From Sign to Signing

From Sign to Signing

Iconicity in language and literature 3

Edited by Wolfgang G. Müller (Friedrich-Schiller-Universität Jena) and Olga Fischer (University of Amsterdam)
Amsterdam/Philadelphia: Benjamins 2003. Hb xiv, 441 pp. 90 272 2593 1 EUR 130.00 1 58811 288 8 USD 130.00


Preface and acknowledgments    
List of contributors    

Olga Fischer and Wolfgang G. Müller: ‘Introduction: From Signing back to Signs’

Part I: Auditory and visual signs and signing     
Klaudia Grote and Erika Linz: ‘The influence of sign language iconicity on semantic conceptualization’

William J. Herlofsky: ‘What You See Is What You Get: Iconicity and metaphor in the visual language of written and signed poetry: A cognitive poetic approach’

Axel Hübler: ‘Spatial iconicity in two English verb classes’

Keiko Masuda: ‘What imitates birdcalls? Two experiments on birdcalls and their linguistic representations’

Part II: Visual iconicity and iconic mapping    
John J. White: ‘Perspective in experimental shaped poetry: A semiotic approach’

Julian Moyle: ‘Where reading peters out: Iconic images in the entropic text’

Andreas Ohme: ‘Iconic repre-sentation of space and time in Vladimir Sorokin’s novel The Queue (Ochered’)’

Matthias Bauer: ‘“Vision and Prayer”: Dylan Thomas and the Power of X’

Christina Ljungberg: ‘Diagrams in narrative: Visual strategies in contemporary fiction’

Part III: Structural iconicity    
C. Jac Conradie: ‘The iconicity of Afrikaans reduplication’

Volker Harm: ‘Diagrammatic iconicity in the lexicon: Base and derivation in the history of German verbal word-formation’

Beate Hampe and Doris Schönefeld: ‘Creative syntax: Iconic principles within the symbolic’ Günter Rohdenburg: ‘Aspects of grammatical iconicity in English’

Wilhelm Pötters: ‘Beatrice: or The geometry of love’

Masako K. Hiraga: ‘How metaphor and iconicity are entwined in poetry: A case in Haiku’

Part IV: Intermedial iconicity    
Werner Wolf: ‘Intermedial iconicity in fiction: Tema con variazioni’

Elzbieta Tabakowska: ‘Iconicity and literary translation’

Part V: New applications of sign theory    
Jørgen Dines Johansen: ‘Iconizing literature’

Piotr Sadowski: ‘From signal to symbol: Towards a systems typology of linguistic signs’

Author index
Subject index


introduction (excerpts)


Olga Fischer and Wolfgang G. Müller                               University of Amsterdam, University of Jena        

Iconicity is one of the few fields of research in which the disciplines of linguistics and literary studies – both of which have regrettably drifted apart as a consequence of specialization – can fruitfully co-operate. We see iconicity (be it of an imagic or a diagrammatic kind) operating in everyday as well as literary language. And it is through the interdisciplinary nature of the iconicity symposia, through approaching iconic phenomena from both a linguistic and literary point of view, that we may develop a keener perception of the pervasive presence of iconicity in all forms of language. This will provide us with a better understanding of how language is structured and at the same time give us a deeper insight into the tools and methods used by poets and writers, leading to a fuller appreciation of the literary text itself.         

    This volume, a sequel to Form Miming Meaning (1999) and The Motivated Sign (2001), offers a selection of papers given at the Third International Symposium on Iconicity in Language and Literature (Jena 2001). The studies collected here present a number of new departures. Special consideration is given to the way non-linguistic visual and auditory signs (such as gestures and bird sounds) are represented in language, and more specifically in ‘signed’ language, and how such signs influence semantic conceptualization. Other studies examine more closely how visual signs and representations of time and space are incorporated or reflected in literary language, in fiction as well as (experimental) poetry. A further new approach concerns intermedial iconicity, which emerges in art when its medium is changed or another medium is imitated. A more abstract, diagrammatic type of iconicity is again investigated, with reference to both language and literature: some essays focus on the device of reduplication, isomorphic tendencies in word formation and on creative iconic patterns in syntax, while others explore numerical design in Dante and geometrical patterning in Dylan Thomas. A number of theoretically oriented papers pursue post-Peircean approaches, such as the application of reader-response theory and of systems theory to iconicity.


