Exploring Language Through Contrast 1443840904, 9781443840903

Table of contents :
TABLE OF CONTENTS
PREFACE
PART ONE
FRIEND OR FOE? ON SOME HORIZONTAL BONDS BETWEEN PHONOLOGICAL PRIMES
THE MANY FACES OF OLD ENGLISH I-UMLAUT
PART TWO
GRINDING
A COMPARISON OF APPROACHES TO DIMINUTIVE MEANING AND DIMINUTIVE FORMATION PROCESSES
THE DISTRIBUTION OF THE SUFFIX-NESS ACROSS REGISTERS
THE MORPHO-SEMANTICS OF EXOCENTRIC SYNTHETIC COMPOUND NOUNS IN POLISH
PART THREE
SECONDARY PATTERNS OF USE WITH LOCATUM VERBS
THE ENGLISH NPN FORMS
THE BENEFITS OF SWEARING
PART FOUR
THE POTENTIAL OF THE CONCEPTUAL INTEGRATION THEORY IN THE DESCRIPTION AND EXPLANATION OF SLIPS OF THE TONGUE PHENOMENA
THE RELEVANCE OF THE JUXTAPOSITION OF MEANINGS IN PUNS
CONCEPTUAL METAPHOR THEORY AS AN ISOLATING MODEL
JUXTAPOSITION AND DIFFERENTIATION WITHIN THE METAPHORICAL SENSES IN OE FYR AND LEOHT CONCEPTS
PART FIVE
TRACING TABLOIDIZATION
PATIENTS, DOCTORS, DISEASES AND CASES
DISCURSIVE STRATEGIES IN THE MEDIA CONSTRUCTION OF POLAND, RUSSIA AND THE USA IN THE CONTEXT OF THE DEBATE ON THE US ANTI-BALLISTIC MISSILE DEFENCE SHIELD IN POLISH AND RUSSIAN QUALITY PAPERS
PRAGMATIC ASPECTS OF THE SELF AS A COMMUNICATOR
DISCOURSE GOALS AND NORMS
PART SIX
STEPHEN FRY AND DAVID CRYSTAL
CONTRIBUTORS

Citation preview

Exploring Language through Contrast

Exploring Language through Contrast

Edited by

Waldemar Skrzypczak, Tomasz Fojt and Sáawomir Wacewicz

Exploring Language through Contrast, Edited by Waldemar Skrzypczak, Tomasz Fojt and Sáawomir Wacewicz This book first published 2012 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2012 by Waldemar Skrzypczak, Tomasz Fojt and Sáawomir Wacewicz and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-4090-4, ISBN (13): 978-1-4438-4090-3

TABLE OF CONTENTS

Preface Waldemar Skrzypczak, Tomasz Fojt, Sáawomir Wacewicz ..................... viii Part One: Phonology Friend or Foe? On Horizontal Bonds between Phonological Primes Anna Bloch-Rozmej .................................................................................... 2 The Many Faces of Old English i-umlaut Jerzy Wójcik.............................................................................................. 19 Part Two: Morphology Grinding: A Case of Lexical Derivation Maria Bloch-Trojnar.................................................................................. 38 A Comparison of Approaches to Diminutive Meaning and Diminutive Formation Processes Dorota Gorzycka ....................................................................................... 53 The Distribution of the Suffix –ness Across Registers Wojciech Guz ............................................................................................ 69 The Morpho-Semantics of Exocentric Synthetic Compound Nouns in Polish Joanna Kolbusz-Buda ................................................................................ 82 Part Three: Syntax Secondary Patterns of Use with Locatum Verbs Tomasz Fojt............................................................................................. 104 The English NPN Forms: Words or Constructions? Wiktor Pskit............................................................................................. 126

vi

Table of Contents

The Benefits of Swearing: Towards an Old English Typology of Old English Verbs Magdalena CharzyĔska-Wójcik .............................................................. 145 Part Four: Semantics The Potential of the Conceptual Integration Theory in the Description and Explanation of Slips of the Tongue Phenomena Izabela SekĞciĔska................................................................................... 168 The Relevance of the Juxtaposition of Meaning in Puns Agnieszka Solska..................................................................................... 184 Conceptual Metaphor Theory as an Isolating Model: a Case for Re-Contextualization Ariadna Strugielska ................................................................................. 201 Juxtaposition and Differentiation within the Metaphorical Senses in OE FYR and LEOHT Concepts Agnieszka Wawrzyniak........................................................................... 219 Part Five: Pragmatics and Discourse Tracing Tabloidization: A Comparative Analysis of Rhetorical Styles in Quality and Tabloid Newspapers Katarzyna Molek-Kozakowska ............................................................... 236 Patients, Doctors, Diseases and Cases: Patient-Centered and Objectifying Discourses in Medical Case Reports Magdalena Murawska ............................................................................. 251 Discursive Strategies in the Media Construction of Poland, Russia and the USA in the Context of the Debate on the US Anti-Ballistic Missile Defense Shield in Polish and Russian Quality Papers Agnieszka SowiĔska and Tatiana Dubrovskaya ...................................... 268 Pragmatic Aspects of the Self as a Communicator Jolanta SzymaĔska................................................................................... 288 Discourse Goals and Norms: An Axiological Perspective Przemysáaw ĩywiczyĔski ........................................................................ 299

Exploring Language through Contrast

vii

Part Six: Normative Linguistics Stephen Fry and David Crystal: A Thespian and A Linguist Writing on Modern-Day Prescriptivism, Language Complaint and Language Change in English Maciej Rataj ............................................................................................ 322 Contributors............................................................................................. 331

PREFACE

The present volume aspires to portray a panorama of recent linguistic research in Poland. The principle that shaped its contents draws upon the concepts of comparison and juxtaposition as driving forces in an attempt to grasp descriptive and explanatory aspects of linguistic use and organization. The spectrum of contributions is broad and spans all the levels of language, from phonology to discourse. The constellation of methodological perspectives is equally varied, as the book includes papers that lean upon some versions of the generative theory, as well the recent developments in cognitive linguistics, synchronic and diachronic perspectives, and a measure of freedom has also been allocated to papers written in a more traditional structurally-oriented and/or eclectic spirit. However, the ideas of comparison and contrast have been the common denominator in the service of explicating the central and the focal from what is fuzzy and nebulous. The section on phonology features two papers. The first concerns horizontal bonds between phonological primes, element-to-element relations within non-linear models of segmental representation. It is argued, on the basis of examples from Bukusu and Sundanese, that inter-element bridging and status alignment help to clarify the consonant harmony phenomena. The second paper focuses on Old English i-umlaut. It shows how the evolution of views and theoretical surroundings on the nature of sound change influences the way in which historical data are handled. The section on morphology and word-formation presents such issues as grinding, diminutives, suffix distribution and compound nouns. The paper on grinding addresses the question of the status of nominal number as an inflectional category, and how recategorization from count to mass nouns is conditioned. The paper on diminutives aspires to grasp the conditions for an unambiguous set of criteria that would satisfy the definition of the issue in question. The paper on the distribution of the suffix –ness across registers calls for the parallels between systematic patterns of variation in language use and proportions in the application of the –ness derivative. Drawing upon corpus data the Author identifies distributional preferences between two stylistically polarized extremes, i.e. fiction and academic prose. The contribution on synthetic compound nouns focuses on the explanatory power of argument assignment in the

Exploring Language through Contrast

ix

hierarchical morphological structure of words and morphology-syntax interface. This detailed account contributes significant observations and addresses new questions in the area of derivational morphology. The section on syntactic phenomena is comprised of three papers, addressing respectively: grammatical constructions with locatum verbs, the status of English NPN forms and typology of Old English verbs. The opening paper is a contribution on secondary patterns with locatum verbs with regard to their distributional characteristics and sheds new light on the relationship between literal and metaphorical senses of fairly stable and predictable grammatical constructions emerging from valency potentials of selected locatum verbs. The result of rigorous and welldocumented data analysis is followed by an inevitable reflection on gradability of metaphor. The second paper discusses the status of NPN forms with regard to categorial status, internal structure and distributional properties. The Author presents two competing views between NPN as a construction (which in fact violates the X-bar schema) and NPN as a lexical item, where the mechanism of syntactic reduplication is assumed to be at play. It is concluded that “comparing two competing approaches can be beneficial”, as such a situation fuels research and exchange of ideas between different linguistic orientations. The next contribution is devoted to some clarifications regarding the typology of Old English verbs with regard to complementation frames. The Author proposes interesting innovations to existing approaches. As a result, the dynamicity of OE verbal complementation can be manifested by a varying number of complements, it exhibits variation with regard to morphological categories of their objects and displays coexistence of more than one argument type, e.g. Case-marked NP or clause and a PP. The section on semantics opens with a discussion of possible explanations regarding the “slips-of-the-tongue” phenomena. The Author elegantly clarifies the issue in view of The Conceptual Integration Theory and Relational Network Model, and provides explanations that account for phonological, lexemic, syntactic and semantic blunders, bringing to attention such elements as competing plans, distractors and forward planning. Juxtaposition of meanings in puns, in the next paper, is yet another attempt to bring focus onto the contrasting principle, where identical forms display dissimilar meanings, which, in turn, produce humorous results in conversational witticisms, newspaper headlines, jokes, etc. It is demonstrated how puns achieve relevance measurable in terms of processing effort and cognitive gains. In the next paper the question of Conceptual Metaphor Theory is revisited – in the spirit of revision. The Author argues against the “classical” model – as it has been known and

x

Preface

championed for decades in its well-established and celebrated form – and proposes a collostructional integrating approach, which does justice to novelty of meaning achieved through collocational/constructional environment of every lexical item plus contextual aspects. Juxtaposition and differentiation arises also in the paper on metaphorical senses in Old English lexical concepts for and FYR and LEOHT. It draws upon the fundamental dualism between the two and aspires to assign to them domains of metaphorical applications reflecting some interesting aspects of the Old English conceptual universe along with axiological implications. The area of pragmatic and discursive phenomena has gained the attention of a number of contributors and contrasts such polarities as: quality vs. tabloid newspapers, subjectification vs. objectification in viewing arrangement (also with regard to political discourse), viewing imposition through the anchoring of “the Self” of the textual narrator, as well as discourse goals and norms. The first paper in the section concerns the choice of rhetorical styles in the aforementioned types of newspapers, The Guardian and The Daily Mail, which are predefined by institutional contexts. The study of stylistic features is based on corpora. Even though the argumentative and generalized rhetoric of the former is juxtaposed against the more emotional and direct reporting of the latter, a gradual infiltration of tabloid style into the so-called quality press is to be observed. The paper on patient-centered and objectifying discourses presents the issue of two distinct strategies in meaning construction in medical case reports. A number of relevant features that display contrasts between the two modes have been identified, such as agent/patient allocation, the use of personal pronouns, the choice between passive and active constructions, etc. The paper addresses a very important issue within the domain of medical discourse speaking in favour of the patientcentered and empathy-driven type of discourse. The next contribution concerns discursive strategies in the media construction of Poland, Russia and the USA in the context of the debate on the US anti-ballistic missile defense shield in Polish and Russian quality papers. By exploring three main types of strategies: referential, predicational and proximization – it was established how the three aforementioned “social actors” vary with respect to changing contextual factors. Authorial persona, the Self, is the central question of the next contribution, where the Author explores the domain of pragmatic realizations of the Self, a semiotic subjective entity, in academic texts, in which the coding of interpersonal relationships and modelling social distance appear to be of fundamental significance. The section is concluded with an important paper on the application of

Exploring Language through Contrast

xi

axiological parameters in conversational analysis. In order to capture highly varied and dynamic phenomena of spoken interaction, the Author employs a complex methodology that stems from a number of disciplines. Politeness and co-operation are but the tip of the ice-berg that in a broader context can shed light on such seemingly distant problems as those regarding the evolution of language. The last paper in this volume seems to elude the architectural design of the book, but is equally important, and it also draws upon the principle of juxtaposition. We place it under the rubric of “normative linguistics”, as it contrasts two individuals – a thespian (Stephen Fry) and a linguist (David Crystal) writing on modern-day prescriptivism. It shows how views and value judgements on both language change and correctness are voiced by these two individuals, and how, despite the two different backgrounds, they both display a comparable measure of sensitivity to linguistic issues. Waldemar Skrzypczak Tomasz Fojt Sáawomir Wacewicz

PART ONE PHONOLOGY

FRIEND OR FOE? ON SOME HORIZONTAL BONDS BETWEEN PHONOLOGICAL PRIMES ANNA BLOCH-ROZMEJ

1. Introduction: the model and the aims The major aim of this article is to provide some insight into the domain of element-to-element relations within one of the non-linear models of segmental representation, namely Government Phonology. It will be demonstrated that in certain systems neighboring elements can be friends, whereas in others, they are foes, both situations having their unique interpretational consequences. Below, a brief outline of the main theoretical concepts will be presented which we find crucial for the discussion of the linguistic phenomena to be analyzed in the remainder of this paper. The phonetic interpretation of a phonological structure is an interplay of both prosodic and subsegmental factors, such as the prosodic strength of skeletal positions,1 inherent nature of phonological primitives, their combinatorial possibilities as well as their status within melodic units. In the forthcoming presentation, we shall employ the framework of Government Phonology (henceforth GP, e.g. Harris 1994, Gussmann 2002, Cyran 2003), its Element Theory in particular, advocating an autosegmental concept of phonological representation and a privative nature of lexical oppositions. According to its compositional view of a segment, elements are the primitive units of melodic structure. They are monovalent2 and autonomous, each having its unique phonetic interpretation. Accordingly, 1

Within the model adopted here, i.e. Government Phonology, the prosodic strength of skeletal slots has to do with their licensing capacity. Licensing, in turn, is defined as an asymmetric relation between two positions of which one licenses, or authorises the presence of, the other. The head of a phonological domain, e.g. a word, constitutes the source of licensing potential for the whole domain. 2 Unlike distinctive features which are binary valued. Elements are pronounceable, both individually and in combinations with other primes.

Friend or Foe? On Horizontal Bonds between Phonological Primes

3

segments consist of either a single element or a combination of primes, where one of them can enjoy the dominant status of the head. A phonological representation is a sequence of onsets and rhymes headed by nuclei. Thus, the model recognizes only three phonological constituents: Onset, Nucleus and Rhyme. These constituents are constructed on the basis of licensing and governing relations contracted by adjacent prosodic positions. Constituent heads are also capable of entering into inter-onset and inter-nuclear relations, the contraction of which requires the fulfillment of more stringent conditions.3 (1) Constituents in Government Phonology4

Cross-linguistic evidence appears to corroborate the view that elements are capable of contracting certain types of relations, not only within segments and in local melody sequences, but also across longer stretches of a phonological representation. Such long-distance alliances happen to involve segments licensed by positions belonging to either inter-nuclear or inter-onset domains.5 The aim of this paper is to depict mechanisms underlying element interplay at a distance. More specifically, we shall dwell on the operation of bridging – a special type of inter-constituent relation that seems to

3

One such condition can be the presence of a nuclear licenser of this relation. In some languages, for instance, this licenser has to be filled with a full vowel, whereas in others, it might be empty, i.e. possessing no active elements. As for the very constituents themselves, the licensing relations therein have to be left-headed. 4 Heads (licensers/governors) are underlined. It has to be clarified here that within branching constituents of onsets and nuclei the relation is stronger and it is called government. Government is a more restrictive form of relation and it often manifests itself as a phonotactic restriction whereby the left-hand position restricts its right-hand neighbour. 5 Notice that such relations involve constituents which are not strictly adjacent but are separated by other constituents. Within the current theoretical framework, these relations involve government or licensing by projection, i.e. one appearing at a higher level of phonological hierarchy.

4

Anna Bloch-Rozmej

foster segmental harmony (Bloch-Rozmej 2008).6 We shall discuss evidence from two languages – Bukusu and Sundanese – to demonstrate the effects of element-to-element bonds. As already mentioned, identical elements licensed by certain positions within inter-constituent domains can be either friends or foes. Our attention will be focused on the behavior of consonants which often appears quite puzzling as their elements contract bonds regardless of intervening melodies that separate them in the melodic plane (i.e. the subskeletal level). We shall employ the tools and mechanisms of Government Phonology to account for prime interactions in consonant harmony phenomena as well as their refusal to submit to harmony in specific contexts (e.g. those pinpointed in Gafos 1999 and Hansson 2001a, 2001b). In search of an adequate framework that is capable of accommodating the linguistic data, we shall propose certain refinements to the classical model of GP.

2. Horizontal bonds between elements As far as the horizontal relations are concerned, we have to consider phenomena that pertain to segments occurring in a sequence. At this stage, it needs to be noted that element spreading falls outside of this category. Frequently when the notion of spreading is evoked, it is stereotypically envisaged as element movement from its host to some other segment, which would result in its duplication. Nevertheless, Government Phonology advocates a totally different understanding of spreading. As proposed in Harris (1994), spreading is defined as an interpretive phenomenon. More precisely, a spreading prime simply extends its range of interpretation to include more than a single segment. Put differently, the element becomes interpreted over a particular stretch of representation. Hence in spreading, quite clearly, there is no inter-element relation to speak of. The relations that we shall analyze in this article obtain between two lexically specified primes (i.e. ones that are independently licensed by different skeletal positions). On the whole, a two-way distinction can be made within the class of horizontal harmony relations: one that is completely identity-based and the other, status-based. The two respective types are schematized in (2a) and (2b) below. 6

The concept of bridging was introduced in Bloch-Rozmej (1998) but its explanatory potential was more thoroughly explored in Bloch-Rozmej (2008). Also the evidence on consonant harmony addressed in this article comes from the latter source.

Friend or Foe? On Horizontal Bonds between Phonological Primes

5

(2)

As depicted in (2a), the very occurrence of identical elements in consecutive segments plays a crucial role in the harmony relation. The same primes can enjoy either different or similar status within their segmental expressions but this has no impact on the nature or consequences of the relation. We shall label this type of bond a substantial identity relation. The latter, status-based, type of relation in (2b) has three sub-cases. In (2b, i., ii.), the identity of melodic material in its substance must additionally be coupled with the same functional status it enjoys within melodies that contract the relation. This means that for a harmonic bond to become established, a particular language will require that the elements not only be identical, but also both function as either heads or operators. Such relations will be labeled as head and operator alignment respectively. The third theoretical option in (2biii.) assumes that primes which are significant for the harmonizing process are heads of their segments. However, it is not their substance that matters but the very headed nature of the melodies in which they reside. Hence, the structure in (2biii.) depicts a headedness alignment relation in which the substantial identity of primes is of minor importance. We shall subsume all the cases schematized in (2b) under the category of functional identity.

3. Consonant harmony: inter-element relations at a distance Consonant harmony constitutes a specific manifestation of inter-element bonds. The harmonic change can be analyzed in terms of the requirement imposed on particular melodies within a given domain to agree with respect to some phonological feature. Within a single morphological domain, it takes on the form of a co-occurrence restriction. In other words, sequences of segments that contain a harmonizing property will be attested in lexical forms, to the absence of disharmonic strings. Across domain boundaries, in turn, harmony manifests itself as a dynamic process of assimilation. The most widely discussed types of melodic harmony

6

Anna Bloch-Rozmej

involve vowel–vowel and vowel–consonant interactions. However, the available studies of inter-consonantal harmony have been based on limited empirical material (e.g., Shaw 1991, Odden 1994, or Gafos 1999). In this context, the work of Hansson (2001a), an OT-based study, is definitely the most extensive in scope, offering the most detailed typological survey of inter-consonantal harmony phenomena. The author presents a wide array of cross-linguistic evidence and formulates generalizations pertaining to consonant harmony. Hansson (2001a) challenges the assumption issued by Gafos (1999), and compliant with his theory of articulatory locality, that the only possible kind of consonant harmony is coronal harmony. Such a premise is claimed to derive from the ability of “the articulatory parameters controlling the shape and orientation of the tongue tip-blade […] to permeate intervening vowels and consonants” without modifying their articulation and acoustics (Hansson 2001a). Hansson brings up apparent counterexamples to Gafos’s analysis, outlining evidence on interconsonantal harmony based on sharing other phonological properties, underlying such processes as stricture, dorsal and labial consonant harmony, secondary articulation harmony, nasal, or liquid and laryngeal harmony. Undoubtedly, the process of melodic sharing that targets two consonants at a distance can also involve elements other than those triggering coronal harmony, for example those specifying the manner of articulation or voicing characteristics of segments. Furthermore, it needs to be observed that the phenomenon of consonant harmony has received various treatments, depending on the framework adopted. Within the model of Generative Phonology, it is consistently assumed that basically the same mechanism motivates both vocalic and consonantal harmony, its nature being determinable primarily on the basis of the vocalic interactions, definitely prevailing in natural languages. Government Phonology offers a range of conceptual devices for analyzing inter-segmental relations. In brief, we should mention interonset and inter-nuclear government, spreading and bridging. In the present article, we intend to focus on the last type, which will be used to express the horizontal relations between consecutive elements at a distance. It will be maintained that bridging, an inter-element bond, turns out to be of utmost importance in the analysis of consonant harmony phenomena. What may seem pretty puzzling about consonant harmony is its selective character. More precisely, the modification targets selected segments across some stretch of representation, even though other realized vowels and consonants intervene between them. What is more, harmony will not affect the intervening melodies in any way (an option depicted in (3b). The harmonizing property modifies a specific expression either to the right or

Friend or Foe? On Horizontal Bonds between Phonological Primes

7

to the left of its segmental host. An alternative to the case just described can be the situation when the onsets whose segments undergo some sort of assimilation are strictly adjacent melodically (as in (3a)). For example, in the case of palatalization spreading in Irish in the form doirse [do:r´‫ݕ‬ԥ] “door, pl.”, the segments [r´] and [‫]ݕ‬, which agree in their slender quality, are prosodically separated by an empty nucleus, which makes them neighbors on the melodic level.7 These two situations can be represented as follows: (3)

D, E, J, M = segments X = element dom-ek ‘little house’ (masc.), ram·a ‘frame’ (fem.) > ram-k·a ‘little frame’ (fem.), jezior·o ‘lake’ (neuter) > jezior-k·o ‘little lake’ (neuter), tom ‘tome’ (masc.) > tom-ik ‘little tome’ (masc.), klucz ‘key’ (masc.) > klucz-yk ‘little key.’ The choice of the exponents of the diminutive in masculine nouns in Polish is to a large extent phonologically conditioned (Szymanek 2010: 204). When it comes to meaning, apart from the denotative meaning, Szymanek claims that each diminutive suffix in Polish conveys a concomitant affectionate or pejorative attitude (2010: 206). Moreover, he notices that “occasionally, the meaning of a morphological diminutive departs quite radically from the notion of smallness” (2010: 210). He mentions several groups of such morphological diminutives, including singulative formations based on mass nouns, like dym ‘smoke’ > dymek ‘a puff of smoke’ or trawa ‘grass’ > trawka ‘a blade of grass,’ or a group approximative pseudo-diminutives based on nouns denoting temporal concepts of quantity, like godzinka < godzina ‘hour’ (2010: 210). A similar remark that diminutives in Polish might have other denotational meanings apart from smallness can be also found in Grzegorczykowa and Puzynina’s (1998) description of diminutives in Polish. Grzegorczykowa and Puzynina, apart from singulative diminutives, mention two other groups: diminutives denoting objects made of the substance indicated by the base word, e.g., Īelazo ‘iron’ > Īelazko ‘an iron;’ and diminutives denoting objects similar to the designatum of the base word, e.g., rĊka ‘hand’ > rączka ‘handle,’ máyn ‘mill’ > máynek ‘grinder.’ However, the formations referred to by Szymanek as pseudo-

56

Dorota Gorzycka

diminutives or morphological diminutives and by Grzegorczykowa and Puzynina as diminutives with secondary mutational meanings are given very little attention in both works. Interesting results of research on derivations with a structure typical of diminutives were presented in articles: Sarnowski (1987, 1991b); and in a monograph by the same author (1991a). In the course of his study concerning Russian and Polish, Sarnowski investigates quasi-diminutives, i.e. nominal formations which maintain the material structure of diminutives, but denote new elements from the extralinguistic reality (1991a: 3), e.g., szpilka ‘spike heel,’ maseczka ‘beauty mask,’ jodek ‘iodide.’ Sarnowski notices that there can be two contradictory explanations for the existence of such words: a morphological one (stanowisko sáowotwórcze) and a lexicological one (stanowisko leksykologiczne) (1987: 59). In the morphological approach, the suffixes used to coin such problematic formations are considered to be homonymous to diminutive suffixes. In the lexicological approach, the words are simply treated as diminutives which have acquired a secondary meaning. Sarnowski favours the latter. He believes that in quasi-diminutives, it is the diminutive sign that obtains new denotational competencies (Sarnowski 1991a: 3). Only later does the new formation undergo lexicalisation, i.e. it becomes a word in its own right in which the suffix is no longer perceived as a separate morpheme isolated from the stem (1991a: 11). In his corpus, Sarnowski identified about 500 Polish quasi-diminutives (1991a: 17). This group includes various words from colloquial speech; botanical and zoological terms, e.g., dzwonek ‘bellflower’ (Campanula), pantofelek (Paramecium); terms from the sphere of technology and sport, Īabka ‘tongue-and-groove pliers,’ páotek ‘hurdle;’ names for foods containing an ingredient denoted by the base word, e.g., migdaáki ‘almond-flavoured hard candies;’ names for various types of entities denoted by the base word, e.g., kolejka ‘narrow gauge railway;’ names for objects which fulfil the same function as the objects denoted by the base word, but different from them in some way, e.g., kapliczka ‘wayside shrine;’ and finally names for chemical compounds, e.g., tlenek ‘oxide.’ The definitions mentioned at the beginning of this paper are believed to describe well diminutives from languages with a rich system of suffixes, such as Polish. Still, even in the case of Polish, there is clearly some disagreement concerning the nature of formations which appear to have a morphological structure typical of diminutives but little or no diminutive meaning. They might be classified as diminutives or excluded from the category, the interpretation is to a large extent complicated by the lack of detailed descriptions in literature. What is more, they can be either

A Comparison of Approaches to Diminutive Meaning

57

interpreted as coinages with suffixes identical to diminutive suffixes and at the same time unrelated to diminutives, or as formations first created as diminutives which have acquired new signification properties. The solution to these problems could only be found after an etymological study concerning all the dubious formations. It is interesting to note that Polish linguists, i.e. Grzegorczykowa, Puzynina, and Szymanek, are willing to expand the morphological category of the diminutive with formations with non-prototypical meanings. Either way, morphological diminutives are not diminutives in the light of traditional definitions. To sum up, there is a problem where to draw the line between diminutives and other derivatives similar to them in form, including quasi-diminutives.

4. The multifunctionality of diminutive suffixes in English Similarly to Polish, in English there are words which might cause difficulties when it comes to their classification; without the knowledge about their origins one cannot decide whether they started as diminutives and acquired new meaning later, or if the fact that they resemble diminutives in form is just a coincidence. Nevertheless, it is easier to study English diminutives because the OED provides detailed etymological information on every entry as well as on individual suffixes. Let us consider two suffixes as examples: -ling and -let. The information provided below on the suffixes and the meanings of derivatives was excerpted from the OED as well as from Marchand (1960) and Schneider (2003). According to the OED, the suffix -ling is a Common Germanic formative. It is appended to nouns, adjectives, verbal stems, and adverbs to form nouns. In Old English, -ling in denominal nouns has the general sense “a person or a thing belonging to or concerned with what is denoted by the base noun,” examples are †earthling “a ploughman” and †gadling “a companion or fellow,” gæd meaning “fellowship.” The derivatives from adjectives have the sense “a person or thing that has the quality denoted by the adjective,” e.g., darling; †evenling “an equal.” Derivatives based on adverbs bear similar meanings, e.g., underling “one who is subordinate.” Some names of birds have the suffix -ling in Old English, e.g., starling “any bird of the passerine genus Sturnus.” Later on, in Middle English and in Modern English, the suffix continued to be used with the same function as in Old English (OED), for instance deathling “one subject to death;” firstling “the first of its kind to be produced, come into being, or appear;” nestling “a young bird which is not yet old enough to leave the nest.” From the 17th century onwards, the

58

Dorota Gorzycka

personal designations in -ling are always used in a contemptuous or unfavourable sense (OED): courtling, popeling “papist,” and worldling. As stated by the OED, the suffix is no longer productive in the uses explained above. There is some uncertainty concerning the emergence of a diminutive force of -ling in English. Marchand (1960: 265) claims that the type youngling “young animal,” which dates back to 1220, gave rise to a strong type wolf-ling “young wolf,” in which -ling was a clipped form of youngling. The meaning of youngness and smallness became so prominent that it extended to other types of coinages, i.e. the suffix -ling began to form words denoting small animals and plants, but the formations had a range of entirely different bases, e.g., fatling. The OED does not comment upon Marchand’s explanation. It only suggests that in Old Norse the suffix -ling had a diminutive force and that there were slight traces of it in the other Germanic languages. In English the earliest certain instance of a diminutive use of this suffix appears to be in codling 1314 (OED). The diminutive force of -ling in English is present in words denoting the young of animals, for example in catling or gosling; and in a few other words, such as bookling, fruitling, and wormling. Moreover, the suffix has been employed in many predominantly contemptuous diminutives designating persons, such as godling, lordling, kingling, and princeling. According to the OED, it is still a living formative in this last use. It needs to be emphasised that -ling has never been extensively used in the formation of diminutives expressing merely smallness of size, although several 19th century writers employed it in nonce-words (OED). When it comes to the suffix -let, Marchand (1960: 263) provides the information that it is appended to nominal bases to coin nouns. Schneider (2003: 96) writes that some linguists regard -let as the purest diminutive suffix in English. Additionally, he cites Cannon’s (1987) statistics according to which -let is one of the most productive suffixes in presentday English, occupying position 17 among the thirty most frequently used suffixes. Similarly, in the OED there is a remark that in the formation of nonce-words -let is now perhaps the most frequent of diminutive endings. The OED states that the oldest words in English with this ending are loanwords from Old French which were formed by adding the diminutive suffix -et, -ete to nouns with the ending -el. Examples include bracelet, chaplet, and hamlet. The OED states that It is somewhat difficult to see how these words gave rise to the English use of -let as a diminutive suffix, as none of them, except the heraldic crosslet, have the appearance of being diminutives of English words; possibly

A Comparison of Approaches to Diminutive Meaning

59

French diminutives like enfantelet, femmelette, osselet, tartelette, were directly imitated by some English writers. (OED)

Marchand puts forward a different explanation (1960: 263). He notices that first, there was a word tartlet 1420, connected with tart 1400. Moreover, there were such loanwords as castelet 1320, connected with castle; gablet 1440 connected with gable; circlet 1481 connected with circle; islet 1538 connected with isle; which etymologically contained -et, but possibly gave rise to -let as a separate suffix. Marchand (1960: 263) mentions that crosslet 1538 could be the first word coined with the suffix let. The OED says that some early diminutives in -let include armlet 1538 “little arm of the sea,” others are ringlet and kinglet. Marchand notices that by 1550, -let had become an established suffix in English (1960: 263). The formative became frequently used in the 18th century and continued to be very productive throughout the 19th century (Marchand 1960: 263). Exemplary diminutives with -let are: booklet, brooklet, jokelet, and keylet. Apart from the purely diminutive meaning of smallness, some formations with -let can indicate young age of living beings. This use could have originated as a semantic pattern from the loan word eaglet 1572, an adapted version of French aiglette, diminutive of aigle. Youngness can be ascribed not only to animals, but to plants and human beings as well. Little wonder that the possible choice of base words was soon enlarged by those two groups. Examples include: toadlet, piglet, froglet, princelet, lordlet, plantlet, and oaklet. It is important to note that this use of -let is productive. Moreover, suffix -let can convey a broadly understood negative assessment of the referent, sometimes even contempt. According to information included in the OED, the first derivation of this type was kinglet 1603, a loan translation of French roitelet. Soon more derivatives followed, including princelet, bardlet, squirelet, queenlet, mayorlet, and priestlet. At the same time, it is possible to encounter derivatives referring to people and expressing a positive attitude or even affection, such as wifelet, which was classified in the OED as a term of endearment. The suffix -let is also used in derivatives based on the names of body parts, referring to articles of jewellery, armour or attire. According to the OED, the starting point from which such formations have been recorded is the introduction of bracelet 1438, a loan from French. Marchand (1960) does not agree. He claims that the starting point is the loan frontlet 1478. The fact remains that other similar formations were soon coined, including armlet, earlet, necklet, anklet, and leglet. What is more, -let is often used in broadly understood scientific terminology: deerlet “the smaller species of Musk Deer, found in South-

60

Dorota Gorzycka

East Asia and the adjacent islands;” fruitlet “a single member of an aggregate fruit;” platelet “thrombocyte;” rootlet “a slender subsidiary root; a branch of the primary root of a plant;” winglet “a small projecting part in a piece of mechanism.” The first conclusion that can be drawn from the analysis of the two suffixes is that different semantic types of formations with one suffix can have separate origins, cf. the semantic types represented by booklet, lordlet or wristlet which emerged as imitations of different loanwords. Some of them are classified as diminutives, while for instance formations referring to articles of ornament or attire are not. Secondly, let us look at the word darling. It is considered to be a diminutive, yet untypical as it was coined from an adjectival base. This word, however, has a different origin than other diminutives with -ling. It dates back to Old English when the suffix -ling was used in words denoting people with particular characteristics, in this case a person dear to somebody, and had no diminutive meaning at all. This particular example proves that even when a formation displays features typical of diminutives, its etymology might prove that it is just a coincidence. Moreover, there are constructions with diminutive suffixes which convey a more complex meaning than that of smallness. Some examples incorporating smallness with the primary meaning of individuation are: bomblet “any of many explosive devices forming part of a larger, complex weapon” or branchlet. In others, e.g., deerlet, goslet “a few diminutive species of geese (Nettepus);” swiftlet “a small species of swift, as those of the genus Collocalia;” the formative -let is used to create names for animals that are smaller than the typical representatives denoted by the base word. The referents of the diminutive may be classified as belonging to the group named by the base or may simply resemble its members in some respect. Regardless of the fact that the meaning of smallness can be traced in the semantics of the abovementioned words, they are not normally classified as diminutives. Summarising, the formative -ling is of Germanic origin and its diminutive force emerged relatively late, when the suffix was productive and already well established. Thus, there are numerous formations with this suffix that do not inform about size assessment, and relatively few with a diminutive force. When it comes to the suffix -let, from very early on it was a diminutive suffix; however, it is very easy to enumerate a plethora of formations with this suffix that are not prototypical diminutives. The investigation proves that in both Polish and English diminutive suffixes appear in formations that do not have a typically diminutive

A Comparison of Approaches to Diminutive Meaning

61

meanings. There is a core of prototypical diminutives, but there are numerous derivatives which share with them only some properties. Those formations, however, cannot be classified as diminutives once we take into consideration the traditional definitions. What is more, drawing distinction between diminutives and non-diminutives not relying on historical data can result in arbitrariness.

5. Approaches based on the priority of form In this section, we intend to present the approaches to diminutive meanings proposed by Taylor ([1989] 2003) and Jurafsky (1996). Both authors are cognitive linguists and wanted to trace the regularities in the meanings of formations with various exponents of diminutiveness. As we have demonstrated, such derivatives exist in Polish and in English, but they can be observed in a number of other languages as well. Most of Taylor’s examples come from Italian. Jurafsky took a much broader scope and explored more than 60 languages. The title of this section the priority of form comes from the starting point for research that both authors took; they first identified the morphological devices denoting smallness. Their second step was to find out which other functions those devices can have. As has been said above, Taylor’s assumption is that diminutive morphemes are polysemous. He writes that “there is considerable agreement across different languages with regard to the kinds of meaning that can be conveyed by the diminutive” (2003 [1989]: 173); however, most of his conclusions were based on the behaviour of Italian diminutive suffixes. His approach disregards the differences between individual suffixes, perceiving them as representations of a single macro-category. The diminutive, according to Taylor, is a polysemous category whose various meanings are linked through metaphor or metonymy to the central sense which is the small size of a physical entity ([1989] 2003: 173). There is no common semantic component between the senses. Taylor notices that along with material nouns, nouns designating more abstract entities can be diminutivized as well, so can be other parts of speech than nouns. He explains this extension of usage by metaphorisation in which the notion of smallness is transferred from the spatial to the nonspatial domain. Therefore, nominal diminutives can express the short temporal duration, the reduced strength, or the reduced scale of an entity: sinfonia > sinfonietta ‘small-scale symphony,’ cena > cenetta ‘small supper,’ pioggia > pioggerella ‘light rain, drizzle.’ (Taylor 2003 [1989]: 173)

62

Dorota Gorzycka

Diminutivized adjectives and adverbs express reduced extent or intensity: bene ‘well’ > benino ‘quite well.’ Verbs can also be used as bases for diminutives and they “usually designate a process of intermittent or poor quality: dormire ‘sleep’ > dormicchiare ‘snooze’” (2003 [1989]: 173). Furthermore, Taylor mentions that diminutives can convey an attitude. According to Taylor, this meaning emerged as a metonymic transfer from the central sense. He writes small animals and small children can be cuddled and caressed without embarrassment or fear. The association of smallness with affection is thus grounded in the co-occurrence of elements within an experiential frame (2003 [1989]: 174).

An attitude of affection is not the only one that can be experientially associated with smallness. Diminutives are sometimes applied to express depreciation, e.g., tesina would be used to convey a low opinion about a thesis. Taylor (2003 [1989]: 174) states that the experiential base for this meaning is evident: “superior worth correlates with increased size, decreased size with diminished worth.” Additionally, Taylor remarks that diminutives in Italian and in other languages as well, are used in the following senses: the dismissive sense, the approximative sense, and for intensification. The dismissive sense of the diminutive resulted from an association that small things are of little importance; therefore, storia ‘story’ > storiella ‘lie, fib’ (2003 [1989]: 174). The approximative use is restricted to expressions of quantity. The speaker who uses a diminutive implies that the exact value of the diminutive is unimportant, he or she excuses himself or herself for being imprecise. Thus the real quantity may in fact often be bigger than the one denoted by the base word (Taylor 2003 [1989]: 174). Diminutives used as intensifiers, e.g., caso ‘state of affairs’ > casino ‘uproar, chaos,’ are a metonymic extension from the central sense; the centre of an entity is necessarily of smaller dimensions than the entity in its totality. Thus the diminutive comes to denote the very essence of a thing, a thing stripped of its non-essential periphery. (2003 [1989]: 175)

Finally, Taylor notices that many diminutives “have a tendency to acquire the status of independent lexical items, with more or less specialized meanings” (2003 [1989]: 176), cf. quasi-diminutives in Sarnowski (1987, 1991a, 1991b) Among the examples, he mentions sinfonietta, not a short symphony, but a different musical form in its own right.

A Comparison of Approaches to Diminutive Meaning

63

The second approach to be discussed is Jurafsky (1996). As has been stated, he studied more than 60 languages from different language families. For Jurafsky the diminutive is “any morphological device which means at least ‘small’” (Jurafsky 1996: 534). He further writes that its realization may take any of the forms in the wide universal repertoire of morphological devices: affixes, shift in consonant, vowel, or lexical tone, and changes in noun-class or gender. (Jurafsky 1996: 534)

Hence the material for his study comprises morphological devices which in most cases are multifunctional; however, the meaning of smallness must be included among their functions. Jurafsky proposes to model the synchronic and diachronic semantics of the diminutive category with a radial category [. . .], a type of structured polysemy that explicitly models the different senses of the diminutive and the metaphorical and inferential relations which bind them. (Jurafsky 1996: 533)

From the synchronic point of view, the model explains the various senses of the diminutive, from the diachronic point of view, the model is supposed to act as an archaeology of meaning, expressing the generalizations of the classic mechanisms of semantic change (metaphor, abstraction and inference) as well as a new one: lambda-abstraction, which accounts for the rise of quantificational meaning and second-order predicates in the diminutive. (Jurafsky 1996: 533)

Unfortunately, due to space constraints we will not discuss the intricacies of the process of lambda-abstraction in this paper. Jurafsky notices that “the varied senses of the diminutive occur with astonishing regularity across languages” (1996: 535). He voices a claim that cross-linguistically the origins of the category lie in the words semantically or pragmatically related to children (1996: 537). To prove it, Jurafsky proposes a new reconstruction of the protosemantics for the PIE suffix *-ko-. He supports his theory concerning the order in which the different senses of the diminutive emerged with unidirectionality hypotheses. According to the first of them, meaning changes from the more specific to the more abstract, this phenomenon is referred to as bleaching, generalization, or desemanticization (1996: 540). The second group of hypotheses focuses on

64

Dorota Gorzycka the tendency of semantic change to proceed from the “real” physical or spatial world or the ideational domain to create more qualitative, evaluative, and textual meanings. (Jurafsky 1996: 541)

Among the senses, the meaning of child or offspring, e.g., duck > duckling, gave rise to the semantic meanings of: smallness, member, and female gender usher > usherette. The last two, as Jurafsky claims, can be interpreted as extensions from the meaning of smallness as well. Among the pragmatic meanings that originated from child/offspring, we have affection, pets, sympathy, and intimacy. In Jurafsky’s model, the stems expanding from the meaning of smallness are: small type-of, here the diminutive object is classified in the same ontological hierarchy as the bigger one, but they differ in something else than size, e.g., cigar > cigarette; individuation/partitive hoof > hooflet; approximation, arguably red > reddish; and intensity/exactness. Among the pragmatic meanings Jurafsky mentions two: contempt and diminutives used as metalinguistic hedges. The use of diminutives as metalinguistic hedges might be also interpreted as extension from the meaning of approximation. Jurafsky writes that “the pragmatic type of hedge is often used in an extended way to soften or weaken the illocutionary force of the entire utterance” (1996: 557). Thus, diminutives are often used in requests, offers, and to elicit sympathy. Finally, the meaning of small type-of expands into two: imitation arrow > arrowlet “the stalked and plumose seeds of the Dandelion and allied plants;” and related-to neck > necklet. The new meanings of imitation and relation no longer indicate any size assessment. Jurafsky stresses the fact that the semantic and pragmatic meanings that he enumerates do not have to be present in all languages since the model was created with the intention to show universal tendencies in the category, not to provide rigid rules in the realm of diminutive meanings. The two cognitive approaches presented in Taylor (2003 [1989]) and Jurafsky (1996) aim at creating the model which would include the senses conveyed by diminutive suffixes that are not mentioned in the traditional definitions. Both authors notice that there is a cross-linguistic regularity in the senses that diminutive can express. Still, Taylor’s (2003 [1989]) model was developed mostly by using data from Italian. Jurafsky takes a broader perspective than Taylor, and tries to put forward a universal model, thus making his approach more controversial. Especially, if we take into consideration the ambitious assumptions that the model represents the order in which senses originated diachronically, and Jurafsky’s belief that once a diminutive in a particular language has a sense, it either will have

A Comparison of Approaches to Diminutive Meaning

65

or have had each prior sense on the same path to the root of the category (1996: 543). It needs to be highlighted that in his model Jurafsky does not pay attention to how meaning extensions are structured for individual suffixes. On the contrary, he is interested in the history of the category as a whole, i.e. in the order in which various morphological devices acquired meanings as well as in the motivations behind those meaning extensions in general. His claims can be juxtaposed with the facts about diminutive suffixes and their etymology revealed in the OED. It is true that the suffix -let started off as a diminutive suffix and acquired new meanings over time. The suffix -ling, on the other hand, first had a more general and vague sense which only in some cases can be labelled as “related-to.” It acquired a diminutive force later and is used with this sense in a relatively small group of derivatives when compared to the total number of formations that are listed in the OED with this suffix. Hence it is important to remember that Jurafsky’s model represents the most probable paths of meaning development in the category, especially as it takes into account the human cognitive factor, but quite likely it will not mirror the development of individual affixes. Moreover, the model lists those meanings of the diminutive that appear with some regularities crosslinguistically, but it does not take into consideration the idiosyncrasies in the meanings of individual suffixes. Thus, drawing conclusions based directly on the model about the meanings or the historical development of particular affixes might result in inaccuracies.

6. The approach based on the priority of meaning (Schneider 2003) The previous section was devoted to the meanings of morphological diminutives. This one describes Schneider’s (2003) attempt to study diminutives from an entirely different perspective. He adopts a formalfunctional paradigm based on the claim that in order to understand the nature of language, grammar, pragmatics, and the interplay between them need to be studied (Schneider 2003: 1). He emphasises the importance of the distinction between diminutive form and meaning and, consequently, between diminutive formation and diminutivization, which are a formal and a semantic process respectively (2003: 4). Schneider believes that prototypical diminutives are found in some languages, but the concept of diminution appears in all languages. The particular linguistic devices used to express diminution depend on the structural make-up of the language in question. For him the diminutive “is not a morphological category in the

66

Dorota Gorzycka

first instance, but a term which refers to all expressions of diminution” (Schneider 2003: 1). For that reason, the starting point for any analysis of diminutives in any given language should not be individual formal devices such as suffixes, but the expressed meaning, i.e. smallness in all of its literal and metaphorical meanings (including closeness as “small social distance”) and the positive or negative attitudes connected to it. Diminutive research should aim at identifying the formal devices available in any given language for expressing this concept and at establishing how these devices are used in communication. (Schneider 2003: 15)

In the course of research Schneider investigates a range of constructions with the meaning of smallness. He distinguishes two very broad types of diminutive formation processes: synthetic diminutive formation and analytic diminutive formation. As for synthetic diminutive formation, he enumerates several types. The first one is suffixation. In the English language, Schneider (2003: 85) identified fourteen diminutive suffixes: {A}, {EEN}, {ER}, {ETTE}, {IE}, {KIN}, {LE}, {LET}, {LING}, {O}, {PEG}, {POO}, {POP}, and {S}. He adopts upper case notation since some of the suffixes can be combined to form complex suffixes, e.g., -kin+s, or have several spelling variations, for instance {IE} might be spelled as -ie, -y, -ee, and -ey. Apart from suffixation, other types of synthetic diminutive formation include: prefixation with mini- and micro-; reduplication as in Annie-Pannie; compounding, e.g. formations with baby or dwarf as in baby plant; and truncation, which Schneider treats as a more predictable type of clipping, like in Dan < Daniel (Schneider 2003: 7–9). Some formation processes con co-occur in one formation, e.g. truncation with suffixation, as in Andrew > Andy. What is interesting about Schneider’s approach is that he discerns analytic diminutives. They are “adjective + noun constructions in which the noun is the base word and the adjective the diminutive marker. Adjectives used in this function belong to the word field SMALL” (Schneider 2003: 122), e.g., little, small, tiny, minute, as in little house, small car. Thus, he expands the group of diminutives with constructions that are not normally treated as diminutives. To prove his point, he investigates how those devices are used in speech acts, and observes that they are applied by English speakers in a similar ways that synthetic diminutives are used by speakers of languages with more complex morphology than English.

A Comparison of Approaches to Diminutive Meaning

67

7. Conclusion The recent studies of the diminutive prove that the traditional view on the category is insufficient. The analyses show that in such languages as English, Italian or Polish, there are derivations with either diminutive suffixes or suffixes identical to diminutive suffixes, but with different meanings than prototypical smallness. Along with that, the meaning of smallness can be conveyed by different devices, not necessarily the prototypical ones, i.e. suffixation. The studies of diminutives presented in this paper reflect these features of natural languages. There are approaches, i.e. Taylor (2003 [1989]) and Jurafsky (1996), which investigate the various meanings that the morphological exponents of diminutiveness are likely to acquire. The other strand of research is centred around the meaning of smallness. This is the path that Schneider took in his study of diminutives in English. As a result he investigated constructions, such as analytic diminutives, which are not classified as the exponents of diminutiveness in any source with a more traditional view on diminutives. As a result, every linguist investigating diminutives is left with some room for interpretation of what he or she thinks diminutives are. Wording a detailed definition in any paper dealing with the subject of diminutiveness is of great importance as it helps the readers interpret and compare the findings. Especially, since all of the new approaches discussed here (Taylor 2003 [1989]; Jurafsky 1996; Schneider 2003) tend to expand the category of the diminutive with new constructions.

References Asher, Ronald and Seumas Simpson (eds.). 1994. The Encyclopedia of Language and Linguistics, Volume 10. Oxford, New York, Seoul, Tokyo: Pergamon Press. Cannon, Garland. 1987. Historical change and English word-formation. New York: Lang. Carter, Ronald and Michael McCarthy. 2006. Cambridge grammar of English. A comprehensive guide. Spoken and written English grammar and usage. Cambridge: Cambridge University Press. Crystal, David. 1991 [1980]. A Dictionary of Linguistics and Phonetics. Third edition. Oxford: Basil Blackwell. Grzegorczykowa, Renata, Laskowski, Roman and Henryk Wróbel (eds.). 1998. Gramatyka wspóáczesnego jĊzyka polskiego. Morfologia. Wydanie drugie [Grammar of modern Polish. Morphology. Second edition] Warszawa: Wydawnictwo Naukowe PWN.

68

Dorota Gorzycka

Jurafsky, Daniel. 1996. “Universal tendencies in the semantics of the diminutive.” Language 72.3: 533–578. Marchand, Hans. 1960. The categories and types of present-day English word-formation. Wiesbaden: Otto Harrassowitz. McArthur, Tom (ed.). 1992. The Oxford Companion to the English Language. Oxford and New York: Oxford University Press. OED. Oxford English Dictionary. http://oed.com. DOA: 15 June 2011 Sarnowski, Michaá. 1987. “Leksykalizacja deminutiwów w jĊzyku rosyjskim i polskim” [Lexicalization of diminutives in Russian and Polish]. Slavica Wratislaviensia XLIV: 59–70. —. 1991a. “Quasi-deminutiwa w jĊzyku rosyjskim i polskim” [Quasidiminutives in Russian and Polish]. Slavica Wratislaviensia LVIII: 1– 94. —. 1991b. “Neutralizacja deminutiwów w jĊzyku rosyjskim i polskim” [Neutralization of diminutives in Russian and Polish]. Slavica Wratislaviensia LXIV: 75–84. Schneider, Klaus P. 2003. Diminutives in English. Tübingen: Max Niemeyer Verlag. Szymanek, Bogdan. 2010. A Panorama of Polish Word-Formation. Lublin: Wydawnictwo KUL Taylor, John R. 2003 [1989]. Linguistic Categorization. Third edition. Oxford and New York: Oxford University Press.

THE DISTRIBUTION OF THE SUFFIX -NESS ACROSS REGISTERS WOJCIECH GUZ

1. Introduction Variation is inherent in language. On different occasions, different forms are used to express the same idea and still different forms may be used in a similar situation by different language users. Speakers constantly switch from one language variety to another depending on situational settings, the kind of discourse they are involved in, the purpose of language production, the intended audience, and the nature of the speaker– hearer relationship. In general, it is commonly acknowledged that such non-linguistic, contextually defined factors underlie the differentiation of distinct language varieties or registers, where register is a “cover term for any variety associated with a particular configuration of situational characteristics and purposes”1 (Biber and Conrad 2001: 175). Thus registers are defined in terms of non-linguistic factors. And yet these situational parameters of language use correlate with linguistically definable properties that are traceable to particular registers. The study of register variation thus focuses on systematic patterns of variation in language use that are instantiated in certain patterns of linguistic features (i.e. grammar constructions, vocabulary items, word-formational elements, 1

The term genre is also used with the same meaning by some authors and corpus compilers. Most linguists use one or the other, or both, interchangeably, on different occasions. However, there have been attempts to differentiate between the two. Lee (2001: 47) argues that “the term genre is used to describe groups of texts collected and compiled for corpora or corpus-based studies, and register has typically been used in a very uncritical fashion, to invoke ideas of ‘appropriateness’ and ‘expected norms,’ as if situational parameters of language use have an unquestionable, natural association with certain linguistic features and that social evaluations of contextual usage are given rather than conventionalised and contested.” Biber (1994) gives a survey of how these and other related constructs (text types, styles, sublanguages) have been variously used and defined.

70

Wojciech Guz

etc.).2 These features also include phenomena relevant to morphology and one such phenomenon will be discussed in this study. Corpus-based studies indicate that, of all morphologically defined features, affixal nominalizations in particular can shed new light on some of the systematic ways in which various language registers differ (cf. Biber 1988). For example, Biber (1988) establishes that a relatively high frequency of occurrence of nominalizations is an important factor in isolating purely linguistic differences between registers such as conversation, panel discussion, prose and academic discourse. What is of interest to us, all nominalizations are superficially taken there as a unified group to contrast the four genres, no allowances are made for individual patterns of particular affixes. Similarly, Chafe and Danielewicz (1987) consider all nominalizations collectively to be more characteristic of one register than another. Paying more attention to individual affixes, Biber et al. (1998: 63) assert that, although nominalizations are by far most common in academic texts overall, “[t]he -ness ending is more important in fiction than in either academic prose or speech.” The suffix -ness is singled out as the only nominalizing suffix that does not follow the general distribution pattern of nominalizations. Claims to the same effect are made by Biber et al. (1999). On closer inspection, however, the accuracy of such statements is undermined by investigations into the morphological make-up of the base form. For example, derivatives of the X-ive-ness type (e.g. selectiveness) from the British National Corpus (BNC) are found in this study to be almost nine times as common in academic prose as they are in fiction (see section 4). Similarly, Biber (2004: 21) makes another sweeping generalization in his claim that nominalizations, along with some other language features such as long words, abstract nouns and relative clauses, are “typical of written non-fictional registers intended for specialist audiences.” As both accounts, i.e. Biber et al. (1998) and Biber (2004), regard -ness words as nominalizations, the two quotations above must now appear confusingly contradictory. Thus, it is the kind of more detailed analysis that we have illustrated with X-ive-ness derivatives, i.e. the scrutiny of the morphological structure of the base form, which will properly describe the distribution of -ness. Our study of the suffix in the BNC indicates that its distribution is not as consistent as has been considered to be. Little if anything is known 2

For example, passive voice tends to co-occur with nominalizations to characterize more formal registers. At the same time, these registers will have markedly few pronouns and contractions (Biber 1988). The identification of such patterns of cooccurrence is mostly attributed to the work of Douglas Biber (especially 1988).

The Distribution of the Suffix –ness

71

from previous research about the influence of base-internal morphological structure on register variation: few analyses extend beyond the rightmost affix.3 We add to the research carried out hitherto by examining the nominalizing English suffix -ness and putting forward claims about its varied distribution across registers. It will be shown that the suffix exhibits preferences as to its occurrence in particular registers and that the internal morphological make-up of the base forms that it attaches to may also have a significant bearing on its quantitative distribution.

2. Methodology and procedure The results of our study are based on the 100 million word BNC corpus (available at http://corpus.byu.edu/bnc/) of contemporary British English. All the texts in the BNC are classified into distinct registers (or genres) according to criteria established by the corpus compilers and this corpus-internal division has been used in this study. The range and sample sizes of distinct varieties of English investigated here are as follows4: Spoken (taken as a whole) – approximately 10.33 million word tokens; Fiction – 16.19 million; Academic – 15.43 million; Newspapers – 10.64 million; Non-Academic (Non-Fiction)5 – 16.63 million; Popular Magazines – 7.38 million. In order to retrieve all eligible nominalizations the entire BNC was searched and the word lists thus obtained were hand-edited, deleting irrelevant items, consulting the Oxford English Dictionary and/or analyzing the words in their context in the BNC when necessary. This yielded 1700 word types in -ness. Afterwards, further queries were run to investigate the distribution and frequency of the suffix across the six sub-corpora. Plural nouns were collapsed under their respective singular forms for the purpose of token frequency measurements. The results are given below in section 4.

3

Base-internal complexity is ignored in, for example, Biber (1988), Biber et al. (1998, 1999) and Plag et al. (1999). Suffix–suffix combinations are considered by Baayen and Renouf (1996) but their analysis is limited to newspaper English only. 4 The discrepancies in sample sizes are largely irrelevant. Frequency counts of word tokens are normalized to a common basis of 1 million words of text. 5 In the BNC, the non-academic genre consists of texts that are essentially academic in content but have been popularized for a general audience.

72

Wojciech Guz

3. Data The subject of this study were de-adjectival nominalizations in -ness (creepiness). This excluded items departing formally from this criterion (for example through affix generalization) such as denominal -ness derivatives (owlness, godness, Guinessness). An additional criterion of selection that was adopted here was that the words qualifying for inclusion had to be clearly first-cycle -ness derivatives and not any instantiations of the suffix, such as prefixed formations whose base forms end in -ness. For example, semi-baldness was excluded because it is best seen as a derivative of semi- prefixation. To count semi-baldness as another word type of -ness would have been to artificially inflate the productivity of this suffix and distort the accuracy of our findings. That is why most prefixed words ending in -ness were excluded. However, exceptions to this rule were made in two cases: (1) when prefixation (or compounding) clearly precedes suffixation in the derivation of a word, e.g. outrageous#ness (not out#rageousness) and wideawake#ness (not wide#awakeness), and therefore the result of the rule that applies last is a -ness nominalization; and (2) when postulating prefixation would have implied the existence of unattested forms that cannot stand on their own, e.g. unruli#ness (*ruliness) and law-abiding#ness (*abidingness). In both cases such items were retained in the sample. Compound nouns were included only when they were eligible candidates meeting the above criteria, i.e. when they were clear cases of last-cycle nominalization (topsy-turviness, soft-spokenness) and when the second constituent of the compound could not stand on its own (*turviness, *spokenness). Items with two free-standing constituents are best viewed as nominalizations-turned-compounds and thus were excluded, e.g. sword-sharpness. All in all, the overriding principle in the selection of data was to include only those items that clearly illustrated ness nominalization. For this same reason blends were excluded altogether. Any unclear items that could not be traced to any possible base forms were also deleted (e.g. chasifness).

4. Results and discussion In this section we discuss register variation by considering token frequencies. We first concentrate on the distribution of affixal nominalizations overall and then look into how -ness departs from this pattern. Below we compare our research results with those reported by Biber et al. (1998), who analyze the distribution of -ness, -ity, -ion and -

The Distribution of the Suffix –ness

73

ment based on samples of academic prose, fiction (both from the Longman-Lancaster Corpus) and spoken English (from the London-Lund Corpus). Below we cite their findings along with those obtained in the present study, given as joint normalized token frequencies per 1 million words. 50000 40000 30000

Spoken

20000

Fict Acad

10000 0

Biber et al. 1998

Presen t study

Fig. 1. Joint token frequencies of the suffixes -ness, -ity, -ion and -ment across registers

Despite the differences in the corpora used, the frequencies are consistent in both accounts. Fiction and spoken language have approximately the same proportion of nominalizations; namely, a quarter that of academic. 6 It is commonly believed that, overall, written registers are lexically richer, not only with respect to nominalizations but indeed as far as the type-token ratio is concerned (Biber et al. 1999). Morphologically speaking, Plag et al. (1999) hypothesize and subsequently show that derivation in spoken English is much less productive. These claims are consistent with our findings above. Most other nominalizing suffixes (-ity, -al, -ance/-ence, etc.) tend to follow the overall pattern in Figure 1, both collectively and individually (see Guz 2010). When considered jointly in token frequencies and across several other registers, abstract nominalizations are distributed in the following sequence of increasing frequencies: Fiction < Spoken < Popular magazines < News < Non-Academic < Academic, as plotted in Figure 2:

6

Note that the numbers reported by Biber et al. (1998) are double those obtained in this study. This may be due to differences in the criteria of data selection adopted by Biber et al. (1998). It is possible, for example, that their frequencies include prefixed formations.

Wojciech Guz

74

35000 30000 25000 20000 15000 10000 5000 0

Fict

Spoken

Pop

News Non-acad Acad

Fig. 2. Normalized joint token frequencies of twelve nominalizing suffixes

The increasing numbers of nominalizations occurring in the pattern noted above are only to be expected. Academic prose seeks to condense as much information – often abstract notions – as possible into the minimum of form. Nominalizations offer such efficiency and condensation of ideas as longer phrases, even clauses, can be effectively replaced by a single complex word (cf. text categorization vs. the manner in which text is categorized, his clumsiness vs. the fact that he is always clumsy). Typically, key words in such structures are shifted, or recategorized, to become complex nominals.7 In a broader perspective of the theory of word formation, this particular effect of syntactic recategorization or transposing is one of the most basic functions of word formation, next to the labelling or referential function (i.e. inventing a word label for a new concept) (Plag 2003: 73–74; Booij 2005: 14; Lieber 2005: 406). Condensation of information is one possible motivation for syntactic recategorization, in which case abstract nominalizations are typically used. However, other syntactic categories can also be involved in similar operations, although possibly for different functional purposes. For example, Booij (2005) and Plag (2003) cite stylistic variation and text cohesion as possible results of recategorization in the following sentences (examples (a), (b) and (c) after Kastovsky 1986: 595 and quoted by Booij 2005: 14): a. He made fists [...] He defisted to gesture. b. If that’s not civil, civilize it, and tell me. c. [. . .] and whether our own conversation doesn’t sound a little potty. It’s the pottiness, you know, that is so awful. 7 Lieber (2005: 406) refers to deverbal nominalizers (-al, -ment, etc.) as transposers of verbs to nouns.

The Distribution of the Suffix –ness

75

Abstract nominalizations are thus to be singled out from among all other classes of complex words as particularly useable in information condensation and syntactic recategorization. This is simply due to their general categorial semantic paraphrase of ‘the act/state of...,’ ‘the quality of....’ We also note that abstract nominalizations in general are insignificant in their use in the labelling (referential) function of word formation. Turning back to Figure 2, we note the disproportionate pattern of distribution across register that is typical of nominalizations overall and most individual suffixes. Still some exceptions are to be observed, notably the suffix -ness, which is the most frequent in fiction (Figure 3). 1000 800 600 400 200 0 Fict Spoken Pop

News

Nonacad

Acad

Fig. 3. Normalized token frequencies of -ness across registers

Clearly, -ness is indeed marked by its preference for fiction texts, as noted by Biber et al. (1998) and (1999) (cf. section 1). A functional interpretation of this finding might be that -ness words – in comparison with words derived with -ity, the functional rival of -ness – seem more speaker-friendly and less technical at the same time (e.g. tenaciousness vs. tenacity). There may be several reasons for this. Firstly, the suffix -ness may be preferred by speakers because it is a ‘safer’ option when little editing time is available in online production: it is easily parsed out, i.e. it is straightforwardly attached to its base with a clearly observable morpheme boundary with no adjustment or truncation operations (again, tenacious-ness vs. tenac(ious)-ity); it has no phonological effect on the base form either (cf. the change in vowel quality in tenacious – tenacity); hence, -ness and its base forms are more easily strung together and the resulting words are more easily decomposed and interpreted by the hearer (cf. Hay 2003; Hay and Baayen 2003). Derivatives in -ity, on the other hand, may be both formally and semantically opaque. This in turn may be because -ity is Latinate and the exact ways in which it combines with

76

Wojciech Guz

(usually non-native) bases may be obscure to speakers of English. The suffix -ness, on the other hand, is native, although it is questionable whether etymological considerations of native or non-native origin in themselves influence the choice of one suffix over another.8 Rather, as mentioned above, decomposability and full predictability of usage will be a more likely explanation.9 Secondly, another reason why -ness words are more informal and particularly common in fiction is the very meaning of many of its bases. Many of high-frequency -ness derivatives denote personal qualities or feelings that rarely surface in technical/scientific texts but often do in fiction, such as happiness, kindness, sadness, tenderness, etc. Such formations will not only be infrequent in technical texts but even more so in spoken language (Figure 2). Biber et al. (1998) argue that spoken language, due to its limited use of nominalizations overall, is more likely to express the same notions by means of adjectives, often attributed directly to the speaker or hearer (e.g. I’m happy, I feel sad, you were very kind). Thus it has been noted that -ness nominalizations are more frequent or “more important” (Biber et al. 1998: 63) in fiction than in other registers, especially academic prose. This is also the picture emerging from Figure 2 in the present study. Yet any such claims must inevitably be rectified once -ness nominalizations are investigated more closely. Below we present our findings of the effect of morphological structure on register distribution. We take into account the internal morphological structure of the base forms themselves. Table 1 shows normalized token frequencies across the registers with reference to the morphological structure of the base form.10

8

Admittedly, free choice between -ness and -ity will only occasionally be an option for one and the same base form (if only because of the selectional restrictions of -ity). Instead, either suffix will usually be the more common alternative or indeed the only reasonable choice. 9 See Hay (2003), Hay and Baayen (2003) for a detailed treatment of the effects of decomposability on productivity. In a nutshell, “any factor which facilitates decomposition of complex forms should also facilitate the emergence of productivity” (Hay 2003: 151). 10 Note that frequencies observed for combinations of suffixes, when added together for a particular register, do not equal their respective total counts from Figure 2. This is because some of the items considered in the total counts did not fit any of the suffix combination templates. For example, we have ignored combinations such as -al+ness, -ary+ness, -ate+ness due to their negligible occurrence in the BNC (around a dozen each).

