Anton Marty and Contemporary Philosophy [1st ed.] 978-3-030-05580-6, 978-3-030-05581-3

Table of contents :
Front Matter ....Pages i-xvii
Introduction (Giuliano Bacigalupo, Hélène Leblanc)....Pages 1-9
Front Matter ....Pages 11-11
Natural Meaning and the Foundations of Human Communication: A Comparison Between Marty and Grice (François Recanati)....Pages 13-31
How a Statement Has Meaning by Expressing a Judgement—Brentano Versus Marty on Utterance Meaning (Mark Textor)....Pages 33-57
Determining and Modifying Attributes (Jan Claas, Benjamin Schnieder)....Pages 59-96
Front Matter ....Pages 97-97
A Presentation and Defense of Anton Marty’s Conception of Space (Ingvar Johansson)....Pages 99-119
Raum and ‘Room’: Comments on Anton Marty on Space Perception (Clare Mac Cumhaill)....Pages 121-152
Experiencing Change: Extensionalism, Retentionalism, and Marty’s Hybrid Account (Thomas Sattig)....Pages 153-171
Front Matter ....Pages 173-173
A Syncretistic View of Existence and Marty’s Relation to It (Alberto Voltolini)....Pages 175-196
Misleading Pictures, Temptations and Meta-Philosophies: Marty and Wittgenstein (Kevin Mulligan)....Pages 197-232
Back Matter ....Pages 233-237

Citation preview

H I S T O R Y O F A N A LY T I C P H I L O S O P H Y

ANTON MARTY AND CONTEMPORARY PHILOSOPHY EDITED BY

Giuliano Bacigalupo Hélène Leblanc

History of Analytic Philosophy

Series Editor Michael Beaney King’s College London Humboldt University Berlin Berlin, Germany

Series editor: Michael Beaney, Professor für Geschichte der analytischen Philosophie, Institut für Philosophie, Humboldt-Universität zu Berlin, Germany, and Professor of Philosophy, Department of Philosophy, King’s College London, UK. The main aim of this series is to create a venue for work on the history of analytic philosophy, and to consolidate the area as a major branch of philosophy. The ‘history of analytic philosophy’ is to be understood broadly, as covering the period from the last three decades of the nineteenth century to the end of the twentieth century, beginning with the work of Frege, Russell, Moore and Wittgenstein (who are generally regarded as its main founders) and the influences upon them, and going right up to the recent history of the analytic tradition. In allowing the ‘history’ to extend to the present, the aim is to encourage engagement with contemporary debates in philosophy, for example, in showing how the concerns of early analytic philosophy relate to current concerns. In focusing on analytic philosophy, the aim is not to exclude comparisons with other earlier or contemporary traditions, or consideration of figures or themes that some might regard as marginal to the analytic tradition but which also throw light on analytic philosophy. Indeed, a further aim of the series is to deepen our understanding of the broader context in which analytic philosophy developed, by looking, for example, at the roots of analytic philosophy in neo-Kantianism or British idealism, or the connections between analytic philosophy and phenomenology, or discussing the work of philosophers who were important in the development of analytic philosophy but who are now often forgotten. Editorial board members: Claudio de Almeida, Pontifical Catholic University at Porto Alegre, Brazil · Maria Baghramian, University College Dublin, Ireland · Thomas Baldwin, University of York, England · Stewart Candlish, University of Western Australia · Chen Bo, Peking University, China · Jonathan Dancy, University of Reading, England · José Ferreirós, University of Seville, Spain · Michael Friedman, Stanford University, USA · Gottfried Gabriel, University of Jena, Germany · Juliet Floyd, Boston University, USA · Hanjo Glock, University of Zurich, Switzerland · Nicholas Griffin, McMaster University, Canada · Leila Haaparanta, University of Tampere, Finland · Peter Hylton, University of Illinois, USA · Jiang Yi, Beijing Normal University, China · Javier Legris, National Academy of Sciences of Buenos Aires, Argentina · Cheryl Misak, University of Toronto, Canada · Nenad Miscevic, University of Maribor, Slovenia, and Central European University, Budapest · Volker Peckhaus, University of Paderborn, Germany · Eva Picardi, University of Bologna, Italy · Erich Reck, University of California at Riverside, USA · Peter Simons, Trinity College, Dublin · Thomas Uebel, University of Manchester, England. More information about this series at http://www.palgrave.com/gp/series/14867

Giuliano Bacigalupo · Hélène Leblanc Editors

Anton Marty and Contemporary Philosophy

Editors Giuliano Bacigalupo Sciences Humaines et Sociales (SHS) Université de Lille Villeneuve d’Ascq Cedex, France

Hélène Leblanc Département de Philosophie Université de Genève Geneva, Switzerland

History of Analytic Philosophy ISBN 978-3-030-05580-6 ISBN 978-3-030-05581-3  (eBook) https://doi.org/10.1007/978-3-030-05581-3 Library of Congress Control Number: 2018964582 © The Editor(s) (if applicable) and The Author(s) 2019 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, express 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. Cover illustration: Ben Horton\Getty Images This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Series Editor’s Foreword

During the first half of the twentieth century, analytic philosophy gradually established itself as the dominant tradition in the Englishspeaking world, and over the last few decades it has taken firm root in many other parts of the world. There has been increasing debate over just what “analytic philosophy” means, as the movement has ramified into the complex tradition that we know today, but the influence of the concerns, ideas and methods of early analytic philosophy on contemporary thought is indisputable. All this has led to greater self-consciousness among analytic philosophers about the nature and origins of their tradition, and scholarly interest in its historical development and philosophical foundations has blossomed in recent years, with the result that history of analytic philosophy is now recognized as a major field of philosophy in its own right. The main aim of the series in which the present book appears, the first series of its kind, is to create a venue for work on the history of analytic philosophy, consolidating the area as a major field of philosophy and promoting further research and debate. The “history of analytic philosophy” is understood broadly, as covering the period from the last three decades of the nineteenth century to the start of the v

vi     Series Editor’s Foreword

twenty-first century, beginning with the work of Frege, Russell, Moore and Wittgenstein, who are generally regarded as its main founders, and the influences upon them, and going right up to the most recent developments. In allowing the “history” to extend to the present, the aim is to encourage engagement with contemporary debates in philosophy, for example, in showing how the concerns of early analytic philosophy relate to current concerns. In focusing on analytic philosophy, the aim is not to exclude comparisons with other—earlier or contemporary— traditions, or consideration of figures or themes that some might regard as marginal to the analytic tradition but which also throw light on analytic philosophy. Indeed, a further aim of the series is to deepen our understanding of the broader context in which analytic philosophy developed, by looking, for example, at the roots of analytic philosophy in neo-Kantianism or British idealism, or the connections between analytic philosophy and phenomenology, or discussing the work of philosophers who were important in the development of analytic philosophy but who are now often forgotten. One group of philosophers who have indeed been relatively neglected in mainstream analytic philosophy, but who arguably constitute their own philosophical tradition—or indeed, traditions—with many points of connection with mainstream analytic philosophy, is the Brentano School, a group of philosophers who were influenced by the work of Franz Brentano (1838–1917). Brentano taught first at Würzburg, where he published his most influential book, Psychology from an Empirical Standpoint, in 1874, and then moved to the University of Vienna, where he taught until 1895. The members of the Brentano School include Edmund Husserl (1859–1938), who went on to found the phenomenological tradition, Kazimierz Twardowski (1866–1938), who established the Lvov–Warsaw School, and Alexius Meinong (1853– 1920), who studied with Brentano in Vienna and taught at Graz from 1882 until his death. There has been a lot of work recently on the connections between analytic philosophy and phenomenology; see, for example, the volume in this series edited by Mark Textor: Judgement and Truth in Early Analytic Philosophy and Phenomenology (2013). The Lvov–Warsaw School forms the core of the Polish analytic tradition; a volume on this, edited by Kevin Mulligan, Katarzyna Kijania-Placek

Series Editor’s Foreword     vii

and Tomasz Placek, has also been published in this series: Studies in the History and Philosophy of Polish Logic (2013). Brentano and Meinong are key figures in what has been called the Austrian analytic tradition. Meinong, most famous in mainstream analytic philosophy for his theory of objects, taken as including nonexistent as well as existent objects, both influenced and was criticized by Bertrand Russell in his early work. But another, less well-known figure is Anton Marty (1847–1914), who studied with Brentano in Würzburg (before he left for Vienna), did his doctorate under Hermann Lotze (1817–81) at Göttingen, and from 1880 taught at the Charles University in Prague, where he was Rector from 1895 to 1897. (Prague at the time was part of the Austro-Hungarian Empire.) His doctoral thesis was on the origin of language, and the philosophy of language was his central focus throughout his subsequent career. Influenced by Brentano’s theory of intentionality as well as his conception of descriptive (as opposed to genetic) psychology, Marty was concerned to develop what he called a “descriptive semasiology”. His main work was Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, the first volume of which was published in 1908. A full account of the “linguistic turn” that took place in the nineteenth and twentieth centuries, which is often but wrongly seen as characterizing analytic philosophy alone, would have to recognize Marty’s contribution. The present volume, edited by Giuliano Bacigalupo and Hélène Leblanc, is devoted to the work of Anton Marty. As they explain in their introduction, their aim is not simply to do greater justice to Marty’s place in the history of philosophy, or to his contribution to the development of analytic philosophy, but to breathe “some new life into his thought” by revisiting and exploring his ideas from a contemporary perspective. Divided into three parts, the first part addresses various issues in the philosophy of language, looking at the connection with the work of H. P. Grice (1913–88), among others. The second part turns to the philosophy of space and time, on which a book by Marty was edited and published two years after his death. We see here some of his ideas on the ontology and phenomenology of space and time. The final part is concerned with two metaphilosophical aspects of Marty’s thought,

viii     Series Editor’s Foreword

on existence and being, and on the way that philosophers can be misled by pictures embedded in our use of language. The first shows how rooted he was in Brentano’s writings on Aristotle’s metaphysics, and the second anticipates Wittgenstein’s later efforts—as part of his own Sprachkritik—to diagnose and combat the effect of misleading pictures. This volume succeeds very well in demonstrating the depth of Marty’s thinking, his place in the history of philosophy and the continuing fertility of his ideas in connecting with present concerns. Berlin, Germany December 2018

Michael Beaney

Contents

1 Introduction 1 Giuliano Bacigalupo and Hélène Leblanc References 8 Part I  Language and Communication 2 Natural Meaning and the Foundations of Human Communication: A Comparison Between Marty and Grice 13 François Recanati 1 Marty, Grice, and the Contrast Between Natural and Non-natural Meaning 13 2 The Natural Foundations of Non-natural Meaning 18 3 Back to Salome 22 References 29 3 How a Statement Has Meaning by Expressing a Judgement—Brentano Versus Marty on Utterance Meaning 33 Mark Textor 1 Introduction 33 ix

x     Contents

2 3

The Right Kind of Meaning 35 Non-natural Meaning Without Communicative Intentions 37 4 Brentano and Marty on Judgement and Utterance Meaning 39 5 Reconsidering the Distinction Between Natural and Non-natural Meaning 44 6 The Spoiler 45 7 Non-communicative Intentions to the Rescue? 47 8 Commitment to the Fore 49 9 Utterance Meaning and the Identity Thesis 50 10 What About Lying? 52 11 Commitment Generalized 53 12 Conclusion 54 References 56

4 Determining and Modifying Attributes 59 Jan Claas and Benjamin Schnieder 1 Introduction 59 2 Adjectives—A Primer 61 3 Determination and Modification 64 4 Modifying Adjectives as Meanings Shifters 70 5 On the Pros and Cons of the Meaning Shift Proposal 72 6 Two Variations of the Meaning Shift Proposal 77 7 Modifying Adjectives and Modes of Combination 82 8 Coda and Conclusion 86 References 94 Part II  Ontology and Consciousness of Space and Time 5 A Presentation and Defense of Anton Marty’s Conception of Space 99 Ingvar Johansson 1 Some Introductory Words 99 2 Marty’s Container Conception of Space in Outline 100

Contents     xi

3

Marty’s Container Conception of Space in Metaphysical Detail 104 4 The Handedness Argument I: Kant 107 5 The Handedness Argument II: Graham Nerlich 109 6 The Handedness Argument III: The Relationists’ Go-Up-One-Dimension Argument Refuted 113 7 Some Concluding Words 116 References 117

6 Raum and ‘Room’: Comments on Anton Marty on Space Perception 121 Clare Mac Cumhaill 1 Space and Place 121 2 Kant and Brentano—Dimensions of a Critique 124 3 Marty’s Radical Conception 129 4 Marty, Husserl and a Contemporary Translation 130 5 Space for Naïve Realism? 138 6 Awareness Versus Correlation 142 7 ‘Raum’ and Room 145 References 150 7 Experiencing Change: Extensionalism, Retentionalism, and Marty’s Hybrid Account 153 Thomas Sattig 1 An Extensionalist Account 154 2 A Retentionalist Account 157 3 Marty’s Hybrid Account 163 References 171 Part III  Meta-metaphysics and Meta-philosophy 8 A Syncretistic View of Existence and Marty’s Relation to It 175 Alberto Voltolini 1 Introduction 175 2 Syncretism About Existence 176

xii     Contents

3 How Marty Situates Himself in This Debate 183 References 194 9 Misleading Pictures, Temptations and Meta-Philosophies: Marty and Wittgenstein 197 Kevin Mulligan 1 Philosophy and Austrian Sprachkritik 197 2 Pictures of Modality 199 3 Pictures of Thinking and Its Object 201 4 Pictures of Logic and Propositions 206 5 How Temptations and Pictures Work 212 6 Meta-Philosophies 218 7 From Reid and Bentham to Brentano and Beyond 225 References 229 Index 233

Notes on Contributors

Giuliano Bacigalupo  has been until recently a Postdoctoral Research Fellow in the SNF project “Meaning and Intentionality in Anton Marty. At the Crossroads of the Philosophy of Language and Mind”, at the University of Geneva, Switzerland and is a member of Inbegriff— Geneva Seminar for Austro-German Philosophy. Previously, he has held positions at Seattle University, USA and at the University of Konstanz, Germany. Jan Claas is a Research Assistant (wissenschaftlicher Mitarbeiter) at Hamburg University in Germany. His main interests lie in the philosophy of language, in particular concepts and conceptual containment, as well as in the history of analytic philosophy, in particular the philosophy of Bernard Bolzano and its influence. Bringing both interests together, he has just finished his dissertation on Bolzano’s conception of concepts. Ingvar Johansson is Emeritus Professor in theoretical philosophy at Umeå University, Sweden. His philosophical interests have been centred round problems in, on the one hand, the philosophy of science, and, on the other hand, in ontology taken in a wide sense that includes problems xiii

xiv     Notes on Contributors

of the existence of space, universals, consciousness and social reality. He published, inter alia, Ontological Investigations. An Inquiry into the Categories of Nature, Man and Society (1989, 2004), and Medicine & Philosophy. A Twenty-First Century Introduction (2008). Hélène Leblanc is a Postdoctoral Research Fellow in the SNF/ANR project “SÊMAINÔ. Archéologie différentielle du signe linguistique” at the University of Geneva, Switzerland. Previously, she worked on the SNF project “Meaning and Intentionality in Anton Marty. At the Crossroads of the Philosophy of Language and Mind”. She is also a member of Inbegriff—Geneva Seminar for Austro-German Philosophy. She earned her Ph.D. in philosophy (Lille/Lecce) with a thesis on semiotic theories in early modern philosophy and has published several papers on this topic. Recently, she edited the thematic issue Beauty and Ugliness in the Austro-German Tradition for the Italian journal Paradigmi. Clare Mac Cumhaill  is Assistant Professor of philosophy at Durham University. She works on topics in the philosophy of perception, action, emotion and aesthetics—with a special focus on space, spatial properties, form and structural and formal explanation quite generally. She is co-editor of Perceptual Ephemera (OUP 2018), and co-director of the In Parenthesis project. Kevin Mulligan is Honorary Professor of analytic philosophy at the University of Geneva, Ordinary Professor of philosophy at the University of Italian Switzerland and Director of Research at the Institute of Philosophical Studies, Faculty of Theology, Lugano. He works on ontology, philosophy of mind and the history of Austrian philosophy. His numerous publications include Wittgenstein et la philosophie austro-allemande (2012), as well as the edited collection, Mind, Meaning and Metaphysics. The Philosophy and Theory of Language of Anton Marty (1990). Together with Laurent Cesalli, he led the SNF project “Meaning and Intentionality in Anton Marty: At the Crossroads of the Philosophy of Language and Mind”, at the University of Geneva (2014–2017).

Notes on Contributors     xv

François Recanati is a Research Fellow at the Centre National de la Recherche Scientifique (CNRS), Institut Jean Nicod, Paris. His work focuses on philosophy of language and mind. Among his numerous publications should be mentioned Meaning and Force (1987), Direct Reference (1993), Oratio Obliqua, Oratio Recta (2000), Literal Meaning (2004), Perspectival Thought (2007), Truth-Conditional Pragmatics (2010), Mental Files (2012) and Mental Files in Flux (2016). He is a co-founder and past president of the European Society for Analytic Philosophy and was elected a Foreign Honorary Member of the American Academy of Arts and Sciences in 2012. Since 2019, François Recanati holds the Chair of Philosophy of Language and Mind at the Collège de France. Thomas Sattig  is Professor of theoretical philosophy at the University of Tübingen. He works in metaphysics, philosophy of language and philosophy of mind. His publications include the books The Double Lives of Objects: An Essay in the Metaphysics of the Ordinary World (2015) and The Language and Reality of Time (2006). Benjamin Schnieder is Professor of theoretical philosophy at the University of Hamburg. His main areas of research are metaphysics, the philosophy of language and logic, and the history of theoretical philosophy. His publications include the monographs Substanz und Adhärenz—Bolzanos Ontologie des Wirklichen (Philosophia 2002) and Substanzen und (ihre) Eigenschaften (de Gruyter 2004), the anthology Metaphysical Grounding (Cambridge University Press, co-edited with Fabrice Correia), and numerous papers in philosophy journals. Mark Textor  is Professor of philosophy in King’s College London. His research interests are in logic and metaphysics, philosophy of language, the history of analytic philosophy and Austrian Philosophy. His publications include the books The Routledge Guide Book to Frege on Sense and Reference (2010) and Brentano’s Mind (2017). Alberto Voltolini  is Professor of philosophy of mind at the University of Turin. His researches aim at the concepts of reference, intentionality, perception, as well as fiction and depiction. His main publications include How Ficta Follow Fiction (2006) and A Syncretistic Theory of Depiction (2015).

List of Figures

Chapter 4 Fig. 1 Detachment of adjective and noun Fig. 2 Detachment of noun but not of adjective Fig. 3 No detachment of adjective, detachment of noun debatable Fig. 4 Detachment of noun but not of adjective (for syntactic reasons) Fig. 5 Neither detachment of adjective nor of noun Fig. 6 Detachment of adjective, entailment of negated noun Chapter 7 Fig. 1 An extensionalist account of our experience of change Fig. 2 A retentionalist account of our experience of change Fig. 3 The extensionalist element of Marty’s account Fig. 4 The retentionalist element of Marty’s account

61 62 62 63 63 63 155 160 164 166

xvii

1 Introduction Giuliano Bacigalupo and Hélène Leblanc

Anton Marty was at the centre of one of the most productive traditions in philosophy, which originated in the works and teaching of Franz Brentano and led to Alexius Meinong’s theory of objects, Edmund Husserl’s phenomenology, Roman Ingarden’s ontology of art, and, via Kasimir Twardowski, to the Lvov-Warsaw School. Indeed, several among the theories being argued for or against in the contemporary debate have their forerunners in this tradition. This surely applies to the question whether intentionality should be deemed the ‘mark’ of the mental, a claim that received its—by now classic—formulation in Brentano’s Psychology from an Empirical Standpoint (1874). The same also holds true of the debate between one-level and two-level theories of self-consciousness, which owes much to Brentano’s attempt to provide a synthesis between these two accounts. Looking beyond Brentano, G. Bacigalupo (*)  UMR 8163 Savoirs, Textes, Langage, University of Lille, Lille, France H. Leblanc  Department of Philosophy, University of Geneva, Geneva, Switzerland e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_1

1

2     G. Bacigalupo and H. Leblanc

we may note that the most influential approach to philosophy of perception, namely the intentional theory, has in Husserl’s writings one of its most sophisticated formulations. The widely discussed realist approach to fictional objects was also developed by Amie Thomasson, following the blueprint of Roman Ingarden’s ontology of literary works. Finally, Meinong was even granted the honour of seeing some of the most distinguished contemporary metaphysicians such as Richard Routley, Terence Parsons, Ed Zalta and Graham Priest (self-)labelled as Meinongians. However, Anton Marty, a student and lifelong friend of Brentano, has remained on the sidelines. Not only would we be hard-pressed to name contemporary philosophers who consider themselves ‘Martyans’. As it happens, many philosophers would even have difficulty in placing Marty within the history of philosophy. The reason for the current neglect is mainly due to the fact that Marty is often considered to have simply played the role of the defender of the Brentanian orthodoxy vis-à-vis the innovations introduced by some of his students, first and foremost Meinong and Husserl. True, several pages of Marty’s writings provide sharp and often harsh criticism of heresies such as the introduction of assumptions (Meinong) and symbolic intentions (Husserl) as new fundamental classes of mental phenomena. Many pages are also focussed on defending Brentano’s revision of logic against the criticism raised by philosophers whose names have long been forgotten. Yet this should not overshadow the fact that Marty provides a, previously unparalleled, rich and careful analysis of language, so that we may speak of one of the first philosophies of language worthy of that name (Cesalli and Mulligan 2017: 251). Nor should his interests be seen as confined to language, since he developed an original ontology to address eternal philosophical problems such as, for instance, the objectivity of knowledge and the nature of time and space. Here is where Marty distances himself most dramatically from Brentano’s doctrines. Finally, one may find a common thread running through Marty’s works in his meta-metaphysical considerations on the notions of existence and reality. This collection of eight original essays pursues the aim of bringing Marty back into the spotlight. It does so by having leading figures in the

1 Introduction     3

contemporary debate confront some of his most significant contributions. This undertaking is not devoid of risk since most contributors are not specialists of Anton Marty, nor do they pretend to write as such. As it happens, we would like here to thank them all for accepting this challenge and not shying away from some very technical aspects of Marty’s doctrines and his at times arduous Swiss-German prose. Thus, the aim of this volume is not a reconstruction of Marty’s theories for historical purposes, but rather, that of breathing some new life into his thought, via the lenses of a contemporary perspective. The book is divided into three parts. The first part is dedicated to themes in philosophy of language (Chapters 2–4), which were at the centre of Marty’s philosophical thinking throughout his life. This interest may be traced back to his dissertation on the question of the origin of language (Marty 1875)—a pivotal topic of the philosophical discussion in the second half of the nineteenth century. A series of articles published in the 1880s and 1890s (Marty 1884, 1894, 1895) focusses on a problem at the crossroads of philosophy of language, logic and linguistics: namely, whether the most general and fundamental form of a judgement should be considered predicative or not. Marty’s opus magnum (Marty 1908) finally provides a full characterization of the scope of the discipline of philosophy of language and its various sub-disciplines. This book also offers Marty’s mature attempt at developing an account of language deeply rooted in the philosophy of mind which he had inherited from Brentano. The second part focusses on the problem of the objectivity and phenomenology of space and time (Chapters 5–7), upon which Marty was working in the final years of his life. The notes he left have been edited and published posthumously by his students under the title Raum und Zeit (Marty 1916). It may be interesting here to point out that even this study was seen by Marty as subordinated to his focus on language and was intended to lead to a further study on the linguistic expressions relating to space and time. As already noted by Simons (1990), this particular focus allows Marty to develop subtle phenomenological observations and distinguish which aspects of space and time are given intuitively or only discursively. However, the heart of Marty’s book is a defence of the view of space and time as mind-independent existing objects, which are only ‘filled in’ by objects and events, respectively.

4     G. Bacigalupo and H. Leblanc

The final part (Chapters 8 and 9) turns to Marty’s meta-metaphysical and meta-philosophical considerations. Brentano’s work The Manifold Meanings of Being in Aristotle (1862) made an everlasting impression on the young Marty and led him to pursue his interests in philosophy under the mentorship of its author (see Fréchette and Taieb 2017: 2). Indeed, Marty’s reflections on the notion of being provide a common thread across all his writings. A leitmotiv of these reflections is that language misleads us into making false assumptions about the nature of existence—to wit, to interpret it as a property of objects. Marty’s theory of metaphors—what he refers to as the ‘inner form of language’—places him in a privileged position to generalize this mistrust of language into a full-fledged meta-philosophy: we are systematically misled by our language when we rely on it to advance philosophical theories. From this perspective, Marty presents a strong affinity with ideas one may find in the writings of Wittgenstein and, more generally, in Ordinary Language Philosophy. 1. Language and Communication The first part of the book opens with François Recanati’s contribution (Chapter 2), which explores the possibility of reinterpreting Marty as providing an alternative account of communication to the one offered by Paul Grice. This is an original undertaking since Marty, notwithstanding some crucial differences, is often considered as a forerunner of Grice (see Liedtke 1990; Cesalli 2013; Longworth 2017). Yet Recanati argues that while Grice has built his account of communication on a strict dichotomy between natural and non-natural meaning, Marty allows us to uphold a continuity between them. Crucial to Recanati’s argument is the consideration of the famous Gricean example of King Herod presenting John Baptist’s head to Salome as a sign of John Baptist’s death. Drawing a strict demarcation line between what pertains to natural and non-natural meaning seems to lead to unwanted results in this most telling borderline case. Of equal importance to Recanati is also the interplay of natural and non-natural meaning in indexicals.

1 Introduction     5

In the second chapter of this section (Chapter 3), Mark Textor, not unlike Recanati, ventures on uncharted territories. While Recanati explored a possible gap between Marty’s account and Grice’s, Textor draws our attention to a crucial difference between Marty and Grice’s accounts on the one hand and Franz Brentano’s on the other, with respect to linguistic meaning: indeed, only the first two rely on the notion of communicative intention, whereas Brentano does not. While it is tempting to see this as an advantage of Marty’s and Grice’s accounts, Textor argues that the Brentanian account is the one that dovetails with our phenomenology of communication. Communicative intentions are often at stake in communication, but they are not strictly necessary and are, indeed, in many an instance completely absent. In the last chapter of this section, Jan Claas and Benjamin Schnieder investigate the notions of determining and modifying adjectives (Chapter 4)—a key distinction introduced by Brentano and further developed by Marty and other Brentanians. The original intuition to distinguish between two kinds of adjectives is derived from examples such as ‘dead man’ or ‘painted landscape’: a dead man is not a man and a painted landscape is not a landscape, so that the adjectives ‘dead’ and ‘painted’ are deemed to modify the meaning of the expression in question. To Brentano and Marty, philosophically loaded notions such as ‘being presented’, ‘being possible’, ‘being past’ or ‘being future’ may also be interpreted along the same lines: they modify the meaning of the expression so that, for instance, a presented man is not a man. After relating the distinction between determining and modifying adjectives to relevant notions in contemporary linguistics, Claas and Schnieder raise some powerful objections to Marty’s framework. Furthermore, the authors provide a second, as yet unexplored path for clarifying the distinction, by referring to a suggestion that may be traced back to Bernard Bolzano. 2. Ontology and Consciousness of Space and Time Ingvar Johansson opens the second part (Chapter 5) with a strong case for the exceptional position of Marty’s conception of space in the

6     G. Bacigalupo and H. Leblanc

history of philosophy. He is deemed to be the first philosopher to have defended what Johansson labels as a ‘container conception of space’, i.e. the view that space is a mind-independent entity which is only ‘filled in’ by objects. Furthermore, the author considers this view not only as being of historical relevance, but also as an approach that may be defended today, even within a relativistic framework. This is especially the case since Graham Nerlich (1976) unintentionally revived Marty’s view and presented a strong argument in its defence. The argument which lies at the heart of Johansson’s discussion revolves on the geometrical phenomenon of handedness, i.e. that 3-dimensional bodies with obvious similarity in shape may nonetheless be incongruent. Clare Mac Cumhaill in her contribution (Chapter 6), is also very sympathetic with Marty’s container view of space, which she labels as the ‘Naive Presupposition’. It is indeed very intuitive, even naive, to consider space to be something different from the objects that may or may not fill it. Mac Cumhaill’s real focus is then an exploration of which theory of perception is best suited to the Naive Presupposition, of course provided we grant that space—as a mind-independent object—may be experienced. The challenge here is that, if we embrace the container view, space is causally inefficacious. This is a crucial trait to both Marty’s and Nerlich’s accounts, as Johansson also pointed out. So, how can we be aware of something which is not causally interacting with us? According to Mac Cumhaill, the best candidate would be Naive Realism. Furthermore, again following the thought of Mac Cumhaill, what Marty himself has to say about perception should not be over-hastily interpreted as a declination of an intentional theory of perception. To the contrary, several of his remarks on this topic open the way to a naive realist interpretation. The final chapter of this section turns to Marty’s philosophy of time and, more precisely, of time-consciousness. Thomas Sattig’s contribution (Chapter 7) is indeed not so interested in the ontology of time developed by Marty, but rather, in the subtle apparatus he develops to cash out our experience of change. Two alternative accounts dominate the contemporary debate: an extensionalist and a retentionalist one. While the former account seems best suited to explain our ordering of events, first of all within the specious present, the latter shows its

1 Introduction     7

strength in upholding the phenomenal difference between perception and retention/memory. In his analysis, Sattig shows how the Brentanian distinction between two essentially different modes of consciousness, namely presenting and judging, is exploited by Marty to develop a hybrid account. While presentations allow us to order the objects of our experience in a given time-order, judgements are crucial for our experience of objects as present or past. 3. Meta-Metaphysics and Meta-Philosophy In his contribution to the third and final part of this volume, Alberto Voltolini draws a comparison between the syncretistic account of existence that he himself developed (Voltolini 2012) and Marty’s account of existence and being(s) (Chapter 8). A first obstacle for such an endeavour is that Marty explicitly rejected any form of syncretism relating to existence: to him there is one, and one only, meaning of the predicate ‘to exist’, to wit: the Brentanian attitudinal approach. To exist is to be correctly affirmed. However, since Marty distinguishes between kinds of beings, a form of pluralism is introduced at least at the level of his ontology. As Voltolini remarks, Marty is not a fundamentalist, since he allows for different kinds of beings. Indeed, next to concrete, real beings, Marty also envisages non-real, in a sense ideal or abstract beings. Moreover such beings are crucial to his epistemology, since one kind of being, namely states of affairs, plays the role of the truth-makers of our judgement. Now, if one considers the interplay between the attitudinal approach to existence and the distinction between different kinds of beings, some deeper similarities and differences with a syncretistic position do indeed come to the fore, and these are carefully explored by Voltolini. In the final contribution, Kevin Mulligan draws attention to Marty’s Sprachkritik as an anticipation of the strategy employed by Ludwig Wittgenstein (Chapter 9). If one considers some of Marty’s and Wittgenstein’s reflections on the topics of modality, intentionality and logic, both of them agree that philosophers have a tendency to fall prey to certain images webbed into our language. It is because of these temptations that some of the most extravagant theories may be accounted

8     G. Bacigalupo and H. Leblanc

for. However, Mulligan concludes by highlighting the divide between these prima facie similar meta-philosophies. Whereas to Wittgenstein philosophy as such is nothing over and above the weakness of falling captive of the images of our language, Marty only considers this to be a characteristic of bad philosophies. Thus, according to the latter but not to the former, there still is room for a cathartic, positive role in philosophical theorizing. Indeed, we hope that this will also be the spirit summoned in the readers of this collection of essays. An essential inspiration for the coming into being of this book came from a conference by the same title which took place at the University of Geneva on June 15–17, 2017. This conference was the closing event of the 3-year SNF project ‘Meaning and Intentionality in Anton Marty: At the Crossroads of the Philosophy of Language and Mind’ led by Kevin Mulligan and Laurent Cesalli at the University of Geneva. We would like to thank the SNF for its financial support. Many thanks to Laurent Cesalli, Guillaume Fréchette, Kevin Mulligan and Hamid Taieb for their help and suggestions during the preparation of this volume.

References Brentano, F. 1862. Von der mannigfachen Bedeutung des Seienden nach Aristoteles. Freiburg: Herder. Cesalli, L. 2013. Anton Marty’s Intentionalist Theory of Meaning. In Themes from Brentano, ed. D. Fisette and G. Fréchette, 139–163. Amsterdam: Rodopi. Cesalli, L., and K. Mulligan. 2017. Marty and Brentano. In The Routledge Handbook of Franz Brentano and the Brentano School, ed. U. Kriegel, 251– 263. London and New York: Routledge. Fréchette, G., and H. Taieb (eds.). 2017. Mind and Language—On the Philosophy of Anton Marty. Berlin: De Gruyter. Liedtke, F. 1990. Meaning and Expression: Marty and Grice on Intentional Semantics. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 29–49. Dordrecht: Kluwer. Longworth, G. 2017. Grice and Marty on Expression. In Mind and Language—On the Philosophy of Anton Marty, ed. G. Fréchette and H. Taieb, 263–284. Berlin: De Gruyter.

1 Introduction     9

Marty, A. 1875. Über den Ursprung der Sprache. Würzburg: A. Stuber. English translation in R.D. Rollinger. Philosophy of Language and Other Matters in the Work of Anton Marty. Analysis and Translation, 133–234. Amsterdam: Rodopi, 2010. ———. 1884. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, I–III. Vierteljahrsschrift für wissenschaftliche Philosophie, 8: 56–94, 161–192, 292–340. Reprinted in A. Marty, Gesammelte Schriften II.1, ed. J. Eisenmeier, A. Kastil, and O. Kraus, 1–35, 36–62, 62–101. Halle: Max Niemeyer, 1901/1920. ———. 1894. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, IV-V. Vierteljahrsschrift für wissenschaftliche Philosophie, 18: 320–356; 421–471. Reprinted in A. Marty, Gesammelte Schriften II.1, ed. J. Eisenmeier, A. Kastil, and O. Kraus, 116–145, 146– 189. Halle: Max Niemeyer, 1901/1920. ———. 1895. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, VI–VII. Vierteljahrsschrift für wissenschaftliche Philosophie, 19: 19–87, 263–334. Reprinted in A. Marty, Gesammelte Schriften II.1, ed. J. Eisenmeier, A. Kastil, and O. Kraus, 189–247, 247– 307. Halle: Max Niemeyer, 1901/1920. ———. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, Bd. 1. Halle: Niemeyer. ———. 1916. Raum und Zeit. Aus dem Nachlass des Verfassers herausgegeben von J. Eisenmeier, A. Kastil, und O. Kraus. Halle a. S.: Niemeyer. Nerlich, G. 1976. The Shape of Space. Cambridge: Cambridge University Press. Simons, P. 1990. Marty on Time. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 157– 170. Dordrecht: Kluwer. Voltolini, A. 2012. All Existences That There Are. Disputatio 32: 361–383.

Part I Language and Communication

2 Natural Meaning and the Foundations of Human Communication: A Comparison Between Marty and Grice François Recanati

1 Marty, Grice, and the Contrast Between Natural and Non-natural Meaning Several authors have noted the proximity of Marty’ ideas to Grice’s (Liedtke 1990; Cesalli 2013; Longworth 2017). Both Marty and Grice distinguish natural meaning (the meaning of natural signs) and the sort of meaning involved in human communication (the meaning of gestures and utterances); and they both attempt to provide a characterization of human communication that does not essentially appeal to the conventional nature of the linguistic devices it standardly uses. Natural meaning is the same thing for Marty as it is for Grice. Both take natural meaning (as illustrated by ‘dark clouds mean rain’, or ‘those dark clouds mean that it will rain’, an example they both use) to be a matter of consequence: x means y just in case y can be (correctly) inferred from x. Natural meaning is ‘factive’, as Grice puts it. If those dark clouds mean that it will rain, then it will rain (Grice 1957/1989: 213). F. Recanati (*)  College de France, Paris, France © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_2

13

14     F. Recanati

If it does not rain, Marty points out, that means that those clouds did not really mean that it would rain but were wrongly taken to mean that (Marty 1908: 281). This feature is shared by Dretske’s notion of ‘indication’, itself patterned after Grice’s natural meaning (Dretske 1988: 55–56). For Grice, natural meaning stands in contrast to ‘non-natural’ meaning. Non-natural meaning is first and foremost what a person means—‘speaker’s meaning’, as it is often called. The Indian guide’s insistent gesturing toward the sky means that it will rain, in the non-natural sense, just in case what the Indian guide (the person) means by this gesture is that it will rain. In such a case, the gesture can be reported as meaning ‘It will rain’,1 and it does not follow that it will actually rain. Non-natural meaning is not factive: The guide may well be mistaken or deceitful. In contrast to natural meaning, Gricean non-natural meaning is not a matter of what follows from what, but is a matter of intention: someone means something by a gesture or an utterance just in case the gesture or utterance is made with a certain communicative intention, where a communicative intention is an intention to achieve a certain effect in the hearer via the recognition of this intention (Grice 1957). Communicative intentions have a nested structure and (as Grice and his followers came to realize) potentially involve an infinite sequence of sub-intentions pertaining to the recognition by the hearer of a previous sub-intention (Grice 1969/1989: 94–99). The intention whose recognition by the audience is intended by the speaker to mediate the fulfilment of the speaker’s communicative aims goes beyond the speaker’s intention i1 to achieve a certain effect in the hearer: it includes her intention i2 that the hearer recognizes intention i1, as well as her intention i3 that the hearer recognizes intention i2, and so on ad infinitum. That series of conditions, required to make the speaker’s communicative intention fully ‘overt’, can be captured by letting the communicative intention be reflexive: it is the intention I to achieve a certain effect in H (the hearer) by means of H’s recognition of I (Searle 1969; Bach and Harnish 1979; Bach 1987).2 In Marty we do not find such a sharp contrast between natural meaning and non-natural meaning. For Marty as for Grice, communicative

2  Natural Meaning and the Foundations of Human …     15

behaviour is guided by what Strawson calls ‘audience-directed intentions’ (Strawson 1964), but Marty construes communicative behaviour as embedding natural meaning somehow, and in any case as continuous with natural meaning. Marty notes that human behaviour itself can be a natural sign; in particular, it can be a natural sign of ‘internal psychological processes’ in the subject who so behaves (Marty 1908: 283). This is particularly true of involuntary behaviour, such as a scream (a natural sign of pain) or tears (a natural sign of sorrow). The relation of the behaviour to the psychological state of which it is a natural sign is traditionally called ‘expression’, and I will retain that name.3 According to Marty, human communication has an expressive core: someone who says that p expresses her belief (judgment) that p, just as someone who cries expresses her sorrow (Marty 1908: 285). To be sure, human communication is a deliberate activity, not an involuntary one (Marty 1908: 284). But the expression of one’s psychological states can itself be deliberate. Suppose that, upon being hurt, the subject screams. If the subject actually is in pain and the scream is caused by the pain, there is no reason to deny it the status of a natural sign, even if the subject could have inhibited the externalization of his or her internal state. The fact that the externalization was within the subject’s control (since he or she could have inhibited it) means that the externalization was wilful or deliberate, to some extent at least. It did not take place against the subject’s will. Still, it remains a natural sign of the pain which caused it. The fact that it is deliberate is not sufficient to turn the expression of one’s psychological state into a communicative act, however. What is distinctive of human communication, for Marty, is the fact that the communicator’s primary intention is to manipulate the mental states of his or her audience, by instilling in him or her a matching attitude towards the object of the expressed thought (Marty 1908: 284–292). The revelation of the communicator’s own psychological state is only a means to that end. If the subject openly reveals her state by behaving in a certain way, that is the deliberate production of a natural sign but does not, or not necessarily, count as an instance of communication proper, because the subject may lack the further intention to induce a particular psychological state in the audience. Thus Wharton distinguishes

16     F. Recanati

behaviours that are deliberately produced and behaviours that are deliberately shown (Wharton 2009: 31). The deliberate production of a natural sign counts as an instance of communication in Marty’s sense if it is accompanied and motivated by the primary intention to affect the audience’s mental life in the relevant way. This shows that for Marty, in contrast to Grice, something can be both an instance of natural meaning and a bona fide instance of human communication.4 A putative example of that sort of case is Grice’s wellknown Salome example (Grice 1957/1989: 218).5 When Herod presents Salome with the severed head of St. John the Baptist on a charger, he produces a natural sign of St. John the Baptist’s death, and thereby conveys to Salome that St. John is dead. This is the wilful exploitation of a natural sign for communicative purposes. Now Grice insists that Herod’s ostensive sign is not an instance of non-natural meaning. Herod intends to induce in Salome the belief that John the Baptist is dead, but he does not intend to induce that belief in her by means of the recognition of that intention. The crucial reflexivity is missing: the severed head of John the Baptist is a natural sign which, by itself, is sufficient to induce the relevant belief in Salome. Marty, however, would presumably construe the Herod example as an instance of human communication exploiting the natural meaning relation. What is distinctive of human communication, for Marty, is the fact that the communicator’s primary intention is to manipulate the mental states of his or her audience, by instilling in him or her an attitude towards the object of thought that (in standard cases at least) matches or replicates the speaker’s own attitude. In the Salome case the condition seems to be satisfied: Herod intentionally shares with Salome his knowledge that St. John the Baptist is dead, by openly producing a natural sign of his death. Distinct from but closely related to that type of case is the simulation of a natural sign: by screaming, as if she were in pain, the subject implies that she is. In this case the scream no longer is a natural sign. It is a faked natural sign, where ‘fake’ is what Brentano and his students, including Marty, called a ‘modifier’.6 A faked natural sign is no more a natural sign than a faked gun is a real gun. The factivity constraint no longer holds: from the fact that the subject screams, it no longer follows that she is in pain. (The pretended scream may still be a natural sign, but of

2  Natural Meaning and the Foundations of Human …     17

something else than the pain: it is a natural sign of the subject’s deceptive intention. More on this below.) For Marty, presumably, both the wilful exploitation of natural signs and the production of fake signs simulating natural meaning are instances of communication, provided the communicator’s primary intention is to manipulate the mental states of his or her audience, by instilling in him or her an attitude towards the object of thought corresponding to the attitude the speaker expresses or pretends to express. The paradigm case of human communication is, of course, linguistic communication. According to Marty, if the speaker says that p, she expresses her belief that p. If the speaker is sincere, the utterance is a natural sign of the belief (just as a scream is a sign of pain). If the utterance is not sincere, it is a faked natural sign. Either way, the speaker deliberately produces the sign to reveal, or pretend to reveal, to the audience her psychological state (the fact, or alleged fact, that she believes that p ), but that is not the speaker’s primary intention. The speaker’s primary intention is to induce the belief that p in the hearer. That intention is (intended to be) fulfilled mediately, through the (alleged) revelation to the hearer of the speaker’s own state. If the hearer takes the speaker to be sincere and well-informed, he or she will take the speaker’s belief that p itself to be indicative of the fact that p. In other words, we have a chain of (alleged) natural signs. In the basic cases, the speaker’s utterance is a natural sign of the psychological state which it expresses, and that state in turn is a natural sign of the state of affairs which is the content of the belief. The basic cases are the cases in which the speaker is sincere and well-informed, and the hearer is trustful (Marty 1908: 286–287). The non-basic cases (e.g. the cases in which the speaker attempts to deceive the hearer or does not know what she is talking about) presuppose the basic cases. Even the liar pretends to be telling the truth. Now we see what the difference is, for Marty, between the involuntary expression of a psychological state (the scream, or the tears) and human communication. Natural meaning involves only one semiotic relation (Marty 1908: 280–281): a natural sign indicates (to use Dretske’s term) what can be correctly inferred from it. In human communication there are three distinct semiotic relations at work, not just

18     F. Recanati

one. An utterance (or gesture—the difference does not matter) expresses the speaker’s psychological state. Through her utterance the speaker reveals (or pretends to reveal) her psychological state to the hearer, but that is not the aim of communication. The aim is to induce a corresponding state in the hearer, via the hearer’s recognition of the speaker’s alleged psychological state. The hearer will be led to infer that p from the fact that the speaker believes that p and intends her to share that belief. In the basic cases, the chain of natural signs makes the utterance itself a sign of the fact that p (Marty 1908: 293–294), and that is the third semiotic relation which Marty considers at work in a linguistic utterance. In his framework an utterance has three kinds of meaning (Marty 1908: 288–294)7: (1) Expressive meaning (Kundgabe): the utterance expresses, and reveals to the hearer, the speaker’s psychological state (the belief that p, in the judicative case); (2) Communicative meaning: the utterance expresses the speaker’s communicative intention to induce in the hearer a psychological state corresponding to that expressed by the speaker. (3) Denotational meaning 8: the utterance means that p.

2 The Natural Foundations of Non-natural Meaning Let us now summarize what Marty’s theory of communication and Grice’s have in common, as well as the differences between them. They have in common two things. First, in contrast to most semantic theories, the relation between a linguistic utterance and its worldly content (the fact that p ) is mediated by the relation between the utterance and various psychological states, one of which is a communicative intention. Second, the speaker’s communicative intention is (intended to be) fulfilled mediately, via the hearer’s recognition of the speaker’s psychological state (and of his or her intention to reveal it openly to the hearer). There are two main differences between Marty and Grice. First, Grice’s communicative intentions have a nested/reflexive structure. That arguably is Grice’s major contribution to the analysis of

2  Natural Meaning and the Foundations of Human …     19

non-natural meaning, and there is no anticipation of that idea in Marty’s work (Longworth 2017). Second, Marty views linguistic communication as continuous with natural meaning, which it embeds within a chain which itself can be construed as a chain of natural signs (even though Marty is not explicit on this issue). Grice, on the other hand, insists on the irreducible difference between natural meaning and non-natural meaning. The second difference is more apparent than real, however. In ‘Meaning Revisited’, published 25 years after his seminal ‘Meaning’, Grice himself argues that natural meaning is the ancestor of non-natural meaning: Here I am interested not so much in the existence of [the distinction between natural and non-natural meaning], which has now, I think, become pretty boringly common ground (or mutual knowledge), but rather in the relationship between the two notions, the connections rather than the dissimilarities between them. […] What I want to do now is look to see if one would represent the cases of non-natural meaning as being descendants from, in a sense of ‘descendant’ which would suggest that they were derivative from and analogous to, cases of natural meaning. I shall also look a little at what kind of principles or assumptions one would have to make if one were trying to set up this position that natural meaning is in some specifiable way the ancestor of non-natural meaning. (Grice 1982/1989: 284–292)

This is very close to the ideas I ascribed to Marty. Nor is this a change of mind on Grice’s part. The stance taken in ‘Meaning Revisited’ is anticipated in ‘Meaning’, where Grice discusses cases of intentional exploitation of natural signs which seem to cast doubt on the sharp contrast he draws between natural meaning and non-natural meaning. In ‘Meaning’, Grice writes: If I frown spontaneously, in the ordinary course of events, someone looking at me may well treat the frown as a natural sign of displeasure. But if I frown deliberately (to convey my displeasure), an onlooker may be expected, provided he recognizes my intention, still to conclude that I am displeased. Ought we not then say, since it could not be expected to make

20     F. Recanati

any difference to the onlooker’s reaction whether he regards my frown as spontaneous or intended to be informative, that my frown (deliberate) does not meanNN anything? I think this difficulty can be met; for though in general a deliberate frown may have the same effect (with respect to inducing belief in my displeasure) as a spontaneous frown, it can be expected to have the same effect only provided the audience takes it as intended to convey displeasure. That is, if we take away the recognition of intention, leaving the other circumstances (including the recognition of the frown as deliberate), the belief-producing tendency of the frown must be regarded as being impaired or destroyed. (Grice 1957/1989: 219)

The important point is that, according to Grice, a natural sign ceases to be a natural sign as soon as the hearer recognizes that it is produced deliberately. From the deliberate scream, one can no longer infer that the screamer is in pain, since an alternative explanation for his or her (deliberate) scream is that he or she intends to persuade the hearer that he or she is in pain. But if the deliberate character of the production of the sign impairs or destroys the belief-inducing tendency of the sign qua natural sign, that tendency can be restored: The import of the recognition by Y that the production is voluntary undermines… any tendency on the part of Y to come to the conclusion that creature X is in pain. So, one might ask, what could be required to restore the situation: what could be added which would be an antidote, so to speak, to the dissolution on the part of Y of the idea that X is in pain? (Grice 1982/1989: 293)

What must be added, Grice says, is Y’s recognition of X’s intention to let Y know that he or she was simulating pain-behaviour in order to induce in Y the belief that X is in pain. Someone who intends to deceive by faking a natural sign must hide his or her intention to do so (Recanati 1979: 175–177). Think of the game of poker, where a large bid induces in the other players the belief that one has a good hand only if the player’s intention to induce that belief is not recognized; or think of the act of leaving the lights on to deceive the potential burglars into thinking that there is someone in the house: the burglars will only be deceived if they don’t recognize the intention to deceive.

