Insistent Images (Iconicity in Language and Literature, Volume 5)

Citation preview

Insistent Images

Iconicity in Language and Literature A multidisciplinary book series which aims to provide evidence for the pervasive presence of iconicity as a cognitive process in all forms of verbal communication. Iconicity, i.e. form miming meaning and/or form miming form, is an inherently interdisciplinary phenomenon, involving linguistic and textual aspects and linking them to visual and acoustic features. The focus of the series is on the discovery of iconicity in all circumstances in which language is created, ranging from language acquisition, the development of Pidgins and Creoles, processes of language change, to translation and the more literary uses of language.

Editors

Olga Fischer

University of Amsterdam

Christina Ljungberg University of Zurich

Volume 5 Insistent Images Edited by Elżbieta Tabakowska, Christina Ljungberg and Olga Fischer

Insistent Images Edited by

Elżbieta Tabakowska Jagiellonian University, Krakow

Christina Ljungberg University of Zurich

Olga Fischer University of Amsterdam

John Benjamins Publishing Company Amsterdam / Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Insistent Images / edited by Elżbieta Tabakowska, Christina Ljungberg and Olga Fischer Symposium on Iconicity in Language and Literature (5th : 2005 : Jagiellonian University) p. cm. (Iconicity in Language and Literature, issn 1873–5037 ; v. 5) Papers from the Fifth Symposium on Iconicity in Language and Literature, organized by the Jagiellonian University in Kraków and held March 17–20, 2005. Includes bibliographical references and index. 1. Iconicity (Linguistics)--Congresses. P99.4.I26 S96

2005

302.21--dc22



2006047947

isbn 978 90 272 4341 6 (Hb; alk. paper) © 2007 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Table of contents

Table of contents Preface and acknowledgements List of Contributors Introduction: Insistent Images Christina Ljungberg and Elžbieta Tabakowska

vii 1x 1

Part I. Iconicity and grammaticalization Putting grammaticalization to the iconicity test Paul Bouissac

17

Iconic thumbs, pinkies and pointers William J. Herlofsky

37

Part II. Iconicity and the aural The physical basis for phonological iconicity Keiko Masuda

57

Reading aloud and Charles Dickens’ aural iconic prose style Tammy Ho Lai Min

73

Iconicity and the divine in the fin de siècle poetry of W.B. Yeats Sean Pryor

91

Is lámatyáve a linguistic heresy? Iconicity in J.R.R. Tolkien’s invented languages Joanna Podhorodecka

103

Part III. Iconicity and the visual The beauty of life and the variety of signs: Guillaume Apollinaire’s ‘lyrical ideogram’ La Cravate et la montre Peter Gahl Forms of restricted iconicity in modern avant-garde poetry John J. White

113 129

v

vi

Table of contents

Eco-iconicity in the poetry and poem-groups of E. E. Cummings Etienne Terblanche and Michael Webster

155

The language of film is a matrix of icons Strother B. Purdy

173

Liberature: A new literary genre? Katarzyna Bazarnik

191

Part IV. Iconicity and conceptualization Meaning on the one and on the other hand: Iconicity in native vs. foreign signed languages Meike Adam, Wiebke Iversen, Erin Wilkinson and Jill P. Morford

211

Iconic text strategies: Path, sorting & weighting, kaleidoscope Friedrich Ungerer

229

‘Damn mad’: Palindromic figurations in literary narratives Christina Ljungberg

247

Part V. Iconicity and structure Iconicity and the grammar–lexis interface Dylan Glynn Iconicity in the coding of pragmatic functions: The case of disclaimers in argumentative discourse Angelika Bergien Double negation and iconicity Ludovic de Cuypere, Johan van der Auwera and Klaas Willems

269

289 301

Part VI. Iconicity and multimedia / intertextuality Iconicity in multimedia performance: Laurie Anderson’s White Lily Sibylle Moser

323

Author index Subject index

347 353

List of contributors vii

Preface and acknowledgements The studies in this volume form a selection of the most relevant papers given at the Fifth Symposium on Iconicity in Language and Literature, organized by the Jagiellonian University in Kraków. It took place on 17–20 March, 2005 and gathered more than 70 scholars from 18 countries all over the world. Following the tradition of earlier symposia, the main topic has remained the same: looking for meaningful similarities between form and meaning. But the Kraków meeting meant a further widening of scope: contributions from music and visual arts, initiated at the fourth symposium in Louvain-la-Neuve in 2003, were now given full attention. Making use of the fact that there are two artistic universities in Kraków – the Academy of Fine Arts and the Academy of Music – scholars from both institutions were invited to present papers at the 2005 symposium. Some of their contributions found their way into the present volume; some more papers – originally written in Polish or translated into Polish – were gathered in a separate volume, to be soon published by the Kraków Academic Publishing House UNIVERSITAS. The organizers of the Kraków symposium would like to express their gratitude to those who helped to make it what it was: a wonderful and inspiring meeting of friends and scholars. Our special thanks go to the Director of the Institute of English of the Jagiellonian University and to the British Council in Kraków, who let us use their premises free of charge, to all student-volunteers who gave us so much of their time and energy, to Marta Bożyk and her friends, who awakened the iconic potential hidden of our bodies during the iconic movement workshop, and to the Chamber Music Orchestra of the Kraków Academy of Music, who delighted our guests with their playing. Much gratitude also goes to John J. White, a frequent participant in the Symposia and a knowledgeable and appreciated contributor to all five publications in the series, for his assistance as a reader and advisor to this volume. We are immensely sad to have to announce that Max Nänny, one of the founding fathers of our symposium (as Wolfgang Müller aptly called him in an earlier volume) very unexpectedly died on February 4, 2006. We remember him with great fondness for the work he did for the symposia, the charm and energy that he put into them, the interesting papers that he delivered, and his care and concern about the publication. We will sorely miss him. We hope that, by dedicating this volume to him, we can posthumously offer Max Nänny a memorial of our love and respect. E.T., C.L. and O. F.

viii List of contributors

List of contributors

List of contributors Meike Adam Collaborative Research Centre "Media and Cultural Communication", University of Cologne, Bernhard-Feilchenfeld-Str. 11, 50969 Cologne, Germany [email protected] Johan van der Auwera University of Antwerp, Campus Drie Eiken, Linguistics Department, Center for Grammar, Cogntion, and Typology, Universiteitsplein 1, 2610 Wilrijk, Belgium [email protected] Katarzyna Bazarnik Department of English, Jagiellonian University, al. Mickiewicza 9, 31-120 Krakow, Poland [email protected] Angelika Bergien Institut für fremdsprachliche Philologien, Otto-von-Guericke-Universität Magdeburg, Zschokkestraße 32, 39104 Magdeburg, Germany [email protected] Paul Bouissac University of Toronto (Victoria College), 253 College Street, Box 429, Toronto, ON M5T 1R5, Canada [email protected]

ix

x

List of contributors

Ludovic De Cuypere Ghent University, Department of German and General Linguistics, Jubileumlaan 33-35, 9000 Gent, Belgium [email protected] Olga Fischer Engels Seminarium, Universiteit van Amsterdam, Spuistraat 210, 1012 VT Amsterdam Th Netherlands [email protected] Peter Gahl FB Literaturewissenschaft, Fach D 169, Universität Konstanz, 784 57 Konstanz, Germany [email protected] Dylan Glynn Departement Linguïstiek, Universiteit Leuven, Blijde-Inkomststraat 21, Leuven 3000, Belgium [email protected] William J. Herlofsky Faculty of Foreign Languages, Nagoya Gakuin University, 1350 Kamishinano, Aichi-ken, Seto-shi, 480-12 JAPAN [email protected] Tammy Ho Lai Min School of English, The University of Hong Kong, Pokfulam Road, Hong Kong [email protected]

List of contributors

Wiebke Iversen Collaborative Research Centre, "Media and Cultural Communication", University of Cologne, Bernhard-Feilchenfeld-Str. 11, 50969 Cologne, Germany [email protected] Christina Ljungberg Englisches Seminar, University of Zurich, Plattenstrasse 47, 8032 Zurich, Switzerland [email protected] Keiko Masuda Faculty of Commerce, Chuo University, 742-1 Higashinakano, Hachioji, Tokyo 192-0393, Japan [email protected], Jill P. Morford Department of Linguistics, University of New Mexico Humanities, Rm. 526, Albuquerque, NM 87131-1196, USA [email protected] Sibylle Moser Institut für Theater-, Film- und Medienwissenschaft, Batthyanystiege, Hofburg, 1010 Wien, Austria [email protected] Joanna Podhorodecka Pedagogical Academy of Krakow, ul. Karmelicka 41, 31-128 Krakow, Poland [email protected]

xi

xii

List of contributors

Sean Pryor Trinity College, University of Cambridge, Cambridge CB2 1TQ, United Kingdom [email protected], Strother Purdy 34 Blacksnake Lane, Bridgewater CT, USA 06752 [email protected] Elzbieta Tabakowska Instityt Filologii Angielskij Jagiellonian University of Cracow Al Mickiewicza 9/11 31-120 Krakow, Poland [email protected] Etienne Terblanche School of Languages (English) North-West University Potchefstroom Campus Private Bag X 6001 Potchefstroom South Africa [email protected] Michael Webster English Department 129 Lake Huron Hall Grand Valley State University Allendale, MI49401-9403 [email protected] John J. White Department of German, King’s College, University of London, Strand, WC2R 2LS, United Kingdom [email protected]

List of contributors xiii

Erin Wilkinson Department of Linguistics, University of New Mexico Humanities, Rm. 526, Albuquerque, NM 87131-1196, USA [email protected] Klaas Willems Ghent University Department of German and General Linguistics Blandijnberg 2 B-9000 Gent, Belgium [email protected]

xiv List of contributors

Introduction

Introduction Insistent Images Christina Ljungberg* and Elžbieta Tabakowska** *University of Zurich; **Jagiellonian University**

[We] are obsessed with correspondences. Similarities between this and that, between apparently unconnected things, make us clap our hands delightedly when we find them out. It is a sort of national longing for form – or perhaps simply an expression of our deep belief that forms lie hidden within reality; that meaning reveals itself only in flashes. (Rushdie 1982:300)

The tendency to perceive iconicity, or relations of similarity between a representation and the object it represents, seems to be characteristic of our biological makeup. Increasing evidence has shown that the “longing for form” that would allow “meaning to reveal itself”, which informs Rushdie’s protagonist Saleem Sinai’s conception of reality, pervades literature as well as other forms of representation. It also permeates everyday speech, which is why research on iconicity is particularly suited for fruitful collaboration between the disciplines of linguistic and literary studies. The conviction that we tend to look for similarities and correspondences between a word or name and the object it represents was the guiding principle in the first volume in the series, Form Miming Meaning (Nänny and Fischer 1999), which primarily dealt with iconicity in written texts and spoken discourse. The range was quickly widened in the subsequent volumes to encompass visual texts, signed and gestural languages, visual and auditory signing, intermedial iconicity, cognitive studies, reader-oriented approaches and music interacting with language. In this fifth volume in the series, Insistent Images, we find yet new and exciting departures, as the project this time spans the bridge to include such genres as ‘liberature’, film and multimedia performance. But let us first have a quick look at the dynamic development of iconicity research over the last few years to which this series testifies. A tracing of the rapid increase in scope from the first volume to the present should give an idea of the extent to which the awareness of the importance of similarities and correspondences between meanings and forms has grown. With their seminal introduction to the first volume, Form Miming Meaning, Max Nänny and Olga Fischer clearly staked out new territory by convincingly arguing the importance of iconicity as a

1

2

Christina Ljungberg and Elžbieta Tabakowska

creative force in both language and literature. Sprinkling their discussion on imagic and diagrammatic iconicity (metaphor being subsumed under the latter, as a form of mapping) with a host of relevant examples, they move from a phylogenesis of language development – from its mythical origins in ‘natural’ language to postmodern literary texts explicitly thematizing iconic features – to ontogenesis. While they repeatedly call attention to the necessity of an interdisciplinary approach, their exemplary intermingling of literary (Norrman, White, Alderson, Bernhart, Nänny, Webster, Bauer and, Halter) and linguistic approaches (Fónagy, Haiman, Fischer, Meier, Wyss, Tabakowska, Ungerer, Piller, Fischer, Kortmann) not only brings out interdisciplinary similarities but also shows different approaches to the same topic. Interesting to note is that, in so doing, they succeed in having the volume diagrammatically display the iconic possibilities and advantages that the interlacing of not only different disciplines but also of various approaches presents. The conception of iconicity as ‘form miming meaning’ was extended in the second volume, The Motivated Sign (Fischer and Nänny 2001), by Winfried Nöth, who proposed two forms of iconicity, endophoric and exophoric. Nöth argues that “form miming meaning” is exophoric and “form miming form”, which occurs in verbal repetition and in symmetrical structures, is endophoric. Exophoric iconicity pertains to the classical, purely semiotic concept of iconicity, which goes back to the ancient theory of mimesis and which Yuri Lotman used to develop his theory of iconicity as a secondary modeling system built on language which functions as a primary modeling system. Endophoric iconicity, on the other hand, concerns the syntagmatic and paradigmatic mapping of language (Nöth 2001:23), which includes features such as repetition, parallelism, alliteration, rhyme and meter. Nöth’s extension of the notion of iconicity from exophoric to endophoric iconicity thus accommodates two conflicting theories within literary aesthetics, namely the theory of literary mimesis and the theory of literary autonomy or self-reference. Thus, these two apparently irreconcilable theories of literary aesthetics have indeed the concept of iconicity in common instead of having it as a dividing line.1 The concept of endophoric iconicity was picked up by several contributors (White, Wolf) to the volume, which also features iconic sound (Sadowski, Norrman, Anderson), iconic features in typography (Henry, Nänny, Goh, Innocenti) and in grammatical (Conradie, Fischer, Jansen and Lentz, Lecercle) and textual (Müller, Ljungberg) structures. Whereas the first two volumes focused mainly on verbal texts, the third volume, From Sign to Signing (Müller and Fischer 2003), presents several new departures while continuing the path laid out by highly diversified articles on iconicity in verbal language (Conradie, Harm, Hampe and Schönefeld, Rothenburg, Pötters). The ‘signing’ in the title of the volume refers to signed and gestural

Introduction

languages, testifying to increasing scholarly interest in the iconic properties of this gestural mode of signification (Grote and Linz, Herlofsky, Hübler). Because signed languages, in which the sequential order, tone and emphasis of spoken language are less prominent, use space, motion and location instead, research in these areas can help to throw some light on the role of iconicity in discourse, since it is more transparent. The reason why signed languages are more iconic than spoken discourse has to do with dynamic forces at play: apart from the fact that they use highly motivated signs, such factors as their visual mode and their relative youth as languages contribute to both the iconicity of gestural signs and its relative lack of conventionalization as well as to its cultural-specificity. The title’s ‘signing’ also extends to investigations of auditory signs such as birdcalls (Masuda), or varieties of visual iconicity and iconic mapping (White, Moyle, Ohme, Bauer, Ljungberg) including conceptual blends (Herlofsky, Hiraga). A particularly interesting new departure is the highly innovative application of reader-response theory in which the reader ‘iconizes’ the literary text by following the instructions that can be found in it (Johansen). Further novel approaches in this volume are explorations into intermedial iconicity, which deals with the interdependence and interaction between different art media and the various forms this takes (Wolf). Intermedial iconicity is also a vital factor in translation in which counterbalancing of preserved or recreated expressions and conventionalized ones involves difficult decision-making that will fundamentally affect the way in which the text will – or will not – work (Tabakowska). Another novelty is the proposal of a ‘systems theory of information’ in which systems interact iconically with one another and change by interchanging information and energy (Sadowski). Cognitive processes involved in iconicity constitute the topic of several contributions to volume four, Outside-In- Inside-Out (Maeder, Fischer and Herlofsky 2005), forming one of the new departures in the series (Bouissac, Hampe, Tabakowska). This is, as the editors point out, reflected in the title itself: what is the relationship between that what happens “inside” the brain and the “outside”, that is, its external expressions of concept and ideas? In addition, the title’s chiastic form iconically enacts the idea of an ongoing, multi-directional activity, constantly blurring the boundaries between the inside and the outside, which several articles on “negative or inverted iconicity” demonstrate (Fill, Wolf, Ljungberg, White). Apart from diverse explorations into various aspects of iconicity and structure (Schönefeld, Kirtchuk-Halevi, Hollman, Slobin and Herlofsky), this volume widens the interdisciplinary field of studies in iconicity to music and its interaction with language (Georis). Music is a particularly interesting semiotic system. It appears to be the prototype of iconicity in its imitation of form such as seemingly representing “a babbling brook”, with the musical pattern graphically

3

4

Christina Ljungberg and Elžbieta Tabakowska

representing a two-dimensional form, though what we perceive as a natural sign is in fact an abstract reduction, which, moreover, is culturally coded: Japanese music, for instance, would represent the sound of rustling water very differently than Western music would. Our perception of a piece of music as performing the sound of a babbling brook is the result of constant learning, both conscious and unconscious, by listening to this over and over again until we register it as such.2 Insistent Images, the fifth volume, continues the series with a selection of the papers given at the Symposium held at the Jagiellonian University in Cracow in 2005. In addition to several theoretical and applied novel approaches to iconicity on the aural, visual, conceptual and structural levels, investigations into the iconic properties of film and into iconicity in multimedia performance form two new fields of inquiry. Film is, through its analogical relationship with the life-world, intrinsically iconic, while at the same time being, like photography, indexical by virtue of its mode of production, which makes it, in C.S. Peirce’s words, “physically forced to correspond point by point to nature” (CP 2. 246). In multimedia performance, on the other hand, iconicity functions intermedially by both integrating and reflecting processes of perception and conceptualization, which would seem to make both film and multimedia performance function as “insistent images” par excellence. But what precisely is an “insistent image”? Let us quickly recapitulate C.S. Peirce’s classification of the relations between a sign and its referent or object. A sign always refers to an object which can both be ‘real’ or imaginary (which means that it can also be a mere thought) and, in so doing, produces or generates a new sign. The relationship between the sign and its object may be iconic, indexical and symbolic, aspects that are always present in any sign process, although in different degrees, as in most cases one particular aspect or characteristic is being foregrounded (comparable to a point of view). Language signs are, as we all know, symbols, even though they may function indexically in e.g. deictic words, or iconically by foregrounding various forms of similarity or correspondence. New research even suggests that the structures of letters and symbols universally match objects in the life-world, which puts forward the importance of iconicity in written communication, as this suggests that there are underlying principles governing the shapes of human visual signs (cf. Changizi et al. 2006). Iconicity would therefore seem to be such an “underlying principle”, although often in ways that are much more complex than ordinarily assumed. In the case of apparently visual signs, such as a photograph, a film frame or a portrait, the relationship between the sign and its referent or object would seem much more directly iconic. A photograph, however, is mainly indexical, and the film image, too, as mentioned earlier. On the other hand, a painted portrait is, despite its uniqueness and presumed correspondence with its object, primarily symbolic

Introduction

and therefore conventional in the sense that it is an interpretation of the person depicted, filtered through the artist’s mind. It also has indexical properties as it evinces an existential connection with its object and, since the brush strokes on the painting refer to the movement of the artist’s hand, it directly refers to its manner of production. Its iconic properties lie in the extent to which the features represented on the portrait share similarities with the object, the person in the portrait. This relationship seems to be what shaped writing systems, too, since there is evidence that visual signs, although they have been culturally selected, have been chosen in order to match the topological contours in natural scenes and on grounds of visual processing, i.e. for optimization and visual recognition (Changizi et al 2006). And it is precisely this immediate recognition of the similarity between sign and object that writing systems obviously once aspired to and that a realistic portrait still does, which, if it were perfect, would have us forget that we are contemplating a sign or a painting. Sign and object, portrait and portrayed, would be one. That would make it an “insistent image …not referring to anything else, and in that sense concrete” (CP 8. 104), which would be the case with a photograph and a film image, too. Such a sign must be predominantly iconic, since only icons can be self-referential; moreover, icons have “the nature of an appearance, and as such, strictly speaking, exist[..] only in consciousness” (CP 4. 447). At the same time, since this sign is identified with its object, it is an indexed icon. In other words, it is a sign that makes us forget that we are looking at a sign, making us take the sign for the object to which it refers. This is of course a hypothetical situation. Yet, knowing the role similarities and correspondences play in all cognitive processes and the performative function they exert in their complex interaction with indexical signs to create new ‘realities’ makes icons indeed “insistent images” that not only spark novel developments but fundamentally influence the way we perceive and interpret the world. *** We have divided the volume into six sections. Part I deals with “Iconicity and Grammaticalization”. Questioning the influence of iconicity in language development, Paul Bouissac starts out by indeed “[p]utting grammaticalization to the iconicity test” and states that the topic of grammaticalization forms one of the most provocative and consistent paradigms in contemporary research on language. Recasting the issues of grammaticalization within the broader framework of iconicity theory, Bouissac discusses the claim that referential words implicitly presuppose a higher degree of iconicity than grammatical morphemes and finds the notion of lexical “concreteness” problematic. He examines the semantic evidence that purports to establish the opposition between “concrete”

5

6

Christina Ljungberg and Elžbieta Tabakowska

lexical items and “abstract” grammatical ones. Arguing for an utterance-based linguistics – since utterances are always context-related, whether real or imagined, which makes the whole lexicon “by necessity indexical or deictic” – he concedes at the same time that morphemes have contrastive or discrete values that are determined by consensual use within a given society at a certain time. In his view, the complexity of natural languages is such that all heuristic deduction of the phenomenology of language is doomed to fail as it is similar to the elusive search for an integrated understanding of the five forces in contemporary physics. Bouissac’s provocative opening, which raises more questions than it answers, is followed by William Herlofsky’s look at grammaticalization in Japan Sign Language (JSL), which displays a high degree of iconicity. In JSL, there are verbs of motion which have complex morphological structure and which are referred to as ‘classifier constructions’, where the movement expressed by the verb is combined with certain ‘animate-entity’ hand shapes called ‘classifiers’. These hand shapes iconically represent certain salient characteristics of their referents, and are said to identify their referents as belonging to certain ‘groups’ or ‘classes’, basing the classification on, e.g. physical properties of their referents, such as size and / or how the referents are handled (imitating the shape of the hand as it does the handling), or what class of animate entity the referent may belong to, such as human or vehicular. Herlofsky investigates the nature of the iconic handshapes in classifier-like constructions in JSL and how their functions and patterns of grammaticalization are generally consistent with those in other languages, with the focus on how the same thumb, pinky and pointer finger handshapes can represent human entities in grammatically different ways. Part II deals with “Iconicity and the aural”. In the first of the four papers in this section Keiko Masuda discusses two types of phonological iconicity (or sound symbolism) – the direct and the indirect – from the acoustic and the phonetic point of view and attempts to unravel the mechanism of phonological iconicity in general. In her argument, direct phonological iconicity pertains to the phenomena where the referent in the iconic ‘referer–referent’ relation is an acoustic signal (a sound/a noise). If the referent is a non-linguistic sound or a noise, the oral sound used to represent it directly models it; i.e. there is a direct mapping of the linguistic sound to the non-linguistic one, and the two display fairly direct acoustic similarity. In indirect phonological iconicity, the correspondence holds between a linguistic expression on the one hand and some non-auditory aspect of the world on the other: an emotion or feeling, or a physical property of an object, such as its colour or size. This type of iconicity can be either associative (e.g. involving the association between front vowels and small size or brightness) or phonaesthetic (where phonemes or phoneme clusters conventionally correspond to particular

Introduction

meanings). Looking for the physical basis of phonological iconicity, Masuda quotes language data to support her hypothesis that the basis for all types of sound symbolism may be the same: the front oral cavity. Tammy Ho’s paper deals with the issue of aural iconicity from an entirely different perspective: she examines the question of how the Victorian practice of reading aloud affected Dickens’s writing style. She argues that Dickens might have been writing his prose with reading aloud in view and that this assumption might have contributed to an aural style in his novels, in which we find a considerable amount of passages which are performance-oriented and particularly aural in nature. Dickens made use of various formal linguistic means, such as typography, onomatopoeia, sound patterning, sentence length and prose rhythm to foreground the sound patterning of his passages. As Tammy Ho points out, Pope’s dictum that the “sound should echo the sense”’ is generally assumed to concern “decorative” uses of literary language rather than raise the question whether there was a practical need for aural reception and appreciation of sound, for instance in novels such as those by Dickens. This is what her paper explores, arguing that the emphasis on sound in Dickens’s novels sometimes also serves the purposes of narrative power, as it has considerable rhetorical impact, which both heightens emotions and highlights key narrative moments. Sean Pryor discusses W.B. Yeats’s use of iconicity to revive the archetypal images and symbols abandoned by nineteenth-century poetry in order to create icons of the unknown and of the ineffable. As Pryor demonstrates, Yeats’s poetry of paradise shows a high degree of iconicity, in particular his two Byzantium poems, which seem both to describe and to embody an earthly paradise of art. Arguing that Yeats’s quest for iconicity began already in the poems that he wrote before the turn of the century, Pryor explores in detail the phonetic and the syntactic iconicity featured in these poems. There is, for example, a strong sense in which the poems are spells or incantations for paradise, at the same time as they express a fear akin to the iconoclast’s dread of idolatry: like God, paradise must always retain an ultimate alterity, and hence remain beyond the compass of human art. This creates a fundamental tension reflected in the iconic practices of Yeats’s manifest early poetry, in the conflict between the divine language of heaven and the earthly language of mortality. Joanna Podhorodecka tackles the question of sound symbolism, or phonetic iconicity, from yet another point of view: that of a linguist looking at samples of an invented language. Her analysis of J.R.R. Tolkien’s linguistic inventions involves the languages that Tolkien created as parts of the world represented in his Hobbit books: the two languages of the Elves, Quenya and Sindarin, and Black Speech – the language of the Orcs. Tolkien’s linguistic knowledge and his attested interest in “word-form in relation to meaning” are shown in his striving to create moti-

7

8

Christina Ljungberg and Elžbieta Tabakowska

vated linguistic forms such as would display the synaesthetic “taste of the sound” – the property of language that he called lámátyave. In agreement with the theory of sound symbolism – and supporting her arguments with those put forward by Fónagy (1999) – Podhorodecka demonstrates that Tolkien’s “linguistic heresy” shows what he indeed believed: that the phonetic shape of words is indicative of meaning associated with emotions and/or aesthetics. The paper presents convincing phonetic analyses, backed up by statistical data, finally leading to the well-evidenced conclusion that phonetic properties of Qenya and Sindarin on the one hand and of Black Speech on the other iconically reflect the nature of their respective speakers: positive emotions as opposed to violence and aggression. Part III, “Iconicity and the visual” includes four papers, each presenting visual iconicity in a different context – from poetry, through film to ‘liberature’, books in which the content and the container form “an organic whole”. Peter Gahl opens this section with his study of Apollinaire’s calligram The Tie and the Watch, which has traditionally been regarded as a relatively straight-forward example of visual poetry. Linking the poet’s preoccupation with verbal and visual form, Gahl argues that Apollinaire’s text is in fact far more complex than the carmen figuratum or pattern poem it appears to be, stating that the poet was not only well aware of the differences between verbal and visual modes of signification but in fact thematizes precisely the problematic relationship between a visual image and a verbal assertion. In fact, as Gahl shows, the poem proves that the author utlilizes the specific characteristics of iconic and symbolic signs and their different modalities to feature the simultaneous presence of two different aspects of human life and two mutually contradictory conceptions of time. John White’s truly innovative study of iconic minimalism in modern avantgarde poetry focuses on forms of iconic minimalism that either involve entire works or various local, often diagrammatic, iconic effects in poems that may not be otherwise mimetic. As he points out, there is a tendency to see iconic effects in experimental poetry against the backdrop of the traditionally shaped or picturepoem, which privileges visual aspects on a macroscopic level rather than consider local iconic elements which are just as meaningful. Iconic minimalism, through its complex mode of semiosis, is an example of how a “reduction of means … leads to textual diversification and enhanced complexity” (White, this volume), which he goes on to illustrate in poems both involving form miming meaning and form miming form. As White convincingly shows, the move towards abstraction and generic referentiality in both traditions of iconicity has lead to increasing intellectualism (contrary to traditional sense-oriented shaped poetry) and to a greater interest in diagrammatic forms of iconicity in the avant-garde period from Dada to Oulipo. An entirely different approach to visual poetry is taken in the next article by

Introduction

Etienne Terblanche and Michael Webster, who approach E. E Cummings’ from an ecological perspective. They define the dynamic complexity in Cummings’ poetry as a dynamism of signs that reflect exemplarily the dynamism of the ecosystem. Analyzing his picture-poems in the sense of ‘form miming meaning’, they argue that formal elements like syntax, word division, visual placement on the page, the use of white space, and what they call a “transformational semantics” all work together to “mime” the dynamic processes of the ecosystem. As a result, the poems iconically enact the ecological, interdependent and co-existent mergence of human and natural worlds. The complexity of the filmic image is the subject of Strother Purdy’s essay which deals with one of the new genres introduced in this volume. Purdy looks at the power and significance of film by analyzing the shot as an icon by approaching film from the iconic aspect of photography in its being similar to the object it represents. Film is then a sequence of photographs, in other words, icons containing other icons linked together to give an illusion of motion, which act on the viewer in combination functioning as a “matrix of icons”. As he argues, form generates meaning as well as effect either by converting icons into symbols by convention or by diagrammatically constituting a greater unit with the support of other icons. Since films refer to other films and most great filmmakers learn from their predecessors, these filmic icons are intrinsically self-referential, which Purdy goes on to demonstrate with the examples of Charlie Chaplin, Sergei Eisentstein and Orson Welles. Another novelty in this volume is Katarzyna Bazarnik’s exploration of ‘liberature’, a term coined in 1999 for a work in which the text, image and physical space of a book are integrated. Liberature is, as she points out, “total art”, or “total literature”, in which the artist assumes control over both the visual design and the content of a work. It is different from the genre ‘artist’s book’, which most often refers to that by a visual artist and which does not have to be read, whereas, in a “liberary” work, it is the form of the book that is dependent on the text, which makes its form and meaning entirely interdependent. Oka-leczenie, the first “liberatic” book by Zenon Fajfer and Bazarnik, does precisely that. It consists of three codices jointed by the covers to form a tripartite concertina. The first and the last take place in hospitals in Poland and England, where an old man is dying (Poland) while a baby is being born in a maternity ward (London), a palindromic event that is verbally and visually sign-posted as an anagrammatic transformation of life and death in this highly aesthetic and intricately conceived work. The text is here seen as a material entity occupying “meaningful space”, a tradition which not only goes back to Blake, Sterne, Mallarmé, and B.S. Johnson but which also is highly iconic, which Bazarnik illustrates. In an analytical tour de force she goes on to examine James Joyce’s “geodetic intention” in Finnegan’s Wake, which inte-

9

10

Christina Ljungberg and Elžbieta Tabakowska

grates the text with the physical space of the volume, which, from a Joycean point of view, is shown to perform an “iconic re-creation of the Earth through word”. The authors of the three papers presented in part IV, “Iconicity and conceptualization”, deal with iconicity as a cognitive mechanism that influences concept formation. The first contribution, by Meike Adam, Wiebke Iversen, Erin Wilkinson and Jill P. Morford, brings us back to the topic of signed languages. This time the authors deal with it from the contrastive perspective, claiming that studies of different signed languages make it possible to establish the proportion between their modality-conditioned and language specific properties. Challenging the claim that iconic signs are self-evident, i.e. that their form is predictable from their meaning and their meaning is predictable from their form, they analyze responses of their German subjects, deaf signers of German Sign Language (DGS), who were asked to judge the iconicity in two sign languages: DGS and ASL (American Sign Language). The experiment clearly shows that iconicity of signs is not self-evident, but depends on language-specific experience of the sign users. Moreover, it is also related to individual experience: for instance, a person who has no experience with computers will not be able to perceive the iconicity of the DGS sign for “office”, which is an imitation of a person writing on a computer keyboard. Drawing a general conclusion, the authors claim that iconicity is not an ontological property of a sign, or, as they say, “there is no ‘shortcut’ between the sign and the referent”. Friedrich Ungerer takes a wider perspective, investigating strategies of text construction and looking for exponents of iconicity in overall text structure. He analyses four basic complex action patterns or strategies that are used to shape experience in a holistic way: the path strategy (a chain of consecutive stages of an event or action), the sorting & weighting strategy (categorization and establishing relative prominence of consecutive parts of a complex whole), which he then contrasts with the fourth type, i.e. the convoluted kaleidoscope patterns, perceived as gestalt experiences. Ungerer makes the claim that it is from those cognitive patternings that text strategies are derived, following the principles of iconicity, which he discusses in relation to the four strategies. His point is that, although all principles of iconicity can be found as constitutive elements of particular strategies, they are not adequate as a tool to distinguish between them. The resulting postulate that iconicity on the level of texts should be treated holistically is corroborated by Ungerer’s analysis of several prototypical genres, exemplifying each strategy, as well as category extensions (strategy blends) and ranging from dictionary through travelogs to lyrical poetry and episode films. Christina Ljungberg’s paper focuses on literary narrative, examining one particular structure, i.e. the palindrome, and its semantic function in a literary text, brought about by associative iconicity. In her discussion, abounding with

Introduction

illustrative examples taken from a selection of contemporary novels, palindromes emerge as going far beyond mere wordplay or simple manifestations of verbal skillfulness. In contemporary prose they are often vital for the overall structure of the text. She shows how these symmetrical structures in fact become multidirectional, multilayered and polytemporal, providing iconic mappings for the notion of notorious unstability of human lives, where the evershifting present oscillates – or vacillates – between the past (or reading the palindrome “backward”) and the future (or reading the palindrome “foreward”). Triggered off by what is usually the narrator’s crisis, in contemporary fiction – and in postmodern novels in particular – such palindromic vacillation becomes an iconic representation of temporal shifting, with an unstable text structuring (thus allowing for multidirectional reading) reflecting the unstable present. This is most convincingly demonstrated in Ljungberg’s detailed analysis of Arundhati Roy’s prizewinning novel The God of Small Things, which is a perfect illustration of the claim that in modern fiction palindromic structures often function as icons of temporal transformation. Part V, “Iconicity and structure”, contains three papers: by Dylan Glynn, by Angelika Bergien and by a team of Belgian scholars: Ludovic de Cuypere, Johan van der Auwera and Klaas Willems. In his fairly technical paper Dylan Glynn chooses the theoretical framework of Cognitive Linguistics to discuss what is called ‘experiential motivation’ or ‘isomorphic iconicity’. To check the validity of the cognitivist postulate whereby principles of iconicity can be used to explain the structure of grammatical categories, he examines the grammatical lexeme-class variation of the concept ‘rain-snow’ in English, Dutch and German. The results of the test demonstrate that iconic motivation behind grammatical class variation interacts with its synchronic arbitrariness. Having discussed the relevant tenets of Cognitive Linguistics, Glenn moves on to his case study, investigating the occurrence, the frequency and the distribution in the three languages of lexemes corresponding to the ‘rain-snow’ concept. The analysis of the data finally leads him to the conclusion that, in all the three languages under scrutiny, certain anomalies occur that seem to question the position of iconicity as the sole factor providing explanation for lexical structure. His plea to acknowledge the existence of both motivation and arbitrariness as principles underlying language structure is a timely warning for linguists of the cognitive persuasion to avoid traps of overgeneralization that await those fascinated with the explanatory power of iconic mechanisms in language development and use. The second paper in this section, by Angelica Bergien, shows some of the workings of iconicity in argumentative discourse. The pragmatic orientation of her argument is established by the very selection of her data: she analyses disclaimers, or constructions used in discourse for the purpose of “strategic self-presentation” and contextually triggered “identity management”. Following

11

12

Christina Ljungberg and Elžbieta Tabakowska

Tabakowska (2003), Bergien claims that iconic devices are language-specific, and the degree of iconicity expressed by linguistic means is to a large extent contextspecific. Thus, what is at stake is a functional explanation, and therefore context becomes indispensable. While stating that disclaimers are largely formulaic in nature and thus belong among symbolic rather than distinctly iconic structures, Bergien nevertheless finds two reasons for their iconic motivation : the presence of indexical reference to the speaker on the one hand and the iconic ordering of syntactic constituents (clauses) on the other. Her case study presents an analysis of disclaimers in written political argumentative discourse, which, as a genre, is not typical for this kind of structure. The claim is that when the disclaimers do occur in such a genre, their iconic force is relatively strong. The discussion based on thirty statements by British and American politicians, all opponents to the war in Iraq, shows the nature and the distribution of disclaimers, pointing to their pragmatic function. It substantiates Bergien’s final conclusion that the iconic perspective enables the pragmatists to gain new insights into phenomena that had been traditionally classified as discourse markers or politeness formulas. In the last paper in this section, de Cuypere, van der Auwera and Willems look at iconic aspects of constructions with double sentential negation (i.e. multiple negation with two negators) in a large array of c. 150 languages of the world. The authors focus on two types of double negation: sentence negation consisting of two negative markers (as in French) and the combination of a sentence negator and a negative indefinite pronoun or adverb (as in some English dialects. A brief typological introduction is followed by a presentation of the diachronic language change known as “Jespersen’s Cycle” and resulting in the emergence of a discontinuous double negation of the type je + ne + dis + pas, where the addition of the second negator is seen as an iconically motivated “strengthener” (emphatic asymmetry). Having analysed three types of motivation for such reinforcement of negation (formal motivation, pragmatic motivation and non-factuality), the authors discuss other sources of the second negator (language contact and reanalysis) and conclude that iconic motivation can be indeed postulated for asymmetric negation (i.e. involving a negator and an additional element, e.g. the auxiliary do in English), with markers other than the traditional reinforcing elements possibly also becoming reinforced through iconicity, defined as extralinguistic analogy (e.g. the fact that negation is prototypically tied to irrealis). Part VI consists of one contribution only but one that is not only a new departure but encompasses iconicity in multiple forms, visual, gestural, verbal and acoustic. The complex combination of these various types of signs in multimedia performance makes an analysis of the semiotic processes at work a challenge, which Sibylle Moser’s fascinating exploration of Laurie Anderson’s White Lily convincingly meets. Moser investigates the role of iconicity in the intermodal integration

Introduction

of verbal forms (lyrics) with nonverbal semiotic systems such as music, dance, cartoons, and their media such as voice, body movement and gestures, computer animation etc. Arguing that intermedial iconicity integrates both different levels of linguistic representation as well as verbal and audiovisual means of communication, Moser discusses Anderson’s multimedia performance from the perspective of cognitive semiotics and ‘embodied cognition’: cognitively, iconic, indexical and symbolic forms of representation constantly interact to form a continuum, along which Anderson’s multimedia performance simultaneously represents different semiotic models of the experience of time. As Moser succeeds in showing, White Lily integrates not only the sensory iconic understanding of the flow of time through sound and rhythm with linguistic metaphors and abstract reflection on time, but Anderson also has it unfold intertextually around the symbol of the white lily. To recapitulate, we hope that the reader finds in the present volume what we intended to offer: a multidimensional selection of insistent images – food for thought and imagination.

Notes 1. A similar distinction has been made by Johansen (1993:98, 2002:179), who distinguishes between first- and second-degree iconicity, although Nöth’s concept of endophoric iconicity is more extensive than that of Johansen’s second-degree iconicity. 2. In Peircean semiotics, however, music is the prototype for the iconic sign as it presents a quality. Whereas our associations are culturally coded, the feeling that music evokes is not and does not have to be. We can listen to Japanese music without any knowledge of the complex cultural patterns involved and yet have a wonderful musical experience.

References Changizi, A., Zhang, Q., Ye, H. and Shimojo, S. 2006. The structure of letters and symbols through human history are selected to match those found in objects in natural scenes. The American Naturalist 167(5): E117–139. Fischer, O., and Nänny, M. (eds). 2001. The Motivated Sign. Amsterdam: John Benjamins. Fónagy, I. 1999. Why iconicity? In Form Miming Meaning, M. Nänny and O. Fischer (eds), 3–35. Amsterdam: John Benjamins. Johansen, J.D. 1996. Dialogic Semiosis: An essay on signs and meaning. Bloomington IN: Indiana University Press. Johansen, J.D. 2002. Literary Discourse: A semiotic pragmatic approach to literature. Toronto: Toronto University Press. Maeder, C., Herlofsky, W. and Fischer, O. (eds). 2005. Outside-In – Inside-Out. Amsterdam: John Benjamins.

13

14

Christina Ljungberg and Elžbieta Tabakowska Müller, W. and Fischer, O. (eds). 2003. From Sign to Signing. Amsterdam: John Benjamins. Nänny, M. and Fischer, O. (eds). 1999. Form Miming Meaning. Amsterdam: John Benjamins. Nöth, W. 2001. Semiotic foundations of iconicity. In The Motivated Sign, M. Nänny and O. Fischer (eds), 17–28. Amsterdam: John Benjamins. Peirce, C.S. 1931–1958. Collected Papers of Charles Sanders Peirce. C. Hartshorne, P. Weiss and A.W. Burks (eds). Cambridge MA: Harvard University Press. Rushdie, S. 1982. Midnight’s Children. London: Picador. Tabakowska, E. 2003. Iconicity and literary translation. In From Sign to Signing, G. Müller and O. Fischer (eds), 361–376. Amsterdam: John Benjamins.

Putting grammaticalization to the iconicity test

part i Iconicity and grammaticalization

15

16

Paul Bouissac

Putting grammaticalization to the iconicity test

Putting grammaticalization to the iconicity test A cognitivist perspective Paul Bouissac University of Toronto, Victoria College

“Drop a word in the ocean of meaning and concentric ripples form”. Robert Bringhurst (2004:1) This chapter questions the premises of the theory of grammaticalization which claims that “abstract” grammatical morphemes derive from “concrete” lexical items through a process of phonetic and semantic attrition. This theory generally assumes that language is grounded in iconicity. Thus, questioning the latter also puts the former to a test. The first section presents arguments against the notion of lexical “concreteness”. The second raises the issue of whether the “abstractness” of grammar is an artifact of pedagogical discourse or truly reflects the nature of grammatical relations. The third part proposes to frame the problems that gave rise to the notion of grammaticalization in an utterance-based perspective inspired by the cognitivist approach to language. The conclusion attempts to explain why contemporary linguistics has taken the form of a mosaic of theories that are often difficult to reconcile.

1.

Introduction

The topic of grammaticalization, or grammaticization as some prefer to call it (e.g., Bybee et al. 1994), occupies a strategic place in the theory of linguistic iconicity, one of the most provocative and consistent paradigms that form contemporary research on natural languages (e.g., Givón 2002, Fischer 2004). Squarely rooted in historical and comparative linguistics, the notion of grammaticalization is based on the premises provided by the principles of linguistic iconicity which assign primacy to the lexicon, and contend that the emergence of words have been fundamentally constrained by at least some of the perceived properties of the objects or processes to which they refer. Similarly, syntactic structures are considered to have developed from word sequences that mimicked analogically the configuration of the scenes or development of actions they were meant to linguistically represent.

17

18

Paul Bouissac This evolution is often conceptualized as a ritualization through which immediate iconic values are lost but not entirely forgotten (e.g., Givón 2002:148–149, 158). Although research on grammaticalization forms a relatively autonomous subset of the diachronic study of languages, most of its proponents also subscribe to the basic tenets of linguistic iconicity with various degrees of qualification and interpretation (eg. Campbell 2001, Fischer et al. 2004). Whatever force may be construed as a cause for the grammaticalization process, let it be frequency or economy, this force is conceived as applying to an originally iconic lexicon. This primary iconicity is at least implied in all research using the notions of concreteness and abstractness. This is clearly evident from Givón’s latest summary of the view of human language as an evolutionary process, in which he refers to all the main researchers in this domain of inquiry in relation to grammaticalization (Givón 2002:152). The common denominator of these various approaches is that “abstract” terms are conceived as having been derived from “concrete” words. It is also claimed that syntax developed over time from paratactic constructions under the influence of diagrammatic iconicity (e.g., Haiman 1985, Givón 1995), although this claim is far from substantiated (cf. Fischer in press, chapter 5). According to the theory of grammaticalization, this process led some lexical items to progressively lose their “concrete” referential values and be used as purely relational linguistic devices, thus becoming “grammatical”, this latter stage being considered as the extreme point on the scale of abstraction following a unidirectional logic of phonetic and semantic attrition often rendered by the chromatic metaphor of “bleaching” and the chemical or culinary notion of “reduction”. Admittedly, such a simplified account may not do justice to the sophistication and complexity of the specialized discourse on grammaticalization, notably when it is confined to a limited period of time and to written evidence in the context of historical linguistics, but it nevertheless accurately sketches the broad lines of the epistemological perspective within which research on grammaticalization makes sense since it assumes, often tacitly, the validity of the iconicity theory although other factors such as economy and frequency are also invoked as fundamental factors. This paper attempts to heuristically question some of the premises of such research by putting, so to speak, grammaticalization to the iconicity test.

2. How “concrete” can a word be? Current usage among proponents of iconicity theory considers some lexical items to be more “concrete” than others, and those items are the ones that are the most likely to exhibit iconic features, such as sharing some common properties with their referents, replicating through their relative position in a sentence the

Putting grammaticalization to the iconicity test

mutual relations of their referents in reality, or showing common properties within a particular semantic field. There is no shortage of examples that appear to illustrate this phenomena. The question, however, is whether these examples can be explained by a law of nature or are cultural artifacts (Bouissac 2005). In other words: Are obvious or suggestive cases of iconicity windows on the very origin of language, or the results of more or less deliberate attempts at using arbitrary language sounds and the flexibility of sui generis sentence structures to model aspects of experience? The former hypothesis, which could be called “the mirror theory” evokes a universal process and must provide explanations for the fact that not all words are iconic. The latter, which considers mimesis as a form of poiesis, must account for the phonological and syntactic plasticity that it presupposes. The notion of linguistic iconicity must indeed explain why there are numerous lexical items that cannot be precisely related to individual objects or processes but indicate abstract grammatical relations of coordination or subordination. Of course, these grammatical tools are only relatively abstract since at least some of them clearly evoke ancient lexical items which were once purely referential rather than syntactical. It is tempting to account for this lesser “concreteness” by invoking a natural wearing out through usage, along the line of the Zipf Law (1932, 1935), in the same manner as artifacts progressively lose their sharpness or specific functionality to become recycled for more general use. Hence, the idea that all lexical items wear out with time. However, why some do and some others do not comply with such a “law” is not clear. And why some words both keep their referential value and lose it to become grammatical tools at the same time (e.g., French pas [step] and pas [negation]), possibly even in the same sentence, requires explanation. This phenomenon is usually described as “divergence” following a process of “layering” but describing a problem in metalinguistic terms does not necessarily constitute an explanation. The very notion of lexical “concreteness” is problematic. A word is a concrete object only in as much it is actualized phonetically or in the form of visual or tactile symbols although, as Fonàgy has shown, the tactile or acoustic qualities of some words may happen somewhat to reflect some perceived properties of the objects to which they refer. But such phenomena can be considered to pertain to the connotations rather than the denotations of the lexical items concerned. From this point of view “should” is not less concrete than “stone”. All words refer to some conceptual categorizations which could be said to be concrete only in as much as we could identify the particular patterns of neuron firings that coincide with their utterance or understanding. But, as a conceptual actualization, the meaning of a word can only be a set of abstract relations whatever its referent may be in the context of an utterance. The work “stone” for instance is so undetermined that if one asks someone to bring a stone without providing any cue concerning its

19

20

Paul Bouissac intended use, the range of objects that would qualify as “stones” between those that would qualify as “pebbles” and those that would qualify as “rock” or “boulders” is indeed very large and diverse. The context of the request might offer some clue as to what kind of stone is required, and the request may specify supplementary qualities such as large or small, round or rectangular, smooth or rough. This is so much true that if a particular stone is needed either a whole text must be attached to the request or a gesture must indicate unambiguously the location of the object or model its shape through a gesture. The concept of “stone” can only be a position in a network of classificatory oppositions relating to physical properties, technical practices, or metaphorical uses. The notion of “radial category”, which may arise through metaphorical and metonymic change, does not really help because there is no obvious reason for assuming that a request necessarily refers to a prototype and because prototypes vary with situations. The prototypical stone of a jeweler is different from the prototypical stones of masons, geologists, or Stone Age hunters. The conundrum of categorization has been amply debated in cognitive linguistics (e.g., Lee 2001). For instance, a questionable use of “concreteness” as a lexical classificatory term is provided by an eminent proponent of the iconicity-inspired grammaticalization theory, Bernd Heine, concerning the directionality of conceptual transfer. Heine (1997:7–8) contrasts the two following sentences: (1) a. They keep the money. b. They keep complaining.

He claims that these sentences show that the verb “to keep” has two meanings in English and that (1a) is concrete or less abstract and (1b) is more abstract to the point of being an auxiliary (grammaticalized). One may wonder on which ground such a distinction can be made. Heine invokes the fact that (1a) is compatible with complements that are “visible and tangible” whereas “complaining” is more abstract. However, the property of being audible such as the process of “complaining” can be assumed to be (either in the form of vocal expression of pain or in the form of articulate language) is not less concrete according to Heine’s criteria than a referent that is “visible and tangible”. Furthermore, complaining can be made in writing or through gesturing. If “keeping the money” were “concrete”, as Heine claims, it could be expected that sentence (1a) could not be ambiguous. However, the following examples show that this is not the case, since the sentence by itself could admit the following sentential contexts in which “to keep” has markedly different meanings: (2) a. Sweepers keep the money they find on the floor. b. Modern farmers keep the money at the bank

Putting grammaticalization to the iconicity test

c. Security guards keep the money (from being stolen). d. A miser keeps the money rather than spending it.

In language use, no verb appears without a context and the context actualizes some semantic properties of the verbs which necessarily are relatively abstract since they can combine with other sets of properties represented in other lexical units to produce complex meanings appropriate to the contextual situation. A verb such as “to keep” is part of a cognitive configuration of values with respect to the possible relations of an object to an agent. If “to keep” is understood as located at one of the poles of a cognitive structure concerned with the movement of objects, which includes for instance “to loose”, “to give back” and “to spend” at the opposite pole, it is possible to analyze this cognitive configuration as follows: Given an agent and a moveable object (or another agent) that is under its control, this object (or other agent) can be let go or not let go, that is, change position in space, and this change can be effected voluntarily or not voluntarily. In its most abstract meaning within this configuration, the verb “to keep” implements the negation of change of location or state since “to change” is also a pole in a cognitive configuration of values which are mutually definable such as, for instance, with respect to “to stay at the same place” or “to persist in a particular state”. It is not necessary to assume a change of semantic status or grammatical category supposedly brought about by the semantic “wearing out ” of a “concrete verb” to explain the meaning of “to keep” in the expression: “They keep complaining” because it is exactly the same abstract or formal values that are actualized as in the sentence “They keep the money”. Other expressions such as “keeping calm”, “keeping one’s cool”, “keeping one’s membership in good standing” and “keeping one’s hat on his head” variously select, through the choice of complements, whether the absence of change applies to location, process or state. Furthermore, if we consider that all actual linguistic communication act is ultimately motivated by the information it provides, and if information is understood as the resolution of lesser or greater uncertainties, the abstract properties selected by the context in the case of “to keep” must be assessed with respect to a horizon of expectation formed by the opposition between discontinuing and continuing a change of location, process or state. In both examples given by Heine, which cannot possibly be uttered out of context in real life, the verb implies that the opposite could have been anticipated either because it is conform to the ethical norm (for instance, in 1(a) they should give back the money) or because the process should have stopped (for instance, in 1(b), they were given what they asked for). The use of “to keep” in those examples without a specification of the contextual constraints is meaningless and cannot prove anything, certainly

21

22

Paul Bouissac not a “semantic wearing out” that is not even indicated, in this case, by a phonetic reduction. Failing to understand that the cognitive structures underlying the lexicon for a community of users are necessarily abstract in the sense that they are sets of cognitively relevant oppositional properties which allow the mapping of particular instances of communication, leads to assigning degrees of concreteness to individual words. Of course, in any culture, some referents are so specifically representative of their category that their image becomes intimately associated with the lexical unit that designates them, thus creating the illusion of concreteness in spite of the fact that this lexical unit may apply to a large number of individual objects that share only a small subset of the properties constituting the semantic configuration of the unit. The claim that semantic structures articulate complex oppositions (e.g., Greimas 1983, 1987) is not incompatible with a naturalistic view of the origin of language. It is indeed possible to construe these binary semantic structures, following the model of phonological oppositions, not as a priori principles of classification or as the “hopeful monsters” produced by a genetic mutation, but rather as the result of the evolutionary changes which molded human perception and cognition. Stereotyping is adaptive in situations when fast decisions have to be reached in order to survive, even if such a strategy at times fails. Hesitating in the face of uncertainties, for instance whether an object is animated or not animated, leads far more often to failure than acting upon snap decisions. Systematically trying to sort out grey areas would be maladaptive in as much as it would take too much time to reach a decision (Glimcher 2003). Such an evolutionary perspective construes semantic structures as robust survival strategies rather than rigid devices, and allows for some degree of plasticity if only because evolutionary and developmental forces are never at rest. The natural process of categorization can be amplified by cultural constraints and is not ontologically divorced from the latter. Abstraction, hence grammaticalization, is not a secondary phenomenon as iconicity theory contends but can only be an adaptive behavior constrained by natural selection. The capacity of thinking analogically may have been enhanced by the emergence of grammar but fundamentally pertains to the evolution of cognition, an adaptation that must predate the earliest forms of language.