The Distribution of the Suffix –ness

77

Fict Spoken Pop News Non-acad simplex root+ness (fakeness) -y+ness (creepiness) -ful+ness (stressfulness) -ish+ness (quirkishness) -ous+ness (curvaceousness) -ed+ness (datedness) -ive+ness (declarativeness) -less+ness (depthlessness) -ing+ness (reassuringness)

Acad

711

148

356

298

393

416

25

3.7

25

11

11

6

18.8

1.8

12.4

6.4

17.7

25.4

3.5

0.3

2.5

1.6

1.7

0.9

43

10

27

19

51

104

15

4

10

5

13

21

12

8

27

18

63

109

29

7

13

17

28

32

5

3

9

10

19

20

Table 1. BNC frequencies of -ness word tokens per 1 million tokens of text Our findings above clearly show that a more fine-grained analysis of ness nominalizations accounts for even more register variation in additional detail. Although fiction has the highest frequency of -ness words overall, the only two morphologically defined templates for which ness words are most frequent in fiction are simplex root+ness and ish+ness. In other cases, -ness is as representative of fiction as it is of some other registers (-y+ness in Pop, -ful+ness, -ed+ness, -less+ness in Acad), although the pattern -y+ness seems to be an important distinguishing feature between fiction and academic prose. In yet other cases, fiction is outnumbered substantially by frequency counts in academic texts (-ous+ness, -ive+ness and -ing+ness). The highest total count of -ness derivatives in fiction is predominantly attributable to items conforming to the morphological template simplex root+ness (711 items out of the total count of 896). The occurrence of -ish+ness words, although highest in fiction (3.5 items), is here negligible. Otherwise, all other instantiations of -ness may be predicted to be equally or less frequent

78

Wojciech Guz

in fiction than in any other of the six registers (see Table 1).11 In view of these facts, our perception of the distribution of -ness across registers needs revision in order to allow for these newly-found patterns. On a more global scale, claims about a universal preference of an affix for any one register need to be thoroughly investigated or else they may be rejected as inadequate and superficial. Admittedly, this inconsistent repartition of -ness is not entirely haphazard. It seems to be the case that register preferences of particular base form types predetermine the varied distribution of -ness. For example, simplex nouns are preferred in less formal registers such as spoken language and fiction (Biber et al. 1999: 322–323) and, presumably, the same also holds true of derived nouns in the sense that, in those registers, simplex adjectival roots are the preferred bases for -ness suffixation. In academic texts, the simplex adjective is still the most frequent base, but the disproportion between the most common and the second and third most common bases (simplex+ness vs. -ive+ness and ous+ness) is much lower than that in fiction (see Table 1). With derived adjectives acting as base forms, the adjectival suffix itself and the type of adjective it derives may be an important factor. The repartition of suffixes such as -ous and -ive, which represent learned vocabulary of Latinate origin and therefore are more frequent in academic writing (Biber et al. 1999: 532), may be the very reason for the high number of words in -ous+ness and -ive+ness in academic texts. In these two cases the learnedness of the adjectives has the greatest impact on the general flavour of the nominalizations that assigns them to a particular register. The suffix -ness, on the other hand, seems to be reduced to a register-neutral nominalizing marker, stripped of its alleged preference for fiction. In fact, the affix combinations -ive+ness and -ous+ness display an even stronger preference for academic texts relative to adjectives in -ive and -ous respectively – compare the relevant ratios in Table 2:

11

Nominalizations in -ari+ness, -al+ness and -ate+ness have not been considered here as separate suffix combinations although they are included in the total -ness counts (Figure 2). Still, we have noted that all three types are most frequent in academic prose.

The Distribution of the Suffix –ness

Fict Acad Fict:Acad ratio

-ive 497 2155 1:4.3

-iveness 12 110 1:9.1

-ous 968 1125 1:1.6

79

-ousness 43 105 1:2.4

Table 2. Normalized token frequencies of -ive/iveness and -ous/ousness in Fict and Acad Similarly, the suffixes -ish and -y may safely be regarded as more characteristic of less formal registers and thus explain the higher frequencies of words in -ish+ness and -y+ness in those registers (see Table 1). Thus the register distribution of -ness can be neatly correlated with the (distinct) distributional patterns of its base forms. However, this correlation does not work without exceptions. Despite the fact that suffixes -less and -ful are by a narrow margin most common in fiction, as shown by Biber et al. (1999), derivatives in -less+ness and ful+ness are found in this study to be somewhat more frequent in academic texts (see Table 1). Both stem-final suffixes (-less and -ful) and the suffix -ness are more characteristic of fiction and thus, theoretically speaking, their combination would be expected to be an even stronger force driving words in -less+ness and -ful+ness towards fiction. This, however, is not the case. Similarly, the affix combinations -ed+ness and ing+ness are more common in academic discourse. One plausible explanation here is that these are cases of several interacting factors: one is the nature of the base form imposing a certain patterning on the part of the derivative (e.g. bases in -less and -ful push -less+ness and -ful+ness words towards fiction); another is the fact that -ness is more common in fiction; yet another is that nominalizations on the whole tend to gravitate towards more formal registers.

5. Conclusion In summary, this paper has analysed the register distribution of -ness nominalizations and shown explicitly that its values of frequency fluctuate across the registers, peaking in the two polarized extremes of fiction and academic prose. Investigating deeper into the morphological composition of these nominalizations, we have shown that particular structural types and suffix combinations contribute differently to the overall distributional pattern of -ness (cf. Figure 2 and Table 1). The suffix is most frequent in fiction clearly due to one particular structural type, i.e. simplex root+ness.

80

Wojciech Guz

Two other types which are also typically fictional are statistically insignificant in this distribution (-y+ness and -ish+ness). The remaining suffixal combinations, contrary to previous assertions, are not most frequent in fiction but, in fact, are most common in academic writing. The combinations -ive+ness and -ous+ness, which have been identified as particularly academic-like, are perhaps the most striking examples of the varied distribution of -ness words across registers. It has also been noted that the distribution of -ness in many cases overlaps with the distributional preferences of the stem-final suffix (e.g. -y, -ish, -ive, -ous), but in other cases it does not (-ful, -less). We have concluded that claims about the suffix’s register distribution, and possibly other suffixes too, must be revised to accommodate finer distinctions of combinations of the affix with distinct types of base forms.

References Baayen, Harald and Antoinette Renouf. 1996. “Chronicling the Times: productive lexical innovations in an English newspaper.” Language 72: 69–96. Biber, Douglas. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. —. 1994. “An analytical framework for register studies.” In Douglas Biber and Edward Finegan (eds.), Sociolinguistic Perspectives on Register, 31–56. New York and Oxford: Oxford University Press. —. 2004. “Conversation text types: a multi-dimensional analysis.” 7es Journées internationales d'Analyse statistique des Données Textuelles (JADT). Biber, Douglas, Susan Conrad and Randi Reppen. 1998. Corpus Linguistics. Investigating Language Structure and Use. Cambridge: Cambridge University Press. Biber Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, Edward Finegan. 1999. Longman Grammar of Spoken and Written English. Pearson Education Limited. Biber, Douglas and Susan Conrad. 2001. Register variation: a corpus approach. In Deborah Schiffrin, Deborah Tannen and Heidi Hamilton (eds.), The Handbook of Discourse Analysis, 175–196. Oxford: Blackwell Publishing. Booij, Geert. 2005. The Grammar of Words. Oxford: Oxford University Press. Chafe, Wallace L. and Jane Danielewicz. 1987. Properties of spoken and written language. In Rosalind Horowitz and Jay S. Samuels (eds.),

The Distribution of the Suffix –ness

81

Comprehending Oral and Written Language, 83–113. New York: Academic Press. Guz, Wojciech. 2010. Register Variation and Lexical Innovation. A Study of English Nominalizations. Lublin: Wydawnictwo KUL. Hay, Jennifer. 2003. Causes and Consequences of Word Structure. New York and London: Routledge. Hay, Jennifer and Harald Baayen. 2003. “Phonotactics, parsing and productivity.” Italian Journal of Linguistics 1: 99–130. Kastovsky, Dieter. 1986. “The problem of productivity in word formation.” Linguistics 24: 585–600. Lee, David. 2001. “Genres, registers, text types, domains and styles: clarifying the concepts and navigating a path through the BNC jungle.” Language Learning & Technology 5.3: 37–72. Lieber, Rochelle. 2005. English word-formation processes. In Pavol Štekauer and Rochelle Lieber (eds.), Handbook of Word-formation, 429–448. Dordrecht: Springer. Plag, Ingo. 2003. Word-Formation in English. Cambridge: Cambridge University Press. Plag, Ingo, Christiane Dalton-Puffer, C. and Harald Baayen. 1999. “Morphological productivity across speech and writing.” English Language and Linguistics 3.2: 209–228.

THE MORPHO-SEMANTICS OF EXOCENTRIC SYNTHETIC COMPOUND NOUNS IN POLISH1 JOANNA KOLBUSZ-BUDA

1. Introduction While English synthetic compounds are described as exclusively endocentric2, Polish reveals a large number of representatives of both kinds, i.e. endocentric and exocentric synthetic compound words. Still, neither class has received the deserved scholarly attention regarding their morphological and semantic make-up in the same way their English counterparts have. Moreover, relevant, but scant publications on compounds proper, obtainable within Polish linguistic literature, were, without exception, written decades ago (cf. Klemensiewiczówna 1951, Kurzowa 1976, Sambor 1976, Grzegorczykowa 1982). Though the mentioned publications constitute a substantial and irrefutable contribution to the field of Polish compounding, they provide strictly descriptive accounts of this morphological process, overlooking the fact that word formation is not merely a linear process of concatenation. They do not raise such substantial questions as the mechanism of argument assignment3, the morphosyntactic feature percolation, the hierarchical 1

This article is part of a larger research project on the morphosemantic contrastive study of nominal root and synthetic compound words in Polish and English, placing a special focus on hierarchical structure, morpho-syntactic feature assignment and argument realisation. 2 The author of this paper is sceptical of the claim that English synthetic compounds are exclusively endocentric. There are reasons to believe that certain exocentric root compounds should in fact receive the structural interpretation of synthetic compounds. This problem, however, falls beyond the scope of our paper and I shall not pursue it further. I discuss it at length in my research project on compounding in Polish and English. 3 The issue of argument realisation is, to a degree, raised by Sambor (1976) and LewiĔski (1993). Unlike LewiĔski, who does not circumvent the gray areas and

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

83

morphological structure of words, and last but not least, the crucial problem of morphology-syntax interface, having a direct bearing on the positioning of compounding on the map of grammar. In this paper I draw on existing literature on English and Polish compounding as well as on my own research proposals to investigate parallel synthetic nominal constructions found in Polish. I choose to focus on both the morphosemantic and morphosyntactic structure of exocentric deverbal4 compound nouns, placing special emphasis on their argument realization. I adopt a modified version of the Weak Lexicalist Hypothesis, in particular Lieber and Scalise’s (2007) Firewall Theory. I argue that the overall evidence clearly weighs in favour of the hierarchical left-branching structure of Polish exocentric synthetic nominal compounds, and that percolation of morphosyntactic features originates from the rightmost head constituent, i.e. the derivational affix. Synthetic constructions in English, i.e. cases of simultaneous compounding and derivation, e.g. truck driver, bookseller, have been widely studied and vigorously debated in linguistic theoretical universe. There is no doubt that the 1978 publication by Roeper and Siegel became a landmark in linguistic theory of compounding and derivation. This new perspective, followed up by Selkirk (1982), Lieber (1983, 1992), Fabb (1984), Sproat (1985), Di Sciullo and Williams (1987), among many others, seeks correspondence between the rules of phrase formation and word formation, recognizing that the latter does not merely consist in a linearly oriented process of concatenation. The contemporary treatments of synthetic compounding range from the lexical, representing both the Strong and Weak Lexicalism, (e.g. Di Sciullo and Williams 1987, Lieber 1983, 2004, 2005, 2009a, Lieber and Scalise 2007) to syntactic (e.g. Fabb 1984, Baker 1988, Lieber 1992, Hale and Keyser 1993, 2002, et alli). Under the lexical treatment derivation is immune to any rules of phrase formation but makes room for some ‘inherited’ syntactic properties, such as argument structure. Under the problematic cases of the semantics of Polish exocentric deverbal compound nouns, Sambor’s analysis of endo- and exocentric deverbal compound nouns is based on limited data. Both scholars, however, abstain from sketching the mechanism of argument assignment per se. Instead, they turn their attention to the semantic role interpretation of the most popular representatives of the Polish deverbal compound nouns. 4 By deverbal we mean, unlike in English, not only these compounds whose lexical heads are derived from verbs (cf. noworodek ‘a newborn’) but also those which have the verb as the head of their internal structure (cf. bawidamek ‘a ladies’ man’).

84

Joanna Kolbusz-Buda

radically syntactic approach word-formation is believed to be the resultant of syntactic operations and syntactic constraints, and, as such, it is positioned in the syntactic component of grammar. Paradoxically, the Lexicalist Theory makes room for two conflicting approaches towards morphology-syntax interface. The one that prevailed since the outset of generative morphology denied all syntactic access to the internal structure of words. The key principles that barred morphology from interacting with the syntactic component of grammar were, and sometimes still are, the so called Lexical Integrity Principle, preventing syntax from ‘looking into’ morphology, and the No Phrase Constraint, frustrating a reverse situation (Halle 1973, Lapointe 1980, Di Siullo and Williams 1987, Williams 2007, et al.). Under such a version of the Lexicalist Hypothesis, blocking possible morpho-syntactic interaction, the analysis of, for instance, root compounds, built on syntactic phrases, or deverbal synthetic compounds would prove highly problematic. In the light of cross-linguistic data amassed, Lieber and Scalise (2007) put forth a less rigid variant of the Lexicalist Hypothesis, the so called Firewall Theory of the Syntax-Morphology Interface (henceforth the Firewall Theory)5. As a foundation for our framework we wish to adopt Lieber and Scalise’s Firewall Theory, being a weaker incarnation of the Lexicalist model of grammar, whose main tenet holds that morphology and syntax do interact and that this interaction, though highly constrained, operates in both directions. On the basis of a rich body of linguistic data coming from English, Italian, Spanish, German, Dutch, Afrikaans and Japanese6 (for detailed discussion, see Lieber and Scalise 2007: 4–12), Lieber and Scalise (2007) subvert the aprioristic separation of morphological and syntactic domains. Although Lieber and Scalise (2007: 21) admit that Syntactic Merge is different than Morphological Merge in that Syntactic Merge produces phrases and sentences, and Morphological Merge produces words, they accentuate a point of contact between them. In effect, ‘[…] languages can allow word formation of certain sorts to Merge syntactic phrases. It is possible, as well, that sentences and phrases can be ‘downgraded’ to words as part of a process of grammaticalization’. 5

The term proposed by Sato (2010: 402) to refer to Lieber and Scalise’s (2007) restatement of the Lexical Integrity Hypothesis. 6 Phrasal compounding in Germanic languages, Italian trasporto latte type construction, conjunction in English derivation and compounding, phrasal derivation in Italian, Kageyama’s (2001) Word Plus cases, prefixation in Spanish, among others.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

85

Lieber and Scalise (2007: 21) propose the so called Limited Access Principle which states that: (1)

Morphological Merge can select on a language specific basis to merge with a phrasal/sentential unit. There is no Syntactic Merge below the word level.

They further define Morphological Merge as follows: (2)

Let there be items Į, ȕ, such that Į is a base and ȕ a base or affix. MM takes Į, ȕ (order irrelevant) and yields structures of the form < Į, ȕ>Ȗ a. where Ȗ is an X0, categorically equivalent to Į or ȕ, and b. Į or ȕ can be null.

What is meant by the above is that in the same way that Syntax feeds on Morphology, Morphology may feed on syntax. In effect, a syntactic construction may become an input to a word-formation process, e.g. [over the fence]PP gossip, [one size fits all]S mass production, [truckdrive7] er], etc. Lieber and Scalise’s (2007) Limited Access Principle may have serious consequences for the way we choose to represent the structure of many products of word formation in which the interaction of Morphology and Syntax seems incontrovertible.

2. The theory – the structure, morphosyntactic feature percolation and argument assignment 2.1 Morphological Structure Traditionally, synthetic compounds are defined as instances of compounding and derivation in that they contain an additional component, namely a derivational affix, attached to the final verbal stem of the formation. In English, the typical synthetic compounds are deverbal nominalizations in -er, e.g. truck driver, shoemaker and -ing, e.g. book selling, lake-swimming. Wiese (2008) explains there are three principal constituents of a synthetic compound, namely A, B and C. A and B represent stems, while 7

The internal root compound, found within the body of a synthetic compound, could be viewed as originating from a corresponding syntactic phrase (i.e. a VP), in this case ‘to drive trucks’.

86

Joanna Kolbusz-Buda

C stands for a derivational affix. A may itself be complex. The status of B is often a matter for debate, which in turn gives rise to contrasting interpretations, i.e. bracketing paradoxes. As Wiese (2008: 254) points out, ‘synthetic compounds are special because the construction [A B C] does not divide into a binary division [[AB] C] or, alternatively, [A [BC]] in an obvious way’. This results from the fact that there is mostly no freestanding complex part [AB], in the same way that the [BC] component is often nonexistent8. Therefore, the arrangement of constituents in synthetic compounds has often been open to debate, giving rise to competing interpretations, such as [[AB] C], proposed by Lieber (1983), Fabb (1984), Sproat (1985), Lieber and Scalise (2007) or [A [BC]], postulated by Di Sciullo and Williams (1987), Lieber (1992, 2004, 2005). Consequently, there are two cardinal theories of synthetic compounding—the Syntactic (also the Weak Lexicalist) approach9 and the Strong Lexicalist approach. While certain generative morphologists, such as Selkirk (1982), Booij (1988), Di Sciullo and Williams (1987), Lieber (1992, 2004, 2005), opt for a NN interpretation of English synthetic compounds, others, e.g. Lieber (1983), Fabb (1984), Sproat (1985), argue that it is the suffix—not the deverbal noun—that assumes the role of the head of the compound. The proponents of NN interpretation hold that the internal compound, such as truckdrive, found within the English synthetic formation is not formed productively in English, arguing that such an interpretation ought to be ruled out. Simultaneously, other generative theorists claim that the interpretation in question is in fact the most plausible as it accounts for the fact that “[…] an obligatory complement of a verb is satisfied in the first projection above the verb” (Lieber 1992: 85). This is because the left-hand stem typically coincides with the object of the verb in a corresponding verb phrase10. Despite substantial differences, the weak lexicalist and syntactic accounts of synthetic compounding11 are united through the structural 8

cf. *jadáodaü ‘to food-give’ (AB) or *dajnia ‘a place where something is given’ (BC) in the Polish jadá-o-daj-nia ‘an eating house (-o- being the intefix). 9 The Weak Lexicalist theory and Syntactic theory of synthetic compounding are, by no means, counterpart approaches. However, despite conflicting tenets, they both give rise to similar left-branching structures of synthetic compounds. 10 ‘[...] unless the verb does not allow an internal argument (for example, ocean swimming) or the internal argument is discharged syntactically (as in hand-making of baskets). In the latter case, the first stem often bears an adjunct relationship to the verb’ (Lieber 2005: 381). 11 ‘Under the lexical treatment the derived word is formed in the morphology but it can ‘inherit’ certain syntactic properties, such as argument structure realization.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

87

common denominator—the internal root compound—as they both propagate the [[AB] C] configuration. In effect, this morphological process begins with the creation of a verbal compound, such as to truckdrive, in which the non-head satisfies the internal argument of the head verb. Next, the compound is subject to derivational affixation. In effect, the word formed will exhibit two parallel structures, i.e. the morphosyntactic [N [V truckdrive] er] and the morphophonological [N [ N truck] [N driver]]. According to Lieber (2004: 48), the [A [BC]] variant is more plausible for two reasons. First, it is odd to attach a derivational affix ‘on the outside’ of a compound (although Lieber and Stekauer (2009) emphasise that in languages such as Czech, Russian or Slovak the derivational suffix will attach to the compound as a whole). Secondly, the [[AB] C] structure “[...] forces us to generate a sort of root compound [...] which in fact is highly unproductive [...]” (Lieber 2004: 48) as this type of root compound rarely functions as an independent word in English (cf. to hand weave – handwoven). It should, however, be noted that although it is possible to offer two different analyses of English synthetic compounds within generative morphology, the same cannot be done with an abundant number of Polish exocentric synthetic formations, built on nonexistent root compounds, in which the verb may either follow its complement/adjunct or precede it, e.g. noworodek ‘a newborn’ (AdvV + ek), bawidamek ‘a lady’s man’ (VN + ek). It seems that the two structures could only be captured by the [[AB] C] analysis12, with the post-compounding derivation being the most plausible sequence, as neither of the combinations of the right-hand root and the derivational suffix that would indisputably support the [A [BC]] configuration exist as an independent word in Polish, e.g. *rod-ek, *krąĪca, *patrz-nia in now-o-rod-ek ‘a newborn’, dom-o-krąĪ-ca ‘a hawker’, gwiazd-o-patrz-nia ‘an astronomical observatory’.

2.2 Percolation An upward movement of morphosyntactic features, such as syntactic category, person, number, gender, and case, from the head morpheme to the node dominating the head, known as percolation, has perennially lain Under the radically syntactic account word formation takes place in the syntax’ (Spencer 2005: 89). 11 The NN analysis may, however, be applicable to endocentric synthetic compounds in Polish, such as skórobicie ‘whipping lit. skin beating’. Nevertheless, the same formation may also receive the structural interpretation of [[AB] C], and an eventive reading, due to the semantics of the suffix -cie.

88

Joanna Kolbusz-Buda

at the heart of the theory of both compounding and derivation (cf. the proposals of Fabb 1984, Sproat 1985, Roeper 1987, Di Sciullo and Williams 1987, Lieber 1983, 1992, 2004, et alli). One of the first statements of feature percolation were probably formulated by Lieber (1980: 49–54), Williams (1981: 247) and Selkirk (1982: 76). Despite differences in technicalities, the scholars generally share the same perspective on the issue in question. However, neither of them specifies exactly which features undergo percolation and how it is attained. In our analysis of Polish synthetic compounding we choose to follow Lieber’s (1992) theory of morphosyntactic feature percolation. Lieber (1992: 88) argues that only morphosyntactic features, such as those marking “[...] agreement across categories [or] possible sequences of clauses [...]”, can percolate. However, percolation will not be allowed cross-categorially as “[...] there is no reason to believe that nouns derived from verbs, for example, are marked for verbal features like tense and aspect”. Lieber (1992: 92) views percolation of features as a frame, the so called categorial signature, which comes in two parts: (3)

I. Head Percolation Morphosyntactic features are passed from a head morpheme to the node dominating the head. Head Percolation propagates the categorial signature. II. Backup Percolation If the node dominating the head remains unmarked for a given feature after Head Percolation, then a value for that feature is percolated from an immediately dominated nonhead branch marked for that feature. Backup Percolation propagates only values for unmarked features and is strictly local.

Head Percolation and Backup Percolation are responsible “[...] for preventing features of one category from passing across categorial lines” to ensure that percolation remains non-cross-categorial (Lieber 1992: 92).

2.3 Semantic Argument Assignment As for inheritance of semantic arguments, Lieber (1992) considers the mechanism hard to define and abstains from explaining the notion. Simultaneously, she argues that one of the essential properties of argument inheritance is the ability to operate cross-categorially.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

89

Lieber (2004) (also Booij and Lieber 2004), on the other hand, designs a mechanism of co-indexation which ensures the proper assignment of arguments within derived and compounded words: (4)

In a configuration in which semantic skeletons are composed, coindex the highest nonhead argument with the highest13 (preferably unindexed) head argument. Indexing must be consistent with semantic conditions on the head argument, if any. (Lieber 2004: 83)14

Co-indexation allows Lieber (2004) to integrate the referential properties of the lexical components used to derive a new word, such as a base and an affix, into one referential unit, i.e. a new complex word. [This referential unit] determines how many arguments are eventually projected into the syntax. Co-indexation [binds] the arguments that come with different parts of a complex word to yield only those arguments that are syntactically active. (Lieber 2004: 45) Even though Lieber’s (2004) proposition is undeniably of substance, it still does not address a critical issue of how to arrive at a specific argument configuration of a synthetic compound. In other words, what is the mechanism responsible for the existence of different referents of the same compound form, manifested by a single, potentially polysemous, affix. For Lieber (2004) and Lieber and Booij (2004) the explanation lies in the polysemy of affixes. They argue that certain derivational affixes, such as the -er affix, may communicate several referents. However, polysemy per se does not explain the choice of one referent over the other in the process of compound formation. As I propose to treat the deverbal compound as a formation morphologically derived from a syntactic construction, I argue that its concrete argument structure is passed on to the created compound. As a result, the referent selected will exactly match the referent of the corresponding syntactic construction. Relying solely on the verb’s thematic role specification contained in the lexical entry could potentially result in multiple or misinterpretations 12

i.e. ‘[...] the highest argument is the argument of the outermost lexical function of the head’ (Lieber 2004: 50). 13 For a recapitulation of Lieber’s (2004) Co-indexation Principle see Lieber (2005) and Lieber (2009a).

90

Joanna Kolbusz-Buda

of the compound’s referent. For instance, if no reference to the underlying syntactic construction of wodo-o-czyszcz-acz ‘water cleanser’ were made, the referent of the compound could easily be interpreted as Agent by analogy to, for instance, spawacz ‘welder’ or wáamywacz ‘burglar’. Similarly, there are many compound forms being assigned more than one referent, e.g. the English lawn mower (Agent or Instrument), the Polish gasiĞwieczka ‘an acolyte who puts/an instrument used to put out candles in church’ (Agent or Instrument), etc. Despite the selfsame morphological make-up, semantically the compounds can only be viewed as separate vocabulary items, originating from two different syntactic structures. Thus, as an alternative to Lieber’s (2004) Co-indexation Principle, I propose what follows: (5)

One-to-One Mapping Principle The deverbal compound is morphologically derived from a syntactic construction. Thus, the unique semantic argument structure of the syntactic input will be mapped onto the relevant morphemes of a newly derived lexical unit.

What is meant by the above is that the two denotations (referents) of, for instance, the English synthetic compound noun lawn mower, the agentive and instrumental ones, could only be derived from two separate underlying syntactic input constructions, namely ‘the one who mows grass’ and ‘this what mows grass’. (6)

lawn mower (Instrument) lawn mower (Agent)

< this what mows grass < the one who mows grass

Thus, the principle of One-to-One Mapping of arguments ensures that the creation of a compound form possessing more than one designatum must always be carried out via two independent morphological processes and utilize separate input constructions. As a result, the unique semantic argument structure of the syntactic input will be mapped onto the relevant morphemes of a newly derived lexical unit.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

91

3. The morpho-semantics of exocentric synthetic compound nouns in Polish 3.1 General characteristics of Polish deverbal compound nouns Polish exocentric synthetic compounds are built on two lexical components (roots), and two formal exponents of composition (affixes). One of the formatives functions as a connecting morpheme (interfix) whose role is to bind the two roots phonologically. The morpheme carries no semantic content and is highly selective with respect to the left-hand root. The interfixes predominantly employed in Polish compounding are o- and -i-//-y- (less typically -u- and ø), the former being utilized postnominally and post-adverbially, the latter following verb roots: (7)

Noun wod-o-lot-ø ‘hydrofoil’ water fly bydá-o-bój-nia ‘abbatoir’ cattle slaughter Adverb sproĞn-o-mów-ca ‘a ribald speaker’ ribaldly speak Verb baw-i-dam-ek ‘lady’s man’ entertain lady mocz-y-morda-ø ‘drunkard’ dunk mug

The suffix, assuming the shape of a derivational affix or a zero affix15 (ø), added compound externally, modifies the word as a whole, determining its morphosyntactic features, such as syntactic category, gender, number and case. 14

The zero affix triggers paradigmatic conversion. In Polish linguistic literature the term is used to refer to an affixless process of word formation in which the difference between the base and the derived word boils down to the difference in their inflectional paradigms. In effect, the main formative assumes the shape of an inflectional paradigm of the derivative replacing the one of the underlying base (e.g. A > N biaáy ‘white’ > biel ‘white’, zdrowy ‘healthy’ > zdrowie ‘health’; V > N biegaü ‘to run’ > bieg ‘a run’, ciągnąü ‘to pull’ > ciąg ‘a pull’). In practice, the root is taken out of a particular paradigm and inserted into another paradigm. For a detailed account of paradigmatic conversion in Polish, see Waszakowa (1996).

Joanna Kolbusz-Buda

92

(8)

a derivational suffix wyrw-i-rącz-ka16 ‘T-bar lift’; -ka: singular, feminine, nominative dom-o-krąĪ-ca ‘hawker’; -ca: singular, masculine, nominative now-o-rod-ek ‘new-born’; -ek: singular, masculine, nominative ø affix – the marker of paradigmatic derivation (conversion) gol-i-broda-ø ‘barber’; -ø: singular, masculine, nominative piorun-o-chron-ø ‘lightning conductor’; -ø: singular, masculine, nominative bruk-o-táuk-ø ‘maunderer’; -ø: singular, masculine, nominative wod-o-ciąg-ø ‘water pipe’; -ø: singular, masculine, nominative

Paradigmatic conversion may involve a shift from one syntactic category to the other or can be effected within the same syntactic category. The verbs chorniü ‘protect’, táuc ‘rove’, ciągnąü ‘pull’ in (9), having undergone paradigmatic conversion, become nouns while the masculine compounds golibroda and liĪyáapa in (10) do not inherit their grammatical gender form the nominal feminine constituents broda and áapa. (9)

15

N V piorun-o-chron-ø lightning protect bruk-o-táuk-ø paving maunder wod-o-ciąg-ø water pull

‘lightning conductor’, piorunochronø = N ‘maunderer’, brukotáukø = N ‘water pipe’, wodociągø = N

Despite the existence of such diminutive word-forms as rączka ‘a little arm’, piórko ‘a little feather’ or gáówka ‘a little head’, it would be highly unconvincing to interpret the derivational suffix as an integral part of the right-hand constituent of such compounds as wyrwirączka, gryzipiórek and áamigáówka. This would give the compound a peculiar reading and falsify its morphological structure. This becomes readily apparent in the paraphrase which elucidates that the formative ought to be viewed as part of the entire compound, thus its head: áamigáówka ‘a puzzle, lit. this over what someone breaks their head (not ‘little head’)’. Nonetheless, Grzegorczykowa and Puzynina (1999) accept both motivations, regarding the one which involves the diminutive as the accompanying motivation.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

(10)

93

V N gol-i-broda-ø (masc.’) ‘barber’; beard (fem.) shave beard liĪ-y-áapa-ø (masc.) ‘toady’; paw (fem.) lick paw

While the combinability of the English synthetic nominal compounds is mostly limited to the sequence of a noun, verb and an affix, in Polish the picture is much more diversified. According to Kurzowa (1976), there are six types of Polish exocentric deverbal compound nouns: (11)

Type 1. Noun + interfix + Verb + suffix (jadá-o-daj-nia ‘eatinghouse’)17 Type 2. Noun + interfix + Verb + ø (piorun-o-chron-ø ‘lightning conductor’)18 Type 3. Adverb + interfix + Verb + suffix (now-o-rod-ek ‘newborn’)19 Type 4. Adverb + interfix + Verb + ø (dalek-o-widz-ø ‘hyperope’)20 Type 5. Verb + interfix + Noun + suffix (gryz-i-piór-ek ‘penpusher’)21 Type 6. Verb + interfix + Noun + ø (mocz-y-morda-ø ‘drunkard’)22

Structurally, the taxonomy proposed by Kurzowa could be reduced to two types, namely type 1 (types 1–4) and type 2 (types 5–6). For reasons of conciseness, we wish to propose such a typology in our article.

16

also chlebodawca ‘employer’, ludoĪerca ‘cannibal’, ojcobójca ‘parricide’, bydáobójnia ‘a slaughter house’, miodosytnia ‘a mead brewery’, gwiazdopatrznia ‘astronomical observatory’, etc. 17 also falochron ‘breakwater’, gazociąg ‘gas pipeline’, kątomierz ‘protractor’, dzieworód ‘parthenogenetic offspring’, dziworód ‘freak’, drzeworyt ‘xylograph’, garnkotáuk ‘bad cook’, listonosz ‘postman’, etc. 18 also pierworódka ‘primipara’, wieloródka ‘multipara’, maáojadek ‘fussy eater’, marnotrawca ‘wastrel’, skrytobójca ‘assassin’, etc. 19 also czarnowidz ‘defeatist’, brudnopis ‘rough copy’, czystopis ‘fair copy’, dalekopis ‘teleprinter’, etc. 20 also áamigáówka ‘puzzle’, bawidamek ‘lady’s man’, etc. 21 also chwalipiĊta ‘bragger’, golibroda ‘barber’, liczykrupa ‘stingy person’, dusigrosz ‘penny-pincher’, káapouch ‘lop-eared person/animal’, pĊdziwiatr ‘fribble, butterfly’, wáóczykij ‘roamer’, etc.

94

Joanna Kolbusz-Buda

3.2 The structure, morpho-syntactic feature percolation and argument assignment of type 1 exocentric synthetic compound nouns The first four types of exocentric deverbal compounds identified by Kurzowa (1976) are combinations of nominal or adverbial roots followed by verb roots. The lexical constituents are bound by means of the interfix o-23. The other formal exponent of composition, i.e. the derivational suffix or the zero affix, constitutes the morpho-syntactic head of the compound. By and large, the left-hand nominal root fulfils the function of the direct object of the verb in a corresponding VP, e.g. lodoáamacz ‘ice-breaker’ – break ice, wodociąg ‘water pipe’ – pull water’, bydáobójnia ‘slaughter house’ – butcher cattle. However, its grammatical function is not restricted solely to a verb-object relation but extends to include adverbials of place, e.g. domokrąĪca ‘a hawker, lit. the one who travels to people’s homes’ (Kurzowa 1976, Grzegorczykowa, Puzynina 1999), and time, e.g. zimorodek ‘kingfisher, lit. the one that is born in winter’. As for adverbial roots, they will naturally carry out the function of adjuncts. As far as the process of percolation is concerned, the formal exponent of composition in the shape of the head suffix (also ø) is the sole source of four morphosyntactic features, namely gender, number and case. The syntactic category assigned by all the head affixes of Polish deverbal compounds discussed in this paper will naturally be N. The spectrum of semantic referents of type 1 compounds is wide and ranges from Agent and Instrument to Location, Result and even Expierencer, e.g. garnkotáukø ‘a bad cook’, gazociągø ‘gas pipeline’, gwiazdopatrznia ‘astronomical observatory’, brudnopisø ‘rough copy’, noworodek ‘newborn’. Interestingly, the list of internal arguments seems even richer, e.g. dobroczyĔca ‘benefactor’ – Result, wodooczyszczacz ‘water cleanser’ – Theme, groszorób ‘capitalist’ – Goal, piorunochron ‘lightning conductor’ – Force, brukotáuk ‘maunderer’ – Location, pierworódka ‘primipara’ – Measure/Manner, etc. 22

It is not clear, however, if the interfix -o- is employed across the entire type of AvdV compounds. In those cases where the underlying form of the adverb terminates in -o (e.g. nowo ‘newly’, maáo ‘little’, Īywo ‘alive’), we may speak of the lack of the connecting morpheme. In other cases, where the adverb component terminates in -e (e.g. sproĞnie ‘ribaldly, pieknie ‘beautifully’), it will most probably admit the interfix: sproĞnomówca ‘a ribald speaker’, piĊknobrzmiący ‘beautifully sounding’. Puzynina (1978) accepts the possibility of a morpheme merger (the interfix -o- and the derivational suffix -o) or a zero interfix as a contextual variant of the -o- interfix.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

95

Again, for reasons of conciseness, I choose to be selective with respect to the examples quoted, leaving aside borderline cases of a more challenging argument structure, so as to avoid a 'book length' treatment of the material. My structural representations of the selected deverbal compound words of both type 1 and type 2 aim at maximal comprehensiveness. Within a single representation I wish to comprise both morphosyntactic and morphosemantic make-up of the compound, namely the morphological structure, percolation and argument structure. i and ii stand for co-indexation of arguments while the arrow represents the source and direction of morphosyntactic feature percolation. For the most part, our percolation representation follows Lieber’s (1992) Categorial Signature. However, under our interpretation, the suffix does not assign the morphosyntactic feature of ‘person’ in both English and Polish. The verb root is given in its infinitive form, the noun root in the nominative singular. The proposed One-to-One Mapping Principle will ensure that the argument structure of every synthetic compound corresponds to one and only one syntactic construction. As a result, in the event of the same compound form being assigned more than one referent, each referent will be derived from a separate syntactic input. This in turn will guarantee that the unique semantic argument structure of a particular syntactic construction will be mapped onto the relevant morphemes of a newly derived compound word. Consider the following morpho-semantic representations of snopowiązaáka, rĊkopis, wieloródka and dalekowidz below. (12)

snop-o-wiąz-aáka ‘sheaf-binder’

this (Instrument) what binds (the stems) into sheaves (Result)

96

(13)

Joanna Kolbusz-Buda

rĊk-o-pis-ø ‘manuscript’

this (Result) what has been written by hand (Instrument) (14)

wiel-o-ród-ka ‘multipara’

the one (Agent) who has given birth many times (Manner) (15)

dalek-o-widz-ø ‘hyperope’

the one (Agent) who sees far (Manner)

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

97

3.3 The structure, morpho-syntactic feature percolation and argument assignment of type 1 exocentric synthetic compound nouns In contrast, the final two types (type 2 in our typology) exhibit a reverse order of roots in that the verb root precedes the nominal one. More importantly, the right-hand root will invariably be realized by a noun. The interfix employed in to combine the verb and the noun is no longer -o-, as in post-verbal positions the connecting morpheme assumes the shape of -i//-y-. The derivational suffix, alternating with the zero affix, constitutes the head of the entire formation and becomes the source of morphosyntactic feature percolation. The discussed compound type falls into two categories of gender, i.e. masculine and feminine. Due to their somewhat archaic and highly expressive colouring, type 2 compounds are an utter rarity in Polish non-artistic texts, simultaneously being abundant in artistic prose and drama. Consider the following examples: (16)

wyrwirączka ‘T-bar lift, lit. this what pulls out one’s arm’ bawidamek ‘lady’s man, lit. the one who entertains ladies’ gryzipiórek ‘pen pusher, lit. the one who bites a pen’ chwalipiĊta ‘braggart, lit. the one who brags his heel’ pĊdziwiatr ‘butterfly, lit. the one who runs like the wind’ wáóczykij ‘rolling stone, lit. the one that drags a stick’

According to Sambor (1976), the type in question is completely unproductive synchronically. First and foremost, it is limited to particular literary genres. Secondly, it has an almost palpable archaic air to it, which in turn makes such compounds an ‘easy target’ for artistic literary texts. The referents of verb-initial compounds are predominantly Agents and Instruments, e.g. dáawidudaø ‘organist’ – Agent, gasiĞwieczkaø ‘tool used to put out candles in church’ – Instrument. However, the more complex the structure of the syntactic input, the more idiomatic the compound and the more diversified its argument structure. In consequence, the function of the right-hand noun will not be confined to that of the object of the verb in a corresponding VP, e.g. wydrwigroszø ‘wangler, lit. the one who wangles a penny’ – Goal, obieĪyĞwiatø ‘globetrotter, lit. the one who traverses the world’ – Location, burczymuchaø ‘grumbler, lit. the one who buzzes like a fly’ – Comparison, pĊdziwiatrø ‘butterfly, lit. the one who runs like the wind’– Comparison, etc.

98

(17)

Joanna Kolbusz-Buda

wyrw-i-rącz-ka ‘T-bar lift’

this (Instrument) what pulls out the arm (Theme) (18)

baw-i-dam-ek ‘lady’s man’

the one (Agent) who entertains ladies (Theme) (19)

gol-i-broda-ø ‘barber’

the one (Agent) who shaves beards (Theme)

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

(20)

99

czyĞc-i-but-ø ‘shoeblack’

the one (Agent) shines shoes (Theme) Notice that paradigmatic conversion does not always trigger an obligatory paradigm shift of the derivative with respect to the right-hand root, e.g. golibroda (masc.) vs. broda ‘beard’ (fem.), but czyĞcibut (masc.) vs. but ‘shoe’ (masc.).

4. Conclusions The purpose of this paper has been to provide a lexicalist account of Polish exocentric synthetic compound nouns. It has been accentuated that the most plausible morphological structure that could be proposed for the discussed compound class is the left-branching structure, with the derivational affix serving the function of the morphosyntactic head. We have argued that deverbal compounds in Polish are morphologically derived from corresponding syntactic constructions. As a result, the unique semantic argument structure of the syntactic input will be reflected in the semantics of a newly derived compound due to one-to-one mapping of the arguments onto the relevant morphemes of this compound word.

References Baker, Mark 1988. Incorporation: A Theory of Grammatical Function Changing. Chicago: University of Chicago Press. Booij, Geert. 1988. The Relation between inheritance and argument linking: Deverbal nouns in Dutch. In Martin Evaraert, Arnold Evers, Riny Huybregts, and Mieke Trommelen (eds.), Morphology and modularity: to Honour Henk Schultink, 57–73. Dordrecht: Foris.

100

Joanna Kolbusz-Buda

Booij, Geert and Rochelle Lieber. 2004. “On the paradigmatic nature of affixal semantics in English and Dutch.” Linguistics 42: 327–357. Di Sciullo, Anna M. and Edwin Williams. 1987. On the definition of word, Cambridge, Mass: MIT Press. Fabb, Nigel 1984. Syntactic Affixation, PhD diss. MIT. Grzegorczykowa, Renata. 1982. Zarys sáowotwórstwa polskiego. Sáowotwórstwo Opisowe [An outline of Polish Word-formation. Descriptive Word-formation]. Warszawa: PWN. Grzegorczykowa, Renata and Jadwiga Puzynina. 1999. Rzeczownik [The Noun]. In Renata Grzegorczykowa, Roman Laskowski and Henryk Wróbel (eds.), Gramatyka wspóáczesnego jĊzyka polskiego. Morfologia [A grammar of contemporary Polish. Morphology.], 389-464. Warszawa: PWN. Hale, Ken and Samuel Jay Keyser. 1993. On argument structure and the lexical expression of syntactic relations. In Ken Hale and Samuel Jay Keyser (eds.), The View from Building 20, 53–109. Cambridge, Mass: MIT Press. —. 2002. Prolegomenon to a Theory or Argument Structure. Cambridge, Mass: MIT Press. Halle, Morris 1973. “Prolegomena to a Theory of Word Formation.” Linguistic Inquiry 4: 3–16. Kageyama, Taro. 2001. Word Plus: the intersection of words and phrases, In Jeroen Van de Weijer and Tetsuo Nishihara (eds.), Studies in Generative Grammar 51, 245–276. Berlin: Mouton de Gruyter. Klemensiewiczówna, Irena. 1951. Wyrazy záoĪone nowszej polszczyzny kulturalnej. Próba systematyki [Compound words of educated social stratum. An attempt at taxonomy]. Kraków: Studium SáowiaĔskie Uniwersytetu JagielloĔskiego. Kurzowa, Zofia. 1976. ZáoĪenia imienne we wspóáczesnym jĊzyku polskim [Nominal compounds of contemporary Polish]. Warszawa: PWN. Lapointe, Steven. 1980. A theory of grammatical agreement, Doctoral Dissertation, University of Amherst. LewiĔski, Piotr. 1993. “Próba klasyfikacji semantycznej compositów polskich, których drugi czáon nie wystĊpuje jako samodzielny wyraz [An attempt at semantic classification of Polish compounds whose second stem does not function as a free-standing word].” Rozprawy Komisji JĊzykowej WTN 19: 57–87. Lieber, Rochelle. 1983. “Argument linking and compounds in English.” Linguistic Inquiry 14: 251–86. —. 1992. Deconstructing Morphology: Word Formation in Syntactic Theory, University of Chicago Press.

The Morpho-Semantics of Exocentric Synthetic Compound Nouns

101

—. 2004. Morphology and Lexical Semantics. New York: Cambridge University Press. —. 2005. English word-formation processes. In Rochelle Lieber and Pavol Stekauer (eds.), Handbook of Word-Formation, Studies in Natural Language and linguistic theory 64, 375-422. Dordrecht: Springer. Lieber, Rochelle and Sergio Scalise. 2007. “The lexical integrity hypothesis in a new theoretical universe.” In Geert Booij, Bernard Fradin, Angela Ralli and Sergio Scalise (eds.), On-line Proceedings of the Fifth Mediterranean Morphology Meeting, http://mmm.lingue.unibo.it. DOA: 20 Jul. 2007 (also published in Lingue e linguaggio 5: 7-32). Lieber, Rochelle and Pavol Stekauer. 2009. Introduction: status and definition of compounding. In Rochelle Lieber and Pavol Stekauer (eds.), The Oxford Handbook of Compounding, 3–18. New York: Oxford University Press. Lieber, Rochelle. 2009a. A lexical semantic approach to compounding. In Rochelle Lieber and Pavol Stekauer (eds.), The Oxford Handbook of Compounding, 78–104. New York: Oxford University Press. Puzynina, Jadwiga. 1978. “Recenzja: ZáoĪenia imienne we wspóáczesnym jĊzyku polskim [A review of Zofia Kurzowa’s Nominal compounds of contemporary Polish].” JĊzyk Polski 58: 219–225. Roeper, Thomas. 1987. “Implicit arguments and the head-complement relation.” Linguistic Inquiry 18, 267–310. Roeper, Thomas and Muffy Siegel. 1978. “A Lexical Transformation for Verbal Compounds.” Linguistic Inquiry 9: 199–260. Sambor, Jadwiga. 1976. Kompozycje rzeczownikowe dwunominalne i nominalno-werbalne w tekstach wspóáczesnego jĊzyka polskiego [NN and VN/NV compounds in corpora of contemporary Polish]. In Maria Renata Mayenowa (ed.), Semantyka tekstu i jĊzyka, 239–256. Wrocáaw-Warszawa-Kraków-GdaĔsk: Ossolineum. Sato, Yosuke. 2010. “Complex phrase structures within morphological words: evidence form English and Indonesian.” Lingua 120: 379–407. Selkirk, Elisabeth. 1982. The syntax of words. Cambridge, Mass: MIT Press. Spencer, Andrew. 2005. Word-formation and syntax. In Rochelle Lieber and Pavol Stekauer (eds.), Handbook of Word-Formation, Studies in Natural Language and linguistic theory 64, 73–97. Dordrecht: Springer. Sproat, Richard 1985. On Deriving the Lexicon, PhD diss. MIT. Waszakowa, Krystyna. 1996. Sáowotwórstwo wspóáczesnego jĊzyka polskiego. Rzeczowniki z formantami paradygmatycznymi [Word

102

Joanna Kolbusz-Buda

formation of contemporary Polish. Nouns with paradigmatic formatives]. Warszawa: Wydawnictwo Uniwersytetu Warszawskiego. Wiese, Richard. 2008. “A two-level approach to morphological structure.” Journal of Germanic Linguistics 20: 243–274. Williams, Edwin. 1981. “On the notions lexically related and head of a word.” Linguistic Inquiry 12: 234–274. —. 2007. Dumping lexicalism. In Gillian Ramchand and Charles Reiss (eds.), The Oxford Handbook of Linguistic Interfaces, 353–381. New York: Oxford University Press.

PART THREE SYNTAX

SECONDARY PATTERNS OF USE WITH LOCATUM VERBS TOMASZ FOJT

1. Introduction The study examines grammatical patterning of locatum verbs used with metaphorical semantic values. It is observed that certain metaphorical uses of locatum verbs are distributionally restricted to particular grammatical patterning. Distinctive grammatical patterns are shown to be correlated disjunctively with either metaphorical or literal semantic values of the verbs, with the result that formal differences come to mark semantic differentiation into figurative and literal verb senses. To demonstrate this, the study undertakes an analysis of the grammatical patterning associated with literal and metaphorical uses of the verb pepper. The relevant aspects of the argument structure of the verb are discussed, together with the semantic properties of the participants interacting in the relationships designated by the verb. The description is completed with an investigation of the relevant collocation patterns. The differences observed in the semantic specifications of the participants of the action designated by pepper and the differences in the surface expression of the locatum argument substantiate the treatment of metaphoricity as a graded phenomenon. The results of the analysis are then extended to the class of locatum verbs, under which pepper is subsumed. The paper is a corpus-assisted study, carried out with the use of the XML Edition of the British National Corpus and the Collins WordBanks Online corpus service. Unless otherwise indicated, all the linguistic samples used in the study are extracted from these corpora.

2. Basic grammatical patterning of the verb pepper The verbal forms of the lexeme pepper pattern in two basic ways shown in examples (1–2) below.

Secondary Patterns of Use with Locatum Verbs

(1)

(2)

105

a. Across River Street and down two houses Nobel Aldrich peppers his macaroni and cheese before he tastes it. (WordBanks) b. The world ‘hope’ peppered his text. (BNC) He peppered his designs with slogans. (BNC)

The verbs in the patterns are used transitively. The pattern NP V(pepper) NP exemplified in (1) admits of both literal and metaphorical uses of the verb pepper. The verb in (1a) designates the activity of seasoning the food with pepper. In (1b) the verb designates a state rather than an action and its meaning pertains to the repeated occurrence of the word hope in a text. The pattern in (2) differs from the above examples in that it includes a prepositional phrase headed by the preposition with. There are substantial grounds for setting the pattern NP V(pepper) NP PP(with) apart from its counterpart without the prepositional phrase. Firstly, barring the exceptions stipulated in section 4, the verb in this construction is used only in metaphorical senses. Secondly, the prepositional phrase does not function as an optional adjunct, but is an obligatory constituent of the construction, as follows from a number of tests discussed in section 3. As the verb in the examples discussed here is used transitively, all the patterns admit of passive alternation, demonstrated in sample (3) (3)

The document was peppered with phrases such as ‘in good faith.’ (WordBanks)

In fact, passive uses of the verb pepper prevail significantly over active ones. A random sample drawn from the BNC shows a 66% to 34% preference for the passive voice in the verbal uses of pepper. Similar values (68% passive to 32% active) were calculated on the basis of a sample extracted from the WordBanks corpus. More significantly, given the construction under examination, the pattern with the prepositional phrase shows an even more marked preference for passive uses, amounting to 83% in the samples from both the corpora used.

3. Argument status of the with-PP This section examines the properties of the with-PP in the NP V(pepper) NP PP(with) construction with a view to establishing its status as an adjunct or argument. The diagnostics employed pertain to the notions of syntactic and semantic obligatoriness of the with-PP, its dependence on the head

106

Tomasz Fojt

verb for interpretation and its distributional properties with regard to the range of heads it co-occurs with. Typically, prepositional phrases headed by the preposition with pattern as adjunts, i.e., a category subsuming more peripheral and optional constituents of constructions (Lyons 1971: 334). This is certainly true in the case of instrumental adjuncts headed by with as in (4) below. (4)

The once-dashing rogue, who now hobbles with a stick and is barely able to talk, is in the prison’s medical unit. (WordBanks)

The extranuclear status of the with-PP in (4) is evident in that its removal does not affect the structural integrity of the sentence, but merely dispenses with some auxiliary information. The same cannot be said about the prepositional phrases in (2) or (3) whose removal would render the sentences semantically anomalous. There is, however, a perennial problem connected with drawing the distinction between adjuncts and arguments with reference to obligatoriness. It has often been observed that there exist syntactically unexpressed arguments1. In the light of this observation Koenig et al. (2002: 225) dismiss syntactic expression as a viable diagnostic for determining argumenthood. They propose instead two complementary criteria that jointly provide the basis for the distinction into arguments and adjunct. The first is semantic obligatoriness of participant information whereby only the mandatory participants of the event designated by the verb can be assigned the status of semantic arguments. Participant information is understood as obligatory if “it is entailed to hold of the class of situations denoted by a word” (Koenig et al. 2002: 226). This criterion is augmented by the requirement of the (relative) specificity of the participant information to the relevant verb, an issue which is addressed in a strictly regimented manner by Schütze (1995)2. Schütze (1995) raises the issue of argument status in connection with his analysis of the phenomenon of the prepositional phrase attachment in sequences of the form VP NP PP. Undertaken to provide theoretical grounds for the empirical claim that human parsing decisions are guided by the so-called argument preference strategy, Schütze’s analysis rests

1 See, for instance, Mauner et al. (2002: 388–389) for the discussion of the syntactically unexpressed, implicit agent argument in passive clauses with sold. 2 The fact that argument roles are only relatively specific to small classes of verbs paves the way for the conception of argumenthood as a graded property, a view embraced in this study.

Secondary Patterns of Use with Locatum Verbs

107

crucially on the distinction between arguments and modifiers as exemplified below: (5)

a. The spy saw the cop with the telescope. b. The spy saw the cop with the revolver. (Schütze 1995: 95)

The prepositional phrase in (5a) clearly modifies the act of seeing and by virtue of its relation to the verb it represents a case of verb-attachment, whereas in (5b) the PP is said to be NP-attached as it modifies the NP. Irrespective of the details of attachment, Schütze shows with a number of tests that the PP in (5a) manifests a number of argument properties, not inherent in the kind of PP represented by example (5b). The distinctions will prove useful in the analysis of the with-PP used as a constituent of the construction with the verb pepper. It seems opportune to approach the semantically-based distinction into arguments and adjuncts from two perspectives: one looking into the way in which a prepositional phrase completes the meaning of the relationship designated by the head verb, and another which determines how the prepositional phrase itself receives its semantic interpretation. Accordingly, Schütze (1995) breaks up his claims concerning the argument status of prepositional phrases as follows: An argument fills a role in the relation described by its associated head, whose presence may be implied by the head. In contrast, a modifier predicates a separate property of its associated head or phrase. (Schütze 1995: 100) A phrase P is an argument of a head H if the semantic contribution of P to the meaning of a sentence in which P is associated with H depends on the particular identity of H. Conversely, P is a modifier if its semantic contribution is relatively constant across a range of sentences in which it combines with different heads. (Schütze 1995: 100)

Taken together, the two claims elaborate the details of the generalization posited by Pollard and Sag (1987) concerning the differences in distributional properties of the constituents whose status varies with respect to their centrality or peripherality: In general, a given adjunct can co-occur with a relatively broad range of heads while seeming to make a more-or-less uniform contribution to semantic content across that range. A given optional complement, by contrast, is typically limited in its distribution to co-occurrence with a small (and often semantically restricted) class of heads (possibly even a

108

Tomasz Fojt single item); in addition, the semantic contribution of the complement is idiosyncratically dependent upon the head. (Pollard and Sag 1987: 136)

Taken separately, the two formulations by Schütze operationalise the notion of the PP argumenthood and make it possible to apply a number of tests that can be used as diagnostics for classifying PPs. The examples in (6) demonstrate the adjunct/argument distinction with regard to the role of the with-PP in completing the meaning of the head. (6)

a. Police have sealed the area off and are searching the countryside with tracker dogs and a helicopter. (WordBanks) b. He peppered his designs with slogans. (BNC)

The clausal head in (6a) does not imply the presence of an instrumental with-PP. The processual relationship designated by the verb search does not require a participant instantiating the role of the instrument. As such, the phrase with tracker dogs and a helicopter predicates an independent property of the process, one that does not follow from the requirements of the head. In (6b), on the other hand, the metaphorical meaning of pepper relies crucially on the presence of the complementation. To construe the process designated by the verb pepper as distinct from the process of applying crushed condiment, it is necessary to recruit a participant, expressed by the oblique phrase, that adequately fills in the role in the relationship designated by the verb. The second postulate of Schütze’s bipartite claim pertains to the constancy of the semantic contribution furnished by the prepositional phrase. A phrase should be appropriately ascribed the status of an argument if it depends for its semantic interpretation on the particular identity of the head it is associated with. Conversely, if its semantic value remains relatively constant in combination with a variety of verbal heads, a phrase is likely to serve as an adjunct. Testing this hypothesis with regard to phrases headed by the preposition with may be problematic because of the notorious heterogeneity of the uses of with (Goldberg 2002: 337–338, 2006: 35). Among the relationships designated by the preposition with Quirk (1985: 684, 698–705) counts accompaniment, instrumentality, means, manner, possession, cause and pervasiveness. Of these relationships, instrumentality is perhaps the most characteristic of prepositional phrases headed by with (Dirven 1993: 89–91). A cursory look at the samples instantiating the instrumental use of with-PP demonstrates the range of verbal heads co-occurring with the prepositional phrase in question.

Secondary Patterns of Use with Locatum Verbs

(7)

109

a. The cupboard doors had all been deeply hacked and scored with a knife. (BNC) b. She just wiped it all off with a tissue. (WordBanks) c. He sat musing and playing with the beer mat. (WordBanks) d. If a corn was offensive it could be tackled with a razor blade. (BNC) e. The nature of the problem can be illustrated with an example. (BNC) f. Adrina replied with a nod. (WordBanks) g. If I don’t get home soon she’ll be after me with a carving knife. (WordBanks)

The verbs in (7a–g) can be at best delimited as expressing actions and activities. Yet, the uses of the prepositional phrase headed by with can be incontrovertibly attributed instrumental meanings in all the cases. The semantic value of the with-PP remains constant in combination with a range of semantically disparate verbal heads. Hence, the instrumental meaning of the with-PPs is relatively independent of the meanings of the verbs, which points to their status as adjunts. The meaning of the with-PP in combination with the verb pepper as exemplified in samples (8a–d) does not express instrumentality. Typically, the referent of the instrument participant is an inanimate thing or body part used volitionally by the agent to implement an action or process (Longacre 1996: 159–161). The role of the instrument as an immediate cause of an event is specified by Langacker (2000: 29, 2008: 355–356) with reference to the archetypal conception of the action chain whereby the instrument constitutes an intermediary in the transfer of energy from the agent to the affected entity. The with-PP in combination with pepper does not conform to the specifications of the instrumental role: its referents often do not represent physical entities and often cannot be conceived of as mediating energy in the interaction designated by the verb3. Its semantic value can perhaps be better characterised as designating an entity which, as a result of the action performed by the agent, comes to be placed at the location designated by the direct object NP. Following Hale and Keyser (1996: 53– 109, 2002), arguments of this kind will be henceforth referred to as locata (for details, see sections 4 and 5).

3

Also, while the corresponding negative meaning of the instrumental with can be expressed with the preposition without (Quirk et al.1985: 699), the NP V(pepper) NP PP(with) construction can only be negated through verb negation (cf. *He peppered his designs without slogans).

110

(8)

Tomasz Fojt

a. He proposed that, instead, we should pepper the skies with computer controlled probes. (BNC) b. There are long disquisitions in Latin peppered with Greek words and phrases. (WordBanks) c. We’re not peppering the world with illegitimate children. (BNC) d. He wrote in a series of short sentences, peppered with question marks. (WordBanks)

Importantly for the argument status of the with-PPs in the constructions examined, their semantic interpretation depends crucially on the identity of the verbal heads with which they are associated. The with-PPs come to designate entities placed at locations only in combination with a relatively small number of verbs, semantically delimitable as locatum verbs (Clark and Clark 1979: 769–771; Hale and Keyser 1996: 53–109, 2002). In distributional terms, it must be concluded that the uses of with-PPs as locata co-occur with a much narrower range of verbal heads than is the case with instrumental adjunts, readily combining with virtually any nonstative verb, which conforms to the distinction postulated by Pollard and Sag (1987: 136). The same observation concerning the idiosyncratic dependence of arguments on their associated heads subserves Dowty’s (2003: 33–66) dual analysis of adjuncts and complements. Comparing PPadjuncts and PP-complements that have parallel ‘surface’ syntax, Dowty concludes that the semantic contribution of adjunct constructions can be determined without recourse to clausal heads, whereas the semantic interpretation of complement constructions can only be established in relation to the verb heads associated with them. In the light of its grammatical and semantic properties, the with-PP in combination with the verb pepper used metaphorically can be adequately viewed as a with-marked complement, to be distinguished from instrumental-with adjuncts. It stands in the (oblique) argument relation to the verb pepper, encoding the participant role of the locatum.

4. The properties of locatum verbs The verb pepper, together with saddle, blindfold or shoe, belongs to the class of verbs termed as locatum verbs (Clark and Clark 1979: 769– 771; Hale and Keyser 1996). The class includes denominal verbs whose derivation was traditionally described in terms of the lexical process of verbalization. In contrast, Hale and Keyser (1996: 62–70) propose that such verbs are formed by a syntactic process of successive incorporation

Secondary Patterns of Use with Locatum Verbs

111

of nominals into abstract prepositions and then into verbs. They further claim that the representation of the lexical relational structure of such verbs can be identified with that of their analytic paraphrases (i.e., ‘to saddle a horse’ and ‘to provide a horse with a saddle’ share the same lexical relational representation). Neither the details of the derivation nor, in fact, its validity, are of direct concern for the present study. A more pertinent observation is that the result of the derivation/incorporation process is a denominal verb that designates not only the action but also the entity that comes to be located at a certain place. The location of this entity is designated by the object of the verb. Thus, in sample (9) the verbs salt and pepper are locatum verbs and the noun phrase the fish designates location. (9)

Lightly salt and pepper the fish. (WordBanks)

Given the designation of locatum verbs (action of applying substances or entities of a particular kind), they normally do not allow a separate expression of the locatum argument. (10)

*Lightly salt and pepper the fish with salt and pepper.

They may, however, participate in the so-called cognate prepositional phrase construction where verbs zero-related to nominals accept prepositional phrases headed by with. (11)

a. Use thickly cut peel from 3 lemons, simmered gently in water that is heavily salted with sea salt. (WordBanks) b. All internal ironwork and insides of gangways were painted with aluminium paint. (BNC) c. Well firstly you tag it with the grammatical tags. (WordBanks) d. Little pieces of chicken are fried and then peppered with red cayenne pepper. (the Internet, DOA: 22 July 2011) e. Saddle your horse with the treeless saddle over your clean, light-colored pad. (the Internet, DOA: 22 July 2011)

The nouns which head the prepositional complement in cognate prepositional phrases are the ones from which the verbs derive. Optionally, the head of the prepositional complement may bear a hyponym relation to the noun from which the zero-related verb is derived (Levin 1993: 96–97). Additionally, cognate prepositional phrases seem felicitous only when they contribute additional information, i.e., when the nouns

Tomasz Fojt

112

heading prepositional complements are modified (note the infelicity of (10) above). Cognate prepositional phrases are more like adjuncts in that they are optional and do not individuate the meaning of the verbs they are related to: the verbs are readily interpretable without recourse to the cognate phrases which merely provide additional information. This is not true when locatum verbs are used in metaphorical senses in the construction NP VP(m-locatum) NP with-PP4. In figurative uses, the prepositional phrases are obligatory and make an indispensable semantic contribution to the meaning of the clause, which warrants their status as arguments (cf. the discussion in section 3).

5. The assignment of participant roles The discussion in section 3 presented the grounds for the argument status of the with-PP in the NP V(m-pepper) NP PP(with) construction. The present section capitalises on these results and aims at examining the assignment of semantic roles to the participants of the event designated by the verb pepper. In this way, besides the specificity of their grammatical patterning, the metaphorical uses of pepper are shown to involve distinctive configurations of semantic roles which set them apart from the literal uses of the verb. Langacker (1991: 284) observes that with the increase in the granularity of analysis, it may be said that “(…) every verb defines a distinct set of participant roles that reflect its own unique semantic properties.” It follows that the semantic properties of the participants of the event designated by the verb contribute to the overall meaning of the construction. Consider the three contrasting samples in (12) below. (12)

a. …Nobel Aldrich NP

peppers his macaroni …(WordBanks) NP VP(l-pepper)

AGENT

b. Such quips and asides NP LOCATUM

4

LOCATION

peppered the concert. (WordBanks) VP(m-pepper) NP LOCATION

In this formula, the letter ‘m’ used in the subscript signals that the relevant verb is used with a metaphorical semantic value. Correspondingly, the letter ‘l’ used in a similar way marks literal uses.

Secondary Patterns of Use with Locatum Verbs

c. We’re not peppering (BNC) NP VP(m-pepper) AGENT

113

the world with illegitimate children. NP

PP(with)

LOCATION

LOCATUM

In (12a) the verb is used with its primary literal semantic value and designates the process of applying condiment. The action is carried out agentively. The subject participant acts volitionally and performs an action which exerts a physical effect. The participant in the object position carries the patientive characteristics of an entity affected by the action; however, in the light of its role in the relationship designated by the verb, it can be more narrowly defined as a location. Furthermore, location participants in the literal uses of pepper are semantically delimitable as names of particular foodstuffs or hypernymic terms such as food or dish. The pattern in (12b) is superficially similar to that in (12a), but involves a distinct configuration of semantic roles assigned to the event participants. The agent remains unexpressed and the subject position is filled in with the locatum argument, i.e., the participant which, through the action described by the verb, comes to be placed at the location specified by the object argument5. The pattern of this kind can be considered as an alternation of the one exemplified by (12c) whereby the complement of the PP headed by with comes to be expressed as an oblique subject (cf. 13a–b below; for a list of verbs participating in the so-called locatum subject alternation, see Levin 1993: 81–82): (13)

a. She

peppered her talk

AGENT

b. Anecdotes LOCATUM

LOCATION

with anecdotes. (WordBanks) LOCATUM

peppered her talk. LOCATION

With the locatum functioning as the oblique subject, as in (13b), the verb may be interpreted as expressing a state rather than an action. In its figurative uses, in both the locatum subject and the prepositional locatum patterns, pepper no longer designates the physical action of seasoning. In fact, the precise semantic value of verb is difficult to pinpoint because it is heavily dependent on the details of the participant roles. For this reason, section 6 offers a review of the most frequent collocation patterns of pepper with a view to determining its metaphorical senses. In general 5

Occasionally, in this pattern, location is encoded in the form of a prepositional phrase, as in French kitchen terms peppered throughout English cookery books (BNC).