2  Natural Meaning and the Foundations of Human …     21

When the deliberate character of the production of a sign is revealed, the sign no longer works as a natural sign, and deception fails. So, by openly revealing your intention to induce a belief in the audience, you show that your intention is not a deceptive one; for if it were, it could not be revealed openly without defeating itself. So here is, in simplified form, the Gricean recipe for restoring the belief-inducing tendency of the faked natural sign: if e.g. the scream is seen as emanating from the subject’s intention to convey that she is in pain, and if that intention itself is openly revealed (rather than hidden, as in cases of deception), then, if certain additional conditions are met, the communicative intention itself will be a reliable sign that the person is really in pain. In such cases of ‘non-deceptive simulation’, the connection between the pain and the scream which was severed by the recognition of the deliberate nature of the production is restored. The scream is no longer directly a natural sign of the pain, but indirectly it is. There now is a chain of signs: from the utterance the communicative intention can be inferred, and from the communicative intention the content (the existence of the pain) can be inferred. Grice’s theory here converges with Marty’s. I talked about the additional conditions which are to be met for the communicative intention to serve as a natural sign of its content. These conditions, which obtain in what I called the ‘basic cases’, are mentioned by both Marty and Grice. Grice: Whether or not in these circumstances Y will not merely recognize that X intends, in a rather queer way, to get Y to believe that X is in pain, whether Y not only recognizes this but actually goes on to believe that X is in pain, would presumably depend on a further set of conditions which can be summed up under the general heading that Y should regard X as trustworthy in one or another of perhaps a variety of ways. For example, suppose Y thinks that, either in general or at least in this type of case, X would not want to get Y to believe that X is in pain unless X really were in pain. Suppose also […] that Y also believes that X is trustworthy, not just in the sense of not being malignant, but also in the sense of being, as it were, in general responsible, for example, being the sort of creature who

22     F. Recanati

takes adequate trouble to make sure that what he is trying to get the other creature to believe is in fact that case, and who is not careless, negligent, or rash. Then […] one would regard it as rational not only for Y to recognize these intentions on the part of X, that Y should have certain beliefs about X’s being in pain, but also for Y actually to pass to adopting these beliefs. (Grice 1982/1989: 294–295)

Marty: If the expression of my own judgments, or what I pretend to be one, should be the medium to suggest to somebody else the same judgement, so must this person on the one hand trust in my truthfulness and be therefore convinced that I did not make a certain statement deceitfully or thoughtlessly (as if I would not understand myself ). Furthermore, however, we also have to presuppose that a being who knows the difference between correct and incorrect judgment is confident that my judgment behaviour is correct. Thus I have to appear to him as an authority and a guarantee of the truth of the judgement. (Marty 1908: 286–287, the translation follows partially Liedtke 1990: 44)

3 Back to Salome What about Salome? I said that there was a difference between Marty and Grice with respect to that type of example, and it is time to revisit the issue. For Grice, the Salome example is an instance of exploitation of natural meaning that does not add up to non-natural meaning, because Herod’s intention is not intended to be fulfilled via the recognition of that intention. I said that for Marty, that should count as a bona fide instance of communication, since the three dimensions of semiosis are exemplified: Herod’s ostensive gesture springs from his knowledge that St. John the Baptist is dead and his desire to share that knowledge with Salome, so (through the indirect route) the gesture means that St. John the Baptist is dead. That makes it a bona fide instance of communication in Marty’s sense. At the same time, it involves a natural sign (the severed head) which signifies St. John’s death through a more direct

2  Natural Meaning and the Foundations of Human …     23

route, independently of Salome’s recognition of Herod’s intentions. That is what prevents it from being an instance of non-natural meaning in Grice’s sense. But shouldn’t that feature prevent it also from being an instance of communication in Marty’s sense? After all, Marty repeatedly says that the communicator’s primary intention (the intention to instil in the audience a certain psychological state) is fulfilled mediately, via a secondary intention which is fulfilled immediately: the speaker’s intention to disclose his or her psychological state, expressed by the utterance. But in the Salome example, as Grice points out, the communicative intention is fulfilled more directly: the sight of St. John the Baptist’s head is sufficient to induce in Salome the belief that he is dead, without the recognition of Herod’s own psychological states playing any significant role. Even though Martyan communicative intentions are not reflexive, contrary to Gricean communicative intentions, still they share the feature that they are intended to be fulfilled mediately, via the audience’s recognition of the communicator’s psychological states. That feature is missing in the Salome case, so Grice and Marty should both rule out that example. What I have just said points to a tension within Marty’s account as I have presented it so far. Marty, I suggested, holds that cases of intentional production of natural signs can be instances of bona fide communication exemplifying the three dimensions of semiosis, while simultaneously claiming that the communicator’s primary intention is to be fulfilled mediately rather than immediately. Now when the sign intentionally produced is a natural sign, it indicates its object (or rather, the existence of its object) in such a way that the audience is led to judge that the object exists, independently of their recognition of the subject’s own judgment to that effect. The inference from the subject’s expressed state to the existence of its object is available to the audience, but a more direct route to the same conclusion is available too: from the natural sign to the existence of its object. So a charitable reading of Marty imposes that we qualify his claim that the communicator’s primary intention must be fulfilled mediately. We should construe him as holding the following, weaker view: in communication the audience must be able to ground her judgement that the object exists (or that the object is such and such) in the recognition of the communicator’s own

24     F. Recanati

judgment, expressed by her words in the linguistic case; but that does not mean that the audience’s judgment cannot also be grounded more directly through the natural connection between the sign and its object. It would be inconsistent for Marty to insist that in all cases of human communication, the communicator’s primary intention is to be fulfilled mediately. Mediate fulfilment must always be possible, in the sense that the expressed psychological state itself is a natural sign of its object, but the utterance may well involve the production of a more direct sign, as in the Salome example.9 In the case of Grice there is no such pressure to weaken his position for the sake of consistency. Grice insists that a natural sign that is deliberately produced (and recognized as such) ceases to be a natural sign and no longer indicates its object. Its meaning can only be restored via the indirect route, by construing the intention of which the deliberate production of the sign is a sign as itself a sign of its object. For Grice, the direct route and the indirect route necessarily compete. So the sharp contrast between natural meaning and non-natural meaning can be maintained, notwithstanding the prior conclusion that nonnatural meaning is itself grounded in natural meaning. In the Gricean framework, either the sign is a natural sign of its object directly, that is, independently of the psychological states which its production expresses, or it only becomes a sign of its object indirectly, because of the psychological states its production expresses. There is non-natural meaning only in the latter case: when the only way for the communicator to fulfil her primary intention is by letting the audience recognize her expressed psychological state (viz. her communicative intention). On this view the Salome example is not an instance of non-natural meaning. Now which view is preferable? I think Marty’s reconstructed position is preferable. A first, prima facie reason for accepting the Salome example as a bona fide instance of communication is that Grice’s own tests for non-natural meaning yield a positive verdict. By his gesture, Herod (the person) means that St. John the Baptist is dead; and his gesture can be reported as meaning ‘St. John the Baptist is dead’. That is the case even though the severed head ‘means that St. John the Baptist is dead’ also in the natural, factive sense. It seems that in this example there is both natural meaning (independent of the recognition of

2  Natural Meaning and the Foundations of Human …     25

the subject’s psychological state) and non-natural meaning in the sense of the informal tests. Still, Herod cannot be said to intend Salome to come to believe that St. John the Baptist is dead via the recognition of Herod’s intention to that effect. Grice’s reflexive analysis seems to be too demanding and to fit only a sub-class of cases of non-natural meaning. A second reason to prefer Marty’s reconstructed position over Grice’s is that many theorists in the Gricean tradition have come to the conclusion that it was a mistake on Grice’s part to rule out cases like the Salome example.10 I explicitly said so in my article ‘On Defining Communicative Intentions’ (1986) and in my book Meaning and Force (1987) which incorporates the same material. There I gave an example of linguistic communication involving indexicals which has the same structure and properties as the Salome example, that is, which qualifies as a mixture of natural and non-natural meaning: My friend and I are walking in a crowded place, and she loses track of me, although I am not at all far away. I tell her: ‘I am here’. I thereby communicate to her that I am here, but to her, my utterance is also a natural sign of the fact that I am here. Even if I had said something different to someone else — with no recognizable intention to communicate to her where I was — the sheer sound of my voice in the vicinity would have indicated my location to her, exactly in the same way as John the Baptist’s head indicates to Salome that he is dead. Moreover, the ‘natural’ meaning of my utterance is not just a matter of fact: I intend my utterance to provide the hearer with some evidence independent of the evidence provided by my recognized intention. This case, it seems to me, is exactly parallel to the Salome case. So if I have performed an act of Gricean communication in the linguistic case, there is no reason to deny that Herod may very well have done so by showing Salome the severed head. (Recanati 1987: 190)

In a footnote, I added: The similarity goes further than I have indicated. To recognize S’s communicative intention, it is necessary, in the case of Herod, to first recognize the natural meaning of the sign (Salome recognizes what Herod means by first recognizing what the severed head of John the Baptist ‘naturally’ means). Similarly, in the linguistic case, the hearer does not know

26     F. Recanati

what is said, and so what the speaker intends to (linguistically) communicate, if she does not grasp the ‘natural meaning’ of the utterance: She does not know who is said to be where if she does not infer, from the familiar voice she is hearing and from its proximity, that I am speaking to her and that I am here. (ibid.: fn.)

Note that, according to Karl Bühler, this mixture of natural and nonnatural meaning is a characteristic property of indexical communication.11 Are we not losing the advantages of Grice’s theory if, to accommodate cases like the Salome example, we give up his central claim pertaining to the reflexivity of communicative intentions? I said earlier that Grice’s reflexive analysis was his major contribution, his great insight, something that is missing from Marty’s account; am I not now saying that we should give up that feature and opt for Marty’s reconstructed account instead of Grice’s own account? No, that’s not, or not exactly, what I am saying. Grice’s insight about the nested/reflexive structure of the intentions underlying human communication is of considerable value, and it should evidently be preserved. But it can be cashed out in many ways. Some of the ways lead to problems and paradoxes, emphasized in the huge literature which Grice’s article ‘Meaning’ gave rise to. In particular, it is no good to ascribe to communicators an infinite sequence of intentions, and it is no good to rule out cases of communication involving the exploitation of natural meaning (as in the Salome example or the ‘I am here’ example). But there are versions of the Gricean story which make it possible to bypass such unwelcome consequences. As many have argued, what we need is a notion of ‘overtness’ through which we can characterize human communication (Strawson 1964). The communicator has audience-directed intentions which she expresses ‘overtly’ by her utterance. Overtness is where the action is, and to spell out what it requires we need something like Grice’s analysis of reflexive intentions, or some equivalent analysis in terms of mutual knowledge.12 But the reflexive intentions (or the mutual knowledge) may be construed as an ‘ideal’ rather than something that actually obtains (Grice 1982, Section 3). One way of doing so is by appealing

2  Natural Meaning and the Foundations of Human …     27

to the notion I introduced in the works referred to above: that of ‘default-reflexivity’. S’s intention is default-reflexive if and only if S has no intention inconsistent with any of the (infinite number of ) intentions that his intention would entail if it were genuinely reflexive. (Recanati 1987: 201)13

The speaker’s communicative intentions are overt, that’s what distinguishes them from the sort of intentions involved in deceptive behaviour; but that only means that these intentions are default-reflexive. We use the notion of a reflexive intention to specify the property of default-reflexivity which communicative intentions actually possess, but the requirement that communicative intentions be overt in the sense of default-reflexive is much less demanding than the requirement that they be reflexive in the full-blooded sense. The important point, as far as this paper is concerned, is that Herod’s communicative intention is ‘overt’; so the case qualifies as a bona fide instance of Gricean communication, despite the involvement of natural meaning.14

Notes 1. The dark clouds cannot be reported as meaning ‘It will rain’, Grice points out. They don’t mean anything in the non-natural sense (but only in the natural sense). 2. Reflexivity already seems to be a feature of Grice’s analysis of communicative intentions in ‘Meaning’ (Grice 1957). There he says that ‘A non-naturally meant something by x ’ is roughly equivalent to ‘A uttered x with the intention of inducing a belief by means of the recognition of this intention ’. Whether reflexivity was actually intended is not totally clear, however: Grice did not explicitly address the issue, except in a brief and puzzling remark (‘This seems to involve a reflexive paradox, but it does not really do so’, Grice 1957/1989: 219). Grice (1969/1989: 97) discusses the threat of an ‘infinitely or indefinitely regressive’ analysis explicitly, but in that paper, instead of advocating a ‘virtuous’ regress in the form of an overtly reflexive analysis, as some of his disciples did, he opts for something along the lines of what I call

28     F. Recanati

‘default-reflexivity’ (see the last paragraph of this chapter and the reference therein). 3. Marty speaks of announcement (Kundgabe ). Cesalli translates Kundgabe by ‘indication’ (Dretske’s term for the natural meaning relation). On the relation between Kundgabe and Ausdruck and terminological variations among Marty’s contemporaries, see Linhaeres-Dias (2006: 127). 4. Of course, Grice was aware that humans often communicate by using natural signs (as in the Salome example I am going to discuss). But what is distinctive of human communication, for him, is the mechanism of non-natural meaning. When I say that, for Grice, nothing can be both an instance of natural meaning and an instance of human communication, I take ‘human communication’ in that distinctive sense: communication involving non-natural meaning. 5. Marty focuses on linguistic communication and rarely mentions instances of nonverbal communication of the sort Grice discusses. Still, on the basis of what Marty says of linguistic communication, we can extrapolate and ask what he would have thought of e.g. the Salome example. 6. See Brentano (1874/1971: II, 62n). 7. Insofar as I can tell, the distinction between these three kinds of meaning appears in Marty’s third paper on subjectless sentences (Marty 1884: 300 sq). It can also be found in Twardowski, who applies it to the semantics of names (Twardowski 1894/1894: 9–10). Van der Schaar suggests that the distinction was common among the students of Brentano (van der Schaar 2013: 65). On Marty’s originality in relation to Brentano, see Cesalli (2013: 155–160, esp. p. 158). 8. What I call denotational meaning Marty calls ‘meaning in a restricted sense’ (Bedeutung im engeren Sinne ), in contrast to communicative meaning which is ‘meaning in a wider sense’ (Bedeutung im weiteren Sinne ). Rollinger speaks of ‘meaning in the ontological sense’ in contrast to ‘meaning in the communicative sense’ (Rollinger 2010: 84). 9. That this interpretation may correspond to what Marty actually had in mind is supported by a couple of passages in which he says that the communicative intention is fulfilled mediately ‘in most cases’ (Marty 1908: 284) or ‘normally’ (Marty 1908: 286). 10. See Schiffer (1972: 57–58), Recanati (1986, 1987: 189–190), Sperber and Wilson (1986: 46–54), Neale (1992, Section 15), Davis (2003: 71), Green (2007: 59), Wharton (2009: 31–33), and Sperber and Wilson (2015).

2  Natural Meaning and the Foundations of Human …     29

11. See, e.g., Bühler (1932/1990: 129) for an example similar to my ‘I am here’. Note that so-called indexicals owe their name to the fact that, while being Peircean symbols, they are also Peircian indices since they carry natural meaning and bear an ‘existential relation’ to their referent (they are ‘indexical symbols’; see Recanati [1987: 6–7]. On Peirce’s classification of signs, see also Burks [1949]). 12. The notion of mutual knowledge (a.k.a. common knowledge) was simultaneously introduced into the philosophical literature by David Lewis in his analysis of conventions and by Schiffer in his discussion of Grice (Lewis 1969; Schiffer 1972). 13. As I pointed out in footnote 2, something like this idea is already present in Grice (1969). 14. The aim of the conference for which I prepared this paper was to assess the contemporary relevance of Marty’s philosophy of language. The organizers asked me to participate not as a Marty expert, but as a philosopher of language, with a genuine interest in some of the ideas which Marty put forward. I am grateful to Giuliano Bacigalupo, coorganizer of the conference, and Kevin Mulligan and Laurent Cesalli, directors of the SNF project “Meaning and Intentionality in Anton Marty” which the conference concluded, for letting me see (the relevant parts of ) the French translation of Marty’s main work, Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie (a translation that was in preparation as part of the project); to Hélène Leblanc, the other co-organizer of the conference, who responded to my talk and, together with Giuliano and two reviewers, provided comments on the manuscript; and to the French Agence Nationale de la Recherche for indirect funding under grant agreement ANR10-LABX-0087 IEC and grant agreement ANR-10-IDEX-0001-02 PSL.

References Bach, K. 1987. On Communicative Intentions: A Reply to Recanati. Mind & Language 2: 141–154. Bach, K., and R.M. Harnish. 1979. Linguistic Communication and Speech Acts. Cambridge, MA: MIT Press. Brentano, F. 1874. Psychologie vom empirischen Standpunkt. Leipzig: Duncker & Humblot. New edition, Hamburg: Felix Meiner, 1971.

30     F. Recanati

Bühler, K. 1932. Sprachtheorie. Jena: Gustav Fischer Verlag. English trans. by D.F. Goodwin, Theory of Language. Amsterdam: John Benjamins, 1990. Burks, A. 1949. Icon, Index, and Symbol. Philosophy and Phenomenological Research 9: 673–689. Cesalli, L. 2013. Anton Marty’s Intentionalist Theory of Meaning. In Themes from Brentano, ed. D. Fisette and G. Fréchette, 139–163. Amsterdam: Rodopi. Davis, W. 2003. Meaning, Expression, and Thought. Cambridge: Cambridge University Press. Dretske, F. 1988. Explaining Behavior. Cambridge, MA: MIT Press/Bradford Books. Green, M. 2007. Self-Expression. Oxford: Clarendon Press. Grice, P. 1957. Meaning. Philosophical Review 66: 377–388. Reprinted with modifications as chapter 14 of Grice (1989). ———. 1969. Utterer’s Meaning and Intentions. Philosophical Review, 78: 147–177. Reprinted with modifications as chapter 5 of Grice (1989). ———. 1982. Meaning Revisited. In Mutual Knowledge, ed. N. Smith, 223– 243. London: Academic Press. Reprinted with modifications as chapter 18 of Grice (1989). ———. 1989. Studies in the Way of Words. Cambridge, MA: Harvard University Press. Lewis, D. 1969. Convention. Cambridge, MA: Harvard University Press. Liedtke, F. 1990. Meaning and Expression: Marty and Grice on Intentional Semantics. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 29–49. Dordrecht: Kluwer. Linhaeres-Dias, R. 2006. How to Show Things with Words. Berlin: de Gruyter. Longworth, G. 2017. Grice and Marty on Expression. In Mind and Language—On the Philosophy of Anton Marty, ed. H. Taieb and G. Fréchette, 263–284. Berlin: De Gruyter. Marty, A. 1884. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, III. Vierteljahrsschrift für wissenschaftliche Philosophie 8: 292–340. ———. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, Bd. 1. Halle: Niemeyer. Neale, S. 1992. Paul Grice and the Philosophy of Language. Linguistics and Philosophy 15: 509–559. Recanati, F. 1979. La Transparence et l’Enonciation. Paris: Seuil.

2  Natural Meaning and the Foundations of Human …     31

———. 1986. On Defining Communicative Intentions. Mind and Language, 1: 213–242. ———. 1987. Meaning and Force. Cambridge: Cambridge University Press. Rollinger, R. 2010. Philosophy of Language and Other Matters in the Work of Anton Marty. Amsterdam: Rodopi. Schiffer, S. 1972. Meaning. Oxford: Clarendon Press. Searle, J. 1969. Speech Acts. Cambridge: Cambridge University Press. Sperber, D., and D. Wilson. 1986. Relevance. Oxford: Blackwell. ———. 2015. Beyond Speaker’s Meaning. Croatian Journal of Philosophy, 44: 117–169. Strawson, P. 1964. Intentions and Conventions in Speech Acts. Reprinted in Strawson (1971: 149–169). ———. 1971. Logico-Linguistic Papers. London: Methuen. Twardowski, K. 1894. Zur Lehre vom Inhalt und Gegestand der Vorstellungen. Vienna: Hölder. English trans. by R. Grossmann, On the Content and Object of Presentations. The Hague: Martinus Nijhoff, 1977. van der Schaar, M. 2013. G.F. Stout and the Psychological Origins of Analytic Philosophy. Houndmills, Basingstoke: Palgrave Macmillan. Wharton, T. 2009. The Pragmatics of Non-Verbal Communication. Cambridge: Cambridge University Press.

3 How a Statement Has Meaning by Expressing a Judgement—Brentano Versus Marty on Utterance Meaning Mark Textor

1 Introduction Some of the breeding and food channels of the bark beetle’s larva look like words.1 What distinguishes them from a word that I carve into the bark of a tree? The philosopher Anton Marty (1847–1917) gave a prima facie plausible answer: I carve the inscription into the bark in order to indicate a mental state or act, the larva doesn’t; it just wants to have a way to transport food. Because of this difference, my inscription is a meaningful sign, while the physically indistinguishable ‘inscription’ produced by the larva is meaningless. For Marty, this example tells us something important about language. The word ‘language’ as it occurs in ‘philosophy of language’ or ‘the science of language’ is synonymous with ‘the intentional indication [absichtliche Kundgabe] of inner life by means of any signs that – as most words of our phonetic languages do – owe

M. Textor (*)  King’s College London, London, UK e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_3

33

34     M. Textor

their significative power to the habit and tradition’ (Marty 1908: 3; my translation). Many will dispute that language is defined the way Marty proposed. But we can treat his definition as a proposal about the nature of language to be confirmed by further argument. Marty’s proposal is a particular form of what is now sometimes called ‘intentionalist semantics’ and mainly associated with the work of Paul Grice.2 Intentionalist ‘semantics’ is a view of the metaphysics of meaning: the meaning of a sign is grounded in or supervenes on facts about people meaning something by the sign (see Grice 1989: 340). Two commitments of intentionalist semantics are worth highlighting: First, mental facts about desire and belief are more fundamental than facts about what signs mean. If they are, the success conditions of desires and the truth-conditions of beliefs are determined independently of language. Second, a speaker only means something by something she does if, and only if, she does it for a purpose, namely to produce an attitude in an audience. According to Marty, the intention that gives meaning to utterances is the intention to indicate [kundgeben] one’s own presentations, judgements, emotions etc. in order to trigger presentations, judgements and emotions in another psychical being, and indeed, ones which are analogous to one’s own. (Marty 1908: 22, translated in Cesalli 2013: 144)

Similarly, Grice held that linguistic meaning is ultimately grounded in behaviour that is done with the purpose of changing the mind of others (see Bennett 1976: 15). A speaker means something by something she does if, and only if, the speaker intends (1) that an audience responds in a particular way because (2) the audience recognizes that the speaker intends (1) and fulfils (1) in virtue of fulfilling (2). Both (1) and (2) have been controversially discussed in the literature, and Grice himself addressed a number of counterexamples.3 Grice refines, but doesn’t change the basic ideas that (a) speaker meaning is the starting point of a theory of meaning and (b) a speaker means something by doing something only if she intends to produce an effect in an audience.

3  How a Statement Has Meaning by Expressing …     35

In this paper, I will focus on the second commitment of intentionalist semantics: meaning facts supervene on facts about someone doing something with a communicative intention. My aim is to explore an alternative view of meaning according to which not speaker, but utterance meaning is the basic notion of a theory of meaning. The alternative is suggested in the work of Marty’s philosophical teacher Franz Brentano (1838–1917). In his lectures on logic, Brentano took some utterances to have meaning in the relevant sense of ‘meaning’ independently of whether they are made in order to influence the thought of others. Brentano therefore prioritized utterance meaning over speaker meaning: the primary source of meaning is not a speaker meaning something by doing something, but what she does has meaning. Brentano’s proposal constitutes a welcome alternative to meaning intentionalism, and I will argue that it solves a number of problems that plague the intentionalist view of Marty and Grice. Hence, while Brentano seems to endorse later elements of meaning intentionalism in unpublished work, he shouldn’t have.4 I will start with some scene setting (Sect. 2) and then argue that non-communicative utterances pose a problem for the meaning intentionalism of Marty and Grice (Sect. 3). I will use the problem to expound Brentano’s theory of meaning and argue that it has the potential to solve the problem of non-communicative utterances (Sect. 4). However, the Brentano’s view faces a different problem (Sect. 5). The remainder of the paper (Sects. 6–11) is devoted to answering the problem and thereby to show that Brentano’s original idea can be defended and developed to yield an insight into speaker meaning.

2 The Right Kind of Meaning Let’s start by getting the notion of meaning Grice et al. are interested in into view. Grice et al. are not trying to understand what is called ‘natural meaning’. Spots naturally mean measles. Why? Because spots of this kind are indicators or symptoms of measles. I will take a basic understanding of natural meaning for granted and focus on non-natural meaning.

36     M. Textor

So what is non-natural meaning? Grice gives his reader two pointers to fix on non-natural meaning: non-natural meaning is non-factive and propositional. We are dealing with non-natural meaning if what the action/event meant can be rendered by a ‘that’ clause and the action/ event can mean that p even if it is not the case that p. These two features make one aware of a distinction between natural and non-natural meaning, and they help us to get an initial grip of the distinction. But for our purposes this is not enough. We need to find the rationale which grounds the distinction. In order to learn more about the distinction, let’s have another look at Grice’s example. Take his (1989: 218) photograph case. If A shows Mr. X a photograph of B eating ice cream in front of Trump Tower with the intention that Mr. X comes to the belief that B ate ice cream in front of Trump Tower, A ’s showing the photograph does only naturally mean that A ate ice cream in front of Trump Tower. In contrast, Grice claims that if A draws for Mr. X a picture of B eating ice cream in front of Trump Tower with the intention that Mr. X comes to the belief that B ate ice cream in front of Trump Tower does mean non-naturally that B ate ice cream in front of Trump Tower.5 Wherein does the distinction between the two cases consist? In the photograph case, merely seeing the photograph and ‘taking it in’ gives Mr. X a reason to form the belief that B ate ice cream in front of Trump Tower, whether A intended Mr. X to acquire the belief or not.6 This is different in the second case. Mr. X needs to recognize the intention with which A draws the picture for him to have a reason to believe that B ate ice cream in front of Trump Tower. According to Grice, such communicative intention-dependent evidence distinguishes natural from non-natural meaning: If the audience could not be expected to arrive at the intended belief apart from the recognition of the speaker’s intention regarding that belief, the speaker must take upon himself the role of providing something with a particular epistemic import that it otherwise would not have, and in this way Grice sharply distinguishes non-natural meaning from the presentation of evidence. (Moran 2005: 14)

3  How a Statement Has Meaning by Expressing …     37

According to Grice and his followers, by doing x the speaker nonnaturally meant that p if, and only if, for some audience A (i) the speaker did x intending that A acquires the belief that p; (ii) that A recognizes the speaker’s intention (i); (iii) that A ’s recognition of intention (i) be part of A ’s reason to acquire the belief. If the speaker does something, but she cannot expect that the audience acquires the belief that p merely in virtue of recognizing her communicative intention, what the speaker does has at best natural meaning. If Herod hands Salome the head of John the Baptist, he cannot expect that she will arrive at the belief that John the Baptist is dead by recognizing his intention in presenting her with the head. Herod cannot have a Gricean communicative intention. For the fact that John the Baptist is dead is so obvious that Herod cannot expect that his audience acquires the belief merely in virtue of recognizing his intention. Hence, he cannot form the Gricean intention and cannot non-naturally mean that John the Baptist is dead by presenting his head. In sum, the rationale of the non-natural/natural meaning distinction is the necessity that we often must be able to ‘vouch’ for the correctness of something for which we cannot provide rational grounds independent of our activities. Grice spells out this basic idea in a particular way: recognizing that the speaker intends to instil a belief in me can make it rational to acquire this very belief. Hence, communicative intentions become central to non-natural meaning. In the next sections, we will see that this proposal to distinguish between natural and non-natural meaning is too restrictive.

3 Non-natural Meaning Without Communicative Intentions It seems that we make utterances and mean something by them even if there is no audience at all or we don’t want to change the beliefs and desires of an audience by our utterance. Bennett relates the following counterexample proposed by Bricke:

38     M. Textor

I sit in my study replaying in my mind a recent conversation with a colleague; it suddenly dawns on me that I have said something stupid or harmful, and that the damage cannot be undone; and I make a slashing movement across my left wrist with my right forefinger, meaning by this something like ‘I’m so angry with myself that I could kill myself ’. The case is not offered as one in which that is literally true, just as one in which it is meant; similarly, one might utter those words meaning something by them yet not speaking truly. (Bennett 1991: 7)

Intuitively, the slashing movement has non-natural meaning, but the agent has no communicative intention to influence the attitudes of an audience by making the movement. The speaker is not communicating to himself: he does not make the utterance with the intention to bring it about that he himself acquires a belief about his emotional state. He does not need to give himself a reason to come to believe that he is angry and upset.7 He already knows this. In Grice’s sense, the speaker does not mean anything by making the slashing movement. Since the movement has no conventional meaning independently of communicative intentions, Grice cannot appeal to an already established meaning that gives the utterance its non-natural meaning. Bennett himself treats this and similar cases as ‘pockets of resistance’, but does not offer a description of these cases that makes them compatible with Grice’s basic thought. Cases like the one above suggest that there can be non-natural meaning without communicative intentions and conventional meaning. But how? Obviously the idea that the slashing movement is the expression of an attitude or emotion must play a central role in an answer. But on the face of it an expression of an attitude or emotion is just that and has no further non-natural meaning. How can there be an expression of an attitude that has also nonnatural meaning although it is not made with an intention to influence the attitudes of an audience? In the next sections, I will draw on Franz Brentano’s work in order to answer this question. Brentano’s main interest is psychology. But his published and unpublished work contains ideas about the expression of mental acts in language that speak to our problem.

3  How a Statement Has Meaning by Expressing …     39

4 Brentano and Marty on Judgement and Utterance Meaning In his unpublished logic lectures (EL 80), Brentano outlined a partial answer to our question. The answer is partial as it deals only with the meaning of utterances that express judgements. But it will yield a model for the connection of judgement and non-natural meaning in general. Brentano writes under the label ‘The meaning of statements’: All our linguistic expressions that mean something are expressions of something that happens in our soul, but mean something different. Because we don’t talk to each other of mental processes in us apart from particular cases when our discourse is directed upon them, but of objects, numbers, spatial magnitudes and all kinds of phenomena of external nature. (EL 80: 13.126, my translation)

Let’s initially set aside the question whether all linguistic expressions that mean something are really expressions of mental events and focus on the case that is of interest to Brentano: statements. He writes: That in our soul whereof the statements [Aussage ] are the expressions are our judgements (because of this one frequently hears people saying that a statement is a judgement expressed in words). If someone makes a statement, this is a sign for the fact that he has a particular judgement. (EL 80: 13.127, my translation)

In what follows, I will focus like Brentano on the meaning of statements, assertoric utterances, and set other kinds of speech acts aside. ‘Statement’ has the notorious product/process ambiguity: ‘John’s statement’ can refer to an act of stating or its product, the statement John made. We may say ‘The statement that John made was true/interesting/that blood is red’. If we do so, we take the statement to refer to a propositional content. This propositional content cannot be the sign of John’s judging. Hence, when Brentano writes about statements, he has the process sense of ‘statement’ in mind. We are concerned with

40     M. Textor

a particular event, a stating that happens at a time and that stands in causal or other relations to a mental event that can cause and justify us to assume the existence of a judging with the same content. Because of these relations, it can be a sign of a particular judging. Just as the spots in my face are a natural sign of measles because they are caused by the processes that are responsible for the illness, my stating is a natural sign for one of my judgings because stating and judging stand in causal or similar relations. The former are natural signs or symptoms of the latter. The crucial question for our purposes is why a stating should be more than a natural sign for a judging as Brentano claims, namely why the stating should also have non-natural meaning. My making a statement is supposed to be a natural sign of my judging and it is supposed to mean that p, where ‘that p’ is a mind-independent state of affairs. How can the stating have such meaning? It is worth quoting Brentano’s answer in full: How do we find the answer to our question what statements mean [bedeuten ]? Let us look at what can be taken as understood, namely that the statement is the expression of our judgement. The affirmation [Bejahung ] is the expression of our acknowledgement, the denial is the expression of our rejection of an object or indicates it. What does someone do who makes a judgement, who acknowledges and rejects? Obviously he treats what he judges as something that is to be judged as he does judge it. If he acknowledges it, he treats it as something that is to be judged, if he rejects it, he treats it as something to be rejected. If we therefore assume that one would be able to immediately perceive the judgement of another person in whose insight one trusted, this judgement would indicate how its object is to be judged, whether it is to be acknowledged or rejected. In the same way the judgement will do this if it is, instead of being immediately recognised, mediately recognised through its linguistic expression. (EL 80: 13.130–13.131; my translation and emphasis)

Brentano has a specific view of judgement: judging is either acknowledging or rejecting an object, not a propositional content. Brentano’s contribution to the discussion about non-natural meaning is independent of this specific aspect of his view of judgement. Hence, I will not make use of it in the following discussion.

3  How a Statement Has Meaning by Expressing …     41

When I acknowledge an object, I have a sense that this is the appropriate or right thing to do. In order to have this sense, I don’t need to judge that my acknowledgement is right. My sense that my judgement is correct is manifest in my emotional responses: if you convince me that the object acknowledged does not exist, I will respond with the same emotions as I respond to doing something incorrectly. I may be angry or blame myself, etc. I can judge that p without judging that it is right to judge that p. But when the question arises, I cannot judge that p and deny that it is right to judge that p. Judging that p commits me in this sense to take that p to be something that is to be judged. Now let us assume for the sake of the argument that one could simply perceive B ’s judgement that p. If we ‘trust B ’s judgement’, our perception of B ’s judgement is awareness of B ’s inner life (his judging) and it has ‘epistemic import’ of the kind discussed in Sect. 2. If A knows what judgements are, A can infer from the fact that B judges that p, together with the knowledge of how they themselves ‘treat’ things they judge and their trust in B ’s insight, that the content that p is to be judged. We know of the thinker’s judgement and thereby acquire knowledge of how the world is. Even if judgement is a spontaneous response and not made for any purpose, it can make it rational to judge likewise. If Brentano is right, the mere fact that someone judges that p will, if the judger is trustworthy, make it rational to judge that p independently of any communicative intention. This idea will be the basis for an account of assertoric utterance meaning. We have now the outline of a Brentanian alternative to Grice’s mechanism of belief acquisition. One reason that can make it rational to make a judgement and thereby to acquire a belief is the fact that someone makes a judgement together with the background assumption that this person is trustworthy. Marty is aware of Brentano’s suggestion for a mechanism of belief acquisition and it is informative to compare and contrast Brentano’s and Marty’s own take on it8: Someone who states “A is” treats, in case he himself judges that way, A as existing and imposes upon [zumutet] his hearer that he, on the basis of trusting the behavior shown by him (the speaker), also treats A as existing. With reference to this one also says that the statement announces

42     M. Textor

[kundgeben ] the being of A or wants to announce it, that is, supposes to announce it and means it in this sense. And since one frequently calls the being of A, or that A is, or the being B of A or, that A is B, the content of the judgement “A is” or “A is B” and, in turn, the nonbeing of A and the not being B of A as the content of the judgement “A is not” or “A is not B” one can therefore also say: the statement announces the content of the judgement and means it in this sense. (Marty 1908: 291–292)9

This passage is rather long and convoluted. Let us take it step by step. First step. Marty starts out by presenting a version of the Brentanian mechanism for belief acquisition. A speaker who voices a judgement by making a statement is committed to the correctness of this judgement and imposes (zumuten) on the audience that they judge like the speaker did. However, Marty’s talk of ‘imposition’ seems misplaced. He refers to his (1884: 70) as an anticipation of his idea. There the speaker’s assertoric utterance is described, among other things, as a call (Aufforderung ) to make a judgement. I prefer Marty’s earlier terminology for the reason just given. Marty added in a footnote that, as a rule, the speaker stating that p judges himself that p. If not—say in case of a lying assertion— the speaker at least makes himself appear as someone who judges that p. By presenting herself as someone who judges that p, the speaker calls upon the audience also to judge that p. Second step. Marty introduces a notion of meaning: meaning in the narrow sense. If we set Marty’s terminology and his different ways of referring to judgement contents aside, we can say that one can learn from a statement a mental fact about the speaker and a non-mental fact about how things are. The latter is what makes making statements valuable and deserves therefore the label ‘meaning’. Now, there seems to be an important difference between Marty and Brentano. Marty adds to Brentano’s initial description of the mechanism for belief acquisition an audience directed invitation. It is not only right to judge like the judger in a particular matter; by manifesting her judgement, the judger asks the audience to judge like herself. But this difference seems to me merely verbal. There is not a further mental act of the speaker. The speaker does not issue an invitation or something like it. The mere fact that the speaker manifests a judgement makes it in the right circumstances reasonable for the audience to judge

3  How a Statement Has Meaning by Expressing …     43

like the speaker. It is the judgement that reveals what is right to judge, and if the audience wants to what is right, they ought to judge like the speaker. Hence, while there may be audience directed attitudes, they are not required for utterance meaning (in the narrow sense). Marty as well as Brentano captures the idea that non-natural meaning requires rational belief formation dependent on a special source of justification: the authority of the speaker. Marty is especially clear on this point: If the utterance or the seeming utterance of my own judgement shall be a means to suggest to someone else a similar judgement he must, on the one hand, trust in my truthfulness and because of this trust be convinced that I am not making a certain statement is thoughtlessly (as if I don’t understand myself ) or mendaciously. In addition we take for granted that a being that knows the distinction between right and wrong judging will be confident that my judging behavior is right. I have to be for him the authority and guarantee for the truth of what is judged. (Marty 1908: 286–287; my translation and emphasis)

Meaning in the narrow sense does not require the speaker to have a communicative intention. What justifies the audience to judge that 7 + 5 = 12 is that the speaker judged that this was so, made her judgement manifest and the audience trusts her to get things right, not that the speaker intended you ‘the audience’ to acquire the belief that 7 + 5 = 12 by recognizing this intention. The manifestation of the judgement need not be a communicative act. To sum up, in coming to know how the thinker judged you not only get knowledge about her mind, but also about how things are. In this sense, knowledge of a judgement is not only knowledge of a mental event but it enables one to come to a view of how things are: the judgement means that things are a certain way. If we could say that assertoric utterances are manifestations of judgements, we could see their meaning as derived from the meaning of the judgement they make manifest without appealing to communicative intentions or audience directed expectations. Unfortunately, as we will see in due course, things are not that straightforward. But let’s first return to the characterization of the distinction between natural/non-natural meaning.

44     M. Textor

5 Reconsidering the Distinction Between Natural and Non-natural Meaning In Sect. 2, the natural/non-natural meaning distinction was based on the feature that something the speaker does has ‘epistemic import’ for (rationally) arriving at the judgement (belief ) that p although there are no reasons available independently of the speaker’s acts and attitudes. Grice gives a specific version of this general idea: if, and only if, the epistemic import of something a speaker does is due to the recognition of a communicative intention, the doing has non-natural meaning. Brentano outlined a view according to which judgements are a source of evidence for the acquisition of a belief, but he denies that this evidence is intention-dependent. The ‘epistemic import’ of a judgement is due to the judgement that p itself. In judging that p, the speaker commits herself to the correctness of her judging. If we have no reasons to distrust the speaker and have experienced what it is to make a judgement, this makes it rational to judge that p ourselves. In short, the work that Grice takes communicative intentions to do is done according to Brentano by the indicated attitude and background conditions. The fact that I judge or judged that p together with the trust you place in me gives you a reason to judge like me. In contrast, the speaker’s judgement plays no evidential role in cases of natural meaning. Consider Grice’s Herod/Salome example. John the Baptist severed head is a natural sign of the fact that he is dead. Neither the fact that Herod judges that John the Baptist is dead nor that he intends to impart this belief to Salome is operative as a reason for the acquisition of this belief. The last observation suggests that the natural/non-natural meaning distinction is not the distinction between events whose occurrence indicates the occurrence of a distinct event or the obtaining of a fact— events that have natural meaning—on the one hand and events that are actions done with a Gricean communicative intention that is a reason to acquire a belief (desire), events with non-natural meaning. An event can neither be an indicator of another event or fact nor something that is done with a communicative intention, and yet, its occurrence can be a reason to acquire a belief. The last possibility opens up the logical

3  How a Statement Has Meaning by Expressing …     45

space for an event to non-naturally mean something although it is not brought about for communicative purposes. According to Brentano, the mental act of judgement occupies this space. If I judge that p, this mental act commits me to its correctness and thereby justifies someone who knows about the judgement and trusts the judger’s faculties to judge likewise. Since the ‘thinker’ does not need to do something in order to instil a belief in an audience, it is not the speaker that means something by what they do, but the basic notion is that of a mental act or, more generally, an event having non-natural meaning. Grice did not engage with the alternative provided by Brentano, I think, because he ignored attitudes that have a correctness commitment. In ‘Meaning Revisited’, Grice points out that some bearers of natural meaning are forms of behaviour: Things like groans, screeches and so on, which mean, or normally mean, that someone or something is in pain or in some other state. Thus special cases of natural meaning are cases in which bits of things like bodily behavior mean the presence of various elements or state of the creatures that produces them. (Grice 1989: 292)

My groan can be a natural sign of pain, but your coming to know that I am in pain does not make any further judgement rational. It does not have the same rationalizing role as coming to know or believe that I judge that p. Grice only considers natural signs of mental acts and states that lack the rationalizing power judgement has. Therefore, he cannot see Brentano’s point and feels compelled to add the intention that the behaviour is recognized as voluntary (see ibid.). In short, Grice ignores the attitude that is Brentano’s building block for the theory of meaning: judgement.

6 The Spoiler How do we get from a judgement meaning something to the meaning of an assertoric utterance? Brentano argued that one cannot perceive the judgements of other people, but that one can infer the existence of

46     M. Textor

such judgements from natural signs/indicator. He takes the difference between direct perception and inference to be inessential for nonnatural meaning: if some behaviour of B is an indicator of B ’s judging that p, and A is disposed to move from perceiving the indicator (natural sign) to the judgement that B judges that p, A is still in a position to arrive at a belief about B ’s mental life as well as a belief about the world. In Brentano’s words: For example, when I say There is a God this statement expresses that I acknowledge God. But its meaning and what it states is that God can be acknowledged in a correct judgement, that he is to be acknowledged. (EL 80: 13.132)

So the core meaning of an assertoric utterance is that it is right to judge that there is a God. If the audience grasps this meaning and wants to do what is right, they will judge that God exists and the acknowledged fact or object can be taken to be a further utterance meaning. However, if we assume like Brentano himself that utterances are indicators of judgements, the correctness commitment of judgement does not help to make progress in developing an alternative to Grice. There are many indicators of judgements that don’t have non-natural meaning. When I judge that it is raining, my right eyebrow raises. It is one of my ticks; I can’t help doing it and I can’t do it intentionally. It simply happens to me. In this situation, my raising of my eyebrow is a natural indicator of my judging that it is raining. People who know about my ticks are able to tell when I make the judgement that it is raining by spotting my eyebrow raising. But the raising of my eyebrow does not meannn that it is raining.10 This event is an indicator of a mental event but nothing more. Brentano is therefore wrong to think that there is no important difference between the case in which one immediately perceives a judgement and the case where one infers a judgement from the existence of a natural sign and additional background knowledge. In sum, Brentano has a promising idea how a judgement can mean a (non-mental) state of affairs, but he has not managed to give an answer to the question how an utterance, whether communicative or not, can non-naturally mean such a state of affairs. More work needs to be done.

3  How a Statement Has Meaning by Expressing …     47

7 Non-communicative Intentions to the Rescue? Our problem is how an utterance that is not directed at a (potential) audience can have non-natural meaning. Brentano has an account of how judgements can mean something, but not how the utterances that make judgements manifest can do so. Can we close the gap? Davis (2003) presents an interesting hybrid position that incorporates elements of both Brentano’s and Grice’s views. Like Brentano and unlike Grice, he dispenses with communicative intentions and the notion of judgement (belief ) is at the core of his account of speaker meaning. Unlike Brentano and like Grice, he makes use of intentions in his account. Davis’s original point is that the intentions with which a speaker does something are non-communicative. His (2003: 57ff) definition of speaker meaning incorporates an intention to express an occurrent belief aka judgement. One of the theorems of his theory is: S meant that p by (producing) e iff S produced e as a direct and undisguised indication that he occurrently believes that p. (Davis 2003: 57)

The variable ‘e ’ ranges over observable actions. One performs an observable action as a direct and undisguised indication entails that one performs e with the intention that e be a direct and undisguised intention (see Davis 2013: 62). So one can mean that p without intending to that the intention is recognized and the recognition serves as reason for a belief: ‘Meaning something does not entail intending to communicate’ (Davis 2013: 62). Davis seems to me on the right track by dispensing with communicative intentions. But his definition begs a crucial question. Why does the speaker S not simply mean that she herself judges that p? After all, that is what S intends to indicate by doing e. Yet, it is not what the utterance (non-naturally) means. Brentano helps to close this gap in Davis’s definition of speaker meaning. If the speaker intends e to be an undisguised indication of her judgement that p and the judgement has the correctness commitment Brentano outlines, the speaker can mean that p by doing something

48     M. Textor

with the intention to make her judgement obvious. For the fact that the speaker judges that p is, given the right background conditions, a reason to judge that p (and not only to judge that the speaker judged that p). The combination of a non-communicative intention to indicate a judgement and Brentano’s account of the correctness commitment looks attractive. But it gives rise to new problems. First. If you do something in order to achieve a goal, the goal must be something that you want to realize for a reason. An obvious question is therefore for which reason someone wants to indicate her belief. Why is it desirable for me to indicate my judging that p independent of any audience directed desires? Davis (2003: 69) responds by rejecting the ‘rationalistic thesis’ that people intend to do something only if they have reason to intend to do it. We often intend to do things because we ‘just feel like it’. Compare this to the Grice’s view. The question ‘Why is it desirable to make an utterance with a communicate intention?’ has a good answer. It is, in general, desirable to share my knowledge with others and, in particular, I take it to be desirable that my audience acquires a particular belief or, more generally, a particular attitude. For this reason, I make an utterance with a particular communicative intention. In contrast, the intention to indicate one’s belief seems unmotivated. Davis’s theory is incomplete as long as we have not given a reason to assume that thinkers like us have often ‘just feel like’ doing something with the intention to indicate their beliefs. Second. Davis characterizes indication as follows: “A indicates B” says roughly that there is a causal and statistical relation between A and B in virtue of which A would give a suitably placed observer a reason to expect B. (Davis 2003: 47)

As a characterization of indication, this seems along the right lines. But it causes a problem for Davis’s definition of speaker meaning. A can only be an indication of B if there is a causal or statistical relation between them. But how can I then intend that something I do is an indication of a belief of mine if indication requires a statistical or causal relation? I cannot perform e with the intention that something stands in

3  How a Statement Has Meaning by Expressing …     49

a causal or statistical relation to a belief of mine because I cannot expect to bring about the obtaining of causal/statistical relations by something I do, at least not by indicating my belief. More is required which is not in power.

8 Commitment to the Fore Frank Vlach gives a different response to the problem of speaker meaning without communicative intentions that does not posit an intention to indicate a judgement. Consider a case of speaker meaning without intention. When the student answers in the oral exam the question ‘When was the Battle of Waterloo?’ by uttering ‘1815’ with assertoric force, she does not intend her examiners to acquire a belief. Nor will she intend that the examiners acquire a belief about her beliefs about the time of the battle.11 For she knows that the examiners as experts in the field know this already. But she does make the utterance in the belief that she is thereby committing herself to the truth of the proposition that the battle of Waterloo took place in 1815. Vlach takes examples like this to point us to the right definition of speaker meaning. He defines, modifying a suggestion from Searle, speaker meaning as follows: U means P by x if and only if U does x in the belief that he is thereby committing himself to P. (Vlach 1981: 382)

Speaker meaning is supposed to be a variety of a person ‘standing behind a proposition’ (Vlach 1981: 369). This is congenial to Brentano. If one judges that p, one is committed to the correctness of the content of one’s judgement. Hence, merely judging that p incurs the commitment that Vlach takes to be fundamental for speaker meaning. The problem is that, so far, Brentano has given us only a reason to hold that judgements have a meaning—they signify something distinct from themselves—but not utterances that manifest judgements. In Vlach’s definition, the idea of commitment to truth plays a central role. But he develops the main idea in a way that works around this problem that

50     M. Textor

arose for Brentano. According to Vlach, the Gricean mechanism only works because the speaker commits himself to the truth of what the utterance says. Doing something in the belief that one thereby commits oneself to the truth of P is necessarily sufficient for meaning that p by it. The notion of ‘standing behind the truth of a proposition’ has a role to play in the definition of non-natural meaning. But Vlach’s definition does not expound this role correctly. A speaker who makes a lying assertion that p, means that p, but does not believe that by making the assertion she commits herself to the truth of p. She only believes that she gives others the impression that she commits herself to the truth of this proposition. Similarly, Vlach’s definition of speaker meaning makes it impossible to mean that p if one is tentative or merely assumes that p (see Davis 2003: 75).

9 Utterance Meaning and the Identity Thesis Brentano has an account of how a mental act can meannn states of affairs. Unfortunately, he assumes that this account would also explain what the meaningnn of utterances that make judgements manifest consists in. We saw that this assumption was mistaken. The problem that spoils Brentano’s strategy to explain speaker meaning without appeal to communication lies in the assumption that an assertoric utterance is just a natural sign of judging: an assertoric utterance, or ‘more generally’ something the speaker does, and the judgement are two distinct events the one causing the other. If the utterance is a natural sign of the fact that a judgement takes place, it has natural, but no non-natural meaning: it does not inherit the non-natural meaning of the judgement of which it is a symptom. The problem is resolved if there are utterances that are not merely (regularly) caused by judgings, but are identical with judgings. Frege, for example, considers the possibility that ‘our thinking was first speaking which then became imaging [vorstellen ] speaking’ (Frege 1924– 1925: 288 [269]).12 It may not be conceptually necessary that we think, as Frege put it, in words. But there are some forms of thought, conscious propositional thoughts, that beings like you and me can only

3  How a Statement Has Meaning by Expressing …     51

have if we imagine uttering sentences.13 Imaging uttering a sentence in the indicative mood with understanding can be a judging. Why? The imagined speaking with understanding can be a making up my mind: the event of imagining can be the initialization of a disposition to bet on the obtaining of a certain state of affairs, to take its obtaining as a premise in reasoning and so on. Now it is contingent that I merely imaged speaking with assertoric force, I could have spoken my mind: the silent speaking might have been a proper speech act. Often, we simply ‘come out’ with our view when we spontaneously make an assertoric utterance. In such cases, it is not warranted that I first judged that p and then went on to make the judgement manifest in an assertoric utterance. A ‘thin description’ (Ryle) of the event is that the speaker produced a sentence with a certain meaning. A richer description is that the speaker judged that p.14 But these are two descriptions of the same event: My uttering s with understanding today is the same event as my judging that something is the case today.

On the basis of this identity thesis, we can define the meaning of an assertoric utterance: S’s doing x meansnn assertorically that p if, and only if, S’s doing x is the same event as S ’s judging that p.

I take a judgement to be an event that either initiates or manifests or activates a belief. Hence, one can mean that p by an utterance although one already knows for some time that p. An event that a speaker brings about has non-natural meaning in virtue of being a judgement and not in virtue of being a natural sign of a judgement. There are natural signs of judgements like my raising my eyebrow (see Sect. 4). These indeed have only natural meaning because in order to have non-natural meaning, an event needs to be identical to a judgement. This move solves the problems created by the incorporation of communicative intentions in a definition of speaker meaning. If your

52     M. Textor

uttering is your judging that p, you mean that p, whether there is someone to hear it or not. There is no audience directed intention required at all. Not even the intention to convey in an undisguised manner a belief is required. In fact, we can and need to understand ‘S ’s doing’ so broadly that it includes events over which I exert no control. Someone may have no control over my vocalizations; I might just come out spontaneously with my judgements. My utterances meannn something, although they are not actions. How can one understand what such an utterance meansnn? Not by recognizing the intention with which the utterance is produced. Let’s go back to Brentano to see how he suggested to answer this question. To repeat: If we therefore assume that one would be able to immediately perceive the judgement of another person in whose insight one trusted, this judgement would indicate how its object is to be judged, whether it is to be acknowledged or rejected. (Brentano EL 80: 13.131, my translation and emphasis)

I think that the assumption Brentano took to be implausible is true. We can immediately perceive judgements as we can immediately perceive events such as sunsets and long jumps. For the event of uttering s with understanding can be perceived and the event of judging can be perceived. Now perceiving an event is not the same as perceiving that the event is of a certain kind. But someone with the right perceptual abilities might simply hear the speaker uttering that p. When I hear someone making an assertoric utterance to the effect that p in a language I master, it is part of how things seem to me that have an experience as of ‘hearing that p’ (see Fricker 2003). This needs further defence and elaboration, but I trust that the basic idea is clear enough for the purposes of this paper.

10 What About Lying? Brentano’s opponents will object that a lying assertion is an assertoric utterance, but it is not a judgement that p. The liar does not judge that p is the case. Yet, a lying assertion that the earth is flat still meansnn that

3  How a Statement Has Meaning by Expressing …     53

the earth is flat. If the lying assertion is not a judgement, one can do something which meansnn that p without judging that p. So the definition of utterance meaning proposed does not work. The first stab response to this objection is that identity thesis concerns particular utterances and judgings, not the types assertoric utterance and judgement. There are assertoric utterances that are not judgements, but there are others that are. However, this seems only to shift the problem. For the assertoric utterances of ‘The earth is not flat’ that are not judgements still mean non-naturally that the earth is flat. They can have this meaning because it has been the case that there were assertoric utterances of ‘The earth is not flat’ that were judgements: S’s doing x meansnn assertorically that p if, and only if (i) S’s doing x is the same event as S ’s judging that p or (ii) S ’s doing x is an event that seems to have the same intrinsic properties as an event that is a judging that p.

The second conjunct is dependent on the first: without there being a regularity of judging that p by doing x, there would be no lying assertions.