3. How “abstract” can grammar be? Discussions of grammar often overlook the fact that they are conducted in a metalanguage that was developed in the context of literacy and as a result of the specific visual patterning that characterizes writing in its many forms. For

Putting grammaticalization to the iconicity test

instance, it seems that the hypothesis that reductions caused by the principles of frequency and economy may be assigned to strategies devised by scribes rather than by purely oral processes would deserve to be examined. The implications of reflecting upon language in general through the mediation of particular languages that are observed and conceptually manipulated in their written forms have not been sufficiently explored although there have been some pioneering efforts in this direction (e.g., Goody 1987:258–289). It is all too easy to take for granted that a written text is a reliable sample of natural language when it is actually a sample of an artificial system initially devised to visually encode auditory signals or other information such as the mapping of a path with respect to some landscape beacons. The information is obviously better protected by a code that is restricted to a linguistic community, or a smaller group within this community, than if it were represented by graphs and images. Any set of symbols to which some distinctive values are assigned according to some conventions operates as a calculus regulated by its own algorithmic laws. A literacy calculus, let it be logogrammatic, syllabic or alphabetic, is used to approximately map language utterances, but is useless if taken independently of the natural language it translates into its visual symbols. Archaeology provides many examples of the impenetrability of such systems because of the incompleteness of the information they encode. In natural language verbal interactions, the situations within which utterances are embedded afford enough semantic constraints for usually (but not always) allowing the effective transfer of relevant information. It is likely that many grammatical constraints which are observed in literate languages were originally parts of the artificial design aimed at disambiguating the graphic encoding of verbal utterances. Moreover, it should be pointed out that the representative function of language is foregrounded in the written medium whereas speech is prominently indexical and performative. This dissymmetry of functions entails consequences and ignoring them can lead to biased linguistic theories. But, irrespective of the object a calculus may map more or less adequately, it is a system in its own right and its formal properties can generate specific transformations in the spatial distribution and transformation of its elements. For instance, the calculus that makes possible alphabetic writing by spatializing speech considerably expands the working memory within which a single utterance can be structured and displayed, hence allowing for a theoretically infinite embedding of propositions within propositions. But this is a geometrical or topological property rather than a linguistic one which depends on the capacity of working memory instead of visual discrimination. Even though, there is of course a necessary limit to such a conceptual “infinity”, since a linearly written text is dependent on the constraints of human visual perception and memory for its functionality. The archaeology of writing and the history of epigraphy show that various conventions

23

24

Paul Bouissac have been introduced over time. Some, such as punctuation, have been developed to specify the melody line that cues listeners of a verbal message about its spatiotemporal organization: others, such as framing or underlining, are specific to the medium in which the writing calculus operates. The latter can artificially structure a text in a visual dimension that is outside the scope of the possibilities offered by verbal utterances alone. In modern western writing systems, the introduction of spaces between words derives from conventional segmentations – and the ensuing categorization of the segments – in spite of the fact that these gaps do not reflect corresponding pauses in speech but result from a secondary cognitive process of artificial segmentation. In verbal interactions, sudden stops and silences occur, but not at the junctures which are created by the metalinguistic knowledge which at some point in time interfered with the initial calculus that made writing possible. From this perspective, it should be clear that grammar conceived as a system of abstract rules derives from the necessity of teaching literacy in societies (or particular groups within societies) and is essentially conventional and normative.1 So-called descriptive linguistics records those rules in the metalanguage of literacy. Linguists who have undertaken to test the “grammaticality” of a variety of possible sentences in any particular language, have tested the level of literacy knowledge of their informants rather than whether or not the meaning of these sentences was understandable in the language in which they were formed. Grammaticality in such a context refers to the conformity with a particular convention (normative grammar) standardized by a social group at a particular point of its historical development. Most of the sentences adorned with the ungrammatical symbol (*) in contemporary linguistic discourse make perfect sense in spite of being unconventional. It is indeed possible to contend that, given a sufficient cognitive development and a basic convention of word order, let it be SOV or SVO, most so-called “ungrammatical” sentences are as efficient as “grammatical” ones for conveying relevant information as long as the verbal interaction takes place in context. Furthermore, most of the sentences artificially constructed such as those that purport to demonstrate the infinite potential of the recursive quality of grammar, strictly result from the combinatory potential offered by the literacy calculus and its arbitrary segmentation and categorization as they are displayed in pedagogical or theoretical texts. The reason for evoking Universal Grammar theory at this point of the argument is that it imposed the idea, beyond the restricted circle of Chomsky’s followers, that grammatical rules are highly abstract and transcendent. The proponents of linguistic iconicity, while refusing to accept the innateness and transcendence of these rules, have nevertheless tended to agree that grammar is abstract and have undertaken to explain this assumed abstractness as a degraded form of concreteness or iconicity. Regarding segmentation and categorization, it is interesting to note that

Putting grammaticalization to the iconicity test

developmental psycholinguists test pre-literate young children from the point of view of their own knowledge of the written language and its conventional graphic representations. Word categorization, notably the distinction between “content words” and “function words”, belongs to the metalinguistic discourse and carries a load of assumptions, among which the most powerful (and perhaps the most invisible) one is that there are such entities as “words” endowed with identities and specialties. As it has been suggested above, language is perceived through the lenses of the written text and centuries of legislating the graphic encoding, segmentation and normalization of verbal flows. One may wonder on which ground, in a phrase such as “going to town”, “to”, which never occurs by itself in discourse, can be considered as an entity of a totally distinct nature with respect to words like “town” or “going” for example, except in the pedagogical discourse that reflects its conventional graphic representation. If, however, for the purpose of analysis, “to” is extracted and if its contribution to the meaning of the phrase is analyzed, it does not seem that it is more abstract or less concrete than “town” or “going” which undoubtedly would be considered to be “concrete” or “content” words according to the proponents of grammaticalization theory. Let us suppose that it makes sense to scrutinize the semantics of “to”, that is, the kind of relevant information it provides to the compounds in which it occurs. Its meaning implies a fundamental distinction between two positions which are bridged by a vector. The context specifies whether these positions are locations, agencies or states of affairs (e.g., going to town, presenting something to someone, going from bad to worse), and, as such, refer to concrete situations. It is only in the metalanguage that “to” can be virtually construed as not referring to any particular situation. Its status is not markedly different from a verb like “going” from a cognitive point of view. The verb “going” implies indeed the same abstract cognitive structure as “to” and can actualize a great diversity of concrete meanings depending on the nature of the disjoined positions (places, times or states of affair) specified by the context of the utterance in which it occurs. When the two, “going” and “to”, are contrasted as “concrete” versus “abstract”, the former is conceived as being perceived in an actual utterance, and the latter is conceived as it is represented in the context of a grammatical list rather than in its conventional use as indicator of the infinitive form of a verb like in “to go”. The fallacy of the concreteness of the lexicon has been addressed in the previous section of this paper. Let us simply note that “town” may evoke a very concrete experience in each urbanite (and functions as a label for this compounded experience) but its meaning fundamentally depends on a configuration of binary oppositions bearing upon the semantic classification of various degrees of density of human settlement and saturation of space as well as relative mutual positions with respect to

25

26

Paul Bouissac center and periphery. But in a given utterance it may refer “concretely” to any city whatsoever by specifying one of the polar positions “abstractly” (or “indexically”) represented by “going to” that could also be written gointo. Actually, “goint’town” or, even, “gointown” [going town] would convey effectively the intended meaning, although in a way that would disclose a lack of adherence to current normative grammatical forms, or even, possibly, evidence of illiteracy. The point made here is that “going” and “to” are somewhat conceptually redundant because they are mutually inclusive. They can be semantically represented by the same structure that consists of positing a disjunction and its virtual resolution. They are cognitively equivalent, equally abstract in themselves (or rather in relation to the other values defining the cognitive system implemented by the English language) and equally concrete in context. To which extent other languages display similar relations is a question that cannot be fully addressed here. However, a brief example taken from the Finno-Ugric linguistic family will provide a brief vista into the possible generality of this phenomenon. Laszlo Fejes (2004) considers compounds verbs in Komi-Zyrian, notably those verbs ending in munni [to go] with a noun, an adjective or an adverb as first component. For instance, nirmunni [come to life, revive, be reborn] or nummunni [start smiling] in which the first constituents respectively mean “sprout, germ” and “smile”. It seems obvious that Komi-Zyrian nummi has the same conceptual properties as English “to go” since the information it conveys in the two compounds is the transition from one state of affair to another, from no-life to life, from no-smile to smile (Fejes 2004:8–9).

4. The notion of grammaticalization path Let us now turn to the morphological and semantic process that the proponents of grammaticalization hypothesize in order to account for the assumed change of status from “content words” to “function words”. The whole theory hangs indeed from the assumption that some concrete words progressively loose their concreteness through a “natural” phenomenon they call semantic “bleaching”, which allows those words to be used as mere syntactic tools. Whether this results from pragmatic inferencing (Hopper and Traugott 2003:67–68) or from “the role played by the overall system of grammar (possessed by the speaker) in which the process takes place” (Fischer in press), the assumption of a progressive transformation from specificity to generality remains the key metaphor through which this phenomenon is modeled. The notion of grammaticalization path must assume that there has been a starting point for the process it claims to describe in each case. This starting point

Putting grammaticalization to the iconicity test

must be arbitrarily chosen, since too little is known about the origin of language for confidently assigning an absolute beginning to any element of the lexicon. Any past written form of an English word and its meaning(s), for instance, can hardly be considered as a starting point. Historical linguistics provides a sense of the infinite regressus toward hypothetical anterior states of any known language. Moreover, the notion of grammaticalization path presupposes that any starting point must be “concrete”, that is, “primitive”, in compliance with a certain representation of human cognitive evolution which is projected upon the axis of language evolution. The model of language evolution that is invoked in this conceptualization is somewhat reminiscent of August Schleicher’s theory of the organic nature of languages, understood as organisms which are submitted to the laws of biological evolution (Schleicher 1863). Perceiving language changes through such a metaphorical filter raises a first conceptual problem: it construes language as an evolving entity and values abstraction as an evolutionary “progress” rather than an adaptive selection with respect to particular environmental or contextual constraints. This metaphorical model conflates biological, cultural and linguistic changes along a virtual scale that is demonstrably fallacious. In addition, if abstract thinking is considered a cognitive achievement supported by complex grammars, this leads to construing as stable entities (states of arrival of the grammaticalization process) the grammatical structures that are transferred to each new generation whereas these structures are themselves transient and only appear stable if they are considered within a relatively short period of historical time to which institutional literacy lends an illusion of permanence (Deutscher 2005). It is more than likely that the barbarisms and solecisms contemporary grammarians denounce in the languages of the Internet’s blogs will be the norms of tomorrow. Moreover, invoking the original meaning of any element of the lexicon forces one to inject dubious etymological knowledge into the semantic description that forms the basis for the theory of grammaticization path. Such speculations imply a semantic theory that is never made explicit and in which original reference to a concrete object is taken for granted. Bybee et al. (1994:10–11) offers telling examples of this approach. In order to account for the grammaticization of the term meaning “face” that may be used in some languages to refer to the “abstract” relation “in front of”, she indicates that a step on this path may have been the metaphorical use of “face” to designate the anterior surface of any object such as a cliff, a house, etc., thus assuming that the meaning of “face” as the part of the human head where the eyes, nose and mouth are located is the point of origin of this path. But at which point in the history of the English language, or any language for this matter, can we securely stop the evolving process in reverse? The case of “face” is particularly revealing. There is nothing in the Latin etymology of facies that indicates a primal connection with the face in the modern sense. The

27

28

Paul Bouissac word is related to facere, whose ancient meaning is “to place”, “to posit”, then “to make”, and the nominal derivation facies meant “form”, “aspect”, “way of doing something”, much later coming to designate the appearance of a building (façade) and the anterior part of the head (Ernout and Meillet 1967:209–213). The popular Latin facia eventually became “face”, attested in English with its current value in 1290 (Murray 1901).There are drastic limitations to the validity of the description of any natural language which relies on historical arguments, such as the proponents of grammaticalization do, by arbitrarily construing a given state as the starting point of a process. The theory of linear grammaticization path might have some validity if unidirectional successive changes in the degree of generality of a word could be observed within the limits of a well documented historical period. In fact, the examples of phonetic and semantic reductions which are usually adduced to drive the grammaticalization point home overlook a well-known feature of phonetic and semantic changes: such changes are systemic and apply to the broad phonological and semantic spectrum of a language, not merely to isolated individual sequences of morphological lineages, although there may be a combination of such changes. Supposedly grammatical words undergo the same phonetic reductions as other elements of the lexicon under the same blind forces with various paths leading to similar results such as, for instance eau (water), aux (garlic, [pl.]) and au (to the [masc.]) in modern French [o]. Since all languages are demonstratively in flux under the pressure of a great variety of constraints (societal, historical, psychological, articulatory, memetic, etc.), assigning any form as a starting point can only be opportunistically arbitrary (see Deutscher 2005 for many examples), and assigning any form as a point of arrival is obviously dangerously relative. The epistemological agenda of most linguists is to uncover regularities and laws in the observable diversity and variations of languages because they assume that there must be such logical abstract constraints. The assumption of rationality as the organizing principle of language might be the fatal flaw of modern linguistics as it restricts a priori the nature of the investigative tools. The claim to the unidirectionality of the grammaticalization path, as it is obstinately expressed by Haspelmath (2004) for instance, bears witness to such a determination to spell out linguistic laws. What is not clear is whether these linguistic laws are conceived as mere statistical regularities constrained by a range of unknown factors or whether they are considered to be compelling necessities that transcend the objects they govern. Do they depend on the vantage point of the observers, who can obviously apprehend languages only partially and through relatively short temporal windows, and daringly extrapolate from such random data? Or is it claimed that these laws belong to the very essence of language, assuming that there is such

Putting grammaticalization to the iconicity test

a grand ontology? Or are languages chaotic phenomena whose complexity is not amenable to any commonsensical logical frame, and which literate societies attempt to domesticate by consensually or authoritatively regulating their lexicon (through dictionaries) and use (through grammars). Social cohesion and control are indeed constantly threatened by the proliferation of changing linguistic forms and functions.

5. Iconicity and grammaticalization in utterance-based linguistics The critical approach that has been outlined above does not exclude the possibility of an evolutionary continuum between the morphology of utterances and prelinguistic or extra-linguistics phenomena that are of particularly important social and cognitive relevance. Unless an arbitrary ontological distinction is posited a priori between language and the life-world, biological and environmental constraints are necessarily con-substantial with language forms and uses and account for their dynamics. But, based on the assumptions of linguistic iconicity, the very concept of grammaticalization might be an abusive generalization of a few examples that can be explained otherwise. For instance, a basic grammatical feature such as negation can hardly be explained as the wearing out or reduction of referentially “concrete” lexical antecedents. It does not require either an elaborate logico-grammatical explanation sui generis if it is related to the signaling of prohibitive warning in primates among which dominant individuals utter a nasal stop toward other individuals’ incipient behaviors that are interpreted as potential or actual challenges to their prerogatives (access to food, females, space). Interestingly, these auditory signals can be produced while the mouth is full and engaged in eating. Nasal stops (m or n) with various vocalizations (ne, no, me, ma, etc) indicate in many languages not only that an object is not present, an action is not being performed, or a statement is not true, but also that a request is denied or that a process should be interrupted or simply not initiated at all. It is obviously easier to explain the former (an informative or redundant statement) as a cognitive expansion of the latter (originating in a pre-linguistic vocal warning) than the reverse. In human multimodal interactions, (English) “No!” can be used to stop incipient gestures or to reverse undesirable behaviors. Depending on the intonation, it can be construed as an affirmation of authority or as an acknowledgement of reluctant submission. Normative grammars, taught or described through paradigmatic examples or diagrams, convey a misleading impression of symmetry between affirmation and negation as if the tools of negation were neutral operators (e.g., Jane sees the cat vs. Jane does not see the cat). Metalanguages of pedagogy and linguistics are biased

29

30

Paul Bouissac by Aristotle’s logical elaborations which suggest such symmetry between assertions and their negation from the point of view of their truth-value. In principle, any assertion can be negated at the cost of creating a few paradoxes. However, the study of the pragmatics of negation shows that the frequency of distribution of affirmative and negative clauses in written English is far from being balanced. Statements of fact that carry functional information largely outnumber the negative forms that merely tend to correct false assumptions or expectations, (Givón 1979:ch. 3; 2002:336–343). Denials and prohibitions are far more prominent in social interactive processes. In bickering, the interacting subjects tend to challenge each other by negating the other’s assertions. These latter uses are consistent with the cognitive value of the nasal stop in primate warning signals. The case of the French negative ne..pas has been evoked above (section 1). It has developed from sentences such as: (3) a. Il ne marche pas encore [he does not walk yet]

In which pas [step] is the smallest unit in the process of walking [marcher] and ne is the grammatical tool to signify negation. This form was parallel to other uses in which the smallest unit was congruent with the process such as in the following: b. Il ne boit [to drink] goutte [drop] c. Il ne mange [to eat] mie [crump] d. Il ne fait [to do or to make] rien [thing, from Latin rem]

And so on. Some have been preserved in archaic expressions, but pas and rien have become negative grammatical tools themselves in modern French as shown in the following examples: (4) a. Vous êtes deçu? – Pas du tout. [Are you disappointed? – Not at all.] b. Que voyez-vous? – Rien. [What do you see? – Nothing]

Two observations are in order here. First, there has been no reduction; secondly, the “original” words have kept their form and meaning as shown in the following:2 (5) a. Elle n’a pas fait un pas. [She has not taken a single step.] b. C’est un rien. [It is a small thing]

It is obvious that, linguistically and cognitively, this homonymy is not a source of ambiguity in the context of an utterance. Native speakers know what means what when and where. Note that there is no “reduction” in this case because formal French demands the presence of “ne” in front of the verb. The mere use of “pas” as negation is a popular alternative form that has captured, so to speak, the negative value of “ne” with which it had become collocated within a relatively stable pattern. This case calls attention to the prevalence of utterances and cognitive

Putting grammaticalization to the iconicity test

context upon the theoretical segmentations of the language chain into arbitrary units. Whereas it may be legitimate to heuristically discriminate components that are functional on various levels and assign them to categories in a particular written metalanguage, it is misleading to assume that such an artificial segmentation and categorization is relevant to the cognitive value of an utterance. The case of French pas shows that the function of a cognitive operator such as negation can be potentially performed by any morpheme, even a morpheme totally alien to the iconic source of the root morpheme of negation. What prevails is the cognitive configuration of relations and values in the interactive, dialogic process of successive (or overlapping) utterances through which meaning is constructed rather than selected from pre-established dictionary definitions. Cognitive structures and relations are adaptive strategies aimed at interpreting patterns, assessing situations and anticipating outcomes. These structures and relations, while being necessarily constrained (or bounded) by the environment in which they evolved, must be sufficiently general (or abstract) to fit a great variety of relevant contextual configurations. Utterances are always embedded in situations (real or imagined) and are part of them, and the whole lexicon is by necessity indexical or deictic, since it is obviously finite but successfully operates in an openended flux of situations. The idea that words have set meanings and functions creates theoretical problems such as grammaticalization and metaphor, because it is assumed that words have a definite sense, a kind of semantic substance, even if it is recognized that these meanings are time-dependent and have a history. But if words are understood as indices of abstract cognitive structures that actualize semantic properties determined by the co-textual and contextual situations of the utterances, there is no need to assume a process of “semantic bleaching” or “grammaticalization path” since the abstract values they are claimed thus to acquire through a loss of “concreteness” were already theirs to start with and are the ones which account for all their meanings over time and across the range of their uses. Of course, if a printed lexicon with standardized definitions (notoriously circular) and examples taken from a given historical period (the contemporary period for instance) is taken as reference and is construed as positively representing the target language, some problems will arise about the assignment of lexical functions and values. Some words will appear to be used to designate objects that are squarely beyond the scope of their standard dictionary definitions (metaphors), some will appear to be less “concrete” than others and will be labeled “grammatical tools”, hence raising the artificial issue of how it is possible to move from the former to the latter. But it is well known that whatever degree of abstraction a word may appear to have acquired on the scale of grammaticalization, it is always possible to nominalize it again as poetic, philosophical and metalinguistic discourses amply

31

32

Paul Bouissac demonstrate (e.g., the “hic and nunc”, the “Who’s who”, the “ça” in French psychoanalytical jargon, a “has been”, and so on). This is not to deny that morphemes have discrete and contrastive values that are determined by consensual use within a given society at a certain time. But cognitive structures are not dependent on the lexicon, even if a lexicon that has been formally institutionalized, as it is the case for most modern literate languages, contributes to stabilize or even freeze some semantic constraints. As was recently shown through cross cultural non verbal testing (Dehaene et al. 2006, Holden 2006), the human brain can use geometrical and topological concepts in the absence of a relevant lexicon, a point that Plato had made long time ago in the Ménon. The mastering of complex games like chess by people who are totally deprived of language abilities by birth demonstrate that notions such as move, turn taking, subjective perspective, intentionality, rule, possibility, prohibition, strategy, gain and loss, presence and absence, values, and the like without which playing chess would be inconceivable, do not require a lexicon to be operational. Such cognitive abilities have evolved long before language emerged among the adaptive competencies of the genus Homo and are not fundamentally dependent upon a linguistic system. So-called linguistic universals, mainly when it comes to grammar, seem to be far less absolute and constraining than it is usually claimed. As more and more languages are described (often as a last chance before they become extinct), outside the scope of literacy-based linguistics, it appears that there is much more flexibility than could be anticipated in the ways in which cognition is mapped upon words to the point that the very notion of linguistic universals can be questioned (Evans 1999, Wuethrich 2000). Differences in approaches, such as the opposing views that have been contrasted in this paper, often depend on the temporal dimension of the target object that is construed by the inquiry. Theories always imply selection of a time frame. Linguistic theories do not always make their time selection explicit, except through the coarse distinction between synchrony and diachrony, which is notoriously fuzzy given the relativity of these heuristic concepts. The language object that is arbitrarily considered over a period of a few centuries through its surviving written texts, or a linguistic corpus that brings together literacy and oral data as much of contemporary linguistics does, is necessarily different from language considered from an evolutionary point of view that encompasses both its absence and its presence, and all the rest. What can be said, and make sense, about a corpus that is limited in time and scope, can only be very tentative. This is why the conclusions proposed by the theories of linguistic iconicity and grammaticalization raise more questions than they bring solutions.

Putting grammaticalization to the iconicity test

6. Conclusion The mosaic of epistemological paradigms that characterizes contemporary linguistics probably results from underestimating the complexity of language. Each linguistic school claims to have uncovered a general principle, sometimes expressed in the form of a law or a set of algorithms that account for most if not all the observables. Typically, each school excludes some of these observables as being outside of its object of study and focuses on one range of problems which it claims both to be essential and to be solvable, that is, which it can abstractly describe in a consistent manner through its metalanguage. Each school provides an abundance of examples that have been selected or elicited for the purpose of demonstrating the legitimacy of its claim to universal validity. All linguistic schools feed, so to speak, on the leftovers of other schools. This socio-epistemological dynamics is patently obvious when functional structuralism, generative grammar, cognitive linguistics and linguistic iconicity theory are compared. Nevertheless all these paradigms have the same object, the languages of the world, which keep changing under their nose, and at times disappear, at various rates, impervious of the metalanguages that purport to conceptually control them but seem to be merely able to make ex post facto predictions. Probably, the fatal flaw of all these theoretical approaches is the assumption of rationality or functionality of their object of study and the fact that, willy nilly, they perceive and observe this object through the literacy filter and within a restricted time frame punctuated by arbitrarily selected historical points of departure, often motivated by nationalistic or ideological obsessions. Perhaps there are so many forces at play on so many levels in live languages that all heuristic reductions of their phenomenology are bound to lead the inquiry astray and to drastically limit its validity. The challenge of a science of language might be similar to the integrated understanding of the five forces that remains the grail of contemporary physicists. From this perspective, it is possible to explain why the various linguistic paradigms are incompatible rather than complementary. They have created disciplinary cultures that are mutually exclusive, both from epistemological and sociological points of view. The agenda of this paper could not do more than merely invoke the possibility of stepping out of the linguistic conundrum and looking afresh at languages as objects which may be neither rational nor functional by essence and to consider them within the epistemological frame of evolutionary time rather than through the narrow temporal focus of historical time, that is literacy.

33

34

Paul Bouissac

Notes The final version of this article has greatly benefited from the critical feedback provided by Olga Fischer and Christina Ljungberg, but they should not be held responsible for any error or defective argument that may still be found in the text. 1. This point has, however, been disputed by Olga Fischer, who argues that each child builds up some grid, some grammar system in learning its language. Referring to Dan Slobin’s argument that children begin with fairly concrete collocations and slowly build up toward more abstract linguistic ‘rules’ or ‘schemas’ based on these collocations, provided that they are frequent enough, she argues that if there is no grammar guild-up, why do English children automatically produce ‘the red car’ and not ‘the car red/car the red/red the car’ etc. 2. As Olga Fischer points out, this is true for ‘pas’ but not for what happened in English, where there is both reduction and bleaching, cf. OE ne + finite verb (+nawiht ‘no wight’= creature) > ME ne + verb + noht >eModE verb + not > PDE not + verb, and where the original ne element has completely disappeared, while nawiht has both been bleached and reduced. Loss, layering and divergence of a variant are all possible in a grammaticalization process.

References Bouissac, P. 2005. Iconicity or iconization? Probing the dynamic interface between language and perception. In Outside-In and Inside-Out, C. Maeder, O. Fischer and W. Herlofsky (eds), 15–37. Amsterdam: John Benjamins. Bringhurst, R. 2004. The Solid Form of Language. Kentville: Gaspereau Press. Bybee, J., Perkins, R. and Pagliuca, W.. 1994. The Evolution of Grammar: Tense, aspect, and modality in the languages of the world. Chicago IL: Chicago University Press. Cambell, L. (ed.). 2001 Grammaticalization. A critical assessment. (Special issue of Language Sciences 23:2–3). Dehaene, S., Izard, V., Pica, P. and Spelke, E. 2006. Core knowledge of geometry in an Amazonian indigene group. Science 311:381–384. Deutscher, G. 2005. The Unfolding of Language. New York NY: Holt. Ernout, A. and Meillet, A. (eds). 1967. Dictionaire étymologique de la langue latine (4th ed.). Paris : Klincksieck. Evans, N. 1999. The Kayardild language. In Voices of Queensland, J. Robinson (ed.), 151–190. Melbourne: OUP. Fischer, O. 2004. Evidence for iconicity in language. Logos and Language. Journal of General Linguistics and Language Theory 5(1): 1–19. Fischer, O. In press. Morphosyntactic Change. Functional and formal perspectives [Oxford Surveys in Syntax and Morphology]. Oxford: OUP. Fischer, O, Norde, M. and Perridon, H. (eds). 2004. Up and Down the Cline: The nature of grammaticalization. Amsterdam: John Benjamins. Fejes, L. 2004. Compound verbs in Komi: Grammaticalisation without a grammatical morpheme. Acta Linguistica Hungarica 51(1–2): 5–43. Givón, T. 1979. On Understanding Grammar. New York NY: Academic Press. Givón, T. 1995. Functionalism and Grammar. Amsterdam: John Benjamins.

Putting grammaticalization to the iconicity test

Givón, T. 2002. Bio-linguistics: The Santa Barbara lectures. Amsterdam: John Benjamins Glimcher, P. 2003. Decisions, Uncertainty, and the Brain. Cambridge MA: The MIT Press. Goody, J. 1987. The Interface between the Written and the Oral. Cambridge: CUP. Greimas, A.J. 1983/1966. Structural Semantics (Trans. D. McDowell, R. Schleifer and A. Velie). Lincoln NB: University of Nebraska Press. Greimas, A. J. 1987/1983. On Meaning: Selected writing in semiotic theory (Trans. P. Perron and F. Collins). Minneapolis MN: University of Minnesota Press. Haiman, J. 1985. Natural Syntax: Iconicity and erosion. Cambridge: CUP. Haspelmath, M. 2004. On directionality in language change with particular reference to grammaticalization. In Up and Down the Cline: The nature of grammaticalization, O. Fischer, M. Norde and H. Perridon (eds), 17–44. Amsterdam: John Benjamins. Heine, B. 1997. Cognitive Foundations of Grammar. Oxford: OUP. Holden, C. 2006. Hunter-gatherers grasp geometry. Science 311:317 Hopper, P. and Closs Traugott, E. 2003. Grammaticalization (2nd ed.). Cambridge: CUP. Lee, D. 2001. Cognitive Linguistics: An introduction. Oxford: OUP. Murray, J.A.H. (ed.). 1901. A New English Dictionary on Historical Principles (Vol. IV), by H. Bradley. Oxford: Clarendon Press. Schleicher, A. 1863. Die Darwinische Theorie und die Sprachwissenschaft. Weimar: H. Bolhau. Wuethrich, B. 2000. Learning the world’s languages – before they vanish. Science 288:1156– 1159. Zipf, G.K. 1932. Relative Frequency, Abbreviation, and Semantic Change: Selected studies of the principle of relative frequency in language. Cambridge: Harvard University Press. Zipf, G.K. 1935. The Psycho-biology of language. Boston MA: Houghton Mifflin.

35

36

Paul Bouissac

Iconic thumbs, pinkies and pointers

Iconic thumbs, pinkies and pointers The grammaticalization of animate-entity handshapes in Japan sign language William J. Herlofsky Nagoya Gakuin University

Sign language research has shown that all sign languages have verbs of motion in which the movement expressed by the verb is combined with an animateentity handshape. These handshapes iconically represent certain salient characteristics of their referents, and are therefore referred to as ‘classifiers’. The objective of this paper is to describe and illustrate the nature of iconic handshapes (especially the thumb, pinky, and pointer) of classifier-like constructions in Japan Sign Language (JSL), and show how their functions and patterns of grammaticalization are in some ways similar to and in other ways different from those identified for classifiers in spoken languages.

1. Introduction In all of the sign languages that have been studied so far, there are verbs of motion which have complex morphological structures that have been referred to as ‘classifier constructions’, where the movement expressed by the verb is combined with certain animate-entity handshapes called ‘classifiers’. These handshapes iconically represent certain salient characteristics of their referents, and are said to identify their referents as belonging to certain ‘groups’ or ‘classes’, basing the classification on, for example, physical properties of the referents, such as size and/or shape, how the referents are handled (imitating the shape of the hand as it does the handling), or what class of animate entity the referent may belong to, such as human or vehicular. Although many questions remain to be answered about these classifier constructions, their iconic motivation is uncontroversial, as is clearly stated by Emmorey and Herzig (2003:221) in the most recent volume on classifier constructions: “All sign language researchers agree that classifier constructions are iconic and not arbitrary in form”. It is not necessary, therefore, for this paper to provide evidence for the existence of iconicity in the handshapes of classifier construc-

37

38

William J. Herlofsky

tions in Japan Sign Language (JSL), and so instead the objective of this paper will be to describe and illustrate the nature of the iconic handshapes in classifier-like constructions in JSL, and how their functions and patterns of grammaticalization are generally consistent with the typology proposed for classifiers in other natural languages. To do this, the analysis will focus on the handshape morphemes used in the formation of JSL signs related to animate human entities, some of which are clearly what are referred to as classifiers, and others that may or may not be strictly of that category. The analysis will illustrate how the same thumb, pinky, and pointer finger handshapes can represent human entities in grammatically different ways, sometimes acting like gender markers, sometimes like person markers, and sometimes like what have been referred to as classifier constructions.

2. Iconicity and language Even at the beginning of the twenty-first century, it is not unusual to find less than accurate statements about arbitrariness and iconicity in language, even from very fine linguists. Consider, for example, the following from Steven Pinker (2003:17): As Ferdinand de Saussure pointed out, a word is an arbitrary sign: a connection between a signal and a concept shared by the members of the community. The word duck does not look like a duck, walk like a duck, or quack like a duck, but I can use it to convey the idea of a duck because we all have learned the same connection between the sound and the meaning. I can therefore bring the idea to mind in a listener simply by making that noise. If instead I had to shape the signal to evoke the thought using some perceptible connection between its form and its content, every word would require the inefficient contortions of the game of charades.

Of course, even if there is a connection between form and meaning, this connection does not have to exist in “every word”, nor does it have to be consciously perceived to “evoke the thought” or meaning of the word. That is, iconicity can be a factor in the formation of a word in a language without the necessity of iconicity being a factor in the interpretation of “every word” in a language. And even if the “shape” of “the signal” evokes the meaning “using some perceptible connection between its form and its content”, it is not necessary to produce “the inefficient contortions of the game of charades”. Consider the Japan Sign Language (JSL) sign for ‘telephone’ below:

Iconic thumbs, pinkies and pointers

Figure 1.

Although there is clearly a motivated connection between form and meaning in this sign, and therefore the connection is not arbitrary but iconic, and although this iconic connection may be consciously perceived by the viewer, the sign itself is not inefficient nor does it require the contortions of a game of charades. Napoli (2003:59) also claims that the signs of sign languages are formed by arbitrary connections: You can convince yourself of the fact that the vocabulary of sign languages has an arbitrary correspondence between structure and meaning just by looking at any two deaf people who are signing together. If signs were truly iconic, we’d be able to guess the meaning of a conversation without knowing the sign language at all. But we can’t.

If by “truly iconic” Napoli means that all lexical items must be one-hundredpercent iconic, then perhaps it would be possible to agree with her. But signs do not have to be one-hundred-percent pure icons to be motivated by similarity. Some signs, like the one in Figure 1, are so transparent that even non-signers can understand them. Others, however, like the sign in Figure 2 below, where the handshape is the same as in Figure 1, but where an additional movement has been added, and the location has been changed slightly, are much less transparent.

Figure 2.

39

40

William J. Herlofsky

The iconicity of this sign is a bit more complicated than the sign in Figure 1, and this results in less transparency, but it is still quite efficient and contortion-free. First, it is necessary to introduce the JSL signs for ‘male’ (Figure 3) and ‘female’ (Figure 4) below:

Figure 3.

Figure 4.

The thumb in Figure 3 represents the (vertical, usually bigger) male in comparison to the little finger in Figure 4 that represents a (vertical, usually smaller) female. The handshape in Figure 2 thus represents a compound of the ‘man’ and ‘woman’ morphemes. The first part of the sign in Figure 2 is a nonproductive ‘prefix’ indicating something like ‘parental relatives’, and so after the prefix, raising only the thumb indicates ‘father’, the prefix plus a raised little finger indicates ‘mother’, and both together as in the sign in Figure 2 indicates ‘motherand-father’. Again, though not at first glance transparent, this non-transparency does not mean that the sign is not ‘truly’ iconic, only that there is a combination of iconic, metonymic, and arbitrary components that make the sign difficult to interpret without some explanation. Once explained, however, this iconicity in the sign becomes a-parent (pun intended). In this way, then, signs can be iconic, but not necessarily transparent to the uninitiated, as Napoli seems to think they should be. In addition, they need not involve inefficient contortions, as Pinker seems to think. In fact, concerning the matter of iconicity in sign languages, as discussed in Herlofsky (2005), there seems to have been a concerted effort by some early sign linguists to avoid references to iconic phenomena, since, as the quotations above suggest, many linguists consider iconicity to play a minor role in natural languages, and since one of the main goals of early sign linguists was to provide data to prove that sign languages were indeed natural human languages, it made sense to them to avoid unnecessary references to the iconic aspects of sign languages. In the case of sign language classifier-like constructions, however, like the animate entity handshapes that are illustrated in the following section, there appears to be no denial of iconic elements. As noted in the introduction, Emmorey and Herzig (2003) claim that all classifier constructions are iconic in form. The following section is devoted to trying to illustrate the function of some of these iconic ‘classifier constructions’, as well as possible paths of grammaticalization. And finally, in the summary, it is suggested that linguists need not ignore or avoid

Iconic thumbs, pinkies and pointers

iconic data, because this data, alone or in comparison with classifier data from other languages, can in fact provide further evidence, if any is still needed, that sign languages are indeed natural human languages.

3. Iconicity and classifier-like handshapes Research on classifiers in both spoken and signed languages can be positively described as interdisciplinary, but describing it a little less charitably, we could say that it is unfocused. Trying to clarify all the problems related to classifier research is not within the scope of this paper, and so instead, for the purposes of the following analysis, we will depend mainly on recent publications on spoken and signed classifiers by two prominent researchers in these fields, Alexandra Aikhenvald (2000, 2003, for spoken language classifiers) and Karen Emmorey (2002, 2003 for sign language classifiers) for our definitions of important terminology and an analytical framework. Aikhenvald (2000:1) uses the term ‘classifier’ as “an umbrella label for a wide range of noun categorization devices”, and Allen (1977:285) provides the definition most often used of classifiers as surface morphemes that denote “some salient perceived or imputed characteristic of the entity to which the associated noun refers”. In other words, many languages have surface morphemes that comprise systems of noun categorization/classification devices that categorize the referents of these nouns according to certain salient characteristics, such as animacy, gender, humanness, size and/or shape, or other perceived or imputed physical properties. And though, as mentioned above, Aikhenvald and Emmorey use the term ‘classifier’ to describe the wide range of noun categorization/classification devices within these systems, in what follows, since there is still considerable doubt about the status of the JSL handshapes to be discussed, I will introduce another term, ‘classifier-like’, to refer to the animate entity JSL handshapes to be analyzed, leaving the decision on whether these handshapes should be termed ‘classifiers’ or not until later. To begin our discussion, let us consider one very common type of classifier in spoken languages, numeral classifiers (see Aikhenvald 2000 for examples of noun classifiers, verbal classifiers, and others). These classifiers attach classifying morphemes to numbers, and these kinds of classifiers categorize their referents according to shape or size or to other inherent properties. For example, in Japanese, flat, thin things like paper are counted with the classifier mai attached to numbers, and long, round, slender things like carrots attach the classifier hon. And so, two (ni in Japanese) sheets of paper are counted as ni-mai, and two carrots are ni-hon. Similar size/shape considerations motivate the formation of hand-

41

42

William J. Herlofsky

shapes in sign languages, and this is one reason why Frishberg (1975) first used the term in reference to sign languages. In JSL for example, flat, thin things like paper can be represented by the handshape in Figure 5, below, which along with the raising movement shown, means ‘to graduate’. For this type of sign the shape of the hands imitates the shape the hands would have handling the graduation diploma, and so they are often referred to as ‘handling-handshapes’.

Figure 5.

The shape of long, round, slender things like carrots can be signed using a handshape like the one in Figure 6a, and these kinds of handshapes are often referred to as ‘size-and-shape-specifiers’, or SASS, and the movement of the hands to trace the shape of the object is called ‘tracing’. To express the JSL sign for ‘carrot’, the signer would first touch his/her lower lip to indicate ‘red’ and then form the long-round-slender SASS-handshape, as in Figure 6b. In this way, then, the flat-thin and long-round-slender SASS handshapes in JSL are not just used with numerals for counting, but are used to form nouns (like noun classifiers) and verbs (like verbal classifiers) as well (It is interesting to note that the ASL sign for ‘diploma’ is similar to the sign in Figure 6a, since diplomas are rolled up in the United States).1.

Figure 6a.

Figure 6b.

One type of verbal classifier that is the focus of much recent research (see Emmorey 2003) are verbs of motion, where classifier-like handshapes form part of the verbs. Sign language verbs of motion usually involve some imitation of the

Iconic thumbs, pinkies and pointers

motion that is referred to, and this motion is usually expressed by some motion of the hand. When this happens, the hand usually assumes a shape that is in some way an iconic representation of the animate entity that moves. For example, though the verb ‘to go’ can be expressed without a classifier-like handshape, as in Figure 7 below, where the hand is more like an index, pointing to the path of motion, in Figures 8 and 9, which also can mean ‘to go’, the handshapes are classifier-like in that they imitate some characteristic of a vehicular animate entity that is going. Figure 8 means to go by airplane, and Figure 9, used often by older deaf people, means to go by train.

Figure 7.

Figure 8.

Figure 9.

Before going any further, let us consider a definition of sign language ‘classifier predicates’ provided by Emmorey (2002:21), where she defines American Sign Language (ASL) classifier predicates as: complex forms in which the handshape is morphemic, and the movement and location of the hand specify the movement and location of a referent. Classifier predicates are generally used to talk about the motion and position of objects and people or to describe the size and shape of objects. For example, to describe a person meandering along a path, a signer would use a classifier predicate in which the 1 handshape (referring to the person) moves forward with a zigzag motion.

In these ASL classifier constructions, then, the handshapes are morphemes that represent certain groups of people or objects, and due to this group-representation, are said to ‘classify’ different types of entities. In the above example, the 1 handshape (which is a fist with only the forefinger extended) is used to represent a person ‘meandering’. Now recall the examples illustrated in the previous section, where gendermarker-like handshapes like those in Figures 3 and 4 can indicate male and female referents, and can combine, as in Figure 2, to form a one-handed compound to indicate a male-female couple. This couple-handshape, a classifier-like hand-

43

44

William J. Herlofsky

shape, can be used in different ways to form other signs as well. For example, the sign in Figure 10, below, changes the orientation of the handshape, and adds a different kind of ‘meandering’ movement to form the sign for ‘date’; in other words the couple is walking along happily together on a date.

Figure 10.

Pluralization is also possible with this handshape, and is illustrated in Figure 11 with the sign for ‘people’, or in some contexts, it could mean ‘society’. For this sign, the movement does not imitate some actual physical movement, as it does in Figure 10, but instead is a kind of reduplication, where the different positions of the hands are intended to imitate more and different stationary men and women, i.e., people.

Figure 11.

A similar but different kind of pluralization occurs for the sign for groups of men and women. The groups are formed in a circle, and again, the movement does not imitate movement, but is a reduplication of the signs for men and women in a circular form.

Figure 12.

Figure 13.

Iconic thumbs, pinkies and pointers

So far we have seen the male-thumb and female-pinky used in gender-markerlike and plural-marker-like constructions. Now let us consider derivational-like, person-marker constructions, where the sign in Figure 14 indicates a ‘policeperson’ or ‘policeman’ (the badge on the hat + person) and Figure 15, a ‘teacher’ (shaking a finger to teach + person).

Figure 14.

Figure 15.

The thumb in these signs, in the unmarked interpretation, can mean ‘person’, but it could also mean ‘male’, in which case it is possible to contrast it with the pinky to indicate female teacher or police. It is also possible that the thumb handshape is beginning to function like a nominalizing derivational suffix like the English suffix ‘-er’ that changes the verb ‘to teach’ into the noun ‘teacher’. Support for this hypothesis is the mouthing of the Japanese word sensei (‘teacher’) that often accompanies the signed expression in Figure 15, with the mouthing perhaps aiding the transition from word to suffix. The head of some group can also be indicated by using the person-markingthumb, as in Figure 16 below, where the sign first imitates a portion of the Chinese character for ‘meeting’, and adds to that a raised thumb, indicating the ‘head of a group’ or ‘company president’ (The sign on the right is a platformed alternativeversion of the raised-thumb sign, and these signs are interchangeable.).

Figure 16.

The sign for ‘assistant head’ is interesting in that it is a compound of two thumbs, as illustrated below.

45

46

William J. Herlofsky

Figure 17.

Now consider the three signs below. These signs are more like the typical classifier constructions discussed in the sign language literature in that they involve an animate-entity handshape combined with a verb of motion. Figure 18 is the sign for ‘commute’, using the thumb for the unmarked or male commuter, but the pinky can also be used for females, as the inset illustrates. Figure 19 illustrates how the forefinger or pointer finger (a more gender-neutral person marker) can also be used as a classifier-like handshape, and the extended pointer finger coming toward the signer as in Figure 19 means ‘ to come’, and an opposite movement with the same pointer finger, away from the signer, means ‘to go’. The sign in Figure 20 means ‘to participate’, and is expressed by representing one ‘person’ (with the pointer finger of the right hand) joining a group or something (represented by the left hand).

Figure 18.

Figure 19.

Figure 20.

What is especially interesting about the following three classifier-like constructions is that they all involve the same or similar movement, but different handshapes. These signs illustrate one of the problems often encountered when discussing the morphology of classifier constructions that involve combinations of handshape morphemes and movement morphemes: trying to decide what the root morpheme is. If the compound signs are verbs, one would expect that the movement would be the root, and the handshapes some kind of affixes. But as the signs below suggest, this solution is unsatisfactory, since there does not seem to be a common identifiable semantic root in the motion of these constructions, except perhaps something like a metaphor-like ‘coming together’.

Iconic thumbs, pinkies and pointers

Figure 21.

Figure 22.

Figure 23.

The first one should be relatively easy to guess. It involves a man and woman coming together, and means ‘to marry’. The second sign, involving similar movement but different handshapes has quite a different meaning. For this sign the thumbs do not specify males only, and are instead the person-like-marker, and when brought together twice express the sign for ‘a meeting’ or ‘conference’. The third sign, in Figure 23, involves almost the same movement as the sign in Figure 21, but the meaning is quite different. The two extended forefingers represent two persons coming together, and express the verb ‘to meet’. In the examples that have been discussed so far, it appears that only subjectlike entities are represented by handshapes in these kinds of constructions. While the subject-like animate entity is what is usually represented by these handshape, object-like entities are also a possibility. Consider the two signs below.

Figure 24.

Figure 25.

Figure 24 means ‘to love’ someone, or to consider them ‘lovable’ (Note that this is an adjective-like function.). The object of the affection, the handshape, could be either the pinky or thumb, but the unmarked form for the verb ‘to love’ uses the thumb. The sign in Figure 25 is a bit more complicated. It is the sign meaning ‘to trick’ or ‘to fool’ someone, where the thumb is the person that is fooled. The right hand imitates a clever fox-like creature (this is in fact the JSL sign for ‘fox’), and the non-manual ‘tongue-in-cheek’ is also used for other signs, like ‘to lie’, and imitates pretending to have something in the mouth.

47

48

William J. Herlofsky

4. Summary What we have seen in the previous sections is how the thumb, pinky and pointer fingers can function as animate-entity classifier-like forms in JSL, but many questions remain. This same sentiment is also echoed by Aikhenvald (2003:89): Many questions remain unanswered at the present stage of research on classifiers in sign languages. Do classificatory verbs in sign languages “categorize” the referent of their arguments? How like – and how unlike – classificatory verbs in spoken languages are they? How do classifiers in sign languages get grammaticalized and what semantic changes take place?

It does seem that the thumb, pinky and pointer handshapes in the above data do, in some ways, categorize their referents into groups or classes of animate entities (see, however, Herlofsky (2004) where the onomatopoeia-like aspects of some of these same handshapes are discussed), and the thumb and pinky can also make this categorization finer, by distinguishing between males and females. To consider grammaticalization, however, it is first necessary to determine the origin of these expressions. Emmorey (2002:22), claims that, “[m]any, if not the majority, of signs in the core lexicon may have an origin in the classifier system and lexical innovation frequently arises from classifier predicates”. From this we might assume that these classifier-like handshapes first entered the lexicon in classifier constructions, and then lexicalized into the core lexicon, which would be the opposite of the route taken by many classifiers in spoken languages, where, Aikhenvald (2003:89) claims, “classifiers of most types come from nouns”. For sign languages, which usually have no writing systems, the determination of the etymology of lexical items is problematic. For the female-pinky handshape, however, the etymology seems to be fairly clear. Like the telephone handshape in Figure 1, the female-pinky seems to have originated as a borrowed gestural ‘emblem’.2 from the majority Japanese-speaking hearing society, where it still means something like ‘girlfriend’. Since no written-language records exist, the next steps can only be speculation, but it appears that the female-pinky handshape began as a borrowed emblem that first became a JSL noun (not unusual for foreign loan words), taking the more general meaning of ‘female’, and then, perhaps because of its vertical, slender, iconic appearance, by contrast, led to the formation of the vertical, thicker, male-thumb handshape (also a possible, but much less frequent emblem in Japanese hearing society), which also seems to have originated as a noun in JSL. Their contrasting iconic appearances, then, along with their ease of formation and mobility (moving along an axis perpendicular to their height), perhaps led to their becoming likely candidates for use in classifier-like functions, affixes and other gender-marker-like functions. These handshapes, then, appear to have entered the JSL lexicon first as

Iconic thumbs, pinkies and pointers

free morphemes, and over time, for some functions, became less free, bound morphemes. And while this grammatical contraction was occurring, the semantic range of the male-thumb expanded, to at times include the more general meaning ‘person’, and in certain verbs, ‘animals’ and even possibly ‘plants’.3 These two signs, then, after apparently entering the JSL lexicon as noun-like free morphemes first, later probably came to be incorporated into classifier-like constructions, and in fact, it appears that something like polygrammaticalization may have been involved, and this process may be similar to what Aikhenvald (2000:353) describes for spoken languages: Classifiers often come directly from open classes of lexical items, or from a subclass of an open class...This involves grammaticalization – from a lexical item to a grammatical marker – or polygrammaticalization, whereby one lexical item gives rise to more than one grammatical marker.

Again, although the male-thumb and female-pinky handshapes apparently entered JSL as open-class lexical items, over time, their grammatical functions diversified, and now, at certain times, these signs seem to be able to function as verb-incorporated classifiers, as in Figure 18, at other times they can also function more like gender markers, as seen in the inset in Figure 18, and at other times like derivational ‘suffixes’, as in Figure 15. This is not the case, however, for the other classifier-like form, the pointer handshape, which appears to be more like the typical classifiers found in other sign languages, in that it seems to have originated as a morphosyntactic portion of classifier predicates, and is still ungrammatical if it replaces the thumb in the grammatical marker-like functions in Figures 14–17. For the study of the grammaticalization of sign language classifiers, it may be important that we can identify classifier-like handshapes with different types of origins, since this may allow us to investigate a handshape that originated as a noun, and follows grammaticalization paths similar to classifiers in spoken languages, and others that originated in a classifier predicate, and exhibit grammaticalization paths similar to those of sign language classifiers. Now let us consider one more aspect of these JSL verbal classifier-like forms in comparison to spoken language verbal classifiers. Let us begin with a definitionlike quote from Aikhenvald (2000:149): Verbal classifiers appear on the verb, categorizing the referent of its argument in terms of its shape, consistency, size, structure, position, and animacy. Verbal classifiers always refer to a predicate argument (usually, S in an intransitive or O in a transitive clause) and can cooccur with it.

It is only necessary to change the word ‘on’ in the first line to ‘with’ or ‘in’ to make this definition applicable to the verb-incorporated JSL classifiers discussed in

49

50

William J. Herlofsky

this paper. That is, the thumbs and pinkies and forefingers in Figures 18–25 all categorize the referent’s shape or size or position or animacy, and, most crucially to this discussion, these forms seem to be (usually) limited to the subject position of intransitive verbs (Figures 18–20) and the object position of transitive verbs (Figures 24–25). What this tells us, then, is that the JSL verb-incorporated classifier-like constructions that we have examined in this paper fit Aikhenvald’s definition. There are some problems, however. Figures 21–23 express mutual action/ motion involving two entities, and the handshapes in these expressions can simultaneously function both as subjects and objects of their verbs. There are also transitive verbs of motion like ‘to follow’ or ‘to stalk’, that can involve the pointer finger or thumb of one hand ‘following’ or ‘stalking’ the pointer or thumb of the other hand, examples where both the object and the subject of the transitive verbs are classifier-like animate-entity handshapes. According to Aikhenvald (2000, p.c.), this type of subject classifier does not occur in spoken languages (see however her footnote, 2000:162). Perhaps these phenomena are the result of sign language’s visual modality, or perhaps due to the fact that there are two manual articulators that can be expressed simultaneously. While a thorough investigation of these phenomena is not within the scope of the present study, from the data a rule-like generalization can be made: Transitive Classifier Subject Rule If the object of a transitive verb is a classifier handshape, it is also possible for the subject to be a classifier handshape for verbs describing motion/action that (a) is mutual (b) involves both the subject and object of the verb

The reasons for these phenomena are not explained by this descriptive rule, and remain to be discovered in future research. Future research should also investigate in more detail the interaction in the contact between spoken and deaf cultures as it relates to language change, as has been only briefly suggested above in the example of the borrowing of emblems from the spoken-language culture to form signs. Interactions among individual signers as they negotiate meaning in conversations are also important, including research on use of mouth-shape and gestures to help convey meaning. In addition, at some point researchers should also consider how deaf children acquire classifier-like forms (see Nakano 2002 for a study of deaf children’s acquisition of SASS-like JSL handshapes), and how they categorize objects when referring to the things that surround them in their environment. Goldin-Meadow (2003:141) provides a hint of how deaf children might accomplish these tasks in the following

Iconic thumbs, pinkies and pointers

example of how deaf children create iconic gesture-like signs for things they have no sign for: [D]eaf children have two ways of referring to objects: (1) diectic gestures, such as pointing at the bird, and (2) iconic gestures, such as flapping two hands in the air, a “fly” gesture used as a noun to refer to a bird. When a deaf child uses a gesture like fly to refer to a bird, that child is, in effect, saying “that’s a bird.” However, the “fly” gesture goes beyond the label “bird” in that it identifies the particular property of the bird the child is focusing on. The gesture consequently reveals what the child means by “bird”. Thus, all of the deaf child’s iconic gestures used as nouns are, in this sense, categoricals. The question is whether any of them can be considered generic.”

We would also be interested to know if these categoricals are the beginnings of classifiers? Another area that should be of interest to future researchers, then, is how deaf children assign salience to objects in their environment and categorize them, how these objects are represented internally, how they are expressed externally as iconic gestures and signs (and classifiers), and, coming full circle, how these internal and external representations correspond with real-world objects. Aikhenvald (2000:5) begins her volume on classifiers by stating that research on classifiers is important because it can “provide a unique insight into how people categorize the world through language”, and, as I hope this paper has shown, extending this research to sign languages, and to the comparison of sign language classifiers and spoken language classifiers, can also provide insights. At the end of her book Aikhenvald (2000:435) states that, “there is an urgent need to provide more good descriptive studies in order to collect additional materials which may then assist us in rethinking the whole framework”. I hope that this paper has been able to make some small contribution to this process, and that it has also illustrated how even iconic data can help to convince skeptics (if some still exist) that sign languages are indeed natural human languages.