114

Tomasz Fojt

terms, the non-stative metaphorical uses of the verb may be said to express what is termed for the purpose of this study the applicative meaning, i.e., the verb designates actions which involve applying some entities (locatum) to another entity (location). Implicit in the meaning of the construction is that the addition of the locatum-argument referent to the location-argument referent endows the latter with certain qualities. Noun phrases functioning as locatum or location arguments are often headed by either abstract nouns or nouns naming activities. The locatum and location arguments receive metaphorical interpretation as places or entities by virtue of the well attested ontological and structural metaphors ACTIONS / 6 STATES ARE LOCATIONS. Also, more generally, the localist conception inherent in the metaphorical conceptualisation of the relationship designated by pepper in the construction examined conforms to the canonical directionality of metaphorical transfer from concrete source domains to abstract target domains put forth within the conceptual metaphor theory. Given the intrinsic asymmetry involved in the relationship in which the locatum and location arguments participate, they may be straightforwardly interpreted at the schematic level as the trajector and landmark respectively (Langacker 2000: 8). In those uses of pepper which can be interpreted as stative, as in (14) below, the verb profiles a relationship whereby the landmark, conceived of metaphorically as a planar region (location), contains within its boundaries the trajector (the entity located). Inherent in this conception is that the trajector (locatum) represents a number of entities of the same kind (expressed linguistically by a plural noun) and that it is dispersed more or less uniformly throughout the location landmark. Alternatively, the trajector encoded as an uncountable noun is conceived of metaphorically as a multiplex entity, often of granular composition. (14)

His talk was peppered with quotations from The Satanic Verses. (BNC)

A more specific semantic characterisation of the constructional meaning of the verb pepper seems possible only when the semantic properties of the participants involved in the relationship are taken into consideration. Hence, the next section looks into the collocational patterns of the verb with the aim of identifying the classes of nouns used as the locatum and location arguments in the construction examined. 6

For an analysis of the relevant conceptual mappings discussed with reference to the LOCATION EVENT-STRUCTURE METAPHOR, see Lakoff and Johnson (1999: 178– 205).

Secondary Patterns of Use with Locatum Verbs

115

6. Collocation patterns and semantic properties of participants The present section investigates collocation patterns characteristic of the metaphorical uses of the verb pepper in the construction examined. Attention is paid primarily to the nominal constituents of the construction in the light of the role that arguments play in individuating the meanings of the verbal heads associated with them (cf. section 3). The collocates taken into consideration are those with the highest statistical significance based on the t-score values. An attempt is then made to group the collocates into possibly narrowly defined semantic classes. In the WordBanks corpus, by far the greatest strength of association is observed between the verb pepper and the preposition with. With the raw frequency of 842, its t-score value amounts to 28.8047. The result is of particular significance for this study if it is remembered that, barring the rather exceptional and restricted prepositional cognate construction, the with-PP occurs only in metaphorical uses of the verb pepper. Thus, this statistic alone suggests that the figurative uses of pepper as a verb prevail over the literal ones. This is corroborated by examining the grammatical preferences of the verb pepper with the use of the Word Sketch function implemented in the WordBanks corpus. With the overall score of 70.3, the combination of pepper with the prepositional phrase headed by with features as a high-salience grammatical relation. Among the statistically significant collocates of pepper there are a number of lexemes that may be semantically delimited as related to the process of communication, both oral and textual. A representative selection of samples is given for illustration purposes in (15a–j). (15)

7

a. Avoid people who pepper their conversation with ‘shoulds, oughts and musts’. (BNC) b. Whitley (…) peppered his talk with literary bon mots … (BNC) c. These types of expressions pepper our language. (WordBanks) d. …his characteristic method was to pepper his interlocutor with questions… (BNC) e. The Chancellor peppered his Commons speech with so many references to Derek Wanless ... (WordBanks) f. He peppered her with questions. (WordBanks)

The lexeme that comes second in terms of association strength is the definite article, which scores only 19.413.

116

Tomasz Fojt

g. The document was peppered with phrases such as ‘in good faith.’ (WordBanks) h. …he’s been peppering his speeches with effusive words about former Tory leaders. (WordBanks) i. Interviews were peppered with New Age buzz-words … (BNC) j. To pepper one’s conversation with French expressions was to show that one was well-bred … (WordBanks) The nouns frequently acting as the location participants in the sentences referring to communicative acts name various modes of oral (cf. 15a, b, e, h, i, j) or written (15g) communication or refer the language system as a means of communication (15c). Alternatively, location nouns designate speech act participants (15d, f). Locata, on the other hand, irrespective of whether they are used as oblique subjects or as prepositional complements, typically name linguistic expressions (words, phrases, etc, cf. 15a, b, c, g, h, i, j) used in communication. They can also refer to illocutionary acts, such as statements or questions (15d, f). Combined with participants of this kind, the verb pepper designates a process of supplementing the message communicated (location) with additional elements of a particular type (locata). If location nouns refer to speech act participants, the verb designates repeated instances of verbal address to the interlocutor; the address is encoded as the locatum (15d, f). In those cases where the verb may receive a stative interpretation (15g, i), it conveys the idea of linguistic expressions of a particular kind (locata) being interspersed throughout a message (location). Another significant configuration of the participants of the event designated by pepper includes location conceived of as the target and locata as projectiles. While the location participants represent a varied category (almost any physical object may become a target), the nouns acting as locata consistently name types of relatively small missiles, such as gunshot, pellets, bullets, etc., as shown in the following citations: (16)

a. …another burst of enemy bullets peppered his motionless body. (BNC) b. The front panel of the medic van was peppered with shrapnel. (WordBanks) c. …they were peppered with pellets as they walked in the garden…(BNC) d. … all the trees to my left were peppered with gunshot. (WordBanks)

Secondary Patterns of Use with Locatum Verbs

117

e. … the resistance that has peppered U. S. forces with gunfire… (WordBanks) f. Those who did were swiftly cut down, peppered with birdshot…(BNC) As is characteristic of the construction under examination, the locatum participant is encoded either by a plural or uncountable noun. In those cases where the noun is uncountable, it designates a multiplex entity composed of small, yet individuated objects (e.g. shot, shrapnel). The uncountable noun gunfire in (16e) designates repeated firings of gun(s) whereby a number of bullets hit the target. The verb pepper used with such categories of participants designates a process of hitting the target with a considerable number of granular projectiles. The missiles may also be more individuated objects, as in the case of bullets, however, their number must be large enough to allow for the multiplex conception of the locatum (for this reason, *peppered with one/two bullet(s) seems infelicitous). The use of pepper in the sense of hitting a target, though distinct from the primary meaning of seasoning food, is significantly less metaphorical than that used in the context of communication. The issue of the gradability of metaphor will be addressed in more detail in section 8 below. The role of the location can be encoded by nouns which are semantically delimitable as designating periods of time (17a–e). In such cases, the construction instantiates the metaphorical conception of time in spatial terms (for a discussion of the spatialization metaphor of time, see Lakoff and Johnson 1999: 139–161). The nouns that name various periods of time are interpreted metaphorically as endowed with extensionality. Events, separated by temporal intervals, are reified as locata. The verb pepper expresses repeated occurrences of certain events in spatial terms as entities which are more or less uniformly interspersed throughout the location participant. (17)

a. At Forkhill, some of those events may pepper the daily cycle of wearying patrols … (BNC) b. ... the current season has been peppered with upsets. (WordBanks) c. The first seven years of his life, then, were peppered with separations. (BNC) d. Rene Tavernier’s life was peppered with great occasions. (BNC)

118

Tomasz Fojt

e. In an afternoon peppered by a spate of track incidents … (WordBanks) In a less specialised sense, the verb pepper designates the action of disseminating objects (locata) throughout a planar region (location) or covering the location with locata. In these meanings, the relationships expressed by pepper can be frequently construed as stative (cf. samples 18a, b, c, f, g, h, i). The degree of metaphorisation involved in the construction is relatively low: similar to the uses of pepper in its primary meaning, the participants in (18a–i) are physical objects (things) and not abstract entities or activities. (18)

a. Casinos pepper the Monte Carlo landscape… (WordBanks) b. … the land around may be peppered with the graves…(BNC) c. Consequently, the surface is peppered with impact craters. (BNC) d. … we should pepper the skies with computer controlled probes. (BNC) e. … a sprinkle of stars coming out shyly to pepper the desert skies. (BNC) f. It was peppered with parasite wasp holes. (WordBanks) g. … the body is peppered with small coloured or black spots…(WordBanks) h. … his hair was light brown…, only peppered with grey… (BNC) i. …there were more than a dozen shops peppering the main street. (WordBanks)

The uses of pepper in the sense of disseminating or covering seem also the least restrictive when it comes to semantic delimitation of the participants. The sole limitation on the location participant is that it represents a planar region or at least that it can be construed as such, as in the case of (18i). As for the locata, the only restriction seems to pertain to their number. Seemingly an exception to the plurality requirement, the noun grey in (18h) is an instance of metonymy whereby the characteristic attribute of the object (greyness) stands for the object (individual grey hairs interspersed throughout the mass of light brown hair).

Secondary Patterns of Use with Locatum Verbs

119

7. Extension of the metaphorical pattern to locatum verbs In their primary senses, locatum verbs designate actions related to putting, placing or scattering objects onto certain locations. Additionally, by virtue of their relation to the nouns from which they are zero-derived (cf. section 4), the verbs also make reference to the entities being located. Typical examples are given in (19a–f). (19)

a. The next day she decided to saddle the horse for a ride. (BNC) b. Sulkily she added milk and buttered a croissant. (BNC) c. They pitched their tent for comfort, and salved the animals’ seeping windburns … (WordBanks) d. These little bites of Chicken (…) and Pork are spiced to please. (WordBanks) e. Blister Padding cut into oval pieces to protect and cushion the bony areas. (BNC) f. She (…) poured her own coffee, sugared it, … (WordBanks)

The transitive verbs in (19a–f) in combination with phrases encoding locations straightforwardly designate the actions of placing a saddle on the horse, spreading butter on a croissant, applying salve to the wounds, padding an area with cushion-like objects, etc. The structure of the construction can be expressed formulaically as V(l-locatum) NP(location). The construction characteristic of locatum verbs should be distinguished from a related, yet distinct construction involving location verbs, exemplified in (20a–b). (20)

a. Seigl had shelved most of his books by subject … (WordBanks) b. My main task was to bottle wine … (BNC)

Structurally similar to the locatum verb construction which is the object of the present study, the location verb construction exhibits significant differences with respect to the configuration of the participant roles and the relationship into which the participants enter. Firstly, location verbs do not incorporate information related to entities being located (locatum participant). Instead, they make reference to the location at which the locata are placed. Consequently, in active-voice clauses locata are expressed separately and overtly as grammatical objects of the verbs. Secondly, as location verbs themselves encode the information about the place at which entities are deposited, they do not allow a separate expression of the location argument (except for the prepositional cognate

120

Tomasz Fojt

construction). Thus, the assignment of participant roles defined by location verbs is a reversal of the pattern characterstic of locatum verbs: V(l-location) NP(locatum). The locatum verbs used in their primary, literal senses in (19a–f) can all be used metaphorically, as shown in the samples below. (21)

a. Young teachers leave university saddled with debts of up to £15,000. (WordBanks) b. Top dealers are inviting big clients into their offices, buttering them up with expense account lunches … (BNC) c. Off duty, he salved his soul with booze and drugs. (WordBanks) d. Sydney also spiced his conversation with indelicate biblical allusions. (WordBanks) e. Many firms (…) cushion the blow with new bonus schemes … (BNC) f. He sugared the bitter pill with a promise … (WordBanks)

Embedded in a specific grammatical environment, locatum verbs accrue a metaphorical semantic value. For instance, the process of saddling in (21a) no longer refers to providing a saddle, but burdening with an (undesirable) liability or obligation; buttering in (21b) comes to express the activity of flattering performed to gain a favour from the referents of the location participant. As to the formal properties of the construction involving metaphorical uses of locatum verbs, it should be noted that the with-PP is an obligatory constituent: (22)

*Young teachers leave university saddled Ø.

Because the verb saddle in (22) no longer incorporates information concerning the entity with which the (metaphorical) saddling is done, it requires a separate expression of the locatum participant, encoded in the form of the with-PP. By virtue of its indispensability for individuating and completing the meaning of the verb, the prepositional phrase in the construction examined is warranted the status of an argument (cf. section 3). Consequently, the constructional pattern identified above for metaphorical senses of the verb pepper may be extended to figurative uses of other locatum verbs and generalised as NP VP(m-locatum) NP(location) withPP(locatum). It is further observed that although locatum verbs in general may participate in the prepositional cognate construction, the same restrictions as those stipulated with reference to the verb pepper obtain. Cognate with-

Secondary Patterns of Use with Locatum Verbs

121

PPs may combine only with locatum verbs used non-metaphorically, where they function as semantically and syntactically dispensable oblique adjunctive phrases, not as arguments. On the other hand, no cognate relation may obtain between a locatum verb used metaphorically and the head of the prepositional complement. In fact, it is precisely the use of non-cognate with-PPs that signals that the verbs should not be interpreted conventionally. The discussion in this section should not be taken to suggest that all locatum verbs are used metaphorically in the pattern VP(m-locatum) NP(location) with-PP(locatum). One of the advantages of the treatment of the construction examined here as a low-level schema is that it allows for lexical idiosyncrasy. Nor is it argued that whenever a locatum verb is used with a metaphorical semantic value, it must be embedded in the pattern with the prepositional phrase encoding locatum. Rather, the argument propounded here is that if a locatum verb is used in the construction of the form VP(locatum) NP(location) with-PP(locatum) and the locatum participant is not a cognate prepositional phrase, then the verb accrues a metaphorical meaning.

8. Gradability of metaphor In affinity with Hank’s proposal (2004: 245–274, 2006: 16–35), the analysis of the metaphorical uses of locatum verbs in this study supports the conception of metaphoricity as a graded phenomenon. A viable measure for determining the metaphoricity of a linguistic expression may be the degree to which the semantic properties of the participants interacting in the relationships designated by figurative uses of locatum verbs differ from those characteristic of the literal uses. For relevant contrasts, consider the citations gathered in (23). (23)

a. …Nobel Aldrich peppers his macaroni (…) before he tastes it. (WordBanks) b .…the current season has been peppered with upsets. (WordBanks) c. He peppered her with questions. (WordBanks) d. To pepper one’s conversation with French expressions was … (WordBanks) e. A volley of bullets immediately peppered the ground… (BNC) f. If they come back, sir, we’ll pepper them, pepper them proper. (WordBanks)

122

Tomasz Fojt

g. Sharpshooters began peppering the building, and in return the automated weapons fired into the woods. (WordBanks) h. The pub was peppered with bullets twice in an hour… (WordBanks) The sample in (23a) exemplifies a literal use of pepper, where the verb designates the activity of seasoning food. The nouns encoding location participants in such uses are concrete nouns, semantically delimitable as names of food or hypernymic terms, and locatum participants typically remain unexpressed. The construction in (23e) represents a minimal extension from the primary meaning of pepper in that the verb still designates a physical activity that involves projecting objects onto targets. The location participant is a planar region characterised by extensionality. The locatum participant shares certain characteristics with crushed pepper by referring to a number of small, granular objects. The uses of pepper with participants characterised by such attributes may be referred to as weakly metaphorical and, given the nature of the action designated, will be dubbed for the purpose of this study as pelting. In (23b–d) the semantic specifications of the participants diverge even more from those typical of literal uses of pepper. Accordingly, the meaning of the verb itself becomes more heavily dependent on the semantic attributes of its arguments. The locatum participants do not represent physical entities, but instead make reference to actions or activities such as disorderly behaviour or asking questions. The location participants may refer to a human being acting as an interlocutor (23c), communicative activity (23d), or even an abstraction, a period of time (23b). In combination with participants thus characterised, the verb pepper may be said to accrue a semantic value which is more strongly metaphorical. Determining the degree of overlap between the semantic properties of the participants interacting in relationships designated by literal and metaphorical uses of the verb is one method of measuring metaphoricity. Another metric proposed here pertains to the formal or structural properties exhibited by constructions. One of the results of the present study is demonstrating that strongly metaphorical uses of pepper require an expression of the locatum argument, either in the form of the with-PP or as the locatum subject. Literal uses of the verb, barring rare cognate prepositional phrase constructions, block the expression of the locatum argument. However, as attested in samples (23f, g), given an appropriate context, the locatum argument may remain unexpressed8. The sentences 8

For the principles of non-focal argument omission, see Goldberg: 2006: 195–197.

Secondary Patterns of Use with Locatum Verbs

123

may be considered elliptical, but their semantic interpretation seems unproblematic. The construction with implicit locatum is felicitous only if the verb is used in the weakly metaphorical sense, referred to above as pelting. What seems to warrant the uses of pepper with the locatum argument unexpressed is the fact that the pelting sense of the verb pepper is semantically closer—and thus less metaphorical—to the literal meaning of the verb than the strongly metaphorical senses. The fact that the locatum with-PP may be left out in weakly metaphorical constructions suggests that its argumenthood, rather than being categorial, is a matter of degree, too. The distribution of the formal patterning with respect to the degree of metaphoricity is as follows: in literal uses of the verb pepper, locatum is typically unexpressed; in the weakly metaphorical sense of pelting, locatum is normally expressed (cf. 23h), but may remain implicit; strongly metaphorical uses of the verb obligatorily require an expression of locatum. In this way, semantic affinity patterns come to be reflected in formal affinity patterns.

9. Conclusions The case study in this paper comprises an examination of the grammatical patterning associated with the verb pepper. Distinct patterns are identified for literal and metaphorical uses of the verb. It is established that the pattern with obligatory expression of the locatum argument in the form of the non-cognate with-PP and the corresponding locatum subject pattern are characteristic of metaphorical uses of the verb. In its nonmetaphorical sense, the verb does not allow a separate expression of the locatum argument (except for the cognate prepositional construction, the use of which is rare and subject to particular semantic restrictions). The results of the analysis are shown to hold, with the necessary modifications due to lexical idiosyncrasy, for the class of verbs termed as locatum verbs. It is further observed that in combination with a subset of metaphorical uses of pepper semantically definable as related to the action of pelting, the with-PP may exhibit adjunct-like characteristics in that its expression becomes optional in favourable contexts. The differing grammatical behaviour of the prepositional locatum sets the grounds for distinguishing weakly and strongly metaphorical senses of the verb. The conception of metaphorisation as a graded phenomenon is further corroborated by evidence adduced from the differences observed in the semantic properties of the participants of the relationships designated by pepper used in various metaphorical senses.

124

Tomasz Fojt

The lexico-grammatical patterning discussed in this study exhibits a number of distinctive properties: its specific syntactic organisation is coupled with a distinct assignment of the semantic roles of the profiled participants; the verbs used in the pattern are restricted to the semantically delimitable class of locatum verbs and are used with metaphorical semantic values. As such they form local patterns of limited scope and productivity, not yielding to generalisation (Langacker 2000: 118). For this reason, the properties of such low-level schemas are best captured in terms of constructions as defined by Goldberg (2006: 5–9). The advantage following from the treatment of the pattern examined here as an integral whole is that not need arises for positing different subcategorisation frames for metaphorical uses of locatum verbs. Their distinct argument structure can be seen as resulting from the specific requirements of the construction (Goldberg 2002: 344–346), which are in turn determined by its function.

References Clark, Eve and Herbert H. Clark. 1979. “When Nouns Surface as Verbs.” Language 55.4: 767–811. Dirven, René. 1993. Dividing up physical and mental space into conceptual categories by means of English prepositions. In Cornerial Zelinski-Wibbelt (ed.), The Semantics of Prepositions. From Mental Processing to Natural Language Processing, 73–97. Berlin: Mouton de Gruyter. Dowty, David. 2003. The dual analysis of adjuncts/complements in Categorial Grammar. In Ewald Lang, Claudia Maienborn and Cathrine Fabricius-Hansen (eds.), Modifying adjuncts, 33–66. Berlin: Mouton de Gruyter. Goldberg, Adele, E. 2002. “Surface generalizations: An alternative to alternations.” Cognitive Linguistics 13.4: 327–356 —. 2006. Constructions at Work. The Nature of Generalization in Language. Oxford: Oxford University Press. Hale, Kenneth and Samuel Jay Keyser. 1996. On Argument Structure and the Lexical Expression of Syntactic Relations. In Kenneth Hale and Samuel Jay Keyser (eds.), The View from Building 20. Essays in Linguistics in Honour of Sylvain Bromberger, 53–109. Cambridge, MA: MIT Press. —. 2002. Prolegomenon to a theory of argument structure. Cambridge, MA: MIT Press.

Secondary Patterns of Use with Locatum Verbs

125

Hanks, Patrick. 2004. “The syntagmatics of metaphor.” International Journal of Lexicography 17: 245–274. —. 2006. Metaphoricity is gradable. In Anatol Stefanowitsch and Stefan Th. Gries (eds.), Corpus-Based Approaches to Metaphor and Metonymy, 16–35. Berlin and New York: Mouton de Gruyter. Koenig, Jean-Pierre, Gail Mauner and Breton Bienvenue. 2002. “Class Specificity and the Lexical Encoding of Participant Information.” Brain and Language 81: 224–235. Lakoff, George and Mark Johnson. 1999. Philosophy In The Flesh. The Embodied Mind and Its Challenge to Western Thought. New York: Basic Books. Langacker, Ronald, W. 1991. Foundations of Cognitive Grammar. Descriptive Applications. Stanford: Stanford University Press. —. 2000. Grammar and Conceptualization. Berlin and New York: Mouton de Gruyter. —. 2008. Cognitive Grammar. A Basic introduction. Oxford: Oxford University Press. Levin, Beth. 1993. English Verb Classes and Alternations. A Preliminary Investigation. Chicago: The University of Chicago Press. Longacre, Robert, E. 1996. The grammar of discourse. New York: Plenum Press. Lyons, John. 1971. Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press. Mauner, Gail, Alissa Melinger, Jean-Pierre Koenig and Breton Bienvenue. 2002. “When is schematic participant information encoded? Evidence from eye-monitoring.” Journal of Memory and Language 47: 386–406. Pollard, Carl and Ivan A. Sag. 1987. Information-based syntax and semantics. Volume 1: Fundamentals. CSLI Lecture Notes Number 13. Stanford: CSLI Publications. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London and New York: Longman. Schütze, Carson. 1995. PP Attachment and Argumenthood. In Carson Schütze, Jennifer Ganger and Kevin Broihier (eds.), MIT Working Papers in Linguistics 26: Papers on Language Processing and Acquisition, 95–151. Cambridge, MA: MIT.

THE ENGLISH NPN FORMS: WORDS OR CONSTRUCTIONS? WIKTOR PSKIT

1. Introduction This paper is concerned with the English NPN forms exemplified by face-to-face or student after student. Three current approaches to these structures are presented. Jackendoff (2008) treats the NPN as a construction, understood as one of stored pieces of structure organized into a hierarchy. Thus viewed NPN forms violate the X-bar schema and constitute a “syntactic nut” problematic from the perspective of generative grammar. Haïk (2009) analyses NPN forms as lexical items, which are, however, split into two classes, viz. lexical small clauses and lexical coordinate nouns. Travis (2001, 2003), in the context of generative grammar, applies the mechanism of syntactic reduplication to the structures in question. The aspects of NPN forms that attract linguists’ attention include their categorial status, internal structure and distributional properties. The accounts in Jackendoff (2008), Haïk (2009) and Travis (2001, 2003) are discussed in terms of similarities and differences between them.

2. Jackendoff (2008): a construction-based approach Jackendoff (2008: 15) refers to the NPN form as construction in the sense of a piece of syntactic structure stored in the lexicon with associated meanings. The relevant construction includes a number of idiomatic expressions (Jackendoff 2008: 8): (1)

cheek by jowl, hand over fist, head over heels, hand in glove, tongue in cheek, hand to mouth

The most prominent feature of the construction is that most of its instances require the two nouns to be identical (Jackendoff 2008: 9):

The English NPN Forms: Words or Constructions?

(2)

127

hand over hand, hand in hand, arm in arm, (tear NP) limb from limb

Another property of the NPN forms emerging from examples in (1) and (2) is that the choice of nouns is relatively limited (e.g. *tongue in tooth, *finger over finger). Yet, the NPN construction appears to be productive with the prepositions by, for, to, after, and upon in that these prepositions do not impose strict constraints on the selected nouns (Jackendoff 2008: 9): (3)

a. b. c. d. e.

day by day, paragraph by paragraph, country by country dollar for dollar, student for student, point for point face to face, bumper to bumper term paper after term paper, picture after picture book upon book, argument upon argument

Nevertheless, the construction is highly constrained in syntactic terms. In general, the nouns in NPNs cannot be uncountable nouns (4a) (with certain exceptions). The nouns are not allowed to have determiners (4b) or plurals (4c), postnominal complements or modifiers (4d) (with the exception of after and upon (4e)). However, with some prepositions, prenominal adjectives occur on either the second noun or both (4f) (Jackendoff 2008: 9): (4)

a. *water after water, *dust for dust b. *the man for the man, *a day after a day, *some inch by some inch c. *men for men, *books after books, *weeks by weeks d. *father of a soldier for father of a soldier, *day of rain to day of rain e. day after day of rain f. day after miserable day, tall boy by tall boy

One of the issues that are problematic is the categorial status of NPNs. Possible clues as to the syntactic category of the relevant structure can be provided by the syntactic functions (understood in terms of positions in the clause) that NPN forms can have. As Jackendoff (2008) observes, NPNs predominantly occupy adjunct positions: sentence-initially (5a), after the subject (5b), and in VP after the complements (5c). They can also occur within an NP, in the position normally occupied by a prenominal adjective (5d). Moreover, specific instances, namely N after N and N (up)on N, can

128

Wiktor Pskit

appear in positions characteristic of NPs, i.e. subject (5e), object (5f), object of PP (5g), possessive NP (5h) (Jackendoff 2008: 19): (5)

a. b. c. d. e. f. g. h.

Page for page, this is the best-looking book I’ve ever bought. John and Bill, arm in arm, strolled through the park. We went through the garden inch by inch. Your day-to-day progress is astounding. Student after/upon/*by student flunked. We filled crack after/*by crack. We looked for dog after/*by dog. Student after/*by student’s parents objected.

The distributional characteristics of NPN forms in examples above suggest that they can be regarded as PPs, because PPs are typical sentential adjuncts and some PPs can appear pronominally. Other possibilities include the categorization of the relevant structures as AdjPs, when dominated by NP, and as AdvP, when dominated by S or VP. Jackendoff (2008: 19) also points out that some NPNs semantically parallel secondary AdjP predicates. Still, there is nothing in the structure of NPNs that makes them resemble canonical AdjPs or AdvPs. Another solution is proposed by Jackendoff’s (2008) referee: NPNs can be treated as NPs marked with the grammatical function oblique. This function can be attributed to the time expression ten minutes in He sat there ten minutes. However, as Jackendoff (2008: 19) argues, this expression can be paraphrased with the PP for ten minutes, whereas no comparable paraphrase is available for adjunct NPNs. According to Jackendoff (2008: 19), the adjunct positions that NPNs occupy are “categorically unselective: they care more about semantics than syntactic category.” Then, there is a possibility that the distribution of NPNs is determined by semantics, leaving open the option that such forms are categoriless in syntactic terms. Although the categoriless status of NPN forms seems to be tempting from the point of view of the Simpler Syntax Hypothesis (Culicover and Jackendoff 2005), where semantics is allowed to override syntax, Jackendoff (2008) decides to classify adjunct NPNs as PPs. The internal structure of English NPNs likewise raises a number of questions. If, as Jackendoff (2008) argues, NPNs are PPs, their internal structure remains unusual. Above all, nouns without determiners do not occur as typical complements of prepositional heads. The same refers to specifiers of prepositions, which are normally realized by degree phrases and quantifiers, not by bare nouns again.

The English NPN Forms: Words or Constructions?

129

For Jackendoff (2008: 21) it is the preposition that determines the interpretation of the whole NPN and that decides which nouns are available. Hence, it is an argument in favour of NPNs as PPs. Jackendoff (2008) offers an account of English NPN forms in the spirit of the construction-based approach and his own model of parallel architecture (Jackendoff 2002, Culicover and Jackendoff 2005). He suggests that each of the prepositions that occur in NPNs has its own associated meanings in the relevant construction. Thus, a number of subconstructions are identified with different degrees of productivity. The subconstructions have the status of lexical entries, because the constructionist approach does not distinguish between words and rules – what matters is the level of productivity or generality/specificity, with some entries being fully specified, i.e. word-like, and others being more variable, i.e. more rule-like. Thus conceived entries form inheritance hierarchies, where what particular instances have in common is accounted for by entries that are higher up in the hierarchy (Jackendoff 2008: 10–18). Based on the idea of the parallel architecture (Jackendoff 2002), Jackendoff (2008: 26) treats the NPN as a reduplicative construction and proposes the following constructional idiom that maps between semantics, syntax and phonology: (6)

Meaning a. Syntax Phonology b. Syntax Phonology

MANY (MODk) Xis IN SUCCESSION [NP Ni Pj (Ak) Ni] Wdi afterj (Wdk) Wdi [NP (Ak) Ni Pj (Ak) Ni] (Wdk) Wdi afterj (Wdk) Wdi

The notation in (6) involves the shared subscript i so that the entity that is counted maps into both nouns in the NPN. Although the preposition of the syntactic component maps into after in the phonological component, it does not map into anything in the semantic component. This emphasizes the idiomatic character of English NPNs, where the preposition determines the meaning of the whole construction. The representation in (6) also accounts for the presence of adjectival modifiers, with two cases (6a and 6b), depending on whether the adjective is reduplicated along with the noun. In (6a), the modifier is mapped into a prenominal adjective preceding the second noun, while in (6b) it is copied before the first noun. This captures the fact that a single adjective before the second noun modifies the whole construction. Indeed, the semantic component contains only one counterpart for the two nouns of the syntactic component, and hence just one modifier as well.

130

Wiktor Pskit

According to Jackendoff (2008), the idiosyncratic behaviour of particular instances of English NPNs renders them “syntactic nuts” that do not lend themselves to the derivational account favoured in generative grammar. The distributional characteristics of NPNs and their corresponding meanings suggest that they are placed on the continuum between idiosyncrasy, semi-productivity and full productivity, which requires explanation in terms of construction grammar.

3. Haïk (2009): lexical small clauses and lexical coordinate nouns Unlike Jackendoff (2008), Haïk (2009) makes a distinction between lexical entities and syntactic entities. NPN forms are lexical items (words) that are divided into two classes: lexical small clauses and lexical coordinate nouns, based on the category of the head and the properties of the construction (Haïk 2009: 2): (7)

a. face-to-face construction (lexical small clause): headed by P face to face, hand over hand, hand in hand, arm in arm, dollar for dollar b. student-after-student construction (lexical coordinate noun): headed by N student after student, (day by day?), (book upon book?)

This classification is based on the following reasoning: face-to-face is an absolute small clause (SC), because its meaning and distribution parallel those of an absolute SC, whereas student after student is a coordinate noun, because it exhibits the meaning and behaviour of a coordinate structure. Following Jackendoff (2008), Haïk (2009) points to the basic properties that the two constructions in (7) have in common. Firstly, the nouns are identical in most cases (counterexamples: head to toe, boy to man, head to tail). Secondly, the relevant expressions have the NPN form, without determiners, plural nouns, complements, or premodifiers, which is the property of words. This means that NPNs are morphological entities. The approach to NPN forms that assumes they are words offers a straightforward explanation of their properties. If, as Williams (1994) claims, word-formation rules cannot take phrases as their inputs, it becomes clear why in NPNs there are no determiners, plural nouns, complements and premodifiers, i.e. elements normally expected in the structure of phrases.

The English NPN Forms: Words or Constructions?

131

Haïk (2009: 3) identifies absolute constructions as small clauses that involve the descriptions of the (physical or mental) posture of humans, as shown below: (8)

a. I found him with his hands up. b. They stood there, their eyes lowered. c. He raised his eyes, his mind racing.

Such absolute constructions function as syntactic modifiers in the form of small clauses. The English face-to-face construction does not occur as the predicative complement of a copular verb (Haïk 2009: 5): (9)

*They seem/appear/became face to face.

Nevertheless, such NPNs can function as NP modifiers and adjuncts of manner (Haïk 2009: 5): (10)

a. b. c. d.

They stood hand in hand. Nose against nose, they were looking at each other. They walked along arm in arm. back-to-back performances

These facts are sufficient for Haïk (2009) to conclude that in terms of meaning and distribution the English face-to-face is an absolute small clause. In Haïk (2009), NPNs are symmetric structures. Given that NPNs are lexical items, they seem to violate the cognitive constraint on ternary structures in the lexicon, ternary structures being apparently excluded since computation only permits pairing. Haïk invokes the general human cognitive capacities to show that there is one way to account for such ternary structures in which two nouns can be associated at the same time with the preposition. The face-to-face construction, not only requires two identical nouns, but also describes a symmetric spatial relation between the relevant parts of body, viz. mirror symmetry, in which one body part is the mirror image of the other. Such mirror symmetry holds for the whole construction (e.g. hand in hand, back to back, side by side, arm in arm, etc.). Haïk (2009) argues that the identity requirement on the nouns is not only lexical but also semantic, i.e. related to mental representation, and that, rather than a condition, this is a consequence of that particular structure and of the fact

132

Wiktor Pskit

that this construction results from a mental operation effected in a symmetric fashion. Haïk (2009) follows Jackendoff’s idea (cf. (6) above) that the numeration of the NPN construction is {P, N}, i.e. it starts with just two elements. There is a single mental operation that involves a symmetric and synchronic merge of the preposition to the left and to the right with the noun. Thus obtained symmetric representation corresponds to symmetric semantics. Both nouns in NPNs are “the structural projections” of the same element, with “twin mental representations, one associated with the N before the P and the other with the N after the P” (Haïk 2009: 10). The preposition is the semantic head of the whole construction, because its meaning determines the setting of the mental representations of the two nouns in mirror symmetry (for the face-to-face construction) or in sequential symmetry (for the student-after-student construction). As far as the student-after-student construction (e.g. house after house, time after time, student upon student) is concerned, it stands for those NPNs in which the preposition indicates succession. The difference between the face-to-face and the student-after-student constructions lies in that while the former are headed by the preposition, the latter have the noun as the head, which is the property of syntactic coordination. The nominal character of the latter type is reflected in its nominal functions. There seems to be a constraint on the prepositions licensed in this kind of coordinated structures, with merely two English prepositions found in such expressions: after and upon. There is some conflict between form (the ternary structure of NPN) and meaning (the subject-predicate interpretation) in the lexical coordinate structures. To solve the problem, Haïk (2009: 12) employs the notion of repetition symmetry, which has to do with “the repetition of the same object on either part of the symmetry axis.” The English prepositions after and upon can appear in lexical coordinate NPNs, because, owing to their meanings, they take on the function of conjunctions: they point to the relation between two nouns interpreted as “a sequencing of the same things” (Haïk 2009: 12). The status of the instances of the student-after-student construction as lexical coordinate nouns stems from two features they share with coordinate structures. Firstly, in NPNs with after and upon, the second noun can take an adjectival premodifier or a complement (where the premodifier or the complement modifies both nouns), as in (11) (Haïk 2009: 14):

The English NPN Forms: Words or Constructions?

(11)

133

a. day after miserable day b. Layer upon layer of mud lay on the seabed.

Secondly, after is the only preposition that can be iterated (Haïk 2009: 15): (12)

day after day after day of unending rain

The structure of such lexical coordinate nouns is as follows (Haïk 2009: 14): (13)

a. [ [ day after day ]N miserable ]NP b. [ [ layer upon layer ]N [ of mud ]PP ]NP

In cases such as (11a), with the structure in (13a), the adjective inserts before the second noun as a result of linearization of the adjective in the coordinate structure. The interpretation of the adjectival premodifier in (11a) or the complement in (11b) “spreads onto the two nouns, as an across-the-board effect of coordinate structures, in which a single surface element is interpreted in the two coordinates” (Haïk 2009: 15). Haïk (2009) stresses that her analysis of N after N forms as truly coordinate structures is supported by the fact that the possibility of an interpretation of a single element as having scope over two phrases characterizes coordinate structures. NPNs classified as lexical small clauses (cf. 7b) do not allow premodifiers or complements with scope over both nouns.

4. Travis (2001): syntactic reduplication Travis (2001, 2003) applies the mechanism of syntactic reduplication to the analysis of NPN forms in English. Although syntactic reduplication seems to be much less common than phonological reduplication, Travis argues that reduplication also occurs in syntax and when it does, it is subject to syntactic constraints. In fact, all cases of reduplication involve phonological copying that is triggered by syntactic environments. This also refers to instances that seem to be morpho-phonological reduplication, where reduplication is triggered by the syntactic configuration created via head movement (Travis 2003). Travis (2001) reconsiders data presented in Pi (1995) in (14) below.

134

(14)

Wiktor Pskit a. b. c. d. e.

The butterfly fluttered from flower to flower. Jon washed plate after plate for hours after the party. The careful artist completed the mosaic tile by tile. Eric can drink mug upon mug of coffee in a single hour. In fairy tale after fairy tale, good triumphs over evil.

Pi (1995) argues that certain prepositions in English like after, upon, and by appear in constituents where the material that appears before the preposition is a copy of (part of) the material in the complement of the preposition. According to Pi, the material preceding the preposition is a linearization of what is generated in the complement position. In a sense, in (14) the material is repeated to encode plurality and the repeated items function like bare plurals, which explains why they occur without determiners. Interestingly, the lack of repetition yields ungrammatical structures, indicating that the repetition is not optional (Travis 2003): (15)

a. *The butterfly fluttered to flower. b. *Jon washed plate.

According to Travis (2001, 2003), NPN forms such as those in (14) are constructed by a form of syntactic reduplication and these structures and phonological reduplication structures should be seen as two examples of the same process. This idea is supported by a number of similarities between phonological reduplication and syntactic reduplication. First of all, in both subtypes of reduplication, different size domains can be reduplicated. Phonological reduplication exhibits variation in the size of prosodic domains that can be copied, ranging from a core syllable to a minimal word. The same holds for syntactic reduplication, where reduplicated items are of varying size/complexity (data from Pi 1995): (16)

a. b. c. d. e.

cup after cup of coffee cup of coffee after cup of coffee cup after steaming cup of coffee steaming cup after steaming cup of coffee steaming cup of coffee after steaming cup of coffee

Further, in both types of reduplication, information must be copied from the rightward host and nothing can be added if it is not included in the host. The following shows that the addition of information beyond

The English NPN Forms: Words or Constructions?

135

what is in the host to the copy results in an ungrammatical expression (Pi 1995): (17)

* steaming cup after cup of coffee

Apart from this, in both types the reduplication mechanism is sensitive to constituency. In Pi (1995) the structure for steaming cup of coffee is represented by (18a), which yields (18b) ungrammatical because it involves copying a non-constituent. (18)

a. [ [ steaming cup ] of coffee] b. * cup of coffee after steaming cup of coffee

Phonological reduplication and syntactic reduplication also share a number of properties that have to do with function. Phonological reduplication is used in a similar manner across languages. According to Moravcsik (1978), this type of reduplication relates to repeated events. Likewise, syntactic reduplication can designate plural nouns or plural events. An iteration of events is illustrated in (19a-b) (data from Pi 1995, after Travis 2001): (19)

a. Student after student visited the professor on Monday. b. Gertrude watched program after program all afternoon.

Travis (2001) discusses some of the details of the mechanism of reduplication. In general, reduplication is assumed to be triggered by the need to check a formal feature. Head movement takes place before copying involved in reduplication occurs, which entails the occurrence of head movement prior to phonological reduplication. By contrast, in syntactic reduplication feature checking does not force movement to the position of specifier. In this case, a copy created by the head occupies the specifier position and checks the relevant feature. In other words, syntactic reduplication requires that copying precede checking. Yet, there are differences between the two subtypes of reduplication as well. First of all, phonological and syntactic reduplication differ with respect to the type of domain that is subject to copying, defined in prosodic terms for the former and in syntactic terms for the latter. Travis (2001) attributes this sort of difference to the difference in syntactic configurations. Then in phonological reduplication, since reduplication follows head movement, a subdomain of X0, which is the sister of the reduplicative head is assumed to be sub-syntactic, and hence invisible in

136

Wiktor Pskit

syntax. On the other hand, in the case of syntactic reduplication, as reduplication precedes head movement, giving rise to XP as the sister of the reduplicative head, a subdomain of XP is determined syntactically. Another source of difference is concerned with the size of reduplicated item, which is fixed for phonological reduplication, by virtue of being reduced to affixal forms with prosodically defined inherent shapes, and variable for syntactic reduplication, because specifiers – hosts for copied material – are not required to have an inherent shape defined in prosodic terms. Furthermore, whereas phonological reduplication involves left-to-right mapping, syntactic reduplication is based on bottom-up derivation. In the case of syntactic reduplication, the head of the complement is included in the copy irrespective of being the left-most or right-most element in the structure. At the same time, syntax permits copying merely a head, but phonology does not allow just a vowel to be copied. Travis (2001) suggests that this difference stems from different default forms in each domain, with a head being a minimal possible default form in syntax, but with CV being a minimal default in phonology. Travis (2001) also points out that contrary to phonological reduplication, which allows only a single copy, syntactic reduplication enables more than one copy to appear in the relevant slot, as shown in (20a). This is possibly due to the general tolerance for emphatic iteration in syntactic structures (20b). (20)

a. He drank cup after cup after cup of coffee. b. He drank many many many cups of coffee.

What is interesting, Pi (1995) observes that the preposition upon is exceptional in that it occurs with plural nouns in NPN forms: (21)

Bags upon bags of marshmallows were stolen this week.

On the other hand, this is not possible with after, a preposition which appears to behave in a way that is very similar to upon: (22)

a. Student after student visited the professor on Monday. b. * Students after students visited the professor on Monday.

According to Travis (2001), this difference in behaviour between the two prepositions is based on the variable nature of the reduplicative morpheme. The morpheme can occupy various head positions: if realized

The English NPN Forms: Words or Constructions?

137

by upon it occurs in a head above Number, thus allowing plural nouns, and if realized as after it appears in the Number head, thus excluding plural nouns. Although it is always a head, the reduplicative morpheme does not have to constitute the final site for the copy. Then, if the morpheme has a head (a minimal projection) as its sister, the copied material may appear in the very position of the morpheme. However, if the morpheme has a phrase (maximal projection) as its sister, then the copied material, probably owing to its structural complexity, has to appear in the specifier of the projection headed by the reduplicative morpheme. From this point of view, Travis (2001) takes after and upon to be reduplicative morphemes since they give rise to a copy of a subdomain of their sisters, but the two prepositions differ from other reduplicative morphemes because they do not constitute the copy themselves. Among major issues that require explanation is the question of the constraint on the domain that is copied. As noticed in (18) above, it seems that syntactic reduplication is sensitive to constituency. However, this generalization is slightly undermined by the data in (23) below, where one can observe that it is easier to copy a domain larger in size than a head noun if the noun functions as a measure phrase (Travis 2001): (23)

a. b. c. d.

I met with student after student. * I met with student of linguistics after student of linguistics. I ate bunch after bunch of bananas. I ate bunch of bananas after bunch of bananas.

Although cup of coffee can be copied as (16b) above demonstrates, copying of student of linguistics in (23b) is not licensed. Travis (2001, 2003) juxtaposes three types of reduplication: phonological reduplication, syntactic reduplication and contrastive reduplication. Contrastive reduplication, analyzed in Ghomeshi et al. (2004)1, exemplified in (24) below, appears to raise similar questions concerning the constraints on what actually can be reduplicated (Travis 2003: 239): (24)

1

a. I want a SALAD salad not a tuna salad. b. They are TOGETHER together.

Travis (2001, 2003) refers to a conference paper, delivered in 2000, that Ghomeshi et al. (2004) is based on.

138

Wiktor Pskit

Contrastive reduplication differs from the other two reduplication types in that it is not associated either with an affixal position, hence it is not phonological, or with the specifier position, hence it is not syntactic. In addition, as examples in (25) show, contrastive reduplication does not seem to be governed by phonological (25a-b) or syntactic (25c-d) constraints (Travis 2003: 240): (25)

a. b. c. d.

She’s a GIRL-girl. * She's not my GIRL-girlfriend. Do you LIKE HIM-like him? *Do you LIKE BILL-like Bill?

Regardless of the type of constituent that is copied, it is possible to define the position that the copy occupies a position which is available for other material in syntactic terms (Ghomeshi et al. 2004: 352) (data from Travis 2003: 240): (26)

a. It is part of the ____ [real/concrete] highway, not the information highway. b. I'll make the tuna salad and you make the _____ [real/green] salad. c. They weren't ____ [really] together, they were just studying together.

Travis (2001, 2003) claims that all reduplication is actually a syntactic phenomenon. One of the reasons for this conclusion is the presence of basic syntactic relations, viz. specifier-head, head-head and modifier-head, in all the three cases investigated as well as the fact that the copy occurs in an independently available syntactic position. This is interpreted as follows: although reduplication is syntactically driven, it may need to resort to phonological terminology whenever the syntactic environment of reduplication involves subsyntactic material. Certain aspects of syntactic reduplication and contrastive reduplication seem to be of significance in a broader theoretical context. For instance, grammatical structures such as (16b) contrasted with ungrammatical ones such as (23b) suggest that syntactic reduplication does not allow reduplication of all complements of heads, which has consequences for the internal structure of the relevant reduplicated expressions. The types of constituents that can undergo contrastive reduplication might reveal some facts concerning structures with modifiers (Travis 2001).

The English NPN Forms: Words or Constructions?

139

As emphasized by Travis (2001), the type of copying observed in syntactic reduplication is essentially different from a similar mechanism found in movement. One difference has to do with the distinction between feature checking by means of head movement and by means of occurrence in the specifier position. Head movement precedes copying whereas setting up the specifier-head relationship follows copying. Also, copying for reduplication appears to be different from copying for movement. The former permits copying subdomains of the complement XP which independently cannot undergo movement. Travis (2003: 240) proposes the following representations for the three types of reduplication: phonological (Fig. 1a), syntactic (or iterative) (Fig. 1b), and contrastive (Fig. 1c):

Fig. 1a. Phonological reduplication

Fig. 1b. Iterative (syntactic) reduplication

140

Wiktor Pskit

Fig. 1c. Contrastive reduplication

All the types in (Fig. 1) involve a syntactic configuration with a head Q (quantity) merged into the structure. If the Q head is realized as a reduplicative morpheme, it creates a copy of its sister or the subpart of its sister. If the relevant sister is a head itself, which is the case in (Fig. 1a) and (Fig. 1c), the copy is realized in the Q head. However, if the sister is a maximal projection (XP), as in (Fig. 1b), the copy is realized in the specifier position. In phonological reduplication (Fig. 1a) the reduplicated item is an affix – possibly conceived of as a feature set – that is generated in a syntactic head. Head movement, triggered by a certain feature of the head, is followed by the copying of the relevant phonological material into the affix. Because in (Fig. 1a) the sister of Q is a head, the subparts of the sister are phonologically determined. On the other hand, in syntactic (iterative) reduplication in (Fig. 1b) no head movement takes place. Instead, the copying of a subpart of the sister of Q means the copying of a subpart of XP, not of X0. The subpart of XP that is copied is determined in syntactic terms. Finally, contrastive reduplication, investigated in detail in Ghomeshi et al. (2004), provides greater difficulty as far as the characterization of its structure is concerned. The syntactic position where the copy appears seems can be treated as a pre-head adjective or adverb. Travis (2003), following insights form Travis (1988), suggests that such pre-head adjectives and adverbs are defective items, since they do not project, but are head-adjoined to modified items. Contrastive reduplication is distinguished from phonological reduplication by the fact that while the sister projects in the former, Q projects in the latter (Travis 2003: 240– 241).

The English NPN Forms: Words or Constructions?

141

5. The three approaches in comparison The three approaches to English NPN forms presented above can be discussed in terms of similarities and differences between them. The thing they most obviously share is the range of data from English they investigate2. Interestingly, the notion of reduplication appears, explicitly or implicitly, in all three approaches. But in each case the phenomenon is differently motivated. For Jackendoff (2008) reduplication results from the mapping between the form, meaning and phonology of NPNs, which reflects the tendency in Simpler Syntax (Culicover and Jackendoff 2005) to analyze a number of issues in terms of semantics as the driving force rather than in terms of syntax, as opposed to common practice in generative grammar. In Haïk (2009) repetition symmetry, understood as a cognitively-grounded mental operation, bears some resemblance to the concept of reduplication. Travis (2001, 2003) takes reduplication to be a syntactically-driven operation to the extent that even phonological reduplication and contrastive reduplication are sensitive to syntactic environments. An issue that resurfaces in all three studies in question is the distinction between grammar (or syntax) and lexicon. Jackendoff (2008) makes no such distinction and employs the concept of construction, which goes across the boundary between linguistic components. Then, for him NPNs are constructions understood as stored pieces of structure with associated meanings organized into an inheritance hierarchy. Haïk (2009) appears to maintain a distinction between words and phrases, i.e. between lexical and syntactic entities. Haïk’s NPN forms are lexical items and they can be small clauses or coordinate nouns. This means that small clauses can be word-level forms, with the consequence that perhaps morphological rules can take syntactic constructs as their input. Although Travis (2001, 2003) does not address the problem explicitly, the crucial role of syntactic environment in all instances of reduplication, including what is traditionally regarded as (morpho)-phonological reduplication, suggests that Travis abandons the distinction between syntax and lexicon at least in its traditional form. This is supported by the idea, put forward in Distributed Morphology (e.g. Embick and Noyer 2007) and recently adopted in much work in minimalist syntax (e.g. Boeckx 2009), that all derivation (i.e. of words and of syntactic objects) reduces to syntactic derivation. The lexicon is limited to a repository of lexical atoms 2

Haïk (2009) devotes some attention to data from French, which is, however, beyond the scope of the present paper.

142

Wiktor Pskit

(category-neutral roots), which are manipulated by syntactic operations. Then, there are at least two different ways – suggested in construction grammar and generative grammar – of avoiding the problem of the status of NPNs as lexical or syntactic entities, i.e. words or constructions in traditional terms. What Jackendoff (2008) and Haïk (2009) have in common is the key role attributed to the preposition as the semantic head of the NPN. This corresponds to the function of reduplication morpheme, possibly hosted by the preposition, which triggers reduplication. Crucially, what distinguishes between the three analyses presented in this paper is the way they explain why English NPNs have the structure and properties they do. Jackendoff’s (2008) explanation, rooted in his theory of parallel architecture, is based on the idea of mapping between the semantics, syntax and phonology of the construction. Haïk (2009) accounts for the structure and properties of NPN forms in terms of wellformedness rules of syntax and the lexicon, coupled with general human cognitive capacities, arguably rewarded by symmetry; hence her proposal to treat NPNs as symmetric structures. In Travis (2001, 2003) NPNs are derived by means of syntactic operations of more general character. English NPN forms are problematic from the point of view of their categorial status. In Jackendoff (2008) NPNs that occur in adjunct positions are treated as PPs, but a position regarding these forms as having no syntactic category is recognized as well. Haïk (2009) distinguishes between NPNs that are lexical small clauses and those that are lexical coordinate nouns. In Travis (2001, 2003) NPNs are maximal projections of the Q head hosting the reduplicative morpheme. One of the issues that remain controversial and require explanation is the fact that in NPNs both nouns appear without determiners even if they are apparently countable singular nouns. Then, from the generativist perspective, these bare nouns are not DPs unless headed by a null determiner. A solution that deserves further exploration is to treat them as NP-shells (nPs) headed by light nouns in the sense of Radford (2000), an idea mentioned by Travis (2003) in the context of parallels between NPNs and pseudo-partitives.

6. Conclusion The paper presents three different analyses of NPN forms in English. It points to the similarities and differences between the works it compares. The areas discussed above have to do with the structure of English NPN

The English NPN Forms: Words or Constructions?

143

forms, their categorial status and the mechanism of their derivation/formation. The study shows that comparing competing approaches can be beneficial. First of all, it demonstrates that the results of the work in one framework often fuel research in other theoretical camps, and usually it is a mutual or multidirectional influence. This is evidenced by references in the works discussed above. Jackendoff (2008) is in many ways a response to minimalist derivational approaches, and Haïk (2009) is a reaction to Jackendoff’s account. Secondly, from a broader perspective, instead of looking at the multitude of frequently opposed proposals as a weakness, it is possible to interpret the current state of syntactic theory, or perhaps even linguistic theory in general – a great number of theoretical frameworks and lack of consensus among their representatives on many basic issues in language study – as a symptom of the rapid development of the discipline. This development is definitely facilitated by exchange of ideas between those who represent different linguistic orientations, a situation recognized as desirable for instance in Culicover and Jackendoff (2005: 546–7) and Boeckx (2009: 250).

References Boeckx, Cedric. 2009. Bare Syntax. Oxford and New York: Oxford University Press. Culicover, Peter W. and Ray Jackendoff. 2005. Simpler Syntax. Oxford and New York: Oxford University Press. Embick, David and Rolf Noyer. 2007. Distributed Morphology and the Syntax/Morphology Interface. In Gillian Ramchand and Charles Reiss (eds.), The Oxford Handbook of Linguistic Interfaces, 289–324. Oxford and New York: Oxford University Press. Ghomeshi, Jila, Ray Jackendoff, Nicole Rosen and Kevin Russell. 2004. “Contrastive Focus Reduplication in English (The Salad-Salad Paper).” Natural Language and Linguistic Theory 22: 307–357. Haïk, Isabelle. 2009. “Symmetric structures.” www.crisco.unicaen.fr/IMG/pdf/Symmetric_structures-TLR.pdf, DOA 08/09/2010. Jackendoff, Ray. 2002. Foundations of Language. Oxford and New York: Oxford University Press. —. 2008. “Construction after construction and its theoretical challenges.” Language 84.1: 8–28.

144

Wiktor Pskit

Moravcsik, Edith. 1978. Reduplicative constructions. In Joseph H. Greenberg (ed.), Universals of Human Language, 297–334. Stanford, CA: Stanford University Press. Pi, Chia-Yi Tony. 1995. “The structure of English iteratives.” Canadian Linguistics Association Proceedings (Toronto Working Papers in Linguistics): 434–445. Radford, Andrew. 2000. “NP shells.” Essex Research Reports in Linguistics 33: 2–20. Travis, Lisa. 1988. The syntax of adverbs. In Denise Fekete and Zofia Laubitz (eds.) McGill Working Papers in Linguistics: Special Issue on Comparative Germanic Syntax, 280–310. Montreal: McGWPL. —. 2001. “The syntax of reduplication.” Proceedings of NELS 31: 455– 469. —. 2003. “Reduplication feeding syntactic movement.” Canadian Linguistic Association: 236–247. Williams, Edwin. 1994. Thematic Structure in Syntax. Cambridge, MA: MIT Press.

THE BENEFITS OF SWEARING: TOWARDS AN OLD ENGLISH TYPOLOGY OF OLD ENGLISH VERBS MAGDALENA CHARZYēSKA-WÓJCIK

1. Introduction Old English verbs exhibit enormous variability as far as their complementation frames are concerned. This variability concerns not only the available object types but also their number, to the effect that one and the same verb may appear with two objects, one object or even without an object at all. The category of the objects may differ as well, i.e. an object may be expressed by means of NPs, PPs or clauses. The variability also affects the choice of a particular type of object so that an NP may appear in the ACC, GEN or DAT Case, prepositions may also differ and as far as the clause is concerned, it can either be finite or non-finite. Clearly, this range of variability is not exhibited by all OE verbs but it has to be emphasised that the majority of OE verbs are characterised by dynamic complementation patterns. In this paper we intend to show that this dynamicity, inherent in OE complementation, is not captured by the existing classifications of OE verbs (cf. Section 2) and thereby indicate a need for a new classification. A set of desiderata for a new typology will be presented in Section 3. For the sake of clarity, the properties of OE verbal complementation will be illustrated here with a single verb swerian ‘to swear (an oath)’ but the conclusions presented in the paper are consistent with the findings following from a research on a much larger scale (cf. CharzyĔska-Wójcik in press, in prep. a., in prep. b).

2. Problems with Visser’s typology The existing classifications of OE verbs are those by Visser (1963–73), Mitchell (1985) and Ogura (1995). Since the latter two draw heavily on Visser, the paper will focus on Visser’s typology. Whenever the

146

Magdalena CharzyĔska-Wójcik

classifications diverge, comments and references will be made to the latter two.1 Let us now proceed to the analysis of the first example. (1)

Gyf man wat þæt oÿer man sweraþ, if one knows that other man swears he biÿ scyldig gyf he hit forhylÿ. he is guilty if he it conceals ‘If somebody knows that somebody else swears an oath, he is guilty if he conceals it.’ Lev; B8.1.4.3 (DOEC)

In (1) above swerian is not accompanied with an object of any kind, i.e. the structure is objectless. Let us see how to accommodate objectless clauses in Visser’s classificatory system. Visser’s (1963–73) typology of verbs draws a dividing line between verbs with complement and verbs without complement. While it is clear that the former are further subdivided according to the number and types of complements, complementless verbs at first glance seem to constitute a type in its own right. However, Visser (1963–73: §129), somewhat counter-intuitively, distinguishes as many as four types of complementless verbs. These are given in (2) below. (2)

a. “Self-sufficient intransitive” verbs which never appear with any kind of object, either direct, indirect, prepositional or causative. b. Intransitive verbs for which there exist etymologically related transitive homonyms. c. Absolute use of “intransitive verbs that are usually construed with an indirect object (in the dative) or with a causative object (in the genitive).” d. Transitive verbs which are used without the direct object they normally require, i.e. the absolute use of transitive verbs.

As can be seen, the above classification makes crucial reference to what other structures (if any) a given verb can appear with. In this sense the classification is dynamic since it takes into account the tendency of an OE verb to suppress an object. As a result, it is impossible to classify the example quoted in (1) above without making reference to alternative 1

The paper concentrates on the practical application of the classifications rather than on a general discussion of their advantages and disadvantages. For more theoretical comments concerning the available typologies of OE verbs, see CharzyĔska-Wójcik (in press, in prep. b).

The Benefits of Swearing

147

structures swerian can appear with. In particular, for a verb used absolutely to be classified as (2c), it must be attested with a single NP object either in the DAT or in the GEN Case. As far as (2d) is concerned, the formulation given above is equivocal: the absolute structure defined there can be interpreted either as resulting from the omission of the single ACC object or the omission of the ACC object which is one of the two objects. In principle, the absolute use of transitive verbs does not exclude ditransitive verbs. However, it transpires from Visser’s full definition given in §129 and §152 that the author does not intend the type to encompass ditransitives into type (2d). In §129 Visser talks about the “‘suppression’ of the direct object,” thus indicating that it is only the direct object that has been suppressed. In §152, where the author expands on the topic of the absolute use of transitive verbs,2 he says that “the object understood is so vague that the verb is virtually an intransitive one.” In view of these facts, it has to be concluded that Visser’s classification of the absolute use of transitive verbs does not include ditransitives, though, as we have indicated above, this exclusion is not a principled one: a transitive verb without complement can, theoretically, be understood as the absolute use of a transitive verb which can take either a single object in the ACC Case or two objects, one of which is direct. However, as is clear from Visser’s description of the type, it was not Visser’s intention to include the absolute use of ditransitives into (2d). This means that for a complementless verb to be classified as (2d), it must additionally be attested in a single-object structure with an ACC NP. Consequently, according to Visser’s typology, in order to classify the absolute use of a verb, one needs to contrast it with a single-object use. Following this logic, below we quote examples of swerian in single-object structures. (3)

2

a. Ne he aÿas ne swerige, ac not he oaths-ACC not should-swear but mid anfealdnysse sprece æfre unleaslice, with simplicity should-speak always truly ‘He should not swear oaths but he should always speak truly with simplicity.’ ÆLet 1 (Wulfsige Xa); B1.8.1 (DOEC)

For the inconsistencies present there see CharzyĔska-Wójcik (in press).

Magdalena CharzyĔska-Wójcik

148

b. Drihten gedeÿ ÿæt ÿu hæfst ælces godes Lord will-cause that you have of-each good genoh, swa swa he swor ÿinum fæderum. enough as he swore your fathers-DAT ‘The Lord will make sure that you have enough of everything just as he swore to your fathers.’ Deut; B8.1.4.5 (DOEC) As can be seen, swerian is accompanied with a single ACC object in (3a), i.e. aÿas ‘oaths’, while in (3b) the single object is a DAT NP, i.e. ÿinum fæderum ‘your fathers’. Since the classification of a single-object use of a verb is a necessary prerequisite for classifying its absolute use, let us now look at Visser’s (§315) classification of single-object verbs given in (4) below. (4)3

a. b. c. d.

indirect object4 (DAT) causative object (GEN) prepositional object direct object (ACC)

This classification, in contrast to the one presented in (1) above, is static and self-sufficient in the sense that it does not make reference to any alternative structures a given single-object verb can be attested with. In effect, (3a) classifies as type (4d), while (3b) represents type (4a). Note that the dual (or even multiple) membership of a verb is not in itself controversial: a verb may have a variety of different complementation patterns, which, depending on the approach, may receive various interpretations, i.e. either each complementation pattern may be classified separately or the whole range of complementation patterns can be viewed collectively in order to classify a verb which allows them. However, 3

Within these basic divisions Visser introduces further sub-classifications which are irrelevant for our discussion. 4 Visser (1963-73: §370) objects against the term “dative object”, “genitive object” and “accusative object,” emphasising that these are not syntactical terms referring “not to the function, but to the form of the object.” He considers these terms unsuitable “especially in a diachronic study of a language in which inflectional forms come and go, but functions remain.” Visser emphasises that “it is essential to bear in mind that the verbs in these constructions [i.e. with the objects in DAT and GEN Case] are intransitive.” Let us note however, that in this paper, despite Visser’s objections, we are going to use the terms DAT, GEN and ACC objects in a purely descriptive sense.

The Benefits of Swearing

149

Visser’s classification of complementless verbs (cf. 2 above) requires that a verb should be uniquely assigned to a single-object type before its membership in the complementless type can be established. Thus, the unavailability of a unique classification of a verb into a single-object type, while not problematic in itself, entails the inability of classifying the absolute use of this verb, which is a serious drawback of Visser’s classification. His classification of single-object verbs is not free from flaws either. Consider the following example: (5)

Þa axode he hi, hwæþer hit swa wære; when asked he them whether it so was þa sworon hi swiÿe, þæt hit swa wære then swore they strongly that it so was ‘When he asked them whether it was so, they swore eagerly that it had been so.’ LS 10.1 (Guth); B3.3.10.1 (DOEC)

Note that swerian appears with one object in (5) and the object is expressed by a clause: þæt hit swa wære ‘that it was so’. However, Visser’s typology of single-object verbs given in (3) above does not admit clausal objects thereby rendering not only the classification of examples like the one given in (5) impossible but it also precludes the classification of the absolute uses of verbs whose single object in alternative structures is expressed by a clause.5 This does not exhaust the inventory of problems posed for Visser’s classification by the complementation patterns available for the verb swerian as we have not yet discussed the double-object use of swerian. Note that so far we have seen swerian used absolutely or appearing with a DAT or ACC NP or a clause. The DAT NP expressed the Ĭ-role of Recipient (cf. 2b), the ACC NP and the clause expressed the Ĭ5

Ogura’s (1995) classification of OE verbs recognises verbs accompanied with “dependent clause” (§4.14) and lists verbs belonging to the type. However, the group encompasses both single-object verbs and verbs which take two objects, one of which is a clause. Mitchell’s (1985: §1090) classification recognises clausal objects but only in the double object use. It has to be emphasised, however, that this is a classification made for the sake of passivisation. Moreover, the way Mitchell classifies ditransitive verbs with clausal objects suffers from important drawbacks. In a nutshell, Mitchell not only recognises there a type of clausal ditransitives which does not exist but also presents as one type structures which should be kept distinct. These problems are discussed in detail in CharzyĔskaWójcik (2009, 2010) and CharzyĔska-Wójcik and Wójcik (2010).

Magdalena CharzyĔska-Wójcik

150

role of Theme (cf. 2a and 5 respectively). These two object types can cooccur in ditransitive structures with swerian, as shown in (6) below. (6)

Se gerefa [. . .] ÿa [. . .] þone aÿ him swor, the steward then the oath-ACC him-DAT swore swa he hyne sylf stafode, be hys sunu wifunge. as he himself dictated about his son’s marriage ‘Then the steward swore him (Abraham) an oath concerning his son’s marriage, as Abraham himself had dictated it.’ Gen; B8.1.4.1 (DOEC)

First, let us see how to accommodate examples of this type into Visser’s (§§676–708) classification of double-object verbs, given in (7) below. (7)

a. b. c. d. e. f. g. h.

indirect object + causative object: DAT + GEN direct object + causative object: ACC + GEN direct object + ablative object: ACC + DAT indirect object + direct object: DAT + ACC direct object + direct object: ACC + ACC indirect object + prepositional object: DAT + PP direct object + prepositional object: ACC + PP two prepositional objects: PP + PP

The example in (6) above with the DAT Recipient him ‘him’ and ACC Theme þone aÿ ‘the oath’ clearly classifies as (7d). However, though the above classification seems to be a straightforward listing of the attested co-occurrence of the available object types, it is, as evidenced by examples like the one in (8) below, not exhaustive. (8)

Þa swor se cyning þæm þæt he then swore the king that-one-DAT that he wolde hyne weligne gedon, gyf he hyne would him wealthy make if he him(self) wolde oncyrran from godes æ. would turn from God’s law ‘Then the king swore to him that he would make him wealthy if he would turn away from God’s law.’ Mart 5 (Kotzor); B19.5 (DOEC)

The Benefits of Swearing

151

In this example swerian is found with two objects: a DAT Recipient, i.e. þæm ‘that-one’ and a Theme expressed by a clause, i.e. þæt he wolde hyne weligne gedon ‘that he would make him wealthy’. The combination of a DAT NP and a clause is not recognised in Visser’s classificatory system of double-object verbs, as shown above. It has to be noted, however, that this omission does not follow from the fact that Visser does not classify the clausal Theme as an object. In further parts of the book, dealing with other aspects of syntax, Visser discusses the modally marked form in object clauses (§§868–873), the infinitive as object (§§924–925) and the form in –ing as object (§§1001–1008), which indicates that the omission of clausal objects in his classification of verbs is not a side effect of the way the author classifies the clausal constituents in the relevant strutures but of the way he classifies the verbs. In conclusion, we have tried to apply Visser’s system to classify six examples featuring the verb swerian. Examples (2a), (2b) and (6) can be classified in Visser’s typology, however, the result is a classification of different types of structures rather than the classification of a verb. This, as we indicated above, is not in itself controversial, though admittedly it does not give us insight into the nature of the verb. Three structures, however, cannot be accommodated into Visser’s system, namely the absolute use of swerian (cf. 1) and the two examples with a clausal Theme, i.e. the singleobject type (cf. 5) and the double-object type (cf. 8). In effect, it has to be concluded that Visser’s typology of verbs suffers from serious drawbacks. Some of them may reside in the way the (sub)types are formulated (for example the formulation of type (2d) and the omission of clausal objects), while others reflect genuine problems inherent in Visser’s approach (the assumption of unique slots for multi-structure verbs). While the problems were illustrated with a single verb only, it is worth noting that if we looked at more verbs, we would see more problems, especially as far as variability is concerned. For example, verbs which can appear with an NP revealing ACC, GEN and DAT Case expressing the same Ĭ-role (cf. wiþsacan ‘to deny, reject’), or verbs which appear with a PP which can be suppressed (cf. sorgian ‘to sorrow’), resulting in the absolute structure which cannot be classified in Visser’s system at all since Visser’s classification given in (2) above does not recognise the absolute use of verbs which take a PP object.6 In view of the above shortcomings we would like to propose our own way of accommodating the data given above into what seems to be the most straightforward classification of the 6

A more detailed discussion of the drawbacks inherent in Visser’s classification can be found in CharzyĔska-Wójcik (in press, in prep. b).