11 Commitment Generalized Now while we are on the right track, we are still not there. Brentano only considers judgements, assertoric utterances and their non-natural meaning. But many utterances are not judgements. How can one handle requests, assumptions and the like? We don’t need to assume that all utterances that carry non-natural meaning are judgements. For other attitudes have the same correctness commitment. Take desire. I follow Stampe who claims: It is intrinsic to a desire, essential to its being a desire, that it should, radical breakdowns apart, comprise an indication that the thing wanted would be good. If, in itself, a desire is such a state, then we may understand why a desire, through itself, comprises a reason to act so as to get the thing one wants. (Stampe 1987: 366)

54     M. Textor

There are attitudes like desire and judgement that have correctness conditions. If one has an attitude of this kind, one is rationally committed to hold that the correctness conditions are satisfied: If I actively desire that p, I am committed to the correctness of my desire and the goodness of the state of affairs that p. If I assume that p, I am committed to the correctness of my assuming and the state of affairs that p is worthy of this attitude.

Hence, while we can see the point of Vlach’s characterization of speaker meaning as ‘standing behind a proposition’, we can now bring out why it is too narrow. There is a broader notion of commitment that holds for a range of attitudes and not just for judgement. According to this notion, there are propositional attitudes φ such that φ-ing that p commits one to take one’s φ-ing to be correct. One does not ‘stand behind the proposition that p’, but one commits oneself to the correctness of one’s attitude. Utterance meaning can now be defined as follows: S’s doing x meansnn that p if, and only if, S’s doing x is the same event as S ’s φ-ing that p and φ-ing that p commits one to its correctness.

12 Conclusion I have outlined a view of non-natural meaning that is rooted in Brentano’s work. According to this view, the basic bearers of natural meaning are events such as judgements that can be identical to utterances. The fact that distinguishes natural from non-natural meaning is that in the former case an event produced by a speaker has epistemic import because the speaker is trustworthy and the audience knows basic facts about judgements or can at least put themselves in the judger shoes. This kind of epistemic import is independent of causal regularities and the like that hold between natural signs and the things they are signs for. There are further details to be filled in to complete Brentano’s picture, but I hope that I gave the reader some reasons to take Brentano’s proposal to be the basis for an alternative to Grice’s and Marty’s that is worth completing further.15

3  How a Statement Has Meaning by Expressing …     55

Notes 1. See Marty (1908: 493). See also Marty (1908: 22), translated in Cesalli (2013: 144). Cesalli (2013) is a detailed account of Marty’s philosophy of language. He (2013: 159–160) comments on the difference between Brentano and Marty, but does not explore it further. On Marty and Grice, see Liedtke (1990), Longworth (2017), and Recanati (this volume). 2. A prominent example is Lewis (1975: 3). Bennett (1976) and Schiffer (1972) are book-length developments of Grice’s ideas. Neale (1992) gives a detailed overview of Grice’s work on meaning and its problems. 3. See Grice (1989: 93–116). See Neale (1992), section 5 for an overview over the discussion and a modification of Grice’s basic definition. 4. In unpublished manuscripts, Brentano endorsed also an intentional view of utterance, see his MS (‘Die Sprache’ Sp 4d). 5. See Sperber and Wilson (2015: 118–120) for a discussion of the photograph example that draws evaluates it in a way that differs from Grice’s. 6. See Bennett (1976: 171) on intention-dependent versus intentionindependent evidence. 7. See also Hyslop (1977: 68–69). Grice (1989: 112ff) gives an overview of prima facie audience-less utterances that nonetheless have nonnatural meaning. He outlines a number of responses. However, none of them—reminding yourself, addressing your future self, pretending to address someone, etc.—gets a grip on utterances. Marty (1908: 495) considers very similar examples and strategies. 8. Thanks to the editors for drawing my attention to this paragraph of Marty (1908). 9. See also Marty (1884: 70). 10. Thanks to Nick Allott and Benjamin Schnieder for pressing this point. 11. I take the example and the response from Grice (1989: 106–108). 12. See Dummett (1981: 362–363) for argumentative support for Frege’s view. 13. See Carruthers (2005: 119ff) for an argument that some ‘imaged sentences’ are identical with mental acts. An imaged sentence is an imagining uttering a sentence. 14. Thanks to David Jenkins for suggesting the identity thesis in discussion and for pointing me to Ryle’s work. 15. My thanks go to the participants of the workshop Meaning and Intentionality in Anton Marty in Geneva in 2017 and the graduate

56     M. Textor

students at the King’s Summer Seminar Series. I am grateful to Laurent Cesalli, David Jenkins, Eliot Michaelson, François Recanati, Thomas Sattig, Benjamin Schnieder and Alberto Voltolini for discussion. Special thanks go to Claudio Majolino for very helpful comments at the workshop, the editors for comments and Nick Allott for two rounds of comments and discussion.

References Bennett, J. 1976. Linguistic Behaviour. Cambridge: Cambridge University Press. ———. 1991. How Do Gestures Succeed? In John Searle and His Critics, ed. E. LePore, 3–15. Cambridge: Blackwell. Brentano, F. EL 80. Ms. Houghton Library. http://gams.uni-graz.at/archive/ objects/o:bag.el.80-html-norm/methods/sdef:HTML/get. Accessed 21 June 2018. ———. Sp 4d. Die Sprache. Ms. Houghton Library. Carruthers, P. 2005. Conscious Thinking: Language or Elimination. In Consciousness: Essays from a Higher-Order Perspective. Oxford: Oxford University Press. Cesalli, L. 2013. Marty’s Intentionalist Theory of Meaning. In Themes from Brentano, ed. G. Fréchette and D. Fisette, 139–163. Amsterdam: Rodopi. Davis, W. 2003. Meaning, Expression and Thought. Cambridge: Cambridge University Press. ———. 2013. Meaning, Expression and Indication: Reply to Buchanan. Thought 2: 62–66. Dummett, M. 1981. Frege: Philosophy of Language, 2nd ed. London: Duckworth. Frege, G. 1924–1925. Erkenntnisquellen in der Mathematik und Naturwissenschaft. In Nachgelassene Schriften, ed. H. Hermes, F. Kambartel, and F. Kaulbach, 286–294. Hamburg: Felix Meiner. Trans. as ‘Sources of Knowledge of Mathematics and Natural Sciences’ In Posthumous Writings, Hermes, F. Kambartel, and F. Kaulbach, 267–274. Oxford: Blackwell. Fricker, E. 2003. Understanding and Knowledge of What Is Said. In Epistemology of Language, ed. A. Barber, 325–366. Oxford: Oxford University Press.

3  How a Statement Has Meaning by Expressing …     57

Grice, P. 1989. Studies in the Way of Words. Cambridge, MA: Harvard University Press. Hyslop, A. 1977. Grice Without an Audience. Analysis 37: 67–69. Lewis, D. 1975. Languages and Language. In Minnesota Studies in the Philosophy of Science, ed. K. Gunderson, 3–35. Minneapolis: University of Minnesota Press. Liedtke, F. 1990. Meaning and Expression: Marty and Grice on Intentional Semantics. In Mind, Meaning and Metaphysics, ed. K. Mulligan, 29–49. Dordrecht: Kluwer. Longworth, G. 2017. Grice and Marty on Expression. In Mind and Language: On the Philosophy of Anton Marty, ed. G. Fréchette and H. Taieb, 263–284. Berlin and New York: De Gruyter. Marty, A. 1884. Über Subjektlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie I–III. Viertelsjahrschrift für wissenschaftliche Philosophie, 3. Reprinted in Anton Marty: Gesammelte Schriften, II.1, ed. J. Eisenmeier, A. Kastil and O. Kraus, 3–102. Halle: Max Niemeyer, 1918. ———. 1908. Untersuchungen zur Allgemeinen Grammatik und Sprachphilosophie, Bd.1. Halle: Niemeyer. Moran, R. 2005. Getting Told and Being Believed. Philosophers’ Imprint 5: 1–29. Mulligan, K. (ed.). 1990. Mind, Meaning, and Metaphysics. The Philosophy and Theory of Language of Anton Marty. Dordrecht: Kluwer. Neale, S. 1992. Paul Grice and the Philosophy of Language. Mind and Language 15: 509–559. Schiffer, S. 1972. Meaning. Oxford: Oxford University Press. Sperber, D., and D. Wilson. 2015. Beyond Speaker’s Meaning. Croatian Journal of Philosophy 15: 117–149. Stampe, D. 1987. The Authority of Desire. The Philosophical Review 96: 335–381. Vlach, F. 1981. Speaker’s Meaning. Linguistics and Philosophy 4: 359–392.

4 Determining and Modifying Attributes Jan Claas and Benjamin Schnieder

1 Introduction Franz Brentano, in his Psychologie vom empirischen Standpunkte: If adjectives are combined with a noun they usually enrich the concept with additional determinations; but sometimes they modify it. Similarly, when a predicate is combined with a subject it usually enriches the concept, but sometimes it modifies it. The former is the case, for example, when I say ‘A man is learned;’ the latter when I say ‘A man is dead.’ A learned man is a man; but a dead man is not a man. Accordingly, what the sentence ‘A dead man is.’ presupposes for its truth is not the existence of a man but merely that of a dead man; and similarly, the sentence ‘A centaur is a fiction.’ does not require that there is a centaur but only J. Claas (*) · B. Schnieder (*)  University of Hamburg, Hamburg, Germany e-mail: [email protected] B. Schnieder e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_4

59

60     J. Claas and B. Schnieder

that there is a fictional centaur, i.e. the fiction of a centaur, etc. (Brentano 1874: 287f., in the footnote that starts on p. 286)

In these remarks, we find an important distinction between two classes of attributive phrases. Usually, we are licensed to judge that an F-y G is a G. But in some cases we are not; in fact, in some cases we are licensed to judge that an F-y G is not a G. Since Brentano’s works, this distinction goes by the labels of determining versus modifying uses of attributive phrases. This is a noteworthy linguistic peculiarity that deserves the attention of a philosopher of language. Moreover, awareness of the distinction seems advisable to philosophers in general; for, separating the two cases is important to avoid philosophical confusions based on wrong linguistic generalizations. As far as we know, Brentano did not elaborate on the distinction in his published work.1 However, his remarks on the topic were taken up by some of his pupils, most notably by Anton Marty and Kazimierz Twardowski.2,3 In this paper, we want to discuss the distinction between modifying and determining qualifications, as it was conceived of in the Brentano School. In our discussion, we will lay a particular focus on how the distinction is applied to adjectives. While we think that the Brentano School drew attention to an interesting and philosophically important phenomenon, we also think that many of their theoretical views about them were on the wrong track. In particular, the way they spelled out the distinction when applied to adjective-noun combinations gets their semantics wrong. Here is how we will proceed: In Sect. 2, we give a very brief overview about how adjective-noun combinations can differ with respect to their inferential potential and we will introduce some technical vocabulary from contemporary linguistics. In Sect. 3, we introduce the distinction between determining and modifying adjectives in the Brentano School, characterize it extensionally, and point out some of its philosophical applications. In Sect. 4, we show how Marty and Twardowski want to account for their distinction in terms of meaning shifts. In Sect. 5, we discuss arguments for and against their proposal; ultimately, we will opt against the proposal since it involves an implausible

4  Determining and Modifying Attributes     61

postulation of multiple systematic ambiguities of natural language nouns. In Sect. 6, we take a look at two refinements of the proposal and argue that they do not fare better. In Sect. 7, we turn to a more promising approach made by Bernard Bolzano: expressed in contemporary vocabulary, the idea is to account for the distinction between determining and modifying adjectives by the different roles they play in a theory of compositional grammar. In Sect. 8, we revisit what we take to be the strongest argument in favour of Marty’s and Twardowski’s meaning shift and show how the discussion in the preceding section sheds some new light on it.

2 Adjectives—A Primer When adjectives are combined with nouns, the resulting phrases can differ with respect to their inferential potential. In what may seem to be the standard case, an adjective-noun combination allows for the detachment of either of the two components:

Fig. 1  Detachment of adjective and noun

But in some cases, detachment of the adjective is barred. In one class of examples, this is because the adjective exhibits a certain sort of relativity. Some adjectives do not simply characterize things in isolation as thus-and-so. Instead, they characterize things as thus-and-so for a specimen of a certain comparison class (so that the standards under which something counts as such-and-such depend on the comparison class). Thus, something may be small for a whale and yet not small for a mammal. In an adjective-noun combination, the pertinent comparison class is introduced by the noun. This is why the detachment of the adjective

62     J. Claas and B. Schnieder

is not a valid move here, because it can result in a change of the salient comparison class (or, alternatively, leave us without any salient comparison class, so that the predication which only involves the adjective is semantically indeterminate):

Fig. 2  Detachment of noun but not of adjective

Another sort of relativity is involved when certain evaluative adjectives are combined with nouns. A skilful cook is not skilful per se, but rather skilful when it comes to cooking. And a good liar clearly isn’t good per se. As before, detachment of the adjective is prohibited in such cases because it can result in a change of the salient aspect under evaluation.

Fig. 3  No detachment of adjective, detachment of noun debatable

Incidentally, in these cases the result of detaching the noun often seems to allow for different readings, e.g. as descriptions of a profession or of a habit; the result of detaching the noun is clearly not always legitimate for any of its possible readings. Whether there always is a reading in which it is detachable may be a matter for a debate. The actress Scarlett Johansson is reported to be a good cook. But to our ears it sounds just wrong to say that Scarlett Johansson is a cook. Finally, in some rare cases, detachment of the adjective is forbidden not because the resulting predication might be false, but rather because the adjective cannot occur in predicate position at all:

4  Determining and Modifying Attributes     63

Fig. 4  Detachment of noun but not of adjective (for syntactic reasons)

Turning to the other half of adjective-noun combinations, we find that there are also exceptions to the standard case in which the noun is detachable. Some potential winners are winners, some are not:

Fig. 5  Neither detachment of adjective nor of noun

Finally, there are even cases in which the detached noun never applies to things to which the adjective-noun combination applies. As Brentano pointed out, a dead man is not a man. And a merely potential winner, one may add, is no winner at all:

Fig. 6  Detachment of adjective, entailment of negated noun

The described distributional profile of adjective-noun combinations has been studied by linguists for some decades. They call an adjective A in an adjective-noun combination A∩N:

64     J. Claas and B. Schnieder

• subsective iff the extension of A∩N is a (proper or improper) subset of the extension of N (the set of human comedians is, for all we know, an improper subset of the set of comedians; the set of funny comedies is, unfortunately, a proper subset of the set of comedies) • intersective iff the extension of A∩N is the intersection of the extension of A and N (the set of lazy monkeys is the intersection of the set of lazy things and the set of monkeys); note that by this definition, intersective adjectives are a special sort of subsective ones • non-subsective iff A is not subsective (e.g. ‘alleged’ in ‘alleged murderer’) • privative iff the extension of A∩N is disjoint from the extension of N (e.g. ‘counterfeit’ in ‘counterfeit money’). We can relate these notions to the distributional features discussed above: • In the case of intersective adjectives, we have detachability both for the adjective and for the noun (Fig. 1). • In the case of adjectives which are subsective while not intersective, the adjective is not detachable, while the noun is (see Figs. 2 and 4; perhaps Fig. 3, if the noun is detachable). • In the case of non-subsective and non-privative adjectives, we have no detachability of the noun, but no entailment of its negation (Fig. 5). • Finally, privative adjectives are those in which the noun is not only non-detachable, but in which the negation of the noun is entailed (Fig. 6).

3 Determination and Modification 3.1 Determination, Modification, and the Contemporary Distinctions of Adjectives Let us now come back to the Brentano School and their distinction between determining and modifying uses of attributive phrases. Among other things, the distinction is meant to apply to adjectives combining

4  Determining and Modifying Attributes     65

with nouns (though it is not limited to such constructions; we will shortly come back to the question of the scope of the distinction). In this application, it is clearly related to the distinctions from linguistics that we mentioned in the preceding section. But how exactly? Brentano, Marty, and Twardowski regularly illustrate the distinction between determining and modifying adjectives by drawing attention to the distributional potential of the phrases they consider, focusing on the detachability of the noun. They stress that in the regular cases, i.e. the determining ones, you can detach the noun, whereas in the more extraordinary cases, i.e. the modifying ones, this detachment is prohibited: • A learned man is a man, a dead man is not (cf. Brentano 1874: 287f.). • A good man is a man, a dead man is not (cf. Twardowski 1894: §4). • A painted man is not a man, but a real man is a man (cf. Marty 1908: 515). The examples they give of determining adjectives are usually intersective; this provides a case for mapping the two categories onto each other. The examples of modifying adjectives, on the other hand, are usually privative,4 which in turn provides a case for mapping these two categories onto each other. In sum, it is therefore tempting to map the distinction between determining and modifying adjectives onto that between intersective and privative ones. What speaks against this mapping, however, is that the Brentano School seems to treat the distinction between determining and modifying adjectives as an exhaustive dichotomy. If that is indeed the intention, then determining adjectives should better be mapped onto subsective ones and modifying adjectives onto non-subsective ones.5 So we have two options of how to map the distinction between determining and modifying adjectives onto the contemporary categories from linguistics. The choice between the two options is hard to make, as it may seem that Brentano and his students were simply not aware that there are more subsective adjectives than intersective ones (they indeed never

66     J. Claas and B. Schnieder

mention the non-detachability of subsective adjectives such as ‘large’), or more non-subsective than privative ones (at least they never explicitly discuss any non-subsective phrases apart from privative ones). We think that the textual basis is too meagre to definitely settle this issue. But since our overall impression is that Brentano and his students rather had an exhaustive distinction in mind, we will work with the assumption that determining adjectives should be mapped on subsective ones, while modifying adjectives should be mapped on non-subsective ones. This assumption gives us an extensional grip on the distinction that Brentano and his students make—but not more than that. It is important to notice that the Brentano School does not propose to define or conceptually understand the distinction via the inferential behaviour of phrases. We will turn to their official analysis of the distinction in Sect. 4.

3.2 Determination and Modification: The Scope of the Distinction As we already mentioned, the Brentano School distinction is not exclusively concerned with adjective-noun combinations, even though they are a central application for it. Still, Brentano and his students take the phenomenon to be much more widespread. Marty, in particular, draws attention to a variety of linguistic constructions instances of which usually have a determining function, while some have a modifying one. Among them are • composite nouns A racehorse is a horse, while a seahorse is not.6 • predications, in particular with adjectives in predicative positions; Thus, a burglar can tell his colleague ‘This dog is fictile!’, conveying that it is merely an apparent obstacle (cf. Marty 1910: 134). • possessive constructions The valley of the Rhine is a valley, the valley of tears is not; ‘of the Rhine’ therefore has a determining function whereas ‘of tears’ has a modifying one (Marty 1908: 501). The case is similar for ‘web of lies’ versus ‘web of silk’ (ibid.).

4  Determining and Modifying Attributes     67

• phrases with ‘in’, ‘on’, ‘at’, etc. The king in Prussia is an actual king, whereas the king on the chess board and the king in a deck of cards are not (Marty 1908: 411; 1925: 56). Marty often makes use of such expressions in order to account for expressions designating fictional entities, e.g. ‘wolf in a fable’ (Marty 1925: 56) or ‘Famulus in Goethe’s Faust’ (Marty 1910: 86). • sometimes also self-standing nouns (‘This is a Goethe’; Marty 1908: 506, 514). Confronted with the range of the phenomenon, it is an open question whether all modifying expressions can receive equal treatment (disclaimer: we think they can’t). As we said in the beginning, our focus will lie on combinations of adjectives and nouns.

3.3 Determination and Modification: Uses vs. Phrases So far, we have talked as if the labels ‘determining’ and ‘modifying’ were supposed to mark (among other things) a distinction between adjectives. This is not completely accurate though, and it is time to correct the inaccuracy. What is really marked is a distinction between uses of adjectives. Twardowski stressed that many adjectives are modifying as part of some and determining as part of other expressions (Twardowski 1894: §4): A false friend is not a friend, so ‘false’ has a modifying function here. But a false statement is a statement; here, ‘false’ has a determining function. An analogous point can be made about one of Marty’s examples; he observes that the adjective ‘fictile’ (meaning made of clay; tönern ) has a modifying character as it occurs in ‘fictile dog’ (Marty 1910: 134). A fictile dog is not a dog but a piece of burned clay with certain perceptual features that make it resemble dogs. However, the same adjective plays a determining role as it occurs in ‘fictile vase’. A fictile vase is a vase; here, the sole function of ‘fictile’ is to specify what material the vase is made of. In ‘fictile dog’, however, it seems to do something in addition to specifying the material of the object; it seems responsible for making the whole phrase refer to statues instead of dogs.

68     J. Claas and B. Schnieder

Relatedly, sometimes an adjective-noun combination has two readings, where the adjective has a determining function in one of them and a modifying function in the other.7 Take the example ‘painted horse’ (Twardowski 1894: §4). When ‘painted’ is understood in a modifying way, a painted horse is not a horse but (a part of ) a painting. Understood in a determining way, Bucephalus, the horse of Alexander, is a painted horse, since it has been painted several times.8 (In yet another determining sense of ‘painted’, a painted horse is what you may get when you leave children alone with a horse and finger paints.)9 So, at bottom level, the distinction between a determining and a modifying role neither concerns adjectives per se, nor does it concern occurrences of adjectives in complex expressions. Instead, it is first of all a distinction between particular interpreted uses of adjectives. For ease of presentation, we will nevertheless often talk of an adjective being determining or modifying, trusting that the context will make clear to the reader what particular use of the adjective we have in mind.

3.4 Determination and Modification: Philosophical Applications The distinction between modifying and determining adjectives is per se an interesting phenomenon for the philosophy of language. But the Brentano School thought that it is of greater overall importance: neglecting the distinction is an apt source for philosophical confusions. While a detailed discussion of such confusions would lead us too far astray, let us at least give a few illustrative examples. A first case in point is the confusion between the content and the object of an idea or presentation, which can arise from not realizing that the adjective in ‘presented object’ (‘vorgestellter Gegenstand’) is ambiguous along the determining/modifying dimension. In order to illustrate this ambiguity, Twardowski (1894: §4) referred to the similar ambiguity of ‘painted’, which we already commented on: by a painted archbishop, one can understand either an actual archbishop who commissioned a portrait of himself (then we make a determining use of ‘painted’); or

4  Determining and Modifying Attributes     69

the resulting portrait that depicts him (then we make a modifying use of ‘painted’). In a similar way, Twardowski holds, by a presented (or: imagined ) archbishop we might understand an actual archbishop of whom someone has a presentation/idea, or we might understand the content of such a presentation. Conflating these usages of ‘presented’ can easily lead to the assumption that the object of an intentional state just is the content of the intentional state. (We only note in passing that Twardowski’s analogy is not perfect. In the modifying use of ‘painted’, he takes the painted landscape to be the painting; in the modifying use of ‘presented’, he does not take the presented archbishop to be the presentation, but rather to be the content of the presentation.) Another case in point is Meinong’s conception of merely possible objects. According to Meinong, a merely possible golden mountain is both golden and a mountain (see, e.g., Meinong 1971: 490). But this contention arguably relies on a wrong understanding of ‘merely possible’. While Meinong construes this expression as a determining phrase, a better reading would construe it as a modifying phrase: after all, a merely possible winner is not a winner; it is someone who could have won but actually had no such luck. But if we understand ‘merely possible’ in this way, a merely possible mountain is not a mountain; instead, it is an object that could have been a mountain.10 This line of reasoning leads to a conception of merely possible entities that crucially differs from Meinong’s; one can find such an account in Bolzano’s writings (see Schnieder 2007), and more recently in the works of Linsky and Zalta (1996) and Williamson (1998). An analogous point can be made about Meinongian conceptions of fictional entities, according to which a fictional detective is both a detective and fictional. Such a conception construes ‘fictional’ as a determining adjective. But a better understanding seems to be a modifying one, so that a fictional detective is not a detective; instead, it is a fictional entity which is described as a detective (an entity which is fictionally a detective; see Schnieder and von Solodkoff 2009: 145f.). This line of reasoning is in harmony with recent conceptions of fictional entities as abstract entities (see, e.g., Kripke 2013; van Inwagen 1977; or Thomasson 1999).

70     J. Claas and B. Schnieder

4 Modifying Adjectives as Meanings Shifters As we pointed out above, inferential patterns of detachability can serve as a criterion (or as necessary and sufficient conditions) for whether an adjective is determining or modifying. But Marty and Twardowski think that the distinction should not be defined in terms of inferential patterns. Rather, they define the distinction in terms of meaning and in particular of shifts of meaning. In this section, we will present the basics of this proposal; the next two sections will be devoted to a critical discussion of the proposal (and two variants of it). The following passage contains Marty’s official explication of modifying adjectives, as he gave it in his major work on the philosophy of language: Adjectival phrases such as: former, possible, merely imagined, merely desired etc. are called (in opposition to enriching ones) modifying ones in the eminent sense because that with which they are combined, [a] changes its meaning so completely [b] that what is designated by the whole combination has a completely different character than what is otherwise named by the respective noun. (Marty 1908: 514; square brackets inserted by JC & BS)

Twardowski gives a similar explication of modifying adjectives: A determination is modifying if it [c] completely changes the original meaning of the name next to which it stands (Twardowski 1894: 46f ).

We will call this the Meaning Shift proposal.11 According to it, • a modifying adjective shifts the meaning of a noun with which it is combined ([a], [c]), • and it shifts it completely ([a], [c]), so that the adjective-noun combination denotes something of a wholly different character than what the noun denotes on its own ([b]). In order to spell out the latter condition, we can make use of the inferential patterns noted earlier: the extension of the adjective-noun

4  Determining and Modifying Attributes     71

combination is not a subset of the extension of the noun, occurring on its own; in that sense, the two expressions denote something of a different character.12 So, the following definition seems to capture the core of Marty’s and Twardowski’s Meaning Shift proposal: A in A∩N is modifying ↔df. A changes the meaning of N, so that A∩N applies to things to which N does not apply in its standard meaning. (As should be clear from the above, our intended reading of ‘applies to things’ is not exclusive; we allow that A∩N applies to some or all the things to which N applies, so that the whole class of non-subsective adjectives is covered.) One can then characterize determining adjectives in a similar spirit: A in A∩N is determining ↔df. A determines the meaning of N further, so that A∩N applies to a (proper or improper) subset of the standard extension of N. Determining adjectives are the standard case which hardly seems puzzling. It is the modifying adjective that deserves our attention, and we will henceforth concentrate its case. The definition of modifying adjectives alone cannot count as a comprehensive account of the semantics of modifying adjectives yet. One might desire two further bits of information: Firstly, one would like to know what exactly the meaning of a noun combined with a modifying adjective is supposed to be. Perhaps, though, nothing general can be said here since it will vary from case to case. So let us take a look at some examples. The idea might be that in ‘wooden dog’ the meaning of ‘dog’ is that of ‘statue of a dog’. Then, the meaning of ‘dog’ in ‘painted dog’ would probably be that of ‘painting of a dog’. This might seem a little odd, though, since then ‘painted dog’ would be equivalent to ‘painted painting of a dog’ in which ‘painted’ is redundant. Perhaps, then, the modified meaning of ‘dog’ in ‘painted dog’ is a more general one, such as representation of a dog; that way, ‘painted’ would not be redundant, since there are painted representations of dogs as well as, for instance, sculpted or linguistic ones. It is not clear, however, that we can always find a modified meaning which

72     J. Claas and B. Schnieder

would not render the modifying adjective redundant. Take ‘winner’ in ‘potential winner’ or ‘murderer’ in ‘alleged murderer’. It seems hard to see what meaning ‘winner’ in ‘potential winner’ might have that has a greater extension than ‘potential winner’. What may be worse, it may even seem hard to explicitly state the alleged modified meaning other than by using the phrase ‘potential winner’ itself. But if we cannot specify the meaning in any other way, the application of the Meaning Shift proposal to this particular case would not only make the semantics of ‘potential winner’ redundant, it even threatens to render it defective. There is a looming regress in the claim that in ‘potential winner’, the word ‘winner’ means as much as ‘potential winner’. So, certainly some additional work would have to be done here to supplement the proposal in a sensible manner. Secondly, for a completion of the proposal one might further desire to see an account of how exactly the meaning shift is brought about by the adjective. Marty is not very explicit on this, and for the nonce, we do not want to delve into this question; but we will come back to it later and discuss an explicit proposal made by Twardowski (see Sect. 6.2).

5 On the Pros and Cons of the Meaning Shift Proposal 5.1 Why Believe the Proposal? One may expect the Brentano School to provide some reasons for why one should accept the Meaning Shift proposal. Actually though, neither Brentano, Marty, nor Twardowski pause to motivate the proposal; they rather present it without much comment, as if it were obviously correct. One train of thought that might lead to this view would be the idea that the combination of an adjective and a noun is a syntactic construction that, by itself, uniformly allows detachment of either component. If that is the accepted starting point, then what to say about the observation that a false friend is no friend after all? A meaning shift would open a way of making this observation compatible with the idea that the syntax of the phrase ‘false friend’ should allow for the detachment

4  Determining and Modifying Attributes     73

of the noun. For, the inference from ‘false friend’ to ‘friend’ would then be syntactically justified but semantically blocked because the detached noun ‘friend’ would in course of it undergo a shift of meaning. The inference would therefore not be syntactically blocked (indeed, if we could keep the interpretation of ‘friend’ constant over the inference, the inference was correct). But the described motivation would seem rather weak to us; it rests on a general belief about the inferential potential of a particular syntactical construction which seems contestable in the light of empirical evidence. Looking at a particular example might provide a better motivation for the proposal. Consider the expression ‘wooden monkey’ which we apply not to monkeys, but to wooden replicas of monkeys. In fact, this is what the phrase apparently means: wooden replica of a monkey. But how can it do that? It is built up from two meaningful components. The first of them is the term ‘wooden’; as occurring in the phrase, it seems to mean exactly what it means in phrases such as ‘wooden cup’ and ‘wooden table’. To wit: made of wood. But then the other meaningful component should be responsible for contributing the remainder of the meaning; since the remainder of meaning is not monkey but rather replica of a monkey, we should conclude that ‘monkey’, as occurring in ‘wooden monkey’, does not mean what it usually means (i.e. it does not mean monkey). We take this to be probably the strongest motivation one can give for the Meaning Shift proposal. But even though this might motivate the account, there are still strong reasons against its correctness, to which we will now turn. (In the final section of the paper, we will come back to this particular motivation and re-evaluate it in the light of our discussion; see Sect. 8).

5.2 Against the Meaning Shift Proposal It is an integral part of the proposal that when a noun is combined with modifying adjectives, it will have a different meaning than it has when it occurs on its own:

74     J. Claas and B. Schnieder

A in A∩N is modifying → N in A∩N has a different meaning from when N occurs on its own. In what follows, we will take issue with this central part of the Meaning Shift proposal. We will present three arguments against it. Firstly, as we pointed out, the Meaning Shift proposal implies that nouns which can be combined with modifying adjectives have multiple meanings, i.e. that they are ambiguous. But many such nouns just do not seem to be ambiguous. We have paradigmatic cases of ambiguous nouns. ‘Bar’ is clearly ambiguous. If the Meaning Shift proposal is correct, however, presumably all nouns are ambiguous. But clearly ‘dog’ and ‘bar’ are not on equal terms, which is why a treatment of them in terms of ambiguity appears undesirable. ‘Bar’ deservedly has different entries in a dictionary, one concerned with long rigid pieces of certain materials and the other one concerned with the counter in a pub. The entry for ‘dog’ does not list meanings concerned with statues and paintings and we do not feel that it should. This suggests that ambiguity is not the right way to go in this case.13 Kripke agrees with our objection. A toy duck is not a duck but a toy resembling a duck. However, it would be wrong to conclude from this that ‘duck’ is ambiguous: ‘[N]o dictionary should include an entry under ‘duck’ with a sense in which ducks may be made of plastic and not be living creatures at all’ (Kripke 2011: 346). Relatedly, the proposal seems at odds with how we learn to use modifying expressions. What we learn, in such cases, appears to be the meaning of the modifying expression, not new extraordinary meanings of a vast amount of terms: When we learn how to use ‘depicted’, we do not learn new meanings of all those general terms which may be qualified by it. Rather, we learn to derive the meaning of ‘depicted F’ from the meanings of its components. The difference from a determining attribute as ‘blue’ in ‘blue fish’ consists in how the meaning of the component is derived, not in what the common component (in our case ‘fish’) means. (Schnieder 2007: 532, note 29)

4  Determining and Modifying Attributes     75

Secondly, ambiguity tests provide evidence against the Meaning Shift proposal. Here is a standard test for ambiguity: we can hark back to expressions by using appropriate pro-forms. For instance, we can hark back to a singular term by using pronouns (‘This is Beth; she is my wife’.) and to a common noun by using ‘one’ (‘Beth has a boat; I’d like to have one too’.). If we use a pro-form, it will receive the same interpretation as the target term. In particular, if a pro-form harks back to an ambiguous term, interpreting the target term in one meaning and the pro-form in a different one results in a zeugma. Thus, the following sentences are both fine: (1) I withdraw my money from a bank. (2) I often sit on the bank (where ‘bank’ denotes a riverbank). If we combine them and replace one occurrence of ‘bank’ with ‘one’, we obtain: (3) # I withdraw my money from a bank and often sit on one. But if (3) is meant to convey the conjunction of (1) and (2) as interpreted above, it sounds zeugmatic. The only non-defective interpretation of (3) would amount to the statement that the speaker often sits at a financial institute, but not that she often sits at on a riverbank. Now let us use this method in order to test the hypothesis that when a noun is combined with a modifying adjective, it has a different meaning than when it occurs on its own or when it is combined with a determining adjective. In the following sentence, ‘dog’ is combined with a determining adjective, which is why it should express its ordinary meaning: (4) Healthy dogs usually have four legs. In the following sentence, however, ‘dog’ is combined with a modifying adjective. So according to the Meaning Shift proposal, ‘dog’ should have a different meaning in (5) than in (4):

76     J. Claas and B. Schnieder

(5) Fake dogs usually have four legs. However, the occurrences of ‘dog’ in the two sentences fail the ambiguity test, since the following combination of (4) and (5) does not sound zeugmatic: (6) Healthy dogs usually have four legs, and so do fake ones. The fidelity of the following sentences (all of which were collected on Coca, the Corpus of Contemporary English ) provides further evidence against the ambiguity claim: (7) Alice [praised] the current President Roosevelt, as opposed to the previous one […]. (8) [A] national security appointment, much less a potential one, should never be turned into a political football. (9) His real eye matched the frown of the painted one. [Context: The character in question wears an eye patch with a painted eye on it.] (10) He lost a leg in World War I, had a wooden one fitted, and practised so well at concealing his limp that he seems to float through a room. (11) … others mix real locations with fictional ones. That such sentences are fine speaks against the hypothesis that the meaning of a noun changes when it is combined with a modifying adjective.14 Thirdly, let us focus on the application of the Meaning Shift proposal to some particular example. An alleged murderer is not always a murderer; so, ‘alleged’ has a modifying function here. Now what is the meaning of ‘alleged murderer’? We take it that the phrase means as much as Analysis AM someone who allegedly is a murderer. In this phrase, the term ‘murderer’ occurs and it clearly has its common meaning, the same it has in ‘Al Capone is a murderer’ as well as

4  Determining and Modifying Attributes     77

in ‘Jean Seberg is not a murderer’. But if the above analysis correctly captures the semantics of ‘alleged murderer’, then it seems that the underlined part captures the meaning of the modifying phrase ‘alleged’. Since ‘murderer’, as it occurs in the analysis, has its ordinary meaning, it must have its ordinary meaning in ‘alleged murderer’ as well. No meaning shift has occurred. The same holds, for instance, for ‘merely possible winner’. The meaning of the phrase can be captured as follows: Analysis MPW someone who is possibly, but not actually a winner. In this phrase, the term ‘winner’ has its common meaning, and so it does in ‘merely possible winner’.15 For the said reasons, we reject the Meaning Shift proposal in the form we presented it. However, there are two variations of the proposal that we will take a look at in the next section. Since we will find them wanting as well, we will then turn to what we take to be a better proposal.

6 Two Variations of the Meaning Shift Proposal 6.1 First Variation: Expression Types and Tokens Let us take a brief look at a variant of the Meaning Shift proposal that Maria van der Schaar defends in a recent book on Twardowski. When she discusses Twardowski’s take on modifying adjectives, she agrees that one can read him as holding the Meaning Shift proposal and she furthermore agrees that this proposal is defective (particularly because of the second argument above). However, she also presents ‘an interpretation of Twardowski that would not be vulnerable to [the same] critique’ (van der Schaar 2015: 46, in footnote 22 that starts on p. 45): According to Twardowski, a “determination is called attributive or determining if it completes, enlarges … the meaning of the expression to which it is attached. A determination is modifying if it completely changes the original meaning of the name to which it is attached.”

78     J. Claas and B. Schnieder

(Twardowski 1894: 11). This does not mean that, according to Twardowski, the meaning of the expression type changes. It is rather that he is speaking of the meaning of a particular occurrence of a term, that is, the meaning of the term in this particular context. The meaning of the expression in the context of a modifying term is different from the meaning it has in more standard contexts. (van der Schaar 2015: 51)

So, van der Schaar thinks that a variant of the Meaning Shift proposal is defendable. That variant is formulated in terms of the meanings of expression tokens, rather than the meaning of expression types: in uttering a noun, we produce a token of a type expression. If the noun is combined with a modifying adjective, the meaning of the token noun will differ from the meaning of the type noun (and hence, from what tokens of that type usually mean). That notwithstanding, the meaning of the type noun itself remains untouched. Let us call this the Token Meaning Shift proposal (for short: the tms proposal). We are not convinced. On the one hand, the argument from ambiguity tests and the argument from meaning analysis count against the tms proposal just as much as against the original Meaning Shift proposal. If the token of ‘eye’ that occurs in an utterance of ‘painted eye’ did not mean the same as the expression type ‘eye’, then an utterance of (9) should still be zeugmatic: (9) His real eye matched the frown of the painted one. And if (and since) a token of ‘alleged murderer’ means someone who allegedly is a murderer, the contained token of ‘murderer’ just means murderer. Moreover, and what is worse, on reflection we find the very idea of the tms proposal puzzling. What should one say, according to it, about the meaning of the expression type ‘painted eye’? It would seem bizarre to hold that the type does not have a meaning in which it denotes paintings of eyes, whereas tokens of it do have such a meaning (and presumably, van der Schaar would not want to hold this bizarre claim). The same holds for ‘fake barn’ and for ‘alleged murderer’. These expressions, as types, are proper members of the English language. As a type expression, the first term has a meaning in which it does not denote barns; the

4  Determining and Modifying Attributes     79

second has a meaning in which it denotes some murderers and a number of innocents as well. So let us agree that the type expressions do have the said meanings. How could van der Schaar’s tms proposal account for this? The meaning shift postulated by the proposal only occurs on the level of expression tokens. So the proposal does nothing to explain the meaning of the type. It seems fair, therefore, to conclude that the proposal is, at best, incomplete. However, we think the incompleteness of the tms proposal actually reveals an even deeper problem. Recall that van der Schaar formulated the proposal so as to immunize it against certain objections raised against the simpler Meaning Shift proposal: Her crucial idea is to deny that a meaning change occurs on the level of type expressions, since she agrees this is implausible. But now we have seen that something happens on the level of type expressions; ‘fake barn’ has a meaning, and one in which it does not denote barns; this is a linguistic datum in need of an explanation. Similarly, the expression type ‘painted eye’ has a reading in which it denotes paintings; linguistic theory has to account for that meaning. Now, however such an account may look like, it seems certain that it will straightforwardly carry over to tokens of those expression types; so the account will eo ipso explain why the tokens have their respective meanings. But the explanation for the type level, by van der Schaar’s admission, should not involve any meaning shift, and so there will also be an explanation for what goes on the level of token expressions in which meaning shifts play no role whatsoever. The tms proposal, therefore, is not only incomplete; it offers an explanation which will be overridden by a better one, once the proposal is completed. So we conclude that van der Schaar’s defence of the Meaning Shift proposal is a failure. The main reason is that any good explanation of how modifying adjectives interact with nouns must take into account that such interaction can take place on the level of expression types.16

6.2 Second Variation: Subtraction and Addition Marty is not very explicit about how exactly a modifying adjective is supposed to affect the meaning of a noun with which it is combined. Twardowski, however, lays out an account of this in his later paper

80     J. Claas and B. Schnieder

‘Issues in the Logic of Adjectives’ (which is only in part available in an English translation): The function of modifying (…) includes, first, the function of a partial removal of the content of the idea expressed by a given noun, and second, the function of replacing that removed part of the content – which is a result of combining a given adjective with a given noun – by other positive or negative characteristics. (Twardowski 1979: 28)

We shall call this the Give and Take proposal. In order to illustrate the main idea, let us take a look at a toy example. Twardowski thinks that the meaning of a noun, for instance ‘octopus’, determines a bundle of attributes that fixes its application conditions; the denotation of the noun is the class of objects that possess all the attributes in the bundle. For our example ‘octopus’, the attributes might be the following: being an animal, being a mollusc, being soft-bodied, having an oval-shaped body, and having eight arms and a beak. Now consider the term ‘woollen octopus’, in which the adjective ‘woollen’ has a modifying function: a woollen octopus is not an octopus but a toy for children. The Give and Take proposal explains how the term works: when the adjective ‘woollen’ is combined with ‘octopus’, it first takes away some of the attributes from its application conditions—namely those of being an animal, being a mollusc, and presumably also that of having a beak—while leaving the other attributes in place. It then adds a further attribute, namely that of being made of wool. ‘Woollen octopus’ then denotes soft-bodied things with an oval-shaped body and eight arms that are made of wool—i.e. woollen toys for children that resemble octopuses. What should one say about Twardowski’s proposal? First of all, even though Twardowski himself describes the Give and Take account in terms of meaning shifts (a characterisation he probably adopted from his former teacher Brentano), we think he shouldn’t. Assume for the moment that modifying adjectives work in the way described in the quotation; if such an adjective is combined with a noun, it operates on the concept expressed by the noun and removes certain components from it while adding others. In that case, it seems, there would be

4  Determining and Modifying Attributes     81

no reason any more to say that the noun combined with a modifying adjective does not have its normal meaning. Instead, the picture suggests that the modifying adjective operates on this very meaning; the result of this operation will be the meaning of the combined phrase. It will contain parts of the meaning of the contained noun (those which the modifying adjective has not removed) and some additional material provided by the adjective. Nothing is gained by adding that the noun in the phrase has a different meaning than normally; however, something is lost, because it makes the account vulnerable by inviting the objections we raised against the meaning shift thesis (see above). So, one can separate what Twardowski says about the twofold function of modifying adjectives from the claim that they affect any shift in meaning, and one ought to do so.17 But the Give and Take proposal is still objectionable for two independent reasons. Firstly, it is based on a picture of meaning according to which the meaning of a common noun assembles various qualities of the things denoted by the noun. While this was a traditional view, it came under attack from various sides in the philosophy of language in the twentieth century and it is clearly not the prevalent view any more, particularly in the case of natural kind terms. This is why we think the account does not even work for the toy example we gave; the meaning of ‘octopus’ does not consist in a bundle of attributes. But if there is no bundle of attributes, there is nothing that a modifying adjective could take away from it; which is to say, the adjective cannot fulfil the function it has according to the Give and Take proposal. Secondly, even if such a theory of meaning were accepted, the Give and Take proposal would only look plausible for selected examples, such as ‘woollen octopus’ or ‘wooden monkey’. But it would still be implausible for other examples. Take an example from above, the phrase ‘alleged murderer’. As we remarked earlier, this expression means as much as ‘someone who is allegedly a murderer’. The meaning of the latter phrase, however, involves the meaning of ‘murderer’, and it indeed involves the full meaning of ‘murderer’; no aspects of that meaning have been removed. So, Twardowski’s Give and Take proposal fails as a general description of how modifying adjectives work.

82     J. Claas and B. Schnieder

7 Modifying Adjectives and Modes of Combination We take the Meaning Shift proposal to be Marty’s considered view; he voices it in what seems to be his official explication of how modifying phrases work, and he often uses formulations in line with the proposal (moreover, in his Untersuchungen he deals with modifying expressions in the context of a larger investigation into meaning change). However, some of his remarks may be taken as a hint towards a different proposal—and in our view a more promising one. Concerning the expressions ‘merely desired A’ or ‘impossible B’ Marty tells us: A and B do not serve here as names, as they do when I talk about a little A or a big B. And even though the idea of A or B is at first evoked in both cases, in the former case [i.e. in ‘merely desired A’, ‘impossible B’] it still has a different function: here it does not serve as a meaning but rather as an inner linguistic form for a concept in which the idea of A or B does not occur in the same way as a constituent as it does in big A or little A. (Marty 1908: 137) Suppose I say: there is a sea lion. I thereby no more assert the existence of a lion than the sentence ‘A horse is painted’ asserts the existence of a horse. In neither case is there a composite concept that corresponds to the composite name in the same way in which such a concept corresponds to the names ‘young horse’, ‘desert lion’, and similar ones. (Marty 1884: 179; our italics)

Marty stresses that where a noun N is combined with a modifying adjective A, the ordinary meaning of N does not occur in the meaning of A∩N in the same way as usual. What he thereby suggests is that it nevertheless occurs in the meaning of the complex phrase—only in another way than when the noun is combined with a determining adjective. This suggestion, therefore, would mean a departure from the Meaning Shift proposal, since the meaning of the modified noun would not change; what would distinguish modifying adjectives from determining ones is not what the noun means, but rather how that meaning enters into the meaning of the whole expression. We can call this the Mode of Combination proposal.

4  Determining and Modifying Attributes     83

As it stands, this should perhaps better be called a hunch than a proposal, since it is all too fragmentary; what is missing is an account of the different ways in which the meaning of a noun can contribute to the meaning of the whole expression. As far as we can see, Marty does not go into any details. He does not tell us how exactly, according to the Mode of Combination proposal, the meaning of the whole expression is built up from the meanings of its parts. Fortunately, with his suggestion Marty is in good company. Bernard Bolzano draws a distinction between different ways in which combinations of adjectives and nouns should be understood which can capture the distinction between determining and modifying adjectives. He moreover offers an account of how these expressions are to be analysed and how they derive their meaning from the meanings of their parts. Thereby, he gives us an answer to the question of how the noun, when prefixed by a modifying expression, expresses its meaning in a different way. Firstly, let us consider expressions containing determining adjectives. Bolzano takes these cases to be the usual ones. It is usually the case that ideas for whose linguistic expression we use a noun and an adjective—e.g., ‘an honest man’, ‘fruit-bearing trees’—are so constituted that the idea designated by the noun is also the main idea, while the idea expressed by the adjective occupies a place in the proposition connected by means of ‘which’. (Bolzano 1837: §59.2 [I, 257])

By ‘idea’, Bolzano here means what he more explicitly calls ideas in themselves: in this sense of the word, ideas are not mental episodes (i.e. acts of thinking of something) but rather the contents of such episodes; he conceives of such contents as abstract entities that are building blocks of propositions and that also serve as the meaning of non-sentential expressions.18 Bolzano believes that the meanings of a variety of expressions are ideas of the form something which has b or A which has b. Brought into this schema, the meaning of ‘colourful fish’ is the idea fish which is colourful. The contribution of the noun ‘fish’ is to express what Bolzano calls the main idea, which governs the whole idea. This main idea already determines what kind of entity is represented. The

84     J. Claas and B. Schnieder

contribution of the adjective is to indicate the further structure of the idea and to express the idea colourful, which is a part of the which is colourful construction suggested by the whole expression. By adding this further specification, the class of things represented by the whole idea is a subclass of the things represented by the main idea. The main idea represents all fish; the remainder restricts the extension of the whole idea to colourful ones. Nouns accompanied by determining adjectives are the standard case. In the continuation of the passage we quoted earlier, Bolzano also offers an analysis of the non-standard case: But one would be wrong to assume that this always happens. In the expression ‘a painted fish’, ‘fish’ certainly does not designate the main idea, since (at least in the ordinary meaning of the expression) a painted fish is not understood to be a fish that is painted, but rather a painting that represents a fish. The main idea here is therefore that of a painting, which is indicated by the adjective. (Bolzano 1837: §59.2 [I, 257])

In ‘painted fish’, at least under the relevant reading of the expression, the adjective ‘painted’ is modifying. In this case, the contribution of the adjective and the noun to the meaning of the whole expression is converse to the one of an expression containing a determining adjective. ‘Painted’, rather than ‘fish’, expresses the main idea, i.e. painting, which tells us what kind of thing is represented by the whole idea, namely paintings. The noun expresses a part of the further specification. ‘Painted fish’ then expresses the idea painting which represents a fish. The meaning of ‘fish’ still occurs in the meaning of the whole expression; it is a part of the further specification that accompanies the main idea. So ‘fish’ still contributes its ordinary meaning. The way in which the expression contributes its meaning to the whole expression is, however, a considerably different one because it figures as a part of the attribute idea of representing a fish, not as the main idea. In both cases, the ordinary meaning of the noun is a part of the meaning of the whole expression, but it is contained in that complex meaning in different ways. That is why Bolzano’s proposal is a fleshed out version of the Mode of Combination proposal. Of course,

4  Determining and Modifying Attributes     85

expressions of this form always need some interpretation. There are different possible connections between the main idea expressed by the adjective and the ordinary meaning of the noun. In ‘painted fish’, it is via the concept of depicting (as it presumably is in ‘fictile dog’); in ‘wooden leg’, it may be via the concept of serving as, for a wooden leg is a piece of wood that serves as a leg. It has been suggested that the meanings of nouns are structured around a few broad categories that can account for the types of the connection, such as resemblance, function and constitution (Del Pinal 2015: 9f ). Can all combinations of adjectives and nouns be treated in one of the two ways described, i.e. with either the noun or the adjective expressing the main idea and the other one contributing in some way their usual meaning to the further specification? Bolzano does not take this to be the case. Sometimes neither of the two words straightforwardly expresses the main idea. His own examples are a bit elusive,19 but there are other, more accessible examples. A case in question is ‘former Prime Minister’. David Cameron is a former Prime Minister, but he is neither a Prime Minister, nor is he a former (whatever that may be). Here, both expressions work together in expressing the qualification while we have to retreat to a very general main idea such as someone: ‘former Prime Minister’ means, roughly, someone who once was a Prime Minister. And an alleged murderer is someone who is allegedly a murderer. He is not, necessarily, a murderer. He is necessarily not an allegation. We conclude that Bolzano’s account of the different functions of adjectives fares much better than the official accounts offered by the Brentano School; moreover, it predates them and it may actually have influenced them. It is documented that Brentano and his students read Bolzano. We do not know for sure, though, whether they read the passages that contain the Mode of Combination proposal. But note that Brentano, in the passage quoted at the beginning of this paper, applies the strategy Bolzano describes for interpreting combinations with modifying adjectives. He tells us that a fictitious centaur is not a centaur but the fiction of a centaur (Brentano 1874, 288, footnote starting on 286). So here, just as Bolzano’s proposal suggests, ‘fictitious’ expresses the main idea fiction while the noun ‘centaur’ contributes the further specification indicating just what kind of fiction it is.20

86     J. Claas and B. Schnieder

Marty often refers to specific passages in Bolzano’s works; we found him referencing §19, 48, 54, 123, 125 of the Theory of Science. While we did not find references to §29, in which Bolzano already sketches the Mode of Combination proposal, or to §59, in which he spells it out, it does not seem unlikely that he read them. Be this as it may, as Edgar Morscher told us, Bolzano’s anticipation of the distinction was certainly acknowledged by someone else in the Brentano School. Benno Kerry, commenting on Brentano’s claim that the distinction has been overlooked by logicians so far, draws attention to the relevant passages in the Theory of Science (Kerry 1891: 136, footnote starting on 135). He also notices that Bolzano uses the example of a dead man not being a man (Bolzano cites the example from Savonarola’s logic; Bolzano 1837: §19, [I, 79]), which was also used (and possibly picked up) by Brentano and his students. Marty’s disciple Hugo Bergmann explicitly names Bolzano as the discoverer of the distinction (Bergmann 1909: 61).