Summaries of papers

    We welcome the attention paid to gestures and signed language in this volume because the research into gestural signs and the structure of signed languages can cast new light on the structure of spoken languages (in both oral and written form), and more particularly, it may tell us more about the iconic foundations of spoken languages and the way iconicity has evolved in them. There are a number of studies in this volume that are directly or indirectly related to sign language or sign language research (gathered together in Part II). It must be noted, however, that the five broad sections distinguished in this book are meant as a loose guide; most of the contributions cannot be limited to the section in which they have been placed. For this reason, we will not always follow these sections closely in this introductory chapter; rather, we will indicate how the various topics discussed in each study refer to and are linked with topics discussed in others.                        
    The most direct investigation into the role played by iconicity in signed language is provided by Klaudia Grote and Erika Linz. Their contribution pays special attention to the fact, already remarked on above, that the interpretation of a sign depends very heavily on its context, especially as concerns the extent to which its occurrence is perceived as conventional. In their study, Grote and Linz first remind us that only certain qualities of a referent can be iconically represented in a sign, i.e. the similarity between a sign and its referent is neither complete nor objective, and the recognition of its iconicity is always filtered by the interpretation of the perceiver. Quite understandably, therefore, the more conventional the immediate context of the sign is, the less likely it is that a conventionalized sign will be interpreted iconically. The main aim of their article is to explore the influence of sign language iconicity on semantic conceptualization. They test whether the iconic quality of a signed word helps to establish the semantic concept more quickly, and whether it influences which quality of the concept stands out more or is more prototypical. This, in turn, may provide information about the semantic structuring of the lexicon. The experiments presented in their study involve confronting both signers and non-signers as well as bilinguals with fully lexicalized (conventionalized) signs (visual or oral ones) as well as with certain qualities connected or unconnected with the referent of the sign. It turned out that signers and bilinguals were faster in judging iconic sign picture relations than the non-signers, who reacted to all qualities connected with the sign in the same way. This was in fact the outcome that the researchers had expected because for the non-signers there was no iconic visual connection between the sign and one of the visual qualities, their sign representing a sound rather than a visual word. This suggests that the iconicity of the sign persists in sign language even when the sign has become lexicalized. The experiments thus show that the semantic organization of the mental lexicon of signers and bilinguals is influenced by the iconicity of the sign. These results additionally suggest that language may influence conceptualization, i.e. that it is not just conceptualization that steers linguistic expression. In other words, the outcome suggests a “moderate version of linguistic relativity”.                    

    A rather different experiment was conducted by William Herlofsky, who shows how research in sign language, in this case the signing of metaphor in Japanese sign language, may be beneficial in helping to develop a comprehensive theory of metaphor and iconicity. Working within a cognitive theory of language where form, meaning, metaphor and iconicity are all equally relevant and fully integrated into the framework of the theory, Herlofsky illustrates how the relationship between real-world space and mental space (which he investigated earlier [Herlofsky 2001] from an evolutionary point of view) may be made more visible by considering the use made of ‘signing spaces’ in signed poetry. He first shows that the iconicity of visual language can best be approached through the analysis of metaphor. Metaphor provides the foundation for our conceptualization of many basic abstract ideas, as the author illustrates by referring to the work of Lakoff and Johnson, who show how complex metaphors arise through primary, cross-domain associations acquired in early childhood (this is called conceptual ‘blending’). After a brief discussion of the types of iconicity that occur in sign language to express concrete objects (these make use of structural correspondences in both ‘form’ and ‘path’ between our conception of objects in the real world and the form and movement of the articulators in the signing space, used to ‘sign’ these concepts), Herlofsky moves on to the signing of abstract concepts, where it is much more difficult to create signs that ‘resemble’ their object. He shows that here the same type of metaphorical blending takes place as is the case in regular, i.e. non-signed language. By illustrating how this blending actually occurs visibly in the signing space in the performance of signed poetry, Herlofsky gives us a better idea of how metaphoric blending may take place in regular language.                    

    At this point we will discuss a contribution by Masako Hiraga (even though we have placed in it Part III because it is also concerned with structural iconicity) since it likewise deals with the notoriously difficult problem of metaphor and iconicity and since it refers to Herlofsky’s study and draws on the same example, a Japanese haiku, which is also available in sign language. In Hiraga’s study the vagueness of the Peircean notion of metaphor is  counterbalanced by a more precise definition derived from cognitive theory. Elaborating and refining Turner's and Fauconnier's model of blending, Hiraga explains the dynamic interplay of metaphor and iconicity from two angles: (1) iconicity manifested as image-schema in metaphor, (2) metaphor giving iconic interpretation to form. The example she uses to illustrate her theoretical model is a haiku by Basho Matsuo whose bipartite metaphorical structure is perfectly suited to substantiate her argument. The essay also demonstrates that in the revision of the poem, kanji (Chinese logographs) effectively strengthen the link between form and meaning. The studies by Herlofsky and Hiraga, which deal with the same problem from  a linguistic point of view on the one hand and a literary  one on the other, exemplify particularly well the advantages of the interdisciplinary approach taken in this and the previous volumes on iconicity.        