152

Magdalena CharzyĔska-Wójcik

verb swerian. This classification, as we intend to show, will imply a set of desiderata for a new classification of OE verbs.

3. A new proposal We have analysed six examples featuring the verb swerian: used absolutely (cf. 1), with a single object in the ACC (cf. 3a), in the DAT (cf. 3b) and expressed by a clause (cf. 5) and, finally, two instances of swerian in ditransitive use: with an NP-DAT and NP-ACC in (6) and an NP-DAT and a clause in (8). In other words, swerian can be accompanied with as many as two objects and as few as none. The two objects are: a Recipient, which is always a DAT NP, and a Theme, which is an ACC NP, clausal objects aside for the time being. These are the indiosyncratic properties of swerian and we suggest that specifying the maximum complementation range from the point of view of the idiosyncratic properties, i.e. Case and Ĭ-role is enough to properly capture the available complementation patterns exhibited by the verb swerian. The fact that OE verbs allow the suppression of either object or even of both (if two are available) is a typical feature of OE syntax, hence it should be accommodated into the system rather than into individual classifications. We intend to demonstrate that, taking this tendency of OE verbs into account, the mere specification of the available Ĭ-roles and Cases without specifying the categories of the objects will be sufficient to project the whole range of available structures. As far as the categories of the arguments are concerned, note that the Ĭ-role of Recipient can generally be expressed by an NP or a PP. The Theme, on the other hand, can be expressed by an NP, a PP or a clause. However, as the choice of the preposition is idiosyncratic, it has to be specified in the lexical entry of any verb and the specification of a preposition naturally implies that a PP is an available complementation pattern. Thus, without any further specification, we are left with NPs and clauses as potential categories expressing the Theme argument. These, however, need not be pre-specified either. Independent research (cf. CharzyĔska-Wójcik 2010, in prep. a) reveals that in OE a clausal argument always alternates with either an NP argument or (marginally) with a PP argument. Hence, a clausal argument never needs to be specified in the lexical entry and its availability in OE (unlike in ModE) can be derived from the presence of an NP (or a PP) object and is conditioned by the meaning of the verb. Therefore, taking into account the tendency of OE verbs to suppress their object(s), a verb specified for the four features given above, i.e. DAT Recipient and ACC Theme is

The Benefits of Swearing

153

expected to project the following range of structures, providing that the semantics of the verb is compatible with a clausal Theme: (9)

a. absolute use—if both objects are suppressed (cf. example 1); b. single NP-ACC Theme—if the Recipient is supressed (cf. example 3a); c. single NP-DAT Recipient—if the Theme is supressed (cf. example 3b); d. single clausal Theme—if the Recipient is supressed and Theme (in the case of swerian the content of the oath) is expressed by a clause (cf. 5); e. ditransitive use with an NP-DAT Recipient and an ACC Theme—if neither object is supressed (cf. 6); f. clausal ditransitive use with an NP-DAT Recipient and a clausal g. Theme—if neither object is supressed and the content of the oath in the case of swerian is expressed by a clause (cf. 8).

The crucial step in this reasoning is showing that the Case specified for the Theme can be realised either by an NP or a clause, i.e. that both can bear Case in Old English. Corroboration for this interpretation of OE clausal objects is provided by three types of arguments. First of all, convincing support for the claim that the clausal Theme is indeed Casemarked with the Case available for an alternative NP object is provided by the data connected with apposition, in particular with anticipation. Secondly, the analysis of passivisation of clausal ditransitive verbs in OE and their subsequent developments in ME indicate that clausal ditransitive verbs pattern with verbs assigning two Cases rather than one as far as passivisation is concerned. Additionally, the fact that in OE a clause is never the only object type but alternates with an NP object or (marginally) with a PP7 object indicates that a clause cannot be a pre-specified object type for OE verbs and has to be interpreted as based on some other structure. For reasons of space, however, we will limit our discussion here to the apposition as it is convincing enough to provide the necessary support for the interpretation of Case-marking of clauses pursued here. Before the data can be adduced, we will present the necessary facts connected with OE apposition.

7

Predominantly, in OE a clausal object constitutes an option to an NP object. In some instances there is yet another structure available, i.e. a PP. Only in some rather marginal cases is an NP not an attested object type, to the effect that a clausal object alternates with a PP object.

Magdalena CharzyĔska-Wójcik

154

3.1 Apposition in OE As reported by Mitchell (1985: §1428) after Pei and Gaynor (1954: 16), apposition is the use of paratactically joined linguistic forms occurring in the same clause or sentence which refer to the same referent and which have the same or similar grammatical form and function, but not the same meaning. Pei and Gaynor (1954: 16)

Mitchell (1985: §1437) further adds that the appositional elements exhibit concord and agree in number and Case as a rule and also in gender. As far as the form and function is concerned, Mitchell (1985: §1445) notes that a pronoun in the appropriate case, gender and number may anicipate a noun without or with qualifiers, [. . .] infinitives and noun clauses. [. . .] [T]he second appositional element may have been explanatory or emphatic, according to the intonation. (Mitchell 1985: §1445)

In sum, in OE all appostional elements share the same set of features by virtue of contracting the relationship of apposition and the appositional group as a whole is seen as expressing one grammatical relation in a clause. The following examples illustrate the point. (10)

a. He cwæÿ se apostol paulus. he-NOM said the apostle Paul-NOM ‘He, apostle Paul, said.’ ÆCHom I, 9; B1.1.10 (DOEC) b. he mid aÿe swor eowrum fæderum, he with oath swore your fathers-DAT Abrahame & Isaace & Iacobe. Abraham-DAT and Isaac-DAT and Jacob-DAT ‘He swore (with) an oath to your fathers, to Abraham, Isaac and Jacob.’ Deut; B8.1.4.5 (DOEC)

The Benefits of Swearing

(11)

155

a. ac hie gewearÿ þæt hie wolden but them was-agreeable that they would to Romanum friþes wilnian. from Romans peace-GEN desire ‘But they agreed that they would seek peace from the Romans.’ Or 4; B9.2.5 (DOEC) b. gif hi þæs wilniaÿ þæt him heora yfel if they that-GEN desire that them their evil unwrecen sie be ÿæs gyltes andefne unpunished should-be by the sin’s proportion ‘If they ask for it that they should not get their just deserts.’ Bo; B9.3.2 (DOEC)

The two examples in (10) illustrate appositional groups consisting of two NPs. In (10a) the NPs are he ‘he’ and se apostol paulus ‘the apostle Paul’. Both are singular NOM masculine NPs and both function as the subject of the verb cweÿan ‘to speak’, with the latter NP expanding on the meaning of the anticipatory pronun. The only way in which the two NPs can receive NOM Case in (10a) is by virtue of contracting the relation of apposition, where all elements share the same set of features.8 In (10b), in contrast, the two appositional NPs, i.e. eowrum fæderum ‘our fathers’ and Abrahame & Isaace & Iacobe ‘Abraham and Isaac and Jacob’ are in the DAT Case and they function as the Recipient object of the verb. Again, the verb can assign DAT Case only once to its object, yet, the object being appositional, all constituents entering apposition share the same Case. The example in (11b) illustrates the apposition of a pronoun and a clause. (11a) shows that the verb wilnian ‘to desire’ when accompanied with an NP assigns GEN Case to its object, i.e. friþes ‘peace’. As shown in (11b), wilnian can also be found with a clause which can (optionally) be anticipated by a pronoun in the appropriate Case. The ‘appropriate’ Case for wilnian is GEN, as evidenced by (11a). The pronoun and the clause stand in apposition and while it is clearly visible in (10) that both elements which form an appositional group exhibit the same Case, here only one of the two elements can overtly reveal Case. However, the relationship of apposition ensures that all memebrs of the appositional group share the same features, hence, it must be concluded that the subordinate clause in (11b) is GEN. The set shown in (11) demonstrates that the Case-assigning properties of a verb remain the same regardless of whether it is 8

An example of the same type of apposition, i.e. two NOM NPs can also be seen in (15) below, where he ‘he’ stands in apposition with Petrus ‘Peter’.

156

Magdalena CharzyĔska-Wójcik

complemented with an NP or a clause, which brings us to the starting point of this discussion, when we set out to show that the ACC Case specified for the Theme argument of swerian can be assigned not only to an NP but also to a clause, cf. (9d) and (9f). In view of the fact that in Old English all elements of the apposition share the same set of features, it now remains to demonstrate a clausal Theme with swerian in monotransitive use (cf. 9d) and in ditransitive use (cf. 9f) standing in apposition with an ACC (pro)noun to show that the clausal Theme in the relevant examples bears ACC Case. Consider the examples below. (12)

a. Soÿlice ÿu lædst ÿis folc on ÿæt land truly you lead this nation into the land ÿe Drihten swor ÿæt he syllan wolde which Lord swore that he give would hyra fæderum, their fathers ‘Truly, you lead the nation into the land which the Lord promised that he would grant to their fathers.’ Deut; B8.1.4.5 (DOEC) b. & eal se here on East Englum him and all the army in East England him-DAT swor annesse, þæt hie eal þæt swore unity-ACC that they all that woldon þæt he wolde, wanted what he wanted ‘And all the army in East England swore him unity: that they would fall in with his plans.’ ChronA (Plummer); B17.1 (DOEC)

In (12a) swerian is used monotransitively, with the object of the oath being the promised land, which is represented in the swerian-clause by the relative particle þe functioning as the Theme. Note that the Theme is further expanded on by the clause ÿæt he syllan wolde hyra fæderum ‘that he would grant to their fathers’. The two elements clearly stand in apposition, however, the relative pronoun appearing in (12a) is indeclinable, hence it does not give us any insight into the Case marking of the appositional group and we can draw no conclusions from this example as far as the Case of the clausal Theme is concerned. Importantly, however, the example illustrates apposition that obtains between the two

The Benefits of Swearing

157

constituents functioning jointly as a Theme complement of swerian. Apposition of the same type can be observed in the ditransitive use of swerian, illustrated by (12b). Fortunately, in (12b) apposition obtains between an element which is overtly Case-marked, i.e. an NP annesse ‘unity’ and the clause þæt hie eal þæt woldon þæt he wolde. As in (12a), the function of the clausal Theme is to expand on the meaning of the first appositional element and, since the first element is ACC,9 so is the second one. From here we can extrapolate the analysis of (12b) to examples of the type given in (12a), where neither element overtly reveals Case. While it would certainly be more convincing to provide an example of the monotransitive use of swerian with a clausal Theme which stands in apposition with an overtly Case-marked NP, providing a complete set of examples with one and the same verb is a task which almost invariably ends in failure, considering that apposition is optional in OE. However, if we did not restrict the set of examples to one verb, we would be able to provide the desirable monotransitive example. This allows us to generalise the conclusion transpiring from (12b) to (12a), namely that while (12a) only shows the apposition of the relevant type, (12b) also confirms the ACC Case-marking of the appositional group, thereby confirming the ACC Case marking of OE clausal objects. In this way, the data given in (12) above support our interpretation of the distributional dependence of clausal objects upon NP objects as following from the fact that the two object types are projected from the same lexical entry. Note that in OE it is an NP not a clause that overtly reveals Case hence the information concerning the Case available for the object can only be read off from an NP object. It can, however, be applied to NPs and clauses alike. If clausal objects alternating with NP objects were not marked for Case, they would have to be projected from a different lexical entry, notably an entry which specifies the morphological category of the object as a clause. If clausal objects were specified in the lexical entry of a verb there would be no way of accounting for the distributional dependence of clausal objects upon NP objects. In this way the distribution of clausal objects upon NP objects provides additional support for the fact that OE clausal objects are Casemarked. In conclusion, we have shown that the clausal Themes accompanying swerian bear ACC Case and that the whole range of structures available 9

The relevant NP in (12b) is morphologically ambiguous between ACC, DAT and GEN. However, as swerian in unambiguous examples with NP Themes is only attested with ACC, it is clear that the NP annesse in (12b) represents ACC. It is a standard procedure in historical studies to interpret ambiguously Case-marked examples against unambiguous ones.

158

Magdalena CharzyĔska-Wójcik

for swerian can be projected from a single lexical entry which specifies the Cases and Ĭ-roles. There is, however, one piece of data which we have not brought to light so far, namely the fact that there are examples of swerian found with a PP Recipient. While Recipients can, as has been mentioned, be expressed by PPs, the status of the PP Recipient with swerian is slightly dubious.

3.2 Prepositional Recipient The Toronto Corpus contains over a dozen examples with swerian in which the Recipient is expressed by a PP, as in (13) below. (13)

a. Þone swerigyndan aÿ þone he swor to the swearing oath which-ACC he swore to abrahame fædyre urum syllynde hyne us Abraham father our giving him us ‘The oath which he swore to Abraham, our father, in which he gave us to him.’ PsCaC (Wildhagen); C11.1 (DOEC) b. Iusiurandum quod iurauit ad abraham patrem nostrum daturum se nobis.

(13a) contains an ACC Theme expressed by a relative pronoun þone and a Recipient PP, i.e. to Abrahame fædyre urum ‘to Abraham, our father’. This example, however, is a translation of the Latin clause given in (13b) and, as is evident, the structure of the OE clause slavishly follows the Latin original. The PP Recipient in OE is modelled on the Latin ad abraham patrem nostrum and it does not seem to represent an active syntactic pattern available in OE for the verb swerian. This conclusion is prompted by the fact that all OE examples where swerian is accompanied with a PP Recipient are translations of the Latin phrase ad Abraham patrem nostrum from the various versions of the Canticles of the Psalter. When the Latin original expresses the Recipient by means of an NP, the OE rendering sticks to the Latin structure as well, as shown below:

The Benefits of Swearing

(14)

159

Innocens manibus et mundo corde qui non Unsceÿþynde hand & clænre heortan se ne innocent hands and clean heart which not acepit in uano animam suam nec onfeng on idylnysse sawle his ne he ne accepted in vain soul his nor he not iurauit in dolo proximo suo. swor on facne þam neahstan his swore in deceit the neighbor his ‘Innocent hands and clean heart which neither accepted his soul in vain nor falsely swore to his neighbour.’ PsGlC (Wildhagen); C7.1 (DOEC)

In view of the above, the PP Recipient in OE clauses with swerian cannot be interpreted as an active syntactic pattern. This conclusion is further strengthened by the fact that the PP Recipient, unlike the NP Recipient, is never found in clauses where the Theme has been supressed. Since examples exhibiting the suppression of either object are extremely well attested with swerian, the PP Recipient with swerian does not seem to be an intergral part of the system in OE. Moving on to the PP expressing the Theme argument, as has already been mentioned, Themes in general can be expressed by a PP. However, swerian does not seem to allow that option. Corpus search yields examples of swerian appearing with a PP containing the noun ‘oath’, which naturally suggests that we might be dealing with a PP Theme. However, the available examples represent the following type: (15)

Þa onsoc he, Petrus, sona eft then denied he Peter immediately again þriddan siÿe & mid aÿum swor þæt he third time and with oaths swore that he his ne wære. his not was ‘Then Peter denied for the third time and swore (with) oaths that he was not one of His apostles.’ HomS 24 (ScraggVerc 1); B3.2.24 (DOEC)

The PP mid aÿum ‘with oaths’ is best interpreted here as expressing manner, i.e. an Instrument not Theme, especially in examples as (15),

160

Magdalena CharzyĔska-Wójcik

where a clausal Theme is present. However, the relevant PP is also encountered in passages where it is found only with a Recipient (cf. example 10b above) and even there it is still best interpreted as an Instrument. As swerian is very frequently attested with PPs, representative examples will be discussed below in order to establish their status. (16)

a. for ÿam ÿe he is mildheort God; ne because he is merciful God not forlæt he eow, ne he eallunga will-abandon he you nor he assuredly ne adylgaÿ eow, ne he ne not will-destroy you nor he not forgyt his wedd, on ÿam he will-forget his covenant on which he swor eowrum fæderum. swore your fathers ‘Because He is mericful God, He will certainly not abandon you, nor will He destroy you or forget His covenant on which He swore to your fathers.’ Deut; B8.1.4.5 (DOEC) b. Ic swerie þurh eallum godum and gydenum I swear by all gods and goddesses þæt ÿu scealt geoffrian. oÿÿe ic þe that you shall offer-sacrifice or I you mid mislicum pinungum acwelle; with various torments will-kill ‘I swear on all gods and godesses that you will offer (sacrifice to our gods) or I will kill you with manifold tortures.’ ÆCHom I, 29; B1.1.31 (DOEC) c. Ic secge eow ne swerige ge I say you not should-swear you þurh nanum þinge by no thing ‘I tell you that you should not swear on anything.’ ÆCHom I, 32; B1.1.34 (DOEC)

The above examples contain prepositional phrases headed by on or þurh: on ÿam in (16a), þurh eallum godum and gydenum in (16b) and þurh nanum þinge in (16c), which, at least theoretically, can be interpreted as

The Benefits of Swearing

161

arguments of the verb swerian. However, the examples given in (17) below show conclusively that the PPs accompanying swerian should be interpreted as adjuncts. (17)

a. Ic on eared bad mælgesceafta, [. . .] I in native-land waited what-fate-would-appoint-me ne sohte searoniÿas, neither sought feuds ne me swor fela aÿa on unriht nor myself-DAT swore many oaths-ACC in injustice ‘At home I awaited what fate might come, I neither sought feuds I nor swore an oath falsely.’ Beo; A4.1 (DOEC) b. & he him geswor on his goda and he them-DAT swore by his gods’ noman þæt he ægþer wolde, ge þæt ærende names that he both would that message abeodan, swa swa hi hiene heton, ge announce as they him ordered and eac him þæt anwyrde eft gecyþan. also them the answer afterwards reveal ‘And he swore to them by his gods’ names that he would both announce the message to them, as they ordered him, and also that afterwards he would bring them the answer.’ Or 4; B9.2.5 (DOEC) c. ic swerige ÿe þurh mines rices I swear you by my kingdom’s mægna þæt swa hwæt swa ÿu on sæ power that whatever you at sea forlure ic ÿe þæt on lande gestaÿelige. lost I you that on land will-restore ‘I swear to you by the power of my kingdom that I will restore to you on land whatever you lost at sea.’ ApT; B4.1 (DOEC)

In (17a) swerian is found with DAT and ACC arguments and additionally the example contains the PP on unriht. Here it is quite clear from the meaning that the function of the PP cannot be that of an object of swerian but, more importantly, the presence of the two objects, i.e. the Recipient and the Theme excludes the possibility of an argument interpretation for the PP. In (17b), the DAT Recipient him ‘him’ and the

162

Magdalena CharzyĔska-Wójcik

clausal Theme þæt he ægþer wolde [. . .] function as arguments hence the on-PP naturally receives an adjunct interpretation. The same can be observed for the þurh-PP in example (17c): since swerian is found there with two constituents whose object status is unquestionable, the þurh-PP naturally receives an adjunct interpretation. Therefore, while the status of the PP in the examples in (16) could be seen as ambiguous between the adjunct and the argument, the presence of two objects in the examples in (17) disambiguates the status of the relevant PPs and it can be concluded that the arguments of swerian can indeed only be expressed by NPs or clauses. In particular, the Recipient in clauses representing the genuine OE syntax is always an NP and the Theme is either an NP or a clause. Having examined the available complementation patterns of the verb swerian, we can now return to the discussion concerning the information necessary to classify an OE verb.

3.3 Desiderata for a new classification We have shown that the verb swerian can appear in six basic types of structures representing variations of the two objects that are available for this verb. The range of attested structures is increased as a result of the tendency of OE verbs to suppress their object(s). In consequence, swerian is attested in two types of ditransitive structures, with the difference between them consisting in the category of the Theme, three types of single-object structures, representing the suppression of either of the available objects and exhibiting variation between an NP and a clausal Theme and, finally, the verb is also attested in absolute use. It seems reasonable that the six uses of swerian do not represent unrelated structures: they are intuitively felt to be different realisations of the same verb. If we pigeonholed each structure into a separate classification, the resulting picture would not reveal their underlying unity. Note that the classification of swerian proposed here captures this affinity. It does so not only by specifying the maximum complementation range while ignoring the categories of the objects but also by recognising the tendency of OE verbs to suppress their object(s) as one of the defining properties of OE verbal complementation. In this way it is capable of correctly predicting the available frames and captures the whole range of complementation in a simple and economical way. Importantly, it not only expresses the underlying affinity of all the six structures but it also avoids Visser’s problems by not classifying the post-suppression structures. Having shown an exemplary classification of one OE verb together with the benefits such a classification entails, we would like to extract a set of desiderata

The Benefits of Swearing

163

following from this classification for the general classificatory system of OE verbs. First of all, suppressions as an inherent feature of OE complementation should be removed from classifications. As a result, a classification of OE verbs should be based on the maximum complementation range including the available idiosyncratic variation without pre-specifying the categories of the object(s), which we have shown to be redundant in the sense of being derivable from the Case specification and the semantics of the verb. This means that complementless verbs, pace Visser, should be understood as verbs which never appear with an object of any kind rather than being subdivided into several types on the basis of the possible pre-suppression structures. The classification of single-object and double-object verbs should follow the same principles, i.e. they should ignore the categories of the object(s) and instead they should be based only on the idiosyncratic information, i.e. Case(s), Ĭ-role(s) and preposition(s). In this way the classifications will subsume clausal objects and will account for the fact that they are always an option to an NP (or a PP). Importantly, a classification based on such principles will accommodate the tendency of OE verbs to co-occur with more than one Case/object type without a change of meaning.

4. Conclusion The paper demonstrates the insufficiency of Visser’s classification of OE verbs. On the basis of the complementation properties of the verb swerian it has been shown that a classification assuming unique slots for non-uniquely behaving verbs fails to capture the inherent dynamicity of OE verbal rection. The dynamicity of OE verbal complementation concerns three aspects. First of all, it can be manifested by the varying number of complements. We have proposed that this type of dynamicity be removed from classifications, i.e. suppressions, some of which Visser’s typology takes into account (in particular when classifying complementless verbs), should be kept away from the typology of verbs. Secondly, OE verbs exhibit variation as far as the morphological categories of their objects are concerned. This parameter, as we have argued, need not be prespecified in classifications as far as variation between NPs and clauses is concerned—an omission which brings about two benefits, i.e. the resulting classification is not only more economical but it also accounts for the dependence of clausal objects upon NP objects. As for PP objects, since the choice of the preposition is idiosyncratic, it has to be specified individually for each verb. This specification naturally

164

Magdalena CharzyĔska-Wójcik

implies the availability of a PP object. Clearly, some verbs do while other verbs do not allow PP objects. The specification of the choice of the preposition for a particular verb will naturally set out verbs which allow PP objects as sole objects or as an option to NP objects. This is another aspect which will have to be accommodated into the classification of OE verbs. The third type of variation exhibited by OE verbs concerns the coexistence of more than one argument type with one and the same verb and more than one Case type being available for an object. Though this variability has not been illustrated in this paper (as it is not exhibited by swerian), it seems clear that it needs to be encompassed by the relevant classifications. The system whose preliminaries have been proposed here specifies the available maximum range of object types and it will naturally accommodate the co-existence of various Cases for one and the same object and the co-existence of a Case-marked NP or clause and a PP.

References Bosworth, Joseph and T. Northcote Toller. 1898. An Anglo-Saxon dictionary based on the manuscript collections of the late Joseph Bosworth, edited and enlarged by T. Northcote Toller. London: Oxford University Press. CharzyĔska-Wójcik, Magdalena. 2009. “On the non-existing type of clausal ditransitives in Old English.” Roczniki Humanistyczne LVII: 123–136. —. 2010. The syntax of Old English clausal ditransitive verbs. In Eugeniusz Cyran and Bogdan Szymanek (eds.), Verb structures: between phonology and morphosyntax,75–109. Lublin: Wydawnictwo KUL. —. in press. Reflections on Structural Variation in Old English Verbs. In Mirosáaw Pawlak and Jakub Bielak (eds.), New Perspectives in Language, Discourse and Translation Studies. Heidelberg/Dordrecht/ London/New York: Springer. —. in prep. a. “The ca(u)se for a clause.” —. in prep b. “Can we suppress more or less?” CharzyĔska-Wójcik, Magdalena and Jerzy Wójcik. 2010. On the necessity of original research. In Ewelina Mokrosz, Ewa Pająk and Sáawomir Zdziebko (eds.), Young Linguists in Dialogue. The 3rd Conference, 25– 40. Lublin: Wydawnictwo KUL. DOEC = Dictionary of Old English corpus in electronic form. Antonette di Paolo Healey (ed.), Distributed by the Oxford Text Archive. Mitchell, Bruce. 1985. Old English syntax. Oxford: Clarendon Press.

The Benefits of Swearing

165

Ogura, Michiko. 1995. Verbs in Medieval English: differences in verb choice in verse and prose. Berlin/New York Mouton de Gruyter. Pei, Mario and Frank Gaynor. 1954. A dictionary of linguistics. New York: Philosophical library. Toller, T. Northcote. 1921. An Anglo-Saxon dictionary based on the manuscript collections of the late Joseph Bosworth. Supplement. London: Oxford University Press. Visser, Frederick, Th. (1963–73) An historical syntax of the English language. Leiden: Brill.

PART FOUR SEMANTICS

THE POTENTIAL OF THE CONCEPTUAL INTEGRATION THEORY IN THE DESCRIPTION AND EXPLANATION OF SLIPS OF THE TONGUE PHENOMENA IZABELA SEKĝCIēSKA

1. Introduction Slips of the tongue have always attracted linguists’ attention because they are believed to reveal the workings and the fallibility of the human mind. They provide an insight into the structure of the mental lexicon and partly disclose discrepancies between speaker’s thoughts and their verbal realization. Though it is possible to induce slips of the tongue by means of scientific experiments, SOTs, which disrupt spontaneous speech, seem to be much more valuable. The assumptions discussed in the present paper are thus based solely upon speech errors which occurred naturally during various conversations held either in the English or in the Polish language.1 The aim of this article is to pinpoint major regularities found within verbal blunders and to ponder upon reasons for the occurrence of those linguistic faux pas. The introductory sections will constitute a brief review of major findings concerning SOTs. A taxonomy of verbal blunders, as well as a list of frequency factors determining the inducement of various slip-ups will be provided, and general features of SOTs will be discussed. Subsequently, a number of verbal blunders will be analyzed in accordance with Fauconnier – Turner’s Conceptual Integration Theory (Fauconnier and Turner 1998, Turner and Fauconnier 1995). It will be shown that two divergent theories, that is mental binding and Lamb’s Relational Network Theory (Lamb 1997, 1999) may be viewed as complementary and may act 1

Most of the SOTs are drawn from my own collection of verbal blunders, and some of them are adopted from Jeri J. Jaeger Adult Database (abbr. AD), available from http://linguistics.buffalo.edu/people/faculty/jaeger/adultSOT.html. A small percent of the English errors are found at http://hotword.dictionary.com/freudian-slip/.

The Potential of the Conceptual Integration Theory

169

in accord in the analyses of verbal blunders. It will also be underlined that there are two main factors facilitating the occurrence of SOTs, i.e. the influence of the immediate or maximal scope (Langacker 2008) upon the target unit and the inner speech imposition upon overt production.

2. Slips taxonomy There are three major groups of verbal blunders (Dell and Reich 1980, Fromkin, ed. 1980): 1. REPLACEMENT SLIPS – in which a target unit is substituted by another unit, which either comes from the target’s immediate or maximal scope or is related phonemically or semantically to the target unit. As a result, the following replacement slips could be distinguished: a) ANTICIPATIONS – when a unit, which is to appear later in a discourse is pronounced earlier, e.g. Can you help you? instead of Can I help you? b) PERSEVERATIONS – refer to errors in which a unit which has already been pronounced perseveres and replaces a nearby unit in the stream of speech, e.g. Sure, of sure – instead of Sure, of course. c) TRANSPOSITIONS – the exchange of positions between two units, e.g. sad back instead of bad sack (Dell and Reich 1980: 20). Metathesis, i.e. the switching of two or more adjacent sounds, and spoonerisms are examples of transpositions. d) SUBSITUTIONS – errors, in which a unit is replaced by another, whose origin is not certain, e.g. When I skate, I wear knee-pads, elpo-pads, and … - /p/ substituted for /b/ (Jaeger Adult Data: AD–192). 2. BLENDS – which incorporate part of the target unit and part of another unit, such as the name of a distracting object, a semantically related word, a synonymous or a competing alternate unit, etc. In consequence, the following verbal blunders may be listed (Fromkin, ed. 1980: 104): a) DISTRACTION BLENDS – appear when speakers are influenced by visual stimuli or their thoughts. Such a blend incorporates part of the distractor and part of the target word, e.g. a waiter jotting down the order of a beautiful woman writes sweet face instead of sweet tea (http://hotword.dictionary.com/freudian-slip/).

170

Izabela SekĞciĔska

b) SEMANTIC GROUP BLENDS – in which two units related semantically, but with opposite meanings, are mixed together, e.g. Don’t worry. The exercises are not easy … I mean not difficult. c) SYNONYM BLENDS – refer to a blend of two synonymous words, which have competed for a single syntactic slot, e.g. czwor – blend of “four” and its Polish equivalent “cztery.” d) COUNTERBLENDS – once a blend appears, the speaker tries to correct himself/herself and in a second attempt to produce a correct utterance makes another blend called counterblend: Can we avord it? I mean, can we affoid it? Target sentences: Can we afford it? Can we avoid it? (Dell and Reich 1980: 20) e) COMPETING PLAN BLENDS – combine fractions of two sentences structuring the same situation but in a different lexico-syntactic pattern: Imagine a room without walls (Imagine a room only with walls / Imagine a room without windows). f) FRAME BLENDING – I have added to the list yet another type of a blend, which involves partial integration of a sentence referring to the ongoing action with a sentence depicting a subsequent event: Put on the lights (I will switch on the light and you will put on your pyjamas). 3. HAPLOLOGIES – encompass such errors in which similar parts of the target unit are omitted and not pronounced: hipotamus instead of hippopotamus. Such SOTs being rather rare, they will not be discussed in details in this article.

3. SOTs frequency factors The presented types of errors may appear at different linguistic levels and units of different lengths may be involved in blundering. Replacement slips are the most common at the phonological and lexical strata, whereas blends usually occur at the syntactic and semantic level of a language. A verbal blunder may, on the surface, differ substantially from the target, but a closer scrutiny reveals its interconnection with the intended unit. A SOT may thus be phonemically or semantically related to the target (Fromkin, ed. 1980), or may bear traces of speakers’ extra linguistic environment or of their “inner languaging.” Noteworthy as it is, SOTs conform to the rules peculiar to a given linguistic stratum and are constrained by the grammar and the sound system of the language in question (Erard 2007). Still, a number of conditions conducive to the occurrence of verbal blunders

The Potential of the Conceptual Integration Theory

171

could be distinguished and according to Dell and Reich (1980: 21), they are as follows: 1. THE TYPE, LOCATION AND STRESS EFFECT – slips usually occur in stressed syllables and affect initial phonemes, and the most common type of a slip of the tongue is the anticipation error. 2. THE DISTANCE EFFECT – transposed units are rarely separated from each other by more than one word. 3. THE LOCATION SIMILARITY EFFECT – phonemes, syllables or words which undergo slipping usually occupy the same location as target units. 4. THE ITEM AND CONTEXT SIMILARITY EFFECT – similar units are more likely to be involved in a slip, as well as those units of which adjacent phonemes or words are akin to each other. 5. THE FREQUENCY EFFECT – a unit which is infrequently used and which is not ingrained in the speaker’s mental lexicon may be changed into a more common one, thus causing a verbal blunder. 6. THE LEXICAL AND SEMANTIC EDITING EFFECT – a slip is more likely to be made when its offshoot is a proper word or a semantically appropriate text rather than a pseudo-word. 7. FREUDIAN SLIPS AND THE SPEAKING RATE EFFECT – Freudian slips are believed to reveal suppressed thoughts of the speaker and the fast rate of speaking may increase the number of speech errors. In my opinion, however, this list should be extended by at least four additional factors: 8. THE STRESS EFFECT – in stressful situations people tend to produce more verbal blunders. 9. THE FOCUS EFFECT – the less focused we are, the more likely a slip is. 10. THE “GROUNDEDNESS” EFFECT – units and routine actions, which are well anchored in the cognitive warehouse, may be the reason for a slip. It is related to the frequency effect, but refers to action slips and slips resultant from the cultural scripts interference. 11. INNER SPEECH IMPOSITION – a great number of slips of the tongue occur as a result of the influence of mental language upon overt production. Inner speech pervades not only Freudian slips, but other verbal blunders as well, and its significance in SOTs will be discussed in the subsequent section of this paper.

Izabela SekĞciĔska

172

4. Lower strata SOTs As mentioned earlier, replacement slips are the most common type of verbal blunders at the phonological and lexemic level of a language. Let us thus consider a phonological substitution error and ruminate upon the reasons for its occurrence. During the campaign for the 2008 presidential election in the USA, the surname of the presidential candidate – Obama was frequently misspelled as Osama. In the vein of stratificational linguistics (Lamb 1999), the word Obama instigates the following linguistic network, which extends from the phonemic up to the lexemic level (Fig. 1): Obama

o-

a-

b-

LEXEMES

PHONEMES

m-

PHONONS Vo

Lb

Cl

Ba Lo

Ns

Fig. 1. Relational network of the surname Obama: Vo – vocalic, Lb – labial, Cl – closed, Ba – Back, Lo – low, Ns – nasal (adopted from Lamb 1999)

It is well attested that “[w]hen a given word is activated words similar in sound to it become activated by the spreading activation” (Dell and Reich 1980: 30). As a result, the surname Obama must have triggered the ignition of a linguistic network for Osama (Fig. 2), which won the race for final production, because it was probably better grounded in the speaker’s

The Potential of the Conceptual Integration Theory

173

cognitive system. Besides, those two surnames differ in only one phoneme and as Jeri Jaeger (2005) claims, similar words or phonemes are very likely to be involved in an error. Osama

s-

Obama

o-

Uv Rt Sp Vo Lb

a-

b-

Cl

m-

Ba Lo

Ns

Fig. 2. Relational network of the surname Obama and Osama: Uv – unvoiced, Rt – retracted, Sp – spirant, Vo – vocalic, Lb – labial, Cl – closed, Ba – Back, Lo – low, Ns – nasal (adopted from Lamb 1999)

Lamb’s Relational Network Theory seems to be highly conducive to the elucidation of phonological and lexemic slips of the tongue, as it shows the interconnection between units evoked by the target concept. What is more, it may feed the conceptual binding process, and provide a more detailed account of verbal blunders. The Conceptual Integration Theory entails the activation of two input spaces, one of which, in the case of SOTs, will embrace the target word and the other one will encompass the linguistic network triggered by the key unit (Fig. 3):

174

Izabela SekĞciĔska

Fig. 3. Phonological and lexemic SOT network

The generic space, which permeates the entire network and constrains projection across spaces, will comprise the phonotactic or syntactic rules peculiar to the language in question, whereas the blend will incorporate a slip of the tongue. The general space network for phonological and lexemic blunders and for the phonological Obama-Osama slip of the tongue could thus be presented as shown in Fig. 4:

The Potential of the Conceptual Integration Theory

Osama

s-

175

Obama

o-

Uv Rt Sp Vo Lb

a-

b-

Cl

m-

Ba Lo

Ns

Fig. 4. Obama-Osama SOT network

However, the Obama-Osama slip of the tongue may be considered as a semantic blunder as well. It should be remembered that this slip was uttered by the would-be president’s opponents. For them Obama was an enemy just like Osama Bin Laden, and that is why those two people may have been placed in the same conceptual packet on a par with each other, because they are cohyponyms, which may also have contributed to the unfortunate misspelling. Besides, what is worth noting is that the US president’s second name Hussein is also phonemically connected to another salient Middle-Eastern enemy of the US – Saddam Hussein. The semantic features of this slip may also be elucidated by means of conceptual integration theory, and the resultant frame network is represented in Fig. 5:

176

Izabela SekĞciĔska

Fig. 5. Frame network for the Obama-Osama SOT

Undeniably, the misspelling of the surname and the frequent emphasis upon Obama’s second name instigated negative associations and may have contributed to an unfavourable portrayal of the presidential candidate. As a result, the slip of the tongue may be viewed as an intentional act of the opposition, which could have aimed at disgracing their rival. However, the interpretation of a SOT depends, to a great extent, upon the context in which it occurs. Recently, during the breaking news that Osama bin Laden was killed by the US forces, Geraldo Rivera said, “Obama is dead, I don’t care” (available from: http://www.youtube.com/watch?vAUFWaYQk6Mg). This phonological SOT may also be regarded as a Freudian slip as it may reveal the journalist’s worries that the death of the Al-Qaeda leader could put the US president in danger of assassination.

The Potential of the Conceptual Integration Theory

177

5. The role of inner speech in SOTs A considerable number of SOTs are caused, however, by inner speech imposition. Our articulators do not keep up with our thoughts, and the words which are to appear later in the discourse leak out and fill in the syntactic slots which are, in fact, not assigned to them. Hasty speech with high information load and without a clear focus is particularly prone to such slips. Introducing a new coursebook to students, for instance, I was trying to express my thoughts as quickly as possible and I fell prey to the following faux pas: You’re going to learn a lot of practical things, for example, how to order a room in a restaurant.

The aforementioned slip arose as a result of merging two frame structures: FRAME 1: Booking a room in a hotel. FRAME 2: Ordering a meal in a restaurant.

The reason for this verbal blunder was probably inner speech imposition. My intention was to produce the following sentence: You’re going to learn a lot of practical things, for example, how to order a meal in a restaurant and how to book a room in a hotel.

According to Norbert Wiley (2006: 321–322), inner speech is devoid of the main parts of speech, and its syntax is extremely condensed and simplified. He claims that it is approximately “ten times as fast as outer speech” (2006: 322) because it relies upon visual imagery to a great extent and draws upon a number of extra-linguistic elements such as: tactile sensations, emotions, kinesthetics, smells, etc. To vocalize the above sentence, we do not have to think about each word separately. The visual glimpses of the situations unfold in front of our mind’s eye, the appropriate frames are elicited from the background knowledge and the vocabulary essential to fill in the slots of a given frame is extracted from our mental dictionary. Probably during the planning stage, two domains structured by the hotel and restaurant frame, were activated by inner speech and the slip was caused by the overimposition of linguistic structures, which were to depict the situations in question. What is worth noting, however, is that, the slot-filling constraints were not violated. Similarly to phonological slips of the tongue, which always conform to the

178

Izabela SekĞciĔska

phonotactic rules, frame blending slips do not violate grammatical constraints. Inner speech interference is also visible in competing plan blends when two sentences, structuring the same scenario, are merged into one. For instance, in a blend OK, just one [mÕmΩt] (Jaeger, AD–496) two phrases are combined: one minute, one moment, and in the blend Bobby needs a headache (Jaeger, AD–667), the following sentences are evoked to structure the same situation: Bobby needs an aspirin, Bobby has a headache. Probably during the process of inner languaging, a given situation evokes two sentences, which are assumedly blended into one during the slot-filling phase. The slip of the tongue must arise in the cognitive background and the articulators receive faulty instructions, which they carry out with perfect fluency. Therefore, Michael Erard (2007) rightly points out that SOTs are, in fact, slips of the brain because they are created in the upper level of linguistic processing. Extra-linguistic environment may also play a significant role in SOTs. Verbal blunders are frequently caused by distractors which are in the nearby surrounding of the speaker. These may be some objects or some actions which interfere with the flow of our thoughts: CONTEXT: A friend of mine (F) was standing in front of the mirror taking her coat off and talking with me. I: What about David? Did he get up? F: Yes, and he’s eating breakfast in front of the mirror,… oh, in front of the TV set. CONTEXT: A mother to her son, while stroking his head (her son frequently gets headaches), We need to get you a headache, uh, a haircut (Jaeger, AD–357).

Those SOTs are probably the result of blending inner speech with our outer speech. It is our inner speech which must attribute labels that are associated with objects in our surrounding and those names may crop up and intervene with overt production. Interestingly, these SOTs never violate grammatical constraints. In Polish, which is abundant in inflectional endings, the transposed or blended words are properly fitted into syntactic slots. Even when words from competing plans, which are evoked to structure a situation, are misplaced, they acquire proper endings and they occupy the place appropriate for the semantic category. Consider the example provided below: An elementary student translates a sentence She goes to college by car saying Ona dojeĪdĪa do koledĪu karem (She goes to college by car-em),

The Potential of the Conceptual Integration Theory

179

where the English word car attains the ending peculiar to a Polish instrumental case.

5.1 Freudian slips Undoubtedly, the interference of inner speech is especially noticeable in Freudian slips, which reportedly reveal hidden thoughts of the speakers. As an example, consider the following sentence and its linguistic context: CONTEXT: A girl and her boyfriend were talking about what animals they look like. Having been asked by her boyfriend which animal he reminded her of, the girl, after some cogitation, replied: If you weren’t a hippo, you’d be a bear (available from: http://answers.yahoo.com/question/index?qid=1006042602827)

The above Freudian slip epitomizes the imposition of inner languaging upon outer production and reveals complex processes, which take place in the cognitive background. The speaker of this sentence must have initially created a blended space, in which her boyfriend gains the attributes of a hippo (the shaded area of the diagram in fig. 6). Almost at the same time, key features of a bear have been brought to light and mapped upon the man. However, due to some interference, the linguistic network for the hippo, which was to remain hidden, slipped into outer speech probably because of a high level of activation. The cognitive configuration behind the slip may have acquired the following form:

Izabela SekĞciĔska

180

LIVING CREATURE

INNER SPEECH HIPPO

THE MAN

Role: THE MAN Vaule of the role: A HIPPO

BEAR

Role: THE MAN Value of the role: A BEAR

Role: THE MAN Value of the role: A HIPPO & A BEAR

OUTER SPEECH

Fig. 6. Frame network for the Freudian slip: If you weren’t a hippo, you’d be a bear

Let us consider another Freudian slip available from http:// hotword.dictionary.com/freudian-slip/: CONTEXT: A person, terrified of sharks, after a lot of coaxing and teasing finally decided to go swimming in the ocean. As she was splashing about in the water, a friend of hers played some rude joke on her and she pretended to be offended. However, instead of saying I’m shocked she said, I’m sharked.

The speaker of this sentence must have been thinking about sharks, and the highly activated network of a shark was implemented in a syntactic frame evoked by the target sentence. The two words shark and shock are phonemically related, so that the slip was even more likely to be made.

The Potential of the Conceptual Integration Theory

181

What is interesting, however, is that the noun shark received even the ending appropriate for adjectives, i.e. -ed so that the sentence could remain grammatically correct.

6. Conclusions The general explanation of SOTs is depicted in the network presented in figure 7:

Fig. 7. The SOT network

All the rules which constrain linguistic expressions in a given language are housed in the generic space, whereas the source space encompasses major SOT triggers. Under the influence of the source space constituents, the intended phoneme, syllable, word or a sentence undergoes transposition

182

Izabela SekĞciĔska

or alternation, and the final result of this process can be found incorporated within the blended space. Thus, phonological and lexemic blunders are most often influenced by the immediate constituents of the evoked linguistic network, whereas semantic SOTs are triggered by cohyponyms or distractors. Syntactic slips of the tongue, on the other hand, are usually triggered by competing plans or by schemas and cultural scripts ingrained in our cognitive system, whereas action slips are frequently caused by forward planning or distractors. Importantly, however, the rules intrinsic to the generic space permeate the entire network, and the blended space as well and the SOT always follow the patterns peculiar to the language in question.

References Dell, Gary S. and Peter A. Reich. 1980. Slips of the tongue: the facts and a stratificational model. In James Copeland and Philip Davis (eds.), Papers in Cognitive-Stratificational Linguistics. Rice University Studies, 19–34. Houston: Rice University Press. Erard, Michael. 2007. Slips, stumbles and verbal blunders, and what they mean. New York: Anchor Books. Fauconnier, Gilles and Mark Turner. 1998. “Conceptual Integration Networks.” Cognitive Science 22.2: 133–187. Fromkin, Victoria A. (ed.). 1973. Speech errors as linguistic evidence. The Hague: Mounton & Co. N. V. —. (ed.). 1980. Errors in linguistic performance: Slips of the tongue, ear, pen, and hand. New York: Academic Press, Inc. Jaeger, Jeri J. 2005. Kid’s Slips: What young children’s slips of the tongue reveal about language development. New Jersey: Lawrence Erlbaum Associates, Inc. Lamb, Sydney. 1997. “Bidirectional processing and expanded relational network Notation.” LACUS Forum 23: 109–124. —. 1999. Pathways of the brain: The neurocognitive basis of language. Amsterdam: John Benjamins B.V. Langacker, Ronald W. 2008. Cognitive Grammar: A basic introduction. New York: Oxford University Press. Turner, Mark and Gilles Fauconnier. 1995. “Conceptual Integration and Formal Expression.” Metaphor and Symbolic Activity 10.3: 183–203. Wiley, Norbert. 2006. “Inner Speech as a Language: A Saussurean Inquiry.” Journal for the Theory of Social Behaviour 36.3: 319–341.

The Potential of the Conceptual Integration Theory

183

Online sources Jaeger, Jeri J. “Adult Data.” www.linguistics.buffalo.edu/people/faculty/jaeger/adultSOT.html. DOA: 5 Jan. 2010. http://answers.yahoo.com/question/index?qid=1006042602827. DOA: 12 Nov. 2010. http://hotword.dictionary.com/freudian-slip. DOA: 3 Feb. 2011. http://www.youtube.com/watch?vAUFWaYQk6Mg . DOA: 6 May 2011.

THE RELEVANCE OF THE JUXTAPOSITION OF MEANINGS IN PUNS AGNIESZKA SOLSKA

1. Introduction The main focus of this paper is on puns, a linguistic phenomenon which depends for its existence on the juxtaposition of identical or similar forms and dissimilar meanings. The comprehender of a pun is forced not only to establish distinct meanings of a single linguistically ambiguous expression (the so-called punning element or the connector) but finds himself unable to choose between them and ends up going back and forth between them. Interestingly, this oscillating effect can be observed not only in the so-called double retentions puns, i.e. puns which carry two equally valid interpretations, such as (1), which owes its effect to the noun lie oscillating between two unrelated senses of “an intentional untruth” and “the position of a golf ball,”. It also extends to single retention puns, such as (2), in which the inadmissible meaning, instead of being abandoned, seems to linger in the interpreter’s mind providing counterpoise which makes it possible to juxtapose readings and which, effectively, creates the pun. For example, the phrase give a thumbs up in example (2) makes sense only when interpreted metaphorically as giving approval, but the pun relies for its rhetorical effect on the fact that it is mentioned in the context of hitch-hikers, who traditionally use a thumbs up gesture to solicit rides from oncoming vehicles. (1)

Being in politics is just like playing golf: you are trapped in one bad lie after another.

(2)

Most state laws don’t give a thumbs up to hitch-hikers.

In the present paper the theoretical framework provided by Sperber and Wilson’s (1986/1995) Relevance Theory will be applied to examine three issues connected with punning: first, to specify which elements of the

The Relevance of the Juxtaposition of Meanings in Puns

185

connector’s meaning are juxtaposed in puns representing different types of ambiguity, second, to present the mechanism behind the oscillating effect characterizing puns, and third, to show how the juxtapositions of meanings allow puns to achieve relevance understood in relevance-theoretic terms as providing the comprehender with sufficient cognitive benefits to justify the processing effort he has exerted.

2. Relevance theory and understanding verbal inputs Relevance Theory endorses the view that all ostensive, i.e. deliberately produced linguistic signals, whether spoken or written, are treated by the addressees as evidence of the meanings the communicator intends to convey. The process of identifying these meanings starts automatically, since as predicted by the so-called Communicative Principle of Relevance (cf. Sperber and Wilson 1995: 260), the hearer assumes that whatever has been presented to him in speech or in writing will be worth his attention. While establishing meanings the comprehender conducts two major activities: linguistic decoding and pragmatic inference. Decoding involves identifying mentally represented concepts associated with the words and expressions used in the utterance as well as identifying the logical form of which they are constituents. The concepts encoded by individual words and expressions give the comprehender access to the three types of information stored in the three entries of each concept. The lexical entry of a concept contains information about the lexical properties of the word or expression, i.e. its phonetic and grammatical properties. The logical entry specifies the inference rules which apply to logical forms of which that concept is a constituent. Finally, the encyclopedic entry of a concept gives access to the knowledge about the objects, events or properties that instantiate the concept and involves both folk and specialist assumptions, cultural beliefs and personal experiences which are stored in the form of propositional representations, scenarios or scripts and mental images. Now, mere recovery of lexically encoded concepts is not enough to work out the meanings of utterances since, as has been noted by numerous scholars (Bach 1994; Carston 2002a, 2002b; Récanati 2002), linguistically encoded meaning radically underdetermines speaker meaning. That is why in his attempt to figure out what is meant at word and utterance level, the addressee has to resort to pragmatic inference. For instance, on encountering an indexical, such as it in example (3), pragmatic reasoning has to be employed to work out what the pronoun refers to, and on encountering a lexically ambiguous word, i.e. one which encodes more than one concept, such as lie in example (1) above, the addressee has to

186

Agnieszka Solska

apply pragmatic reasoning to decide which of the word’s senses is the one intended by the addresser. (3)

Reading Kant shouldn’t be hard, it comes with Immanuel. [Immanuel/a manual]

Moreover, since speakers can entertain many more concepts than there are words in the language, linguistic expressions used in utterances often communicate diverse unlexicalised, occasion-specific concepts, whose meanings are narrower or broader than the concepts they encode lexically. For instance, while deriving the contextually appropriate meaning of the predicate give a thumbs up in the single retention pun in (2), the addressee will have to modify the category of things to which the predicate can apply so as to exclude people who use the gesture to indicate approval or to flag down a vehicle and to include institutions which authorize or disallow various activities by legislative means. However, interpreting utterances involves not only recovering the explicit content of the utterance. Equally crucial is determining its implicit content: its so-called strong implicatures, i.e. meanings which are emphatically implied by the producer and its so-called weak implicatures, i.e. wider possibilities of meaning that the addressee may conclude from the utterance. The heuristic which the hearer follows while working out meanings at word and utterance level, which is vindicated by the Communicative Principle of Relevance, involves formulating and testing hypotheses about the explicit and implicit content of the utterance in order of accessibility and stopping as soon as the first acceptable hypothesis has been found or giving up as soon as the hearer has found himself unable to find one (cf. Carston 2002b: 45). In doing so, the interpreter is guided by the Cognitive Principle of Relevance (cf. Sperber and Wilson 1995: 260), i.e. the tendency of the human mind to derive as many benefits for as little mental effort as possible. On the one hand, this so-called relevance-theoretic comprehension procedure stipulates that multiple meanings must be derived consecutively (though it does not preclude parallel processing of the meanings which have been derived); on the other hand, it predicts when it will need to be reapplied, namely when the assumptions making up the hearer’s cognitive environment give him reason to believe that the interpretation he has identified, though perfectly valid, does not give him the full range of benefits he can expect. In other words, when it does not meet his expectations of optimal relevance.

The Relevance of the Juxtaposition of Meanings in Puns

187

Now, in Sperber and Wilson’s model of utterance comprehension, relevance is seen as a function of two key factors, namely the mental effort the comprehender has to expend in order to interpret an utterance and the cognitive effects he can derive from interpreting that utterance. Relevance is thus a matter of degree, in that the more effort has to be exerted or the fewer benefits the utterance brings, the less relevant it is, and conversely, the less effort has to be exerted or the more benefits the utterance brings, the more relevant it is. Thus the discussion of the relevance of the juxtaposition of meanings in puns effectively boils down to establishing what factors contribute to making a pun more difficult or easy to process and what cognitive effects can be gained by processing a pun. Let us consider then what this model of utterance comprehension can tell us about the three issues we wish to address, i.e. the conceptual content of the punning element, the source of the oscillating effect and the relevance of punning utterances.

3. The concepts juxtaposed in the pun’s connector As has been said above, puns rely for their effect on using one linguistic form to link two dissimilar, i.e. two unrelated or very loosely related meanings. This is achieved in a rather straightforward way in the so-called horizontal, or syntagmatic puns, such as (5), in which the punning effect is the result of the pivotal expression surfacing more than once, on each occasion giving access to a different set of encyclopaedic assumptions and leading to a different set of cognitive effects, each of which has to be used while working out the meaning of the entire punning utterance. Such puns are sometimes described as explicit since the repeated occurrence of the connector makes the punning intention of the addresser evident. (4)

What to Expect When You’re Expecting. (Title of a book for expectant mothers)

A more subtle way of juxtaposing meanings can be observed in vertical, or paradigmatic puns, such as (1) – (3) above, in which a single lexicalization carries more than one meaning. Since the connector’s double meaning may escape the comprehender’s notice, such puns are sometimes described as implicit. As I argued elsewhere (Solska forthcoming), depending on the actual type of ambiguity exhibited by the connector, it is possible to distinguish

188

Agnieszka Solska

five possible pairings of concepts which the connector makes available and which, despite being distinct, share part of their content. One kind of pairing can be seen in puns based on homonymy, such as the vertical pun in (1) and the horizontal pun in (5), or on polysemy, such as the vertical pun in (6) and the horizontal pun in (4). In puns of this kind the comprehender is forced to juxtapose two distinct concepts encoded by a single lexical form, both of which have the same the lexical entry, but different logical addresses and which provide access to different sets of encyclopaedic data. (5)

We must, indeed, all hang together, or, most assuredly, we shall all hang separately. (Attributed to Benjamin Franklin, at the signing of the Declaration of Independence in 1776)

(6)

When it rains it pours. (The advertising slogan for Morton salt)

In puns juxtaposing the connector’s metaphorical and literal reading, such as the vertical pun in (7) and the horizontal pun in (8), or juxtaposing its idiomatic and literal reading, such as the vertical pun in (2) and the horizontal pun in (9), the concepts brought into contrast are also identical in terms of their lexical entries. However, only one of them is lexically encoded, the other one has to be inferentially constructed in the process of meaning adjustment, i.e. broadened and/or narrowed, to convey the unlexicalised concept(s) conceived of by the speaker. As a result, the logical entries of the two concepts overlap as do their encyclopedic entries. (7)

The Alpine Skiing competition started poorly and went downhill from there.

(8)

He’s too much of a chicken to kill a chicken.

(9)

If you don’t kick the habit, you’ll kick the bucket.

In puns based on homography, such as the vertical pun in (10) and the horizontal pun in (11), the two juxtaposed concepts are in fact encoded by different words, whose lexical, logical and encyclopedic entries are different by definition but whose graphic form happens to be the same or similar. (10)

Buff And Polish. (A headline of an article about Poland’s museums getting an overhaul)

The Relevance of the Juxtaposition of Meanings in Puns

(11)

189

I enjoy bass fishing and playing the bass guitar.

Puns based on perfect homophony, such as the vertical pun in (12) and the horizontal pun in (13), allow the comprehender to juxtapose concepts which share only a section of their lexical entries, namely the part specifying the phonetic form of their linguistic counterparts, but which have different logical and encyclopedic entries. (12)

When In Roam. (A headline of a newspaper article on the cost of making cell phone calls overseas)

(13)

Literacy Hour gives pupils the right to write.

Finally, in puns based on imperfect homophony, such as the vertical pun in (14) and the horizontal pun in (15), and in puns based on paronymy, such as the vertical pun in (16) and the horizontal pun in (17), the distinct concepts involved have even less in common than in the case of perfect homonyms, as they share only a fragment of the phonetic form of the words or expressions associated with them. (14)

Merchant of Tennis. (The name of a shop selling sports equipment)

(15)

Potable and portable? (Caption under a picture showing a 3-D rendition of the process of towing an iceberg in the North Atlantic so that it can be transported to countries where water is scarce)

(16)

I used to be a doctor, but then I lost patients.

(17)

Contraceptives should be used on every conceivable occasion.

It has to be noted that in vertical puns based on homography, in puns based on both perfect and imperfect homophony, and in puns based on paronymy, the punning element does in fact encode only one concept which, if the pun is to be detected, has to be contrasted with the concept encoded by another identically or similar sounding yet unexpressed “target” expression, which the hearer first has to identify. Thus the addressees of utterances (10), (12), (14) and (16) will only see the puns if they realize that the connectors polish, roam, tennis and patients serve as primes leading to the target words Polish, Rome, Venice and patience respectively.

Agnieszka Solska

190

4. The role of context in triggering the oscillating effect Obviously, it is not the potential ambiguity of the punning element that creates a pun. After all, though ambiguities are a pervasive feature of natural languages, in most cases they pass unnoticed as the hearer effortlessly selects or constructs the intended speaker meaning. What triggers the punning effect is the interplay of the assumptions brought to bear not only by the connector but also by another key element of a pun, a so-called disjunctor. Defined by Attardo (1994) as the element causing passage from one interpretation to another, the disjunctor may be linguistic or non-linguistic. A non-linguistic disjunctor can be observed in the advertising slogan in (18), where it takes the form of the visual material accompanying the punning advertising slogan, showing a car parked on a long driveway leading to a mansion house and foregrounding the “driveas-driveway” reading of the key word. A similar role can be played by the physical setting in which the pun appears, as is the case with punning names of shops or businesses, such as (14) above. (18)

The perfect car for a long drive – Mazda car. (Tanaka 1992: 77)

As for linguistic disjunctors, they may come before or after the connector forming part of the connector’s co-text. The former configuration can be seen in example (1), the latter in (19). (19)

The lecturer has exhausted both the topic and the audience.

A linguistic disjunctor can also coincide with the conjunctor, which is what happens in horizontal puns and in vertical puns in which, in addition to the linguistic material surrounding the pivotal expression, the conjunctor itself doubles up as the disjunctor and functions as a prime allowing the interpreter to identify the intended target. For instance, the pun in (14) will only be noticed by language users familiar with the title of Shakespeare’s tragic comedy The Merchant of Venice. In general, it would seem that the disjunctor’s function is to perform one of two jobs: (i) In vertical puns based on homonymy, polysemy or metaphor it forces the comprehender to perceive and retain the ambiguity of the connector, which linguistically encodes more than one concept. (ii) In vertical puns based on homography, homophony and paronymy it actively creates the ambiguity where none exists by allowing the

The Relevance of the Juxtaposition of Meanings in Puns

191

hearer to identify the “target” expression with which the encoded concept is to be contrasted. The first situation can be observed in the double retention pun (1), in which the concept POLITICS, encoded by the word politics, gives the addressee access to the encyclopedic information about the sphere of life connected with activities people resort to in order to gain power, which may include being “economical with the truth.” At the same time the presence of the concept of GOLF, encoded by the disjunctor word golf, gives the comprehender access to the encyclopedic information about a game called golf. Involving putting balls into holes in the ground, golf is an activity in which the position of the ball, i.e. its lie, can be a challenge to the golfer. The resulting semantic conflict makes it impossible to disambiguate the word lie and compels the addressee to assume that it encodes not one but two concepts which share the same lexical entry causing the utterance to convey two propositions containing these two concepts. Since both meanings are not only equally relevant but also equally prominent at the cognitive level, the interpreter has no reason to abandon either and ends up oscillating between them. The second situation can be seen in example (14), which is a name of a shop selling sports equipment and which activates in the reader’s mind the concepts pertaining to conducting business (via the concept MERCHANT) and the game called tennis (via the concept TENNIS). The name’s main function is to allow potential buyers to identify what kind of business is conducted on the premises. The last word of the name, i.e. the word tennis, sounds almost exactly like the word Venice, the final item in the title of a well-known play of Shakespeare’s. The play has nothing to do with either tennis or sports equipment and seems totally irrelevant but because of the high accessibility its title enjoys in the memory of most, if not all speakers of English, it is highly likely to get strongly activated in the reader’s mind triggering so many contextual effects that the comprehender will be unable to totally ignore it. As can be seen, it is the special conceptual make up of the punning utterance which causes the comprehender of a pun not only to derive more than one meaning of a single chunk of text but prevents him from (totally) rejecting either and, as Attridge put it (1988: 141), leaves “the reader or hearer endlessly oscillating in semantic space.”

192

Agnieszka Solska

5. How puns achieve relevance 5.1 Processing effort in pun comprehension Following the Cognitive Principle of Relevance, i.e. attempting to obtain the greatest number of benefits for the smallest amount of mental effort, the hearer expects the speaker to make the utterance as easy as possible to understand (within the limits of her abilities and preferences). Puns, however, often foil such expectations. There can be little doubt that interpreting an utterance whose full import depends on deriving two interpretations is anything but costly for the comprehender. It is not surprising then that language users do not always succeed in comprehending a pun and sometimes may even have trouble working out any of the meanings involved. This may happen if the comprehender lacks the required encyclopedic knowledge connected with the concepts involved. For example, the lack of encyclopaedic knowledge connected with the game of golf can make the pun in (1) hard to understand. Similarly, when a pun contains references to proverbs or titles of books and films, it may pass unnoticed if the addressee is not familiar with them. Problems may also appear at an even more basic level of utterance comprehension, i.e. the level of processing the acoustic or visual input. This is particularly striking in puns making use of non-standard spellings, such as (20), puns containing unexpected or unknown abbreviations, such as (21), or puns resorting to non-standard grammar, such as (22), when presented in writing. Though unlikely to totally defy the processing efforts of a native speaker, such puns definitely disrupt the normally smooth process of utterance comprehension since at first it is not clear how the punning element is to be integrated into the explicature of the utterance. (20)

Bingo hall worker B-10 and robbed. [B-10/beaten]

(21)

When the smog lifts in Los Angeles, UCLA. [UCLA/you see L(os) A(ngeles)]

(22)

Immanuel doesn’t pun, he Kant. [Kant/can’t]

Examples such as (20) – (22) highlight yet another difficulty awaiting the addressees of puns, namely parsing problems, which appear when the punning utterance exploits not only the lexical but also structural ambiguity. This kind of ambiguity underlies the pun in (23), in which the

The Relevance of the Juxtaposition of Meanings in Puns

193

alligator is understood to both end up with feathers in its mouth and to become sad. (23)

After he ate the duck, the alligator got a little down in the mouth.

However, recovering the utterance’s syntactic structure may also pose a problem when an addressee is presented with a pun exploiting a homophony between a polysyllabic word and a homophonous string of words. Again, this is unlikely to happen in horizontal puns, such as (24), in which the two predicates justify sin and just defy sin are presented to the hearer one after another, especially if they are presented in writing. Nonetheless, even if the channel of communication is oral, differences in pronunciation (different stress patterns, the actual realizations of the weak and strong syllables in each of the two conjunctor strings) will not lead to any parsing problems. (24)

Don’t justify sin, just defy sin.

The situation is different in vertical phrasemic puns, in which the role of the conjunctor is the same as in other puns based on homophony, described above, namely the conjunctor serves as a prime to be contrasted with an unexpressed target. Hempelman (2003) identifies three different possible connector-target configurations: (i) multi-word connectors mapped onto single-word targets, as in (25), (ii) single-word connectors mapped onto multiword targets, as in (26), and (iii) multi-word connectors mapped onto multi-word targets, as in (27). (25)

Which president was least guilty? Lincoln. He is in a cent. [in a cent/innocent]

(26)

Metaphors be with you! [Metaphors/May the force]

(27)

Why did the cookie cry? Because her mother has been a wafer so long. [a wafer/away for] (Pepicello and Green 1983: 59)

With the exception of example (27), which relies on linguistic knowledge alone, the essential factor enabling the hearer to perceive the pun is extralinguistic encyclopaedic knowledge. The comprehender who does not know the famous line May the force be with you from the Star Wars movies is unlikely to detect the pun in (26) and those oblivious to

194

Agnieszka Solska

the fact American one cent coins feature the profile of Abraham Lincoln’s face may fail to make any sense of utterance (25).

5.2 Cognitive gains Puns may be costly to process yet at the same time they allow the comprehender to derive a whole array of meanings which offset the extra effort. Consider again the punning fragment you are trapped in one bad lie after another in utterance (1). Once the undeterminacies connected with the intended meaning of such items as you, trapped, bad and especially lie have been resolved, what the utterance explicitly communicates is the two propositions listed below: (28)

(a) The speaker believes that politicians get repeatedly hampered because of the falsehoods they said with intent to deceive, which they cannot get away with. (b) The speaker believes that golf players get repeatedly hampered because of the positions of the golf ball they find hard to cope with.