Notes I would like to express my gratitude to the Deaf group ‘Kusa-no-ne’ (‘Grassroots’) for their permission to use JSL illustrations from their book, ‘Akusesu! Rou-sha-no- Shuwa’(‘Access! Deaf Sign Language’ (Akaishi-shoten, Tokyo, 1998)) in this paper. I am also grateful to Sumiko Saito and Kazumi Muto for their kind instruction and patient answers to my questions about JSL, and Alexandra Aikhenvald for her many insightful and valuable comments on this and earlier versions (Herlofsky 2004) of this paper. 1. In sign languages, there are a variety of handshapes that appear to function in a way similar to noun classification devices in spoken languages. Although Emmorey (2003) discusses these similarities (and differences) in detail (see especially Schembri 2003, Aikhenvald 2003 and

51

52

William J. Herlofsky

Grinevald 2003 in that volume), in the present paper the analysis focuses only on a few classifierlike animate-entity handshapes in JSL. It should be mentioned, however, that handshapes like those in Figures 5 and 6 must also be included in a thorough analysis of classifier-like handshapes in JSL. These handshapes, the handling-handshape in Figure 5, and the SASShandshapes in Figures 6a and 6b, along with the classifier-like thumb and pinky and pointer handshapes, are representative examples of the three major classifier-like handshape groups in sign languages. The handling-handshapes focus mainly on the shape and function of the objects they represent, and are usually lexical items. The SSAS-handshapes as well are usually lexical, and focus on the physical size and shape of their referents. The classifier-like handshapes of classifier-predicates are morpho-syntactic and usually take into account not only physical characteristics of their referents, but social and cultural factors as well. In addition, there are combinations of these three types, like Figure 24, that combines the classifier-like female-pinky with a handling/touching-like shape and movement (also see the signs in note 3 below). Again, there is not enough space to discuss all of these important factors in the present analysis, and those needing more information should consult Emmory (2003). 2. Beattie (2004:39) defines emblems as “gestures with a direct verbal translation” that can be used as a “substitute for words”. 3. In JSL, it is now possible for the thumb (and pointer) to represent vertical living things other than human beings in certain contexts. To illustrate, Figure I below is the sign for ‘to raise’ a child, and enacts something like the giving of nourishment to the child, which is represented by the thumb-handshape. It is also possible to use the thumb-handshape in Figure I to refer to a domestic animal, like a dog, or perhaps something like the sign like Figure II, which could also mean something like ‘to take care of’ a pet dog. Figure IIIa illustrates how, in the proper context, it is possible to use the thumb-handshape (and the pointer-handshape as in Figure IIIb as well) to represent a non-animate but ‘vertical’ living thing, like a plant, as the sign means ‘to take care of a plant’, in this case imitating the watering of a plant.

Figure I.

Figure II.

Figure IIIa.

Figure IIIb.

And on an even more abstract level, the sign in the first portion of Figure 17 (the portion in the inset) has developed into an independent free morpheme with the meaning of ‘subordinate’, and in certain contexts can mean ‘the subordinate food’ that accompanies the main food, rice, at a meal. The thumb, therefore, has come to mean the main (or the subordinate) ‘anything’. It can also be used for some ‘main’ grammatical terms. The JSL expression for ‘subject’ is formed with the thumb as in Figure 3, and then the sign for ‘word’, which is formed with the pointer finger held up to the lips and then brought forward, which is also the sign for ‘say’. In other words, the thumb plus the pointer moving away from the mouth means ‘the main word’, or the ‘subject’ of a sentence.

Iconic thumbs, pinkies and pointers

References Aikhenvald, A. 2000. Classifiers: A typology of noun categorization devices. Oxford: OUP. Aikhenvald, A. 2003. Classifiers in Spoken and in Signed Languages: How to know more. In K. Emmorey (ed.), 87–90. Allen, K. 1977. Classifiers. Language 53:285–311. Beatie, G. 2004. Visible Thought: The new psychology of body language. London: Routledge. Christiansen, M. and Kirby, S. (eds). 2003. Language Evolution. Oxford: OUP. Emmorey, K. 2002. Language, Cognition, and the Brain: Insights from sign language research. Mahwah NJ: Lawrence Erlbaum. Emmorey, K. (ed.) 2003. Perspectives on Classifier Constructions in Sign Languages. Mahwah NJ: Lawrence Erlbaum. Emmorey, K. and Herzig, M. 2003. Categorical versus gradient properties of classifier constructions in ASL. In K. Emmorey (ed.), 221–246. Frishberg, N. 1975. Arbitrariness and iconicity: Historical change and American Sign Language. Language 51:676–710 Goldin-Meadow, S. 2003. The resilience of language: What gesture creation in deaf children can tell us about how children learn language. New York NY: Psychology Press. Grinevald, C. 2003. Classifier systems in the context of a typology of nominal Classification. In K. Emmorey (ed.), 91–109. Herlofsky, W. 2004. Aikon-sei to shizengengo: Nihon shuwa no [Classifier] no Baai (Iconicity and natural languages: The case of “classifiers” in Japan Sign Language). Paper presented at the 30th JSL Conferences, Nagoya, Japan. Herlofsky, W. 2005. Now you see it, now you don’t: Imagic diagrams in the spatial mapping of signed (JSL) discourse. In C. Maeder, O. Fischer and W. Herlofsky (eds), 323 – 345. Maeder, C., Fischer, O. and Herlofsky, W. (eds). 2005. Outside-In – Inside-Out. Iconicity in language and literature 3. Amsterdam: John Benjamins. Nakano, S. 2002. Otona no shuwa, kodomo no shuwa: Shuwa ni miru kuukan ninchi no hattatsu (Adult JSL, Child JSL: The development of spatial cognition seen in JSL). Tokyo: Akashi Shoten. Napoli, D. 2003. Language matters: A guide to everyday thinking about language. Oxford: OUP. Pinker, S. 2003. Language as an adaptation to the cognitive niche. In Language Evolution, M. Christiansen and S. Kirby (eds), 16–37. Oxford: OUP. Schembri, A. 2003. Rethinking ‘Classifiers’ in Signed Languages. In K. Emmorey (ed.), 3–34.

53

54

William J. Herlofsky

The physical basis for phonological iconicity

part ii Iconicity and the aural

55

56

Keiko Masuda

The physical basis for phonological iconicity

The physical basis for phonological iconicity Keiko Masuda Chuo University

This paper examines the physical basis of two types of phonological iconicity from a phonetic viewpoint and attempts to unravel the mechanism of phonological iconicity in its entirety. Many researchers consider the basis of association in indirect phonological iconicity (e.g. high-front vowels being associated with smallness) to be acoustic resonance frequency, size of oral cavity, and kinesthetic sensation. The first two bases appear closely related when examined in the light of the general principle in direct phonological iconicity (e.g. bow-wow). The bases both involve the front oral cavity (FOC), the resonance frequency of which is also the key factor in direct phonological iconicity (Masuda 2003; 2005). Therefore FOC may be one of the most influential bases of phonological iconicity.

1. Introduction As is generally agreed, phonological iconicity, which is also called ‘sound symbolism’, is a phenomenon covering a limited proportion of language; it does not affect all the words in a given language. Even in this small proportion of language, however, many words or expressions can be claimed to be iconic or sound-symbolic, and their nature and/or degree of phonological iconicity may vary greatly. For instance, cock-a-doodle-doo, bzzz, boing, weeny, and slip may be all sound-symbolic to some extent, but they are different in many aspects. Cocka-doodle-doo is a common expression which is meant to imitate the cockerel call, i.e. the speech sound of cock-a-doodle-doo imitates the sound produced by a cockerel. Bzzz may be the same as cock-a-doodle-doo in that it imitates the sound produced by a bee (or bees). It does differ from cock-a-doodle-doo, however, in that bzzz has an unusual structure of CCCC and lacks a vowel that can bear a syllable nucleus, therefore making itself not standard in English phonology. Boing is not standard in English phonology either, although its structure conforms to it, as a diphthong [fI] does not occur before a velar nasal []]. Many people may find phonological iconicity in weeny, as frequently claimed in literature, since the high front vowel [i:] is often associated with smallness. In this phonologically standard

57

58

Keiko Masuda

word, the speech sound of weeny does not imitate any sound but represents a sense of smallness, and the mechanism of representation or association has also been investigated. Some people may feel that slip, also a phonologically standard word, shares a common meaning or sense – namely, ‘smoothness’ – with other words beginning with sl- such as slime, slush, slop, and slide. Although the word does not directly imitate anything, it may be hard to deny that the word is sound-symbolic to some extent. Phonological iconicity may be divided into two major categories: direct phonological iconicity and indirect phonological iconicity. As one side (referrer) of the iconic ‘referrer–referent’ relation (or ‘signans–signatum’ in Jakobson and Waugh 1979) is an acoustic signal, it may be reasonable to divide phonological iconicity into two major types by whether or not the referent (non-linguistic reality) is also an acoustic signal (sound/noise), in which case the relation would be relatively transparent. If the referent is sound/noise, the oral sound represents it by directly modelling it; i.e. there is a direct mapping of linguistic sound to nonlinguistic sound. Expressions such as cock-a-doodle-doo and bzzz may belong to direct phonological iconicity. If the referent is not an acoustic signal, on the other hand, the process of iconisation or symbolisation does not involve such direct mapping. In other words, the process is acoustically indirect. Words belonging to this category may include boing, weeny and slip. Then, how is the referent iconised or symbolised into linguistic sounds? Are there any differences between the mechanisms of iconisation or symbolisation of direct phonological iconicity and indirect phonological iconicity? This paper examines the mechanisms of the two types of phonological iconicity from an acoustic and phonetic viewpoint and attempts to unravel the physical basis for phonological iconicity as a whole.

2. Direct phonological iconicity 2.1

Lexical and non-lexical phonological iconicity

Words or expressions belonging to the category of direct phonological iconicity may be most frequently found in comic strips, and are probably more commonly known as ‘onomatopoeia’. They include linguistic representations of animal sounds such as bow-wow, meow, cock-a-doodle-doo, or bzzz, of sounds produced by human beings such as atchoo or hic, and of other external sounds/noise such as bang, knock, ring, vroom, or thud. Direct phonological iconicity may be divided further into two categories according to the structure of the sound-symbolic expressions: non-lexical and lexical. Quoting some famous examples from Joyce’s Ulysses such as ‘pprrpffrrppffff’ or ‘kraaaaaa’, Attridge defines non-lexical onomatopoeia as

The physical basis for phonological iconicity

‘the use of phonetic characteristics of the language to imitate a sound without attempting to produce recognizable verbal structures’ (1988:136). Fischer (1999) gives an example of non-lexical onomatopoeia, ‘shshsh’, by contrasting it with its more lexicalised versions including ‘shoo’, ‘swoosh’, and ‘whoosh’. Non-lexical onomatopoeia lacks the structure to be a word of a language (English, in this case). Every English lexical word contains at least one long vowel, diphthong, or short vowel (followed by one or more consonants when the vowel is short), which can bear a stress (Giegerich 1992). One of Attridge’s examples from Ulysses, pprrpffrrppffff, and Fischer’s example, shshsh, lack a vowel and cannot therefore be lexical. Kraaaaaa, on the other hand, has either too many vowels or too long a vowel to be an English lexical item. Other such non-lexical onomatopoeia includes bzzz, brrr, aaaagh, and even some orthographically unrepresentable expressions. Lexical onomatopoeia, on the other hand, complies with the standard phonology of the language – well-formed in terms of the structure, use of phonemes, and phonotactics. However, there may also be sound-symbolic expressions which fall between them. While almost everybody would agree that pprrpffrrppffff is nonlexical, vroom, which has an unphonotactic sequence vr- and where r could be pronounced as a trill, may be non-lexical or lexical depending on how strict the criteria are. After all, lexical status is perhaps not binary but gradient.

2.2 Principle of direct mapping In direct phonological iconicity, the referent is an acoustic signal (non-linguistic sound or noise) such as animal sounds or environmental sounds/noise, and there should therefore be a direct mapping between the referrer and the referent, between the linguistic sound and the non-linguistic sound. Non-linguistic sound is interpreted into or replaced by the most appropriate (or, possibly in some cases, the least inappropriate) linguistic sound. What is the principle of mapping then? Brown (1958:116) states that the initial sounds of onomatopoeic names copy the stimulus gradients of the noises they represent, giving examples of voiceless plosives and fricatives. A noise with an abrupt onset seems to be given a name beginning with a voiceless plosive, while a noise with a gradual onset is usually named by a word beginning with a fricative such as [s] or [z]. Oswalt’s account of obstruents (1994:297–298) seems to be in line with the suggestion by Brown. Oswalt adds that fricatives represent sounds of abrasion or air turbulence produced in or out of the vocal tract, which indeed have a gradual onset. Rhodes claims that expressions with [I] or [i:] signify high-pitched sounds (1994:283–284). Bladon (1977) attempted a cross-linguistic investigation with five languages and also presented acoustic data of some of the modelled non-linguistic sounds and their linguistic representations. His analysis confirms the assumption that

59

60

Keiko Masuda

obstruents tend to be used to express an abrupt onset of the sound and high front vowels to express high-pitched sounds, and he provides other tentative observations. In addition to the general tendency of obstruents already mentioned above, it was observed that nasals seem to correlate with the presence of an overall weak formant structure in the modelled sound. The presence of periodic source spectra in the modelled sound seems to correlate well with the presence of a vowel in the linguistic representation, and the second formant (F2) of each vowel seems to be the one most influenced by the presence of a band of energy in the modelled sound spectrum. The second point appears be in line with the claim by Rhodes (1994:279) that the second formant pattern of the bird name arising from onomatopoeia mimics the call of the bird. Masuda (2003) implemented an experiment using ten birdcalls and their linguistic representations to reveal the general principle of interpreting nonlinguistic sound to linguistic sound. The research found that what tracked the main component of a birdcall was the lowest front cavity resonance frequency (FCR) of the realisation of the linguistic representation of the call. It had previously been pointed out that F2 of the linguistic representation matched the main component of a birdcall fairly well (e.g. Rhodes 1994), and in fact FCR is often represented by F2. In Masuda’s experiment, however, the analysis of representations with a sequence like [i]-[u] or [u]-[i], between which the two main resonance frequencies converge and change their customarily labelled names (i.e. either F2 or F3), demonstrated that it was the FCR dynamics, rather than the F2 dynamics, that tracked the main birdcall component. F2 is a conventional name used to refer to the second lowest formant and can represent different entities in some cases, whereas FCR represents one single entity. Therefore FCR appears to be the most important factor in the mapping principle. Employing the same method, Masuda (2005) conducted an experiment with six animal sounds (including one avian, the cockerel) and their linguistic representations to test if the FCR hypothesis by Masuda (2003) could be applied to animal sounds as well. The analysis of the experiment showed that the FCR dynamics of a linguistic representation tracked the ‘formant’ dynamics of the animal sound. It also found that other factors played a secondary role in choosing a particular vowel when the ‘formant’ dynamics of the animal sound were static. In the cases of the cow, for instance, the cow sound has a static formant pattern and some relatively clear formants. Of these, the lowest two formants are distributed at frequency levels similar to those of the first and second formants of the vowel [u:] used in its linguistic representation.1 It may therefore be suggested that the frequency levels of lower formants play a role in choosing vowels when the trajectory of the energy concentration is static in the original source sound. Nevertheless, taking into consideration the other cases of the experiment and other studies including

The physical basis for phonological iconicity

Masuda (2003), it seems likely that there is now sufficient evidence that FCR plays the most important role in the mapping principle.

3.

Indirect phonological iconicity

3.1 Associative and phonaesthetic phonological iconicity In contrast to direct phonological iconicity, where non-linguistic sounds are symbolised by linguistic sounds with fairly direct acoustic correspondence, expressions in indirect phonological iconicity represent non-linguistic, non-auditory realities such as sensation (e.g. pain), movement, feeling, size, or colour. This type of phonological iconicity can be further divided into two subcategories according to its nature: associative and phonaesthetic. Certain vowels, consonants, and suprasegmentals seem to represent visual, tactile or proprioceptive properties of objects, such as size or shape (Hinton et al. 1994). It was Socrates, according to Plato’s dialogue, who first noted that the high front vowels might denote diminutiveness, while the low back vowels may express augmentativeness (Jakobson and Waugh 1979:183). This type of phonological iconicity is a popular research topic and has been repeatedly investigated by both linguists and psychologists ever since Sapir’s pioneer empirical work (1929). In this paper this type of phonological iconicity will be called ‘associative phonological iconicity’ because linguistic sound and non-linguistic reality seem to be related by association. The other subcategory, phonaesthetic phonological iconicity, is a rather conventional submorphemic symbolic relation between a phoneme or cluster and particular meanings, which concerns positive, negative, and neutral meanings. A typical example of phonaesthetic phonological iconicity includes initial clusters such as gl- [gl] of glitter, glisten, glow, glimmer, etc. Phonaesthetic phonological iconicity may be the most lexicalised and conventional type of phonological iconicity. The relation between sounds and meanings is more indirect, arbitrary, and incorporated into the lexicon of the language. Phonaesthetic phonological iconicity seems to be considered generally as language-dependent and its level of iconicity may be lower than that of associative phonological iconicity. What will be compared, in this paper, with direct phonological iconicity as a representative of indirect phonological iconicity is associative phonological iconicity. In the field of associative phonological iconicity, which has been commonly referred to as ‘phonetic symbolism’ among the researchers, quite a lot of studies have been made following Sapir (1929), many of which were accompanied by similar experiments. Those studies can be divided into two types according to

61

62

Keiko Masuda

the kind of sound stimuli used in the experiments. One type of experiment uses nonsense (mostly monosyllabic) stimuli manipulated by combining vowels and consonants of a language, while the other type uses stimuli sampled from natural languages. Under both types, some experiments attempted cross-linguistic investigation.

3.2 Indirect phonological iconicity in nonsense words In many experiments on associative phonological iconicity with nonsense words, subjects were asked to assign one member of the minimal pairs of nonsense stimuli consisting of CVC (e.g. mil and mal) to one end of the semantic scales such as size, brightness, or distance. For example, Sapir (1929) used five vowels: [i], [e], [ε], and [æ] in preparing minimal pairs. The subjects were asked to assign one of the names, e.g. mal and mil, to either a small table or a large table and assign the other name to the other table. Newman (1933), inspired by Sapir, added three back vowels, [f], [o], and [O], some consonants and another scale ‘dark–bright’, in addition to the further analysis of Sapir’s. Bentley and Varon (1933) added some stimuli to the nonsense words used by Sapir and Newman. Upon hearing the stimuli, their subjects were simply asked to come up with a free association for each item on a mixed list, i.e. without a scale of magnitude or brightness or anything. Miron (1961) prepared pairs of nonsense CVC stimuli for his subjects, about two thirds of whom were native speakers of American English with no knowledge of Japanese, and the rest of whom were native speakers of Japanese whose proficiency in English varied. The subjects were asked to put each member of the pairs, which were auditorily presented, on 15 semantic differential scales such as good–bad, beautiful–ugly, high–low, heavy–light, thick–thin, large–small, cold–hot, etc. More recently, Tarte (1982) examined associative phonological iconicity in tones as well as vowels and consonants, using nonsense CVC word stimuli. Although the results of all those experiments differ a little from one another reflecting their varied settings and methods, the following observations can be made in general: (i)

(ii)

(iii)

high front vowels and high tones tend to be associated with qualities such as smallness, sharpness, and brightness. The front vowels can be put into a magnitude scale in the order: [i] < [e] < [”] < [œ] < [a]. Consonantal associative phonological iconicity does not seem to be evident. The only exception may be voiceless and voiced consonants of the same place of articulation (e.g. [p] and [b]), where voiceless consonants tend to be significantly ‘smaller’.2 The semantic scales which have significant correlation with vowels tend to be

The physical basis for phonological iconicity

(iv)

quantitative (e.g. small–large) rather than evaluative (e.g. beautiful–ugly). In matching nonsense stimuli and meanings, a binary linguistic reference or reference scale must be given to observe associative phonological iconicity.

3.3 Indirect phonological iconicity in natural language Associative phonological iconicity in natural language has been claimed with such examples as little–large (English: size), petit–grand (French: size), mikro–makro (Greek: size), and chiisai–ookii [tfiisai]–[ookii] (Japanese: size), ici–la (French: distance), and hier–da (German: distance) (Tanz 1971; Woodworth 1991; etc.). Associative phonological iconicity in natural language may also exist in grammar; the past tense of many English irregular verbs with the high front vowel takes a similar form with apophony (e.g. sing–sang, drink–drank, think–thought). This contrast of [I]–[æ] (or [f:]) may symbolise ‘present–past’, where ‘past’ could be considered as ‘distant from the present’ or ‘a large amount of time lying before the present’. Although not all the words indicating size, proximity, or temporality contain either [i] or [a] (or [o]), there seems to be at least some evidence from natural language. Most of the empirical researches on phonological iconicity in natural language involve experiments using antonym pairs in the subjects’ native language and the corresponding antonym pairs in a foreign language which the subjects do not know, or experiments similar to that. Such researches include Tsuru and Fries (1933), Maltzman et at. (1956) and Klank et al. (1971). Tsuru and Fries (1933) used pairs of antonyms in English and Japanese such as new–old, and asked monolingual native speakers of American English to guess which member of the Japanese pair meant which member of the English pair. While Tsuru and Fries (1933) tested only with the ‘English–foreign’ pairs, Maltzman et al. (1956) included ‘foreign–foreign’ pairs using two foreign languages (Japanese and Croatian) with English. Also in contrast to the preceding study, Maltzman et al. (1956) used only one word, not a pair, as a stimulus. That is, the subjects were visually presented with a word in one of the languages and a pair of words in another of the languages where one member of the pair meant the same as the stimulus. Then they were asked to underline one member of the pair which they thought meant the stimulus. In their attempt to investigate the determinants for successful word-pair matching in tests such as those introduced so far, Klank et al. (1971) conducted an experiment using English and five foreign languages (Czech, Hindi, Chinese, Japanese, and Tahitian). Eight pairs of antonyms from three semantic categories (potency, activity, and evaluative) and another eight pairs of antonyms from outside the three categories were prepared in English and translated into the

63

64

Keiko Masuda

five languages. They hypothesised that members of a pair must be on opposite ends of the same semantic category for successful word-pair matching. While the settings and methods vary from research to research, the results of the experiments in natural language seem to be in general less significant and convincing than those of the experiments with nonsense stimuli. This may not be very surprising since the sound shape or meaning of words of natural language can be influenced by many factors, while manipulated nonsense words are not. Words in most languages, probably in all languages, undergo sound change over a long period of time and/or lose or alter their original meanings. Such changes occur in each language, and the time, speed, degree, and direction of the changes may vary from language to language. Consequently, words of a similar phonological structure that might have once had a similar meaning across languages might have very different sounds and/or meanings after such changes. It may therefore be difficult to obtain convincingly significant results in cross-linguistic experiments to test the universality of associative phonological iconicity. Likewise, it is not surprising that universal associative phonological iconicity is more likely to be observed in simple and basic words such as those concerning size or proximity, the concept or value of which may be relatively stable over a period of time and not very different across languages. The concepts or value of other words might be more susceptible to historical change and/or culture connotation. Therefore it can be said that associative phonological iconicity in natural language may be observed in a limited group of words with simple and basic meanings such as size or proximity.

3.4 Basis of association It seems safe to say that across languages people generally tend to associate high front vowels and/or voiceless consonants with smallness/proximity and low back vowels and voiced consonants with largeness/distance in nonsense words. It has been also confirmed by some studies that words meaning ‘small’ or ‘near’ in many languages have a tendency to have high front vowels, voiceless consonant(s), and/or fewer syllables than those meaning ‘large’ or ‘distant’, which often seem to have low back vowels and/or voiced consonants. These two seem to correlate closely. Associative phonological iconicity concerns a relation between sound and meaning through association. The mechanism of association may be different from person to person. Nonetheless there seems to be considerable agreement on the manner and direction of association between some sounds and some meanings. What is the basis of association given that the phenomenon is shared with other people in the same linguistic community or even with people from different linguistic communities? In the case of vowels, the general view seems to be that the basis for associative

The physical basis for phonological iconicity

phonological iconicity may be acoustic resonance-frequency, size of oral cavity, and kinaesthetic sensation (Sapir 1929; Newman 1933; Brown 1958; French 1972; Fischer-Jørgensen 1978; Woodworth 1991). Brown suggests that humans might learn the extralinguistic mechanism or rule of association such as large objects making low pitched sounds and may use such prior knowledge for guessing the meanings of new names (1958:134). While some like Weiss (1968) support the learning theory, whether such prior knowledge is acquired or innate is not known. Some (Fischer 1999; Bentley and Varon 1933) seem to consider that articulatory sensation (size of oral cavity and kinaesthetic sensation) in particular is a key for association of vowels with meanings, while some such as Jespersen (1922), Jakobson and Waugh (1979), and Tarte (1982) seem to consider that frequency is a critical factor among other potential candidates. Ohala (1984; 1994), in accounts of the cross-linguistic use of fundamental frequency (F0) for the sound symbolism (phonological iconicity) in intonation, suggests a ‘Frequency Code’; high acoustic frequency is associated with smallness (as small objects tend to emit high-frequency sounds) while low frequency is associated with largeness (as large objects tend to emit low-frequency sounds). He claims that the utilisation of the Frequency Code may not be only cross-linguistic but also cross-species (vocalising species) such as dog, chickadee, frog, etc. The sounds made by a confident aggressor (or one who wants to appear so) are typically rough and have a low F0, as the aggressor tries to appear as large and intimidating to its competitor(s) as possible. On the other hand, to appear small and non-threatening may be necessary when in need of pacifying a potential aggressor. Producing high-frequency sounds to appear small could represent a form of infant mimicry, which may inhibit the potential aggressor from attacking since ‘natural selection has left most species with a very strong inhibition against harming conspecific infants’. He therefore considers that the Frequency Code is innate rather than acquired (Ohala 1984:4–6). Although he does not imply that acoustic frequency is the only factor, the Frequency Code could indeed be one of the most convincing bases for associative phonological iconicity. Eberhardt (1940) carried out interesting experiments with deaf and hearing subjects, which may shed light on the issue of the potential bases of association. Her study involved three experiments, the first two of which were preliminary ones with deaf subjects only, using stimuli from natural language, and the other involved both deaf and hearing subjects, using nonsense stimuli.3 The results of the preliminary experiments indicate that deaf children do feel the non-arbitrary, symbolic relation between experimental words and their meanings. The third experiment compared responses by deaf and hearing subjects directly, using the same methodology as Sapir (1929) and Newman (1933) used. The result indicates

65

66

Keiko Masuda

that there is a rough agreement between deaf and normal subjects, but there are also differences. For instance, both sets of subjects seem to agree that the high front vowels are ‘smaller’ and the low back vowels are ‘larger’. However, the deaf subjects rated [æ] between [I] and [I] ([I] < [æ] < [I]), while the hearing subjects rated ‘[I] < [I] < [æ]’. The deaf subjects also rated [u] smaller than the hearing subjects did. These differences could be attributed directly to the lack of auditory factor of the deaf subjects and to the greater significance of kinaesthetic factors (Eberhardt 1940:41). There may also be a visual factor; the deaf children saw the protruding lips and ‘small’ opening of the mouth of the person who taught the pronunciation of [u]. This might also have affected their judgement. Although the deaf subjects may have been able to feel the size of oral cavity when they pronounced the stimuli, which may be related to the resonance frequency, they may have been more influenced by the kinaesthetic factor, such as the degree of lip rounding, and possibly visual factor. This may indicate that the basis for associative phonological iconicity may be a combination of acoustic (auditory) and kinaesthetic (articulatory) factors while it might be able to exist solely on the basis of kinaesthetic sensation without acoustic sensation.

4. Physical basis for phonological iconicity We have so far considered the general principle of mapping in direct phonological iconicity and the basis of association in indirect phonological iconicity. When these two are taken together, there may be a great parallelism between the hypothesis of a general principle for creating a linguistic representation in direct phonological iconicity and the grounds for indirect phonological iconicity. What are referred to as ‘frequencies of vowels’ in giving an acoustic basis for associative phonological iconicity are, more precisely, resonance frequencies of front and back oral cavities, which are analysed as formants. Of the two main resonance frequencies, the one in question that is high for high front vowels is the front cavity resonance frequency. It can therefore be said that the acoustic basis for associative phonological iconicity is the front cavity resonance frequency, which may also be the basis of imitation in direct phonological iconicity. On the other hand, the ‘oral cavity’ whose size is considered to be one of the articulatory bases of associative phonological iconicity is, precisely speaking, the front oral cavity. High front vowels, where the size of the front oral cavity is small, tend to be associated with smallness or a small object. Here again, there is a parallelism between the basis for associative (indirect) phonological iconicity and that for direct phonological iconicity. And, of course, the front cavity resonance frequency is high when the front oral cavity is small. These parallelisms can be

The physical basis for phonological iconicity

described as in Figure 1. From the diagram of the parallelism, it may be said that the front cavity, whether its size or resonance frequency, is primarily responsible for creating a linguistic form in phonological iconicity. This will be called the FOC (Front Oral Cavity) theory. The mechanism shown in the diagram cannot accommodate all the bases of phonological iconicity as a whole. In associative phonological iconicity, kinaesthetic factors such as the movement of the tongue may also contribute to the mechanism of association. It has also been pointed out and supported by empirical studies that high tones, which are virtually unconnected with the front cavity, are likely to be associated with smallness.4 In this wider sense, Ohala’s Frequency Code (1984) can give a better explanation of the bases of phonological iconicity. The Frequency Code, however, cannot explain the results of experiments with totally deaf children by Eberhardt (1940). The results showed that the deaf children successfully associated high front vowels with smallness, which may indicate that phonological iconicity can exist without an acoustic basis. The FOC theory, on the other hand, can explain the results well. The deaf children may have felt the size of the front oral cavity and associated it with meanings. The FOC theory as shown in the diagram (Figure 1) might therefore give better explanations in the case of segmental phonological iconicity.

Figure 1. Diagram showing the parallelism between the basis of association in indirect phonological iconicity and that of representation in direct phonological iconicity. The example is for high front vowels such as [I].

67

68

Keiko Masuda

The FOC theory can be well accommodated in Ohala’s Frequency Code with only a slight change to his explanation (1984), and indeed these two theories together can explain a large part of the bases of phonological iconicity complementarily. In applying his Frequency Code to sound symbolism (phonological iconicity), Ohala states that ‘high front vowels have higher F2 and low back vowels the lowest F2’ and that the tendency to associate the former to smallness and the latter to largeness can be explained by the Frequency Code (1984:9–10). Although his explanation is not incorrect, it would be more suitable to say that the FCR of the different vowels should be compared, rather than F2, because F2 may represent different entities in different vowels. F2 represents the front cavity resonance for low back vowels and the back cavity resonance for high front vowels. It may therefore be more appropriate to consider that what the Frequency Code refers to as ‘frequency’ in segmental phonological iconicity is the (lowest) front cavity resonance frequency, which reflects the same entity regardless of whether it is represented by F2 or by F3, rather than the second formant, which is sequentially numbered. This modification does not influence the Frequency Code at all, but may help support it by giving a biologically and acoustically more legitimate explanation to its application. The two theories together may go some way toward explaining phonological iconicity by strengthening and complementing each other, and may be further strengthened with other yet unknown theories (Figure 2).

Figure 2. Diagram showing how the two theories, the Front Oral Cavity theory and the Frequency Code, could account for phonological iconicity together.

5. Conclusion No one could deny that there are some corners of language where sound and meaning are closely related in a non-arbitrary way. Such a phenomenon, called ‘phonological iconicity’ or ‘sound symbolism’, has attracted many scholars including linguists, psychologists, and literati. When considered in terms of phonological iconicity as a whole, Masuda’s FCR theory (2003; 2005) may highlight a great parallelism between the general principle of imitating a referent in

The physical basis for phonological iconicity

direct phonological iconicity on the one hand and the basis of association in indirect (associative) phonological iconicity on the other. This paper suggested that the FCR theory could bridge the acoustic and articulatory bases for associative phonological iconicity, since the size of the front oral cavity, which is the articulatory basis, correlates with the resonance frequency of the cavity (i.e. FCR), which is the acoustic basis. Therefore what constitutes the key basis of association in indirect phonological iconicity may be the front oral cavity. The paper also proposed that the basis for direct phonological iconicity and that for indirect phonological iconicity may indeed be the same, namely the front oral cavity (FOC theory). The FOC theory could help complement the Frequency Code (Ohala 1984), which cannot cover phonological iconicity felt by deaf children, who cannot perceive the acoustic frequencies. Together with the Frequency Code, which can account for tonal and prosodic phonological iconicity, the FOC theory may explain the physical basis for phonological iconicity in its entirety.

Notes 1. Before the Great Vowel Shift, the Present-day English [u:] was pronounced as [o:] in the Middle English. There might be a possibility that the cow sound moo [mu:] was pronounced as [mo:] before the Great Vowel Shift. This possibility, however, does not conflict with the analysis by Masuda (2005), since the frequency range of BCR (F1 in this case) of the cow sounds used in the experiment is 300–350 Hz, which would presumably not have been so deviant from that of F1 of the Middle English [o:]. 2. This may not be very surprising since vowels are acoustically and articulatorily less complicated than consonants; it may not be difficult to put various vowels on a comparable scale according to the height or frontness of the tongue, or first, second, or third formant. Such ordering of all the consonants seems to be impossible, although it is possible to put some of the consonants on a certain scale (e.g. putting voiceless plosives on a scale according to the place of articulation, or putting two consonants of the same place and manner of articulation on a voicing scale). 3. Deaf children normally learn to speak in such a way as to conform as nearly as possible to standards set by hearing persons. They must make speech movements of which they cannot experience the full sensory results. Although their speech is never exactly like that of hearing people, they get some very definite sensory impressions from speech through their remaining senses. The subjects used in the experiment, who were considered ‘definitely deaf’, had already mastered the fundamentals of pronunciation. The pronunciation of the stimuli in all the experiments were taught to the subjects, and their pronunciation was also checked, so that they understood exactly what sounds were involved in the stimuli (Eberhardt 1940). 4. The only exception where the front cavity could be connected to tones or pitch is the ‘intrinsic F0’ effect of high vowels.

69

70

Keiko Masuda

References Attridge, D. 1988. Peculiar Language: Literature as difference from the Renaissance to James Joyce. London: Methuen. Bentley, M. and Varon, E.J. 1933. An accessory study of ‘phonetic symbolism’. American Journal of Psychology 45:76–86. Bladon, R.A.W. 1977. Approaching onomatopoeia. Archivum Linguisticum 8:158–166. Brown, R. 1958. Words and Things. Glencoe IL: The Free Press. Eberhardt, M. 1940. A study of phonetic symbolism of deaf children. Psychological Monographs 52:23–41. Fischer, A. 1999. What, if anything, is phonological iconicity?. In Form Miming Meaning: Iconicity in language and literature, M. Nänny and O. Fischer (eds), 123–134. Amsterdam: John Benjamins. Fischer-Jørgensen, E. 1978. On the universal character of phonetic symbolism with special reference to vowels. Studia Linguistica 32:80–90. French, P.L. 1972. Toward an explanation of phonetic symbolism. Word 28:305–322. Giegerich, H.J. 1992. English Phonology: An introduction. Cambridge: CUP. Hinton, L., Nichols, J. and Ohala, J.J. 1994. Introduction: Sound-symbolic processes. In Sound Symbolism, L. Hinton, J. Nichols and J.J. Ohala (eds), 1–12. Cambridge: CUP. Jakobson, R. and Waugh, L.R. 1979. The Sound Shape of Language. Brighton: Harvester Press. Jespersen, O. 1922. Language: Its nature, development and origin. London: George Allen & Unwin. Klank, L.J.K., Huang, Y.H. and Johnson, R.C. 1971. Determinants of success in matching word pairs in tests of phonetic symbolism. Journal of Verbal Learning and Verbal Behavior 10:140–148. Maltzman, I., Morrisett, L. Jr. and Brooks, L.O. 1956. An investigation of phonetic symbolism. Journal of Abnormal and Social Psychology 53:249–251. Masuda, K. 2003. What imitates birdcalls? – Two experiments on birdcalls and their linguistic representations. In From Sign to Signing: Iconicity in language and literature 3, W.G. Müller and O. Fischer (eds), 77–102. Amsterdam: John Benjamins. Masuda, K. 2005. A phonetic study on animal sounds and their linguistic representations. English Language and Literature (The Society of English Language and Literature, Chuo University) 45:211–232. Miron, M.S. 1961. A cross-linguistic investigation of phonetic symbolism. Journal of Abnormal and Social Psychology 62(3): 623–630. Newman, S.S. 1933. Further experiments in phonetic symbolism. American Journal of Psychology 45:53–75. Ohala, J.J. 1984. An ethological perspective on common cross-language utilization of F0 of Voice. Phonetica 41:1–16. Ohala, J.J. 1994. The frequency code underlies the sound-symbolic use of voice pitch. In Sound Symbolism, L. Hinton, J. Nichols and J.J. Ohala (eds), 325–347. Cambridge: CUP. Oswalt, R.L. 1994. Inanimate imitatives in English. In Sound Symbolism, L. Hinton, J. Nichols and J.J. Ohala (eds), 293–306. Cambridge: CUP. Rhodes, R. 1994. Aural images. In Sound Symbolism, L. Hinton, J. Nichols and J.J. Ohala (eds), 276–292. Cambridge: CUP. Sapir, E. 1929. A study in phonetic symbolism. Journal of Experimental Psychology 12:225– 239.

The physical basis for phonological iconicity

Tanz, C. 1971. Sound symbolism in words relating to proximity and distance. Language and Speech 14:266–276. Tarte, R.D. 1982. The relationship between monosyllables and pure tones: An investigation of phonetic symbolism. Journal of Verbal Learning and Verbal Behavior 21:352–360. Tsuru, S. and Fries, H.S. 1933. A problem in meaning. Journal of General Psychology 8(1): 281–284. Weiss, J.H. 1968. Phonetic symbolism and perception of connotative meaning. Journal of Verbal Learning and Verbal Behavior 7:574–576. Woodworth, N.L. 1991. Sound symbolism in proximal and distal forms. Linguistics 29:273– 299.

71

72

Keiko Masuda

Reading aloud and Charles Dickens’s aural iconic prose style

Reading aloud and Charles Dickens’s aural iconic prose style Tammy Ho Lai Ming University of Hong Kong This paper explores how the Victorian practice of reading aloud affected Dickens’s writing style. My argument is that the practice contributed to an aural prose style in the author’s novels. In the novels, a considerable amount of passages which are particularly aural in nature and performance-oriented can be found. Dickens made use of various formal linguistic means such as typography, onomatopoeia, sound patterning, sentence length and prose rhythm to foreground the sound portrayed in the passages. Also, I argue that the emphasis on sound in Dickens’s novels sometimes also serves the purposes of narrative power, that is, it results in rhetorical impact, emotional heightening, and highlighting of key narrative moments.

1. Introduction

Dickens’s writing style was no doubt influenced by the Victorian practice of reading aloud, an activity significant of the period, and one which he himself indulged in both privately and publicly. Dickens knew his writings would be read aloud and my claim is that, as a result, he gave his characters’ speech and his prose narratives a notably oral-, aural- and performance-oriented style. In several places across his novels, his writings even both presuppose and encourage an oral recitation. I describe this as an aural prose style, a style meant to be heard. Iconic features are especially relevant to the analysis of this style, being “devices which depend on some perceptible resemblance between the physical properties of language and external reality” (Attridge 1982:287). Formal linguistic features such as sound patterning, onomatopoeia, repetition (lexical and syntactic) and sentence length help create iconicity in Dickens’s prose. I will argue that the aural elements in the prose are designed to boost narrative power by building up rhetorical impact, heightening an audience’s emotional responses and highlighting key narrative incidents.

73

74

Tammy Ho Lai Ming

This kind of performance-oriented style is especially prominent in certain thematic contexts. For example, sounds are an important part of Dickens’s narratives. He highlights them by using auditory signals such as contrastive phonetic textures, words that imitate sounds (in particular onomatopoeias), rhythmic patterns and lexical repetition. Another thematic topic lending itself to performative representation is the experience of repetition. Dickens seems particularly interested in using iconic features to transmit the physical and psychological turmoil of characters to a listening audience, which makes an analysis of these features helpful to plausibly reconstruct the way a reader might perform the texts aloud.

2. The Victorian practice of reading aloud Commenting on the success of Pickwick Papers, the Metropolitan Magazine remarks “Boz had completely taken possession of the ear […] of his countrymen” (James 2004:34). Ward (1970:339) also declares that “to understand Dickens you must think with the ear”. Altick (1980:81), by contrast, stresses that “[i]t was the combined theatrical and oral quality of Dickens’s novels, […], which so admirably fitted them to be read aloud”. Dickens’s own choice of title for his first ‘conducted’ weekly magazine, Household Words, taken from a speech from Shakespeare’s Henry V, “Familiar in their mouths as Household Words”, also hints at his personal motivations for having his words rolled on the reader’s tongue. Reading aloud has a long history. In their introduction to A History of Reading in the West, Cavallo and Chartier (1999:4) state that In the ancient world, in the Middle Ages and as late as the sixteenth and seventeenth centuries, the sort of reading implicit in many texts was oralized (as was their actual reading). The ‘readers’ of those texts were listeners attentive to a reading voice. The text, addressed to the ear as much as to the eye, played on forms and formulas that adapted writing to the demands of oral performance.

Reading aloud has two functions, namely communicating the written word to those who are unable to decipher writing and reinforcing the interlocking forms of sociability – within the family circle, in convivial social intercourse or in literary discussion among like-minded persons – that make up private life. (ibid.)

Though written in a time period after the one discussed by Cavallo and Chartier, there is ample evidence that Dickens’s novels were read aloud as the traditional reading practice seems to have continued from the Middle Ages to the Victorian period. His works were quite often read out loud in the family circle, especially in

Reading aloud and Charles Dickens’s aural iconic prose style

that of the middle class. Indeed, Dickens was one of the favourite writers in family reading throughout the Victorian period, the other writers being Shakespeare and Scott (Flint 1993:193). The practice of reading aloud enabled family members to engage in a simultaneously intellectually stimulating and relaxing activity, which also strengthened their sense of belonging together. Dickens himself read to his family occasionally when he was in the mood; he also read to his friends so that afterwards they could have discussions regarding the content and composition of the writing. Moreover, his works were often read aloud in public to large audiences who were not trained to read, making him one of the very few Victorian writers who could reach the non-middle classes. This was also due to his decision to publish his works in an affordable weekly or monthly serial format.

3. Reading aloud and Dickens’s writing: content and style Given what we know about Victorian reading practices, it is very likely that Dickens was aware of the way his works were read aloud. Indeed, Shelston (1970:78) says that Dickens was “conscious that his instalments were read, as they appeared, at family gatherings”. Perkins (1982:25) also perceives that “[t]he novels of Dickens are peculiarly fitted to be read aloud […] Dickens himself ultimately recognized this”. As a matter of fact, during his farewell reading tour, Dickens advertised the forthcoming new story The Mystery of Edwin Drood, so that the listeners to his readings could “enter upon a new series of readings, in their own homes at which his assistance would be indispensable” (Dexter 1932:253). The knowledge that his works were read aloud would have a significant effect on Dickens’s writing on at least two levels which call for notice: content and prose style. In terms of content, Dickens was cautious not to include indelicate material which could hurt the sensibilities of his readers when read aloud in a family context or at a large gathering where all ages and sexes could be potential listeners. House (1976:216), for example, believes that Dickens constantly censored his work due to “the common practice in Victorian society, in all classes, of reading books and periodicals aloud in the family circle”. Reading aloud also played a significant role in Dickens’s literary craftsmanship. His daughter Georgina (1897:62–63) recollects how Dickens dictated Household Words to her while walking about his room. It is also reported that he would dictate a sentence or a paragraph to the assistant who did shorthand for him, and afterwards he would ask the assistant to read aloud the sentence or the paragraph. In an interview which eventually developed to John Forster’s article “Dickens’s Amanuensis: An Interesting Talk with the Man who Used to Write for Charles

75

76

Tammy Ho Lai Ming

Dickens”, this assistant recalls that “[Dickens] would walk up and down the floor several times after dictating a sentence or a paragraph, and ask me to read it. I would do so, and he would, in nine cases out of ten, order me to strike out certain words and insert others” (Collins 1981:193). It is as if, to Dickens, reading aloud was an integral part of his writing and the two could not be divorced: he would like to judge how well the written word would translate into an auditory experience. This writing process seems to be in line with what Trickett (1976:37) says: “Novelists must hear what they are writing, at least in the act of writing, as poets more especially hear what they imagine in the act of conceiving it”. What I am after here is the effect of reading aloud on Dickens’s prose style. Because he knew that, at a technical and stylistic level, he was writing to an audience who not only read his novels but also listened to them, he employed what I will term an aural prose style. By an aural prose style I mean one that focuses particular attention on the physical, aural reception of a text that is read aloud. It utilises heightened phonetic textures and overtly patterned and foregrounded syntactic structures. It is highly iconic in the sense that it emphasises the sounding of the writing and reflects the writer’s desire and design of making the narrative heard instead of simply being read silently. While not dismissing the aesthetic motivation behind the construction of some sound-expressive passages, I would like to argue that, when there is a practical need for aural reception and appreciation of some literary work as is the case with Dickens’s fiction, it is possible to look at the sound-sense relationship beyond that of the aesthetic. Listening was vital to understanding Dickens’s fiction, as some critics have noted. Perkins (1982:25), for example, says “[i]t is interesting to note that those who experience difficulty in reading Dickens’s novels are able to listen with delight to their being read by someone who understands and loves the word”. Roseberg (1960:11) even declares outright that “[a]nyone who has not heard Dickens read aloud has probably never savoured the novels to the full”. Dickens opted for exploring the possibilities of iconic sound and syntax precisely because of their expressive effectiveness in the ‘performance’ context of reading aloud. This iconic aspect is an important but much neglected aspect of his narrative style.

4. Sounds in Dickens’s prose narratives Sound patterning is used extensively and frequently across Dickens’s novels to recreate and enhance some key narrative moments for the person reading aloud and for the listening audience. These moments mainly include environments

Reading aloud and Charles Dickens’s aural iconic prose style

describing natural forces and human sounds. It is natural that sound assumes a significant role in Dickens’s prose narrative since he was writing against a general background of reading aloud; and the portrayal of sounds is doubtlessly the most direct manifestation of sound-sense correspondences.

4٫1 Natural forces Rain is one of the natural elements Dickens aurally represents. It is meant to be heard, which the utterance “Mrs Rouncewell might have been sufficiently assured by hearing the rain” (1852–53/2003:105, emphasis mine) makes clear. The following passage from Bleak House shows how sound patterning is able to recreate aspects of the rain pouring on the Dedlock house in Lincolnshire: (1)

The vases on the stone terrace in the foreground catch the rain all day; and the heavy drops fall, drip, drip, drip, upon the broad flagged pavement, called, from old time, the Ghost’s Walk, all night. (1852-53/2003: 21, emphasis and boxes mine)

The contrastive repetition of two sets of vowels (high front vowel and low back vowel) points to two contrapositive aspects of the rain in the passage above. Whereas the repetition of the fully-lexicalised onomatopoeia ‘drip’, a word with a high front vowel, highlights some qualities of the rain: sharp and quick,1 the subsequent short phrases “upon the broad flagged pavement, called, from old time, the Ghost’s Walk, all night” with their dismantled syntax imitate the dismal and irregular rain rhythm. The assonance or repetition of the long back vowel sound /f:/ in ‘all’, ‘fall’, ‘called’ and ‘walk’ (‘foreground’ and ‘broad’ also have the dominant /f:/ element in them), on the other hand, builds up a sense of monotonous repetition and symbolises something that keeps going on. This aural quality of rain accords with the thematic structure of the novel: the monotonous rain is closely related to the bleakness of the Ghost Walk. This example from Bleak House shows the repetition of onomatopoeia and vowels to create expressive and contrastive sound patterns that diagrammatically enact the rain rhythm. This conforms with C. S. Peirce’s view that “[m]any diagrams resemble their objects not at all in looks; it is only in respect to the relations of their parts that their likeness consists” (CP 2. 282). This is obvious in the passage above, as the contrasting relation of two sets of sounds is diagrammatically reflected by the motivated structural contrasts of phonetic forms in language. The rain in Lincolnshire is mentioned again five chapters later. The elements in (1) such as the onomatopoeia ‘drip’ and the long back vowel /f:/ (‘falling’,

77

78

Tammy Ho Lai Ming

‘upon’, ‘broad’, ‘Walk’) are repeated to describe the dreary rain in a way which is strongly reminiscent of the previous description: (2) The rain is ever falling, drip, drip, drip, by day and night, upon the broad flagged terrace-pavement. The Ghost’s Walk. (1852–53/2003:103)

This proves that Dickens indeed intended to remind the reader of the unique style of the rain associated to the Dedlock House. The almost exact repetition of the lexical items and sounds suggests an unchanging pattern of the rain throughout the span of several chapters.

4٫2 Human sounds Dickens also used sound patterning to mime human sounds such as whispering and human noises which involve special vocalisation. Distinguishing the characteristics of a sound, he elaborated it across a long string of words. This has the person reading aloud to some degree repeat the sounds when he or she is orally pronouncing the sentences in which the listener-reader can recognise the phonetic patterns and thus acknowledge the sound described. Moreover, he often wrote about whispering and rendered with care a narrative depicting a genuine context of whispering, which the scene below from Bleak House exemplarily shows: (3)

Both sit silent, listening to the metal voices, near and distant, resounding from towers of various heights, in tones more various than their situations. When these at length cease, all seems more mysterious and quiet than before. One disagreeable result of whispering is, that it seems to evoke an atmosphere of silence, haunted by the ghosts of sound — strange cracks and tickings, the rustlings of garments that have no substance in them, and the tread of dreadful feet, that would leave no mark on the sea-sand or winter snow. So sensitive the two friends happen to be, that the air is full of these phantoms; and the two look over their shoulders by one consent, to see that the door is shut. (1852-53/2003: 514, emphasis and boxes mine)

In this passage in which two characters are whispering to one another, there is repetition of mainly the fricative consonants [s], [z] and [ƌ]. Ullmann (1964:69) points out that words connected with whispering contain the phonemes /s/, /ƌ/ and /tƌ/. The collocation of [s], [z] and [ƌ] in the above passage iconically mimes the sound and act of whispering between the two characters and recreates a phonetic equivalence of whispering in the voice of the one reading aloud. Some of the words used in the passage are also related to the environment when whispering

Reading aloud and Charles Dickens’s aural iconic prose style

typically takes place: ‘mysterious’, ‘quiet’ and ‘silence’. References to ‘ghosts’ and ‘phantoms’ are pertinent, too, since they refer to bodies without substance, that is, something of an immaterial nature, which is in a sense like what whispering is. Significantly, many words in the passage contain the high front vowel /i/ and its longer partner /i:/ (see boxes). These front vowels share the characteristic that their pronunciation requires a minimum opening of the oral cavity with the tongue positioned in the pre-palatal region. This way, the lips can be quite close when producing these vowels. I would argue that their use is a form of kinaesthesia2 and reinforces the physical aspect of whispering since when people are whispering they often attempt to be not heard and not seen. Here, Dickens’s text presupposes that the one who is reading aloud will be trying to modulate the voice to indicate the act of whispering to an audience. The narrative boosts the reader’s ability to do this by concentrating additional phonetic effects of whispering in the passage. Whispering is associated with soft sounds and usually involves only a handful of people in an enclosed space. To represent harsher noises in a larger area, Dickens employed another set of sound patterns. In Barnaby Rudge, there is a scene where some unpleasant and loud street noises are drawn to the reader’s attention: (4) Tink, tink, tink – clear as a silver bell, and audible at every pause of the streets’ harsher noises, as though it said, ‘I don’t care; nothing puts me out; I am resolved to be happy.’ Women scolded, children squalled, heavy carts went rumbling by, horrible cries proceeded from the lungs of hawkers; still it struck in again, no higher, no lower, no louder, no softer; not thrusting itself on people’s notice a bit the more for having been outdone by louder sounds – tink, tink, tink, tink, tink. (1841/2003:337, emphasis mine)

The ‘streets’ harsher noises’ include women scolding, children squalling, heavy carts rumbling and hawkers crying. The crowding of words with long back vowels, which are associated with heaviness (‘scolded’, ‘squalled’, ‘carts’, ‘rumbling’, ‘horrible’, ‘cries’, ‘hawkers’) recreates the collective noise that prevails on the streets. Most of these words also share the consonant /k/ (‘scolded’, ‘squalled’, ‘carts’, ‘cries’ and ‘hawkers’), an unvoiced stop which according to Anderson (1998:103) is the hardest and the noisiest in the sonority-periodicity spectrum. The progressive increase in length of the clauses (‘women scolded’ (4 syllables)  ‘children squalled’ (3 syllables)  ‘heavy carts went rumbling by’ (7 syllables)  ‘horrible cries proceeded from the lungs of hawkers’ (13 syllables)) suggests an increase in intensity of the sounds. Also, the repetition of the syntactic pattern NP (‘Women’, ‘children’, ‘heavy carts’, ‘horrible cries’) + VP (‘scolded’, ‘squalled’, ‘went rumbling by’, ‘proceeded from the lungs of hawkers’) in sequencing creates a sense of volume or multiplicity.

79

80

Tammy Ho Lai Ming

The passage also mentions another type of sound which contrasts with the harsh sounds resulting from the various human activities happening on the streets, namely the pleasant ‘tinking’ sound of Mr Varden, the locksmith, working in his workshop. The sound is likened to that of a silver bell, which draws the attention to Dickens’s use of the word ‘silver’. But not only Dickens: when commenting upon several lines from Edith Sitwell’s Colonel Fantock in which the word ‘silver’ appears, Anderson (1998:75) writes: The expressive value of silver remains dormant in neutral contexts, but their onomatopoeic and chromaesthetic potentialities are brought out in this case through visual and aural association (moonlight and music box), and through collocation with tinkling, an onomatope that echoes the high front vowel of silvery. Resonance is the point where linguistic and aesthetic resources converge.