    To return to the topic of Part I, Axel Hübler’s contribution to this section does not actually make use of sign language proper but of gestures used by speakers of spoken language. He is interested in the connection between gestures and linguistic signs, and he shows how the loss of the one may lead to change in the other, thus giving an indication of how originally iconic gestures may emerge in  a different form linguistically. By offering a glimpse of how such gestures may have been translated into spoken language, we may acquire a clearer idea about what links there are between spoken and signed languages and about the structure ultimately underlying both. Hübler first develops the idea that there is an iconic relation between the verbal and the gestural mode in the expression of so-called ‘redundant phrasal verbs’ (i.e. expressions in which the particle does not alter the propositional content of the verb, as in ‘swallow down’ for ‘swallow’) and in the expression of ‘pure spatial verbs’ (such as ‘to up’). Thus, the accompanying gestures not only resemble but highlight a certain aspect of the spatial meaning inherent in the verb and/or the adverbial particle. In other words, a phrasal verb like ‘lift up’ accompanied by an upward movement of the hand highlights a part of the event itself. Similar gestures could be used in a metaphorical spatial way in verbs such as ‘yell out’. Hübler suggests further that this cross-modal form of iconicity is the result of the link that exists between these verbs and gestures on the operational level; i.e. the spatial concepts, whether expressed verbally or gesturally, are linked to the same part of the brain. These observations are then used to explain a rather interesting historical-linguistic development: the rise of most of these redundant and pure phrasal verbs in the sixteenth and seventeenth centuries are linked to the great efforts courtly society took in “subjecting the body to a rigorous control”. Too much gesturing became frowned upon. Hübler’s (still tentative) suggestion is that the resulting reduction in gestures, was compensated for verbally by means of an additional but semantically redundant particle.         

    Keiko Masuda’s investigation into birdcalls has a link with the above studies because it likewise involves the more direct mimicking of a real-world object, but this time by oral rather than visual means: she investigates in how far the sounds produced by birds are actually reflected in the linguistic signs that we use to refer to those sounds. The experiments show how close the phonetic word of a particular language, in this case English, is to the actual sound made, but the investigation also indicates what phonetic or phonological constraints of the language in question are in force, or indeed relaxed. It is well-known that even in onomatopoeia, which is considered to be one of the most direct forms of imagic iconicity (i.e. a type of icon that comes close to the ‘real’ thing), the conventional phonological system of language plays a role, and it is to be expected that the choices made to represent birdcalls are somehow constrained by the phonological framework of the language in question. Iconicity, then, is a creative device that to some extent does and does not follow the conventions of language (cf. Lecercle 1990, Fónagy 2001: 2ff.). Sound combinations that do not occur in the phonemic inventory of English such as ts, pf, ps, may still crop up in onomatopoeic expressions, as in tse tse, phft, phsst (the last two are examples from comic books given in Crystal 1995: 250). However, the more conventional the onomatopoeic words become, the more likely it is that they follow the normal phonological rules. In other words, exclamations or interjections in comic books may disregard these rules more easily than verbs which are next formed from those exclamations, simply because verbs are more rule-governed than interjections.      

            Similarly, the description of birds’ sounds may stretch the system of the language more than the actual names given to birds on the basis of these sounds because again these names as nouns – fulfilling a regular part in the lexicon and grammar – will tend to conform to the regular pattern. Still, they remain distinctly iconic, and therefore preserve some of the exceptional behaviour. Marian Klamer has shown in a number of studies (e.g. 2001, 2002) that “expressives” in a language (viz. “lexicalization[s] of vivid sense impressions”, which include “names, and morphemes with negative connotations or referring to undesirable states” [2001: 166]) are usually distinguished formally from non-expressives. This formal distinction could involve some direct or imagic iconicity, but in established or conventional lexicalizations, according to Klamer, it is the diagrammatic iconicity that has been best preserved: the formal complexity of these expressives resembles their semantic complexity, i.e. the marked phonetic form is iconic of the marked semantic content.                         
            By means of an acoustic analysis of the actual sounds of birds, Masuda shows in her experiments which parts of the sounds are used in the linguistic realization of it. She finds that the front cavity resonance (in which both the second and third formants in the spectogram play a role) is most crucial in the imitation of a call, i.e. the selection of the vowels for a linguistic representation seems determined by the frequency of the front cavity resonance. As far as consonant selection is concerned, it appears that especially plosives are selected, which are very well suited to express both the abrupt onset and the extremely short duration of most calls. There is also evidence that further acoustic factors may influence both the place and voicing of the plosive. In addition, Masuda also conducts some experiments the other way around, i.e. from the perceptual point of view, by confronting human subjects with other linguistic, but deviant, signs, which likewise imitate the birdcall. These latter experiments show conclusively that indeed the front cavity resonance pattern is crucial for the type of vowel used. As far as the use of deviant initial and final consonants is concerned, the matter was less clear. It seems that in the selection of consonants, other factors, apart from pure perception, play a role as well. It is quite possible that the language-specific phonological system is involved here. Although this was not part of the present study, it would be interesting to find out by an investigation of the same birdcalls in other languages in how far the signs used there differ from the ones found in English, and thus in how far the phonological system of the language itself determines the shape of the linguistic sign.                     