Since we are considering the utterance out of context, it is hard to imagine what kind of strong implicatures (if any) it may trigger. However, supplying implicated premises such as the highly accessible assumptions given below: (29)

People who get repeatedly hampered may find it hard to succeed. In all spheres of life some people are better at tackling difficulties than others. Getting away with a lie can be difficult. People who lie cannot be trusted. Some people are better at lying than others. Tackling a difficult position of a golf ball requires skill.

Some golfers are better at tackling bad lies than others. would allow the addressee to derive a whole range of weakly implied meanings, some more determinate, some less, each of them connected with the different reading of the key word and all of them adding to the relevance of utterance (1). Some of these weak implicatures the comprehender may derive include those listed in (30a) and (30b):

The Relevance of the Juxtaposition of Meanings in Puns

(30)

195

(a) Being repeatedly caught lying can damage a politician’s career. To achieve success in politics depends on the ability to tell lies. Politicians cannot be trusted. (b) Being good at golf depends on learning how to tackle bad lies. Playing golf can be frustrating. Golf is not a game for everyone.

It would seem that the pun in (1) achieves relevance through a wide array of such weak implicatures, in fact two sets of weak implicatures, each set engendered by one of the meanings of the punning element. However, it might be argued that a pun, at least this particular pun, is more than a sum total of these meanings. As noted by Ariel (2010: 139), “formal (dis)alignments can help establish relevant connections.” When a communicator creates a pun he does so to encourage the addressee to view its two interpretations and the two sets of meanings they trigger as somehow relevant to each other simply because they are triggered by the dual sense of one expression. We can see this happen in (14), where by evoking associations with high culture the seemingly irrelevant name of a play creates the impression that the shop owners and anyone patronizing it is well-educated and refined. Thus juxtaposing the lowly game of tennis and the exalted title of a classic play seems to give the shop the aura of a place that serves the people who in addition to being physically active, are well-red and sophisticated. Similar connections can be perceived in (16), whose addressee will probably surmise that a doctor’s losing patients resulted from his losing patience, and there is no doubt that the addressee of (19) will perceive a connection between a speaker’s discussing every aspect of a topic and his wearing out his listeners in the process. As for example (1), it is quite unique in that its introductory sentence Being in politics is just like playing golf actually makes the connection between two domains explicit and steers the comprehender towards deriving some totally emergent meanings, such as the ones in (31), which increase the relevance of the pun even further: (31)

A successful politician must be good at lying. Politicians enjoy lying. For politicians lying is a game. Politics is a game.

None of these assumptions is directly stated by utterance (1) nor is any of them directly derivable from it, yet emergent meanings of this sort seem

196

Agnieszka Solska

to arise in interpreting some puns in much the same way as the emergent meanings that can be encountered in the comprehension of metaphors and complex concepts1. Unquestionably it requires talent to be able to find a word that can link two unrelated domains of meaning and to be able to contrive a context that will bring this link out and trigger such a plethora of meanings, and it is not surprising that puns of this sort are often perceived as witty. Though all puns bring together two meanings in one form, not all of them are characterized by commonality of meaning. In some the relation between the two meanings is best described in terms of incongruity or opposition, which helps explain the most easily perceived quality of puns, namely their humorousness. The view that the presence of incongruity is what makes humor possible has been voiced by quite a few researchers (Raskin 1985; Pollio 1996; Attardo and Raskin 1991). Summarizing the necessary and the sufficient conditions a text must meet in order to be perceived as humorous, Raskin (1985: 99), the founder of the Semantic Script Theory of Humor, contends that a text can be described as “a single-joke-carrying text” if it satisfies both of the following conditions: “(i) The text is compatible, fully or in part, with two different scripts. (ii) The two scripts with which the text is compatible are opposite.” The first condition is met by all puns, whose connectors make available two meanings which are either fully compatible with two scripts (as is the case with double retention puns) or partially compatible with two scripts (as is the case with single retention puns). The second condition is not met by all puns, which might explain why not all puns are perceived as funny. Specifying the kind of links that may constitute oppositeness Raskin (1985: 113) points to such binary categories as “real vs. unreal” (instantiated by actual/non-actual, normal/abnormal, and possible/impossible oppositions) and to concrete pairs of oppositions, such as DOCTOR vs. LOVER, seen for instance in his own example, given in (32): (32)

1

“Is the doctor at home?” the patient asked in his bronchial whisper. “No,” the doctor’s young and pretty wife whispered in reply. “Come right in.” (Raskin 1985: 117)

For the discussion of emergent meanings in metaphors see Vega Moreno (2004, 2007). For the discussion of emergent meanings in literal language see Franks (1995), Rips (1995), Hampton (1997) or Estes and Glucksberg (2000).

The Relevance of the Juxtaposition of Meanings in Puns

197

Both types of opposition can be observed in puns. For instance, in (33) we have an impossible scenario of drowning in a bowl of cereal after being pulled in by a strong piece of fruit set against a possible scenario of drowning in a body of water after being pulled in by a strong current. (33)

My friend drowned in a bowl of muesli. A strong currant pulled him in.

In (1) we find a pair of concrete opposing scripts POLITICS vs. GOLF and in (14) high culture of Elizabethan drama is contrasted with the mundane business of engaging in sports activities and selling sports equipment. On the other hand, no two opposing scripts are introduced by the “drive-as-driveway” and the “drive-as-travelling” reading of the key word in the witty but not funny pun in the Mazda ad in (18). Finally, it has to be noted that in some puns the two juxtaposed meanings exhibit neither commonality nor opposition of meaning. Depending on the conceptual content of the pun’s pivotal element, the connection between their two meanings may be seen as forced, unrealistic, nonsensical or non-existent and the only commonality to be found is often merely the commonality of form. This seems to be true of many puns used in newspaper headlines, such as (34) and in some jokes, such as (25), (27) or (35): (34)

Dutch Take Courage And Prepare For The Euro. (Headline of a newspaper article about the introduction of the Euro into the Netherlands)

(35)

The dolphins at the aquarium are having a whale of a time (not accidentally, but on porpoise).

In (32), Dutch courage, i.e. courage inspired by drunkenness, has nothing to do with the Dutch nation’s decision to introduce the new currency in their country. In (25), having one’s profile engraved on one cent coins is in no way linked to being innocent. Wafers, mentioned in (27), have nothing to do with being away, and while porpoises, mentioned in (35), have a lot to do with dolphins and whales, they can in no way be associated with doing things on purpose. Puns like these are often regarded with scorn and are perceived as puerile for the simple reason that they merely exploit the phonological (or graphic) similarity or identity of the ostensive signal and the meaning(s) they carry are often bizarre or absurd. Nonetheless, even puns of this sort exhibit two important qualities

198

Agnieszka Solska

which they share with other puns. On the one hand, they focus the attention of the addressee on the form of the message, which is what makes all kinds of puns easy to spot and helps explain why they are used in texts where standing out and being noticed is of the essence (e.g. advertising slogans, newspaper headlines or titles of books). On the other hand, like other puns they compel the addressee to exert extra effort to derive their meanings, which is makes whatever message they convey harder to forget.

6. Concluding remarks Puns perform a variety of different functions which are specific to the communicative setting in which they are used and which have been identified and discussed by numerous researchers. The poetic function of puns, particularly striking in works of literature, has received a lot of attention in literary studies (cf. Redfern 1984; Culler 1988; Delabastita 1993, 1997). Tanaka (1992, 1994) discussed at length the role of puns in advertising and Alexander (1997) concentrated on their role in newspaper headlines. Puns used as conversational witticisms have been discussed by Norrick (1984), while Chiaro (1992), Attardo (1994) and Yus (2003) offered numerous insights into the connection between puns and humor. Finally, the social dimension of punning has been the focus of attention of KosiĔska (2005) and has been noted by Crystal (1998). Presenting the findings emerging from this extensive research was beyond the scope of this paper, whose aim was to show how the juxtapositions of meanings allow puns to achieve relevance measurable in terms of processing effort and cognitive gains. Nonetheless, by exploring the nature of the difficulties involved in pun comprehension and the way cognitive effects of punning utterances are derived I hope to have shed some light on the kind of qualities puns possess, such as being memorable and eye-catching, and emanating humor and wit.

References Alexander, Richard J. 1997. Aspects of verbal humour in English. Tubingen: Gunter Narr. Ariel, Mira. 2010. Defining pragmatics. Cambridge: Cambridge University Press. Attardo, Salvatore. 1994. Linguistic theories of humor. New York: Mouton.

The Relevance of the Juxtaposition of Meanings in Puns

199

Attardo, Salvatore and Victor Raskin. 1991. “Script Theory revis(it)ed: joke similarity and joke representation model.” Humor 4.3/4: 293–347. Attridge, Derek. 1988. Unpacking the portmanteau, or who’s afraid of Finnegan’s Wake. In Jonathan Culler (ed.), 140–155. Bach, Kent. 1994, “Conversational Impliciture.” Mind & Language 9: 124–162. Carston, Robyn. 2002a. “Linguistic meaning, communicated meaning and cognitive pragmatics.” Mind and Language 17: 127–148. —. 2002b. Thoughts and utterances: the pragmatics of explicit communication. Oxford: Blackwell. Chiaro, Delia. 1992. The Language of jokes: analyzing verbal play. London: Routledge. Culler, Jonathan (ed.). 1988. On puns: the foundation of letters. Oxford and New York: Basil Blackwell. Crystal, Daid. 1998. Language play. London: Penguin Books. Delabastita, Dirk. 1993. There’s a double tongue: an investigation into the translation of Shakespeare’s wordplay, with special reference to Hamlet. Amsterdam and Atlanta: Rodopi. —. 1997. Traductio: essays on punning and translation. New York: St Jerome Estes Zachary and Sam Glucksberg. 2000. “Interactive property attribution in concept combination.” Memory and Language 28.1: 28–34. Franks, Bradley. 1995. “Sense generation: a ‘quasi–classical’ approach to concepts and concept combination.” Cognitive Science 19: 441–505. Hampton, James A. 1997. Emergent attributes in combined concepts. In Thomas B. Ward, Steven M. Smith and Jyotsna Vaid (eds.). Creative thought: an investigation of conceptual structures and processes, 83– 110. Washington D.C.: American Psychological Association. Hempelman, Christian F. 2003. Paronomasic puns: target recoverability. Towards automatic generation. http://homepage.mac.com/hempelma/puns.pdf. DOA: 15 Feb. 2011. KosiĔska, Katarzyna. 2005. Puns in relevance. In Aniela Korzeniowska and Maágorzata Grzegorzewska (eds.). Relevance studies in Poland, 75–80. Warsaw: University of Warsaw Press. Norrick, Neal. 1984. “Stock conversational witticisms.” Journal of Pragmatics 8: 195–209. Pepicello, William J. and Thomas A. Green. 1983. The language of riddles. Columbus: Ohio State University Press. Pollio, Howard R. 1996. Boundaries in humor and metaphor. In Jeffery Scott Mio and Albert N. Katz (eds.). Metaphor. Implications and applications, 231–253. Mahwah, New Jersey: Lawrence Erlbaum.

200

Agnieszka Solska

Raskin, Victor. 1985. Semantic mechanisms of humor. Dordrecht: Reidel. Récanati, Francois. 2002. “Does linguistic communication rest on inference?” Mind and Language 17: 105–126. —. 2004. Literal meaning. Cambridge: Cambridge University Press. Redfern, Walter. 1984. Puns. Oxford and New York: Basil Blackwell. Rips, Lance J. 1995. “The current status of research on concept combination.” Mind and Language. 10: 72–104. Solska, Agnieszka. forthcoming. “Relevance–theoretic comprehension procedure and processing multiple meanings in paradigmatic puns.” Sperber, Dan and Deirdre Wilson. 1986/1995. Relevance: communication and cognition. Oxford: Blackwell. Tanaka, Keiko. 1992. “The pun in advertising.” Lingua 87: 91–103. —. 1994. Advertising language: A pragmatic approach to advertisements in Britain and Japan. London: Routledge. Vega Moreno, Rosa Elena. 2004. “Metaphor interpretation and emergence.” UCL Working papers in linguistcs 16: 297–322. —. 2007. Creativity and convention: the pragmatics of everyday figurative speech. Amsterdam: John Benjamins. Yus, Francisco. 2003. “Humor and the search for relevance.” Journal of Pragmatics 35: 1295–1331.

CONCEPTUAL METAPHOR THEORY AS AN ISOLATING MODEL: A CASE FOR RE-CONTEXTUALIZATION ARIADNA STRUGIELSKA 1. Introduction Although context can be defined in a number of ways, each reflecting the theoretical commitment of the analyst (see Poulsen (2005: 69) for an overview), its importance was emphasized as early as in 1923, when a Polish anthropologist, Bronisáaw Malinowski, observed the fact that words take on particular meanings by virtue of the context in which they occur. Naturally, context can and should be viewed as a plethora of factors and conditions influencing a specific semantic interpretation. A number of these influences come from the language system itself in the form of, for instance, syntagmatic and paradigmatic relations. While the linguistic context can be a convenient starting point for a semantic analysis, it is not encapsulated from other related phenomena. Thus, according to Jensen (2007: 15), we can embed grammar in contexts ranging from “human cognition and knowledge, culture and society, and discourse and communication, presenting social code, individual system, and performance as equally integrated aspects of language.” Depending on the number of admitted perspectives and their relative significance in a particular model of language, Geeraerts (2003) distinguishes between isolated and integrated grammars. The former are those paradigms which tend to constrain the number of possible contexts influencing linguistic theorizing, while integrated approaches allow the researcher to “describe the entirety of language from discourse through lexis and culture to syntax and cognition” (Glynn 2004: 198). Contexts are thus competitive in nature by virtue of enabling alternative construals of a particular situation. Simultaneously, though, isolation and integration need to be seen as continua rather than dichotomous categories, and thus it could perhaps be more suitable to posit isolating and integrating paradigms. While the former are approaches in which a particular

202

Ariadna Strugielska

perspective is, often arbitrarily, taken as superior, the latter would advocate a dialogic methodology. In the following section, specific isolating tendencies and their instantiations in selected linguistic theories will be presented. The dimensions of isolation thus established will serve as a ground for evaluating standard Conceptual Metaphor Theory (henceforth also CMT) against the isolating/integrating continuum.

2. Isolation across linguistic theories It is often argued that Noam Chomsky’s generative grammar is the classic example of an isolating approach, thus providing an initial inventory of parameters according to which modular models can be defined (cf. Geeraerts 2003; Jensen 2007). Firstly, in Chomsky’s (1965, 1995) model, language is seen as an idealized autonomous faculty, and within the language system itself an important postulate refers to the primacy of competence over performance. Consequently, usage-based aspects of meaning are overlooked in favor of universal principles and parameters. Next, autonomous syntax, which is granted a unique status, is defined in isolation from context and hence language is a set of specifications for parameters in an invariant system of principles of universal grammar (Chomsky 1995). The system is thus a static model where idiosyncratic elements are ruled out for the sake of generalizations supported by introspective data. The inventory of dichotomies proposed within generative grammar can, to various degrees, be traced in a number of approaches. In addition, methodology-specific segregations can also be found. For instance, in structuralist semantics linguistic meaning is favored over encyclopaedic knowledge, onomasiological perspectives over semasiological potentials and langue over parole, while in Goldberg’s construction grammar, “clear constructions” are separated from those in which the idiosyncratic element is missing (cf. Goldberg and Casenhiser 2008). Naturally, exclusionary fallacies result in positing a specific profile determinant – a linguistic unit characterized by particular salience – which entails a localized view upon meaning (cf. Langacker 2008: 29). Profile determinants, as well as other linguistic units acknowledged as meaningful within a theory, can be represented by means of grounded or abstract semantic primitives, where the latter are characteristic of isolating approaches. However, even when an inventory of basic concepts is declared as anchored in, for instance, the conceptual system, its validity can be questioned, depending on the definition of cognition adopted. As

Conceptual Metaphor Theory as an Isolating Model

203

Langacker (2008: 31) rightly observes, “[a]dmitting that meaning resides in conceptualization does not itself solve anything but merely lets us formulate the problem.” Essentially, non-integrating approaches are built on a rather reduced definition of cognition, instead of adhering to its maximum designation (cf. Langacker 2008: 4), and consequently, they employ semantic primitives which are atomic, universal, topological and imagistic rather than complex, situated, functional and propositional (cf. Hampe and Grady 2005; Hanks 2006). An immediate consequence of adopting an abstract metalanguage is the fact that definitions of categories do not capture important aspects of meaning, i.e. features which are relevant from the human point of view. Thus, generalizations tend to violate the content requirement, whereby “the linguistic knowledge we ascribe to speakers should be limited to elements of form and meaning found in actually occurring expressions” (Langacker 2008: 25), and can therefore be classified as expert models (cf. Ungerer and Schmid 1996). As I have already argued elsewhere (SiekPiskozub and Strugielska 2010), expert taxonomies are characterized by perfect coherence relations imposed by the analyst. Instead, then, of taking into account a language user’s perspective, arbitrary intuitions are promoted (cf. Evans 2003; Sandra and Rice 1995). In the same vein, Tyler and Evans argue that “just because a linguist can come up with a highly elaborate, and indeed logically possible, semantic network for a particular lexical form does not entail that this is how language users represent the meanings associated with such forms” (2003: 40). Thus, a number of schematic concepts come to be interpreted as “a product of language – in probably both phylogenetic and ontogenetic terms” (Zlatev 2005: 14) once the extended embodiment position is adopted. Consequently, within a theory’s heuristic apparatus, representations à la Johnson’s (1987) image schemas need to be viewed as inventions rather than discoveries (cf. Hudson 2006: 22). Isolating paradigms can then be further characterized as those in which meta-categories are isolated terms, i.e. valid for out-of-context examples, whose par excellence instantiation is the notion of the prototype (cf. Geeraerts 2006a; Ungerer and Schmid 1996). Consequently, in an integrated model a number of meta-constructs pertaining to an isolating paradigm may need to be either altered or abandoned. The first tendency can be illustrated by the idea of principled polysemy proposed by Tyler and Evans (2003) or the concept of anthropocentric cognitive features posited by Hanks (2006), which have been developed in lieu of the notions of rampant polysemy and topological schemas respectively. The other propensity can be exemplified by the notion of the profile determinant in

204

Ariadna Strugielska

Cognitive Linguistics, which becomes an epiphenomenal category in a usage-based model (see Ungerer and Schmid 1996; Taylor 2002; Croft and Cruse 2004 and Langacker 2008 for discussion). Adopting atomic and static definitions of relevant terms entails concentrating on a limited number of data, which is another characteristic of isolating methodologies. For instance, various approaches to conventionality result in defining the basic unit of analysis, e.g. a construction, in alternate ways. Consequently, Langacker (2005: 140) notes that “[i]n Construction Grammar, a construction is recognized only if some aspect of it is unpredictable from its component parts or other, independently established constructions.” Thus, I love you would not be a construction in Goldberg’s (1995) model despite its high degree of entrenchment and familiarity within a speech community. All in all, as already signalled in Section 1, isolation and integration are tendencies rather than dichotomies. Consequently, a linguistic model should be carefully evaluated against the extent to which it displays the above-discussed parameters. In other words, “declared” integration needs to be verified at the level of data analysis in order to determine the predominant cline of a methodology. In the next section, I will argue that standard CMT is a principally isolating approach, and is thus more significantly related to Chomsky’s model than to second-generation cognitive linguistics (also: Cognitive Linguistics).

3. The context of CMT Conceptual Metaphor Theory is rather customarily situated within the framework of Cognitive Linguistics (henceforth also CL). For instance, Dirven (2005: 17–18) sees CL as a diverse paradigm with several ramifications, among which there is “a phenomenology-based strand […] explored by George Lakoff and Mark Johnson in the direction of ‘embodied realism’ […] Here belong prototype theory, lexical network theory, conceptual metaphor theory, and conceptual metonymy theory with inroads into cognitive pragmatics.” Simultaneously, Geeraerts (2006b: 3–6) argues that CL is a model permeated with meaning, which is perspectival, dynamic, non-autonomous, and grounded in the “experience of actual language use” (2006b: 6).The other common tendency is to place Conceptual Metaphor Theory against the background of metaphor theories, where, as Steen (2000: 262) observes, CMT presents itself as a rather revolutionary paradigm.

Conceptual Metaphor Theory as an Isolating Model

205

The consequence of the first classification is that CMT is expected to share the integrating tendencies of Cognitive Linguistics, while placing CMT within (selected) metaphor theories arbitrarily stresses its innovative character. In fact, the latter assumption of CMT has already been succinctly criticized (see, for instance, Haser 2005; Geeraerts 2010), and the conclusion is that CMT is far less ground-breaking than proclaimed by its founders (cf. Givón 2005: 75). The above criticism is assumed to be corroborated and extended if the isolating criteria developed in Section 2 can be validated in the case of CMT.

4. Conceptual Metaphor Theory as an isolating model In Section 2 above, it has been argued that isolating tendencies can be defined with reference to (at least) six interrelated parameters, whose presence within CMT will now be demonstrated. The thrust of this exposition should be seen against integration “declared” by the proponents of standard CMT. In other words, isolation in the model is to be demonstrated contrary to a usage-based propensity pronounced by, for instance, Kövecses (2000: 22), whereby metaphorical generalizations are assumed to be grounded in real linguistic occurrences (cf. Section 3), and thus “[i]t can be safely suggested that most of this language is the normal conventional way of talking about [for instance] anger in English.”

4.1 Dichotomies in CMT To begin with, the theory of cognitive metaphor perpetuates a number of dichotomies, e.g. “use/concepts, form/function, objective/experiential, [and] literal/figurative” (Nerlich 2007: 95), and among which the source/target distinction is foundational and implies that the source concept is “typically a better understood and more concrete domain than the target domain” (Kövecses 2008: 381). However, judgments about the level of abstractness of a domain are reflections of the analyst’s competence. In other words, a particular linguistic expression may well be classified as belonging to alternative source domains (cf. Glynn 2002; Stefanowitsch 2006). Thus, as Haser (2005: 154) concludes, “[i]n resorting to metaphorical phrases that allow only a single interpretation […] Lakoff commits the very mistake he accuses objectivist philosophers of.” A conventional/novel distinction is another illustration of how declared integration is violated at the level of empirical studies. The proponents of

206

Ariadna Strugielska

CMT “say that a metaphor is highly conventional or conventionalized (i.e. well established and deeply entrenched) in the usage of a linguistic community. Since there are both conceptual metaphors and their corresponding linguistic expressions, the issue of conventionality concerns both conceptual metaphors and their linguistic manifestations” (Kövecses 2002: 30). Conventionality, however, is defined in a minimalist way since “well established” and “deeply entrenched” are isolated parameters, not supported by “the analysis of large amounts of data and the collection of numerous examples in order to determine whether an individual metaphorical expression is simply a one-shot metaphor, or is part of a systematic, frequently-occurring pattern, and therefore conventional” (Semino et al. 2004: 6). Hence, the effects of frequency and productivity, fundamental in a usage-based model, are ignored in CMT. To illustrate, the concept of ANGER is comprehended via 24 source domains (cf. Stefanowitsch 2006: 74, 76). Of these, as many as 10, i.e. A SUPERIOR, AGGRESSIVE ANIMAL BEHAVIOR, A BURDEN, INSANITY, A CAPTIVE ANIMAL, A PLANT, GORGE, A SLEEPING ORGANISM, A DISEASE, A SHARP OBJECT, are instantiated by maximum six distinct expressions. Hence, there is a measurable difference between, for instance, ANGER IS AN OPPONENT IN A STRUGGLE, instantiated by 22 different linguistic patterns, and ANGER IS A SLEEPING ORGANISM, with just two combinations. However, in CMT both truly conventional and rather unmotivated mappings are classified as productive conceptual metaphors. The conventional/novel dichotomy at the level of linguistic examples is problematic as well since, while CMT researchers declare that they concentrate on conventional, i.e. literal expressions (Lakoff and Johnson 1980: 5), this is not confirmed at a closer scrutiny. For instance, Deignan (2005: 95) observes that in Yu’s (1998) analysis of the ANGER IS HEAT mapping, some linguistic examples are less frequent, i.e. more novel than others. To conclude, the dichotomies proposed within CMT demonstrate that the model upholds “a rather antiquated distinction similar to how some view the distinction between langue (convention, supraindividual, static) and parole (dynamic, individual, discourse)” (Nerlich 2007: 97). Most importantly, however, the notions proposed within CMT are inconsistently applied within the paradigm itself, which may well indicate that they are inventions rather than discoveries (cf. Section 2).

Conceptual Metaphor Theory as an Isolating Model

207

4.2 CMT and profile determinance The previous section concentrated on highlighting a number of dichotomies inherent in CMT. As argued above, arbitrary oppositions entail that a particular construct is viewed as the profile determinant. In CMT, content words establish the meaning of a metaphorical expression, while conceptual mappings are privileged at the representational level. With reference to the first type of dominance, “semantics [is] called on to designate, through abbreviation, lexical semantics alone” (Ricoeur 2004: 121). Consequently, content words are seen as more salient than function items and morpho-syntactic patterns. For instance, in “She was engulfed by panic,” “My mother was tormented by fear,” or “His actions were dictated by fear” (Kövecses 2000: 23) only the words in italics determine the activation of the source concepts, while the semantic contribution of, for instance, the prepositional phrases is overlooked. The other type of profile determinance to be observed in CMT is related to the sense-imposing role of generalizations. In other words, conceptual metaphors seem to provide an interpretative lens through which the meaning of linguistic expressions should be seen since “people’s tacit knowledge of conceptual metaphors, such as ANGER IS HEATED FLUID IN A CONTAINER, partly motivates how they make sense of idiomatic phrases like blow your stack and flip your lid” (Gibbs et al. 1997: 141). To illustrate, the FEAR IS COLD mapping, as proposed by Stefanowitsch (2006: 81), facilitates the understanding of icy/cold fear or land of cold and fear against the domain of “negative valuation” (Kövecses 2000: 44). In the same vein, Cortazzi and Jin (1999: 165) argue that the LANGUAGE IS CLOTHES metaphor explains the following linguistic expression: “One size fits lots of different people and you can stretch it to fit your own shape or identity.” Moreover, it is argued that the aspect emphasized by the source domain, or the main meaning focus of metaphor (Kövecses 2002: 110), is cognitively motivated since it reflects the primary human concern related to a given source and is supported by the majority of linguistic expressions for a particular metaphor. However, studying re-contextualized examples of mappings reveals aspects which may well constitute alternatives to CMT interpretations. This can be illustrated on the basis of BNC citations of cold fear: (1) (2)

Cold fear filled every heart. (existence) As he dressed for dinner in his room, Dorian remembered what he had seen and cold fear ran through him like a knife. (existence/duration; availability to senses)

208

(3) (4) (5) (6) (7)

Ariadna Strugielska

The icy fear which showed in the older man's eyes cut through Vologsky's mild concern like a knife, chilling him to the bone. (existence/duration; availability to senses; control) A cold fear gripped Connon's stomach. (control) She struggled with the cold fear that had laid its hand on her. (control) Nuadu did not move, but the icy fear closed about him again. (control) Jack grunted, conscious only of the icy fear that gripped him. (availability to senses; control)

While the feature of negative valuation, proposed by CMT, may well transpire in the above examples, it can hardly be denied that the aspects implied in brackets are evoked as well. The above phenomenon, clearly related to the notion of perceived metaphoricity discussed in Deignan (2005) and Hanks (2006), could have a further, more fundamental implication. Namely, if profile determinance cannot be felicitously ascribed to conceptual mappings, it may well be the case that metaphor as a mode of semantic representation is in itself unfounded. In fact, there is no conclusive experimental evidence which demonstrates that a conceptual metaphor is indispensable for interpreting metaphorical language, particularly its conventional instantiations. For instance, Stern (2000: 187) argues that “it is far from evident that the ‘mappings’ to which the theory appeals are specifically metaphorical, [and] that the principles that govern [the] mappings are really different from those of opposing accounts.” The redundancy of conceptual metaphors is also discussed by Givón (2005: 80), who argues that in the example: They were much younger than us, but still they raced from youth to senility in record time, “race from […] to” automatically activates ‘motion in space’, while “much younger than us” triggers the node ‘time’, and the phrase “from youth to senility” activates ‘age’. “‘Motion’, ‘space’ and ‘time’ are thus all automatically activated in the metaphoric usage without any need to invoke a conceptual metaphor.”

4.3 CMT and its metalanguage If CMT were an integrating model, its conceptual primitives would be non-atomic, adaptive, and relevant from the human perspective (cf. Section 2). Instead, two sets of basic concepts can be distinguished within CMT. The first one is specifically proposed by Kövecses (2000) to discuss the domain of emotions, while the other is a general set of image-schemas, as stipulated by the Invariance Hypothesis (Lakoff 1990: 39).

Conceptual Metaphor Theory as an Isolating Model

209

With reference to the former inventory, Kövecses (2000: 40–47) posits that the main meaning focus of emotion concepts should encompass the following aspects: existence, intensity, passivity, control, evaluation, difficulty, desire/need, nonphysical unity, progress, and harm. At the same time, the linguist admits that “there is a clear set of correspondences between Wierzbicka’s categories of semantic primitives and the aspects of emotion that are focused on by emotion metaphors” (Kövecses 2000: 47). Importantly, Wierzbicka’s (1972) Natural Semantic Metalanguage includes a number of abstract constructs which belong to “an idealistically isolated realm of language, cut off from the extralinguistic world” (Geeraerts 2010: 134). Thus, the above-noted correspondences between CMT and Wierzbicka’s model imply the cognitively unmotivated status of the metalanguage of the standard, i.e. Lakoffian, metaphor approach. The other set of basic concepts within CMT, image schemas, are topological elements preserved in the mapping. However, as Jäkel (2002: 37–38) convincingly argues, “the inventory of image-schemata is anything but agreed upon, and […] not all image-schemata are […] clearly delineated in their internal structure […]. [Thus], the ‘invariance hypothesis’ is not an empirical hypothesis whatsoever. The Cognitive Theory of Metaphor would be better off without it.” In the same vein, Pawelec (2006: 47) notes that the list of gestalt-like units proposed by Lakoff (1987: 267) is unbounded since any time a new communicative situation presents itself, a new basic element of meaning can in fact be posited.

4.4 CMT as an expert category In the previous Section, I have observed that the inventory of aspects proposed by CMT cannot be limited in any motivated way. Likewise, Haser (2005: 234–235) argues that, due to the effect of family resemblances, the expression win an argument could be potentially linked to the source domains of WAR, GAME-PLAYING, or A JOURNEY. In the same vein, Ritchie (2003) claims that the ARGUMENT IS WAR mapping is only one of the possible formulations that could capture relations between domains underlying linguistic metaphors apparently pertaining to wars or arguments. Other fields include: chess, fencing, boxing or shouting matches. Consequently, as Cameron (2003: 62) observes, in a family-resemblance approach to metaphor, “[w]e are no longer counting instances of something cut-and-dried.” Thus, by arbitrarily attributing

210

Ariadna Strugielska […] actual metaphors to their more ‘generic conceptual’ metaphors […] [Lakoff and Johnson employ the method] of the logician-philosopher. The vast collection of out-of-context metaphors, together with their classifiers, are obtained through the study of competence rather than of performance. And as such, it makes intuitive sense in much the same way as syntactic deep structure or transformations made sense in another ubiquitous framework that relied primarily on the native’s proverbial intuition (Chomsky 1965). (Givón 2005: 79)

In other words, conceptual metaphors are expert taxonomies, by which they resemble other isolated categories, e.g. constructions or lexical units. More specifically, the postulate to discuss conceptual metaphors in terms of scientific taxonomies is based on the following characteristics of the model. Firstly, as already argued above, mappings are not derived from linguistic examples and hence the central attributes proposed for the metaphor are imposed rather than discovered. Furthermore, the onomasiological salience of basic level categories is overlooked. Instead, superordinate-level constructs seem most prominent although, as emphasized by members of the MetNet project (2006), “[i]t is by no means obvious that this level and type of categorization is appropriate in social sciences research concerned with the thinking and attitudes of individuals as members of socio-cultural groups. Important detail may disappear in abstracting too far from the actual words used.” Moreover, the expert-like nature of metaphor categories transpires from the fact that the taxonomies proposed are internally complex to the effect that the grouping is neither mind-sized nor mind-oriented (cf. Ungerer and Schmid 1996). Thus, conceptual metaphors are probably best defined as categories which “linguists live by” (cf. Wierzbicka 1986).

4.5 Isolated theoretical and linguistic categories In the preceding sections, I have demonstrated that four isolating tendencies, as defined in Section 2, are clearly observable in CMT. The remaining two parameters, i.e. isolated meta-categories and limited linguistic instantiations can also be detected in CMT, and have in fact already been alluded to above. For instance, the source/target dichotomy implies constructs which can only be maintained if a particular set of data is considered. Thus, on the one hand, CTM is important for investigations into the phenomenon of figurative language because it provides the researcher with a well-developed metalinguistic apparatus, including heuristically significant concepts such as source domain, target domain, metaphoric model, conceptual mapping,

Conceptual Metaphor Theory as an Isolating Model

211

conceptual correspondence, metaphoric entailment, etc. (Dobrovol’ skij and Piirainen 2005: 29)

On the other hand, though, the meta-categories proposed by CMT are most felicitous in the case of novel metaphorical expressions (see, for instance, Dobrovol’skij and Piirainen 2005; Tendahl 2009). Rather paradoxically, then, the activation of conceptual metaphors, mappings or entailments is not necessary for the interpretation of conventional linguistic metaphors – the primary evidence which CMT evokes to support the validity of conceptual mappings. The final parameter of isolation distinguished in Section 2, i.e. the lack of systematicity in both selecting and analyzing the data, can be illustrated by the notion of metaphorical linguistic expressions (Kövecses 2002: 251) and metaphorical patterns (Stefanowitsch 2006: 66). The former are instantiated by: (8) (9) (10) (11) (12) (13) (14)

He’s without direction in life. I’m where I want to be in life. She’ll go places in life. I’m starved for affection. He thrives on love. We had to cheer him up. She lit up.

On the first definition, it seems, linguistic metaphors are either single content words or their arbitrarily delimited combinations. What seems of particular relevance in examples 8 – 14 above is the little and unsystematic semantic contribution of function words, e.g. prepositions and particles (cf. Section 4.2). At the same time, however, Kövecses (2000: 75), discussing: She was consumed by passion or I am burning with emotion, as instances of the EMOTION IS FIRE/HEAT mapping, claims that “[i]t is the prepositions with and by that indicate that there is a causal link between certain emotional responses and emotion as fire.” In other words, the preposition which highlights the main meaning focus of the metaphor is excluded from the linguistic example. To recapitulate, in metaphorical linguistic expressions the distinction between content and function words is not systematic. Hence, the degree of schematicity of a unit and the ensuing amount of context needed for its interpretation are seen as irrelevant. Moreover, the problems already discussed elsewhere (see, for instance, Deignan 2005 or Steen 2007), i.e. the fact that the linguistic examples are predominantly invented by the researcher and encompass both novel and conventional expressions as well

212

Ariadna Strugielska

as the observation that numerous linguistic expressions are ambiguous, should not be overlooked. The other example of the language of metaphors are patterns derived from naturally-occurring data. Undoubtedly, Stefanowitsch’s proposal solves the problem of ambiguity inherent in Kövecses’ idea, since the target word is necessarily included within the expression, e.g. He shot down all my arguments (see Stefanowitsch 2006: 67). At the same time, however, metaphorical patterns seem to perpetuate a number of unmotivated generalizations already present in metaphorical expressions. In order to illustrate the idea, let us look at the juxtaposition of four sets of metaphorical patterns for the domain of fear, extracted from Stefanowitsch (2006: 81–83), and their full realizations arrived at via a web search using the search engine Google: Set 1 Fear fill X e.g. An unreasoning fear filled him, his heart rate had doubled Fear torment X e.g. The fear tormented me but I couldn’t let go. Fear occupy X’s mind e.g. Think of a fear that currently occupies your mind Fear weigh heavily on X e.g. This fear weighed heavily on her. Set 2 Fear peak e.g. Fear peaked during the stock panic in October and November. Fear rise e.g. Bird flu fear rises in Indonesia's Bali with another patient. Set 3 Fear evaporate e.g. Fear and terror are starting to evaporate in Iraq. X conquer fear e.g. If they can conquer the fear of public speaking. X curb fear e.g. GDP figures curb fear of interest rise. Set 4 Fear outweigh X e.g. We can't let intimidation and fear outweigh our commitment to justice. X fill Y with fear e.g. Nicole Kidman says acting fills her with fear. X feed fear e.g. If we feed fear for a lifetime, fear becomes very strong. X’s actions be dictated by fear e.g. It becomes apparent that much of an individuals actions are dictated by fear of judgment. X be driven away by fear e.g. As soon as people are trained, they leave the province–driven away by fear and poverty. X‘s face be frozen in fear e.g. At the center’s inaugural Clean Planet conference, one of the delegates is found dead, his face frozen in fear.

Stefanowitsch admits that the patterns “are presented in a form that is somewhat abstracted from the actual citations: verbs are shown in the infinitive, slots for participants are shown as X or Y, and similar patterns are collapsed into compact form using slashes for alternatives and

Conceptual Metaphor Theory as an Isolating Model

213

parentheses for optional elements” (2006: 73). This is evident in the pattern/citation juxtaposition above. However, the level of generalization proposed by Stefanowitsch has important consequences for the meaning of metaphorical language since a number of phenomena pertaining to the semantic pole of a linguistic unit are overlooked. As illustrated by Set 1 above, metaphorical patterns are abstracted in a way which renders grounding elements, e.g. determiners, meaningless. In Set 2, another grounding component, i.e. verbal inflections, is neglected, while Set 3 illustrates that metaphorical language abstracts over the semantic contribution of phrase structures within which fear is situated. Likewise, in Set 4 metaphorical patterns hide the semantic details provided by syntactic constructions, e.g. the impact of embedding or subjectification. In other words, metaphorical language in the forms of patterns posited by Stefanowitsch (2006) considerably departs from the definition of linguistic expressions as symbolic units proposed in integrating approaches, e.g. Cognitive Grammar, in that metaphorical patterns do not take into account the semantic contribution of numerous linguistic categories. Thus, they are not “revealed,” and hence, metaphorical patterns may well constitute inventions rather than discoveries.

5. Conclusion Conceptual Metaphor Theory is often placed within the context of Cognitive Linguistics – a predominantly integrating paradigm. However, in view of the fact that a systematic analysis of CMT at the level of both theoretical assumptions and empirical studies reveals a number of isolating tendencies within the model, it has been argued that the cognitive theory of metaphor should be seen in the context of modular approaches rather than dialogical and competitive methodologies. This re-contextualization of CMT has important consequences for the descriptive terms proposed within the approach, including source and target concepts, mappings and entailments, or the main meaning focus, whose limited applicability has now become transparent. In the same vein, metaphorical language may need to be discussed in a new, integrated context, which will ultimately result in the necessary reinterpretation of such ubiquitous phenomena as headship or conventionality.

214

Ariadna Strugielska

References Cameron, Lynn. 2003. Metaphor in educational discourse. London and New York: Continuum. Chomsky, Noam. 1965. Aspects of the theory of syntax. Cambridge Mass.: MIT Press. —. 1995. The minimalist program. Cambridge Mass.: MIT Press. Cortazzi, Martin and Lixian Jin. 1999. Bridges to learning: Metaphors of teaching, learning and language. In Lynne Cameron and Graham Low (eds.), Researching and applying metaphor, 149–176. Cambridge: Cambridge University Press. Croft, William and D. Alan Cruse. 2004. Cognitive Linguistics. Cambridge: Cambridge University Press. Deignan, Alice. 2005. Metaphor and corpus linguistics. Amsterdam and Philadelphia: John Benjamins Publishing Company. Dirven, René. 2005. Major strands in Cognitive Linguistics. In Francisco J. Ruiz de Mendoza Ibáñez and M. Sandra Peña Cervel (eds.), Cognitive Linguistics: Internal dynamics and interdisciplinary interaction, 17–68. Berlin and New York: Mounton de Gruyter. Dobrovol’skij, Dmitrij O. and Elisabeth Piirainen. 2005. “Cognitive theory of metaphor and idiom analysis.” Jezikoslovlje 6.1: 7–35. Evans, Vyvyan. 2003. The structure of time: Language, meaning and temporal cognition. Amsterdam: John Benjamins. Geeraerts, Dirk. 2003. Decontextualizing and recontextualizing tendencies in 20th century linguistics and literary theory. In Ewald Mengel, HansJorg Schmid and Michael Steppard (eds.), Anglistentag 2002 Bayreuth, 369–379. Trier: Wissenschaftliger Verlag. —. 2006a. Words and other wonders. Berlin and New York: Mounton de Gruyter. —. 2006b. Introduction: A rough guide to Cognitive Linguistics. In Dirk Geeraerts (ed.), Cognitive Linguistics: Basic readings, 1–28. Berlin and New York: Mounton de Gruyter. —. 2010. Theories of lexical semantics. Oxford: Oxford University Press. Gibbs, Raymond W., Jody Bogdonovich, Jeffrey Sykes and Dale Barr. 1997. “Metaphor in idiom comprehension.” Journal of Memory and Language 37: 141–154. Givón, Talmy. 2005. Context as other minds. Amsterdam and Philadelphia: John Benjamins Publishing Company. Glynn, Dylan. 2002. “Love and anger: The grammatical structure of conceptual metaphors.” Style 36.2

Conceptual Metaphor Theory as an Isolating Model

215

http://www.findarticles.com/p/articles/mi_m2342/is_3_36/ai_9477563 0/?tag=content;col1 (accessed 18 April 2009) —. 2004. “Constructions at the crossroads: The place of Construction Grammar between field and frame.” Annual Review of Cognitive Linguistics 2: 197–233. Goldberg, Adele E. 1995. Constructions: A Construction Grammar approach to argument structure. Chicago: University of Chicago Press. Goldberg, Adele E. and Devin Casenhiser. 2008. English constructions. http://www.uni-mainz.de/FB/PhilologieII/fb1414/lampert/download/so2008/English%20Constructions.pdf (accessed 2 October 2010) Hampe, Beate and Joseph. E. Grady (eds.). 2005. From perception to meaning: Image schemas in Cognitive Linguistics. Berlin: Mouton de Gruyter. Hanks, Patrick. 2006. Metaphoricity is gradable. In Anatol Stefanowitsch and Stefan Th. Gries (eds.), Corpus-based approaches to metaphor and metonymy, 17–35. Berlin and New York: Mounton de Gruyter. Haser, Verena. 2005. Metaphor, metonymy, and experientialist philosophy. Berlin and New York: Mounton de Gruyter. Hudson, Richard. 2006. Word Grammar and Construction Grammar. http://www.phon.ucl.ac.uk/home/dick/papers/texts/wg&cg.pdf (accessed 17 November 2010). Jäkel, Olaf. 2002. “Hypotheses revisited: The cognitive theory of metaphor applied to religious texts.” metaphoric. de 02: 20–42. http;//www.metaphorik.de/02/jaekel.pdf (accessed 12 December 2010). Jensen, Kim Ebensgaard. 2007. A construction-based study of English predicators. Odense: University of Southern Denmark PhD dissertation. http://static.sdu.dk/mediafiles//Files/Information_til/Studerende_ved_S DU/Din_uddannelse/phd_hum/afhandlinger/2007/JensenKimEbensgaa rd.pdf (accessed 15 February 2011) Johnson, Mark. 1987. The body in the mind: The bodily basis of meaning, imagination and reason. Chicago: University of Chicago Press. Kövecses, Zoltán. 2000. Metaphor and emotion: Language, culture, and body in human feeling. Cambridge: Cambridge University Press. —. 2002. Metaphor: A practical introduction. Oxford and New York: Oxford University Press. —. 2008. Metaphor and emotion. In Raymond W. Gibbs (ed.), The Cambridge handbook of metaphor and thought, 380–396. Cambridge: Cambridge University Press.

216

Ariadna Strugielska

Lakoff, George. 1987. Women, fire, and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press. —. 1990. “The Invariance Hypothesis: is abstract reason based on image schemas?” Cognitive Linguistics 1.1: 39–74. Lakoff, George and Mark Johnson.1980. Metaphors we live by. Chicago and London: The University of Chicago Press. Langacker, Ronald. W. 2005. Construction Grammars: cognitive, radical, and less so. In Francisco J. Ruiz de Mendoza Ibáñez and M. Sandra Peña Cervel (eds.), Cognitive Linguistics: Internal Dynamics and Interdisciplinary Interaction, 101–162. Berlin and New York: Mounton de Gruyter. —. 2008. Cognitive Grammar: A basic introduction. Oxford: Oxford University Press. Malinowski, Bronisáaw. 1923. The problem of meaning in primitive languages. Supplement to Charles Kay Ogden and Ivor Armstrong Richards (eds.), The meaning of meaning: A study of the influence of language upon thought and the science of symbolism, 451–510. London: Routledge and Kegan Paul. MetNet Project. 2006. Theoretical and methodological challenges in using discourse data. A research project funded by ERSC National Centre for Research Methods. http://www.education.leeds.ac.uk/metaphor/aims.htm (accessed 20 February 2011). Nerlich, Brigitte. 2007. “A Review of Zoltán Kövecses. 2006. Language, mind and culture: A practical introduction.” metaphoric.de 12: 94–99. http:// www.metaphorik.de/12/rezensionkovecses.pdf (accessed 18 January 2011). Pawelec, Andrzej. 2006. Metafora pojĊciowa a tradycja [Conceptual metaphor and tradition]. Kraków: Universitas. Poulsen, Sonja. 2005. Collocations as a language resource: A functional and cognitive study in English phraseology. Odense: University of Southern Denmark PhD dissertation. http://static.sdu.dk/mediafiles//Files/Om_SDU/Fakulteterne/Humanior a/Phd/afhandlinger/2005/Afhandlinger%2048_s_poulsen%20pdf.pdf (accessed 12 January 2011). Ricoeur, Paul. 2004. The rule of metaphor: The creation of meaning in language. London and New York: Routlege Classics. Ritchie, David. 2003. “ARGUMENT IS WAR – Or is it a game of chess? Multiple meanings in the analysis of implicit metaphors.” Metaphor and Symbol 18: 125–146.

Conceptual Metaphor Theory as an Isolating Model

217

Sandra, Dominiek and Sally Rice. 1995. “Network analyses of prepositional meaning: Mirroring whose mind – the linguist’s or the language user’s?” Cognitive Linguistics 6.1: 89–130. Semino, Elena, John Heywood and Mick Short. 2004. “Methodological problems in the analysis of metaphors in a corpus of conversations about cancer.” http://www.lancs.ac.uk/staff/eiaes/Cancer_metaphors.pdf (accessed 23 November 2010). Siek-Piskozub, Teresa and Ariadna Strugielska. 2010. “Metaphors as signs: between cognitive attitudes and cognitive standpoints.” Interstudia: Cultural Spaces and Identities in (Inter)Action. Alma Mater Bacău 6: 288–299. Steen, Gerard. 2000. “Metaphor and language and literature: a cognitive perspective.” Language and Literature 9.3: 261–277. http://vunl.academia.edu/GerardSteen/Papers/309968/Metaphor_and_language _and_literature_A_cognitive_perspective (accessed 17 January 2011). Steen, Gerard J. 2007. Finding Metaphor in Grammar and Usage. Amsterdam and Philadelphia: John Benjamins Publishing Company. Stefanowitsch, Anatol. 2006. Words and their metaphors. In Anatol Stefanowitsch and Stefan Th. Gries (eds.), Corpus-based approaches to metaphor and metonymy, 63–105. Berlin and New York: Mounton de Gruyter. Stern, Josef. 2000. Metaphor in context. Cambridge, MA: MIT Press. Tendahl, Markus. 2009. A hybrid theory of metaphor: Relevance Theory and Cognitive Linguistics. Basingstoke: Palgrave Macmillan. Wierzbicka, Anna. 1972. Semantic primitives. Frankfurt: Athenaeum. —. 1986. “Metaphors linguists live by: Lakoff and Johnson contra Aristotle. (Review of George Lakoff and Mark Johnson.1980. Metaphors we live by.” Papers in Linguistics 19.2: 287–313. Taylor, John R. 2002. Cognitive Grammar. Oxford: Oxford University Press. Tyler, Andrea and Vyvyan Evans. 2003. The semantics of English Prepositions. Spatial scenes, embodied meaning and cognition. Cambridge: Cambridge University Press. Ungerer, Friedrich and Hans-Jörg Schmid. 1996. An introduction to Cognitive Linguistics. London: Longman. Yu, Ning. 1998. The contemporary theory of metaphor: A perspective from Chinese. Amsterdam and Philadelphia: John Benjamins Publishing Company.

218

Ariadna Strugielska

Zlatev, Jordan. 2005. What’s in a schema? Bodily mimesis and the grounding of language. In Beate Hampe and Joseph E. Grady (eds.), From perception to meaning: Image schemas in Cognitive Linguistics, 313–342. Berlin and New York: Mouton de Gruyter.

JUXTAPOSITION AND DIFFERENTIATION WITHIN THE METAPHORICAL SENSES IN OE FYR AND LEOHT CONCEPTS AGNIESZKA WAWRZYNIAK

1. Introduction The aim of the present paper is to analyse the concepts of fire and light as mental objects and socially constructed ones. The analysis of these concepts is vital for light and fire carried great cultural significance for the Anglo-Saxons, which is reflected in the rich lexicon in the field. The present paper will analyse the concepts of fire and light on two planes, cultural and semantic, which are closely intertwined and should, by no means, be viewed as separate criteria. These concepts, imbued with religious as well as secular undertones, were indispensable for the people in the Anglo-Saxon period. They were believed to be equipped with mysterious, divine forces and constituted the core of the Universe. The aims of the present paper are the following: firstly, the paper will illustrate varieties of metaphors instantiated by the domains of fire and light. Thus, the initial aim is to discuss different types of metaphors as reflected by the conceptual nature of the respective concepts. The metaphors evoked by the domain of light will reflect highly ennobling qualities, such as nobleness, divinity and magnificence. The metaphors, however, instantiated by the domain of fire will be grouped along positively and negatively loaded semantic poles. Hence, light and fire could be assumed to constitute a kind of dualism between the directed towards heaven, divine and ennobling light, on the one hand, and more human-like and linked with earthly world fire, on the other. Light was associated for the Anglo-Saxons with the ideal, the belief and the dream, whereas fire was frequently perceived as the symbol reflecting the nature of humans. It seems that the need to draw on dualisms and to conceptualise and categorise the world in terms of the binary opposites is entrenched in our minds and also reflected in the nature surrounding us.

220

Agnieszka Wawrzyniak

For this reason, Hertz, approaching the vivid and ever-present concept of dualism claims: All the oppositions presented by nature exhibit this fundamental dualism. Light and dark, day and night, east and south in opposition to west and north, represent in imagery and localise in space the two contrary classes of supernatural powers: on one side life shines forth and rises, on the other it descends and is extinguished. The same with the contrast between high and low, sky and earth [ . . .]. The same contrast appears if we consider the meaning of the words ‘right’ and ‘left’. The former is used to express ideas of physical strength and ‘dexterity’, of intellectual ‘rectitude’ and good judgment, of ’uprightness’ and moral integrity, of good fortune and beauty, of juridical norm; while the word ‘left’ evokes most of the ideas contrary to these. (Hertz 1960: 96)

The paper, by drawing on the analogy between light and fire, illustrates the dualism that constituted the part of the belief system of the AngloSaxons. Furthermore, the analysis will raise such issues as what should be considered as central, metaphorical or marginal with regard to the domain of light bearing in mind the changing priorities and the perception of the world. Moreover, the domain of fire and light will be also differentiated with respect to the difficulty in delineating conceptual borders between the central and marginal senses. The paper will explain why the border between the central and marginal is more discernible in the domain of fire than in the domain of light. The paper thus invites the notion of vagueness, which is the strategy of cognition and contributes to the constantly changing shape of lexical items and relationships among them. The idea related to the vagueness of categorical boundaries has been vastly discussed (Cruse 1986; Langacker 1987; àozowski 2000). Langacker advocates the notion of vagueness in approaching the perception and conceptualisation of linguistic entities: Much in language is a matter of degree. Linguistic relationships are not invariably all-or-nothing affairs, nor are linguistic categories always sharply defined and never fuzzy around edges. This is perhaps unfortunate from the analytical standpoint—discrete entities are easier to manipulate, require simpler descriptive tools, lend themselves to stronger claims, and yield esthetically more pleasing analyses—but it is a true nonetheless. (Langacker 1987: 14)

Juxtaposition and Differentiation within the Metaphorical Senses

221

The present study, based on Thesaurus of English Dictionary, is an interdisciplinary study based on religion, medicine, as well as social values and evaluative norms. In order to achieve maximum accuracy of the semantic contexts, the study draws on the texts from the Toronto Corpus. The analysis of the metaphors of fire is partially based on my earlier analysis (Wawrzyniak 2010). The comparison of the two domains, namely of fire and light, shows that fire and light were not highly related concepts for the Anglo-Saxons. Thus fire was imbued with both religious and secular undertones, while the light was perceived as the embodiment of God and was thereby imbued with divine qualities.

2. The metaphorical aspects of the lexemes linked with fire Lexemes that referred to the central senses in the domain of fire were numerous, namely fyr, bryne, bæl,glƝd, blæst, blæd, æled, brynewylm, fyrwylm, heaþuwylm, wannfyr and heahfyr (Wawrzyniak 2010). Nevertheless, only a few lexemes developed metaphorical senses in the domain of fire, namely fyr, bryne and blæd. The reason for this is that most lexemes related to the concept of fire were infrequent and used with a certain underlying purpose in specific genres and sometimes not even properly contextualised. Therefore, the only lexemes that could serve as a basis for metaphorisation are those approaching the ideal: the best exemplar of a category in the mind of the speaker. In other words, the components referred to as central ones and embracing a great number of nuances, in the core of the Conceptual Domain, developed metaphorical senses. FYR FYR is traced back to the Indo-Germanic root PU accounting for fire and purification (CEDEL, sv. fyr). According to the Anglo-Saxon Dictionary, fyr denoted the following metaphorical senses: - curiosity, fervour, strong feelings - material, substance As for curiosity, the following contexts from the Toronto Corpus can be exemplified: (1)

fricgan þurh fyrwet (Christ: Krapp and Dobbie 1936: 3–49) (to know through curiosity)

222

(2) (3)

Agnieszka Wawrzyniak

fyrwet frinnan (Elene: Krapp 1932a: 66–102) (to ask/ enquire out of curiosity) fyrwet brecan (Beowulf: Dobbie 1953: 3–98) (to get incited)

The metaphorical senses of fyr can also imply fierceness and sinful character, which is recorded in the examples: (4) (5)

fyrenfulra fyhtenhornas (The Paris Psalter : Krapp1933 b: 3–150) (fierce fights) ic eardige on fyrenfulra folce (The Paris Psalter: Krapp 1932 b: 3–150) (I live among fierce people)

The following contexts evince that fyr abounds in negative undertones, thereby emphasising inevitable oblivion, torture, annihilation or indulgence. Apart, however, from senses evolving around emotional states, fyr denotes also material/substance, which can be illustrated by the examples: (6) (7) (8) (9) (10)

fyr wudu byrneþ (The Paris Psalter: Krapp 1932b: 3–150) (the cross burns) fyrena flan (Genesis: Krapp 1931: 1–87) (arrow, dart) fyrena strælas (Beowulf: Dobbie 1953: 3–98 (arrow, dart) fyr bend (Beowulf: Dobbie 1953: 3–98) (lock) fyr gemylted (Wednesday in Easter Week: Godden 1979: 164–8) (iron; the melted material)

Juxtaposition and Differentiation within the Metaphorical Senses

223

BRYNE Bryne evokes a wide range of emotions, such as curiosity/burning desire, fervour, irritation as well as overwhelming sadness. Thus, contrary to fyr, whose marginal senses instantiated emotional states and a variety of tools, the semantics of bryne is restricted to emotional states, yet incompatible and diversified. Apart from the “seemingly” shared senses with fyr, such as curiosity and fervour, bryne is semantically related with two more aspects, namely irritation and overwhelming sadness. All in all bryne is used to emphasise varieties of emotions at a peak level. According to the Toronto Corpus, the following contexts are provided with respect to the following emotional states: Curiosity/burning desire (11)

(12)

(13)

(14)

(15) (16)

heortan hædre wealleþ hwilum me bryne stigeþ (Bede, History of English Church and Nation, Book 4; Miller 1890: 980: 252–384) (The anxiety increases at heart while the curiosity progresses) byrnende mode (Fourth Sunday after Pentecost: Text from Clemoes 1955–6: 347– 59) (the burning spirit) hata bryne for healicum gymstanum (Saint ǘthelryth: Skeat 1881–1900: 432–40) (the ardent burning for valuable gems) Zealot/ bad feeling þe bryne of ande of byternesse heore agene sawla (Scragg 1992: 52–64) (the burning of zealot, of bitterness of individual souls) Sadness bryne tear (Christ: Krapp and Dobbie 1936: 3–49) (the burning tear) tear tunbryne (Saint Sebastian: Skeat 1881–1900, I: 116–46) (the enclosed land overwhelmed in tears)

The juxtaposition of contexts exemplifying emotional states, in the cases of fyr and bryne, shows semantic differences even in the apparently similar attributes, namely curiosity. It should be emphasised that any

224

Agnieszka Wawrzyniak

similarity related to curiosity, as expressed by fyr and bryne is only superficial, which precludes treating these two lexemes as synonymous. First of all, whereas fyr denotes curiosity in the generic sense, bryne, when used in the sense of curiosity, stands for the specific sense, namely burning desire implying something one strives for. The discrepancies can be explained by re-approaching the central senses of fyr and bryne. In short, fyr projects rather the visual image of fire, whereas bryne signifies the mere process of burning, the activity rather than the image. It should also be emphasised that burning in itself always needs an entity for the mere process to occur, whereas fire can represent the natural phenomenon on its own. Therefore, in metaphorical terms, bryne always implies an entity to strive for. Moreover, the semantics of bryne signifies a movement, direction towards an entity, towards a particular point or a track to be covered for a specific point to be achieved. Furthermore, apart from emotional states, bryne is also used with reference to illnesses and medications, which can be exemplified by the contexts: (17) (18)

(19)

Se ilca bisceop from bryneadle wæs gehæled (Bald’s Leechbook: Cockayne 1864–6, II: 18–156) (The same bishop recovered from a fever) Wiþ maganbryne, wlaco wæter menge wiþ þone gelestan ele, sele drincan (Bald’s Leechbook: Cockayne 1864–6, II: 174–298) (In case of the stomach- ache, mix lukewarm water with the oil, drink from the vessel) smyrian wiþ bryne sealfe (Bald’s Leechbook: Cockayne 1864–6, II: 304–58) (smear with the burning ointment)

2.1 The metaphorical aspects of lexemes that gained the sense of fire via the process of metaphorisation The aim of the present section is to deal with the metaphorical aspects of the lexeme that gained the sense of fire via the process of metaphorisation, namely blæd. According to the Anglo-Saxon Dictionary, (sv. blæd,), this lexeme possesses a cognate OHG blat, which corresponded to Latin flatus, denoting ‘blowing, blast’. Thus, the central OE senses linked with fire in blæd are underpinned by the sense of wind,

Juxtaposition and Differentiation within the Metaphorical Senses

225

which highly affects the nature of their central as well as metaphorical senses. The analysis of blæd indicates that its metaphorical sense (success) is highly influenced by its etymology (flatus – ‘wind’), where the fleeting character, changeability and passing are reflected in the metaphorical sense of blæd by implying a momentary state and transience. Some analogies can be drawn between blæd denoting fire and success. Both senses are of a fleeting and ephemeral character. Although they signalise the peak, the highest status achievable, both also carry the implication that such illuminating moments are just temporary and that both may lose their brightness. The above implications are well visible in the following contexts recorded in the Toronto Corpus: (20) (21)

wesan blæd mid (Maxims 11: Dobbie 1942: 55–7) (to be proud with) he heold a rice, eþeldreamas, blæd mid bearnum (Genesis: Krapp 1931: 1–87) (He held a kingdom, he felt happy and proud with children)

Thus, approaching the structure and the semantic undertones of blæd, one can point to the primary and secondary metaphorisation that affected blæd. Namely, the central sense (flatus – ‘wind’) contributed to another sense, that is fire. This process signals the primary metaphorisation. Yet, as the fire played the pivotal role for the Anglo-Saxons, the new sense, namely, the fire superseded in the usage its former central sense, namely, the wind. In other words, what might take place is the change of the status whereby the actual metaphorical function takes over the role of central attributes and acts accordingly. Given this approach, one can explain the further metaphorical usage of blæd, namely success, which operates on the same attributes as blæd denoting fire. Thus, one can refer to the process of transfer of the semantic traits from the category of fire upon the category of success as a secondary metaphorisation.

3. The analysis of the concept of light The concept of light belongs to the oldest concepts created in the human mind. Nevertheless, its conceptualisation and perception have been subject

226

Agnieszka Wawrzyniak

to constant re-considerations, which accompanied its reconceptualisation in the human mind. Therefore, defining certain attributes of light as either central or peripheral, the contemporary linguist should rely on recorded sources even though the inferences may run counter to his/her established perception of the light. According to Dyszak (1999: 14), initially, there were two different concepts of light due to the source the light stemmed from. Namely, there was the primary and secondary light. Following this approach, the primary light was the one not connected with any source, such as the Sun, the Moon or stars, but which had its roots directly in God. As he maintains, the light was identified with God, and the divinity of the physical world stemmed from the similarity to its Creator. The secondary light, on the other hand, was the one emitted by any source, such as the Sun or the Moon. Therefore, the properties such as glory, splendour or magnificence should be viewed as central senses as they directly referred to the primary source. In other words, the centrality of these properties was the reflection of the predominant view of light, which was the only primeval one and only later on started to be juxtaposed with the emitted source. Dyszak (1999: 13) maintains that initially the primary source was the only concept, which points to the primacy of all attributes of light identified with God and not connected with any physical, tangible source. Therefore, while defining certain attributes of light as either central or peripheral, the contemporary linguist should rely on recorded sources, the culture and a system of beliefs of the speakers even though the inferences may run counter to his/ her established perception of the concept of light with its central and metaphorical aspects. Consequently, the division in the category of light should not proceed along the central and metaphorical but rather along the central, which were metaphorical, and the marginal senses which evoked the literal conception of light. Moreover, some lexemes in the conceptual category of light were of a bipolar nature showing the link with the two sources. In such cases, the borderline between the central and marginal senses becomes more vague and thus more difficult to delineate. LEOHT According to the Anglo-Saxon Dictionary (sv. leoht), leoht possessed the following cognates: Icel. logi – ‘flame’, Lat. Luna – ‘moon’ and Greek eukos meaning ‘white, bright’. Leoht can be conceptualised on various planes. Firstly, it can be viewed as synonymous with the day as it operates after the Sun. Furthermore, leoht can constitute a semantic bridge between the two sources, the primary and the secondary source. In other words,

Juxtaposition and Differentiation within the Metaphorical Senses

227

though triggered off by the secondary source, the word acquired implicit undertones and became identified with God rather than with pure physical presence of the light from the Sun, thus its underlying beliefs are those of primary source. Moreover, leoht can be documented as the light from the primary source as it is completely cut off from the natural source and not conceptualised in terms of physical presence of light. Day (22) Leoht wæs ærest þurh drihtnes word dæg genemned, white beorhte gesceaft, þa Com oþer dæg, leoht æfter þeostrum. (Genesis: Krapp1931: 1–87) (At the very beginning, the light was named a day through God’s word; the white, bright creation. Then another day came; light after darkness). There are also contexts, in which leoht functions as a semantic bridge; Semantic bridge (23) Bletsige sunne and mona, leohte leomanum, God þonne swegles leoht, gamma gladest. (The Rune Poem; Dobbie 1942: 28–39) (The blessed Sun and moon/ light to people, life-bringing God, embodiment of the light of the sky, the brightest of all gems). (24) Ĉær is engle song, eadigra blis, þær is dyre dryhtnes, sunnan leohte. (The Meters of Boethius: Krapp 1932b: 153–203) (There is the angels’ song, the divine bliss, there is divinity, light from the Sun) Primary source (25) Leohte ic eom. (Genesis: Krapp 1931: 1–87) (I am the light) (26) In heofenes leoht þær heo mid wudolre cyninge wuniaþ. (The Phoenix: Krapp and dobbie 1936: 94–113) (She lives with the King of Creation in Heaven’s light). BEORHT, TORHT Beorht as well as torht are entrenched in two sources. They refer to the primary source of light as well as to the secondary one. Nevertheless, while the secondary source is not underpinned by the primary one in

228

Agnieszka Wawrzyniak

beorht, the traces of the primary source are visible in the secondary source in case of torht, which can be exemplified by the contexts: Primary source (27) beorht hiofu (Beowulf: Dobbie 1953: 3–98) (the bright heaven) (28) beorhte fæder, God, cyning, biscop (Krapp and Dobbie1936: 3–49) (the bright Father, God, king, bishop) (29) torhta frumlicne ham (Christ and satan: Krapp 1931: 135–58) (the bright home of the origin) Secondary source (30) beorhte sunne (Genesis: Krapp 1931: 1–87) (the bright sun) (31) beorhte tungel (Daniel: Krapp 1931: 111–32) (the bright star) Secondary source underpinned by the primary source (32) Ĉa com morgen torht beacna beorhtest; halig of heolstre (Andreas: Krapp 1932a: 3–51) (Then came the morning light; the brightest of all signals; the holiness in the darkness) (33) heofon torht scineþ (The Order of the world: Krapp and Dobbie 136: 63–6) (the Heavenly light shines) All in all, while beorht could denote the purely secondary source, torht was not recorded as emitted from the secondary source with no religious undertones exhibiting the direct link with the primary source.

