift, A Tale of a Tub. Written for the Universal Improvement of Mankind. To Which Is Added, an Account of the Battel between the Antient and Modern Books in St. James Library (London: Printed for John Nutt, 1704), 41–43.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

285

contemporary writers that had caused Pope much angst due to accusations of being the “Trumpeter” of his own praise.37 Within Dennis’s social network, the relationship with Jacob is an outlier; that is, it stands out as atypical given the high frequency with which the two are associated. As seen in Figs. 8.4 and 8.6, the association between the two is one of similarity and support: Jacob wrote a letter to Dennis in which he disparaged Pope, a letter which Dennis published in his Remarks to the Dunciad and, subsequently, Dennis defended Jacob against Pope. A close analysis of these graphs also shows the strength of their relationship: the tie is bidirectional (the two characters interact with and support each other), and it occurs more than once (Fig.  8.6). Indeed, as revealed by this graph, Dennis and Jacob are the most vocal critics of the first three books of The Dunciad, being mentioned 92 and 25 times, respectively, in the apparatus. Together with Dennis, Jacob plays the role of the indiscriminate critic, being constantly quoted in relation to authors considered by Pope of little consequence: Laurence Eusden, Ned Ward, Lewis Theobald, John Ozell, Eustace Budgell, John Oldmixon, and Susannah Centlivre. In other words, the most important critic of the first half of the century is associated with a would-be critic in the poems’ textual underground in an effort to diminish Dennis’s inflated persona as a cultural guardian and suggest their similar ineptitude.

37  Giles Jacob, The Poetical Register: Or, The Lives and Characters of All the English Dramatick Poets. With an Account of Their Writings, 2 vols. (London: Printed by E. Curll, 1719–1720). Although he wrote favorably about Pope, whom he praised for his “great Ease” and “Strength of his Compositions,” Jacob criticized here Three Hours after Marriage (1717), a comedy authored by “three mighty Bards” (John Gay, Pope, and Arbuthnot), for containing “some extraordinary scenes … which … trespass on Female Modesty” (1:115). Even worse, Jacob praised The Confederates (1717), a production of the phantom poet Joseph Gay (aka John Breval), an author hired by Curll to attack Pope and his allies. Consequently, Pope included Giles Jacob in the 1729 edition of The Dunciad to punish him for his attacks against his friend, John Gay, and Scriblerians at large. Jacob immediately allied himself with John Dennis, who included Jacob’s letter attacking Pope in his notorious Remarks upon the Dunciad (1729); here, Jacob revealed “the true secret History” of the “selfish Mr. Pope,” who had written the “high Praises and Commendations” contained in The Poetical Register himself. Jacob is also the author of The Rape of the Smock: An HeroiComical Poem (London: Printed for R. Burleigh, 1717), a rewriting in scatological register of Pope’s masterpiece, The Rape of the Lock. Jacob’s poem was republished or anthologized frequently at the time, together with writings by Pope.

286 

I. BAIRD

Fig. 8.6  All relations for Giles Jacob as NEATO (spring-model layout). This graph highlights Jacob’s close relationship with critic John Dennis, with whom he is in a relation of similarity, and with Edmund Curll, whom Jacob supported in his campaign against Pope

Another interesting observation involving Dennis and Jacob sheds light on Pope’s strategy of revising the poem after the change of kings and addition of a fourth book in 1742. When tabulating Dennis’s and Jacob’s connections, I realized that the two were not mentioned even once in the footnotes to the fourth book of The Dunciad. As we know, The Dunciad in Four Books is a significant revision of the previous text that replaces playwright Lewis Theobald with actor and Poet Laureate Colley Cibber as King of Dunces. In light of this major change, Pope made substantial revisions to the poem’s argument to accommodate the new hero. Dennis’s and Jacob’s places are consequently taken in Book IV by Richard Bentley, “the era’s most formidable annotator,”38 and Martinus Scriblerus, a fictional alter ego of Bentley’s; in other words, a pedant (Bentley) replaces a mad critic (Dennis), and a fictional character (Scriblerus) replaces an amatory critic (Jacob). This suggests a new pairing of Dennis and Bentley over  Zerby, The Devil’s Details, 54–55.

38

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

287

their shared acribia and of Jacob and Scriblerus over their similar lack of authority in the literary field. These associations may indicate that Pope targeted these dunces not only because of personal bouts against the poet but also because of reasons that were more comprehensive in nature and involved both their moral and cultural incompetence. The most revelatory finding of this analysis is Dennis’s prominent role in the social network of The Dunciad. Although Dennis is mentioned in the poem itself in only three brief instances, he is one of the best-connected characters in the text as a whole, being in a relationship of support, similarity, or antagonism with no less than 29 other protagonists. Using Malcolm Gladwell’s suggestive term, I have described such an individual as a “connector,” that is, a protagonist who has a high number of ties with other characters.39 Connectors are important not only because of their ability to bring together a large number of individuals from different walks of life but also because, by doing so, they have the capacity to spread rumors and gossip, disseminate innovation, and start cultural trends. As Mark Granovetter pointed out, their central position within a network gives the information spread by group members the needed authority.40 Indeed, as shown by these graphs, Dennis is one of the three connectors of the poem, together with Colley Cibber, the last King of Dunces, and with publisher Edmund Curll (Figs.  8.7 and 8.8). The Shiva graph, in particular, is very useful in highlighting the poem’s three connectors and the spread of their influence along their respective networks: as made obvious by this graph, each dunce can reach another dunce through his or her acquaintances in two to five steps,41 and a connector in only one to two steps. We may hypothesize, therefore, that Dennis’s position as one of the poem’s connectors is granted by his acting as a liaison between various characters belonging to the cultural, political, and religious spheres of the time.42 Thus, the Shiva graph illustrates beautifully the relatedness and 39  Malcolm Gladwell, The Tipping Point. How Little Things Can Make a Big Difference (New York: Back Bay Books, 2000), esp. 30–59. 40  Mark S. Granovetter, “The Strength of Weak Ties,” The American Journal of Sociology 78, no. 6 (1973): 1360–80, http://www.jstor.org/stable/2776392 41  Stanley Milgram, “The Small World Problem,” Psychology Today 1, no. 1 (1967): 65. 42  As Gladwell explains, connectors are “people with a special gift for bringing the world together” because of their ability “to occupy many different worlds and subcultures and niches” and effectively navigate among them. See Gladwell, The Tipping Point, 38 and 48, respectively. A similar argument was made later on by Barabási, who pointed out that “the

288 

I. BAIRD

Fig. 8.7  Graph showing the social networks of all the six dunces considered as NEATO (spring-model layout). This graph highlights the three connectors of the poem: John Dennis, Edmund Curll, and Colley Cibber

multiplicity of spaces—social, political, cultural—that describe eighteenthcentury London, spaces which “attain ‘real’ existence by virtue of networks and pathways, by virtue of bunches or clusters of relationships.”43 These graphs also make evident the nature of Dennis’s connections and thus the role played by the critic in the cultural space of eighteenth-century London. For instance, Dennis has strong, bidirectional ties (i.e., ties with individuals in his close circle of friends or that involve “reciprocal services”44) not only with Giles Jacob but also, importantly, with the other two connectors of the poem, Colley Cibber and Edmund Curll. Dennis’s connections with Cibber are especially interesting given the 1742 revision truly central position in networks is reserved for those nodes that are simultaneously part of many large clusters” due to their ability to be at home in various spheres, from arts to sciences. See Albert-László Barabási, Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life (New York: Basic Books, 2014), 61. 43  Henry Lefebvre, The Production of Space, trans. Donald Nicholson-Smith (Oxford: Blackwell Publishers, 1995), 86. 44  Granovetter, “The Strength of Weak Ties,” 1361.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

289

Fig. 8.8  Shiva Graph. This graph gives the viewer a keen sense of the relatedness of all the characters in the poem and makes visible the poems’ three connectors: Colley Cibber, Edmund Curll, and John Dennis

of the poem, which involved a change of kings: the critic is mentioned twice in relation to Cibber, once in a context suggesting their similar blindness to personal satire (“Testimonies”) and once in a context suggesting their sycophancy toward the ruling class (Appendix VI, “Of the Poet Laureate”). Dennis’s relations with Curll, on the other hand, highlight their similar animosity toward the poet (they both are authors “whose

290 

I. BAIRD

wrath is perilous”45), cowardice (they both feared Pope’s “slander and poison”46), and malicious spread of false rumors (they both wrote about Pope’s “profaneness and immorality”).47 These positionings are important not only because they give the examiner a new way of looking at the argument of the poem by identifying the main protagonists involved in articulating this argument but also because they highlight actors with similar roles, or actors who are “regularly equivalent” within the network due to having similar patterns of relationships.48 Conversely, Dennis has weak ties (i.e., occasional ties with individuals who are remote acquaintances, or, in our case, peripheral within the poem’s social network) with authors like George Duckett, S.  Popping, and Thomas Rhymer, to give just a few examples. Although these characters are connected with Dennis, they are not directly connected with any other members of his network; however, they all belong to the same camp and/or support the same connector.49 The importance of these weak ties should not be underestimated: the connections with people who are not members of one’s close group of friends are important because they provide a bridge, that is, “the only route along which information or influence can flow”50 from one group to another. In other words, the weak ties in a network expand the radius of the connections within the network and, by doing so, they are best placed to diffuse innovation and spread opinion. Indeed, when looking closely at Dennis’s connections, they reflect the critic’s involvement in some of the most heated debates of the epoch: Curll’s defamatory campaign against Pope, Jeremy Collier’s pamphlet war against the “profaneness” of the stage, the argument about the institution of laureateship, the “Ancients” versus “Moderns” debate, the defense of classical rules and the sublime in art, the influence of political conditions upon the production of letters, the role of religion as a social unifier, and so on (Table 8.1).  Pope, Dunciad in Four Books, 60.  Ibid., 113. 47  Ibid., 188–89. 48  As explained by Hanneman and Riddle, “actors that are regularly equivalent do not necessarily fall in the same network positions or locations with respect to other individual actors; rather, they have the same kinds of relationships with some members of other sets of actors.” See Robert A. Hanneman and Mark Riddle, “Network Positions and Social Roles: The Idea of Equivalence,” in Introduction to Social Network Methods (Riverside, CA: University of California, Riverside, 2005), https://faculty.ucr.edu/~hanneman/nettext/ C12_Equivalence.html 49  Another interesting area of investigation would be to compare the membership of these three social networks, identify similarities and/or differences, and hypothesize on the power dynamics each of them suggests. 50  Granovetter, “The Strength of Weak Ties,” 1364. 45 46

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

291

Finally, this analysis of Dennis’s social network reveals the competing stories told by the poem and the apparatus. As already indicated, Dennis is mentioned in only three brief instances in the poem: in Book I, l. 106, as “the mighty Mad” for his perpetual “rage” against his opponents, in Book II, l. 239–42, as an inept critic and a highly reactive individual, and in Book III, l. 173, where he is associated, “fool with fool,” with Charles Gildon, a hack writer with whom he allegedly wrote A True Character of Mr. Pope (1716). These attacks are neither particularly developed nor do they give Dennis a prominent role in the poem’s plot network. However, when looking at the poem’s apparatus, Dennis is mentioned no less than 92 times in the prefatory material, Testimonies, Notes to Testimonies, List of Abusers, appendices, and footnotes, becoming the uncrowned king of The Dunciad’s textual periphery. The story told by the apparatus is in stark contrast with the story told by the poem, involving Dennis in a rich network of relations with a large number of characters of loose morals, little talent, or suspect associations. Pope’s method of disparaging Dennis is, thus, deeply subversive: he describes Dennis as doing his work of cultural policing from a textual underground where the value criteria are fundamentally flawed. The findings resulting from performing this type of analysis are even more interesting when working with larger datasets. As seen in Figs. 8.9 and 8.10, the intersections of the six social networks created during this project have led to revelatory results: these visualizations single out the seventeen dunces (out of many dozens) who are central to the poem’s plot network, or the poem’s “hall of infamy,” and the four authors epitomizing “good writers” (i.e., Alexander Pope, John Gay, Joseph Addison, and John Dryden), or the poem’s “hall of fame.” Most importantly, these social networks show that Pope’s dunces are a cultural category that, far from representing a marginalized, minority, or disempowered group, participates in a forceful way in shaping public opinion. Therefore, unlike Nancy Fraser, who argues for the inclusion of “subaltern” groups in the public sphere, I contend here that Pope’s dunces are engaged in discourses that place them in the same realm of public debate with cultural authorities, rather than in a subaltern position. While “counterpublic”51 is a term that suggestively encapsulates the idea of competing interests—be they divided along class, gender, or political lines—, a networked public may better describe the collaborative nature of the public sphere of the time, which I see not as a conglomerate of divergent ideologies, but as an organic whole. 51  Nancy Fraser, “Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy,” Social Text 25/26 (1990): 61, https://doi.org/10.2307/466240

292 

I. BAIRD

Fig. 8.9  The Inner Circle. Full view of all networks of the six dunces considered as CIRCO (circular layout). This graph highlights the central and the peripheral characters of the poem and apparatus

Conclusion Acknowledging Dennis as the uncrowned king of the Dunciad’s textual periphery needs to be understood less as Pope’s victory cry against his most spiteful attacker and more as an expression of his concern with the critic’s role as a cultural and social regulator. As England’s leading critic of the first decades of the century, Dennis held a central position in the literary landscape of the time: his contributions to establishing an English canon by reassessing the influence of the “Ancients” on modern thought and emphasizing the importance of rules in art in The Impartial Critick

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

293

Fig. 8.10  The Inner Circle. Detail view of all networks of the six dunces considered as CIRCO. This detail view highlights characters who appear in more than one network, or the main protagonists of the poem and the apparatus. The hall of fame/“good writers,” includes four authors: Alexander Pope, John Gay, Joseph Addison, and John Dryden. The hall of infamy/“bad writers” includes seventeen authors: Colley Cibber, Edmund Curll, Eliza Haywood, John Henley, John Ozell, John Oldmixon, Lewis Theobald, Giles Jacob, Laurence Eusden, Elkanah Settle, Thomas Cooke, John Dennis, Bernard Lintot, Charles Gildon, George Duckett, Leonard Welsted, and Richard Blackmore

(1693), The Advancement and Reformation of Modern Poetry (1701), and The Grounds of Criticism in Poetry (1704) still pass the test of time. Better known today for his theory of enthusiasm and the sublime as “rational” delight,52 Dennis is also a precursor of the Romantic movement, having a direct influence on William Wordsworth and Samuel T. Coleridge. More importantly, though, Dennis epitomized for Pope the seminal role played  The Critical Works of John Dennis, ed. Edward Niles Hooker (Baltimore: The Johns Hopkins Press, 1941), 2: 381. 52

294 

I. BAIRD

by a critic in shaping public opinion, acting as a liaison among diverse social spheres. This is reflected both by his oeuvre53 and, as this analysis has attempted to demonstrate, by his social network, which highlights the critic’s involvement with individuals occupying a variety of political, religious, and cultural spheres. Dennis’s, Cibber’s, and Curll’s roles as the poem’s connectors are thus explained by the “multiple components of their identities and engagements within society.”54 This type of social network analysis also answers Martin Paul Eve’s important question: “what can the computer see, in its repetitive and unwavering attention to minute detail, that is less (or even invisible to human readers?”55 As this case study demonstrates, Dennis’s social network enhances camp visibility and clarifies the nature of his associations, highlights the critic’s central role as one of the three “connectors” of the poem, calls attention to the competing stories told by the poem and the apparatus, identifies in Jacob a network outlier that sheds light on Pope’s elaborated defamatory campaign against the critic, and singles out Dennis as the uncrowned king of The Dunciad’s textual periphery. Although Dennis’s associations are often meant to suggest the critic’s blurry sense of cultural value and tempestuous character, by assigning him the main role in the apparatus, Pope implicitly acknowledges Dennis’s centrality in the cultural landscape of early eighteenth-century London. Pope’s argument against Dennis seems to question, therefore, less his ideas and more his moral competence: a committed Whig with a lifelong service to various political patrons, an individual with a highly volatile temper, and a critic with a tendency “to crack nuts with a sledgehammer,”56 Dennis does not meet the impartiality requirement of his job description. 53  Dennis published extensively on the benefits of the government’s regulation of the theater (The Stage Defended, 1726), as well as on other issues of public interest, such as foreign influence on local culture (An Essay on the Operas after the Italian Manner, 1706), or public morality (Vice and Luxury Publick Mischiefs, 1724). He also called attention to the impact of political and social conditions on the production of letters (A Large Account of Taste in Poetry, 1702) and to the importance of religion in “cementing Societies” (The Grounds of Criticism, 1704). 54  Dan Edelstein and Chloe Summers Edmondson, “Introduction: Historical Network Analysis and Social Groups in the Enlightenment,” in Edmondson and Edelstein, Networks of Enlightenment, 11. 55  Martin Paul Eve, “Close Reading with Computers: Signals, Parts of Speech, and David Mitchell’s Cloud Atlas,” SubStance 46, no. 3 (2017): 77, http://www.muse.jhu.edu/ article/676240 56  James R.  Sutherland, review of The Critical Works of John Dennis, ed. Edward Niles Hooker, The Review of English Studies 18, no. 69 (January 1942): 118, www.jstor.org/ stable/509884

Prefatory material Notes to Testimonies

Notes to Testimonies

Notes to Testimonies Testimonies Notes to Testimonies

Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Dennis, John

Notes to Testimonies Testimonies Notes to Testimonies

Prefatory material

Dennis, John

Dennis, John Dennis, John

Address

Character

Curll, Edmund

Pope, Alexander

Pope, Alexander Pope, Alexander

Blackmore, Richard

Person referred to

Dennis on Windsor Forest Letter to B.B. at the end of Remarks on Pope’s Homer, 1717

Pope, Alexander B.B.?

Dennis on Essay on Criticism Pope, Alexander Reflections Critical and Satyrical Pope, Alexander on a Rhapsody Called An Essay on Criticism, printed for Lintot Lintot, Bernard

Dennis, Remarks on Pr. Arthur by R. Blackmore Dennis, Character of Mr. P. 1716 Dennis on Pope’s education in Reflections on the Essay on Criticism Dennis the author of a libel against Pope (Character of Mr. P and His Writings, 1729), according to Curl in the Curliad

Reference

Attacked Akin to

Akin to

Attacked Attacked

Akin to

Attacked

Attacked Attacked

Akin to

Relationship

Critic and playwright

Furius, “a dry old gentleman”

(continued)

Who’s who

AKA

Table 8.1  Spreadsheet that captures the salient information about John Dennis in the poem and apparatus: the character’s address in the text, the reference to the character (in bold when in the poem, in regular font when in the apparatus), the name of the person he is related to, the description of their relationship (similarity, dissimilarity, character attacked, or character defended), alternative references to Dennis, and authorial id

Address

Testimonies

Testimonies

Testimonies Testimonies Testimonies Testimonies

Testimonies Note to Testimonies

Note to Testimonies Note to Testimonies Testimonies

Note to Testimonies

Note to Testimonies

Character

Dennis, John

Dennis, John

Dennis, John Dennis, John Dennis, John Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John Dennis, John

Dennis, John

Dennis, John

Table 8.1 (continued) Person referred to

On Pope’s Character, in Pope, Alexander Reflections on the Essay of Criticism and Character of Mr. Pope and His Writing (1716) and in his anonymous A True Character of Mr. Pope (1716) Mr. Dennis and Gildon’s Gildon, Charles contradictory statements in True Character of Pope Pope, Alexander Pope compares Cibber to Dennis Cibber, Colley Pope, Alexander Dennis and Gildon on Pope’s Gildon, Charles rhymes in A True Character of Mr. Pope (1716) Pope, Alexander Dennis proves he did not work Gildon, Charles with Gildon on A True Character of Mr. Pope Pope, Alexander Remarks on the Dunciad Pope, Alexander Dennis on Pope’s “depravity of Pope, Alexander genius and taste” in Essay on Criticism Dennis’s Preface to His Pope, Alexander Reflections on the Essay on Criticism Dennis’s Preface to his Remarks Pope, Alexander on Homer

Reference

Attacked

Attacked

Attacked Attacked Attacked

Attacked Akin to

Attacked Akin to Attacked Akin to

Akin to

Attacked

Relationship

AKA

Who’s who

Testimonies

Testimonies Testimonies

Testimonies Martinus Scriblerus of the Poem Appendix II List of Abusers

Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John

Appendix II List of Abusers

Appendix II List of Abusers Appendix II List of Abusers Appendix II List of Abusers

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Appendix II List of Abusers

Dennis, John

Dennis, John

Testimonies

Dennis, John

Dennis ascribes Pope two farces Pope, Alexander (probably What D’ye Call It, and Three Hours After Marriage) Dennis considers Pope below Pope, Alexander D’Urfey in drama D’Urfey, Tom Dennis thinks What D’ye Call It Pope, Alexander is not Mr. P’s but Mr. Gay’s Gay, John Rymer and Dennis become poets Rymer, Thomas in their later years Dennis, Reflections Critical and Pope, Alexander Satirical on … An Essay on Criticism A True Character of Mr. P. and Pope, Alexander His Writings, in a Letter to a Friend. Anon [Dennis] Remarks upon Mr. Pope’s Pope, Alexander Translation of Homer; with two letters concerning the Windsor Forest and the Temple of Fame Remarks on Mr. Pope’s Rape of the Pope, Alexander Lock, in Letters to a Friend, 1728 A Letter against Mr. P. at Large. Pope, Alexander Anon. [John Dennis] Remarks on the Dunciad Pope, Alexander (dedicated to Theobald) Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked Akin to

Akin to Attacked

Attacked

Attacked

(continued)

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Appendix VI Poet Laureate

Appendix VI Poet Laureate Appendix VI Poet Laureate

Appendix II List of Abusers Appendix VI Poet Laureate Appendix VI Poet Laureate

Appendix II List of Abusers Appendix II List of Abusers Appendix II List of Abusers Appendix II List of Abusers

Appendix II List of Abusers Appendix II List of Abusers

Dennis, John

Dennis, John

Address

Character

Table 8.1 (continued)

Duckett, George

Pope, Alexander

Cooke, Thomas

Concanen, Matthew

Moore, A.

Theobald, Lewis

Theobald, Lewis

Person referred to

Dennis, “the worthy president” of the Grub Street Journal society of authors Dennis and Anstis as organizers Anstis, John of laureateship ceremony

If Dennis is chosen, he should be given a mixture of brassica Cibber and Theobald don’t have Theobald, Lewis a good stomach; Dennis (“a dry old gentleman”) does Cibber, Colley

Pope Alexander’s Supremacy and Infallibility Examined [with Ducket]

Member of the Club of Theobald, Moore, Concanen, Cooke

Reference

Akin to

Akin to

Akin to

Akin to

Attacked

Akin to

Akin to

Akin to

Akin to

Akin to

Relationship

AKA

Who’s who

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope Appendix VIII Dryden and Pope

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Dennis, John

Dennis, John

Appendix VIII Dryden and Pope

Appendix VII Advertisement 1730 Appendix VIII Dryden and Pope

Dennis, John

Dennis, John

Dennis, John

Lintot, Bernard

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

(Pope) has no admirers, Pope, Alexander according to Dennis, Remarks on the Rape of the Lock [Pope] has no genius, he is a Pope, Alexander little author according to Dennis, Remarks on Homer and Character of Mr. P.