8 Coda and Conclusion In Sect. 4, we presented Marty’s and Twardowski’s account of modifying adjectives in terms of shifts of meaning. According to their proposal, a modifying adjective changes the meaning of the noun with which it is combined; hence, the noun is ambiguous: it would have one meaning when it is prefixed by a modifying adjective and another one when it occurs on its own. But we argued that this is an untenable view. Evidence against it is firstly provided by general reflections on ambiguity, secondly from ambiguity tests, and thirdly from a meaning analysis of test cases such as ‘alleged murderer’ or ‘potential winner’. So we rejected the Meaning Shift proposal and found a better account in Bolzano’s Mode of Combination proposal. But we are not quite done, yet. Before we presented arguments against the Meaning Shift proposal, we also presented what seems to be a strong argument in its favour, namely the wooden monkey argument (see Sect. 5.1): In ‘wooden monkey’, ‘wooden’ clearly means wooden. So the other meaningful component of the expression,

4  Determining and Modifying Attributes     87

‘monkey’, should contribute the remainder of the meaning, which is replica of a monkey. So in ‘wooden monkey’, the noun ‘monkey’ does not mean what it ordinarily means. Has something gone wrong? Or is the Meaning Shift proposal vindicated after all? There are mainly two things to be said here. Firstly, whatever else its merits may be, the wooden monkey argument just does not generalize. No parallel argument works, for instance, for ‘alleged murderer’. To the contrary, as we showed earlier, an analysis of its meaning provides an argument against a meaning shift: the phrase means someone who is allegedly a murderer. Here, ‘murderer’ contributes its ordinary meaning; ‘alleged’ contributes the rest. So, if the wooden monkey argument established that a meaning shift is involved in the case of ‘wooden monkey’, the lesson to draw would be that modifying adjectives do not allow for a uniform treatment. We are happy to concede this; we think that, on reflection, this claim should indeed not come as a big surprise. When the members of the Brentano School give examples of the distinction between determining and modifying expressions, they cast a very wide net. Just consider Marty’s examples of modifying phrases we presented earlier (see Sect. 3.1). They constitute a mixed bag of very different linguistic phenomena. They include, for instance, cases of deferred reference (‘This is a Goethe’) and cases of unbreakable idioms (‘vale of tears’), but also straightforward cases of adjective-noun combinations that do not involve any non-literal speech (‘alleged murderer’). It would be surprising if all such cases could receive a uniform treatment. And importantly, the reason is not just that Marty’s examples belong to a range of syntactically diverse expressions. Even if we concentrate on adjectives, his diet of examples is too diverse to allow for a uniform treatment. Here are two he would be happy with: a Lazy Susan is neither lazy nor a Susan; it is a convenient tool for sharing dishes, typically found in Chinese restaurants. And a mountain chicken is not a chicken; it is a frog (which actually lives in mountainous regions, though not exclusively on mountains). The two examples satisfy the typical inferential pattern of modifying adjectives. However, none of the proposals we discussed is applicable to them; this holds equally for the Meaning Shift proposal in all its variants and for the Mode of Combination

88     J. Claas and B. Schnieder

proposal. For, these proposals are concerned with phrases that have a compositional meaning. But where the terms ‘Lazy Susan’ and ‘mountain chicken’ are used to denote trays and frogs, their meaning is not a function, not even partially, of the meanings of their components. In these uses, the phrases do not work compositionally at all; they function as proper names, and as such the literal compositional meaning they otherwise have is interesting only from an etymological point of view. But even if we concentrate on cases of compositional adjective-noun combinations, there is an important difference to make (which we haven’t mentioned yet). The meaning of ‘alleged murderer’ and ‘fake gun’ is a straightforward product of the meanings of their components and the way they are combined. But what about the term ‘wooden monkey’? There is a straightforward product of the meanings of its components and the way they are combined. However, this compositionally determined meaning is monkey made of wood. Not only is this the straightforward compositional meaning of the phrase; we think it is its only semantic meaning. wooden replica of a monkey is not a semantic meaning of the expression in English. It is just that in any ordinary situation, the semantic meaning of the term is immediately ruled out, since there are no monkeys made of wood (though the meaning can be relevant in extraordinary contexts, such as fictions in which some animals consist of wood). Hence, we realize we have to reinterpret the expression and so we reach the alternative reading. But this reinterpretation is, in our view, clearly a pragmatic phenomenon. It is triggered because the semantic meaning of the term does not make sense in the cases we are concerned with.21 So there is an important difference between two classes of modifying adjectives. On the one hand, there are those which play a modifying role by virtue of their semantics; on the other hand, there are those which play a modifying role only in a pragmatic reinterpretation of an expression in which they occur. For the semantic cases, we stand by our endorsement of the Mode of Combination proposal, and accordingly by our rejection of the Meaning Shift proposal. The wooden monkey argument is concerned with a pragmatic case. But what exactly does the wooden monkey argument show then? Does it establish the meaning shift thesis for this particular case? While

4  Determining and Modifying Attributes     89

we think that this is a viable view, we do not think it is mandatory. To explain why, let us first reformulate the argument in pragmatic terms: when we interpret ‘wooden monkey’ as a term for wooden replicas of monkeys, we clearly take ‘wooden’ to mean wooden. So, the argument continues, we must take ‘monkey’ to mean replica of a monkey in order to arrive at our interpretation of the whole phrase. This is in fact how Kamp and Partee (1995) reason about the analogous case of the term ‘stone lion’. But we think this is a non sequitur. While the diagnosis that we reinterpret the noun is a genuine option, it is not the sole option. For, when we (whether pragmatically or semantically) assign a meaning to a complex phrase consisting of two terms t1 and t2, that meaning can involve elements that do not belong to the meaning of either t1 or t2. The syntax of the phrase can introduce such elements. And as a matter of fact, it sometimes does. By way of illustration, consider German composite nouns consisting of two nouns, such as the terms ‘Babybrei’ (literally: baby-mash) and ‘Kartoffelbrei’ (literally: potato-mash). In its standard usage, the latter term means mash made of potatoes. But ‘Babybrei’, in its standard usage, does fortunately not mean mash made of babies. Instead, it means mash for babies. In these cases, we need not, and should not, say that any of the terms ‘baby’, ‘potato’, or ‘mash’ gets reinterpreted. It is rather that the concatenation of two nouns in German can introduce different aspects of meaning. And something analogous could be said about our interpretation of ‘wooden monkey’. We might take both terms in their standard meaning and add some element of meaning by way of reinterpreting the concatenation function. This is a viable alternative to the view that we reinterpret ‘monkey’. Let us wrap up now. Firstly, we argued that the wooden monkey argument does not show anything about modifying phrases in general. Secondly, we argued that it does not show anything about the semantics of the phrase ‘wooden monkey’—that we assign the phrase the meaning wooden replica of a monkey is a pragmatic phenomenon. Thirdly, we agreed that one possible diagnosis of our pragmatic reinterpretation of ‘wooden monkey’ is that we assign ‘monkey’ the meaning replica of a monkey. But, pace Kamp and Partee, we do not think this is the

90     J. Claas and B. Schnieder

only possible diagnosis. An alternative view is that we keep the meaning of ‘monkey’ constant and reinterpret the concatenation function (as German speakers do in the case of Babybrei versus Kartoffelbrei ). Which option is the better diagnosis? For the moment, we do not know. So we end with the concession that there might be a quantum of truth in Marty’s and Twardowski’s views after all.

Notes 1. The distinction is repeatedly mentioned in Brentano’s notes for his Würzburg lectures on logic, but without further elaboration (in our bibliography, see the entry Notes ). 2. A note on the pertinent text basis: Marty takes the distinction between determination and modification to be important and gives it a considerable weight in his overall views on language, which is evident in his main book on the philosophy of language, the Untersuchungen. Nevertheless, he did not treat it at great length. In particular, he did not make it the topic of a whole chapter in the book (nor did he dedicate a chapter-long treatment to it elsewhere); instead, he repeatedly comes back to it in a number of smaller remarks here and there, where the longest treatment has just about the length of a page (it is the second half of §126 in the Untersuchungen, Marty 1908: 514f.). Twardowski explicates and uses the distinction in §4 of his ‘Zur Lehre vom Inhalt und Gegenstand der Vorstellungen’. He elaborates on it in a paper ‘Issues in the Logic of Adjectives’; unfortunately, only an excerpt of the paper is available in an English translation. 3. The distinction is also invoked by another prominent student of Brentano, namely Edmund Husserl (1921, section IV, §11). However, we will not discuss Husserl here, since he concentrates on cases of deferred reference (in particular on quotational uses of words as in: If is a connective), on the application of the distinction to predicates (as in: the centaur is a fiction), and on nominalizations (as in: Green is a colour). (Husserl also mentions the distinction in some posthumously published texts, see Husserl 2009, in particular 77, 99, 160, 182, 328.) 4. The possible exceptions we found occur in lists that Marty presents without mentioning a difference to his other examples; they are ‘alleged’ (‘angebliches’; 1916: 198) and ‘possible’ (‘mögliches’; 1908:

4  Determining and Modifying Attributes     91

514; 1916: 198) (in Brentano’s notes on logic, one also finds the example ‘möglicher Thaler’ meaning possible thaler; see Notes 13.311[2]). However, while we take those adjectives not to be privative, Marty may well have classified them differently. This is clearly suggested by the context of the second occasion at which he mentions ‘possible’ (we maintain, however, that he is mistaken about this; some possible winners actually are winners). 5. We should add that in his paper ‘Some Issue in the Logic of Adjectives’, Twardowski claims that besides determining and modifying adjectives, there are others. But the examples he mentions are somewhat opaque; in any case, they have nothing to do with subsectivity without intersectivity, or with non-subsective, non-private adjectives. (We decided not to discuss his examples here, since—as already remarked—they are somewhat opaque. They are moreover heavily dependent on his peculiar account of the determining/modifying distinction; we criticize that account below, in Sect. 6.2.) 6. Marty’s own example (1884: 179) only works in German: A Jagdhund (a hunting dog) is a Hund (a dog) whereas a Seehund (a seal) is not. 7. In the Coen brothers’ adaptation of No Country for Old Men, Carla asks sheriff Bell: ‘Sheriff, was that a true story about Charlie Walser?’ The sheriff answers: ‘Oh! Well … uh … a true story? I couldn’t swear to every detail but it’s certainly true that it is a story’.   Incidentally, the Coen brothers’ movie Fargo begins with the disclaimer ‘This is a true story’. 8. 1st Theorem: No painted horse without a painted horse, but not vice versa. 9. 2nd Theorem: A horse painted on a horse gives you a painted horse and a painted horse though not a painted horse unless the horse is painted on itself. 10. Marty (1908: 341–349) also employs the distinction between determining and modifying phrases in a critique of Meinong, though his criticisms differ from the one indicated above. 11. Some of Brentano’s remarks point towards this way of understanding the distinction between determining and modifying adjectives and may well have influenced how his students thought about the matter. Thus, the quotation from Brentano with which we started this paper says that if a modifying adjective is combined with a noun N it modifies the concept expressed by it, which may mean it changes the meaning

92     J. Claas and B. Schnieder

of N. But it is a somewhat condensed passage that probably allows for different readings. In Brentano’s notes for his logic lectures, however, he is more explicit when he remarks that in the case of a complex name consisting of two parts, ‘the relation between the one part and the other part can differ. Sometimes, the first part just adds a new attribute, but sometimes it modifies the sense of the other part; for instance “former king”, “future battle”, “dead man”, “imagined castle”, “desired money”, “fictitious Jupiter”, “possible thaler” etc.’ (Notes 13.211[2]; our italics). In the same spirit, he also remarks: ‘Sometimes it can happen that one name (expression of a concept) receives an addition in a separate expression, and through this a modification of its meaning’ (Notes 13.293[3]). 12. Our account does not capture every aspect of Marty’s and Twardowski’s proposals, since they certainly mean something stronger by ‘wholly different character’ than an extensional condition. They mean, roughly, that objects of a wholly different character have a different ontological status (as, for instance, concrete versus abstract objects). But for two reasons we concentrate on the simpler account we stated in the main text: firstly, it isn’t trivial to spell out the required notion of an object’s ontological status; secondly, the objections we will raise against the proposal equally apply to the weaker version and to stronger ones.   Relatedly, a decision we made earlier creates a further tension with Marty’s and Twardowski’s words here: we decided to include under the category of modifying adjectives not only privative adjectives but all non-subsective adjectives (see Sect. 3.1). Now if A is non-subsective but also non-privative, the extension of A∩N overlaps with that of N; here, it sounds more than odd to say that A∩N designate something of a wholly different character N. One could avoid this if one identified the class of modifying adjectives with that of privative adjectives, since for them the extension of A∩N never overlaps with that of N. Then, however, the distinction of determining vs. modifying adjectives would not be an exhausting alternative. As we said earlier, we think one can go either way. Nothing in our main discussion (see below, Sects. 5 and 6) depends on this issue. For, we will criticize the core idea of the Meaning Shift proposal and argue that a modifying adjective A does not change the meaning of a noun at all (no matter how exactly the alleged meaning change is said to bear on the extension of A ∩N ).

4  Determining and Modifying Attributes     93

13. One could try to distinguish standard ambiguity from non-standard ambiguity, where the latter is relevant for an account of modifying expressions. This, however, merely seems to be labelling the phenomenon in quite misleading terms rather than explaining it. 14. It is contestable how much evidence the ambiguity tests provide. Someone might, for instance, reply that in the examples we reinterpret the antecedent after the pro-form has been introduced (this was suggested to us by François Recanati). While the reply would have to be worked out in more detail, we agree that there is room for debate, which is why we do not rest our case against the Meaning Shift proposal solely on the application of ambiguity tests. 15. Another example would be ‘wrongly accused murderer’, where it is understood as meaning someone who is wrongly accused to be a murderer. In the analysis, as in the original expression, ‘murderer’ just means murderer. (Incidentally, Kevin Mulligan told us that the phrase cannot be understood that way. Since his English is excellent, we are confident that he is right—if, that is, ‘cannot’ is understood in a normative way. But as an empirical statement about how the phrase is actually understood, his claim seems false. To wit, see, for instance, the following articles in the New York Times: ‘So You Think You Want to Direct’, by Stephen Manes, the 29th of October 1996; or ‘At the Rialto’, author’s initials A.M., the 8th of April 1944.) 16. Let us add that we are open to the idea that there may be special cases of modifying adjectives whose behaviour indeed has to be accounted exclusively for on the level of token expressions. We come back to this point in the concluding Sect. 8 of this paper. 17. Note that the same does not hold for Marty. His account of modifying phrases essentially consists in the thesis that they effect a shift of meaning. He indeed deals with modifying phrases in the context of a larger investigation into meaning change and equivocation (Marty 1908: §§122–130). And his account of modifying phrases merely consists in the meaning shift thesis; unlike Twardowski, he does not offer a further characterization that would be separable from the meaning shift claim. 18. On Bolzano’s conception of ideas in themselves, see Künne (2001). 19. One of his examples is the locution ‘formal truth’. The phrase was often used in logical treatments of Bolzano’s times, but as Bolzano (1837: §29.4.b [I, 138f.]) rightly complains, the explications given varied and were often rather unclear. But one recognizable meaning of the phrase

94     J. Claas and B. Schnieder

was non-contradictory proposition. A formal truth, in this sense, is not always a truth then (since many falsehoods are non-contradictory), and it is not a form of anything either (since a formal truth is a proposition ). 20. Similarly, Husserl (2009: 328) mentions the idea that ‘presupposed’ has a modifying function in ‘presupposed state of affairs’ (‘vorausgesetzter Sachverhalt’) and that a presupposed state of affairs is not a state of affairs but rather a presupposition. 21. Recall van der Schaar’s Token Meaning Shift proposal according to which modifying adjectives effect a meaning shift, though only on the level of token expressions, not on the level of types. We rejected the view because of examples such as ‘alleged murderer’ and ‘fake gun’ which, as types of the English language, have a meaning in which the adjectives play a modifying role. But as we just explained, we don’t think ‘wooden monkey’, as a type expression, has such a meaning. So, our argument against the proposal does not work here. This is why (as we announced in footnote 16) we think that the proposal may be defensible for certain special cases. We can say this more precisely now: the proposal may be defensible for cases in which the modifying function of an adjective is not a semantic phenomenon but only a pragmatic one.

References Bergmann, H. 1909. Das Philosophische Werk Bernard Bolzanos. Reprinted in 1970. Hildesheim: Georg Olms. Bolzano, B. 1837. Wissenschaftslehre: Versuch einer ausführlichen und grösstentheils neuen Darstellung der Logik mit steter Rücksicht auf deren bisherige Bearbeiter, 4 vols. Sulzbach: Seidel. Reprinted 1970 and 1981 in Aalen: Scientia. Trans. as Theory of Science by P. Rusnock and R. George. Oxford: Oxford University Press, 2014. Brentano, F. 1874. Psychologie vom empirischen Standpunkt. Leipzig: Duncker & Humboldt. ———. [Notes ]: Notes for a lecture on logic. A transcription of the notes can be found on the webpage of the Brentano Archiv: http://gams.uni-graz. at/archive/objects/context:bag/methods/sdef:Context/get?mode=edition. Accessed 25 June 2018. Del Pinal, G. 2015. Dual Content Semantics, Privative Adjectives, and Dynamic Compositionality. Semantics and Pragmatics 8: 1–53.

4  Determining and Modifying Attributes     95

Husserl, E. 1921. Logische Untersuchungen II. Untersuchungen zur Phänomenologie und Theorie der Erkenntnis, 2nd ed. Halle: Max Niemeyer. In Husserliana XIX, ed. Ursula Panzer. The Hague: Martinus Nijhoff, 1984. ———. 2009. Untersuchungen zur Urteilstheorie: Texte aus dem Nachlass (1893–1918), ed. Robin Rollinger. Dordrecht: Springer, 2009. Kamp, H., and B. Partee. 1995. Prototype Theory and Compositionality. Cognition 57: 129–191. Kerry, B. 1891. Über Anschauung und ihre psychische Verarbeitung VIII. Vierteljahresschrift für wissenschaftliche Philosophie 15: 127–167. Kripke, S. 2011. Unrestricted Exportation and Some Morals for the Philosophy of Language. In Philosophical Troubles. Collected Papers, vol. 1, 322–350. Oxford and New York: Oxford University Press. ———. 2013. Reference and Existence: The John Locke Lectures for 1973. New York: Oxford University Press. Künne, W. 2001. Constituents of Concepts: Bolzano vs. Frege. In Building on Frege, ed. A. Newen, U. Nortmann, and R. Stuhlmann Laiesz, 267–287. Stanford: CSLI. Linsky, B., and E. Zalta. 1996. In Defense of the Contingently Nonconcrete. Philosophical Studies 84: 283–294. Marty, A. 1884. Über subjektlose Sätze und das Verhältnis der Grammatik zur Logik und Psychologie I–III. Vierteljahrsschrift für wissenschaftliche Philosophie 8: 59–94, 161–192, and 292–340. Reprinted in Gesammelte Schriften, vol. II/1 (1918), ed. J. Eisenmeier, A. Kastil, and O. Kraus, 3–101. Halle a. S.: Max Niemeyer. ———. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, vol. I. Halle a. S.: Max Niemeyer. ———. 1910. Zur Sprachphilosophie. Die “logische”, “lokalistische” und andere Kasustheorien. Halle a. S.: Max Niemeyer. ———. 1916. Raum und Zeit, ed. J. Eisenmeier, A. Kastil, and O. Kraus. Halle a. S.: Max Niemeyer. ———. 1925. Satz und Wort: Eine kritische Auseinandersetzung mit der üblichen grammatischen Lehre und ihren Begriffsbestimmungen, ed. O. Funke. Reichenberg: Stiepel. Meinong, A. 1971. Über Gegenstandstheorie. In Gesamtausgabe II, ed. R. Haller, R. Kindinger, and R.M. Chisholm, 481–535. Graz: Akademische Druck- u. Verlagsgesellschaft. Schnieder, B. 2007. Mere Possibilities: A Bolzanian Approach to Non-Actual Objects. Journal of the History of Philosophy 45 (4): 525–550.

96     J. Claas and B. Schnieder

Schnieder, B., and T. von Solodkoff. 2009. In Defence of Fictional Realism. Philosophical Quarterly 59: 138–149. Thomasson, A. 1999. Fiction and Metaphysics. Cambridge: Cambridge University Press. Twardowski, K. 1894. Zur Lehre vom Inhalt und Gegenstand der Vorstellungen: Eine psychologische Untersuchung. In Kasimir Twardowski: Gesammelte deutsche Werke, ed. A. Brożek, J. Jadacki, and F. Stadler, 39–122. Dordrecht: Springer, 2017. ———. 1979. Issues in the Logic of Adjectives. In Semiotics in Poland 1894– 1969, trans. O. Wojtasiewicz and ed. J. Pelc, 28–30. Synthese Library 119. Dordrecht: Reidel. van der Schaar, M. 2015. Kazimierz Twardowski: A Grammar for Philosophy. Leiden: Brill/Rodopi. van Inwagen, P. 1977. Creatures of Fiction! American Philosophical Quarterly 14: 299–308. Williamson, T. 1998. Bare Possibilia. Erkenntnis 48: 257–273.

Part II Ontology and Consciousness of Space and Time

5 A Presentation and Defense of  Anton Marty’s Conception of Space Ingvar Johansson

In order to come to conceptual grips with the world in which we live and perceive, we need to posit two kinds of spaces, physical space and perceptual spaces, i.e., the space talked about in physics and those talked about in perception theory. This paper is concerned with physical space.1 The expression “physical space” is in the singular, since I neglect the many-worlds interpretation of quantum mechanics.

1 Some Introductory Words Famously, Newtonian mechanics has a container conception of space. Space is regarded as an empty receptacle in which all the material bodies are contained. The philosophical question what kind of existence such a space can be ascribed, was for a long time mostly answered by saying

I. Johansson (*)  Umeå University, Umeå, Sweden e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_5

99

100     I. Johansson

that it must be regarded as being dependent on some kind of mind. Anton Marty is of a different opinion. He argues that the physical space of Newtonian mechanics should be regarded as mind-independent, and he works out a number of interesting metaphysical implications of his view. This fact alone, ought to give his posthumously published and much-neglected book Raum und Zeit (1916; henceforth RZ) a prominent place in the history of different conceptions of space. However, by combining Marty’s views with some arguments from Kant (Sect. 4) and Graham Nerlich (Sect. 5), and adding some reflections of my own (Sect. 6), my paper tries also to show that a container conception of space and space-time should be taken seriously even in the contemporary philosophy of physics. In other words, I regard Marty’s views on space as being of interest not only for historians of philosophy interested in presenting his RZ-criticism of the conceptions of space of Leibniz, Berkeley, Kant, Lotze, and Brentano, but also of being of interest in themselves. I disregard his views on time.2 It is important to note that Marty’s term “non-real” (“nichtreal,” “irreal”) does not mean non-existing, and that his term “subsisting” (“subsistierend”) does not mean what it means to his more famous contemporary Alexius Meinong, namely a way of existing that differs from that of ordinary objects.3 Meinong is not mentioned in RZ, and I take Marty to be of the opinion that everything that exists exists in the same way (RZ: 4, 52n1).

2 Marty’s Container Conception of Space in Outline According to Marty’s metaphysical conception of physical space, space is not a web of relations (RZ: 72); whatever the relata are. In contemporary terminology, Marty is not (like Leibniz, Berkeley, and Lotze) a relationist with regard to space. Possible relata such as bodies, properties, events (Machian sensations), spatial points, and spatial relations are regarded as contained within an ontologically preexisting space. Moreover, regarded as necessarily so contained, since space is a

5  A Presentation and Defense of Anton Marty’s Conception of Space     101

presupposition for their existence. Marty has, to employ a term he does not himself use, a container conception of space.4 Marty’s container conception of space is a special case of what today is mostly talked about under the label “substantivalism,” i.e., the general view that space exists in itself in addition to material objects. Barry Dainton, for example, distinguishes between three kinds of substantivalism (Dainton 2001: ch. 9.4), but none of them subsumes Marty’s conception. In both Dainton’s relational substantivalism and container substantivalism space and material objects are regarded as equally basic, but in Marty’s conception space is the more basic entity. Marty’s conception also differs from Dainton’s super-substantivalism, which holds that spatial objects are identical to or “adjectival” on regions of space. According to Marty’s conception, objects and spatial regions are colocated; that is, an object is always located at a specific spatial region, and this region is, in turn, located in a certain place in the all-encompassing space that contains all the regions. This registers yet another difference from Dainton’s container substantivalism; according to the latter “there is space only outside and between material things.” In other words, Marty’s conception is not present in the contemporary discussion of physical space. It conforms to the general definition of substantivalism, and had Dainton not already introduced the term “container substantivalism” to refer to another kind of space conception, this would have been a good label. However, if I may replace Dainton’s term “container substantivalism” by “symmetric container substantivalism” (since space and material objects are regarded as equally basic), then Marty’s view can be called “asymmetric container substantivalism” (since space is more basic), but I will call it simply the container conception of space. To repeat, Marty thinks that space is ontologically more basic than the entities contained in it. The contrary opposite of “substantivalist space” is, as already noted, the term “relationist (or relational) space” (e.g., Leibniz); the contrary opposite is neither “relative space” nor “relativistic space.” A container conception of physical space, be it symmetric or asymmetric, is compatible not only with an absolute space (think of Newton, i.e., certain kinds of motion between a body and space can be empirically detected), but also with a relative space (think of Ernst Mach, i.e., only motions

102     I. Johansson

between bodies can be detected) and a relativistic space (think of Einstein’s special theory of relativity, i.e., coordinate transformations between the contained inertial frames are Lorentz transformations); on these distinctions, see Johansson (2004: chs. 10 and 12.5). Of course, Newton and Kant also had a container conception of space, but Marty differs from both in regarding physical space as having a mind-independent existence. The famous Kant of the critical period, as is well known, regarded space as being mind-dependent in the sense that space is a form of intuition created by one of his posited transcendental faculties, but even the precritical Kant regarded space as subjective.5 More about this in Sect. 4. Newton, as is less well known, had a similar conception. To those who are familiar with only Newton the physicist, his and Marty’s conceptions of space look very much the same. Marty remarks, however, that Newton the philosopher seems to have regarded space as an attribute of God, and that in this sense space is for Newton, but not for him, a mind-dependent entity (RZ: 5–6). Today, we know that Newton described space as a sensorium of God. Giordano Bruno—who is reckoned to be the creator of the modern view that space is infinite, where space is easily thought of as a container—was a pantheist. Marty claims not only that space is mind-independent, but also that in principle it could have been empty. This is the main reason why I have chosen the label “container conception”; what is normally called a container is never regarded as being by necessity filled. His antiKantian teacher, Franz Brentano, also regarded physical space as mind-independent, but he thought empty space to be an impossibility. As one Brentano expert says: “Marty seeks a position more straightforwardly commonsensical than that of his master […] Thus for Marty […] space can exist even in the absence of bodies and of all qualitative determinations” (Smith 1994: 102). Marty himself says: “Briefly put, also according to him [Brentano] there is indeed real space or real places only where there are bodies” (RZ: 87; my translation). One of Brentano’s anti-Kantian teachers, Adolf Trendelenburg, might also be mentioned, even though Marty does not refer to him in RZ. Trendelenburg became famous for claiming that Kant made a mistake when claiming that space, since its structure can be known a

5  A Presentation and Defense of Anton Marty’s Conception of Space     103

priori, must be subjective; there can very well, he argued, be a mindindependent space, too. Nonetheless, neither Brentano nor Marty adopted Trendelenburg’s specific view of objective space. He regarded space (and time) as an effect of what he took to be ontologically absolutely basic, namely motion (“Bewegung”); see Beiser (2014: Part I, chs. 3 and 6). Marty, then, is the first modern philosopher who explicitly claims that physical space is in itself a mind-independent existing container. If “the void” of the ancient atomists (Leucippus, Democritus, et al.) should not be conceived of as a container, then he might be the first philosopher ever to put forward such a claim.6 I have not seen this fact noted anywhere. His only philosophical precursor is perhaps Pierre Gassendi (Crombie 1969: 167). It should be said, though, that the famous eighteenth-century Swiss mathematician and physicist Leonhard Euler once, in a paper directed against “the metaphysicians,” claimed that space is just as real as bodies are (Euler 1748). It seems likely, that some of his colleagues in physics implicitly had the same view. At least in RZ, Marty mentions Euler only once, and only in passing (RZ: 70). However, none of these possible philosopher or physicist precursors worked out the metaphysical implications of the container conception in the detailed way that Marty did, as we shall see in the next section. Marty ought, therefore, to be given a prominent place in the history of different conceptions of space. Of course, a new idea can later be created anew by thinkers who know nothing about its first appearance. And this fate has befallen Marty. He has had an eminent “follower” in Graham Nerlich. But Nerlich has done much more than merely repeat Marty’s position. Marty knew nothing about non-Euclidean geometries, the theories of relativity, and modern geometric topology. Nerlich does. And he takes them all into account when he, just like Marty, defends a realist container conception of physical space/space-time. I will return to Nerlich in Sect. 5. On the other hand, Nerlich does not discuss all the metaphysical issues that Marty takes pain to set out. Some years ago Nerlich claimed that “A large majority of ontologists are relationists [with respect to physical space]” (Nerlich 2009: 600), which I doubt holds true today. But, like Dainton, all ontologists seem to dismiss the kind of substantivalism that Nerlich and I defend.

104     I. Johansson

Convinced by Nerlich’s arguments, I came, already in the mid-1980s, to the conclusion that a container conception must be the most reasonable conception of mind-independent physical space. My view is presented in Johansson (2004: ch. 10), which relies on Nerlich but contains a reference to Marty. Unhappily, Nerlich’s philosophy of space seems very much to be forgotten or neglected.7

3 Marty’s Container Conception of Space in Metaphysical Detail The editors of RZ have done a good job in summarizing paragraph by paragraph what Marty says (RZ: VII–XIII). I will present and comment on parts of their summaries of what are, for the purpose at hand, the four most essential paragraphs. My translations are sometimes rather free, since I hope that some of the English terms that I have chosen are, compared to literal translations, more easily understandable by contemporary analytic metaphysicians and philosophers of physics. § 18 Space is neither a substantial difference between bodies nor a contingent property (“Akzidens”) inhering in these. The true-and-actual (“wirkliche”) existence of space is by nature ontologically prior (“ist ein Prius”) to the world of bodies, as also to its voids and unboundedness (“Lücken und Grenzenlosigkeit”). (RZ: IX)

If we take away the word “unboundedness,” what is said in the paragraph fits well our conception of ordinary rooms. Since normally we do not perceive air and electromagnetic fields, we perceive rooms as to a larger or lesser extent containing a void, and usually we think of them as existing prior to the things that have been put into them. A room is regarded neither as some kind of difference between the things in the room, nor as a contingent structure inhering in the properties of the things there. Moreover, rooms can always easily be thought of as being completely empty, as three-dimensional holes so to speak, and some of them are also so perceived.

5  A Presentation and Defense of Anton Marty’s Conception of Space     105

§ 30 What exists (“das Seiende”) is 1. either causally efficacious (“real”) or causally non-efficacious (“nichtreal”), 2. either absolute or relative, 3. either subsisting (“subsistierend”) or inhering (“inhärierend”); these classifications overlap. […] Space is something absolute and causally nonefficacious, which does not inhere in something else, but subsists for itself. (RZ: X)

Ad point 1. In contemporary analytic metaphysics, there is a discussion about whether or not one should subscribe to the so-called Eleatic principle, which claims that a necessary condition for an entity’s existence is that it is capable of making a difference to some causal process. Marty rejects such a principle, but not in the way it is sometimes rejected in analytic metaphysics, namely by a reference to the existence of abstract objects or theoretical entities; see, e.g., Colyvan (1998). According to Marty, existing space is neither an abstract object nor a theoretical entity, but it does nonetheless lack causal power.8 I would say that this point is basic to his view on space; without it, it would have been hard to regard a container space as being mind-independent. Ad point 2. Mostly, one should be cautious before one thinks one has understood what an author means by the term “absolute,” and this is especially true of Marty. According to him, the classification of existing entities into absolute and relative ones is exhaustive, and “relative” is the positive term of the opposition (RZ: 78). Therefore, the term “absolute” is the contradictory opposite of “relative,” and means simply non-relative. The term “relative” is understood in the pre-Russellian old Aristotelian sense of the Greek word “pros ti ” (RZ: 4), which means that Marty makes no distinction between relation predicates such as “longer than” and relational property predicates such as “longer than A.” If I understand Marty correctly, this means that, for example, monadic property instances are absolute even if they are regarded as being for their existence dependent on a property bearer (RZ: 177). Ad point 3. In the opposition between the terms “subsisting” and “inhering,” the latter seems to be the positive term. That something subsists, means according to Marty, only that it does not inhere in anything in the way properties are often regarded as inhering in things.

106     I. Johansson

Substances both subsist and can have causal powers. Space subsists, too, but lacks causal power. §31 That space is homogenous means only that all the different spatial points are equal species of a genus [namely, ‘spatial point’]. (RZ: X)

This view means that Marty does not think of spatial points as in themselves being scalar or vectorial physical point-properties. Such properties are regarded as fillings (“Erfüllungen”) of the spatial points; see §33 below. Using the distinction between names and predicates, I think Marty’s view is that “space” is a name and “spatial point” a predicate. “Space” denotes a single particular, and “spatial point” denotes an indefinite number of particulars that are exactly alike. In my opinion, Marty’s view that space is homogenous is implied by his view that it may be empty; where there is emptiness, there can be no heterogeneity. §32 Space cannot be understood as a) a detachable part of bodies, b) nor as something that inheres in bodies, c) and nor as a universal. (RZ: X)

Points a) and b) do not add anything to what has already been said. Point c) makes it explicit that Marty regards space as a particular, not a universal—in the ordinary philosophical senses of these terms. Particular spatial relations between bodies or spatial points are regarded as—to use a term used by Husserl and Marty—grounded in the allembracing particular that is space (Johansson 1990: 151). Space is a particular that contains other particulars such as bodies and their properties; the latter are not parts of space. §33 Space-filling (“Raumerfüllung”) is a relation sui generis. (RZ: X)

For a property instance or trope to have a position in space, is for Marty something other than for it to fall under a predicate or to stand in a relation of inherence to a substance. The term “space-filling” is a primitive term. One consequence of Marty’s view is, that one can say that two qualitatively and/or quantitatively identical bodies can, despite this kind of identity, occupy different places in space and be numerically different.

5  A Presentation and Defense of Anton Marty’s Conception of Space     107

Marty is of the opinion that the conceptions of space he is criticizing are due to an inability among philosophers to keep distinct all the concepts and distinctions he has put forward. I will not here rely on this view, even though in the main I accept his distinctions. Instead, I will defend the container conception by briefly presenting the argument for it that seems to me to be the strongest one; it was never noticed by Marty. It was first suggested by Kant, but Marty is, despite his knowledge of Kant, seemingly not aware of it.

4 The Handedness Argument I: Kant Most arguments about the container conception of space have been concerned with the question whether in relation to such a space motion can be detected or not. The best argument I know of in favor of a container conception has quite another character, it is geometrical. I call it the handedness argument, and I regard it as an argument only for a container conception of space, not for an absolute space in Newton’s sense. It was first put forward by Kant in his precritical period (Kant 1991 [1768]; 1894 [1770]); it is not used in his Critique of Pure Reason, but later in the critical period he uses it again; see Broad (1978: 40–41).9 Kant asks his readers to think of a lonely hand, and then note that, surely, it must be regarded as being either a left hand or a right hand. On a relationist conception of space, a lonely hand could be neither, since there is nothing to compare it with. This fact raises this question: how then should we explain the handedness fact? Kant’s answer is that there must be something distinct from the hand to which the hand has a relation, even if not an ordinary material thing. In fact, he claims, the hand is contained in a space that is distinct from the hand, and handedness is a relation between the shape of the hand and the space that contains it. In other words, where a handedness shape is possible, there must be a container space, too. Let it be noted, before I proceed, that the feature I call handedness is a feature of all shapes that, like hands, lack a center point, line of symmetry, and plane of symmetry. For example, so-called chemical enantiomers and chiral molecules have by definition this feature.

108     I. Johansson

The essence of the argument comes out more easily if, as Kant also did, one considers two equally large shapes that have the feature of handedness. Before I bring in hands again, look at the two twodimensional shapes b and d. If they are confined in a two-dimensional space, a surface, they cannot, despite their obvious similarity, by any kind of movement in the surface be made to occupy exactly the same spatial region where one of them was earlier. Figures that can occupy in this way a space are, to use a classic geometrical term, congruent. In the surface in which they exist, b and d are incongruent. Kant would call them incongruent counterparts, stressing the fact that they are incongruent despite having the same size and an obvious similarity in shape; in a mirror, a b becomes a d. However, in a three-dimensional space they are congruent. Rotate b 180 degrees in a perpendicular third dimension and place it down again, and then you will have the shape of d. The shapes b and d are two-dimensionally incongruent (I will define this notion in Sect. 6), but they are congruent in all spaces that have more than two dimensions. This latter fact, by the way, also eloquently shows that handedness cannot possibly be a brute intrinsic feature; such features stay the same independently of the dimensionality of the space in which they exist. Think now again of a right hand and a left hand. They are in a three-dimensional space what b and d are in a two-dimensional. In whatever way they are moved in a three-dimensional space, they cannot possibly be made to occupy exactly the same region where the other hand earlier was. So: how best to explain this peculiar similarity and dissimilarity between the two hands? It cannot be explained by the intrinsic features such as part sizes, part shapes, and angles between different corresponding parts of the hands. Therefore, the most reasonable explanation is to say that the hands are located in a container space to which they have different relations. Metaphorically speaking, they have been differently inserted in space. Literally speaking, the shapes b and d have been differently inserted on the surface in which they exist. In a mirror image, the shape of a left hand becomes the shape of a right hand, and vice versa; a mirror changes the mirrored thing into, in Kant’s terminology, an incongruent counterpart.10 The hands relations to space must involve more than mere occupancy.

5  A Presentation and Defense of Anton Marty’s Conception of Space     109

5 The Handedness Argument II: Graham Nerlich Both Kant and Marty took the following three views for granted: space is in its essence distinct from time, is three-dimensional, and has a Euclidean geometrical structure. Therefore, a question that has to be answered before one can claim that Marty’s conception of space is still of interest is this: has the combined emergence of the relativity theories (the special and the general), the non-Euclidean geometries (elliptic and hyperbolic) and geometric topology made a container conception of space and space-time obsolete? In my opinion, Graham Nerlich has convincingly demonstrated that this is not the case (Nerlich 1994, 2009). Mainly, I can here only recapitulate Nerlich’s conclusions, since the non-Euclidean geometries, the relativity theories, and geometric topology bring with them quite other kinds of complexities than Kant’s handedness argument does. For Nerlich’s detailed arguments and his counter-objections to objections, I can do no more than to refer to his own writings. First, in the wake of the scientific developments mentioned, Nerlich lays bare a missing presupposition in Kant’s argument, but he also describes how to take account of it. Geometric topology and elliptic non-Euclidean geometry have shown, in a purely mathematical way, that it is possible to think about spaces as being closed. That a space is closed implies that if a body in the space moves in what is called a geodesic (in Euclidean geometry, the geodesic is a straight line), then the body may come back to the point it started from. And the general theory of relativity has shown that such mathematical constructions can be of relevance in physics, too. And, as Nerlich notes, in some such closed spaces handedness disappears. The pedagogical device used in order to understand the notion of a closed space is to look at finite and closed two-dimensional spaces (surfaces) from a three-dimensional perspective, and then make the thought experiment that one is a two-dimensional creature living in the surface. Two such surfaces are by Nerlich used as illustrations. First, the so-called Möbius strip; take an ordinary strip cut it, turn one of the ends

110     I. Johansson

around 180 degrees, and then connect the ends again and you have a Möbius strip. When a b in such a strip is moved along the surface of the strip until it comes back to its starting point, then it has become a d. In such a strip, a b and a d are still locally two-dimensionally incongruent, but they are globally congruent. Such a space, however, can be dismissed as a physical space, because it has edges, i.e., lines where movements (those orthogonal to the strip) have to come to a stop. Physical space cannot be regarded as having such boundaries, since then one has to answer the question “what is on the other side, mustn’t there be some kind of space there, too?” Nerlich calls the Möbius space a pathological space, and if spaces like this would be the only ones where handedness does not exist, the handedness argument would not be threatened. However, the surface and two-dimensional space of the so-called Klein bottle cannot in the way mentioned be dismissed. This finite surface has no edges; in whatever direction a particle in the surface is moving, it can continue to move. And here, as in the Möbius strip, a b can be moved in such a way that it comes back as a d. That is, even in the two-dimensional surface of the Klein bottle, a b and a d are locally two-dimensionally incongruent but globally congruent. From a four-dimensional point of view, there can be threedimensional Klein bottle “surfaces.” And in these a left hand can be moved in such a way that it becomes a right hand. If our world is like this, it would in principle be possible to make a journey in which one becomes turned into one’s own mirror image where one’s old right hand has become one’s left hand. Nerlich calls spaces like the surfaces of the Möbius strip and the Klein bottle for non-orientable spaces, whereas Euclidean space is orientable. In this terminology, it has to be concluded that the handedness argument is valid only in orientable spaces. Whether handedness is possible or not depends on what kind of spatial paths the structure of the space in which the handedness exists permits. True as this conclusion is, it does not show that the container conception should be deleted. It seems reasonable to think that the spaces talked about in geometric topology are all of the same kind. It is odd to think that non-orientable spaces would be relationist spaces, but

5  A Presentation and Defense of Anton Marty’s Conception of Space     111

orientable spaces be container spaces; and as far as I know, no scientist or philosopher has put forward such an idea. Therefore, since orientable spaces are possible and by means of the handedness argument are shown to imply a container conception and refute a relationist conception of space, even orientable spaces should be regarded as container spaces. As Nerlich remarks: “we do not know whether the space we live in and take for granted is orientable or not” (Nerlich 2009: 610). Kant talks about incongruent counterparts. When it is claimed that they cannot possibly occupy the same spatial region, it is assumed that their sizes are not changed and that their shapes are not changed by transformations such as stretching, compressing, bending, and twisting; not even if the transformations are continuous. Such transformations are studied in geometric topology. For example, an ellipse can by means of continuous compressing be turned into a circle, but a square cannot. Kant’s notion of incongruent counterparts does not allow topological operations. A fact which raises a new question: can perhaps geometric topology in some other way than that already discussed—and dismissed—threaten the handedness argument? Of relevance for the handedness argument is that if the hands are allowed to be stretched, then a right hand can be turned into a left. “For instance, both a left and a right hand will snugly fit a latex glove since it will stretch to fit either” (Nerlich 2009: 606). Taking such possible shape changes into account, and calling the incongruence spoken of so far geometric incongruence, I would like to define a notion of topological incongruence as follows: Two figures are topologically incongruent = def they can by means of no continuous changes of size or shape be made congruent. In my opinion, the geometric notion of incongruence is sufficient for the handedness argument to be valid. But, of course, the argument becomes stronger if it can allow topological incongruence, too. And Nerlich has pointed out that it can. There are shapes, e.g., the trefoil knot, which preserve their handedness in any kind of continuous distortion. A trefoil knot will in orientable spaces never, independently of any topologically smooth distortion, match its mirror reflection. Because of this, Nerlich distinguishes between two kinds of handedness: ordinary and deep handedness. The trefoil knot, he says, is deeply handed

112     I. Johansson

(Nerlich 1994: 54–58, 2009: 606–607). Ordinary handedness rests on geometric incongruence, but deep handedness rests on topological incongruence. Nerlich casts new light also on the view that physical space is causally non-efficacious. According to him (and I agree), a distinction has to be made between two kinds of explanations; he calls them causal and geometrical, respectively. His use of the term “causal” conforms both to the Aristotelian notion of efficient cause and the Newtonian notion of force. In both kinds of cases, the cause and the effect belong to different spatial regions; be they in contact as in pulls and pushes, or at a distance as in the Newtonian explanation of the planetary movements. In geometrical explanations, this is not the case. In geometrical explanations, the explanandum (changes in a material entity) exists in the same spatiotemporal region as the explanans (a region of space/space-time). In the relationist conception of space/ space-time such explanations make no sense, since according to relationism space/space-time is nothing in itself. But in the container conception they are meaningful, since here bodies are always co-located with a spatial region. Next comes a brief attempt of mine to present Nerlich’s argument for the distinction. Explicitly, Newton’s first law says that the center point of a body in motion not affected by causes will continue in a straight line, but implicitly it is taken for granted that without causes neither will its shape have to change only because of the motion. As Nerlich shows, this is the case only because of the flat structure of Euclidean space.11 In nonEuclidean spaces, different parts of the space may have different curvatures, as for instance in the Klein bottle surface. Here, different regions of space can be differently structured, and because of this not be exchangeable with each other. In turn, this means that a body that fits one region need not fit another. Here is a central quotation from Nerlich: Helmholtz showed that only spaces of constant curvature permit free mobility; that is, if the space is variably curved then a thing would have to change its shape in order to move from one region to another of different curvature. It would not be freely mobile, therefore. […] Since we ourselves are reasonably elastic we could move about in a space of variable

5  A Presentation and Defense of Anton Marty’s Conception of Space     113

curvature, but only by means of distorting our body shape into nonEuclidean forms. (Nerlich 1994: 39)

According to Nerlich, it is not a causal factor but the geometry of space that explains why we have to distort our body in the way mentioned. Using Nerlich’s terminology, Marty’s view that space is causally non-efficacious can be retained in non-Euclidean spaces/space-times; even though Marty’s view has to be amended by a notion of non-causal explanations. The quotation above continues as follows: We would have to push to get our bodies into these [unoccupied] regions, for only forces will distort our shape. […] Let us suppose that the changes are noticeable and the effort to move into the hole perceptible too. Then we could feel non-Euclidean holes. […] If that does not show the hole concrete, what could possibly count as concrete. (ibid.)

Nerlich’s observation refutes Ernst Mach’s and the positivists’ view that all non-relationist conceptions of space are unscientific; meaning that they cannot possibly have any observable effects. Marty was well aware of Mach as a philosopher, but he never discusses him in RZ.

6 The Handedness Argument III: The Relationists’ Go-Up-One-Dimension Argument Refuted Put in very general terms, both Kant and Nerlich take it for granted that physical spatial and temporal dimensions are of a finite number. In this section, I will argue that this assumption seems to be true independently of possible developments in physics. The assumption has to be backed by arguments, since one line of defense from relationists is to say: handedness in an n-dimensional space always disappears in an (n + 1)-dimensional space. Let me present this relationist defense, and then refute it. I start by quoting the early Wittgenstein, who is a relationist, and who is aware of Kant’s handedness argument. He says:

114     I. Johansson

Kant’s problem about the right hand and the left hand, which cannot be made to coincide, exists even in two dimensions. Indeed, it exists in one-dimensional space - - - o–––x - - x –––o - - a      b in which the two congruent figures, a and b, cannot be made to coincide unless they are moved out of this space. The right hand and the left hand are in fact completely congruent. It is quite irrelevant that they cannot be made to coincide. A right-hand glove could be put on the left hand, if it could be turned round in four-dimensional space. (Wittgenstein 1963 [1921]: 6.36111)

I give this argument the label “the go-up-one-dimension argument.” But let it be noted that Wittgenstein makes a mistake, even though it can be rectified. The lines a and b are not truly one-dimensional, since both the endpoints o and x are two-dimensional. However, in order to get a true one-dimensional example, Wittgenstein could have exchanged the o:s for, say, a short yellow ending, and the x:s for a short blue ending. Then the figures with their internal color features could not be made to coincide by merely being moved along the line in which they are placed; in order for them to become congruent, one of them would have to be turned round 180 degrees in the two-dimensional surface in which they exist. In order to make my point conceptually more clear, I will now say some further words about the notion of two-dimensional incongruence that I used in passing in Sect. 4. As said, according to a classic definition in geometry, two geometrical figures that differ only in location are congruent. Two figures of equal size and shape (and with no relevant internal structure such as color spots) are congruent. If two congruent figures could be co-located in the same spatial region, they would occupy exactly the same region. There are, however, two kinds of such geometric incongruence: (i) between figures that are incongruent simpliciter and (ii) between figures that are n-dimensionally incongruent. The first incongruence relation is a two-term relation, but the second is a three-term relation; apart from the relation between the figures there is also a relation to a space.

5  A Presentation and Defense of Anton Marty’s Conception of Space     115

A square and a circle are incongruent simpliciter. According to geometric topology, a square cannot be turned into a circle unless three mathematical points are taken away. In whatever kind of n-dimensional space, they are placed they are incongruent. But, as said in Sects. 4 and 5, the two two-dimensional shapes b and d are two-dimensionally incongruent, but congruent in all spaces that have more than two dimensions. Let me return to the shapes of the left and right hands discussed. There are no motions that in three-dimensional space can turn the shape of the one into the shape of the other. The two hands are in the sense presented three-dimensionally incongruent, but, as Wittgenstein notes, they are four-dimensionally congruent; the hands can by “rotation” in a fourth dimension be made congruent. The argument can be generalized. In whatever n-dimensional space there is handedness, there will always be a (n + 1)-dimensional space that removes the necessarily n-dimensional incongruence. Therefore, many relationists conclude, whatever the true explanation of handedness is, positing a container space cannot be the explanation. My antirelationist reply builds on two observations. First observation, we are discussing physical space, not the abstract n-dimensional space constructions of pure mathematics. So far, no physical theory has—to put it very mildly—not even been close to assume that physical space or space-time contains an infinite number of dimensions. The general theory of relativity posits four spatiotemporal dimensions, ordinary string theory posits ten or eleven; and in the early very special bosonic string theory, which posits more dimensions than any other physical theory so far, there are twenty-five spatial dimensions and one temporal. I think that in all probability physicists will always regard physical space as having a finite number of spatiotemporal dimensions. Second observation, I am unable to prove it mathematically, but it seems to me clear, that the three-dimensional handedness shape can have four-, five-, and n-dimensional analogues. That is, in any ndimensional physical space there might be shapes that are necessarily n-dimensionally incongruent. Therefore, combining the observations, each future physical theory will work with a physical space that contains the possibility of housing entities with shapes that are necessarily incongruent. This, in turn,

116     I. Johansson

means that to regard physical space as a container is a much more reasonable option than to view it in relationist terms.

7 Some Concluding Words By bringing Anton Marty and Graham Nerlich together, and adding some reflections on the go-up-one-dimension argument, I hope the paper shows that a container conception of space/space-time has to be taken seriously even in the contemporary philosophy of physics. Acknowledgements   This paper was once part of a paper where also perceptual spaces were discussed. For a number of important comments of all kinds on any of the earlier versions, I would like to thank Jan Almäng, Ann-Sophie Barwich, Clare Mac Cumhaill, Olivier Massin, Kevin Mulligan, Graham Nerlich, Ingemar Nordin, and Kristoffer Sundberg. The participants in the conference “Anton Marty and Contemporary Philosophy” (Geneva, June 15–17, 2017) made me realize some pedagogical problems. For some appropriate critical comments on the penultimate version I would like to thank an anonymous referee. At last, a large extra gratitude goes to Kevin for thirty-five years of intermittent philosophical exchange. Without these discussions, not even the first version of this paper could have been written.

Notes 1. With respect to perceptual spaces, I have recently argued that they are sense-modality-neutral; see Johansson (2018). Marty has the traditional view that each sense modality has a perceptual space of its own: “The intuition of space [“Raumanschauung”] or – better phrased – the intuitions of spaces, since each sense has its own” (RZ: 234). 2. In many respects, Marty states his views on time clearly, but they are not as well worked out as those of space. For a presentation see Simons (1990), who regards Marty’s philosophy of time as “a torso” (ibid.: p. 169). Some of the ideas are unusual, for instance, that “time is itself a change” (ibid.: p. 162), and that “a temporal vacuum is possible” (ibid.: p. 160). 3. The conflation exists only in English. Meinong’s original German term for “subsist” is “bestehen.”