In the quoted passage from Barnaby Rudge, ‘silver’ and ‘tink’ are indeed related to each other as different aspects of the bell: the material from which it is made, and the sound it makes, which foregrounds the sound expressive nature of ‘silver’. The continuous linear appearance of the word ‘tink’, a non-lexicalised onomatopoeia, at the beginning and at the end, creates a sense of rhythmic repetition of the sound, and encloses the unpleasant street sounds. The words ‘clear’ (/kli6/), ‘silver’ (/silv6/) and ‘bell’ (/bel/), which are associated with the ‘tinking’ sound, all have high front medial vowels and are phonemes that invoke positive ideas and images (Anderson 1998:229). Low back vowels and high front vowels are used against each other in the passage to signify two contrasting types of sounds: the unpleasant street noises and the agreeable ‘tinking’ sounds. This is an example of diagrammatic iconicity similar to the rain example from Bleak House discussed above (cf. ex.1). The contrast is not only felt by the person who reads this passage aloud, but also acoustically and perceptually by the audience member who is listening to the passage. The single sound ‘tink’ and its associated words with high front vowels such as ‘clear’, ‘silver’ and ‘bell’ are being placed in contrast with a syntactic unit multiplied four times to create ‘volume’. But this loudness of the noise is overpowered by the still louder ‘tinking’ sounds, represented by five repetitions of the word ‘tink’ at the end of the paragraph, signalling a long duration of the sound. The person reading aloud may be slowing down the utterance at this point, making the sound of each of the five ‘tinks’ clear and sharp. There is also a clear opposition between the hammer’s sound and the amorphic and irregular noises from the street. The ‘tink, tink, tink’ not only conveys an impression of a clear, neat and repeated sound, but also of order and regularity which cannot be found in the amalgam of disordered acoustic impressions from the street. Judging from this example Dickens seems to have been

Reading aloud and Charles Dickens’s aural iconic prose style

particularly interested in iconic effect when he is creating sound contrasts for dramatic result.

4٫3 Echoes The linguistic technique of repetition iconically mimes a sound which repeats itself. Repetition of a lexical unit in terms of rhetoric can suggest emphasis, but if the word hints at a sound, or is related to a sound, the repetition of the item can evoke the idea of that sound being repeated, as in the following example: (5)

“Murder,” said Nadgett, looking round on the astonished group. “Let no one interfere”. The sounding street repeated Murder. Barbarous and dreadful Murder; Murder, Murder, Murder. Rolling on from house to house, and echoing from stone to stone, until the voices died away into the distant hum, which seemed to mutter the same word. (1843-44/2004: 736, emphasis and box mine)

The narrative describes that the whole street is repeating the word ‘Murder’. This is mirrored precisely by the repetition of the word in the passage. According to Nänny (1997:122), “the repetition of words, phrases and whole lines (as a kind of refrain) is typical of a large number of echo-passages, some of which even acquire a performative character, that is, they do what they say”. Once the word ‘murder’ is uttered by Nadgett, it seems to take off and bounce back and forth, creating echoes. The echoing effect in the passage is reinforced by the phrases ‘house to house’ and ‘stone to stone’: the second ‘house’ and ‘stone’ are echoes and shadows of the first ‘house’ and ‘stone’. Although it is said that “the voice died away into the distant hum”, the sound of the word ‘murder’ lingers and reappears under the disguise of the similar sounding word ‘mutter’. Both ‘murder’ and ‘mutter’

begin with the bilabial consonant /m/ and end with the consonant cluster that is a combination of a plosive (/d/ and /t/) and the unstressed syllable /6/, which makes this a marginal case of pararhyme. The passage evolves around the word ‘murder’ and, in the end, when the word ‘mutter’ appears, Jonas has the illusion that the same word is repeated. As a result, ‘murder’ becomes a prominent word in the passage, in which the sound of the word ‘murder’ is heard repeatedly. Iconic repetition here is used both to create mimetic and expressive representation of Jonas’s inner emotional state and to involve the audience in his experience, thereby at the same time creating suspense and excitement.

81

82

Tammy Ho Lai Ming

4٫4 Sounds of machines or tools Dickens frequently uses an onomatopoeia or repeats it in order to imitate the sound a machine or a tool produces to iconically mime the monotonous nature of the sound as, for instance, the hammering sound in Barnaby Rudge represented by the repetition of ‘tink’ (cf. example 4); another example is the use of ‘whirr’ in the following passage: (6) “I’ll do nothing to-day,” said Mr Tappertit, dashing it down again, “but grind. I’ll grind up all the tools. Grinding will suit my present humour well. Joe!” Whirr-r-r-r. The grindstone was soon in motion; the sparks were flying off in showers. This was the occupation for his heated spirit. Whirr-r-r-r-r-r-r. “Something will come of this!” said Mr Tappertit, pausing as if in triumph, and wiping his heated face upon his sleeve. “Something will come of this. I hope it mayn’t be human gore”. Whirr-r-r-r-r-r-r-r. (1843–44/2004:49, emphasis mine)

‘Whirr’ is a lexicalised onomatopoeia which refers to “a continuous vibratory sound, such as that made by the rapid fluttering of a bird’s or insect’s wings, by a wheel turning swiftly, or by a body rushing through the air” (OED). In the above passage, ‘whirr’ is used to describe the sound of the wheel going round and round, its repetition drawing special attention to the existence and sound of the grindstone and the process of sharpening. Dickens applies innovative typographical means in his usage of the word ‘whirr’: the word is made longer in length through the repetition of the hyphen and the additional use of the letter , thus turning the word into a pure imitative construction and requiring the reader to interpret a conventional onomatopoeic form from a new perspective. Moreover, every whirr is longer than the previous one, which iconically represents an increase of volume of the grinding noise, at the same time as it is displayed iconically and visually on the page. Repeating a letter many times with the effect of stretching a corresponding phoneme is a technique also used by James Joyce in Ulysses in for instance, ‘frseeeeeeeeeeeeeeeeeeeefrong’ and ‘pfuiiiiiii’ (quoted in Wales 1992:108–109). That makes Ward (1970:229) conclude that “the writer in English most like Dickens is Joyce” because both writers are concerned with “words heard over words seen”. According to Anderson (1998:123), consonant extension like this is “synaesthetic as well as onomatopoeic”. The /r/ is supposed to be an “effortful noise” (Ehrenpreis 1970:232); Robson (1959:31) describes it as “the harsh raucousness and the oral roar of the vocal chords and the throatal resonance” which, when repeated, articulates the lingering noise created by the grinding tool. The sentence after the

Reading aloud and Charles Dickens’s aural iconic prose style

first ‘whirr-r-r-r’ (“The grindstone was soon in motion; the sparks were flying off in showers”) is loaded with fricative consonants (/s/, /f/ and /ƌ/) which further enhance the sound caused by the friction between the tool and the grindstone. Anyone reading this passage aloud is forced to prolong the pronunciation of the word ‘whirr’ and imitate the grinding sound. The reader also faces the challenges to reproduce this sound in a way that is distinctive, authentic and maybe even exciting to someone listening. Onomatopoeias, in particular those not yet fully lexicalised, are easily recognisable. Dickens often foregrounds them with peculiar spellings or typographical arrangements, as in the following example from David Copperfield: (7) At the same time there came from a workshop across a little yard outside the window, a regular sound of hammering that kept a kind of tune: RAT—tattat, RAT—tat-tat, RAT—tat-tat, without any variation. (1849–50/20049:136, emphasis mine)

‘Rat-tat’, like ‘whirr’, is also a sound imitative expression, but it is a non-lexicalised linguistic item. It refers to “a sharp rapping sound, esp. of a knock at a door” (OED). In the quoted sentence, ‘rat-tat’ refers to the hammering sound produced when a coffin is being made. It is made longer with the repetition of ‘tat’ and is repeated three times to form the refrain RAT—tat-tat, RAT—tat-tat, RAT—tat-tat which in total appears three times across the chapter (1849–50/20049:136–137) “without any variation”, as the narrative points out. The refrain represents a continuous flow of regular and rhythmic hammering which is enhanced by the use of the dash and the alternation of capitalisation of the word RAT with the lower cased tat-tat, indicating that a strong and loud sound is followed by weaker ones. The two intervening hyphens in each unit, one long and one short, suggest the possibility of prolonging the word RAT longer and the short duration of the first tat respectively. Providing the reader hints about the volume and duration of the utterance, Dickens is in this case obviously trying to encourage a dramatic or expressive performance of the hammering sound.

5. The experience of repetition in Dickens’s prose narratives Dickens also makes use of repetition to diagrammatically display his characters’ physical and psychological turmoil when they are experiencing an event or an emotion more than once. In some instances, he iterates a lexeme to reflect the intense mental or physical action, for example: “Then I went on, thinking, thinking, thinking” (1852–3/2003:47, emphasis mine); in others, he repeats the whole clause to signify an action or feeling either repeated or experienced as

83

84

Tammy Ho Lai Ming

repeated, as in the following examples from two novels as diverse as Little Dorrit and A Tale of Two Cities: (8) Once, Little Dorrit knocked with a careful hand, and listened. Twice, Little Dorrit knocked with a careful hand, and listened. All was close and still. (1855–57/2003:189, emphasis mine) (9) Once, he went aloft, and the rope broke, and they caught him shrieking; twice, he went aloft, and the rope broke, and they caught him shrieking; then, the rope was merciful, and held him, and his head was soon upon a pike, with grass enough in the mouth for all Saint Antoine to dance at the sight of it. (1859/2003:234, emphasis mine)

Though in (8) there are full stops between each part and in (9) only commas, suggesting a different pace, (8) and (9) are structurally similar, since both involve exact repetition of every level except for the word ‘once’ and ‘twice’. Since an audience expects written or spoken narrative to be communicated in a varied language, the repetition of identical clauses shows that there is something other than simple description going on. Therefore, when the clause is read aloud the second time, the audience is forced to re-experience the original situation over again. In (8), Little Dorrit is looking for a sleeping place. When she locates one, she tells Maggy, “This is a good lodging for you, Maggy, and we must not give offence. Consequently, we will only knock twice, and not very loud; and if we cannot wake them so, we must walk about till day” (1855–57/2004:189). Little Dorrit keeps her words and knocks twice only. The two knockings are the same: there is no variation in the sequencing of the words describing them. Consider the experience of a listener in hearing the repeated sentences read aloud, knowing the situation Little Dorrit and Maggy are in and the importance of there being an answer to prevent the two girls from wandering on the streets late in the evening. A good reader would pause for a considerable time after the first sentence, creating suspense and tempting the listener to guess if the knocking would by any chance be heard. When the same sentence is repeated word for word, the listener might then realise that fate is unkind to Little Dorrit and Maggy after all, and come to sympathise with their misfortune. Other instances in Dickens’s novels where a door is knocked on twice show that Dickens could narrate the event in a plain descriptive manner in a few words such as “The door was opened, after he had knocked twice; and he went in, leaving us standing in the street” (1952–53/2003:906, emphasis mine) or “It seemed that M. Todgers was not up yet, for Mr Pecksniff knocked twice and rang trice, […]” (1843–44/2004:127, emphasis mine). The almost exact repetition of the two sentences in (8) indicates that Dickens deliberately chooses to forsake economy of

Reading aloud and Charles Dickens’s aural iconic prose style

words and the principle of elegant variation in order to heighten the monotony of the knocking and its significance in the narrative. My example (8) shows how Little Dorrit takes the initiative to knock twice, with the linguistic repetition as a reflection of an external physical action and event. (9), on the other hand, describes a scene of a man involuntarily being lynched in public: the repeated clause unfolds his repeated experience of being hung, as if being punished once is not enough. When the same clauses are repeated, the listener knows that the man had to be hung again and most likely hopes that, after the second repetition, the torturous incident will be finished, which involves him or her in the character’s experience. My last example of this section is from Martin Chuzzlewit: (10) Up they came with a rush. Up they came until the room was full, and, through the open door, a dismal perspective of more to come was shown upon the stairs. One after another, one after another, dozen after dozen, score after score, more, more, more, up they came: all shaking hands with Martin. Such varieties of hands, the thick, the thin, the short, the long, the fat, the lean, the coarse, the fine; such differences of temperature, the hot, the cold, the dry, the moist, the flabby; such diversities of grasp, the tight, the loose, the short-lived, and the lingering! Still up, up, up, more, more, more: and ever and anon the captain’s voice was heard above the crowd… (1843-44/2004: 350, emphasis and boxes mine)

When Dickens visited America for the first time in 1842, “he was constantly writing about the fatigue and strain of spending a number of hours each day shaking hands with five or six hundred people” (Stone 1957:472). In Martin Chuzzlewit, Dickens transferred a similar experience to the eponymous character. In (10), Martin is greeted by a roomful of people and is shaking hands with them. In the description there is repetition on almost every linguistic level: there is lexical repetition because the words ‘more’ and ‘up’ are repeated. There is syntactic repetition because the pattern x after y (“One after another, one after another, dozen after dozen, score after score”) and such a of b (“such varieties of hands”, “such differences of temperature”, “such diversities of grasp”) are repeated. There is phonetic repetition because the vowel /f:/ is repeated: ‘door’, ‘more’, ‘score’, ‘coarse’. There is also the repeated use of monosyllabic adjectives which describe the different qualities of hands Martin shook (‘thick’, ‘thin’, ‘short’, ‘long’, ‘fat’, ‘lean’, ‘coarse’, ‘fine’, ‘hot’, ‘cold’, ‘dry’, ‘moist’, ‘tight’, ‘loose’, ‘short-lived’ (a compound made of two individual monosyllabic words). Dickens intentionally chooses to use monosyllabic antonyms, which, in a sense, is also repetition of linguistic form, making repetition the underlying linguistic principle organizing the words in the passage which describes an event of repetition, an event of people

85

86

Tammy Ho Lai Ming

doing the same activity (i.e. shaking hands). Therefore, the language pattern in the passage recreates the experience of the character for the listener, making the listening audience an active participant in the scene. There are other interesting linguistic details in this passage such as the syntactic repetition of the clause: “One after another, one after another, dozen after dozen, score after score” provides a sequencing of number from small to large (one  dozen  score), which mirrors Martin’s expectation of the number of people getting larger. Despite the fact that more people are coming, the number of adjectives used to describe their hands decreases, from eight (“the thick, the thin, the short, the long, the fat, the lean, the coarse, the fine”) to five (“the hot, the cold, the dry, the moist, the flabby”) and finally four (“the tight, the loose, the short-lived, and the lingering”), indicating that the more hands Martin shakes, the less he feels. However, the antonym pairs, which are linguistic examples of ‘varieties’, ‘differences’ and ‘diversities’, signify the variety of handshakes, and it can be argued that Martin still experiences each hand as physically different. But the fact that almost all the adjectives are monosyllabic shows that the brevity of each handshake fails to register a more elaborate sensation in Martin. The word ‘lingering’ with its three syllables stands out, implying that something lasts longer than one would expect. It refers to the overall monotonous experience of shaking hands. The fourth sentence in the passage is a long sentence packed tightly with linguistic information and structured into three separate parts, none of which has a verb. A verb in a sentence brings action, moving the narrative or creating some kind of direction, either forward or backward movement. However, we are given an inactive sentence in which nothing actually happens except that monotonous gesture of handshaking. Finally, “Still up, up, up, more, more, more” is the climax of the passage when read aloud. We have a clause of seven words with only three lexical items; practically it consists of nothing but repetition. So by the end of the passage, the accumulation of repetition has overwhelmed everything else. Judging from this handshaking episode, it is evident that Dickens is taking a great deal of care in organising words to recreate the experience of repetition both through repetition itself and through a meticulous word choice, involving the reader and the audience in the psychologically and physically charged incident. It is instructive to compare (10) with another passage in the novel which also involves repeated shaking hands but in which there is no significant syntactic repetition: (11) Mr Pogram shook hands with him, and everybody else, once more; and when they came back again at five minutes before eight, they said, one by one, in a melancholy voice, “How do you do, Sir?” and shook hands with Mr Pogram all over again, as if he had been abroad for a twelvemonth in the meantime, and they met, now, at a funeral. (1843–44/2004:510, emphasis mine)

Reading aloud and Charles Dickens’s aural iconic prose style

At the outset, Mr Pogram is the subject (“Mr Pogram shook hands with…”) and in the second half of the sentence he is the object (“they … shook hands with Mr Pogram”). This shift already implies variety, unlike the passage in which Martin is shaking hands, which is almost completely focalised on his perspective. Another difference between the two passages is that, while repetition is used lavishly in the Martin description, it is almost entirely absent in the passage about Mr Pogram, apart from a few lexical repetition (‘shook hands’; ‘they come back again’, ‘all over again’; ‘one by one’). The use of syntactic repetition makes the reader assume that an action or a state of mind takes place more than once. But Dickens’s selective use of this technique informs us that there is more repetition can do than simply denote an event being repeated. Indeed, linguistic repetition can also help register the feeling of monotony. This is maybe what Monod (1968:348) means when she says, “some of these repetitions can provoke a literary and psychological effect”. In the paragraph about Martin, the abundant use of linguistic repetition not only reflects the repeated action of handshaking, but it also mirrors how Martin feels about shaking hands as something boring, monotonous and torturing. Mr Pogram does not share his perception, and to him it may even have been a delight to shake hands with all those people. Hence, though Dickens has Mr Pogram repeat the action of handshaking, he chooses to abstain from using repetition in order to create a different effect.

6. Conclusion The passages discussed in this paper demonstrate Dickens’s intentional exploitation of the aural, iconic and expressive potentialities of language. The basic linguistic features Dickens used to create strong mimetic and performance-oriented passages include sound patterning, onomatopoeia and lexical and syntactic repetition. That Dickens used these features intentionally becomes clear when we see that he employs iconic features for certain kinds of evocative scenes and events while opting for a direct and descriptive representation for others. The non-universality of the performance-oriented style in Dickens’s prose indicates precisely that he recognised the potential of iconicity to achieve narrative power or to highlight key narrative incidents. While the first set of examples (cf. section 4) I analysed shows how Dickens stresses some sounds or aspects of sounds by highlighting the phonetics, the last set of examples (cf. section 5) has less to do with sounds than enacting the thoughts and feelings of the characters involved. Thus, when discussing the function of iconic form in Dickens’s fiction in the context of reading

87

88

Tammy Ho Lai Ming

aloud, the focus is not merely on the primary power of language to imitate sounds; the linguistic ability to mirror different psychological and emotional experiences of the characters through physical iconic sound textures of the language is also significant. Dickens often diagrammatically represents moments of suspense, or of physical and psychological turmoil. It also seems as if the use of iconic strategies is one of his most favoured ways of handling lengthy non-dialogue narrative that involves representing the combination of place and character psychology. This may have been prompted by his commitment to maintain and engage the interest of the reader and listeners. It has been a long-held opinion that Dickens’s strength lies in the dialogue which is entertaining and can easily be dramatised since, when reading aloud, the dialogue would capture the most attention. However, I would like to argue that attention should be paid to Dickens’s own prose as well, since he developed his iconic prose style in response to his awareness of the widespread habits of reading aloud and his own interests in the power and potential of the spoken word for injecting narrative with dramatic and psychological force.

Notes 1. Sadowski (2001:72) comments: It is possible to talk about high vowels as being ‘bright’, ‘sharp’, ‘cold’, ‘hard’, ‘light in weight’, ‘quick’, ‘narrow’ etc., and about low vowels as being ‘dark’, ‘blunt’, ‘warm’, ‘soft’, ‘heavy’, ‘slow’, ‘wide’ and so on. Similar sound-symbolic qualities can be argued for consonants with, for example, stops characterized as ‘hard’ and sibilants as ‘soft’, or the liquid [r] described in sound-symbolic experiments as ‘rough’, ‘strong’, ‘heavy’, ‘bitter’ etc., and [l] as ‘tame’, ‘peaceful’, ‘smooth’, ‘light-weight’, ‘clear’, ‘weak’ and so on.’ 2. “Kinaesthesia is iconism of a special kind, based upon correspondences of meaning and the physical attributes of articulation” (Anderson 1998:167).

References Altick, R. 1980. Varieties of readers’ response: The case of Dombey and Son. The Year of English Studies 10:70–94. Attridge, D. 1982. The Rhythms of English Poetry. London: Longman. Anderson, E. R. 1998. A Grammar of Iconism. Madison NJ: Fairleigh Dickinson University Press. Cavallo, G. and Chartier, R. (eds). 1999. A History of Reading in the West (Trans. L.G. Cochrane). Cambridge: Polity Press. Collins, P. (ed). 1981. Dickens: Interviews and recollections, Vol. 2. London: Macmillan. Dexter, W. (ed). 1932. Letters to His Oldest Friend: The Letters of a Lifetime from Charles Dickens to Thomas Beard. London: Putnam.

Reading aloud and Charles Dickens’s aural iconic prose style

Dickens, C. 2003/1859. A Tale of Two Cities. R. Maxwell (ed). London: Penguin. Dickens, C. 2003/1841. Barnaby Rudge. J. Bowen (ed). London: Penguin. Dickens, C. 2003/1852–53. Bleak House. N. Bradbury (ed). London: Penguin. Dickens, C. 2004/1849–50. David Copperfield. J. Tambling (ed). London: Penguin. Dickens, C. 2003/1855–57. Little Dorrit. S. Wall and H. Small (eds). London: Penguin. Dickens, C. 2004/1843–44. Martin Chuzzlewit. P. Ingham (ed). London: Penguin. Dickens, M. 1897. My Father as I Recall him. London: Dutton and Co. Ehrenpreis, I. 1970. The style of sound: The literary value of Pope’s versification. In The Augustan Milieu: Essays Presented to Louis A. Landa, H.K. Miller, E. Rothstein, G.S. Rousseau (eds), 232–246. Oxford: Clarendon Press. Flint, K. 1993. The Woman Reader, 1837–1914. Oxford: Clarendon Press. House, H. 1976/1941. The Dickens World. London: OUP. James, E. 2004. Charles Dickens. London: British Library. Monod, S. 1968. Dickens the Novelist. Norman OK: University of Oklahoma Press Nänny, M. 1997. Textual echoes of echoes. In Repetition in Language and Literature [SPELL Swiss Papers in English Language and Literature 7], A. Fischer (ed), 115–143. Tübingen: Narr. Oxford English Dictionary www.oed.com Peirce, C.S. 1932. Speculative grammar. In Collected Papers of Charles Sanders Peirce, Vol.2, C. Hartshorne and P. Weiss (eds), Elements of Logic, 129–269. Cambridge MA: Harvard University Press. Perkins, D. 1982. Charles Dickens: A New Perspective. Edinburgh: Floris Books. Robson, E. . 1959. The Orchestra of the Language. New York NY: Yoseloff. Roseberg, M. 1960. The dramatist in Dickens. JEGP 59:1–12 Sadowski, P. 2001. The sound as an echo to the sense: The iconicity of English gl- words. In The Motivated Sign, O. Fischer and M. Nänny (eds), 69–88. Amsterdam: John Benjamins. Shelston, A. 1970. Dickens. In The Victorians, A. Pollard (ed), 74–106. London: Barrie and Jenkins. Stone, H. 1957. Dickens’ use of his American experiences in Martin Chuzzlewit. PMLA 72:464– 478. Trickett, R. 1976. Vitality of language in nineteenth-century fiction. In The Modern English Novel: The Reader, the Writer, and the Work, G. Josipovici (ed), 37–53. London: Open Books. Ullmann, S. 1964. Language and Style. Oxford: Blackwell. Wales, K. 1992. The Language of James Joyce. New York NY: St. Martin’s Press. Ward, W.A. 1970. On Dickens. In Literary English since Shakespeare, G. Watson (ed.), 229–346. London: OUP.

89

90

Tammy Ho Lai Ming

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats Sean Pryor Trinity College, University of Cambridge This article examines the iconicity of the divine in the early poetry of W. B. Yeats. It begins with a central difficulty: how can poetic form mime meaning when that meaning, the divine, is unknown or even ineffable? Some of Yeats’s early poems respond to this difficulty with an ironic form of iconicity, mourning the fact that the more a poem seems like the divine, the less it is divine. The article then explores the ways in which Yeats’s early poetry tests this distinction between likeness and identity. It closes by examining certain poems which transform iconic effects into acts of creation (so that it is as if meaning mimes form), and by exploring the troubling implications of this inversion.

1. Introduction W. B. Yeats brought the 1890s to a culmination with the publication in April 1899 of a new volume of verse, The Wind Among the Reeds. The penultimate poem of the volume, and probably the last one written before publication, was “Aedh wishes for the Cloths of Heaven” (Yeats 1899b: 60). The poem encapsulates and epitomizes that fin de siècle fascination for grimoires and talismans, for a supernatural art about supernatural things: Had I the heavens’ embroidered cloths, Enwrought with golden and silver light, The blue and the dim and the dark cloths Of night and light and the half light, I would spread the cloths under your feet: But I, being poor, have only my dreams; I have spread my dreams under your feet; Tread softly because you tread on my dreams.1

The verse is self-consciously mellifluous and incantatory. Its rich textures of sound and syntax seem a deliberate attempt to copy those heavenly cloths. The poem weaves and reweaves a few specially coloured threads – the word dreams,

91

92

Sean Pryor

for example, or the sound of I in night and light – and so it makes a glimmering, otherworldly tapestry. The form of the poem mimes its meaning. Or, rather, this miming is deeply ironic, for the poem laments a horrible difference between its heavily embroidered cloths and the heavens’. In fact, it’s as if the more its cloths are like the cloths of heaven, the more iconic the poem is, the less its cloths are the cloths of heaven. Yeats’s early poetry continually tests this distinction between similarity and identity, and “Aedh wishes for the Cloths of Heaven” is an epitaph for the desire to find similarity and identity converge. It also reveals the crux which complicates this kind of iconicity. How can form mime meaning when we don’t really know what that meaning is, when we’ve never been to heaven or seen its cloths or heard the music of its spheres? Can iconicity be seen as a kind of imagination, a creation as much as a reflection, and what might be the repercussions? I want to explore the ways in which the young Yeats, spellbound by all things supernatural, sought to write icons of the unknown and even of the ineffable.

2. Iconicity and the divine in “The Song of Wandering Aengus” Another poem in The Wind Among the Reeds, “The Song of Wandering Aengus” (Yeats 1899b: 15–16), gives a good example of the kind of iconic effects with which Yeats experimented. Like “Aedh wishes for the Cloths of Heaven”, this is a poem of rich verbal texture signalling reverie and revelation. It describes Aengus’s vision of a miraculous girl: I went out to the hazel wood, Because a fire was in my head, And cut and peeled a hazel wand, And hooked a berry to a thread; And when white moths were on the wing, And moth-like stars were flickering out, I dropped the berry in a stream And caught a little silver trout. When I had laid it on the floor I went to blow the fire a-flame, But something rustled on the floor, And someone called me by my name: It had become a glimmering girl With apple blossom in her hair Who called me by my name and ran And faded through the brightening air.

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats

Though I am old with wandering Through hollow lands and hilly lands, I will find out where she has gone, And kiss her lips and take her hands; And walk among long dappled grass, And pluck till time and times are done, The silver apples of the moon, The golden apples of the sun.

There is a basic similarity established here between the mysterious and magic metamorphosis of the girl on the one hand, and the mysterious and incantatory metamorphoses of the poem on the other. This is clearest in the transformation of “something” into “someone”, emphasized by the parallel syntax of their lines. But such verbal shiftings actually appear everywhere: from “hazel wood” to “hazel wand”, from “hollow lands” to “hilly lands”, and so on. Single sounds weave in and out: “when white moths were on the wing”, for example, or “When I had laid it on the floor / I went to blow the fire a-flame”. This shifting of sound, word, and phrase is complemented by strange shifts in sense. The first major verbal phrase in the first and second stanzas, “I went”, changes from a sense of simple movement in line 1 to a sense of tried or attempted in line 10 (a difference stressed by parallel placement at the head of a line), and each instance echoes in a subclause: “when white moths” and “When I had laid”.2 The phrase then metamorphoses into the “I will” of line 19, which anchors the third stanza. The wholesale repetition of the phrase “called me by my name”, on the other hand, is iconic less because it participates in metamorphosis than because it mimes the real-time progress of the vision. Aengus is so caught up in reliving his memory that the second “called me by my name” is as revelatory as the first. (Similarly, the poem’s only identical rhyme gives the second “floor”, in line 11, a magic immediacy, as if freshly perceived.) The flowing enjambment and the running conjunctions of lines 15 and 16 – “called me by my name and ran / And faded through the brightening air” – trace the disappearing girl moment by moment. The trout transforms from “something” to “someone” as the very words are spoken. We can see further iconicity at work in the poem’s fluid figuration. In the first stanza, for example, Aengus sings metaphorically of a fire in his head; he likens the stars above him to the moths around him in a simile; and he catches a symbolic silver trout. There is a restless movement from one figurative dimension to another, bringing the poem into the vision’s dreamy world. Concrete details like the cutting and peeling of a hazel wand take on symbolic significance: the hazel is also a magic wand in Irish tradition.3 And through slippery use of the definite and indefinite articles, the metaphoric fire in Aengus’s head magically becomes the “real” fire on the floor, and then shifts immediately into a flame in “a-flame”

93

94

Sean Pryor

(leaving grammatical strictures in metamorphic uncertainty and suspension).4 “The Song of Wandering Aengus”, therefore, establishes similarities between its voice and its vision. And it is largely for this reason that Yeats can attribute divinity to his poem, can claim this voice is the voice of a god. For, as he explains in a note, Aengus is the Irish “god of youth, beauty, and poetry. He reigned in Tirnan-Oge, the country of the young [an Irish paradise]” (Yeats 1899a: 292; Yeats 1957:794). Yeats used this device throughout his career, from “The Stolen Child” (1886), written in the voice of the Irish faeries, to “The Delphic Oracle upon Plotinus” (1932), which channels the very god Apollo. And so it seems as though the rich embroideries of Aengus’s song are the cloths of its heaven. It may help to understand the combination of iconicity and divinity which “The Song of Wandering Aengus” promises, and the absence of which “Aedh wishes for the Cloths of Heaven” laments, in terms of Yeats’s symbolist aesthetic.5 In The Symbolist Movement in Literature, another seminal work of 1899, Yeats’s close friend Arthur Symons struggles with the arbitrariness of language and the desire for iconicity, for a motivated language. This inexorably leads him towards the divine and the unknown. The absent signified of any signifier is drawn into the realm of the supernatural: Symbolism began with the first words uttered by the first man, as he named every living thing; or before them, in heaven, when God named the world into being. And we see, in these beginnings, precisely what Symbolism in literature really is: a form of expression, at the best but approximate, essentially but arbitrary, until it has obtained the force of a convention, for an unseen reality apprehended by the consciousness. It is sometimes permitted to us to hope that our convention is indeed the reflection rather than merely the sign of that unseen reality. We have done much we if have found a recognisable sign. (Symons 1899:3–4)

Yeats is much less circumspect than his friend. For Yeats, though convention may play a part, art is ideally more than a reflection; it is a revelation. In an essay on Blake’s illustrations to Dante, first published in 1896, he declares: a symbol is indeed the only possible expression of some invisible essence, a transparent lamp about a spiritual flame; while allegory is one of many possible representations of an embodied thing, or familiar principle, and belongs to fancy and not to imagination: the one is a revelation, the other an amusement. (Yeats 1961:116)6

Of course, the distinction between symbol and allegory derives from Coleridge, who stresses the iconic participation of form in meaning which he finds in symbols. For Coleridge, the correspondence between “The Song of Wandering Aengus” and the metamorphic world of its vision is fundamental. In The Statesman’s Manual he writes that a symbol is

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats

characterized by a translucence of the Special in the Individual or of the General in the Especial or of the Universal in the General. Above all by the translucence of the Eternal through and in the Temporal. It always partakes of the Reality which it renders intelligible; and while it enunciates the whole, abides itself as a living part in that Unity, of which it is the representative. (Coleridge 1972:30)7

It’s important that for Symons, writing at the same time as Yeats, this translucence means a prelapsarian, Edenic language, or even the language of God, the Word or Logos. (In a very early poem, “The Song of the Happy Shepherd”, Yeats also invokes the equation of the Word with Creation: “The wandering earth herself may be / Only a sudden flaming word” [Yeats 1957:65].) Similarly, the eighteenthcentury scientist, philosopher, and spiritualist Emanuel Swedenborg, whom Yeats read and reread as a young man, describes the partaking or participation of form in meaning, which Coleridge celebrates and Symons tentatively imagines, as the heavenly language of angels: The sound of their speech corresponds to their affection, and the articulations of sound, which are words, correspond to the ideas of their thought derived from their affection; and since their language corresponds to their thoughts and affections, it is spiritual, for it is audible affection and speaking thought. (Swedenborg 1875:109)8

So, in order to explore these issues – the iconic likeness of form and meaning; the role of poetry as a living part in some greater, divine unity; and the language of heaven or God – I want to turn to “To the Rose upon the Rood of Time”, a poem which Yeats chose to place as the first of the lyrics in The Countess Kathleen and Various Legends and Lyrics (1892) and at the head of his early collection of verse, The Rose.9 It forms a kind of programme for his poetry.

3. Irony and desire in “To the Rose upon the Rood of Time” The Rose was the major symbol of Yeats’s early verse and he returned to it in poem after poem. In a sense it became his symbol of symbols, able, as Daniel Albright observes, to “stand for any desirable thing” (Yeats 1994:427) – even for the kind of symbolism which Yeats desired to write. Whether suggesting beauty, unity, poetry, Ireland, the divine, or a beloved, its origins may be traced in a range of literary and philosophical precedents, from the Rosicrucian philosophy with which Yeats was absorbed in the early 1890s to the celestial rose of Dante’s Paradiso. “To the Rose upon the Rood of Time” centres upon a paradisal vision of “Eternal Beauty”. Constructing a symmetrical structure, the poem flowers ornately from this centre to form its own rose in language (Yeats 1899a: 101–2):

95

96

Sean Pryor

Red Rose, proud Rose, sad Rose of all my days! Come near me, while I sing the ancient ways: Cuhoollin battling with the bitter tide; The Druid, gray, wood-nurtured, quiet-eyed, Who cast round Fergus dreams, and ruin untold; And thine own sadness, whereof stars, grown old In dancing silver sandalled on the sea, Sing in their high and lonely melody. Come near, that no more blinded by man’s fate, I find under the boughs of love and hate, In all poor foolish things that live a day, Eternal Beauty wandering on her way. Come near, come near, come near – Ah, leave me still A little space for the rose-breath to fill! Lest I no more hear common things that crave; The weak worm hiding down in its small cave, The field mouse running by me in the grass, And heavy mortal hopes that toil and pass; But seek alone to hear the strange things said By God to the bright hearts of those long dead, And learn to chaunt a tongue men do not know. Come near; I would, before my time to go, Sing of old Eire and the ancient ways: Red Rose, proud Rose, sad Rose of all my days.10

In his investigation of iconicity and symmetry, Ralf Norrman (1999:61–2) argues that symmetry offers “a magic vision of restored wholeness”: “Through asymmetry paradise was lost; through the magic of symmetry wholeness might be restored, and paradise regained.” This seems an apt account of the symmetry in “To the Rose upon the Rood of Time”, which is framed by the opening couplet and its inverse in the final couplet. It is as if the poem circles or spirals in towards Eternal Beauty in its first stanza, and circles or spirals back out again in the second. This rather spatial iconicity, depending upon the structure of the whole poem, is neat and perfect. It shifts the subject of the poet’s song, governed by the colon which follows each instance of “ancient ways”, from the legendary Irish past (Cuhoollin, the Druid, and Fergus) to the Rose itself. In this way, the second stanza’s retreat from the Rose, begging for yet a little distance, is encompassed in a greater unity figured by the Rose of the final line. The poem, we might say, “abides itself as a living part in that unity” which it invokes and celebrates. This roseate structure suggests that the poem’s most paradisal point – that point from which, as Dante says, “depende il cielo e tutta la natura” (Paradiso, 28.42) – comes between its two stanzas, in the silent white space which follows the

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats

vision of Eternal Beauty. In this way, “To the Rose upon the Rood of Time” promises the inviolate completion of Keats’s Grecian Urn, whose unheard melodies are sweetest. But the silent melody which wafts behind Eternal Beauty doesn’t bring the poem and the Rose together. If it did, if silence and white space could speak a heavenly language, then there would be no need to continue the beseeching chant “Come near” after the stanza break. Instead, Beauty’s eternal “wandering” may be better mimed by the potentially infinite incantation “Come near, come near, come near”. And since this incantation is motivated by absence and desire, its resemblance to Beauty’s wandering is ironic. This form of iconicity is irreconcilable with the self-contained, self-sufficient perfection of a Grecian Urn; it depends on imperfection and temporality. Each of its finite words passes away, just as the poet is spurred into song, he says, because he too will pass away in time. In turn, transience alerts us to a crucial difference between the first and final lines: the substitution of a full stop for an exclamation mark. Whether read as a sign of disappointment, so that the last lines become mere hollow repetition of a mortal incantation that can never reach eternity, or as a sign of hushed awe, or perhaps of quiet acceptance, this substitution is like a crack in the Urn, a tear in a rose-petal, preventing paradisal symmetry and seamless heavenly iconicity. (Even the threefold invocation “Red Rose, proud Rose, sad Rose”, which wills the Rose into being through the figure of apostrophe, betrays the absence which motivates the threefold incantation “Come near, come near, come near”, and so participates in that incantation’s earthly transience.) This fundamental tension within the poem’s iconic practice illuminates the conflict between the divine language of heaven (“the strange things said / By God to the bright hearts of those long dead”) and the earthly language of mortality (“heavy mortal hopes that toil and pass”). The incompatibility of these languages is an age-old theme. In his second letter to the Corinthians Paul writes that he once knew a man who “was caught up into paradise” and who “heard unspeakable words, which it is not lawful for a man to utter” (2 Corinthians 12.4). Swedenborg (1883–1902:1.9) writes that the “speech of celestial angels amongst themselves is incomprehensible” to man, and that it “involves more things in a moment than can be expressed on many sheets of paper”. That is, the cloths of heaven are beyond our temporality. Moreover, if the resemblance of wandering words to wandering Beauty is a sign of desire rather than of true likeness, this also seems true of the poem’s paradisal symmetry and wholeness. It may be true of every iconic feature which looks to mime heavenly language with earthly language, no matter how successful the resemblance may seem. This internal tension or paradox also complicates “The Song of Wandering Aengus”. We have seen the way in which the running repetition of and in lines 15 and 16 mimes the real-time progress of the girl’s disappearance. But these running

97

98

Sean Pryor

conjunctions also appear in the first and third stanzas, and to quite different effect. In the first stanza, with its conspicuous anaphora, there is another mysterious shifting of senses, this time between linear narrative (“And cut and peeled a hazel wand”) and simultaneous accompaniment (“And moth-like stars were flickering out”), so that the lines and the experience they describe share a hypnotic, dreamlike flow. In the third stanza the same form mimes a quite contrary meaning, reaching for an eternal simultaneity which escapes linear narrative: “And kiss her lips and take her hands; / And walk among long dappled grass”. But doesn’t this therefore establish another ironic conflict, setting mortal language against the divinity it desires, even in the case of the second stanza’s real-time progress? If the movement of the vision really matches the movement of the poem, does the utterance of the poem actually make the girl disappear, orchestrating loss and absence, and making that asymmetry which impels a poem to long to mime the world of long dappled grass?

4. Iconicity and solipsism in “The Lake Isle of Innisfree” The notion of making, of the creative agency of language, is important here. Iconicity assumes a meaning independent of the form, what Symons calls the “unseen reality apprehended by the consciousness”. But divine Roses and glimmering girls and heavenly cloths can’t be granted that reality without complication and doubt. Yeats’s early poetry explores this tension by tempting the convergence of similarity and identity. Heidegger, who sees art as a marriage of allegory and symbol which surpasses those very categories, believes that ancient statues aren’t representations of the form of a god, but actually are gods themselves (Heidegger 1971:19–20, 43).11 For some this convergence is the beginning of idolatry, and advocates of religious icons developed a range of strategies to avoid it without relinquishing the reality of the transcendent, unknown, or divine. In the Middle Ages certain images were worshipped as divine impressions made directly from Christ’s body, and legends arose which accounted for icons as the works or copies of the works of St Luke, who was thought to have painted portraits of Mother and Child in the flesh (Belting 1994:47–77). This divine inspiration parallels the attribution of poems to divine voices, except that the poems seem to understand their rhetorical claims as such. The poetry is also about the desire and the struggle to mime what cannot be mimed, and about the kinds of iconicity this produces. In this way, iconoclasm becomes equally relevant (Belting 1994:144–63), for fears of idolatry are objections to a world in which the icon is self-sufficient and in which man may imagine or even create God, dreaming divinity up out of nothing. This is ultimately a fear of solipsism, of iconicity so turned in upon itself that the form no longer needs the meaning.

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats

We can see this clearly in a poem which does not obviously claim a divine voice or look to represent a divine subject, though its slow and sonorous incantation and its liturgical echoes do cast a magic spell. “The Lake Isle of Innisfree” (Yeats 1899a: 117–18) was first published in December 1890, at the beginning of the decade which closed with The Wind Among the Reeds: I will arise and go now, and go to Innisfree, And a small cabin build there, of clay and wattle made; Nine bean rows will I have there, a hive for the honey bee, And live alone in the bee-loud glade. And I shall have some peace there, for peace comes dropping slow, Dropping from the veils of the morning to where the cricket sings; There midnight’s all a glimmer, and noon a purple glow, And evening full of the linnet’s wings. I will arise and go now, for always night and day I hear lake water lapping with low sounds by the shore; While I stand on the roadway, or on the pavements gray, I hear it in the deep heart’s core.

Perhaps the first iconic features to strike the reader here are the deceleration produced by the poem’s famous metrical trick12 and by the overlapping syntax of the second stanza, in which the poem itself comes dropping slow: “some peace there, for peace comes dropping slow, / Dropping from the veils”. This slow lingering affects the narrative of the poem, paradoxically deferring the desired journey from London to Innisfree: lingering over the “I” in “arise” and the repeated “go”, the poet goes nowhere. In fact, the poet returns to his opening chant in the third stanza because it is the chant rather than any physical arrival that matters. (Notice the way in which the third line’s desiring verb “have” is like but not identical to its object, the “hive”.) In a poem intoxicated by sound, from the bees’ loud hum to the cricket’s singing, the sound of Innisfree seems already present. In listening to this sound the poet is magically transported: the repetition of “hear” in lines 10 and 12 inverts the earlier insistence upon “there” in lines 2, 3, and 5, and makes Innisfree here. But really what the poet hears is the sound of his own voice, under the illusion of iconic similarity. He alliterates l sounds and creates ripples on a lake. He delights in the adjacent long vowels and cloying consonants of the words bee-loud glade, and imagines the silence-filling hum of bees. And, in a play of sound which moves from the iconicity of form miming meaning to the solipsism of self-referentially miming itself, he intones the final line such that hear-ing and here-ing are found and resound in the “deep heart’s core” (not to mention the line’s orthographic rhyme). Like “The Song of Wandering Aengus”, “The Lake Isle of Innisfree” is anchored

99

100 Sean Pryor

in each stanza by a narrative verb (“I will”, “I shall”, “I will”) that concedes the lack which iconicity desires to satisfy, the asymmetry which a perfect symbol would restore by partaking of and manifesting the unknown and divine. Instead the poem’s only narrative is a journey into its own voice. Every line of the final stanza begins with “I” (doubled in lines 9 and 11 in “arise” and “While”). The poem could almost be described as the chant of a tongue men do not know, for it is no longer about communication. There is no explicitly divine language involved because this solipsistic voice presumes its own overwhelming divinity and its own Creation. It is the presumption of Milton’s Satan, who defiantly declares that “The mind is its own place, and in it self / Can make a heaven of hell, a hell of heaven” (Paradise Lost, 1.254–5). And yet, finally, the poem’s retreat into the self is haunted by the bitter irony of an imitative caw in “core”, the all too earthly, anguished, and inarticulate cry of a raven, a bird that has no place on Innisfree.

Notes 1. Here and in the case of “The Song of Wandering Aengus” I have used the text which appeared in The Wind Among the Reeds (Yeats 1899b). Yeats later made minor revisions, and variants may be traced in the Variorum edition of Yeats’s poems (Yeats 1957:149–50 and 176). In a note at the end of The Wind Among the Reeds Yeats explains his use of personas such as Aedh: “I have used them in this book more as principles of the mind than as actual personages. […] Aedh is the myrrh and frankincense that the imagination offers continually before all that it loves” (Yeats 1899b: 73–4; Yeats 1957: 803). 2. When the poem was first published in The Sketch on 4 August 1897 (under the title “A Mad Song”), the tenth line read “I bent to blow the fire aflame” (Yeats 1957:149). In revising to “went” Yeats swapped the alliteration on “b” for the shift in the sense of the verb. 3. Cf. “Mongan laments the Change that has come upon him and his Beloved” (Yeats 1899b: 22–3): A man with a hazel wand came without sound; He changed me suddenly; I was looking another way; And now my calling is but the calling of a hound. In a note Yeats remarks that “The man in my poem who has a hazel wand may have been Aengus, Master of Love” (Yeats 1899b: 94; Yeats 1957:807). 4. In later printings “a-flame” appears as “aflame” (Yeats 1957:149), and the revision is marked in some of the extant manuscripts (Yeats 1993:57). The change has no great effect on the pun. 5. The following account of Romantic and of spiritualist theories of language is inevitably brief and incomplete. I have tried only to give a sketch of a few ideas which informed Yeats’s poetic. 6. The essay is dated 1897 in Essays and Introductions (Yeats 1961:145), but was first published in the July, August, and September 1896 issues of The Savoy. 7. I would like to thank Leon Surette, whose fine paper at the Kraków symposium reminded

Iconicity and the divine in the fin de siècle poetry of W. B. Yeats 101

me of Coleridge and of Yeats’s debt to Coleridge. Yeats would also have known Browning’s introduction to Shelley’s Letters, in which Browning declares (Browning 1852:43): I would rather consider Shelley’s poetry as a sublime fragmentary essay towards a presentment of the correspondency of the universe to Deity, of the natural to the spiritual, and of the actual to the ideal, than I would isolate and separately appraise the worth of many detachable portions which might be acknowledged as utterly perfect in a lower moral point of view, under the mere conditions of art. 8. In the introduction to his play The Cat and the Moon, Yeats recalls reading Swedenborg’s Heaven and Hell and Spiritual Diary when a young man (Yeats 1962:404). 9. In 1895 Yeats reorganized his lyrics into two retrospective collections: Crossways, containing his earlier work, and The Rose, containing more recent work (Yeats 1957:64). 10. In order to have a text contemporary with The Wind Among the Reeds, and so as to position this article at the end of the 1890s, I have used the text which appears in the revised edition of Yeats’s Poems (1899a), a collection first published in 1895. I have also used this edition for “The Lake Isle of Innisfree”. All variants may be found in the Variorum Poems (Yeats 1957:100–1 and 117). 11. In his essay on “The Origin of the Work of Art” (Heidegger 1971:43), Heidegger writes that “the sculpture of the god” is not a portrait whose purpose is to make it easier to realize how the god looks; rather, it is a work that lets the god himself be present and thus is the god himself. The same holds for the linguistic work. In the tragedy nothing is staged or displayed theatrically, but the battle of the new gods against the old is being fought. The linguistic work, originating in the speech of the people, does not refer to this battle; it transforms the people’s saying so that now every living word fights the battle. 12. Albright describes this trick as the substitution of a bacchius for an iamb in the third foot of the first three lines of each stanza (Yeats 1994:437). I would qualify this by saying that Yeats plays off stress and length: “now” in the first and ninth lines, “there” in the second, third, and fifth lines, and “way” in the eleventh line are long but unstressed vowels. Their juxtaposition with the previous stressed syllable creates the sense of deceleration. This is effective even when the stressed syllable is short (“build”, “have”), but not when the following syllable is short (“morning”, “glimmer”, “lapping”). In the latter case it might be better to describe the lines as substituting an anapaest for an iamb in the fourth foot, though any strict insistence upon classical scansion is probably misplaced.

References Belting, H. 1994. Likeness and Presence: A history of the image before the era of art (Trans. E. Jephcott). Chicago IL: University of Chicago Press. Browning, R. 1852. Introductory essay. In P.B. Shelley, Letters. London: Edward Moxon. Coleridge, S.T. 1972. Lay Sermons. London: Routledge and Kegan Paul. Heidegger, M. 1971. Poetry, Language, Thought (Trans. A. Hofstadter). New York NY: Harper and Row. Norrman, R. 1999. Creating the world in our image: A new theory of love of symmetry

102 Sean Pryor

and iconicist desire. In Form Miming Meaning, M. Nänny and O. Fischer (eds), 59–82. Amsterdam: John Benjamins. Swedenborg, E. 1875. Heaven and Hell; Also, the intermediate state, or world of spirits; A relation of things heard and seen. London: Swedenborg Society. Swedenborg, E. 1883–1902. The Spiritual Diary, 5 Vols. (Trans. G. Bush and J. H. Smithson). London: James Speirs. Symons, A. 1899. The Symbolist Movement in Literature. London: William Heinemann. Yeats, W.B. 1892. The Countess Kathleen and Various Legends and Lyrics. London: T. Fisher Unwin. Yeats, W.B. 1899a. Poems. London: T. Fisher Unwin. Yeats, W.B. 1899b. The Wind among the Reeds. London: T. Fisher Unwin. Yeats, W.B. 1957. The Variorum Edition of the Poems of W. B. Yeats. P. Allt and R.K. Alspach (eds). New York: Macmillan. Yeats, W.B. 1961. Essays and Introductions. London: Macmillan. Yeats, W.B. 1962. Explorations. London: Macmillan. Yeats, W.B. 1993. The Wind among the Reeds: Manuscript materials. C. Holdsworth (ed). Ithaca: Cornell University Press. Yeats, W.B. 1994. The Poems. D. Albright (ed). London: J. M. Dent.

Is lámatyáve a linguistic heresy? 103

Is lámatyáve a linguistic heresy? Iconicity in J.R.R. Tolkien’s invented languages Joanna Podhorodecka Pedagogical Academy of Kraków The aim of this paper is a brief study of iconic effects in the phonology of J.R.R. Tolkien’s invented languages. Tolkien’s notion of lámatyáve or ‘phonetic fitness’ is here explained in reference to Ivan Fónagy’s theory of symbolic vocal gestures – systematic, meaningful distortions of speech sounds that convey emotive messages. By analyzing several samples of Tolkien’s artificial languages, the author proves that effects similar to those that Fónagy describes on phonetic level appear on phonological level in the structure of those languages, most notably in quantitative proportions of particular sounds.

1. Introduction One of the basic tenets of structuralist linguistics, strongly emphasized by de Saussure, was the notion of the linguistic sign as conventional and arbitrary. Iconicity was regarded as a marginal phenomenon, limited to a few onomatopoeic words. Now we know that iconic effects are present in many ways in many areas of language. The question I would like to address here is whether iconicity could be present as well at the general phonetic level of language. Using as my examples J.R.R. Tolkien’s invented languages, I would like to see if elements of the very phonological structure of a language can be regarded as iconic. J.R.R. Tolkien is widely known for his imaginative writing, but he was a linguist before he became an author. Indeed, few of his numerous admirers realize that the complex fictional world that Tolkien created was primarily linguistic in its inspiration. Middle-earth originated in Tolkien’s invented languages, a kind of “an intellectual game of Patience” (Carpenter 1992:102) that he carried on throughout his life, starting from his early childhood. “The invention of languages was the foundation,” he explained. “The ‘stories’ were made rather to provide a world for the languages than the reverse.” (Letters: 219–220) “Nobody believes me,” he went on to complain, “when I say that my long book is an attempt to create a world in

104 Joanna Podhorodecka

which a form of language agreeable to my personal aesthetic might seem real. But it is true.” ( Letters: 264.) The construction of languages was for Tolkien a form of art, a private symphony having for its instruments the system of sounds. His interest was therefore primarily in the word-form itself, but also in the relation between the sound-symbol and the notion signified. In the invented languages, even more than in a natural language, that relation is not purely arbitrary, but derived from certain phonetic fitness based on an individual phonetic taste and contributing to actual aesthetic pleasure in ‘‘contemplating the new relation established’’(Tolkien 1997:206).