A Certificate that one is no Wit from Mr. Dennis singly Mr. Pope is a mortal enemy to his country according to Dennis, Remarks on Rape of the Lock (Pope) both a Whig and a Tory, according to Dennis, Character of Mr. Pope (Pope) as a popish rhymester, according to Dennis, Remarks on Homer (Pope) has a notable knack of rhyming, according to Character of Mr. P. and Dennis on Homer Lintot’s Homer does not talk like Homer, but like Pope, according to Dennis, Remarks on Homer

Attacked

Attacked

Akin to

Attacked

Attacked

Attacked

Attacked

Attacked

(continued)

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope

Appendix VIII Dryden and Pope Appendix VIII Dryden and Pope Appendix VIII Dryden and Pope Appendix VIII Dryden and Pope I, Note to 63 I, Note to 63

I, 10

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Address

Character

Table 8.1 (continued)

[Pope] doesn’t know either English or Greek, according to Dennis, Remarks on Homer [Pope] has a notable talent at burlesque, according to Dennis, Remarks on H. [Pope] as an Ape, according to Dennis, Daily Journal, May 11, 1728 [Pope] as an ass, according to Dennis, Preface to Remarks on Homer [Pope] as a frog, according to Dennis, Remarks on the Rape of the Loc [Pope] as a coward, according to Character of Mr. P. [Pope] as a knave, according to Character of Mr. P. [Pope] as a fool, according to Dennis, Remarks on Homer [Pope] as a thing, according to Dennis, Remarks on Home Dennis’s dislike of puns Dennis on Homer and Daily Journal, June 11, 1728 “and all the mighty Mad in Dennis rage”

Reference

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Pope, Alexander

Person referred to

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Attacked

Relationship

AKA

Who’s who

I, Note to 106

I, Note to 106

I, Note to 106

I, Note to 106 I, Note to 106

I, Note to 106 I, Note to 106 I, Note to 106

I, Note to 106 I, Note to 106

I, Note to 106

I, Note to 286

I, Note to 286 II, Note to 118

Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John Dennis, John

Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Pope, Alexander

Pope, Alexander

Theobald, Lewis

Curll, Edmund Reference to A True Character of Pope, Alexander Mr. Pope and His Writing, printed for S. Popping (1716) Popping, S. Reference to Remarks on Homer Pope, Alexander Reference to Dennis’s Preface to Blackmore, Richard Remarks on Prince Arthur Dennis son of a sadler in London Mr. Dennis excellent at Pindaric verses Giles Jacob’s account of Dennis Jacob, Giles in his Lives of Dramatic Poets (Dennis on himself) Dennis on Theobald, in Remarks Pope, Alexander on Pope’s Homer Theobald, Lewis Re Homer’s poverty, in Dennis’s Pope, Alexander Preface to Remarks on the Rape of the Lock

Theobald calls Dennis “Furius” (The Censor, vol. 22, 33) Reference to Reflections on the Essay on Criticism Reference to A Full and True Account of a Horrid and Barbarous Revenge, by Poison, on the Body of Edmund Curl (1716)

Akin to Attacked

Attacked

Akin to

Akin to Attacked Akin to

Akin to Attacked

Attacked

Attacked

Akin to

(continued)

Address

II, Note to 140

II, Note to 142

II, Note to 142

II, Note to 142 II, Note to 207

II, Note to 207 II, Note to 226

II, 239–42

II, Note to 268

II, Note to 268

II, Note to 268

II, Note to 268 II, Note to 268 II, Note to 268

Character

Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John Dennis, John

Table 8.1 (continued)

Law, William Welsted, Leonard

Pope, Alexander

Pope, Alexander

Person referred to

Dennis’s remarks on Blackmore’s Blackmore, Richard Prince Arthur and fustian writers Dennis’s remarks on Dryden and Dryden, John Pope in Preface to Remarks on Prince Arthur Pope, Alexander Blackmore, Richard Dennis’s relationship with Curll Curll, Edmund

Pope, Alexander Dennis invented a new device to make Thunder on the stage “Dennis and Dissonance, and captious Art, And Snip-snap sort, and Interruption smart, And Demonstration thin, and Theses thick, And Major, Minor, and Conclusion quick.” Dennis’s friendship with Gildon Gildon, Charles

“Teach more my half than Dennis’ rules” Dennis re deformity, Character of Mr. Pope Dennis thinks Pope’s original is from the Devil Dennis against Mr. Law Dennis identifies Welsted as the eel from Peri Bathous.

Reference

Attacked Akin to Akin to

Attacked

Akin to

Akin to

Attacked

Attacked Defended

Attacked

Attacked

Relationship

AKA

Who’s who

II, Note to 283

II, Note to 413

II, Note to 413 II, Note to 413 II, Note to 413

II, Note to 413

II, Note to 413

III, Note to 24

III, Note to 36

III, Note to 36 III, Note to 36

III, Note to 36 III, Note to 36

Dennis, John

Dennis, John

Dennis, John Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John

Ref. to Dennis’s Preface to Rem. on Homer

Dennis about Settle as rival to Dryden

Dennis, Stage defended against Mr. Law Dennis re Bavius in Remarks on Prince Arthur Dennis, Oldmixon and Welsted didn’t notice a spelling mistake (length of ears instead of years) in the line about Settle

Dennis about a true play Dennis’s reference to Collier’s Short View on the … Dennis’s reference to Bedford’s Serious remonstrance

Mr. John Oldmixon, next to Mr. Dennis, the most ancient Critic of our Nation Dennis—his exchange with William Law and A. Boyer re stage and state

Dryden, John Pope, Alexander

Welsted, Leonard Settle, Elkanah

Oldmixon, John

Bavius

Bedford, Hilkiah (should be Bedford, Arthur) Law, William

Collier, Jeremy

Abel Boyer

Law, William

Oldmixon, John

Attacked Attacked

Akin to Akin to

Akin to

Defended

Attacked

Attacked

Attacked

Attacked

Attacked

Akin to

(continued)

Dennis, John Dennis, John

Dennis, John Dennis, John

Dennis, John Dennis, John Dennis, John

Dennis, John

Dennis, John

Dennis, John

Dennis, John

III, Note to 149–50 Jacob’s letter to Dennis printed in Dennis’s Remarks on the Dunciad III, Note to 149–150 III, Note to Dennis re Pope’s repentance 149–150 III, Note to Dennis’ name crept into the 149–150 poem by mistake III, 173 “Ah, Dennis! Gildon, ah!” His anger at Pope’s claim that he had written A True Character of Mr. Pope (1716) together with Gildon. III, 173 III, Note to 173 Dennis’s interest in our author III, Note to 173 Dennis’s own account of himself in Jacob’s Lives III, Note to 173 D’Urfey’s senior III, Note to 179 Dennis defends G. Duckett’s III, Note to 179 heterosexuality in his Dedication of Remarks on Pope’s Rape of the Lock (1728) III, Note to 179 III, Note to 179 Dennis, Stage Defended against Mr. Law

Dennis, John

Reference

Address

Character

Table 8.1 (continued)

Pope, Alexander Law, William

D’Urfey, Tom Ducket, George

Pope, Alexander Pope, Alexander Jacob, Giles

Gildon, Charles

Pope, Alexander

Pope, Alexander

Jacob, Giles

Person referred to

Attacked Attacked

Akin to Defended

Attacked Attacked Akin to

Akin to

Attacked

Attacked

Akin to

Relationship

AKA

Who’s who

III, Note to 330

Index of Persons Index of Matters Illustrations to 1728, 1729 editions, headpiece to 1729 edition

Dennis, John

Dennis, John Dennis, John Dennis, John

The great Critic Mr. Dennis attacked (unsuccessfully) the Italian Opera 3 entries 17 entries 3 illustrations

306 

I. BAIRD

Bibliography A Collection of Miscellany Letters, Selected out of Mist’s Weekly Journal. 1732. Vol. 2. London: Printed by N. Mist. Algee-Hewitt, Mark. 2019. The Principles of Meaning: Networks of Knowledge in Johnson’s Dictionary. In Edmondson and Edelstein, Networks of Enlightenment, 251–277. Baird, Ileana. 2011. Visual Paratexts: The Dunciad Illustrations and the Thistles of Satire. In Book Illustration in the Long Eighteenth Century: Reconfiguring the Visual Periphery of the Text, ed. Christina Ionescu, 329–366. Newcastle upon Tyne: Cambridge Scholars Publishing. ———. 2019. The Strength of Weak Ties: Eliza Haywood’s Social Network in The Dunciad in Four Books (1743). ABO: Interactive Journal of Women in the Arts, 1640–1830 9 (2): 1–36. https://doi.org/10.5038/2157-7129.9.2.120. Barabási, Albert-László. 2014. Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life. New York: Basic Books. Burrows, Simon, and Glenn Roe, eds. 2020. Digitizing Enlightenment: Digital Humanities and the Transformation of Eighteenth-Century Studies. Liverpool: Voltaire Foundation in association with Liverpool University Press. Cannan, Paul D. 2015. John Dennis. In The Encyclopedia of British Literature 1660–1789, ed. Gary Day and Jack Lynch. Oxford: Oxford University Press. Cuccuru, Kathrine. 2016. That ‘Tremendous’ Mr. Dennis: The Sublime, Common Sense, and Criticism. In Passions, Sympathy, and Print Culture: Public Opinion and Emotional Authenticity in Eighteenth-Century Britain, ed. Heather Kerr, David Lemmings, and Robert Phiddian, 105–121. London: Palgrave Macmillan. Dennis, John. 1711. Reflections, Critical and Satyrical, upon a Late Rhapsody, Call’d, An Essay upon Criticism. London: Printed for Bernard Lintott. ———. 1716. A True Character of Mr. Pope, and His Writings. In a Letter to a Friend. London: Printed for S. Popping. ———. 1717. Remarks upon Mr. Pope’s Translation of Homer, with Two Letters Concerning Windsor Forest, and the Temple of Fame. London: Printed for E. Curll. ———. 1728. Remarks on Mr. Pope’s Rape of the Lock. In Several Letters to a Friend. With a Preface Occasion’d by the Late Treatise on the Profound, and the Dunciad. London: Printed for J. Roberts. ———. 1729. Remarks upon Several Passages in the Preliminaries to the Dunciad, Both of the Quarto and the Duodecimo Edition. And upon Several Passages in Pope’s Preface to His Translation of Homer’s Iliad. In Both Which Is Shewn the Author’s Want of Judgment. With Original Letters from Sir Richard Steele, from the Late Mr. Gildon, from Mr. Jacob, and from Mr. Pope Himself, Which Shew the Falsehood of the Latter, His Envy, and His Malice. London: Printed for H. Whitridge. Edelstein, Dan, and Chloe Summers Edmondson. 2019. Introduction: Historical Network Analysis and Social Groups in the Enlightenment. In Edmondson and Edelstein, Networks of Enlightenment, 1–20.

8  OUTLIERS, CONNECTORS, AND TEXTUAL PERIPHERY 

307

Edmondson, Chloe, and Dan Edelstein, eds. 2019. Networks of Enlightenment: Digital Approaches to the Republic of Letters. Liverpool: Voltaire Foundation in association with Liverpool University Press. Eve, Martin Paul. 2017. Close Reading with Computers: Signals, Parts of Speech, and David Mitchell’s Cloud Atlas. SubStance 46 (3): 76–104. http://www. muse.jhu.edu/article/676240. Fraser, Nancy. 1990. Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy. Social Text (25/26): 56–80. https://doi. org/10.2307/466240. Gay, John. 1717. Three Hours after Marriage: A Comedy. London printed; reprinted in Dublin by S. Powell. Gladwell, Malcolm. 2000. The Tipping Point. How Little Things Can Make a Big Difference. New York: Back Bay Books. Grafton, Anthony. 1997. The Footnote: A Curious History. Cambridge, MA: Harvard University Press. Granovetter, Mark S. 1973. The Strength of Weak Ties. The American Journal of Sociology 78 (6): 1360–1380. http://www.jstor.org/stable/2776392. Hanneman, Robert A., and Mark Riddle. 2005. Introduction to Social Network Methods. Riverside: University of California. https://faculty.ucr.edu/ ~hanneman/nettext/C12_Equivalence.html. Hooker, Edward N. 1940. Pope and Dennis. English Literary History 7 (3): 188–198. https://doi.org/10.2307/2871490. Jacob, Giles. 1719–1720. The Poetical Register: Or, The Lives and Characters of All the English Dramatick Poets. With an Account of Their Writings. 2 vols. London: Printed by E. Curll. ———. 1717. The Rape of the Smock: An Heroi-Comical Poem. London: Printed for R. Burleigh. Johnson, Samuel. 1832. Pope. In The Works of Samuel Johnson, LL.D, vol. 2. New York: John Dearborn. Law, William. 1726. The Absolute Unlawfulness of the Stage-Entertainments Fully Demonstrated. London: Printed for W. and J. Innys. Lefebvre, Henry. 1995. The Production of Space. Trans. Donald Nicholson-Smith. Oxford: Blackwell Publishers. Meulen, David L. Vander. 1991. Pope’s Dunciad of 1728. A History and Facsimile. Charlottesville: University of Virginia Press. Milgram, Stanley. 1967. The Small World Problem. Psychology Today 1 (1): 60–67. Moretti, Franco. 2005. Graphs, Maps, Trees: Abstract Models for a Literary History. New York: Verso. Morillo, John. 2000. John Dennis: Enthusiastic Passions, Cultural Memory, and Literary Theory. Eighteenth-Century Studies 34 (1): 21–41. https://doi. org/10.1353/ecs.2000.0063. Murphy, Avon Jack. 1984. John Dennis. Boston: Twayne Publishers. Pearl, Lisa, Kristine Lu, and Anousheh Haghighi. 2017. The Character in the Letter: Epistolary Attribution in Samuel Richardon’s Clarissa. Digital Scholarship in the Humanities 32 (2): 355–376. https://doi.org/10.1093/llc/fqw007.

308 

I. BAIRD

Peri Bathous, or The Art of Sinking in Poetry. 1727. In Miscellanies. The Last Volume. London: Printed for B. Motte. Pope, Alexander. 1999. The Dunciad in Four Books. Edited by Valerie Rumbold. New York: Pearson. ———. 1711. An Essay on Criticism. London: Printed for W. Lewis. Rumbold, Valerie, ed. 2007. The Poems of Alexander Pope. Volume III: The Dunciad (1728) & The Dunciad Variorum (1729). New York: Pearson. Sherburn, George, ed. 1956. The Correspondence of Alexander Pope. Vol. 2. Oxford: Oxford Clarendon Press. Smallwood, Philip. 2017. Petty Caviller or ‘Formidable Assaillant’? Johnson Reads Dennis. The Cambridge Quarterly 46 (4): 305–324. https://doi. org/10.1093/camqtly/bfx025. Sutherland, James R. 1942. Review of The Critical Works of John Dennis. The Review of English Studies 18 (69): 115–118. www.jstor.org/stable/509884. Swift, Jonathan. 1704. A Tale of a Tub. Written for the Universal Improvement of Mankind. To Which Is Added, an Account of the Battel between the Antient and Modern Books in St. James Library. London: Printed for John Nutt. The Popiad. 1728. London: Printed for E. Curll. Weber, Harold. 1999. The ‘Garbage Heap’ of Memory: At Play in Pope’s Archives of Dulness. Eighteenth-Century Studies 33 (1): 1–19. https://doi. org/10.1353/ecs.1999.0060. Woloch, Alex. 2003. The One vs. the Many: Minor Characters and the Space of the Protagonist in the Novel. Princeton: Princeton University Press. Zerby, Chuck. 2002. The Devil’s Details. A History of Footnotes. New York: Simon & Schuster. Open Access  This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

CHAPTER 9

Publishing Music by Subscription in Eighteenth-Century Britain: The Concertos of Charles Avison Simon D. I. Fleming Publication by subscription was a method commonly employed by numerous authors, editors, composers and publishers in eighteenth-century Britain as a way to issue their latest musical or literary works.1 Through This chapter is dedicated to the memory of my mother, Elizabeth Josephine Fleming (d. 2019). I am particularly grateful for the assistance of the staff at the British Library, Durham University Library, the Dean and Chapter Library, Durham Cathedral, the East Riding Archives, Beverley, the National Archives of Scotland, Edinburgh, Newcastle City Library, the Literary and Philosophical Society, Newcastle, and Otto Haas Music. Individual thanks are also due to Helen Clark, Colin Coleman, Gordon Dixon, Simon Heighes, H.  Diack Johnstone, Martin Perkins, Tim Rishton, James Smith, and Michael Talbot. This chapter is primarily derived from my previous articles “Avison and His Subscribers: Musical Networking in Eighteenth-Century Britain,” Royal Musical Association Research Chronicle 49, no. 1 (2018): 21–49, https://doi.org/10.1080/14723808.2018.1363210, and “The Gender of Subscribers to Eighteenth-Century Music Publications,” Royal Musical Association Research Chronicle 50, no. 1 (2019): 94–152, https://doi. org/10.1080/14723808.2019.1570752. 1  Books on a diverse range of subjects were also published by subscription, including works on mathematics, science, and fiction. See, for example, Ruth Wallis and Peter Wallis, “Female

© The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_9

309

310 

S. D. I. FLEMING

subscription, it was possible for an individual to sell enough copies of a book or musical score in advance to meet the costs involved in engraving the plates, completing any necessary typesetting, and the printing itself. As a result, it was of particular benefit to those who did not have the financial means to undertake such an expensive endeavor themselves. Publication by subscription was additionally attractive to potential subscribers, who would frequently receive a discount off the full sale price and, furthermore, their name would often be included in a printed list attached to the work. These subscriber lists provide us with a unique insight into the social history of Britain and the changes that took place over the course of the Georgian Period. Nevertheless, even though the late Stanley Sadie, over thirty years ago, observed that an “analysis by social class [through subscription lists] of the music-buying public would no doubt be revealing,”2 a study of the patterns that emerge from searching a large number of lists has not been attempted until recently.3 This chapter examines the subscription lists attached to the published musical works produced by the Newcastle-based composer Charles Avison (1709–1770), arguably the most significant British-born composer of Philomaths,” Historia Mathematica 7, no. 1 (1980): 57–64, https://doi. org/10.1016/0315-0860(80)90064-6; and P. D. Garside, “Jane Austen and Subscription Fiction,” Journal of Eighteenth-Century Studies 10, no. 2 (1987): 175–88, https://doi. org/10.1111/j.1754-0208.1988.tb00014.x 2  Stanley Sadie, “Music in the Home II,” in The Blackwell History of Music in Britain: The Eighteenth Century, ed. H.  Diack Johnstone and Roger Fiske (Oxford: Basil Blackwell, 1990), 315. 3  Earlier articles that deal with the topic of subscription lists include David Hunter and Rose Mason, “Supporting Handel through Subscription to Publications: The Lists of Rodelinda and Faramondo Compared,” Notes 56, no. 1 (1999): 27–93, https://doi. org/10.2307/900471; Margaret Seares, “The Composer and the Subscriber: A Case Study from the 18th Century,” Early Music 39, no. 1 (2011): 65–78, https://doi.org/10.1093/ em/caq111; Michael Talbot, “What Lists of Subscribers Can Tell Us: The Cases of Giacob Basevi Cervetto’s Opp. 1 and 2,” De Musica Disserenda 10, no. 1 (2014): 121–39, https:// doi.org/10.3986/dmd10.1.08; Michael Kassler, “The Bachists of 1810: Subscribers to the Wesley/Horn Edition of the ’48,” in The English Bach Awakening, ed. Michael Kassler (London: Routledge, 2016), 315–40.

S. D. I. Fleming (*) Durham University, Durham, UK e-mail: [email protected]

9  PUBLISHING MUSIC BY SUBSCRIPTION 

311

concerti grossi. This work is part of an ongoing project to create a database of all subscribers to every music-related publication issued in Britain by 1820, which I am compiling in partnership with Martin Perkins of Birmingham Conservatory. By mapping the geographical locations of Avison’s subscribers, it is possible to explore the popularity of his music, and how it changed over an extended period of time. Data related to title, gender, or profession extracted from these subscription lists has also allowed me to draw conclusions on the types of subscribers that purchased Avison’s music and their social status.4 Finally, these lists highlight patterns of dissemination and connections (professional, of locality, of patronage, etc.) previously unknown to students of his work.

Publication by Subscription: A Data-Driven Approach The earliest known book to have been issued by subscription in England was John Minsheu’s Ductor in Linguas (1617). By the end of the seventeenth century, publication by subscription had become a familiar method, although still something of a rarity in music.5 The earliest music-related work so far identified for which a subscription list survives is Thomas Mace’s Musick’s Monument (1676); this is also the only musical work issued in the seventeenth century for which a subscription list is known to exist. In the eighteenth century, it became increasingly common for musical works to be issued by subscription, so much so that by the year 1820 there had been around 750 music or music-related works issued this way for which a copy of the subscription list has survived to the present day. In order to reveal patterns of subscription to Avison’s work, I transcribed in a Microsoft Excel spreadsheet data containing the names and titles of his subscribers along with, where possible, their place of residence and/or position. I extracted this information from my personal collection of antiquarian music, held by Durham University Library, as well as from several other collections located at Durham Cathedral’s Dean and Chapter Library, Newcastle City Library, the Literary and Philosophical Library, Newcastle, and the British Library. I then used this data to produce  Fleming, “The Gender of Subscribers,” 131–152.  A good starting point for anyone wishing to investigate subscription lists are the indexes compiled by Peter Wallis. See Peter Wallis and Francis Robinson, Book Subscription Lists: A Revised Guide (Newcastle upon Tyne: H. Hill, 1977), and Peter Wallis and Ruth Wallis, Book Subscription Lists: Extended Supplement to the Revised Guide (Newcastle upon Tyne: Phibb, 1996). 4 5

312 

S. D. I. FLEMING

tabulations, pie charts, and scatter maps meant to translate into a visual form this information. Although, as Julia Flanders and Fotis Jannidis observe, “there is a long and rich tradition of gathering and modeling information as part of humanities research practice,”6 this type of visual approach is rare in musicology. Nevertheless, extracting data from paratextual material, such as subscription lists, and presenting it in visual form comes with undisputed benefits: such an approach “shows immediately the possible patterns the data can assume” and allows researchers to infer “the cultural significance that informs the data we create” by using digital tools.7 In the case of Avison and the concerto grosso, this approach has allowed me to conclude that Avison’s location in the North-East of England did not significantly impact on his ability to forge connections across Britain and beyond and revealed some interesting individual connections so far unknown to researchers of his work. Ultimately, this type of analysis has brought to light interesting patterns of music purchase and performance in eighteenth-­century Britain.

Avison’s Subscription Lists and His Popularity In total, Avison published six works by subscription, five of which are collections of concerti grossi. He also issued, as part of his Two Concertos, a keyboard concerto, his only published example in this genre. Avison was furthermore involved in preparing for publication the eight-volume English version of the Psalms of Benedetto Marcello, a project that he passed on to his friend and colleague, the Durham-based organist and composer John Garth (1721–1810).8 As a whole, the subscription lists attached to these works shed considerable light on Avison’s growth as a composer and the links he forged with a diverse range of individuals across Britain and beyond. As seen in Table 9.1, the first work that Avison issued by subscription was his 1740 op. 2 concertos. His subscription lists are, for the most part, attached to his early works, with the 1755 op. 4 being the last of these “early” publications (Figs. 9.1a–c). There is, then, a gap of over a decade 6  Julia Flanders and Fotis Jannidis, “Data Modeling in a Digital Humanities Context: An Introduction,” in The Shape of Data in Digital Humanities. Modeling Texts and Text-Based Resources, ed. Julia Flanders and Fotis Jannidis (Abingdon: Routledge, 2017), 3. 7  Ibid., 7, 10. 8  Although each volume in Garth’s edition of Marcello’s Psalms had a different list, only the list attached to the final volume has been included in Table 9.1. Avison also issued his Twenty Six Concertos ... in Score (Newcastle: Printed for the Author, 1758) by subscription. He had planned, according to the “Advertisement” in volume one, to include a subscription list in the fourth and final volume, but no copy with a list has been located.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

313

Table 9.1  Subscription lists to Avison’s publications (1740–1767). When the year printed on the title page was not, in fact, the year when the work was issued, the actual year of publication has been given in square brackets List Title Ref. A

B

C

D

E

F

G

H

I

Publisher

Newcastle, Joseph Barber Two Concertos the Newcastle, First for an Joseph Organ or Barber Harpsichord Twelve London, Concertos... Done for the from the Two Author Books of Lessons... by Sigr Domenico Scarlatti Six Concertos... London, Opera Terza John Johnson Eight London, Concertos…. John Opera Quarta Johnson Twelve London, Concertos... for the Opera Nona Author [Book 1-Issue 1] Twelve London, Concertos... for the Opera Nona Author [Book 1-Issue 2] Twelve London, Concertos... for the Opera Nona Author [Book 2] The First Fifty London, Psalms Set to John Music by Johnson Benedetto Marcello... Adapted to the English Version by John Garth. Vol. VIII

Six Concertos... Opera Secunda

Year published

Total Male Female Instit. No. no. of subscr. subscr. subscr. of subscr. copies

1740

170

149

14

7

173

1742

144

119

17

8

163

1744

151

125

20

6

157

1751

174

136

33

5

192

1755

198

150

38

10

218

1766

209

168

32

9

273

1766

229

184

35

10

293

1766 [1767]

253

194

48

11

317

1757 [1764]

122

86

30

6

148

314 

S. D. I. FLEMING

Fig. 9.1  (a–c) Subscription list to Charles Avison’s Eight Concertos, op. 4, 1755. Print from the author’s collection held at Durham University’s Palace Green Library

9  PUBLISHING MUSIC BY SUBSCRIPTION 

Fig. 9.1  (continued)

315

316 

S. D. I. FLEMING

Fig. 9.1  (continued)

9  PUBLISHING MUSIC BY SUBSCRIPTION 

317

until the appearance of his twelve op. 9 concertos, published in two parts in 1766 and 1767. Of the works issued during this “gap” period, none of Avison’s three sets of six accompanied keyboard sonatas (the opp. 5, 7, and 8, which were published respectively in 1756, 1760, and 1764) has lists, nor does his set of twelve op. 6 concertos from 1758.9 It appears that Avison, feeling financially stable and relatively affluent after the publication of his op. 4 concertos, had the means to finance the publication of some of these works himself. This certainly appears to have been the case with the opp. 6, 7, and 8, which were “Printed for the Author.” However, in the case of op. 5, it appears that publisher John Johnson was willing to assume, due to Avison being by the mid-1750s a well-known composer, that any publication under his name would be profitable, so he paid to produce this set himself. The fact that the two subsequent sets of sonatas were printed for Avison suggests that this first set did not meet Johnson’s expectations. As the op. 6 set contains re-workings of his op. 2 concertos, it could be Avison’s choice to finance this publication himself to avoid any disappointment to potential subscribers. As already mentioned, Avison’s final subscriber list was attached to the op. 9 concertos. This work was published in two sets, both of which have an individual list.10 This work is also something of an anomaly as the first set has two different subscription lists. What appears to have happened is that the first volume was published by John Johnson’s widow, Ruth, in March 1766. It was then reissued by her, probably later that same year, with an updated list. By the time the second volume appeared, in March 1767, she had been replaced as publisher by Robert Bremner, who brought out the second set with yet another updated list; he additionally reissued the first set, presumably once all copies of Johnson’s imprint had been sold.11 When comparing these three lists, it is clear that Avison 9  Charles Avison, Six Sonatas for the Harpsichord … Opera Quinta (London: Printed for John Johnson, 1756); Six Sonatas for the Harpsichord … Opera Settima (London: Printed for the Author, 1760); Six Sonatas for the Harpsichord … Opera Ottava (London: Printed for the Author, 1764); Twelve Concertos … Opera Sesta (Newcastle: Printed for the Author, 1758). 10  Set 1 contains the first six concertos and set 2 the remaining six (concertos 7 to 12). 11  I am grateful to Gordon Dixon who provided me with a copy of the list to the second issue of set 1, which he sourced from Newcastle’s Literary and Philosophical Society’s collection. I am also grateful to James Smith who allowed me to consult the original document. We do not know why Avison decided to switch publishers midway through the production of his op. 9, although it is possible that Ruth Johnson may have decided for some unknown reason not to produce the rest of the set. Alternatively, Avison himself might have been unhappy with some part of the publication process.