5  A Presentation and Defense of Anton Marty’s Conception of Space     117

4. This term is neither used in the expositions of Marty’s views on space given by Egidi (1990) and Smith (1990: Section 6), but my views are consistent with theirs. 5. He put forward this “space subjectivism” in his Inaugural Dissertation of 1770, Kant (1894: §15: D). 6. The notion of a receptacle that plays an important role in Plato’s Timaeus cannot be regarded as referring to a container space. Plato is of the opinion that voids are metaphysically impossible. 7. There is, though, apart from me, at least one other exception: Mac Cumhaill (2011, 2015). Let it be noted, though, that Nerlich’s views are discussed in Dainton (2001: ch. 14). 8. Nerlich has a similar but more qualified view, since he also discusses non-Euclidean spaces. He argues that non-Euclidean spaces can in a way affect bodies, but that this kind of affection is not causal; see Sect. 5. Euclidean spaces, however, cannot even on his account affect bodies. 9. What next follows in this section can be found spread out in Broad (1978: ch. 2.2.3), van Cleve and Frederick (1991), Dainton (2001: ch. 14.4–5), Nerlich (1994 [1976]: ch. 2), and Nerlich (2009). 10. It should be noted that if an outside-inside distinction is allowed for the shapes discussed, then, to take an example, a right glove can be made into a left glove by being turned inside out. In geometric topology, surfaces are regarded as infinitesimally thin, i.e., it makes no sense to speak of them as has having two sides such as an outside and an inside. 11. However, let it be noted that in the special theory of relativity shape is not a frame-invariant property. “What is a circle in one specific inertial frame is a non-circle in all other inertial frames” (Johansson 2011: Section 6). Nonetheless, there is what might be called a rest shape or intrinsic shape (Almäng 2015: 470).

References Almäng, J. 2015. A Note on Shapes. Journal of Philosophical Research 40: 469–471. Beiser, F.D. 2014. Late German Idealism. Trendelenburg and Lotze. Oxford: Oxford University Press.

118     I. Johansson

Broad, C.D. 1978. Kant. An Introduction. Cambridge: Cambridge University Press. Colyvan, M. 1998. Can the Eleatic Principle Be Justified? Canadian Journal of Philosophy 28: 313–336. Crombie, A.C. 1969. Augustine to Galileo 2. Harmondsworth: Peregrine Books. Dainton, B. 2001. Time and Space. Chesham: Acumen. Egidi, R. 1990. Marty’s Theory of Space. In Mulligan (1990), 171–180. Euler, L. 1748. Reflections on Space and Time. http://eulerarchive.maa.org/ docs/translations/E149tr.pdf. Accessed 14 September 2017. Translation of Réflexions sur l’espace et le temps, Histoire de l’Académie Royale des Sciences et des Belles Lettres, 4: 324–333. Johansson, Ingvar. 1990. Marty on Grounded Relations. In Mulligan (1990), 151–156. ———. 2004 [1989]. Ontological Investigations (2nd enl. ed.). Frankfurt: Ontos Verlag. ———. 2011. Shape is a Non-Quantifiable Physical Dimension. In Proceedings of the First Interdisciplinary Workshop on Shapes, ed. K. Hastings and B. Bhatt. http://ceur-ws.org/Vol-812/invited1.pdf. Accessed 14 September 2017. ———. 2018. Perceptual Spaces Are Sense-Modality-Neutral.  Open Philosophy 1: 14–39. https://www.degruyter.com/view/j/opphil.2018.1.issue-1/issue-files/opphil.2018.1.issue-1.xml. Kant, I. 1991 [1768]. On the First Ground of the Distinction of Regions of Space. In van Cleve and Frederick (1991), 27–33. ———. 1894 [1770]. Kant’s Inaugural Dissertation of 1770. New York: Columbia College. https://archive.org/details/kantsinauguraldi00kant. Accessed 14 September 2017. Mac Cumhaill, C. 2011. Seeing Through. PhD thesis at the University of Edinburgh. ———. 2015. Perceiving Immaterial Paths. Philosophy and Phenomenological Research, XC: 687–715. Marty, Anton. 1916. Raum und Zeit. Aus dem Nachlass des Verfassers herausgegeben von J. Eisenmeier, A. Kastil und O. Kraus. Halle: Niemeyer (Referred to in the text as RZ). Mulligan, K. (ed.). 1990. Mind, Meaning and Metaphysics. The Philosophy and Theory of Language of Anton Marty. Dordrecht: Kluwer.

5  A Presentation and Defense of Anton Marty’s Conception of Space     119

Nerlich, G. 1994 [1976]. The Shape of Space (2nd rev. ed.). Cambridge: Cambridge University Press. ———. 2009. Incongruent Counterparts and the Reality of Space. Philosophy Compass 4: 598–613. Simons, Peter. 1990. Marty on Time. In Mulligan (1990), 157–170. Smith, B. 1990. Brentano and Marty: An Inquiry into Being and Truth. In Mulligan (1990), 111–149. ———. 1994. Austrian Philosophy. The Legacy of Franz Brentano. Chicago: Open Court. van Cleve, J., and R.E. Frederick (eds.). 1991. The Philosophy of Right and Left: Incongruent Counterparts and the Nature of Space. Dordrecht: Kluwer. Wittgenstein, L. 1963 [1921]. Tractatus Logico-Philosophicus. London: Routledge & Kegan Paul.

6 Raum and ‘Room’: Comments on  Anton Marty on Space Perception Clare Mac Cumhaill

1 Space and Place At the very outset of his discussion of space, in the first part of Raum und Zeit, a work composed in haste at the end of his life and published posthumously by his students—the ‘feather’ fell from the author’s hand mid-composition (RZ: iv),1 we are told—Anton Marty introduces a number of ways in which ‘Raum’ (space), and the related term ‘Ort’ (place) can be used. At least some of these uses are centrally relevant to his exploration. More than that, they set the data for his treatise in a way to be made clear. Marty at once dispenses with two uses; a ‘social’ use which he concedes may even be primary, whereby ‘place’ is understood to be an ‘inhabited space’; and one which might be expressed in English as ‘terminus’—the end of some process or thing (for instance, the ground may be the ‘place’ for fallen things). He observes that sometimes ‘Ort’ C. Mac Cumhaill (*)  Durham University, Durham, UK e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_6

121

122     C. Mac Cumhaill

and ‘Raum’ are used interchangeably, sometimes synonymously. The example he gives is striking: “Der Körper ist an dem Orte” heißt ganz dasselbe wie “er ist in dem Raume”. (RZ, §1: 3)

There is a like parallel in English. ‘an dem Orte’ is naturally translated in English as ‘at the place’. But while ‘at’ may be distinguished from ‘in’ insofar as the former refers to a place idealised as a zero-dimensional topological point and the latter a place idealised as two-dimensional area or a three-dimensional volume (Wesche 1986: 385), we typically tend to think, and say, that things ‘in’ space, in being in space, are ‘at’ places. On this intuitive conception, a place is understood to be a spatial region of limited extent, a use of ‘place’ that Marty also catalogues and which goes hand in hand with the conception of space that I note in closing. For now, I can be brief. Places are countable and ‘spaces’ may be too. ‘Space’, however, is often used as a mass noun like sugar or rain; we may speak of there being more or less ‘space’ (notice, it is less felicitous to say more or less ‘place’). In everyday English, we also sometimes speak of there being more or less ‘room’, a term cognate with the German term ‘Raum’. Marty’s exploration evidences that his ‘Raum’ encompasses what we call ‘room’—a space for a purpose or a person, a space which is often enclosed. But the English word ‘space’, and certainly in recent analytic philosophical use, is less embracive. Finally, Marty introduces two analogous uses where ‘place’ and ‘space’ overlap. Empty places are those that can be easily passed through and filled, while ‘the space of a body’ refers to its cubic content or three-­ dimensional extent. Critically, Marty writes that the latter concept ‘presupposes a local positivity that is not identical with it [the body]’ (RZ, §1: 4). Let us call this the Naive Presupposition—‘naïve’ since it tallies, I think, with our pre-theoretic ontology of perceptual space. Or so I shall assume. An important assumption of Marty’s is the following. Marty takes it that what exists is the subject of true affirmative judgement—viz. a judgement with the content that ‘x is’ or, in this case, ‘space exists’. Such pronouncements are hardly every day. Less

6  Raum and ‘Room’: Comments on Anton Marty …     123

mysteriously, it seems we can truly judge of things that they are ‘in space’ or ‘at a place’—that an acrobat is suspended in mid-air ‘over there’, say. Marty’s project is to uncover the nature of this ‘what is’— space—and the spatial relations that, as he argues, it grounds, and he does so by arguing against a series of philosophical fathers, among them Descartes, Berkeley, Kant and his former teacher Brentano, a dialectical strategy that, as Simons (1990: 157) notes is by no means advantageous. In particular, Marty’s positive view, which is interpolated with polemic, is motivated mostly negatively. By sifting a number of positive claims from the earlier sections of Raum und Zeit—the section entitled ‘vom Raume’—I have a go at setting a Martyan picture of space perception against the backdrop of contemporary philosophy of perception. The paper unfolds as follows. In Sect. 2, I sketch two charges that Marty raises against Kant and Brentano. Both charges are descriptive and conceptual. Kant’s descriptive phenomenology is inadequate, Marty urges. ‘Form’ is in no sense prior to ‘matter’. But it follows from this exploration, in ways I will explain, that Brentano’s empirical psychology is also inadequate, and this despite its anti-Kantian flavour. From there, I outline Marty’s unusual ontology of space (Sect. 3). Marty has it that spatial relations are non-real but existent, causally inert relations that are grounded in space, which is itself non-real but existent. Objects do not inhere in space in the way properties inhere in substances. Rather, there is a primitive non-real relation of ‘fulfillment’ (Erfüllung) that holds between objects and places in space, which itself subsists.2 In Sect. 4, I consider whether any contemporary philosophy of perception is equipped to make sense of Martyan space perception, and I suggest that the most promising conception is Naïve Realism. I support my proposal by drawing on some limited remarks that Marty makes in a short correspondence with Husserl.3 I then outline a difficulty for this theoretical translation. Naïve Realism is a direct theory of perception which is often cast as relationalist: Perceptions are fundamentally conscious experiences in which the perceiver is directly acquainted with mind-independent worldly objects, events, and, for some, regions of space and intervals of time. Thus, S perceives O, just in case S stands in such a psychological relation of acquaintance. This relation is non-representational,

124     C. Mac Cumhaill

primitive, and, typically, is conceived as a relation of perceptual awareness. As we shall see, the worldly objects of one’s acquaintance partly, though constitutively, determine conscious character.4 For Marty, however, all relations are non-real. Further, insofar as they are grounded, they are not fundamental. But to this extent, it might be supposed that they ought not to be construed as brute or primitive either. With this in mind, we might wonder in what sense, if any, the perceptual relation that the Naïve Realist envisages might be conceived as non-real and what its real grounds could thereby be. I explore these matters in Sect. 5, before going on to describe a distinctively Martyan form of Naïve Realism, one which preserves the central theoretical tenet that phenomenal character is fundamentally constituted by worldly objects—Call this the Assimilation Thesis—but which regards the apparent relational structure of awareness as derived, for reasons to be made plain. There are two routes to explain the derivation of the structure that the standard Naïve Realist invokes in her talk of a relation of awareness: one is psychological and necessary, the other is artefactual. I make headway in spelling out the latter by bringing Marty into fleeting conversation with another Thomist—G.E.M. Anscombe (Sect. 6).

2 Kant and Brentano—Dimensions of a Critique Marty’s Raum und Zeit has two parts; the first dealing with space, of which there are 33 numbered sections, the latter with time. My reconstruction considers only the first part. Of those 33 sections, a substantial number deal with the shortcomings, as Marty sees them, of central canonical figures—most notably, for my purposes, Kant and Brentano. Both of the charges that I see as relevant to my exploration (there are many others), Marty erects primarily on descriptive phenomenological grounds. Let’s start with the first objection.5 Kant is correct, says Marty, in maintaining that if anything sensory is given then so is space. For Kant, notoriously, this is since the form of the receptive faculty is such that, necessarily, anything that

6  Raum and ‘Room’: Comments on Anton Marty …     125

is phenomenally given is given in space. But Marty wonders whether from this descriptively true claim, Kant’s treatment of space as a subjective form of intuition is invited. Marty reads Kant’s transformation of Newtonian space into a subjective form of intuition as preserving Newton’s insistence on the unity, infinity and independence of space from that which fills it. In Kant’s case however, the independence criterion is reconceived in the following way: the form of intuition is ‘pure’—a priori and not dependent on quality. Marty questions this from a descriptive phenomenological perspective. It is ‘undeniable’, he says, that we can only abstract spatiality from quality; when we see colour we see extent, and it is only by abstraction that we can ‘abstract’ away from the peculiarities of colour and conceive of spatiality as distinct from quality. Further: Immediately and vividly we are only given the presentation of a space filled with quality. (RZ, §4: 8)

But, this being so, it is ‘factually above all doubt’, says Marty, that our intuition of space is not quality-less, or ‘pure’ in the sense specified above, and nor are we given an infinite space—Marty takes it as just obvious that our spatial vision is finite, and even very limited.6 Again, his point is descriptive, but he augments it with a genetic or causal claim: no purely empty space could affect our senses. This does not show that space is a subjective form of intuition however. It only shows that just as colour cannot be sensed without extent, the converse also holds. Space cannot be intuited without localised quality.7 Prima facie, this might be thought to suggest that Marty must deny that intuited space retains aspects of the Newtonian conception—most modestly, the thought that space is something over and above its contents, or that which fills it. As we will see however, though Marty agrees with Kant that whenever something sensible is given so is space, and while he argues that space cannot be given without quality—deriding, as his friend Carl Stumpf also does, Kant’s imaginative ‘subtraction argument’, something which I leave aside8—he rejects a relationist conception of space, and not only on descriptive phenomenological grounds, but also,

126     C. Mac Cumhaill

it seems, on conceptual grounds too. Thus, we find Marty writing in an intriguing footnote: Kant considers the presentation of an ‘empty space’ as a posteriori. If one understands by this the presentation of an empty space outside the world, this only follows. For the presentation of a world is certainly empirical, and so also is the [presentation] of adjacent space, not filled with bodies, that would have common boundaries with that world, since the adjacent, as such, cannot be presented without what borders it. (RZ, §6: 17, fn. 1)9

It is notable that Marty refers here to the presentation of an empty space, the singularity of which must, in Kantian terms, be a posteriori. This claim is difficult to make sense of, but one suggestion is the following: such a presentation must be a posteriori since an empty space is one that is not filled with body. It is hence not ‘empty’ in the sense that a mere form of intuition is, viz. contentless, or without sensible matter. If correct, this suggests a further tempting line of thought. C. B. Martin holds that the provision of the limits of the being of presences (things like pens and bicycles) requires the presence of absence outwith those limits—the presence of absence at places where those things are not or empty regions.10 But if so, and if part of the concept of a body is that of a limited whole, then the concept of an empty space in the first sense (viz. not in the sense of a form of intuition) attends, or is part of the structure of, our concept of a body. More explicit is Marty’s critique of Brentano, also on conceptual grounds—and here, as Smith puts it, ‘Marty seeks a position more commonsensical than that of his master, even at the price of a certain sort of theoretical inelegance’ (1990: 129). I detail Marty’s ‘inelegant’ theory in Sect. 3. First, some comments on his divergence from Brentano. Like Marty, Brentano also maintains a form of spatial nativism— namely, the thought that space is given as part of the originary content of experience: it is not a form of intuition, as Kant thought, and nor is form (typically the form of objects) constructed from bundles of sensation, as an empiricist might hold. Rather, colour continua are founded on spatial continua, the spatial extents they ‘colour’. Here ‘founded’ is a technical term that we can gloss simply, for the purposes of this paper,

6  Raum and ‘Room’: Comments on Anton Marty …     127

as one-sided dependence.11 Since colour is founded on spatial continua, we cannot see colour without seeing extent. For most of his life, Brentano maintained that spatial continua are in turn secondary to, or founded on, temporal continua, where this entails a concomitant rejection of spatial absolutism; what is spatial coincides with what is corporeal and only insofar as bodies persist through time can we say that the three-dimensional spatial continuum that they constitute persists also.12 In later writing on space, however, specifically in the Categories, Brentano appears to grant that places are substances which may or may not be filled. For Smith (1990: 128), Brentano makes this shift so as to be able to give an account of what individuates otherwise qualitatively identical things. Thus, ‘[t]wo dots of identically the same red are individually different only because one is here, and the other there’ (Kat, 247, Eng tran., p. 177). But, unsurprisingly, such an individuation condition makes movement impossibly rare; if things are individuated by the places at which they are, no thing can move place and yet be the same.13 Marty rejects both these Brentanian theses, the earlier and the later. Philosophy, he says, has to ‘exercise its office by stubbornly and repeatedly, pointing out the questionable, even the impossibility and absurdity of certain conditions’ (RZ, §1: 72), words he borrows from Hermann von Lotze, his former dissertation supervisor (an appeal that is striking in the context of his dispute with his other mentor, Brentano). For Marty, that space is nothing but the ordering of bodies is one such absurdity. But so is the thought that movement is something merely, and only relative or, worse, a fiction. I will assume that this rejection is something that Marty thinks follows from the very concept of movement. But this is not all. Recall the Naïve Presupposition has it that we take bodies to be at places of three-dimensional extent, where such places are positivities that we may say are filled—indeed, as we shall see, Marty takes the relation of ‘fulfilment’ as basic. The relationist about space can only reductively identify such places with respect to other bodies in the web of corporeal bodies that constitute the spatial continuum. But, as such, the place at which an object is, for the spatial relationist, somehow ‘external to it’; it is individuated with respect to other bodies in the nexus, something that, for Marty, seems to run counter

128     C. Mac Cumhaill

to how things strike us. How so? For if this were the case, he seems to ask, what could be the meaning or essence of what might be called the in-dwelling, or in-space being of a body (Innewohnen oder Im-RaumSein der Körper [RZ, §15: 76])—the very space or place which the Naïve Presupposition presupposes that things in space occupy or fill? Marty characterises the alternative, reductive strategy of his opponent as follows: To say that between this and another body (atom) is an empty space is only a pictorial way of speaking which, by a fiction of inner form of language, puts something positive in the place of something negative. The correct negative form of expression would be that no other body is between bodies X and Y at any distance from them. Similarly, instead of: from a certain place, empty space expands (in the infinite), the facts can be stated more reasonably as: beyond a certain distance, no body is found anymore. (RZ, §18: 88)

Marty does not accept this negative thesis. I sketch his positive proposal below. For now, it is worth noting that Brentano appears, if not to revert to his former position at the end of his life, then to equivocate. Only weeks before his death, on 23 February 1917, a year after the publication of Raum und Zeit, Brentano dictated the essay ‘What we can learn about space and time from the conflicting errors of philosophers’. Brentano, it seems, wants to reclaim the thesis that empty space is a fiction: It has been said that if a body is to move then there must exist an empty space into which it moves. This is just as compelling as if someone were to say that, if something to change colour, there must already exist a colour which it then takes on. (quoted in Smith 1990: 169)

Adding in a footnote: And indeed why not also: if someone is to enter into marriage, then this marriage must already exist beforehand? (ibid.: fn. 3)

Marty’s name is mentioned six times in this discussion.

6  Raum and ‘Room’: Comments on Anton Marty …     129

3 Marty’s Radical Conception We have seen that for Marty space is not a subjective form of intuition. Mind-independent space would exist in the absence of anyone perceiving it (see Johansson, this volume). Nonetheless, when space is perceived, it is given with the sensible. Further, just as one cannot see colour without seeing extent, one cannot see space without seeing sensible bodies. Importantly, however, this does not entail, pace the early Brentano, that space is to be identified with sensible bodies and the relations in which they stand to one another—it does not mean that ‘empty space’ is a fiction. So, what, then, is Marty’s radical alternative? For Marty, empty space, at least on one understanding, is space that is not filled, where space is a positivity, or, more precisely, as I explain below, non-real existent (see also Johansson, this volume). Prima facie this might seem to align with Brentano’s later position, but this is not so. For Marty, whatever can be the subject of a true affirmative judgement exists, where what is non-real in addition is non-real insofar as it cannot enter into causal relations. After Smith, let us call those objects that can enter into causal relations, and which Marty designates as ‘real’, energetic objects. Non-real entities are, by contrast, anergetic objects, among which are included collectives, states of affairs, values and, as we shall see, relations and space. Now, that space is anergetic might tempt one into supposing that it is ideal and subjective. But Marty holds that it cannot be subjective for the nativist reasons he offers in critique of Kant. Yet nor can it be objective only through being identified with real bodies, as the earlier Brentano proposes. This not only runs counter to what I am calling the Naïve Presupposition, but leads too to an unacceptable scepticism, or at least reductionism, about movement. And once space is admitted instead as a substance that individuates property instances, this scepticism only grows. It is this latter Brentanian proposal that Marty’s proposal definitively wants to avert. Substances are real for Marty, and they individuate property instances. Since movement is possible however—and here is Marty’s common sense in action—space cannot be said to individuate otherwise

130     C. Mac Cumhaill

identical things. Space then is not a substance, so defined. Rather it is, says Marty, a subsistent. Subsistents are like substances insofar as they do not ‘inhere’ in anything, but unlike substances subsistents are non-real. Thus, while property instances (accidents) inhere in substances, objects cannot be said to ‘inhere’ in space. Rather, Marty takes the relation of objects to space as brute and basic—they are said, again in the line with common parlance, to ‘fill’ it (we have called this relation ‘fulfilment’; Johansson, this volume, calls it ‘space-filling’; the German term is ‘Raumerfüllung’). With this much spelt out, we can finally detail out the nature of spatial relations for Marty. Spatial relations hold between positions in space: As far as location [Ortlichkeit ] by itself is concerned it is uncontroversial that the relations of being outside and of being side by side are grounded local relations that presuppose absolute places as their grounding fundamenta. (RZ, §7: 24)14

Like all relations, spatial relations are themselves non-real. Nonetheless, they are grounded in space, itself an objective non-real existent (see Johansson 1990 for discussion).

4 Marty, Husserl and a Contemporary Translation Let us gather together some claims that have been sketched, and grant that, from the contemporary perspective, a rather exotic picture emerges. – Space is existent; it is the subject of true affirmative judgement. – Its manner of existence is that it subsists; nothing inheres in it. – Space is a positivity. – Space is anergetic. – Space is non-real. – Space is not a quality-less form of intuition. – When something sensible is given so, necessarily, is form.15

6  Raum and ‘Room’: Comments on Anton Marty …     131

– Objects fill space, where the relation of objects to space is sui generis—it is one of fulfilment. – Space cannot be intuited without the presentation of something sensible. To make this latter point a little clearer, let us distinguish subsistent space from substantial form. Unlike Brentano and Stumpf, Marty doesn’t think that we only perceive substantial form. The relation of substantial form to subsistent space is, recall, one of fulfilment (space-filling), where empty space is subsistent space that is unfilled. Can we assume, therefore and further, that Marty supposes that empty space can be perceived? Two considerations might be offered here. First, Marty’s opening exploration appeals directly to empty space. The concept of an empty space is of a region that can be passed through. Such spaces can be subjects of true affirmative judgement. Second, the concept of an empty space is not the concept of a pure form of intuition, but, arguably, the concept of a region that is empty of body. Since, however, Marty denies that space is to be identified with sensible bodies and the relations in which they stand to one another, even while he holds that we cannot perceive space without also perceiving something sensible, it is arguable that he ought to be prepared to grant that we perceive empty space. Why? Since he allows that we perceive space tout court; both filled and unfilled (to wit: empty) spatial regions. I consider this line of thinking attractive, and it guides what follows.16 However, further work is needed to establish Marty’s final view on the possibility of our perceiving empty space—a question over which Brentano, Stumpf and Husserl all waver. To this extent while the explorations that follow are Martyan insofar as they are in the spirit of Marty, they might not be wholly to the letter. My task in the remainder of the paper is to consider Martyan space perception, thus understood, in the light of contemporary philosophy of perception. I take this to be a useful exercise since Marty, as I read him, offers a host of conceptual resources that suggest ways of reframing questions about space perception and perceptual experience more generally, in ways I hope to showcase, if only in a schematic way.

132     C. Mac Cumhaill

The explorations that follow then are rough. Even so, it strikes me that two lessons already follow from what I have drawn out so far. Setting these out helps frame the course of the rest of the paper. First, for Marty, space is a non-real existent, not a subjective form of intuition. This being so, it might be thought that any philosophy of perception that endorses the thought that space itself is a subjective form of intuition—rather than something that can be presented under a subjective or ideal mode of presentation say—is non-Martyan.17 Second, since space is non-real on this Martyan picture, it is anergetic. Hence any philosophy of perception which necessitates a causal theory of perception, even where this is treated counterfactually, or one whereby our receptivity extends only to sensible matter, is also at odds with Marty. These two lessons touch on Marty’s anti-Kantian misgivings on the one hand, and his anti-Brentanian considerations on the other. Space is not ideal (anti-Kantian), and nor can it be identified with corporeal, substantial bodies (anti-Brentanian). On the assumption that space can be an object of experience, however, rather than a subjective form of intuition, and on the assumption that space is something anergetic that exists in addition to the objects it fills, a question is suggested. What philosophical theory of perception could embrace a Martyan space perception, this much assumed? Begin with orthodoxy. Many forms of standard representationalism, where this is taken to be a thesis concerning the fundamental nature of experience, should run into difficulty on the second count—viz. where the anergetic nature of space is acknowledged. This is since veridical experience is typically taken to be experience that has been caused in the right way, where here the appeal is to experiential states and their causes. In addition, externalists emphasise a history of causal interaction with properties or kinds represented. But this being so, the inefficaciousness of space should make problematic the individuation of certain aspects of spatial content.18 These twin difficulties suggest to me that we should leave representationalism aside, and I do so for the most part of the remainder paper. But the standard palette of philosophical theories of perception is more colourful—it includes forms of relationism, as well as adverbialism. In

6  Raum and ‘Room’: Comments on Anton Marty …     133

what follows, I will suggest that it is plausible to think that Martyan perception embraces some form of Naïve Realist relational theory; how we should characterise this is something I turn to in the next section. Before that however, it is worth querying whether the kind of structure of experience that the sense-datum theorist envisages, also a relationist of sorts, could plausibly apply to experience of space, including, critically, experience of empty regions. For the purposes of this paper, I will treat the philosophical concept of a sense-datum functionally. Call a sense-datum whatever it is that we are immediately aware of in experience, in virtue of which, in non-hallucinatory perception, we are mediately aware of worldly objects and their properties.19 Assume, here, that experience has an act/object structure and, further, that demonstrative reference to worldly objects is secured in virtue of the experience of or sensing of sense data. After H. H. Price, call the relation between the sense-datum and that which is indirectly or mediately perceived in virtue of the sensing of the sense-datum ‘belonging’. Take it that belonging is a non-causal relation.20 Now, supposing that sense-datum ‘belongs to’ the worldly object sensed in virtue of the sensing of it (the sense-datum), we might ask: Could there be a space-presenting sense-datum in virtue of which nonreal objective space is perceived and could be demonstrated? If my reading of Marty is on the right track, the relevant sense-datum could not be ideal or subjective, though it could plausibly be private (cf. Johansson 2018). Importantly, however, it cannot be entirely insensible. The notion of an insensible sense-datum is hardly a happy one, but, as we have seen, Marty anyway insists that space cannot be perceived in the absence of the perception of some local quality. Yet even if this much is granted we might wonder: How could a space-presenting sensedatum belong to its object? Belonging is perhaps no more mysterious in the case of space perception than it is in the case of objects. Still, from the functional perspective we have adopted above, belonging, whatever its nature, must allow for the kind of cleavage on which arguments from conflicting appearances to the existence of sense-data spin. That is, belonging ought to be consistent with the putative immediate objects of experience having properties that, as we may say, ‘conflict’ with the properties of the

134     C. Mac Cumhaill

worldly items mediately perceived in virtue of the sensing of sense data. Thus, we want to allow that an elliptical sense-datum can ‘belong’ to a circular coin. It is not clear that Marty would be willing to admit such disparity or conflict between entities that can be said to putatively ‘belong’ together. Why so? As Smith (1990: 137) explains ‘consciousness, for Marty, is itself just a variety of assimilation of mental processes to (real or nonreal) objects in the world’. He likens Marty’s position to that given by Aristotle in De anima, quoting the following passages: What has the power of sensation is potentially like what the perceived object is actually; that is, while at the beginning of the process of its being acted upon the two interacting factors are dissimilar, at the end of the process the one acted upon has become assimilated to the other. (De anima, 418 a 2ff.) Within the soul the faculties of cognition and sensation are potentially these objects, the one what is knowable, the other what is sensible. These faculties, then, must be identical either with the things themselves, or with their forms. Now they are not identical with the objects; for the stone does not exist in the soul, but only the form of the stone. (De anima, 431 b 26ff.)

Commenting on the parity, Smith writes: Similarly, now, for Marty, all psychic activity is a process which has as its consequence that the psychic activity comes into a certain sui generis sort of conformity with something other than itself. (Ibid.)

Smith’s exegesis makes it plausible to suggest that the metaphors of ‘conformity’ and ‘assimilation’ are more aptly applied to Martyan perception than the metaphor of ‘belonging’, where conflict or non-conformity is even implied or permitted. But, if so, then this suggests that Marty’s position is quite different from those versions of sense-datum theory that lend themselves to being characterised functionally in the way I have above. Instead, it seems closer to a distinct form of relationism: Naïve Realism. This is (in part) the view that:

6  Raum and ‘Room’: Comments on Anton Marty …     135

[T]he objects we are consciously acquainted with in perceptual experience are constitutive of conscious character: the aspects of the world that we are acquainted with in perceptual experience constitutively ‘shape the contours of the subject’s conscious experience’ (Martin 2004: 64). Perceptual experiences have the characters they have because of, or in virtue of, the nature and character of the mind-independent objects they involve. (French, 2018)

Beck (2018) notes a list of adherents: Brewer (2011: 100), Campbell (2002: 116), Fish (2009: 49–50), French (2014: 395–396), Logue (2012, 212), Martin (1998, 173–175) and other naive realists all hold that the items a subject perceives in having a perception constitutively shape the perception’s phenomenology.

There are, I propose, two features of this understanding of Naïve Realism as a thesis about perceptual phenomenology that bring us closer to a Martyan space perception (for further discussion see Beck [2018]). First, on Naïve Realism conceptual space may be made for anergetic objects. This is since there is no requirement that the aspects of the world that shape the contours of consciousness be real or energetic. Thus, arguably shadows can shape the contours of consciousness. Second, it appears that insofar as worldly objects ‘shape the contours of the subject’s conscious experience’, conscious experience can be understood to ‘assimilate’ its objects, where here assimilation means the ‘taking in’, or ‘picking up’, or as Naïve Realists sometimes put it, the involvement of objects in experience. Call the Assimilation Thesis the central theoretical Naïve Realist tenet that the aspects of the world that we are acquainted with in perceptual experience constitutively shape the contours of the subject’s conscious experience. Strictly speaking, we can make sense of involvement without Assimilation and a theorist may be committed to the former but not the latter; the metaphysical structure of experience is such that it just ensures that objects are involved. Nonetheless, if we read Naïve Realism as a thesis about not only the nature of experience but the nature of conscious character too, and if it

136     C. Mac Cumhaill

is granted that Assimilation entails involvement, then it is fair to pit the Assimilation Thesis as a central tenet. Now, there is reason to think that the theoretical puzzle that I used to partly frame this chapter can be dissolved on such a Naïve Realism. I asked: If space can be an object of experience rather than a subjective form of intuition, and if it is something anergetic that exists in addition to the objects it fills, what philosophical theory of perception could embrace a Martyan space perception? Naïve Realism is a contender since on Naïve Realism it can be granted that non-efficacious entities, including empty regions, can shape the contours of consciousness, such that the phenomenal character of experience is fundamentally constituted by those entities and regions, recognising, as Marty does, that we cannot perceive space in the absence of the perception of local quality. Necessarily, when we see empty regions we see sensible objects. Both are, to borrow a term from Husserl, co-seen.21 This congruence notwithstanding, however, the theoretical translation is not seamless. Let us explore why. For most Naïve Realists, the Assimilation Thesis goes hand in hand with a further doctrine. Cast neutrally, this is the idea that perceiving involves the obtaining of a perceptual relation whereby the perceiving subject ‘stands’ in a perceptual relation to mind-independent worldly objects. For most Naïve Realists, it is in virtue of the perceiver standing in a perceptual relation to mind-independent worldly objects that those objects can play their character constituting role—to wit, by shaping the contours of consciousness. Different theorists have different ways of articulating this idea. Some speak of a relation of ‘conscious acquaintance’; others refer to the obtaining of a ‘psychological relation of acquaintance’—both technical notions. Sometimes ‘conscious attention’ is appealed to, but very often the perceptual relation is spelt out as a relation of awareness. For instance, Soteriou (2013) writes (commenting on a passage from Moore’s ‘Refutation of Idealism’): Sensory experience somehow involves some kind of psychological relation of awareness, and not simply some psychological event, process, state, or property. In this context, to say that the relation of awareness is a psychological one isn’t simply to say that one of the relata of the relation is a psychological subject—a bearer of psychological properties—for there are

6  Raum and ‘Room’: Comments on Anton Marty …     137

also non-psychological relations that psychological subjects can stand in to things. The suggestion appears to be that when one has a sensory experience there obtains a distinctive psychological relation that one stands in to some sensory quality, where that sensory quality is not a quality of the psychological relation. The awareness of blue is not itself blue. (Soteriou 2013: 13)

Importantly, for many theorists, this relation is primitive. Now, there is reason to think that Marty would partly concur with this picture. In a letter to Husserl, written in 1901 from a ‘well-wooded and charming area of the Bohemian Mittelgebirge’, he iterates the following two points, responding to objections from Husserl22: 1. You [Husserl] argue: if an act is a relation to an object and if the object, as is normally the case, is transcendent, then on my view it would follow that the transcendent object exists necessarily. Thus, this view (that if a presenting exists then a presented something exists too) cannot be correct. But what does transcendent object mean? Doubtless an object that does not exist merely in consciousness. But a presented something exists. Whether this something exists outside consciousness has nothing to do with the presentation as such […] 2. Now another of your objections seems to be: according to me the object of a presentation of blue would not be ‘blue’ but ‘presented blue’. But the object presented in a presentation is in fact ‘the same object that is judged about in the corresponding judgment and loved in the corresponding state of love’. This object is just blue not the presented blue. I entirely agree with this assertion of yours. The object of the presentation of blue is: blue, not: presented blue. I suggest that 2 might be recruited against reading Marty’s position along sense-datum lines—after all, it is ‘the same object that is judged about in the corresponding judgement and loved in the corresponding state of love’. While 1 can be read as gesturing at the Assimilation Thesis and an associated relationism. This is so at least in the following sense. For the Naïve Realist, in order that worldly objects can play a character constituting, consciousness-shaping role, the perceptual relation must obtain. In such cases, the perceptual relation obtaining entails that those

138     C. Mac Cumhaill

objects that partly shape phenomenal character exist, but not that they exist necessarily. I pick up the further matter of ‘presenting’ below. For all that, however, there are significant differences between Marty’s position and the Naïve Realism I have articulated above. These can be brought into view by sketching a challenge for the Naïve Realist attracted by Marty’s unusual take on space. For the Naïve Realist, perception involves a primitive perceptual relation. But for Marty all relations, and not only spatial relations, are unreal. If there is such a perceptual relation then, it is a non-real relation. What kind of Naïve Realism can be developed, starting from this assumption? I make some tentative suggestions in the next section.

5 Space for Naïve Realism? To begin, on a Martyan Naïve Realism, the perceptual relation ought not to be conceived as primitive in the sense of brute or unanalysable. The relation is non-real, as all relations are. Further, it is grounded. In Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, Marty writes that ‘the peculiar relationship of consciousness to its object’ is a special kind of grounded relation (quoted in Johansson 1990: 191). Johansson reads Marty as supposing that nonreal relations can be grounded in non-real entities like space and time, as well as in real entities. Start by considering only real worldly relata of the putative perceptual relation—for instance, a cherry tree. All Naïve Realists teach that when a perceiver sees something, a particular cherry tree say, she stands in a perceptual relation to that thing, where standing in this relation explains why the cherry tree, in this case, can play a consciousness-shaping role. It should be plain by now, I hope, that merely being a psychological subject isn’t sufficient to field the subject-end of the relation since such a subject can stand in manifold non-psychological relations to things—as when asleep and the cheery tree is outside the window. Rather, the relevant relatum must be a psychological subject in a certain conscious state or in whom certain conscious occurrences are unfolding.23 I suggest that Marty assumes this much with the concept of ‘presenting’:

6  Raum and ‘Room’: Comments on Anton Marty …     139

Presenting […] is a real process in the mind. In case there exists that which one calls the presented, then as a non-real consequence of the process it follows that the presenting mind stands to this thing in a peculiar relation, which might be described as an ideal similarity or adequacy. (U: 406, quoted in Smith 1990: 125)

And when ‘presenting’ occurs Marty writes: What really exists within us is not a peculiar, modified double of the real object, but only the real psychic process to which in certain circumstances there becomes attached as consequence an ideal similarity with something other, existing independently of this process. (ibid.: 415f., quoted in Smith, ibid.)

Importantly, on this view then, to say that the experience ‘has an object’—the presented—is not to suppose that the presented is a ‘double’, some kind of sensory image of the real object say. For in circumstances where ‘there exists that which one calls the presented ’, the object that the experience has, is the object that exists independently of the presenting—to wit, the worldly object. Accordingly, the experience ‘has an object’ necessarily in the sense that the present existence of the object is a requirement on the existence of process that is the presenting of an existing object.24 Support for this reading comes, I think, from a distinction that Marty makes between the relations of correlation and relative determination. The relation of correlation entails the coexistence of its relata (Mulligan 1990b: 19)—though, as we have seen, in the perceptual case, the subject-end of the relatum cannot simply be a psychological subject, but one for whom there is a presenting of an existing object, a presenting which thus has an existent object necessarily but which does not in turn necessitate the existence of its object. But experiences can also ‘have an object’ in a different sense (see Egidi 1990 for discussion). Experiences are often said to ‘have an object’ insofar as they are typically taken to be intentional. Insofar as experiences are taken to be intentional however, they may also seem to involve a relational structure, at least insofar as they are said to be objects of

140     C. Mac Cumhaill

mental states of various sorts. It is notable that for Brentano this apparent relational structure is only apparently relational; in fact, it involves a ‘complex of presentations’ which involves instead ‘relative determinations’ (Smith 1990: 126), where the complexity of the relevant presentations is relative in the following sense. Brentano supposes that where one thing is thought of (or experienced) relative to another, that which is thought of or experienced directly is said to be thought of or experienced ‘in modo recto ’, while that which is thought of or experienced relative to what is thought of or experienced directly, is thought of or experienced ‘in modo obliquo ’—viz. relative to that which is thought of in the direct mode. Importantly, in cases of relative determination that which is thought of or experienced in modo recto must exist, if that which is relative is to exist. But that which is presented in modo obliquo need not exist (except in certain cases). But as such, on this understanding, an experience can have an object in the second sense (the intentional sense), while not having an object in the first sense (the transcendent sense). Chisholm (1990: 2) details Brentano’s application of this distinction to sensation in a 1914 manuscript: [Brentano] makes two remarkable statements. The first is: ‘In sensing I am the sole object that is presented in recto [das einzige in recto vorgestellte Objekt] ’. The second is: ‘The thing that we have as external object is sensed only in obliquo…. It is sensed as sensed by us’.

As we have seen however, Marty appears to insist that presenting of an existent object involves correlation—the necessary coexistence of its relata. As such, this demands the existence of the object of experience in the first, transcendent sense, and not merely as a relative determination within a complex presentation. Now, it might be objected that the obtaining of a relation of correlation is consistent with contemporary forms of representationalism that likewise deny the existence of an internal ‘double’ of the real object— I am thinking here of what Fish designates as ‘strong’ versions of representationalism (2010: 67)—and insofar as Marty often speaks of correlation in terms of ‘adequation’ or ‘correctness’ this reading might seem supported. Once it is recalled, however, that Marty seems to hold that

6  Raum and ‘Room’: Comments on Anton Marty …     141

space, as a non-real objective entity, is a possible object of experience in the first sense, it is, I think, possible to resist this interpretation. Why? Space is anergetic. There is then a question for the representationalist as to how space can ‘get into experience’—viz. be represented—without itself having any causal impact on the subject.25 One promising strategy, surely, is to appeal, simply, to relations between things and to represent those relations. But it not clear that Marty would find the mere representation of spatial relations between things satisfactory. This is because he seems to think that the spatial relations are themselves also presented to us in experience (bearing in mind, recall, that these are grounded in space). He writes: We find them [spatial relations] there before us, and if this were not so, if they were a product of our psychic activity, then how would things stand with regard to the objectivity of our entire knowledge of nature […]? (U: 468, quoted in Smith 1990: 126)

However, it is plausible to think that, for the representationalist, spatial relations are precisely not ‘there before us’ but are only represented as ways things are arrayed or stand relative to each other and the perceiver. But, as such, spatial relations cannot be presented in experience directly and nor is there any possibility of their playing a character-shaping role. This former point is in essence the criticism that Marty wages against Brentano’s reduction of non-real spatial relations to relative determinations in complex (intentional) presentations (see Egidi 1990 for discussion). Johansson (2014) suggests a way of transposing this objection to the contemporary scene, formulating Naïve Realism as the conjunction of the following four broad theses: (a) the perceiving subject and the perceived object are two distinct entities where none is part of the other26; (b) there is a distance between the subject and the object, but the subject and the object are nonetheless in some sense connected; (c) the distance between the subject and the object is empty; (d) there is a relation of directedness (the arrow) from the subject to the object.

142     C. Mac Cumhaill

He notes: The term ‘distance’ adds something important to the view that, necessarily, consciousness is consciousness of something. The conjunction of the statements (a) and (d) says that perception is always perception of something, but it does not bring in any notion of distance; this is done in (b) and (c). (RZ, 3)

But, for a Brentanian, arguably, (b) and (c) are subsumed into (a) and (d) inasmuch as the ofness (a) and directedness (d) of intentional experience might be supposed sufficient to capture the phenomenology that (b) and (c) articulates. For the contemporary representationalist in addition, experience might be said to be ‘of ’ or about distance—the empty distance between the subject and object say. Such regions must be represented. But for Marty, who resists the ideality or subjectivity of space as a form of intuition, insisting too that space is anergetic, such regions are, and are presented as being, there, before us.27 Now, so far, I have suggested reasons for thinking that the Martyan position I am constructing should be found sympathetic to contemporary Naïve Realism, specifically insofar as it seems to endorse a correlational account of the relation which requires the existence of, and is dependent on, the worldly consciousness-shaping objects that such experiences thereby have. On this understanding, perceptual correlation involves Assimilation. As noted, correlation is a non-real relation. Earlier, however, I noted that most Naïve Realists characterise the perceptual relation as one of awareness. I now want to suggest that this spells trouble for the attempt at a theoretical transposition.

6 Awareness Versus Correlation To begin, we ought to grant that perceptual awareness requires more than perceptual correlation. It requires relative determination insofar as it requires that the experience or episode of perceptual awareness have an object in both senses detailed above. Second, perceptual awareness is typically understood as reflexive in the sense best characterised by quoting directly from Moore’s ‘Refutation of Idealism’:

6  Raum and ‘Room’: Comments on Anton Marty …     143

To be aware of the sensation of blue is not to be aware of a mental image of a ‘thing’, of which ‘blue’ and some other element are constituent parts in the same sense in which blue and glass are constituents of a blue bead. It is to be aware of an awareness of blue; awareness being used, in both cases, in exactly the same sense. (Moore 1903: 25)

That perceptual awareness involves awareness of awareness explains why sensory experience, so understood, can be meaningfully cast as a kind of knowing, as the Naïve Realist often insists—this follows from the claim that experience is direct acquaintance with mind-independent worldly things and events, etc. But that awareness could be a kind of knowing is also, one would think, explained by the fact that genuine perception involves what Marty calls correlation: it involves the presentation of an existent object.28 Now, although perceptual awareness requires more than perceptual correlation insofar as the former but not the latter is conceived to have an object in the second sense as well as the first, and insofar as the latter, but not necessarily the former, involves reflexive awareness, it is also the case that many Naïve Realists deny that we are aware of anything in hallucination. This might seem to tally with Marty since although perceptual awareness requires more than correlation, correlation demands the coexistence of its relata. If this requirement were ‘inherited’ by awareness then, it would follow that in the absence of the existence of the object apparently presented, as in hallucination, there could be no awareness. Marty, I suspect, would say differently. On the Martyan picture I am sketching, in hallucination, there may be relative determination without correlation. Cast differently, the hallucinatory experience does not ‘have an object’ in the first, transcendent sense, but it does in the second, intentional sense. It is worth harnessing G. E. M. Anscombe’s criticism of the Ordinary Language philosopher in her difficult paper ‘The Intentionality of Sensation: A Grammatical Feature’ to make sense of this. In saying that Marty seems to allow that hallucination involves the presenting of a non-existent object, it might be thought that the use of ‘object’ deployed here is that which is also admitted by the sense-datum theorist or Meinongian when they wish to preserve the intuition that in

144     C. Mac Cumhaill

such cases (or in thinking of entities like unicorns say) we are aware of, in experience (or in our thinking of ), some thing. But Marty does not seem to have this particular understanding of ‘object’ in mind, at least in the case of hallucination. This is because he seems to think that in these cases the meaning of ‘object’ is transferred from cases where predicates are applied to real objects; there is a kind of linguistic Bild, or fiction of inner linguistic form, as though: one were to allow the person portrayed to inhabit the portrait because of the similarity between the contours and colours of the painted bit of canvas and those of the portrayed body or face. (RZ, §12: 58)

This recalls G. E. M. Anscombe’s critique of both the Ordinary Language philosopher and the sense-datum theorist. Both theorists offer an ontological response to the question ‘what do you see?’ That is, they respond by giving the name of entities, things—everyday objects or, alternatively, sense data—as things seen. But while the Ordinary Language theorist denies that we see anything in hallucination—since what we see are ordinary things—the sense data insists that we see some thing, some data of sense! In contrast, Anscombe recommends a grammatical approach. The objects of sensation are not things, in the weighty ontological sense but direct objects of the sensation verb. Thus, we can intelligibly say that we see things even in cases where it is evident to all, including the speaker, that no such thing exists in the perceiver’s vicinity. How does this shift from ontology to grammar help the Martyan theorist, who is nonetheless tempted by Naïve Realism with respect to the nature of phenomenal character? Notice that when the Naïve Realist denies, as she typically does, that in hallucinatory experience we are aware of anything, in doing so, she preserves the sense of ‘thing’ that applies in the good, non-hallucinatory case—this is precisely why she denies that we are aware of anything in the hallucinatory case. It strikes me, however, that a Martyan Naïve Realist should want to urge that genuine perception involves correlation and that this is what grounds phenomenal character, but without deploying what is ostensibly a linguistic picture to articulate the metaphysical structure of experience (and which may in turn lead

6  Raum and ‘Room’: Comments on Anton Marty …     145

one to insist that in hallucinatory experience we cannot say that something is seen). To clarify: sensation verbs take direct objects. The verb ‘to be aware’ though not a sensation verb strictly speaking also takes a grammatical object. The Naïve Realist, in noting that in hallucination, by definition, there is nothing in the world that answers the subject’s description of what they ‘see’, and in supposing that phenomenal character is partly constituted by the worldly objects of which the perceiver is putatively aware, may be tempted to suppose both that nothing is seen—i.e. the subject is mistaken in her use of that term, that she is not thereby aware of anything—and, further, that the experience thereby lacks phenomenal character. The Martyan picture is more circumspect. It says only that hallucination involves the presenting of a non-existent object. But, if I am right, Marty does not want to read the notion of object here—an intentional notion—ontologically. Scholarly work is required to establish the scope and correctness of this claim.

7 ‘Raum’ and Room To conclude, I want to sketch briefly two positive lessons that fall out of my attempt to provide for a Martyan Naïve Realism. At the outset I noted what I called the Naïve Presupposition, which is just the perceptual datum that we do not take the spaces that bodies fill to be identical with them; and this is partly why it makes sense to say that things appear to be located at places. The Assimilation Thesis, recall, is the idea that perceptual conscious is fundamentally shaped by (or constituted by) worldly objects. We might not ask: Are these two conceptually or theoretically related? I think it is plain that they are not conceptually related. Nonetheless, if the generic characterisation I have given of Naïve Realism above is accurate, they are, at least, theoretically related. How so? For those theorists who think the perceptual relation is a relation of awareness, and where the possibility of Assimilation is explained by the subject’s standing in such a relation of awareness to the objects

146     C. Mac Cumhaill

of her experience, there may be a tendency to think that the Naïve Presupposition is likewise assured. Why so? Awareness is a non-solipsistic relation. P. F. Strawson famously proposed that a requirement on non-solipsistic consciousness is that a subject be able to conceive of things existing in the absence of her experience of them where this requires a conception of space. If having a conception of space is partially what grounds the phenomenology that the Naïve Presupposition is supposed to capture however, and where the possibility of awareness is secured when there is such a conception, awareness being non-solipsistic, then the Assimilation Thesis and the Naïve Presupposition should go hand in hand. If we take Naïve Realism to be a thesis, only, about the fundamental nature of phenomenal character however, a weaker form of relationism may suffice. What form this may take, I leave for some other occasion, suffice to say that unyoking the Naïve Presupposition from the Assimilation Thesis opens up explanatory paths not much explored in contemporary analytic philosophy of perception. This is one reason for taking seriously Marty’s peculiar take on space—one which, as I understand it, is nonetheless supposed to honour perceptual and linguistic data. Another is the following: I began this paper by detailing a variety of ways in which the words ‘place’ and ‘space’ can be used. I noted Marty’s inclusion in this list of the English word ‘room’, completely untheorised in contemporary analytic philosophy of perception—for various reasons, not unconnected with the point above, ‘room’ just vanishes from contemporary analytic theorising (the Brentanian school is thereby an informative counterpoint). Johansson, this volume, explains in what sense the concept of ‘a room’ helps isolate Marty’s conception of ‘Raum ’, with which it overlaps: A room is regarded neither as some kind of relations between the things in it, nor as a contingent structure inhering in the properties of the things there. Also, rooms can always easily be thought of as being completely empty, as three-dimensional holes so to speak; and some of them are also so perceived. Such a room is homogenous in the sense that all its different parts are regarded as exactly similar in their emptiness. (This volume, p. 104 )

6  Raum and ‘Room’: Comments on Anton Marty …     147

This gloss is illuminating. Empty rooms can be perceived—the empty space they enclose is co-seen with the walls which ‘line’ the room (recall, for Marty, we cannot see space without seeing localised quality). But there are a host of other uses that are relevant to the concept of space that contemporary analytic philosophers have almost wholly ignored. In English, ‘room’ can sometimes mean dimensional extent—the amount of space that is or may be taken up by a thing, event or process. In such cases, there may be ‘more’ or less ‘room’. Sometimes ‘room’ refers not to some dimensional extent but the capacity to accommodate a person or thing or to allow a particular kind of action; ‘room’ in this sense appears with modifying words such as ‘ample’, ‘enough’, ‘plenty of ’; sometimes there can be ‘no room’. Marty takes it that spatial relations are there: ‘We find them there before us’. Space is not a nexus of spatial relations between real things on this view, a conception of which is a condition on the possibility of awareness and so, on certain views, experience. Nor are experienced spatial relations determined relative to ourselves, where we are experienced in modo recto, and everything else in the oblique mode. Rather, on a Martyan understanding, as I have been telling it, there are existent ways in which our world is shaped, ways which in turn shape our consciousness. Anscombe famously thought that no action could fail to have moral significance—even the plucking of a single flower. Assuming perceiving involves perceptual activity, it is hardly plausible that this demanding and austere thought could apply to Raum qua ‘space’. Yet although, as Johansson urges, Raum is homogeneous in its emptiness, ‘room’ in some of the above senses, which is also there—room for a person, or for an event—is not. And this suggests, at least to me, an unexplored place for the theorist of perception to retreat to, and ‘dwell in’ with her theories.