2. Lámatyáve Among several languages that Tolkien invented there were two that reached the greatest degree of elaboration and complexity: Quenya and Sindarin, within Tolkien’s mythology – languages of the Elves. Both appear in The Lord of the Rings in the form they reached in late 1930s. Quenya was begun as early as the year 1912 and by 1917 it was fairly complex, though it never reached a complete form – Tolkien kept changing and eleborating it even after the publication of The Lord of the Rings. Quenya grew out of Tolkien’s fascination with the language of Kalevala, the Finnish national epic, and it is modelled on Finnish especially in its phonetics and noun case system, but it was not the only source of inspiration. Tolkien described it as “the archaic language of lore meant to be a kind of ‘Elven-latin’, (…) composed on a Latin basis with two other (main) ingredients that happen to give me ‘phonaesthetic’ pleasure: Finnish and Greek. It is however less consonantal than any of the three” (Letters: 176). Sindarin, on the other hand, borrows from Welsh noun plurals based on vowel mutations and its characteristic phonology with frequent nasals and voiced fricatives. While Quenya is a language of lore and ritual, Sindarin is the living Elvish vernacular, the lingua franca of Middle-earth. It was derived from an origin common to it and Quenya, but the changes have been deliberately devised, in Tolkien’s own words, “to give it a linguistic character very like (though not identical with) British-Welsh: because that character is one I find, in some linguistic moods, very attractive; and because it seems to fit the rather ‘Celtic’ type of legends and stories told of its speakers”. (1997:197). In the process of elaborating his linguistic systems Tolkien discovered that “the making of language and mythology are related functions” and that “your language construction will breed a mythology” (Tolkien 1997:210). Language needs history to grow and change, a background against which it can develop, so

Is lámatyáve a linguistic heresy? 105

Quenya and Sindarin became languages of the Elves, and Tolkien, having begun with purely linguistic invention, found himself creating ‘legends’. Tolkien’s main interest as a linguist was in “word – form in itself, and in word – form in relation to meaning” (Tolkien 1997:211). He described the aesthetic pleasure he derived from language as resulting from his sensitivity to linguistic pattern affecting him emotionally like colour or music (Tolkien 1981:212). Tolkien’s sensitivity to language was not limited to the word-form. He argued that what makes a language beautiful is not only its phonetic pattern, but also its combination with meaning. Could that relation be limited to convention only, as the notion of arbitrary linguistic sign would entail ? Or on the contrary, was it motivated ? Discussing the problem, Tolkien referred to his somewhat intuitive notion of “phonetic fitness”, which in Quenya is termed lámatyáve – ‘the taste of the sound’. It typically refers to individual phonetic taste, appreciation or dislike for particular combinations of sounds, but Tolkien distinguished in it two closely connected aspects – the personal and the traditional. He believed that lámatyáve has an intersubjective component. According to T.A. Shippey, the notion of lámatyáve lies at the heart of Tolkien’s “linguistic heresy” (1983:87). Namely, Tolkien believed that out of the phonetic shape itself one can extract a certain kind of meaning related to the emotive and aesthetic sphere. He insisted on placing several untranslated Elvish passages in The Lord of the Rings so that the readers would perceive some kind of sense in the very sound, uncover certain ‘feel’ or ‘style’ that English translation could not convey. He asumed that judgements could be communicated through the sound. Consequently, the aesthetic and axiological aspects of Tolkien’s mythology are embedded in (or indeed grew out of) the language – Elves are noble and beautiful, Orcs harsh and hideous, which is visible in their names and languages. Tolkien described Quenya, the “Elven-latin”, as “phonetically and semantically sentimental” (1997:213). How exactly can emotive attitudes such as sentimentality be present at the phonetic level ? I would like to argue that there is something specific and fairly substantial behind the seemingly subjective notion of lámatyáve. A “linguistic heresy” similar to Tolkien’s was propagated by Ivan Fonágy, a Hungarian linguist, who suggested that speech acts result from double encoding procedure. They consist of two levels: a primary speech act based on an arbitrary linguistic code and a secondary speech act founded on the universal paralinguistic code and conveying attitudinal message, whose content is subconscious and preconceptual (Fonágy 1999:3).

106 Joanna Podhorodecka

3. Vocal style in Tolkien’s invented languages One of the ways in which the patterns of pre-verbal communication can be embedded in speech acts is what Fonágy termed vocal style (1991:495). It consists in the transmission of complementary pre-verbal messages by means of meaningful distortions of sounds connected with similar emotive attitudes. On the basis of labolatory analyses of Hungarian and French emotive speech, Fonágy argues that there are clear differences in articulation of sounds influenced by positive and negative emotions. Agressive emotions, such as hatred and anger, are characterized by highly increased muscular tension, the withdrawal of the tongue during the articulation and its fast spasmodic movements – the transitions are short, the tongue is briefly stiffened in extreme positions. In consequence, the relative duration of vowels decreases and that of consonants is increased. This is the most clearly noticeable in the articulation of plosives (sounds such as /p, b, t, d, k, g /, where the passage of the air through the vocal tract is first briefly blocked and then suddenly released). On the other hand, in speech expressing tenderness the articulation is smooth and continuous, speech organs are more relaxed, and the tongue is more advanced in the articulation of vowels. (Fonágy 1991:495–496). The core of Fonágy’s argument is the idea that secondary speech acts are essentially motivated, i.e. iconic (1999:7). He calls them symbolic vocal gestures and argues that from an evolutionary point of view they are primary speech acts, which brings to mind Vico’s idea of primeaval human communication by means of hieroglyphs and symbolic gestures. Apart from quantitative isomorphism (e.g. the speed of the utterance corresponding to the degree of excitement), Fonágy distinguished two basic principles of encoding vocal gestures. In my view, the two correspond very closely to the notions of metonymy and metaphor, therefore I shall describe them as metonymic and metaphorical vocal gestures. The former consist in the reproduction of symptoms associated with the vocal apparatus, e.g. pharyngal contraction or laryngal closure in angry speech can be considered as biological metaphors of nausea and cough, which serve to eliminate harmful substances from the body. By extension, they indicate the attitude of refusal and rejection (Fonágy 1999:7). Metaphorical speech acts take place when speech organs are identified, on the basis of resemblance or functional analogy, with some other objects, or when articulation accompanies or replaces bodily gestures. For instance, forwarding of the tongue in speech dominated by positive emotions corresponds to symbolic approach and backwarding of the tongue in hostile emotions represents withdrawal (Fonágy 1999:8). Emotive attitudes result in similar changes of the phonetic shape of words. Can phonetic effects of a particular emotion metonimically represent the emotion? Can an emotion be evoked or referred to by the reproduction of its phonetic

Is lámatyáve a linguistic heresy? 107

symptoms? According to Fonágy, the effect produced by meaningful distortions of sounds in speech can also be achieved by deviances in the frequency of particular sounds (1999:11). For instance, the pre-verbal message transmitted by fronting vowels in speech influenced by positive emotions can also be conveyed by increasing the number of front vowels within a particular text. The changes take place not in the quality of a particular sound, but in the quantity of sounds. From the level of phonetics they are transferred to the level of phonology. Such meaningful changes in the typical distribution od sounds usually take place in poetry, but the same mechanisms can be used in individual linguistic invention, and then one could actually talk about a phonetic structure of a word or a language as ‘‘aggressive’’ or ‘‘sentimental’’. I think that such phonetic metaphors underlie Tolkien’s idea of lámatyáve and phonetic fitness; and indeed make the very form of his invented languages meaningful. Under 1, 2 and 3 below, there are three samples of Tolkienian languages. I take it as a consequence of my thesis that distinguishing out of them the Black Speech, the negatively charged language spoken by the evil servants of the Dark Lord Sauron in The Lord of the Rings, will not constitute a problem. Tolkien himself described Black Speech as “so full of harsh and hideous sounds and vile words that other mouths found it difficult to compass, and few indeed were willing to make the attempt” (Tolkien 1996:35). 1.

A Elbereth Gilthoniel, silivren penna míriel o menel aglar elenath ! Na-chaered palan-díriel o galadhremmin ennorath, Fanuilos, le linnathon nef aear, sí nef aearon ! (Tolkien 1991, I: 311)

2.

Ai ! Laurie lantar lassi súrinen, Yéni únótime ve rámar aldaron ! Yéni ve linte yuldar avánier mi oromardi lisse-miruvóreva Andúne pella, Vardo tellumar nu luini yassen tintilar i eleni ómaryo airetri-lírinen. (Tolkien 1991, I:490–491)

3a.

Ash nazg durbatulûk, ash nazg gimbatul, ash nazg thrakatulûk agh burzum-ishi krimpatul.

(Tolkien 1991, I: 332)

108 Joanna Podhorodecka

3b.

Ugluk u bagronk sha pushdug Saruman-glob bubhosh skai (Tolkien 1991, II: 53

The Black Speech is passage 3, where 3a. is a sample of literary language and 3b a specimen of its slang form, used by the Orcs. Passages 1 and 2 are Sindarin and Quenya, the languages of the Elves. The table below presents the specification of the number of vowels ( V ) and consonants ( C ) within 100 sounds in the samples presented above. It also includes the percentage of particular vowels in the overall number of vowel sounds. I am interested here mainly in the quality and number of the vowel sounds, so I treated diphthongs as separate vowels. I also ignored the length of the vowels, since due to the unavailability of native speakers it is difficult to establish the exact duration of the sound. I am well aware that this is too small a sample to draw conclusions, but even in such a cursory analysis very interesting regularities can be noticed. Language 1. Sindarin 2. Quenya 3.Black Speech

V C C per V / I: /machine / e / end 45 55 1.22 22,5% 35,5% 48 52 1.08 28% 23% 37 63 1.7 14% –

/ % / cut 29% 29% 38%

/ f: / for / u / brute 11% 2% 10% 10% 8% 40%

In articulation distorted by aggressive emotions, consonants are lengthened and vowels shortened. In consequence, angry speech is characterized by a smaller proportion of vowel sound, in comparison with unmarked speech. A very interesting thing can be observed in the quantitative proportions of consonants and vowels within the analysed samples (the third column of the table). In the Elven languages – Sindarin and Quenya – a smaller number of consonants correspond to a single vowel (1.22 in Sindarin and 1.08 in Quenya), while in negatively charged Black Speech the proportion is greater – 1.7 consonants per one vowel. As far as the place of articulation is concerned, negative emotions cause withdrawal and positive emotions forwarding of the tongue. The vowels in Elven languages, in Tolkien’s own description, correspond approximately to the sounds represented in the English words machine, end, cut, for, brute. Out of these, /i/ and /e/ are articulated at the front of the mouth, / f: / and /u/ at the back and / % / in the centre. The most frequent vowels in Sindarin and Quenya are /e/ , / i / – front vowels, and / % / – a central one. Black Speech is dominated by / % / and / u / sounds, articulated respectively at the centre and at the extreme back of the mouth. So, in the elven languages the articulation of vowels takes place more at the front of the mouth, while Black Speech utilizes sounds articulated at the back of the mouth, with front vowels being quite marginal. The distortions of consonants in aggressive speech occur mostly in the articulation of fricatives (e.g. /f, s, z/) and plosives, particularly voiceless ones (i.e. /p,

Is lámatyáve a linguistic heresy? 109

t, k/), while sounds such as / l,j,n,m,w /, where the articulation is smooth and sonorous, remain relatively unaffected (Fonágy 1991:497). In other words, the sounds most affected by aggressive speech are those in which the passage of the air through the mouth is obstructed by friction or complete closure. The sounds that are not distorted are the unobstructed sounds, also called sonorants, in which the passage of the air through the vocal tract is relatively unhindered. The table below presents the percentage of particular types of consonantal sounds in the overall number of consonants within examined samples. Language 1. Sindarin 2. Quenya 3. Black Speech

voiced plosives / b, d, g / 10,9% 5,8% 29%

voiceless plosives /p, t, k / 3,6% 5,8% 19%

unobstructed / m, n, w, j, l / 52,7% 46,1% 20%

The disproportions in the number of plosives between the languages are significant: plosives, both voiceless and voiced, consitute almost half of the consonants in Black Speech and at the same time they are quite marginal in the Elven languages. In Quenya, stops appear almost exclusively in clusters with sonorants: sounds such as nasals /n, m/ or a semi-vowel /w/, smoothing their articulation. This phenomenon is so frequent that in the Elvish alphabet there are single letters for combinations such as /mb/, /kw/ or /nd/. In Sindarin voiceless plosives, particularly prone to emotive distortions, never occur after a vowel – they undergo mutation into their voiced counterparts, which are articulated less violently. What is quite prominent in Quenya and Sindarin is the large proportion of unobstructed sounds, relatively unaffected by emotive distortions: the sonorants / l,j,m,n,w /. They share between themselves about half of the consonantal sounds, while in Black Speech they comprise only 20% of consonants – less than half of the number of plosives. Generally speaking, the articulation in Black Speech is quite violent. Diphthongs are practically absent and due to the relatively small number of sonorants, the transitions between particular sounds are sharp and abrupt. Many sounds are articulated at the back of the mouth. On the other end of the linguistic-aesthetic spectrum there are the Elven languages, with their dominating front vowels, frequent sonorants occuring mostly as long consonants or in clusters, smoothing the articulation of plosives. Additionally, a large number of diphthongs make the transitions gentle and the speech sonorous and flowing. To conclude, Black Speech displays – at the level of phonology – effects similar to those characterizing – at the level of phonetics – speech distorted by aggressive emotions. It has a large proportion of consonants, particularly plosives, in relation

110

Joanna Podhorodecka

to other sounds, and the preference for back vowels over font vowels. The Elven languages, on the other hand, imitate the characteristic features of speech affected by positive emotions: a large proportion of vowels, with preference for front vowels, and a considerable number of sonorants.

4. Conclusion Tolkien described his imaginative writing as “an essay in linguistic aesthetics”(1981:220). His invented languages, as well as The Lord of the Rings, were thus an exercise of their maker’s lámatyáve, individual, and yet founded on paralinguistic universals of pre-conceptual communication. Tolkien’s response to linguistic patterns resulted from his keen sense of phonetic metaphor, making his own invented languages contribute in a very subtle way to the aesthetic and axiological aspects of his mythology.

References Carpenter, H. 1992. J.R.R. Tolkien. A biography. London: Grafton. Fonágy, I. 1991 Paralinguistic universals and preconceptual thinking in language. In New Vistas in Grammar: Invariance and variation, L. Waugh and S. Rudy (eds), 495–515. Amsterdam: John Benjamins. Fonágy, I. 1999. Why iconicity? In Form Miming Meaning: Iconicity in language and literature, M. Nänny and O. Fischer (eds), 3–36. Amsterdam: John Benjamins. Shippey, T.A. 1983. The Road to Middle-earth. Boston MA: Houghton Mifflin Company. Tolkien, J.R.R. 1981. The Letters of J.R.R.Tolkien. Carpenter, H., (ed). Boston MA: Houghton Mifflin. Tolkien, J.R.R. 1991. The Lord of the Rings. London: Grafton. Tolkien, J.R.R. 1996. The Peoples of Middle-earth. London: HarperCollins. Tolkien, J.R.R. 1997. A Secret Vice. In The Monsters and the Critics and Other Essays, 198–223. London: HarperCollins.

The beauty of life and the variety of signs

part iii Iconicity and the visual

111

112

Peter Gahl

The beauty of life and the variety of signs

The beauty of life and the variety of signs Guillaume Apollinaire’s ‘lyrical ideogram’ La Cravate et la montre* Peter Gahl University of Konstanz

This article offers a close reading of Guillaume Apollinaire’s ‘lyrical ideogram’ La Cravate et la montre (The Tie and the Watch, 1914): an analysis of the poem’s specific form of interaction between words and images which shows how the author probably intended it to be read, and a systematic approach to the semantic structure. ‘Time’ and ‘the human body’ are identified as the main isotopies and ‘life’ vs. ‘death’ and ‘linear’ vs. ‘circular’ conception of time as the most prominent oppositions, the latter being reflected also by the poem’s graphical structure which ‘quotes’ Marc Chagall’s Hommage à Apollinaire. A second important topic is the co-presence of different modes of signification – iconic and symbolic – which corresponds to Apollinaire’s metasemiotical reflections of that period, namely to his conception of ‘surréalisme’.

1. Introduction Not too many of the contributions to past volumes in this series have considered visual poetry in general and even less – so far as I can see: none at all – have dealt with Guillaume Apollinaire’s ‘Calligrammes’ or ‘lyrical ideograms’, as he himself termed them. This may be due to the apparent lack of challenge: no particular interpretative skills are required to reveal the iconic element, seemingly a less subtle variety of ‘form miming meaning’.1 However, in the specific case of La Cravate et la montre (The Tie and the Watch), things are far from being as simple as that, and this is what the following analysis and interpretation of the poem will try to show. La Cravate et la montre was written in early 1914. It first appeared together with three other visual poems in the last issue of Les Soirées de Paris only a few weeks before the outbreak of World War I; later it became a part of the volume

113

114

Peter Gahl

The beauty of life and the variety of signs

Calligrammes (1918).2 Since Apollinaire published his first ‘lyrical ideograms’, their artistic value and originality have often been questioned. In fact visual or shaped poetry has a long history, which goes back to the technopaignia of the hellenistic age. The genre was particularly popular in the sixteenth and seventeenth centuries. Many of the renaissance and baroque carmina figurata deal with religious matters, whereas numerous others are purely ornamental divertissements; one of the best-known poems of this kind being Charles-François Panard’s oftenquoted Que mon flacon / me semble bon…, an ode to the comforting effect of wine, aptly disposed in the shape of a bottle.3 When Apollinaire’s first Calligrammes appeared, more than one reader considered them a rather pointless revitalisation of an outmoded facetious tradition. More recent critics have suggested instead that these texts were inspired by the Free-word tables by futurist artists such as Francesco Cangiullo or Ardengo Soffici, whilst others regard them as a reaction to contemporary works of cubist painters who inserted single letters or fragments of text in their canvasses. But such comparisons have rather poor explanatory value, for Apollinaire had no doubt a great ability to synthesize various influences and create something new and original on the basis of inspirations drawn from artists of his own age as well as of the past. In his volume Reading Apollinaire, published in 1987, Timothy Mathews mentions a “lack of critical interest” (Mathews 1987:161) with regard to Apollinaire’s ‘lyrical ideograms’. Although some substantial work has been done since then, most contributions deal with these poems as a whole rather than offering closer readings of single texts. Those few critics who have worked on La Cravate et la montre in particular have supplied some lucid and interesting interpretations, albeit concentrating mostly on the verbal message and paying too little attention to the specific form of interaction between words and image, which will be a major focus of the present essay. I shall deal with three main questions: What does the poem’s structure tell about how it is meant to be read and looked at? What is the poem about? And finally, how does it signify, how is its meaning constituted?

2. Words, images, and their perception Let us begin with a closer look at the structure of the text, and by considering the question of how the author wants us to ‘read’ it. Like every traditional carmen figuratum, La Cravate et la montre is made up of words that at the same time form a text and create the visual image of an object. That is to say, in this poem we have two distinct sections, two different stanzas or drawings, one representing a tie (rather wide and short, by present-day standards), the other an old-fashioned pocket-watch. In fact, most of Apollinaire’s Calligrammes show more than one

115

116

Peter Gahl

image, and this is a noteworthy difference from most of the traditional technopaignia.4 Just like La Cravate et la montre, several other titles of Apollinaire’s visual poems consist merely in the naming of the two or three depicted objects.5 Many of these combinations may seem at first somewhat gratuitous – we will see that this is not at all the case for La Cravate et la montre. How is the verbal message structured? Whereas the tie is formed by one single sentence, the watch is slightly more complicated and heterogeneous: we have a brief exclamation in the winding-crown – or is it a ring? –, two slightly longer ones, one of them in the shape of a semicircle forming the right rim, the other one distributed into two half-verses forming the two clock-hands, and finally we find a series of twelve noun-phrases of different length in the positions of the twelve figures of the clock-face. There is, however, a substantial difference between these ‘figures’ and the other elements, for the former do not represent their object iconically, by imitating the shape of the figure in question, but by verbal paraphrase: “mon cœur” stands for the number one, for one has only one heart; “les yeux” is obviously for the two, while others are less evident and some – most of all, the ‘figure’ 4, “Agla” – have proved to be truly challenging, even though, nota bene, each of them is identifiable from its position.6 So this series of phrases is not subject to the principle of the carmen figuratum. We will return to that later on. It is worth noting that the ‘reader’ (as in any technopaignion) will usually recognize the two depicted objects at first glance, before even starting to decipher the words they are made of.7 The verbal message, on the other hand, is harder to read than a ‘non-visual’ poem: you have to determine where to start, and then you have to find the path through it, which can in some cases be fairly complicated. Actually, the sequence of words is always somewhat delinearized. In our poem, the tie section is not really challenging: one starts at the highlighted words at the top, and then reads it syllable by syllable, maybe a bit more slowly than if it was printed in a single horizontal line. The order of the elements, however, is obvious. In the watch stanza, it is not: most readers will start with “Comme on s’amuse bien”, but then the decision where to continue is always arbitrary: first the figures; or first “La beauté de la vie…?”, and the reader will probably be conscious of that. And this holds even for the succession between the stanzas: we are likely to consider the ‘tie’ first, but the order of the two parts is not as determined as in a linear text.8 Looked at in this way, the lyrical ideogram seems to be something like a more radical version of a conventional poem set out in lines. A fragment of prose – apart from its possible division into chapters and paragraphs – is basically a single string of words, for mere convenience printed in different lines.9 A poem, on the other hand, introduces pauses and turning-points by means of the line breaks, according to the author’s decisions or to a given metrical pattern – you can tell a sonnet from a limerick at first sight. In the technopaignion we also have a disposi-

The beauty of life and the variety of signs

tion of the words according to an external structure, with the important difference that now this structure is meaningful in itself. The reader of such a poem, even when concentrating on the words, will therefore hardly forget that they also form an image. On the other hand, when he directs his attention to the picture, he will keep in mind that it is made of letters and words. Besides, it is schematic and oversimplified, compared to a normal drawing. In addition, most of Apollinaire’s visual poems – and La Cravate et la montre is no exception – definitely lack unity, insofar as they show more than one image: the depicted objects often appear as merely juxtaposed, they remain isolated from each other and do not form one coherent image (a fact due, last but not least, to the notable difference of scale). In any case, the reader of La Cravate et la montre cannot but focus attention on only one of the two levels of expression: he will inevitably either ‘read’ or ‘look at the picture’, even though always conscious of the poem’s hybrid nature. Technopaignia are thus characterized by a superimposition of the iconic and the symbolical mode of signification: the graphemes become also ‘iconemes’, meaningful in their sheer materiality as marks of ink which compose an image. To be more precise: in La Cravate et la montre the smallest graphical unit – the iconeme, what in painting would be the single stroke of the brush – is not the letter, but the syllable. In the minute-hand of the watch, where two syllables claimed by semantics would have been too long (or too large) to fit into the silhouette, the author replaces them by the mathematical symbol ‘-’ and the figure ‘5’.10 There are, however, two exceptions to this principle: the letters of “LA CRAVATE” which as such form the shape of the tie’s upper part, and the syllable “bien” of the watch’s winding crown which is split into halves. Note that the author makes use of different letter sizes (“La Cravate” vs. the rest of the text), of capitals (in the entire tie-stanza and in one part of the watch) and even of different typefaces: the exclamation “Come l’on s’amuse bien” is set in a sans serif font, unlike the rest of the poem. Syllables (which in La Cravate et la montre become iconemes) are usually the units of counting in French versification. But do we have anything like a metrical structure in this text? Considering also that it is not possible to read it aloud, at least without sacrificing the iconic component? Of course, not on the whole, but there are fragments where this makes sense: the first six ‘figures’ form a perfect Alexandrine line, as do the two clock-hands. Even though the poem’s two parts, the tie and the watch, seem somewhat isolated one from the other – an effect which is enhanced by the aforementioned typographical device –, some structural connections between them can be found. One is the anaphorical use of ‘douleur’ – the lexeme appears in the tie-stanza as well as in the statement forming the right rim of the watch. Another – more striking

117

118

Peter Gahl

– element is the similarity between the tie’s two ribbons and the two clock-hands which Alain-Marie Bassy (cf. Bassy 1973–4:189) has long ago noted: they show the same angle, just upside down. Maybe we could call this a visual rhyme, because it can be seen as perfectly analogous to rhymes in versification: the formal similarity (but not identity) of two signifiers in different but parallel positions which does not affect the level of their meanings. In addition, this twofold 30° angle is also a kind of quotation, as Kurt Raible (cf. Raible 1972:94) has pointed out: the same element also appears in a painting by Marc Chagall, aptly titled Hommage à Apollinaire (1911–12). The canvas shows Adam and Eve (distinguished by the famous apple) growing out of a single abdomen and a pair of legs they have in common, and in the background a sort of sketched section of a clock-face can be seen. The analogy between Apollinaire’s poem and Chagall’s Hommage becomes even more evident when we look at a preparatory sketch for the painting, which can be found in Meyer (1961:152).

3. Life, death, and renewal Let us now have a closer look at the semantic level. Again, we have to be aware of the double nature of Apollinaire’s poem and take account of the words as well as of the images. In an essay on Magritte, Michel Foucault claims that “the calligramme is tautology” (“Le calligramme est donc tautologie”; Foucault 1994:638), thus suggesting that the visual and the verbal level convey basically the same meaning. Of course he is referring mainly to traditional technopaignia like Panard’s Que mon flacon... quoted above,11 but can a visual image and a verbal assertion really ‘mean’ the same? This is far from being straightforward. However, in La Cravate et la montre Foucault might be right when we look at the clock-hands, for the longer one is formed by a description of what we see: “it is five minutes to...” – “Il est – 5 enfin”.12 But the words which on the visual level form the image of a tie syntactically compose the exhortation to take off this annoying garment. We usually teach our students to approach a modern lyrical text by searching for the main isotopies and oppositions which constitute the text’s semantic structure. In other words: which – if any – are the salient series of elements which seem to be related to the same topic or opposed to each other in a significant way? There is no reason why this method should not be applicable to our text. Two main isotopies seem to recur here: time and the human body. Time is evoked by the watch as such, by the message of the hands (‘five minutes, and all is over’), but also by the ‘figures’ 7 and 12 which name units of time.13 Even more elements are related to the human body and its basic functions: the ‘figures’ 1, 2, and 5 which can be interpreted respectively as symbolising the human feelings, the senses and

The beauty of life and the variety of signs

the human activity, three central aspects of life in general and of artistic creation in particular. The tie stanza mentions respiration, the ‘figure’ 3 refers to procreation. Then we have the ‘nine doors of your body’ which can also be found in two poems (both written in 1915) dedicated by Apollinaire to his war-time fiancées and which make clear that this is to be understood as an allusion to sexuality.14 This subject is also evoked by the ‘figure’ 6, “Tircis”, which contains an erotic pun included in an anonymous 19th century Salade mythologique (“J’en ai Tircis, c’est Baucis, mais je ne puis Alexis...”). The two main isotopies – time and the human body – are linked to each other by the sentence on the watch’s rim, which is a declaration of touching simplicity: “the beauty of life surpasses the sorrow of dying”. Here we have a semantic opposition of central importance, perhaps the most elementary one, that between life and death. We have seen that the vital functions are distributed all over the poem, but we also find the evocation of death: the ‘figure’ 9 contains the adjective ‘cadavérique´ and mentions Dante, author of a famous epic description of the other world. More importantly, we have the sinister message of the clock-hands with its insistence on the lexeme ‘fin’: everything is bound to end soon. And if we think that the declaration which opposes life and death is a simple one, we are, alas!, mistaken, for the typographical distribution of the words makes us read the sentence syllable by syllable, which brings out its inherent ambiguity: read as a whole, it declares the victory of life’s beauty over death; if we read only the first eight syllables, the message changes: “la beau-té de la vie passe” – ‘the beauty of life passes’. This ambiguity remains unresolved, either message stands there in its own right.15 The ‘figure’ 8 deserves some closer consideration, for it implies another basic semantic opposition. By ‘erecting’ the mathematical symbol of the infinite (∞), i.e. by turning it by a right angle, thus obtaining the grapheme ‘8’, a ‘mad philosopher’ (whose identity remains unspecified)16 changes the infinite into something finite. If we call the elements of the ‘time’ isotopy back to our minds, we now see that they refer to two different conceptions of time, one linear and finite, the other cyclic and potentially infinite. The succession of the hours of the day or of the seven days of the week is essentially cyclic, the idea that all will be over in five minutes is linear. Human life, on the other hand, is linear and finite, too (and the text has the courtesy of reminding us of this rather sad fact), but it is equally true that the succession of the generations evoked in “l’enfant” implies a cyclic renewal of life. We should at this point return briefly to Chagall’s Hommage à Apollinaire: the painting refers to the story of Adam and Eve as we know it from the Book of Genesis (a story to which ‘semaine’ alludes). But why does the artist place Adam and Eve in front of a clock-face? According to Franz Meyer’s (1961:161) most convincing interpretation, he does so because he wants to suggest that the fall of

119

120 Peter Gahl

man and the loss of innocence also meant the beginning of time. In fact, the idea of a linear succession of time which runs towards an end also implies the idea of a beginning, which we can situate in this Edenic moment. And we now see what it is that links the two objects depicted in Apollinaire’s poem, the tie and the watch, with each other: not only are both components of a rather elevated male dress code, they both have to do with this very moment in paradise when time and civilisation (and the history of clothing!) started. The watch is an instrument for the measurement of time, and the tie is presented to us as a purely ornamental symbol of civilisation (which we ought to take off from time to time in order to breathe more naturally), an equivalent of the famous biblical fig leaf Adam and Eve covered their pudenda with. In Chagall’s painting, the leaf does not appear, but Adam holds his left hand before his member – forming an acute angle analogous to that of the tie in Apollinaire’s calligram. The watch, on the other hand, shows an even stronger similarity of shape to the apple from the Tree of Knowledge (with peduncle and leaf) as depicted in the Hommage. Thus the tie and the watch allude to two prominent elements of Chagall’s pictorial representation of the fall of man, and the juxtaposition of these two objects is definitely not as arbitrary as it may have seemed at first. But what is still more important, the geometrical structure of Apollinaire’s poem reflects the fundamental opposition between two possible conceptions of time, the linear and the cyclic – and here it becomes iconic in quite a substantial way: We see in fact a basically linear object and a round one. And inside the watch, we have the clockhands being again basically linear and finite (and speaking of the end soon to come) and the winding-crown and most of all the clock-face being circle-shaped – a fact which is underlined by a linguistic device, the rhyme between “les heures” and “Mon cœur”. So we have an iconic representation strictu senso of the two different conceptions of time by the salient features of the represented objects. Moreover, these two conceptions are here represented as parallel and perfectly free of hierarchical order, as alternative ways of looking at time, none of which has more dignity or more justification than the other. And only in a visually figured poem like this one is such an effect possible. Nearly all of the essays on La Cravate et la montre focus on the watch section, sometimes almost entirely neglecting the tie. Pénélope Sacks-Galey (1988:111) claims explicitly what several other authors seem to tacitly assume: “The verbal expression of the watch […] contains, so to say, the heart and the soul of the calligram. […] The tie, on the other hand, does not enrich at all this poetics, nor does it open the poem to new implications”.17 The results obtained so far in the present essay show that this line of interpretation is one-sided and reductive, mainly because it does not take into account sufficiently the calligramme’s inherent geometrical complementarity between linear and circular forms, which

The beauty of life and the variety of signs

has proven to be highly meaningful. Apart from that, we have seen that the tie stanza does indeed contribute significantly to the poem’s richness, in particular by both verbal and visually imaged allusion to the biblical fig leaf.

4. Metasemantical aspects From a semiotic point of view, the poem’s most interesting feature is probably the series of expressions for the figures from 1 to 12. As we have already seen, these somehow evade the technopaignion principle, for they do not form an image or a part of it. Moreover, they do not just name the numbers but somehow allude to them, more or less cryptically.18 They are not puzzles in the normal sense that you have to guess what they mean, for this is clear regardless of anything else from their position on the clock-face. The problem is rather of understanding why one phrase or another represents a certain number. And this is exactly what the text invites us to reflect upon. Some are straightforward, others are not – as if Apollinaire had wanted to demonstrate that language in itself can be both clear or obscure, with plenty of shades of difficulty in between. In two cases, the phrase alludes to the shape of the figure (the eight, for reasons we have seen, and the ten, a capital X in roman figures); in others, to a cultural fact of more or less common knowledge: seven days make one week, we usually count the hours from one to twelve, there are nine muses, and Dante’s Divine Comedy is entirely composed of hendecasyllables. In more than one case, we even have an accumulation of different forms of relation between signifier and meaning: we have two eyes, but “les yeux” also rhymes with “deux”; a week has seven days, and the word “semaine” has seven letters. “Tircis” is based on partial homonomy, but again the number of letters, six in this case, corresponds to the figure in question.19 And the ‘figure’ eight is represented by a phrase which consists of eight words, the nine by one of nine syllables. Those latter cases constitute an interesting phenomenon of diagrammatic iconicity: form miming meaning; the meaning here being a number, and this being conveyed by a number inherent to the expression’s form. In no other of Apollinaire’s Calligrammes is the mechanism of visual poetry similarly contaminated with ‘non-figurative’ elements (the purely verbal signifiers representing the clock’s figures) – a feature which makes La Cravate et la montre quite unique.20 Moreover, with this combination of different modes of signification Apollinaire seems to anticipate a thought developed three years later, in his preface to Les Mamelles de Tirésias, subtitled “Drame surréaliste”. In this play, strange characters appear and bizarre things happen, so in fact it can’t possibly be regarded as ‘realistic’. This is how the author explains what he means by his newly coined term of surréalisme:

121

122

Peter Gahl

In order to characterize my play I have made use of a neologism, coining the adjective ‘surrealist’. […] I have thought that one must return to nature in itself, but without imitating it like photographers do. (“Pour caractériser mon drame je me suis servi d’un néologisme […] et j’ai forgé l’adjectif surréaliste […]. (J)’ai pensé qu’il fallait revenir à la nature même, mais sans l’imiter à la manière des photographes”; Apollinaire 1965:865)

Unfortunately, this definition is basically negative: it is difficult to imagine how this ‘return to nature, but without imitating it’ is intended. To make things clearer, Apollinaire resorts to an example: When man wanted to imitate walking, he created the wheel, which does not resemble a leg. By doing so, he did surrealism without knowing it. (“Quand l’homme a voulu imiter la marche, il a crée la roue qui ne ressemble pas à une jambe. Il a fait ainsi du surréalisme sans le savoir”; Apollinaire 1965:865f.)

The most interesting aspect of this image is the fact that Apollinaire stresses the question of similarity. This proves that his concept of a ‘higher realism’ (‘sur-réalisme’) has to be read semiotically: the significant difference between the wheel and for instance a pair of stilts is that the wheel does not imitate legs in shape and therefore does not resemble them. The invention of the wheel can thus be compared to the passage from icons to symbols (whose relation to possible referents is no longer based on similarity). Now, the words on the clock-face of La Cravate et la montre operate in quite the same way: they serve as signs for the figures without resembling them in shape, unlike the other elements of the poem. So they belong to another category of signs: the various components of the drawing are ‘realist’ (in the sense in which Apollinaire understands this term: with a good dose of indulgence they imitate nature ‘à la manière des photographes’), the series “Mon cœur / les yeux / ...” is ‘surrealist’, semiotically analogous to the wheel which imitates the act of walking without bearing any similarity to the leg. It is not only the poem we have looked at so far which shows Apollinaire already concerning himself with semiotic questions three years before formulating his idea of a surréalisme: the last issue of Les Soirées de Paris (the same in which La Cravate et la montre appears) includes an article signed Gabriel Arbouin and bearing the title Devant l’idéogramme d’Apollinaire.21 For a long time this signature has been generally accepted as a pseudonym of Apollinaire’s (note the identical initials), but in his Introduction to the English translation of the Calligrammes Lockerbie confutes this point of view (cf. Apollinaire 1980:10). Yet, even if Apollinaire is not materially the author of the text – which in fact differs notably in style from his other essays – there can nevertheless be little doubt that it expresses basically his own position.22 The article begins by claiming that Apollinaire’s ‘ideograms’ – and the first

The beauty of life and the variety of signs 123

one, Lettre-Océan (Ocean Letter) had just appeared in the previous number of the Soirées de Paris – mark a revolution in poetry, for these fragments of language which form the image are no longer linked according to the logic of grammar, but to that of an ideographic logic which leads to an order of disposition in space which is exactly the opposite to that of discursive juxtaposition – (“le lien entre ces fragments n’est plus celui de la logique grammaticale, mais celui d’une logique idéographique aboutissant à un ordre de disposition spatiale tout contraire à celui de la juxtaposition discursive”). (Arbouin 1914:384)

Poetry of this new kind demands a specific form of reception, which the author defines as ‘synthetic-ideographic understanding’.23 This thought seems to quote and elaborate elements borrowed from the Technical Manifesto of Futurist Literature (1912), in which Filippo Tommaso Marinetti introduces his poetics of the ‘words-in-freedom’ (‘Parole in libertà’). Marinetti, too, postulates the abolition of the traditional logic of concatenation, still preserving however the text’s linear structure – the ‘free-word tables’ (‘tavole parolibere’) were still to come, and they were created by other artists. And Marinetti also speaks of a new and essentially different modality of perception which he characterises as “perception by analogy” (“percezione per analogia”) claiming that this is engendered by recent technical achievements, namely the invention of aircraft: “as aerial speed has multiplied our knowledge of the world, the perception of analogy becomes ever more natural for man” (Marinetti 1972:85).24 Thus, one obvious difference from Arbouin/Apollinaire’s position is that the leader of the Futurismo movement believes this new and a-logical kind of perception to be a consequence of changes of the outside world, whereas the French essay regards them as produced by intraartistic innovations. In overt opposition to Marinetti’s ‘modernolatry’, Devant l’idéogramme d’Apollinaire denies that this ‘nature of human existence’ could never have existed before: “Has this kind of human intelligence never existed? Yes, it has. It even seems to have characterised the first phase which ended with the invention of alphabetical characters” (“Est-ce que cette nature d’intelligence humaine n’a pas existé? Si. Elle semble même avoir constitué la première phase à laquelle mit fin l’invention de l’écriture alphabétique”; Arbouin 1914:384). The text therefore partially withdraws the initial claim and states that “attempts like Apollinaire’s published in Les Soirées de Paris do not constitute a revolution but rather a regression” (“des tentatives comme celle d’Apollinaire dans les Soirées de Paris représentent non précisément une révolution, mais bien une régression”; Arbouin 1914:384f.), for they lead us back to a primordial phase of human culture, thus suggesting that the Calligrammes resemble somehow the first graphemes – which are, as we know, iconic in nature – before they developed to those plainly abstract and conventional or, in Peircean terms, symbolic signs. Again, as

124 Peter Gahl

in La Cravate et la montre, the opposition between iconic and non-iconic signs becomes a central issue.25

5. Conclusion We have seen that La Cravate et la montre, in spite of the apparent simplicity of some of its elements, reveals a high degree of semantic complexity. A central feature of the text is the simultaneous presence of two different aspects of human life and of two mutually contradictory conceptions of time. The second result of the present analysis is that the fusion of poetry and drawing is far from being as ingenuous as Apollinaire’s declaration ‘I, too, am a painter’ (“Et moi aussi je suis peintre”) may suggest. In the pre-war period Apollinaire was – more than ever before or afterwards – intensely concerned with aesthetic and semiotic reflections, as texts like Les Peintres cubistes, Zone or the ‘conversation poems’ prove. And this holds, too, for La Cravate et la montre proves the author’s awareness of the specific characteristics of iconic and symbolic signs and their different modalities of perception. And finally, we can now see more clearly what forced Apollinaire to create his ‘lyrical ideograms’ (a choice which stirred the disapproval of readers and critics, given the doubtful reputation of the genre of shaped poetry): the possibility of quoting a work of visual art directly, without having to resort to the comparably reductive expedient of ekphrasis, as well as the possibility of combining different forms of semiosis and putting them into a fertile tension. And La Cravate et la montre explores these specific possibilities probably more than any other of Apollinaire’s lyrical ideograms.

Notes * Ι wish to express my gratitude to everyone who partecipated in the lively and inspiring discussion which followed the presentation of this paper on occasion of the Krakow conference, in particular to John J. White for his precious comments on an earlier version and to Robert R. Calder for his helpful revision of the manuscript. 1. In fact the obvious pertinence of Apollinaire’s shaped poems to the issue of iconicity persuaded the organizers of the fourth conference (Louvain-la-Neuve 2003) to choose one of them – the Lunettes (1917) dedicated to Léopold Survage – as an illustration for the symposium’s presentation on the internet. 2. An English translation by Anne H. Greet can be found in Apollinaire 1980:79. All quotations in the present essay follow this edition, unless otherwise stated. 3. The model for this and several other texts in the same vein is François Rabelais’ Dive Bouteille, which figures in the posthumously published Fifth Book of his Pantagruel.

The beauty of life and the variety of signs 4. The author himself underlines the significance of this element in an article published in Paris-midi (22th July 1914): “The solitary shapes by Rabelais and Panard lack expressiveness […], whereas the relations between the juxtaposed figures in one of my poems are as expressive as the words that compose it. And this at least, I think, is a new invention.” (“Les figures uniques de Rabelais et de Panard sont inexpressives […], tandis que les rapports qu’il y a entre les figures juxtaposées d’un de mes poèmes sont tout aussi expressifs que les mots qui le composent. Et là, au moins, il y a, je crois, une nouveauté”, Bohn 1986:47f., Bohn’s translation) 5. Other examples, all taken from the Calligrammes, are: Cœur couronne et miroir (Heart Crown and Mirror), La Colombe poignardée et le jet d’eau (The Stabbed Dove and the Fountain), and La Mandoline l’œillet et le bambou (The Mandoline the Violet and the Bamboo). 6. Interpretations of the whole series can be found in Chevalier 1966:65–67, Bassy 1973– 4:191–193, Raible 1972:92–94, Sacks-Galey 1988:111–115, and Nykrog 1989:117–119. 7. It is perhaps an interesting question to what extent this recognition is facilitated by the poem’s title – perhaps a suggestion for future research for our colleagues from the department of cognitive psychology. 8. In the first version of La Cravate et la montre the tie was not right but left of the watch. Probably this modification is due to the author’s intention to make the succession between the elements slightly less obvious. 9. The Latin adverb ‘prorsus’ from which the term ‘prose’ is derived, means something like ‘in a forward direction’, ‘right onwards’. The origin of the word ‘verse’, on the other hand, is the verb ‘vertere’, ‘ to change the direction’ e.g. of a plough at the end of the furrow (on a corn field) in order to start a new one. 10. The insertion of a mathematical symbol follows a suggestion given in Marinetti’s Technical Manifesto of Futurist Literature, and critics like Raible who favour the idea of a futurist influence on Apollinaire have not failed to point this out (cf. Raible 1972:92). 11. Even among the carmina figurata of the baroque age, though, many counterexamples can be found, in which the verses are arranged in the shape of a horse, a rose, a heart etc. for mere ornamental purposes. 12. Greet’s very accurate word-for-word translation “It’s – 5 at last” (Apollinaire 1980:79) does not render exactly the sense of the original. 13. Note that both are not related to some astronomical reality – unlike the day or the month or the year – but merely conventional. 14. Both are enclosed in private letters and published only posthumously (Apollinaire 1965:459 and 619 respectively). The second one, dedicated to Madeleine Pagès, bears a title which is nearly an exact quotation from our poem: Les neuf portes de ton corps – The nine doors of your body. 15. Several critics have pointed out the importance of the semantic opposition ‘life’-’death’, always preferring, however, one of the two elements. Antoine Fongaro regards the message of the winding-crown (“How we amuse ourselves!”) as “the calligramme’s true title” (“Cela constitue le titre même du calligramme”; Fongaro 1988:207), thus interpreting it as an appeal to enjoy life. Willard Bohn, on the other hand, states that La Cravate et la montre is “concerned with human mortality. Indeed it belongs to a traditional subgenre of visual poetry devoted to this very theme. La Montre [sic!] is the latest in a long series of poems, going back to the

125

126 Peter Gahl

Renaissance, that call on sinners to prepare themselves for heaven. Whereas traditionally these were composed in the shape of an hourglass, Apollinaire has chosen a more modern form.” (Bohn 1993:76) Either position tends to reduce the ambiguity and polyvalence made possible by the poem’s delinearized and therefore de-hierarchized structure. 16. Hartmut Heep proposes to read this “fou de philosophe” as Virgil, “who leads Dante through ‘Inferno’ and ‘Purgatory’ into the ‘infini’ ‘Paradise’” (Heep 1993:93), thus regarding the two ‘figures’ 8 and 11 as closely connected. It may be objected against this interpretation – in favour of which no evidence is put forward – that Dante and his leader separe already in the midst of the Purgatory because the Latin poet, who is not baptized, is not allowed to go further. 17. “L’expression verbale de la montre […] contient pour ainsi dire le cœur et l’âme du calligramme. […] La cravate, par contre, n’enrichit pas du tout cette poétique pas plus qu’elle n’ouvre le poème è de nouvelles implications.” 18. The notable difference in transparency between these ‘figures’ and the apparently simple statements of the rest of the poem have induced more than one critic to limit his interpretation to a comment on this series of elements. 19. Moreover, Tircis is also the common French name for the speckled wood (pararge aegeria) butterfly. The males of certain varieties bear six conspicuous specks on the upper side of the wings. 20. Only Cœur couronne et miroir, published in the same issue of Les Soirées de Paris as La Cravate et la montre, contains an element which can be interpreted as somehow similar: the mirror-section is made of letters which form an oval, representing the mirror’s frame, and in the midst of it, the author’s name appears, instead of the human face we are used to seeing when we look into a mirror. 21. This title is followed by the indication “(1)”, so obviously further considerations were intended to follow. 22. There is not much evidence about how familiar Apollinaire and Arbouin were in the period in question, but it is known that they met at the regular dinners of the collaborators of ParisJournal; both were present at the meeting of 5th of July 1914 and had exchanges of views already on preceding occasions (cf. Bohn 1985:5f.). 23. “(C)omprendre synthético-idéographiquement” (Arbouin 1914:384). 24. As I have pointed out in another occasion (cf. Gahl 2005:87), R. W. Flint’s translation – perception of analogy – is misleading. 25. This confirms that there is a categorical difference between La Cravate et la montre and traditional shaped poetry, which, according to Foucault, “pretends to erase playfully the oldest oppositions of our alphabetical civilisation: showing and naming, figuring and saying, reproducing and articulating, imitating and signifying, looking and reading” (“Ainsi, le calligramme prétend-il effacer ludiquement les plus vieilles oppositions de notre civilisation alphabétique: montrer et nommer; figurer et dire; reproduire et articuler; imiter et signifier; regarder et lire”, Foucault 1994:638).

The beauty of life and the variety of signs 127

References Apollinaire, G. 1965. Œuvres poètiques [Bibliothèque de la Pléiade 121]. Paris: Gallimard. Apollinaire, G. 1980. Calligrammes. Poems of Peace and War (1913–1916) (Trans. A.H. Greet, with an introduction by S.I. Lockerbie and commentary by A.H. Greet and S.I. Lockerbie). Berkeley CA: University of California Press. Arbouin, G. 1914. Devant l’idéogramme d’Apollinaire (1). Les Soirées de Paris 26/27 Aug. 1914): 383–385. Bassy, A.-M. 1973–4. Forme littéraire et forme graphique: Les schématogrammes d’Apollinaire. Scolies. Cahiers de recherche de l’École Normale Supérieure 3–4:161–207. Bohn, W. 1985. Sur la Butte. Apollinaire et Savinio. Que vlo-ve. Bulletin de l’Association Internationale des Amis de Guillaume Apollinaire 2(13): 5–9. Bohn, W. 1986. The Aesthetics of Visual Poetry 1914–1928. Cambridge: CUP. Bohn, W. 1993. Apollinaire, Visual Poetry and Art Criticism. Lewisburg PA: Bucknell University Press / London, Toronto: Associated University Press. Chevalier, J.-C. 1966. La poésie d’Appolinaire et le Calembour. Europe 451/452 : 56–76. Fongaro, A. 1988. Apollinaire poète: Exégèses et discussions. Toulouse: Presses universitaires du Mirail-Toulouse. Foucault, M. 1994. Dits et écrits 1954–1988, Vol. I, 1954–1969. Paris: Gallimard. Gahl, P. 2005. The beginnings of iconicity in the work of F. T. Marinetti. In Outside-In – Inside-Out, C. Maeder, O. Fischer, and W.J. Herlofsky (eds), 79–93. Amsterdam: John Benjamins. Heep, H. 1993. Apollinaire’s visual poetry: The case of La cravate et la montre. Dalhousie French Studies 24:87–98. Marinetti, F.T. 1972. Selected Writings (Trans. R.W. Flint and A.A. Coppotelli, ed. and with an introduction, by R.W. Flint). New York NY: Farrar, Straus and Giroux. Matthews, T. 1987. Reading Appollinaire: Theories of poetic language. Manchester: Manchester University Press. Meyer, F. 1961. Marc Chagall. Leben und Werk. Cologne: DuMont Schauberg. Nykrog, P. 1989. A la Veille du grand Adieu. Sur les ‘idéogrammes lyriques’ d’Apollinaire. Romanic Review 80(1): 109–123. Raible, W. 1972. Moderne Lyrik in Frankreich. Darstellung und Interpretationen. Stuttgart: Kohlhammer. Sacks-Galey, P. 1988. Calligramme ou écriture figurée: Apollinaire inventeur de formes. Paris: Minard.

128

Peter Gahl

Forms of restricted iconicity in modern avant-garde poetry 129

Forms of restricted iconicity in modern avant-garde poetry John J. White King’s College London

This paper explores some of the forms minimal iconicity has taken in European poetry. Covering the period from early modernism to the experiments of the Paris Oulipo group and recent concrete poets, it offers close readings of work by Man Ray, Christian Morgenstern, Ernst Jandl, Jiří Kolář, Harry Mathews, François Le Lionnais and Otto Nebel. Illustrating the inventiveness of minimalist iconicity, it shows that a reduced iconic repertoire can actually multiply the interpretive options available. Types of complexity analysed include: the interplay between visual and acoustic iconicity and instances of unstable referentiality in examples of form miming meaning and form miming form and hybrid combinations thereof. As visual and acoustic iconicity is superseded by diagrammatic forms, a process of intellectualization is shown to set in.

1. Introduction In “A Cubist Historiography”, in semiotic terms the most rewarding section of her study The Colors of Rhetoric, Wendy Steiner addresses the possibility that certain kinds of poetic experimentation could lead to a “denial of the semiotic status of the concrete work of art” (Steiner 1982:197). The subtext here is that while the discipline’s foundational taxonomies were designed to categorize all signs in terms of their relationship to an “object”, some modern literary experiments are characterized by something that appears to verge on an abrogation of referential function. Concrete poetry, in particular, creates poetic contexts in which iconic, indexical and symbolic modes of sign-object relation (CP2:277) are more or less absent and hence even the less differentiating distinction between arbitrary and motivated signs (Saussure 1988:100) no longer seems applicable. Nevertheless, Steiner goes on to modify her original hypothesis to emerge with the more cautiously formulated possibility of a “near denial” of semiotic status and to speak of the “apparently non-semiotic properties” of certain experimental works (Steiner 1982:197). This allows for situations where even seemingly abstract works still invoke established poetic forms and conventions as well as occasions

130 John J. White

when radical poetry resorts to visual and acoustic iconic minimalism. With Steiner’s hypothesis in mind, I wish to explore a series of poetic phenomena situated in the disputed no-man’s land somewhere between the territory occupied by uncontentious examples of minimalist iconicity and works of “concrete art, rather than […] concrete reference” (Steiner 1982:197), the endproduct of which sometimes does seem perilously close to the achievement of a zero semiotic status.1 I intend to concentrate on two forms of iconic minimalism: (1) those involving entire works2 and (2) various local, often diagrammatic, iconic effects found in poems that are not in any overall respect mimetic, but which nevertheless resort to various innovative forms of micro-iconicity.