318 

S. D. I. FLEMING

received a good number of extra subscribers in the time between each set’s publication.12 It is also interesting to note that, in some cases, a name may change between lists. Mary Eleanor Bowes, who had married the 9th Earl of Strathmore and Kinghorne on 26 February 1767, subscribed in her maiden name to set 1, issue 2, but in her married name to set 2. Additionally, it was not unusual for subscriptions to be taken after the list of subscribers had been produced; thus, several lists, including the British Library copy of the first issue of set 1,13 have names added by hand. The four scribal names in that list were all incorporated into the printed list attached to the second issue. What is perhaps so unusual about the op. 9 is that there are not only three lists printed at different times for one single opus, but that there was also an impressive 20% increase in the number of subscribers between the first and final lists. The data in Table  9.1 also reveals a general growth in sales between 1740 and 1767, during which time Avison had almost doubled the number of copies he was selling by subscription. Avison’s ultimate success with op. 9 appears to have been due to several factors. First, by 1767 Avison’s reputation had grown significantly, so there were more who were willing to take a chance on the quality and usefulness of his music. Second, in his desire to increase the number of copies sold, Avison set out to make these works adaptable. They could be performed as keyboard solos, as Avison provided a partially realized harpsichord part, as well as quartets, so they were ideal for domestic purposes. Of the works with subscription lists, the earliest three were published in Newcastle while the later examples were produced in London. Avison presumably knew the publisher Joseph Barber personally, which may have led to his decision to have these editions published locally. The difficulty was, of course, that Newcastle was a considerable distance from London. The British capital was then one of the biggest markets for music and, if Avison wanted to increase his sales and make himself more widely recognized as a

12  Only one name, John Cuthbert of Newcastle, who was included in the list to the first issue of set 1, was removed from the second issue, presumably as he had reneged on his promised payment. It is possible that Cuthbert had died but, if he had, he had almost certainly not paid for his copy. It was not unusual for the names of the recently deceased to appear in subscription lists, sometimes with the addition of the designation “late.” This happens in the lists to Marcello’s Psalms, to which Sir John Dolben subscribed. Dolben died in 1756, before the first volume was issued, but his name appears in all the eight lists. 13  GB-Lbl: Music Collections g.256.h.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

319

composer, it was to London that he needed to look.14 The earliest two works issued by Barber were, according to the title pages, also available for purchase from the London publisher Benjamin Cooke, who took subscriptions for both collections.15 Cooke died c.1743, so the Scarlatti concertos were simply recorded as “Sold by the Musick Shops in Town [London].”16 By the time Avison’s op. 3 appeared, he had switched allegiance to John Johnson who, as well as publishing all of Avison’s new works and Marcello’s Psalms, also reissued the op. 1 sonatas.17 Avison’s final publisher, Robert Bremner, produced the complete op. 9 and the op. 10 concertos.18 Bremner furthermore reissued the op. 7 sonatas, but it is unclear whether his imprint appeared before or after Avison’s death. The growth in Avison’s popularity in Britain can be seen not only in the number of subscribers or copies purchased but also in the spread of his work, which can be estimated by considering the geographical places that some subscribers provided alongside their names (Figs. 9.2a–c).19 Where a place name is provided, it was most commonly the town in which the subscriber lived, or the name of a larger town near where the subscribers lived or worked. Organists tended to give the town in which the church where 14  The population of London c.1750 was around 675,000. The next biggest town, Norwich, had a population of around 50,000. For details, see E. Anthony Wrigley, “Urban Growth and Agricultural Change: England and the Continent in the Early Modern Period,” in The Eighteenth-Century Town: A Reader in Urban History 1688–1820, ed. Peter Borsay (London: Routledge, 2013), 42. The book trade was London-centric, with well-developed links that enabled the spread of books out into the provinces, but it was not geared for transportation in the opposite direction. See also David Shaw, “Canterbury’s External Links: Book-Trade Relations at the Regional and National Level in the Eighteenth Century,” in The Mighty Engine: The Printing Press and Its Impact, ed. Peter Isaac and Barry McKay (Winchester: St Paul’s Bibliographies, 2000), 110–11. 15  Daily Gazetteer, 7 January 1742. Cooke also published the first edition of Avison’s op. 1 trio sonatas. Charles Avison, VI Sonatas (London: Printed by Benjamin Cooke, c.1737). 16  William Smith and Peter Ward Jones, “Cooke, Benjamin (i),” Grove Music Online, Oxford Music Online, Oxford University Press, http://www.oxfordmusiconline.com/subscriber/article/grove/music/06393 17  Charles Avison, Six Sonatas … Opera Prima (London: Printed for John Johnson, c.1757). 18  Charles Avison, Six Concertos … Opera Decima (London: Printed for Robert Bremner, 1769). 19  It should be noted that, in a significant number of cases, the subscriptions give no indication of where an individual lived. This makes it particularly difficult to ascertain who the subscribers with a common name were, although, in Avison’s case, many would have resided within Newcastle’s environs.

320 

S. D. I. FLEMING

Fig. 9.2a  Location of subscribers to Avison’s op. 2 (1740 - List A)

they were employed was located. For example, in Avison’s lists, Garth generally describes himself as the “Organist of Sedgefield.” A few are more precise and give the name of the house in which they lived, while others simply give the county. When these place names are plotted on a series of scatter maps (Figs. 9.2a–c), it becomes apparent just how much Avison’s reputation as a composer grew between 1740 and 1767: As expected, there was a core of Avison’s subscribers who resided in the north of England. Nevertheless, by the publication of the op. 2, Avison’s reputation had spread a significant distance from Newcastle, with subscribers located as far north as Edinburgh and as far south as Lincolnshire and Norfolk. Although London was not recorded in that list, there were subscribers, such as Benjamin Cooke, who lived there. Additionally, as seen in Fig. 9.2a, there is a noteworthy anomaly created through the subscription of Thomas Chilcot, who resided in Bath. By the time of the op. 4, Avison’s reputation had spread further north to Aberdeen and had greatly expanded across England (Fig. 9.2b), but it is with the publication

9  PUBLISHING MUSIC BY SUBSCRIPTION 

321

Fig. 9.2b  Location of subscribers to Avison’s op. 4 (1755 - List E)

of op. 9 that a significant growth of Avison’s reputation as a composer occurred. As seen in Fig. 9.2c, he had a sizeable number of subscribers from Scotland and significantly more from across England, with copies purchased by subscribers as far away as Devon and the south coast. A few subscribers gave their place of residence as being located outside the British mainland. This includes William Avison, possibly Charles’ brother, who, in the mid-1760s, resided at the Baltic seaport of Danzig.20 Many of these subscribers were merchants and presumably traded in Newcastle, which might be how Avison came into contact with them; the fact that they individually subscribed to relatively few works indicates that Avison might have sought their subscription during an opportune visit to 20  Southey believes that William was Avison’s brother, but another possibility is that he was Avison’s cousin. See Roz Southey, Margaret Maddison, and David Hughes, The Ingenious Mr. Avison: Making Music and Money in Eighteenth Century Newcastle (Newcastle upon Tyne: Tyne Bridge Publishing, 2009), 24, and G. H. Smith, A History of Hull Organs and Organists (London: A. Brown and Sons, n.d.), 9.

322 

S. D. I. FLEMING

Fig. 9.2c  Location of subscribers to Avison’s op. 9, set 2 (1767 - List H)

Newcastle. The prevalence of merchants in Newcastle is evident from the following description, which records that St. Nicholas’ Church, where Avison was organist, had “one of the most numerous Congregations in the kingdom; consisting greatly of seamen, who, if they ever learned to sing in parts, could probably only learn it in Dutchland [sic].”21 Another indication of the popularity of Avison’s concertos across Britain lies in the number of musical societies that subscribed to his works. During the eighteenth century, there was a significant growth in the number of such groups. These societies, which tended to include a mix of professional and amateur musicians, came together on a regular basis to make music, often at a local tavern. Avison was himself a member of the Newcastle Musical Society and it was at their meetings that his op. 1 trios were performed. They even encouraged Avison to have these pieces published: as mentioned in the dedication, he “had then no thoughts of their 21  The Works of William Mason, M.A. (London: Printed for M.  Cadell and W.  Davies, 1811), 3: 385.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

323

Table 9.2  Musical societies which subscribed to Avison’s publications. The figure in brackets refers to the number of copies the society purchased Musical society

List Ref.

Aberdeen Cambridge (Master of Arts’ Club) Carlisle Darlington Derby Dublin (The Philharmonic Society) Dundee Durham Edinburgh

D (2), E, F, G, H E B, C, D F, G, H F, G, H A G, H A, B A, B, E, F (4), G (4), H (4), I D (2) D, E C A, B, E E, H A, B F (3), G (3), H (3), I (2) C (3)

Edinburgh (at St. Mary’s Chapel) Fakenham Glasgow Hull Lichfield Lincoln London (The Academy of Ancient Music) London (The Philharmonick Society at the Crown and Anchor in the Strand) Newcastle Newcastle (in the Close) Norwich Nottingham (The Senior Musical Society) Oxford Spalding Whitehaven York York (in Blake Street)

A, B F, G, H A (3), B (2), C (2), E F, G, H C, D, E (2), I (2) B F, G, H C (2) D, E (2)

ever being made Publick, but beyond expectation meeting with some applause in private and being importun’d by the Musical Society in Newcastle to publish them, I could no longer refuse to comply.”22 The index of musical societies from the subscription lists is given in Table 9.2, which outlines not only how far Avison’s reputation spread but also how highly he was regarded by some individual groups who subscribed to more than one work and purchased multiple copies of some sets.  See Avison, Six Sonatas … Opera Prima, n.p.

22

324 

S. D. I. FLEMING

The Aberdeen Musical Society, for example, subscribed to everything Avison published from op. 3 onward. That town had a vibrant musical society, founded in 1748; one of its founding members, the dancing master Francis Peacock, was an individual subscriber to Avison’s music.23 An earlier musical society had existed at Edinburgh since at least 1727 and was formally established the following year.24 Avison corresponded with this group, writing a letter in 1756 in which he revealed that it gave him “a particular satisfaction to find by the Favor of yours that my Compositions are still approved in Edinburgh.”25 Edinburgh appears twice in Table 9.2, but it is likely that the society based at St. Mary’s Chapel is the same as that which subscribed to his other works; the addition of a name would have been a way to distinguish itself from another similar group. The Edinburgh Society subscribed to all of Avison’s publications, except for the Scarlatti concertos, probably because they found Avison’s music, with its easier ripieno parts, ideal for their players.26 One of the most notable societies to subscribe was the one in Oxford, which had existed since the seventeenth century. By 1740, this musical society had grown to such an extent that they required their own purpose-­ built accommodation, the Holywell Music Room, which opened in 1748.27 The Oxford group was usually led by a Professor of Music who, from 1741, was William Hayes.28 Hayes had a notoriously low opinion of Avison’s music, going so far as to savage the op. 3 concertos in his 1753 23  Henry George Farmer, Music Making in the Olden Days: The Story of the Aberdeen Concerts 1748–1801 (London: Peters-Hinrichsen, 1950), 13–15. 24  Martin Hillman, Thomas Sanderson’s Account of Incidents: The Edinburgh Musical Society 1727–1801 and Its Impact on the City (Edinburgh: Friends of St. Cecilia’s Hall, 2017), 15–16. 25  The National Archives of Scotland: GD113/5/208/10, a letter from Charles Avison to the Edinburgh Musical Society, dated 16 July 1756. Reference sourced from the introduction to Charles Avison in Context: National and International Musical Links in EighteenthCentury North-East England, ed. Roz Southey and Eric Cross (Abingdon: Routledge, 2018), 12. 26  The concerto grosso was a popular form in eighteenth-century Britain due to the nature of concert orchestras, which would perform with little rehearsal time beforehand. The main body of the orchestra would play the simpler ripieno parts leaving the more difficult concertino parts to the concert organizers or other billed performers. 27  Susan Wollenberg, Music at Oxford in the Eighteenth and Nineteenth Centuries (Oxford: Oxford University Press, 2001), 44–49. 28  See Wollenberg, Music at Oxford, and Peter Ward Jones and Simon Heighes, “Hayes Family,” Grove Music Online, Oxford Music Online, Oxford University Press, http://www. oxfordmusiconline.com/subscriber/article/grove/music/12621pg1

9  PUBLISHING MUSIC BY SUBSCRIPTION 

325

pamphlet, Remarks on Mr. Avison’s Essay on Musical Expression.29 The Oxford Musical Society subscribed to a number of Avison’s works, including the Scarlatti concertos, the op. 3, and the 1755 op. 4 concertos. As the society’s rules set out that “[t]he Steward … shall not … subscribe to any Books to be published … without an Order from a Meeting,” it is unlikely that their subscription would have been placed unless a consensus had been reached among the society’s members.30 The fact that they subscribed to op. 4, the collection issued in the wakes of Hayes’ attack, gives some indication that a sizeable number of Hayes’ immediate musical circle rejected Hayes’ opinions. For Avison, if anything, the dispute with Hayes made him even more widely known, with the Spalding Gentlemen’s Society commenting on the matter at one of their 1753 meetings.31 The debate appears to have also boosted sales of his music, with Avison reflecting, in 1756, that “my own Countrymen would have obstructed them in their way … but as they have failed in their Design, I have reaped the Benefit of a considerable Sale, which is at once the most certain and substantial Proof of the unprejudiced Approbation they have at last met with.”32 It is perhaps no coincidence that, in the mid-1750s, Avison capitalized on his increasing notoriety by reissuing his earlier works, with the op. 4 concertos selling well enough to require a reprint. As seen in Table 9.2, some musical societies subscribed to more than one set of an individual publication and there are several reasons why this might have been the case. A few may have been large enough to warrant more than one desk to some parts. Of course, multiple copies could have been purchased if some of the society’s members wanted their own copy, but the purchase was made through the society itself; likewise, some individuals who subscribed might have used their parts at a society meeting. Several musical societies subscribed to more than one copy of some publications, including those at Aberdeen, Edinburgh, Oxford, and York. The Norwich Musical Society stands out as they purchased three sets of op. 2, 29  William Hayes, Remarks on Mr. Avison’s Essay on Musical Expression (London: Printed for J. Robinson, 1753). 30  John Mee, The Oldest Music Room in Europe: A Record of Eighteenth-Century Enterprise at Oxford (London: John Lane Company, 1911), 49–50. I am grateful to Simon Heighes for his advice and for pointing out this reference. 31  Simon D.  I. Fleming, “The Musical Activities of the Spalding Gentlemen’s Society,” Royal Musical Association Research Chronicle 48, no. 1 (2017): 78, https://doi.org/10.108 0/14723808.2016.1271572 32  The National Archives of Scotland: GD113/5/208/10.

326 

S. D. I. FLEMING

two sets of the Two Concertos and the Scarlatti concertos, and one set of op. 4; this could be taken as evidence of a reduction in their numbers. Edinburgh, in contrast, purchased an impressive four copies of op. 9. This society was particularly large: in 1728 it had 70 members, a number which had risen to 195 by 1775.33 Presumably, many were there to hear the music and socialize, rather than to play, but their subscription to four copies indicates that they may potentially have had in the region of eight players on some parts if, as per modern practice, two performers shared one part placed on a stand.34 Other societies subscribed to a single publication, and for several it was the op. 9 that they purchased. It is possible that some, such as Derby, may not have encountered Avison’s music until a later date and certainly saw no need, if they were aware of him before then, to subscribe to anything earlier. However, some societies might not even have existed when the earlier works were issued. For example, the musical society at Whitehaven was probably organized, if not established, by William Howgill. He was appointed organist at the town’s St. Nicholas’ Church in 1756, the same year an organ was installed. Howgill appears to have known Avison personally, having formerly spent time in Newcastle, on which instance he subscribed to Avison’s op. 4. He presumably used his personal copies of these concertos in his Whitehaven concerts.35 Table 9.2 also highlights that some societies subscribed to an early publication but not to any later works by Avison. This could have been due to changes in taste, but these groups could possibly have experienced a drop in their numbers, which meant that they were either unable to afford the subscription or no longer had the numbers required to perform concertos. This appears to have been the case with the musical society at Spalding, which subscribed to Avison’s Two Concertos, but, by 1747, had been “reduced to very few performers.”36 In the case of Durham, their musical society subscribed to Avison’s first two works but to none of his later works. This may have been due to the longstanding dispute that existed 33  Peter Holman and Richard Maunder, “The Accompaniment of Concertos in 18th-Century England,” Early Music 28, no. 4 (2000): 646, http://www.jstor.org/stable/3519000 34  It is also possible that, if more parts were required, these could be produced in manuscript. See Holman and Maunder, “The Accompaniment of Concertos,” 645–46. 35  For details, see Simon D.  I. Fleming, “The Howgill Family: A Dynasty of Musicians from Georgian Whitehaven,” Nineteenth-Century Music Review 10, no. 1 (2013): 61–62, https://doi.org/10.1017/S1479409813000049 36  Fleming, “The Musical Activities of the Spalding Gentlemen’s Society,” 74.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

327

between Avison and the organist of Durham Cathedral, James Hesletine. The source of this argument appears to originate in 1742, when Avison proposed a series of concerts that coincided with Durham’s race week.37 Hesletine, who had a fiery temperament and was, in the words of Durham’s Dean, Spencer Cowper, unable “to bear a Competitor,” subsequently refused to participate in any musical event that involved Avison.38 Although Hesletine never personally subscribed to any of Avison’s publications himself, he did subscribe to Marcello’s Psalms, presumably after Garth took over their publication from Avison. Since they were both resident in Durham, Helestine’s relationship with Garth appears to have been less strained.

Individual Subscribers: Categories and Gender Besides institutional subscribers, Avison also attracted a good number of individual subscribers to his published concertos. A significant proportion of these were, understandably, professional musicians, with all those identified given in Table 9.3. These musician subscribers can be subdivided into three distinct groups. First, there are those who appear to have subscribed reciprocally to Avison’s works.39 In the second group are those who worked or studied in the North-East and knew Avison personally; they were probably asked to subscribe and may even have felt obligated to do so whether they had any use for his music or not. Finally, there are those with no known connection with Avison. Among them, the first category is of interest as it reveals at least one previously unknown relationship: that with Thomas Chilcot. Thomas Chilcot, the organist at Bath Abbey, subscribed to all of Avison’s publications, except for op. 3, and Avison returned the favor by subscribing to two publications by Chilcot. What is unusual is that Avison  Newcastle Journal, 17 July 1742.  Edward Hughes, ed., Letters of Spencer Cowper, Dean of Durham, 1746–74 (Durham: Andrews, 1956), 159; Simon D. I. Fleming, “A Century of Music Production in Durham City 1711–1811: A Documentary Study” (PhD diss., University of Durham, 2009), 61–70, 112. 39  Of the 91 musicians who subscribed, Avison subscribed to works by 13 of them. I am grateful to H. Diack Johnstone, who provided references to most of the works that Avison subscribed to. Other references were sourced from Southey, Maddison, and Hughes, The Ingenious Mr. Avison, 106–108, and from Otto Haas, Recueil de Different Airs, May 2017, http://www.ottohaas-music.com/SO_2018.May.pdf, accessed 4 July 2018. 37 38

328 

S. D. I. FLEMING

Table 9.3  Musicians who subscribed to Avison’s publications Musician subscriber

List Ref.

Thomas Arne William Ayrton, organist of the cathedral, Ripon Edward Avison, Newcastle [brother of Avison] William Avison, organist at Hull [brother of Avison?] [Francesco] Barsanti [Thomas] Beilby, organist at Scarborough [Edward] Bets, organist at Manchester [Peter] Blenkinsop, of Durham [cathedral lay-clerk] [Capel] Bond, organist at Coventry William Boyce, composer, and one of the organists of his Majesty’s Chapel-Royal Ralph Brockett, of Durham [former cathedral chorister] [William] Broderip, organist at Bristol [Charles] Burney [John] Camidge, organist of the cathedral, York Mr. Carr, organist at Leeds [Giuseppe] Cattanei Mr. Charles and Son, French Horns Thomas Chilcot, organist at Bath [John] Clark, of the Choir of Durham James Clarke, organist of St. John’s in Newcastle Stephen Clark, organist at Dundee Walter Cottingham, organist at Newark upon Trent William Cowley [played organ at Sedgefield] [Miles?] Coyle, York Mr. Denby, organist at Wakefield [William] Denby, organist at Derby [Abraham] Dobinson, organist of the cathedral, Carlisle William Douglas, Edinburgh [Thomas] Ebdon, organist of the cathedral, Durham William Felton Michael Christian Festing John Garth, organist at Sedgefield Cornforth Gelson, Durham [cathedral lay-clerk] [Francesco] Geminiani [Felice] Giardini Maurice Greene [Barnabus] Gunn, organist at Birmingham Henry Hargrave, Nottingham Matthias Hawdon, Newcastle then organist at Hull [Thomas] Haxby, York

A, B F, G, H, I A, B, C, D, E, F, G, H A, B, C, D B, D F, G, H A D H B, I D, E, I F, G, H G, H F, G, H B D, E E A, B, C, E, F, G, H F, G, H A G, H A, B, C E D A F, G, H A D F, G, H C A, B, C A, B, C, D, E, F, G, H D, E A, C, D E (2), F (2) G (2), H (2), I C, D B F, G, H D, E, F (2), G (2), H (2), I F (7), G (7), H (7) (continued)

9  PUBLISHING MUSIC BY SUBSCRIPTION 

329

Table 9.3 (continued) Musician subscriber

List Ref.

John Hebden, York [Musgrave] Heighington Claudius Heron, [Newcastle then] London [William] Herschel, organist at Halifax William Howgill, Newcastle then organist at Whitehaven Mr. Humberston, organist at Yarm [William] Jackson, Exeter [Robert] Jobson, organist at Wakefield John Jones, organist of the charter-house Thomas Jones, dancing-master, in Newcastle [Richard] Justice, organist at Hull Mr. King, organist at Stockton [Thomas] Linley, Bath William Linley, Bath [Joseph] Mahoon, harpsichord-maker to his Majesty [Bailey] Marley, organist at Hull John Marshall, Durham [cathedral lay-clerk] [John] Matthews, of the Choir of Durham [Benjamin] Milgrove, Bath [James] Nares, organist of the cathedral, York [Giovanni] Noferi [Niccolo] Pasquali, Cambridge Henry Ogle, organist at Liverpool [Thomas] Orpin, organist at Devizes [Robert] Page, Newcastle [Stephen] Paxton, London [formerly of Durham] [William?] Paxton, Durham [Francis] Peacock, Aberdeen [Thomas] Perkins, of York John Randall, organist of King’s College, Cambridge [David] Richards, Bath [William] Rogers, Bath Thomas Roseingrave [George] Rush, composer of music [John] Scamardine, Grantham Francis Sharp, Stamford Joseph Shaw, York, [Thomas] Shaw, Bath Burkat Shudi Joshua Shudi John Simpson, Newcastle

A, B C A, B, C, D, E, F, G, H F (4), G (4), H (4) D, E, F, G, H, I F, G, H F (2), G (2), H (2), I F, G, H E, I A E F, G, H F, G, H F, G, H B F, G, H F, G, H F, G, H F, G, H A, B, C, D, E F, G, H B E F, G, H F, G, H F, G, H E F, G, H D E, I F, G, H, I F, G, H A, D F, G, H F, G, H F, G, H F (2), G (2), H (2) F, G, H C, E, F (3), G (3), H (3), I D F, G, H (continued)

330 

S. D. I. FLEMING

Table 9.3 (continued) Musician subscriber

List Ref.