    In the next section (part II), there are a number of studies that all concentrate on visual iconicity, a type, as we have mentioned above, that is more common in literary texts, and of course in signed languages, as we have just seen. The study by John J. White is innovative in several respects. With its exploration of instances of perspective in shaped poetry – on the page as a two-dimensional surface – and its semiotic effects, it opens up a new field to iconicity research. Albeit rooted in the Peircean tradition, it is post-Peircean in that it utilizes theoretical approaches derived from the visual arts (E. H. Gombrich, Umberto Eco, Nelson Goodman), approaches which are concerned with the role played by perceptual conventions and cultural codes in processes of iconic signification. The central part of his essay deals with Italian Futurists' experiments with typographical iconicity. Starting with minimalist Futurist examples of shaped poems by Bruno Sanzin, which iconize perspective in a rather simple, albeit instructive way, he passes on to more sophisticated examples such as Francesco Cangiullo’s free-word collage poem “Milan-Demonstration” (1915), which is shown to evince both visual and acoustic depth. In the paper’s last section, White moves from avant-garde poetry of the early twentieth century to one of the most significant new forms of shaped poetry at the end of the century, Eduardo Kac’s “holopoems”. These ‘poems’ manifest forms of signification that have emancipated from the static renderings of perspective in earlier avant-garde poetry using the protean effects of perspective in a holograph.            
    An equally complex form of visual iconicity is explored by Julian Moyle, namely the iconic use of corrupt and blurred and the partly illegible typing of poems in Peter Reading’s Last Poems. His essay attempts to enrich the iconic image by showing that it may express much more than just a one-to-one relationship between word and thing. Moyle rejects a simplistic interpretation of Reading’s poem, which would understand its fragmented and illegible typographical form as iconically reflecting the idea of a text under erosion. By relating the poem to other poems in the volume – the “untitled final two pages” (presenting a text even more corrupt than “Erosive”) and “[Untitled]” (which looks like an uncorrupted version of “Erosive”) – and to the wider context of Reading’s work, Moyle calls in question  its conception “as a totalising ‘Entropicon’”. Referring to Michel Serres’ complex model of entropy, which includes ‘negentropy’, i.e. negative entropy, he proposes an interpretation of the contemporary British poet, which sheds light both on individual poems and his whole vision as an artist.             
    A much less intricate instance of iconicity ─ or so it seems at least at first sight ─ is the topic of Andreas Ohme’s analysis of Vladimir Sorokin’s The Queue (1985), a generically unclassifiable text, which represents the incessant polylogue of people in an endless queue – an everyday Soviet experience – waiting to buy consumer goods. While the text’s dialogue is radically desemanticized, its iconicity is foregrounded, the typography miming the length of the queue, the similarity and even identity of the replies and the roll-calls as well as the monotony and repetitiveness of the whole process. A similar technique is to be found, at least in the English translation of the text, in the use of blanks and blank pages to iconize pauses and phases of sleep, and in the representation of sexual encounters which is typographically reduced to moaning sounds followed by dots. What may on the whole seem to be a rather straightforward satiric treatment of an unpleasant part of Soviet reality is shown to be an extremely effective subversion of the ideology and practice of Soviet Realism. The very iconicity of the text challenges and subverts the authority of the ideologized doctrine of art.            
    The subject of Matthias Bauer's essay also concerns a deceptively simple iconic representation, i.e. Dylan Thomas’ poem “Vision and Prayer”, which, with the rhomboid and triangular arrangement of its lines, seems to refer explicitly to the classical tradition of shaped poetry. But Bauer shows that there is no straightforward relationship between the poem’s topic – as expressed in the title – and its pictorial form. He makes it plausible that the poem’s pattern is based on the form of the letter X, an elementary geometrical form which, as he argues, evokes shapes profoundly significant in mystical religious thought such as the double pyramid, the legs of a stork, the cross, a kiss, and the figure of a naked man with the arms and legs spanning the globe in what is a combination of the letters X and O. In the poem’s mystical geometry the X (and derivative forms such as triangle, diamond, double pyramid, and cross) emerges as an icon of the poet’s aiming for the creative word. The iconic use of mystical geometry is here oriented to forms or shapes without numerological implications as expounded in Pötters’ essay on Dante’s geometry of love in this volume (see further below).            
        The iconic significance of visual elements in postmodernist fiction is the topic of Christina Ljungberg's study, which succeeds in exemplifying the heuristic value of Peircean thought even with regard to recent experimental literature. Ljungberg is concerned with diagrammatic iconicity, focussing on the interaction between visual artifacts – such as photographs and maps – and the verbal level of expression. In texts such as Ondaatje’s prose poem The Collected Works of Billy the Kid and Atwood’s The Blind Assassin, photographs are used not only to reflect the structure of the narrative itself and intra-textual relationships (e.g. relationships within a group or those of characters to their surroundings), but they are also employed as devices for self-reflexive comment on representation in general, and on the art of writing in particular. Maps are shown to be used by a postcolonial writer such as Merlene Nourbese Philip (Looking for Livingstone: An Odyssey in Silence) as visual correlates of their fragmented cultures. In self-reflexive fiction such as Paul Auster’s City of Glass, maps iconize the problem of representation.                    