2. Restricted iconicity at the macro-level In May 1924, Francis Picabia’s Paris-based journal 391 published a concrete poem by Man Ray (Figure 1). On first encounter, most readers are likely to assume that the work (Ray 1924:119) is fully worthy of its generic title: “Poème optique”. In many respects Man Ray’s minimalist poem also rates as an example of “form miming form” (Nänny and Fischer 1999:174) or what some linguists would regard as a case of “endophoric iconicity” (Nöth 2001:23). It is an iconic artefact, not simply standing for a conventional poem made up of words, but also itself functioning as an abstract poetic construct. In order to guide its reception, the work evidently still requires an explanatory title in conventional language, but the two short dashes followed by one longer one at the top of this visual pattern also represent a title, one in the semblance of a code that the rest of the work employs. Man Ray’s sparse replica poem is divided into stanzas ranging in length from three to five lines, which in turn vary in length, the longest and shortest to be found in the first stanza. Indeed, for all its optical austerity, Man Ray’s poem is in other respects constructed according to familiar literary conventions. It is reproduced black on white, its component lines are still (despite recent advances in Futurist and Dada typography) horizontally arranged and centred on the page using generous margins. Moreover, the material is left-aligned, although, as befits the lines of a poem, not right-justified. The work also pays homage to an overall linearity involving a beginning and an end. Yet while the French poet Eugène Edme Pottier had already, in 1887, blazed the trail by producing a Morse Code version of “Die Internationale” (see Thalmayr 1985:436), Man Ray’s visual poem is clearly not in Morse Code, though the indecipherable sequence of dashes still gives the impression of being the encrypted version of some conventional utterance. An alternative title proposed for the work (“Dada phonetic poem without words” (Adler and Ernst 1987:260)) places it unequivocally in a contemporary

Forms of restricted iconicity in modern avant-garde poetry

Figure 1. Man Ray: “Poème optique”

131

132

John J. White

Figure 2. “Fisches Nachtgesang”

context, perhaps implying that if it were to be declaimed in true Zurich Cabaret Voltaire fashion, the performers would find themselves having to emit appropriate long and short sounds to replicate the work’s visual pattern acoustically, just as Pottier’s earlier Morse poem had been performed with a combination of “doo”-sounds standing for dashes and “di”s for dots. Despite certain conventional features, Man Ray’s poem clearly goes well beyond the language-based experiments of the time, above all thanks to “a pronounced curtailment of its semantic potential” (Adler and Ernst 1987:260). A further corollary of such extreme minimalism is the fact that it is impossible to pigeon-hole the work using conventional criteria for distinguishing between visual and phonetic concrete poetry and even difficult to assign it to a specific hybrid form. My second illustration, “Fisches Nachtgesang” (Night Song of a Fish, written 1905) (Figure 2) is taken from Christian Morgenstern’s Galgenlieder (Gallows Songs, Morgenstern 1947:31). It is better known than my first, which was a rare excursion into the field of concrete poetry on the part of someone primarily known as a Dada artist and photographer. Because of its notoriety one can safely assume that Man Ray must have been familiar with Morgenstern’s pioneering example of visual minimalism when he created his own wordless ‘poem’. Morgenstern’s intriguing title puts the stress squarely on the acoustic aspect:

Forms of restricted iconicity in modern avant-garde poetry

The work’s title identifies the structure as being underwater music made visibly palpable, music not intended, in other words, for human ears […]. If “song” is language sung to music, then here the reference is to a submarine language that remains incomprehensible. [It has] no connotations, no words, merely signs for their absence. Perhaps signs for breathing, fish mouth open and closed[…]. (Neumann 1973:61)

“Fisches Nachtgesang” has been hailed as “the crowning achievement of all sound poetry, for it represents sounds emanating from a mute creature” (Liede 1963, vol. 2:292). In semiotic terms, what Morgenstern has created is a work of almost total acoustic isomorphism. This is in part achieved through an ingenious innovation that evokes the medium of silence by using two types of scansion sign normally employed to indicate long and short syllables within classical metrical feet, but here isolated as virtual superscripts in sequences deprived of any poetic language for them to quantify.3 In his early commentary, Leo Spitzer offered the canonical reading of Morgenstern’s filled-figure fish as composed of signs for long and short feet in poetic metre (Spitzer 1921:96–97); and one anthology (Thalmayr 1985:205) follows this reading by putting the poem in the section headed “Metrisches Schema”. Spitzer’s pioneering reading was also the first to present the equally bold thesis that these indexical-cum-iconic signs could also represent the fish’s closed or open mouth (ibid.). Yet while helpfully interpreting the scansion signs in this way, Spitzer makes nothing of the fact that they lack any text to quantify in metrical terms or even of the fact that the scansion signs are not organized in such a way as to suggest any recognizable classical metre. Readers with a pronounced sense of the absurd might wish to pursue the logic of Spitzer’s interpretation a few steps further. For if the lines of the poem are read as alternating between iconic signs for a closed and an open mouth, we find ourselves confronted by a further paradox (apart from the Kafkaesque equation of a fish’s “night song” with silence). Tracing the pattern of the poem’s alleged visual iconicity line-by-line from top to bottom, we discover that our fish does not open and close its mouth continually, but, rather, opens its mouth more frequently than it closes it and, paradoxically, even opens it when it is already open and closes it when it is closed, with the result that the mouth is more often open than closed. Evidently Spitzer’s original suggestion that the poem represents the changing shape of a fish’s mouth either needs reformulating or it must be rejected as too wishfully logical. An alternative solution would be to read multiple occurrences of the same scansion sign as signifying continuing silence or continued singing. Prolonged silences were sometimes indicated in this way in German Expressionist drama, with repeated dashes standing for extensive pauses, e.g. in the works of Reinhard Sorge and Georg Kaiser. But whether its mouth is open or closed – or open or closed for substantial periods of

133

134

John J. White

time – cannot detract from the fact that the fish in question always remains silent. The referential object of Morgenstern’s poem is, of course, not any fish we know; it is, as the Dadaists recognized, a proto-Dada fish! As such it serves as an illustration of that penchant for “visual paradoxes” that has been seen as a hallmark of concrete poetry (Steiner 1982:214). Focusing on the function of these orphaned scansion signs that make up Morgenstern’s text entails positing a situation where the most the reader’s “inner ear” can be expected to register is a sequence of silences of longer or shorter duration, in their turn separated by silences. What initially looked like a modest effect in fact creates a complex mode of semiosis involving iconicity (inasmuch as the contrasting metrical signs can signify open or closed fish-mouth), indexicality (“open mouth”-sign stands for “singing fish”) and symbolicity (on the grounds that scansion signs are part of a longstanding only partly motivated convention for signifying syllable-value in poetic metre). In other words, it is a textbook illustration of concrete poetry’s stressing “a conception of language as a dynamic interplay of signs and things rather than a stable system of names” (Steiner 1982:208). Although Morgenstern’s poem has been judged difficult to recite (Neumann 1973:61), the critic Hans Mayer recalls someone at his school who rose to the challenge: “a boy in my class was able to perform ‘Night Song of a Fish’ using his over-sized mouth. He was much admired and imitated” (Jandl 1973:190). The schoolboy’s party piece involved treating the poem as an acoustic score, thus referring the sequence of signs not just to the fish’s mouth, but also to that of the performer.4 Many readers are content to treat “Fisches Nachtgesang” as a paradoxically “silent” acoustic poem. Yet this by no means exhausts its potential. Spitzer also detected in the work’s expanding and subsequently diminishing shape an analogy to the poetic structure of German Minnesang (consisting of conventionalized opening, middle and closing sections (Spitzer 1921:96)). Others have interpreted the overall configuration not just as a minimalist gesture towards some familiar poetic structure or the genre suggested by the term “Night Song”, but also as an example of “form miming content”. Quite understandably, for it is surely not beyond the bounds of the imagination to see in the poem’s overall structure the outline of a fish or even its scales in the configuration of short scansion signs. This effect is more easily appreciated if one turns the page 90º anti-clockwise (according to Neumann 1973:61). Before casting my poor fish back into a sea of competing readings, I want to recapitulate by highlighting two features of the works looked at so far. The first is the fact that at issue is not whether we are dealing with visual or acoustic concrete poetry or a picture rather than a picture-poem,5 but the extent to which each work is a complex amalgam of iconic, indexical and symbolic elements. And

Forms of restricted iconicity in modern avant-garde poetry

second, the equally important point that many of the features considered are not givens, unequivocal properties there in the poem for all to see (or in one case not hear). Peircean semiosis seldom works that way. Rather, they are sign-aspects some readers will wish to foreground and others may choose to discount.6 A consequence of this is that, far from simplifying its material, iconic minimalism can even multiply the interpretive pathways available. While the macro-iconic economy of Man Ray’s and Morgenstern’s poems cannot be accommodated within the categories of acoustic or shaped poem, one could easily make a case out for both works being examples of the growing exploitation of ‘diagrammatic iconicity’ by avant-garde poets. In their study Signs in Use, the authors point to certain features of diagrammatic iconicity which in the present context have fruitful implications, starting with the fact that “images share sensory qualities with the objects they represent, while diagrams share relations and structures” (Johansen and Larsen 2002:38). Even such a Peircean distinction suggests that there is a prima facie case to be made for positing diagrammatic iconicity in our first two examples. Both are diagrammatically iconic representations of a poem, not because dashes of contrasting length or a combination of dashes and hook shapes automatically suggest a poem or a fish – they would be unlikely to – but because their overall sign-structure (title, stanza-like units, left-alignment, in the one case, shape, scaly texture, changing mouth-shape, in the other) indicates their object synoptically in much the same way as most diagrams do.7 Some diagrams represent specific objects, of course, whereas most function in a non-specific way. Maps and piecharts, to be helpful, must be specific, whereas the syllogisms of formal logic can often lay claim to universal validity. Likewise, both of the abstract patterns we have considered so far are essentially diagrammatic, not by virtue of standing for specific, existing poems or for a particular fish, but because they invoke generic templates in structural terms. In this respect, diagrammatic iconicity’s tendency towards abstraction and the generic makes the object of such experiments closer to type than token in Peirce’s sense (CP8:313), This feature is not unrelated to the way in which such works tend to make their component parts simultaneously understandable. Synoptic reception is also a feature of both of the following works involving diagrammatic iconicity: first, a combination of visual-cum-acoustic mimesis in a minimalist work by the Austrian writer Ernst Jandl, and second, a sophisticated fusion of diagrammatically shaped abstraction and acoustic iconicity in a more stylized concrete poem by the Czech poet and visual artist Jiří Kolář. Their diagrammaticity differs in as much as one relates to a specific geographical phenomenon, whereas the other appears to be non-specific (unless, that is, it is read, as it has been, as a pre-1989 Czech political parable). In “niagaaaaaaaaaaaaaaaarafelle” (Jandl 1966a: 78) (Figure 3), the compound

135

136

John J. White

Figure 3. Ernst Jandl: “niagaaaaaaaaaaaaaaaaarafelle”

noun that gives the poem its title is expanded elastically to create a visual silhouette of the Niagara Falls while simultaneously suggesting the volume of water rushing over the precipice and collecting to form the fragmented sequence “ra felle” in the whirlpool below. Jandl has taken a densely textured pictorial image (familiar from myriad paintings and photographs) and drastically reduced it to its crude outline form. The resultant poem combines visual iconicity with an intensely acoustic performative text. The loudness of Jandl’s legendary public recital of “niagaaaaaaaaaaaaaaaarafelle”8 added a further vivid dimension to the work’s dynamism, for whereas the printed poem’s visual delineation of the Falls creates a relatively frozen impression of cascading torrents of water seen from a distance, Jandl’s voice adds a powerful indexical-acoustic sense of the majesty of the Niagara River’s current as it sends the water surging towards and over the Horseshoe Falls’ precipice and thundering down to the rocks below. On the page the image begs to be registered synoptically, but in performance optical simultaneity is broken down into its constituent parts. As a result, the declaimed version introduces a temporal linearity that draws a parallel between performance time and the passage of the water over the Falls. Kolář’s “kapka” (Figure 4) consists of the Czech words pak (then) and kapka (drop). It evokes the sound of liquid, after repeatedly coalescing, falling in measured drops to form a pool on reaching the ground (Kolář 1971:32). The main diagrammatic feature of the resultant poem is the relationship between its vertical and horizontal parts. Starting top left, our eyes inevitably follow the downward

Forms of restricted iconicity in modern avant-garde poetry 137

Figure 4. JirZ í KolárZ : “kapka”

138

John J. White

movement of the Czech word for “then”, repeatedly split as “p-ak” to suggest the sound of drips of water individually landing with a plop. The transition from the repeated letters “p” and “ak”, positioned vertically above one other, but separated by a line-space, and the later vertical “pak” (now without a space) imitates the increasing acceleration of a falling object. One of the poem’s many structural peculiarities is its ironic refusal to resort to what would have been the more obviously iconic splitting of the word “kapka” (rather than “pak”) to signify the dripping sound. There may be more than one reason for such displaced iconicity. Clearly, the split nature of the originally monosyllabic word “p-ak” is physically better able to capture the rhythm of a globule of liquid gradually accumulating until its volume forces it to fall downwards, with the sharp “k”-consonant imitating the sound of the drop landing, in a way that the open “ka”-ending to “kapka” could not. Moreover, since the sound can also be thought to resemble the ticking of a clock’s second-hand, a link between measured time passing and water dripping is also established. A further irony derives from the fact that the word “kapka” in the title is deferred and only reappears in the lower part of the layout. Here the pool of liquid created by the accumulating repetitions of “kapka” is actually made from the individual drops, but at the same time, because the repeated instances of the word for “then” run into one another, the word for “drop” only appears when the individual droplets have merged into the greater verbal conglomeration at the bottom of the poem. A visual-temporal iconicity thus gives way to the static visual image of collected liquid, at which point vertical dynamism is replaced by horizontal, static word-pool. In ways more complex than those to be found in Jandl’s “niagaaaaaaaaaaaaaaaarafelle”, Kolář’s shaped poem brings together local visual and acoustic mimetic effects within an overall diagrammatic iconicity. Although both works thematize falling liquid, the sheer volume of water evoked in the Niagara Falls poem requires a more powerful acoustic effect than that of the sequential drip-drip-drip structure of Kolář’s “kapka”. The Czech work achieves its effect through multiple repetitions, whereas Jandl’s delineates the shape of the Falls through the splitting of the single word. In both works, the iconicity is still predominantly sensory, whereas in the examples that follow it will be more intellectual.

3. Diagrammatic macro-iconicity My next illustration (Figure 5), Harry Mathews’ “Liminal Poem” (Collective Oulipo 1981:11), can only be fully understood in terms of diagrammatic iconicity. Such a reading is no longer merely one of a number of available options, it offers the principal semiotic key, although the operative factor in this case is not the

Forms of restricted iconicity in modern avant-garde poetry 139

Figure 5. Harry Mathews: “Liminal Poem”

degree of abstraction or any overriding synoptic reception, but the ingeniously referential relationship between the work’s two halves – and even between its individual lines. It is from the outset obvious that the lozenge formed by the words displays neither the visual nor the acoustic iconicity encountered in the works looked at hitherto. Mathews’ poem consists of two halves (or matching isosceles triangles) standing in an equation-like relationship to one another. Technically, the work is an example of a generative paradigm invented by the French Oulipo group (the Ouvroir de Littérature Potentielle). The exercise is known within Oulipo as a “boule de neige” (snowball), though the way in which the two parts of “Liminal Poem” relate to one another is untypical of this playfully idiosyncratic

140 John J. White

genre.9 The minimum definition of an Oulipo “boule de neige” is “a poem the first line of which consists of a word made up of one letter, the second two letters and so on” (Oulipo 1 [1975], 107, cf. Collective Oulipo 1981:194–210). There exist a number of variations on the “snowball” method: for example, involving adding a syllable instead of a single letter of the alphabet to each subsequent line or working with larger units (whole words or complete stanzas) that systematically make each unit longer than its predecessor or by working with an incremental increase rather than the simpler letter +1 method. The rules of some variations stipulate that previously used vowels have to be retained in initial position, with further ones then being inserted from line to line. Sometimes the repertoire is subject to further work-specific limitations and on occasions, as is the case here, the initial additive procedure can then also be reversed to produce a “melting snowball”. Clearly, the iconicity of such a work is diagrammatic rather than imagic. “Liminal Poem” is not about a snowball; nor does the diminution in the second half refer to the image’s literary associations in any metaphor-oriented sense of diagrammatic iconicity. The relationship between the poem’s two parts is in thematic terms one between wish and fulfilment or hypothesis and illustration. Hence, there is a clearer motivation to Mathews’ “snowball” construct than is usually the case, though it operates at a high level of generality. The expanding contour of the first half of the poem, which describes the kind of avant-garde literature the poet longs to be part of, is subsequently counterbalanced by a diminishing shape intended to evoke the “aesthetic of formal constraint” (Motte 1986:10–12; Bénabou 1986:40–47 that has now been central to over four decades of Oulipo creativity. The glossary of technical terms at the end of Warren Motte’s anthology of Oulipo writings identifies the main creative paradigms experimenters have evolved, exercises entailing the imposition of chosen formal constraints on the conventional language repertoire, grammatical structure or text-layout.10 Of course, none of these methods is actually used in Mathews’ “Liminal Poem”, but the group’s various favourite processes are nevertheless alluded to iconically through the gradual diminution the construct undergoes to create the shape of the lozenge and (diagrammatically) by the difference between the top triangle and the bottom one. Appropriately, Mathews’ name occurs in the middle of the overall lozengeshape: i.e. he forms the pivotal point of transition between his expressed wishes and their coming to fruition through the ongoing work of Oulipo. The names of Oulipo writers (all well-known proponents of formal constraint) are set out in the lower half of the poem in a way that diagrammatically replicates the dominant Oulipian aesthetic of formal constraint through the gradually diminishing shape of the enclosing frame. The effect is achieved by having certain characters have their first names replaced by initials, later ones (Perec and Bens) being deprived of their initials and the final two being represented by initials only. Since the initials

Forms of restricted iconicity in modern avant-garde poetry

in question, FLL and RQ, refer to the founding members of Oulipo (François Le Lionnais and Raymond Queneau) they are logically positioned at the inverted fulcrum of the lozenge: i.e. the locus of the greatest degree of formal constraint, while at the same time implying that they are so famous that their initials are sufficient to evoke their status as Oulipo’s gurus. The word “Liminal” in the work’s title is, perhaps not coincidentally, a near-anagram of “minimal”.

Figure 6. François Le Lionnais (1958): “Poème basé sur la ponctuation”

Like Man Ray’s and Morgenstern’s wordless poems, another Oulipo work, Le Lionnais’s “Poème basé sur la ponctuation” (Poem based on punctuation, Figure 6) from the aptly entitled “Enchaînements et tentatives à la limite” (Constraints and radical experiments, Collective Oulipo 1973:174)) also relates wordlessness to familiar verbal contexts and hence still possesses an abstractly intertextual referentiality. The poem is made up of a progression of numbers, something to which the title, with its emphasis on punctuation, mischievously fails to refer. The difference between the repeated commas in the long line after the solitary colon and the semi-colons in the subsequent line is significant. Because of the punctuation, the pauses thus signalled between numbers 1 to 5 are shorter than those between 6 and 10, as if the poem were becoming increasingly more halting. The concluding minimalist joke depends on a subordination of punctuation to content, an operative feature of most conventional discourse, although in this case the sign-system is mathematical and can only be transformed into conventional language if one reads the numbers out loud: “un, deux, trois”, etc. (For full effect, the punctuation should perhaps also be read aloud à la Victor Borge.) The combination of the second full-stop with the greater gap – and hence longer pause – between numbers 10 and 12 brings the previous rhythm virtually to a standstill, as if which number should continue the sequence were uncertain: “12?” of course not, it must be “11”. But now that number cannot occur in its rightful place in the sequence because the number which should have been inserted after the long lines in fact breaks the established sequence to become the answer to the question and hence has to appear in twelfth position. At which point, the poem prudently comes to an abrupt conclusion with the visual equivalent of a joke’s punch-line. As

141

142 John J. White

was the case with our earlier examples, the relationship between the various parts of Le Lyonnais’s poetic exercise is essentially one of diagrammatic iconicity. The relationship of punctuation to numbers here serves as a striking de-familiarization of the conventional interaction between punctuation and words. The next stage in such an experimental progression would be Günter Eich’s poem “Satzzeichen” (Punctuation, Eich 1966:42) where the reader is confronted with a poetic layout consisting purely of punctuation marks minus any intervening words.

4. Iconic minimalism and referential complexity Despite its stringently reductive strategies, my final example is by far the most innovative and complex of those under review. It consists of a 4-page sequence retrospectively generated from a longer cycle entitled UNFEIG (UNCOWARDLY), a project the German Expressionist Otto Nebel worked on between 1924 and 1956.11 In the 1960s Nebel was re-discovered by the Oulipo group who were particularly interested in his use of the lipogram and his general concern with the materiality of language. The relationship of all lipogrammatic writing to conventional language is iconic in the sense that whereas standard language works with the full alphabetical repertoire, a modest lipogram would omit certain letters, severe ones, such as Nebel’s, might operate with only a third of those at its disposal. My focus will be less on the hilariously exploited lipogrammatic medium (on which see Jones 1982 and 1984) than on one minimalist transformational game played with the language, once the overall content of the segment in question has been decided on and the lines written. Looked at in isolation, the multi-shaded sequence that forms part of this work (Figure 7) might in some respects seem to recall the stark abstraction of Man Ray’s poem. But in place of a series of long and short dashes, we now find clusters of squares and disks arranged in a shape that mirrors a type of early modernist poem symmetrically positioned on a central axis. In Germany such a layout received a powerful impetus from the typographical experiments of Arno Holz, although the layout receives more iconic motivation here than it did in Holz’s Phantasus (published 1924, the year Nebel was working on the first draft of UNFEIG). Figure 7 displays the rudimentary outline shape of a poem, though this time, unlike Man Ray’s “Poème optique”, without stanzas. It also retains the horizontality of conventional poetic layout; but in this case (although the illustration cannot, for cost reasons, show this) simple black-on-white effects have been abandoned in favour of a palette of nine primary and secondary colours (reflected here, faute de mieux, by various graduated shades of grey). Unlike “Poème optique” and “Fisches Nachtgesang”, Nebel’s bold coloured pattern has been generated from his

Forms of restricted iconicity in modern avant-garde poetry 143

own pre-existent poetry. Once again, minimalism has been achieved by denuding the source-poem of words, letters of the alphabet, punctuation, syntax and thus ostensibly any semantic import. Seen in the context of Figure 8 and then Figures 9 and 10, however, the abstract coloured sequence reproduced in monochrome in Figure 7 is revealed to be merely the second stage in a process of de-semanticizing the original lipogrammatic material and superimposing on one another various further visual versions of it, a process which now needs to be examined in detail. Figure 7 is an encoded version of the “Runenring” (Ring of Runes) section of UNFEIG, sub-titled “Eine Neun-Runen-Fuge” (A Nine-Rune-Fugue). The poem works exclusively with words constructed from nine letters of the alphabet, here called “runes”, representing three vowels (E, I, U) and six consonants (N, F, G, R, T and Z). Unlike the early Scandinavian runes from the third century AD, Nebel’s invented ones stand in a much clearer relationship to our own alphabet; the ones in Figure 7 are still relatively recognizable iconic versions of familiar letters. This is easily tested by comparing the invented runes in Figure 8 with the text of which they are a version (Nebel 1960:3) and, for contrast, with real Nordic runes only meaningful to an expert.12 Even though the individual letters are thicker, squarer and more vertically structured than real runes, Nebel’s so-called “runes” still correspond recognizably to our own alphabet. (Employing the same poetic licence, Nebel’s conception of the work’s fugal structure is just as figurative.) Since all editions of UNFEIG are for economic reasons printed in conventional typography and it is only in the version on Figure 8 that a more primitive form of writing in what Nebel called “schwarzen Runen oder Lese-Zeichen” (dark runes or reading signs, Nebel 1979:146) is evoked, it would be best to posit a four-stage evolutionary process: from standard typography (Nebel 1960:3) to runes (Figure 8), then to their iconically minimal coloured equivalent (Figure 7) and finally to two further syntheses considered shortly. The most radical stage of the poem’s experiment comes with the transformation of the rune-stage (Figure 8) to a corresponding system of coloured shapes (Figure 7). Each of the nine runes has been colour-coded so that the abstract visual design is correspondingly made up of nine colours standing for the poem’s restricted alphabet of nine “runes”. (Even the colour-patterning is ‘lipogrammatic’.) The scheme is as follows: Rune Rune Rune Rune Rune

e becomes an orange disk i becomes a yellow disk u becomes a blue disk n becomes a grey square g becomes a red square.

Rune Rune Rune Rune

f becomes a green disk z becomes a black square t becomes a pale green disk r becomes a brown square

144 John J. White

Figure 7. Otto Nebel: UNFEIG

Forms of restricted iconicity in modern avant-garde poetry 145

Figure 8. Otto Nebel: UNFEIG

146 John J. White

Figure 9. Otto Nebel: UNFEIG

Forms of restricted iconicity in modern avant-garde poetry 147

Figure 10. Otto Nebel: UNFEIG

148 John J. White

Although the “runification” of the conventional printed word is intended to suggest some Ur-proto-shape underlying the modern alphabet and thereby to convey an aura of primitive wisdom, its subsequent transformation into coloured elements (elsewhere Nebel describes them as glass pieces in a mosaic, beads, pieces of crystal, and leaves on a language-tree (Nebel 1979:137, 139, 149)) goes beyond a mere revitalizing of language’s texture. It is an attempt at returning the word to its cultic origins. Despite the fact that, when viewed in isolation, the gradual transition from modern alphabet to rune to monochrome and coloured tokens looks like a pattern of primitivization, Nebel then moves on from pure colour to a modernistic form of black-on-white hieroglyphs (Figure 9) which take the visually iconic language beyond colour-coding to a more cryptic form of coding and, in the final version (Figure 10), he superimposes the material from the last three phases in a (once more coloured) sequence intended to transcend the earlier primal “languages” of the text. But there is more to the function of this cultic set of phased artificial sign-systems than the four stages of signification. For ultimately the material in Figures 7 to 10 is intended to be translated into the reader’s own body-language. The four stages we have been examining in printed form on the sober pages of a book did not begin their lives as such. They were inscribed – in some cases painted – on massive banners. In this form they had an even more cultic purpose than that of reprimitivizing typography and re-energizing the reading sensitivities of philistine modern man, referred to satirically in Nebel’s writings as “der verbildete Europäer, Fritz Krittelgrips” (the miseducated European, Fritz Krittelgrips, Nebel 1973:126). In a fictive conversation between a poet and a painter who are said to have created the banners, the poet describes what he and his friend tried to achieve (the analogy with Buddhist prayer-flags in what follows is deliberate). The basic purpose of the four banners, we hear, is to express the unsayable (or as Nebel puts it in his invariably idiosyncratic Expressionist German: “bestimmtes Unsagbares durch Entsprechendes bildhaft-sinnfällig auszudrücken” (to express something specific [yet] Unsayable by means of some metaphorical equivalent in terms meaningful to the senses): Your decision to express the primal dance of the nine UNFEIG-runes by means of a vivid evocation of the dance’s rhythms using a vast set of visual images on four massive word-banners […] resulted in a series of embryonic sequences of the Nine-Rune-Fugue being rendered more visible to the eye[…]. All four paper banners are equally large, […] each is one metre across and two and a half metres high. This is a crucial, vital piece of information, from which one can appreciate that you didn’t simply construe the physicality of the fugue-form as being that of a huge book, something that would appeal to both the human intellect and

Forms of restricted iconicity in modern avant-garde poetry 149

the eye, but rather as […] some vast spectacle, as the encounter […] with FOUR GIANTS – with a series of erect image-trees, arranged in parallel, constructed out of printed words and signs. (Nebel 1979:145–46)

The metaphors come thick and fast here. The four massive language-banners not only resemble Buddhist prayer flags, they are also GIANTS to whom the observer must look up in awe. They are at the same time word-trees, an image which not only relates to the impression of being tree-shaped which the work’s central-axis typography produces (with individual runes resembling leaves on a branch), but is expanded to include associations of word-trees reaching up to new meanings, language’s roots burrowing down into an etymological past and generative poetry’s creative growth from its basic source-elements (Nebel 1979:140). Treating these structures as a form of diagrammatic form-to-form iconicity would, I suggest, account for the shared quality of some of these living metaphors. For they all (banner, giant, tree) require the resultant writing to be apprehended as a single towering entity. Nebel’s word-banners were evidently intended as the focus of some primitive form of pagan worship. The purpose of these ritually displayed runes is to make the person encountering (not just reading) them become ecstatic to the point where his or her bodily stance gradually transforms from that of upward-looking worshipful beholder to ecstatic Dionysian dancer. This is how Nebel describes the intended effect: The observer with his eyes and ears pauses in front of each of the four banners in order to de-cipher the runic word-sound in the four different modes of presentation, each time depicted in such a way as to let him once more simultaneously experience the ORDER of the STRUCTURE as he raises his body up erect. When he has felt his way into the spirit of the four representations, the mystical stimulus that the encounter gives him, he will respond to the gestures of the runic words with his own body gestures and dance patterns provided that his body is sufficiently relaxed. Twice already what we call Chance has brought about a situation where deaf and dumb men have responded in an appropriate way to the four image-banners by breaking into dance. The one man […] at first merely moved his lips and tried to create sounds, but he then proceeded to perform dance steps in the space in front of the banners. The other man […] was catapulted into a state of violent physical excitement upon seeing the mosaic of coloured signs: he started to breathe more rapidly, he gesticulated with his hands and arms as well as moving his lips in the manner of oriental dancers. In the case of both of these deaf and dumb men, I would claim that they had grasped the authentic MEANING of the TRUTH visually revealed to them and had done so in a manner that lies beyond the REALM OF CONCEPTS. (Nebel 1979:146)

While it is clear that all modern philistine readers are figuratively “deaf and dumb”

150

John J. White

in Nebel’s eyes and that these two accounts are apocryphal, what is being described is not simply the impact of the word-banners on the person encountering them, but essentially the aesthetic underlying Nebel’s entire rune-fugue project. Instead of mere linguistic reduction, the end-effect was to be the total transformation of the beholder: It is a matter of leading the person seeking knowledge, by deploying the CREATIVELY UNIQUE, towards something un-familiar! That is: to lead him away from the long-familiar to such a degree that he is able, as it were of his own accord, both to recognize the UNIQUELY-NEW as a quintessentially primal value and also to acknowledge it as a fresh meaning. (Nebel 1979:175)

These images generated by patterns of extreme diagrammatic iconicity go some way to corroborating Nebel’s claim that in art the part is greater than the whole.13 In the modern world, diagrams are usually associated with rational explication and cognitive processes rather than with the cultic or with epiphany and mysticism. Yet it was a Polish semiotician, Mieczyław Wallis, who was one of the first to point to the connection between iconicity and magic (Wallis 1973:494–97). As my references to runes, cultic banners, sacred trees and primitive Dionysian gestures suggest, and as the first half of the twentieth century has reminded us with a vengeance, diagrammatic iconicity has by no means always possessed the enlightening capacities one might ideally expect of it. However, it is worth putting on record that Otto Nebel was, to his eternal credit, a vociferous pacifist during the First World War, a voluntary exile from Hitler’s Germany from 1933 onwards, a lifelong enemy of cant and philistinism and one of the most civilized of ambassadors of the avant-garde in later life.

5. Conclusion The discussion of “Fisches Nachtgesang” showed that there is still a tendency to see iconic effects in experimental poetry exclusively against the backdrop of the traditional shaped or picture-poem. As a result, the visual aspect often becomes privileged, with concentration all too often focusing on iconicity at a macroscopic level – i.e. usually involving synoptic reception – rather than paying attention to local iconic elements. However, like demarcations between indexical and iconic signs, the micro-/macro-level distinction can at times be something of a simplification. Even in minimalist works, local imitative visual and acoustic effects often appear within the framework of some larger iconic paradigm. And as our corpus suggests, they are sometimes not so much unequivocal features of a work’s content as potential associations. Hence, such works display a less stable

Forms of restricted iconicity in modern avant-garde poetry

referentiality. They challenge the reader to decide which sign-aspects (Liszka) or object-relationship (Peirce) to emphasize in any attempted reading. The tendency we observed towards abstraction and generic referentiality, in examples of both form miming meaning and form miming form, has on the whole led to an increasing intellectualism (in contrast to the senses-oriented shaped poetry of earlier periods) and to a greater interest in diagrammatic forms of iconicity in the fertile period from Dada to Oulipo. Although most of the material looked at in the present paper has in some sense been either minimalist or reductive, the result has seldom been simplification in a pejorative sense. Instead, iconic minimalism has the ability to create situations where a reduction of means leads, as Nebel observed, to textual diversification and enhanced complexity. As we have seen, semiosis as a consequence becomes more richly unstable, in contrast to the controlled sign-object relation in much non-literary discourse.

Notes 1. Another way of expressing this would be as a zero iconicity-rating in A. A. Moles’s proposed scheme for calibrating degrees of similarity (Moles 1972:52). 2. Literature’s equivalent of “iconic supersigns” in Martin Krampen’s sense (Krampen 1973: 116). 3. Neumann considers this possibility, only to reject it: “If, however, the metrical scheme of a verse of poetry is meant to be adumbrated in the ‘Night Song’, then, in any case, it remains that of a poem that cannot be filled with the words available to us. On the contrary, it is not a metrical scheme, but a symmetrical ornament” (Neumann 1973:61). All translations from German and French in the present contribution are my own. 4. Although Mayer does not appear to be aware of the fact, what the schoolboy in his class had unwittingly created was a sub-genre of performance poetry that Ernst Jandl would later go on to write and call the “visual lip-poem”: “the visual lip-poem is the reversal of the visual paperpoem, the reciter is the paper of the visual lip-poem. the visual lip-poem is spoken without intonation. it is written in the air by the lips. the untrained reader speaks the visual lip-poem in front of a mirror; with a trained reader the movements of the mouth will suffice to create the impression of the poem. whoever knows visual lip-poems by heart will never grow totally blind. deafness, muteness and deafmuteness are suspended by the visual lip-poem. those born blind are the only ones beyond its reach” (Jandl 1965:n.p.). On the score as a class of iconic sign, see Greenlee 1973:119 and Goodman 1968:200. 5. After an excursus on Baroque-shaped poetry, Neumann concludes: “‘Night Song of a Fish’ is situated on the borderline that divides image from picture-poem” (Neumann 1973:63). 6. According to Peirce, “nothing is a sign unless it is interpreted as a sign” (CP2:308). On some of the implications of this for semiotic literary interpretation, see White 1999:84–85 and 104– 05; and on the choice to emphasize certain sign-aspects rather than others, Liszka 1996:46. 7. Cf. the entry on “Diagrams” in Bouissac 1998:189:“Visuo-spatial representations of concepts and their relations, diagrams offer the opportunity to regard all parts of a system

151

152

John J. White

simultaneously, which is one key to their cognitive power. Simultaneity is a perceptual construct of consciousness”. 8. Recorded at the 1965 “Poetry International” reading at the Albert Hall London (Jandl 1966b). 9. The vast majority of Oulipo creative techniques work on a principle of serendipity. Various stochastic paradigms are devised to randomly generate new configurations of words and phonemes, often with little semantic coherence. Mathews’ “Liminal Poem”, by contrast, is rigorously organized around the idea of a new aesthetic and systematically identifies the proponents of this new way of writing. 10. The 55 strategies of “formal constraint” itemized in Motte’s Glossary (Motte 1986:197– 202) include: antonymic translation (the replacement of the words of a poetic text by their opposites), the chronogram (a language-game in which certain letters of the alphabet, when placed together, form a date in Roman numerals), haikuization (involving the retention of the rhyming parts of a pre-existing poem to stochastically create a new one), various sub-categories of palindrome, the application of the S + 7 formula (whereby words in a text are replaced by the word coming seventh after them in the dictionary) and the tautogram (a work with all words beginning with the same letter). 11. Nebel made a recording of UNFEIG at his home in Berne on 8 April 1970 (Nebel 1970). 12. Nebel’s eccentric conception of the “rune” is explained in “Einweisende und abschließende Worte über Sinn, Wert und Tragweite der Neun-Runen-Fuge UNFEIG” (Nebel 1979:130–51) and “Von der Erhabenheit der RUNE” (Nebel 1979:166–70). The following passage is typical of the ponderously pseudo-mystical conception of runes on which the witty language-games of UNFEIG are predicated: “The rune has a magic, incantatory effect: an ordinary letter of the alphabet in everyday language is not a rune in the terms of the ART OF THE FUGUE and THE RUNE POEM. A letter only truly becomes a RUNE when it belongs, as co-fabulator, as part of a supra-sensual, incantatory group of primal forms of the word – when it is elevated to the unique status and pristine unity of a creative group of meaning-bearing shapes, which are able to reveal the innermost structure of our being by means of the divine primal forces which flow into them” (Nebel 1979:134). Even the lipogram is mystified in this way to show “how a poet can make do with a minimum of letters of the alphabet, when he attempts to write PRIMAL POETRY” (Nebel 1979:135). 13. “In the realm of art half is always more than the whole. […] The absolute minimum of means has the maximum effect” (Nebel 1979:153).

References Adler, J. and U. Ernst (eds). 1987. Text als Figur: Visuelle Poesie von der Antike bis zur Moderne. Wolfenbüttel: Herzog August Bibliothek. Bénabou, M. 1986. Rule and constraint. In Oulipo: A primer of potential literature, W. Motte, (ed.), 40–47. Lincoln NB: Nebraska University Press. Bouissac, P. (Editor in Chief). 1998. Encyclopaedia of Semiotics. Oxford: OUP. Collective. 1973. Oulipo: La littérature potentielle. (Créations, Re-créations, récréations). Paris: Gallimard. Collective. 1981. Oulipo: Atlas de littérature potentielle. Paris: Gallimard.

Forms of restricted iconicity in modern avant-garde poetry

Eich, G. 1965. Anlässe und Steingärten. Frankfurt am Main: Suhrkamp. Fischer, O. and M. Nänny (eds). 2001. The Motivated Sign. Amsterdam: John Benjamins. Goodman, N. 1968. Language and Art. Indianapolis IN: Bobbs-Merrill. Greenlee, D. 1973. Peirce’s Concept of Sign. The Hague: Mouton. Jandl, E. 1965. Three visual lip-poems. In mai hart lieb zapfen eibe hold. Unpaginated. London: writers forum poets. Jandl, E. 1966a. Laut und Luise. Olten-Freiburg im Breisgau: Walter. Jandl, E. 1966b. Laut und Luise. [recording] (Quartplatte 2) Berlin: Klaus Wagenbach. Jandl, E. 1973. Dingfest: Gedichte mit einem Nachwort von Hans Mayer. Darmstadt-Neuwied: Luchterhand. Johansen, J. and S. Larsen. 2002. Signs in Use: An introduction to semiotics. London: Routledge. Jones, M. 1982. A Study of Otto Nebel’s Major Works: Zuginsfeld and the ‘Runen’. German Life and Letters 35:253–66. Jones, M. 1984. ‘Der Sturm’: A focus of expressionism. Columbia SC: Camden House. Kolář, J. 1971. Das sprechende Bild: Poeme – Collagen – Poeme. Ed. K. Schäuffelen. Frankfurt am Main: Suhrkamp. Krampen, M. 1973. The role of signs in different sign processes: Towards a basis of generative semiotics. Versus 4:101–08. Liede, A. 1963. Dichtung als Spiel: Studien zur Unsinnspoesie an der Grenze der Sprache, 2 Vols. Berlin: De Gruyter. Liszka, J. 1996. A General Introduction to the Semeiotic of Charles Sanders Peirce. Bloomington IN: Indiana University Press. Moles, A. 1972. Vers une théorie écologique de l’image. In A.-M. Thibault-Lalan (ed), 49–73. Image et Communication. Paris: Editions Universitaires. Morgenstern, C. 1947. Alle Galgenlieder. Wiesbaden: Insel. Motte, W. 1986. Oulipo: A primer of potential literature. Lincoln NB: Nebraska University Press. Nänny, M. and O. Fischer, (eds). 1999. Form Miming Meaning. Amsterdam: John Benjamins. Nebel, O. 1960. UNFEIG: Eine Neun-Runen-Fuge. R. Radrizzani (ed.). Zurich: G. Blumer. Nebel, O. 1970. UNFEIG: Eine Neun-Runen-Fuge. Zur Unzeit gegeigt. S Press Tonband Nr. 2. Hattingen-Buchholz: Edition S Press. Nebel, O. 1979. Das dichterische Werk, Vol. 3: Frühwerke, Schriften zur Sprache und zur Kunst [Frühe Texte der Moderne], R. Radrizzani (ed). Munich: edition text + kritik. Neumann, P. 1973. Morgensterns Galgenlieder als poetologische Modelle betrachtet. Sprachkunst 4:53–64. Nöth, W. 2001. Semiotic foundations of iconicity in language and literature. In The Motivated Sign, O. Fischer and M. Nänny (eds), 17–28. Amsterdam: John Benjamins. Peirce, C. 1960. The Collected Papers of Charles S. Peirce, 8 Vols. C. Hartshorne and P. Weiss (eds). Cambridge MA: Harvard University Press. (References to this edition are abbreviated in the customary way, by volume and paragraph number only). Ray, Man. 1924. Poème optique. 391, XVII, 2:119. Saussure, F. de. 1916 (1988). Cours de Linguistique Générale. Tullio de Mauro (ed.). Paris: Payot. Spitzer, L. 1921. Zur Interpretation Christian Morgensternscher Gedichte. Euphorion 23:95– 99.

153

154

John J. White

Steiner, W. 1982. The Colors of Rhetoric: Problems in the relation between modern literature and painting. Chicago IL: Chicago UP. Thalmayr, A. 1985. Das Wasserzeichen der Poesie oder die Kunst und das Vergnügen, Gedichte zu lesen. Nördlingen: Franz Greno. Wallis, M. 1973. On Iconic Signs. In Recherches sur les systèmes signifiants: Symposium de Varsovie 1968, J. Rey-Debove (ed), 481–98.The Hague-Paris: Mouton. White, J. 1999. On semiotic interplay. Forms of creative interaction between iconicity and indexicality in twentieth-century literature. In Form Miming Meaning, M. Nänny and O. Fischer (eds), 83–108. Amsterdam: John Benjamins.

Eco-iconicity in the poetry and poem-groups of E. E. Cummings

Eco-iconicity in the poetry and poem-groups of E. E. Cummings Etienne Terblanche* and Michael Webster** *North-West University, Potchefstroom; **Grand Valley State University, Allendale

If in iconicity “form mimes meaning”, then in eco-iconicity formal elements like syntax, word division, visual placement on the page, the use of white spaces, and what might be called a transformational semantics all work together to mime the dynamic processes of the ecosystem. The poet’s lower-case persona inhabits insubstantial air as a voice, and then sinks in a moment of transcendence into a star. In addition, the star is iconically present within the air and on the page. Moreover, the poem ends not with the period, but with white space, resisting closure. In this paper, we will show how Cummings uses devices like these to emphasize and enact dynamic transformations in, between, and among poems, positing “meaning” as a continual process, transformation, and co-incidence of “now” moments of being. By means of a heightened iconic precision, Cummings persuades the reader to become aware again of his or her original ecological being.

1. Introduction Reading E. E. Cummings’ poetry from an ecological perspective brings some of its most powerful values to the surface for analysis and discussion. Roughly a century of critical engagement with Cummings’ work has seen numerous appraisals of his singular ability to combine modernist poetic devices and a deep awareness of nature: one thinks instantly of critics in this respect such as Norman Friedman, Martin Heusser, Michael Welch, Pushpa Parekh, and more. Ecology is a vast and intricate subject, but it is in the nature of dynamic complexity rather than static complication (see Naess 1995:5). Of the modernist poets perhaps Cummings knew this most clearly. A working definition of the ecology to be found in his poetry could therefore read as follows: a dynamism of signs that corresponds maximally with the dynamism of the ecosystem. In other words, Cummings refreshes one’s moving perceptions that one lives radically and

155

156

Etienne Terblanche and Michael Webster

even originally within a larger natural context in which interaction, co-existence and continuation or open-endedness are vital. This dynamism has important implications for the reading of Cummings’ poetry, especially on the levels of iconicity and the sequencing or grouping of one’s reading. In fact, Cummings’ iconicity and his dynamic ecology blend to a remarkable degree in his poems, and this blending facilitates, to our mind, a better understanding not only of the poems, but also of the poem-sequencing or poem-groups. With ‘ecology’ we have in mind a profound awareness of wholeness: the necessary poetic recognition that events, elements and creatures on earth hang together interdependently and interactively within a simultaneously vast, detailed and delicate, ongoing and changing equilibrium of which the limits are probably the endless universe in space and the enormous and intricate scale of ecological and evolutionary events in time; a space and time that Cummings overcomes on the poetic level with mere words and supreme skill into (at least) glimpses of one’s whole place within things, and timelessness. That is, his poems have the capacity to revitalize the continuing given of continuation itself, a continuation that includes, ultimately, all those events, forces, elements and creatures. It follows that eco-iconicity (or holo-iconicity) is that intense level of iconicity which manages to dissolve the perceived “solid” boundaries or limits of signs to such an extent that a glimpse or more opens up of wholeness in being (existing) on and within earth: wholeness of one’s physical being within physical (everchanging and stable) nature. The dissolution does not occur randomly; rather it traces precisely the edges where the sign reaches its limit into the natural world. As such, Cummings makes no less than an entirely new sign: a holo-logos and not logocentrism. Perhaps it boils down simply and complexly (but neither in simplistic nor in a complicated, interfering or ignorant manner) to an active human acceptance of the concrete, changing nature of being. In its simplest form, this is to see the birth of a little horse or the nest of a hummingbird in all its profundity, without complication or interference either of subjective sentimentality or of objective rationalism (distance of hardness). For a moment the sheer marvel and joy of natural (physical, concrete) being is enough, as mediated by poetic signs. It is therefore a unique form of humility, staying close to humus, and Cummings thus reveals a unique variant of iconicity to be called perhaps humble iconicity. For example, Cummings breaks open the word “growing” to show that it contains an “o” as well as the sign “Wing” (1994:657). The reader who can follow his ungrammar will know that the “o” signifies in all kinds of ways in Cummings’ poetry overall: as a moon, as eyes, as an entrance, as an icon of completeness and circular motion, as a marker of the female force or tendency in collaboration with the male “i”, and more. In this instance the reader is free to view the “o” either as an entrance or a pivot from which the dreamlike perception of the newborn

Eco-iconicity in the poetry and poem-groups of E. E. Cummings

little horse is growing or “gro// Wing” like a wing – from an eye or a certain way of perception, as a process with a center of one. Thus the signs involve not merely a relatively static or “solid” designation (such as “o” for moon), but also an iconicity of or into dynamism along with the active reader, a moving iconicity that mimes movement or change in time (into timelessness) within the world. In Max Nänny’s terminology, this combines imagic iconicity and diagrammatic iconicity, not to the effect of a “somewhat more abstract nature” as Nänny (2001:210) says, but rather to the effect of a revitalization of being aware that one is part of an ongoing natural process, also in its concrete aspect of being a physical part of a physical natural process with dreamlike qualities. Part of the combination and dynamism is Cummings’ dialectical play that reduces the world into opposite pairs, creating a simplified avenue into the complexity of contexts that rejuvenates being within nature. Nature for him is a “ground zero” not merely in the sense of an absolute limit of thought, but in the sense of a nothingness within which living feet on living soil or living eyes within living air embody a profound unity of continuity and being. As Richard D. Cureton (1980:250) says, Cummings creates an interlevel iconicity, such as “o” for moon, in which signs mime the natural world. Combine this with Norman Friedman’s (1996:55) statement that a Cummings poem can open out the implications faster than it closes them down: the poem creates contexts with increasing openness. The result is an approximation of what we mean with eco-iconicity: poems that open out into contexts of the natural world with such precision and “speed” that they become strikingly moving icons of nature in motion.

2. Iconicity in a poem-group One way of reading this dynamism is to follow the dance of the lower case “i” and the lower- or upper case “o” in poems and groups of poems. However, one can hardly read this i/o dance in isolation. As the poem-groups in Cummings indicate copiously, it forms part of the much larger contextual dancing of Cummings’ poetic signs, namely the dance between male, female, and nature. To illustrate this dancing or movement among poems, we have decided to focus mainly on the consecutive pair of poems, “hate blows a bubble of despair into” and “air,”. Obviously this group resonates with other Cummings poems and we shall also look at some of them. First consider this: hate blows a bubble of despair into hugeness world system universe and bang – fear buries a tomorrow under woe and up comes yesterday most green and young

157

158

Etienne Terblanche and Michael Webster

pleasure and pain are merely surfaces (one itself showing,itself hiding one) life’s only and true value neither is love makes the little thickness of the coin comes here a man would have from madame death neverless now and without winter spring? she’ll spin that spirit her own fingers with and give him nothing(if he should not sing) how much more than enough for both of us darling. And if i sing you are my voice, (Cummings 1994:531)

Numerous Cummings poems ranging from the serene to the playful enact a mergence of male and female, and this sonnet is another example. It begins with a positive denial of a man-made, artificial “world” – the world of wage slavery, monotony, signing on the dotted line, and so on, a world which is the target of Cummings’ satirical poetry. In particular, the poem begins by positively, actively denying the dualities that underpin or appear to underpin this “world”. As the sonnet unfolds, it takes dualities into a sense of their interactive, clear and precise merging. So in line 8, “love makes the little thickness of the coin” – love combines two sides or opposites into one, and this is its worth. The predominant iconic aspects of the poem center on the almost hidden distribution of “i” and “o” within it. Careful reading sees a number of interesting developments in this respect. Every line contains at least one “i” or one “o”, and most lines contain a distribution or dance of the two. For instance, line 6 – “(one itself showing,itself hiding one)” – distributes the two as follows: o i o i i i o, so that the “o”s contain the various “i”s in this instance, and the whole line is contained within the parentheses. Consider that this i/o dance is simultaneously iconic on three levels: first, the level of sound in which the “o” is lower and fuller than the perhaps more optimistic-sounding and up-turning “i”; second, the level of vision in which the “o” represents seeing, cyclical wholeness, and entrance while the “i” presents uprightness and a vertical movement upwards and/or downwards (with the dot acting as the first indication of a return to the cyclical, full form of the “o”); third, the level of gender in which the “o” presents femininity and the female while the “i” presents masculinity and the male. It is therefore all the more striking that the distributional pattern of these characters (or micro-icons or micro-ideograms) maintains a certain separation with the exception of two vital moments within the sonnet: the conclusion of the sestet, when the two combine visually and audially in the diphthong “oi” within the word “coin”, and the confirmation of this combining movement in the last word of the sonnet, “voice”. This is how intricately and precisely Cummings

Eco-iconicity in the poetry and poem-groups of E. E. Cummings 159

will steer iconicity on various levels to achieve a renewed sense of active malefemale balance. Not only do the letters “i” and “o” interact iconically and ecologically across spatial and lexical boundaries within the sonnet, but Cummings extends these interactions across the white paper-space to the much more spatial (and iconic) poem on the next page. And these interactions occur not only at the letter level, but at various semantic and thematic levels as well. But before we look at these cross-poem interactions, we first need to unravel some of the dualities of the first stanza, dominated by the image of hate blowing a bubble of despair. There are at least two ways of looking at this bubble: either the hot air of hate inflates a bubble-gum balloon that splats despair over and into the world when it pops, or some sort of bubble of anxiety pushes within the “world system” an anxiety bubble which bursts into fear and “buries a tomorrow”. With the mention of “tomorrow”, we might expect to see the word “hope”, but hope is not mentioned in the poem because tomorrow is buried, and “yesterday”, or the nightmare of history, has been reborn. The missing third term in the first stanza is “now”, which stands between or beyond the today of woe, the yesterday of fear, and the denied hope of tomorrow. (The poem’s date – 1940 – allows us to think of Hitler and Stalin here.) This bleak picture may be thought to extend to the second stanza, where the surface emotions of pleasure and pain have their backs turned to the true value of love. Which means that we’ll have to consider the rather mind-boggling third stanza. One way of reading it is to follow the movement of its uniquely Cummingsian arithmetic. On the negative end, the stanza seems to suggest a danger, especially among men, of a one-sided life that will end up on one face of a static coin. This would be concomitant with an attitude that would expect spring without winter, whereas it is the dynamic integrity of those and other opposites that make a full, or full-empty, life. On the positive end, it seems that death will keep a lifetime in motion in any event: madame death will spin the coin with her own fingers if the male persona would not do so himself. The last line of this stanza – “and give him nothing (if he should not sing)” – should perhaps be read against the background of the negative end and the positive end cancelling each other out to lead to an overall positivity of “nothing”. On the one hand: should he not sing in response to life or spinning, he will receive nothing from her in the negative sense of missing all that a lifetime is about, namely that nowhere from which everywhere, everything and everyone continue to become, beyond all static opposites of mind. On the other hand: he will have received the great gift of nothing already, merely by being alive. In short: in view of the dying-spinning of one’s lifetime one cannot do else than to move beyond the static, even beyond madame death. (The notion of madame death seems to us somewhat different, with something of a more static

160 Etienne Terblanche and Michael Webster

and stifling air, than the concept of dying usually found in Cummings’ oeuvre. Consider, for instance, the lines “dying is fine(but Death// ?o/ baby/ i// wouldn’t like// Death if Death/ were/ good” (Cummings 1994:604).) As with the image of the bubble, the image of “madame death” may be read in at least two ways: she is either spinning the coin on a table, in which case it is bound to fall on one side or another, or we may see her flipping the coin in the air and calling it, heads or tails, pleasure or pain. No matter which side comes up, the spirit loses this coin toss with death and ends up with nothing. However, if we imagine the sestet of this sonnet as countering the octave or changing the metaphor, then madame death may not be spinning a coin at all, but a metaphorical top; or perhaps she spins the man’s spirit as the Fates do, spinning yarn, weaving it into a life-tapestry, and then snapping the thread, giving one the emptiness and nothing of death. But as we have already pointed out, this nothing is its own third term, beyond the pairs of opposites, combining both the nothingness of death and the nothing of the now (one without nevers and winter) that lovers inhabit. It’s also the nothing of a breath, necessary for life, singing, and speaking. But what to make of the comma at the end of the poem? The compliment in the last line, “if i sing you are my voice”, encodes a rather usual poet-muse relationship, and it also pays tribute to the third “term” of the two lovers’ whole now, the song or voice or breath that is made of nothing. Such a nothing also forms a “little thickness” (the life beneath the voice?) between or beyond the surface thinness of pleasure and pain. This nothing is “more than enough for both of us” since it produces a unity, an “oi”, which contains and transcends gender doubleness, and it is actually something, one dynamic, which is worth more than everything. The only possible conclusion here is transcendental, a resonance: if the male sings, the unity with the female is his voice, and no other singing is possible. And the conclusion cannot end in a full stop of closure – it opens up into blank space with a comma. The open-ended and dynamic “voice” that ends the “hate” poem dissipates into the circumambient “air” in the following poem (Cummings 1994:532). According to Cummings, the miraculous breath or nothing of being cannot be captured: the bubble of despair ultimately can only burst – all is movement. It has no other choice, so to speak, since being and now always continue to be more than any conceivable form of confinement, including a bubble-gum bubble. Air will not remain trapped within such a bubble. Which brings us to the next poem, #44 of this volume entitled 50 Poems, a poem which, at first glance, appears to be quite different in form and theme:

Eco-iconicity in the poetry and poem-groups of E. E. Cummings

air, be comes or (a) new (live) now ;& th (is no littler th an a: fear no bigger th an a hope)is st anding st a.r (Cummings 1994:532)

This poem formulates the recognition that as a star first appears and increases in brightness, the air becomes new with aliveness. The very being or essence of the star increases – it is in the process of “anding” – and it remains or “stands:” it embodies a stability. Its essence is also extremely delicate, especially from a subjective or human perspective. The increasingly participating observer feels that the star continues to find itself somewhere between the closure of fear and the expansiveness of hope. One implication of this is that the star is at the ongoing, moving center of (the subject’s) experience and that this continues to move through and beyond opposites such as fear versus hope and natural object versus observing subject. Shifting the full stop from the end of the poem to the middle of its intriguing final sign, “a.r”, could therefore amount to saying that full stops or ends are pivots or beginnings, not final closures. Something else has happened in the movement from voice to air: we have moved from a poem designed to be read aloud, a poem of “voice”, to one which,

161

162 Etienne Terblanche and Michael Webster

as Cummings put it, is “to be seen & not heard” (Cummings 1969:267) – a poem of “air”, an iconic poem. This poem is constructed on a number of seemingly arbitrary rules. There are 11 “stanzas” in the poem, with a line count pattern of 1-3-1-3-1-3-1-3-1-3-1. After the first, each of the next four stanzas is pushed one typewriter space to the right; and each stanza after the sixth or middle stanza is backspaced one space to the left. Perhaps the middle stanza is pushed out to intimate pregnancy, as if the poem gradually became big with child, giving birth to the star-dot in the last line. In addition, the shape of the poem may mimic a crescent moon, the new moon that Cummings represents elsewhere by a right-facing parenthesis: [ ) ]. And, among other things, the comma after “air” is part of one punctuation pattern usual with this poet: (, / ; / : / . ) or comma, semicolon, colon, period, moving from the most hesitant mark to the most emphatic. The word “air” at the top of the poem transforms into a star, or “a.r” at the bottom – an “are” with the dot of a star in the middle. As we have seen, the comma at the end of “voice”, reappears at the end of “air”, an indication of the dissipation of voice into ambient air. Cummings expects us to feel the comma more as gesture, as index, than as grammatical punctuation. The comma after “air” signals a further process of becoming or transformation: through a slight vowel shift, the vocable “air” transforms into “or” – it moves from “a”[n] “i” to an “o” – so the male “i” becomes the more complete but more indecipherable female “o”. On a semantic level, “or” may mean at least three things: (1) the simple English meaning of “or”, indicating another possibility, which in this case is “(a) // new / (live) / now”; (2) “air becomes gold” [“or” means “gold” in French], i.e., sunset; and (3) “or” may also mean “now” in French., so air becomes “(a)(live)” or a “new now” (referring back to the “neverless now” of the previous poem – now is always new). The multivalence of this one small vocable indicates that Cummings is following Keats’ advice to Percy Shelley to “‘load every rift’ of your subject with ore”.1 In addition, the first three lines, “air, // be /comes” can mean “air is” (it “be”s), and then “be -comes”: begins to take on solidity and color and depth as dusk falls. Numerous iconic and semiotic devices emphasize that “meaning” is the continual process and transformation of moments of “now”. But Cummings’ air poem involves a further fragile triumph of the lower case “i”. Compare the first and last lines of this poem. One sees that the upright little body of the lower case “i” physically, visually disappears on the page behind the readerly horizon in the last line. Only the dot remains, suggesting that the subject and the star “a.r” – that is, they are in the process of becoming one. The vertical line of the sign “i” literally disappears to highlight only the remaining dot. This embodies a process of simultaneous appearance and disappearance, decrease and increase. In other words, the isolated subjectivity of the subject disappears in correspondence with the increasing appearance of the star: the two horizons of

Eco-iconicity in the poetry and poem-groups of E. E. Cummings 163

experience and being, the being of a star or the realms out there and the being of the human or the world in here are merging, their boundaries are in the process of dissolving. Here is a childlike innocence, a moment of being lost in the most positive sense, a moment in which there is no difference between I and it, inside and outside. One cannot but appreciate the precision and crispness of Cummings’ manner of presenting such a moment in poetry. Incidentally, the dot of the “i” therefore also acts as an icon or a micro-ideogram of the star, just as the imaginary and imaginative line below the dot acts as an icon of the meeting and blending of horizons. This meeting and blending is similar to the transformation of voice and air that occurs between the two poems. The poems “hate blows a bubble of despair into” and “air” should therefore be read with each other in mind, in accordance with the open-endedness that the lack of titles and the blank spaces between them suggest. The poems share a charged comma each, as well as the motifs of “now” and “fear”, and they also pun on “air” turning into either “despair” or a star. In short, just as the male and the female voices are cross-stitched into oneness through poetry, so are the existence of a human subject and a star. Like male and female, the human subject and the natural star-object are supposedly on opposite sides of a divide. But in Cummings, this divide is permeable in the most moving or actual sense, which permeability does not merely allow a transgression of categories and their transfusing saturation, but actively invites and steers the process, the readerly dance of iconicity and ecology. In addition, the transformations in the first poem of nothing into “more than enough” and of voice into song are echoed in this poem by the transformation of “air” into the star of “a.r” and of “i” into a dot. I think we might consider that this little dot at the end could also stand for the disappearing sun; thus readers are asked to transform one sign three times: into sun, the disappearing “i” of the poet, and emerging star. The stuttering effect of the repeated vocables “th” “an a” and “st” mimes these linguistic transformations of sounds, creating new words and meanings from nothing, or from old vocables. In addition, these stuttering syllables reveal a poet suddenly inarticulate in front of the emerging (but now standing) star. This “alive” star is, as the words within the parentheses read, “no littler than a: fear no bigger than a hope”. Even though both poems share iconic and semantic transformations that mirror and present the ecological, interdependent, and co-existent mergence of human and natural worlds, there are transformations that cross the blank space between the poems as well. For example, the air in the changing sky of “air”, contrasts sharply with the stale bubble-air of “despair” in the previous poem. By simply lopping off four letters, this “despair” becomes “air”, which bursts, not into fear but a star. In the second poem, we need find only one more vowel for “or” to

164 Etienne Terblanche and Michael Webster

become “our” – air becomes our voice, the voice of the last line of the preceding poem. The nothing of the lovers, of the voice, of air, of the “i” becomes the virtual nothing of a star as an iconic dot on the page that is nevertheless “st / anding” and “anding” within and without the poet and his/her voice/song. It becomes clear, therefore, that Cummings employs his ungrammar with great effect to indicate a direction of signs inwardly that corresponds with their direction or purpose outwardly into nature. These corresponding movements achieve a high degree of precise simultaneity in his work, and they ultimately allow a reciprocal influence or interlevel sign-nature interaction between the poem and natural actuality, while the reader acts as the essential go-between or collaboratorcreator of the poem, as is often the case in modernism. We saw how the comma at the end of the previous poem refers to “breath” or “voice” or “life”, a kind of silent singing. The poem that follows “air” picks up the motif: enters give whose lost is his found leading love whose heart is her mind) supremely whole uplifting the, of each where all was is to be

The (comma or the) breath and life of two lovers named “give” and “love” is lifted up by the whole the two make together.