[Timothy?] Smart, London [John] Snetzler [organ builder] Mr. Stobbs, organist, Birmingham Abraham Taylor, Durham [cathedral lay-clerk] William Tireman, [a York wait then] organist of Trinity College, Cambridge [John] Travers, one of the organists of his Majesty’s Chapel-Royal [Bernard] Turner, organist of St. John’s, Cambridge John Valentine, Leicester [John] Wainwright, organist in Manchester [John] Worgan [organist in London]

F, G, H F (3), G (3), H (3) H A, B, C A, B, F, G, H B B F, G, H E, I G, H

All biographical data included in this table has been derived from subscription lists and has been standardized. In most cases, the lists omit Christian names; where this is the case, they have been added in square brackets with the title. If only “Mr.” or “Segnor” is given in the source, the reference is omitted. Finally, surnames occasionally spelt inconsistently have been made consistent

took five copies of Chilcot’s Twelve English Songs, which not only raises the question as to why Avison would require so many prints but also why he would purchase so many copies of a work he had not seen. It seems possible that Avison may have taken the extra copies to support a fellow musician, perhaps one he knew quite well; it is also possible that Chilcot may have asked Avison to provide feedback on a manuscript copy of these songs before they were published, a not uncommon practice at the time.40 According to Tim Rishton, Chilcot rarely travelled far from Bath, so it is unlikely that he would have visited the North-East, but a link may be established through Masonic circles.41 Chilcot was certainly a freemason and evidence suggests that Avison may have been one as well.42 What is 40  For example, according to the preface, William Flackton sent a manuscript of his Six Solos. Op. 2 (London: Printed for the Author, 1770) to Carl Friedrich Abel to “inspect” before publication. 41  Tim Rishton, “Chilcot, Thomas,” Grove Music Online, Oxford Music Online, Oxford University Press, http://www.oxfordmusiconline.com/subscriber/article/grove/ music/05581 42  Gwilym Beechey, “Thomas Chilcot and His Music,” Music & Letters 54, no. 2 (1973): 184, https://www.jstor.org/stable/734363. I am grateful to Tim Rishton who suggested the Masonic link. There is no firm evidence that Avison was himself a freemason although Garth certainly was, as was Avison’s youngest son, Charles. See Simon D.  I. Fleming,

9  PUBLISHING MUSIC BY SUBSCRIPTION 

331

perhaps even more striking about Bath is the number of subscribers from that place who purchased a copy of Avison’s op. 9. As a major spa town, Bath had one of the most vibrant musical scenes in provincial Britain. Many of those who visited the town would take the waters in the “Pump Room,” where music was provided as entertainment. The musicians who formed the “Pump Room Band” were also the core of Bath’s concert orchestras, with 63% of this rather small group subscribing to op. 9.43 The subscribers were Thomas Linley (who played violin), Benjamin Milgrove (French horn), Thomas Shaw (clarinet/viola/violin), David Richards (cello/violin), and William Rogers (trumpet/horn).44 No Bath-based group subscribed to anything by Avison, yet there was clearly a relationship of sorts between Avison and Bath’s musicians. It is within the realms of possibility that Avison visited the town, during which time he acquired these subscriptions. Although there is no concrete evidence to confirm this, research has underscored just how well-traveled Avison was, having taken excursions to Carlisle, London, Derwentwater, Edinburgh, and Stirling. Bath would have been a more challenging endeavor, but he may have made the journey if, in addition to visiting Chilcot, he or his wife, Catherine, wished to take the waters for health reasons.45 Another notable musician with whom Avison appears to have had some contact is Carl Friedrich Abel. Although Abel did not subscribe to any of Avison’s publications, Avison did subscribe to Abel’s op. 2 sonatas, purchasing six copies.46 Other subscribers to this work from the North-East include Garth, who purchased four copies, and Lady Milbanke. As Abel “Charles Avison Jnr and his Book of Organ Voluntaries,” The Musical Times 153, no. 1918 (2012): 99, http://www.jstor.org/stable/41442085; and Simon D.  I. Fleming, “John Garth and His Music: An Important Provincial Composer from 18th-Century Britain,” The Musical Times 153, no. 1921 (2012): 66, https://www.jstor.org/stable/41703543 43  See Robert Hyman and Nicola Hyman, The Pump Room Orchestra Bath: Three Centuries of Music and Social History (Salisbury: Hobnob Press, 2011), 3–4. 44  In 1767 this orchestra consisted of eight musicians. For details, see Hyman and Hyman, The Pump Room Orchestra Bath, 16–19. For more on Milgrove, see Matthew Spring, “Benjamin Milgrove, the Musical ‘Toy Man’ and the ‘Guittar’ in Bath 1757–1790,” Early Music 41, no. 2 (2013): 14, https://doi.org/10.1093/em/cat044 45  Catherine had a “lingering indisposition” which ultimately led to her death in 1766. See Simon D. I. Fleming, “Charles Avison (1709–1770): An Important and Influential English Composer, Musician, and Writer” (MMus diss., University of Liverpool, 1999), 48. 46  Charles Frederick Abel, Six Sonatas for the Harpsichord … Opera II (London: Printed for the Author, 1760). A copy of this list was supplied with the courtesy of Otto Haas. I am also grateful to Colin Coleman who arranged for this copy to be made.

332 

S. D. I. FLEMING

had relocated to London by 1759,47 one wonders if he may have solicited Avison’s and Garth’s subscriptions during a visit to the north of England at around that time. Of the musicians based in the North-East, there were a few who subscribed to all of Avison’s publications, including Garth, Edward Avison, and Claudius Heron, the last of whom was a Newcastle-based cellist.48 Hestletine’s successor as organist at Durham Cathedral, Thomas Ebdon, did not share his predecessor’s prejudices and had no qualms about subscribing to op. 9. Other Durham Cathedral musician subscribers include Peter Blenkinsop and Cornforth Gelson; their subscription preceded a later decision in which they switched allegiance from the cathedral choir’s concert series to a rival Durham series run by Garth in partnership with Avison.49 There were also subscribers from York, including two of the cathedral’s organists, James Nares and John Camidge, the concert promoter John Hebden, and the instrument-maker Thomas Haxby. The future astronomer William Herschel also appears, having been for a time bandleader of the Durham Militia.50 There were also a few musician subscribers who were not based in the North-East but who visited to perform at concerts. Felice Giardini, for instance, played violin at Avison’s Newcastle concerts during the 1750s; in the 1760s, he was succeeded by another subscriber, Giovanni Noferi. Another subscriber was Mr. Charles, the Hungarian horn player, who had passed through Newcastle in 1755, at which point Avison must have used this opportune moment to seek his subscription to op. 4.51 Finally, there is a sizable number of musician subscribers who have no known connection with Avison, such as the Newark organist Walter Cottingham. Unless he was particularly enamored with Avison’s music, one suspects that there was a relationship of sorts between the two men, since he purchased the earliest three sets issued by subscription. One likewise suspects that Avison was in contact with the Shudi family of 47  Walter Knape, Murray R. Charters, and Simon McVeigh, “Abel Family,” Grove Music Online. Oxford Music Online. Oxford University Press, https://doi.org/10.1093/ gmo/9781561592630.article.00035 48  Edward was presumably Avison’s elder brother rather than his eldest son. See Roz Southey, Music-Making in North-East England during the Eighteenth Century (Aldershot: Ashgate, 2006), 211. 49  Fleming, A Century of Music Production in Durham, 66, 69. 50  Ibid., 164, 171. 51  Ibid., 97–99; Newcastle Courant, 28 December 1754.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

333

instrument-­makers, with either Burkat or Joshua Shudi subscribing to everything from the Scarlatti concertos. Another subscriber was Francis Sharp II of Stamford. Sharp’s subscription is of interest as he is unlikely to have known Avison well, even though Avison would certainly have passed through Stamford on his way to and from London. However, Sharp acted as an agent in Stamford, taking subscriptions for various musical publications, including Henry Hargrave’s Five Concertos, and may have acted in this capacity for Avison’s op. 9 as well.52 Employing agents was just one way to attract subscriptions to new publications. A composer or publisher would often place an advertisement in a newspaper, naming the agent. The following example, which appeared in the London press, announced the publication of Avison’s Two Concertos: In February next will be published, THE First Yearly Number of CONCERTO’s (Compos’d by Mr. CHARLES AVISON, of Newcastle upon Tyne.) Containing one for the Harpsichord or Organ, the other for Violins, &c. The Price of those that do not subscribe will be Four Shillings; and Subscribers at 3s. 6d. will be taken in by Mr. Benj[amin] Cooke in New-­ street Covent-garden, and Mr. Walmsley in Piccadilly. N.B.  As in the Course of this Work there will be Concerto’s for a Harpsichord, Violoncello, German Flute, &c. These will consist of Eight Parts, and those for Violins of Seven.53

Another method was to produce a printed notice that could be forwarded to potential subscribers to encourage them to pre-order a copy. Avison is known to have used this method for op. 9, as a rare example survives in the papers of John Grimston of Kilnwick. Avison was clearly successful in this endeavor as Grimston subscribed to the second set: To be Published, by SUBSCRIPTION, a NEW WORK, of TWELVE CONCERTOS, In FOUR PARTS; Viz. For two VIOLINS, one ALTO VIOLA, and a VIOLONCELLO, figured for the Harpsicord.  London Chronicle for the Year 1763, 446. See also Simon D.  I. Fleming, “Music and Concert Promotion in Georgian Stamford,” The Consort 73 (2017): 61–83. 53  Daily Gazetteer, 7 January 1742. Avison also placed advertisements for works to be published by subscription in the local press. See, for example, the Newcastle Courant, 1 September 1750, where there is an advertisement for op. 3. 52

334 

S. D. I. FLEMING

Composed by CHARLES AVISON, ORGANIST, in Newcastle. OPERA NONA. These CONCERTOS will be divided in two Sets: The first to be published in March, 1766; and the other, the Year following. The SUBSCRIPTION (Half a Guinea each Set) to be paid on Delivery of the Books. This WORK is calculated for the Service of both public and private Concerts; and may be also adapted to the Practice of the Organ and Harpsicord. SUBSCRIPTIONS will be taken by Mrs. JOHNSON, Mr. WALSH, and Mr. BREMNER, in London; Mr. Haxby, in York; and by the Author, in Newcastle. March, 1765.

To the foot of this printed announcement, Avison added the following handwritten note: Sir, I beg Pardon for ye Liberty of this Application for ye Honour of your name among my Subscribers to ye last Work I shall probably offer to ye Public, the first Part of which will be published ye latter End of this Month. I am, Sir, yo[u]r most obed[ien]t humble Serv[ant]t. Cha[rle]s Avison Newcastle. March 3d. 1766.54

Besides musicians, a considerable number of clergy were also subscribers to Avison’s works. In addition to being university-educated, many were not only able musicians but also drawn toward musical activities as amateurs. Additionally, given the high incomes that many received, they could afford to indulge in musical pursuits and subscribe to the latest published works; some, such as William Felton, were also active as composers and had their music published. Many of the clergy subscribers would have known Avison personally. A few were also associated with Durham Cathedral, making it clear that Hesletine’s opinions had little or no influence on them, although it is notable that Spencer Cowper, a staunch supporter of Hesletine and the cathedral choir’s concerts, only subscribed to Marcello’s Psalms. John Sharp, a Prebendary at Durham and then 54  East Riding Archives: DDGR/42/16/23. I am grateful to Helen Clark who kindly provided me with a copy of this document.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

335

Archdeacon of Northumberland, subscribed to all of Avison’s publications; John’s father, Thomas, was also a subscriber. Additionally, there was the prebendary, Sir John Dolben, and the minor canon, Jonathan Branfoot. Other vicar subscribers were involved with a literary club at Carlisle, a group Avison was also associated with. They include James Farish and John Brown, both of whom subscribed to all of Avison’s publications; another member of this group, Captain Leonard Smelt, subscribed to op. 3 and op. 9.55 One also suspects that Avison met Abraham Dobinson, the organist at Carlisle Cathedral, on one of his visits there, when he subscribed to Avison’s op. 2. Other clergy to work in the North-East include two vicars choral from York Minster, Bryan Allet and Joseph Bridges, along with William Pawson, who had been an assistant curate at York, Spencer Madan, a vicar at Bossall near York, and William Becher, a deacon and then priest at Brodsworth, South Yorkshire. There was also John Darch, a vicar at Long Benton, Newcastle, Nathaniel Clayton, a “lecturer” at Newcastle’s St. John’s Church, and Utrick Lowthian, a vicar of Whitfield, Northumberland.56 Some clergy subscribers were authors themselves and not only published their own writings but also may have contributed to Avison’s important 1752 treatise, An Essay on Musical Expression.57 William Mason, the vicar of Hull, claimed a significant passage of the Essay, substantiating both William Hayes’ accusation and Avison’s response that the work was the product of a “junto.” Norris Stephens also suspects that John Brown may have been another contributor, although the Scottish antiquarian, John Callander, a close friend of Avison and a subscriber to his op. 9 concertos, denied Brown’s involvement.58 Another subscriber and, according to Stephens, possible contributor to Avison’s Essay, was John Jortin, who

55  William Jackson, ed., Memoirs of Dr. William Gilpin (London: Bernard Quaritch, 1879), 74–81. 56  The positions of these subscribers were ascertained through the Clergy of the Church of England Database, accessed 10 December 2019, http://theclergydatabase.org.uk/ 57  Charles Avison, An Essay on Musical Expression (London: C. Davis, 1752). 58  Norris Stephens, “Charles Avison: An Eighteenth-Century English Composer, Musician and Writer” (PhD diss., University of Pittsburgh, 1968), 61; Simon D. I. Fleming, “John Callander and the Avison Connection: A Recently Rediscovered Letter,” Eighteenth Century Music 11 (2014): 285, https://doi.org/10.1017/S1478570614000086. Callander purchased two copies of op. 9.

336 

S. D. I. FLEMING

penned an account on the “Music of the Ancients” that was attached to the second and third editions of Avison’s Essay.59 Avison, like other composers, attracted a large number of subscribers from the aristocracy and gentry as well (Table 9.4). What is perhaps most interesting about this group is that most of them only subscribed to one or two of his works. Subscribing to a musical publication was, for many of the upper class, an act that demonstrated their patronage of the arts, and some would certainly have subscribed for outward show. Indeed, many did not have a particularly strong attachment to Avison, as indicated by their subscription to only one or a small number of his works, although some subscribers in this group are known to have been closely associated with Avison. Walter Blackett, for instance, supported Avison by subscribing to everything except for the Two Concertos. Blackett was an Alderman and the MP for Newcastle; his wife, Lady Elizabeth Blackett, was the dedicatee of Avison’s op. 5 sonatas. Another committed subscriber was Sir Ralph Milbanke, whose wife, Lady Anne, was the dedicatee of op. 4. Yet another devotee of Avison’s was Walter Scott, MP for Roxburghshire, a man with whom Avison, as far as we are aware, had no other connection.60 He purchased copies of all Avison’s concertos except op. 2. Ralph Jenison, for whom Avison had worked in his youth and who had provided opportunities for him to study music in London, was a subscriber and the dedicatee of Avison’s op. 1.61 There was, furthermore, Robert Shaftoe, an MP whose family estate was located at Whitworth, near Durham,62 and George Bowes, who lived at Gibside near Newcastle; Bowes’ wife was the 59  Stephens, Charles Avison, 60. Charles Avison, An Essay on Musical Expression, 2nd ed. (London: Printed for C.  Davis, 1753); Avison, An Essay on Musical Expression. 3rd ed. (London: Printed for Lockyer Davis, 1775). Other clergy subscribers with no known association with Avison include Robert Eden of Winchester and Edward Howkins of Cambridge. Richard Fawcett, even though he became vicar at St. Nicholas’ Church, Newcastle, in 1766, had no apparent connection with Avison before that time; he subscribed to everything from the Two Concertos to the op. 4. There was also Joseph Brackenbury, who was chaplain to Peregrine Bertie, the third Duke of Ancaster; Bertie’s younger brother, Brownlow, was yet another subscriber. 60  Charles Rogers, Genealogical Memoirs of the Family of Sir Walter Scott, Bart. of Abbotsford (London: Printed for the Grampian Club, 1877), xix. See also the subscription list to Francis Hutcheson, A System of Moral Philosophy in Three Books (Glasgow: Printed and Sold by R. and A. Foulis, 1755), i. 61  Fleming, “John Callander and the Avison Connection,” 287–88. 62   Jessica Kilburn, “Shafto, Robert (c.1732–1797),” Oxford Dictionary of National Biography, Oxford University Press, https://doi.org/10.1093/ref:odnb/75159

9  PUBLISHING MUSIC BY SUBSCRIPTION 

337

Table 9.4  Members of the aristocracy and gentry who subscribed to Avison’s publications Subscriber

List Ref.

Sir Edmund Anderson, Bart. Sir William Anderson, Bart. Sir Jacob Astley, Bart, of Melton-Constable in Norfolk The Right Honourable Lord Bellfield The Right Honourable Lady Bellfield The Honourable Lord Brownlow Bertie The Honourable Lady Rob. Bertie Lady Bewick Capt. Bisset The Right Honourable Lady Bingley Walter Blackett, Esq Lady [Elizabeth] Blackett The Right Honourable the Lord Blantyre The Honourable Colonel Blathwayt George Bowes, Esq. Mrs. [Mary] Bowes The Right Honourable the Earl of Buckinghamshire The Right Honourable Lord Buttivant The Right Honourable the Lord Byron The Right Honourable the Countess of Carlisle The Honourable Lady Mary Carre William Carre, Esq; F. R. S. Sir Thomas Clavering, Bart. His Grace the Duke of Cleveland Her Grace the Duchess of Cleveland Sir Bryan Cooke, Bart. The Right Honourable Earl Cowper The Right Honourable the Earl of Darlington The Right Honourable the Earl of Dartmouth The Right Honourable the Countess of Dartmouth George Shaftoe Delaval, Esq Sir Robert Eden, Bart. The Right Honourable the Earl of Effingham The Honourable Lady Sophia Egerton Captain Edward Fage, York The Honourable Mr. Fitzpatrick, London The Honourable Mr. Fitzwilliams, London Captain Robert Franklin of Lynn-Regis Captain Gilpin, Carlisle The Honourable Alexander Gordon

B F, G, H A, D, E A A G, H H F, G, H F, G, H F, G, H A, C, D, E, F (2), G (2), H (2) D, E C A A, C, D, E C, D, E, F, G, H, I F, G, H, I A C E, F, G, H E, I A, B D, E A, D, E, I A F, G, H D, E E, I (2) G, H G, H A, C A C E D F, G, H F, G, H A, B F, G, H G, H (continued)

338 

S. D. I. FLEMING

Table 9.4 (continued) Subscriber

List Ref.

The Honourable Mr. Baron Grant Captain Grant, Monymusk Captain Nathaniel Green of Preston The Right Honourable the Marchioness of Grey Sir Henry Grey, Bart. The Right Honourable the Marchioness of Hartington Sir Robert Hildyard, Bart. Lady Hilton The Right Honourable the Earl of Holderness The Right Honourable the Earl of Home Colonel Honeywood The Honourable Captain [Colonel] George Howard Sir Richard Hylton, Bart Lady Hylton Capt. Ralph Jenison Charles Jennens, Esq. Nicholas Lambton, Esq; of Biddick Sir Robert Lawly Bart. Sir Digby Legard, Bart. Captain Leigh, York Sir Henry Liddell Bart The Honourable Thomas Liddell, Esq. Lady Liddell The Right Honourable the Earl of Lichfield The Honourable Lord George Manners Sir Henry Manwaring Bart. Sir William Middleton, Bart. Captain Edward Milbanke Sir Ralph Milbanke, Bart. Lady Milbanke The Right Honourable Lord Viscount Molesworth The Honourable Miss Molesworth Lady Grace Montgomery Colln Noel Captain O Carrol, York [Anne] Ord The Honourable Miss Paulet Sir Lionall Pilkington Bart. The Right Honourable the Earl of Plymouth The Right Honourable Lord Ravensworth, The Right Honourable Lady Ravensworth

F (2), G (2), H (2) F, G, H A D D, E E F, G, H D E (2), I C D B, E E E A, C (2), D, E, I F, G, H E, I C D D C E C C E C, D, E A, C D A, C, D, E (3), F, G, H A, C, D, E (3), F, G, H, I A G, H D C D D(3), E, F, G, H, I F, G, H C G, H D (2), E (2) D, E (2) (continued)

9  PUBLISHING MUSIC BY SUBSCRIPTION 

339

Table 9.4 (continued) Subscriber

List Ref.

The Honourable Miss Anne Rochfort The Honourable Miss Tamazon Rochfort The Honourable William Rochfort The Right Honourable the Marchioness of Rockingham The Right Honourable Lord St. John, of Bletsoe Sir George Saville Bart The Right Honourable the Countess of Scarborough The Honourable Lady Jane Scott Mr. Walter Scott, of Harding, then Mertoun, Esq. The Honourable Miss Fitzroy Scudamore Robert Shaftoe, of Benwell, Esq. The Right Honourable the Earl of Strathmore The Right Honourable the Countess of Strathmore The Right Honourable Lady Grace Vane The Honourable Lady Harriot Vane The Honourable Henry Vane, Esq; The Honourable Charles Hope Weir The Right Honourable the Earl of Wemys The Honourable Mr. Widdrington The Right Honourable the Lord Widdrington The Right Honourable Lady Widdrington The Honourable Lady Frances Williams The Right Honourable the Lord Viscount Windsor The Honourable Mr. Withrington Sir Marmaduke Wyvil, Bart

A A A E, I D C, E, I E E B (2), C, D, E (3), F, G, H F, G, H D, E, F, G, H H H A E, I A, D E A C, D D (2), E, F, G, H, I D, E E A, E A, B F, G, H

dedicatee of the Scarlatti concertos while his daughter, Mary Eleanor, was the dedicatee of op. 8. Anne Ord of Fenham was the dedicatee of op. 3, of which she purchased three copies; the Ord family were important supporters of Avison’s and members of this family appear in all of Avison’s lists. Yet another subscriber was Colonel John Blathwayt, the dedicatee of Avison’s op. 2. A final important subscriber, although having no known connection with Avison, was Charles Jennens, also famous for his association with Handel. A final but important group of subscribers are the music publishers themselves, with a significant number subscribing to works that they had themselves printed (Table 9.5). A few publishers did, however, subscribe to works with which they had no direct association; some, such as John

340 

S. D. I. FLEMING

Table 9.5 Publishers who subscribed to Avison’s publications

Publisher

List Ref.