    Moving on from visual iconicity, which is closer to the imagic type, we now turn to Part III and to more diagrammatic forms of iconicity. These play an important role in the way natural languages are structured. But here too the iconic patterns range from the more concrete to the more abstract. Jac Conradie’s contribution to this section concerns repetition, a very common device in language, which is often used in a concrete iconic way. Indeed, it is likely that all conventional repetitive patterns in language were originally iconic. Conradie considers the device of reduplication in Afrikaans, where it is a much more frequent phenomenon than in Dutch, which is the base language from which Afrikaans is derived. It is very likely that the higher number of reduplicative forms in Afrikaans is due to Malay influence, one of the languages that play a role in the creolization of Afrikaans. It is well-known that reduplication is a common feature in Malay (for instance the plural is formed by repeating the stem), as it is in many languages deriving from a pidgin (cf. Tabakowska, this volume, and Kouwenberg and LaCharité 2001).                         
    Reduplication is a versatile and multifarious device in that it can be used in many functions. The same is true for repetition as a literary device. In literature, it is commonly used to express “similarity, continuity, regularity, monotony, emphasis” (Nänny 1986: 205). Because the reduplication in Afrikaans is fairly recent,it is still more clearly iconic, i.e. there has been little grammaticalization. Not surprisingly therefore, the functional categories distinguished for Afrikaans by Conradie, closely resemble the categories found to be relevant for literary language, where the repetition is not usually conventional. The most concrete function of reduplication in Afrikaans is indeed the suggestion of ‘repeated action’. The mechanism is especially encountered in the names of games, which, as Conradie states, “may be regarded as essentially repetitive”. It is also used to suggest repeated efforts, which comes close to another concrete function of reduplication, i.e. ‘intensification’ (similar to Nänny’s “emphasis”). Another function Conradie illustrates, again close to the pure notion of repetitiveness, is its use to suggest intermittent activities, which in turn is related to the idea of interruption or discontinuity. Discontinuity, of course, can only exist in a situation in which “continuity” is foregrounded, so again we have a link here with one of Nänny’s functions of repetition in literary language, and a link with the notions of ‘continuation’ or ‘extension’, which Conradie has also established for Afrikaans. What makes the categorization of the functions of reduplication in Afrikaans so convincing is the fact that all functions clearly ‘hang together’, as well as the fact that they are used creatively in precisely the same way as in literary language. As with all types of iconicity, the key to repetition is similarity. In reduplication, the repeated form is literally similar to the first form, which probably accounts for its simple but strong iconic quality, and for its natural presence in both everyday language and literary language.                    

    The other contributions placed in this section show structural iconicity of a more abstract type. When the iconic relation is more abstract, it is more difficult to observe it directly. At the same time, however, it is also more frequently present in language. It is a well-known fact that iconicity increases with an increase in structure. Thus, compounds and derivations are more iconic than simple stems, and syntactic structures are more iconic than words. Volker Harm shows that there may be another iconic tendency in the lexicon which not so much concerns a diagrammatic relationship between morphemes within words and the way they share morphemes with other words, but which concerns an increase in isomorphism between the meaning and form of a whole (simple or complex) word. He investigates a development in the history of German whereby in a class of verbs derivationally related to one another (e.g. hören, erhören, verhören, gehören), which were once all more or less synonymous (and therefore non-isomorphic between their form and their meaning), the one-to-many relations between form and meaning change slowly into a one-to-one (i.e. isomorphic) relation, so that each individual verbal form ends up with only one of the meanings that they previously shared. What makes this isomorphic tendency even more interesting from an iconic point of view, is the fact that the central, most prototypical meaning gets attached to the most central sign or form, i.e. the unprefixed stem, while the more peripheral meanings fall to the prefixed forms. Thus we see an (iconic) equivalence between a morphologically marked form and a semantically marked meaning. This is rather similar to the type of iconicity that Klamer (2002) discovered in two unrelated languages between phonologically marked forms (of the phonaesthetic type, such as flimmer, flicker) and their semantically marked meanings.                         
    The study by Beate Hampe and Doris Schönefeld deals with structures larger than the word. What Hampe and Schönefeld investigate is verb phrases which are combined with arguments that they do not normally subcategorize for. These new argument frames, however, can be understood by virtue of similar frames used with other, more general verbs. The reason  why they can be understood is that the verbs – now sharing the same pattern – begin to share a number of conceptual properties. This type of iconicity works a little bit like metaphor in that the verbal concept, normally expressed by these rather general and frequently occurring verbs, is now expressed more concretely and vividly by another verb using the same subcategorization frame as the old verb, a frame that it did not possess before. In all cases the new verbal construction highlights a number of expressive qualities that the general verb did not possess. Thus, the general verb put is associated with the following argument frame:                                        