3.1 Lexemes of rare frequency linked to light It is worth emphasising that more central, more widely used lexemes were of a bipolar nature. In other words, they exhibited the link with the two sources, while lexemes whose application was limited were attributed to either of the two sources. Therefore, the strict semantic division

Juxtaposition and Differentiation within the Metaphorical Senses

229

between the primary and the secondary sources could be applied but to rare, more marginal lexemes, created with a view to highlighting a certain set of attributes needed for the text’s sake. The analysis covers the following lexemes of low frequency: lihting, leoma, scir as well as lexemes with the lowest frequency: bliccetung, glæterung, glitenung. Lihting, leoma Lihting and leoma have been documented as encompassing a secondary source of light. They are neutral terms and not implying any moral or judgmental undertones. They apply to natural objects that either generate or reflect the light but not to artificial objects that evoke the effect of brightness. The Toronto Corpus records no single case where lihting or leoma could account for light identified with God. Contrary, however, to leoma, which applies only to the day time, lihting refers to the effect of lightening generated by both the Sun at day time and the Moon/stars at night. Therefore, though used primarily in the central sense, lihting needs identifying the source to define its nature. In other words, it stands for visible, tangible, well-identifiable light which can be defined by its source. It should also be emphasised that lihting is not specific about sharpness, vividness or quality of light and can be used equally well to apply to light at day or at night time. (34) (35) (36) (37)

Ĉa steorran sind mannum to nihtlicre lihtinge (Genesis: Krapp 1931: 1–87) (The stars bring the light at night to people) Hi sum leoht habban oþ þes dæges leoht oþþe lihtinge niht. (Hexameron: Crawford 1921: 33–74) (They have some light from the day or night’s light) Sunnan leoma scineþ (Gregory the Great, Dialagues, Praface and Book 4, 260–265) (The Sun’s light shines) Seo sunne ne scinþ þonne blindan and se blindan ne gesihþ þære sunnan leoman. (Exodus: Krapp 1931: 90–107) (The Sun does not shine to the blind and the blind do not see the light of the Sun)

SCIR Another rare lexeme related to the semantic field of light and entrenched in the secondary source is scir. Contrary, however, to leoma and lihting, which apply only to natural objects, scir can be used with reference to both natural and artificial objects. According to the Toronto

230

Agnieszka Wawrzyniak

Corpus, there has been one documented context where scir functions as the brightness from the Sun. (38)

On hate sunnan on sumera synd þa sciran dagas (Virgins: Godden 1979: 327–34) (During the Sun’s heat in Summer’s time there are shiny days)

The mere glittering, clear and transparent attributes of scir found also their application in artificial objects to highlight their glittering appearance: (39) (40)

scire helmes (The Paris Psalter: Krapp 1932: 3–150) (the bright/ glittering helms) scir swyrd (Saint George: Skeat 1881-1900, I: 306–16) (The glittering sword).

BLICCETUNG, GLǘTERUNG, GLITENUNG According to a Thesaurus of Old English (TOE, sv. bliccetung, glæterung and glitenung), bleccetung, glæterung and glitenung are perceived as synonyms and rendered into glittering and effulgence. Nevertheless, approaching these lexemes in context, one notices sharp differences among them, which precludes embracing them as synonyms. As for bliccetung, the lexeme occurred only once in glosses in the following context: (41)

Inlihton bliccetunge þine eorþan ond onstyred wæs eorþe. (Psalms, Ms Cotton Vespasian, A.I. Kuhn 1965, I: 146) (The glittering lit Your Earth and the Earth was in motion).

Thus, bliccetung can be viewed as related to the primary source and identified with God. Above all, it highlights the act of Creation thus being imbued with positive undertones. Contrary to bliccetung, glæterung evolves around the negative semantic pole and denotes the destructive light in hell: (42)

Swa swa scearp on helle se deaþ on glæterunge. (Psalms, Ms Royal 2 B. V.: Roeder 1904, I: 274) (The death is so sharp in hell in the destructive light).

Juxtaposition and Differentiation within the Metaphorical Senses

231

From the above context, one can assume that the mere form glæterung could perform the role of metonymy where it stood for fire. Light, by itself, cannot be destructive. What, however, glæterung brings to mind is death, destruction and the light being the result of flames. Moreover, due to the implications glæterung is equipped with, one should postulate the primary source, which evokes the notion of glæterung. First of all, the properties inextricably bound with the secondary source are objectivity, neutrality and well-defined source. Glæterung, on the other hand, should be located in the primary source by dint of denying divine properties. Though it is not identified with God-like qualities, it still should be placed in the same domain as it evokes, by the mere contradiction, the divine properties. Furthermore, the fire which gives rise to glæterung is far from being a neutral, natural element but the fire bound with hell, thus highlighting the religious domain. The third lexeme rendered into glittering/effulgence is glitenung, which was linked with the secondary source, thereby implying neutrality, objectivity and a well-defined source: (43)

Ligræscas mid glitenunge (Psalms, Ms stowe 2. Kimmens 1979, I: 280) (Lightings with glittering)

The juxtaposition of bliccetung, glæterung and glitenung reveals their connotative shades. Bliccetung and glæterung are identified with the primary source but the former one brings to mind highly positive, divine properties whereas the latter one highlights the absence of such attributes. Glitenung, on the other hand, is associated with the secondary source thus being neutral with respect to religious connotations. The analysis of fire and light domains seems to show vivid differences in the scope of application of the above concepts. The fire domain can be used to conceptualise various emotions and states that centralise intensity and a seeming incompatibility, whereas the domain of light applies to decent, divine and immaculate qualities. Light, hence, is linked with values and phenomena of a more lasting character, which seem not to change abruptly and which give the sense of safety and eternity. The source of fire, however, builds target concepts, which are rather shortlived. They come unexpectedly and may be self-destructive or destructive to others. Such built target concepts make the impression of being intense but this intensity is all the time on the verge of either decreasing its strength or even being quenched. It is by no means permanent as in the

232

Agnieszka Wawrzyniak

case of concepts instantiated by light. Consequently, the source of fire is far from creating the sense of safety and is rather closer to the sense of unpredictability, change and anxiety.

4. Conclusions The analysis of the concepts fire and light leads to the following conclusions. Firstly, the metaphors instantiated by fire are linked with emotions and tools. They are grouped along positively and negatively loaded semantic poles. The metaphors, however, that evolved in the category of light give rise to divine qualities, such as nobility, glory and magnificence. Secondly, the borderline between central and metaphorical senses is easier to delineate in the category of fire than in the category of light, as many lexemes in the category of light were of a bipolar nature. Moreover, the division in the category of light proceeds along the central (metaphorical) and the marginal. The properties, such as glory, splendour and magnificence (primary source) constituted the basis underlying the conceptualisation of light as opposed to the more literal conception of light (secondary source), the instances of which were less frequently recorded. Thirdly, only prototypical and frequently used lexemes in the category of fire (fyr, bryne, blæd) developed metaphorical senses as opposed to the category of light where most lexemes instantiated metaphorical, central senses.

References Baldinger, Kurt. 1980. Semantic theory. Towards a modern semantics. Oxford: Basil Blackwell. Bosworth, Joseph and Toller Northoote (eds.). 1898. An Anglo-Saxon Dictionary. London: Oxford University Press. Cruse, Alan. 1986. Lexical semantics. Cambridge: Cambridge University Press. Dyszak, Andrzej. 1999. JĊzykowe wyraĪenia zjawisk emisji Ğwiatáa [Linguistic expressions of light emission phenomena]. Bydgoszcz: Wydawnictwo Uczelniane WSP. Healey, A di Paolo (ed). 1986. Dictionary of Old English Corpus. Toronto: The University of Toronto Press. Hertz, Robert. 1960. Death and the right hand. Aberdin: Cohen and West. Langacker, Ronald. 1987. Foundations of cognitive grammar, vol. 1: Theoretical prerequisites. Stanford: Stanford University Press.

Juxtaposition and Differentiation within the Metaphorical Senses

233

àozowski, Przemysáaw. 2000. Vagueness in language. Lublin: Wydawnictwo Uniwersytetu Marii Curie Skáodowskiej. Roberts, Jane and Christian Kay (eds.). 1995. A Thesaurus of Old English. London: Kings College; Centre for Late Antique and Medieval Studies. Wawrzyniak, Agnieszka. 2010. The role of the semantic space in the analysis of the FYR concept in Old English. In Andrzej Kciuk and Katarzyna Molek Kozakowska (eds.), Exploring the notion of space in cultural, literary and language studies, 104–116. New Castle upon Tyne: Cambridge Scholars Publishing.

PART FIVE PRAGMATICS AND DISCOURSE

TRACING TABLOIDIZATION: A COMPARATIVE ANALYSIS OF RHETORICAL STYLES IN QUALITY AND TABLOID NEWSPAPERS KATARZYNA MOLEK-KOZAKOWSKA

1. Introduction Since the advent of electronic media, the press, an erstwhile leader in breaking news, has needed to re-invent itself in the competitive media market. To do so, even mainstream quality press outlets have set on the route of tabloidization in order to attract audiences. Yet, tabloidization would not be an analytically useful concept were it not traceable in terms of the changing rhetorical features of newspaper texts. This paper applies the category of rhetorical style in order to identify the most pervasive differences, but also increasing correspondences, in quality and tabloid press coverage of clerical sex abuse in Britain. Rhetorical styles are defined as clusters of linguistic devices used strategically in order to attract readers. Although rhetorical styles are pre-defined by institutional contexts, news genres and communicative objectives, there still remains a plethora of linguistic choices that can be made to engage audiences. Such choices, when taken systematically, can cumulatively build a specific rhetoric that characterizes a given news outlet. Obviously, an analysis of all rhetorical features of tabloid and quality journalism is far beyond the scope of this study; thus, the focus here is on the stylistic features that tend to originate in tabloid press and may have been strategically used to increase audience appeal in quality press. This is because stylistic variation usually becomes evident in a comparative enterprise. Therefore, in its analytical part, the paper juxtaposes predominant features of the rhetorical styles identified in two special-purpose corpora (each over 10 000 words) of newspaper texts on clerical sexual abuse that were published in the Guardian and the Daily Mail throughout 2010. The

Tracing Tabloidization: A Comparative Analysis

237

analysis draws on quantitative data for reliability, but its rhetorical orientation requires adopting an interpretive approach.

2. On the downward route to tabloidization? The Media Studies literature on the tabloidization of print media is extensive and envisions it as a predominantly negative process. Tabloidization is often approached from the position of cultural critique, despite the fact that much media research is in fact devoted to validating the uses of popular culture (Turner 1999). Tabloid outlets are often blamed for undermining the public sphere, as they tend to dwell on dramatically bad news, such as corruption, conflict, crime, violence, disaster and death, or sensationalize the coverage with the focus on the scandalous and the saucy. It is often noted that citizens in democratic systems, instead of being informed of the issues of public concern, are flooded with cheap entertainment or episodic coverage that causes confusion, alienation and overload (Postman 1986; Blumler and Gurevitch 1995). And yet, despite much theorizing about the grave social consequences of the “media malaise,” for example in engendering cynicism (Cappella and Jamieson 1997), there is little empirical evidence of tabloids significantly depoliticizing public life (Newton 1999). In the midst of a “moral panic,” to use another Media Studies concept, with regard to falling journalistic standards, it is rarely noted that tabloids may well have opened the public sphere to masses of people, who would otherwise be excluded from democratic deliberations. Tabloid press offers them assessable frameworks for interpreting political events and social processes. In fact, recent research into tabloid reception demonstrates the complexity of patterns of their consumption and the positive role of the tabloids in helping readers to make sense of the meanings of the social, the national and the moral (Johansson 2007). As a result, the ultimate social effects of tabloidization are yet to be determined. Tabloid news presentation is often juxtaposed with the so-called “broadsheet” or quality press journalism, which is often said to be serious and reliable with its toned-down emotions and rather observant separation of “fact” from “opinion” pieces. Credited with the sense of responsibility for the society, quality press outlets are said to be “objective,” but the notion of media objectivity is problematic. For example, Richardson (2007: 86–88) claims that objectivity is but one of textual practices rather than a defining property of quality journalism, which, even in the case of news reporting, relies on value judgments at every stage of news production: from news gathering, to news selection, writing, editing and

238

Katarzyna Molek-Kozakowska

presenting. To appear objective and neutral, quality press texts often feature verbalizations of competing truth claims without resolving the controversy, refer to institutional authors as sources of information, or use “scare quotes” to indicate contested terminology. All such textual devices are applied strategically to produce an impression of objectivity, which contributes to the building of specific “quality” rhetoric. By contrast, tabloid journalism tends to be explicitly engaged politically and engaging emotionally and does not have pretensions to objectivity. This is evident in the tabloids’ preference for explicit evaluation—a judgmental rhetoric appealing to ethos, and theatrical emotionality produced by highly charged words and structures. In many cases, tabloid news-breaking is predicated on exposing “the hidden,” that is anything from the details of private life of known people, to finances of public institutions, to violent and illegal activities. Such revelations are often expressed with the air of moral superiority. The indispensable visuals provide tabloid readers with a chance to experience voyeuristic pleasure, or, at least indulge in escapism. Characteristically, tabloid texts exploit stereotypes and cater to prejudices by linguistically perpetuating such elementary social divisions as the government vs. the citizens, the immigrants vs. the nationals, or the working class vs. the middle class. Tabloid texts are highly concentrated in terms of information provided and highly repetitive as far as their structure is concerned. They use “transparent” everyday idiom compatible with lived experience, with presupposed values (often culturally and morally conservative), in which direct quotes function as emblems of authentic and immediate rendering of the news. In this they aim to target all audiences, including the uneducated (Biressi and Nunn 2007). Hence, reading is often facilitated by the frequency of narrative sequences, records of conversation (e.g. exclusive interviews) and pre-digested conclusions announced in headlines or leads.

3. Newspapers and their rhetorical styles In order to operationalize the concept of tabloidization for the purpose of this study, the category of rhetorical style has been proposed. Rhetorical styles can be defined as clusters of linguistic devices used strategically in order to attract readers. This is because one way to study tabloidization is to identify (and compare) sets of specific linguistic options which tend to be persistently applied in the so-called tabloid newspapers (as opposed to the so-called quality press). However, the category of style, which conjoins formal and functional aspects of texts, is still contested and notoriously difficult to apply for larger comparative

Tracing Tabloidization: A Comparative Analysis

239

analyses. Within communication studies, semiotics, pragmatics and discourse studies, the notion of style has recently been variously labeled and described, for example as communicative style (Selting 1999), rhetorical style (Fairclough 2000), conversational style (Tannen 2004), language style (Machin and van Leeuwen 2005), or sociolinguistic style (Coupland 2007). As styles can be applied to signify meanings, project identities and express values verbally, Machin and van Leeuwen (2005), for example, see them as tools indicating not only unique personal qualities but also group ideologies and even consumption patterns typical of particular lifestyles. Language styles tend to be heterogeneous categories— “composites of connotations” (2005: 587)—that may well be deliberately contrived. However, as media discourse tends to be not only expressive but also strategic and goal-oriented, which is reflected in the pursuit of linguistic options that have the largest appeal to the public, it is the notion of rhetorical style that best encapsulates this property. What is more, “a rhetorical style is not an invariable way of using language, it is rather a mixture of different ways of using language, a distinctive repertoire” (Fairclough 2000: 96). Yet, those “mixtures” of linguistic devices are not at all combined out of random and incidental elements, since many stylistic features (e.g. registers, semantic macrostructures, syntactic patterns, lexical choices) are constrained by institutional contexts and prescribed by the texts’ generic requirements. Also, the linguistic resources strategically selected for a newspaper’s rhetorical style must be pervasive to achieve desired effects. The linguistic devices singled out for stylistic analyses, according to Jeffries (2010), could be “conceptual” rather than “technical,” in order to better approach the form-function mapping of textual choices. These include such linguistic aspects as naming and describing (through reference and modification for example); representing actions/events/states (through verbal or nominal preferences or various syntagmatic choices); equating and contrasting (through establishing lexical or syntactic equivalence and opposition); exemplifying and enumerating (through the number, type and arrangement of specific elements in a clause); prioritizing (through word order manipulation); implying and assuming (through presuppositions and implicatures); negating and hypothesizing (through lexical items or grammatical structures such as particles, adverbials or modals); presenting others’ speech and thought, as well as representing the temporal and spatial setting of the discourse. This conceptual framework seems to constitute an expandable and applicable toolkit for identifying rhetorical styles.

240

Katarzyna Molek-Kozakowska

4. Rhetorical styles juxtaposed: samples and method The rationale behind this comparative study is to shed light on some of the characteristic features of newspaper rhetorical styles applied in two comparable samples of press coverage. Thematically, both samples include articles devoted to clerical abuse and have been deliberately restricted to the coverage of paedophile priests in 2010 and limited to the British quality daily the Guardian and the mid-market tabloid the Daily Mail in order to highlight the most conspicuous differences as well as striking correspondences between the outlets’ rhetorical styles. The analysis is necessarily limited to the most pervasive stylistic aspects of the coverage. The material subjected to this qualitative analysis are two corpora of articles which emerged during a web search of the Guardian’s and the Daily Mail’s archives with the keyword “paedophile priest”. As far as the Guardian is concerned, the search at the beginning of 2011 yielded 441 hits. Out of 106 articles on the subject published in 2010 (comparing to 27 in 2009, 28 in 2008, 23 in 2007, 25 in 2006, 38 in 2005, 25 in 2004, 41 in 2003, 55 in 2002, 38 in 2001 and 32 in 2000), 14 have been assigned to the corpus until the threshold of 10 000 words was reached. The criterion for inclusion was the search’s category of relevance. All the articles in the corpus include the words “paedophile” and “priest” in the headline and were authored by nine different journalists so as to exclude personal stylistic preferences. As regards the material from the Daily Mail, the sampling procedure was the same; however, the search yielded only 71 hits altogether (35 articles in 2010, 7 in 2009, 9 in 2008, 6 in 2007, 4 in 2006, 2 in 2004, 4 in 2003, 3 in 2002, and 1 in 2001), which indicates that the coverage of clerical sex abuse has not been as extensive and multifaceted as the Guardian’s. The number of articles needed to reach the 10 000 word limit was only 10, which means that the Daily Mail’s texts on paedophile priests tend to be slightly longer. Four texts were written by anonymous correspondents, while the remaining six were authored by four different journalists. Closer analysis has revealed that the Daily Mail’s texts are much more repetitive (with the same facts and events mentioned twice or three times within one article), and feature longer quotations from sources. The corpora are first scrutinized quantitatively for repeated lexical items, collocations and salient patterns of textual organization. The thematic focal points of the coverage (semantic macrostructures) are thus delimited. Then the qualitative method is used to specify characteristic aspects of coverage and evaluate the stylistic choices applied, sometimes

Tracing Tabloidization: A Comparative Analysis

241

with reference to extant literature on current journalistic practices. The way these choices contribute to the overall rhetorical style is discussed and sometimes the explanation as to the possible effects on the readers is offered. As the project is comparative in its methodology, the description of the second rhetorical style—the Daily Mail’s—is often conducted vis-àvis the features identified in the first corpus pertaining to the rhetorical style of the Guardian.

4.1 The Guardian’s rhetorical style The Guardian’s coverage of clerical sex abuse has intensified in 2010 due to the Pope Benedict XVI’s planned visitation to Britain. The main trend of that coverage is a re-orientation of the thematic focus from reporting new cases of paedophile priests to scrutinizing the ways the Catholic Church has been trying to handle the problem. Thus, the most frequently repeated words in the headlines and leads of the 14 articles studied, namely “fail(ure)” (4 times), “(mis)handling” (4 times), “accusations/allegations” (3 times), concern the Catholic hierarchy’s, including the Pope’s, alleged inability to respond appropriately to the cases of child molestation by priests. The articles do not expose new instances of child abuse, but new documents or statements which indicate that both local bishops and the Vatican were likely to have known about abuse and yet chose not to take any decisive measures to deal with it in order to preserve the Church’s reputation. In this sense the current coverage of clerical sexual abuse is an example of “easy journalism” (Woffinden 1996), which constantly recycles controversies and moral panics over cases of paedophilia (and over their suppression) that date back to the 1970s or 1980s. The Guardian’s rhetorical style seems to be designed to encapsulate a negative assessment of the situation in the Catholic Church. First of all, the rhetoric of previous hysteria connected with exposing paedophiles is still detectable in the articles, particularly in such exaggerated, yet naturalized, metaphors as “predatory1 priests,” “putting children in danger,” “abuse survivors.” However, the paper’s critique is no longer directed at individual abusive priests, but rather at those who condoned their practices, notably local bishops and Vatican administrators. This is revealed through the foregrounding of such nominal labels as “scandal” and “crisis”, and specifically through the reoccurrence of the negatively 1

All italics are mine and are used to highlight certain stylistic features of the analyzed material

242

Katarzyna Molek-Kozakowska

charged colloquialisms “cover-up” and “hush up,” which stand out in the paper’s rather formal register. Due to such labeling devices, the subsequent coverage is likely to be already “pre-defined” for the readers. Moreover, the choice of adjectives that modify the paper’s revelations is clearly aimed at sensationalizing its own coverage: “allegations are devastating for the authority and reputation of the church,” “fresh revelations have been made directly implicating Pope Benedict XVI,” “in the wake of damaging disclosures of clerical sexual abuse and cover-ups,” or “a calamitous week for the Irish hierarchy.” Hence, in terms of naming and describing, the paper explicitly reveals its stance, situating itself in the position of a morally superior body entitled to evaluate an institution that no longer can be trusted to be accountable. The newspaper’s otherwise relatively detached and neutral way of reporting is “compromised” stylistically when the coverage concerns the cases when senior Catholic leaders failed to refer the paedophiles to the police or forced vows of silence on the victims, as well as when they delayed decisions of defrocking or transferred “problem priests” to other dioceses, enabling them to go on abusing minors. The paper’s critical stance transpires through the frequent repetition of predicates that indicate the Church’s inaction or passivity, such as “ignoring complaints,” “allowing to continue,” “remain in priesthood,” “failing in their duty,” “kept quiet,” “failing to report,” “was mostly silent,” or “submit the incidents of this sort to very careful consideration.” There are a few rhetorically potent verb phrases that seem to have been coined or selected for the purpose of underlining the seriousness of the situation: “revelations which plunged the Catholic Church worldwide into turmoil,” or “the actions of paedophile priests have left ordinary Catholics demoralized and confused.” In terms of predication, the representations of actions, events or states are often metaphorical or metonymic (or indirectly negated, as shown above) and an analysis of the material, verbal, mental or relational nature of actionals used here does not reveal consistent patterns. What is also characteristic of this kind of sustained coverage is the relative generalization of the phenomenon of clerical sex abuse. News of cases of specific paedophile priests is scarce, and their abusive practices are rarely detailed. The “distilled” image of clerical abuse is achieved through frequent nominalizations, e.g. “systemic abuse,” “tide of accusations,” “collusion, cruelty and endemic abuse” and multiple premodification, as in “clerical child abuse victims,” “developing child protection policies.” This evidences a discursive shift in the current coverage from exposing specific activities to registering a general trend. It also demonstrates that, as such, paedophilia among priests is taken by the

Tracing Tabloidization: A Comparative Analysis

243

Guardian’s editors to be well-known among its readership. This is further reinforced by the paper’s “profiling” of a “typical” paedophile priest: “the classic perpetrator was a priest in his thirties who spent some time, mostly less than a year, sexually involved with a boy in his teens,” which acknowledges that “almost all the accused are alleged to have molested one minor; only three percent […] had more than ten alleged victims.” Inevitably, such overgeneralized depiction reproduces a stereotype and influences the readers’ interpretations of the following coverage. Rhetorical styles may be designed in relation to conventionalized notions of specific institutionalized registers. For example, the language of the Guardian’s articles in the corpus abounds in examples of legal jargon: priest paedophilia is not only represented as moral or social problem but as a legal one as well. A priest is “arrested and charged with sex abuse of minors,” another one becomes “a registered sex offender,” yet another “pleaded no contest to charges of lewd behavior.” Such terminology is not only used when quoting official sources (i.e. spokespeople for the police, the prosecution, or the court), but also in the articles’ narratives. Arguably, this feature of rhetorical style increases the perception of authoritativeness of the coverage and, by implication, its being objective. Yet, legal pronouncements are known to be highly evaluative, as this supposedly matter-of-fact report indicates: “Passing the sentence, Judge […] described the defendant as devious and manipulative. […] Offences committed were unimaginably wicked and caused immense and long-lasting […] damage to the victims.” Arguably, such choice of terminology doubly affects the audience, since, while appearing to be neutral, the outlet is passing judgment that is congruent with its interests, namely to represent the issue as socially relevant and emotionally and morally involving for the readers. Legal language adopted by the Guardian also stands in sharp contrast to the euphemisms used by Catholic bishops, whose statements with regard to the clergy’s abusive practices are mostly quoted verbatim. Some of the ways of referring to priests’ paedophilia are the following: “evil deeds,” “sinful acts,” “succumbing to abnormal tendencies,” “flaws and personal failings,” or through the circumlocution “influence of the powers of darkness.” What is also characteristic, the Guardian’s quoting of senior Church officials at some length tends to expose their studied aloofness and visible discomfort when having to comment on the issue. This is also why, in terms of framing, much of the Guardian’s coverage is devoted to the question of “responsibility.” Although the paper is reluctant to directly put blame on anybody, with its careful qualification of its coverage with the words “alleged” or “allegation,” the choice of news items and the selection of quotes, particularly from abuse victim support groups, does not leave

244

Katarzyna Molek-Kozakowska

many doubts about who is to be held responsible. A rhetorically sharp critique of the Church is often embedded in the excerpts from statements of victim groups’ representatives. For example, much attention is devoted to the groups’ often futile efforts to arrange meetings with various Vatican officials who have promised to “demonstrate the Church’s concern for its wounded congregation.” Yet even such examples of goodwill are often dismissed by the paper as “secret meetings [of Vatican officials with abuse victims] that had been clandestine and orchestrated.” The paper’s preoccupation with the question of legal and moral responsibility is also evident in its sustained interest in whether some senior clergymen, as well as Vatican administrators, should consider “accepting full responsibility for the mistakes” and stepping down. The paper criticizes bishops who are reluctant to abdicate their positions in response to revelations about their participation in (or at least knowledge of) covering up the cases of molestation in their dioceses. It highlights that “more than half” of the Catholic priests jailed for paedophile activity in the UK still “remain in the priesthood […] receiving financial support from church authorities.” The Guardian seems to dismiss the senior clergymen “profuse apologies” as insincere, demonstrating that the Church is not likely to actually reach out to victims of child abuse or to punish the paedophile priests in its ranks. Thus, by bringing Church’s “words” in sharp contrast with its “deeds,” the paper is able to prove the institution’s hypocrisy and disdain for the congregation. Concluding, it can be noted that clerical child abuse is no longer presented in the Guardian through individual cases, but rather as a trend, which is evidenced through striking statistics: “abusing as many as 200 deaf boys at a school where the chaplain taught between 1950 and 1974,” “abused by a Boston-based priest who molested at least 40 boys in the area,” “charged in 2002 with 13 counts of child molestation from the 1970s,” which are then harshly condemned with the aid of the paper’s lexical choices and syntactic patterns. Also, the trend is often represented through overgeneralization: “anger mounted over cases involving thousands of children across Europe,” “the crisis will not alone consist in mass defections,” “The Catholic Church may well be facing the biggest crisis since reformation.” Such speculations work towards projecting a view that the Catholic Church in the UK is an institution rife with abuse and that the papal visit should be devoted to dealing with the problem categorically. While acceding that only a small percentage of priests may in fact have engaged in abuse, the paper succeeds in discrediting the Church’s moral authority through the cumulative effect of such narrowly focused negative coverage.

Tracing Tabloidization: A Comparative Analysis

245

4.2 The Daily Mail’s rhetorical style As was the case with the Guardian, the coverage of clerical sex abuse, as well as other Church-related issues, in the Daily Mail intensified in 2010. Yet, the main tendency of this coverage is to expose, or at least revisit, some of the most scandalous instances of paedophilia among priests, with particular attention to cases when the perpetrators have managed to escape punishment. The paper spotlights the ways in which the Church may have deliberately “shielded” the priests from justice. It also devotes much more attention to how some abuse victims have dealt, or have yet to deal, with what had been done to them. In contrast to the Guardian, the Daily Mail focuses much more on the people involved in clerical abuse rather than on the practice, its attempted concealment and its consequences for the Church’s reputation. The recurring words in the Daily Mail’s headlines and leads, apart from “paedophile” and “priests,” are “abuse” (4 times), “shame” (3 times), “victim” (3 times), “(sickening) catalogue” (twice), “hell” (twice), and “shield” (twice). What is characteristic of the way clerical sex abuse of children is covered in the Daily Mail is the extent of repetition of the same “revelations” (police records, Church documents, statistics) within and across the articles in the corpus. This may be explained as the tabloid’s attempt to introduce the readers to the legal and ethical context of the situation the Catholic Church has been put into with the rise in the number of complaints about molestation by priests. Repetitiveness of the Daily Mail’s rhetorical style can be attributed to its audience’s assumed ignorance as to the current scope of critique the Church, including the Pope, has been forced to confront. For example, the article “Paedophile priest victim to go on trial for ‘revenge attack’ 35 years later” about a specific instance of an assault on a retired priest by a now-adult victim is closed with additional details about “several other cases of violence, sometimes fatal, against priests accused of abuse since the Roman Catholic clergy abuse scandal unfolded in 2002.” Such apparently irrelevant information can be justified only in terms of the paper’s attempt at contextualizing the issue, showing its scope, or clarifying the psychological mechanism of revenge attacks; however, one cannot resist the feeling that this only thinly disguises the tabloid’s preoccupation with topics involving sex and violence. With respect to paedophile priests, the Daily Mail tends to use the metaphor of predation, according to which abusers “prey on new victims,” while victims are sometimes conceptualized as “survivors.” Such action verbs as “rape,” “sodomize,” “molest,” “grope,” “fondle” can often be found alongside the more general “sexually assault” and “abuse.” Also,

246

Katarzyna Molek-Kozakowska

the paper is extraordinarily specific in depicting the acts of sexual abuse: “convicted of tying up and molesting two boys in San Francisco,” “they were sexually abused and made to dress up in women’s clothing,” “he asked the little girl to lie over his lap.” In most cases it reveals the names of priests and their victims and specifies the circumstances of abusive behavior (e.g. “Lindner abused them in 1975 during camping trips in the Santa Cruz Mountains. The boys [William Lynch and his brother], who were seven and five at the time, were raped in the woods and forced to have oral sex with each other while Lindner watched”). In the Daily Mail the details of instances of abusive practices are not left without judgment, although in many cases the judgment comes from those involved: victims, their relatives, their lawyers, victim support group members, rather than the paper itself (e.g. “‘I wanted to exorcise all the rage and anger and bitterness he put into me,’ Lynch said; “‘I felt physically sick,’ said Joan. ‘I couldn’t believe that this was Bill doing all these vile, awful things to these poor boys.’”). As a result, the Daily Mail’s rhetorical style tends to be fairly colloquial and emotional, mainly due to many recounts of events by those closely involved in the cases (e.g. “Ms McEntire said her brother molested her starting when she was five. She last spoke to him in 2009. ‘Jerry’s violent and I would not be surprised if he did get beat up. I could understand somebody getting that mad,’ she said. ‘I’ve often said myself that I don’t trust myself around him. I would likely wind up in jail because I’d probably kick him somewhere the sun doesn’t shine.’”). However, it would be wrong to assume that the Daily Mail avoids formal legal terminology and jargon; on the contrary, in many specific cases the tabloid uses legal language to precisely label the lawsuit charge, the settlement outcome or the defense strategy used, as in “during his time as priest in Dublin, Carney pleaded guilty to two counts of indecent assault and he was given probation. Six families were paid compensation by the Catholic Church and, in 1992, Carney was defrocked after church authorities convicted him under canon law of child sex abuse.” Presumably, the Daily Mail is concerned about its credibility as an outlet and is prepared to offer the readers hard facts as well as their evaluations. The Daily Mail’s articles are relatively long, but, unlike is the case with many types of current newspaper coverage, they do not overload the reader with random bits of information. On the contrary, they are easy to follow, mainly due to the predominantly narrative format employed for the sequencing of events. For example, the stories of abusive (ex-)priests Jerry Lindner, Patrick O’Donell or Aquilar Riviera are documented from the beginning of their church duties till present time with occasional

Tracing Tabloidization: A Comparative Analysis

247

explanatory flashbacks. The accounts are complete with how the abusive priests were shifted from parish to parish to avoid responsibility and to protect the Church’s reputation. Such narratives help to depict how the mounting number of complaints from parents and victims has paralleled the local bishops’ failures to prevent abuse, either due to lengthy internal procedures or because of the clandestine nature of the measures undertaken (as teenage victims of abuse were forced to sign vows of silence). Narrative and deictic elements applied in such articles place the reader in the position of hind-side reflection facilitated by the paper’s sensation-oriented stance. In a similar vein, the narratives of abuse victims, such as Marie Collins, William Lynch or Andrew Madden unfold to illustrate how the failures of justice, on the part of the Church and the police, broke their lives, despite offered compensations. For instance, the most elaborate and sophisticated use of the narrative is the article “Mum… your husband is a paedophile priest: Joan Clayton spent 13 years with ‘kind, gentle’ Bill Carney until her son confronted her with some damning evidence.” Not only is the storyline deliberately contrived to create suspense, but also the article provides descriptions of locations, employs primary and secondary characters, includes dialogues and internal monologues, “slows the narration” at climactic moments, offers a final resolution and finishes with a coda. Thus, the extensive use of the narrative format, even to chronicle Pope Benedict XVI’s Malta visit, seems to be a characteristic feature of the Daily Mail’s rhetorical style. This stands in contrast to the Guardian’s synthetic and predominantly argumentative compositional technique. Also in contrast to the Guardian, whose primary frame for interpreting clerical sex abuse is demarcating “responsibility,” the Daily Mail frames its coverage of paedophile priests in terms of “shame” and “pain.” This can be observed for example by noting the frequency of such words as “regret,” “failure,” “hurt,” “apology,” “apologize,” as well as “scandal” and “crisis.” The Church itself is often demonstrated to acknowledge the gravity of the situation, as in “the Vatican said the Pope had ‘expressed his shame and sorrow’ to the men.” or “Cardinal Brady’s apology came as Pope Benedict confirmed that he would sign a letter to Irish Catholics expressing his ‘profound concern at the painful situation in Ireland.’” Not surprisingly, the churchmen’s references to sex abuse are mainly euphemistic: cases of molestation are repeatedly called “abominable acts,” whereas revelations about attempts at concealing abuse are “failures to safeguard young people,” “painful episodes,” or “letting [Catholics] down.” Such formulaic expressions from the Church, even if sincere, contrast with the direct and emotionally charged expressions preferred by

248

Katarzyna Molek-Kozakowska

the paper: “sickening catalogue of abuse,” “a generation betrayed by priests, bishops and police,” “disturbing reading,” or “the Pope was reduced to tears.” Also, in tune with the typical tabloid rhetoric, the Daily Mail’s framing of news tends to be simplistic or exaggerated. For example, the clipped announcement “Pope Benedict has come under fire over child abuse” is exaggerated, taking into consideration the fact that the item covered is whether local bishops could have prevented child abuse if they had reported complaints to church authorities and the police immediately. Likewise, writing that “the Vatican has been rocked by a series of abuse cases in Ireland, America, Germany and Austria” on the basis of the fact that “since 2001 it has dealt with 3000 abuse cases dating back 50 years” is hardly putting the issue into perspective.

5. Conclusions The above analysis has revealed some specific differences between the rhetorical styles used by the Guardian and the Daily Mail in the coverage of clerical sex abuse. However, it has also demonstrated some overlapping features of quality and tabloid reporting, which may testify to the gradual infiltration of tabloid rhetorical and linguistic devices into the so-called quality press. Perhaps not surprisingly, considering the painfulness of the issue, both outlets studied here equally strongly condemn the practice of abuse and its attempted cover-up, which is reminiscent of stylistic choices characteristic of earlier “moral panics” over child abuse. In terms of their rhetorical styles, both outlets subscribe to the explicit negative judgment of the issue presented, even though the Guardian does it through relatively formal, argumentative and generalized rhetoric, whereas the Daily Mail achieves it by means of more emotional and direct reporting. As both outlets are prone to using legal language and resort to extensive quotations from elite sources, both may seem authoritative for their respective audiences. In a contrastive analysis, it emerges that the Guardian is credible to its readers by virtue of its focus on new factual details, its argumentative structures for coverage, and its critical stance (which apparently excludes naivety). Meanwhile, the Daily Mail is likely to be read for its first-hand “witness” reporting, its narrative-like compositional formats, as well as its visual illustration of newsworthy items. Notwithstanding the systematic differences in the rhetorical styles employed to appeal to their respective audiences, it may be observed that both outlets tend to indulge in overgeneralizing, simplifying or sensationalizing the issue. The fact that the Pope’s responses to and statements on clerical abuse are so intensely covered also indicates the

Tracing Tabloidization: A Comparative Analysis

249

application of the discursive strategy of personalization. Human interest stories are, after all, almost always more appealing to audiences than the coverage of institutional processes. Finally, both papers seem to depict the problem of priest paedophilia as much more threatening to the society and as much more damaging to the Catholic Church as may be the case. They both intensify the general perception that abusive priests have not been brought to justice as a result of Church’s persistence in “shielding them” (the Daily Mail) or reluctance in admitting “responsibility” (the Guardian). Indeed, in a broader perspective, what the papers insist on depicting as the “gravest concern in the Church” is only a marginal issue for the Vatican. Yet, it would be contrary to the press’s interests to cool emotions and put the issue into perspective. Instead, it appears that even quality press seems to occasionally indulge in the rhetoric of exaggeration and moral panic to attract audiences. The above analysis of quality and tabloid journalism brings them into sharper contrast, but the aim of this study has not been to perpetuate further oppositions between broadsheets and tabloids. It has been assumed that, due to a growing competition on the market, quality print media in Britain are likely to attract consumers by “tabloidizing” their news coverage. Through a predominantly qualitative analysis, the study focused on examining the scope of application of typically tabloid stylistic properties in quality news coverage. It has demonstrated that news outlets may strategically resort to linguistic devices that maximize the appeal to audiences, for example by sensationalizing, simplifying or personalizing news stories. The paper has also proposed a method to trace tabloidization by characterizing a rhetorical style of quality newspaper texts by juxtaposing it with tabloid rhetoric.

References Biressi, Anita and Heather Nunn. 2007. The tabloid culture reader. London: Open University Press. Blumler, Jay and Michael Gurevitch. 1995. The crisis of public communication. London: Routledge. Cappella, Joseph and Kathleen H. Jamieson. 1997. Spiral of cynicism: The press and the public good. New York/London: Oxford University Press. Coupland, Nikolas. 2007. Style: Language variation and identity. Cambridge: Cambridge University Press. Fairclough, Norman. 2000. New Labour, new language? London: Routledge.

250

Katarzyna Molek-Kozakowska

Jeffries, Lesley. 2010. Critical stylistics: The power of English. Basingstoke: Palgrave Macmillan. Johansson, Sofia. 2007. Reading tabloids: Tabloid newspapers and their readers. Stockholm: Sodertorns hogskola. Machin, David and Theo van Leeuwen. 2005. “Language style and lifestyle: The case of a global magazine.” Media, Culture and Society 27: 577–600. Newton, Kenneth. 1999. “Mass media effects: Mobilization or media malaise.” British Journal of Political Science 29: 577–599. Postman, Neil. 1986. Amusing ourselves to death. New York: Metheuen. Richardson, John. 2007. Analysing newspapers. Basingstoke: Palgrave Macmillan. Selting, Margaret. 1999. Communicative style. In Jeff Verschueren (ed.), Handbook of Pragmatics, 1–33. Amsterdam: John Benjamins. Tannen, Deborah. 2004. Conversational style: Analyzing talk among friends. 2nd edn. Oxford: Oxford University Press. Turner, Graeme. 1999. “Tabloidization, journalism and the possibility of critique.” International Journal of Cultural Studies 2.1: 59–76. Woffinden, Bob. 1996. “The easy journalism of child abuse.” British Journalism Review 7: 46–51.

PATIENTS, DOCTORS, DISEASES AND CASES: PATIENT-CENTRED AND OBJECTIFYING DISCOURSES IN MEDICAL CASE REPORTS MAGDALENA MURAWSKA

1. Introduction Case reports are “the briefest and simplest category of article” (Adams Smith 1984: 27), and a “reportage, a piece of scientific puzzle” (Hunter 1991: 94), which, after being solved by a physician, is recorded for other members of the medical community with a view to sharing a new piece of information. Taavitsainen and Pahta (2000) define the genre of case report in the following way: In its typical form, the case report records the course of a patient’s disease from the onset of symptoms to the outcome, usually either recovery or death. The background and a commentary on the disease are also given, but their scope may vary. Often a limited review of the literature is added and the number of known cases stated (2000: 60).

The aim of this genre is to describe a new disease or a disease which is already known but which has unusual manifestations. As regards their structure, case reports usually consist of three parts: “a short Introduction, a more detailed Case Report body, and a brief Comment or Discussion section” (Rowley-Jolivet 2007: 185), yet it is not a unified composition. Consequently, case reports start with Introduction which, as the name suggests, orients readers to the patient whose case is described. This section, which may also include some demographic facts about the patient such as sex, age, sometimes occupation or ethnic origin, briefly explains the reason of his/her presentation. What follows is Case Report body. It is the most extensive part, as it features not only the descriptions of diagnostic procedures of various kind but also the very treatment. Lastly, Comment or Discussion, which, often combined, evaluate the actions performed by a physician and discuss the outcomes and effectiveness of

252

Magdalena Murawska

treatment. The section may also include implications for further assessment and treatment procedures. This paper focuses on patient’s reference in medical case reports. Firstly, the theoretical background will be presented, i.e. the relation between scientific discourse in general and the context of its production – models of medical practice – as well as some analytical tools. Then the results of the analysis of patient’s reference will be given and the characteristic of the two discourses indentified in the texts at hand.

2. Theoretical background In this section, the theoretical background as well as some analytical tools will be presented. According to Bazerman (1988), scientific discourses are shaped by given disciplines (1988: 47). It follows that the ways in which academics inform about their scientific activities are influenced by modes of reasoning, methodologies, objectives, etc. of a given area of study (cf. Atkinson 2001; Taavitsainen and Pahta 2000). In other words, how researchers argue in scientific papers and the theories and methods they choose depend on a particular model which is practiced at a particular moment in a particular discipline. Following this line of reasoning, features of medical texts might be conditioned by the nature of medicine both as an area of study and of practice. As regards medical practice, the framework that has outlined the premises of how medicine has been practiced in western societies since the mid nineteenth century is the biomedical model, which views illness as a direct consequence of the diseased body and patients as mere recipients of treatment (cf. Wade and Halligan 2004: 1398). Therefore, the model is reductionist because it limits the understanding of disease only to its biological manifestations excluding social and psychological aspects. Yet, as Armstrong observes (1984), such one-dimensional perception of the patient has been changing from the fifties (1984: 739). Apart from the bodily sphere where diseases manifest themselves, also the social sphere and its influence upon the patient’s health came to be acknowledged. The author of this “boldest attempt” (Armstrong 2002: 66) was Balint (1956) who maintained that localising bodily manifestations was not enough to diagnose the patient. He claimed that what was missing was the elicitation of information about symptoms which then had to be placed against the background of the patient’s experience (Balint 1956, as cited in Armstrong 2002: 66). Consequently, observation began to be accompanied by patients’ interviews which were to “illuminate the dark spaces of the mind

Patients, Doctors, Diseases and Cases

253

and social relationships” (Armstrong 1984: 739). While diagnosing, physicians examined patients’ bodies to discern any signs of a disease and listened to the patients’ accounts of symptoms. “The patient was beginning to have a voice – and an existence – independent of pathological lesion” (Armstrong 2002: 65). Such a change in perception paved the way not only to the acknowledgement of the role of the patient’s account but also to subsequent improvement in the quality of doctor-patient communication (cf. Armstrong 2002: 67–70). Eventually, the change gave rise to two conceptual frameworks, namely “illness-centred medicine and personcentred” approach which marked the shift from the treated illness to the treated patient (Balint et al. 1970: 25–26). This clear differentiation led to further formulations which, although in various forms, shared the underlying criticism of the biomedical model and advocated the openness to the patient’s experience. For example, the biopsychological model (Engel 1977) is based upon the premise that the diagnosis of a disease is incomplete without the consideration of the psychological and social aspects of the patient’s life (Pereira and Smith 2006: 455). Such a holistic approach was in direct opposition to the reductionist biomedical model which took only biological facts as pertinent in the treatment of a disease (cf. Pellegrino 2001: 83). It will be demonstrated in the analysis that the two identified discourses in the texts at hand correspond to these two models of medicine.

3. Analytical tools In this subsection some analytical tools necessary for the analysis of patient’s reference will be introduced.

3.1 Isolatable entity and nominalist account of a disease The sociology of medicine, which approaches medicine critically, provides a number of notions and distinctions which prove to be instrumental in examining patient reference. According to the sociologists of medicine, the biomedical model (cf. section 2 above) conceptualises diseases as “it,” i.e. as an “isolatable entity” (Blois 1984), which manifests itself unchangingly in all patients. Associated with Plato, whose aim was to classify diseases, this mode of disease presentation is referred to as the nominalist mode. It allows to describe a disease as a purely abstract concept and separate from its context, i.e. the patient (cf. Blois 1984: 92). The following sentence exemplifies this mode:

254

(1)

Magdalena Murawska

A nine-year-old girl presented to the National Institutes of Health Clinical Center with weight gain, growth arrest, hypertension, abdominal striae, acne, hirsutism, proximal muscle weakness, mood swings, and increasing skin pigmentation. NEJM10

Although example (1) includes a direct reference to the whole patient, her symptoms and reactions (in bold) are enumerated as if they were entities which are not part of the patients’ experience of illness.

3.2 Metaphor Another analytical tool which allows to examine spatial configurations of patient imaging is the metaphor of a container. Introduced by Lakoff and Johnson (1980), it presents objects or notions as having an inside and outside and as being capable of holding something. As Lakoff and Johnson (1980) explain, “[w]e are physical beings, bounded and set off from the rest of the world by the surface of our skins, and we experience the rest of the world as outside us. Each of us is a container, with a bounding surface and an in-out orientation” (1980: 28). In medical discourse, the concept of disease in the patient is utilised to describe medical procedures or to give an account of medical facts. From this perspective, the patient’s body tends to be viewed as a container in which diseases are localised and particular treatment is performed. As regards example (1) above demonstrating the nominalist account of a disease, such a conceptualisation is referred to as an ontological metaphor, which through the “understanding of our experiences in terms of objects [...]” helps to describe them in a variety of ways (Lakoff and Johnson 1980: 25). In medical discourse, this conceptualisation facilitates the presentation of decontextualised information about diseases. Yet another metaphorical tool found in medical texts is metonymy which concerns either referring to wholes in terms of their parts and vice versa or “naming by association” (Knowles and Moon 2006: 37). From the perspective of the present paper, referring directly to the patient’s bodyparts as representing him/her can be given as an example pervasive in medical discourse. This way, the body-part is figuratively separated from its owner and “the sufferer is excluded from the ensuing treatment, which is directed toward the synecdochic sign […] [and] from the health professional’s perspective the patient becomes the affected body part” (Fleischman 1999: 22).

Patients, Doctors, Diseases and Cases

255

3.3 Sentence grammar As patient reference will be examined on the basis of exemplary sentences, a few terms regarding the structure and function of sentential elements need to be discussed. Traditionally and from a syntactic point of view, a sentence is thought to consist of two parts, namely the subject and the predicate, the latter comprised of object, complement and adverbial. As regards the subject and the object, position-wise, they precede and follow the verb respectively (in the case of a declarative sentence) (Greenbaum and Quirk 1990: 207). Functionally, the subject is “prominent positionally” in that it is a “perspectival centre” and “the starting of the communication of a sentence” (Smith 2003: 192–193). Supporting evidence for the prominence of the subject can also be found in psychological studies (Paivio 1979). Chafe (1976) compares the structure of a sentence to a package of information which is “unwrapped” step by step by a reader. Although it consists of many elements, “knowledge directly attached to the subject may be most immediately accessible” (1976: 44). Semantically, given a prototypical Active Voice sentence, the Agent, “the ‘doer’, or instigator of the action denoted by the predicate” (Aarts 1997: 88), is “the most topical participant,” “the apex of the topic hierarchy” (Givón 1990: 566), being the first candidate for the subject position. In contrast, in a prototypical Passive Voice sentence it is the Patient, “the ‘undergoer’ of the action or event denoted by the predicate” (Aarts 1997: 88), who/that is granted the title of “the most topical participant” (Givón 1990: 566) and the position of the subject. Indeed, if a given situation with participants involved is described, “they rank on a scale according to their importance” (Givón 1984: 137). Consequently, the subject and the direct object are primary and secondary clausal topics respectively (Givón 1984: 138). Clearly then, grammatical relations have their pragmatic consequences for the prominence of specific participants with their thematic roles (cf. Van Dijk 1980: 95–96). Underlying this reasoning is the approach to a sentence introduced by the linguists of the Prague School – Functional Sentence Perspective – in which semantic, grammatical and functional levels are considered (cf. Firbas 1974).

4. Data and methods The corpus for this study comprises 56 case reports taken from four international medical journals aimed at health professionals – The Lancet (15), The Journal of American Medical Association (13), The New England Journal of Medicine (16) and The British Medical Journal (12).

256

Magdalena Murawska

The rationale behind the choice of the genre of case report was the fact that this text-type does not present general medical knowledge or detailed results of clinical research, but it discusses particular patients suffering from particular diseases. In other words, case reports give the account of diagnosis and treatment but always in the real context referring to a given person. What is more, it is the genre which brings together all the elements of the medical management of the patient – recounting his/her history, performing physical examination/tests, as well as therapy and the results of treatment which are described in respective sections of the genre. This way, all the past and present aspects of the patient’s health are composed into a report of a case. In the analysis, each article was carefully read in search of any words that referred to the patients described there. Next, the examples containing references to the patients were isolated by means of WordSmith 5 and further examined. However, the examples which did not refer to the patients directly, yet concerned various aspects of their treatment, were also taken into consideration. In the following discussion of the results, first, the list of identified modes of patient reference in the texts at hand will be demonstrated. Then they will be classified as belonging either to patient-centred or objectifying discourses. Their characteristic features will also be discussed.

5. Results and discussion 5.1 Types of patient’s reference Depending on the type of patient reference and its sentential position, the following scale of modes of patient imaging in the examined corpus can be proposed: Patient’s perspective The patient underwent thymectomy, which did not lower her (2)

plasma corticotropin concentration. NEJM10 Patient as the object of medical procedure (3) We gave her intravenous heparin, then warfarin, and stopped the oral contraceptives. LA15

Patients, Doctors, Diseases and Cases

257

Patient as a location (4) Whipple's disease was first described in 1907 in a patient with migratory polyarthritis, cough, diarrhea with malabsorption, weight loss, and mesenteric lymphadenopathy. NEJM5 Patient’s body-part/aspect of condition/treatment (5) To treat her nonfunctional kidney and scarred bladder, we then did a left nephrectomy and caecocystoplasty. LA6 (6) I needed to establish the extent to which her management could be evidence based and, where possible, to determine the optimum management using evidence based medicine. B3 (7) His urinary symptoms settled after a two week course of ciprofloxacin and diclofenac. B12 Indirect reference – body-part (8) Imaging studies of the neck, chest, abdomen, and pelvis were normal. NEJM10 Patient as a case (9) Two of the 18 US cases of inhalational anthrax reported prior to the recent bioterrorism-related outbreak had underlying lung disease; one had beryllium exposure and chronic pulmonary fibrosis and the other had underlying pulmonary sarcoidosis. JA3 Abstraction from the patient (10) Repeated vitreous examinations revealed lymphocytes, neutrophils, and foamy macrophages and slender rod-shaped structures suggestive of bacteria that were positive on periodic acid–Schiff (PAS) staining, but no malignant cells. NEJM5 The above-given scale begins with three sentences referring to whole persons (2–4) which occupy three different positions: the subject, the object and the complement of a prepositional phrase. In the case of (4), the patient functions as a location (the container metaphor, cf. section 3.2. above) where a disease is found. Moving down the scale, more impersonal are sentences which refer only to various aspects of the patient’s health and treatment (6 and 7) or to body-parts (5). In (8) only the definite article points to whose body-parts are mentioned. Lastly, various diagnostic or treating procedures can be described without those to whom they apply (10). Consequently, language-wise, the scale starts with the nouns and a personal pronoun referring to the whole person with varying degrees of

258

Magdalena Murawska

sentential prominence (cf. 2– 4), then moves to possessive pronouns (or nouns in the possessive) referring only to patients’ bodies, symptoms or the aspects of treatment (cf. 5–7), and ends with sentences in which individual body-parts (8) and the treatment (10) are described in abstraction from the person. The patient may also be referred to as a case (9). The word case is used very frequently in medical case reports to refer to individual occurrences of particular diseases, but it may also be used to refer to patients. These varying forms of reference can be classified as various modes of patient imaging, ranging from patient-centred to objectifying ones.

5.2 The patient-centred discourse As there is a mode of practicing medicine which encourages patient’s active participation in the process of diagnosis and treatment and takes into consideration his/her experience of illness (cf. section 3.1. above), so there is a patient-centred discourse which reflects the model linguistically. The acknowledgement of the patient’s presence in medical texts is achieved in two ways, namely by foregrounding them as well as recognising them as experiencing individuals. Firstly, the qualitative analysis has shown that patients come to the fore when they are presented as active participants in the process of diagnosis and treatment. Such a mode of writing is possible when their volitional actions are described: (11) (12)

At that time, the patient reported difficulty in concentrating, memory loss, and occasional loose stools. NEJM5 A 64-year-old woman presented to the emergency department with a stiff painful jaw. LA14

Sentences (12) and (11) portray patients as visiting the health unit and reporting their health problems respectively. The patients function here as the Agents of the sentences, i.e. the instigators of the actions denoted by the verbs. As has already been mentioned, participants in a given sentence can be compared to actors in a little drama (cf. Aarts 1997: 80). The roles they play are hierarchically ranked, and the Agent is the most prominent participant. What is more, being the Agents of the sentences, patient references hold the most prominent sentential position, namely the Subject position, which makes them the referents of the rest of sentence (cf. section 3.3 above). As a result, their significance in terms of semantics and syntax foregrounds them in the texts. Such examples can be found in the first

Patients, Doctors, Diseases and Cases

259

section of the case reports, i.e. Introduction, where the account of the patient’s presenting condition is given, though from the physician’s perspective. Therefore, such a mode of imaging patients is conditioned primarily by the aim of the section of a case report. Yet, as case reports overwhelmingly describe medical procedures as performed on patients and not their experience of illness, the type of information as presented above is limited. Nevertheless, even if a patient is described as undergoing treatment, which means that he/she is not the Agent, it is still possible to foreground him/her in a sentence. This can be done by keeping him/her in the Subject position (cf. 11 and 12 above) or using the Passive Voice (cf. 15 below). (13) (14) (15)

He underwent tracheostomy and was ventilated mechanically for the next few days. B5 The patient deteriorated rapidly with swinging pyrexia of 41ºC and died 2 weeks later, in September, 2003. LA3 He was treated with broadspectrum antibacterial agents (i.e., vancomycin, ceftriaxone, and metronidazole) and antivirals (i.e., acyclovir and foscarnet). JA2

As a result of the meaning of the verbs in (13), (14) and (15), the patients take the thematic role of Patient. Nevertheless, their position is still central, as they are those to whom the sentences refer in the context of particular treatment (13 and 15) and condition (14). This effect is possible thanks to the location of patient references in the Subject positions, as in (13) and (14). In the case of (15), where the patient, due to the meaning of the verb, cannot be the Agent and is not in the Subject position in the Active Voice, the Passive Voice can be used. This option enables Patients to be the sentential topics while the Agents, in this case physicians who perform the described procedures, are textually absent. Consequently, although patients’ status in examples (13), (14) and (15) is different from, for example (11) or (12) above, the readers’ attention is still drawn to the patients as being managed or experiencing changes of their conditions and not exclusively to treatment or diseases. It follows that even if a given section of the case report does not deal with patients’ volitional actions, linguistic means make it possible to concentrate on the patients. The foregrounding of patients in sentences seems also a matter of the author’s choice of linguistic resources and how he/she wishes to phrase specific information. By contrast, the second element of the patient-centred discourse, i.e. the presentation of the patient’s experience of illness is a categorical choice.

Magdalena Murawska

260

Such information is either present, i.e. thought as relevant, or not. Table 1 below lists selected textual references to the patient’s mental or bodily experience of illness, divided according to the parts of speech, whereas the subsequent examples demonstrate these references in their particular sentential co-texts. Part of speech Nouns

Verbs Adjectives/phrases

Examples anxiety, pain, complaint, problem, fatigue, ache(s), difficulty, disorientation, confusion, discomfort, malaise, weakness, soreness, distress, satisfaction; complain, suffer from, feel, experience; confused, disoriented, painful, painless, worried, depressive, surprised, determined, searing, out of sorts, groggy, sore;

Table 1. References to the patient’s experience of illness. (16) (17) (18) (19)

Although the man complained of a headache for the next 24 hours, he recovered completely. JA6 Having sustained two fractures from simple falls, my patient was understandably worried about further fractures. B3 He immediately felt rectal pain; four hours later he developed dysuria, frequency, and frank haematuria. B12 He developed acute pelvic discomfort at the time of injection, followed by urgency, dysuria, fever, and aching in the left testis. B12

As the examples in Table 1 above demonstrate, the authors of medical case reports under study acknowledge the way patients react to illness in a two-fold manner. It can be presented either as an “isolatable entity” (cf. section 3.1 above) associated and, therefore, mentioned with a particular condition (cf. 19) or described as a state that can be experienced (cf. 16, 17 and 18). In the former case, the means of expression are nouns which allow to list symptoms, while in the latter case the verbs which denote particular experience, and adjectives which describe the state in which the patient finds himself/herself. Additionally, there is one adjective and a phrase that seem to belong to lay lexicon – groggy and out of sorts – rendering information closer to the patient’s account.

Patients, Doctors, Diseases and Cases

261

The sentences below Table 1 describe patient’s discomfort. The verbs: complain, feel, ache and the adjectives: worried on the one hand, and nouns: pain, discomfort, difficulty on the other, denote the experience of particular reactions and states or refer to them as abstracted elements of a disease respectively (cf. Table 1 above). Yet, regardless of whether listed or described as an experience, all these manifestations of how patients feel are mentioned in the co-texts where patients are referred to as whole persons (in contrast to their body-parts only) and where they are the subjects of the sentences, i.e. the point of reference of the sentence. Such instances can be found primarily in the Introduction section where the patient’s account of his/her health is presented. Occasionally, similar examples may also appear in the Case Report body, where some references can be made to how the patient feels during, for example, surgery, and in Comment/Discussion, where the outcome of treatment is evaluated and patient’s feelings about his/her condition may be included. Furthermore, with reference to the above-presented two models of medical practice (cf. section 2 above) patient-centred discourse, referring to the patient as an active participant in the process of diagnosis and treatment and acknowledging his/her experience of illness, corresponds to patient-centred medicine. However, as some counterexamples below demonstrate, in some cases the authors may choose to convey given information in a different way.

5.3 The objectifying discourse It has been said in section 2 above that biomedical model is reductionist, because it focuses solely on the biological manifestations of diseases and treatment procedures and not on the patient’s experience of illness. The discourse which corresponds to this model is the objectifying discourse, which objectifies patients in three respects. Firstly, patients’ mental/somatic experience is given without the reference to the experiencers. Secondly, the texts separate patients from their bodies, either writing solely about the affected/treated body-parts or abstracting from their owners. Thirdly, also medical procedures are either foregrounded or dealt with without mentioning those who undergo them. For one thing, as has already been stated, the textual presence of the patient’s experience of illness is one of the characteristic features of the patient-centred discourse. Examples (11–19) above clearly portray this experience as contextualised in particular patients who are referred to as whole persons. In contrast, mentioning pain or any other sensations without reference to an experiencer may not be treated as the patient-focused discourse.

262

(20) (21)

Magdalena Murawska

We started a magnesium infusion to maintain ionised magnesium levels of 1·5–2·0 mmol/L, as muscle spasms were consistently worse once the serum magnesium fell below 1·5 mmol/L. LA14 The development of pelvic pain at the time of injection and urinary symptoms within a few hours was highly suggestive of a misplaced injection into the prostatic or periprostatic tissue. B12

The above-given sentences enumerate two bodily reactions – muscle spasms and pelvic pain – and abstract them from their experiencers. Instead, the focus falls here on the treatment procedure while patients’ bodily responses are described as accompanying biological conditions. Such instances treat the patients themselves and their reactions or symptoms separately and, consequently, do not acknowledge patients’ suffering or any other discomfort stemming from the abnormal condition. As a result, the reactions and symptoms seem to have value in themselves without the need of contextualising them in patients. The second group of examples includes sentences in which it is the patient’s body-parts that take centre stage: (22) (23)

One month later the spleen was no longer palpable; at six months, the size of the spleen was normal (13 cm) on ultrasonography. NEJM2 The left side of the chest was not explored. NEJM10

What these sentences concern is the condition of the body-part (organ, etc.) itself (22) or its treatment (23) and not the patient whom this surgery concerns. These examples, i.e. where parts stand for wholes – patients, represent a metonymic type of reference (cf. section 3.2 above). Here bodyparts which are treated or whose condition is assessed seem to exist individually in the text. Their separation is an assumed feature, firstly, by virtue of the specificity of the genre that describes diseases in particular patients, and, secondly, of the local character of treatment. It follows that even if the patient is not mentioned, every time patient-related issues are dealt with, readers know who they are reading about. What is more, admittedly, when a disease is described, reference will be made more frequently to the treated body-parts than to the whole persons. Consequently, the patient becomes objectified as he/she is reduced to his/her body, on which medical procedures are performed. It needs to be added, however, that if the body-part is accompanied by a possessive pronoun, then a link between the body and its owner is established.

Patients, Doctors, Diseases and Cases

263

Thirdly, there are also sentences which, again, abstract from the patient but this time while describing various medical procedures – from tests to surgery. Regardless of whether they concern the whole patient or his/her body-part, patients may be textually absent. (24) (25)

Physical examination confirmed the previous findings and did not show any warning signs or indications of progressive involvement. B10 On the evening of admission, transplantation of a cadaveric kidney into the left iliac fossa was performed without incident, except that endarterectomy of the left common iliac artery was necessary before creation of the anastomosis; […] NEJM13

Admittedly, the sentences describe medical procedures and tests by means of which information concerning patient’s health is obtained. Therefore, specific diagnostic devices are presented as showing (24) whereas the operation is the primary topic of sentence (25). Yet, the textual absence of patient references contributes to the instructive character of these examples which seem to be telling how to perform a given procedure rather than describing patient management.

6. Conclusion The aim of this paper has been to present the many ways in which patients can be referred to in medical case reports. In the study, first, the list of possible reference modes has been compiled, ranging from the patient’s perspective through body parts to total abstraction from the patient. Next, these different types have been classified into two discourses that the authors of the texts avail themselves of. All in all, the patient-centred discourse is clearly in line with the broadly understood patient-centredness in medicine. Firstly, the focus falls on the whole person of the patient, which is achieved by the use of personal pronouns or nouns referring to people and not to the parts of their bodies/aspects of condition/treatment, and by placing the words indexing patients in the Subject position. The latter is possible both in active and passive sentences, depending on the semantic features of verbs. As a result, regardless of the patient’s thematic status as Agents or Patients, they are the topics of the sentences. Secondly, patient’s experience of illness is presented, which, though filtered through the author’s/physician’s lenses, clearly acknowledges the psychological (patient’s feelings) and social (patient’s inability to perform certain activities) aspects of illness. Consequently, rather than presenting them as

264

Magdalena Murawska

entities separated from an experiencer, conditions, symptoms, or reactions are literally felt by the patient, therefore, incorporated into patient’s experience. It follows that the patient-centred discourse focuses on the whole person and recognises other than biological aspects of ailment. The result is a narrated story of illness which is contextualised in an experiencing individual, something that is not possible in the biomedical model. As regards the objectifying discourse, it derives its name from the fact that this mode of imaging renders patients invisible while giving the account of their somatic/mental reactions and medical procedures. In other words, the discourse does not provide a textual link that would connect patients to their bodies and would index patients as experiencing individuals. As a result, this discourse can correspond to the biomedical model.

References Aarts, Bas. 1997. English syntax and argumentation. Houndmills: Macmillan Press. Adams Smith, Diana E. 1984. “Medical discourse: Aspects of author’s comment.” English for Specific Purposes 3: 25–36. Armstrong David. 1984. “The patient’s view.” Social Science & Medicine 24.8: 651–657. —. 2002. A new history of identity. A sociology of medical knowledge. Basingstoke: Palgrave. Atkinson, Dwight. 2001. Scientific discourse across history: A combined multi-dimensional/rhetorical analysis of the Philosophical Transactions of the Royal Society of London. In Douglas Biber, Susan Conrad and Randi Reppen (eds.), Corpus linguistics: Investigating language structure and use, 45–65. Cambridge: Cambridge University Press. Balint, Michael. 1956. The doctor, his patient and the illness. London: Pitman. Balint, Michael, John Hunt, Dick Joyce, Marshall Marinker and Jasper Woodcock. 1970. Treatment or diagnosis: A study of repeat prescriptions in general practice. London: Tavistock. Bazerman, Charles. 1988. Shaping written knowledge. The genre and activity of the experimental article in science. Madison: The University of Wisconsin Press. Blois, Marsden S. 1984. Information and medicine. Berkeley: University of California Press.