Joseph Barber [Robert] Bremner Benjamin Cooke [John] Johnson [Charles] Thompson [John] Walsh

A, C F (7), G (7), H (7), I (3) A, B B (12), I (12) F (14), G (14), H (14) B (6), D, E (7), F (14), G (14), H (14), I (6) F, G, H

[Peter or John] Welcker

Walsh and Robert Bremner, did so with op. 9, for which they also acted as agents by taking subscriptions.63 The most notable person in this group is Walsh, who subscribed to everything but op. 2 and the Scarlatti concertos. He was, nevertheless, clearly aware of op. 2 as he brought out a set of Eight Concertos for the Organ or Harpsichord, six of which were based on the concertino parts from this opera and designed to work with the already available ripieno parts.64 In addition, several publishers subscribed to a large number of copies that were presumably sold in their shops. John Walsh, for instance, purchased six copies of the Two Concertos, seven copies of op. 4, and 14 copies of op. 9. Subscribing would certainly have made financial sense if, by doing so, they were provided with a discount: they could then sell these copies in their shops at full price. In addition, the appearance of their names in a subscription list could potentially provide them with some free publicity. What is perhaps most interesting about Avison’s concerti grossi is the number of women who subscribed to them, especially given that, at the time, such works were viewed as unsuitable for women musicians (Fig. 9.3). In the eighteenth century, there was a reluctance not only to allow women to perform in public, which made large orchestral works such as these inappropriate, but also to play on stringed instruments, such

 East Riding Archives: DDGR/42/16/23.  Of the two remaining “Walsh” organ concertos, one was based on the second concerto from Avison’s Two Concertos and the other one on an apparently now lost published concerto. Revised versions of all eight concertos were included by Avison in his op. 6. 63 64

9  PUBLISHING MUSIC BY SUBSCRIPTION 

341

4.2% 15.9%

79.9%

Male

Female

Institutional

Fig. 9.3  The proportion of subscribers to all of Avison’s concerti grossi by gender

as the violin, viola, and cello.65 The prejudices of the time are evident from the writing of the dancing master, John Essex, who wrote that: The Harpsichord, Spinet, Lute and Base Violin, are Instruments most agreeable to the LADIES: There are some others that really are unbecoming the Fair Sex; as the Flute, Violin, and Hautboy; the last of which is too Manlike, and would look indecent in a Woman’s Mouth; and the Flute is very improper, as taking away too much of the Juices, which are otherwise more necessary employ’d, to promote the Appetite, and assist Digestion.66

In spite of such prejudices, some ladies are known to have played the violin in private, such as Lady Sophia Hope (d. 1813) and Elizabeth Rose of Kilravock (1708–1789).67 Also, given that keyboard instruments were 65  Jean-Jacques Rousseau clearly thought that a woman’s role lay within the home when he wrote that “any woman who shows herself off disgraces herself.” Qtd. in Joan B. Landes, Women and the Public Sphere in the Age of the French Revolution (Ithaca, NY: Cornell University Press, 1988), 72. 66  John Essex, The Young Ladies Conduct: Or, Rules for Education (London: Printed for J. Brotherton, 1722), 84–85. 67  Elizabeth Cary Ford, “The Flute in Musical Life in Eighteenth-Century Scotland” (PhD diss., University of Glasgow, 2016), 71; Helen Goodwill, “The Musical Involvement of the

342 

S. D. I. FLEMING

viewed as more suitable for a lady, women could have played the harpsichord part, although it is unlikely that they would have done so in public. Grisie, the eldest daughter of Lady Grisell of Mellerstain House, Berwickshire, is known to have been able to play a thorough bass.68 In addition, women account for 25.2% of the subscribers to George Jackson’s A Treatise on Practical Thorough Bass … Op. 5 (1791), which indicates that a significant number of women could play from a figured bass. Across all concerti grossi issued before 1820, the average percentage of women subscribers per list, as my earlier study has shown, comes to 16%.69 In this regard, Avison’s string concertos are quite typical, with his average coming to 15.9% (Fig. 9.3). Presumably, many of the unmarried ladies in his lists, including the aforementioned Mary Eleanor Bowes, were his students. When one looks at non-aristocratic female subscribers who gave a place of residence, it is also quite striking that many lived relatively close to Newcastle. Male subscribers could be located much farther afield, which implies that it was they who responded to published advertisements. By implication, this means that female subscribers to Avison’s concertos may not have performed the music themselves but purchased a copy in support of the composer. Avison’s overall percentage of 15.9% is a slight increase for his op. 2, where 8.2% of subscribers were female; this percentage had risen by the time of his op. 4, when 19.2% were of this gender. Given that the op. 9 includes a realized keyboard part, and that he was making these pieces more appropriate for the use of a female domestic performer, one might have expected Avison to see an increase in the proportion of this type of subscriber. Although he did receive, when compared with op. 4, an extra 55 subscribers, the list to Book 2 records that still only 19.2% of the total subscribers were women, indicating that the inclusion of a partially realized keyboard part did not make a significant impact. This may be why Avison opted to return to a more standard thorough bass part when he issued his op. 10 concertos in 1769. Although intended for orchestral performance, keyboard concertos, like concerti grossi, tended to be arranged with a self-sufficient keyboard Landed Classes in Eastern Scotland, 1685–1760” (PhD diss., University of Edinburgh, 2000), 45, 58. 68  David Johnson, Music and Society in Lowland Scotland in the Eighteenth Century (London: Oxford University Press, 1972), 26–28. 69  Fleming, “The Gender of Subscribers,” 114.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

343

part that enabled their performance as either keyboard solos or with one or more of the accompanying string parts. One would expect such works to attract more female subscribers and this is certainly true, with the average across the genre coming to 34.9%.70 However, Avison’s only published keyboard concerto, issued in the Two Concertos, did not come even close to this: only 11.8% of the subscribers to this work were female. Other composers did much better at appealing to female subscribers; for example, the proportion of female subscribers to the two sets of keyboard concertos by Chilcot is 30.1%.71 However, this figure should not be taken at face value as, even if Chilcot did attract more subscriptions from women, he was not able to attract anywhere near as many subscribers in total for his first set (101) as Avison did for his op. 9 (253). When one looks at the breakdown of subscribers by title or type (Fig. 9.4), some interesting observations can be made. Of the subscribers to op. 4, only 5.1% were institutional. This percentage is higher than the average across all lists issued before 1820, which stands at under 2%.72 Nonetheless, if Avison had been reliant on musical societies and other such groups, then publication by subscription would never have been a viable option. Individuals with aristocratic titles account for 18.2% of subscribers and vicars for 10.6%. However, the biggest proportion of subscribers are individuals with commonplace titles such as “Mr.,” “Mrs.,” or “Miss,” with the total of these three groups accounting for 66.2% of all subscribers. What this indicates is that publication by subscription relied heavily on the support of ordinary people. That is not to say that those in the other groups were not important, and Avison would certainly have valued their patronage, but this method of publication was only feasible through the rise of the new middle class. If we incorporate the gender of subscribers into these results, then, as expected, a total of 55.6% of subscribers gave a title of “Mr.,” the biggest group of all. This clearly indicates that untitled males were the main source of Avison’s subscriptions.

 Fleming, “The Gender of Subscribers,” 115.  I am grateful to Timothy Rishton, who provided me with copies of the subscription lists attached to Chilton’s published works. The character of Chilcot’s concertos is discussed in Timothy J. Rishton, “The Eighteenth-Century British Keyboard Concerto after Handel,” in Aspects of Keyboard Music: Essays in Honour of Susi Jeans, ed. Robert Judd (Oxford: Positif Press, 1992), 126–28. See, also, Rishton, “Chilcot, Thomas.” 72  Fleming, “The Gender of Subscribers,” 105. 70 71

344 

S. D. I. FLEMING

Female Miss 8.6%

Institutional 5.1%

Female Mrs 2.0%

Male Titled 9.6% Male Vicar 10.6%

Female Titled 8.6%

Male Mr 55.6%

Fig. 9.4  Types of subscribers to Avison’s op. 4 concertos (1755—List E)

Conclusion Any investigation that uses subscription lists to reveal trends and connections that existed in the eighteenth century among artists and patrons can only provide a partial picture of their interactions. Nevertheless, the study of the subscription lists attached to Avison’s concertos not only provides a significant amount of new information about the Georgian society but also reveals the range of connections an individual composer could form. Avison may have spent most of his life in the northern town of Newcastle, but this location did little to impair his growth as a composer: as the subscription lists to his works show, he had an increasing spread of subscribers across Britain and sold a growing number of copies by this method. Likewise, the subscription of numerous musical societies confirms the popularity of Avison’s concertos with professional musicians across the country. There is also evidence that Avison’s music was a popular choice with members of the Oxford Musical Society, a fact that somewhat tempers Hayes’ virulent attack on Avison’s music. Avison also attracted a large number of individual subscribers, of which his musician peers account for a significant proportion. Some of these connections are already known from other sources, but some, such as his links with musicians from Bath, are otherwise unrecorded. These subscription

9  PUBLISHING MUSIC BY SUBSCRIPTION 

345

lists also indicate that Avison was in contact with a sizable number of the clergy, some of whom contributed to his important Essay, and with several merchants and music publishers. He was also well supported by members of the aristocracy and gentry, although many of them did not have a particularly strong attachment to Avison himself. It was, however, the sale of his new works to ordinary people that made publication by subscription a success. Even the copies purchased by Britain’s musical societies account for a very low proportion of those sold. Subscribers had a range of reasons for subscribing to Avison’s music. For some, it was clearly a reciprocal process. Others had a personal connection with the composer as they lived in the North-East or were members of a local or a national organization, such as the freemasons. Some may have had no connection with Avison at all, with their subscription being elicited through a printed advertisement. Some would certainly have subscribed only to have their names appear in print. Finally, Avison also had a core of supporters who were willing to subscribe to most, if not all, of his publications due to their relationship with him; some were fellow musicians, while others might have been patrons of Avison’s concerts or his pupils. As this study reveals, subscription lists are important historical documents that can provide a significant amount of information about the Georgian society. Regarding Avison, they underscore the importance of his extensive networking through subscription as he grew into one of Britain’s leading composers of the mid-eighteenth century. Data visualization tools help highlight the extent of his influence, the broad dissemination of his work, and period-specific purchasing patterns, shedding new light on the role played by gender and social status in the process of building prestige. By including other types of documents, such as letters, diaries, and concert details, this type of analysis can provide a more detailed picture of the musical life in the Georgian era than ever before.

Bibliography Abel, Charles Frederick. 1760. Six Sonatas for the Harpsichord … Opera II. London: Printed for the Author. Avison, Charles. 1737. VI Sonatas. London: Printed by Benjamin Cooke, c. ———. 1753. An Essay on Musical Expression. London: C. Davis. ———. 1756. Six Sonatas for the Harpsichord … Opera Quinta. London: Printed for John Johnson.

346 

S. D. I. FLEMING

———. 1757. Six Sonatas … Opera Prima. London: Printed for John Johnson, c. ———. 1758. Twenty Six Concertos … in Score. 4 vols. Newcastle: Printed for the Author. ———. 1760. Six Sonatas for the Harpsichord … Opera Settima. London: Printed for the Author. ———. 1764. Six Sonatas for the Harpsichord … Opera Ottava. London: Printed for the Author. ———. 1769. Six Concertos … Opera Decima. London: Printed for Robert Bremner. ———. 1775. An Essay on Musical Expression. London: Lockyer Davis. Beechey, Gwilym. 1973. Thomas Chilcot and His Music. Music & Letters 54 (2): 179–196. https://www.jstor.org/stable/734363. Essex, John. 1722. The Young Ladies Conduct: Or, Rules for Education. London: Printed for J. Brotherton. Farmer, Henry George. 1950. Music Making in the Olden Days: The Story of the Aberdeen Concerts, 1748–1801. London: Peters-Hinrichsen. Flackton, William. 1770. Six Solos, Op. 2. London: Printed for the Author. Flanders, Julia, and Fotis Jannidis. 2017. Data Modeling in a Digital Humanities Context: An Introduction. In The Shape of Data in Digital Humanities. Modeling Texts and Text-Based Resources, ed. Julia Flanders and Fotis Jannidis, 3–25. Abingdon: Routledge. Fleming, Simon D.  I. 1999. Charles Avison (1709–1770): An Important and Influential English Composer, Musician, and Writer. MMus diss., University of Liverpool. ———. 2009. A Century of Music Production in Durham City 1711–1811: A Documentary Study. PhD diss., University of Durham. ———. 2012a. Charles Avison Jnr and His Book of Organ Voluntaries. The Musical Times 153 (1918): 97–106. http://www.jstor.org/ stable/41442085. ———. 2012b. John Garth and His Music: An Important Provincial Composer from 18th-Century Britain. The Musical Times 153 (1921): 63–93. https:// www.jstor.org/stable/41703543. ———. 2013. The Howgill Family: A Dynasty of Musicians from Georgian Whitehaven. Nineteenth-Century Music Review 10 (1): 57–100. https://doi. org/10.1017/S1479409813000049. ———. 2014. John Callander and the Avison Connection: A Recently Rediscovered Letter. Eighteenth Century Music 11: 283–290. https://doi.org/10.1017/ S1478570614000086. ———. 2017a. Music and Concert Promotion in Georgian Stamford. The Consort 73: 61–83. ———. 2017b. The Musical Activities of the Spalding Gentlemen’s Society. Royal Musical Association Research Chronicle 48 (1): 65–90. https://doi.org/10.108 0/14723808.2016.1271572.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

347

———. 2018. Avison and His Subscribers: Musical Networking in Eighteenth-­ Century Britain. Royal Musical Association Research Chronicle 49 (1): 21–49. https://doi.org/10.1080/14723808.2017.1363210. ———. 2019. The Gender of Subscribers to Eighteenth-Century Music Publications. Royal Musical Association Research Chronicle 50 (1): 94–152. https://doi.org/10.1080/14723808.2019.1570752. Ford, Elizabeth Cary. 2016. The Flute in Musical Life in Eighteenth-Century Scotland. PhD diss., University of Glasgow. Garside, P.D. 1987. Jane Austen and Subscription Fiction. Journal of Eighteenth-­ Century Studies 10 (2): 175–188. https://doi.org/10.1111/j.1754-0208. 1987.tb00014.x. Goodwill, Helen. 2000. The Musical Involvement of the Landed Classes in Eastern Scotland, 1685–1760. PhD diss., University of Edinburgh. Hayes, William. 1753. Remarks on Mr. Avison’s Essay on Musical Expression. London: Printed for J. Robinson. Hillman, Martin. 2017. Thomas Sanderson’s Account of Incidents: The Edinburgh Musical Society 1727–1801 and Its Impact on the City. Edinburgh: Friends of St. Cecilia’s Hall. Holman, Peter, and Richard Maunder. 2000. The Accompaniment of Concertos in 18th-Century England. Early Music 28 (4): 636–650. http://www.jstor. org/stable/3519000. Hughes, Edward, ed. 1956. Letters of Spencer Cowper, Dean of Durham, 1746–74. Durham: Andrews. Hunter, David, and Rose Mason. 1999. Supporting Handel Through Subscription to Publications: The Lists of Rodelinda and Faramondo Compared. Notes 56 (1): 27–93. https://doi.org/10.2307/900471. Hutcheson, Francis. 1755. A System of Moral Philosophy in Three Books. Glasgow: Printed and Sold by R. and A. Foulis. Hyman, Robert, and Nicola Hyman. 2011. The Pump Room Orchestra Bath: Three Centuries of Music and Social History. Salisbury: Hobnob Press. Jackson, William, ed. 1879. Memoirs of Dr. William Gilpin. London: Bernard Quaritch. Johnson, David. 1972. Music and Society in Lowland Scotland in the Eighteenth Century. London: Oxford University Press. Jones, Peter Ward, and Simon Heighes. 2017. Hayes Family. Grove Music Online. Oxford Music Online. Oxford University Press. http://www.oxfordmusiconline.com/subscriber/article/grove/music/12621pg1 Kassler, Michael. 2016. The Bachists of 1810: Subscribers to the Wesley/Horn Edition of the ’48. In The English Bach Awakening, ed. Michael Kassler, 315–340. London: Routledge.

348 

S. D. I. FLEMING

Kilburn, Jessica. n.d. Shafto, Robert (c.1732–1797). Oxford Dictionary of National Biography. Oxford University Press. https://doi.org/10.1093/ ref:odnb/75159. Knape, Walter, Murray R. Charters, and Simon McVeigh. n.d. Abel Family. Grove Music Online. Oxford Music Online. Oxford University Press. https://doi. org/10.1093/gmo/9781561592630.article.00035. Landes, Joan B. 1988. Women and the Public Sphere in the Age of the French Revolution. Ithaca: Cornell University Press. Mee, John. 1911. The Oldest Music Room in Europe: A Record of Eighteenth-­ Century Enterprise at Oxford. London: John Lane Company. Rishton, Tim. n.d. Chilcot, Thomas. Grove Music Online. Oxford Music Online. Oxford University Press. http://www.oxfordmusiconline.com/subscriber/ article/grove/music/05581 Rishton, Timothy J. 1992. The Eighteenth-Century British Keyboard Concerto after Handel. In Aspects of Keyboard Music: Essays in Honour of Susi Jeans, ed. Robert Judd, 126–128. Oxford: Positif Press. Rogers, Charles. 1877. Genealogical Memoirs of the Family of Sir Walter Scott, Bart. of Abbotsford. London: Printed for the Grampian Club. Sadie, Stanley. 1990. Music in the Home II. In The Blackwell History of Music in Britain: The Eighteenth Century, ed. H.  Diack Johnstone and Roger Fiske, 313–354. Oxford: Basil Blackwell. Seares, Margaret. 2011. The Composer and the Subscriber: A Case Study from the 18th Century. Early Music 39 (1): 65–78. https://doi.org/10.1093/ em/caq111. Shaw, David. 2000. Canterbury’s External Links: Book-Trade Relations at the Regional and National Level in the Eighteenth Century. In The Mighty Engine: The Printing Press and Its Impact, ed. Peter Isaac and Barry McKay, 107–119. Winchester: St Paul’s Bibliographies. Smith, G.H. n.d. A History of Hull Organs and Organists. London: A. Brown and Sons. Smith, William, and Peter Ward Jones. n.d. Cooke, Benjamin (i). Grove Music Online, Oxford Music Online. Oxford University Press, http://www.oxfordmusiconline.com/subscriber/article/grove/music/06393 Southey, Roz. 2006. Music-Making in North-East England during the Eighteenth Century. Aldershot: Ashgate. Southey, Roz, and Eric Cross. 2018. Introduction. In Charles Avison in Context: National and International Musical Links in Eighteenth-Century North-East England, ed. Roz Southey and Eric Cross, 1–34. Abingdon: Routledge. Southey, Roz, Margaret Maddison, and David Hughes. 2009. The Ingenious Mr. Avison: Making Music and Money in Eighteenth Century Newcastle. Newcastle upon Tyne: Tyne Bridge Publishing.

9  PUBLISHING MUSIC BY SUBSCRIPTION 

349

Spring, Matthew. 2013. Benjamin Milgrove, the Musical ‘Toy Man’ and the ‘Guittar’ in Bath 1757–1790. Early Music 41 (2): 317–329. https://doi. org/10.1093/em/cat044. Stephens, Norris. 1968. Charles Avison: An Eighteenth-Century English Composer, Musician and Writer. PhD diss., University of Pittsburgh. Talbot, Michael. 2014. What Lists of Subscribers Can Tell Us: The Cases of Giacob Basevi Cervetto’s Opp. 1 and 2. De Musica Disserenda 10 (1): 121–139. https://doi.org/10.3986/dmd10.1.08. The Works of William Mason, M.A. 1811. 4 vols. London: Printed for M. Cadell and W. Davies. Wallis, Peter, and Francis Robinson. 1977. Book Subscription Lists: A Revised Guide. Newcastle upon Tyne: H. Hill. Wallis, Peter, and Ruth Wallis. 1996. Book Subscription Lists: Extended Supplement to the Revised Guide. Phibb: Newcastle upon Tyne. Wallis, Ruth, and Peter Wallis. 1980. Female Philomaths. Historia Mathematica 7 (1): 57–64. https://doi.org/10.1016/0315-0860(80)90064-6. Wollenberg, Susan. 2001. Music at Oxford in the Eighteenth and Nineteenth Centuries. Oxford: Oxford University Press. Wrigley, E. Anthony. 2013. Urban Growth and Agricultural Change: England and the Continent in the Early Modern Period. In The Eighteenth-Century Town: A Reader in Urban History 1688–1820, ed. Peter Borsay, 39–82. London: Routledge.

CHAPTER 10

Afterword: Novel Knowledge, or Cleansing Dirty Data: Toward Open-Source Histories of the Novel Emily C. Friedman

Introduction In this chapter I discuss the most important, most under-rewarded, and most unsexy aspect of data visualization: the production and/or usage of reliable underlying data. Indeed, visualizations are only as good as their underlying evidentiary base. As Lauren Klein noted at the 2018 meeting of the Modern Language Association, “[w]e need to assemble more corpora—more accessible corpora—that perform the work of recovery or resistance.”1 This goes for metadata as well—something that, in theory, we do not lack for in the eighteenth-century novel: massive multigenerational bibliographies of the novel, for example. And more data is coming: for example, The Cambridge Guide to the English Novel, 1  Lauren Klein, “Distant Reading after Moretti,” Arcade (blog), 2019, https://arcade. stanford.edu/blogs/distant-reading-after-moretti

E. C. Friedman (*) Auburn University, Auburn, AL, USA e-mail: [email protected] © The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8_10

351

352 

E. C. FRIEDMAN

1660–1820 will provide synopses of every surviving English novel, produced by expert readers from within the field. But as Laura Mandell has noted, while we have a lot of data, what it represents is still—and will always be—partial: the works of women writers are disproportionately lost compared to those by their male counterparts, to give just one glaring example. And this makes for potentially disastrous effects when creating visualizations from that partial data. Where the sources for those visualizations are clear (and they are not always), they were already obsolete at the time of their construction. But the main thrust of this chapter is to imagine universal standards for this work, which I argue must be at the center of any future reliable visualizations about novel history. I propose guidelines for best practices in creating new data so that amendable, transformable visualizations can be produced, built on collective knowledge. I note the contributions of digital projects which have laid the foundation for such practices, including (though not limited to) massive multi-institution projects like Orlando2 and small to mid-sized projects like The Early Novels Database (END).3 My own small-scale project, Manuscript Fiction in the Age of Print, 1750–1900, creates meaningful metadata about unprinted manuscript fiction during the period, creating a parallel corpus to those of published fiction. Because I work with never-printed fiction, there are unique challenges in identification, classification, and dissemination that are beyond the scope of this chapter, but that I have written about elsewhere.4 As an active practitioner in the field of eighteenth-century fiction studies, I am aware of the very real obstacles to full implementation of any potential standards, even without the challenges relating to unprinted texts.

2  Orlando: Women’s Writing in the British Isles from the Beginnings to the Present, Cambridge University Press, accessed April 15, 2019, http://orlando.cambridge.org/ 3  The Early Novels Database, accessed April 15, 2019, https://earlynovels.github.io 4  See Emily C. Friedman, “Amateur Manuscript Fiction in the Archive: An Introduction,” in After Print: Eighteenth-Century Manuscript Cultures, ed. Rachael Scarborough King (Charlottesville: University of Virginia Press, 2020), 217–36; “Must Anonymous Be a Woman? Gender and Anonymity in the Archives,” Tulsa Studies in Women’s Literature, Special Issue on “Women in Archives” (forthcoming 2021); “Ownership, Copyright, Ethics of the Unpublished,” in Access and Control in the Digital Humanities, ed. Richard Mann and Shane Hawkins (forthcoming 2021); and “Eluding Print: Manuscript Fiction and the Survival of Scribal Practices in the Age of Print,” Special Issue of Huntington Library Quarterly by the Women in Book History Research Group (forthcoming 2021).

10  AFTERWORD: NOVEL KNOWLEDGE 

353

I am completing this chapter in the shadow of Nan Z. Da’s recent critique of computational literary studies,5 which appeared first in Critical Inquiry and then in a more mainstream variation in The Chronicle of Higher Education.6 The former publication almost immediately hosted an entire preplanned suite of responses within a week of the essay’s publication. There is much to say about the circumstances and the circulation of Da’s argument, much of which lies outside the scope of this chapter. And Da’s essay is, in a sense, just one more in a very long line of critiques of the emerging discipline of digital humanities, both by those who wish to create best practices and by those who wish to discredit it utterly. Excavating all the layers of power, risk, and citational strategy present in very public fora is beyond my scope here. Nevertheless, what Da and I—and all of us—share is an investment in the verifiability of claims that are numerically based—either via counting or communicated via visualization. We will be the better for publicly stated, and preferably publicly verified, forms of accountability.

Dreaming of Ideal Data Before I begin, I would like to briefly imagine the perfect conditions for the study of the eighteenth-century novel at scale. In that alternate universe, a copy of every work of fiction—at least!—was preserved from the flames and the privy. While we are dreaming, go one better: a copy of every edition of every work was saved. Or, if we would like to be profligate with our wishes, every copy. In this dream world, the Stationer’s Company requested and required the author’s name, even if the work was pseudonymously or anonymously published. Expand the vision further: their address was also required, and maybe some demographic data (age, gender identity, occupation). Moreover (to ride a personal hobbyhorse), every manuscript of fiction, published or unpublished, was somehow preserved according to the same principles. Perhaps that goes too far—we do not want to intrude on the actual material conditions of production too far, and manuscripts were 5  Nan Z. Da, “The Computational Case against Computational Literary Studies,” Critical Inquiry 45, no. 3 (2019): 601–39, https://doi.org/10.1086/702594 6  Nan Z.  Da, “The Digital Humanities Debacle: Computational Methods Repeatedly Come Up Short,” The Chronicle of Higher Education, March 27, 2019, https://www.chronicle.com/article/The-Digital-Humanities-Debacle/245986

354 

E. C. FRIEDMAN

often simply a casualty of the ways books became printed books. But, at least, we can dream that no manuscript was ever disconnected from its author or authors, nor crossed borders without documentation. We can dream that all of this information has been consolidated in one place: a single complete bibliography and checklist of fiction, produced through unprecedented collaboration, available not only in the durability of print, but in an open access relational database, allowing users to search by all of these different known fields of information, creating book lists researchers could use to dive into a given author, publisher, neighborhood, profession, or more. In this world of unlimited—enthusiasm? money? time? investment?—, optical character recognition of both eighteenth-century type and handwriting is perfect, and at least minimally encoded digital surrogates of each work are available. Not only the bibliography but also the digital corpus itself would then be truly comprehensive and reflective of the fiction produced. Dream still bigger and imagine every circulating library, private library, bookseller’s records or catalogs also survived, preferably with the kind of detailed demographic information of, for example, the New York Society Library’s records, now digitized in the City Readers database.7 Imagine every mention of any literary work from correspondence, reviews, and the like somehow gathered together, a macro version of Cardiff University’s British Fiction, 1800–1829: A Database of Production, Circulation, and Reception.8 Luxuriate in what we might be able to know, to glean. What we might be able to ask such a dataset, how we might represent different aspects of it.