Notes 1. Page numbers refer to the original edition text. 2. For discussion, see Johansson, this volume, as well as Johansson (1990), Rosaria Egidi (1990), and Barry Smith (1990).

148     C. Mac Cumhaill

3. Trans. by K. Mulligan and K. Schuhmann in Mulligan (1990a: 225–236). 4. Though intentional theories are sometimes presented as relational— experience involves an intentional relation to sensible properties (that may or may not be instantiated)—I leave explicit consideration of this construal aside. See, for example, Pautz (2010). 5. Marty’s work is untranslated. The exposition that follows is far from scholarly. I have been helped in my comprehension by the work of Chisholm, Johansson, Mulligan, Simons, Smith, and, particularly, Egidi in the 1990 volume ‘Mind, Meaning and Metaphysics’, edited by Kevin Mulligan. Hélène Leblanc and Mark Textor helped me with the translation of some puzzling passages. Special thanks to Craig French and Ingvar Johansson for insightful comments on an earlier draft. 6. He supplements this point with an interesting and early discussion of modality-specific spatial fields, which I leave aside. 7. Husserl is more explicit on this point in his 1907 Thing and Space lectures: ‘It must be noted, however, that a given empty space is necessarily an empty space between given things or phantom of things. If nothing spatial at all is given, then neither is any space’ (Husserl 1997/1973: 323, fn. 1). 8. See instead Mac Cumhaill (in preparation). 9. Thanks to Mark Textor and Hélène Leblanc for suggestions as to how to translate this passage. 10. See C. B. Martin (2006). 11. See O. Massin (forthcoming) for a discussion of the conception of ‘founded’ relations. 12. See Barry Smith (1988) for discussion of Brentano’s notion of a continuum, and the distinction, and relations, between spatial and temporal continua. 13. Ingvar Johansson rightly points that if things are partly individuated with respect to places, there is no problem with movement. 14. This translation appears in Johansson (1990: 153). Johansson notes that this characterisation leads to a peculiar consequence: there must be at least two different kinds of grounded relations—those grounded in something real (e.g. colour resemblance and difference of weight) and those grounded in something non-real (e.g. spatial relations). 15. We can now see perhaps why Marty mostly favours the term ‘ausgebreitet’ (spread out) to characterise the relation of sensible matter to form, over ‘durchdringend’ (interpenetrated), used by Stumpf.

6  Raum and ‘Room’: Comments on Anton Marty …     149

6. See, however discussion at §§17–18. 1 17. Johansson (2018) argues that space is objective in Marty’s sense, but that the space of perceptual experience is nonetheless subjective. Since a careful elaboration of Johansson’s subtle and distinctive position requires the introduction of a number of concepts that are peculiar to Johansson (for instance the notion of intentional-logical distance), I treat his view directly and in detail elsewhere. 18. Prima facie, Chalmers recent development of spatial functionalism (Chalmers 2006, 2012, forthcoming), a species of phenomenology-first representationalism where veridical space perception involves a Fregean manner of presentation, might seem to escape this difficulty. I leave the reader to follow this position up. I would suggest however that Fregean representationalism is also at odds with Marty’s picture, albeit for another reason: it makes visible space essentially ideal. 19. I recognise that this functional treatment may be viewed as inadequate in a number of respects. It defines sense-data in terms of non-hallucinatory experience rather than, as is more common, by appealing to historical responses to the argument for hallucination and illusion (see Fish 2010 for discussion). I do this for dialectical reasons. Second, my formulation may prompt epistemic worries—for instance, it might be thought that whether or not an entity is a sense datum cannot be determined until it is known whether experience is non-hallucinatory. I don’t think this is a difficulty for a functional treatment for the kinds of reasons spelt out in Martin (2004). Further, it does not strike me that saying that a sense-datum is whatever one is aware of in non-hallucinatory perception in virtue of which objects (say) are indirectly sensed, does not preclude the further claim that hallucination involves entities of the same kind but in virtue of which nothing is sensed. Thanks to Ingvar Johansson for raising these worries. 20. For discussion, H. H. Price (1932). 21. See fn. 7. 22. Translated by K. Mulligan and K. Schuhmann (1990). 23. It is plausible to think that, for the world to play a consciousness-shaping role, the subject must be awake. Merely being awake isn’t sufficient for perception however—one could be awake in a sensory-deprivation tank. 24. To be clear: both according to Marty and Brentano, a presenting may be a presenting of an existing or non-existing object. But according to

150     C. Mac Cumhaill

Marty, only the presenting of an existing object really has an object. This constitutes a difference between Brentano and the late Marty: Brentano maintained that even the presenting of a non-existing object has in a sense an object. Thanks to Giuliano Bacigalupo for emphasising this point. 25. I leave aside the question as to how space could be phenomenally represented. 26. An important though not unproblematic exception is of course when we perceive ‘ourselves’. 27. Johansson distinguishes between material-logical and intentional-­ logical distance, an important and productive distinction that I discuss in Mac Cumhaill (in progress). 28. Cf. Johansson (2014, Section 4).

References Beck, O. 2018. Rethinking Naïve Realism. Philosophical Studies (Online First). Brentano, F. 1988. Philosophical Investigations on Space, Time and the Continuum, trans. B. Smith. Oxford: Routledge. Brewer, B. 2011. Perception and Its Objects. Oxford: Oxford University Press. Campbell, J. 2002. Reference and Consciousness. Oxford: Oxford University Press. Chalmers, D. 2006. Perception and the Fall from Eden. In Perceptual Experience, ed. T. Gendler and J. Hawthorne, 49–125. Oxford: Oxford University Press. Chalmers, D. 2012. Constructing the World. Oxford: Oxford University Press. Chalmers, D. Forthcoming. Three Puzzles About Spatial Experience. In Themes from Ned Block, ed. A. Pautz and D. Stoljar. Cambridge: MIT Press. Chisholm, R. 1990. Brentano and Marty on Content: A Synthesis Suggested by Brentano. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 1–11. Dordrecht: Kluwer. Egidi, R. 1990. Marty’s Theory Space. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 171– 179. Dordrecht: Kluwer. Fish, W. 2009. Perception, Hallucination, and Illusion. Oxford: Oxford University Press. Fish, W. 2010. Introduction to Philosophy of Perception. London: Routledge.

6  Raum and ‘Room’: Comments on Anton Marty …     151

French, C. 2014. Naive Realist Perspectives on Seeing Blurrily. Ratio 27: 393–413. French, C. 2018. Naïve Realism and Diaphaneity. Proceedings of the Aristotelian Society 118 (2): 149–175. Husserl, E. 1997/1973. Thing and Space: Lectures of 1907. Dordrecht: Kluwer. Johansson, I. 1990. Marty on Grounded Relations. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 151–157. Dordrecht: Kluwer. Johansson, I. 2014. Triple Disjunctivism, Naive Realism, and AntiRepresentationalism. Metaphysica 15: 239–265. Johansson, I. 2018. Perceptual Spaces Are Sense-Modality-Neutral. Open Philosophy 2 (1): 14–39. Logue, H. 2012. Why Naive Realism? Proceedings of the Aristotelian Society 112: 211–237. Mac Cumhaill, C. In Progress. Co-Seeing and Seeing-Through. Martin, C.B. 2006. How It Is: Entities, Absences and Voids. Australasian Journal of Philosophy 74 (1): 57–65. Martin, M.G.F. 1998. Setting Things Before the Mind. Royal Institute of Philosophy Supplement 43: 157–179. Martin, M.G.F. 2004. The Limits of Self-Awareness. Philosophical Studies 120: 37–89. Marty, A. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, Bd. 1. Halle: Niemeyer (Referred to in the text as U). Marty, A. 1916. Raum und Zeit. Aus dem Nachlass des Verfassers herausgegeben von J. Eisenmeier, A. Kastil und O. Kraus. Halle: Niemeyer (Referred to in the text as RZ). Massin, O. Forthcoming. ‘Brentanian Continua’. Brentano Studien. Moore, G.E. 1903. The Refutation of Idealism. In Philosophical Papers, 1–30. New York: Harcourt, Brace & Co Inc. Mulligan, K. (ed.). 1990a. Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty. Dordrecht: Kluwer. Mulligan, K. 1990b. Marty’s Philosophical Grammar. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 11–29. Dordrecht: Kluwer. Mulligan, K., and Schuhmann, K. 1990. Two Letters from Marty to Husserl by Kevin Mulligan and Karl Schuhmann. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 225–237. Dordrecht: Kluwer.

152     C. Mac Cumhaill

Pautz, A. 2010. A Simple View of Consciousness. In The Waning of Materialism, ed. G. Bealer and R. Koons, 25–67. Oxford: Oxford University Press. Price, H.H. 1932. Perception. London: Metheun & Co. Simons, P. 1990. Marty on Time. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 157– 169. Dordrecht: Kluwer. Smith, B. 1990. Brentano and Marty: An Inquiry into Being and Truth. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 111–143. Dordrecht: Kluwer. Soteriou, M. 2013. The Mind’s Construction. Oxford: Oxford University Press. Stumpf, C. 1891/2017. Psychology and Epistemology. Trans. J. Leech and M. Textor. Manuscript. Originally Published as ‘Psychologie und Erkenntnistheorie’ in Abhandlungen der Königlich Bayerischen Akademie der Wissenschaften 19(2): 465–516. Wesche, B. 1986. At Ease with ‘At’. Journal of Semantics 5: 385–398.

7 Experiencing Change: Extensionalism, Retentionalism, and Marty’s Hybrid Account Thomas Sattig

Something changes in virtue of having different, incompatible attributes at different times. How do we experience episodes of change? This is a central part of the larger question as to how we experience time. Anton Marty addresses this issue in his posthumously published book Raum und Zeit (1916). In this paper, I shall reconstruct Marty’s position and relate it to contemporary work on our experience of change. In Sects. 1 and 2, I shall introduce and criticize a typical extensionalist account and a typical retentionalist account of our experience of change. In Sect. 3, I shall present Marty’s hybrid account as an original response to certain problems with the accounts of Sects. 1 and 2. I shall close by raising three challenges for Marty’s picture.

T. Sattig (*)  University of Tübingen, Tübingen, Germany e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_7

153

154     T. Sattig

1 An Extensionalist Account As a preliminary, I shall follow Marty and many others by adopting the common view that short episodes of change through time, such as the movement of a falling leaf or the frequency shift of a tone over the period of a second or less, can be experienced ‘immediately’. In order to motivate this view, compare looking at a falling leaf and looking at a wilting leaf. It seems that in the case of the falling leaf we can see the leaf ’s movement just by looking at it, whereas we cannot just see the wilting leaf ’s change in shape and colour. What explains this contrast? A natural thought is that there is a maximal period of time that an individual experience can span, a Jamesean ‘specious present’ (James 1890), such that the leaf goes through a sequence of locational states that are different enough to be visually discriminable, and that fall within a specious present, whereas no visually discriminable differences in the leaf ’s shape and colour fall within a specious present, and hence its wilting fails to be visually perceivable (under normal circumstances). Instead, the wilting of the leaf is inferred on the basis of memories of past leafstates of shape and colour.1 How do we experience episodes of change (in the scope of a specious present)? In order to set the stage for my reconstruction and discussion of Marty’s view, let me sketch an extensionalist account that is fairly common in the contemporary debate about the issue. The account rests on the metaphysical block-universe-theory of time, according to the simplest version of which time is a one-dimensional structure of moments that stand in an absolute relation of temporal priority. In the block universe, all moments of time are equally real—i.e. none of them is metaphysically privileged over the others. In this framework, an episode of change in the material world is standardly understood as a temporal sequence of incompatible states of the same material object. The movement of a leaf is thus understood as a temporal sequence of incompatible locational states of the same leaf, none of which is metaphysically privileged. Suppose that a leaf moves from location l1 to l2 to l3 across moments t1, t2, and t3. Suppose, moreover, that I experience this particular episode of change, a diachronic succession of leaf-states, ‘immediately’.

7  Experiencing Change: Extensionalism, Retentionalism …     155

How do I manage to experience such a succession? According to extensionalism, the explanation of the experiential representation of succession rests on the assumption that my token experience is itself temporally extended through the interval from t1 to t3, and that it has temporal parts at t1, t2, and t3, that correspond to the temporal parts of the leaf ’s movement.2 How the explanation proceeds from here is a matter of debate within the extensionalist camp. One approach is to hold that the temporally extended token experience represents the different states of the leaf as having a certain temporal order R, because the temporal parts of the diachronic token experience themselves have temporal order R. Another approach is to hold that the temporally extended experience just represents the movement of the leaf, and that the token experience has temporal parts with temporal order R, where these temporal parts represent temporal parts of the movement, because the movement of the leaf has temporal parts with order R. Whichever the direction of dependence, the relations of temporal priority holding among the temporal parts of the diachronic experience typically match the relations of temporal priority holding among the represented leafstates, as illustrated in Fig. 1, where p1–p3 are momentary perceptual

time

experiential content

t3

p3

t2

pp22

t1

p1 p1

l1

l2

l3

space

Fig. 1  An extensionalist account of our experience of change

experience

earlier than

156     T. Sattig

states whose cross-temporal sum is an ordinary experience of a contemporaneous movement of a leaf. This is an account of the contents of experiences of change in the scope of a specious present. How can the account be extended to address not only individual experiences of change during a specious present but also connected experiences constituting a continuous stream of consciousness over longer periods? Suppose that a leaf moves from location l1 to l2, … to ln, in thirty seconds. While this episode is too long to be immediately experienced as a whole, I experientially track the whole movement, representing the locational transition of the leaf from each li to li+1. How do my immediate experiences of the leaf ’s movements during various specious presents connect to yield a stream of consciousness that tracks the leaf ’s whole movement? A simple response is that my stream of consciousness is the sequence of my immediate experiences of the leaf ’s movements during various ‘adjacent’ specious presents, where the first experience in the sequence represents the leaf ’s movement from l1 to l3 (as in Fig. 1), the second experience in the sequence represents the leaf ’s movement from l4 to l6, and so on. This response is unsatisfactory, however, since I end up having no immediate experience of the leaf ’s transition from l3 to l4, contrary to our assumption. So my immediate experiences must be connected in such a way that all transitions in the leaf ’s movement are represented. This may be achieved by letting the contents of the immediate experiences making up the stream of consciousness ‘overlap’, such that the first experience in the sequence represents the leaf ’s movement from l1 to l3, the second experience in the sequence represents the leaf ’s movement from l2 to l4, and so on. Extensionalists can implement this response elegantly by allowing the immediate, temporally extended token experiences in the stream of consciousness to overlap by sharing temporal parts, in a way that matches the overlap of the represented segments of the leaf ’s movement. While the experiential tracking of longer episodes of change raises further philosophical challenges, I shall leave this issue here, since my discussion will focus on experiences of change during a specious present.3 Let us now turn to the phenomenal character of such experiences. In order to focus the discussion, I shall assume representationalism,

7  Experiencing Change: Extensionalism, Retentionalism …     157

according to which phenomenal aspects of mental states are grounded in (or are identical with) representational aspects—aspects of the contents—of these mental states. Given representationalism, extensionalists face the following question: Can the extensionalist content of our experience of the leaf ’s movement explain the experience’s phenomenal character? That is, can the content explain how experiencing the movement feels to us? It is hard to see how such an explanation could go. According to extensionalism, we experience the leaf ’s movement as having the shape of a four-dimensional spacetime-worm reminiscent of a comic strip. Prima facie, this shape is miles from how an experience of a moving leaf normally feels like to us. Something seems to be missing from this picture. One aspect of our experience of change that seems to be missing is the following. While we seem to experience several three-dimensional locational leaf-states together, these different states do not all seem real. At most one leaf-state seems real. Moreover, which leaf-state appears real seems to change. We seem to experience one leaf-state after the other entering and leaving reality. What seems real never includes non-simultaneous states. Experienced reality never becomes temporally extended, four-dimensional. Reality rather remains temporally unextended, yet in constant flux, with one leaf-state in the bounds of reality being replaced by another, and that one by another, and so on. Such constant replacement seems integral to the sense of temporal flow, the sense of whoosh, in the experience of change. While the sketched extensionalist picture captures our experience of the occurrence and temporal order of multiple locational leaf-states, it misses the apparent differences concerning the states’ reality-status. What makes our experiences of change feel that way?

2 A Retentionalist Account Many philosophers have thought that an adequate explanation of our experience of change, which includes the sense of differences in realitystatus, requires as its basis a different metaphysical picture of time. Friends of ‘dynamic’ theories of time—flow-lovers, for short—have

158     T. Sattig

traditionally viewed their accounts of the nature of time as providing ideal foundations for explaining the phenomenal character of our experiences of change. They have argued that, unlike block-lovers—friends of the block-universe-theory of time—they can give a simple explanation. According to flow-lovers, we seem to experience a change in reality when we experience a change in objects, because our experience represents the change in objects as involving a change concerning the reality of the experienced object-states. And the experience represents the objectual change in this way, because the objectual change in fact involves a difference and change concerning what is real. This approach looks simple and elegant. But it raises two obvious questions. What does a difference and change concerning what is real consist in? And how do we experience this sort of difference and change? One starting point is to recognize a metaphysically robust notion of the present, or presentness, and to think of object-change, such as the movement of our leaf, along the following lines. The present undergoes a change in which three-dimensional states of our leaf it contains. The present currently contains the leaf ’s l3-state, say, but the present used to contain the leaf ’s l2-state instead, and even longer ago the present contained the leaf ’s l1-state. When the leaf undergoes locational change, then one three-dimensional locational state contained in the present is replaced by a different three-dimensional locational state as the present changes. I shall call this the replacement model of change. Flow-lovers may welcome the replacement model, while block-lovers must reject it, since they do not recognize a metaphysically robust present. Intuitively, the replacement model of change seems to correspond to what our experiences of change feel like. We do not seem to experience undifferentiated four-dimensional episodes of change in objects. We rather seem to experience successions of three-dimensional object-states, where each state is immediately replaced by another one. Given the replacement model of change, the experienced difference and change in what is real may be understood as change concerning which object-state is present. The replacement model of change may be based on different flow-loving metaphysical pictures of time. The usual suspects are presentism, the moving-spotlight theory, and the growing-block theory.

7  Experiencing Change: Extensionalism, Retentionalism …     159

All three views may be portrayed as saying that when our leaf undergoes locational change, then one three-dimensional locational state contained in the present is replaced by a different three-dimensional locational state as the present changes. According to presentism, very roughly, each locational leaf-state comes into existence as it enters the present and goes out of existence as it leaves the present, since only presently obtaining states exist. Thus, different states of the leaf do not ‘add up’ to yield a four-dimensional sequence, and there are no states that stand in the absolute earlier-than relation or the absolute later-than relation to present states, as is the case in the block universe. According to the growing-block view, by contrast, each locational leaf-state comes into existence as it enters the present but remains in existence as it leaves the present. Thus, different states of the leaf add up over time to yield a four-dimensional sequence—the sequence ‘grows’—such that past states stand in the earlier-than relation to present states. Finally, according to the moving-spotlight view, no leaf-state comes into or goes out of existence as it enters or leaves the present, since all leaf-states exist absolutely, composing a four-dimensional sequence of states atemporally. The non-existential change with respect to which of these locational states is present may be compared to a change with respect to which of a sequence of objects is illuminated by a spotlight. The formulation of these views raises many questions that cannot be addressed here. So let us move on. Can we experience a change concerning which object-state is present, in the flow-lover’s robust sense of ‘present’? On the face of it, we cannot. For the flow-lover’s fundamental property of presentness does not seem to be a property that is fit to be represented in immediate experience along with an object’s shapes and colours, as Brentano and Marty, among others, have pointed out. Likewise for the property of pastness. The good news, though, is that there is another way for the difference between presentness and pastness to become manifest in experience. According to the view that is now typically called retentionalism, we can distinguish between a present-directed experiential state, which I shall call a perception, and a past-directed experiential state, which Husserl called a retention. Husserl got the idea from Brentano, who called retentions proteraestheses. I shall stick with the now-common Husserlian

160     T. Sattig

term.4 Brentano, in his later period, thought that perceptions and retentions are individuated by different temporal modes of presentation (Vorstellung ). We may think of perceiving an object-state as experiencing-present the state, and we may think of retaining an object-state as experiencing-past the state. Brentano also thought that there are different degrees of retaining a state, so that a subject can retain a state more or less strongly. Suppose that in our leaf case a subject S perceives (in my strict sense of the term) the leaf ’s l3-state, which is the state currently contained in the present. Moreover, S retains the leaf ’s l2-state to a high degree, which is a state that was contained in the present in the most recent past. Moreover, S retains the leaf ’s l1-state to a slightly lower degree, which is a state that was contained in the present a tiny bit longer ago. It is important that each of these experiential states represents a leafstate without representing presentness or pastness of the state, and without representing any temporal relations between the leaf-state and any other object-states. S’s perception and retentions are illustrated in Fig. 2. For ease of illustration, the different ‘versions’ of the present to which S’s three intentional states, namely perception p and retentions r1 and r2, ‘reach out’ are depicted in a single diagram. WLPH H[SHULHQWLDOFRQWHQW

S

QRZ

U

U

QRZ

WHPSRUDOPRGHV RISUHVHQWDWLRQ

QRZ

O

O

O

VSDFH

Fig. 2  A retentionalist account of our experience of change

7  Experiencing Change: Extensionalism, Retentionalism …     161

The psychological difference between perception and retention c­orresponds to the worldly difference between presentness and pastness. While presentness and pastness are not represented in experience, they nevertheless make their mark in experience through these modes of presentation. But how does this difference rise to consciousness, given that temporal modes are mere ways of representing? Brentano’s answer, as I understand it, is that the past-present divide becomes conscious through the subject’s awareness of being in a state of perception and in a state of retention, which awareness accompanies each such state. In our case, S is currently aware that she is perceiving the leaf ’s l3-state, S is aware that she is retaining to degree n the l2-state, and S is aware that she is retaining to degree n* the l1-state, where n* is lower than n. Furthermore, S’s current states of awareness are phenomenally unified, which allows S experientially to ‘compare’ her perception and her retentions, and hence to be aware of their differences. Through a subject’s awareness of differences in temporal mode of presentation, those differences become conscious. As pointed out at the beginning, when we experience episodes of change, such as the leaf ’s movement, we seem to experience a succession involving a difference concerning which object-state is real: when several, non-simultaneous object-states are experienced, only one of them feels real, whereas the others do not. On the sketched retentionalist model, neither the differences in reality-status between, nor the temporal order among, leaf-states are represented by the content of the experience. What these phenomenal aspects consist in are aspects of the awareness of perceptions and retentions. What makes me say, intuitively, that in experiencing change one object-state feels real, while another one does not, is that I am aware of perceiving the first state, while I am aware of retaining the second. And what makes me say, intuitively, that the experienced object-states are temporally ordered, is that I am aware of the differences in type and degree of temporal modes of presentation involved in the experience of change, which differences give rise to a unique order. This picture aims to explain more aspects of our experiences of change than the simple extensionalist model of Sect. 1. While the latter only captures the experienced temporal order among object-states,

162     T. Sattig

the retentionalist model outlined here also captures the experienced differences in reality-status between object-states. It should, furthermore, be indicated briefly how the retentionalist account may be extended to yield a basic picture of how immediate experiences of change in the scope of a specious present connect to yield a stream of consciousness over a longer period of time. Suppose, again, that a leaf moves from location l1 to l2, … to ln in thirty seconds, and that while I lack an immediate experience of the movement as a whole, I track the whole movement experientially, representing the locational transition of the leaf from each li to li+1. As we saw in Sect. 1, one challenge of this case (among others) is that my immediate experiences must be connected in such a way that all transitions in the leaf ’s movement are represented. This may be achieved by letting the contents of the immediate experiences making up the stream of consciousness ‘overlap’. In the retentionalist framework sketched above, this desideratum may be satisfied by letting S’s perception of the leaf ’s l3-state and S’s simultaneous retentions of the leaf ’s l2-state and of its l1-state (as illustrated in Fig. 2) be immediately followed by S’s perception of the leaf ’s l4-state and S’s simultaneous retentions of the leaf ’s l3-state and of its l2-state, and so on. As a result, the subject does not miss any locational transition in the leaf ’s movement from l1 to ln. Is the retentionalist approach plausible? One might object that it is implausible to move all temporal aspects of an experience of change from the content of a first-order experience to the content of a higher-order awareness of the temporal modes involved in the first-order experience. Many contemporary philosophers share the view that our temporal experience is ‘transparent’ at least to the extent that we represent things in the world—objects, object-states, property-instances, or the like—as being ordered by temporal priority. The sense of temporal priority comes from first-order content; worldly temporal order is experienced ‘directly’. Marty expresses this view when he says: Direct experience alone seems to show me that we have an intuition (Anschauung ) of such non-modal temporal differences. (1916: 209)

7  Experiencing Change: Extensionalism, Retentionalism …     163

The Brentanian retentionalist lacks the resources to capture this common transparency intuition. Furthermore, Marty complains about Brentanian retentionalism that retention is taken to come in degrees (1916: 208). In experiencing a leaf ’s motion over a second, say, we seem to experience a continuum of different leaf-states. In order to capture this impression, the retentionalist must admit indefinitely many different retentional states of the subject of experience, which are individuated by different degrees of retention. Marty, among others, finds it hard to accept that our experiences of change involve this sort of mental complexity. The extensionalist of Sect. 1 seems to be in a better position than the Brentanian retentionalist to capture our impression of temporal order, because the extensionalist simply takes such order to be represented in first-order experience. As we saw, however, the extensionalist lacks the resources to capture our impressions of differences in reality-status among object-states, whereas the retentionalist can explain these differences by recourse to temporal modes of presentation. Might there be a position that combines these two explanatory strengths?

3 Marty’s Hybrid Account The heart of Marty’s own account of our experiences of change is to combine an extensionalist element with a retentionalist element into a hybrid account. I take Marty to develop his account in a presentist framework, since he states, at the beginning of his considerations on temporal experience in Raum und Zeit: Only the present is actual (wirklich ) in the proper and relevant sense, while the past and the future is not. (1916: 197)5

Let me begin with his extensionalist element. Marty holds that when we immediately or intuitively (anschaulich ) experience change a temporal sequence of object-states, within the scope of a specious present, is represented as temporally ordered in the experience. He does not

164     T. Sattig

consider or take seriously the option of allowing a temporally extended token experience to do the job of representing a temporally extended sequence. So he claims that a momentary, temporally unextended, experiential state represents a sequence of object-states as temporally ordered, which sequence includes object-states having occurred before the experiential state itself. Applied to our case, a momentary experience of subject S represents the leaf ’s l1-state, its l2-state, and its l3state, as well as the earlier-than relation holding between these states, as illustrated in Fig. 3. (As in Fig. 2, different ‘versions’ of the present are depicted in a single diagram.) Marty’s approach is extensionalist, in so far as he allows the content of an experience of change to represent a temporally extended sequence of object-states. He does not, however, join the straightforward extensionalists of Sect. 1, because he denies that the token experience itself has a temporal extension that matches the extension of the sequence it represents. Instead, the token state itself is, according to Marty, temporally unextended.6 Marty does not stop here. For he realizes that there is more to the experience of change than the representation of a sequence of objectstates and its temporal order. As I put it earlier, the different states time

experiential content presentation (Vorstellung)

p

now

now

now earlier than

l1

l2

l3

space

Fig. 3  The extensionalist element of Marty’s account

7  Experiencing Change: Extensionalism, Retentionalism …     165

experienced do not all feel real. As he puts it, the states do not all feel actual, or factual (aktuell, tatsächlich ). There is a phenomenal aspect that still needs to be accounted for. In response, Marty incorporates a retentionalist element into his picture. Marty is convinced, following Brentano, that we must distinguish between different psychological attitudes to what is presented in an immediate experience of change. In contrast to Brentano’s later view, however, he holds that an experience of change involves different temporal qualities, or forms, of the mode of judgment without also involving different temporal modes of presentation. (For Brentano temporal modes of judgement are derived from temporal modes of presentation.) Given that for Marty the temporal horizon of a presentation is extended, presenting a plurality of object-states, such a presentation cannot be partly individuated by a specific temporal mode of presenting a momentary object-state, in the way Brentano advocates. As regards the mode of judgement, Marty admits two temporal forms of this mode, in addition to its positive form (accepting) and its negative form (rejecting). Judgements thus come in three forms: accepting-present, accepting-past, and rejecting.7 The present-directed form of judgement and the past-directed form accompany each of our immediate presentations of temporally extended sequences of object-states. As Marty is sceptical about the idea that retentions come in degrees, he does not admit different degrees of judging in the past-directed form. Nor does he admit a future-directed form of judging. Marty summarizes the two components he finds in an experience of change as follows: The intuition (Anschauung ) of constancy and change involves: 1. that a continuum of temporal positions and their fillers is presented to us; and 2. that a judging consciousness is connected with the presentation of each of those positions and their fillers, which currently judges the presented as present and then, for a while, as past. (1916: 210)

In the leaf case, subject S has a presentation of three temporally ordered leaf-states. About these different leaf-states S makes mental judgements in different temporal forms. S mentally judges in the present-directed form—that is, S accepts-present—that the leaf is in l3.

166     T. Sattig

As regards the leaf ’s l1-state and its l2-state, Marty holds that S mentally judges in the past-directed form—that is, S accepts-past—that the leaf is in l1 and l2. Notice that he does not hold that S makes distinct past-directed judgements, one concerning the leaf ’s l1-state and one concerning the leaf ’s l2-state. He rather holds that one past-directed judgement spans a whole manifold of past states (1916: 211). Marty’s take on S’s judgements in the leaf case is illustrated in Fig. 4, where j1 and j2 are a present-directed and a past-directed mental judgement, respectively. As on the retentionalist account of Sect. 2, a present-directed mental judgement does not represent an object-state as being present. Nor does a past-directed mental judgement represent an object-state as being past. A subject is aware of judging-present an object-state while making this mental judgement, and she is aware of judging-past an object-state while making that mental judgement, and she is aware of the difference in temporal form of judgement when she is judging-present and judgingpast. These moves put Marty in the position to explain that S has the sense that the leaf ’s l3-state is actual, whereas she has the sense that the leaf ’s l2-state and its l1-state are both non-actual. The difference consists in S’s awareness of judging-present that the leaf ’s l3-state obtains, time

j1

now

now

j22 p

temporal forms of judgment

now

l1

l2

l3

space

Fig. 4  The retentionalist element of Marty’s account

7  Experiencing Change: Extensionalism, Retentionalism …     167

and S’s awareness of judging-past that the leaf ’s l2-state and its l1-state obtain. By combining an extensionalist element with a retentionalist element Marty can offer an account of our immediate experience of change, in which both the impression of temporal order and the impression of differences in reality-status, or, using his term, in actuality-status, among experienced object-states are captured. The impression of temporal order is grounded in an aspect of experiential content, whereas the impression of differences in reality/actuality comes from the awareness of differences in temporal forms of judgement.8 Marty’s hybrid view is original and deserves more discussion in the literature on temporal experience than it has received so far. I presented it in a way that is meant to highlight its advantages over the simple extensionalist view of Sect. 1 and the Brentanian retentionalist view of Sect. 2. In contrast to simple extensionalism, the hybrid view captures the impression of differences in reality-status—that is, the impression of a past-present divide—in addition to the impression of temporal order. In contrast to Brentanian retentionalism, the hybrid view captures the impression of temporal order by respecting the common intuition of transparency—that is, the intuition that we represent temporal order as occurring out there in the world—and without massively increasing the complexity of temporal modes of presentation by admitting degrees of retention. I shall close by pointing to three challenges for Marty’s account. First, Marty does not offer much by way of resolving the mystery of how an experience manages to represent a temporally extended sequence of object-states. He just takes for granted that a present experience can represent a chunk of the past in addition to the present. Contemporary extensionalists claim to offer the most straightforward account of the representation of temporally extended sequences by holding that token experiences are themselves temporally extended and represent an extended sequence in the world atemporally. The experiences just ‘mirror’ the world. (In Sect. 1, I distinguished two implementations of this explanatory strategy.) On Marty’s account, however, a token experience is supposed to represent, at a moment, a temporally nonmomentary sequence of object-states. This seems to be a way of seeing

168     T. Sattig

(not remembering, and not retaining) the past from the present. Many find this sort of representational power mysterious. Second, while Marty adopts presentism and presumably expects our experiences as of change to represent episodes of change veridically, an experiential representation of a sequence of object-states as ordered by temporal priority cannot be veridical in a typical presentist world. The reason is that in such a world no states stand in an absolute relation of temporal priority. There are no object-states or objects or any other things earlier than or later than the present. This is the key assumption of the view. While there were other object-states and other things, as ‘contents’ of the present as it used to be, none of these states or things are earlier than the present, in the sense of ‘earlier-than’ that Marty has in mind. So our experiences as of change, as Marty construes them, are all illusory, if presentism is true. Most friends of ‘dynamic’ theories of time, however, wish to avoid the apparent consequence of ‘static’ theories that we are stuck with systematically illusory temporal experiences. Flow-lovers impressed with the explanatory potential of Marty’s hybrid view might consider cutting its presentist roots and reconstructing it within a different flow-friendly metaphysical framework instead. According to Marty, a present experiential state represents a sequence of object-states that reaches into the past. Since both the growing-block view and the moving-spotlight view accept the existence of past objectstates, Marty’s presently obtaining, past-directed experiential states representing object-states as ordered by temporal priority might have a chance of being veridical in these frameworks. Third, while Marty’s account goes further than the other accounts reviewed here in explaining how we experience change, it shares with the others the significant defect of omitting the most difficult aspect of experiencing change, namely, the sense of flow, as I call it. Recall my initial phenomenological description of experiencing the movement of a leaf (Sect. 1). It seems that we experience the replacement of one three-dimensional locational state of the leaf by another locational state, and the latter by another one, and so on, such that one leaf-state after the other enters and leaves reality. When experiencing the occurrence of a plurality of leaf-states, we seem to experience a change in

7  Experiencing Change: Extensionalism, Retentionalism …     169

which each of these states is real. What Marty’s hybrid account gives us is an experience of a temporal sequence of states, such that one state in the sequence feels real, or actual, and the others feel unreal, or non-actual. The source of this phenomenal difference concerning reality/actuality-status is Marty’s distinction between two temporal forms of judgement. But the hybrid account does not yet tell us how we get to experience a change concerning which state feels real, or actual, and which states do not. One might attempt the following response on behalf of Marty. Suppose that at one time, S has a momentary experience that represents that the leaf ’s l1-state obtains before the leaf ’s l2-state, and that the leaf ’s l2-state obtains before the leaf ’s l3-state. Moreover, at this time, S accepts-present that the leaf ’s l3-state obtains, while accepting-past that the leaf ’s l1-state and its l2-state obtain. Suppose, further, that a moment later, S has an experience that represents that the leaf ’s l2-state obtains before the leaf ’s l3-state obtains, and that the leaf ’s l3-state obtains before the leaf ’s l4-state obtains. Moreover, S now accepts-present that the leaf ’s l4-state obtains, while accepting-past that the leaf ’s l2-state and its l3-state obtains. In this example, different leaf-states feel real/actual to S at different times. At one time, S has an experience of a sequence of object-states, where one state feels real/actual and the others do not, and at a later time, S has a different experience of almost the same sequence of object-states, where a different state feels real/actual and the others do not. This is a Marty-style account of how different immediate experiences of change, each in the scope of a specious present, connect to yield a stream of consciousness over a longer period of time. While the picture involves a change in which object-state feels real/actual, it is insufficient to explain how we immediately experience the actuality/ reality flux. The subject may have one experience of a sequence where one state feels real/actual, and she may remember (in the standard episodic sense) that she had an experience a moment ago where a different state felt real/actual, and then she may infer that a change has occurred in which state feels real/actual. But this is not a model of experiencing a change in reality/actuality of the sort required by a phenomenally adequate account of temporal experience.9

170     T. Sattig

Notes 1. See, inter alia, Broad (1923: 351). 2. Extensionalism’s main defenders are Dainton (2006), Foster (1991), and Phillips (2014). 3. The issue of how experiences of change in the scope of a specious present are connected to yield a stream of consciousness is discussed in detail in Dainton (2006). 4. See Brentano (1988) and Husserl (1991). 5. As Simons (1990) points out, Marty’s notion of being actual (wirklich, aktuell ) is different from his notion of being real, by which he means being causally efficient (wirkungsfähig ). The present, for him, is actual but not real. 6. This is, in essence, the view of experiencing change put forth by C. D. Broad in his (1923). 7. Cf. Simons (1990: 167). 8. According to Brentano, it would be implausible to hold that all of our time consciousness requires temporal forms of judgement. For we may simply think about states in the past or in the future without accepting or rejecting them, thereby taking a non-judgemental intentional attitude towards them. Likewise for various other intentional attitudes, including emotional ones. In response to Brentano’s worry, Marty admits a mediate type of time consciousness in addition to the immediate type that is involved in our experiences of change (1916: 214ff). Since his account of mediate time consciousness remains fragmentary, and since the focus of my discussion is on immediate time consciousness, I shall not address the nature of the mediate form here. See Simons (1990: section 4.3) for a reconstruction. 9. I would like to thank the participants at the conference ‘Anton Marty and Contemporary Philosophy’ (University of Geneva 2017) and an anonymous referee for helpful discussion of this paper. I am especially grateful to Guillaume Fréchette for his extensive comments, and to Kevin Mulligan for introducing me to Anton Marty’s work.

7  Experiencing Change: Extensionalism, Retentionalism …     171

References Brentano, F. 1988. Philosophical Investigations on Space, Time and the Continuum, trans. B. Smith. Beckenham: Croom Helm. Broad, C.D. 1923. Scientific Thought. London: Routledge & Kegan Paul. Dainton, B. 2006. Stream of Consciousness, 2nd ed. London: Routledge. Foster, J. 1991. The Immaterial Self. London: Routledge. Husserl, E. 1991. On the Phenomenology of the Consciousness of Internal Time (1893–1917), ed. and trans. J.B. Brough. Dordrecht: Kluwer Academic. James, W. 1890. The Principles of Psychology. New York: Dover. Marty, A. 1916. Raum und Zeit. Aus dem Nachlass des Verfassers herausgegeben von J. Eisenmeier, A. Kastil und O. Kraus. Halle a. S.: Niemeyer. Phillips, I. 2014. The Temporal Structure of Experience. In Subjective Time: The Philosophy, Psychology, and Neuroscience of Temporality, ed. D. Lloyd and V. Arstila. Cambridge, MA: MIT Press. Simons, P. 1990. Marty on Time. In Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty, ed. K. Mulligan, 157– 170. Dordrecht: Kluwer Academic.

Part III Meta-metaphysics and Meta-philosophy

8 A Syncretistic View of Existence and Marty’s Relation to It Alberto Voltolini

1 Introduction In this paper, I want to present, first, a syncretistic account of existence, which tries to show not only that two first-order properties and one second-order property of existence are compatible, but also why we need their cooperation in order to properly understand what existence overall amounts to. Second, I want to discuss up to what extent Marty’s account of existence, which inter alia mobilizes Brentano’s attitudinal approach to it, can be legitimately considered to be a syncretistic account as well. In theory, the answer to this question should altogether be negative. Marty may certainly be taken as a, admittedly non-standard, pluralist about existence, yet pluralism is not syncretism. Yet one might still wonder whether the three notions Marty mobilizes in order to account for existence play the same role as the three aforementioned properties play in the above syncretistic account, so as to A. Voltolini (*)  University of Turin, Turin, Italy e-mail: [email protected] © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_8

175

176     A. Voltolini

configure a specific form of syncretism. But also once so reformulated, the question is to be answered in the negative. First, if Marty’s account were an even specific syncretistic account, it would hardly be tenable. Second, if it were nonetheless a syncretistic account, it would just be the previous syncretism in disguise, so as to lose the specificity of an attitudinal approach.

2 Syncretism About Existence Syncretism about existence (Russell 19372; Voltolini 2012) says that the tension between alternative conceptions of existence—a first-order conception according to which existence is a property of individuals and a second-order conception according to which existence is a property of properties—can be reconciled.1 For according to syncretism, there are at least three related notions of existence: (i) a second-order property, instantiatedness (Frege 19742; Russell 1905), which properties exemplify; (ii) a first-order, yet universal, formal property, i.e., a property that is exemplified by all items in the overall ontological domain by however making no substantial difference to them: existence in a formal sense (what Russell 19372 labelled being and Williamson 2002 relabelled logical existence; alternative formulations may be being an object [Berto 2013], being identical with something [Lewis 1990]); (iii) a first-order, apparently nonuniversal, substantive property, i.e., a property that is apparently exemplified by just a subset of the individuals in the overall ontological domain so as to make a substantial difference to them: existence in a substantive sense (what Meinong 1960 labelled Sein ). And perhaps other, even more substantive yet also apparently nonuniversal, first-order properties are around, what the tradition has called modes of being.2 For instance for Meinong (1960), Sein can be further split into nonspatiotemporal existence (subsistence ) and spatiotemporal existence (physical existence, according to Williamson 2002, or having the potential to interact causally, the old Platonic idea revived by Berto 2013 and Priest 20162). This is for Routley the property that makes a real difference for the things that have it: ‘existence often makes a substantial difference to an object and to its character; e.g., removal of existence by

8  A Syncretistic View of Existence and Marty’s Relation to It     177

death or destruction can make the difference between a lively energetic creature and a lifeless object that was (even briefly), before, that creature’ (Routley 1980: 51). Clearly enough, syncretism is not only motivated by the fact that the two aforementioned conceptions are clearly defective, but also and more importantly by the fact that we need all the above notions in order to properly understand what existence overall amounts to. To begin with, as is well known, the second-order conception of existence, the one mobilizing instantiatedness, has been systematized by Quine (1952).3 Instantiatedness has a genuine existential nature insofar as it affects ontological commitment. Indeed, ontological commitment is captured in terms of being the value of a bound variable, hence in terms of the second-order property of existence: we ontologically commit to Fs iff a sentence logically reconstructed as ‘There are Fs’ is true, hence iff the property of being F is instantiated. Thus, in a nutshell, that a certain property F is instantiated is existentially relevant for it fixes that there are Fs. Yet on the one hand, this second-order property of existence does not suitably capture all the ways in which we pretheoretically speak of existence. For example, ‘Trump [helas] exists’ is not suitably captured as ‘There is a (just one) Trumpian individual’. For it is not clear not only what being a Trumpian individual is, but also whether the fact that an individual (uniquely) satisfies the Trumpian features is enough in order for Trump to exist (for instance, as Kripke [1980] has reminded us, it seems that Trump might not have existed despite the fact that the Trumpian features are still possibly [uniquely] satisfied).4 Thus, it would be hard to understand how one may account for the anaphoric link in ‘Trump exists yet the round square does not’, if the second (implicit) occurrence of the existence predicate meant, as it seems, the secondorder property of existence.5 Nevertheless on the other hand, also the first-order property of existence seems to be defective, for how can we legitimately state in a first-order sense of existence that, unlike Trump, there are individuals that do not exist, e.g. Vulcan? Is it not like contradictorily saying that there exist individuals that do not exist?6 And even if such a statement were perfectly meaningful, as many people agree on (cf., e.g., Crane 2013), how could that property be an existential property in an interesting sense, rather than be a property of individuals like

178     A. Voltolini

any other (e.g. a property alluding to some contingent facts of production, which explains for instance why a counterfactual of the kind ‘Had this man’s parents not met, this man would have never existed’7 is true)? At this point, syncretism comes into the stage by claiming not only that there are three properties of existence, a second-order one and the two aforementioned first-order ones, but that the two first-order properties of existence are existentially genuine for they are related to the second-order one, insofar as we need all of them if we want to understand any ontological disagreement. To begin with, ontological disagreement is disagreement as to what we allow in the overall ontological domain. For example, unlike realists, nominalists about universals do not allow for universals in that domain; likewise as to antirealists about ficta, and so on. That means that a disagreement as to what we allow in the overall ontological domain is captured in the above Quinean terms, hence in terms of the second-order property of existence. For one to ontologically commit to ficta is to truly say that there are fictional entities, for one to ontologically commit to universals is to truly say that there are universals, etc., yet antirealists and nominalists, respectively, deny such sayings. If this is the case, then an ontological disagreement consists in the fact that for people committed just to Fs, sentences of the kind ‘there are no nonFs’ are true sentences, whereas for people not so committed they are false, so that sentences like ‘There are nonFs (notably, Gs)’ are true. For instance, nominalists hold that ‘There are no nonparticulars’ is true, while realists hold that it is false, so that ‘There are nonparticulars (notably, universals)’ is true. Likewise, ficta antirealists hold that ‘There are no unreal individuals’ is true, while realists hold that it is false, so that ‘There are unreal individuals (notably, fictional ones)’ is true. As hinted at before, this situation already shows that the second-­order notion of existence is relevant for ontological matters. Hence, it is relevant in order to understand what existence overall amounts to, since ontological commitment to certain items, or better, their figuring in the overall ontological domain, captures a fundamental sense of existence. Yet moreover, once it is understood in the above terms, an ontological disagreement is not a verbal disagreement, a disagreement in the meaning of the relevant notion of existence, but is a disagreement as to the

8  A Syncretistic View of Existence and Marty’s Relation to It     179

truthvalue of the relevant existentially quantified sentences. Taken in this way, one may say that ontological disagreement is genuine (for it does not depend on a meaning, but on a truthvalue, shift of the relevant sentences) and yet it is not faultless (for it depends on the reasons one must suitably fix the relevant domain). Now, how can we assess that this disagreement only concerns the truthvalues of the above sentences (and not their meaning)? For the disagreement has to do with the fact that the overall ontological domain is differently fixed by how one differently considers to be the extension of the above first-order properties of existence. As a result, the overall scope of the existential quantifier is differently fixed in the different circumstances of evaluation of the relevant quantified sentence, since these different circumstances mobilize an extended or a shrunk overall ontological domain. Thus, also the above two first-order notions of existence are relevant for ontological matters, hence for understanding existence as well. For they also have a role in the fixation of what figures in the overall ontological domain. Let me clarify what I mean by means of the most relevant case of disagreement. The paradigmatic sample of an ontological disagreement is that holding between (late) Russellians and (neo)Meinongians. In my terms, Russellians say that ‘There are only existing things’, or ‘There are no nonexistent things’, are true, while Meinongians say that such sentences are false; for them, ‘There are things that do not exist’ is true. The former conceive the substantive first-order property of existence, Meinong’s (1960) Sein, as universal, hence as determining what there is in the overall ontological domain. Alternatively put, the former take the substantive first-order property of existence as coextensional with the formal universal first-order property of being, both thereby determining what there is in the overall ontological domain. For them, therefore, the substantive first-order property taken as universal, or as coextensional with the first-order former universal property, fixes the overall scope of the existential quantifier in the above sentences. As a result, in the circumstances of evaluation that are determined by such a fixation, the above two sentences are true and the third one is false. Yet the latter deny all that. For them, the substantive first-order property is not universal; thus, it does not fix the overall scope of the existential quantifier in the above sentences. What fixes that scope is merely the first-order

180     A. Voltolini

formal universal property, which for them is not only intensionally but also extensionally different from the substantive first-order property. As a result, in the different circumstances of evaluation that are determined by this alternative fixation, the above two sentences are false and the third one is true. Thus, in order to understand this ontological disagreement by assessing it as a mere disagreement on the truthvalues of the relevant existentially quantified sentences, one needs all the three aforementioned properties of existence, all of which thereby play an ontologically relevant role.8 A caveat. Meinong (1960) literally said that non-existents do not exist in any sense, hence there is for them no quasi-Sein. So, for Meinong they do not even possess the first-order formal universal property of existence. This is official Meinongianism, echoed by many of its scholars, principally Routley (1980) and Priest (20162). To be sure, in his unpublished (1978) Meinong ascribed all objects the feature of being given, which precisely plays the role of the first-order formal universal property of existence.9 Yet not only exegetical, but also verbal issues do not have to lead us astray. Saying, e.g., that non-existents are non entia (Routley 1980; Priest 20162) is fine, if it simply means that such items lack the first-order nonuniversal property of existence while they merely possess the first-order formal universal one. The point indeed is that Meinongians need the first-order formal universal notion of existence that fixes the extension for them of the overall ontological domain, hence the overall scope of the existential quantifier, in order for the ontological disagreement with non-Meinongians to concern only the truthvalues of the relevant existentially quantified sentences.10 Put alternatively, what Meinongians à la Routley (1980) and Priest (20162) call the neutral quantifier is (means) the same as the existential quantifier of non-Meinongians. The disagreement between the two parties just concerns the overall scope of that quantifier, which makes for the latter but not for the former a sentence such as ‘There are things that do not exist’ true, where the first-order substantive property of existence is mobilized (actually, for those Meinongians an even more substantive property of existence is mobilized: see below).11 Once things are so seen, the disagreement issue arises again and again, by showing that even more properties of existence must be

8  A Syncretistic View of Existence and Marty’s Relation to It     181

mobilized, in order for one to understand what the relevant ontological disagreement amounts to. So, fundamentalists say that ‘There are only fundamental things’ and ‘There are no nonfundamental things’ are true, while non-Fundamentalists say they are false; for the latter but not the former, ‘There are nonfundamental things’ is true. The former conceive an even more substantive first-order property of existence (spatiotemporal existence, as Meinong 1960/Williamson 2002 would put it; or being endowed with causal properties, as Berto 2013 and Priest 20162 would put it) as universal, hence as determining what there is in the overall ontological domain. Alternatively put, the former take the more substantive first-order property of existence as coextensional with the formal universal first-order property, both thereby determining what there is in the overall ontological domain. For them, therefore, the new more substantive first-order property taken as universal, or as coextensional with the first-order former universal property, fixes the overall scope of the existential quantifier in the above sentences. As a result, in the circumstances of evaluation that are determined by such a fixation, the above two sentences are true and the third one is false. Yet the latter deny all that. For them, the more substantive first-order property is not universal, so it does not fix the overall scope of the existential quantifier in the above sentences. What fixes that scope for them is the previous first-order substantive first-order property of Sein taken as universal, or as coextensional with the formal universal first-order property. As a result, in the different circumstances of evaluation that are determined by this alternative fixation, the above two sentences are false and the third one is true. In this respect, an ontological disagreement is structurally not different from a disagreement between people who implicitly restrict the scope of the existential quantifier as to the circumstances of evaluation of the relevant quantified sentence and people who do not so restrict it, so that what is true for the former is false for the latter. For the restriction, criterion is given in terms of a first-order supersubstantive property that for the restrictionists suitably restricts the scope of the existential quantifier. Taken in this way, one may again say that the disagreement between those people is genuine (for it does not depend on a meaning, but on a truthvalue, shift) and yet it is not faultless (for it depends on

182     A. Voltolini

the reasons one must suitably fix the relevant domain). However, I want to stress that there is just a structural similarity between this case and the cases where ontological disagreement is involved. For, literally speaking, in such cases there is no restriction of the overall scope of the existential quantifier, which is a contextual matter, but rather a noncontextual fixation of such a scope. In other terms, the latter kind of disagreement concerns how to determine the extension of the overall domain of the existential quantifier, whereas the former kind of disagreement takes for granted that there is such a determination, by only wondering whether a selection in that extension must be performed by means of a contextual restriction of that extension. This is why, unlike the previous first-order properties, the supersubstantive first-order property helping one to do the contextual restriction is not taken to be ontologically, hence existentially, relevant. For instance, ‘There is no beer’ is true for those who restrict the scope of the existential quantifier to the things located in the fridge, false for those who do not so restrict that scope. One may say that the former people just contextually take the first-order supersubstantive property of being located in the fridge as a universal property, by so restricting the scope of the existential quantifier, hence the circumstances of evaluation of the above sentence, while the latter people deny all that (perhaps because the former remark, while the latter reject, that the overall discourse topic is focalized on the fridge). Before leaving this section, a caveat. Since syncretism is clearly a compromise between views of existence as a first-order property and views of it as a second-order property, one may wonder how it sits with respect to other such conciliatory positions, such as deflationism about existence, at least as it figures in Bacigalupo (2016) version. According to this version of deflationism, existence is a redundant notion, for one may eliminate it away by means of the following (conceptual) schemas: (EES) n exist(s) if and only if sn (NES) n do(es) not exist if and only if it is not the case that sn For example, ‘Something red exists if and only if something is red’ and ‘Something red does not exist if and only if it is not the case that something is red’.