3. Now-here osmosis There are other examples where Cummings indicates transformative processes by revealing signs that hide within other signs. For instance, he shows how the word “in” resonates within the word “arriving” and how the words “now” and “here” find themselves within the word “nowhere”, as the following poem illustrates. this man’s heart is true to his earth;so anyone’s world does -n’t interest him(by the

Eco-iconicity in the poetry and poem-groups of E. E. Cummings 165

look feel taste smell & sound of a silence who can guess exactly what life will do)loves nothing as much as how(first the arri -vin -g)a snowflake twists ,on its way to now -here (Cummings 1994:676)

This poem in its entirety is about an acute awareness of natural movement, expressed in moving poetry. In the third to the last stanza, the isolated “-v-” that finds itself favourably within the word “arriving” acts as an icon of the simultaneity of movements: downward, an arrow pointing to earth, of the snowflake falling, and upward, the two hyphens not unlike two arms opening up to the above, iconic of joy or an upward funnelling of the speaker’s feeling, something which the poem as a whole illustrates. The arrival of the snowflake is also a “coming in:” the poem exploits the coincidence of the word “in” within the word “arriving”. The appearance of the snowflake is the very epitome of that nowness and being-here that continues to emanate from the miraculous nowhere as much as it will go back again into that nowhere upon melting or upon becoming part of a white blanket of snow. Words like “nowhere” point in other directions as well. Thus the “here”

166 Etienne Terblanche and Michael Webster

and “now” in the word “nowhere” refer to place and time, while contradicting the emptiness and timelessness of “nowhere”. The word “nowhere” contains its opposites and means both at the same time, and as such is a perfect example of what Cummings termed the “the secret every mystic tries to tell” (Cummings 1969:261), the coincidentia oppositorum [coincidence of opposites]. As he says (underlined and in red ink) in the notes now at the Houghton Library, Harvard University: “P[oetry] is that temperature at which opposites fuse”.2 In addition, the last three one-line stanzas read “nothing . . . in . . . -here” – in here, there is nothing, in the snowflake is nothing. So as the line breaks indicate, the man “loves // nothing”. And this nothing is on its way – in fact, between the “in” and the “-here” we see in the second-to-last stanza -g)a snowflake twists ,on its way to now

– so that the comma we saw before has become a snowflake, on its way, a word that we believe Cummings intends to be taken in its Taoist sense as well. (For Cummings’ interest in Taoism, see Terblanche 2002.) The Way or the Tao is a process of becoming, of coming out of nothing, becoming, going into being and then going back into nothing. This view is implicit in the Chinese word for nature, tzu-jan, which translated literally means something like “self thus”, or “occurrence appearing of itself”. According to David Hinton (2002:xiv), the word itself encodes an entire cosmology of “the ten thousand things emerging spontaneously from the generative source, each according to its own nature”. As the Zen monk Shunryu Suzuki (1999:116) said, “I discovered that it is necessary, absolutely necessary, to believe in nothing”. This nothing is not “voidness”, however. “There is something [in this nothing] but that something is something which is always prepared for taking some particular form, and it has some rules, or theory, or truth in its activity. This is called Buddha nature, or Buddha himself” (ibid.: 117). Cummings’ snowflake which comes out of nothing and enters into the “now / here” of nowhere enacts this “self-thus” of nature. And this multi-morphous and intensely vibrant something always emanates from and returns to nothing as is suggested in the poem “hate blows a bubble of despair” (as we have seen). It seems to say that the so is just so, a form of active and humble acceptance. Perhaps the colloquialism to “go with the flow” is an equivalent to this, but then it must be seen to mean that one goes thus in full faith and in full self-being. It is also to see this very so-ness (so to speak) dance in the miracle of being: at the still point there is only the dance. Moreover, its unifying dynamism is utterly interactive, a dance of opposite forces. Indeed, the dance of the micro-ideogram “i” and its other self, the you, the female, the “o” can be

Eco-iconicity in the poetry and poem-groups of E. E. Cummings 167

found in a dazzling range of configurations in Cummings’ oeuvre. (We base this notion of an i/o dance in Cummings’ oeuvre roughly on work done within feminist studies – see Dullaart 1998:155, 156.) This article cannot hope to do justice to this sheer range of the dance. It has not been mentioned that the “o” acts as an attractor point and an entrance into a sense of mergence or completeness in the much-discussed leaf poem. l(a le af fa ll s) one l iness (Cummings 1994:673)

In haiku manner (Welch 1995:117) this poem states with insightful simplicity that loneliness is a falling leaf, or that a falling leaf is loneliness. The modernist device of fragmentation and recombination (Friedman 1984:174) gives rise to a complexity that matches this profound simplicity: for the falling leaf also turns out to be subject-hood (the “i-ness” of the final line) and unity (the word “one” of line 7). The 1–3–1–3–1 line pattern (similar to that of the star poem above) as well as the ever-spiralling alternation of consonants and vowels contribute to the increasing awareness that this poem enacts a mergence of subject (lonely human) and object or other (falling leaf). There is only one “o” in this poem and this is a telling feature. Knowing Cummings, one wants to find its partner, the “i”. It finds itself in the last line where, as has been demonstrated in other poems, the shell of a signifier breaks co-incidentally to deliver the speaker. The poem overall shows that this subjecthood can only or one-ly be understood in terms of the natural other, the falling, twirling leaf. Consider here that the male subjecthood of the “i” presents itself visually and metaphorically in the uprightness and linearity of that sign with its dot jumping out, or its dot already halfway towards the cyclical. Now in the leaf poem uprightness resonates within the ample distribution of “l’s” which can be read either as numbers one or fluid, rich and soft letters “l”. Uprightness is to be found further in the overall figure 1 shape of the poem. Given all this, it is all the more striking that there is only one cyclical figure in

168 Etienne Terblanche and Michael Webster

this poem: the “o” of the word “one”. Indeed, a number of things come together or coincide once this recognition dawns on one. As Norman Friedman (1996:55) implies so rightly about Cummings’ poetry in general, here the possibilities open faster than the reader can close them down – giving rise, to my mind, to an ecological sense of dynamism, mergence, becoming part of a larger context through a poetic arrangement and movement of signs. The “o” is also part of the only word that is immediately discernible to the English reader on the surface of the poem: the word “one”. So, on these levels and with a more or less simultaneous occurrence, the cyclical “o” attracts and focuses the reading of this poem, and leads like an entrance to a sense of mergence between word and context. We must infer that reading Cummings’ poems as separate little rooms or monuments arranged in volumes – although this is an important value in his case as in those of other poets – should not blind one to another equally important value when it comes to reading his poems in groups. It is our belief that the blank space and generic lack of titles in his poetry are important overall indications that (or invitations to) the fact that Cummings’ poems can and should be read “osmotically” – like cells whose membranes invite and precisely steer various complexes of integration and integrity, of heightened interactivity also across the usual limits of poem-sequencing . For instance, we have travelled here in this “osmotic” manner from a pair of poems on love and nature – a pair that touches and intersects in various ways with other poems as we have indicated – into a poem (the leaf poem) that may seem quite separate in many respects (a different volume, a different time of Cummings’ poetic development and so on), especially to our minds trained to be hungry for differences and distances. Yet, as we have demonstrated perhaps all too briefly, looking for an equal precision of “osmotic” continuation in Cummings is almost always rewarding, startling or dazzling. Looking for the i/o dance in the leaf poem furthers the complexity, depth and joy of one’s engagement with the poem.

4. Eco-iconicity on the street Consider one further poem, one that is not about nature in any direct way. Nevertheless, we think that this poem will illustrate Cummings’ eco-iconicity (the linguistic miming of natural processes) very well. The poem depicts some men waiting, spitting, and coughing: fl a tt ene

Eco-iconicity in the poetry and poem-groups of E. E. Cummings 169

dd reaml essn esse s wa it sp i t) (t he s e f ooli sh sh apes ccocoucougcoughcoughi ng with me n more o n than in the m (Cummings 1994:488)

Flattened into horizontal prose, the poem reads: “flattened dreamlessnesses wait spit) (these foolish shapes ccocoucougcoughcoughing with men more on than in them”. The “flattened” men (called “dreamlessnesses” in the poem) wait, spit, and cough while standing “more o / n than in” their shadows (“foolish shapes”). The poem is 25 lines long, and the parentheses in the middle of the center line (13) divide the men from their shadows. Despite this division, however, each half of the poem refers to the other. For example, both men and shadows are flattened, the men possibly by life and alcoholism. As we see it, the poem depicts a scene in the Bowery in New York City on a sunny day: men are waiting against a wall for a mission or soup kitchen to open, or perhaps for someone to come by and hire them for day labor. Simply by isolating and repeating letters in the short lines of the poem, Cummings creates imagic icons in which the letters of the poem seem to shadow or even mirror one another, reinforcing the notion that the men are shadows as well. Some examples of this letter-shadowing occur in lines one (“fl”),

170 Etienne Terblanche and Michael Webster

three (“tt”), four (“ene”), and five (“d d”). Lines 7 and 8 are near mirrors, with “esse” (Latin for “to be”) shadowing the “essn” (“essence”?) above it. Line nine (“s wa”) can be read from middle to front as “was”. (The word is so “was” that its final letter has migrated to the front of the line.) But what “was”? Something has flattened the men’s souls or personalities, for they are defined by their lack of dreams; each man in his wait has been reduced to an “it” (line 10); and their “sp / i // t)” (lines 11–13) cannot quite amount to “spirit”, though somewhere in their “it” and spit lurks an “i” (line 12), an individual, just as their dreamlessness hides an “esse” (line 8), potential being. Looking further, we find there are three more “i”s that begin or end a line (“it” “ooli” “coughi”), so there must be four men and four shadows. The shadows below seem to have more life in them than the men: each shadow becomes a “he” (line 14) and “apes” (line 20) the man above. Not only that, but the shadows also drool (“ooli” – line 18), caution silence (“sh sh” – line 19), and yet cough (lines 21–22). A line break (line 22) turns the word “men” into “me”, the poet who shadows forth both men and shadows in verse while coughing with them. Lines 21–22 (“ccocoucougcoughcoughi // ng”) contain a new sort of shadowing: the first letter of the word coughs out the first and the second letters, which then cough out the first, second, and third letters, and so on. Here, instead of the imagic iconicity of shadow-letters, a diagrammatic iconicity builds a hacking cough in which letters seem to dynamically generate further letters. As we have seen, Cummings also diagrammatically splits the poem in two at line 13 – “t) (t” – men above and shadows below. But like the hate that exploded, turned into voice, and escaped into the air of another poem, the letters and signs of the two halves of this poem interact and interpenetrate as well. Shadows, men, and poet all cough together. The men stand “more on than in” their shadows because they cannot inhabit the shadow. They lack fully-developed souls: flattened shadows mirror flattened men. These are hollow men in and on their shadows, their identities and individualities obscured. The last line of the poem is the single letter “m”, which mirrors the three letter n’s above it, with a shadow added. This last letter confirms the identical flatness of men and shadows. It also points to another meaning in the poem. For the poem is not only about men, shadows, and poet, it is also about language and its ability to shape (iconically, phonetically, and semantically) sound and meaning. As the five other single letter lines (“a” “i” “s” “e” and “f”) indicate, these “foolish shapes” or “foolish apes” are also letters which ape or represent sounds and, when put together, shadow forth meanings. They are cautioned to silence (the onomatopoetic “sh sh”), yet these letter-shapes cough themselves forth into sound. These letter shapes spit, wait, and cough on the page because they were trained to do so and arranged into iconic shapes and patterns by the poet: they cough “with me”,

Eco-iconicity in the poetry and poem-groups of E. E. Cummings

creating a kind of half voice half picture – perhaps something like the sight and sound of shadows coughing. (We are now allowed to see the “i” of line 12 and the “i” at the end of the first coughing line (21) as representing the poet along with the men, and perhaps the reader as well.) By engineering a complex array of semantic and iconic signs, Cummings creates a dynamic and complex interplay (or ecology) among sign, world, poet, and reader. All work together in a dynamic ecology of meaning and movement.

5. Conclusion We have seen that in the world according to Cummings, spring cannot be imagined or lived without winter, the male voice can only find itself in the female voice, and existence can only or one-ly occur with a view to a star. On the factual level this is so, of course: should the universe with all its enormous forces, elements and worlds not remain stable as it continues to change, existence would not be possible. But Cummings continues to reveal the miracle that goes along with this seemingly mundane fact or truism. By means of a heightened iconic precision as has been demonstrated with a view to the role of i/o on various levels, he persuades the reader to become aware again of his or her original ecological being. The masculine and feminine tendencies of natural existence dance and combine to show change and wholeness. In the sonnet “hate blows a bubble of despair” the laden “i” and “o” combine in “oi”; in the air poem the “i” disappears into the horizon; and in the leaf poem “o” is an entrance through which the subjective speaker and the objective leaf become one, while the speaker finds in this manner his “iness” or individuation. In the coughing poem, shadows, men and poet cough together, (sh)aping a meaning of charged iconicity on various levels once more. On a linguistic level, we’ve shown how Cummings evokes a plurality of meanings by finding one sign within another (the “in” in “arriving”), by moving one sign throughout a poem-group (the comma we saw in the first three examples), and by gendering certain signs (what we call “the i/o dance”). (Though Cummings uses “i” and “o” in the flattened shadows poem, the lower case “o”s are not gendered as female, and there is no dance between the two, since the poem depicts an exclusively male world.) In addition, Cummings’ dynamic conception and deployment of signs enact meaning as a process of multiplication, transformation, and becoming: the female provides a voice for the male’s song; “despair” bursts into “air”, which transforms into “or” or “ore” or gold; a lower case “i” (individual) becomes one with a star; a snowflake twists into nowhere, or “now // -here”; a falling leaf re-enacts the dance of male and female, the end of its fall identical with “iness” or the individual; and men and shadows interpenetrate, mirroring

171

172

Etienne Terblanche and Michael Webster

lost souls and individualities. In this last poem the lost souls, shadows, and poet cough out an “i” – a letter, a foolish shape and ape, a poet, a man, and perhaps the soul of the reader. What we call “dynamic unity” occurs when a committed reader activates these multivalent transforming signs. Cummings’ linguistic games with commas, letters, and letter-sh(apes) in many ways embody and reflect in their processes the dynamic, emergent, co-incident, cyclical, symmetrical, and interconnected ecology of the natural world.

Notes 1. Letter to Percy Bysshe Shelley, August 16, 1820. The expression comes from Spenser’s Faerie Queene 2.7.28: “With rich metal loaded every rifle”. 2. E.E. Cummings Papers. Houghton Library, Harvard University, call numbers bMS Am 1892.7 (90). folder 42, sheet 34.

References Cummings, E.E. 1994. Complete Poems, 1904–1962. George J. Firmage (ed). New York NY: Liveright. Cummings, E.E. 1969. Selected Letters of E. E. Cummings. F.W. Dupee and George Stade (eds). New York NY: HBJ. Cureton, R. D. 1980. On the Aesthetic Use of Syntax: Studies on the syntax of the poetry of E. E. Cummings. London: University Microfilms International. Dullaart, G. 1998. Feminisme. In Alkant Olifant: ‘n Inleiding tot die literatuurwetenskap. C. N. Van Der Merwe and H. Viljoen (eds), 152–160. Pretoria: J. L. Van Schaik. Friedman, N. 1984. Post Script to ‘E. E. Cummings and the Modernist Movement’. In Critical Essays on E.E. Cummings. G. Rotella (ed), 174–175. Boston MA: G.K. Hall. Friedman, N. 1996. (Re) Valuing Cummings: Further essays on the poet, 1962–1993. Gainesville FL: University Press of Florida. Hinton, D. (ed. and trans.). 2002. Mountain Home: The wilderness poetry of ancient China. Washington DC: Counterpoint. Naess, A. 1995. The shallow and the deep, long-range ecology movement. In The Deep Ecology Movement: An introductory anthology, A. Drengson and Y. Inoue (eds), 3–9. Berkeley CA: North Atlantic Books. Nänny, M. 2001. Iconic features in E. E. Cummings’ poetry. In The Idea and the Thing in Modernist American Poetry, C. Giorcelli (ed), 209–234. Palermo: Ila Palma. Suzuki, S. 1999. Zen Mind, Beginner’s Mind. Trudy Dixon (ed). New York NY: Weatherhill. Terblanche, J. E. 2002. E. E. Cummings: The ecology of his poetry. PhD dissertation, Potchefstroom University. Welch, M.D. 1995. The haiku sensibilities of E. E. Cummings. Spring: The Journal of the E. E. Cummings Society. New Series 4:95–120.

The language of film is a matrix of icons 173

The language of film is a matrix of icons Strother B. Purdy Marquette University

The powerful and near-universal effect of 100 years of motion pictures on human culture has naturally led to considerable study of how this effect is brought about, including an effort to break it down into its operative units. That they are not linguistic units, as held by regnant theories of the “language of film” but representational, and as such best analyzed by application of Charles Sanders Peirce’s concept of the icon, is the argument of the present paper.

1. Introduction “The language of film,” “film” meaning the cinema, or motion pictures, is a phrase I have never been happy with. A generation ago it signified a serious attempt to locate and define the meaning-bearing, or better “effect-bearing” elements of the motion picture that could be compared to the already codified meaning-bearing elements of language, particularly in the structural linguistics of Ferdinand de Saussure. It was by definition a semiological effort in an area of human society generally agreed to have a profound communicative power (after Leni Riefenstahl’s Triumph des Willens on the 1934 Nuremberg NSDAP Reichsparteitag no one could harbor a doubt about that). That attempt at a language based cinematic semiology was not successful, leaving us without anything resembling a phoneme, or a morpheme, or a sememe, even an alloseme of film that made a useful tool in the film student’s armament.1 Nevertheless, the phrase “language of film” lingers on, with all its tacit assignment of a linguistic nature to film. It should have a meaning, but it doesn’t, since there is finally an irreconcilable difference between linguistic nature, which is language, by which is meant human speech, and anything else. While films are filled with language – the speech of the actors, the lettering on signs, headlines of newspapers, or the titles of the silent era, they are not themselves speech, and can only be called “language” by the same absence of semantic precision that enables phrases such as “the language of flowers” or “the language of painting.” In its

174 Strother B. Purdy

present lingering state the cinema-language link has come to be no more than shorthand for the techniques of film, the tools used in order to make a motion picture – the studio system, the actor, the director, the camera, the boom, the traveling, the close up, the cut, and so forth.2 All of this is important, since films could not be made without tools, but it is no more than the means to manufacture a film, not an analysis of how film is an art – an art of communication, or art tout court. When that art leaves a powerful impression on you, it is not simply a matter of fooling you into thinking a painted backdrop is a landscape or a computer image is reality. Special effects have a very low shelf life. The rocket ship in Fritz Lang’s Die Frau im Mond (1928) was breathtaking, but made to look like cardboard by that in Menzies’ Things to Come (1936) which, along with every other such effort, was blown out of memory by Kubrick’s stupendous 2001 in 1968, then reduced to trivia by the Star Wars series (1977 onward). Something deeper is involved. It can be approached in a way by negatives: As far as the actors and their lines go, it is what makes a film not like a stage play, whether or not a stage play is in the film – see L’Année dernière à Marienbad. As far as photography goes, it is in what makes the photographic images not like those taken, say, on a family picnic, whether or not family picnic is in the film as in Heimat or Picnic at Hanging Rock. As for sound, meaning music and anything else other than speech, it is in what is not like the experience of listening to a phonograph record or attending a concert or simply hearing noises in the street – see Meeting Venus. All of which seems simple – we all know how it works – but it has so far eluded satisfying analysis and may, indeed, escape words altogether .3

2. The shot and the icon Is there a common element in these locations-of-film-art? Clearly enough it is photography. What then is the unit of photography in film? Most often is called the shot. Every film is made up of shots, individual camera operations, generally a large number. To call a shot the filmic equivalent of a word – or even a sentence – in a literary work of art has tempted some. However, not only does it suffer from the metaphoric dead end I have already mentioned, the gulf between language and the visual, but also the problem remains that any shot may contain a profusion of the artistic elements as well as the techniques of film. In addition, while a shot in a filmed conversation, for instance, may last only a few seconds, one actor in closeup speaking and then the camera making another shot to show another replying, a determined director can, like Aleksandr Sokurov of Russian Ark (2002), make an entire ninety-six minute film in one shot, never turning off the camera. That could be said to have worked, if barely; Andy Warhol’s 8 hour single shot Empire (1960)

The language of film is a matrix of icons

is a famous example of how to render the concept absurd. The point is this: if we are to have a usable basic unit with which to analyze film, it must be far smaller than the shot. It must be a universal visual unit, and not a limited visual technique, like the close-up or jump cut or traveling pan. It must be a semiotic unit, whether or not it is so named, which inevitably points us to Peirce’s icon, described by W. B. Gallie as “a sign whose...mode of signifying depends on some likeness, qualitative or structural, holding between it and its object. ”4 A film is a sequence of photographs, each one an icon and containing icons. As photographs linked together to give an illusion of motion, they are irreducibly cinematic. They act on the viewer in combination; that combination I call a “matrix of icons”.5 By “matrix” I mean “a place or enveloping element within which something originates, takes form, or develops.”6 It could well be argued that this introduction of a “matrix” into which the icon is drawn has the effect of gratuitously complicating the already difficult matter of the interrelation icon – index – symbol. Let me hold off a defense of that until I have introduced another abstraction, which is convention . It is by convention that filmgoers take a two-dimensioned variegated shadow on the screen to be a three-dimensioned object, or a human being. These shadows are icons by their (conventional) likeness and their cultural coding; otherwise filmgoers would not know what they mean. But it doesn’t stop there: we suspend the disbelief that common sense dictates, but there are limits. When a screenplay dictates that an actor draws a gun and fires it, we agree for the purpose of watching the film that the icon-shadow on the screen signifies a gun. But we add a further convention of our own, which is that we agree that we are not threatened when it is pointed our way, a matter problematic for early cinema audiences, among whom, as legend has it, cowboys would draw their non-shadow guns and fire back at cowboys on screen. And this basic realization coincides perfectly with one of Peirce’s strongest and most suggestive definitions of the icon, by which it has both likeness and convention: “Anything... is an Icon of anything, in so far as it is like that thing and used as a sign of it.”7 The result is, of course, a blurring of the boundaries icon –index – symbol, awkward but necessary, for us if not for Peirce, to avoid the argument of regression (cf. Sonesson 1997:740). Convention is by definition something made, in the sense by human agency, not natural, in the sense occurring in nature without human agency, and thus cannot be “regressed,” and protects the icon from falling into a universal signing, by which everything is a sign of everything else. Such an icon has its conceptual purity compromised while its usefulness in cinematic analysis greatly enriched. We need only recognize that film is constituted by icons in enormous profusion, and that only certain ones interrelate. It is therefore I propose this extra term, matrix, for nothing more than “related filmic icons,” as well as

175

176 Strother B. Purdy

proposing that we identify “unrelated filmic icons” as “empty” or “unused.” In both cases the setting or frame is contextual, “in the film we are now looking at” and not elsewhere. Thus it is that the filmaker is free to make new, and what in another context would be called unconventional , but are really “re-conventional-ized” uses of the icons he films. A use he wishes us to adopt, as indeed we must if we are to fully appreciate his film. A photograph of a glass ball is an icon of a real glass ball, as well as, potentially, an icon of a balloon or of anything spherical, each link to its object having a touch of convention to it.8 In the hands of Orson Welles, by his own invented convention, as we shall see, it becomes an icon of a soul’s core. And on the point of that tiny icon rests the whole dramatic weight and, I think, a major part of the international fame, of Citizen Kane, regularly nominated as one of the 10 greatest films of all time. That film is so advanced in the historical development of film that I wish to cover two earlier stages before returning to it.

3. The example of Chaplin Now for some examples of “analysis by icon”. I’d like to start chronologically, since the filmmakers of genius learn from their predecessors. One of the greatest of these was Charlie Chaplin. His stunning accomplishment was to make film universal, mocking the idea that cultures, like their languages, severely impede communication. His title cards had to be translated to make as good sense in Yokohama as in Manchester, but almost without exception the great moments in his films need no non-visual prompting – his struggle to get up while stepping on his cape in The Rounders, his acceptance of a flower from the blind girl in City Lights; his doomed attempt to catch up with the assembly line in Modern Times, are as fresh and true and affecting today as they were ninety years ago. They are internationally recognized icons, pure film, as well as indices of the human condition. In their independence from language they are not so much comparable to mime, for Chaplin asks us to imagine no material object we cannot see, as they are comparable to music.9 At the same time Chaplin is the master of the icon that appears to resemble nothing but its object, stripped of any potentiality, but in effect acts quite otherwise. It merges into and displays all the imaginative power of a matrix, in the most remarkable cases outstripping the observer’s powers of analysis. Take his trousers, surely the most famous in film – as worn by The Tramp. There, with the sadly inappropriate dignity of the bowler hat and the boulvardier’s cane, they project a negation of potentiality, of ruin. Yet they are the locus of mysteries the hat and cane never share. No one to my knowledge has managed a coherent analysis,

The language of film is a matrix of icons 177

semiotic or otherwise, of one of the most affecting moments in The Great Dictator (1940) where Chaplin as Hitler is giving a screaming speech in the best Teppichfresser manner. At the end of a rhetorical high point he pauses for breath, steps back from the microphones and takes a sip from a glass of water. Then he pulls out the front of his trousers and pours the rest of the water down inside (see Fig. 1). There is not the requisite convention here, so no symbol, just an icon of – film magic. We may well say it is the defiance of convention , the impossibility of symbol, that gives the scene its power. What is more unconventional than pouring water down your trousers? It is an index of nothing beyond the act of pouring, yet something else is being communicated, as unanalyzable, and as profoundly stirring, again, as music. “Ah, wait,” you say, “Hitler is a madman, since only a madman would pour water down his trousers”. That is too simple a dismissal of the evocative power of the gesture – in Chaplin’s hands. Why is it so evocative? Not only because of its odd appeal in itself, but because it belongs to the ‘trousericon-matrix’, well established in the early Mutual and Essanay silent shorts as the modus operandi of The Tramp, then brought forward again by Chaplin as the little Jewish barber in The Great Dictator, where its iconic availability is reinforced in the opening World War One sequence. There Charlie, a German soldier at the front, and no madman (Hitler was also an obedient and dutiful German soldier at the front) dutifully pulls the pin on a grenade and goes to throw it at the enemy, only to lose his grip on it and have it slide down the sleeve of his uniform and end up – inside his trousers! The chastened but not yet subdued Freudianist critics could have – probably did have – an exciting time with this, allying it to their established cliché of the Little Tramp as infantile eunuch. We can only note how far from cinematic such a laid-on analysis is. There are no Freudian icons, or, as the story of the cigar established;10 once you assign one, the argument of regression takes over and every icon is laden with the same charge. Because the Dictator is played by the Tramp, the Dictator must be somehow part-Tramp, and that part is the trousers. The hat and cane are ruled out by the Dictator’s costume, but even a dictator depends on trousers, with their sadly human contents. Once the trousers come to the fore (opening up their potential of signification, which is typically iconic), all Chaplin’s magic follows. A matrix easier to deal with is that of One A. M. (1916). Drunken Charlie staggers home to go to bed, only to find he can’t get there, thanks to the determined opposition of what in daily life is jokingly called “a conspiracy of inanimate objects,” as when one’s efforts to get to work on time have to face a bar of soap that gets under foot in the shower, a missing cuff link, a coffee maker that refuses to boil, and a set of car keys that have hidden themselves behind the sofa. Cinematically such objects are unremarkable as icons, the “conspiracy” an absurd assumption that defies common sense. Here the clock can’t know what the bottle

178

Strother B. Purdy

is doing, nor the bottle the bed, the table the door, the coat rack the stairs, yet they all work together to frustrate Charlie’s efforts. His progress to bed is converted to running a gauntlet, each step of the way diagrammatically reflected by a contrary object. Progress must be made upward, first from street level to the front door, which once reached is found to be locked. The key? Only after climbing through the window does he find it in his pocket. He can then properly enter the house, only to be attacked by the rug inside the door. Climb the stairs? Impossible. In the end the upper floor can only be reached by climbing the coat rack (see Figure 2). Finally the tragic reversal as the bed, the fixed goal, turns out to have motion and comes down, crushing Charlie with it. The physical ascent and then falling back is imitative form of the human effort to rise, to win through, to conquer and to survive; the contest with the falling bed an imitative form of the endless setbacks, the Sisyphean blows received and the rising from the floor to take another blow: that is the appeal of the Tramp. The filmic art of One A.M. is a matrix of object icons enabling a diagrammatic enactment of iconic downwardness – hang up cloak, it falls down, fall down stairs, fall down drunk, stay down, beaten down, done. Yet key is key, rug is rug, stairs are stairs, bed is bed, pure icons that Chaplin converts to mental states of incapacity. Only Laurel & Hardy’s The Music Box could ever compete, and that at a distance.

4. The example of Eisenstein: October Next after Chaplin, I see – as the world sees – Eisenstein as a filmaker of enormous importance; one who has much to teach us about the elements of film art. He is arguably the greatest of all theorists of film on account of his matching extensive theoretical writing with a practice that gave to the world masterpieces like Potemkin (1926), October (1928), Alexander Nevsky (1938) and Ivan the Terrible (Pts. I& II 1945, 1958). He wrote of “the language of film,” but by that he meant something specifically and only cinematic, namely montage, which is the setting of images into creative opposition, putting side by side apparently unrelated and contrasting images that create a third image that cannot be seen on the screen but is produced in the viewer’s mind. His description of the process could serve equally well as a description of the signifying process instigated by the icon: What is this [montage] method remarkable for? First and foremost, for its dynamics. For the fact that the desired image is not ready-made, but arises, comes into virtual being. The spectator witnesses the birth of an image as conceived by the author, the director and the actor....The spectator is made to traverse the road of creation which the author traversed in creating the image....(Eisenstein 1946:77)

The language of film is a matrix of icons 179

Can the spectator be made to traverse such an inner road, belonging to the mind of another? Eisenstein could not always manage it. His enormous influence laid out a terrain that would be littered for decades with the wreckage of innumerable art films, even ambitious studio films, rendered mute by uninspired application of montage theory. Harps and bicycles can’t be relied upon to speak for themselves, any more than Michael Snow’s windowsill in Wavelength (1967 ) or the Guggenheim museum ramp in Matthew Barney’s Cremaster 3 (2002). The so-called Kuleshov Experiment could serve as a warning as well as an inspiration.11 When the method works, and it does work, it creates an iconic matrix. Eisenstein’s 1928 film of the Bolshevik revolution in St. Petersburg, October (shown in the West as Ten Days That Shook the World), is very much a case in point. Let us take the arrival of Kerensky at the Winter Palace, specifically his ascent of the great staircase that leads to the imperial apartments. Eisenstein makes sure we see the ascent as diagrammatically reflecting Kerensky’s doomed rise to power by intercutting between the titles he assumed before his overthrow – “Dictator” “Generalissimo”; “Minister of Navy – and of Army” (oddly reversed in most prints of the film), each one appearing after a few more steps. By the appearance of the last of these, the viewer may have come to suspect Kerensky is doing too much climbing, which of course he is, from Eisenstein’s Soviet point of view, but in a diagrammatically iconic way, for there aren’t that many stairs from ground level to the piano nobile in the Winter Palace. He is being refilmed, forced to climb stairs in the director’s mind (see Figure 3). I take this as an Eisensteinian equivalent to Chaplin’s trousers, amounting to a “stairs-icon matrix.” Eisenstein was fascinated by Piranesi’s famous Carcere series of horrific etchings, in which he noted the “runaway staircases” in those ghastly prison interiors where “some planes...carry the eye into unknown depths, and the staircases...extend to the heavens, or in a reverse cascade...rush downward” (Eisenstein 1977:91). There he saw something comparable to “the emotional image of the elemental headlong descent of human masses down the Odessa staircase.” (ibid.) That last is a reference to the sequence in Battleship Potemkin, where the crowd greeting the revolutionary sailors is attacked by soldiers in a great rank, rifles levelled, descending the steps on which they are gathered. Repeated shots make the steps into an endless downward progression, indexically pointing to the Czarist government’s descent into depravity and murder. So do the Winter Palace stairs now leave their iconic relation to the real ones, and become an index of a mental state, the insane grasp at the autocrat level of power by the upstart Kerensky. Without Eisenstein’s own notes it would be difficult to realize that he sees Kerensky at least partly as a comic figure: he appears to be ascending on these stairs, but he is not. Here the influence of Chaplin, so clear in general, may be

180 Strother B. Purdy

suggested as specific. These are the stairs from One A.M. but unreally doubled, one staircase on each side of the same surreal room, suggesting (drunken) illusion; and, while they can be repeatedly climbed on, they are never climbed to the point of reaching the top, attaining the icon of success (diagrammatic, since he is trying so hard but not succeeding). In this light the stairs of the Winter Palace become at once the nightmare stairs of Piranesi and the nightime stairs of Chaplin. To tread on them is to insure a fall. So can iconiity enable a complex metaphoric comparison. Once Kerensky reaches the top of the stairs, he is at his goal, the doors of the Czar’s private apartments, where Czar-dom awaits him. Repeated shots of his well shined boots are by now drawing us into the creative function Eisenstein describes, and what I call drawing of icons into a matrix; it is up to us to match those polished boots to two separate and equally simple iconic images of boots elsewhere in the film, one a seemingly endless sequence of booted legs going back and forth through a swinging door at the Smolny Institute, headquarters of the revolutionaries; the other a close-up of the shapeless and dusty boots of the soldiers of a pro-Bolskevik machine gun battalion, being led through the streets under arrest. All of these boots show wear; they have seen service. What is Kerensky’s service? We are teased into proximity of a mental image of nothing. The next image is of a great metal bird revolving, perhaps a silver peacock, with shining silver feathers spreading (see Figure 4).12 The doors before which the polished boots await now open. Kerensky enters – the Imperial apartments! No collapsing bed for him! The bird, again a separate image, now turns back, then twists its head about, showing one eye. Is it a mechanism for opening and closing the doors? We never see it triggered. In the absence of any help the film gives us in interpreting its road to creation, we are left with magic. A miraculous metal bird, seemingly endowed with life, an icon of a bird, its turning an index of turning. For Eisenstein a symbol of – Tsarist vanity? Waste of the people’s money on costly baubles? Eisenstein does descend to such Marxist-Leninist bathos, of necessity we might say, considering the world he lived in, but typically he has more up his sleeve. This image has too much mystery, too much of just the shock of oppositions Eisenstein wanted in montage. Ascent / polished boots / door / glittering metal bird that seems alive. What Eisenstein wants, or supposes he wants here, is meiosis, the diminution of Kerensky, no more than his shiny boots, as vain as a peacock, but it is not all we receive, for Eisenstein’s irrepressible cinematic genius and his vivid imagination have far overshot the Marxist-Leninist script. The Winter Palace becomes a domain of wonder, its ingredients darkness and light, polished metal, doors that magically open, the work of art that has taken life, Yeats’s immortal and scornful bird of the Byzantine emperor to readers of English poetry, or Ali Baba’s cave. It is a silent montage of icons that transmits something

The language of film is a matrix of icons

more than a critique of the March revolution, something like the core of “a stirring work of art”.13 Krakow’s emblematic cockerel of Poland’s Golden Age is much more similar to this bird. The process is as Eisenstein imagined it, but it doesn’t keep to the rails he laid down. Once again, there are no symbols, only icons. Under their combined stimulation our imaginations carry us off into symbol and story, no more responsible to the filmmaker who assembled them than we are to Beethoven for what we imagine as we listen to the Seventh Symphony. Such is the power, and weakness, of montage.

5. The example of Welles For my last example I turn to Orson Welles and his Citizen Kane (1941). Its technical hallmark is its depth of field; its narrative innovation the mixing of time frames and differing points of view. Its nucleus lies, once more, with the icon, deeper than the shot, and thus unconcerned with the greater shot length that depth of field enables. That lack of concern trivializes much of the once raging controversy among film critics over montage vs. depth of field and mise en scène. Depth of field is here made memorably effective: when Kane enters the cathedralsized living room of his home and sits in his usual chair, he is about fifty feet away from his unhappy wife Susan, yet they are both in focus. Talk about diagrammatic positioning! a bit obvious, but “he was distant with her” and “they lived together but apart”; all put into one image, an effect that would require numerous shots, non-visuals, spoken lines, if done in any other film style – this is wonderfully cinematic, and has no need for speech. What the deep focus wide angle single shot does is make things easier, and, at the same time, harder for the viewer, who is once more immersed in what Eisenstein called the creation of the image. Easier because the puzzlingly abrupt juxtaposition of contrasting icons in montage, such as stairs to boots, door to bird in October, is absent; harder because the constituent icons are spread across a wide visual field containing a great number of non-constituent icons, belonging to no matrix, and now their identification, the sorting-out, is up to the viewer. In montage the viewer must work out the indices and trace their development into symbols; in deep field he must locate them before he can start, but they are in a sense “laid out for him”. The emotional crisis in the film comes when Susan leaves Kane. In a twoshot he stands in her bedroom arguing with her, trying to convince her not to leave him, and our attention is naturally drawn to the two actors and their words. But their heads are high in the frame, and the deep focus encourages our eyes to wander. Low down in the frame, and off to one side, is the head of a doll lying on a

181

182

Strother B. Purdy

shelf. High up lies the ceiling of the room, also in focus and thereby made equally available to our gaze. Now the doll, being only a part of the shot, is easier to handle than Eisenstein’s bird, for we may presume nothing is lost if we ignore it and watch the stage-play part of the scene, Kane and Susan speaking their lines just as they would on stage. It is easy to “locate”, as an object just where it is, lying on a shelf. In fact my illustration (see Figure 5), as a still photo, is much more conducive to its coming into awareness than viewing the film with its distractions of motion and speech. Nevertheless it is a critical iconic element, reflecting the whole from its position at the edge of the frame. It grows and looms, dramatizing the the fact, and the facing of the fact, that Kane has always treated Susan not as an equal or even as a fully human being, but as a doll, a pretty thing to own and to decorate his life, to entertain him. It is also an icon, a likeness of a second Susan, the one she is now separating herself from by leaving, leaving it in the room and becoming one, or whole, again.14 That leaves a baffled Kane alone, his riches and his power for the first time useless to him. He starts to tear the room apart. How is the matrix to be identified? In one sense, it is the bedroom, the place where an image is taking form, the secret of the film, so teasingly kept from the viewer until the last scene. Kane’s rage of destruction is pure imitative form; he enacts diagrammatically the scenario of a man destroying a bedroom from pique. If he can obliterate the bedroom he will have rid himself of the woman’s existence and thus her rejection of him, which in her absence has been transferred to her place and her possessions. They have joined in Kane’s mind to become her icon, his very map of her. But this is not all, as indeed it is the stuff of the most ordinary and lamentable human drama (“ Bitch walked out on me – I’ll show her!”), which is hardly a triumph of film art. What there is beyond lies in the matrix of icons: cued by the head of the doll he is left with. On this level the imitative form of the effort to rip, tear, lift and throw the furnishings is a struggle: a struggle with himself, a furious effort to defend himself from the vista of self-awareness, the look into his soul that Susan has opened up before him. Charles Foster Kane, master of the universe, wrestles with the reality of his puny moral being. The diagrammatically enacted scene comes to a sudden end as Kane discovers a paperweight with a country snow scene inside, a small glass ball amid the wreckage (see Figure 6). Kane picks it up, then leaves the room in retreat, with the lumbering gate and the stunned look of a broken man. Somehow he has faced the truth. As he carries the ball he shakes it, and floating material inside mimics a snowstorm. It is a scene to amuse a child. He murmurs “Rosebud”. And now it becomes clear that there was a third level of form enacting meaning in this scene: first the effort to destroy, take a symbolic revenge; second the effort to defend himself from a painful truth; now third, a search. There is a visual clue to it in that

The language of film is a matrix of icons 183

when Kane is done the bedroom looks exactly as if it has been “tossed” by burglars in search of valuables. A search stops when its object is found: Kane stops when he finds the glass ball. In order to interpret this we need to know how it came to be in Susan’s bedroom, and we don’t know. One possibility is that she has kept it for its association with him, in which case we have an explanation of the dramatic effect its discovery has on Kane. For he has discovered it in the deepest recesses of her space, an icon and an index by association of her mind and heart, which mimes that “she held him in her heart” purely visually, and that, too late, shows him his terrible loss, that she loved him, was not rejecting him in walking out but only giving in to her despair at his inaccessibility to her. And indeed the whole film is cast as a search, that of the reporter trying to get a “handle” on Kane for his illustrated obituary article, a search that comes to center on the mystery word “Rosebud”. That search ends for the viewer when Kane’s childhood sled is tossed onto the (funeral) pyre of his possessions and we see it was named Rosebud, lost during play in a snowstorm when the heavy weight of his inheritance fell upon him and took him from his parents. This together with the links to his childhood – the lost love of the mother, his lost homeland, his effort to find love by giving at the same time as his giving was undermined by a wish for revenge for lost love, a self-deceptive refuge taken in memory of his childhood – are brought around to make a full circle. Kane forging these links is what we watch through the film. We don’t see the whole design until the last scene; Kane sees it all here in the bedroom. He picks up the ball and takes it to his own bedroom, where he will hold it in his hand the day he dies. There is a little boy in it, carrying a sled, hastening toward the cabin that is his home. That boy is Kane.

6. Conclusion Since film is essentially visual form, using the icon to analyze film has a natural advantage. The icon is an image of likeness. It can be made to generate meaning in film as well as to render effect, in two ways: 1) by conversion to symbol by convention, or 2) with the support of other icons constitute a greater unit, a diagrammatic collocation of iconical features that I call a matrix. Hynkel-Hitler’s trousers to those of the Tramp; the great metal bird to the doors and the stairs of the Winter Palace; the glass ball to the doll and the child’s sled called Rosebud. It is by such collocation of icons that film diagrammatically and metaphorically takes on that critical part of its identity, and its power, which is visual. As a sidelight, adoption of iconicity to film theory would clear up a continuing problem, the misapplication of linguistics. For while film contains language, its artistic core of generating meaning escapes it altogether. There dwells thought too deep for words.

184 Strother B. Purdy

Notes 1. The main efforts, all in the 1960s and 1970s, were those of Christian Metz (1971a: 502–515; 1971b; 1977), Yuri Lotman (1977) and Umberto Eco (1968). A later critique of note is that of Gilles Deleuze (1985). 2. Many textbooks and general film studies cling to the language metaphor in one way or another, and their contents, being on the whole capable surveys of technique, seem indistinguishable whether or not their titles feature it. See, e.g., Rod Whitaker ( the former name of Nicholas Seare, 1970); Gianfranco Bettetini (1973); Daniel Arijon (1991); Inez Hedges (1991); Elaine Scarratt (2005). 3. ....escape words altogether...A daunting prospect, but no excuse to abandon the task. After all, look at the centuries of study devoted to music and painting, both of application to film, and their general inability to get beyond descriptions of technique failing to discourage their authors. 4. W. B. Gallie, Peirce and Pragmatism (Harmondsworth: Penguin Books, 1952), p. 116. As Peirce puts it in CP 1.369 “three kinds of signs which are indispensable in all reasoning; the first is the diagrammatic sign, or icon, which exhibits a similarity or analogy to the subject of discourse”, the other two being the index and the symbol; in CP 2.276 “An Icon is a representation whose Representative Quality is a Firstness of it as a First. That is, a quality that it has qua thing renders it fit to be a representamen.” [ CP 2. 274 “A Sign, or Representamen...”; CP1.25 “Firstness is the mode of being which consists in its subject’s being positively such as it is regardless of aught else.”] ; “a sign may be iconic, that is, may represent its object mainly by its similarity, no matter what its mode of being”; and in CP 2.279 “a great distinguishing property of the icon is that by the direct observation of it other truths concerning its object can be discovered than those which suffice to determine its construction. Thus, by means of two photographs a map can be drawn, etc.” 5. That combination is as well diagrammatic, since it has a strong indexical element. Film is also, like photographs, indexical, due to its process of production; there are also intrafilmic indices and connectors linking shots and sequences; cf. Nöth 1990:467–68. 6. “Matrix” in Webster’s New Collegiate Dictionary, 1953 ed. This corresponds to Peirce’s definition of a diagram as a spatial figure that enables the creation of something new--which only icons and symbols have the capacity to do. 7. CP 2.247, my emphasis, as quoted from Jappy 2004:9. 8. W. B. Gallie, op. cit., pp. 116–17, “ thus a soap bubble is naturally, or one might say accidentally, suited to serve as an Icon of an air-balloon....neither an Index nor an Icon nor a Symbol can actually function as a sign unless or until it is interpreted,” to which I would add, “by convention,” there being no other way in (natural , thus shared) language. Gallie’s image appears to come from CP 2.293 “ a constituent of a Symbol may be an Index, and a constituent may be an Icon. A man walking with a child points his arm up into the air and says, ‘There is a balloon.’ The pointing arm is an essential part of the symbol without which the latter would convey no information. But if the child asks, ‘What is a balloon,’ and the man replies, ‘It is something like a great big soap bubble,’ he makes the image a part of the symbol.” 9. Notice the remarkable similarity as to linguistic silence in these word of Wu Han, pianist of the Emerson String Quartet, “[-there is]...a feeling of universality. Beethoven’s music is not about himself. His music is about everything that moves and lives and breathes and feels, and

The language of film is a matrix of icons 185

has proven to be a great connector among humanity....Beethoven says everything that needs to be said, without a single word spoken.” (emphasis mine, quotation from Shattuck 2005:34). 10. There are many versions to this probably fictional story, by which when the master heard that it had been suggested that his fondness for puffing on cigars was due, according to the master’s own law of universal sexual symbolism, to an unconscious desire to suck on a likeshaped part of the male anatomy, he angrily replied, “there are times when a cigar is a cigar”. 11. According to film legend, Kuleshov took a single image of an actor wearing a neutral expression, then intercut it with widely differing images – a child’s coffin, a plate of soup, a girl, and asked for audience reaction. The audience was sure the actor’s expression differed each time, revealing sorrow, hunger, amorousness, and so forth. This is cited as an illustration of the power of montage, but surely illustrates as well how easily film viewers are led astray by images, and not always in predictable directions. 12. Actually the crowning figure of the great automaton clock made in London by the amazing James Cox (1723–1800), or by his artificer, appropriately named Merlin, that by a series of stranger-than-fiction-adventures came into the possession of the fabulous Grigory Potemkin, or into that of Catherine the Great and then given to Potemkin, and kept in the Hermitage, adjoining the Winter Palace 13. “We must once more command a method, a directive guide to embody in stirring works of art...” Sergei Eisenstein (1949:85). 14. It is more a matter of potentiality, the possibility of a different life, that is more dramatically relevant here than the outcome assigned to her, for, deprived of Kane, she descends into alcoholism and a pathetic career as a second rate night club singer.

References Arijon, D. 1991. Grammar of the Film Language. Los Angeles LA: Silman-James. Bettetini, G. 1973. The Language and Technique of the Film (Trans. D. Osmond-Smith). Den Haag: Mouton. Deleuze, G. 1985. Cinéma 2 L’Image Temps. Paris: editions de minuit. Eco, U. 1968. La Struttura Assente. Milano: Bompiani . Eisenstein, S. 1946 Notes of a Film Director. Moscow: Foreign Languages Publishing House, n.d. (Foreword dated August 1946). Eisenstein, S. 1949 Film Form: Essays in film theory (Trans. J. Leyda (ed.)). New York NY: Harcourt, Brace. Eisenstein, S. 1977. Piranesi, or the Fluidity of Form. Oppositions: A Journal for Ideas and Criticism in Architecture XI (Winter 1977): 83–103. Gallie, W. B. 1952. Peirce and Pragmatism. Harmondsworth: Penguin. Hedges, I. 1991. Breaking the Frame: Film language and the experience of limits. Bloomington IN: Indiana University Press. Jappy, T. 2004 (downloaded 10/17/2004). Iconicity, hypoiconicity. http:/www.univ-perp.fr/lsh/ens/anglo/PDE.html. Lotman, Y. 1977. Esthétique et sémiotique du cinema (Trans. S. Breuillard). Paris: Éditions Sociales.