Waking Up And then, reluctantly, wake up. Face what we actually have: a fraction of what we know to have been produced. We are haunted by a Great Forgetting (to put a new spin on Clifford Siskin’s phrase) that we cannot possibly recover from. We know 7  City Readers. Digital Historic Collections of the New York Society Library, The New York Society Library, accessed April 15, 2019, http://cityreaders.nysoclib.org/ 8  British Fiction1800–1829. A Database of Production, Circulation & Reception, Cardiff University Centre for Editorial and Intertextual Research, accessed April 15, 2019, http:// www.british-fiction.cf.ac.uk/

10  AFTERWORD: NOVEL KNOWLEDGE 

355

this, for example, from periodical advertisements for books that we have yet to recover copies of. For what we do have, we face incomplete records, innumerable challenges of attribution, and varying levels of reliable transcriptions and scans. Enormous amounts of digitized microform surrogates are still only partially representative of what survives, let alone what once existed. And that’s what is available to those with institutional access to proprietary databases. It is worth noting, for the record, that these are all known losses from print fiction; the losses from manuscript culture during the age of print are still more profound because they are largely unknowable. Unlike the print marketplace, which was largely controlled by the Stationer’s Company, works of manuscript fiction in the eighteenth-century and thereafter were not necessarily known, named, and documented for commercial exchange. The nature of their survival varies wildly from that of the print work. For that reason, we do not yet, nor may we ever know how much manuscript fiction survives in inaccessible or nearly inaccessible private collections. We certainly will never have a full sense of how much literary work was produced in manuscript and subsequently intentionally destroyed or accidentally lost. Critical consensus for a substantial amount of time was that these were not significant losses: they were unimportant precisely because manuscripts with no connection to print culture have no value. While attempting to create a database and digital corpus of manuscript fiction, I have seen the difficulties associated with such critical dismissal. Such dismissal, and by extension poor reckoning, also occurred to fiction that appeared in so-called “ephemeral” forms: in periodicals, pamphlets, cheap editions, and so on. From this perspective, one’s work is tinged by the regret that no matter how long or how hard we labor, we will still be faced with what has been obliterated. To dream of the perfect dataset is to dream away the very real working conditions we hope to study: the fact that a manuscript of a printed work is nearly always destroyed in the process of production, the fact that selection and disposal themselves are part of history. I am reminded of Aden Evan’s description of the digital as “calculably imprecise,” by which he means the digital is capable of weighing and measuring “to a given level of accuracy and no more”—that “no more” being the very fuzziness of objects and individuals whose borders are never entirely

356 

E. C. FRIEDMAN

demarcated.9 We can enumerate and visualize and analyze more texts than ever before, but unless the end results show us something in soft-focus, slightly blurred, we have missed a key aspect of our knowledge, that is, what we cannot know. Even with all such caveats, we have access to spectacularly more information, that is more rapidly retrievable, than any prior scholarly generation ever had to hand. It is important to recall that we are luckier than some other fields: we do not face the challenges attendant with, for example, reassembling page by page (often scattered across the globe) the surviving Coptic corpora described by Schroder and Zeldes.10 Moreover, in addition to what has been digitized, we have enormous datasets of surviving eighteenth-century prose fiction produced in the twentieth century. We tend to call these datasets “bibliographies”—the monumental works of Beasley, Forster, Garside, Letellier, McBurney, Raven, Schöwerling— understandably, perhaps, given that they officially exist exclusively in print editions.11 These works of enormous intellectual labor are preserved and stabilized by their instantiation in print, but to use them as something other than a reference requires laborious transformation, which due to copyright regulations cannot be then shared publicly. I often wonder how many times researchers (or their student assistants) have privately done this kind of transformative work that allows for dynamic analysis and visualization of these bibliographies, and how many sub rosa exchanges of that information occur. In an ideal world, these bibliographies would be undergoing the same kind of digitization project as that currently led by Mattie Burkert to transform The London Stage from print to a truly usable database.12 But, as Burkert has noted of The London Stage and Katherine Bode

9  Aden Evans, Sound Ideas: Music, Machines, and Experience (Minneapolis: University of Minnesota Press, 2005). 10  Caroline T.  Schroeder and Amir Zeldes, “Raiders of the Lost Corpus,” Digital Humanities Quarterly 1, no. 2 (2016), http://www.digitalhumanities.org/dhq/ vol/10/2/000247/000247.html 11  The University of Michigan has digitized The English Novel 1770–1829: A Bibliographical Survey of Prose Fiction Published in the British Isles (Oxford: Oxford University Press, 2000) and placed it on HathiTrust, allowing for search-only access by non-UM users. 12  The project is currently in beta at https://londonstagedatabase.usu.edu/. Burkert wrote about the project’s origins in the 1960s London Stage Information Bank in “Recovering the London Stage Information Bank: Lessons from an Early Humanities Computing Project,” Digital Humanities Quarterly 11, no. 3 (2017), http://www.digitalhumanities.org/dhq/vol/11/3/000321/000321.html

10  AFTERWORD: NOVEL KNOWLEDGE 

357

has reminded us about the fiction bibliographies of the last half-century, these datasets are themselves still partial and decontextualized.13 From one perspective, eighteenth-century scholars of the novel are swimming in data: massive periodical runs survived in some form or another full of reviews and advertisements about new fiction, more novels survived than can possibly be read by one human, and the Stationer’s Company trade records provided enormous amounts of information about the marketplace for commercial fiction in the period. Another essay could easily be written about the challenges attendant in wrangling the “big data” aspects of eighteenth-century studies. Certainly, the challenges attendant with our search-driven research milieu have been discussed before. As Ted Underwood has noted, most literary scholars have transitioned into practices of searching, often using tools with proprietary algorithms that are opaque even to those with the technical expertise to understand them. The implicit “understanding” of what search is and how it operates is now so established that there are tenured faculty whose careers have taken place entirely within its mindset.14 Our scholarly habits have been transformed by a practice that seems so simple that the vast majority of practitioners, even very sophisticated ones, have not built up a theoretical lens through which to understand what is “found” and what is overlooked in such practices—shockingly considering this is a now decades-old research practice. But as Underwood’s work with classification algorithms shows, there are many places where algorithmic search and the substantial capacities of the analysis break down: while they are good at making broad category distinctions between “fiction, drama, poetry, and nonfiction prose,” they are poor at detecting parody, fine distinctions between subgenres, or genres that exist across a long period of time (and thus across large shifts in generic expectations).15 Thus, researchers of eighteenth-century fiction face challenges of simultaneously not

13  Katherine Bode, A World of Fiction: Digital Collections and the Future of Literary History (Ann Arbor: University of Michigan Press, 2018), 20–21. 14  Ted Underwood, “Theorizing Research Practices We Forgot to Theorize Twenty Years Ago,” Representations 127, no. 1 (2014): 64–72, https://doi.org/10.1525/ rep.2014.127.1.64 15  Ted Underwood, Michael L.  Black, Loretta Auvil, and Boris Capitanu, “Mapping Mutable Genres in Structurally Complex Volumes,” Proceedings of the 2013 IEEE International Conference on Big Data, ed. Xiaohua Hu et al., 95–103. Silicon Valley, CA: IEEE, 2013. https://doi.org/10.1109/BigData.2013.6691676

358 

E. C. FRIEDMAN

enough and too much, which are also the challenges attendant with fields from particle physics to genomics and beyond. That said, to throw up our hands, turn our backs on large-scale collaborative projects, and cultivate our small, impressionistic gardens seems to be falling off the other side of the horse. If computational work and the visualizations used to communicate that information to larger audiences do not save us, they probably will not damn us either. My call here is twofold: to afflict the comfortable and comfort the afflicted, as it were. We should absolutely challenge and critique work that makes too sweeping claims based on macro-analysis from already available data and text sets and call, instead, for more work in making such sets more robust, and representing their outputs more transparently and with more attention to the unknowable. Simultaneously, the perfect cannot be the enemy of the good: we should recognize and reward more robustly digital editorial work that refines our existing knowledge and allows for future iteration and improvement.

Case Studies: Where We Are Now In this section, I want to survey the landscape of available data and related public-facing projects that tackle enumerating, describing, analyzing, and/or visualizing works of eighteenth-century fiction. It is worth noting at this point that the vast majority of such projects are in article or monograph form, and either do not make their data or work materials publicly available, or work from curated samples of one of two major data and text sets, which I will describe below. Leah Orr’s groundbreaking Novel Ventures: Fiction and Print Culture in England, 1690–1730 is an excellent example of the former.16 Her work considers 475 works of fiction printed between 1690 and 1730 in order to assess more precisely the fiction market of the period. Many of the texts Orr considers had not been previously included in any of the standard bibliographies; she lists those titles within an appendix in her monograph. Her monograph also includes tables and charts describing trends in genre over the period. But it would be difficult to test her arguments because there is no space to reproduce her entire list of works, alongside how she coded them by genre. For those (like me) who would love to see other scholars adopt Orr’s devotion to precision 16  Leah Orr, Novel Ventures: Fiction and Print Culture in England, 1690–1730 (Charlottesville: University of Virginia Press, 2017).

10  AFTERWORD: NOVEL KNOWLEDGE 

359

and careful assessment decade by decade throughout the eighteenth century, this is a disappointment. But this is not just Orr’s challenge: it is a challenge that all of us face when we work (as Orr did) with bibliographies and other very large datasets. Almost all such projects were produced at least in part within for-­ profit publishing structures. I have already touched on the nature of bibliographies in this respect, which are all printed books of recent production, published by major presses, unlikely to take on an open access transformation of the data within those volumes, and even less likely to do so together. Could Orr have released her spreadsheets? It seems unlikely. Nor is this simply the challenge attendant when one’s research is extending existing print resources. Proprietary databases have significant affordances and limitations of their own. Many research projects, including the ones that would not identify themselves as “digital,” are built on data that emerges from Gale Cengage’s Eighteenth-Century Collections Online (ECCO), a subscription database that claims to include “every significant English-language and foreign-language title printed in the United Kingdom during the 18th century, along with thousands of important works from the Americas.”17 Even for scholars who will never attempt to visualize large amounts of data, ECCO has become an essential component of academic work in eighteenth-century studies, at least for those who work or study at institutions that can afford its cost. ECCO’s own platform allows for full-text searching of the corpus, but because the full text is derived from optical character recognition (OCR) of digitized black and white copies of microform surrogates, there are limitations to the capabilities of search due to errors.18 In order to improve the full-text searchability of their database, Gale Cengage partnered with the Text Creation Partnership (TCP),19 which also partners with ProQuest’s Early English Books Online (EEBO) and Readex’s Evans Early American Imprints. In exchange for human labor providing human-transcribed texts, these publishers allow the TCP to 17  This claim appears on the Text Creation Partnership’s page describing “ECCO-TCP: Eighteenth-Century Collections Online,” 2019, https://www.textcreationpartnership.org/ tcp-ecco/ 18  This has been ably discussed by Patrick Spedding in “‘The New Machine’: Discovering the Limits of ECCO,” Eighteenth-Century Studies 44, no. 4 (2011): 437–453, https:// www.jstor.org/stable/41301590 19  “EEBO-TCP Phase I Public Release: What to Expect on January 1,” TCP, December 24, 2014, https://www.textcreationpartnership.org/

360 

E. C. FRIEDMAN

make the data and metadata of these texts publicly available in XML-­ encoded electronic editions and plain text. In a related endeavor, Typewright, a tool hosted by the aggregation site 18thConnect.org, allows users to correct the OCR-generated transcriptions in ECCO. If any users contribute to the correction of a given text, they will be given access to the completed transcription and XML file for their own use, even if they do not have institutional access to ECCO. Gale Cengage can afford this generosity because ECCO-TCP is unlikely to put ECCO out of business any time soon: at last publicly reported count, the ECCO-TCP corpus included 2231 texts—roughly 1.5% of the over 136,000 titles in ECCO. Thus, while there is a pathway to facilitate access to a human-scaled corpus of texts (say, all editions in ECCO of the work of Samuel Richardson), ECCO-TCP is still comparatively tiny. Depending on the kind of questions you want to ask—for example, about the frequency of a word in print usage across the century—you would find ECCO-TCP both too small and too unrepresentative of the whole of ECCO, to say nothing of print culture, to be very certain of your findings. More texts, transcribed and ideally encoded, are needed for such work. This is the goal of the HathiTrust (HT), a collective project to create a “shared repository of cultural heritage materials” by combining efforts of various library and collection digitization projects. HT facilitates what it calls “non-consumptive research,” including data analysis and visualizations, from its corpus of millions of titles, of both in-copyright and out of copyright works. “Non-consumptive” means that the researcher is conducting computational analysis without reading or displaying a “substantial portion” of a volume if it is in copyright.20 In other words, HathiTrust Research Center (HTRC) walks a very careful line that avoids any suggestion of copyright infringement by restricting the kinds of data exports that can be performed by a user. Anything that could recreate an in-copyright text would be verboten, understandably, while summaries, token counts, topic models, and the like can be exported. Datasets and text sets have been created for users, but more importantly users can create work sets that can be shared and cited. Compared to ECCO-TCP, HTRC’s corpus is both smaller (only 505 volumes from ECCO directly) and far larger: over 34,000 books in English produced between 1700 and 1800, roughly 26,182 that describe 20   “Non-Consumptive Use Research Policy,” HathiTrust Research Center, accessed February 20, 2019, https://www.hathitrust.org/htrc_ncup

10  AFTERWORD: NOVEL KNOWLEDGE 

361

themselves as either prose, fiction, novel, tale, or story. This is an imperfect reckoning of what is available in a few ways: first, not all items are available full-text for all users, and second, the only metadata connected to genre for the user interface is based on subject headings.21 HTRC is a model of transparency, keenly aware of the gaps in its underlying scope. As the documentation for one project, Word Frequencies in English-Language Literature, 1700–1922, notes, “[c]ontributing institutions are mainly located in the United States. So, while the collection contains volumes from around the globe, coverage of works published in the U.S. is more complete. Also, because books before 1800 may be held closely in Special Collections, digitization of that period is less predictable. We don’t necessarily recommend this dataset as a source for literary research before 1750.”22 Thus, projects using HTRC’s metadata or text sets must be exceptionally careful in their construction of work sets, or else their conclusions (here, visualizations) will reflect more the holdings of contributing institutions than anything else. Both the HTRC and ECCO-TCP emerge from massive ongoing digitization and transcription efforts. They allow for their material to be cited and used at enormous scale. But while both have text and data sets that can be used, they are not extensible by a given user. Researchers can carefully select which items go into their corpus as they prepare to do visualization work within the HTRC ecosystem, but should they transcribe and encode their own additional items, ingestion (as far as I understand it) would not be possible. And, as I will argue in the final section of this chapter, extensibility is absolutely essential to the robustness of our research moving forward. Mark Algee-Hewitt et al. remind us to distinguish between “the published, the archive, and the corpus,” that is, “the totality of the books that have been published,” “that portion of published literature that has been preserved,” and “that portion … that is selected, for one reason or another,

21  Word Frequencies in English-Language Literature, 1700–1922 is specifically not a text, but a dataset concerned with providing word frequencies at the volume and page level for fiction, drama, and poetry. For their documentation, see Ted Underwood, Boris Capitanu, Peter Organisciak, Sayan Bhattacharyya, Loretta Auvil, Colleen Fallaw, and J.  Stephen Downie, Word Frequencies in English-Language Literature, 1700–1922, [Dataset], HathiTrust, 2015, https://doi.org/10.13012/J8JW8BSJ 22  Ibid.

362 

E. C. FRIEDMAN

in order to pursue a specific research project.”23 While the authors imagine a convergence of all three that “may soon be a reality”—a theoretical imaginary that they themselves walk back as they move further into their argument, they do not indicate explicitly that their sampling procedure is “an ideal model of research” but note that “dirty hands are better than empty.”24 Dirty data is better than no data, but it is potentially misleading in the same way that a dirty window not only obscures what is present, but might suggest shadows of things not present at all. Given the right partnership, and a different intellectual framework, it is possible to create a meaningful corpus without impossible comprehensiveness. In her analysis of the Stanford Lab model, Katherine Bode notes that without a robust knowledge of the “bibliographic and editorial practices” of the historical moment in which texts were produced, there is an “inadequate foundation” upon which to build one’s argument.25 In response, her latest book, A World of Fiction: Digital Collections and the Future of Literary History, walks through her work focusing on a subset of the 16,500 works of fiction identified in The National Library of Australia’s TROVE database. This work is meant to create what she calls a “scholarly edition of a literary system”26 that emphasizes the constructedness of any corpus, paralleling the traditional scholarly edition, which is an “argument” present through a “curated text” that is “designed to enable and advance rather than to decide or conclude—investigation.”27 To that end, her dataset, Reading by Numbers, is openly available on several locations on the web, including TROVE’s site, so that her choices can be fully understood and engaged with.28 As a sampling method, it has the virtue of being transparent and thus correctable/iterative over time. Creating cleaner, open-source, extensible data is also more achievable when one turns to projects that do not digitize, transcribe, or encode, but focus on the production of metadata. The Early Novels Database (END),29 23  Mark Algee-Hewitt, Sarah Allison, Marissa Gemma, Ryan Heuser, Franco Moretti, and Hannah Walser, “Canon/Archive. Large-Scale Dynamics in the Literary Field,” Literary Lab Pamphlet 11 (January 2016): 1–13, https://litlab.stanford.edu/LiteraryLabPamphlet11.pdf 24  Ibid. 25  Bode, A World of Fiction, 5. 26  Ibid., 4, and further described on 6. 27  Ibid., 6–8. 28  Reading by Numbers, AusLit, accessed April 15, 2019, https://www.austlit.edu.au/specialistDatasets/ResourcefulReading/ReadingByNumbers 29  The Early Novels Database, accessed April 15, 2019, https://earlynovels.github.io/

10  AFTERWORD: NOVEL KNOWLEDGE 

363

led by Rachel Sager Buurma and Jon Shaw, has produced over 1800 records in a MARC-based schema,30 representing novels published between 1660 and 1850. Its object of focus is individual copies of a given novel, describing a variety of features that require close human attention: prefaces, introductions, tables of contents, and so on. END generates new metadata for existing collections, which creates a beneficial and mutual relationship between librarians and archivists, faculty specialists in literature, and their student collaborators. That literary scholars and their students have learned, in essence, one of the fundamental structuring data schemas of librarianship (i.e., MARC), means that the project is far more likely to grow beyond its original scope. And it has: while its own dataset is largely bounded by the institutional and geographic limitations of its Philadelphia-based team, the data is openly available and the schema extensively documented on GitHub, providing a model for other research teams to add on their own, site-specific data. I have experimented with having my own students in eighteenth-century novel courses prepare descriptions of Auburn’s small collection of relevant eighteenth-century printed texts for incorporation into the END, and I have found it takes only a minimal amount of ramp-up time to do so. A project need not be massive or even large to be open, extendable, and useful. A good example is Reading With Austen (RWA),31 a project led by Peter Sabor and managed by Catherine Nygren and Megan Taylor, which creates a visual representation of the shelves in the library of Godmersham Park, the estate of Jane Austen’s brother, Edward Austen Knight. They were particularly fortunate to have an 1818 handwritten catalog32 that included the shelf locations of each book. The majority of the books, collectively named The Knight Collection, are still at the estate (now Chawton House Library). The longer-term goal of the project is to find the “lost sheep” via the distinctive Knight family bookplates, and either purchase them for reincorporation into the Knight Collection or virtually “reunite” them through photographs of the book’s spine, bookplate page, title page, and any pages with marginalia. This project requires support from the 30  MARC, or MAchine-Readable Cataloging Record, was developed by Henriette Avram, Sally McCallum, and Lucia Rather at the Library of Congress in the 1960s to create shared standards for libraries. While it has been revised over the years, MARC continues to be the lingua franca of structured data for libraries. 31  Reading With Austen, accessed April 15, 2019, http://www.readingwithausten.com/ 32  This catalog was turned into a spreadsheet by Deborah Bygrave and Hugh MacKay under the direction of Stephen Bending and Stephen Bygrave.

364 

E. C. FRIEDMAN

public, and thus it is extremely transparent in its process. While its core dataset has not yet been published, the project website notes that they will release that clean dataset under a Creative Commons License, understanding that many queries can be run with such a dataset that would not be possible with their library-shelf visualization or catalog search interface. RWA is able to be open with its data because of two reasons. First, it received funding from Social Sciences and Humanities Research Council (SSHRC) Canada Research Chairs Program, which did not put any requirements of exclusivity or monetization on the output. Second, the Knight family already had an entire apparatus of agreements that made their family home and the Knight Collection accessible to scholars through the creation of Chawton House Library and a long-term loan of the Knight Collection. Without the cooperation of Richard Knight, it would have been far more difficult, if not impossible, to have created and disseminated the underlying spreadsheet, to say nothing of the visualizations and photographs that are part of the front end. Other projects on historical libraries face such restrictions as they come to fruition. Orlando: Women’s Writing in the British Isles from the Beginnings to the Present (Orlando), led by Susan Brown, Patricia Clements, and Isobel Grundy since 2008, is a rich database on the history of women’s writing in Britain. Its encyclopedia entries and other data allows users to search for women across an enormous sweep of interpretive tags, to move between and among them through densely hyperlinked entries on people, primary and secondary material, and contextual information, and to create timelines for their own use. It is also an important example of a massive scholarly project that is largely controlled by a publisher (i.e., Cambridge University Press) that requires that it be held behind a paywall, and while its tag structure can be revealed through the user interface, the user interface cannot be superseded: the site’s data cannot be analyzed outside of the framework of the interface. While the latter obstacle is unavoidable, Orlando provides two workarounds for the challenge of the paywall: free access during March (Women’s History Month), and a scheme in development that allows contributors access to the project in exchange for their labor. This is an important initiative, and one that is very much in line with the ideals of the editorial team. It is important to note that Orlando here stands in for innumerable other projects, large and small, that now try to balance the needs of the editorial and technical team with those of the end users. It is easy to call for open access, but someone must pay for the labor that goes into the

10  AFTERWORD: NOVEL KNOWLEDGE 

365

work of transcribing, scanning, encoding, coding, building, and then sustaining and continually updating a large-scale text set or database. As scholarly publishing must increasingly justify its expenditures using the value systems of the market, it is little surprise that publishers would look to support subscription-based digital projects or create public interfaces but retain valuable back-end data. As I will note in the next section, we should not necessarily assume that this is needed when it comes to smalland medium-scale datasets.