                NPsubject/agent  put  NPobject/patient  into  NPobject/locative                                

as in, She put the child into his high chair. Next, this same frame comes to be used with the verb wrestle, even though wrestle normally occurs only in the following frame:                                             

                NPsubject/agent wrestle  with NPobject/recipient                                                                  

    When a language user produces an utterance such as, She wrestled a screaming Dudley into his high chair, he is using the frame creatively, fixing the pattern of put onto that of wrestle, thereby not only changing the subcategorization frame of wrestle, but also giving it a new meaning, i.e. ‘to put somebody/something into a place with great difficulty’. This new construction can only be understood by reference to the old one. In other words, it is the diagrammatic iconic link between the old and the new, that gives the new construction its meaning. It works like metaphor, but the basis of comparison is structural and not conceptual. What is transferred from ‘tenor’ to ‘vehicle’ is what Hampe and Schönefeld call a “schematic icon” (p. 2 [in ms!]), and not a particular conceptual quality. In both cases, however, we have the same result, i.e. the expression of the familiar by the unfamiliar, making the utterance new and fresh. In both cases, too, it is the context that helps to make the correct interpretation. So this study is another illustration of the fact that the recognition of iconicity very much depends on clues provided by the context, as was also emphasized by Grote and Linz. Language users may creatively manipulate the argument or complementation patterns in which verbs appear provided that enough clues are present to interpret the manipulated complementation structure correctly. It is important to note too that the iconic diagram in this study is different from metaphor in that it is motivated intra-linguistically, it is a form of ‘endophoric’ iconicity (cf. Nöth 2001). Metaphor is motivated by qualities of the external world (‘exophoric’) and is therefore more likely to occur cross-linguistically (more universal). In other words, the use of the same metaphor in different languages depends on cultural phenomena, whereas the use of the same iconic schemes or diagrams depends on the grammatical system of a particular language.                        
    Another study that is interested in the kind of (diagrammatic) iconic principles that play a role in grammatical structuring is the contribution by Günter Rohdenburg. Rohdenburg especially addresses the question of what determines grammatical form in cases of variation. One of the most salient aspects of iconicity in this respect is ‘isochrony’, i.e. the phenomenon that the order of linguistic elements referring to events in the real world mimic the real-world order (cf. Tai 1985). This indeed constitutes Rohdenburg’s “first principle” of linear order. The other two major principles which he distinguishes are the quantity principle and the distance principle (the distance principle is similar to Givón’s [1985] ‘proximity principle’ and also incorporates Bybee’s [1985] ‘relevance principle’[5]). In his essay for this volume Rohdenburg is concerned only with the latter two, but it is interesting to note in connection with the first principle that, although the temporal order of events is indeed frequently mimicked in language, it is difficult, if not impossible to represent the simultaneity of events iconically in this way (cf. Haiman 1985). We have briefly noted above that this is not the case in signed languages where the iconic expression of simultaneity plays a most important role. Clearly the possibilities for iconic forms are constrained by the different modes of communication (gestural/visual vs. oral/aural). Nevertheless it is clear that in spite of the problem of representing simultaneity iconically in terms of order, spoken languages still make use of iconic means to convey the occurrence of simultaneous events. Jansen and Lentz (2001), for instance, have shown that simultaneity can be iconically suggested by intertwining the two simultaneous events, or by embedding one in the other. In this way the linear distance between the sub-event and the main event is reduced (and so the proximity principle is called upon). Another possibility is using the principle of quantity. Discussing two simultaneous events, Jansen and Lentz show that the one that is of minor importance, is given ‘minor’ linguistic form – ‘minor’ in the sense that explicit verbal and nominal markers are lacking – while semantically too the form is less specific than the form which represents the ‘major’ event.                        
            The principle of quantity, as Rohdenburg briefly notes, is also at work in repetition and reduplication, as is indeed shown in detail in Conradie’s contribution to this volume. In his own investigation, Rohdenburg shows how both the principle of quantity and of distance plays a role in the determination of a number of grammatical variants in Present-day English. He looks at the role of quantity in the expression of verbs which may function as auxiliaries as well as full verbs, and finds that only auxiliaries can be shortened and phonetically reduced. Similarly contracted forms with not occur only with auxiliaries, and dialect evidence shows that uninflected forms of do or be (so forms with less quantity) are far more frequent in the case of auxiliaries.                         The quantity principle also plays a role in areas of greater syntactic complexity. For instance, in cases of variation, such as between the comparative form with -er and the periphrastic form with more, the longer form is often used in syntactically more complex constructions. In such cases, extra ‘quantity’ is used, as it were, to indicate the markedness of the total structure. In this same case of variation, the distance principle is also at work in that in adjective-noun phrases where adjective and noun belong together conceptually, the use of more is found to be more frequent. Using -er in such cases would create distance between adjective and noun; so we find ‘more high-minded’ rather than ‘higher-minded’. Other examples where the choice of inflexion is influenced by the distance principle concerns the variation in voicing present in the plural formation of words like hoof (hoofs vs. hooves) and in the adjectival derivation of phrases like loud-mouthed, where the choice is between [θt] and [ðd]. In each case reduced quantity is iconic  of reduced referential meaning.                        