8  A Syncretistic View of Existence and Marty’s Relation to It     183

Yet to begin with, syncretism about existence is not deflationism about it, for it goes precisely in the opposite direction, by inflating the properties of existence. Moreover, it seems that deflationism does not provide the right account of sentences in which there are anaphoric links wrt existence. Consider for instance the following triad: ‘Both something red and something round exist’, ‘Both something chimerical and something unicornic do not exist’ and ‘Unlike something red, something chimerical does not exist’. Intuitively speaking, the use of ‘both’ in the first two sentences seems to indicate that the items involved share something, while the use of ‘unlike’ in the last sentence seems to indicate that the items involved do not share that very something. Yet no such indications appear to be captured by the following triad ‘Something is red and something is round’, ‘It is not the case that something is chimeric and it is not the case that something is unicornic’ and ‘Something is red yet it is not the case that something is chimerical’, whose members are supposed by the deflationist to be, respectively, equivalent with the corresponding members of the first triad. Finally, one may wonder whether the above schemas always work. For there are cases in which not only it may be disputed whether what instantiates the schema is true, but also such a dispute depends on a different precomprehension of the notions of existence involved. For instance, what about something like ‘Something dreamt-of exists if and only if something is dreamt-of ’, as to which people may dispute whether it is true, depending on how ‘exists’ is meant in the LHS of the biconditional (either as the first-order universal property or as the first-order nonuniversal property).

3 How Marty Situates Himself in This Debate Anton Marty definitely figures as a pluralist on being: for him as well as for Aristotle, being is said in many ways (Smith 1990: 113). In particular, being can be articulated into the property of reality or being real, which just some entities possess, and a property of existence (Smith 1990: ib). This latter property must be meant à la Brentano (20122) in terms of the property of being a fitting object of acceptance (see later).

184     A. Voltolini

To be sure, in virtue of this conception of existence, Marty is taken to be a monist about it; Marty considered himself to be such (cf., e.g., 1884: 173). Yet one may take the aforementioned property of being real to play an existential role as well, by playing the role of a mode of being that only real things possess. This move would one legitimate to rank Marty, his own conviction notwithstanding, as a pluralist about existence; from now on, I will work under the assumption that Marty is such. Yet being a pluralist, at least in spirit if not in letter, does not make one eo ipso a syncretist. For syncretism requires not only a plurality of notions of existence, but also that such notions are interconnected. So, is there any reason to hold that Marty is also a syncretist? In order to answer this question, to begin with, one may stress that Marty is a nonfundamentalist, i.e., someone who does not take spatiotemporal or causal existence, roughly coinciding with the aforementioned property of being real, as the universal property of beings. For him, there are both real things, those that enter into causal relations, and nonreal things, those that do not enter in such relations. The latter things are grounded in the former ones (Marty 1908). According to the list Smith (1990: 119) provides, the following are at least the paradigmatic cases of nonreal entities for Marty: • past and future realia • collectives • relations • space • states of affairs [Sachverhalte ] or ‘contents of judgments’ • values [Wertverhalte ] or ‘contents of phenomena of interest’. If we put things in terms of the framework I provided in the previous section, it turns out that Marty first of all allows for a first-order very substantive nonuniversal notion of existence: being real. For being real, which is at least coextensional with spatiotemporal existence, qualifies just a subset of the overall ontological domain, the subset of the fundamental things.12 To be sure, he ranks merely possible objects and even impossible objects, i.e. things that either do not exist (either spatiotemporally or non-spatiotemporally) or cannot exist (either

8  A Syncretistic View of Existence and Marty’s Relation to It     185

spatiotemporally or non-spatiotemporally), within the subdomain of the nonreal things (Smith 1990: 119), hence within the overall domain as well. Yet Marty is not a Meinongian strictly speaking, for he does not allow for Meinongian objects that are qualified by Meinong’s Principle of Freedom of Assumption (1960: 82), reprised nowadays as the Characterization Postulate: ‘According to the Characterization Postulate objects, whether they exist or not, actually have the properties which are used to characterize them, for example, where f is a characterizing feature, the item which fs indeed fs’ (Routley 1980: 46). In a nutshell, Marty allows for the merely possible golden mountain, as well as for the impossible round square, but not for the golden mountain or for the round square. In this respect, his ontological commitment is similar to that provided by a modal Meinongian, who believes in an amended version of CP. According to this version (Priest 20162; Berto 2013), the modal Meinongian replaces ‘actually’ with ‘(im)possibly’ in the Routley formulation, so that where f is a characterizing feature, the item which fs indeed (im)possibly fs. This version of the principle precisely lets one allow both for the merely possible golden mountain and for the impossible round square, which are, respectively, golden as well as mountainous and round as well as square yet not in the actual but in a possible vs. an impossible world. Moreover, qua nonfundamentalist, theoretically speaking at least Marty must allow for a further first-order property of existence that must however be taken as universal; namely, what I have labelled the substantive first-order property of existence (Meinong’s Sein ). For, as we saw in the previous section, this is the property that for nonfundamentalists captures what all things share in the overall ontological domain. Indeed, as Smith puts it, for Marty ‘non-real entities exist, no less than real entities, and these two categories together exhaust the totality of all the entities that there are in any sense at all ’(Smith 1990: 118; my italics). Yet clearly enough, Marty does not buy the Frege-Russell secondorder notion of existence, for, as I hinted at above, he shares and extends Brentano’s (20122) attitudinal account of ontological commitment. According to this account, to say that X exists is to say that X is a fitting object of acceptance, or better put, that the correct attitude to take towards X is that of accepting or believing in it (Kriegel

186     A. Voltolini

2015; Mulligan 2013).13 As is well known, Marty (1884, 1894, 1895) extended this account, which Brentano applied to categorical judgements as a whole, to subjectless statements as well. In this respect, first of all, ‘The tree is green’ means ‘The green tree exists’, as well as ‘It’s raining’ means ‘Rain exists’ (Schumann 1990: 199). Moreover, such existential sentences, respectively, mean that we have to accept the green tree as well as rain. To be sure, this sketch on Marty’s position about existence already shows that although in my theoretical reconstruction Marty counts as a pluralist, he cannot be seen as a standard triadic pluralist. For although he appears to be committed to two first-order properties of existence, a universal and a nonuniversal one, he does not believe in a second-order property of existence that captures the most fundamental sense of existence, the sense that captures ontological commitment. For this commitment has to be cashed out for him in attitudinal terms, as we have just seen. Nevertheless, one may say that although Marty does not believe in the idea that a property triad captures what existence overall amounts to, he still shares an actually non-standard triadic account where two first-order properties as well as an attitude mode are mobilized to provide such a capturing. At this point, one may further question whether Marty’s account may also count as a syncretistic account of existence. In theory, the answer to this question should be straightforwardly negative: in itself, pluralism, even in a non-standard form, is not syncretism. Yet one might rebut that, in order to properly answer this question, one must see whether also this new triad is made by properties and notions that, as a syncretist wishes, are all really ontological ones; that is, they are all relevant in order to understand what existence overall amounts to, insofar as they are interconnected. Thus, one must also check whether and how they interact in a Martyan perspective. In this respect, let us start again from the issue of ontological disagreement. How can one describe ontological disagreement in Marty’s account? Quite likely, in that account an ontological disagreement may be understood as a disagreement as to the scope of an attitude of acceptance vs. the scope of an attitude of rejection. Depending on the specific ontological approach one appeals to, the domain of objects that

8  A Syncretistic View of Existence and Marty’s Relation to It     187

may be affected by either attitudes vary. As we have seen, Marty himself accepts whatever is either real or nonreal. Yet unlike Meinong, he rejects whatever is beyond the real and the nonreal. In this respect, one may well say that, once again, the relevant first-order properties of existence serve to fix the scope of ontological acceptance, hence the overall ontological domain. Ontological disagreement, in fact, turns out again to be captured as a disagreement as to which first-order property of existence plays that fixation job. While for Marty, the substantive first-order property of existence, Sein, fixes that acceptance scope, for Meinong it is the formal first-order property of existence, being (an object), that does that job. Incidentally, note that if the notion of acceptance (as well as the corresponding notion of rejection) were not primitive, but it had to be cashed out in terms of the notion of representing something as existing, this would straightforwardly show that, in order to qualify his very notion of acceptance, Marty needs a substantive first-order property of existence, Sein, which he takes to be universal. Maybe this is however not the case. As Kriegel (2015) holds, accepting is not constituted by such an independent notion of existence, since it is just representing-as-existing, where this is just a single notion. Anyway, since the attitudinal mode of acceptance and the first-order properties of existence seem again to interact for ontological purposes, one might say that, once reconstructed along the above lines, Marty’s account also qualifies itself as a legitimate but specific syncretistic account of existence. Yet at this point, there are two problems as to whether in this framework the attitudinal mode may play the role the Frege-Russell secondorder property of existence plays in the syncretistic account of existence we saw in the previous section. First of all, it is unclear whether by means of the attitudinal approach one can suitably describe, hence understand, the particular aforementioned disagreement between a Meinongian and a non-Meinongian. This disagreement is very important for syncretistic matters, since it is the form of disagreement that should certainly mobilize the first-order formal universal property of existence. For, as we saw before, the Meinongian accepts whatever the non-Meinongian rejects. Yet in Brentano-Marty’s terms, acceptance and rejection are polar attitudes; that is, there cannot be acceptance of something if there is not also a rejection of something else (cf. Mulligan 2013). So, even if one

188     A. Voltolini

said that the scope of Meinongian acceptance is fixed by the first-order formal universal property of existence—as one might say, Meinongians accept absolutely everything—how can one understand the Meinongian’s overall acceptance as an acceptance? Thus, it seems that the disagreement between a Meinongian and a non-Meinongian cannot be reconstructed in Martyan attitudinal terms. If this is the case, Marty’s account cannot count as syncretistic, even in a sui generis form. For the first-order formal universal property of existence cannot play the same fixation role as in a syncretistic account like the one I sketched in the previous section. Moreover and more problematically, it is unclear whether appealing to attitudes of acceptance or rejection can really dispense a Martyan with the standard second-order property of existence, instantiatedness. As Kriegel (2015) states, what one accepts or rejects are intentional objects, namely the targets of the relevant attitude of acceptance or of rejection. For instance, a Meinongian accepts the intentional objects a Martyan rejects, e.g., as I said above, the golden mountain and the round square. Yet the Meinongian may then say that what prompts her to justify her overall acceptance is the fact that the sentence ‘There are intentional objects that are rejected by Martyans’ is true and thereby à la Quine commits everyone, Martyans included, to such objects. Now, if the Martyan were forced to accept this move, this would let the standard second-order property of existence come back from the rear door as far as ontological matters are concerned. So, if this were the case, modulo a positive solution to the previous problem, Marty’s allegedly non-standard syncretist account of existence should be rather traced back to the previous form of syncretism. For appearances notwithstanding, it would have to ultimately appeal to that second-order property in the existential role the aforementioned form of syncretism ascribes to it. (Clearly enough, if that solution could not be found, then Marty’s account could not be counted as syncretist, but it would have to be considered not only as pluralist as I said before, but as a standard form of pluralist appealing both to first-order and to second-order properties of existence. For the role played by the attitudinal conception of existence would still be resumed by the second-order property of existence.) What can the Martyan reply to these problems? Let us start from the ‘polarity’—problem. First of all, a Martyan may say that precisely

8  A Syncretistic View of Existence and Marty’s Relation to It     189

because acceptance is polar, the Meinongian overall acceptance is unintelligible. Yet this is unfair to the Meinongian. As we have seen before, her saying that there are things that do not exist, as expression of her overall acceptance, produces a sentence that is utterly intelligible. Or the Martyan may say that the fact that attitude and rejection are polar attitudes does not prevent one from adopting just one such overall attitude, just as the fact that love and hate are polar does not prevent one either from just loving everyone (the philanthrope) or from just hating everyone (the misanthrope). Yet in the case of love and hate, what saves their polarity is the fact that it is just contingently the case that one loves vs. hates everyone (compare the situation that holds with propositions: the fact that there is a possible world in which all propositions are true (false)14 does not prevent them from being polarly true or false). For the Meinongian, however, overall acceptance is not a contingent matter, since she accepts absolutely everything as a matter of necessity. So, her commitment to a nonpolar form of acceptance remains for a Martyan to be explained. Moreover, however the Martyan addresses the first problem, the second problem is even trickier. How can the Martyan avoid relying à la Quine on the second-order property of existence as a mark of ontological commitment, so that, oddly enough, modulo a solution to the previous problem, Marty’s alleged syncretism would fall back into the aforementioned form of syncretism? First of all, the Martyan may say that, appearances notwithstanding, the sentence ‘There are intentional objects that are rejected by Martyans’ is false, for according to the Martyans the overall ontological domain is exhausted by the accepted objects, i.e., the objects that are represented-as-existent (where existence, as a first-order universal property, amounts in this framework to be either real or nonreal). But even if she says this, she must acknowledge that there is an ontological disagreement with a Meinongian that is treated in terms of the ascription of different truthvalues to the simpler sentence ‘There are intentional objects that are rejected’. For Meinongians, the simpler sentence is false, yet for Martyans it is true, for it is a way of stating their ontologically negative position, e.g. about Meinongian objects—the golden mountain, the round square—that are thought-of. Hence once again, the Martyan

190     A. Voltolini

must acknowledge that matters of ontology mobilize the second-order property of existence, for they regard the issue of whether this property holds or not of the property of being a rejected intentional object.15 To be sure, the Martyan may try to say that even the simpler sentence ‘There are intentional objects that are rejected’ is false for another reason. Since for her the overall ontological domain is exhausted by the accepted objects, rejection of (some) intentionalia cannot for her be given a relational, but at most an adverbialist, hence monadic, construal à la Brentano in the second volume of his (20122) (on which cf., e.g., Kriegel 2007, 2011): there is no something to be rejected; rather, one is rejecting-somehow (e.g. one is rejecting-golden mountainwise). Incidentally, Marty’s (1908) conviction that (rejected) intentionalia must be dispensed with points precisely to this direction. Yet a Martyan cannot interpret acceptance in the same way, since for her acceptance establishes the ontological commitment. Hence, how can rejection be the attitude opposite to acceptance? Indeed, when two attitudes are polar, as is the case, e.g., with love and hate, they must be given the same construal, either relational or monadic. As a second option, the Martyan may say that the sentence ‘There are intentional objects that are rejected’ is true but not committal, for again the overall ontological domain is exhausted by the accepted objects (for a similar move, cf. Crane 2001, 2013, and Sainsbury and Tye 2012).16 But it is unclear how she can say both that the sentence is true and that the overall ontological domain is exhausted by the accepted objects. For given such an exhaustion, what is really true must be the contradictory of that sentence, namely ‘There are no rejected intentional objects’, viz. ‘It is not the case that there are rejected intentional objects’.17 As a third option, the Martyan may say that the sentence ‘There are intentional objects that are rejected’ is true but not committal even if the first-order property that is predicated in the above sentence, being rejected, is a nonuniversal one: some intentional objects are accepted and some others are rejected, just as some English kings died violently and some other did not (for a similar move, cf. again Crane 2013). But then it is hard to see how this option is not Meinongianism in disguise. For this is precisely what the Meinongian would say, by simply adding that the truth of the sentence commits one to intentional objects that are rejected.18

8  A Syncretistic View of Existence and Marty’s Relation to It     191

As a fourth option, the Martyan may say that the sentence ‘There are intentional objects that are rejected’ is just fictionally true; hence, it is not committal just as any other such sentence. Clearly enough, that ‘Holmes is a detective’ is fictionally true makes one commit to no entity whatsoever, let alone a fictional one (for a similar move, cf. Kroon 2011, 2013, 2015). That sentence is indeed true just in a fictional world where ‘Holmes’ refers to someone figuring just in that world. Yet this fact has no bearing at all as to the overall ontological domain; the fictional referent of ‘Holmes’ does not figure in that domain. Yet being merely fictionally true cannot be what the Martyan is looking for. For what she is not committed to must be something that she really rejects. For instance, if she does not believe in God or in fictional characters either, she cannot say that sentences like ‘There is something that is identical with God yet she is rejected’ and ‘There are fictional objects yet they are rejected’ are merely fictionally true; such sentences must be really true. To be sure, at this point the Martyan may play the radical fictionalist move that holds that all such sentences are really true iff they are fictionally such: being fictionally true is the truthmaker that is both necessary and sufficient in order to make those sentences also really true (for this move, cf. Walton 1990 and Crimmins 1998). Yet this move does not help, for being fictionally true cannot be sufficient in order for something to be also really true. As a matter of fact, being fictionally true is something that affects a sentence that is also interpreted in a fictional narrow context of interpretation (a context that provides that sentence with truthconditions and that is given in terms of a set of determinate parameters: agent, space, time and world [cf. Predelli 2005]). This context is different from a real context, for unlike the latter, its ‘world’ parameter is saturated by a fictional, not by the actual world. Yet that a sentence so interpreted is true in a fictional world says nothing as to that sentence being true in the actual world when interpreted in a real context of interpretation whose ‘world’ parameter is saturated by the actual world. The situation is structurally similar to that of a sentence that was interpreted in a possible context of interpretation, i.e., a context whose ‘world’ parameter is saturated by a merely possible world. If a sentence so interpreted were true in such a world, this would have no bearings as to the fact that such a sentence, when interpreted in a real

192     A. Voltolini

context of interpretation, was true in the actual world.19 In a nutshell, there is no crossworld relation that lets a sentence in our use be true in our world for in another use of theirs is true in their world. Thus, it seems that Marty’s appeal to acceptance cannot ultimately dispense with the second-order property of existence. As an overall result, if it turned out that, in my theoretical reconstruction, Marty’s allegedly distinctive pluralist approach to existence is, modulo a positive solution to the first aforementioned problem, a syncretistic approach, then, given the second problem, this would merely amount to its disguisedly being the form of syncretism I presented in the first section of this paper. Or, to put it alternatively, in order for Marty’s account to legitimately amount, given the first problem, to a form of syncretism and not to a mere (albeit non-standard) pluralist account of existence, given the second problem it must manage to show how it is not the above syncretism in disguise.20

Notes 1. The formulations of existence I am going to present accept that there are properties. If one is a nominalist about properties, one should obviously put forward corresponding nominalist formulations of existence that just mobilize predicates instead of properties. For instance, Quine (1952) version of the second-order approach to existence is surely a nominalist version of that approach. On this cf., e.g., van Inwagen (2008: 37–38). Yet such a version is hardly an alternative to the traditional Frege-Russell account if existentially quantified sentences can figure as LHS of biconditionals of the kind ‘there is a F iff the property of being F is instantiated’, what Schiffer (2003) calls ‘“something from nothing”- transformations’. As the RHS of those biconditionals is committed to second-order properties, it is hard to deny that the same holds to the LHS of such biconditionals. 2. For a recent revival of this idea, cf. McDaniel (2009, 2017). 3. See however fn. 1. 4. On this problem, cf. Kriegel (2015). 5. On this problem, cf. Voltolini (2012). For other critiques against the second-order conception of existence, cf. McGinn (2000).

8  A Syncretistic View of Existence and Marty’s Relation to It     193

6. On this problem, cf. again Kriegel (2015). 7. The example comes from Evans (1982: 372). 8. As Lewis (1990) originally intuited: for him, a Meinongian such as Routley ‘affirms the existence of all the controversial entities (as we may call them). He does not join us when we dodge questions about some of these alleged entities by denying that they exist’ (1990: 29). ‘If “existence” is what [Routley] thinks it is—a distinction among the items we are committed to—then we dispense with existence’ (1990: 30). 9. For a reconstruction of Meinong’s own position on this matter, cf. Marek (20132). 10. As Lewis (1990) originally suggested. See note 8. 11. Priest (20162) denies this interpretation of the neutral quantifier. But then he is forced to admit that the disagreement between the non-­ Meinongians and the Meinongians concerns the different meanings they give to their respective quantifiers. 12. This is an idea reprised by Fine (2009) along with many others. 13. The idea that believing in something captures ontological commitment has been recently reprised by Szabo (2003). According to Szabo, believing in something amounts to thinking of something while having a conception of it. So conceived, however, ‘believing in’ can hardly amount to a Martyan form of acceptance, since it is something also Meinongians may endorse. For instance, a Meinongian may well think of the round square qua round square, even though the round square is something a Martyan would reject. 14. This is the case that Wittgenstein (1961: §2.013) described as the possibility of logical space’s being empty. On this, see, e.g., Frascolla (2006). 15. Incidentally, clearly enough the sentence ‘There are intentional objects that are rejected’ cannot be paraphrased by a Martyan in her attitudinal terms roughly as follows: ‘Intentional objects that are rejected are accepted’. For instead of being true, this paraphrase comes close to a contradiction. 16. In comparing Szabo (2003) ‘believing-in’ approach to existential matters, which is a very recent version of the attitudinal approach, Crane says that it is quite compatible with his attempt at taking existentially quantified sentences of the above kind as being both true and noncommittal ontologically. Cf. Crane (2013: 50). To be sure, Crane might be right, if the ‘believing in’ approach were not Meinongianism in disguise (see the problem raised in note 13).

194     A. Voltolini

1 7. For a similar problem in Crane, cf. Voltolini (2016, 2018). 18. On this problem, cf. again Voltolini (2016, 2018). 19. Cf. again Voltolini (2018). 20. I warmly thank Laurent Cesalli, who discussed a preliminary version of this paper at the conference Anton Marty and Contemporary Philosophy, June 15–17, 2017, University of Geneva, for his important comments to such a version. I am also very grateful to Giuliano Bacigalupo for having insightfully commented and edited the paper.

References Bacigalupo, G. 2016. Whose Existence? A Deflationist Compromise to the Fregean/Neo-Meinongian Divide. Argumenta 3: 5–24. Berto, F. 2013. Existence As a Real Property. Dordrecht: Springer. Brentano, F. 20122. Psychology from an Empirical Standpoint. London: Routledge. Crane, T. 2001. Elements of Mind. Oxford: Oxford University Press. ———. 2013. The Objects of Thought. Oxford: Oxford University Press. Crimmins, M. 1998. Hesperus and Phosphorus: Sense, Pretense, and Reference. The Philosophical Review 107: 1–47. Evans, G. 1982. The Varieties of Reference. Oxford: Clarendon Press. Fine, K. 2009. The Question of Ontology. In Metametaphysics: New Essays on the Foundations of Ontology, ed. D.J. Chalmers, D. Manley, and R. Wasserman, 157–177. Oxford: Oxford University Press. Frascolla, P. 2006. Understanding Wittgenstein’s Tractatus. London and New York: Routledge. Frege, G. 19742. The Foundations of Arithmetic. Oxford: Blackwell. Kriegel, U. 2007. Intentional Inexistence and Phenomenal Intentionality. Philosophical Perspectives 21: 307–340. ———. 2011. The Sources of Intentionality. Oxford: Oxford University Press. ———. 2015. How to Speak of Existence: A Brentanian Approach to (Linguistic and Mental) Ontological Commitment. Grazer Philosophische Studien 91: 81–106. Kripke, S. 1980. Naming and Necessity. Oxford: Blackwell. Kroon, F. 2011. The Fiction of Creationism. In Truth in Fiction, ed. F. Lihoreau, 203–222. Frankfurt: Ontos.

8  A Syncretistic View of Existence and Marty’s Relation to It     195

———. 2013. The Social Character of Fictional Entities. In From Fictionalism to Realism: Fictional and Other Social Entities, ed. C. Barbero, M. Ferraris, and A. Voltolini, 87–110. Newcastle upon Tyne: Cambridge Scholars. ———. 2015. Creationism and the Problem of Indiscernible Fictional Objects. In Fictional Objects, ed. S. Brock and A. Everett, 139–171. Oxford: Oxford University Press. Lewis, D. 1990. Noneism or Allism? Mind 99: 23–31. Marek, J. 20132. Alexius Meinong. In The Stanford Encyclopedia of Philosophy, ed. E.N. Zalta (Fall 2013 Edition). https://plato.stanford.edu/archives/ fall2013/entries/meinong/. Marty, A. 1884. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, I–III. Vierteljahresschrift für wissenschaftliche Philosophie 8: 56–94, 161–192, 292–340. ———. 1894. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, IV–V. Vierteljahresschrift für wissenschaftliche Philosophie 18: 320–356, 421–471. ———. 1895. Über subjectlose Sätze und das Verhältnis der Grammatik zu Logik und Psychologie, VI–VII. Vierteljahresschrift für wissenschaftliche Philosophie 19: 19–87, 263–334. ———. 1908. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, vol. I. Halle: Niemeyer. McDaniel, K. 2009. Ways of Being. In Metametaphysics: New Essays on the Foundations of Ontology, ed. D.J. Chalmers, D. Manley, and R. Wasserman, 290–319. Oxford: Oxford University Press. ———. 2017. The Fragmentation of Being. Oxford: Oxford University Press. McGinn, C. 2000. Logical Properties. Oxford: Clarendon Press. Meinong, A. 1960. On the Theory of Objects. In Realism and the Background of Phenomenology, ed. R. Chisholm, 76–117. New York: Free Press. Meinong, A. 1978. Über Inhalt und Gegenstand. In Kolleghefte und Fragmente. Ergänzungsband zur Gesamtausgabe, 145–159. Graz: Akademische Druckund Verlagsanstalt. Mulligan, K. 2013. Acceptance, Acknowledgement, Affirmation, Agreement, Assertion, Belief, Certainty, Conviction, Denial, Judgement, Refusal and Rejection. In Judgement and Truth in Early Analytic Philosophy and Phenomenology, ed. M. Textor, 97–137. Basingstoke: Palgrave. Predelli, S. 2005. Contexts. Oxford: Oxford University Press. Priest, G. 20162. Towards Non-Being: The Logic and Metaphysics of Intentionality. Oxford: Oxford University Press.

196     A. Voltolini

Quine, W.V.O. 1952. On What There Is. In Semantics and the Philosophy of Language, ed. L. Linsky, 189–206. Urbana: University of Illinois Press. Routley, R. 1980. Exploring Meinong’s Jungle and Beyond. Canberra: Australian National University. Russell, B. 1905. On Denoting. Mind 14: 473–493. ———. 19372. The Principles of Mathematics. Cambridge: Cambridge University Press. Sainsbury, M., and M. Tye. 2012. Seven Puzzles of Thought and How to Solve Them. Oxford: Oxford University Press. Schiffer, S. 2003. The Things We Mean. Oxford: Clarendon Press. Schuhmann, K. 1990. Contents of Consciousness and States of Affairs: Daubert and Marty. In Mind, Meaning and Metaphysics, ed. K. Mulligan, 197–214. Dordrecht: Kluwer. Smith, B. 1990. Brentano and Marty: An Inquiry into Being and Truth. In Mind, Meaning and Metaphysics, ed. K. Mulligan, 111–149. Dordrecht: Kluwer. Szabo, Z. 2003. Believing in Things. Philosophy and Phenomenological Research 66: 584–611. van Inwagen, P. 2008. McGinn on Existence. The Philosophical Quarterly 58: 36–58. Voltolini, A. 2012. All Existences That There Are. Disputatio 32: 361–383. ———. 2016. Tim Crane. The Objects of Thought, Dialectica 70: 245–252. ———. 2018. Can One Refer to and Quantify Over Nonexistent Intentional Objects (of Hallucination)? In Objects of Inquiry (in Philosophy of Language and Literature), ed. P. Stalmaszczyk, 97–115. Bern: Peter Lang. Walton, K.L. 1990. Mimesis as Make-Believe. Cambridge, MA: Harvard University Press. Williamson, T. 2002. Necessary Existents. In Logic, Thought and Language, ed. A. O’ Hear, 233–251. Cambridge: Cambridge University Press. Wittgenstein, L. 1961. Tractatus Logico-Philosophicus. London: Routledge & Kegan Paul.

9 Misleading Pictures, Temptations and Meta-Philosophies: Marty and Wittgenstein Kevin Mulligan

1 Philosophy and Austrian Sprachkritik The first three substantive contributions to the philosophy of language in the Brentanian tradition are (parts of ) Husserl’s 1900–1901 Logische Untersuchungen, Martinak’s 1901 Psychologische Untersuchungen zur Bedeutungslehre and Marty’s 1908 Untersuchungen zur Grundlegung einer allgemeinen Grammatik und Sprachphilosophie.1 They contain many good ideas which only began to be taken seriously much later during the twentieth century. One is Husserl’s account of the nature of and differences between proper names, occasional expressions (demonstratives and indexicals) and what were later called definite descriptions, in particular his account of the difference between proper names which “name K. Mulligan (*)  University of Italian Switzerland, Lugano, Switzerland e-mail: [email protected] K. Mulligan  University of Geneva, Geneva, Switzerland © The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3_9

197

198     K. Mulligan

directly” and proper names which refer thanks to definite descriptions. Another is the distinction drawn by Martinak and Marty between what is now called natural and non-natural meaning and Marty’s account of meaning in terms of a speaker’s intentions to modify the doxastic and conative states of his interlocutor and of linguistic rules. Marty’s treatise also contains the first systematic development of the idea that philosophers are often misled by pictures or images (Bild ). Wittgenstein’s later claim that philosophers who think of philosophy as a type of science are always the victims of such misleading pictures is a more ambitious version of Marty’s view. But both Marty and Wittgenstein attach great importance to the claim that philosophers are tempted into error by pictures which lie ready in language (cf. Mulligan 1990b: 23; 2012: 39–42). Their reflections on misleading pictures are a part, perhaps the most important part, of a much more general Austro-Hungarian phenomenon—the critique of language (Sprachkritik ), which is to be found in the writings of Brentano, Karl Kraus, Hugo von Hofmannsthal and, for example, Fritz Mauthner, amongst many others. “All philosophy”, Wittgenstein declares in the Tractatus, “is critique of language (But not in Mauthner’s sense)” (Wittgenstein 1998: 4.0031). Marty does not agree that all philosophy is critique of language. For “everything which can be called critique of language” an important preliminary is “descriptive semasiology”, that is to say, the philosophy of language (Marty 1950a: 23) and “philosophical grammar” (Marty 1920b: 136). If Wittgenstein is to be believed, philosophy can aspire only to the critique of language, in particular, to the task of identifying the pictures which tempt us to philosophise. If his Swiss predecessor is right, the important task of identifying the pictures which mislead philosophers must be based on and cannot replace philosophy. I shall return to this central difference between Marty and Wittgenstein in Sect. 6. But I begin by looking at three families of misleading pictures which Marty and Wittgenstein identify as such: pictures of modality (Sect. 2), of thinking and its object (Sect. 3), of logic and propositions (Sect. 4). I then consider their accounts of the temptations to which philosophers who are misled by pictures succumb and of the ways pictures and temptations work (Sect. 5). I conclude with some remarks about the predecessors of Marty and Wittgenstein (Sect. 7).

9  Misleading Pictures, Temptations and Meta-Philosophies …     199

2 Pictures of Modality There is, Wittgenstein says, a misguided way of thinking of, of looking at modal facts. His description of it contains distinctions which recur in many of his descriptions of misleading pictures: There are…various reasons which incline us to look at the fact of something being possible, someone being able to do something, etc., as the fact that he or it is in a particular state. Roughly speaking, this comes to saying that “A is in the state of being able to do something” is the form of representation we are most strongly tempted to adopt; or, as one could also put it, we are strongly inclined to use the metaphor of something being in a peculiar state for saying that something can behave in a particular way. And this way of representation, or this metaphor, is embodied in the expressions “He is capable of…”, “He is able to multiply large numbers in his head”, “He can play chess”: in these sentences the verb is used in the present tense, suggesting that the phrases are descriptions of states which exist at the moment when we speak. (Wittgenstein 1972: 117)

In 1908, Marty refers to a misguided way of thinking of possibilities, all possibilities, not merely abilities, and, unlike Wittgenstein, indicates the culprits he has in mind. He refers to the common perpetration of the fiction which consists in bringing what is not real…, in thoughts about it, far too close (heranrücken ) to the real. Think, for example, of the important roles played by the possibilities of sensation in the metaphysics of a man such as J. St. Mill and by the doctrine of forces and energies…in modern philosophy of nature. Forces, that is, possibilities of acting, and…possibilities of undergoing and indeed all possibilities are, of course, merely something non-real. But they are treated as things, which have effects and are effected….And one proceeds in the same way in a thousand other cases. (Marty 1976: 354)

Marty’s “real” refers to what is causally efficacious. How do we treat possibilities and impossibilities as things and as being in time? We employ what Marty calls an inner linguistic form, that is to say, a conceptual presentation which has a certain function, that of directing an interlocutor’s attention to what the speaker has in mind:

200     K. Mulligan

All our names have as their inner linguistic form either the presentation of a substance or of an accident, thus of something real. But we always designate the non-real, too, indeed even what is completely fictitious, with the help of a substantive (such as:….possibility, impossibility etc.) and then there is the presentation of a thing as a pictorial, inner linguistic form, or with the help of an adjective which is attributed (beilegt ) to a real or apparent subject as a predicate or attribute, and then the presentation of inherence is in play, as is in fact the case with real accidents and substances (Marty 1976: 354–355)

To talk about possibilities is to use names or adjectives and to think of possibilities as substances or as inhering in substances like real accidents. To think in this way is to think with the help of pictorial, inner linguistic form, for no possibilities are substances and no possibilities inhere in substances. An accident is either permanent or variable (Marty 1916: 258), it is, for example, a rising from some position or it is a being seated. Thus not seriously but pictorially, we conceive of a subject as a substance, the predicate as something which inheres in it (as…a state, doing, or undergoing). (cf. Marty 1950b: 51; 1965: 147; my emphasis—KM, )

The assimilation of possibility to something like a state of sadness, Marty asserts, has disastrous consequences: ….language uses expressions for what is real also for what is non-real. And taking these pictures for what is really meant is here, in the field of metaphysics, much more disastrous, because of its much further-reaching consequences, than when a physicist takes for the truth what is, in his field, merely a picture and an attempted illustration. (Marty 1976: 356)

Wittgenstein’s account of the disaster resembles Marty’s. Wittgenstein gives toothache as an example of a state (Wittgenstein 1968: 96) and says, as we have seen, that possibilities are not states. “Where”, he asks, “did our queer ideas come from?” Well, I show you the possibility of a movement, say by means of a picture of the movement: ‘so possibility is something which is like reality’.

9  Misleading Pictures, Temptations and Meta-Philosophies …     201

We say: “It isn’t moving yet, but it already has the possibility of moving”——‘so possibility is something very near reality’. (Wittgenstein 1968: §194)

Wittgenstein’s misguided interlocutor thinks of possibility as like or near reality. Marty’s misguided metaphysician brings possibilities, in thought, far too close to reality. She is misled by a picture of possibility. Marty thinks that the picture may mislead us in thinking about all possibilities. He would thus perhaps have agreed with Wittgenstein that it is wrong to call the ability to solve a mathematical problem, the ability to enjoy a piece of music states of mind.

3 Pictures of Thinking and Its Object A family of closely related pictures of thinking or consciousness systematically misleads philosophers, including the philosopher in all of us, and the resulting errors, Marty and Wittgenstein say, are ancient: The secret of consciousness stimulated reflection in the old Ionians and, trying to get to the bottom of the puzzling fact that something which is not identical with the presenter or knower can be present to her, they hit upon the solution that something which is not what is presented or known but essentially similar…to it in the soul must be given if presenting and knowing are to be possible. (Marty 1976: 385)

The ancient mistake Marty refers to is the claim that presentations or knowledge requires that something similar to what is presented or known be in the soul. Theaetetus, quoted by Wittgenstein, on the other hand, thinks that whatever is thought must be real: Socrates to Theaetetus: “And if someone thinks (vorstellt ) must he not think something?” – Th.: “Yes, he must”. Soc.: “And if he thinks something, must it not be something real (Wirkliches )?” Th.: “Apparently”. (Wittgenstein 1968: §518)

202     K. Mulligan

The German locution “etwas vorstellen ” used in the translation of Plato quoted by Wittgenstein is just the locution employed by Brentano and all his pupils: to be presented with something, to have a presentation of something ((sich ) etwas vorstellen ). It is a locution Marty prefers to what he thinks of as the “confused” notion of thinking. Wittgenstein does not share this preference: …the question: “How can one think what is not the case?” - This is a beautiful example of a philosophical question. It asks “How can one…?” and while this puzzles us we must admit that nothing is easier than to think what is not the case….the difficulty which we are in… arises when we look at the facts through the medium of a misleading form of expression (Wittgenstein 1972: 30–31)

Marty’s 1908 diagnosis of the sources of certain puzzles about intentionality strikes a confessional note (which is also present in some of Wittgenstein’s identifications of misleading pictures): Even today, many a philosopher distinguishes between the real object of our presenting or of our thinking and an object immanent in the thinker (I myself used to do this) and in so doing appeals to the claim that without this distinction our thinking could not be essentially a “relation to an object” (Objektsbeziehung ). For it is not always the case that a real object is given;…in other words, the judgment that what is presented really exists is not always correct. If therefore presenting is to have something presented as a correlate, this correlate which is always required can only be an immanent object (Marty 1976: 387–388)

This apparent correlation may seem to be the relation of representation: One can think (vorstellen ) that something is the case which is not the case: quite remarkable! For it is not remarkable that the thought (Vorstellung ) does not agree with reality; what is remarkable, is that it then represents (repräsentiert ) it. (Wittgenstein 2005: 363)

The conclusion of the almost fifty pages Marty devotes to examining the philosophy of immanent objects (which critically discuss the views of Aristotle, Anselm, Aquinas, Ockham, Kant and others) runs as follows:

9  Misleading Pictures, Temptations and Meta-Philosophies …     203

One wanted to get to the bottom of the secret of consciousness and in so doing took more or less seriously a linguistic picture used in the description of the peculiar process. The more abstract locution of the immanence or Innewohnung of what is thought in the thinker (and likewise of what is felt in one who feels), like the more concrete locution of having what is thought or felt in the head or heart is, in my opinion, only justified as a fiction of pictorial, inner, linguistic form, which may offer certain stylistic advantages (concision, comfort, variety) but leads to a falsification as soon as it is taken more seriously. (Marty 1976: 397)

Wittgenstein notes one of the absurd consequences of the picture: I see someone pointing a gun and say ‘I expect a bang’. The shot is fired.—Well, that was what you expected; so did that bang somehow already exist in your expectation? (Wittgenstein 1968: §442)

A claim central to Brentano’s philosophy of mind, which Marty accepts, is that emoting, conating and judging contain and presuppose presentings. Thus, to be the victim of misleading pictures of presentings is eo ipso to be misled about emotions and desires. “Like the immanent object of the presentation, its analogue, the immanent object of judgment and of interest [emoting and conating] is a fiction” (Marty 1976: §94, 399). Wittgenstein agrees: To what extent can one call wishes, expectations, beliefs “unsatisfied”? What is our ur-picture (Urbild, prototype) of non-satisfaction? Is it a hollow space? And would one call that unsatisfied? Wouldn’t this be a metaphor too?… (Wittgenstein 1968: §439)

If non-satisfaction is thought of as a hollow space, satisfaction may well be thought of as filling empty space. There is, says Marty, a “picture of a vessel (Gefäß, container)…into which the object of consciousness is so to speak absorbed” (Marty 1976: 117). Wittgenstein’s question about the extent to which one can call wishes, expectations, beliefs “unsatisfied” is not a question asked by Marty. It is a question perhaps first asked, and answered to his own satisfaction, by Husserl, in his account of “the satisfaction (Erfüllung, fulfilment) of an…expectation, of a hope, of a wish” (Husserl 1984a: V §29).

204     K. Mulligan

In their discussions of thinking or presenting and its object, Marty and Wittgenstein identify two misleading pictures: of Doppelgänger (Marty) and shadows (Wittgenstein). Friends of such abstract, non-real, entities as propositions, Marty says, think of these as Doppelgänger. Bolzano’s Sätze an sich and their non-propositional parts, Vorstellungen an sich, Husserl’s ideale Aussagebedeutungen, non-propositional meanings, his states of affairs (Sachverhalte ) and other ideal objects, as well as Meinong’s objectives, are some of Marty’s exhibits for the way misleading pictures lead to Platonism. But friends of private psychological objects,2 Marty (1976: 392–394) argues, are also misled by the picture of the Doppelgänger, as are those who think of mental objects, contents or propositions as intermediaries between thinkers and the world, and those who attribute a particular mode of being to immanent objects: Is it not…an illusion to think that the so called mental object mediates consciousness of a real object? The thought of such a representative or mediating role is based on the presupposition that [the mental object] is a Doppelgänger of [the real object]… (Marty 1976: 395) …‘immanent’ objects are simply a fictitious Doppelgänger of objects tout court (Marty 1976: 410) There is…a sense in which one can speak of a mental existing-in (Inexistenz ) of the objects of our consciousness. But this way of talking is merely a picture. What really exists is not a peculiarly modified Doppelgänger of the real object which inhabits us…(Marty 1976: 415–416)

According to Wittgenstein, when we note that “the object of our thought” is not always a fact, this may lead us to think that “it is a shadow of the fact. There are different names for this shadow, e.g. ‘proposition’, ‘sense of the sentence’”. This then leads us to wonder: “How can something be the shadow of a fact which doesn’t exist?” (Wittgenstein 1972: 32, cf. 36). To the temptation to see the object of thought as a shadow, there corresponds the temptation to see the object of expectation as a shadow:

9  Misleading Pictures, Temptations and Meta-Philosophies …     205

The way out of this difficulty seems to be: what we expect is not the fact, but a shadow of the fact;…we say “Surely two sentences of different languages can have the same sense”; and we argue, “therefore the sense is not the same as the sentence”, and ask the question “What is the sense?” And we make of ‘it’ a shadowy being, one of the many which we create when we wish to give meaning to substantives to which no material objects correspond (Wittgenstein 1972: 36)

Wittgenstein, then, sees the picture of a shadow at work where Marty sees the picture of a Doppelgänger. Already in 1901, Husserl rejects the accusation of an imaginary interlocutor to his account of presentation or thinking that it is committed to shadows: to say that the object of a presentation is only an intentional object is not to say that there is in the presentation “any sort of shadow of the object” (Husserl 1984a: V §21, Appendix). According to Marty, just as it is wrong to think of the objects of presentations or thoughts as Doppelgänger, so too, it is wrong to think of the objects of emotings and conatings as Doppelgänger. This is indeed a consequence of the already mentioned view Marty inherits from Brentano according to which all emotings and conatings (which Marty subsumes under the heading “phenomena of interest”) inherit their objects from the presentings on which they depend: We have seen that the assumption of immanent judgment-contents and interest-contents, like that of immanent objects of presentation turns out to be a fiction. In the attempt to get to the bottom of consciousness in its how and its what one should have been on one’s guard against taking linguistic pictures for conceptual clarification and this is just what happened when one attributed to what is presented, judged etc. a sort of existence or immanence in the presenter or judger. – Yet the question arises of the inducement (Anlass ) to such a picture…Even if there is not, as one believed, a subject-object relation in the sense of a correlation in all consciousness, is there not something similar which was the inducement for believing in such a correlation in all cases? (Marty 1976: 405–406)

Wittgenstein suggests that the tendency to assume a pure intermediary (Mittelwesen ) “between the propositional signs and the facts” stems from

206     K. Mulligan

a pursuit of “chimeras” induced by “our forms of expression”, which “prevent us in all sorts of ways from seeing that nothing out of the ordinary is involved” in our use of signs (Wittgenstein 1968: §94). There are, Wittgenstein thinks, many shadowy beings which we create “when we wish to give meaning to substantives to which no material objects correspond” (Wittgenstein 1972: 36). As Marty puts it, “we always designate the non-real,…even what is completely fictitious, with the help of a substantive” (Marty 1976: 354–355). We create “nonthings” (Marty), “Undinge ”, chimeras (Wittgenstein).

4 Pictures of Logic and Propositions One picture of logic in the Tractatus is that of a framework or scaffolding (Gerüst ). “The logical framework around the picture determines logical space” (Wittgenstein 1998: 3.42). “The proposition constructs a world with the help of a logical framework” (4.023). “Logical propositions describe the framework of the world or rather they represent it. They ‘treat’ of nothing” (6.124). It is a picture to which Wittgenstein returns in the Investigations: …logic presents an order, in fact the a priori order of the world: that is, the order of possibilities, which must be common to both world and thought…. the order existing between the concepts of proposition, word, proof, truth, experience, and so on. This order is a super-order between– so to speak–super-concepts… (Wittgenstein 1968: §97)

As we shall see, if logic is to be thought of as a framework in this way, then a certain picture of propositions is required, a picture which springs to mind when we ask about the essence of propositions and the question “sees in the essence….something that lies beneath the surface” (Wittgenstein 1968: §92). Marty, too, was familiar with the picture of logic as a framework and did not approve of it. In 1908, he quotes Husserl’s 1900–1901 philosophy of logic:

9  Misleading Pictures, Temptations and Meta-Philosophies …     207

Within pure logic a first, basic sphere delimits itself, the pure theory of meaning-forms; that is, the theory of pure categories of meaning and the laws of compounding or modification which are grounded a priori in these. It lays bare the ideal framework (Gerüst ) which every actual (factische ) language fills up and clothes differently, following common human motives or variable empirical motives. To whatever extent the actual content and grammatical forms of historical languages are thus empirically determined, each is bound to this ideal framework…. (Husserl 1984a: IV §14; emphases mine—KM)

The first, basic sphere of logic referred to by Husserl is that of sense and nonsense, the sphere of what was to be called later categorial grammar. Marty does not think Husserl has really been taken in by the picture of a framework. Nevertheless, he comments, the picture of a framework which the different languages, following variable empirical motives, are supposed to have filled out and dressed in different ways does not seem to me to be suitable (Marty 1976: 59–60; emphases mine—KM)

He proposes an alternative: The tissue of the elementary meaning-categories stands to real languages and their makers more like a pattern (Vorlage, model), which they try to trace,…not as a frame (Rahmen ) which would stands before the consciousness of all in the same way and which they would merely fill out in different ways (Marty 1976: 59)

But in 1913, in the second edition of his Logical Investigations, Husserl’s formulation of his view is even more emphatic, although in a concession to Marty, he italicises the expression ideal framework: the basic forms of propositions, the categorical proposition with its many particular patterns and forms of members, the primitive types of complex propositions, e.g., the conjunctive, disjunctive, and hypothetical propositional unities, the differences of universality and particularity,….— all

208     K. Mulligan

these matters are entirely a priori, rooted in the ideal essence of meaning as such….an absolutely fixed “ideal framework”. Grammarians….must keep the…system of pure meaning-forms in mind, if they want to ask significantly how German, Latin, Chinese, etc. express “the” existential proposition, “the” categorical proposition, “the” hypothetical antecedent, “the” plural, “the” modalities of “possible” and “probable”, the “not”, and so on. (Husserl 1984a: IV §14B; emphases mine—KM)

The idea of logic as a framework of which Marty and the later Wittgenstein disapprove is bound up with conceptions of propositions some elements of which we have already come across: the conceptions of propositions as ideal entities, as intermediaries, as shadows, as immanent or private objects. It is also, as we shall see, bound up with the idea that propositions enjoy a special sort of unity. Wittgenstein presumably has in mind some of these conceptions in the following passage: One person might say “A proposition is the most ordinary thing in the world” and another: “A proposition—that’s something very queer!”– And the latter is unable simply to look and see how propositions really work. The forms that we use in expressing ourselves about propositions and thought stand in his way - Why do we say a proposition is something remarkable? On the one hand, because of the enormous importance attaching to it. (And that is correct). On the other hand this, together with a misunderstanding of the logic of language, seduces us into thinking that something extraordinary, something unique, must be achieved by propositions.–A misunderstanding makes it look to us as if a proposition did something queer (Wittgenstein 1968: §93)

A misunderstanding makes it look as though propositions represent states of affairs or the world in a sui generis and irreducible way. A way which is made possible by the special sort of unity they enjoy: We recognize that what we call “sentence” and “language” is not the formal (formelle ) unity I imagined, but is the family of more or less related structures (Wittgenstein 1968: §108)

9  Misleading Pictures, Temptations and Meta-Philosophies …     209

Part of what friends of the formal unity of the proposition overlook is the fact that they illustrate “what a proposition is…by a few commonplace examples” and then presuppose it “as understood in full generality”; this is the “basic evil of Russell’s logic, as also of mine in the Tractatus ” (Wittgenstein 1984a: I §38). Wittgenstein quotes and comments on his earlier endorsement of this kind of unity: (Logico-Philosophical Treatise 4.5): “The general form of propositions is: This is how things are (Es verhält sich so und so ).”–That is the kind of proposition that one repeats to oneself countless times. One thinks that one is tracing the outline of the thing’s nature over and over again, and one is merely following the form through which we look at it (Wittgenstein 1968: §114) A picture held us captive (Wittgenstein 1968: §115)

Husserl, too, had endorsed the idea of a general form of the proposition, “the ultimate point of unity to which all things logical must be traced back” (Husserl 1984a: V §40); a proposition “is ‘directed’ towards the things themselves; it says, things are so (so ist es )” (Husserl 1984b: VI §39). To employ “the form of the proposition (Form des Aussagesatzes ) is to say…things are so” (Husserl 1984a: I §11). This is “the categorial idea of a proposition” (Husserl 1984a: IV §13). If I assert something, I think of things, and that things stand thus and so (dass sich die Sachen so und so verhalten ): this is what I express, and perhaps also know. (Husserl 1984a: VI §67)

Marty, on the other hand, rejects the formal unity of the Satz, of verbs and of predication: The term Satz, if taken as widely as is usually the case, obviously does not correspond to a natural class (Marty 1950b: 18) …a unified description of the so called verb in terms of its function is impossible. This part of speech corresponds not to one but to many functions;

210     K. Mulligan

but one or more of these functions may be present and we still do not speak of a verb. (Marty 1950b: 61) It is simply wrong to think that the sense of predication is always the same. (Marty 1965: 139)

According to Marty, natural languages abound in what Wittgenstein was later to call family resemblance terms (cf. Mulligan 1990a). Marty himself has no term for such terms. But the philosopher who did more to develop Marty’s philosophy of language than any other, Karl Bühler (1982), takes over (from von Kries) the barbarous term “synchytic (synchytische ) concepts” for concepts the extensions of which contain objects between which only partial, overlapping similarities hold. The term derives from the Greek for fusion or merging. Wittgenstein was familiar with this way of thinking of family resemblance terms: And there are words of which one might say: They are used in a thousand different ways which gradually merge into one another. No wonder that we can’t tabulate strict rules for their use (Wittgenstein 1972: 28)

Bühler formulates Marty’s view in the following reference to the very ramified and highly synchytic central concepts of a domain, the sort of concepts that are formed in everyday language and remain undefined even when they reach deep into the sciences. As a paragon of this sort of concept, the concept of the sentence deserves particular attention from the logician of the human sciences. It involves a multiple synchysis that must not be dissolved as long as it is a philological concept and is supposed to remain one. (Bühler 1982: 356)3

Perhaps the most famous picture in Wittgenstein’s writings is what has been called the Augustinian picture of the essence of human language at the beginning of the Investigations. The main elements of this picture had been described and rejected by Bühler in 1909, the year in which he published his long and positive review (Bühler 1909b) of Marty’s Investigations:

9  Misleading Pictures, Temptations and Meta-Philosophies …     211

The old view was based essentially on two assumptions, that are internally connected. - It was believed that the functions of language could all be traced back to the naming function of words: every word is a name for something, its meaning (Bedeutung)…. And it was thought that the sentence contains essentially an aggregate (Inbegriff) of names (Nennungen). And in accordance with this first assumption the processes of language learning were accounted for as [a process of ] learning to name objects. - Both claims are false; the function of naming is only one of several functions of words and the fact that language learning is not based only on acquisition of the naming function is being shown more and more by systematic observation of children. Matters are essentially more complicated than they seemed to the first simple theory; just how complicated they are cannot be somehow deductively inferred but must be grasped on the basis of systematic observation of concrete cases of linguistic understanding. (Bühler 1909a; italics mine—KM)

Perhaps the main difference between Bühler’s description of what he takes to be a pervasive misunderstanding of the functions of language and Sect. 1 of Wittgenstein’s Investigations is that Wittgenstein, unlike Bühler, calls the misunderstanding a picture. Another is that Bühler quotes Hobbes as a friend of the view of language he criticises, whereas Wittgenstein quotes and comments on Augustine: a particular picture of the essence of human language…: the individual words in language name objects–sentences are combinations (Verbindungen) of such names (Benennungen).–In this picture of language we find the roots of the following idea: Every word has a meaning (Bedeutung). The meaning is correlated with the word. It is the object for which the word stands.- Augustine does not speak of there being any difference between kinds of word. If you describe the learning of language in this way you are, I believe, thinking primarily of nouns like ‘table’, ‘chair’, ‘bread’, and of people’s names, and only secondarily of the names of certain actions and properties; and of the remaining kinds of word as something that will take care of itself (Wittgenstein 1968: §1; italics mine—KM)

212     K. Mulligan

5 How Temptations and Pictures Work The temptations to which philosophers succumb and the temptation to philosophise are omnipresent in Wittgenstein’s reflections. Cavell even says that the “voice of temptation and the voice of correctness are the antagonists in Wittgenstein’s dialogues” (Cavell 1968: 183). If Marty is to be believed, bad philosophy often involves succumbing to a temptation: “where a figural inner linguistic form plays the sort of role described, the temptation will be greater to think the meaning in an improper way” (Marty 1950b: 43). A philosophical puzzle may arise because we give into the temptation to use a certain picture but it may also lead us to use a certain picture. As we have already noted, Wittgenstein thinks that the difficulty of a philosophical puzzle may arise when we look at the facts through the medium of a misleading form of expression. And Marty refers to “all the aporiae by which one was tempted to think that an object without a subject is absurd” (Marty 1916: 58–59). Temptations, our two philosophers claim, lie in language. This is what makes the critique of language necessary. Of the already mentioned “temptation (Verführung )” to assimilate in thought modality to what is real, Marty says that “it lies certainly not least again in language” (Marty 1976: 354). Similarly, It can…happen that language, thanks to the fictions of pictorial, inner form, seems to display something as what ought be recognized whereas in fact something else should really be recognised (Marty 1976: 356)

Wittgenstein, as always, is more emphatic and more eloquent: The problems which arise through a misinterpretation (Missdeuten ) of our forms of language…. [have] roots [which] are as deep in us as the forms of our language… (Wittgenstein 1968: §111). A simile (Gleichnis ), that has been absorbed into the forms of our language, produces a false appearance (Schein ),… (Wittgenstein 1968: §112)

9  Misleading Pictures, Temptations and Meta-Philosophies …     213

Of “a picture which has held us captive” he says “we could not get outside it, for it lay in our language and language seems to repeat it to us inexorably” (Wittgenstein 1968: §115). Grammatical problems are “connected with the oldest habits of thought, i.e. with the oldest pictures, which are engraved in our language itself….” (Wittgenstein 2005: 423). Perhaps the most striking formulation of this idea is the claim that there are pictures which force themselves on us: ‘The picture forces itself on us…’ (Es drängt sich uns das Bild auf…). It is very interesting that pictures do force themselves on us…. (Wittgenstein 1984b: I §14) In this case, we already have a picture, which forces itself on us at every turn,… (Wittgenstein 1968: §425, cf. §607, §178)

Marty, too, employs this locution and, like Wittgenstein, identifies certain pictures as the result of habits: But it is certain that the presentation of thing and property always and everywhere forces itself on us (aufdrängt ), if not as a consequence of an innate necessity, then thanks to the power of a strong and general linguistic habit (Marty 1976: 355)

The locution, as Marty notes, is employed by Kant, who, in a famous thought-experiment, refers to “the necessity with which this concept [the concept of substance] forces itself upon you” (Kant, Critique of Pure Reason, B5, and elsewhere). What, then, is it for something to force itself upon one in a non-physical way? The question is raised and answered by Scheler, who distinguishes between a type of active attention and passive attention and identifies the latter with the phenomenon of something forcing itself on one: Insofar as attention appears “in” striving, active attention is given most clearly in the phenomenon of “searching for” and passive attention in the phenomenon of undergoing a “forcing itself on one” (Sichaufdrängens ).