Patients, Doctors, Diseases and Cases

265

Chafe, Wallace. 1976. Giveness, contrastiveness, definiteness, subject, topic and point of view. In Charles N. Lee (ed.), Subject and topic, 25– 55. New York: Academic Press. Engel, George. L. 1977. “The need of a new medical model: A challenge for biomedicine.” Science 196: 129–136. Firbas, Jan. 1974. Some aspects of the Czechoslovak approach to problems of functional sentence perspective. In Frantisek Danes (ed.), Papers on Functional Sentence Perspective, 11–37. Prague: Academia. Fleischman, Suzanne. 1999. “I am…, I have…, I suffer from…: A linguists reflects on the language of illness and disease.” Journal of Medical Humanities 20.1: 3–32. Givón, Talmy. 1984. Syntax. A functional-typological introduction. Vol. 1. Amsterdam: John Benjamins Publishing Company. —. 1990. Syntax. A functional-typological introduction. Vol. 2. Amsterdam: John Benjamins Publishing Company. Greenbaum, Sidney and Randolph Quirk. 1990. A student’s grammar of the English language. Harlow: Longman. Hunter, Kathryn Montgomery. 1991. Doctors' stories. The narrative structure of medical knowledge. Princeton: Princeton University Press. Knowles, Murray and Rosamund Moon. 2006. Introducing metaphor. London: Routledge. Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago: Chicago University Press. Paivio, Allan. 1979. Imagery and verbal processes. Hillsdale: Lawrence Erlbaum. Pellegrino, Edmund D. 2001. Physician and philosopher. Charlottesville, VA: Carden Jennings Publishing. Pereira, M. Graca and Thomas Edward Smith. 2006. “Evolution of the biopsychological model in the practice of family therapy.” International Journal of Clinical and Health Psychology 6.2: 455–467. Rowley-Jolivet, Elizabeth. 2007. A genre study of If in medical discourse. In Kjersti Fløttum (ed.), Language and discipline perspectives on academic discourse, 176–201. Newcastle: Cambridge Scholars Publishing. Smith, Carlota S. 2003. Modes of discourse. The local structure of texts. Cambridge: Cambridge University Press. Taavitsainen, Irma and Päivi Pahta. 2000. “Conventions of professional writing: The medical case report in a historical perspective.” Journal of English Linguistics 28.1: 60–76.

266

Magdalena Murawska

Van Dijk, Teun A. 1980. Macrostructures. An interdisciplinary study of global structures in discourse, interaction, and cognition. Hillsdale: Lawrence Erlbaum Associates, Publishers. Wade, Derick T. and Peter W. Halligan. 2004. “Do biomedical models of illness make for good healthcare systems?” The British Medical Journal 329: 1398–1401.

Primary sources Arioglu, Elif, John Doppman, Mario Gomes, David Kleiner, David Mauro, Carolee Barlow and Dimitris A. Papanicolaou. 1998. “Cushing’s syndrome caused by corticotropin secretions by pulmonary tumorlets.” The New England Journal of Medicine 339.13: 883–886. NEJM10 Ash, Samanta and Jo Beardsley. 1999. “Sciatica: Which intervention?” The British Medical Journal 319: 302–303. B10 Barakat, Lydia A., Howard L. Quentzel, John A. Jernigan, David L. Kirschke, Kevin Griffith, Stephen M. Spear, Katherine Kelley, Diane Barden, Donald Mayo, David S. Stephen, Tanja Popovic, Chung Marston, Sherif R. Zaki, Jeanette Guarner, Wun-Ju Shieh, H. Wayne Carver II, Richard F. Meyer, David L. Swerdlow, Eric E. Mast and James L. Hadler. 2002. “Fatal inhalational anthrax in a 94-year-old Connecticut woman.” The Journal of American Medical Association 287: 863–868. JA3 Bernstein, R. and S. Futterer. 2004. “Venous unresponsiveness.” The Lancet 363: 368. LA15 Bullock, Nigel. 1997. “Impotence after sclerotherapy of haemorrhoids: Case reports.” The British Medical Journal 314: 419. B12 Lo, Steven, Jeremy Noble, Ian Bowler and Brian Angus. 2004. “Dysuria and a headache.” The Lancet 364: 1554. LA6 Mak, Ryan, G.C. Gray, M.D. Malasig, L.N. Binn, L.V. Asher, D. Cute, S.C. Kehl, B.E. Dunn, J. Zablocki and A.J. Yund. 2001. “Two fatal cases of adenovirus-related illness in previously healthy young adults – Illinois, 2000.” The Journal of American Medical Association 286: 553– 556. JA2 Mulrow, Cynthia D. 1998. “Helping an obese patient make informed choices.” The British Medical Journal 317: 266–267. B5 Parker, Martyn, J. 2000. “Managing an elderly patient with a fractured femur.” The British Medical Journal 320: 102–103. B3 Rickman, Leland S., William R. Freeman, W. Richard Green, Sandy T. Feldman, Janet Sullivan, Valentina Russack and David A. Relman.

Patients, Doctors, Diseases and Cases

267

1995. “Uveitis caused by tropheryma whippelii (Whipple's Bacillus).” The New England Journal of Medicine 332: 363–366. NEJM5 Sayegh, Mohamed H. and Robert B. Colvin. 2003. “Case 8-2003: A 35year-old man with early dysfunction of a second renal transplant.” The New England Journal of Medicine 348.11: 1033–1045. NEJM13 Tucker, M., B. Eichold II, K. Micher, J.P. Lofgren, J. Schier, M. Belson, M. Patel and C. Rubin. 2005. “Brief report: Acute illness from dry ice exposure during hurricane Ivan – Alabama, 2004.” The Journal of American Medical Association 293.12: 1585–1586. JA6 Vinetz, Joseph M., Jun Li, Thomas F. McCutchan and David C. Kaslow. 1998. “Plasmodium malariae infection in an asymptomatic 74-year-old Greek woman with splenomegaly.” The New England Journal of Medicine 338: 367–371. NEJM2 White, J.M., R.D. Barker, J.R. Salisbury, A.J. Fife, S.B. Lucas, D.C. Warhurst and E.M. Higgins. 2004. “Granulomatous amoebic encephalitis.” The Lancet 364: 220. LA3

DISCURSIVE STRATEGIES IN THE MEDIA CONSTRUCTION OF POLAND, RUSSIA AND THE USA IN THE CONTEXT OF THE DEBATE ON THE US ANTI-BALLISTIC MISSILE DEFENCE SHIELD IN POLISH AND RUSSIAN QUALITY PAPERS AGNIESZKA SOWIēSKA AND TATIANA DUBROVSKAYA

1. Introduction This is the first study to analyse comparatively the media construction of Poland, Russia and the USA within the context of the debate on the US anti-ballistic missile (ABM) defence system in the Polish and Russian press.1 The issue of the deployment of the ABM system in Central and Eastern Europe (CEE) arose publicly in Poland in late 2004 (Wągrowska 2010). The negotiations concluded with an agreement signed on August 14, 2008. According to Hynek and Stritecky (2010: 179), the choice of the Czech Republic and Poland for hosting the ABM system “was justified by the favourable location of the two countries for a defence against Iranian ballistic missiles.” As they further maintain, the very timing of the conclusion of the agreement that took place after the Russia-Georgia conflict over South Ossetia flared up was not coincident. In his initial reaction, Polish PM Donald Tusk linked the conflict’s outcome to the fear that Poland could be the next country to be invaded by Russia. (Hynek and Stritecky 2010: 183) 1

This study is part of a larger research project on the discursive construction of “us” and “them” in the newspaper coverage on the ABM issue in Poland and Russia. The preliminary results were first presented at the CADAAD conference held in àódĨ in 2010.

Discursive Strategies in the Media

269

It seems that the Polish government’s legitimisation of the need for the ABM system through promotion of the alleged Russian threat to Poland was also strategically linked to the need for the modernisation of Polish armed forces (cf. Ministry of Foreign Affairs 2008). The negotiations and the US-Polish agreement caused a backlash among Russian leaders, military experts and the mass media (cf. Kunze and Bonet 2008). There were claims that Washington’s anti-missile shield was potentially intended to destroy Russian strategic missiles, could lead to the increase in ABM systems in Europe and, by the same token, ruin the established military balance in the world (Zolotarjov 2008). However, what was rather at stake was Russia’s public perception. As remarked by Moravcsik (2007): American weapons in Russia’s former satellite Poland offer Putin a tailormade opportunity to build himself up with high-profile speeches criticising the United States as ‘provocative’ and ‘imperial.’ Moravcsik (2007)

Barack Obama’s accession to the US presidential post marked a sharp U-turn in the ABM issue. In summer 2009 the US President visited Russia, initiating a new stage in relations between Russia and the USA, referred to as a “reset.” On September 17, 2009 Obama withdrew from the agreement, cancelling the project negotiated by the Bush administration and suggesting a new scheme for developing a missile defence architecture in CEE. The cancellation was followed by visits of American high officials to Poland and Russia: the US Vice-President Joe Biden’s visit to Poland and the US Secretary of State Hillary Clinton’s visit to Russia in October 2009. At present the ABM, though suspended, remains an important issue in Polish-US-Russian relations. The role of media in perpetuating particular representations while discrediting others cannot be underestimated. Regrettably, comparative analysis of media discourse is still a relatively unexplored research area in Russia and Poland, and there are rare studies in English coming from this part of the globe. As far as the authors are aware, there is a study by Turaev (2009), which centres on the framing analysis of the anti-missile shield issue in one American and one Russian newspaper. No other studies on that topic have been found. For that reason, it is believed that the presented analysis is particularly relevant. Specifically, the study aims to explore discursive strategies in the construction of Poland, Russia and the USA in the context of the debate on the US anti-missile shield in one Polish and two Russian papers. As such, it provides a picture of chronological transformations of discursive representations of Poland,

270

Agnieszka SowiĔska and Tatiana Dubrovskaya

Russia and the USA in the investigated papers from two different perspectives.

2. Method In the study, we draw on typical strategies involved in the discursive construction of social actors distinguished by critical discourse researchers, and their specific, context-sensitive linguistic realisations. The critical research enterprise, which originates in French Marxist structuralism and the Critical Theory of the Frankfurt School associated with Jürgen Habermas ([1981] 1984) in particular, has a long-standing record in research on media representations (cf. Wodak and Meyer 2009). Specifically, it is the discourse-historical approach (DHA) of Critical Discourse Analysis (CDA) that has extensively been applied to analysing media discourse (e.g. Reisigl and Wodak 2001; SowiĔska 2009). The DHA features five strategies used to construct the in-group and the outgroup: referential or nomination, predicational, argumentation, perspectivation, and intensifying and mitigation. More recently, Hart (2010) has proposed an alternative approach to the analysis of media discourse, which draws, inter alia, on referential and predicational (evaluative) strategies identified by CDA and, additionally, proximisation —originally discussed by Chilton (2004) and developed by Cap (2006). These strategies, namely, referential, predicational and proximisation, are explored in the present study. After Reisigl and Wodak (2009: 94), we take strategy to be “a more or less intentional plan of practices (including discursive practices) adopted to achieve a particular social, political, psychological or linguistic goal.” Specifically, referential or nomination strategies discursively construct social actors, objects, phenomena or events and processes or actions. They are accomplished mainly through pragmalinguistic (and linguistic) means, such as membership categorisation devices, deictic expressions, metaphors or metonymies. Predicational strategies positively or negatively evaluate social actors, objects, phenomena, events or processes and actions. As pointed out by Reisigl and Wodak (2009: 94), this evaluation may involve, inter alia, assigning evaluative attributions of negative or positive traits in the form of adjectives, prepositional phrases, relative clauses, conjunctional clauses, infinitive clauses or participial clauses, explicit predicates or predicative nouns, explicit comparisons, metaphors or other rhetorical figures (cf. Hart 2010: 49–88). It must be noted too that predicational strategies may also be realised by particular referential

Discursive Strategies in the Media

271

expressions and, conversely, particular predications can at the same time serve a referential function through semantic prosody (Hart 2010: 61). Finally, proximisation strategies presuppose referential and nomination strategies (Hart 2010: 84). While referential and predicational strategies are essentially semantic-level strategies, proximisation ones lend themselves to be operationalised on the conceptual level. Crucially, proximisation, as proposed by Cap (2006: 11), is a cognitive and sociopsychological concept. It draws on deixis and the speaker’s ability to present the events on the discourse stage as directly affecting the addressee, usually in a negative or threatening way, which, in turn, prompts immediate reaction. To put it differently, proximisation consists in the conceptualisation of antagonistic and alien entities (so-called outside-the-deictic-centre elements) entering the territory of the inside-thedeictic-centre (IDC) elements. Hart (2010: 84) contends that “spatial proximisation occurs when the threat depicted in the predication is presented as close to or approaching to the text consumer.” Conversely, in positive spatial proximisation, which is also possible, the IDC-instigated events would be construed as promising and beneficial (cf. Cap 2006). Temporal proximisation, as indicated by Cap, involves construing the events which take place in the spatial dimension as momentous and historic and hence of central significance to the discourse addressee, as well as to the speaker. (Cap 2006: 5)

In particular, it is representations of the threat in the predication as already taking place, having just taken place, about to take place or taking place regularly that define temporal proximisation (Hart 2010: 84). Much as proximisation is achieved through language constructs, it must be remembered that it cannot be equated with any particular linguistic forms (Cap 2006: 11). These strategies, we claim, are employed to effectively lay the ground for a dynamic negotiation of different ideas and multiple social actors’ standpoints (including a newspaper’s own viewpoint) and thereby their subsequent legitimisation or delegitimisation.

3. Data The data for the analysis come from one Polish centrist quality paper Gazeta Wyborcza (GW), and two Russian popular quality newspapers: liberal Izvestiya (IZV) and pro-government Rossijskaya Gazeta (RG). Articles were retrieved from the papers’ electronic archives using the keyword: ɉɊɈ ‘anti-rocket defence’ in Russian and tarcza ‘shield’ in

Agnieszka SowiĔska and Tatiana Dubrovskaya

272

Polish. Two Russian newspapers were selected in order to achieve the approximate number of articles available in the Polish press. In our view, there are two reasons for a smaller number of texts in Russian. Firstly, the Georgia-Russia conflict in South Ossetia in August 2008 distracted the mass media attention from the shield issue. Secondly, Russia was only indirectly involved in the events, while Poland was one of the key actors, which predetermined extensive newspaper coverage on that issue. The data cover only a few selected stages in the debate over the shield: signing the US-Polish agreement on August 20, 2008 (Stage I); the US President Barack Obama’s visit to Russia in July 2009 (Stage II); Obama’s decision to withdraw from the agreement announced on September 17, 2009, the US Vice-President Joe Biden’s visit to Poland on October 20, 2009 and the US Secretary of State Hilary Clinton’s visit to Russia on October 14, 2009 (Stage III). All articles pertaining to the subject of the study that covered the three stages as well as those that were published shortly before and after the events were selected for the analysis (see Table 1). Data

Stages Stage I

Russian corpus

Stage II Stage III Stage I

Polish corpus

Stage II Stage III

Total:

Dates August 17– 27, 2008 July 5–9, 2009 September 9– 24, 2009 October 13– 26, 2009 August 21– 27, 2008 July 4–8, 2009 September 18–25 and October 7–23, 2009

Number of Articles 7 6 18 8 4 23 66

Table 1. Description of the Polish and Russian corpora: the number of articles, their dates of publication and selected stages of analysis.

Discursive Strategies in the Media

273

Textual in-depth analyses of the retrieved Russian and Polish articles were carried out independently by two authors: the Polish texts were analysed by the Polish author and the Russian ones by the Russian author. Examples from the articles were subsequently translated into English by the authors, and after each stage the results obtained were discussed by the two researchers, bringing out similarities and differences between the representations of the selected social actors and discursive strategies.

4. Results The exploration of both Russian and Polish papers demonstrated that there are three main social actors who participated in the events and whose political interests were involved. These were Poland, Russia and the USA. In the texts, the actors are denoted by a number of nominations that are mostly neutral and based on metonymy: either geographical names or names of political figures and institutions which stand for the countries as social actors, e.g. Russia, the Kremlin, Poland, Warsaw, President KaczyĔski, the US, Washington, etc. Within the group of predicational strategies we further distinguished: – emotionalisation, which is accomplished lexically through explicit references to emotions; – deprecation, which involves negative evaluation and, more specifically, disapproval of actions performed by a social actor. In the case of deprecation, the fact that the action took place is not questioned; – accusation, which ascribes particular negatively evaluated actions to a social actor. More precisely, in the case of accusation, negative evaluation of a particular action prompted by the other social actor (or actors) is presupposed. However, the veracity of the action can be questioned; – inferiorisation, which consists in the presentation of the other social actor as inferior to the self. A necessary element of inferiorisation is comparison. Inferiorisation can only be accomplished in the context of explicit or implicit juxtaposition of two or more social actors; – positive self-presentation, which presupposes positive evaluation of one’s own actions and qualities; – negative foretelling, which involves giving a negative forecast for other social actors, foreseeing failures in or negative consequences of their actions.

274

Agnieszka SowiĔska and Tatiana Dubrovskaya

In what follows, we focus on both similarities and differences in the ways the social actors are represented discursively, starting with the description of Poland, then Russia and finishing with the USA. The data are considered chronologically.

4.1 Stage I: setting the scene and establishing relations The most salient strategy used in the construction of Poland in Stage I in the Polish paper is deprecation. Provincial and childish political elites (e.g. “a cabaret for two actors” [kabaret rozpisany na dwóch aktorów] (GW 26/8/08) – about the Polish President and Prime Minister’s behaviour during the signing of the US-Polish agreement) are juxtaposed with ordinary Poles – anxious and fearful of the shield. Giving the emotional account of the issue, i.e. employing emotionalisation, reinforces deprecation directed at the political elite and allows to highlight tensions within the Polish society caused by the ABM issue. The representation of Poland in the Russian media is mainly built on accusation and inferiorisation strategies. Poland is accused of being a mercenary social actor, who is ready to sell itself to the USA. The metaphoric description of the US-Polish agreement conjures up the picture of a market, of which bargaining is an essential part: (1)

Of course, almost no one had doubts about the strategic deal between the USA and Poland. However, the bargaining proved to be long – the Poles turned out to be less submissive than the Czechs.2 (IZV 17/8/08) [Ʉɨɧɟɱɧɨ, ɦɚɥɨ ɤɬɨ ɫɨɦɧɟɜɚɥɫɹ ɜ ɫɬɪɚɬɟɝɢɱɟɫɤɨɣ ɫɞɟɥɤɟ ɦɟɠɞɭ ɋɒȺ ɢ ɉɨɥɶɲɟɣ. Ɉɞɧɚɤɨ ɬɨɪɝ ɡɚɬɹɧɭɥɫɹ – ɩɨɥɹɤɢ ɨɤɚɡɚɥɢɫɶ ɤɭɞɚ ɦɟɧɟɟ ɫɝɨɜɨɪɱɢɜɵɦɢ, ɧɟɠɟɥɢ ɱɟɯɢ.]

The strategies of inferiorisation and negative foretelling are also employed in constructing the portrait of Poland. Specifically, Poland is depicted as unimportant (e.g. “a small Eastern European country” [ɧɟɛɨɥɶɲɚɹ ɜɨɫɬɨɱɧɨɟɜɪɨɩɟɣɫɤɚɹ ɫɬɪɚɧɚ] (RG 18/8/08)) and bound to lose:

2

For matters of clarity, phrases of interest are in bold print.

Discursive Strategies in the Media

(2)

275

In any case, constructing [Poland’s, T.D.] own ABM system, which now consists of 99% of outdated Soviet arms, will apparently cost Poland a fortune. (IZV 21/8/08) [ȼ ɥɸɛɨɦ ɫɥɭɱɚɟ ɫɨɡɞɚɧɢɟ ɫɨɛɫɬɜɟɧɧɨɣ ɫɢɫɬɟɦɵ ɉȼɈ, ɤɨɬɨɪɚɹ ɧɚ 99% ɫɨɫɬɨɢɬ ɢɡ ɭɫɬɚɪɟɜɲɟɝɨ ɫɨɜɟɬɫɤɨɝɨ ɜɨɨɪɭɠɟɧɢɹ, ɹɜɧɨ ɜɫɬɚɧɟɬ ɉɨɥɶɲɟ “ɜ ɤɨɩɟɟɱɤɭ”].

Negative foretelling in Example (2) is realised grammatically through the use of the future tense and emphasised with an evidential discourse marker “apparently”. This also serves the strategy of positive selfpresentation as it represents Russia as a source of knowledge and authority. Although there are scarce references to Russia in Stage I in the Polish paper, it is mostly negative representations that surface: Russia as a neoimperial superpower, an aggressor and a former Big Brother to Poland. These are accomplished by strategies of emotionalisation (Examples (3) and (4)) and negative proximisation (Example (5)): (3)

The nearest future will show whether the shield will increase Poland’s security or whether it will only be another justification for Moscow to lead even more aggressive politics towards its neighbours […]. (GW 21/8/08) [NajbliĪsze lata pokaĪą, czy tarcza zwiĊkszy bezpieczeĔstwo Polski, czy teĪ bĊdzie jedynie dodatkowym uzasadnieniem dla Moskwy, by prowadziü wobec sąsiadów jeszcze bardziej agresywną politykĊ [...].]

(4)

Everybody fears that the Russians [Polish ‘Ruscy’ expresses negative attitude, A.S.] are rutheless. (GW 23/8/09) [Wszyscy boją siĊ, Īe Ruscy są bezwzglĊdni.]

Emotionalisation and negative temporal and spatial proximisation emerge in small narratives of the inhabitants of Redzikowo (the would-be site for the shield), quoted by the paper. Temporal proximisation is realised by tense aspect, taken as an indicator of situations that are about to occur, at the same time leaving the analogy of a similar situation in the past in which Russia acted as an aggressor open for inference. On the other hand, spatial proximisation is accomplished by the nominal phrases expressing distance (e.g. “13 km” in Example (5)) and nouns that prompt the conceptualisation of movement, spread (e.g. “fire”) or location (e.g. “at her workplace”):

276

(5)

Agnieszka SowiĔska and Tatiana Dubrovskaya

Teresa thinks she will be threatened even at her workplace in Sáupsk. – Because the field of fire will be at least 13 km. […] When they attack the shield, it won’t hold out. (GW 23/8/08) [Pani Teresa uwaĪa, Īe bĊdzie zagroĪona nawet w pracy w Sáupsku. – Bo pole raĪenia wyniesie co najmniej 13 km. […] Jak zaatakują tarczĊ, nie przetrwa.]

Unlike the Polish paper, the Russian media construct a positive image of Russia as a strong-willed and decisive social actor. The media quote President Medvedev, who expresses Russia’s negative attitude to the placement of the shield in CEE: (6)

Medvedev: As for the rocket missiles and the ABM, we, of course, don’t like when new bases are heaped around Russia and new rockets appear [...]. We don’t like all this [...]. (IZV 27/8/08) [Ɇɟɞɜɟɞɟɜ: ɑɬɨ ɤɚɫɚɟɬɫɹ ɪɚɤɟɬɧɵɯ ɫɪɟɞɫɬɜ ɢ ɉɊɈ, ɧɚɦ, ɤɨɧɟɱɧɨ, ɧɟ ɧɪɚɜɢɬɫɹ, ɤɨɝɞɚ ɜɨɤɪɭɝ Ɋɨɫɫɢɢ ɜɵɪɚɫɬɚɸɬ ɧɨɜɵɟ ɛɚɡɵ, ɩɨɹɜɥɹɸɬɫɹ ɧɨɜɵɟ ɪɚɤɟɬɵ […]. ɇɚɦ ɜɫɟ ɷɬɨ ɧɟ ɧɪɚɜɢɬɫɹ […].]

The strategy of emotionalisation is complemented with the strategy of positive self-presentation, which draws the readers’ attention to Russia’s ability “to take decisive measures” ([ɩɪɟɞɩɪɢɧɹɬɶ ɠɟɫɬɤɢɟ ɦɟɪɵ] (RG 21/8/08)) as well as negotiate and overcome problems in relations with other social actors: (7)

Russia’s mission with the NATO pointed out that in the Russian capital all issues in further development of relations between Russia and the North Atlantic Alliance would be considered. (RG 21/8/08) [ȼ ɦɢɫɫɢɢ Ɋɨɫɫɢɢ ɩɪɢ ɇȺɌɈ ɭɬɨɱɧɢɥɢ, ɱɬɨ ɜ ɪɨɫɫɢɣɫɤɨɣ ɫɬɨɥɢɰɟ ɛɭɞɭɬ ɪɚɫɫɦɨɬɪɟɧɵ ɜɨɩɪɨɫɵ ɞɚɥɶɧɟɣɲɟɝɨ ɪɚɡɜɢɬɢɹ ɨɬɧɨɲɟɧɢɣ Ɋɨɫɫɢɢ ɫ ɋɟɜɟɪɨɚɬɥɚɧɬɢɱɟɫɤɢɦ ɚɥɶɹɧɫɨɦ.]

The image of the USA in the Polish paper is rather ambivalent. On the one hand, positive evaluation of the social actor is recorded, which involves allusions to Western values and, at the same time, exclusion of Russia and its construction in opposition to the rest of the world:

Discursive Strategies in the Media

(8)

277

By signing the agreement we have tightened the bond with America and the West, because all NATO […] no longer opposes the shield. […] We are a step closer to America and one away from Russia, which after its recent excesses in Georgia reminded the world that it is a threat to its neighbours. (GW 21/8/08) [Podpisując umowĊ o tarczy, związaliĞmy siĊ jeszcze mocniej z Ameryką i z Zachodem, bo caáe NATO […] juĪ siĊ tarczy nie sprzeciwia. JesteĞmy o krok bliĪej Ameryki i krok dalej od Rosji, która po ostatnich wyczynach w Gruzji przypomniaáa Ğwiatu, Īe jest zagroĪeniem dla sąsiadów.]

On the other hand, the strategy of negative proximisation is deployed in a similar vein to that used in the depiction of Russia in GW. In particular, the analogy between Russian and US aggression and imperialism is accomplished either explicitly (Example (9)) or, more implicitly, through the Big Brother metaphor, used with reference to both social actors (Example (10)). Example (9) reveals negative proximisation, realised through evoking the image of one of Poles’ favourite leisure activities in an inferential chain (having a plot of land, growing raspberries and other fruit and vegetables and doing jars), which now may be threatened by the USA: (9)

On the cupboard in jars – raspberries from garden plots. They may be one of the last ones because how to do jars when Americans will confiscate the plots? (GW 23/8/08) [Na szafce w sáoikach maliny z dziaáki. Byü moĪe jedne z ostatnich, bo z czego robiü weki, gdy Amerykanie zarekwirują dziaáki?]

Example (10) shows how the difference between the representation of the USA as a Strong Father and that of the USA as Big Brother is strategically obliterated. On the whole, these metaphors, whether used for the positive or negative evaluation of the USA, successfully communicate not only the US superiority and Poland’s dependence on the USA, hence inferiority, but may further trigger the inference of a specific complementarity of the two social actors. Such representation also enters into a dialogue with the presentation of Poland’s political immaturity and childishness, as discussed above.

278

(10)

Agnieszka SowiĔska and Tatiana Dubrovskaya

Mother Europe is too weak […] and we, Poles, need a Strong Father or Big Brother as usual. It doesn’t matter that the Brother doesn’t love us and won’t let into his house, it’s enough that we have here an element symbolising his power. (GW 27/8/08) [Matka Europa jest zbyt sáaba […], a nam, Polakom, potrzeba jak zawsze silnego Ojca lub Wielkiego Brata? To nic, Īe Brat za nami nie przepada i do wáasnego domu nie wpuszcza, grunt, Īe mamy u siebie jakiĞ element symbolizujący jego potĊgĊ.]

Similarly, in the Russian papers, negative proximisation emerges in the portrayal of the USA in Stage I: (11)

Judging by the content of the agreement and the rush with which it was signed, the project has nothing to do with Iranian nuclear threat and is directed against Russia. (IZV 17/8/08) [ɋɭɞɹ ɩɨ ɫɨɞɟɪɠɚɧɢɸ ɫɨɝɥɚɲɟɧɢɹ ɢ ɩɨ ɫɩɟɲɤɟ, ɜ ɤɨɬɨɪɨɣ ɨɧɨ ɡɚɤɥɸɱɚɥɨɫɶ, ɩɪɨɟɤɬ ɧɟ ɢɦɟɟɬ ɨɬɧɨɲɟɧɢɹ ɤ ɢɪɚɧɫɤɨɣ ɹɞɟɪɧɨɣ ɭɝɪɨɡɟ, ɚ ɧɚɩɪɚɜɥɟɧ ɩɪɨɬɢɜ Ɋɨɫɫɢɢ.]

Negative proximisation is accomplished through verb phrases and nouns of movement that conceptualise the shield as the US threat approaching Russia’s territory (“the American ABM movement towards our borders” [ɩɪɨɞɜɢɠɟɧɢɹ ɚɦɟɪɢɤɚɧɫɤɨɣ ɫɢɫɬɟɦɵ ɉɊɈ ɤ ɧɚɲɢɦ ɝɪɚɧɢɰɚɦ] (IZV 18/8/08); see also Examples (6), (11)).

4.2 Stage II: political and discursive rapprochement between the USA and Russia In both the Russian papers and the Polish GW, the focus is almost exclusively on Russian-American relations as a default option. Poland, if mentioned, occurs only in references to “Poland and the Czech Republic” or “Central and Eastern Europe” as a site for the shield. In the Russian papers, the strategy of emotionalisation is exploited to depict Russia as exuding optimism (e.g., “obamamania” or “obamaoptimism” (IZV 8/7/09)) and to construct Russia as a strong and influential actor: (12)

There is no doubt that at the Big 8 summit, which opens today in Italy, Russia will not feel uncomfortable. (RG 8/7/09) [ɇɟɬ ɫɨɦɧɟɧɢɣ, ɱɬɨ ɧɚ ɨɬɤɪɵɜɚɸɳɟɦɫɹ ɫɟɝɨɞɧɹ ɜ ɂɬɚɥɢɢ ɫɚɦɦɢɬɟ “ɛɨɥɶɲɨɣ ɜɨɫɶɦɟɪɤɢ” Ɋɨɫɫɢɹ ɧɟ ɛɭɞɟɬ ɫɟɛɹ ɱɭɜɫɬɜɨɜɚɬɶ ɧɟɭɸɬɧɨ.]

Discursive Strategies in the Media

279

The Russian press creates an ambivalent representation of the USA. On the one hand, through negative proximisation the USA is depicted as a social actor who may still present a threat. Moscow is said to still believe that “the American ABM (…) can be used against Russian strategic forces” [Ⱥɦɟɪɢɤɚɧɫɤɚɹ ɉɊɈ…ɦɨɠɟɬ ɛɵɬɶ ɢɫɩɨɥɶɡɨɜɚɧɚ ɩɪɨɬɢɜ ɪɨɫɫɢɣɫɤɢɯ ɫɬɪɚɬɟɝɢɱɟɫɤɢɯ ɫɢɥ] (RG 8/7/09). On the other hand, rapprochement between Russia and the USA is constructed discursively through the “step towards each other” metaphor (e.g. “Because this important, though very modest, step has been made, it was a win-win situation” [ɂ ɩɨɷɬɨɦɭ ɜ ɬɨɣ ɦɟɪɟ, ɜ ɤɨɬɨɪɨɣ ɷɬɨɬ ɜɚɠɧɵɣ, ɯɨɬɹ ɢ ɞɨɫɬɚɬɨɱɧɨ ɫɤɪɨɦɧɵɣ ɲɚɝ ɫɞɟɥɚɧ, ɜɵɢɝɪɚɥɢ ɨɛɟ ɫɬɨɪɨɧɵ] (IZV 08/7/09)). Finally, positive forecasts, conceptualising a promising future awaiting both Russia and the USA, accomplish the strategy of positive proximisation: (13)

The USA President started talking about most serious matters. He believes that the Russians and the Americans have ‘a huge potential’ for economic development and the construction of ‘a new security structure in Europe and the world.’ (RG 8/7/09) [ɉɪɟɡɢɞɟɧɬ CɒȺ ɡɚɝɨɜɨɪɢɥ ɨ ɤɪɚɣɧɟ ɫɟɪɶɟɡɧɵɯ ɜɟɳɚɯ. Ɉɧ ɫɱɢɬɚɟɬ, ɱɬɨ ɭ ɪɨɫɫɢɹɧ ɢ ɚɦɟɪɢɤɚɧɰɟɜ ɟɫɬɶ “ɨɝɪɨɦɧɵɣ ɩɨɬɟɧɰɢɚɥ” ɞɥɹ ɷɤɨɧɨɦɢɱɟɫɤɨɝɨ ɫɨɬɪɭɞɧɢɱɟɫɬɜɚ ɢ ɫɨɡɞɚɧɢɹ “ɧɨɜɨɣ ɤɨɧɮɢɝɭɪɚɰɢɢ ɛɟɡɨɩɚɫɧɨɫɬɢ ɜ ȿɜɪɨɩɟ ɢ ɦɢɪɟ”.]

The discursive rapprochement of these actors is also evidenced in the depiction of the Russian-American relations of Stage II in the Polish paper, where infatuation metaphors, making up the strategy of emotionalisation, are used: (14)

[…] there is some ‘chemistry’ between them [between Medvedev and Obama, A.S.]. (GW 4/7/09) [[…] jest miĊdzy nimi jakaĞ ‘chemia’]

4.3 Stage III: the USA as a dubious social actor Characteristic of Stage III in the Polish paper is a blend of emotionalisation (Example (15)) and deprecation strategies (Example (16)), evident in the depiction of Poland. Interesting is the frequent reference to Poland and the Czech Republic as a collective social actor to present the close affinity between both countries in their initial reaction towards the Obama administration’s cancellation of the project. On the

280

Agnieszka SowiĔska and Tatiana Dubrovskaya

one hand, there is disappointment present in both states (e.g. “felt put aside by America […]” [poczuáy siĊ odsuniĊte na bok przez AmerykĊ […]] (GW 7/10/09)); on the other hand, ordinary citizens show quite opposite emotions (e.g. “And in Redzikowo they are happy” [A w Redzikowie siĊ cieszą] (GW 18/9/09)). Further, the strategy of deprecation is potentially intended to communicate a constructive criticism of US-Polish relations. It is hoped that the fiasco of building the anti-missile shield “will act like a cold shower on Poland” and that Polish relations with the USA will become “normal,” and no longer characterised by idealism (“ […] fiasko planów budowy tarczy antyrakietowej podziaáa jak zimny prysznic i stosunki z USA staną siĊ normalne, a nie nacechowane idealizmem” GW 21/9/2009). The picture of naïve and gullible Polish political elites (Example (15) is juxtaposed with that of an American unreliable and self-interested ally (see also Example (19)): (15)

Americans are kidding us and we are believing. (GW 7/10/09) [Amerykanie bajają, a my wierzymy.]

However, Poland is also portrayed as an actor capable of playing an important and constructive role in Europe: (16)

Poland should play a role of a country, which pushes both giants [the USA and Europe, A.S.] towards one another. Such rapprochement is in the interests of both the USA and Europe. (GW 18/9/09) [Polska powinna odgrywaü rolĊ kraju, który oba kolosy popycha ku sobie. Takie zbliĪenie leĪy w interesie i USA, i Europy.]

The similarity between the strategies of presenting Poland in the Polish and Russian newspapers is discovered with respect to emotionalisation. This is used to construct an unfavourable image of Poland as a disappointed, dependent and insufficient social actor: (17)

[…] Washington, which disappointed the Poles with its decision to postpone the construction of the ABM on the country’s territory. (RG 22/9/09) [[..] ȼɚɲɢɧɝɬɨɧ, ɤɨɬɨɪɵɣ ɪɚɡɨɱɚɪɨɜɚɥ ɩɨɥɹɤɨɜ ɫɜɨɢɦ ɪɟɲɟɧɢɟɦ ɨɬɥɨɠɢɬɶ ɫɬɪɨɢɬɟɥɶɫɬɜɨ ɉɊɈ ɧɚ ɬɟɪɪɢɬɨɪɢɢ ɫɬɪɚɧɵ.]

Discursive Strategies in the Media

281

It is noteworthy that in Example (17) the predicate “disappointed” grammatically refers to the USA, but semantically – to Poland. In the sentence, the Poles play the role of the ‘affected’ by Washington’s actions. It can be inferred from this statement that Poland is a weak actor, subject to the US decisions and will. Feelings and emotions of CEE after the suspension of the agreement on the ABM are painted exclusively in dark colours. In some contexts the political insufficiency of the Poles and Czechs is emphasised by a reference to their jealousy (“the former countries of the socialist camp are too jealous about developing the dialogue between Russia and the USA” [ɛɵɜɲɢɟ ɫɬɪɚɧɵ ɫɨɰɥɚɝɟɪɹ ɫɥɢɲɤɨɦ ɪɟɜɧɢɜɨ ɨɬɧɨɫɹɬɫɹ ɤ ɪɚɡɜɢɬɢɸ ɞɢɚɥɨɝɚ ɦɟɠɞɭ Ɋɨɫɫɢɟɣ ɢ ɋɒȺ] (RG 18/9/09)). This reference presupposes the seeming impossibility of building satisfactory three-side relations between CEE, Russia and the USA. The image of Russia in Stage III in the Polish paper is decidedly more indistinct than in the previous stages. No longer is the representation of Russia completely negatively-loaded as it was in Stage I. Rather, even though the previously assigned role of a protester is retained (e.g. “Russia has always protested against permanent US bases in the former Eastern Block countries [Rosja zawsze protestowaáa przeciw staáym bazom USA w krajach dawnego bloku wschodniego]; GW 25/9/09), it is the representation of Russia’s distrust towards other social actors that becomes prominent (Example (18)). Significantly, this overlaps with the presentation of Russia in the Russian press. (18)

So what that Polish and American politicians declared that the shield does not threaten Russia. Moscow did not believe them. (GW 21/9/09) [CóĪ z tego, Īe politycy polscy i amerykaĔscy deklarowali, Īe tarcza Rosji nie zagraĪa. Moskwa im nie wierzyáa.]

Importantly, the social actor that is the most frequently referred to in Stage III in the Polish media is the USA. The Polish paper deploys the strategy of accusation to present the USA as a dishonest actor, looking after its own interests only: (19)

The pattern is always the same. At first Americans pull virtual proposals of strengthening Poland out of a hat in order to receive something concrete in exchange. (GW 7/10/09) [Schemat jest zawsze taki sam. Najpierw Amerykanie wyciągają jak z kapelusza wirtualne propozycje wzmocnienia Polski, aby otrzymaü coĞ konkretnego w zamian.]

282

Agnieszka SowiĔska and Tatiana Dubrovskaya

However, it should be noted that this negative evaluation of the USA is mitigated by the strategy of deprecation used to negatively evaluate Polish political elites. In Stage III, emotionalisation is practised intensely in the Russian papers to construct not only Poland, but also Russia and the USA. However, emotions and feelings referred to differ considerably. Unlike Poland, Russia is represented as filled with hopes and positive spirit: (20)

The leaders of Foreign Ministries have not started their meetings in such a good mood for many years. During the 3 hours of negotiations they did not give each other any reasons to lose the positive spirit. (RG 13/10/09) [ɋ ɬɚɤɢɦ ɯɨɪɨɲɢɦ ɧɚɫɬɪɨɟɧɢɟɦ ɪɭɤɨɜɨɞɢɬɟɥɢ ɜɧɟɲɧɟɩɨɥɢɬɢɱɟɫɤɢɯ ɜɟɞɨɦɫɬɜ Ɋɨɫɫɢɢ ɢ ɋɒȺ ɧɟ ɧɚɱɢɧɚɥɢ ɫɜɨɢ ɜɫɬɪɟɱɢ ɦɧɨɝɢɟ ɝɨɞɵ. Ɂɚ ɬɪɢ ɱɚɫɚ ɩɟɪɟɝɨɜɨɪɨɜ ɫɬɨɪɨɧɵ ɧɟ ɞɚɥɢ ɞɪɭɝ ɞɪɭɝɭ ɩɨɜɨɞɨɜ ɪɚɫɬɟɪɹɬɶ ɩɨɥɨɠɢɬɟɥɶɧɵɣ ɧɚɫɬɪɨɣ.]

The commonality of positive emotions experienced by Russia and the USA can be construed as an element of rapprochement. Russia’s positive self-presentation contributes to the overall picture of a powerful and influential social actor: (21)

...Washington and its allies seem to have come to the conclusion that it will not be possible to do without Russian means in the sphere of the ABM. (RG 26/10/09) […ȼɚɲɢɧɝɬɨɧ ɢ ɟɝɨ ɫɨɸɡɧɢɤɢ, ɩɨɯɨɠɟ, ɩɪɢɲɥɢ ɤ ɜɵɜɨɞɭ, ɱɬɨ ɛɟɡ ɪɨɫɫɢɣɫɤɢɯ ɜɨɡɦɨɠɧɨɫɬɟɣ ɜ ɨɛɥɚɫɬɢ ɉɊɈ ɨɛɨɣɬɢɫɶ ɧɟ ɩɨɥɭɱɢɬɫɹ.]

Although the Russian press points to a possible reconciliation in Russian-American relations, there are still indications of a threat from the USA. A new configuration of the ABM is claimed “to become for us much more dangerous” [ɤɭɞɚ ɛɨɥɟɟ ɨɩɚɫɧɨɣ] (IZV 13/10/09). Moreover, the USA is explicitly accused of numerous heinous actions with dramatic consequences for the whole world:

Discursive Strategies in the Media

(22)

283

America is guilty of many sins: excessive consumption, aggravating the economic problems of the rest of the world, arrogant aggression of their policy, and direct aggressions against Yugoslavia, Iraq. Guilty of claiming to itself […] the role of the world’s policeman, which it disastrously failed, aggravating international instability and chaos. (RG 13/10/09) [Ⱥɦɟɪɢɤɚ ɜɢɧɨɜɚɬɚ ɜɨ ɦɧɨɝɢɯ ɝɪɟɯɚɯ – ɢ ɜ ɱɪɟɡɦɟɪɧɨɦ ɩɨɬɪɟɛɥɟɧɢɢ, ɭɫɭɝɭɛɢɜɲɟɦ ɷɤɨɧɨɦɢɱɟɫɤɢɟ ɩɪɨɛɥɟɦɵ ɨɫɬɚɥɶɧɨɝɨ ɦɢɪɚ, ɢ ɜ ɜɵɫɨɤɨɦɟɪɧɨɣ ɚɝɪɟɫɫɢɜɧɨɫɬɢ ɩɨɥɢɬɢɤɢ, ɢ ɜ ɩɪɹɦɵɯ, ɢ ɜ ɩɪɹɦɵɯ ɚɝɪɟɫɫɢɹɯ ɩɪɨɬɢɜ ɘɝɨɫɥɚɜɢɢ, ɂɪɚɤɚ. ɂ ɜ ɬɨɦ, ɱɬɨ ɜɡɹɜ ɧɚ ɫɟɛɹ […] ɪɨɥɶ ɦɢɪɨɜɨɝɨ ɠɚɧɞɚɪɦɚ, ɨɧɢ ɫ ɝɪɨɯɨɬɨɦ ɩɪɨɜɚɥɢɥɢɫɶ, ɭɫɭɝɭɛɢɜ ɦɟɠɞɭɧɚɪɨɞɧɭɸ ɧɟɫɬɚɛɢɥɶɧɨɫɬɶ ɢ ɯɚɨɫ.]

The structure “is guilty of” establishes a direct connection between the social actor and its actions and characteristics, and indicates the strategy of accusation. On the whole, the strategy of accusation used in the Russian media to represent the USA overlaps with a similar strategy in the Polish newspaper.

5. Conclusions By exploring three main types of strategies (referential, predicational and proximisation) of CDA in the analysis of our data, we demonstrated how the representations of three social actors—Poland, Russia and the USA—vary with respect to changing contextual factors. The dynamics of various strategies for the construction of the social actors within the context of the ABM issue are apparent. This testifies to the changes in attitudes of the social actors to each other and to the ABM issue at different stages of the negotiation process. The type of representation of a particular social actor is predetermined by its strategic and political purposes at a particular moment. Presenting Poland as an insufficient and dependent social actor by the Polish media seems to legitimise the necessity of the protective shield. Conversely, presenting the shield as a threat and the USA as an enemy in the Russian press legitimises the necessity to take some protective measures. As one can observe in Table 2, which summarises the results of the analysis, more dramatic transformations in representations of the social actors occur in the Polish media. In Stage III, self-presentation becomes more self-promoting, which is made manifest through the strategy of positive self-presentation. This is missing in Stages I and II. Pointing to

284

Agnieszka SowiĔska and Tatiana Dubrovskaya

Poland’s ability to play a key role in European security relocates the focus of the debate from the ABM system to a broader issue of the European architecture. The representation of Russia in the Polish media becomes more neutral and less negatively-loaded, although the same negative characteristics are retained. The Russian representation of Poland through different stages is rather stable. Poland is characterised, through the strategy of inferiorisation, as an unimportant and insufficient actor. The self-portrayal of Russia does not undergo any considerable transformations: Russia is consistently presented as a powerful and influential actor. For both the Russian and Polish media ambivalent representations of the USA are typical, especially in Stages II and III. The negative presentation of the USA in the Polish paper significantly overlaps with that in the Russian press, but there are key differences. In the Polish paper, the attitude to the USA changes from positive to negative, which is indicated by the emergence of the accusation strategy in Stage III. In the Russian press, discursive transformations indicate rapprochement, although negative proximisation and accusation still serve to signal Russia’s distrust towards the USA. It is noteworthy that the strategy of emotionalisation appears in most cases to construct Poland and Russia – not the USA. We believe that this fact can be accounted for in two different ways. Firstly, the employment of emotionalisation is predetermined by the type of discourse under analysis. The media language is supposed to construct and influence readers’ opinions, for which the strategy of emotionalisation seems to be one of the most effective ways as it induces an emotional response and side-taking on the readers’ part. Since the texts under analysis are addressed to readers in Russia and Poland, references to Russia’s and Poland’s emotions prevail. Secondly, the lack of references to emotions with regard to the USA may contribute to the overall construction of this actor as more politically mature and, hence, less dominated by emotions than the other two. The present study opens new perspectives in Polish-Russian comparative analysis of media discourse, which, in the long run, can give a key to understanding the specifics of international relations between Poland and Russia and social processes within the two countries.

Discursive Strategies in the Media

Country presented S T A G E I S T A G E

POLAND RUSSIA

USA POLAND RUSSIA USA

Strategies in the Polish paper - deprecation - emotionalisation - emotionalisation - negative proximisation - positive proximisation - negative proximisation no specific strategy, neutral presentation - emotionalisation - emotionalisation

II S T A G E

POLAND

III

USA

RUSSIA

- emotionalisation - deprecation - positive selfpresentation no specific strategy, neutral presentation - accusation and neutral presentation

285

Strategies in the Russian papers - accusation - inferiorisation - negative foretelling - emotionalisation - positive selfpresentation - negative proximisation no specific strategy, neutral presentation - emotionalisation - negative proximisation - positive proximisation - emotionalisation - inferiorisation - positive selfpresentation - positive proximisation - negative proximisation -accusation

Table 2. Discursive strategies used in the construction of Poland, Russia and the USA in all the investigated stages.

References Cap, Piotr. 2006. Legitimization in Political Discourse: A CrossDisciplinary Perspective on the Modern US War Rhetoric. Newcastle: Cambridge Scholars Press. Chilton, Paul A. 2004. Analysing Political Discourse. Theory and Practice. London: Routledge.

286

Agnieszka SowiĔska and Tatiana Dubrovskaya

Habermas, Jürgen. [1981] 1984. Theorie des kommunikativen Handelns. Vol. 1: Handlungsrationalität und gesellschaftliche Rationalisierung. Frankfurt am Main: Suhrkamp [The Theory of Communicative Action: Volume I. Reason and the Rationalization of Society]. Translated from German by Thomas McCarthy. Boston: Beacon. Hart, Christopher. 2010. Critical Discourse Analysis and Cognitive Science: New Perspectives on Immigration Discourse. Basingstoke: Palgrave Macmillan. Hynek, Nik and Vit Stritecky. 2010. “The rise and fall of the Third Site of Ballistic Missile Defense.” Communist and Post-Communist Studies 43: 179–187. Kunze, Thomas and Henri Bonet. 2008. Rosyjskie reakcje na planowane w Europie wschodniej bazy wojskowe [Russian reactions towards planned military bases in Eastern Europe]. In Stephan Raabe (ed.), AmerykaĔska tarcza antyrakietowa w Europie [American anti-missile shield in Europe], 57–73. Warsaw: Konrad Adenauer Stiftung. Ministry of Foreign Affairs of the Republic of Poland. 2008. “Signing of the agreement regarding the placement in Poland of anti-ballistic defensive missile interceptors as well as the adoption of a Polish-US declaration of strategic cooperation.” http://www.msz.gov.pl/Signing,of,the,agreement,regarding,the,placem ent,in,Poland,of,antiballistic,defensive,missile,interceptors,as,well,as,the,adoption,of,a,Poli sh-US,declaration,of,strategic,cooperation,20784.html. DOA: 30 July 2011. Raabe, Stephan (ed.). 2008. AmerykaĔska tarcza antyrakietowa w Europie: koniecznoĞü, warunki, akceptacja [American anti-missile shield in Europe: necessity, conditions and acceptance]. Warsaw: Konrad Adenauer Stiftung. Reisigl, Martin and Ruth Wodak. 2001. Discourse and Discrimination. Rhetorics of Racism and Anti-Semitism. London: Routledge. Reisigl, Martin and Ruth Wodak. 2009. The discourse-historical approach. In Ruth Wodak and Michael Meyer (eds.), Methods of Critical Discourse Analysis, 87–121. London: Sage Publications. SowiĔska, Agnieszka. 2009. “A European identity on the periphery. A comparative study of the representations of Europe in the awkward squad’s press.” CADAAD Journal, 3.1: 21–35. Turaev, Abdufarrokh. 2009. “U.S. missile shield in Europe: framing in the New York Times and Izvestiya.” Master’s Theses. Paper 3708. http://scholarworks.sjsu.edu/etd_theses/3708. DOA: 30 July 2011.

Discursive Strategies in the Media

287

Van Dijk, Teun A. 1998. Ideology: A Multidisciplinary Approach. London: Sage Publications. Wągrowska, Maria. 2008. “Tarcza antyrakietowa z polskiej perspektywy” [Anti-missile shield from a Polish perspective]. In Stephan Raabe (ed.), AmerykaĔska tarcza antyrakietowa w Europie [American anti-missile shield in Europe], 39–47. Warsaw: Konrad Adenauer Stiftung. Wodak, Ruth and Michael Meyer (eds.) 2009. Methods of Critical Discourse Analysis. Second Revised Edition. London: Sage Publications. —. 2009. Critical Discourse Analysis: History, agenda, theory and methodology. In Ruth Wodak and Michael Meyer (eds.), Methods of Critical Discourse Analysis, 1–33. London: Sage Publications. Zolotarjov, Pavel. 2008. Ɂɨɥɨɬɚɪɺɜ ɉ.ɋ. “ɉɪɨɬɢɜɨɪɚɤɟɬɧɚɹ ɨɛɨɪɨɧɚ: ɢɫɬɨɪɢɹ ɢ ɩɟɪɫɩɟɤɬɢɜɵ” [Anti-rocket defence: history and perspectives]. Ɋɨɫɫɢɹ ɜ ɝɥɨɛɚɥɶɧɨɣ ɩɨɥɢɬɢɤɟ [Russia in the global politics]. ʋ3. http://www.globalaffairs.ru/number/n_10948. DOA: 30 July 2011.

PRAGMATIC ASPECTS OF THE SELF AS A COMMUNICATOR JOLANTA SZYMAēSKA

1. Introduction The idea of the Self has been taken from the domain of grammatical person (pronominal grammatical categories) and transferred into the domain of its pragmatic realizations in the written academic text. Authorial persona, encoded in the lexical structure of the text may be reconstructed from lexical items among which, in this paper, the most significant are grammatical personal pronouns, which reveal the social functions of coding interpersonal relationships. In this sense, the Self may be discussed in terms of a semiotic subjective entity reproducing social communicative practices realized in the language of the written academic text. This paper refers to the discussion of the results of the pragmatic interpretative analysis of twenty volumes on linguistics published by Cambridge Textbooks in Linguistics in the last two decades of the twentieth century. More specifically, it focuses on the pragmatic reconstruction of the social communicative roles encoded in the pronominal address and its distribution in written academic discourse. As we would wish to propose, the pronominal address there, rather than realizing traditional writing convention or reflecting scholarly rhetoric, reveals functions of modeling social distance, becoming a source of knowledge about academic communicative practices and the character of cognitive structuring and restructuring of communicating knowledge. Pragmatics as a discipline concentrating on the meaning of the sign in context, depending on the speaker’s meaning and the cooperative principle, puts emphasis on the speaker’s meaning and underlines communication focused on human intentions. Intentional meaning is directed towards being recognized by the receiver according to the speaker’s intentions (van Dijk 1989). The writer (speaker) must, therefore, design his communicative strategies to make contact with the addressee in order to achieve a desired

Pragmatic Aspects of the Self as a Communicator

289

effect. The goal of the writer is to create a common ground between himself and the reader. A good piece of writing closes the gap between them. Concentrating on the language user, the pragmatic approach gives the speaker’s intentions as well as the context of the utterance the greatest importance. The speaker’s intentions and the context of the utterance are necessarily interrelated phenomena.

2. Denotation and reference During an interaction a speaker and a hearer may assign (intend, interpret, infer) different meanings to the same expression and so, the same expression may mean different things in different contexts (van Dijk 1998: 204–206). Hence, meanings of discourse are indeed contextual or situated and depend wholly on the interpretation of the communication participants. Following this, we view the meaning of a word as an actualization of its intentionality (van Dijk 2008). The function of reference makes it different from denotation in that it cannot meet true or false relationship to any denotata independently of any context. Reference, being a fundamental component of any speech act (see e.g., Searle 1969: 23–26) and so attached to a specific time and place, is a property of an individual utterance. As a consequence, reference is context-dependent, while denotation is not. Reference, together with the human factor, that is “the intention of the speaker to pick up some entities from those which could possibly qualify” (Strawson 1950: 327), constitutes meaning as a relation between a linguistic expression and entities in the minds of discourse participants conceptualized in the course of communication. In this sense, denotation remains in the domain of logic, while the context-dependency of reference situates it within the domain of pragmatics. Following Krzeszowski, for a linguistic expression to denote, a chain of knowers is required. If a linguistic expression is to refer, it necessitates a chain of interpreters or conceptualizers (Krzeszowski 1997). An expression will gain several reference meanings, depending on the number of interpreters. Successful communication, then, will reflect a skillful use of contextual features aimed at filling the expression with appropriate reference and will be treated as a human rather than systemic property – negotiation of intersubjective meaning. Accordingly, here, the lexical items discussed do not denote participants but refer to them in terms of their roles assigned for communication. The idea of written text as an interactive social event is now well established in the literature (see e.g., Duszak 1998, Thompson 1994, Hoey 1991, Tadros 1994, Hyland 2002, van Dijk 1989, 2008, and others). In line

290

Jolanta SzymaĔska

with this approach, written texts may be seen as a stage-managed form of dialogue between the writer and the reader, in which writers enact readers by assigning particular speech roles to the readers and to themselves (Thompson 2001). The text is featured by an underlying interaction between the reader and the writer, which is revealed in regular patterning of lexical and grammatical material.

3. Grammatical and pragmatic person According to the typology of pronominal forms prevailing in the theory of English grammar (see e.g., Quirk et al. 1983), the first person singular pronoun “we” obligatorily denotes the originator(s) of messages and optionally the receiver(s) or addressee(s) and/or others, that is, the referents excluded from the definition of the originator or the addressee. As we briefly demonstrate here, the pronominal address “we” reveals particular pragmatic functions and does not necessarily refer to the writer. Pronominal address has been discussed, though not extensively within the domain of academic writing, in the literature on the subject (see e.g., Cygan 1979, Lalewicz 1983, Brown and Gillman 1975, Lubecka 1993, De Vito 1976, Dickey 1997), where the grammatical person is transferred into the domain of pragmatic interpretation. The terms of address constitute a linguistic domain in which the lexical items are assigned functions and meaning depending on the situation. Address forms are most essential in creating relationships between speakers and shaping the course of interaction as regards the goals of the interlocutors. Their distribution in texts reveals social relationships, though in writing they are restricted by convention more than in speech. Lalewicz points out that the categories of person naturally reflect the needs of oral communication as the paradigms of grammar developed at the time when language functioned in a spoken form only (1983: 273). The written communication lacks the aspect of reversing the speaker and hearer turns and so only in certain cases can the spoken categories of person be employed in writing with a similar deictic effect. In printed academic text there is no proper equivalent to express the relation between the writer and the reader(s). Such texts would resemble public speech, and at the same time, the receiver is an individual. The first analogy suggests the use of you collective, the second, you individual. None of them is fully justified but when they are used, they introduce a set of additional implications and are rhetorically marked.

Pragmatic Aspects of the Self as a Communicator

291

4. Persuasion in writing – academic texts The main parts of academic publications, the chapters proper, are marked by aiming at objectivity, less personalization and more persuasion. The communicating person, the writer, presents the facts to speak for themselves. There are fewer sentences with pronouns as subjects, and passive structures are generally favoured. For the sake of accuracy of stylistic expression, the communication in these parts may at times become difficult to process where too many passives are used by the authors, e.g.: for a typological study, the two most basic assumptions (or, perhaps, observations) are, first, that the concept of predicate structure is applicable to all languages, and, secondly, that the arguments both (i) differ in their semantic relationships to the predicator and (ii) are clearly distinguished from one another through grammatical marking. GRRP: 2 most of the issues that are to be considered in this book have been mentioned, often very briefly, in this chapter, although the discussion will, inevitably, introduce others that are related to those under discussion, notably promotion to object (6.6) and incorporation (7.5). GRRP: 21

In contrast to the introductory parts of written publications, the communication in chapters proper assumes two participants, the writer and the reader. This involves persuasive transmission of knowledge, whose effectiveness may be achieved by reducing social distance between the writer and the reader, which, in turn, is encoded in the lexical patterning of texts regarding pronominal address. The analysis of the distribution of personal pronouns in this communication allows for the reconstruction of the roles of the sender and receiver and the social distance between them. The personal pronouns “we” and “I,” as regards their distribution and reference meaning reconstructed in the course of contextual interpretation reveal the function of signaling prominence of particular roles enacted by the writer of the academic text.

5. The personal pronoun “we” After Lalewicz (1983: 270), the addressee of a printed text is comparable to the audience listening to a speech being delivered. At the same time, the reader or readers are reminiscent of an individual. The reference meanings of “we” may be multiple, as both the author and the reader play different roles.

292

Jolanta SzymaĔska

Most often, where the pronoun “we” prevails, it is the distribution and interpretation of other pronouns that will reveal the reference of “we”: If some group of segments persistently figures in phonological process (either synchronic or historical), especially in unrelated languages, an adequate feature system should specify it in a natural way. I will illustrate this point with one particular class; and this will help us focus on the problem of adequacy […]. Consider the class consisting of labials and velars. PL: 97.

Here, “us” does not refer to the author, as he points himself with “I.” “Us” refers to the reader(s) and may or may not include the author. If it does not, it could be substituted by “you.” Here, such possible reference is called “we-you” meaning “we” referring to “you,” the reader(s). We will see shortly that this is oversimple. We observe further that certain morphemes can occur in various context […]. We must now refine our use of ‘morpheme’. (We will include morphemic representations in braces here to indicate their status; for other uses of braces see below). PL: 12 We are still some way from understanding how gender systems arise. What follows is not the whole story, nor the only story. But we now have sufficient evidence for the different stages in the rise of gender systems to be able to offer a plausible account. GGC: 310

The two samples above compared, it appears that “we” pronouns used in the former, (PL: 12), in different sentences in the same text, cannot be referring to the same person. The first “we” does not include the author as he has obviously seen how simple “this” is. The second “we” in the sentence with the instantaneous verb form refers to the author himself and is a regular “we” editorial. The third “we” refers to both the author and the reader(s) as a team. The last “we” refers again to the author. Here, as in most similar texts, the changing of the roles of the sender and receiver may be described in terms of continuity or polarization rather than inclusiveness or exclusiveness. Hence, the “we” pronoun may refer to both the author and the reader on a continuum, depending on the context. The roles then are not reversed but mitigated or negotiated. In the latter, (GGC: 310), the role of the communicator is that of a scientist, a member of a group, which suggests polarization towards the sender and does not include the reader(s). Among many language signals in the written texts, it is often the use of tensed verb forms that determines the interpretation of particular pronominal items. Let the following text samples illustrate the point.

Pragmatic Aspects of the Self as a Communicator

293

We saw above that Spanish is a language with word-stress which is more or less fixed. But we also saw that there were exceptions even in Spanish, and such exceptions can lead to distinctive pairs […]. In the next section I give a brief introduction to the predictability of English word-stress. LAC: 15

Here, if “I” refers to the author, then the first “we” refers to the reader(s) only. It is not only the verb “see” that would suggest such reference, though, quite logically, the fact that Spanish is a certain type of language must have been known by the author before he even started to write the book. Rather, it is the tense form of the verb “to be” in the two first sentences that matters here and consequently, the common or uncommon time reference. The role of the author in “I,” as contrasted to that of “we” has changed from the instructor to the omniscient narrator. “I” referring to the author is often used together with the verb forms referring to the past time, placing the communication participants in the same time perspective and reducing the distance: Here we have a falling tone.[…] However, I mentioned earlier that on some occasions it seems sensible and productive of a simpler analysis if we take syntactic or semantic factors into account. LAC: 36

This sample comes from the same author. The referent of “I” is the author himself. Generally, it may be said that the writers, reluctant to use other addresses than the conventional “we,” are more inclined to avoid “you” than “I.” Another illustrative example where the prominence of particular reference is moderated by the author: We have seen therefore that even within a lexical class it may be necessary to label certain members as exceptionally agreeing in gender or failing to agree in gender. In concluding this section we should bear in mind that the possibility of marking agreement in gender may depend on any of the grammatical categories we have discussed. When in the next chapter we come to the analysis of gender in particular languages, we shall naturally look to those agreeing elements and combinations of categories which give the greatest range of possibilities. But, as we shall see in section 5.5, some of the surprising facts here receive a partial explanation in diachronic terms. Before looking at gender from this perspective, however, we should consider classifiers. GGC: 135

The first “we” refers to both the writer and the reader, the tense form suggesting the same time reference for both. The second “we” refers to the reader(s) only, as it is obviously the readers who should bear in mind some

294

Jolanta SzymaĔska

necessary facts. In the sentence beginning with “but,” however, “we” refers to the reader(s) again. The tensed form of the verb “receive” suggests a present time perspective for the writer, as contrasted to the future for the reader. We saw how in Archi insects belong in gender III. This criterion does not fit in any obvious way with the other criteria by which nouns are assigned to gender III; we might rather have expected insects to be assigned to gender IV, along with the smaller wild animals and birds. Given that apparently unrelated criteria may assign nouns to the same gender, it is tempting, when nouns do not fit into a semantic assignment system, to look for more and more semantic criteria which could account for them. In some languages, however, it soon becomes evident that this approach fails, and that there is instead a formal rather than semantic criterion which will account for some or all of the nouns in the semantic residue. […] We have seen how in some languages the meaning of noun always or virtually always determines its gender. […] In both cases it is important to bear in mind that the world view of the speakers determines the categories involved. GGC: 32

If we compare (GGC: 135) to (GGC: 32), we notice that the referents of the first “we,” as contrasted to the first sentence in the former (GGC: 135) and the first sentence of the second paragraph in the latter, (GGC: 32) is “you, the reader(s),” which is implied by the verb tense. Should both verb forms in the first sentence in (GGC: 32) be in the same tense – the past – the time perspective for both participants would be the same, just as is the case of the second paragraph in (GGC: 32). Also, the first “we” in the second paragraph here refers to both the reader(s) and the author, while the second, only to the reader(s). In yet other patterns, “we” refers to the author himself as the scientist and the distance thus signaled becomes considerable. ‘Culture’ is what ‘others’ have, what makes them and keeps them different, separate from us. […] a criticism of such uses is valuable, among other things, in making us aware of the role of academic discourse in the production and legitimation of marginalization; a role that academic personnel engage in often without an awareness of it. […] At the same time, new generations of students of human social conduct need to have an historical understanding of our root metaphors and concepts, if they want to attempt new theoretical elaborations and syntheses. Whatever problems […], they are small compared with the danger of avoiding defining the concepts that can help us understand similarities and differences in the ways in which people around the world constitute themselves in aggregates of various sorts. LAD: 23

Pragmatic Aspects of the Self as a Communicator

295

Here, “we” refers to the author as a scientist possibly including the other scientists, reader(s) excluded. In the passage, the author enacts two roles, the teacher and the scientist. The problem is that if we assume (as we do) that phonemic representations are only specified in terms of non-redundant features, then only one version of these two statements can be made; […] the problem that we are here encountering is really this. Given that we require phonemic representations to express contrast as economically as possible, the notion of economy itself is ambiguous and gives rise to two alternative interpretations. Do we mean by this that each phoneme is to be represented by the smallest possible number of features? Or do we mean that we want to employ the smallest possible number of features? […] It suffices to make the reader aware of the problem; we neither can nor need to solve it in a principled way. […] When we discuss the regularities of the concrete phonetic level of representations (in chapter 8 below), we shall no longer need to distinguish between redundant and non-redundant features simply all features will be specified on that level; in particular, the redundant ones will by then have been supplied by redundancy rules. We do not discuss the complete set of redundancy rules but simply assume that by the time the phonetic level of the derivation is reached, they have all operated. EPG: 126

In the above, (EPG: 126), “the reader” and “we” are used instead of direct “you” and “I.” The use of the Present Progressive Tense in the first part of the passage determines the time reference – as here and now at the moment of reading. The “we” in the first sentence may refer to both the writer – instructor and the reader, the exception here being the “we” in parentheses, which refers to the writer as a scientist. In the second part all the instances of “we” refer to the author himself in the role of the omniscient narrator and the scientist.