Conclusion: Toward Best Practices In short, our work would be much easier if we were more able to work together. Here, I propose what I believe are reasonable best practices for those of us that are working with or creating large datasets connected to Anglophone fiction in the long eighteenth century. I suspect I am not saying anything new to the majority of practitioners, but unlike (for example), the Text Encoding Initiative (TEI), which makes collective standards for the sustainable and interoperable encoding of individual texts, there is no collective or regulatory body for the interoperability of datasets in a similar fashion. Thus, it is worth reiterating and making explicit practices that make our work iterative—not just for today, but for generations to come. It is my hope that this is only the starting point of a larger conversation about creating durable and interoperable datasets, and that refinements and additions will come to address omissions or new challenges or affordances that come with new technology. Clarify the Definition of “Novel” and “Fiction” in Our Work.  As a collective action, I realize this may well be as much of a pipe dream as any articulated in this chapter. I can barely imagine a singular agreed-upon definition that all scholars of the novel or even prose fiction would agree upon. But ensuring that we articulate our individual project’s boundaries in this respect is critical; also, ensuring that the datasets, bibliographies, or other sources of lists employ the same definition is absolutely vital. Acknowledge Sample Size and Sources.  As literary scholars, we need to get into the habit of thinking about how our evidence arrives to us. We need to be explicit early and often about where we get our data, especially when we are extrapolating them into numbers. Most (if not all) of us work from either small text sets we have created ourselves, or from preexisting

366 

E. C. FRIEDMAN

large-­scale corpora, and neither is without its attendant blind spots and challenges. As I have discussed already, large-scale corpora give the illusion of comprehensiveness, which can be misleading. And small text sets potentially suffer from selection bias. It is almost surprising that Da does not take on the question in her tests of statistical methodology from extant projects, but it is possible that she sees that as well-trod territory. I want to affirm that there is not one standard threshold to determine the appropriateness of a conclusion. If we deal with eighteenth-century fiction, we always deal with a small slice of the printed material produced in the period—a reality we do not always acknowledge, but we should. As Leah Orr notes in her study of the early eighteenth-century print fiction market, when we talk about new titles in a given year, we talk about very small numbers: even including reprints, translations, and the like, just 475 different works of fiction were printed in England in the period 1690–1730.33 If you chose to focus on new works each year, you would be referring to a much smaller set still. There’s lots we can say about fiction and its readers, but it is crucial to keep scale in mind. Contextualize.  As noted above, gaps and silences in data are nearly inevitable when dealing with eighteenth-century fiction. Those who read our work or engage with our data should have a clear understanding of the curatorial choices that went into selection and those choices should be defensible through evidence drawn from knowledge about the production of texts in the period. To the extent possible, a project should note the gaps in data. As Nathan Yau has pointed out, missing data is itself a useful piece of information; therefore, visualizations should use white space, variable scale, and treat absence itself as a category in order to represent this absence effectively.34 Open Dataset.  Ideally, every peer-reviewed publication that relied on enumerative or visualized datasets would require either a publicly available attachment of the corpora and tools used or, at minimum, to have that information stored in institutional repositories (even if not publicly available). As I have discussed in this chapter, some do—many don’t. One of the most searing footnotes in Da’s essay for me was the following: “the process of requesting complete, runnable codes and quantitative results  Orr, Novel Ventures, 103.  Nathan Yau, “Visualizing Incomplete and Missing Data,” FlowingData, accessed April 15, 2019, https://flowingdata.com/2018/01/30/visualizing-incomplete-and-missing-data/ 33 34

10  AFTERWORD: NOVEL KNOWLEDGE 

367

(tables, output data, matrices, measurements, and others) took me nearly two years. Authors and editors either never replied to my emails, weren’t able or willing to provide complete or runnable scripts and data, or gave them piecemeal only with repeated requests.”35 Da’s experience is common across disciplines. Digital humanities are not out of step with all other data-heavy fields in this respect: while collaboration and data sharing is common in a handful of fields, a 2002 study published in The Journal of American Medical Association showed that 45% of geneticists withheld data because of the attendant expenses, and 80% noted that the effort required to share data made them unlikely to do so.36 A 2012 study that interviewed researchers in various fields revealed that lack of disciplinary standards and repositories shaped whether an individual researcher was likely to share his or her data with others.37 For a variety of reasons that I have discussed in this chapter, publicly accessible datasets or text sets are not always achievable. The major bibliographies are all decades away from entering the public domain. The costs of permissions of various sorts become unwieldy at anything approaching large scale. While the published fiction we work with is officially out of copyright by every conceivable measure, the photographs and scans of those texts are not. Moreover, not all fiction produced in the period was “published”: as of the time of this writing, never-published literary manuscripts produced before the mid-twentieth century will remain in UK copyright for another twenty-one years and declaring such a work “orphaned” requires up-front fees with no assurance of success. Even when institutions waive such fees and grant permissions, making these images available to a larger public still incurs the recurring costs of server space and maintenance. Even among scholars, data is power—but power of a strange kind. Because we, as a scholarly community, do not value dataset production in the ways we would monographs: publication of data comes with no reward or incentive. Thus, data is too often only valued as 35   See note 2  in Da, “The Computational Case against Computational Literary Studies,” 602. 36   E.  G. Campbell, B.  R. Clarridge, M.  Gokhale, L.  Birenbaum, S.  Hilgarten, N. A. Holtzman, and D. Blumenthal, “Data Withholding in Academic Genetics: Evidence from a National Survey,” Journal of American Medical Association 287, no. 4 (2001): 473–80, https://doi.org/10.1001/jama.287.4.473 37  Youngseek Kim and Jeffrey M.  Stanton, “Institutional and Individual Influences on Scientists’ Data Sharing Practices,” Journal of Computational Science Education 3, no. 1 (2012): 47–56, https://doi.org/10.22369/issn.2153-4136/3/1/6

368 

E. C. FRIEDMAN

the raw material from which publications come—publications which cannot be thoroughly vetted, peer-reviewed, understood, much less built upon, without that data being available. Moreover, because there may be another publication that might emerge from it, there is every reason to guard one’s data until a sufficient number of publications have been wrung out of it—if indeed the data is ever made public. This is a value system that we must collectively transform if we are ever to succeed. Extendable.  Some gaps are the product of data that can never be retrieved. But gaps like those of ECCO-TCP (and, indeed, ECCO itself) can ultimately be filled in, though the scale of the task is daunting. No one research team, no matter how well-funded, is likely to transcribe all known fiction from the period. Instead, we should work and create documentation for our data that begins with the assumption that our work will be picked up and continued by others. And we must make space—physical, digital, conceptual—for that work to occur in. Indeed, data visualization is not a static process: today’s network visualization or graph can and should be expected to be superseded by new information gained by the team, by new ways of asking questions of existing data or text sets, or ideally by both. Many of our current ways of building data collections and conducting data analysis still fail to wholly tackle these challenges.

Bibliography Algee-Hewitt, Mark, Sarah Allison, Marissa Gemma, Ryan Heuser, Franco Moretti, and Hannah Walser. 2016. Canon/Archive. Large-Scale Dynamics in the Literary Field. Literary Lab Pamphlet 11 (January): 1–13. https://litlab. stanford.edu/LiteraryLabPamphlet11.pdf Bode, Katherine. 2018. A World of Fiction: Digital Collections and the Future of Literary History. Ann Arbor: University of Michigan Press. Burkert, Mattie. 2017. Recovering the London Stage Information Bank: Lessons from an Early Humanities Computing Project. Digital Humanities Quarterly 11 (3): n.p. http://www.digitalhumanities.org/dhq/vol/11/3/000321/ 000321.html Campbell, E.G., B.R.  Clarridge, M.  Gokhale, L.  Birenbaum, S.  Hilgarten, N.A.  Holtzman, and D.  Blumenthal. 2001. Data Withholding in Academic Genetics: Evidence from a National Survey. Journal of American Medical Association 287 (4): 473–480. https://doi.org/10.1001/jama.287.4.473.

10  AFTERWORD: NOVEL KNOWLEDGE 

369

Da, Nan Z. 2019a. The Computational Case against Computational Literary Studies. Critical Inquiry 45 (3): 601–639. https://doi.org/10.1086/702594. ———. 2019b. The Digital Humanities Debacle: Computational Methods Repeatedly Come Up Short. The Chronicle of Higher Education, March 27. https://www.chronicle.com/article/The-Digital-Humanities-Debacle/245986 Evans, Aden. 2005. Sound Ideas: Music, Machines, and Experience. Minneapolis: University of Minnesota Press. Friedman, Emily C. 2020. Amateur Manuscript Fiction in the Archive: An Introduction. In After Print: Eighteenth-Century Manuscript Cultures, ed. Rachael Scarborough King, 217–236. Charlottesville: University of Virginia Press. Garside, Peter, James Raven, and Rainer Schöwerling, eds. 2000. The English Novel 1770-1829: A Bibliographical Survey of Prose Fiction Published in the British Isles. Volume I: 1770–1799. Oxford: Oxford University Press. Kim, Youngseek, and Jeffrey M.  Stanton. 2012. Institutional and Individual Influences on Scientists’ Data Sharing Practices. Journal of Computational Science Education 3 (1): 47–56. https://doi.org/10.22369/ issn.2153-4136/3/1/6. Orr, Leah. 2017. Novel Ventures: Fiction and Print Culture in England, 1690-1730. Charlottesville: University of Virginia Press. Schroeder, Caroline T., and Amir Zeldes. 2016. Raiders of the Lost Corpus. Digital Humanities Quarterly 1 (2): n.p. http://www.digitalhumanities.org/ dhq/vol/10/2/000247/000247.html Spedding, Patrick. 2011. ‘The New Machine’: Discovering the Limits of ECCO. Eighteenth-Century Studies 44 (4): 437–453. https://www.jstor.org/ stable/41301590. Underwood, Ted. 2014. Theorizing Research Practices We Forgot to Theorize Twenty Years Ago. Representations 127 (1): 64–72. https://doi.org/10.1525/ rep.2014.127.1.64. Underwood, Ted, Michael L.  Black, Loretta Auvil, and Boris Capitanu. 2013. Mapping Mutable Genres in Structurally Complex Volumes. In Proceedings of the 2013 IEEE International Conference on Big Data, ed. Xiaohua Hu et al., 95–103. Silicon Valley, IEEE. https://doi.org/10.1109/BigData.2013. 6691676. Underwood, Ted, Boris Capitanu, Peter Organisciak, Sayan Bhattacharyya, Loretta Auvil, Colleen Fallaw, and J. Stephen Downie. 2015. Word Frequencies in English-Language Literature, 1700–1922. [Dataset], HathiTrust. https:// doi.org/10.13012/J8JW8BSJ Yau, Nathan. 2019. Visualizing Incomplete and Missing Data. FlowingData. https://flowingdata.com/2018/01/30/visualizing-incomplete-and-missingdata/. Accessed 15 Apr 2019.

370 

E. C. FRIEDMAN

Open Access  This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Index1

A Abdul-Rahman, A., 13n47 Abel, Carl Friedrich, 331, 331n46 The Aberdeen Musical Society, 324 Actor data, 96–97 Addison, Joseph, 88, 109, 270, 273, 273n19, 291, 293 Aesthetics of the grid, 223 Agent-based modeling, 12 Aldridge, Alfred Owen, 207n26 Algee-Hewitt, Mark, 15n57, 16n60, 266n2, 361, 362n23 Algorithmic manipulation, 20, 158, 162–163 Algorithmic search, 357 Allahyari, Mehdi, 165n26 Allison, Sarah, 362n23 Altieri, Charles, 65n9 Altman, Janet Gurkin, 233–235, 233n3, 235n6 Alvarado, Rafael, 269 Ambiguous data, 43 Anderson, Benedict, 68, 68n20, 114

Arch, Stephen Carl, 234, 234n5 Armstrong, Isobel, 183, 183n46 ARTFL Encyclopédie, 16 ARTFL Project, 14, 14n55 Assefi, Mehdi, 165n26 Associated lexis, 141, 148 Atteveldt, Wouter Van, 164n22 Austen, Jane, 81, 363 Author names, 158, 169, 173, 173n35, 183 Author networks, 48, 68n18 Authorship, 12, 41–42, 88–89, 102n77, 266, 273n12, 276, 284 attribution, 12 Auvil, Loretta, 357n15, 361n21 Avison, Charles, 22, 309–345 B Backscheider, Paula, 183, 183n45 Bag-of-words (BoW) approach, 164, 164n24 Bailey, Jefferson, 5n13, 10, 10n34

 Note: Page numbers followed by ‘n’ refer to notes.

1

© The Author(s) 2021 I. Baird (ed.), Data Visualization in Enlightenment Literature and Culture, https://doi.org/10.1007/978-3-030-54913-8

371

372 

INDEX

Baird, Ileana, 12n45, 22, 270n9, 276n27 Baker, Houston A., 66n14 Bamman, David, 102n77 Bandry-Scubbi, Anne, 12n42 Banned books, 40, 57 Bar chart, 5, 38 Bar graph, 8, 254 Barabási, Albert-László, 288n42 Barash, Carol, 183, 183n46 Barber, Joseph, 318–319 Barbeu-Dubourg, Jacques, 6, 21, 199, 207, 207n26, 207n29, 208, 208n31, 209–211, 216, 218, 224, 227 Barnard, John, 71n27, 84n54 Bayly, C. A., 141n31 Bell, John, 95, 232, 234, 249, 251, 251n25, 257–260 Bell, Maureen, 71n27, 84n54 Bender, John, 7, 8n19, 204, 205n16, 216n48 Benedict, Barbara M., 67n17 Beniger, James R., 8n20 Benjamin, Walter, 200n2 Benoit, Kenneth, 164n22 Bentham, Jeremy, 203 Bentley, Richard, 286 Berry, Michael W., 163n22 Bertin, Jacques, 10, 11n35 Best practices, 23, 352–353, 365 Bestsellers, 18, 35, 37, 45, 51, 51n34, 79, 90 Bevilacqua, Alexander, 122, 122n3 Bhattacharyya, Sayan, 361n21 Big data, 2, 18, 64, 232, 357 Bigram, 189, 189n51 Binder, Jeffrey M., 12n43 Binding lists, 126–127, 132, 134n21, 139 Bing Maps, 209n36 Black, Michael L., 357n15 Blain, Virginia, 183, 183n46

Blei, David M., 163, 164n23 Block, Sharon, 12n43 Bloom, Harold, 66, 66n14 Bode, Katherine, 356, 357n13, 362, 362n25 Boileau-Despréaux, Nicolas, 267, 275 Bolla, Peter de, 127n13–14, 130, 130n17 Bonnell, Thomas F., 67n17, 96n71, 98n74, 107n83, 108, 109n85 Book trade, 18–19, 23, 34, 37, 39, 48, 54–57, 68, 73n38, 74, 95–97, 99, 101–102, 105, 112, 114, 319n14 Bookseller, 35–36, 36n12, 40, 43, 45, 56–57, 95–97, 269, 354 Bourdieu, Pierre, 67n16, 114, 114n92 Boyd-Graber, Jordan L., 164n23 Bradford, Richard, 66n15 Bredvold, Louis I., 186n50 Bremner, Robert, 317, 319, 319n18, 334, 340 British corpus, 123, 127–129, 131–132, 137, 142, 145 British Fiction, 1800–1829: A Database of Production, Circulation, and Reception, 354 Brooke, Frances, 13n50, 21, 102, 231–232, 232n1, 233, 235, 235n7, 236, 238, 240, 245, 247, 249–250, 253, 253n27–28, 257–258, 258n32–33, 259–260, 260n34, 261n35 Brown, Laura, 169n32, 187 Brown, Susan, 364 Bruni, Flavia, 72n33 Brydone, Patrick, 48n31 Buache, Philippe, 8 Buci-Glucksmann, Christine, 200, 203, 203n11–12, 204 Buck-Morss, Susan, 211, 211n39–40, 213, 213n41 Burke, Edmund, 136, 136n24–26, 137

 INDEX 

Burkert, Mattie, 356, 356n12 Burney, Frances, 48n31, 102, 233, 233n2, 262 Burrows, John, 12n41, 12n47 Burrows, Simon, 3n4, 16, 18, 34n4, 55n39, 265n1 Büsching, Anton-Friedrich, 48n31 Buurma, Rachel Sager, 363 Byron, Lord, 142 C Cambridge School, 69 Camp visibility, 281, 294 Cannan, Paul D., 280n29 Canon, 17–20, 35, 54, 63, 64, 64n3, 65, 65n4, 66, 66n15, 67, 67n16, 67n18, 68–69, 69n22, 70, 70n25, 71, 74, 75, 75n41, 76–78, 78n45, 79n48, 79n50, 80–87, 87n60, 88n63, 89–90, 93–95, 97–98, 100–105, 104n80, 108–109, 111–112, 162, 183, 187, 189, 191, 292 accessible, 68, 69 commercial, 69 critical, 66, 68, 69, 189 data-driven, 18–19, 63, 71, 77–78, 81, 86–87, 93–95, 97, 101, 112–113, 113n91 dataset, 75 early modern, 18, 63–114 formation, 17, 19, 66–68, 66n15, 67n17–18, 93, 112–114, 266, 268 index, 64n1–2, 78, 78n45, 113 literary, 80, 82–83, 95, 162 official, 68–69 personal, 68–69 possible, 68, 80, 187 potential, 68–69, 97 selective, 68–69 See also Fowler, Alastair

373

Canonical authors, 23, 91, 93–94, 101 Capitanu, Boris, 357n15 Carroll, Noel, 65n7 Cartographic grid, 209 Catchword, 139n27, 241–243 Centlivre, Susanna, 102, 285 Chambers, Ephraim, 7 Chang, Jonathan, 164n23 Character associations, 281 Character relationships, 268, 271 Character writeprints, 266 Charmantier, Isabelle, 213n43 Chart, 1–2, 4, 5n10, 6, 8, 8n20, 9, 16, 22, 38, 57, 96, 99, 159, 205n19, 207n29, 211, 224, 231–232, 240, 254, 257, 270, 312, 358 Chawton House Library, 363–364 Cheema, M. F., 20n63 Chilcot, Thomas, 320, 327, 330, 330n41–42, 331, 343, 343n71 Chronological machine, 207, 209 Chronological visualization, 252 Cibber, Colley, 267n5, 269, 272, 283, 286–289, 293–294 Circle graphs, 5 Circo, 270, 270n8, 292–293 Cirrus, 236 visualizations, 243 City Readers. Digital Historic Collections of the New York Society Library, 354n7 Ciula, Arianna, 162n18 Classification algorithm, 357 Clements, Patricia, 364 Clergy of the Church of England Database, 335n56 Clergy subscribers, 334–335, 336n59 Close reading, 13, 17, 20, 275 Clusterability, 169, 169n33 Clustering algorithm, 163, 166, 169, 173 Coding in R, 231

374 

INDEX

Coleridge, Samuel T., 293 Collaborative writing, 88 Collier, Jeremy, 276, 281, 290 Collins, Irene, 81n51 Comédie Française Registers Project (CFRP), 14 Community-detection algorithm, 134n22 Computational literary studies, 353 Computational manipulation, 156 Computational modeling, 158, 192 Computational stylistics, 12 Computer-assisted text analysis, 158 Computer graphics, 10 Concanen, Matthew, 283 Connectors, 22, 265, 268, 270n8, 287, 287n42, 288–289, 294 Conrad, Sebastian, 4n7, 123n4 Conroy, Melanie, 15n57 “Containing” concept, 129, 130n17 See also Bolla, Peter de Contour map, 6 Cooke, Benjamin, 319–320, 333 Cooke, Thomas, 281, 283, 293 Coordinate grid, 205 Coordinate system, 200, 205, 205n20–21 Corpus-time, 137, 142 Correspondence networks, 13–14, 16, 33 Counterpublic, 291 See also Fraser, Nancy Crawford, Robert, 67n17 Crome, August Friedrich Wilhelm, 9 Cuccuru, Kathrine, 280n30 Culture of diagram, 7. See also Bender, John; Marrinan, Michael Cultures of Knowledge, 13 Curll, Edmund, 267, 269, 273, 276, 281–283, 285n37, 286–290, 293–294 Curtius, Ernst Robert, 65

D Da, Nan Z., 353, 366–367 d’Alembert, Jean-Baptiste Le Rond, 7 d’Argens, Marquis, 35 Darnton, Robert, 35–36, 36n12, 40, 45, 45n21, 46, 53–55, 57n42 Data analysis, 266, 360, 368 Data collection, 72n32, 205, 262, 368 Data distortion, 43 Data-driven index, 78 Data integrity, 43 Data mapping, 4 Data mining, 236 Data modeling, 162 Datapoint, 242, 252–255, 257 Data quality, 64n3, 71n29, 95 Data scraping, 231 Dataset, 10, 15, 17, 18, 23, 33, 39, 43, 45, 72, 73n36, 74, 75, 75n40, 77, 87, 95, 98, 125, 156, 161–163, 246, 256, 262, 291, 354–357, 359–367, 361n21 Data sharing, 265, 367 Data source, 125 Data visualization, 2–6, 7n17, 9–11, 13–16, 18, 20–22, 153, 158, 161, 163, 165, 191–192, 199–200, 231–232, 235–236, 245, 247, 253, 257, 262, 265, 345, 351, 368 Data visualization programs, 235 Data visualization tools, 14, 235, 345 Davis, Stephen Boyd, 207n29, 209n33 Dawson, Robert L., 40n15 de Boyer, Jean Baptiste, 35 de Calonne, Charles-Alexandre, 51n34, 52n35 de Fourcroy, Charles, 9, 9n25 de Mairobert, Mathieu-François Pidansat, 35 de Montaigne, Michel, 87 de Saint-Vincent, Pierre-Augustin Robert, 51n34

 INDEX 

Demographic data, 9, 353 Dendogram, 12 Dennis, John, 22, 265, 268–270, 270n8, 271–272, 272n16, 273, 273n17–20, 274, 274n22, 274n24, 275–276, 276n26, 277–285, 285n37, 286–294, 294n53, 295–305 Depledge, Emma, 94n69 Descartes, René, 205, 205n20, 205n21, 223 Dewey Decimal Classification (DDC) system, 77 d’Herbelot, Barthélemy, 122 d’Holbach, Paul-Henri Thiry, 42, 51n33, 54 Diagonal grid, 211, 227 Diagram, 7–9, 12, 15, 19, 124, 134, 136, 144, 200, 203, 203n13, 204, 216n48, 221–222, 227 Diderot, Denis, 7, 42, 207 Digital archive, 2, 13 Digital book, 232 Digital humanities, 2, 4, 16–17, 23, 32–33, 77, 162, 231, 262, 353, 367 Digital pedagogy, 260 Digitization project, 14, 356, 360 Digitized dataset, 161 Digitized text collection, 162 Dirty data, 22–23, 351, 362 Disraeli, Isaac, 65, 65n11 Dissemination patterns, 17, 38 Distant reading, 17, 122 Distributional probability factor (dpf), 124n7, 125, 125n10, 126 Document clustering, 158, 165n26, 167n30, 173 Document-level metadata, 176, 176n37 “Do-it-yourself” diagram, 7 Dot algorithm, 270 Downie, J. Stephen, 361n21

375

Drucker, Johanna, 3, 11 Dryden, John, 67, 87, 95, 180, 186, 188, 281, 291, 293 Duara, Prasenjit, 123n6 E Eagleton, Terry, 65n6 Early English Books Online (EEBO), 3, 158n14, 359 Early Modern Letters Online, 14 The Early Novels Database (END), 23, 352, 362–363 Economic chart, 6, 211 Edelstein, Dan, 12n45, 15, 161, 246n20, 265n1 Edmondson, Chloe, 12n45, 15, 265n1 Edwards, Edward, 40n14 Ehret, Georg Dionysius, 214–215 Eide, Oyvind, 162n18 Eighteenth-Century Collections Online (ECCO), 3, 19, 121, 123, 359–361, 368 Eighteenth-century corpus, 123, 145 Eighteenth-century poetry, 20, 153–192 The Eighteenth-Century Poetry Archive (ECPA), 13 Eighteenth-Century Studies (ECS), 2, 10–23, 153–156, 156n8, 158, 161, 167, 174, 176, 180, 180n43, 186–187, 191–192, 232, 265, 357, 359 The Eighteenth Century: Theory and Interpretation (TEC), 20, 154–156, 174, 176, 180, 180n43, 186, 191 18thConnect.org, 13, 360 Eisenstein, Elizabeth, 68, 68n19 Electronic Enlightenment (EE), 14, 16 Elkins, James, 214, 214n44

376 

INDEX

The English Short Title Catalogue (ESTC), 18–19, 63–114 Epistolary novel, 232–233, 247, 262 Essex, John, 341, 341n66 Evan, Aden, 355, 356n9 Evans Early American Imprints, 359 Eve, Martin Paul, 294, 294n55 Explanation graphics, 10 Exploratory pattern recognition, 174 Extracted canon, 80 Ezell, Margaret J. M., 183, 183n46 F Fairer, David, 67n17, 186, 186n50, 187, 191 Fallaw, Colleen, 361n21 Farmer, Henry George, 324n23 Feather, John, 86n59 Febvre, Lucien, 68n19 Female authors, 101–102, 183, 188–189 Female poets, 183, 185, 186n50, 187, 189, 191 Female subscribers, 342–343 Ferguson, Stephen, 207, 207n26, 207n28, 208, 208n30, 209, 209n32, 209n34–36 Ferris, David S., 216n48 Fielder, Leslie A., 66n14 Fielding, Henry, 87 Finch, Anne, 187–189 Fischetti, William, 5n9 Fishkeller, M. A., 10 Flanders, Julia, 162, 162n18, 164n19, 165n25, 180n42, 312, 312n6 Fleming, Simon D. I., 22, 325n31 Forensic linguistics, 266 Foucault, Michel, 200, 203, 203n13, 204, 214, 216, 216n49, 217, 217n50, 218, 218n52, 219, 219n55, 219n57–58, 222, 226 Foulis, Andrew, 107, 107n84, 108

Foulis, Robert, 107, 107n84, 108 Fowke, Martha, 189 Fowler, Alastair, 68, 68n22, 69n23, 75n41, 81, 82n53 Fowlkes, E. B., 10 Franklin, Benjamin, 87, 207, 207n29, 247 Franzini, G., 20n63 Fraser, Nancy, 291, 291n51 Frautschi, Richard, 35n9 Freedman, Jeffrey, 34n5 Freemason, 330, 330n42, 345 French Book Trade in Enlightenment Europe, 1769–1794 (FBTEE), 14, 14n52, 18, 31–32, 32n2, 33, 33n3, 34, 34n536, 37, 46, 55n39, 56–58 French Revolution, 36, 54–55 Friedman, Emily C., 22–23, 352n4 Friendly, Michael, 9n30, 10n32 Funkhouser, H. Gray, 8n21, 9n27 Fussell, Paul, 186n50 G Gants, David. L., 77n43 Garside, P. D., 310n1, 356 Gaskell, Philip, 71n30 Gavin, Michael, 1, 12n44, 158n14 Gay, John, 273, 281, 285n37, 291, 293 Gemma, Marissa, 362n23 Gender distribution, 183, 187 Genre Group Corpus (GGC), 166–167, 169 Geographic patterns, 200, 223 Geographic visualization, 207 Geospatial analysis, 46, 49 Geospatial mapping, 12, 16, 33 Geospatial plotting, 248 Gephi, 231 Gerard, Christine, 186, 186n50, 187, 191