        A literary example of structural iconicity which assumes the form of mathematical or geometrical analogies with profound poetological implications is treated by Wilhelm Pötters in a large-scale exploration of the intertextual relation between Dante's two main poetical works, Vita Nova and Divina Commedia. The connection between these texts is shown to consist in a specific form of iconicity which relates different levels of textual organization by mathematical strategies and geometrical designs. Central to the texts' mathematical designs is Beatrice, whose name is interpreted as a motivated sign. The essay demonstrates an  extraordinary knowledge and use of mathematics in Dante's construction of his poetic universe. Beatrice is given a numerical identity, and the numbers relating to her inform the chronological structure of the romance in Vita Nova and the geometrical conception of the Commedia. The mathematical design is revealed to be a key to the hidden meaning of the two works. The study is amply provided with figures, clarifying the intricate design. It closes with a scheme which summarizes the whole mathematical conception of Dante's poetic cosmos, whose spiritual corner-stones are love of Beatrice, love of God, and love of philosophy. Pötters’ elucidation of the mystical and poetological significance of numerological and geometrical correspondences in Dante can be related to Matthias Bauer’s contribution in this volume which examines the mystical use of geometry in Dylan Thomas’ poem “Vision and Prayer”.                         
    An entirely new departure in iconicity research is to be found in the studies contained in part IV which deal with the interdependence and interaction between different art media on the level of form, a phenomenon called intermedial iconicity. This type of iconicity emerges (1) when a work of art is transferred from one medium to another one and in this process retains formal features inherent in the source medium and (2) when a work of art adopts or imitates formal features characteristic of another medium. The latter type of intermedial iconicity is systematically explored in Werner Wolf’s stimulating contribution, which is devoted to the phenomenon that form in literature can mime other arts and media. It discusses three types of intermedial iconicity in which literary form imitates other media, without ever actually incorporating these media in the text: (1) pictorialization of fiction, illustrated by Thomas Hardy’s Under the Greenwood Tree, a novel which is subtitled A Rural Painting in the Dutch School, (2) filmicization of fiction, exemplified by the last chapter of David Lodge’s Changing Places, and (3) musicalization of fiction, represented by Nancy Huston’s The Goldberg Variations. This paper, which opens new perspectives for further research, also addresses the problem of reception, i.e. the question of how the reader may be induced to identify intermedial references. It is, however, not as radically reader-response oriented as Johansen's contribution to this volume, which is focussed on the construction of iconicity in the reader's mind during the process of reception.            
    Although Elżbieta Tabakowska’s study seems to have a narrower scope than Wolf’s, it examines a highly fruitful topic, namely the problem of how to preserve or recreate iconicity in translation. This also involves interaction between one linguistic medium and another. In her theoretical approach Tabakowska proceeds from the axiom that similarity is basic to iconicity, but that in natural language use, in the processes of lexicalization and grammaticalization, the earlier transparency may disappear, causing the iconic features to change into conventional or symbolic ones. It is interesting to explore the boundaries between these expressive and conventional stages, and Tabakowska shows that a knowledge of iconic practices which are conventional in one language but may still be expressive in another is important in translation. Only if iconic devices are indeed expressively used, i.e. used in order to achieve “a particular communicative purpose” (p. 4 [in ms!]), should they also be translated in such a way that the special purpose comes out in the other language. Obviously, one cannot always use the same devices in the other language because the same device may be part of that other language’s conventional system. In other words, the choice made by the translator needs to be new and expressive for that language in order to be effective. Tabakowska shows how this can be done in practice by comparing a Polish poem by Wisława Szymborska and an English translation of it, and by showing how even in a more conventional prose text, ad hoc forms of iconicity may be spotted, which therefore deserve to be preserved in the translation. Especially in the latter case, Tabakowska makes clear how difficult it is to make a decision about whether the iconicity is intentional or incidental. Although it is true that the intentionality of it could be checked against the background of the use of language in the rest of the book, or in the language in question as a whole (as for instance Shapiro [1998] has done concerning the use of particular sounds in the sonnets of Shakespeare), ultimately the decision of whether something is intentional or not depends on the knowledge of the translator of the two languages involved, as well as on his/her imagination, intuition and acumen. The volume closes with two theoretically-oriented contributions (part V), one looking at iconicity in terms of reader-response theory, and one relating it to systems-theory. Jørgen Dines Johansen’s contribution is  rather exceptional in this volume – and in iconicity research in general – in that its approach is entirely reader-oriented. (To some extent reader response is also considered in the essays of White and Wolf.) It shifts attention from iconic signs as constituents of the text to the reader’s iconization of the text in the process of reception. The study in fact views the literary text as a set of instructions for different ways of iconization. On the one hand being innovative, post-Peircean in fact, in the application of reader-response aesthetics to iconicity, Johansen’s essay is on the other hand firmly grounded in Peircean semiotics. In accordance with Peirce’s categories of image, diagram, and metaphor, it deals with imaginative, diagrammatic, and allegorical iconization occurring in the reading process. Each of these categories is amply illustrated by examples from literary texts. The Peircean axiom of the interdependence of the iconic, indexical, and symbolic dimensions of signs, which has been neglected in recent research, informs the whole discussion. (This trend also emerges in the linguistic contribution of Grote and Linz in the present volume). Johansen notes that even though there is a constant interplay of the three modes of iconization in literature, texts may receive their iconic profile by the predominance of one mode.                     