186 Strother B. Purdy

Metz, C. 1971a. Propositions méthodologiques pour l’analyse du film. In Essays in Semiotics, J. Kristeva, J. Rey-Debove and D. Umiker (eds), 502–515. Den Haag: Mouton. Metz, C. 1971b. Essais sur la signification au cinéma. 2 Vols. Paris: Klincksieck. Metz, C. 1977. Le signifiant imaginaire: Psychanalyse et cinema. Paris: Union générale d’éditions. Nöth, W. 1990. Handbook of Semiotics. Bloomington IN: Indiana University Press. Peirce, C. S. 1931–1958. Collected Papers of Charles Sanders Peirce. C. Hartshorne, P. Weiss and A.W. Burks (eds). Cambridge MA: Harvard University Press. Scarratt, E. 2005. Teaching Analysis of Film Language and Production. London: BFi. Shattuck, K. 2005. Still immortal, still beloved, still heard everywhere. New York Times, Jan. 30, 2005, 34–26 AR. Sonesson, G. 1997. The ecological foundations of iconicity. In Semiotics around the World: Synthesis in diversity. Proceedings of the Fifth International Congress of the IASS, Berkeley, June 12–18, 1994, I. Rauch and G. F. Carr (eds), 739–742. Berlin: Mouton de Gruyter. Webster’s New Collegiate Dictionary. 1953. Whitaker, R. 1970. The Language of Film. Englewood Cliffs NJ: Prentice-Hall.

The language of film is a matrix of icons 187

Figure 1.

Figure 2.

188 Strother B. Purdy

Figure 3.

Figure 4.

The language of film is a matrix of icons 189

Figure 5.

Figure 6.

190 Strother B. Purdy

Liberature: a new literary genre?

Liberature: A new literary genre? Katarzyna Bazarnik Jagiellonian University

The paper defines liberature (Pol. liberatura) as a distinct literary genre whose constitutive feature is an organic unity of the linguistic content with its material form. It discusses the difference of the postulated genre from “the artist’s book” in which the emphasis lies on the visual rather than textual component. Through a discussion of works of Bryan Stanley Johnson, Stéphane Mallarmé, James Joyce and contemporary Polish writers, Zenon Fajfer and Katarzyna Bazarnik, it addresses the question of iconic qualities of liberature, and places the concept in the context of earlier theoretical reflection, especially, Carl Darryl Malmgren’s notion of iconic space in the novel. In particular, iconic compositional space can be identified as a touchstone for the liberatic character of a literary work.

1. Introduction: What is liberature? Liberature (Pol. liberatura) is the term proposed by Zenon Fajfer to describe literary works which consciously employ their material space in such a way that the text and its physical container – the book – form an organic whole. It was first used in this sense1 in an essay published in 1999 in Dekada Literacka (a Polish literary magazine) to accompany an exhibition of what its curators called “unconventional books” for want of a better term.2 Fajfer proposed that liberature should be distinguished as a separate literary type, beside the lyric, epic and dramatic. Its generic distinctness should not be associated with the kind of the speaker but with ways of presentation of the fictional world. These would embrace visual qualities of writing and of material on which it is inscribed. By consciously putting them to use, the author would shape the physical space of the work so that it would become a meaningful component. As Fajfer put it, “the architecture and the visual aspect of the work are no less important than the plot and style” – they also carry a message (Fajfer 2002:234; trans. K. B.). The term draws on Latin liber, i.e., “book”, to stress that it is a kind of literature which involves not only letters and words as an expressive medium, but the whole book in its material existence. The shape and structure of the book, its format and

191

192 Katarzyna Bazarnik

size, layout and kind of typeface, kind and colour of paper, illustrations, drawings and other graphic elements can be valid means of artistic expression (Bazarnik 2002:v). In such a framework the book is not, as Milton called it, “a transparent violl”, or, as Ingarden insisted, a negligible material foundation, but an integral part of the literary work, a visible and palpable text occupying a certain physically delimited space. As Fajfer points out: “[the] physical object ceases to be a mere medium for the text – the book does not contain a literary work, it is the literary work itself” (Fajfer 2002:234; trans. K. Bazarnik). Fajfer applies the term also to shorter works that display a comparable control of the physical space, such as, for example, Mallarmé’s Un Coup de Dés, visual poetry, and his own mobile poem “siedemnaście liter” (“7 letters”). What seems to matter is a “wholistic” approach to the piece, be it a poem, a collection of poems or prose texts, or a narrative. So liberature is the project of “total art”, or “total literature” in which artists-writers exert control over the whole design of their work. This does not necessarily mean that they must do the printing and bookbinding themselves, or that their work must necessarily divert from the traditional codex form, but if it assumes such a form, it is the author’s deliberate choice and not thoughtless following of the dominating convention. And the writer is free to choose whatever means and materials he or she finds appropriate (which draws on the other meaning of Latin liber) (Bazarnik and Fajfer n.d.: para. 2). Despite postulated freedom of expression, liberature, as proposed by Fajfer, has little in common with liberature propagated by a Spanish writer, Julián Ríos, beside its homophonous name. For Ríos emphasis is on a liberating force of literature, on expanding its thematic and expressive scope, and on breaking taboos; he derives it from “liberation” (Gazarian Gautier 1990:para 37–38). Fajfer accentuates the book as an expressive medium, and a conscious, wholistic design of the work. Ríos’s liberature is a subjective category, dependent on an individual taste and viewpoint, whereas for Fajfer it embraces a certain group of works that can be identified on an objective basis. When we consider derivate terms, Ríos calls someone who writes liberature a “liberator” and his work liberating; for Fajfer he is a liberary writer who writes liberary works. Yet derivate terms point to another dilemma: that of generic affiliations. Clearly, “liberary” is not the same as “literary”. By exchanging the transparent “t” for the bodily “b”, liberature seems to place itself in an opposition to traditionally conceived literature and in the realm of the visible and the tangible. The “b” stuck inside the well-known word, is taken for a misprint by those unfamiliar with the term. Hence, it makes the reader stop, look more carefully and see the alien in a familiar name – an intruding, conspicuous, bold presence within a transparent, abstract idea. Indeed, because of its emphasis on the spatial, visual and material

Liberature: a new literary genre? 193

liberature might be seen as another attempt at a synthesis of arts, or perhaps a distinct genre of visual arts. If so, is it not a fancy term for an old phenomenon known as “the artist’s book”? What, if any, is the difference between these two?

2. Liberature versus the artist’s book The artist’s book emerged in the early decades of the twentieth century in consequence of avant-garde experimentation with the visual qualities of writing, typography and layout, different ways of combining text and image, as well as with the form of the book itself. The name was first used to refer to carefully edited publications (often in limited editions) which were designed by a visual artist or co-designed by him/her and a writer. Max Jacob’s Saint Matarel, a prose poem illustrated by Picasso and published by Kahnweiler in 1911 in Paris, and Blaise Cendrars’ 2–meter-long leporello La Prose du Transsibérien et de la petite Jehanne de France (1913) with Sonia Delaunay’s graphic design, can serve as examples. In Poland the poet Julian Przyboś and the painter Władysław Strzemiński cooperated on a few collections of poems, the best know example of their innovative typography being Z ponad (1930). Later the term was extended to embrace other artistic projects in the form of or inspired by the book. A variety of forms covered by it was so great that it became necessary to invent more precise labels. Clive Phillpot, an American art critic and a curator at the Museum of Modern Art in New York, distinguished three types of artist’s books: just books, book objects and bookworks (Rypson 1992: para. 3–4). The first category embraces the “beautiful book”, that is, a carefully designed and edited book, usually illustrated, written and made by artists (or such books about artists), but which does not usually bring about any new formal solutions.3 Although it is definitely an aesthetic object with visual appeal, a paragon of designer’s, printer’s and bookbinder’s craft, it is still a traditional book. Book objects are various kinds of visual artworks inspired by the idea of the book. They may assume different “unbookish” shapes; this category includes installations and artistic projects which refer to the book as a symbol or a metaphor. The third of Phillpot’s types, the bookwork, is an intermedial form that stretches over the borders of plastic arts and literature. Bookworks are artistic objects in the form of the book which, however, experiment with the book’s structure, shape, and meaning. They tend to employ the paginal space as the smallest narrative unit, but the order of arrangement usually breaks with traditional conventions. This third category overlaps with liberature insofar as it emphasizes the unity of the plastic vision and the textual component.

194 Katarzyna Bazarnik

But what is the status of text in the bookwork and in the artist’s book in general? Phillpot settles this question by answering an objection raised to the artist’s book by writers. The objection consists in the fact that the term: defines [these] books exclusively in relation to the profession of the visual artist. While this may annoy writers who experiment with the form of the book, alternatives like “writers’ books”(?), or “musicians’ books” can still be swept up into the all-embracing category of book art: art dependent upon the book form. In any case, it is clear that visual artists have contributed most to the revitalization of the books as art over the last twenty-five years, and to the development of visual and verbi-visual languages articulated within the book form (Phillpot 1986/87:para.1; emphasis mine).

Evidently, text in the artist’s book is but one of several components, definitely not the most essential, and in book objects it may not appear at all. The artist’s book is made by an artist, or to be more precise – by a visual artist, and does not have to be read. It needs to be looked at and contemplated as a sculpture or an installation. Whereas a liberary work may have visual appeal, but unlike in Phillpot’s definition, in such a work it is the form of the book that is dependent on text, and not vice versa. Here lies the fundamental difference between the artefact of plastic arts in the book form and the liberary book. In liberature the material book is subservient to the word – the text dominates and determines the shape and structure of the work. B. S. Johnson, a truly liberary writer, stresses this point when he explains that whenever he departs from conventional forms, it is because the conventional form failed him: “because it is inadequate for conveying what I have to say. [...] So for every device I have used there is a literary rationale and a technical justification” (Johnson 1973:19–20). Liberature is thus something that Phillpot implied in the abortive coinage “the writer’s book”. It is created by the man of letters, not a graphic designer, painter, or sculptor. When Andrzej Bednarczyk brought out The Temple of Stone, a collection of poetry bound in concrete covers with a tiny piece of genuine rock placed inside the pages, he did it as a poet, not as a painter and sculptor, though by profession he is both. And Fajfer formulated his concept precisely as a counter-reaction to attempts at classifying Oka-leczenie (written by himself and the present author) as a work of visual arts.

3. Oka-leczenie – the first liberatic book In fact, the concept of liberature grew out of this book, the first one to be deliberately called a “liberary” work. We had to come up with an appropriate term to describe it, to give critics an appropriate tool to handle it if we wanted them

Liberature: a new literary genre? 195

to take it seriously. Otherwise it would have been labelled “the artists’ book” or a typographic happening, as someone called it, and relegated to the margins of literature. Instead of getting to libraries and bookshops, it would have ended up in galleries and exhibitions. But we wanted it to be read. Our priority in writing and designing it was not to make it visually appealing, but to find an appropriate form that would suit its subject: dying, gestation and birth, and the journey of consciousness through an intermediate region between death and another incarnation or nirvanic liberation. Admittedly, Oka-leczenie has a visually attractive shape: it consists of three codices joined by the covers in such a way that the covers form a three-part concertina (see Figure 1). The peripheral codices are set in hospitals: one in Poland at a man’s deathbed, the other in an English maternity ward where a Polish woman is in labour, accompanied by her husband and two midwives. The middle part contains two kinds of handwritten texts forming 64 hexastiches. One type of handwriting consists of one words written continually so that it resembles an electrocardiogram, the other has readable words that turn out to form a short poem revealed in full only on the last page of this part (see Figure 2). Oka-leczenie makes use of other visual devices, too. For example, it features several different typefaces, each type associated with a particular character; a series of anagrammatic sign-poems that show a transformation of “DENAT” (Pol. a dead body) into “DANTE” (the name of a child to be born) (see Figure 3), and a figural PolishEnglish palindrome.

Figure 1. The threefold concertina of Oka-leczenie by K. Bazarnyl and Z. Fajfer. Photos J. Śliwa

196 Katarzyna Bazarnik

Figure 2. The middle part of Oka-leczenie.

Figure 3. Anagrammatic transformation of DENAT-DANTE, Oka-leczenie (p. -9 – -8)

Liberature: a new literary genre? 197

But all such devices have been used before. The true innovation of this book is connected with a special technique of writing used in the two peripheral codices. Fajfer calls it emanational, because the text is derived from one “bottom” word in such a way that each of its letters “issues” a new word beginning with this letter. The resultant text “issues” another text and the procedure is repeated several times. Such an emanational text must be read in the order reverse to the order of writing it. Thus, the reader should read the initials of each surface word to recover the first underneath layer and then repeat the procedure until she gets to the “bottom” word. This resembles notaricon, only that the emanational text is a multi-layered notaricon (also called a multi-layered or Chinese-box acrostic; Mirkowicz 2003:80).4 This is supposed to convey how the old self (of the dying man) is shrinking into a single-word core from which a new self (of the baby) is developing, or how the posthumous consciousness journeys through an intermediate sphere between death and another incarnation. Finally, the text boils down to a single word “zodiak”, which functions as the karmic element. Its English equivalent gives rise to the text about gestation and birth. “Zodiak/Zodiac” emphasises the cyclicality of the presented world and indicates the settings of each outer part. The outer codices are interlaid with the middle codex reverse in relation to them. It could be seen as an analogon of nirvana: “to, o którym brak już słów” (that of which one is lost for words; Bazarnik and Fajfer 2000:LXIV). The reader can “break free” of the samsaric cycle of the two outer parts if he discovers a textual thread and a visual fissure leading to the middle part (see Figure 3). In fact, this image remains in an iconic bond with the title of the book, and the method of reading it. It points out to the cleft that cuts the eponymous “okaleczenie” (Pol. maiming, inflicting a wound) into its reverse “oka leczenie” (Pol. healing of the eye). The black page is ripped through to reveal a text hidden beneath it, just as the title word is split in order to reveal a phrase with the opposite meaning. Similarly, the reader is invited to cut off the initials of the surface texts in order to notice the invisible texts encoded in them. Or, to remain within the Buddhist framework, the reader needs to tear through the veils of maya in order to realise an illusory nature of surface reality. The structure of the book and the method of reading it is an iconic enactment of how Buddhist conceive of the world. (But it must be stressed that this is not “the canonical reading”; other ingenious interpretations have also been offered.) Some readers have been mislead by what they perceive as visual extravagance of Oka-leczenie. But its unconventional shape is inherently connected with the story delivered through the text and only partly through non-verbal images. Although very visual, the work hides its greatest

198 Katarzyna Bazarnik

secret and its essential meaning in invisible multilayered structures and cannot be adequately analysed if it were merely taken as an artist’s book, or as a purely literary phenomenon.

4. Liberature as a literary genre In fact, some artists who previously called their works “artists’ books” (e.g. Radosław Nowakowski, Andrzej Bednarczyk, and Marek Gajewski) welcomed “liberature” as a more suitable name. As long as their books, though clearly intended for reading, were called “the artists’ books” they hardly ever reached any reader. The half-willingly adopted label automatically located them in the sphere of fine and graphic arts where, unread, they fell into oblivion. Displayed in the gallery space, they were perceived as strangers in the world of literature and ignored by readers and literary critics, for, as I wrote elsewhere, “although we live in the age of ubiquitous Text, we do not actually read works of visual arts” (Bazarnik and Fajfer 2005:10). It is telling that since Nowakowski has started calling himself a “liberary writer”, he has become “readable”: his books have been noticed, read and reviewed, even though some of them exist only in a few copies (Fajfer 2003:9).5 Blake’s case is different, of greater calibre, and equally symptomatic. His works could not have functioned as examples of “the artist’s books”; the term did not exits in his times, though seemingly it would fit him very well. Therefore, the visual component of his works has been treated as mere illustrations. Thus, as additional and redundant, it could be removed from the proper literary essence. Blake’s illuminated poems are typically reprinted as texts expurgated of their original “pictorial excess”, sometimes accompanied by an odd, black-and-white reproduction of a few chosen plates. But since his “illustrations” have a value in their own right, they are displayed in art galleries as independent works of plastic arts. And yet it is clear that Blake conceived of his books as wholes. If he wanted to divorce the text from the image, he would have done so. Why does the rift continue if modern printing technologies have finally made it possible to publish his works in the form he intended them to have? Why do most editors ignore his intentions? Blake’s visual literature (or should we rather call it liberature?) demands respect for its integrity and calls for a critical approach that would employ the tools of both literary and art criticism. But this will be possible only if literary studies acknowledge a distinct and specific nature of such works. Neither “the artist’s book” nor concrete literature pay heed to writers such as Blake, B. S. Johnson, Mallarmé, and Sterne to name only a few. Distinguishing

Liberature: a new literary genre? 199

“liberature” as a separate genre could enhance research on their work. But to stress liberature’s literary affiliations one should perhaps use liberatic as the derivate adjective,6 to imply a generic similarity with the lyric, epic and dramatic. Because liberature does not want to alienate itself from literature. Rather, it wants to point out to its other roots associated with the picture and the development of writing. But the dominant tradition is the one which sees literature as the art of the ear, and of the spoken word. This goes back to ancient times when poets were singers, and their works were transmitted orally. This tradition has had some influential proponents from Plato, Horace, Lessing, the Romantics, to Ingarden, to name only a few. In it writing has been considered merely an imperfect record of a more sophisticated artistic oral expression. As McKenzie pointed out, Plato dramatised this in Symposium, which is an account of what Apollodorus remembered from what Aristodemus repeated about what Socrates’ and his interlocutors had said (McKenzie 2004:32–33). Here successive stages of corrupted transmission find their anti-climax in the written form of the dialogue. Wordsworth shared this attitude; in The Prelude he called the book a “[p]oor earthly casket of immortal verse” (V 165) and was irritated by “this vile abuse of pictured page!”. “Must eyes be all in all, the tongue and ear/ Nothing, he complained in the sonnet “Illustrated Books and Newspapers” (1846). Classification of literature as the art akin to music, corroborated by Lessing, resulted in marginalising such genres and works which exploited its visual and spatial potential. They have often been regarded as eccentric or experimental, the latter usually burdened with a pejorative overtone. This has been inevitable since, as C. D. Malmgren noticed, “fictions that capitalise on the materiality of the discourse, by exploring aspects of textuality, violate some of the strongest and most honoured textual conventions and [therefore] are often dismissed as ‘gimmicky’” (Malmgren 1985:46). So, typically, in his reflection on the nature of the literary work Roman Ingarden claims that the only “true realizations” of the literary work are those read aloud, since they truly embody an ideal which the creative will of an author can suggest intentionally (Ingarden 2000:72). For him written signs are merely an ignorable material foundation. If the text draws the reader’s attention to its visual qualities, this is detrimental to the reading process. It disturbs the correct, i.e. aural perception of the work, and places it outside “literature proper” (Ingarden 1988:447–452; 1981:277–8). However, to say that language is music is a metaphor, whereas to say that language is visible can be a statement of fact. Though usually taken as another paradoxical phrase implying a vivid depiction of something we “see” in our mind’s “eye”, it has a literal meaning if it refers to writing. W. J. T. Mitchell points out that in “visible language” we encounter

200 Katarzyna Bazarnik

the point at which seeing and speaking, painting and printing converge in the medium called “writing”. (…) “Writing”, as Plato suggested in the Pheadrus, “is very like painting”, and painting, in turn, is very like the first form of writing, the pictogram. The history of writing is regularly told as a story of progress from primitive picture-writing and gestural sign language to hieroglyphics to alphabetic writing “proper”. Writing is thus the medium in which the interaction of image and text, pictorial and verbal expression, adumbrated in the tropes of ut pictura poesis and the sisterhood of the arts, seems to be a literal possibility. Writing makes the language (in the literal sense) visible; it is (...) not just a supplement to speech, but a “sister art” to the spoken word, an art of both language and vision (Mitchell 1994:113).

Mallarmé expresses this in a more poetical way in L’après-midi d’un faune when he writes that the melody of Faun’s pipes “is the visible, serene and fictive air” (Mallarmé 1994:38). L’après-midi dramatises duality of language on several levels. It is printed in roman type and in italics and contains two words printed in capitals: “SOUVENIRS” (memories) and “CONTEZ” (to tell, a tale); the Faun plays twin pipes and chases two nymphs. The goddesses are usually read as symbols of the sister arts of painting and music (Weinfield 1994:179), but I would like to read them as incarnations of speech (singing) and writing. In chasing them, the poetfaun chases the essence of language through which he can perpetuate himself. Though dual, the nymphs are inseparable: the faun encounters them intertwined in sleep. To disunite them is to inflict harm, bruise them in the languor of duality. It is an ungodly act to split what is metaphysically united. The faun confesses that he sinned when he divided “the dishevelled/ Tuft of kisses that the gods had ravelled” (Mallarmé 1994:40). Initially, he is even unsure if they are phantoms of his imagination, they seem so immaterial to him. But these textual beings, “the women of [his] glosses” leave their signatures of “an unsolved mystery.../ The toothmarks of some strange, majestic creature” on Faun’s body. Similarly, the poet leaves the page with words imprinted on it, “distill[ing] sonorities/ From every empty and monotonous line”. Having separated the naiads, the faun succumbs, defeated, to the silence of the body and oblivion of his sin. The music gone with the nymphs, the poem ends with a note of confirmation that what is left is merely MEMORIES (“SOUVENIRS”) of the TALE (“CONTEZ”). The awareness of a tension between the inherent visuality and physicality of the written language and its musical qualities may be seen as a tragedy of a poetsinger. Perhaps this is what Mallarmé depicts in the shipwrecked language of Un Coup de Dés. But this can also be seen as a source of renewed inspiration, an impulse to seek ways of revitalising poetic language. L’après-midi d’un faune prefigures radically visual and spatial poetry of Un Coup de Dés, in which Mallarmé, trying to

Liberature: a new literary genre? 201

offer a musical score to guide the oral performance, concludes that “NOTHING/ WILL HAVE TAKEN PLACE/ BUT THE PLACE” (Mallarmé 1994:142; see Figure 4).

Figure 4. S. Mallarmé Un Coup de Dés (p. 18–19)

Liberature does not see this as an ultimate tragedy. It rather attempts to turn this into an advantage. It seeks its roots in forms of visual communication and happily embraces the memory of pictorial qualities of early systems of writing.7 It subscribes to the other tradition, which is about seeing the text as an image: a material entity that occupies a meaningful space. If writing is a visual medium, why not explore possibilities entailed in this?

5. Iconic spaces of liberature One of those few critics who have accounted for the materiality of the literary medium is Carl D. Malmgren. He believes that experimentation with the pictorial qualities of writing and the book is as valid as experimenting with other aspects of literary language and “constitutes an attempt to multiply the types of space avail-

202 Katarzyna Bazarnik

able for signification (Malmgren 1985:46). In Fictional Space in the Modernist and Postmodernist American Novel he provides an elaborate classification of fictional space (Malmgren 1985:60; cf. Figure 5) in which he distinguishes the iconic space which, according to him, is the aspect of the literary work that exposes or foregrounds materiality of literature. The iconic relation shapes “the sign-vehicle of fiction in such a way that the signified practice either resembles or becomes part of the signified reality” (Malmgren 1985:45). He differentiates four levels of iconic space: alphabetic, lexical, paginal and compositional. Alphabetic and lexical spaces are build up by devices that highlight the smallest units of writing, i. e., letters and words. For example, the anagram contributes to both these spaces, as it accentuates letters as constitutive elements of words, and foregrounds the words made of the same letters, thereby implying a meaningful relationship between them. It can express the idea of change and transformation, of “the same anew”, as in the above mentioned pair “DENAT – DANTE” in Oka-leczenie, in which the same components point to reincarnation and a metaphysical identity of the dead man and the baby. Paginal space is constructed when the page becomes a meaningful, narrative or compositional, unit. This pertains to Mallarmé’s typographical arrangement in Un Coup de Dés, Blake’s plates with illuminated poems, and Sterne’s black and marble pages. Finally, compositional space embraces practices that expose the way in which meaningful elements (units of signification) are put together to form a whole. This may result in some radical solutions, such as placing unbound sheets in a box, as did Saporta, Johnson, and Nowakowski. But it can entail some subtler ways of controlling the space of the volume. Herbert’s The Temple in which the position of every poems is determined by its function in an overall scheme is an instance of that. Dante’s carefully calculated Divine Comedy is another. Perhaps less known, but equally astounding is Joyce’s control of the physical volume in Finnegans Wake. The novel intended to embrace “the whole world” is an iconic model of the earthly globe. A narratorial comment promises: “I’ll make you to see figuratleavely the whome of your eternal geomater” (Joyce 1975:296.30)8 and later reminds the reader that “once done, dealt and delivered, tattat, you’re on the map (623.33). The text has “this tendency (...) to expense herselfs as sphere as possible, paradismic perimutter” (298.27–28). But the passage that most explicitly defines a topographic nature of the book comes from the chapter discussing characteristic features of a mysterious letter (identified with Finnegans Wake) discovered by a hen on a midden heap. One cannot help noticing that rather more than half of the lines run north-south in the Nemzes and Bukarahast directions while the others go west-east in search from Maliziies with Bulgarad for, tiny tot though it looks when schtschupnistling alongside other incunabula, it has its cardinal points for all that. These ruled

Liberature: a new literary genre? 203

Figure 5. A Map of Fictional Space by C. D. Malmgren (1985:60), with an inclusion of Liberature suggested by K. Bazarnik (2005)

204 Katarzyna Bazarnik

barriers along which the traced words, run, march, halt, walk, stumble at doubtful points, stumble up again in comparative safety seem to have been drawn first of all in a pretty checker with lampblack and blackthorn. (...) It is seriously believed by some that the intention may have been geodetic, or, in the view of the cannier, domestic economical (114.02–15).

The “geodetic intention” implies that the cardinal points of a textual map could be determined by measurements and calculations. In fact, it is striking that 628, i. e., the number of pages of this “circular” book, can be obtained from a neat equation for the circumference of a circle and sphere, where 2xΠx 100 = 628. If the opened book were visualised as a sphere whose surface were the text, its poles would fall in the middle, and at its beginning and end, which would be in close physical proximity. A concentration of allusions to Southern Polar expeditions in the arithmetical middle of the book suggests that this must be the Antarctic circle with the South Pole on pp 314–15. The Arctic should be then sought at the other “end”, between pages 628 and 3. Interestingly, half of the last page as well as pages 1–2 and a part of page 3 are left blank as if to suggest visually that this expanse of whiteness is the “cold mad feary father” (628.02), the Arctic Ocean. Further calculations and investigations of textual hints indicate that the Equator should be located on pp. 157–8 and 471–2, whereas the Tropics of Cancer and Capricorn on pp. 115–117 and 511–513, and 197–9 and 429–431 respectively. For example, “ver grosse O arundo of a long one in midias reeds” (158.07) may be the Equator; “Maraia” (158.18) possibly Marajo, the island in the mouth of the Amazon, and “she was a Black” (158.26) the Rio Negro (Bazarnik 1998:147–154, 2000:6–9). Thus, geographical allusions to particular areas are coordinated with the numbers of pages and function as “landmarks” allowing to navigate Joyce’s textual map. Gea-Mother-Earth is indeed displayed “figuratleavely”: through numbers on the leaves of the book, and though figurative descriptions of geographical topoi. Clive Hart, who provided an insightful analysis of spatial schemata in Finnegans Wake, is convinced that whenever Joyce had an intention to represent a spatial image he made every effort to “lay out the evolving material in a structural pattern which will reproduce [it] as closely as possible in terms of the physical disposition of the pages of the book” (Hart 1962:129). Thus, the textual map of the “gllll (... (...) ... (...) ...) lobe” (54.29 – 55.1) may be represented as the following diagram (see Figure 6). This mental as well as actual map integrates the text of Finnegans Wake with the physical space of the volume. As the narrator comments: “[t]o book alone belongs the lobe” (305.31), where “lobe” is read as “globe”. In Joyce’s ultimate blasphemy his book is an iconic re-creation of the Earth through word. Its seemingly neutral codex form is a meaningful component of the whole work, cocreating the iconic compositional space.

Liberature: a new literary genre? 205

Figure 6. gllll (... (...) ... (...) ...) lobe, 54.29 – 55.1 (Bazarnik 1998:153, 2000:9)

6. Critical perspectives In Malmgren’s map of fictional space liberature can feature as one of possible types of iconic texts. And the iconic compositional space can be used as a touchstone of liberature as it involves shaping of the book. Distinguishing this type could help to solve some critical dilemmas and be “particularly useful in understanding (...) why some problematical works have been ignored or misunderstood on account of some confusion of their category” (Harman and Holman 1996:231). Take, B. S. Johnson, for example. He has been labelled “a novelist”, but his consequent employment of the compositional space and a tendency to limit his themes to the autobiographical material placed him at the margins of literature. As Jonathan Coe (1999:xv) writes about The Unfortunates (the book in the box), “Johnson’s unclassifiable book – novel, memoir, call it what you will – was accorded at best a sort of grudging respect, tempered with a palpable, barely disguised disdain for its pretensions to originality”. But he puts Johnson’s marginalisation down to the writer’s embarrassing frankness and emotional engagement rather than to his formal inventiveness. Though an ardent admirer of Johnson’s work, the critic wonders how serious the writer was in playing with the material space of his books:

206 Katarzyna Bazarnik

whether there wasn’t some corner of his mind which recognized that the format of The Unfortunates was, apart from anything else, an excellent way of attracting publicity. Are we also allowed to entertain the notion that it might be fun, just for once, to publish a novel with unbound pages, held together in a box? I think we are. (Coe 1999:x)

To support his point he quotes Johnson’s preface to the Hungarian edition of the novel, which was published in the traditional codex form, where the writer humorously encourages the readers to place the graphic symbols identifying each section of the book in a hat and then by drawing them determine the order of reading. This is supposed to make up for the experience of handling loose sheets of the original. But it seems that Johnson is dead serious in the conclusion in which he recognises that this would entail an irretrievable loss, not only of the amused initial surprise, but of meaning conveyed solely in a direct contact with the original book. This procedure, does of course, involve a certain amount of clerical and administrative work on the part of the reader. But the amount is surely not excessive, and the lazy reader may of course proceed in his normal manner and accept the binder’s order. If he does choose not to join in the fun in this way that is, of course, his inalienable right; but then he will, however, be missing an experience not commonly (if at all) to be had: and perhaps the point, too. Which is also his inalienable right. What all Hungarian readers cannot help but miss is the physical feel, disintegrative, frail, of this novel in its original format; the tangible metaphor for the random way the mind works (Johnson 1999:xiv).

To read The Unfortunates in a bound form is like watching a video of a theatrical performance. This book is as much about chaos of random memories as it is about physical disintegration of the dying body. It literally falls apart in our hands, as if enacting the process it describes. The remains of the story are laid in the box as if in the coffin. Johnson found a powerful non-verbal iconic form to render a deeply distressing experience in whose face people are lost for words. Since all words fall flat when we watch our friends dying, and memory fails us when we try to remember the life before. Coe (1999:xii) picks up the melancholy and “intensity of [Johnson’s] remembered grief” in the style. But he seems to play down the emotional force that made the writer maim his work by separating it from the safety of its covers, and possibly risk losing some of its components by a careless reader. Is it because, unconsciously, he feels that applying such “tricks” in a literary work is outrageous or gimmicky – that it interferes with the purity of the oral medium? That it is too blatantly corporeal? But literature does not only consist in telling poignant or humorous stories; it also consists in inscribing them in books. That is why sometimes it should be called liberature.

Liberature: a new literary genre? 207

Notes 1. The term was earlier used by the Spanish writer Julián Ríos in his novel Larva: A Midsummer Night’s Babel (1983, English translation 1990). As I explain below, for Ríos any good literature is “‘liberature’, i.e., literature that strives for freedom” (Gazarian Gautier 1990: para. 37–38). 2. The exhibition, entitled “Booksday” was held from 16 June to 16 July 1999 as a part of Kraków’s Bloomsday celebrations. Its curators included Z. Fajfer, K. Bazarnik, and R. Nowakowski and was coorganised with the Wexford Arts Centre, Ireland. 3. But it may. Some of the Tryznas’ ingenious publications of Muzeum Książki in Łódź, Poland, experiment with the form of the book. E.g., Tomasz Tryzna’s redesigning of Blake’s The Marriage of Heaven and Hell in the form of the K-dron is definitely innovative in the formal sense, but one wonders what the point of doing that is. To display the printer’s and bookbinder’s mastery? To interpret Blake’s work? Or perhaps to produce one’s own work of art that enters into an intertextual or intermedial game with the original? 4. Fajfer has also written shorter poems using the same technique. For instance, “Ars poetica” is available in an animated on-line version, in which the evolution and involution of the text can be watched in motion. 5. Liberature Reading Room opened by K. Bazarnik and Z. Fajfer in Małopolski Instytut Kultury Kraków in 2002 has made many of rare books available for the reading public. It hosts a collection of works that can be classified as liberature as well as criticism on the spatial and visual aspects of literature. Its on-line catalogue is available at http://www.liberatura.art.pl. 6. Suggested by Finn Fordham in the private correspondence with the author. In fact, Fajfer’s initial suggestion for a noun describing “liberature” was liberyka (translated as “the liberatic”), by analogy to epika (the Polish name for the epic poetry and prose) and liryka (lyric poetry and prose) (Fajfer 1999:9; Bazarnik and Fajfer 2005:7). 7. Pondering on the nature of language some scholars go even so far as to claim that writing is a form of communication independent from speech. In fact, there are languages (kawi, the classic Tibetan, and wenyan, the classic Chinese) that exist only in writing. In many languages, e.g. in Japanese, the link between the written and the spoken forms is so close that it is writing that determines the development of speech (Majewicz 1989:236). 8. As all editions of Finnegans Wake have the same number of pages, the references are given in the following format: a page number followed by a dot and a line number; thus 296.30 means page 296, line 30.

References Bazarnik, K. 1998. Globalne spojrzenie na Finnegans Wake. In Bazarnik, K. and F. Fordham, (eds), 143–155. Bazarnik, K. 2000. Looking at Finnegans Wake from a Polar Perspective. The Abiko Quarterly 19 (winter-spring, 2000): 6–9. Bazarnik, K. (ed). 2002. Od Joyce’a do liberatury. Kraków: Universitas. Bazarnik, K. 2005. What is Liberature. Liberature Reading Room website, http://www.mik. krakow.pl/liberatura/1_idea_ENG.html, August 14 2005.

208 Katarzyna Bazarnik

Bazarnik, K. and Z. Fajfer 2000. Oka-leczenie. Kraków: prototype edition of nine copies. Bazarnik, K. and Fajfer, Z. 2005. Liberature. Kraków: Artpartner. Bazarnik, K. and F. Fordham (eds). 1998. Wokół Jamesa Joyce’a. Kraków: Universitas. Coe, J. 1999. Introduction. In B. S. Johnson, v-xv. Fajfer, Z. 1999. Liberatura. Aneks do słownika terminów literackich. Dekada Literacka 5/6 (153/154): 8–9. Fajfer, Z. 2002. liberatura. In Bazarnik (ed, 233–239). Fajfer, Z. 2003. Nie(opisanie) liberatury. Ha!art 2 (15): 9–15. Fajfer, Z. 2005. Ars poetica. Liberature Reading Room website, http://www.mik.krakow.pl/ liberatura/images/niby_czytelnia/en.html, (English version tr. K. Bazarnik and Z. Fajfer), August 14 2005. Gazarian Gautier, M.-L. Interview with Julián Ríos. The Review of Contemporary Fiction, Fall 1990, http://www.centerforbookculture.org/interviews/interview_rios.html, March 5, 2005. Harmon, W. and C. H. Holman 1996. A Handbook to Literature. 7th ed. Upper Saddle River: Prentice Hall. Hart, C. 1962. Structure and Motif in Finnegans Wake. London: Faber. Ingarden, R. 1981. Wykłady i dyskusje z estetyki. Warszawa: PWN. Ingarden, R. 1988. O dziele literackim (Das Literarische Kunstwerk, 1931). Trans. M. Turowicz. Warszawa: PWN. Ingarden, R. 2000. Szkice z filozofii literatury. Kraków: Znak. Johnson, B. S. 1973. Aren’t You Rather Young To Be Writing Your Memoirs? London: Hutchinson. Johnson, B. 1999/1969. The Unfortunates. London: Picador. Joyce, J. 1975/1939. Finnegans Wake. London: Faber. Majewicz, A. F. 1989. Języki świata i ich klasyfikowanie. Warszawa: PWN. Mallarmé, S. 1994. Collected Poems. Trans. H. Weinfield. Berkeley: University of California Press. Malmgren, C. D. 1985. Fictional Space in the Modernist and Postmodernist American Novel. Lewisburg: Bucknell UP. McKenzie, D.F. 2004. Bibliography and the Sociology of Texts. Cambridge: Cambridge University Press. Mirkowicz, T. 2003. Oka-leczenie (an excerpt from an e-mail message of 20 July 2000). In Ha!art 2 (15): 80. Mitchell, W. J. T. 1994. Picture Theory. Chicago: University of Chicago Press. Phillpot, C. Artists’ Booklets. (Printed Matter 1986/87 Catalog), http://www.printedmatter. org/about/booklets.cfm. February 25, 2005. Rypson, P. 1992. Space and personal time. (An exhibition catalogue, BWA Gallery, Szczecin), http://free.art.pl/at/teksty/tk14_an.htm, February 25, 2005. Weinfield, H. 1994. Commentary. In Mallarmé, 1994, 147–275. Wordsworth, W. 1846. Illustrated Books and Newspapers. In The Complete Poetical Works. London: Macmillan and Co, 1888; Bartleby.com, 1999, http://www.bartleby.com/145/. February 25, 2005. Wordsworth, W. 1799–1805; 1850/ 1973. The Prelude. In The Oxford Anthology of English Literature, F. Kermode and J. Hollander, (eds), 188–229. New York: Oxford University Press.

Meaning on the one and on the other hand 209

part iv Iconicity and conceptualization

210 Meike Adam et al.

Meaning on the one and on the other hand

Meaning on the one and on the other hand Iconicity in native vs. foreign signed languages1 Meike Adam*, Wiebke Iversen*, Erin Wilkinson** and Jill P. Morford** *Collaborative

Research Centre: Media and Cultural Communication, University of Cologne; **Department of Linguistics, University of New Mexico

The present study investigates the effect of language-specific knowledge on iconicity ratings of native and foreign signs. German signers judged the iconicity and similarity of DGS (German Sign Language) and ASL (American Sign Language) signs. We found that iconicity ratings were higher for DGS than for ASL signs, and that DGS signers perceived the ASL signs to be less iconic than ASL signers. Further, iconicity judgements of foreign signs were mediated by the phonological and conceptual similarity of those signs to DGS signs. Overall our results indicate that iconicity is not an ontological attribute of the sign itself, no ‘objective iconicity’ exists. We conclude that the perception of iconic reference depends on an interpreter and is shaped by language-specific experiences.

1. Introduction Signed languages are the natural languages used in deaf communities. Like spoken languages, signed languages differ across language communities in morphology, syntax and lexicon. Most signed languages possess a rich contingent of iconic signs, which might be related to their visual-gestural modality. After a controversial research history concerning the topic of iconicity in signed languages, a consensus has been reached that the investigation of the function and processing of iconic signs in signed languages constitutes a fruitful topic of research for gaining a better understanding of iconicity in general. Cross-linguistic studies of different signed languages in particular provide an opportunity to evaluate to what extent iconic functions are modality-specific and to what extent they are language-specific. In the present study we focus on the often stated claim that iconic signs are

211

212

Meike Adam et al.

self-evident, which means – as Wilcox (2004:140) formulates it – “if a form is iconic […] we should be able to predict its form from its meaning, and vice versa.” In contrast to this view, Peirce (see Elgin 1996:183) assumes that the meaning of an iconic sign is not self-evident but depends on an interpreter, who needs specific cultural and language knowledge to be capable of comprehending the iconic aspect of a sign. The assumption of self-evidence of iconic signs implies that iconicity is an ontological attribute and therefore its comprehension depends on the sign and not on the recipient. Thus, there should be little inter-individual variability in iconicity judgements regardless of who is making those judgements – people of the same language community, or people of different language communities. We conducted a study to contrast these views by comparing the respective responses of German deaf participants who were asked to judge the iconicity of signs from German Sign Language (DGS) and from American Sign Language (ASL). Before we describe this study in more detail, Peirce’s basic semiotic assumptions as well as the results of several previous studies concerning iconicity and signed languages are reviewed providing the theoretical and empirical motivations to challenge the concept of self-evidence and serving as a starting point for our study.

2. Iconicity needs an interpreter Peirce proposed a triadic relation between Sign, Object and Interpreter: Now a sign has, as such, three references: 1st, it is a sign to some thought which interprets it; 2nd, it is a sign for some objects to which in that thought it is equivalent; 3rd, it is a sign, in some respect or quality, which brings it into connection with its object. (EP 1, 38; CP 5.283)

The ‘sign’ constitutes the material quality of a sign. Its properties belong to it in itself and have nothing to do with its representative function; it does have a material form independent of its existence as a sign for an object. The ‘sign’ is not identical to the signified object but it refers to an entity. This referential relation between a sign and an object is always generated by an ‘interpreter’ and is not a pre-existing relation. The experience of this interpreter with ‘objects’ is necessarily mediated by signs, which construct the object in a specific way. A sign as a whole only exists in the interplay of its three references. Peirce also differentiated three kinds of signs2 that differ in the manner that they refer to an object: Firstly, there are likenesses, or icons; which serve to convey ideas of the things they represent simply by imitating them. Secondly, there are indications, or

Meaning on the one and on the other hand 213

indices; which show something about things, on account of their being physically connected with them. [...] Thirdly, there are symbols, or general signs, which have become associated with their meaning by usage. (EP 2:5 )

According to Peirce the qualities of iconic signs resemble those of an object. Examples for this kind of relation are onomatopoetic words or diagrams. An index refers to an object by which it is affected thus there is a physical connection not existent for symbols. Examples are weathervanes or fingerprints. A symbol is an arbitrary sign, like many words or signs in human languages. For all three kinds of signs – indices and icons are no exceptions – the referential relation is not selfevident but depends on an interpreter, or as Elgin puts it: “Something is an icon or an index only if it functions as such” (Elgin 1996:183). Although Peirce stressed the importance of an interpreter construing the reference, it has often been stated that the meaning of iconic signs is self-evident. That implies that the signified is unquestionable so no language-specific knowledge is required to understand the reference of an iconic sign. It may be tempting to assume, on first reflection, that the iconic signs of a signed language are interpreted in the same way regardless of an individual’s experience with the language. This assumption has been challenged by several signed language researchers, for example Grote & Linz (2003), who propose the concept of ‘arbitrary iconicity’, which emphasizes the role of construal in language (Grote/Linz 2003). Wilcox expresses a similar idea with the term ‘cognitive iconicity’ (Wilcox 2004; see also Taub 2001). He proposes that both the semantic and the phonological aspects of a linguistic sign are represented in a single multi-dimensional conceptual space. The construal of phonological form drives the perception of the iconic relation of form and meaning. The basic assumption is that “in all cases the perceived iconicity is not an objective likeness between a referent and a linguistic form but a mentally constructed correspondence between two cognitive products,“ (Grote/ Linz 2003:25) or as Wilcox puts it: “the iconic relation is between construals of real-world scenes and construals of form,” (Wilcox 2004:123). These concepts imply that iconicity and arbitrariness are not mutually exclusive properties of language systems but rather endpoints of a referential continuum (see Adam 2004, Grote/Linz 2003, Pelc 1986, Taub 2001 and Wilcox 2004).

3. Signed languages and iconicity Iconicity has not always been an unchallenged topic in signed language research. Hockett’s (1960) widely accepted claim that arbitrariness is a defining characteristic of language provided an impetus for the first linguists advocating the linguistic status of signed languages to argue against any function of iconicity in

214 Meike Adam et al.

signed languages (see Wilcox 2004:121). With the broader acceptance of signed languages as natural languages, current research focuses more on the contribution signed language research could make to fundamental questions on the topic of iconicity and language processing. For example, do iconic signs activate languagespecific neural networks or brain areas associated with the encoding of visual material? Are there differences in the processing of more and less iconic signs? Brain imaging studies have already shed some light on these issues. Klann, Kastrau and Huber (2004) asked if the perception of iconic signs requires specific neural substrates, for example, areas typically recruited for mental imagery. They found that the activation patterns for deaf participants did not differ for more and less iconic signs. Emmorey et al. (2004) examined whether the production of signs with a high degree of motor-iconicity3 engages different neural systems compared to non-iconic signs. These authors also observed no difference in neural activation for both sign classes. This was true even if the shape of a specific sign is indistinguishable from the shape of a pantomimic gesture expressing the same action. The neural systems underlying sign production mirror those engaged when hearing speakers name tools or tool-based actions in speech, but not those engaged when performing the manual action. These results indicate that signed language users process motor-iconic signs as part of a language system. It may be tempting to conclude that iconicity does not influence language representation and processing since distinctive neural networks have not been identified for signs differing in their degree of iconicity. However, there is behavioural evidence to the contrary. Grote and Linz (2003) demonstrated an influence of iconicity on the organisation of the semantic network of deaf and hearing signers (i.e., hearing cross-modal bilinguals4). Participants were asked to judge the presence or absence of a semantic relation between a DGS sign (deaf and bilingual participants) or a German word (hearing and bilingual participants) and a picture. The DGS signs were categorized into two groups on the basis of their iconicity – high vs. low iconicity. Some of the pictures highlighted the iconic base of the signs (the visual aspect of the referent encoded by the sign form), whereas the others did not. For example, the DGS sign for cow is made by placing a hand at the forehead where a cow’s horns appear. A picture of a cow’s horns highlights the iconic base of this sign, whereas a picture of a cow’s hoof does not. Nevertheless, both pictures are related semantically to the concept of cow and thus should have received the same response on the task. Deaf and hearing signers responded faster to pictures that highlighted the iconic base of the sign compared to the neutral pictures. In contrast, a hearing control group with no knowledge of DGS showed no differences in mean reaction times for both picture types. Assuming that reaction time is an indicator of the strength of a semantic relation, these results demonstrate that the iconicity of linguistic symbols influences the organization

Meaning on the one and on the other hand

of the conceptual system. A similar association of iconic aspects in signs and the semantic network organization was also reported for British Sign Language (BSL) by Vigliocco et al. (2005). Pizzuto and Volterra (2000) conducted a study to “clarify the role that culturespecific, language-specific, or language-universal, presumably iconic and/or perceptual features, may play in determining the relative transparency, opacity, or both of a set of signs of a specific sign language,” (Pizzuto/Volterra 2000:272). They used 20 transparent5 and 20 non-transparent Italian signs. Within the set of transparent signs one subgroup was very similar or identical in several different signed languages. Deaf 6 and hearing participants from 6 different national origins and language communities were asked to guess the meaning of the 40 signs. Both hearing and deaf participants guessed the meaning of some transparent signs correctly. Pizzuto and Volterra concluded that “there are some language- and culture-free, presumably universal iconic-transparent features of signs that may be perceived in the same manner by both speakers and signers.” However, all deaf participants scored higher than all hearing participants. Thus the authors propose the “existence of potential universals across sign languages.“ (Pizzuto/Volterra 2000:283) Finally, within the hearing group, the Italians scored higher than hearing participants from any other nation. This result indicates that there are also some culture-specific factors relevant in the perception of signs’ iconicity. An alternative explanation for the better performance of deaf participants relative to hearing participants is that they were able to guess the meanings of signs that were similar to signs in their native sign language. Pizzuto and Volterra did not control this factor, thus there is no information about the correlation of similarity and correct identification of the Italian signs. However, even if similarity is the crucial factor, one would have to ask why iconic signs are similar in different signed languages? In sum the evidence indicates that iconic signs are processed via the same neural structures known for language processing in general, but that iconicity has an impact on semantic organization. Individuals with more experience using signed languages exhibit superior performance in identifying the referents of iconic signs. However, we do not yet know whether this influence of iconicity is a language-general modality effect or whether it depends on language-specific knowledge. To pursue this line of questioning, we investigated iconicity judgements of signs drawn from two signed languages, controlling for the overlap in form and meaning between the two languages.

215

216 Meike Adam et al.

4. Cross-linguistic iconicity judgements The assumption of self-evidence not only means that one should be able to guess the correct meaning of a sign independent of language-specific knowledge, but also implies as a precondition that one is able to perceive the iconic features of a sign without language-specific knowledge. In order to evaluate the second aspect of self-evidence, a group of German deaf signers who had no competence in American Sign Language (ASL) judged the iconicity of ASL signs, as well as the iconicity of their DGS translations. The results were compared to iconicity ratings of four native ASL signers from the U.S. We predicted that if the influence of iconicity in signed languages is a general effect of the modality, the ratings of German and U.S. participants should not differ to a major extent. Such a result would indicate that the perception of iconic aspects of a sign is evoked by something like an ‘objective degree’ of iconicity independent of an interpreter knowing the specific sign. This pattern of results would corroborate Pizzuto and Volterra’s assumption of universals across signed languages. On the other hand, if the perception of iconicity is driven by language-specific knowledge, and in particular by the construal of form and meaning as inhabiting a similar conceptual space, a number of interesting predictions can be made. First, U.S. participants should rate the ASL signs as more iconic than the German group. Likewise, Germans should rate the DGS translations of the ASL signs as more iconic than the ASL signs. Third, the German signers’ iconicity ratings of ASL signs might be mediated by their knowledge of DGS. This question has not yet been investigated empirically, and so we included several conditions to explore what type of language-specific experience could influence judgements of iconicity, including (1) the degree of iconicity of the DGS translations of the ASL signs, (2) the perceived phonological similarity of a DGS and an ASL sign, (3) overlap in form, meaning or both, and (4) frequency or familiarity of the DGS translation. To sum up, we addressed the following major questions: – – – –

–

Do German and U.S. signers make similar iconicity ratings of ASL signs? Do German signers perceive DGS signs as more iconic than ASL signs? Does the iconicity judgement of a foreign sign depend on the perceived iconicity of that sign’s translation into the native sign language? Does the iconicity judgement of a foreign sign depend on either perceived or actual phonological similarity between the target sign and its translation into the native sign language? Does the frequency and familiarity of a sign influence the perception of iconicity?

Meaning on the one and on the other hand 217

5. Method Participants: Twenty deaf German adults (women 9, men 11, mean age 31 years [range: 23, 47]) were paid for participation. All had normal or corrected to normal vision and used DGS as their preferred language. One participant was a native signer; all others learned DGS in pre-school (age 3 or 4). Stimuli and design: Fifty ASL signs and the respective DGS translations were used. The stimulus material was based on a previous study from Grote and Linz (2003). In that study the frequency, familiarity and iconicity of 150 DGS signs were rated by a group of German deaf participants. Those 150 signs were translated into ASL. From this sample all those signs were excluded which were (1) fingerspelled7 in ASL or (2) deictic or (3) depicted basic human functions in a similar manner (e.g., EAT, DRINK, etc.).8 For example, the sign for nose in ASL and DGS is expressed by pointing with the index finger to the nose, and the sign for eat in ASL and DGS involves moving the hand to the mouth as if eating. Moreover those signs that have a low frequency in ASL or are classifier constructions were excluded. The degree of iconicity of the remaining 50 ASL signs was evaluated by 4 native ASL signers. Those signs were assigned to one of three groups by a native DGS signer according to their form-meaning overlap with DGS: –

–

–

ASL signs with the same form and a similar meaning in DGS. The ASL sign for banana for example is phonologically similar in DGS and ASL (see Figure 1). ASL signs with the same form but a different meaning in DGS. The hand configuration of the ASL sign for meaning for example is used in DGS to indicate the sign for technik (technology) (see Figure 2). ASL sign forms with no meaning in DGS. There is for example no sign in DGS which has the same form like the ASL sign for strict (see Figure 3).

Figure 1. The sign for banana in DGS (left) and ASL (right). The handshapes of both signs are different but movement and location are the same. Thus, both signs are phonologically similar.

218

Meike Adam et al.

Figure 2. Handshape, movement, orientation, and location of the DGS sign for technik (left) are congruent to those used in the ASL sign for meaning. Thus, both signs are of the same form but are different in meaning.

Figure 3. The ASL sign for strict. There is no similar sign in DGS.

In the experiment the German participants sat in front of a computer monitor. In the middle of the monitor a video with a DGS or an ASL sign was presented. The signs were presented five times with five different questions: 1. 2. 3. 4. 5.

Do you use this sign? How familiar is this sign to you? How iconic is this DGS sign? How iconic is this ASL sign? How similar is the form of the ASL sign and the corresponding DGS sign?

In questions 4 and 5 participants saw the ASL as well as the corresponding DGS sign in a video clip. Answers were given by moving a scrollbar with the computer mouse on a continuum from 0 to 100 (see Figure 4). The location of the scrollbar was associated with numbers, which were used for the data analyses.

Meaning on the one and on the other hand 219

Figure 4. Screenshot of a test item. The participants were asked “how familiar is this item to you?” and responded by moving the scrollbar.

6. Results The ratings of one participant were excluded, because he judged all DGS signs as not iconic at all (zero). Moreover eight items were not considered in any analyses, because the DGS translations were judged as unfamiliar or infrequent in the German deaf community.

6.1 Iconicity ratings by German and US signers compared Overall, the German signers tended to judge ASL signs to be less iconic on average (31) than the US signers (42), but the ratings did not differ significantly (F(1,21) = 2.40; p = .14). However, the German group judged the DGS translations of the ASL signs as more iconic (58) than the ASL signs (F(1,18) = 116.23; p < .001). This bias towards finding iconicity in native language signs is the first indication that the perception of iconicity is dependent upon language-specific knowledge (see Figure 5).

6.2 The relation of DGS and ASL iconicity ratings Although the mean iconicity rating of the DGS translations of ASL target signs by German signers was higher than the ASL iconicity ratings, it is still possible

220 Meike Adam et al.

that there was a systematic relationship between the two sets of ratings. However, an analysis of covariance revealed only a weak correlation between the two types of iconicity ratings (r = .32; p < .01). These results indicate that signers neither judge iconicity via a direct translation, nor does their perception of the iconicity of translation equivalents indirectly influence the perception of iconicity of a foreign sign.

6.3 Form similarity and iconicity ratings Two measures of form similarity were used in the study. First, a subjective measure of form similarity was gathered by asking participants how similar they thought the ASL sign was to its DGS translation on a scale of 0–100. The mean similarity rating of the ASL target signs and their DGS translations was 26. These similarity judgements were strongly correlated with the ASL iconicity ratings (r = .63, p