6. “We” in patterns discussed in this paper There is interdependence of the lexical context and reference meanings of “we” in complex sentences, which may be summarized as follows: We + past tense verb + present tense verb = you, the readers (students) e.g. We saw how in Archi insects belong in gender III. We + past tense verb + past tense verb = you, the readers (students) and I, the author (instructor) e.g. But we also saw that there were exceptions even in Spanish

296

Jolanta SzymaĔska

As far as the distribution of “we” in other contexts referring to the writer himself, the scientist, what has so far been perceived as “we editorial” denoting the author only, might be said to refer to different roles: it may refer to both the author – instructor and the reader – student involved in the process of sharing knowledge. Also, in other contexts, it may be said to refer to the writer – linguist or the writer himself as the authorial persona. This analysis demonstrates how pronominal addresses in written texts reveal not only their possible referents but the roles enacted by the sender and receiver.

7. Concluding remarks Readers understand deictic referring as cognizing subjects and makers of meaning. It appears a highly regular process wherein linguistic items referring to the reader and writer are distributed in the written academic text. As a text type, a linguistic publication in teaching is comparable to the scientific article (Hyland 2002), whose persuasiveness will depend on the reader’s satisfaction with the author’s argument. One of the possible ways to approach the issue of predominating social communicative practices there is to look at the address in writing in terms of enhancing the persuasive rhetorical effect. Other possible way is to explain the regular pronominal patterning in texts as regards social cognitive processes of communication participants, that is, regarding communicational behaviors and products of communicating individuals, who transmit and interpret pre-constructed meanings. Meaning, then, as a linguistic theoretical phenomenon, would not reside in verbal means by which individuals communicate and cannot be passed on in the same way as meaning bearers, which trigger only the communicating activities within the cognitive domains of individuals. Communicating in language is based on continuous mental processing and interpreting of meaning bearers. Analyzing texts for meaning bearers can help explain this processing. As far as possible teaching implications are concerned, in the perspective of functional concepts of the self as communicator, “writing requires conventional forms of arrangement and style-including forms of self – representation that often reduce the “I” of their discourse to a pronoun without any definite referent” (Buley-Meissner 1990: 5). Yet, academic writing seems not a mere attempt to mirror the objective reality, but reflects the author’s efforts to produce convincing accounts of ideas and events, of which there are many possible versions. Through replicating the communicative writing practices typical of particular

Pragmatic Aspects of the Self as a Communicator

297

categories of discourse, the students benefit from the process of creating the authorial persona or communicating voice by the language of the text.

References Brown, Roger and Albert Gillman. 1975. The pronouns of power and solidarity. In Thomas Sebeok (ed.), Style in Language, 253–276. Cambridge, MA: Massachusetts Institute of Technology. Buley-Meissner, Mary L. 1990. “Rhetorics of the Self.” Journal of Education 172. 1: 47–65. Cygan, Jan. 1979. Strukturalne podstawy gramatyki angielskiej [Structural basics of English grammar]. Warszawa: Wydawnictwa Szkolne i Pedagogiczne. De Vito, Joseph. 1976. The Interpersonal Communication Book. New York – Hagerstown – San Francisco – London: Harper & Row Publishers. Dickey, Eleanor. 1997. “Forms of address and terms of reference.” Journal of Linguistics 33.2: 227–685. Duszak, Anna. 1998. Tekst, dyskurs, komunikacja miĊdzykulturowa [Text, discourse and intercultural communication]. Warszawa: Wydawnictwo Naukowe PWN. Hoey, Michael. 1991. Patterns of Lexis in Text. Oxford: Oxford University Press. Hyland, Ken. 2002. “Options of Identity in Academic Writing.” ELT Journal 56.4: 351–358. Krzeszowski, Tomasz P. 1997. Angels and Devils in Hell. Elements of Axiology in Semantics. Warszawa: Wydawnictwo Energeia. Lalewicz, Janusz. 1983. Retoryka kategorii osobowych [Rhetoric of personal categories]. In Teresa DobrzyĔska and ElĪbieta Janus (eds.), Tekst i zdanie. Zbiór studiów [Text and sentence. Collection of studies], 267–280. Wrocáaw: Polska Akademia Nauk. Lubecka, Anna. 1993. Forms of Address in English, French and Polish. A Sociolinguistic Approach. Kraków: Uniwersytet JagielloĔski. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A Comprehensive Grammar of the English language. London: Longman. Searle, John. 1969. Speech Acts. Cambridge: Cambridge University Press. Strawson, Peter. 1950. “On Referring.” Mind. New Series 59.235: 320– 344. van Dijk, Teun A. 1989. “New developments in discourse analysis 1978– 1988.” Journal of Interdisciplinary Studies 1: 119–145.

298

Jolanta SzymaĔska

—. 2008. Discourse as Structure and Process. London: Sage. Tadros, Angele. 1994. Predictive Categories in Expository Text. In Malcolm Coulthard (ed.), Advances in Written Text Analysis, 69–82. London: Routledge. Thompson, Geoff. 2001. “Interaction in academic writing: Learning to argue with the reader.” Applied Linguistics 22.1: 58–78.

Text sources GRRP Palmer, Frank Robert. 1994. Grammatical Roles and Relations. Cambridge: Cambridge University Press. PL Lass, Roger. 1991. Phonology. Cambridge: Cambridge University Press. LAC Cruttenden, Alan. 1997. Intonation. Cambridge: Cambridge University Press. GGC Corbet, Greville. 1991. Gender. Cambridge: Cambridge University Press. LAD Duranti, Alessandro. 1997. Linguistic Anthropology. Cambridge: Cambridge University Press. EPG Giegerich, Heinz J. 1992. English Phonology. Cambridge: Cambridge University Press.

DISCOURSE GOALS AND NORMS: AN AXIOLOGICAL PERSPECTIVE PRZEMYSàAW ĩYWICZYēSKI

1. Introduction Research to be presented here is part of a larger project, which consists in applying an axiological format of analysis to the study of conversation data. The ultimate goal of this project is to elaborate an axiological model of spoken interaction,1 whose preliminary sketch was given in The Axiology of Spoken Interaction (ĩywiczyĔski 2010) – the work I will refer to throughout this paper. A necessary step in building such a model, which was taken in The Axiology of Spoken Interaction, was to determine the role which interactional, value-related events – most importantly, conflicts and conflict-resolutions – play in conversational discourse. This enquiry led to addressing a number of general questions about the nature of discursive interaction, one of the most important of which was concerned with discourse goals and the role they play in the organisation of discourse. Next, the consideration of teleological aspects of spoken interaction uncovered a domain of normative phenomena, which – as will be shown – is crucial to understanding conversational behaviour.2 In this paper, I will attempt to give an axiological account of goals pursued by conversants and norms regulating their discursive conduct. Subsequently, I will set out to specify the relation between these two planes of conversational interaction. The starting point for my project is the axiological conception of discourse proposed by Tomasz P. Krzeszowski in Angels and Devils in Hell (1997). I believe that the framework for the analysis of discourse he presents touches upon the fundamental processes governing discursive 1

Following Hausenblas, Widdowson, Krzeszowski, and Brown and Yule, “spoken interaction” – alternatively designated as “spoken discourse,” “discursive interaction,” or “conversational interaction” – is used with reference to the process of creating conversation, where “conversation” is understood as the end product of this process. For more on this problem see ĩywiczyĔski (2010). 2 Behaviour as understood in ethology – see e.g. Eibl-Eibesfledt (1989).

300

Przemysáaw ĩywiczyĔski

interaction – processes which in his view, and mine, have a valuative character. I also agree with Krzeszowski that the model put forward in Angels and Devils has a “highly tentative status,” mainly due to a limited treatment which discourse phenomena received in his book, where out of twelve chapters only one was devoted to the axiology of discourse (i.e. Chapter XI “Discourse Phenomena”). It seems that the only theoretically sound way of developing Krzeszowski’s idea is to apply his format of axiological analysis to the study of extensive language material. Accordingly, this work will present or report analyses of spoken discourses, mainly taken from conversational analysis (CA) corpora of contemporary English,3 carried out along the lines sketched by Krzeszowski.

2. Introductory notes on the axiology of discourse phenomena Krzeszowski uses two relatively simple conversations to introduce the basic elements of discourse axiology: (1)

Turn 1 A: I hope I haven’t interrupted. Turn 2 B: Of course not. Have a drink? Turn 3 A: Thanks. Beer? Turn 4 B: Oh yes. Certainly. Here you are. Turn 5 A: Thanks a lot.

(2)

Turn 1 A: I hope I haven’t interrupted... Turn 2 B: Of course not. Have a drink? Turn 3 A: Thanks. Beer? Turn 4 B: Sorry. We haven’t a fridge – we send out for ice. What about a Scotch? Turn 5 A: A small one, if you don’t mind. I’m not very keen on hard liquor. Turn 6 B: On the rocks? Turn 7 A: Plenty of soda – if you aren’t short. Turn 8 B: Here you are. Turn 9 A: Thanks a lot. (Krzeszowski 1997)

3

These include the corpora gathered by Atkinson and Drew (1979), Davidson (1984), Pomerantz (1984), Sacks (1992), Sacks, Schegloff, and Jefferson (1974), Sacks and Schegloff (1973), Sacks, Schegloff, and Jefferson (1978), Stenström (1994), and Wootton (1981).

Discourse Goals and Norms: An Axiological Perspective

301

He explains that the conversations represent two axiologically distinct formats. The conversation (1) is axiologically simple4 in the sense that the interaction between A and B develops in an unperturbed way towards the achievement of the discourse goal, that is, A’s procuring the drink of beer from B. It stands in contrast to the “axiologically complex” conversation (2), where an obstacle arises to the achievement of the discourse goal, that is, A’s procuring the drink of beer from B. The resultant axiological clash triggers off the reparatory sequence, in effect of which B offers and A accepts the surrogate drink (of Scotch). (Krzeszowski 1997: 249)5 To describe axiologically complex discourses such as (2), Krzeszowski elaborates a specialised terminology. Accordingly, the reparatory sequence instigated by the appearance of an axiological clash is termed a Metaaxiological Interlude (MAIN) – it involves a negotiation of values with a view to resolving the clash (e.g. offering an alternative drink in (2)). A MAIN culminates in either a Clash Resolving Segment (CREST): a discourse point where the clash is effectively eliminated, or a Clash Holding Segment (CHOST): a discourse point which marks a failure to eliminate the clash. Thus, in sample dialogue (2), the MAIN terminates in a CREST – A’s acceptance of Scotch. The configuration of an axiological clash with the ensuing MAIN is termed an Axiological Cycle (AC) (Krzeszowski 1997: 248–251). The overall axiological structure of conversation can be graphically represented as follows:

4

It should be acknowledged that the terms “axiologically simple” and “axiologically complex” (to be introduced) are my own terminological inventions; however, they don’t seem to distort Krzeszowski's presentation and capture the essential – axiological – difference between discourses which contain axiological clashes and those that do not. 5 Krzeszowski’s exposition of discourse axiology is based on the DISCOURSE IS A JOURNEY metaphor. This metaphor, identified as the fundamental conceptualisation of the discursive process, is structured as a cluster of preconceptual image schemata, including SOURCE-PATH-GOAL, FORCE, PART-WHOLE, BALANCE, UP-DOWN, LEFT-RIGHT, STRAIGHT-CROOKED, CONTAINER, CENTER-PERIPHERY, RHYTHYM, AND CYLE (1997: 245).

302

Przemysáaw ĩywiczyĔski

Fig. 1. The axiological model of discourse adapted from Krzeszowski (1997: 257)

According to the research proposal sketched in the introduction, I will seek to develop Krzeszowski’s model by applying it to the study of actual conversational material. For this happen, first a few introductory remarks about models of conversation analysis must be given.

3. Models of conversation analysis As regards the format of describing conversational behaviour, two theoretical positions will be consulted: the model of spoken interaction and the classical model of conversation analysis.6 The choice is not arbitrary as these two positions represent the principal approaches to the study of conversation – the former is derived from “discourse studies” as the notion of “discourse” is commonly understood; the latter, from the ethnomethodological tradition. The model of spoken interaction is based on John Sinclair and Malcolm Coulthard’s conception of discourse analysis (Sinclair and Coulthard 1975; Coulthard 1992), later developed by Anna-Brita Stenström (1994) and, independently, by Suzanne Eggins and Diane Slade (1997). 6

For more on the approaches relevant to the study of conversation see Eggins and Slade (1997: 23–66).

Discourse Goals and Norms: An Axiological Perspective

303

The major objective of the “spoken interaction model” is to describe structural properties of conversation; it is a reductionist attempt to present the hierarchy of conversation components: from small organisational units – turns, moves, acts – to more and more complex ones – exchanges, transactions, and conversations (Stenström 1994: 30–67). The key element of the conversational structure is identified as “transaction:” a component delimited by a single topic, which constitutes the primary unit for transferring information (Stenström: 1994: 55–58). Following Stenström, I will henceforth refer to this approach as the Transactional Model of Conversation (TMC). The classical model of conversation analysis was proposed by Harvey Sacks, together with his collaborators – Emmanuel Schegloff and Gail Jefferson (1974, 1978); later, it came to be practised by a number of sociologically and anthropologically oriented scholars (e.g. Wootton 1981, Pomerantz 1984, Davidson 1984, Coates 1986, Drew and Heritage 1992, Tannen 1989, 1994). The researchers from the tradition of Conversational Analysis appeal to the idea of “ethnicity” adapted from Harold Garfinkel (1967), whereby the analyst’s task consists in uncovering participants’ own (i.e. “ethnic”) rules for producing and comprehending conversational texts (Garfinkel 1967, 1972; Levinson 1983: 295). Sacks and collaborators may be interested in the structure of conversation; however, in contrast to TMC, the emphasis is here on an interactional aspect, that is, on how participants use conversational mechanisms (e.g. turn-taking system, preference organisation, repair system, etc.) to accomplish social goals. If TCM is discourse-centred, in the sense that conversation is treated as a closed text thereby allowing the researcher to dissect it into components, conversation analysis is participant-centred, its focus being on the ongoing management of conversational interactions by conversation participants. To underline a contrast between TMC and the approach advocated by conversation analysts, I will designate the latter position as the Interactional Model of Conversation (IMC).

4. Discourse axiology and TMC It should be observed that a TMC analysis does not reveal major structural differences between the axiologically simple conversation (1) and the axiologically complex conversation (2). Specifically, with regard to the organisation of exchanges (an exchange being a minimal interactive unit which involves the negotiation of a single piece of information

304

Przemysáaw ĩywiczyĔski

through a sequence of acts7), both of them exhibit the unmarked – chaining – exchange format characterised by an uninterrupted flow of exchange initiations and responses with speakers alternating the roles of the exchange initiator and the exchange respondent. The only major difference is that if the chaining sequence in (1) is consistently of the coupling type (i.e. such in which A’s initiating move is followed by B’s turn containing both a response and an initiation of a new exchange), the coupling sequence is in (2) suspended between Turns 5 and 6: Turn 5 A: A small one, if you don’t mind. I’m not very keen on hard liquor. Turn 6 B: On the rocks?

In Turn 5, A accepts B’s offer (to take Scotch) but fails to initiate a subsequent exchange – this is done by B, who goes on to put forward another offer (that is, to serve Scotch on the rocks). From this point, the exchanges resume the usual coupling course. It is interesting to see that the suspension of the coupling sequence does not coincide with the axiologically important moments, that is, the appearance of the axiological clash (Turn 4) or its resolution (CREST in Turn 9); it happens during the phase of negotiation (MAIN) and, possibly, is the result of A’s unwillingness to take the initiative after her original request for beer has been turned down. Is it, then, the case that there is no correlation between the exchange format and the axiological structure? So far, I have only been concerned with the chaining type of the exchange model. In an attempt to answer the above question, I will need to refer to conversational samples containing embedded exchanges. As noted before, chaining represents the prototypical, unmarked exchange format of a transaction. Embedding, on the other hand, is marked in the sense that it discontinues a sequence of initiations and responses; the function of a hold-up which thus emerges is to allow for the completion of a prior exchange:

7

An act signals a speaker’s communicative intention, for example , , (Stenström 1994: 38–48); exchanges are conventionally quoted in square brackets while acts are in triangular brackets. An exchange should also be differentiated from a turn, that is, “everything the current speaker says before the next speaker takes over” Stenström 1994: 4). In both the TMC and IMC approaches, a turn constitutes the basic unit of conversation analysis.

Discourse Goals and Norms: An Axiological Perspective

(3)

305

Turn 1 A: That wasn’t the guy I met, was it, when we saw the building? Turn 2 B: Saw it where? Turn 3 A: When I went over to Chetwynd road. Turn 4 B: Yes (fall). (adapted from Stenström 1994: 53)

The completion of the first exchange (Turn 4) is delayed until B has collected the requisite information in the auxiliary question-answer exchange (Turns 2 and 3); the discontinuity in chaining is easy to notice: Turn 1 A: Exchange 1 Turn 2 B: Turn 3 A: Turn 4 B: Exchange 1.

Exchange 2 Exchange 2

Conversation (3) testifies to the existence of an intriguing parallelism between the exchange organisation and the axiological structure. The manifestation of the exchange discontinuity in Turn 2 coincides with the appearance of the axiological clash: B is unable to answer the question owing to her imperfect knowledge of the context, whereby the conversants’ progress towards the discourse goal (identifying “the guy” from Turn 1) is stalled. The enunciation of the clash in the question format automatically triggers off the reparatory MAIN scenario (Turn 2), which in Turn 3 brings about the resolution of the clash. The MAIN phase of the discourse is thus co-extensive with the embedded sequence (Turns 2 and 3). Furthermore, the axiologically important moments (the clash, the MAIN, the CREST) coincide with the transactional configuration based on the embedding: the clash and the onset of the MAIN concur with the beginning of the embedded sequence, while the CREST is contemporaneous with its termination. The match between the two planes is so suggestive that the axiological description seems to repeat the TMC analysis. The analysis of further 25 conversations containing embeddings supported the view that embedded sequences coincide with axiologically complex structures.8 In all these cases, the axiological clashes introduced in the embedded sequences concerned a lack of knowledge on the part of one of the participants; this cognitive deficit resulted in withholding the progression of proper conversational acts, such as accepting, refusing, etc. Limited to fragments of transactions, hold-ups of this sort play but a subsidiary role in the development of conversation, and their origin as

8

For details see ĩywiczyĔski 2010 Chapters II–IV.

306

Przemysáaw ĩywiczyĔski

informational gaps explains why they exhibit the format.

5. Local and global discourse phenomena According to Krzeszowski's nomenclature, embedding-related axiological phenomena should be defined as “local axiological cycles,” that is, such where clashes (local axiological clashes) do not upset the axiological balance of an entire discourse and where MAINs (local MAINs) are limited in scope only to parts of discourse (1997: 252–253). In all of the studied conversational samples with embedded sequences, the axiological clashes related to embeddings did not compromise the accomplishment of the global discourse goals, while the ensuing MAINs were confined to the boundaries of the embeddings (as was the case in (3)). Unlike these, the clash in (2) is a global one: it pertains to the overall goal of the discourse (i.e. getting A a drink), whereby its appearance upsets the balance of the entire interaction.

5.1 Discourse goals and discourse domains I will designate the type of discourse goals pertaining to local – embedding-related – clashes and MAINs as intra-discursive goals. My motivation is to indicate that these are structurally and topically tied to discourse itself: as regards the structure, intra-discursive goals are signalled at the level of discourse organisation by transactional embeddings; as regards the topical range, they concern the informational continuity of discourse. In contrast, discourse goals related to global axiological clashes and global MAINs will be referred to as extradiscursive goals. This is to bring out the fact that they are not focused on discourse as such but, rather, that discourse is here used as a means of accomplishing certain extra-discursive states, such as getting A a drink in (2). The relation between the two types of goals is best seen in terms of enablement, whereby the accomplishment of intra-discursive goals acts as an enabling condition to the successful pursuit of extra-discursive goals.9 The local status of clashes related to intra-discursive goals explains why their solution does not automatically lead to the accomplishment of extra9

This is true at the basic level of analysis; a more thorough examination reveals a considerable complexity of the relation between the two sets of goals is much complex; for details see ĩywiczyĔski 2010: 97–103.

Discourse Goals and Norms: An Axiological Perspective

307

discursive goals. This is exemplified by the interaction (4), where the solution of the local clash pertaining to B’s ignorance of A’s age does not promote A’s extra-discursive goal of procuring the bottle of Mich (Levinson 1983: 304): (4)

Turn 1 A: May I have the bottle of Mich? Turn 2 B: Are you twenty one? Turn 3 A: No Turn 4 B: No



To reiterate the findings made so far, the juxtaposition of the axiological description of conversation with the TMC mode of analysis has rendered a two-fold distinction of axiological clashes: discourseexternal clashes related to participants’ extra-discursive goals and discourse-internal clashes related to participants’ intra-discursive goals. Discourse-external clashes are not reflected in the transactional organisation of conversation, that is, the segments of conversation which contain them have a simple – unmarked – format consisting of coupling exchanges. By way of contrast, discourse-internal clashes are signalled in the transactional organisation of conversation by the appearance of marked structures in the form of embedded exchanges. It was noted that clashes of this first type are global in the sense that they upset the axiological balance of entire conversations, whereas clashes of the latter type are local, that is, their scope is restricted to isolated fragments of conversations, wherein participants upgrade their discourse models. The fact that discourse-external clashes are directly related to the evaluation of participants and their plans intensifies the negative charge of discourses (or segments of discourses) in which they appear. In contrast, the negative charges generated by internal clashes are significantly weaker. In this way, the marked and unmarked transaction formats brought out by the TMC analysis capture the non-trivial distinction between types of axiological clashes – strongly charged global clashes and comparatively weak local clashes. Based on these findings, I propose to distinguish two levels in the axiological structure of discourse: the transactional plane, which describes discourse from the perspective of speakers’ intradiscursive goals and the interactional plane, which is related to speakers’ extra-discursive goals. The terms are derived from the two principal approaches to the study of conversation: the Transactional Model of Conversation analysis and the Interactional Model of Conversation analysis (see Section 3).

308

Przemysáaw ĩywiczyĔski

The above proposal focuses on the role of conversants’ goals for the organisation of discourse in general and its axiological dynamics in particular. The importance of discourse goals is evident in Krzeszowski’s own mode of analysis. He may not often explicitly mention the term “discourse goal,” yet his the line of the argument clearly indicates that axiological phenomena, in particular axiological balance, are in a direct way related to the pursuit of goals by interactants. This is how Krzeszowski describes the axiology of the “beer dialogues” – the axiologically simple (1) and the axiologically complex (2):10 (1) contains no clashes since the interaction between the speakers at no point deviates from the straight path leading to the GOAL, viz. having a drink. Therefore, the axiological balance is maintained throughout the discourse. In (2) there is a clash and it consists in the impossibility of satisfying A’s desire to have beer. […] At this point the axiological balance of the discourse is upset. A finds himself in a negatively valued situation […]. B considers himself partly responsible for the situation in which it may be impossible for A to have a drink. (Krzeszowski 1997: 249)

However, on closer inspection, an axiological account of discursive process reveals a different dimension of conversational interaction. To uncover it, I will now present in more detail phenomena arising on the two planes of discourse – the transactional plane, related to the pursuit of intra-discursive goals, and the interactional plane, related to the pursuit of extra-discursive goals.

6. Transactional organisation of discourse As already noted, intra-discursive goals, which pertain to the transactional domain, are concerned with the informational processing of discourse. In particular, their operation seems best explained with reference to the concept of information accessibility and continuity; that is, discourse is informationally continuous if its informational components are accessible to all discourse participants. Using Brown and Yule’s terminology (1983), it could be said that discourse is informationally continuous if there is concordance between participants’ discourse

10

(1) corresponds to (12a) and (2) to (12b) in Krzeszowski’s numerical arrangement (1997: 248).

Discourse Goals and Norms: An Axiological Perspective

309

models.11 Accordingly, an intra-discursive goal consists in having or acquiring a discourse model which is adequate for processing information provided during a conversational interaction. For example, in the following fragment (5), B possesses a discourse model which is adequate for processing all the information supplied by A in the opening turn: (5)

Turn 1 A: Why don’t we all have lunch Turn 2 B: Okay ... (Levinson 1983: 316)

Specifically, B’s discourse model is able to accommodate the potentially ambiguous – exophoric – use of the pronominal expression “we all,” together with the idea of “having lunch.” In contrast, the interaction (6) illustrates a processual problem related to the pursuit of intra-discursive goals: (6)

Turn 1 A: Look John! There’s going to be an interesting discussion at the meeting this afternoon. Will you be there? Turn 2 B: What time? Turn 3 A: Five o’clock Turn 4 B: Ok.

Here in Turn 2, B signals that because her discourse model lacks information about the exact time of the meeting, she is unable to process A’s initial contribution. After upgrading B’s discourse model in Turn 3, the processing problem disappears. Intra-discursive goals are characterised by a local scope of operation. Typically, intra-discursive goals operate on a turn-to-turn basis, that is, an intra-discursive goal is related to processing information supplied in the immediately preceding turn. Thus, B’s response “Okay” in (5) Turn 2 indicates only that she is able to process the information given by A in the preceding turn. It does not follow that her discourse model accommodates all aspects of the situation discussed. For example, B probably does not know where the prospective lunch is going to take place;12 were A to 11

In this context, Brown and Yule’s “discourse models” seem roughly equivalent to de Beaugrande and Dressler’s “discourse representations” (Yule and Brown 1983: 206–207; de Beaugrande and Dressler 1981: 194). 12 Actually, in the original sample B guesses that the lunch is going to take in St Jude’s; A confirms this in the following turn: Turn 1 A: Why don’t we all have lunch Turn 2 B: Okay so that would be in St Jude’s would it? Turn 3 A: Yes.

310

Przemysáaw ĩywiczyĔski

follow in a successive turn with: “Let’s meet at Charles’s,” then B would have to consult her discourse model again, to see whether she knows which Charles is being referred to and, if so, whether she knows where this Charles happens to live. On analysing the 81 interactions which comprised my corpus, it turned out that none of them contained an unresolved local axiological clash. In the prototypical MAIN scenario in transactionally complex exchanges, a local clash is solved in the turn directly following the turn in which it has been introduced. It seems as if the very act of participating in conversational interaction makes it incumbent on conversants to solve (or attempt to solve) informational gaps that block the progress on the path of discourse. Of course, it is possible to envisage a situation in which a discursant called on by her fellow participant to supply information, may refuse to do so; for example, A in (3) may refuse to specify the place where she saw “the man,” thus upholding the clash which arose in Turn 2: (3a)

Turn 1 A: That wasn’t the guy I met, was it, when we saw the building? Turn 2 B: Saw it where? ((LOCAL CLASH)) Turn 3 A: When I went over to Chetwynd Road ĺ Sorry but I won't tell you ((LOCAL CHOST))

Yet, A’s behaviour in (3a) (or comparable courses of action) does not appear as an interactionally viable option – it is likely to be interpreted as intentionally obtrusive and result in terminating discursive interaction. It is then cooperation in facilitating the informational continuity of discourse that constitutes the expected course of action on the transactional plane of discourse, and given the statistical extent of this cooperation, I conclude that interactants’ conduct can only be explained with reference to some normative element. Specifically, the norm – which will be referred as the norm of transactional cooperation (or the transactional norm) – requires that interactants make elements of discourse accessible to all discourse participants and, furthermore, in the event that informational gaps appear in participants’ discourse models, the norm commands that their fellow discursants’ should help bridge these gaps as soon as possible. Interestingly, the type of informational cooperation related to the pursuance of intra-discursive is strongly reminiscent of Paul Grice’s notion of conversational cooperation. In Gricean terms, the norm of conversational cooperation commits speakers to abide by the Cooperative Principle (the CP), the original definition of which reads as follows:

Discourse Goals and Norms: An Axiological Perspective

311

Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged. One might label this the COOPERATIVE PRINCIPLE. (Grice 1975: 41)

It should not distort Grice’s idea if the above is rephrased as: “Make your conversational contribution such that it is informationally continuous and that it prevents the emergence of informational gaps in the listener's discourse model.” There is however a limit to the similarity between Grice’s and my notion of conversational cooperation. For Grice, the formulation of the CP constitutes a point of departure for laying out the conception of implicature and the system of conversational maxims, which – in very general terms – are designed to explain how utterances communicate meanings intended by speakers. In this work, cooperation is understood in informational terms – interactants cooperate, i.e. abide by the transactional norm, insofar as they try to (i) make their contributions informationally accessible to fellow interactants and (ii) bridge informational gaps which might have arisen in the course of interaction. Important as the contrast between the two approaches may be, it should be stressed that they both view discourse as an essentially cooperative process and attempt to explain its nature in normative terms. As will be shown in the next section, cooperation and normativity are also crucial to the understanding of the interactional plane of discursive interaction.

7. Interactional organisation of discourse The interactional plane of discourse has been defined as the domain of extra-discursive goals. Axiological phenomena arising at this level have a global character; in particular, clashes which block the accomplishment of extra-discursive goals disturb the overall balance of an interaction. It was also noted that axiologically global phenomena (i.e. global clashes and global MAINs) are not signalled by embeddings but exhibit the unmarked, chaining organisation of exchanges. The obvious question is then about the way in which global axiological phenomena are reflected in the structure of discourse. As will be shown, they are strongly linked to the interactional organisation of discourse described by the classical model of conversational analysis (or IMC, see Section 3). In order to give a full account of this relation, the two concepts essential to the IMC approach must be presented – the concept of adjacency pairs and that of conversational preference.

312

Przemysáaw ĩywiczyĔski

7.1 Adjacency pairs and preference phenomena In the tradition of conversation analysis, adjacency pairs are often described as the fundamental unit of conversation. The idea that conversation is organised into pairs of functionally related turns, such as question-answer, greeting-greeting, or offer-acceptance, was first suggested by Erving Goffman and developed by Sacks (1992) and Sacks and Schegloff (1973). Banal as it may seem, the concept of conversational adjacency prepared the ground for a much more significant proposal – the preference organisation view. This was put forward in Anthony Wootton’s study of grantings and rejections in parent-child interactions (1981) and elaborated by Anita Pomerantz in her breakthrough work on the management of agreements and disagreements (1984). In basic terms, the discovery of preference organisation was related to the observation that the structure of second pairs in adjacencies correlates with their function. Preferred second pairs, for example, the acceptance of a request or the consent to an assessment, are structurally simple and delivered without hesitation: (7)

Turn 1 A: Why don’t you come up and see me some // times. Turn 2 B: I would like to. (Atkinson and Drew 1979: 58)

On the other hand, dispreferred second pairs (refusal or disagreement) are structurally complex and are delivered with delays, for instance: (8)

Turn1 A: Uh if you’d care to come and visit a little while this morning I’ll give you a cup of coffee Turn 2 B: > hehh ((DELAY)) well ((ANNOUNCER)) that’s awfully sweet of you ((APPRECIATION)), I don’t think I can make it this morning. ((DECLINATION)) I’m running an ad in the paper and-and uh I have to stay near the phone. ((REASON)) (Atkinson and Drew 1979: 58)

Levinson notes that dispreferred turns characteristically contain delays (e.g. pauses), prefaces (e.g. the use of announcers, token agreement, or appreciation markers), accounts and the declination component itself, which however is in many instances omitted.

Discourse Goals and Norms: An Axiological Perspective

313

7.2 Concurrence between preference and axiological phenomena An analysis of the request sequence (7) and the invitation sequence (8) is illustrative of the relation between the conversational and axiological organisation. The preferentially formatted (7) is axiologically simple – it is free from global axiological clashes pertaining to the participants’ pursuit of extra-discursive goals. By way of contrast, the dispreferential invitation sequence (8) not only exhibits axiological complexity, but what’s more the emergence of the axiological clash seems to be signalled by the dispreferential signals in Turn 2, in which A’s extra-discursive goal to have B for coffee is blocked: Turn 2 B: > hehh ((DELAY)) well ((ANNOUNCER)) that’s awfully sweet of you ((APPRECIATION)), I don’t think I can make it this morning. ((DECLINATION)) I’m running an ad in the paper and-and uh I have to stay near the phone. ((REASON))

The corpus material confirmed a strong correlation between dispreferential organisation and axiological complexity. Out of the 34 dispreferentially organised interactions that were studied, all contained global axiological clashes in the dispreferred turns, and furthermore the overwhelming majority of them (i.e. 27) showed accumulation of dispreferential signals (more than 3) in these turns. A detailed discussion of the role that dispreferential signals play in the axiological organisation of conversation requires a separate treatment.13 With regard to the subject of this paper suffice it to say that they serve two basic functions – the function of cushioning the impact of a global axiological clash which is being introduced into a discourse and the function of facilitating a MAIN sequence.14 As to the former function, dispreferential signals can be characterised as exponents of negative politeness (Brown and Levinson 1987) or sociofugal politeness (ĩywiczyĔski 2010),15 which refers to strategies promoting or highlighting the distance between interlocutors and/or between interlocutors and the described situation. For example, in Turn 2 of the interaction (8) the announcer “well” and the qualifier “I don’t think” together with the modal “can” remove the discourse from the deictic 13

Which is given in ĩywiczyĔski 2010: 213ff. For a detailed description of these two functions see ĩywiczyĔski 2010. 15 For details on the similarity between negative politeness and sociofugal politeness see ĩywiczyĔski 2010: 404–417). 14

314

Przemysáaw ĩywiczyĔski

centre of the here and now (realis region)16 into the region of irrealis – the politeness effect is thus achieved by expressing the refusal of the invitation by means of the hedge on the speaker’s ability to fulfil the action mentioned in the prior turn (i.e. the speaker does not express a straightforward refusal to come and have a cup of coffee with the inviter but does so by means of the hedge on the speaker’s ability to come and have a cup of coffee with the inviter). Secondly, the combination of the token appreciation (“that’s awfully sweet of you”), the delayed and mitigated declination (“hehh well [...] I don’t think I can make it this morning”) and the account component (“I’m running an ad in the paper and-and uh I have to stay near the phone”) does seem to create a context conducive to negotiating the initiator’s original extra-discursive goal. In this way, the dispreferred organisation of Turn 2 in (8) suggests B’s openness to accept a rescheduled version of the invitation and – in axiological terms – to engage in MAIN negotiations.

7.3 Axiological organisation of adjacency sequences Furthermore, the axiological study of the corpus uncovered two basic classes of adjacency pairs – one class, typified by invitations, offers, requests, and questions, is related to transfer of goods, services, and information; the other class, represented by assessments and selfdeprecations, concerns aligning interactants’ view on a particular aspect of reality.17 Out of these two types, which have respectively been designated as economic and phatic exchanges, I will – for want of space – focus only on the former. In my view, such a limitation should not, however, hinder identifying a number of essential characteristics of the spoken discourse. It should be observed that in economic exchanges, transfer of goods, services, etc. can proceed in two opposite directions: either from the initiator of an exchange towards the respondent, as is the case in invitations and offers, where the invitee or the offeree is the intended beneficiary of transfer; or from the respondent towards the initiator, which happens for examples in requests and questions, where it is the requester or the questioner who is the intended beneficiary of transfer. Thus, the 16

For details on the concept of deictic centre see Bühler (1934) and Miller and Johnson-Laird (1976). 17 As explained by Pomerantz (1984), self-deprecations represent a special type of assessments, in which the initiator produces a self-denigrating comment and the respondent preferentially disagrees and dispreferentially agrees. Thus, the organisation of self-deprecations is opposite to that of ordinary assessments, where agreement is preferential and disagreement, dispreferential.

Discourse Goals and Norms: An Axiological Perspective

315

directionally of transfer leads to the sub-classification of economic adjacencies into other-benefiting exchanges (e.g. invitations and offers) and self-benefiting exchanges (e.g. requests and questions). Given the above characterisation, axiological clashes arsing during other-benefiting exchanges should be seen as blocking the initiator’s extra-discursive goal of effecting economic transfer towards the respondent, whereas in selfbenefiting exchanges axiological clashes block the initiator’s extradiscursive goal to receive economic transfer from the respondent. Next, the study of 20 interactions containing global clashes revealed that economic organisation has consequences for axiological structure. Namely, when a dispreferred scenario is instigated in other-benefiting exchanges, a reparatory MAIN action is started by the initiator (i.e. inviter or offerer), who proposes a modified or an altered version of an original invitation or request as illustrated below: (9)

Turn 1 A: Uh: would it be alright if we came in a little early

Turn 2 B: (0.2)

Turn 3 A: Would that upset //you?

Turn 4 B: I: don’t think so (Davidson 1984: 115)

This pattern was consistent in all of the eleven other-benefiting exchanges that were studied. On the other hand, MAINs in self-benefiting sequences are usually introduced by respondents – in our sample, 8 out of 9 exchanges showed this pattern. For example in request exchanges, respondents’ most common tactic is to delay the granting of a request ((10) Turn 2) or to grant an altered version of an original request ((11) Turn 2): (10)

Turn 1 A: Will you read me this story Turn 2 B: Well after I’ve washed the dishes I’ll read you that story (.) yes

(11)

Turn 1 A: I want a’apple –hh two p-hh apples each (1.6) Turn 2 B: Half and apple each = I don’t think we’ve got so many apples

316

Przemysáaw ĩywiczyĔski

Hence, it can be argued that economic exchanges have an “altruistic” character in the sense that initiative for reparatory steps following a clash is taken by the source rather than the target of transfer of goods. To reiterate, in other-benefiting exchanges, where the objective is to bring benefit to the respondent (i.e. intended beneficiary of an exchange), MAIN initiative is taken by the initiator with the aim of effecting some form of economic transfer to the respondent (such is, for example, the role of modified invitations or offers put forward by the inviter or offerer); in selfbenefiting exchanges, where the objective is to bring benefit to the initiator (i.e. intended beneficiary of an exchange), MAIN initiative is taken by the respondent with the aim of effecting some form of economic transfer to the initiator (for example, by suggesting altered versions of original requests).

8. Discourse norms In my view, the above study demonstrates the inadequacy of teleological accounts of conversational interaction. Rather than adhering to individualistic agendas, discourse participants help each other in accomplishing their respective extra-discursive goals and, in doing so, are prepared to resign from their own economic gains (in economic games) – the fact that was illustrated by the analysis of dispreferred sequences in the preceding section. These phenomena suggest the existence of a normative element which seems to override the order of participants’ individual extra-linguistic goals. In the most general terms, this normative element is expressed as interactants’ commitment to cooperate in forwarding each other’s extra-discursive goals and, in the event that the pursuit of these goals is blocked, as the commitment to collaboratively solve the emergent interactional crisis. To underscore the affiliative nature of cooperation which characterises conversational interaction, I will appeal to the ethological concept of “synagonal behaviour” – that is, the type of behaviour which is geared towards friendship, bonding, and group harmony (Eibl-Eibesfeldt 1989: 520–521) – and refer to this normative element as the synagonal norm of discursive cooperation.18 More specifically, discursive cooperativeness discussed here should be seen as comprised of two dimensions. First, interactants cooperate – or to put it in normative terms: are committed to cooperating – in the accomplishment of each other’s extra-discursive goals 18 For more on the synagonal nature of discursive cooperation see ĩywiczyĔski 2010: 154–160.

Discourse Goals and Norms: An Axiological Perspective

317

and, when the accomplishment of these is blocked, interactants are committed to setting and pursuing surrogate goals (by e.g. altering the original versions of invitations, offers, requests, etc.). On the other hand, interactants are obliged to cooperate in upholding the format of exchanges in which they are engaged. This is done by: correctly identifying and sticking to their roles in exchanges (e.g. the initiator of in an invitation exchange sequence should assume the role of the inviter and adhere to it throughout an exchange), using dispreferential signals when they introduce global clashes into discourse, and observing rules related to taking MAIN actions. The last of these points pertains to the discovery of different axiological configurations in axiologically complex exchanges. Accordingly, in economic exchanges participants who are the source of goods transfer are expected to show MAIN initiative; that is, in otherbenefiting games (e.g. invitations and offers) the responsibility to instigate a reparatory action is vested in the initiator of an exchange (i.e. inviter and offerer), while in self-benefiting ones (e.g. requests) it is primarily the duty of the respondent (i.e. requestee). The two dimensions of discursive cooperativeness described above will respectively be designated as “material cooperativeness” (i.e. cooperativeness in the pursuit of extra-discursive goals) and “formal cooperativeness” (i.e. cooperativeness in upholding the format of exchanges). The existence of these modes of cooperativeness allows me to analyse to the synagonal norm of discursive cooperation into two component norms: (i) the material norm, which commits interactants to cooperation in accomplishing extra-discursive goals; and (ii) the formal norm, which commits interactants to cooperation in maintaining the format of exchanges. It should be recalled that the operation of a norm of cooperation was also discovered on the transactional plane of discourse – the transactional norm, as it was termed, committed participants to upholding the informational continuity of discourse (see Section 6). I argue that it should be viewed as yet another manifestation of the synagonal norm of discursive cooperation, which specifically governs interactants’ behaviour on the transactional plane. In this way, the axiological study of conversational exchanges leads to the view of interaction as organised by the synagonal norm of discursive cooperation consisting of the three component norms: the transactional norm operates on the transactional plane and regulates the pursuit of intra-discursive goals; the other two – the material and formal norm – pertain to the interactional place and regulate interactants’ conduct in the manner explained above.

318

Przemysáaw ĩywiczyĔski

In this respect, it is interesting to note that the initial stages of my project were dominated by a discussion of discourse goals, in particular intra- and extra-discursive goals (see Section 5.1). The goal-orientated approach was very much in line with Krzeszowski’s own analyses of axiological phenomena in discourse.19 However, when this original analytical framework was applied to the study of conversational data, it turned out that the axiological structure of discourse points to the interactional order which transcends the goal-oriented account and that axiological phenomena remain largely inexplicable if such an order is not posited. In particular, the goal-oriented analysis is unable to explain the appearance of MAIN sequences – if interactants are primarily concerned with their individualistic goals, why should they help overcome problems in accomplishing fellow interactants’ goals? Without the normative order which commits individuals to discursive cooperation, it is difficult to see how they could successfully overcome the obstacles which appear on the path of discourse. From this perspective, positing the discursive norms does not seem to arise only from the necessity of giving the axiological model of discourse a firmer theoretical footing, but mainly from the fact that such norms are indispensable in explaining the very nature of discursive interaction.

9. Concluding remarks In view of the corpus analysis the results of which have been presented here,20 there is ample evidence to confirm that Krzeszowski’s account of discourse-axiological phenomena, on which this project was founded, captures the essential properties of discursive interaction – the generic structure of the axiological cycle, consisting of an axiological clash, MAIN, and CREST, seems to be a universal characteristic of spoken discourse. More importantly, the value of this study consisted in explaining the discursive role of such conversational phenomena as chaining sequences, embeddings, and dispreferential signals. Important correspondences were noted between the axiological approach to interaction, on the one hand, and Gricean pragmatics, politeness research, and human ethology, on the other. However, the most significant achievement of this project was uncovering the teleological and normative structures of discourse. The identification of two sets of goals – intra- and extra-discursive – and the synagonal norm of discursive cooperation with 19 20

Krzeszowski 1997: 249–250. For an extensive presentation see ĩywiczyĔski 2010.

Discourse Goals and Norms: An Axiological Perspective

319

its three component norms – transactional, formal, and material – may contribute to a better understanding of conversational discourse.

References Atkinson, J. Maxwell and Paul Drew. 1979. Order in Court: The Organization of Verbal Interaction in Juridical Settings. London: Macmillan. Beaugrande, Robert de and Wolfgang Dressler. 1981. Introduction to Text Linguistics. Harlow: Longman. Brown, Gillian and George Yule. 1983. Discourse Analysis. Cambridge: Cambridge University Press. Brown, Penelope and Stephen Levinson. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Bühler, Karl. 1934. Sprachtheorie. Die Darstellungsfunktion der Sprache [Theory of language : the representational function of language]. Jena: Gustav Fischer. Coates, Jennifer. 1986. Women, Men and Language: A Sociolinguistic Account of Gender Differences in Language. London: Longman. Coulthard, Malcolm. 1992. The Significance of Intonation in Discourse. In Malcolm Coulthard (ed.), Advances in Spoken Discourse Analysis, 123–161. London: Routledge. Davidson, Judy. 1984. Subsequent Versions of Invitations, Offers, Requests, and Proposals Dealing with Potential or Actual Rejection. In J. Maxwell Atkinson and John Heritage (eds.), Structures of Social Action: Studies in Conversation Analysis, 102–128. Cambridge: Cambridge University Press. Drew, Paul and John Heritage (eds.) 1992. Talk at Work: Interaction in Institutional Settings. Cambridge: Cambridge University Press. Eggins, Suzanne and Diane Slade. 1997. Analysing Casual Conversation. London: Equinox. Eibl-Eibesfeldt, Irenäus. 1989. Human Ethology. New York: Aldine de Gruyter. Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs: Prentice-Hall. —. 1972. Remarks on Ethnomethodology. In John Gumperz and Dell Hymes (eds.), Directions in Sociolinguistics: The Ethnography of Communication, 301–324. New York: Holt, Reinhart and Winston.

320

Przemysáaw ĩywiczyĔski

Grice, Herbert Paul. 1975. Logic and Conversation. In Peter Cole and Jerry Morgan (eds.), Syntax and Semantics, 41–58. New York: Academic Press. Krzeszowski, Tomasz P. 1997. Angels and Devils in Hell: Elements of Axiology in Semantics. Warszawa: Energeia. Levinson, Stephen. 1983. Pragmatics. Cambridge: Cambridge University Press. Miller, George A. and Philip Johnson-Laird. 1976. Language and Perception. Cambridge, Massachusetts: Harvard University Press. Pomerantz, Anita. 1984. Agreeing and Disagreeing with Assessments: Some Features of Preferred and Dispreferred Turn Shapes. In J. Maxwell Atkinson and John Heritage (eds.), Structures of Social Action. Studies in Conversation Analysis, 57–101. Cambridge: Cambridge University Press. Sacks, Harvey. 1992. Lectures on Conversations. Volumes I and II. Oxford: Blackwell. Sacks, Harvey and Emmanuel Schegloff. 1973. “Opening up Closings.” Semiotica 7.4: 289–327. Sacks, Harvey, Emmanuel Schegloff, and Gail Jefferson. 1974. “A Simplest Systematic for the Organization of Turn-Taking in Conversation.” Language 50.4: 696–735. Sacks, Harvey, Emmanuel Schegloff, and Gail Jefferson. 1978. A Simplest Systematic for the Organization of Turn-Taking in Conversation. In Jim Schenkein (ed.), Studies in the Organization of Conversational Interaction, 7–55. New York: Academic Press. Sinclair, John and Malcolm Coulthard. 1975. Towards an Analysis of Discourse: The English Used by Teachers and Pupils. Oxford: Oxford University Press. Stenström, Anna-Brita. 1994. An Introduction to Spoken Interaction. London: Longman. Tannen, Deborah. 1989. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse. Cambridge: Cambridge University Press. —. 1994. Gender and Discourse. Oxford: Oxford University Press. Wootton, Anthony. 1981. “The Management of Grantings and Rejections by Parents in Request Sequences.” Semiotica 37:1/2: 59–90. ĩywiczyĔski, Przemysáaw. 2010. The Axiology of Spoken Interaction. An Essay on the Conversational Behaviour. ToruĔ: Wydawnictwo Naukowe UMK.

PART SIX NORMATIVE LINGUISTICS

STEPHEN FRY AND DAVID CRYSTAL: A THESPIAN AND A LINGUIST WRITING ON MODERN-DAY PRESCRIPTIVISM, LANGUAGE COMPLAINT AND LANGUAGE CHANGE IN ENGLISH MACIEJ RATAJ

1. Introduction Comparison may be interpreted by a linguist in a variety of ways. It may mean comparing two languages, two dialects or the views of two linguists or linguistic schools on the same issue. This paper, however, transcends the boundary of linguistics proper, in that it focuses on comparing the ideas of two famous people who are both concerned with the English language, albeit from different points of view: Stephen Fry and Professor David Crystal. Before we proceed to sociolinguistic analysis, let us briefly consider what the two people in question have in common. First, Stephen Fry is not only a drama and comedy actor famous for putting his voice and his Received Pronunciation to good use (as he wrote, “My vocal cords are made of tweed. I give off an air of Oxford donnishness and old BBC wirelesses,” (see “Keeping up appearances” at the Icons website)) but also the author of a play, some novels, diaries and a two-volume autobiography. In all his endeavours he is praised for showing an excellent command of English vocabulary and style. Although he holds a degree in English Literature from Cambridge University (see “Stephen Fry” at the BHA website) he has never been an academic linguist. On the other hand, we have Professor David Crystal, arguably Britain’s most famous living linguist. Since 1964, Crystal has written, edited or co-authored over 100 books on a wide variety of linguistic and other subjects (see David Crystal’s website). What the actor and the linguist share is involvement in memoir-writing and blogging, interest in the theatre, as Crystal has worked on staging Shakespeare’s plays in the original pronunciation, and,

Stephen Fry and David Crystal on Modern-Day Prescriptivism

323

as will be shown, also tolerance of non-standard usage and intolerance of linguistic purism. The analysis conducted in this paper is based on “Don’t Mind Your Language,” Fry’s blog entry which is almost identical to an episode of his podcast; the views presented there are compared to several books and articles written by Crystal in recent decades.

2. Language pedants and their complaints Fry discusses the widespread phenomenon of language complaint and criticism of non-standard English in general, a trend well known to linguists and discussed in numerous works, e.g. Milroy and Milroy (1998) and Pinker (1994). Incidentally, this trend is by no means uniquely British, and analyses of language complaint among speakers of other languages have also been done (see e.g. the examples from French and Polish in Rataj 2010). To resume, this is what Fry has to say about purists: They write letters to broadcasters and newspapers in which they are rude and haughty about other people’s usage and in which they show off their own superior ‘knowledge’ of how language should be. I hate that, and I particularly hate the fact that so many of these pedants assume that I’m on their side. When asked to join in a “let’s persuade this supermarket chain to get rid of their “five items or less” sign” I never join in. (Fry 2008).

Purists write to Fry for support probably because he is known for his standard accent and mastery of English style in general. A more likely reason is that he used to be more of a pedant himself. Referring to an earlier sentence he writes, “”None of these are of importance,” I wrote here, you’ll notice – the old pedantic me would have insisted on “none of them is of importance.” Well I’m glad to say I’ve outgrown that silly approach to language,” (Fry 2008). Fry goes on to provide more examples of the lexical pet hates of purists, i.e. the confusion of “uninterested” and “disinterested” as well as “infer” and “imply.” He also mentions the nonstandard use of apostrophes, sentence-final prepositions and so-called split infinitives, where an adverb or an adverbial phrase is used between the particle “to” and the verb proper. Besides, Fry writes, “The worst of this sorry bunch of semi-educated losers are those who seem to glory in being irritated by nouns becoming verbs. How dense and deaf to language development do you have to be?” He supports this argument by reminding his readers that Shakespeare made use of noun-to-verb conversion very frequently, hence this process is neither new nor highly original in English.

324

Maciej Rataj

As far as Crystal is concerned, linguistic complaint and the people he calls “pundits” feature prominently in his 2006 book The Fight for English, which is a history of British English from the viewpoint of standard dialect ideology (a term from Milroy and Milroy 1998), and some of his other works, e.g. The Stories of English (2004), contain analyses of the prescriptivism of the past and current popular ideas of correctness or what is perceived as purity in Standard English. Like Fry, Crystal mentions pundits who express their anger via the mass media. He recounts the angry letters sent to the BBC Radio when he was the host of a programme on the English language in the 1980s. It emerged that the listeners were against his pronunciation of vowels and criticised some widespread non-standard forms in general (Crystal 2006: 182–183; also 2004: 249). Worse still, pedants actually won a battle when Susan Rae, whose Scottish accent was frequently complained about by BBC listeners, lost her job as an announcer in 1980. Fortunately, writes Crystal, she returned to radio announcing many years later (2006: 185–186). As for particular forms that are objects of complaint, in Crystal (2006: 95) one may find some of Fry’s examples of words which according to pedants are misused, namely “disinterested” and “infer.” Crystal disapproves of complaints about these lexical items as they are uncommon, so that using them in the original or prescribed meaning is hardly an indicator of a person’s command of Standard English: we are far more likely to decide whether or not someone can use the standard dialect by judging other factors of their speech, such as pronunciation. Misplaced or missing apostrophes, split infinitives and prepositions finishing sentences and relative clauses are also discussed by Crystal. Firstly, the infamous greengrocer’s apostrophe has been a subject of derision for a long time. Crystal mentions the historical uncertainty regarding the use of this punctuation sign, which began its career as the possessive marker in the 1700s (2006: 140). The disparagement of the error occurring on shop signs alarms Crystal, who quotes Lynne Truss (2003: 15): “The only illiteracy that stirs any sympathy in me is the greengrocer’s variety. First, because greengrocers are evidently hornythumbed people who do not live by words.” This he regards as continuation of eighteenth-century social discrimination based on linguistic grounds (Crystal 2006: 137). Secondly, split infinitives are mentioned six times in Crystal’s 2004 book and six times in his 2006 book. He is against the proscriptive rule for several reasons, including the natural rhythm of stressed and unstressed syllables and the cases where not putting the adverb between the two parts of the infinitive alters the meaning, as in “They failed to completely understand the problem” versus

Stephen Fry and David Crystal on Modern-Day Prescriptivism

325

“They failed completely to understand the problem” (2006: 126–127). He also writes about the diminishing importance of the ban from the sociolinguistic viewpoint, “In the 2000s, the contentiousness surrounding this construction is thankfully diminishing, for it was never sanctioned by the leading usage pundits [...] But it is still one of the most commonly cited prescriptive rules. Purists have surprisingly long memories sometimes,” (2004: 465–466). Thirdly, Crystal devotes a subchapter of his 2004 book to the history of clause-final prepositions and their stigmatisation, quoting the famous Enlightenment grammarians Archbishop Robert Lowth and Lindley Murray and their impact on the prescriptivism and purism of today (Crystal 2004: 400–401). Needless to say, he shows the error inherent in this proscriptive rule and in another book of his he quotes The Queen’s English by Henry Alford (1869), wherein the author confesses that he cannot avoid sentence-final prepositions himself (Crystal 2003: 367). By proving that a classic prescriptive grammarian is unable to comply with this arbitrary rule Crystal shows the fallacy of prescriptive ideas even better. What language pundits often claim to be fighting for is clarity instead of such intangible and subjective qualities as euphony or beauty. This is yet another facet of standard language ideology, owing to which speakers regard clarity as a major aim of standardisation and something objective that can be fully attained in speech and writing. Consider how Fry dismisses their argument. Pedants will also claim, with what I am sure is eye-popping insincerity and shameless disingenuousness, that their fight is only for “clarity.” This is all very well, but there is no doubt what “Five items or less” means, just as only a dolt can’t tell from the context and from the age and education of the speaker, whether “disinterested” is used in the “proper” sense of nonpartisan, or in the “improper” sense of uninterested. No, the claim to be defending language for the sake of clarity almost never, ever holds water. Nor does the idea that following grammatical rules in language demonstrates clarity of thought and intelligence of mind. (Fry 2008)

It is clear from his words that he finds the argument concerning supposed clarity invalid. So does Crystal (2006) in an entire chapter entitled “Clarity.” There apart from the above-mentioned discussion of split infinitives, he also considers another favourite subject of pedants, that is the position of the adverb “only” (2006: 127–128), and elsewhere in the book the “shall” and “will” distinction. He claims that in the past only a handful of principles differentiating standard or “polite” speech were found and that “these rules were propounded with maximum authority and

326

Maciej Rataj

severity, and given plausibility by the claim that they were going to help people to be clear and precise” (2006: 122). In other words, the fight for clear English is a legacy of age-old arguments thanks to which standard dialect speakers were able to impose their native speech upon others, all in the name of facilitating communication.

3. Language correctness and language change In the blog entry discussed here Fry presents a statement which is obvious to the present-day linguist and at the same time controversial to many a layperson: objective language correctness does not exist. He connects this, however, with the view that using Standard English is nonetheless important when the social context so requires. I admit that if you want to communicate well for the sake of passing an exam or job interview, then it is obvious that wildly original and excessively heterodox language could land you in the soup. I think what offends examiners and employers when confronted with extremely informal, unpunctuated and haywire language is the implication of not caring that underlies it. You slip into a suit for an interview and you dress your language up too. You can wear what you like linguistically or sartorially when you’re at home or with friends, but most people accept the need to smarten up under some circumstances – it’s only considerate. But that is an issue of fitness, of suitability, it has nothing to do with correctness. There no right language or wrong language any more than are right or wrong clothes. Context, convention and circumstance are all. (Fry 2008)

As regards Crystal’s views of correctness, they naturally agree with Fry’s and are evident for instance in his mocking of eighteenth-century English grammarians (2006: 105–106) or simply stating their negative impact on the correctness ideology of the Anglophone community but at the same time praising the prescriptivists’ accuracy of grammatical description (2004: 396ff). Interestingly, the fashion analogy is used by Crystal in at least two of his books. First, he compares the range of formality in language to clothing choices at the beginning of his history of English (2004: 9). Then, in his book on linguistic pedantry, he places the analogy in his discussion of education: Appropriateness in language is the same as appropriateness in other walks of life. Take clothing. If you looked into your wardrobe and found there only one suit of clothes, or one dress, how prepared would you feel to face the sartorial demands made upon you by society? [...] The more types of

Stephen Fry and David Crystal on Modern-Day Prescriptivism

327

clothing we have, the better. But having a large and varied wardrobe is only useful if we have developed a “clothes sense.” (2006: 102)

He goes on to write that children who are familiarised with a wide range of language varieties at school are better equipped for living in their linguistic community, or, in Crystal’s own words, “they leave school linguistically fully dressed,” and that language education needs to teach schoolchildren when it is appropriate to employ one variety rather than another (2006: 103–104). From Fry’s and Crystal’s ideas we can thus abstract at least three similes: 1. “Formality in language is like formality in clothing,” 2. “One’s linguistic repertoire is like one’s wardrobe,” and 3. “Language appropriateness is like a dress code.” The use of fashion as a frame of reference matches the idea of language use perfectly as both domains possess a social dimension: in this day and age, just as clothing no longer just protects us from the elements but also gives others social signals in a given situational context, the language we use also makes an impression on the addressees of our texts or utterances, marking class, education and regional origin, to mention but a few. This is why both Fry and Crystal emphasise the importance of knowing how to use one’s language in a given situation, which obviously is not equal to complying with prescriptive rules. One may note in passing that the fashion analogy in reference to language has been used by at least one more scholar, namely LubaĞ, who states just as that clothes demonstrate one’s customs, regional origin or lifestyle, language allows a speaker to show their belonging to or distance from a social group (1979: 202–203). To summarise, prescriptivists and pedants can be compared to people who wish that everyone invariably wore very similar elegant clothes instead of dressing differently according to a variety of factors and circumstances. Another sign of Fry’s linguistic tolerance is his acceptance of language change: he compares it to evolution in nature, something beneficial and indeed necessary to the life of any language. He thus addresses his readers, “Imagine if we all spoke the same language, fabulous as it is, as Dickens? Imagine if the structure, meaning and usage of language was always the same as when Swift and Pope were alive. Superficially appealing as an idea for about five seconds, but horrifying the more you think about it,” (Fry 2008). Although one cannot deny that languages need to change as the needs of their speakers alter, Fry’s attitude is consistent with the viewpoint of many sociolinguists. Consider the first of the ten principles of modern sociolinguistics presented by Crystal (2004: 529): “Language change is normal and unstoppable, reflecting the normal and unstoppable processes of social change.” The only difference between the opinion of Fry and that of sociolinguists lies in the fact that “evolution” is not a very

328

Maciej Rataj

fortunate word, as it implies progress. In fact, as Aitchison (2001: 253) concludes in her important work on the subject, language change leads it neither to improvement nor to deterioration.

4. Conclusions: do Fry and Crystal have identical views? At this point one may wonder whether Fry really behaves like an academic sociolinguist in his blog entry. In fact, as was mentioned earlier, he used to be a purist himself and admits that he dislikes certain words and features of contemporary English. He describes his aversion to the use of reflexive pronouns instead of plain personal pronouns and adds that to him pronouncing the name of the letter “aitch” with an initial /h/, as in “Haitch Eye Vee,” sounds like “nails on the blackboard” (Fry 2008). At the same time he is aware of the fact that some other people may dislike his way of speaking and his upper-class accent. Fry also voices criticism of the lack of a more expressive word for CCTV and suggests that his neologism “scunt” for “Surveillance Camera Universal Network” might be used instead, being more indicative of the sinister connotations of CCTV. Crystal, as a descriptive linguist, does not allow himself to show his subjective ideas; however, at one point in one of his books discussed here he does recommend shorter subjects as preferable to constructions in which a long subject is followed by a short predicate (Crystal 2006: 128– 130; for more on this stylistic principle see Macpherson 2001: 103–104). All this goes to show the basic difference between Fry’s blog and Crystal’s books: although both authors write about the issues of correctness for the benefit of the general reader, Fry does not attempt to make a piece of real linguistics out of his blog entry. He may cite de Saussure and claim to have been inspired by The Language Instinct by Steven Pinker (1994), who writes about language complaint from an American point of view, but as an actor is not limited by the rules of objectivity which are a must in any genuine research conducted in the descriptive linguistics tradition. His blog contains opinions, hence linguistic facts are of secondary importance there. By contrast, Crystal needs to be objective in his books and articles, and cannot criticise language pedants without providing sound evidence. Needless to say, in order to be taken seriously by contemporary linguists he cannot show negative attitudes to selected language features such as changes in progress in Standard English. All in all, the thespian and the scholar display strikingly similar modes of thinking even though their knowledge of language comes from different kinds of schooling and experience.

Stephen Fry and David Crystal on Modern-Day Prescriptivism

329

References Aitchison, Jean. 2001. Language Change: Progress or Decay? 3rd ed. Cambridge: CUP. Alford, Henry. 1869. The Queen’s English: Stray Notes on Speaking and Spelling. London: Strahan. Crystal, David. 2003. The Cambridge Encyclopedia of the English Language. 2nd ed. Cambridge: CUP. —. 2004. The Stories of English. London: Penguin. —. 2006. The Fight for English: How Language Pundits Ate, Shot and Left. Oxford: OUP. —. David Crystal’s website. http://www.davidcrystal.com/index.htm DOA: 29 April 2011. Fry, Stephen. “Don’t mind your language.” Blog entry of 4 Nov. 2008. The New Adventures of Mr Stephen Fry website. http://www.stephenfry.com/2008/11/04/dont-mind-your-language %E2%80%A6/ DOA: 29 April 2011. LubaĞ, Wáadysáaw. 1979. Spoáeczne uwarunkowania wspóáczesnej polszczyzny: Szkice socjolingwistyczne. [The social conditions of Present-Day Polish: Sociolinguistic drafts] Kraków: Wydawnictwo Literackie. Macpherson, Robin. 2001. Advanced Written English. Warsaw: PWN. Milroy, James and Lesley Milroy. 1998. Authority in Language: Investigating Standard English. 3rd ed. London: Routledge. [no author 1] “Stephen Fry.” British Humanist Association website. http://www.humanism.org.uk/about/people/distinguishedsupporters/stephen-fry DOA: 29 April 2011. [no author 2] “Keeping up appearances.” Icons: the Portrait of England website. http://www.icons.org.uk/theicons/collection/blackcab/features/keeping -up-appearances DOA: 7 May 2011. Pinker, Steven. 1994. The Language Instinct. London: Penguin. Rataj, Maciej. 2010. On the nature of linguistic complaint. In Danuta Stanulewicz, Tadeusz Z. WolaĔski and Joanna Redzimska (eds.), Lingua terra cognita: A festschrift for Professor Roman Kalisz, vol. 2, 619–636. GdaĔsk: Wydawnictwo Uniwersytetu GdaĔskiego. Truss, Lynne. 2003. Eats, Shoots and Leaves. London: Profile Books.

CONTRIBUTORS

Anna Bloch-Rozmej – John Paul II Catholic University of Lublin Maria Bloch-Trojnar – John Paul II Catholic University of Lublin Magdalena CharzyĔska-Wójcik – John Paul II Catholic University of Lublin Tatiana Dubrovskaya – Penza University of Ecology and Political Science Tomasz Fojt – Nicolaus Copernicus University, ToruĔ Dorota Gorzycka – Nicolaus Copernicus University, ToruĔ Wojciech Guz – John Paul II Catholic University of Lublin Joanna Kolbusz-Buda – University of Natural Sciences and Humanities, Siedlce Katarzyna Molek-Kozakowska – Opole University Magdalena Murawska – Adam Mickiewicz University, PoznaĔ Wiktor Pskit – University of àódĨ Maciej Rataj – University of GdaĔsk Izabela SekĞciĔska – Nicolaus Copernicus University, ToruĔ Waldemar Skrzypczak – Nicolaus Copernicus University, ToruĔ Agnieszka Solska – University of Silesia, Katowice Agnieszka SowiĔska – Nicolaus Copernicus University, ToruĔ Ariadna Strugielska – Nicolaus Copernicus University, ToruĔ Jolanta SzymaĔska – Opole University Sáawomir Wacewicz – Nicolaus Copernicus University, ToruĔ Agnieszka Wawrzyniak – Adam Mickiewicz University, PoznaĔ Jerzy Wójcik – John Paul II Catholic University of Lublin Przemysáaw ĩywiczyĔski – Nicolaus Copernicus University, ToruĔ