214     K. Mulligan

The latter breaks down into the qualities of “being attracted” and “being repelled”…“Passive attention”, the forcing themselves on one of objects, with their qualities of attraction and repellence,…presupposes an objective factor in the object, its “conspicuousness” (Auffälligkeit ). (Scheler 1966: 159)

Whatever one may think of the merits of Scheler’s identification of passive attention and something imposing itself and his distinction between two forms of the latter, his distinction is nicely illustrated by Wittgenstein’s descriptions of the ways in which the philosopher is attracted—pulled and pushed—by some pictures and put off by others, which he resists or rebels against. In a description of our “wavering between logical and physical impossibility”, Wittgenstein distinguishes the case where a use of words “for some reason strongly recommends itself to us” and the phenomenon of being “alternately tempted to use a word in several different ways”: The man who says “only my pain is real”, doesn’t mean to say that he has found out by the common criteria - the criteria, i.e., which give our words their common meanings - that the others who said they had pains were cheating. But what he rebels against is the use of this expression in connection with these criteria. (Wittgenstein 1972: 57)

And at one point, he says of himself: “We merely rejected the grammar which wants to force itself on us here” (Wittgenstein 1968: §304). Temptation is an affective and conative phenomenon. In his most detailed account of the affective roots of the role of a misleading picture in philosophy, Marty refers to an instinctive urge which, he thinks, plays a role in visual experience and, he goes on to suggest, a very similar role in the types of make-believe peculiar to the use of pictures in philosophy: Here we touch on a point which is what above all has led to the promulgation (Statuierung ) of an “immanent object”. One took the so called relation to an object (Objektsbeziehung ) for a relation in the strict sense of the word, i.e. for a correlation which presupposes the real existence of all its terms. But since the real object is often missing one helped oneself

9  Misleading Pictures, Temptations and Meta-Philosophies …     215

to a feigning (fingieren ) of an “immanent” or mental object. This fiction was encouraged by something else. An instinctive urge (Drang ) leads us, at first, to take whatever appears in a sensory fashion to be real, i.e. to ascribe “external” reality to it; to the colours, sounds, places and changes of place which are present for us in sensation or hallucination…But in spite of the…scientific conviction of the non-existence of sensed colours, absolute places the blind belief in these, based on instinct, remains as vivid as ever…Something similar is the case with hallucinations and the hallucinatory elements which the so called activity of phantasy brings with it. Contradictions with other sorts of experience prohibit us from taking them to be externally real. We therefore transfer (verlegen ) what in them is intuitive and forces itself upon us as real, in an instinctive belief, into the mind. (Marty 1976: 396, cf. 372)

Wittgenstein’s portrait of the affective and conative dimensions of philosophical temptations (the starting point of which may well have been some remarks by the German physicist, Heinrich Hertz) is much richer than that given by Marty. Indeed, he often seems to be describing a Viennese melodrama. There is “a drive (Trieb ) to misunderstand [the] working of our language” (Wittgenstein 1968: §109). Misleading pictures tempt many into repeating the same questions and this satisfies “a longing for the transcendental” for in believing that they see the ‘limit of human understanding’ they of course believe that they can see beyond it (Wittgenstein 2005: 424). Philosophical puzzlement is associated with difficulty, discomfort (Wittgenstein 1972: 26), discontentment (Wittgenstein 1972: 56) and deep disquietude (Wittgenstein 1968: §111). The philosopher or metaphysician who craves generality is contemptuous towards the particular case (Wittgenstein 1972: 18). There is a fascination which the analogy between two similar structures in our language can exert on us (Wittgenstein 1972: 26). Once a false analogy is accepted into language, the result is constant battle and constant uneasiness (Wittgenstein 2005: 409). Grammatical illusions lead to superstition (Aberglaube ), which is not to be identified with error (just as, according to Marty, such illusions take on the character of instinctive or blind beliefs) but have a pathos all their own which redounds to them (Wittgenstein 1968: §110). Deeply entangled as we are in grammatical confusions, only for those “who live in a state of instinctive dissatisfaction

216     K. Mulligan

with language”, for whom, as we might say, Sprachkritik is almost as natural as breathing, is extrication possible; it is not possible for “those who, according to all their instincts, live in the very herd that has created this language as its proper expression” (Wittgenstein 2005: 423). This aristocratic dissatisfaction with language and the rather different dissatisfaction and anger with language of Hofmannsthal’s Lord Chandos are two of the motors of Austrian Sprachkritik. Wittgenstein occasionally indicates that the therapy he offers those who are tempted to philosophise by succumbing to misleading pictures shares some features with Viennese psychoanalysis: One of the most important tasks is to express all false lines of thought in a way so true to character that the reader says: “Yes, that is exactly the way I meant it”. To trace the physiognomy of every error (Wittgenstein 2005: 410) Indeed, we can only prove that someone has made a mistake if he (really ) acknowledges this expression as the correct expression of his feeling. (Wittgenstein 2005: 410) For only if he acknowledges it as such, is it the correct expression. (Psychoanalysis) (Wittgenstein 2005: 410)

Properly understood, philosophy involves overcoming a resistance of the will and it leads to resignation, of feeling not of understanding (Wittgenstein 2005: 406).4 One of Nestroy’s characters answers a question asked by one of Shakespeare’s characters, “What is my nation?”, with the word: resignation. Once cured of philosophy, thanks to Wittgenstein’s patient identification of the pictures we have taken too seriously, we, too, should experience resignation. Wittgenstein’s anatomy of the affective and conative life of the (would-be) philosopher and of her therapy has almost no counterparts in the writings of Brentano and Marty. But the three thinkers all agree that, as Marty’s editor puts it: the task of description has above all to lie in eliminating the errors which have their origin in the imperfections of the vernacular and the pictures of “inner linguistic form” and the philosophy which takes its first orientation from this. (Kraus 1973: xxxv; emphasis mine—KM)

9  Misleading Pictures, Temptations and Meta-Philosophies …     217

As is by now well known, Wittgenstein’s descriptivism shares many features of the descriptivisms of Brentano and his heirs. But it is in Marty’s writings that we find for the first time in descriptivist philosophy the insistence that description’s first task is to identify the misleading pictures which vitiate any philosophy built on them, which takes its first orientation from them. As Wittgenstein was to put it, in philosophy, the first step, the one which commits us to a particular picture or analogy, is the most important one (Wittgenstein 1968: §308). How do pictures, misleading or not, work? Marty and Wittgenstein have surprisingly little to say by way of an answer, as is already apparent from their oscillation between “analogy”, “metaphor”, “simile”, “fiction”, “picture”, “illusion”, “appearance”, “forms of representation”, “form of expression”, “way of looking at a matter” and “symbol”. Marty writes: If I speak of…“a cool deliberation”, then – if the picture has not faded and become ineffective – there are two very different presentations in the consciousness of the speaker and the person who understands him. On the one hand, the presentations which constitute the real meaning (in the mentioned case, presentations of psychological processes and states) and others which have either the goal of serving as an associative link for the meaning or of making the entire presentation…aesthetically more appealing. (Marty 1910: 85)

Sometimes a picture is opposed to what is seriously meant: When we speak of….the suddenness of a decision…the presentation of a thing, which the grammatical category of the substantive involves in the case of words such as…suddenness, is not a matter of the seriously meant meaning but is only a picture. (Marty 1910: 87) Only pictorially, only as a fiction, do I have the presentation of a thing when I speak of “movement”, “rest”, “love” and “hate”. What is really meant is not a thing, but a process or change. (Marty 1950a: 63)

For a conceptual presentation to be a picture is for it to play a role:

218     K. Mulligan

The confusion of inner linguistic form and meaning, taking one to be the other, is…possible because in each case one has to do with similar (gleichartige ) phenomena. What on one occasion is a mere “inner form” can on another occasion be or have been the real meaning of an expression. (Marty 1950a: 43)

Marty points out that “the expression figural use” is itself “a picture” (Marty 1910: 85) and says that the “boundary between error, unconscious fictions, and conscious fictions is a fluid boundary (fliessende Grenze )” (Marty 1910: 99; 1965: 161, cf. 197, 211). Wittgenstein says the same of the boundary between innocent and misleading analogies: When we say that by our method we try to counteract the misleading effect of certain analogies, it is important that you should understand that the idea of an analogy being misleading is nothing sharply defined. No sharp boundary can be drawn round the cases in which we should say that a man was misled by an analogy. The use of expressions constructed on analogical patterns stresses analogies between cases often far apart. And by doing this these expressions may be extremely useful. It is, in most cases, impossible to show an exact point where an analogy begins to mislead us. (Wittgenstein 1972: 28)5

6 Meta-Philosophies Let us consider a handful of further examples of what our two philosophers think of as the havoc wrought in philosophy by misleading pictures before turning to the disagreement between them about the nature of philosophy. Marty quotes the locution dear to friends of universals—“what is coloured is coloured through (durch, thanks to) colour,…what is similar is similar through similarity, what is beautiful is beautiful through beauty etc.” and says of it that “it employs a form of expression and understanding which has no ground in reality” since it is “modelled (nachgebildet ) on the genuine parthood relation” expressed, for example, by “the bird is winged thanks to its wings” (Marty 1910: 95). And Wittgenstein famously says:

9  Misleading Pictures, Temptations and Meta-Philosophies …     219

The idea of a general concept being a common property of its particular instances connects up with other primitive, too simple, ideas of the structure of language. It is comparable to the idea that properties are ingredients of the things which have the properties; e.g. that beauty is an ingredient of all beautiful things as alcohol is of beer and wine, and that we therefore could have pure beauty, unadulterated by anything that is beautiful. (Wittgenstein 1972: 17)

“Our language”, he says, has remained constant and keeps tempting (verführen, seducing) us into asking the same questions. So long as (1) there is a verb ‘be’ that seems to function like ‘eat’ and ‘drink’, so long as (2) there are adjectives ‘identical’, ‘true’, ‘false’, ‘possible’, so long as there is talk about (3) a flow of time and (4) an expanse of space, etc., etc. humans will continue to bump up against the same mysterious difficulties, and stare at something that no explanation seems able to remove (Wittgenstein 2005: 424; numbering mine—KM)

Marty completely agrees that expressions of all four classes function and have functioned as misleading pictures in philosophy (cf. Marty 1916: 253–261; 1918: 33–34, 39, 200–202; 1976: 316–317) just as he agrees that a very common form of Platonism is made possible by misleading picture of parthood. But unlike Wittgenstein he thinks that the best therapy is a true, philosophical theory, a system of truths, of which the most important are non-contingent truths. His account of the misleading pictures mentioned above goes hand in hand with a positive, philosophical alternative: his philosophy of intentionality, of states of affairs (judgement-contents), of the exemplification of value (interest-contents), of modality, of sentences and of the relation between elementary meaning categories and natural languages, and his philosophy of mind or descriptive psychology.6 The latter, in particular, contains many claims which, if Wittgenstein is to be believed, could only be made by one who is in the grip of misleading pictures. For Marty, like Brentano and all his heirs, thinks that there are mental events and processes which display distinctive types of structure and his account of these processes is central to his intentionalist, proto-Gricean account

220     K. Mulligan

of communication and meaning. His view belongs to just the family of views Wittgenstein describes as follows: It seems that there are certain definite mental processes bound up with the working of language, processes through which alone language can function. I mean the processes of understanding and meaning. The signs of our language seem dead without these mental processes; and it might seem that the only function of the signs is to induce such processes, and that these are the things we ought really to be interested in. (Wittgenstein 1972: 3)

Wittgenstein’s meta-philosophy is rarely mentioned in c­ontemporary discussions of meta-philosophy, meta-ontology and meta-metaphysics (unlike the meta-philosophy and meta-ontology of one of his followers, Carnap). It is consistently negative or sceptical about philosophy, and in particular, metaphysics conceived of as a positive, systematic, theoretical enterprise. Already in the Tractatus, the reader is told that “the right method of philosophy would be this….when someone…wished to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his propositions” (Wittgenstein 1998: 6.53). Wittgenstein’s condemnation of the striving for systematicity in philosophy is formulated in the language of temptation, and although no misleading picture is identified in the following passage, he is perhaps thinking of the temptation to see philosophy as science: Our craving for generality has another main source: our preoccupation with the method of science. I mean the method of reducing the explanation of natural phenomena to the smallest possible number of primitive natural laws; and, in mathematics, of unifying the treatment of different topics by using a generalization. Philosophers constantly see the method of science before their eyes, and are irresistibly tempted to ask and answer questions in the way science does. This tendency is the real source of metaphysics, and leads the philosopher into complete darkness. I want to say here that it can never be our job to reduce anything to anything, or to explain anything. Philosophy really is ‘purely descriptive’ (Wittgenstein 1972: 18)

9  Misleading Pictures, Temptations and Meta-Philosophies …     221

That description is the method of philosophy is, as we have noted, a conviction common to Brentano and nearly all his heirs, as is the distinction between description, on the one hand, and genetic or causal explanation and hypotheses, on the other hand (cf. Mulligan 1993, 2012: 21–48). But Brentano and his heirs did not think that description in philosophy excluded systematicity. Marty, as we have seen, thought that the first task of description is to identify and eliminate misleading pictures. Wittgenstein, on the other hand, seems often to have thought of this as his first and last task: If I rectify a philosophical mistake and say that this is the way it has always been conceived, but this is not the way it is, I must always point out an analogy according to which one had been thinking, but which one did not recognize as an analogy (Wittgenstein 2005: 302; emphasis mine—KM)

We encounter philosophical problems, he says, “only when we are guided by…certain analogies within language” rather than by practical purposes (Wittgenstein 2005: 314; emphasis mine—KM). An evaluation of the views of Wittgenstein about misleading pictures and philosophy, of his meta-philosophy, might proceed in different ways. Let me mention briefly and illustrate three such approaches. The first approach is suggested by the role of the category of fictions in the writings of our two Austrian philosophers. Suppose you are convinced, by Marty or by Wittgenstein, that a piece of philosophy employs fictions which are not recognised as such. You might then think that candid recognition of fictions and of their philosophical use is in order. The contemporary meta-philosophy called fictionalism proceeds in this way.7 An example might be a fictionalist version of Bolzano’s abstract propositions. They are, on this view, mere fictions. But they are fictions which enable one to formulate important discoveries about analyticity, logical consequence and the logical theory of probability. These discoveries, however, form a system and for this reason alone (and for many other reasons), could not be countenanced by a friend of Wittgenstein’s meta-philosophy. And it is difficult to imagine what purpose might be served by an anti-systematic fictionalism.

222     K. Mulligan

A second approach would consider, one after another, the different candidates for the role of misleading pictures in philosophy put forward by Wittgenstein and Marty and ask: Must a philosopher fall into the traps identified by our two philosophers? In the case of Wittgenstein’s accounts of these candidates, the task of giving an answer to our question is complicated by the therapeutic and so not wholly theoretical context to which Wittgenstein’s accounts belong. Let us nevertheless, and by way of illustration, consider the first of the misleading pictures described above. Can the ontologist or metaphysician think of modality without falling into the traps identified by our philosophers (Sect. 2)? She may well agree that possibilities are not individual accidents (tropes) or individual states because possibility is a formal property and accidents are material properties or because possibility is the semantic value of a connective unlike accidents or because states and accidents are temporal, real entities, unlike possibility. A philosopher who clearly thinks of de dicto possibility as de dicto is not obviously thinking of it as a state of anything at all. Similarly, a philosopher who thinks of de re possibility as de re, in particular, one who takes “possibly” to modify a predicate, is not thereby assimilating possibility to any sort of first-order property. Indeed, Wittgenstein himself at one point calls belief a state of the soul which endures and a disposition (Wittgenstein 1968: II 191–192). A state of this kind is presumably not to be assimilated to the sort of state exemplified by a toothache. Whatever we should think of Wittgenstein’s assurance that whenever the figure he calls a metaphysician makes a modal, philosophical claim she is being misled by a picture and about the relation between this assurance and philosophical arguments against modality, a third way of evaluating Wittgenstein’s meta-philosophy is to consider alternatives to what has been called Wittgenstein’s endorsement of the autonomy of grammar. On the basis of remarks to the effect that the use of language is “in a certain sense autonomous” because the rules of grammar differ from the rules of activities defined by their ends, such as cookery (Wittgenstein 1973: 55; 1988: §320), Wittgenstein is often taken to endorse the autonomy of rules for the use of language. Baker and Hacker formulate two aspects of this view:

9  Misleading Pictures, Temptations and Meta-Philosophies …     223

The questions of the source of necessity and how we recognize it leads us astray before we have begun. The prior question is: what is it for a proposition to be a ‘necessary proposition’? If this is answered by examining the roles of such propositions in our linguistic transactions and describing them properly the traditional questions shrink to vanishing point. (Baker and Hacker 1988: 263–264) For the rôle of necessary propositions is not descriptive at all…Merely to insist that the propositions of metaphysics describe ‘the necessary features of the world’…is to do no more than advance a picture instead of an elucidation. (Baker and Hacker 2009: 248)

What would it be to deny the autonomy of grammar? What forms might the heteronomy of grammar take? As far as I can see, the only worked out philosophy of the heteronomy of grammar is suggested by a view set out in the 1900 Prolegomena to Husserl’s Logical Investigations (a philosophy which Frege seems to endorse in passing). In this philosophy, as in Wittgenstein’s writings, early and late, the category of essence looms large. But whereas Husserl takes the modal truths forming systems of necessarily true philosophical propositions to be grounded in essences, Wittgenstein tells us that essence is specified by grammar and that all truth-bearers are contingently true or false. According to Husserl, logical laws, considered in and of themselves, are not at all normative propositions,…in the sense of prescriptions, i.e. propositions to whose content it belongs that they state how one should judge. One must indeed distinguish: laws that serve as norms for our epistemic activities and laws which contain the thought of normative regulation and state this to be generally binding. (cf. Husserl 1975: 129, 159)

No logical law, Husserl says, “contains the faintest thought of normativity”. We can employ such a law “for normative purposes, but it is not therefore a norm. We can also ground an explicit prescription on it, e.g. “Anyone who judges that every A is also B, and that a certain S is A, must (ought) also to judge

224     K. Mulligan

that this S is B ”. Everyone sees, however, that this is not the original logical proposition but one which has grown out of it by bringing in the thought of normativity. (Husserl 1975: 160) The same obviously holds of all […] “purely logical” propositions. But not of such laws alone. The ability to be normatively transformed belongs also to the truths of other theoretical disciplines… (Husserl 1975: 161; emphasis mine—KM)

Thus, for example, deontic laws about how we ought to think about colours follow from the non-deontic laws of “colour geometry”, and deontic laws about how we ought to think about emotions or belief follow from the non-deontic laws of descriptive psychology. The deontic rules grounded in necessary truths and connexions, logical and nonlogical, are, Husserl thinks, “noetic norms”, which are neither hypothetical nor instrumental. Of course, they are not rules for using words, the rules which make up grammar according to Wittgenstein and which, according to Husserl, merely express noetic norms which are in turn grounded in theoretical truths. Some of Wittgenstein’s formulations suggest that his account of the autonomy of grammar may be seen as the result of turning a view like Husserl’s on its head: Here I am on the verge of inventing a mythology of symbolism. It looks as if one could infer from the meaning of negation that “~ ~ p” means p. As if the rules for the negation sign follow from the nature of negation. So that in a certain sense there is first of all negation, and then the rules of grammar (Wittgenstein 1973: II §15)

That the rules for using the negation sign and for thinking with the concept of negation follow from the nature of negation is Husserl’s view. The view that the rules follow from the meaning of negation could mean two different things. If “meaning” here means the semantic value of the negation sign, the formulation would resemble Frege’s view: “the rules follow necessarily from the meanings (Bedeutungen ) of the signs” (Frege 1998: II §158, cf. §90). But if, as is more likely, “meaning” here means something like sense, the claim is a Husserlian one.

9  Misleading Pictures, Temptations and Meta-Philosophies …     225

Although Wittgenstein on the autonomy of grammar and Husserl on its heteronomy may be seen as diametrically opposed philosophical theses, Wittgenstein does not, of course, take himself to be defending any positive, philosophical claim about modality or rules or their relation. Rather, he thinks he knows that anyone who puts forward a modal, metaphysical claim is in the grip of a picture and that he knows how to disabuse him. Marty, too, in his own modest, Swiss fashion, thinks, as we have seen, that philosophies of necessary truths like Husserl’s rely on inappropriate pictures.8

7 From Reid and Bentham to Brentano and Beyond Brentano and Marty are by no means the only philosophers to have explored the roles of misleading pictures in philosophy although Marty has a strong claim to have provided the first systematic account of the phenomenon. Mach, Nietzsche’s discussions of fictions and Bergson on spatial symbols in thought about time all belong here.9 So too, do Vaihinger and the already mentioned Mauthner.10 Marty’s account of the roots of the philosophy of immanent objects in a misleading picture was anticipated by Reid (as Spinicci 1991: 152 notes). Husserl, who had noted and rejected the figural nature of the philosophy of immanent objects before Marty, refers in this connexion to Reid (Husserl 1984b: 889). And one of the many things Brentano finds to praise in Reid is his recognition that “if one says that what one thinks is in the mind of the thinker, then one is employing a very improper locution. What is in the mind in this way is, properly speaking, not”, although he disagrees with some of Reid’s attributions of the doctrine he criticises (Brentano 1975: 2).11 Perhaps the most striking anticipation of Marty and Wittgenstein on misleading pictures is Bentham’s account of fictions.12 Peter Hacker has pointed out some of the similarities between what Wittgenstein says about misleading pictures and Bentham’s account of fictions, the “material”, “emblematic” images of real things associated with names

226     K. Mulligan

“of ” ficta (Baker and Hacker 2005: 279). Marty’s editor, Oskar Kraus, noted that the Brentano-Marty campaign against misleading pictures has an important predecessor in Bentham’s account of fictions. In 1937, Kraus mentions that a friend in England had sent him Ogden’s (1932) book Bentham’s Theory of Fictions and refers to its “very valuable Introduction” (Kraus 1937: 111).13 Referring to Brentano’s late reism, the view that there are only things and thus no properties, relations, states of affairs, events or processes, and its corollary, that apparent reference to such non-things is due to the role of misleading pictures, Kraus writes: The real, immediate precursor of reism is Bentham, who has visibly influenced J. St. Mill’s theory of language (cf. Logic, Book I, Ch. 2, §3f. and Book V, Ch. 3, §4). Brentano in turn presumably knew Mill’s Logic well and even if Bentham’s theory of fictions remained unknown to him, its sprachkritische influence had its effect via Mill. But there are many differences between the Sprachkritik of Bentham and Brentano which are not minor. The most general term in Bentham’s doctrine is “entity”, an expression which is supposed to embrace both real things and fictions. The first difference is hence that when Bentham defines these entities he reflects on their existence and non-existence, whereas according to Brentano, the concept of what is thing-like (essential, real) contains no reference to existence, non-existence, not even to possibility or impossibility – there is no consciousness that is not consciousness of something thinglike, there is no concept of entity under which that which is thinglike and that which is not falls. (Kraus 1937: 272)14

Brentano was in fact quite familiar with Mill’s Logic and Mill does indeed there condemn the “natural prejudice” whose “most remarkable manifestation” is “the personification of abstractions” (Mill 1892: Book V, Ch. 3, §4, 450). Kraus, who follows Brentano not Marty, also notes a strange inconsistency in the writings of the logical positivists and the “Logistiker ”, that is to say, the adherents of the new logic of Frege-WhiteheadRussell. They reject out of hand the naive realism about values espoused by some of Brentano’s heirs but happily endorse the naive realism about entities such as the states of affairs and objectives of these same heirs, in

9  Misleading Pictures, Temptations and Meta-Philosophies …     227

particular, Husserl and Meinong. Were they consistent, they would see that the same sprachkritische considerations which lead to the rejection of naive realism about value lead, too, to rejection of states of affairs: The theory of “states of affairs” (Sachverhalte ) has penetrated the most recent psychology and linguistics…even the Logistiker have usurped the expression, and Russell makes use of Meinong’s objective in the same carefree way as Carnap and other neo-positivists. These “Logistiker ” do not see how little inconsistent and logical is their approach when, on the one hand, they denounce as senseless metaphysics the realms of value of modern Platonists yet themselves operate with objectives and states of affairs as though these fictions were more tolerable than the former. (Kraus 1937: 27)

Kraus might have added that Wittgenstein, too, had dismissed value but embraced states of affairs, obtaining states of affairs, situations and facts. But his references to Wittgenstein insist rather on the latter’s indifference to his sources and failure to recognise his precursors in the struggle against language, Brentano and Marty.15 Work on this paper was made possible by the 2014–2017 Swiss FNS project Signification et intentionnalité chez Anton Marty. Aux confins de la philosophie du langage et de l’esprit. Thanks to Ingvar Johansson for his comments and to an anonymous referee.

Notes 1. Marty’s 1908 book builds on a series of earlier publications. Two excellent accounts of Marty’s philosophy of language are Spinicci (1991) and Egidi (1992), see also Mulligan (1990a). 2. This is the first of the two criticisms of the very idea of private objects, in particular private psychological objects, in the Brentanian tradition. The second is due to Max Scheler, cf. Mulligan (2012: 55–59). 3. Cf. “The processes of sentence formation are so various that it is impossible to find a satisfying genetic definition of the Satz ” (Bühler 1918: 14). 4. On the role of feeling in “genuine resignation”, cf. Scheler (1966: 58).

228     K. Mulligan

5. Wittgenstein (1972: 28). What exactly did Wittgenstein take the pictures that mislead philosophers to be? For a discussion of some answers, cf. Egan (2011). 6. Marty thinks that his account of states of affairs, unlike those given by Husserl and Meinong, relates these firmly to the sphere of what is real and temporal, for they are temporal but non-real. On his philosophy of intentionality, cf. Cesalli and Taieb (2012); on his philosophy of states of affairs, cf. Morscher (1990). 7. On fictionalism and other meta-ontological views, cf. Berto and Plebani (2015). 8. The diagnoses of misleading pictures by Marty mentioned so far are to be found in his 1908 Investigations or in later writings. But, for example, already in 1895 Marty has this to say: “one who thinks that in sentences such as ‘all the angles in a semi-circle are right angles’, ‘no tone is a colour’ etc. something is seriously stated about angles or tones… is as much deceived as one who seriously takes every so called verb as the designation of a doing or a suffering or…‘have’ for the expression of owning. In all these cases, a change of function has taken place and the inner form…should not be confused with the meaning”. The inner form here is “the presentation of the being predicated of a determination of something…already accepted”. The latter form is very common but in the cases quoted this form functions “symbolically or pictorially ” (Marty 1918: 267–268; emphases mine—KM). 9. Nietzsche’s views about fictions were criticised by a philosopher influenced by Brentano, Del Negro (1923: 13, 140, 158–161, 190). Marty criticises Nietzsche’s failure to understand inner, linguistic form (Marty 1920a: 128). 10. On Mauthner, cf. Weiler (1970), who also discusses Marty’s views. Vaihinger’s Die Philosophie des Als Ob appeared in 1911. In 1897, he had published a brief review of Marty’s account of impersonal sentence and the relation of grammar to logic and psychology. The review sketches Marty’s complete rejection of Kant’s account of categorical judgements on the basis of Brentano’s theory of judgement. In the fourth edition of Die Philosophie des Als Ob Vaihinger refers to Marty’s account of the non-real as a confirmation of his own views, as Raynaud (1982: 279) points out; cf. Vaihinger (1920: xvi). Marty mentions Leibniz’s rejection of abstract entities as an anticipation of his and Brentano’s philosophy of misleading pictures (Marty 1910: 95–100), cf. Brentano (1977: 249–250) on Leibniz.

9  Misleading Pictures, Temptations and Meta-Philosophies …     229

11. In the light of Mill’s role in the history of appeals to misleading pictures, described below, it is of interest that Brentano also notes some anticipations by Reid of Mill’s views. 12. On this account, cf. Bouveresse (1993). 13. Ogden’s book appeared a year after another book by a Cambridge philosopher, Wisdom (1931), on Bentham. 14. Kraus’ (1901) study of Bentham’s theory of value from the point of view of Brentano’s theory does not discuss his account of fictions. Brentano formulates many of the aspects of the philosophy of misleading pictures mentioned above in the second volume of his Psychology (Brentano 1971: 275–277). 15. Cf. Mulligan (2017), which also presents the meta-ethical and metaaxiological positions of Brentano and his heirs, on the one hand, and those of Wittgenstein, Russell and the early emotivists, on the other hand, in particular Scheler’s 1913–1916 analysis of what he calls ethical nominalism and later became famous as “emotivism”. Scheler, who died in 1928, introduced the terms “meta-ethics” (Meta-Ethik ) and “meta-axiology” (Meta-Axiologie ) in late manuscripts, published in 1979; he does not use the term “meta-metaphysics” but refers to “the theory of the essence…of metaphysics” (Scheler 1979: 14, 54). Scheler, like Marty, thinks that many philosophical views he disapproves of are due to misleading pictures. On empty time as a fictitious picture which leads to the idea of the flow of time, cf. Scheler (1995: 231); on misleading pictures of the mind and selves, cf. Scheler (1966: 421, 415).

References Baker, G., and Hacker, P. 1988. Wittgenstein. Rules, Grammar and Necessity. Oxford: Blackwell. ———. 2005. Wittgenstein: Understanding and Meaning. Part I: Essays, 2nd ed. Oxford: Blackwell. ———. 2009. Wittgenstein. Rules, Grammar and Necessity, 2nd ed. In Baker and Hacker (1988). Oxford: Blackwell. Berto, F., and M. Plebani. 2015. Ontology and Metaontology: A Contemporary Guide. London: Bloomsbury. Bouveresse, J. 1993. La théorie des fictions chez Bentham. In Mulligan and Roth (1993): 87–98.

230     K. Mulligan

Brentano, F. 1971. Psychologie vom empirischen Standpunkt, II, Von der Klassifikation der psychischen Phänomen. Hamburg: Meiner. ———. 1975. Was an Reid zu loben. Grazer Philosophische Studien 1: 1–18. ———. 1977. Die Abkehr vom Nichtrealen. Hamburg: Meiner. Bühler, K. 1909a. Über das Sprachverständnis vom Standpunkt der Normalpsychologie aus. In Bericht über den dritten Kongress für experimentelle Psychologie, 94–130. Leipzig: Barth. ———. 1909b. (Rezension von) Anton Marty. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie. Niemeyer: Halle. 1908. Göttingische gelehrte Anzeigen 12: 947–979. ———. 1918. Kritische Musterung der neueren Theorien des Satzes. Indogermanisches Jahrbuch 6: 1–20. ———. 1982 [1934]. Sprachtheorie. Die Darstellungsfunktion der Sprache. Jena: Fischer. Cavell, St. 1968. The Availability of Wittgenstein’s Later Philosophy. In Wittgenstein. The Philosophical Investigations, ed. G. Pitcher, 151–185. London: Macmillan. Cesalli, L., and H. Taieb. 2012. The Road to ideelle Verähnlichung. Anton Marty’s Conception of Intentionality in the Light of Its Brentanian Background. Quaestio 12: 25–87. Del Negro, W. 1923. Die Rolle der Fiktionen in der Erkenntnistheorie Friedrich Nietzsches. Munich: Rösl. Egan, D. 2011. Pictures in Wittgenstein’s Later Philosophy. Philosophical Investigations 34: 55–76. Egidi, R. 1992. Anton Marty. Eine Sprachphilosophie in der Nachfolge Brentano. International Bibliography of Austrian Philosophy 1984 (85): 23–104. Amsterdam: Rodopi. Frege, G. 1998 [1903]. Grundgesetze der Arithmetik, 2 vols. Hildesheim, Zürich and New York: Georg Olms Verlag. Husserl, E. 1975. Logische Untersuchungen, vol. I, Prolegomena zur reinen Logik, ed. E. Holenstein, Husserliana XVIII. The Hague: Martinus Nijhoff Publishers. ———. 1984a. Logische Untersuchungen, vol. II, 1, Husserliana XIX/1. The Hague: Martinus Nijhoff Publishers. ———. 1984b. Logische Untersuchungen, vol. II, 2, Husserliana XIX/2. The Hague: Martinus Nijhoff Publishers. Kraus, O. 1901. Zur Theorie des Wertes. Eine Bentham-Studie. Halle: Niemeyer. ———. 1937. Die Werttheorien. Geschichte und Kritik. Brünn: Rudolf M. Rohrer.

9  Misleading Pictures, Temptations and Meta-Philosophies …     231

———. 1973. Einleitung. In Psychologie vom empirischen Standpunkt, vol. I, ed. F. Brentano, xvii–xciii. Hamburg: Meiner. Marty, A. 1910. Zur Sprachphilosophie. Die “logische”, “lokalistische” und andere Kasustheorien. Halle: Niemeyer. ———. 1916. Raum und Zeit. Halle: Niemeyer. ———. 1918. Schriften zur deskriptiven Psychologie und Sprachphilosophie. Gesammelte Schriften, II, 1. Halle: Niemeyer. ———. 1920a. Selbstanzeige. In Schriften zur deskriptiven Psychologie und Sprachphilosophie. Gesammelte Schriften, II, 2: 123–128. Halle: Niemeyer. ———. 1920b. Über Begriff und Methode der allgemeinen Grammatik und Sprachphilosophie. In Schriften zur deskriptiven Psychologie und Sprachphilosophie, Gesammelte Schriften, II, 2: 129–172. Halle: Niemeyer. ———. 1950a [1926]. Über Wert und Methode einer beschreibenden Bedeutungslehre. Berne: Francke, vol. III of the Nachgelassene Schriften, ed. by O. Funke. ———. 1950b [1925]. Satz und Wort. Berne: Francke. ———. 1965 [1940]. Psyche und Sprachstruktur. Berne: Francke. ———. 1976 [1908]. Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie, vol. I (Only Volume Published). Hildesheim and New York: G. Olms. Mill, J. St. 1892. A System of Logic. London: George Routledge & Sons. Morscher, E. 1990. Judgement-Contents. In Mulligan (1990): 181–196. Mulligan, K. (ed.). 1990a. Mind, Meaning and Metaphysics: The Philosophy and Theory of Language of Anton Marty. Dordrecht: Kluwer. ——— 1990b. Marty’s Philosophical Grammar. In Mulligan (1990): 19–28. ———. 1993. Description’s Objects: Austrian Variations. In Themes from Wittgenstein, ed. B. Garrett and K. Mulligan, 62–85. Canberra: RSSS, Australian National University. Working Papers in Philosophy 4. ———. 2012. Wittgenstein et la philosophie austro-allemande. Collection Problèmes et Controverses. Paris: Vrin. ———. 2017. The Origins of Emotivism, Expressivism and the Error Theory: Marty, Scheler, Russell, Ogden & Richards. In Mind and Language. On the Philosophy of Anton Marty, ed. G. Fréchette and H. Taieb, 149–168. Berlin: de Gruyter. Mulligan, K., and Roth, R. (eds.). 1993. Regards sur Bentham et l’utilitarisme. Recherches et Rencontres 4. Geneva: Droz. Ogden, C.K. 1932. Bentham’s Theory of Fictions. New York: Harcourt, Brace and Co.

232     K. Mulligan

Raynaud, S. 1982. Anton Marty, filosofo del linguaggio. Uno strutturalismo presaussuriano. Roma: La Goliardica. Scheler, M. 1966 [1913–1916]. Der Formalismus in der Ethik und die materiale Wertethik. Neuer Versuch der Grundlegung eines ethischen Personalismus. Gesammelte Werke II. Berne: Francke (M. Frings and R.L. Funk. 1973. trans. Formalism in Ethics and Non-Formal Ethics of Values. Evanston: Northwestern University Press). ———. 1979. Schriften aus dem Nachlass. Erkenntnislehre und Metaphysik. Gesammelte Schriften XI. Berne: Francke Verlag. ———. 1995. Späte Schriften. Gesammelte Werke IX. Berne: Francke. Spinicci, P. 1991. Il Significato e la forma linguistica. Milan: Franco Angeli. Vaihinger, H. 1920. Die Philosophie des Als Ob. Hamburg: Meiner. Weiler, G. 1970. Mauthner’s Critique of Language. Cambridge: Cambridge University Press. Wisdom, J. 1931. Interpretation and Analysis in Relation to Bentham’s Theory of Definition. London: Kegan Paul, Trench, Trubner and Co. Wittgenstein, L. 1968. Philosophical Investigations. Oxford: Blackwell (Philosophische Untersuchungen. Kritisch-genetische Edition, ed. J. Schulte, Frankfurt: Suhrkamp. Wissenschaftliche Buchgesellschaft, 2001). ———. 1972. The Blue and the Brown Books. Oxford: Basil Blackwell. ———. 1973. Philosophische Grammatik. Frankfurt: Suhrkamp. ———. 1984a. Bemerkungen über die Philosophie der Psychologie, 2 vols. Frankfurt: Suhrkamp. ———. 1984b. Bemerkungen über die Grundlagen der Mathematik. In Ludwig Wittgenstein Werkausgabe, vol. 6, ed. G.E.M. Anscombe, R. Rhees, and G.H. von Wright. Frankfurt: Suhrkamp. ———. 1988. Zettel, ed. G.E.M. Anscombe and G.H. von Wright. Oxford: Blackwell. ———. 1998. Logisch-philosophische Abhandlung. Tractatus logico-philosophicus, Kritische Edition, ed. B. McGuinness and J. Schulte. Frankfurt: Suhrkamp. ———. 2005. The Big Typescript. TS 213, ed. C.G. Luckhardt and M.A.E. Aue. Oxford: Blackwell.

Index

A

Allott, Nick 55, 56 Almäng, Jan 116, 117 Anscombe, G.E.M. 124, 143, 144, 147 Anselm of Canterbury 202 Aquinas, Thomas 202 Aristotle 4, 134, 183, 202 Augustine 211 B

Bach, Kent 14 Bacigalupo, Giuliano 1, 29, 150, 182, 194 Baker, Gordon 222, 223, 226 Barwich, Anne-Sophie 116 Beck, Ori 135 Beiser, Frederick D. 103 Bennett, Jonathan 34, 37, 38, 55 Bentham, Jeremy 225, 226, 229

Bergmann, Hugo 86 Bergson, Henry 225 Berkeley, George 100, 123 Berto, Francesco 176, 181, 185, 228 Bolzano, Bernard 5, 61, 69, 83–86, 93, 204, 221 Bouveresse, Jacques 229 Brentano, Franz 1–5, 16, 28, 35, 38– 50, 52–55, 60, 63–66, 68, 72, 80, 85–87, 90–92, 100, 102, 103, 123, 124, 126–129, 131, 140, 141, 148–150, 159–161, 165, 170, 175, 183, 185–187, 190, 198, 202, 203, 205, 216, 217, 219, 221, 225–229 Brewer, Bill 135 Broad, C.D. 107, 117, 170 Bruno, Giordano 102 Bühler, Karl 26, 29, 210, 211 Burks, Arthur W. 29

© The Editor(s) (if applicable) and The Author(s) 2019 G. Bacigalupo and H. Leblanc (eds.), Anton Marty and Contemporary Philosophy, History of Analytic Philosophy, https://doi.org/10.1007/978-3-030-05581-3

233

234     Index C

Campbell, John 135 Carnap, Rudolf 220, 227 Carruthers, Peter 55 Cavell, Stanley 212 Cesalli, Laurent 2, 4, 8, 13, 28, 29, 34, 55, 56, 194, 228 Chalmers, David 149 Chisholm, Roderick 140, 148 Claas, Jan 5, 59 Colyvan, Mark 105 Crane, Tim 177, 190, 193, 194 Crimmins, Mark 191 Crombie, Alistair Cameron 103 D

Dainton, Barry 101, 103, 117, 170 Davis, Wayne A. 28, 47, 48, 50 Del Negro, Walter 228 Del Pinal, Guillermo 85 Democritus 103 Descartes, René 123 Dretske, Fred 14, 17, 28 Dummett, Michael 55 E

Egan, David 228 Egidi, Rosaria 117, 139, 141, 147, 148, 227 Einstein, Albert 102 Euler, Leonhard 103 Evans, Gareth 193 F

Fine, Kit 193 Fish, William 135, 140, 149

Frascolla, Pasquale 193 Fréchette, Guillaume 4, 8, 170 Frederick, Robert E. 117 Frege, Gottlob 50, 55, 176, 185, 187, 192, 223, 224, 226 French, Craig 29, 135, 148 Fricker, Elisabeth 52 G

Gassendi, Pierre 103 Green, Mitchell S. 28 Grice, Paul 4, 5, 13, 14, 16, 18–29, 34–38, 41, 44–48, 54, 55 H

Hacker, Peter 222, 223, 225, 226 Harnish, Robert M. 14 Hertz, Heinrich 215 Hobbes, Thomas 211 Husserl, Edmund 1, 2, 90, 94, 106, 123, 130, 131, 136, 137, 148, 159, 170, 197, 203–209, 223–225, 227, 228 Hyslop, Alec 55 I

Ingarden, Roman 1, 2 J

James, William 154 Jenkins, David 55, 56 Johansson, Ingvar 5, 6, 102, 104, 106, 116, 117, 129, 130, 133, 138, 141, 146–150, 227

Index     235 K

Kamp, Hans 89 Kant, Immanuel 100, 102, 107–109, 111, 113, 114, 117, 123–126, 129, 202, 213, 228 Kerry, Benno 86 Kraus, Oskar 198, 216, 226, 227, 229 Kriegel, Uriah 185, 187, 188, 190, 192, 193 Kripke, Saul 69, 74, 177 Kroon, Frederick 191 Künne, Wolfgang 93 L

Leblanc, Hélène 1, 29, 148 Leibniz, Gottfried Wilhelm 100, 101, 228 Leucippus 103 Lewis, David 29, 55, 176, 193 Liedtke, Frank 4, 13, 22, 55 Linsky, Bernard 69 Logue, Heather 135 Longworth, Guy 4, 13, 19, 55 Lorentz, Hendrik Antoon 102 M

Mac Cumhaill, Clare 6, 116, 117, 121, 148, 150 Mach, Ernst 101, 113, 225 Majolino, Claudio 56 Marek, Johann 193 Martinak, Eduard 197, 198 Martin, C.B. 126, 135, 148, 149 Martin, M.G.F. 135 Marty, Anton 1–8, 13–19, 21–26, 28, 29, 33–35, 39, 41–43, 54,

55, 60, 61, 65–67, 70–72, 79, 82, 83, 86, 87, 90–93, 100–107, 109, 113, 116, 117, 121–134, 136–150, 153, 154, 159, 162–170, 175, 176, 183–190, 192, 194, 197–210, 212–219, 221, 222, 225–229 Massin, Olivier 116, 148 Mauthner, Fritz 198, 225, 228 McDaniel, Kris 192 McGinn, Colin 192 Meinong, Alexius 1, 2, 69, 91, 100, 116, 176, 179–181, 185, 187, 193, 204, 227, 228 Michaelson, Eliot 56 Mill, John Stuart 199, 226, 229 Moore, George Edward 136, 142, 143 Moran, Richard 36 Morscher, Edgar 86, 228 Mulligan, Kevin 2, 7, 8, 29, 93, 116, 139, 148, 149, 170, 186, 187, 198, 210, 221, 227, 229 N

Neale, Stephen 28, 55 Nerlich, Graham 6, 100, 103, 104, 109–113, 116, 117 Nestroy 216 Newton, Isaac 101, 102, 107, 112, 125 Nietzsche, Friedrich 225, 228 Nordin, Ingemar 116 O

Ockham, Wilhelm 202 Ogden, Charles Kay 226, 229

236     Index P

Parsons, Terence 2 Partee, Barbara 89 Pautz, Adam 148 Peirce, Charles Sanders 29 Phillips, Ian 170 Plato 117, 202 Plebani, Matteo 228 Predelli, Stefano 191 Price, Henry Habberley 133, 149 Priest, Graham 2, 176, 180, 181, 185, 193 Q

Quine, Willard van Orman 177, 188, 189, 192

Schnieder, Benjamin 5, 55, 56, 69, 74 Schuhmann, Karl 148, 149 Searle, John 14 Shakespeare 216 Simons, Peter 3, 116, 123, 148, 170 Smith, Barry 102, 117, 126–129, 134, 139–141, 147, 148, 183–185 Soteriou, Matthew 136, 137 Sperber, Dan 28 Spinicci, Paolo 225, 227 Stampe, Dennis 53 Strawson, Peter Frederick 15, 26, 146 Stumpf, Carl 125, 131, 148 Sundberg, Kristoffer 116 Szabo, Zoltán 193

R

Raynaud, Savina 228 Recanati, François 4, 5, 20, 25, 27–29, 55, 56, 93 Reid, Thomas 225, 229 Rollinger, Robin 28 Routley, Richard 2, 176, 177, 180, 185, 193 Russell, Bertrand 176, 185, 187, 192, 209, 226, 227, 229 Ryle, Gilbert 51, 55

T

Taieb, Hamid 4, 8, 228 Textor, Mark 5, 33, 148 Thomas Aquinas 202 Thomasson, Amie 2, 69 Trendelenburg, Friedrich Adolf 102, 103 Twardowski, Kasimir 1, 28, 60, 61, 65, 67–72, 77–81, 86, 90–93 V

S

Sainsbury, Mark 190 Sattig, Thomas 6, 7, 56, 153 Savonarola, Girolamo 86 Scheler, Max 213, 214, 227, 229 Schiffer, Stephen 28, 29, 55, 192

Vaihinger, Hans 225, 228 van Cleve, James 117 van der Schaar, Maria 28, 77–79, 94 van Inwagen, Peter 69, 192 Vlach, Frank 49, 50, 54 Voltolini, Alberto 7, 56, 176, 192, 194

Index     237

von Helmholtz, Hermann Ludwig Ferdinand 112 von Hofmannsthal, Hugo 198, 216 Von Kries, Johannes 210 von Lotze, Hermann vii, 100, 127 von Solodkoff, Tatjana 69

Whitehead, Alfred North 226 William of Ockham 202 Williamson, Timothy 69, 176, 181 Wilson, Deirdre 28 Wisdom, John 229 Wittgenstein, Ludwig 4, 7, 8, 113– 115, 193, 198–206, 208–225, 227, 229

W

Walton, Kendall L. 191 Weiler, Gershon 228 Wesche, Birgit 122 Wharton, Tim 15, 16, 28

Z

Zalta, Ed 2, 69