 INDEX 

Gerrish, Sean, 164n23 Gessner, Salomon, 48n31 Gies, David T., 4n8, 123n4 Gildon, Charles, 276, 291, 293 GitHub, 363 Gladstone, Clovis, 12n43 Gladwell, Malcolm, 287, 287n39, 287n42 Goldstone, Andrew, 158n13, 161, 161n15 Google Maps, 209n36 Google Web Mercator, 210n36 Gorak, Jan, 66n15, 70n25 Govindasamy, P., 165n26 GPS coordinates, 246, 248, 251, 254 Graff, Gerald, 158, 158n11 Grafton, Anthony, 267n6 Granovetter, Mark, 287, 287n40, 288n44, 290n50 Graph, 3–5, 8, 9n27, 10n32, 20, 22, 38, 154, 165–166, 176–178, 176n39, 177n40, 180, 185–189, 191, 205, 205n20, 205n22, 213, 222, 231–232, 238, 253–254, 257n30, 258, 260, 269–270, 270n8, 271, 282–289, 292, 368 data structure, 269 of barometric variations, 9 paper, 9n27, 205, 205n20 visualizer, 269 Graphical excellence, 11 Graphic design, 20–21, 200, 205–206, 213, 218, 222 GraphViz, 22, 270 Green, James, 103n79 Greetham, David, 32n1 Gregg, Melissa, 249, 249n24 Grid, 9n27, 20, 199–227 as organization, 213–215 pattern, 199–202, 202n8, 205, 213, 216, 223 as relation, 211–213 as time, 207–210

377

Gridded graph paper, 205n20 Gridded graphic method, 6, 207 Grid-pastiche, 222 Grimoard, Philippe-Henri, 51n34 Grolemund, Garrett, 164n22 Grundy, Isobel, 364 Guillory, John, 66n15 Gutierrez, Juan B., 165n26 H Haacke, Hans, 21, 224–225 Habermas, Jürgen, 68, 68n21, 114 Hackler, Ruben, 4n6 Haghighi, Anousheh, 12n47, 266n3 Halasz, Alexandra, 67n16 Halley, Edmond, 6, 6n16 Handel, George Frideric, 339 Hankins, Thomas L., 8n21 Hanneman, Robert A., 290n48 Harding, Edward, 95 Harmonized metadata catalogs, 19, 64 Harris, Wendell V., 64n4, 68–69n22, 114, 114n93 Harvey, David, 211n39 HathiTrust Research Center (HTRC), 360, 360n20, 361 Haywood, Eliza, 269–270, 270n9, 278, 281, 293 Heaney, Peter J., 226n70 Hecht, Jacqueline, 211n38 Helsinki Computational History Group (COMHIS), 14, 18, 64 Helsinki Centre for Digital Humanities (HELDIG), 14, 18 Henley, John “Orator,” 269, 293 Hesletine, James, 327, 334 Heuser, Ryan, 362n23 Higgins, Hannah B., 206, 206n25 Hill, Mark J., 18, 70n25, 73n37, 85n56, 85n58, 103n78 Histograms, 12, 16 Historical-biographical database, 19

378 

INDEX

Historical corpus, 19, 123–124 Historical text network, 158n14 Hodgson, Marshall G. S., 129n16 Holman, Peter, 326n33–34 Holmes, Nigel, 10, 10n33 Hooker, Edward N., 273n20 Hotson, Howard, 15, 15n58 Hoxby, Blair, 66n15 Hughes, David, 321n20 Hume, David, 63, 65, 65n7, 66, 131 Hunter, David, 310n3 Hunter, J. Paul, 67n17 Hyman, Nicola, 331n43–44 Hyman, Robert, 331n43–44 I Ideal data, 353–354 Illegal bestsellers, 18, 45 Illegal book trade, 37 Ilomäki, Niko, 70n27, 71n30, 109n87 Image-schema, 222, 222n65 Indicator diagrams, 8 Individual subscriber, 324, 327, 344 Infographics, 10, 232, 254 Informational illustration, 214 Informational images, 214, 218 Information analysis, 10 Information graphics, 10, 215 Information visualization, 10, 10n32, 218 Ingrassia, Catherine, 183, 183n45, 183n47 Interactive timeline, 253 See also Tableau Interoperable datasets, 365 Interpretational-relational database, 18, 37 Interpretive complexity, 3 Intertopic Distance Map, 177–178 Isopleth map, 15

J Jacob, Giles, 269–270, 274, 282–285, 285n37, 286–288, 293–294 Janicke, S., 20n63 Jannidis, Fotis, 162, 162n18, 163n19, 165n25, 174n42, 312, 312n6 Jennens, Charles, 339 Jennings, Collin, 12n43, 200n2 Jin, Rong, 164n24 Jockers, Matthew, 161, 161n16, 163–164n22 Johnson, John, 8n20, 317, 317n9, 319, 319n17 Johnson, Samuel, 15, 65, 66, 87–88, 95, 173, 187–189, 266, 266n2, 272, 272n13, 272n16, 280n30 Jones, Ewan, 127n13–14 Joselit, David, 227n71 Juratic, Sabine, 56n41 K Kant, Immanuel, 216n48, 222 Karian, Stephen, 70n27, 72n31 Kassler, Michael, 310n3 Keith, Jennifer, 154, 154n1, 191 Kirk, Andy, 2, 2n3 Kirsten, Guido, 4n6 Klein, Lauren, 351, 351n1 K-means algorithms, 158 K-means clustering, 4, 20, 163–183 Knight, Edward Austen, 363–364 The Knight Collection, 363–364 Knowledge network visualization, 15 Kochut, Krys, 165n26 Korshin, Paul J., 36n12 Kramnick, Jonathan, 65, 65n5, 67, 67n16–18 Krauss, Rosalind, 200–201, 201n3, 209 Kwartler, Ted, 164n22

 INDEX 

L Lahti, Leo, 18, 70n27, 71n28–30, 73n35–37, 85n56, 85n58, 103n78, 109n87 Lakshmanaprabu, S. K., 165n26 Lambert, Johann Heinrich, 8, 9n24 Landes, Joan B., 341n65 Landor, Walter Savage, 142 Latent discourse patterns, 158 LDA topic modeling, 20, 158, 163, 173, 176 LDAvis, 177, 177n40, 178n41 Leapor, Mary, 189 Lee, Sabrina, 102n77 Lefebvre, Henry, 288n43 Levin, Harry, 66n15 Levi-Strauss, Claude, 201 Lexical binding, 19, 129, 132, 134n21, 137, 143, 145, 148 Lexical co-association, 123, 125n10, 132, 135, 138–141 Lexical environment, 127, 132, 149 Lexicographic network, 15 Library metadata, 15 Liebersohn, Harry, 136n23 Lily, William, 92, 92n65, 93, 93n66, 94n68 Line graph, 5, 154, 180, 187, 189, 238 Linnaeus, Carl, 6, 21, 48, 48n30, 199, 213–219, 216n49, 218n52, 224, 226–227 Lippard, Lucy, 223n67 Locke, John, 93, 93n67, 247 The London Stage, 356 Lonsdale, Roger, 183, 183n44, 187, 189 Lopez, Jeremy, 67n16 Low, Gail, 66n14 Lu, Kristine, 12n47, 266n3 Lydia, E. Laxmi, 165n26

379

M Mace, Nancy A., 93n66 Mace, Thomas, 311 Machine chronologique, 6, 21, 199, 207–208, 210 Machine clustering method, 192 Machine learning technique, 266 Machine-Readable Cataloging (MARC) 21 standard, 72, 363 Maddison, Margaret, 321n20, 327n39 Male canon, 191 Male poets, 184–185, 186n50, 187–189, 191 Male subscribers, 342 Malherbe of Loudun, 45 Manipulation of models, 162 Manuscript Fiction in the Age of Print, 1750–1900 (MFAP), 23, 352 Mapping the Republic of Letters, 6, 9, 10n32, 12–16, 12n46, 18, 21–22, 31–33, 33n3, 36, 38n13, 39, 45, 50, 105n81, 156n6, 163, 167–168, 177–178, 189n51, 209–210n36, 210–211, 216n48, 222, 222n65, 224, 245, 245n19, 246, 246n20, 247, 249, 252, 262, 266, 270–271, 312, 320 Marche, Stephen, 17n62 Marjanen, Jani, 71n28–30, 71n28, 73n35, 73n36, 109n87 Marker of illegality, 40, 43 Marrinan, Michael, 7, 8n19, 204, 204n16, 216n48 Marshall, Julia, 214, 214n45 Martin, Angus, 35n9 Martin, Henri-Jean, 68n19 Mason, Rose, 310n3 Masonic circles, 330 Materiality, 19, 64, 109–112, 154 Maunder, Richard, 326n33–34 McCandless, David, 11, 11n38

380 

INDEX

McCarty, Willard, 156, 156n5, 161, 161n16, 162n18, 192 McNally, David, 211n40 McNeill, William H., 124, 124n9 Meaningful metadata, 23, 352 Meltzer, Eve, 222–223, 223n66–67, 226 Mercator, Gerardus, 105n81, 210, 210n36 Mercier, Louis-Sébastien, 35, 51–52n34 Metadata, 15, 19, 23, 64, 71n29, 73, 73n37, 95, 158n14, 176, 176n37, 351–352, 360–363 Meteorological data, 9 Meulen, David L. Vander, 267n5, 278–279 Meunier, Luce, 21, 224, 226 Milgram, Stanley, 287n41 Miller, Thomas P., 35n7, 67n17 Minard, Charles Joseph, 9 Minard diagram, 15 Mineral table, 226 Minsheu, John, 311 Missing data, 366, 366n34 Montagu, Lady Mary Wortley, 187–189, 191 More, Hannah, 87, 102 Moretti, Franco, 4, 4n6, 17, 17n61, 271, 271n12, 351n1, 362n23 Morillo, John, 280, 280n30 Morrissey, Robert, 12n43 Mukherjee, Ankhi, 66n14 Multidimensional scaling, 176n39, 178n41 Multivariate data, 5, 11 Munch, Thomas, 68n19 Muri, Allison, 12n46 Murphy, Avon Jack, 48n31, 272n14 Murthy, Viren, 123n6 Mylne, Vivian, 35n9

N Naranch, Bradley, 122, 122n2 Navigational maps, 211 Neato, 270, 270n8 Necker, Jacques, 51, 51–52n34, 52n35 Network, 203 connector, 22 diagrams, 12, 19, 124, 134, 144 of epistolary exchange, 233 graph, 4, 20, 165, 166, 176, 176n39, 177n40, 271 visualization, 12n45, 15, 134n22, 246–248, 262, 368 Networked public, 268, 291 Networking through subscription, 22, 345 Newman, David J., 12n43 Nicolai, Friedrich, 48n31 Non-western Enlightenments, 122, 125 Norwich Musical Society, 325 Nulty, Paul, 127n13–14 Nussbaum, Felicity A., 142n32, 169n32, 187 O O’Brien, Karen, 98n75, 131 Ohmann, Richard, 158, 158n11 Oldmixon, John, 278, 285, 293 Oostrom, Frits van, 65n4 Open access, 262, 354, 359, 364 Open dataset, 366–368 Optical character recognition (OCR), 157n9, 354, 359 Organisciak, Peter, 361n21 Orlando, 23, 352, 352n2, 364 Orr, Leah, 358, 358n16, 359, 366 Osterhammel, Jürgen, 128n15 Othering, 17, 20, 123n5, 124, 137, 145, 150

 INDEX 

Outlier, 22, 90, 251, 255, 265, 268, 285, 294 Outram, Dorinda, 49, 49n32 Oxford Musical Society, 325, 344 P Pagden, Anthony, 132n19, 140, 140n28, 149 Paine, Thomas, 87 Palladio, 21, 236, 245–246, 246n20, 247–254, 260, 262 Panofsky, Erwin, 200, 200n1 Paratext, 276, 280 Passeron, Jean-Claude, 67n16 Patey, Douglas Lane, 67n17 Patron, 88, 294, 344–345 Patterns in attention, 20, 186 of dissemination, 311 of large-scale cultural production, 66 of relationships, 266, 290 of subscription, 311 in word use, 127 Pearl, Lisa, 12n47, 266n3 Pettegree, Andrew, 72n33–34 Philosophie, 49, 54 Pie chart, 5, 22, 312 Piper, Andrew, 155n4, 158, 161, 161n16, 162n18, 163, 164n23 Pitts, Jennifer, 123n6, 335n58 Place names, 15, 19, 127–129, 147, 319, 320 Playfair, William, 1–2, 2n1, 5, 5n9, 5n12, 5n14, 6, 6n15, 7–8 Plot network, 22, 266, 271, 275, 291 Poetic canon, 20 Poetry criticism corpus (PCC), 157, 159, 168, 168n31, 169, 174, 176–181, 183–191

381

Pope, Alexander, 22, 81n52, 87–88, 95, 173, 187–189, 262, 266–267, 267n5–6, 268–270, 270n8–9, 271–272, 272n15–16, 273, 273n19, 274, 274n24, 275–276, 276n26, 280, 280n28, 281, 283–284, 284n34–35, 285, 285n37, 286–287, 290, 290n45, 291–304 Popkin, Jeremy D., 44n20 Posthumous publication frequency, 91, 93 Pouchet, Louis-Ézéchiel, 9, 9n28 Pouriyeh, Seyedamin, 165n26 Pregill, Lily, 5n13, 9n28, 10, 10n34 Prestige building, 17, 20 Priestley, Joseph, 8, 20, 207, 207n29 Print capitalism, 68, 114 Print culture, 12n46, 19, 71, 114, 179, 182, 355, 360 Printed graph paper, 9n27 Printing practices, 111, 139n27, 241 Print marketplace, 355 Proprietary databases, 355, 359 Proudfoot, Richard, 66, 66n13 Ptolemy, 210 Publication by subscription, 309–312, 343, 345 Publication count, 78, 78n45, 92, 106 Publication frequency, 78, 78n45, 81, 91, 93 Publication networks, 13 Publication output, 72, 78, 96 Public domain, 85–86, 91, 367 Public opinion, 280n30, 291, 294 Public sphere, 55, 68, 291 Publishing patterns, 64, 99 Purchasing patterns, 345

382 

INDEX

Q Quantitative analysis, 66, 72, 77, 158, 183, 186 Quantitative data, 4 Quantitative methods, 15–17, 20, 114, 158, 161, 191, 266, 268 Quesnay, François, 6, 21, 199, 211, 211n40, 212–213, 216, 218, 224, 227 R Ramsey, Stephen, 161, 161n15 Ramya, D., 165n26 Rare topics, 77 Raven, James, 12n46, 95n70, 356 Raw term frequency, 173n35 Reading by Numbers, 362 Reading Experience Database (RED), 57, 57n44 Reading habits, 111 Reading With Austen (RWA), 363–364 Reccia, Gabriel, 127n13 Regan, John, 19, 124n8 Relational database, 354 Relationality, 199, 216 Relational network, 219 Relative frequency, 84, 184–185, 187, 189, 189n51, 190 Reliable underlying data, 22, 351 Rhetorical visualizations, 227 Richardson, Samuel, 232–233, 266, 360 Riddle, Mark, 290n48 Rishton, Timothy J., 330, 330n42, 343n71 Robert Plot, 206n23 Roberts, Margaret E., 176n37 Robinson, David, 164n22 Robinson, Francis, 311n5 Robinson, Lillian, 66n14 Robinson, Mary, 189

Robinson, Peter, 65n10 Robyn, Dorothy L., 8n20 Rockwell, Geoffrey, 236, 241, 243 Roe, Glenn, 3n4, 16, 34n4, 265n1 Roivainen, Hege, 71n28 Root, Robert K., 186n50 Ross, Trevor Thornton, 64n4, 67n16 Rounce, Adam, 88, 88n62 Rousseau, Jean-Jacques, 54, 65, 87, 207, 341n65 Rowe, Elizabeth Singer, 102 Rumbold, Valerie, 268n7 S Sabor, Peter, 363 Sadie, Stanley, 310 Safaei, Saied, 165n26 Sahami, Mehran, 163n22 Said, Edward, 122 Säily, Tanja, 73n37 Sales statistics, 56–57 Sales trends, 50 Salkever, Stephen G., 207n27 Salzman, Paul, 66n14 Sample size, 248, 365–366 Sartori, Andrew, 123n6 Savard, Francine, 21, 223–224 Scatter map, 22, 312, 320 Scatterplot, 12, 100, 189–190 Scheuerman, G., 20n63 Schlup, Michael, 34n5 Schmidt, Benjamin M., 180n42 Schroeder, Caroline T., 356n10 Scientific visualization, 10n32 Scottish Enlightenment, 104 Seares, Margaret, 310n3 Seaward, Louise, 55n39 Seigworth, Greg, 249, 249n24 Selection bias, 366 Semantic diagram, 136 Semantic network diagrams, 19, 124, 134, 144

 INDEX 

Semantic network, 19, 123–124, 134, 134n22, 135–138, 140, 140n30, 141–144, 148–149 Separate markets hypothesis, 46 Seward, Anna, 187–189 Shakespeare, William, 66, 69, 81n52, 94–95, 114n92 SHANTI Interactive Visualization Application, 270 Shared lexical space, 134–137, 139, 145 Sharman, J. C., 124n9 Shaw, David, 319n14 Shaw, Jon, 363 Sher, Richard, 71n30 Sherburn, George, 186n50, 266n4 Shevlin, Eleanor, 58, 58n47 Shirley, Kenneth E., 178n41 ShivaGraph, 22, 270 Shuttleton, David, 183, 183n47 Sievert, Carson, 178n41 Silge, Julia, 164n22 Simsion, Graeme, 162n18 Sinclair, Stéfan, 236, 236n8–10, 241, 241n13, 243, 243n17 Siskin, Clifford, 67n17, 354 Sitter, John, 154, 154n3, 191 Skinner, Quentin, 69, 70n25 Smallwood, Philip, 280n30 Smeall, Cheryl, 15n57 Smith, R., 94n68 Smythe, James Moore, 283 Social network, 15–16, 22, 245–248, 262, 265–305 analysis, 12, 22, 245–252, 268, 271, 294 graph, 4, 271 Société typographique de Neuchâtel (STN), 14, 18, 33–40, 34n5, 42–46, 46n25–27, 47, 47n28, 48, 48n31, 49, 51, 51n33, 52, 52n34, 53, 55, 55n39, 56, 56n42 Sood, Gagan D. S., 133

383

Southey, Robert, 142, 321n20 Southey, Roz, 321n20, 327n39 Spatial-temporal analysis, 54 Spence, Ian, 5n9 Spencer, Jane, 67n18 Srivastava, Ashok N., 163n22 Stafford, Barbara Maria, 221–222, 222n65, 226 Stallybrass, Peter, 70n27 The Stanford Lab, 362 The Stationer’s Company, 101–102, 353, 355, 357 Statistical chart, 5 Statistical graphics, 9–10, 10n32, 205n20 Statistical graph, 5, 9 Statistical map, 9 Steele, Richard, 87–88 Steinberg, Leo, 214 Steinberg, Sigfrid H., 79n47 Stephens, Norris, 335, 335n58, 336n59 Stewart, Brandon M., 176n37 Stewart, Mary Beth, 92n65 Stm model, 178, 180 Stm package, 174, 180 STN database, 39, 49, 51–52 Structural topic modeling (stm), 176, 176n37 Structures of knowledge, 127, 156 Stylometry, 12, 266 Suarez, Michael F., 70n27 Subject-topics, 19, 75–77, 78n46, 79–86, 90–91, 99–100, 109–110, 112 Subject-topic distribution, 85 Subscription lists, 20, 22, 310, 310n3, 311, 311n5, 312–327, 340, 343n71, 344–345 Superbooks, 42, 42n16, 43 Sutherland, James R., 294n56 Swift, Jonathan, 81n52, 82n82, 87–88, 173, 187–189, 266, 284 Swiss-Parisian Enlightenment, 48 Symanzik, Jurgen, 5n9

384 

INDEX

T Table, 9, 10n32, 51, 64n2, 73, 77, 124, 143, 186, 199–200, 216, 218–219, 219n55, 219n57, 221, 224, 226, 232, 254, 270, 358, 363, 367 Tableau, 21, 36, 200, 219, 236, 246, 253–260, 262 Tabular notation, 6, 213 Tabulated data, 12, 22, 137, 147–148 Tabulation, 4, 242, 244, 312 Talbot, Michael, 310n3 Temporal analysis, 49 Temporal grid, 209 Temporal polyvalence, 235 Term frequency-inverse document frequency (tf-idf), 157, 157n10, 167 TermsBerry makers, 236 Terry, Richard, 67n16–17 Text Creation Partnership (TCP), 359, 359n17 Text Encoding Initiative (TEI), 365 Text mining, 158, 161, 163, 164n22, 164n24, 191–192 Text sets, 358, 360–361, 365–368 Textual analysis, 21, 231, 236–245, 257, 260–262 Thematic cartography, 10 Theobald, Lewis, 267n5, 274, 278, 280, 283, 285–286, 293 Thomson, Alex, 70, 70n26 Tillotson, Geoffrey, 186n50 Timeline, 94, 200, 207, 207n29, 209, 224, 234, 245–246, 252–260, 262, 364 bar graph, 8 charts, 8, 8n20, 9 of letter writing, 254 Tingley, Dusting, 176n37 Todd, Janet M., 183, 183n46 Tolonen, Mikko, 18, 70n27, 71n28–30, 73n36–37, 85n56, 85n58, 103n78, 109n87

Tonson, Jacob, 95 Topic cluster, 180, 182 Topic distribution, 16, 20, 165, 175, 182 Topic frequency, 182 Topic model, 165–166, 174n36, 176–178, 182, 186, 360 as network, 4, 12, 20, 176 visualization, 4, 12, 163, 177, 182 Topic modeling, 4, 12, 20, 158, 163–183 algorithms, 163, 164n23, 165, 169 Topological graph, 22, 270 Trending topics, 20, 182 Trend tracking, 236 Trippe, Elizabeth D., 165n26 TROVE database, 362 Tufte, Edward R., 5n11, 9, 9n29, 11, 11n36, 206, 206n24 Tukey, John W., 1–2, 2n2 Typewright, 360 U Underwood, Ted, 161, 161n15, 163n22, 164n23, 165, 165n25, 166n28, 176n39, 357, 357n14–15, 361n21 Underwood, William, 102n77, 158n13 Unigram, 189n51 Unique book trade actors, 96 V Vaara, Ville, 18, 73n36–37, 85n56, 85n58, 103n78, 109n87 Veuve Desaint, 56 Vicker, William, 80, 92 Visual data, 214 Visualization, 2–23, 33, 38n13, 105n81, 133–134, 134n22, 137, 144, 148, 153–192, 176n39, 199–227, 231–262, 265, 269,

 INDEX 

270, 281, 291, 345, 351–353, 356, 358, 360–361, 364, 366, 368 of information, 2, 4, 10, 10n32, 199–227, 249, 253, 261–262, 265, 358, 366, 368 literacy, 10 technologies, 13 tools, 3, 13–16, 21–22, 33, 192, 235–237, 253, 345 Visual noise, 206 Voltaire, 15, 42, 54, 87, 247 Voyant, 13, 21, 236–245, 257–260, 262 W Wainer, Howard, 205, 205n19–20, 205n22, 206 Waingrow, Marshall, 186n50 Walker, Robert, 95 Wall, Cynthia, 4n8, 123n4 Wallis, Peter, 309n1, 311n5 Wallis, Ruth, 309n1, 311n5 Wallnig, Thomas, 15, 15n58 Walser, Hannah, 362n23 Walsh, John, 334, 339–340, 340n64 Wang, Chong, 164n23 Warburton, William, 267, 267n5 Ward, Ned, 87, 278, 285 Watt, James, 8 Weak ties, 290. See also Granovetter, Mark S. Web-based visualizations, 13 Weber, Harold, 267n6 Weedon, Alex, 70n27 Weinbrot, Howard, 67n17 Weingart, Scott, 164n23 Welbers, Kasper, 164n22 Welsh, Jennifer M., 136n25 Welsted, Leonard, 278, 281–282, 293 Werner, Sarah, 72n30

385

Wickham, Hadley, 164n22 Wickman, Matthew, 205n21 Wise, Josh, 203, 203n12 Wittgenstein, Ludwig, 202, 202n6 Wittmann, Reinhard, 111n89 Wolff, Charlotta, 12n45 Wollenberg, Susan, 324n27–28 Woloch, Alex, 271, 271n11 Women subscribers, 342 Word cloud, 243 Word co-association, 19 Word Frequencies in English-Language Literature, 1700–1922, 361, 361n21 Word frequency, 125n10, 164n24, 173n36, 240, 242–243, 245, 257, 257n30, 259–260, 361n21 analysis, 158 trend, 257n30, 260 visualization, 260 Word-node, 134n22, 142 Wordsworth, William, 293 Word-use frequency, 21 Work-field dataset, 74–75, 98 Work-field harmonization, 75 Work-field identifier, 75 Work set, 360–361 Wrigley, E. Anthony, 319n14 Wynne-Davies, Marion, 66n14 Y Yau, Nathan, 366, 366n34 Yearsley, Ann, 187–188 Z Zeldes, Amir, 356, 356n10 Zerby, Chuck, 267n6 Zhang, Yin, 164n24 Zhou, Zhi-Hua, 164n24