Piotr Sadowski’s study intends to provide a classification of linguistic signs that diverges from Peirce’s, using the framework of the so-called ‘systems theory of information’. As in Peirce and in other semiotic theories, a distinction is made between ‘information’ as purely physical facts, and ‘para-information’, the interpretation or processing of these facts by animate beings, which turns the purely physical ‘signals’ into (meaningful) ‘signs’. An important part of the theory is that systems interact with one another by exchanging information and energy, and in this way the systems undergo change. In the course of this interaction, language as a system of communication continually evolves. Sadowski considers the different signs that exist in language in terms of ‘information’: emotive, indexical, iconic and arbitrary signs. It is clear that emotive signs represent the most simple type of para-information in that the interpretation that makes them emotive signs does not so much take place by means of association, but by means of instinct, and this is almost direct, subject to physiological conditions only, and not to cultural ones. When such emotive signs are displaced in time or space, they become indexical, and involve para-information of a somewhat higher order. Here the associations performed are not instinctive but acquired behaviour. The next step in the order of signs, are signs which are no longer physically copresent with their signals, but which resemble them only in their structure: iconic signs. The imitations of sounds are probably the earliest iconic signs, as well as the use of gestures. The association here works by means of analogy and is acquired through observation and experience. In all these cases, then, the association retains a physical connection between signal and sign, but one that becomes less and less direct. In  their development towards arbitrariness, signs have become so conventionalized that this direct link is lost. Such signs, in other words, have to be acquired purely by learning. From an evolutionary point of view, this development was highly effective in that it enabled speed of communication and cultural group cohesion; it liberated humans from the constraints laid down by nature. Sadowski adds one further sign in this development, a sign which is no longer prompted by the perception of the signal (as the others are), but which derives from the para-informational level of the sign; he calls this ‘meta-informationally’ derived sign, a symbol (i.e. a symbol in the usual poetic sense, not in the Peircean sense). The meta-informational level serves to distinguish between ‘literal’ and ‘metaphorical’ meaning, between ‘denotation’ and ‘connotation’, between ‘truth’ and ‘fiction’. Meta-information is a specifically human development, and a later stage in the evolution of language. Sadowski applies the notion of meta-information especially to literary language, but the distinction would be equally useful for an understanding of the development that takes place in grammaticalization processes (the development of grammatical systems), which according to Sweetser (1990), proceeds from the socio-physical domain through the epistemic domain into the speech-act domain, i.e. from para-informational to meta-informational.


A review of From Sign to Signing by Elisabeth Leinfeller (University of Vienna) has appeared in 
Studies in Language 29:iii: 739-748