Version classiqueVersion mobile
OpenEdition Books

Recent Trends in Narratological Research

John Pier

Three Dimensions of Space in the Narrative Text

John Pier


This article sets out a syntagmatics of the narrative text that departs from the notion of the "linearity" of the signifier. The space concerned, which is also not that of description, for example, is bound up with graphic space as it is deployed in texts such as the Dos Passos "Newsreel" or Nabokov’s discontinuously articulated Pale Fire. While space as it is manipulated in these texts corresponds to an internal syntagmatics that accentuates the iconic, in Barth’s Lost in the Funhouse, composed of parallel and partially overlapping "series" of narratives, textual space is developed according to an external syntagmatics and follows an indexical principle. With Melville’s Moby-Dick, space is related less to the graphic features of the text than it is to the symbolic function of the sign and the network of semantic relations woven out of the whale as "representative," the whale being a "vehicle" that acts both as a model for the analogy of relations and as a metasign, thus serving as the principle structural metaphor in the novel. The forms of space as conceived here are elaborated along the juncture between the intratextual and the intertextual

Texte intégral

  • 1 I wish to thank Thomas Dutoit for his thoughtful reading of this article and for his perceptive su (...)

1If the evolution of literary theory and criticism over the past few decades were to be summarized succinctly, one conclusion would be that we have witnessed a growing awareness of the problematic nature of the signifying text and, concomitantly, that whatever it is that lies beyond the text is neither the final guarantor nor the ultimate source of meaning in literary expression. While for some, the effet de réel has undermined any possibility of realistic portrayal, for others, it has led to new ways of qualifying the imaginary universe of the literary work (e. g., a narrative semantics based on possible worlds logic) and yet others to subordinate textual production to a wealth of contextual and pragmatic factors (new historicism, cultural criticism). And where authorial intention has survived the death of the author, it has done so in connection with various reconceptualizations of the self (Lacanian psychoanalysis, gender criticism). The text itself, to no negligible degree deconstructed and disseminated in the colloquy of critical discourses and further destabilized in the wake of hypertext technology, hardly seems to fulfill the role of arbiter that it did in times when philological and, later, new critical orientations held sway. Even the book, that depository of cultural memory, has come into question, for as Jacques Derrida has stated: “La fin de l’écriture linéaire est bien la fin du livre...” (1967, 129).1

2The object of this paper is to map out some of the ways in which the space of the written text – and not merely the supposed linearity of the signifier – contribute to and, indeed, are indissociable from the formation of meaning in narrative. As an initial example of this aspect of narrative, we shall consider an extract from John Dos Passos’ novel, The Big Money (1936):

Newsreel 68
[2] This is not Thirtyeight but it’s old Ninetyseven You must put her in Center on time
[6] sympathizers appeared on the scene just as thousands of office workers were pouring out of the buildings at the lunch hour. As they raised their placard high and started an indefinite march from one side to the other, they were jeered and hooted not only by the office workers but also by workmen on a building under construction
[9] He looked ’round an’ said to his black greasy fireman
Jus’ shovel in a little more coal
And when we cross that White Oak Mountain
You can watch your Ninety-seven roll
[10] I find your column interesting and need advice. I have saved four thousand dollars which I want to invest for a better income. Do you think I might buy stocks?
(Dos Passos, 1966, 1151)

3This textual “montage” (only the first third of the Newsreel is reproduced here) reveals the convergence of a number of features that depart markedly from the conventional presentation of prose fiction, including many works of a highly polyphonic nature and other experimental forms of writing. Each of the three typographical forms is associated with a text type: capital letters with newspaper headlines; Roman type with extracts from newspaper articles and letters to the editor; italics with popular songs. Thematically, the first group is doubly subdivided: the shock of the Wall Street crash (1, 4) versus faith in the system (3, 7) which, together, are opposed to the downside of capitalism (5, 8, 11, 12, 13), further divided into the repression of the working classes (5, 8) as contrasted with criminal justice (11) and the seedier side of the entrepreneurial spirit (12, 13). While all of the popular songs reflect the hardships of the working classes, the thematics of the segments in Roman type lies somewhere between the thematics of the songs and that of the headlines, going from the neutral journalistic report of a workers’ demonstration (6) to a letter in the financial column requesting investment advice (10). However, as this latter segment shows, the distribution of the three typographically-determined groups produces networks of meanings that do not exist within the groups themselves – hence, the highlighted naïveté of the well-intentioned but potentially ill-fated investor whose letter is sandwiched between a ballad about work on the railroad and the impersonal headlines reporting unsavory social and economic behavior.

  • 2 See Fludernik, 1996, 178-221, where reflectorization, derived by Franz Stanzel from his earlier “f (...)

4A full analysis of the complete newsreel along these lines would reveal a complex typology of semantic, pragmatic and thematic links among the typographical groups and the various segments. What I would like to emphasize here, however, is the highly schematic manner in which this text is constructed, and particularly the ways in which its graphic features – the “form of the expression,” so to speak – contribute to the elaboration of this textual montage. One cannot fail to be struck by the total lack of punctuation or of any form of syntactic coordination and subordination between the segments. Yet, segmentation of the text into units does occur, and this is accomplished partly by graphic means. Were the entire text presented in, say, Roman type, following the standard line-by-line succession of written discourse but with no other changes to the original, we would have a very different sort of discourse – possibly an interior monologue, in which case the disembodied “voices” that are so cacophonically marked off from one another by typographical means as they are in the newsreel would be filtered through a centralizing reflector figure that is glaringly absent from the text as we have it here; presented on the stage as a soliloquy, or even with a different actor for each of the segments, still fewer aspects of the textual montage would remain intact. Far, too, from the so-called “camera-eye” technique whereby the events “just happen” before our eyes, the passage remains inconclusive as to the positioning of a possible observer and that of the reporting instance. This is due in part to the absence of deictic expressions, notably temporal and spatial deictics in the graphicallymarked newspaper headlines, even though some trace of personal deictics can be recovered through a process of deduction: 3, 4 and 7 are evidently taken from the financial page and thus intended for a specific readership distinct from that of 5, 8, 11, 12 and 13; and while the readers of these various headlines as they appear in their sources are necessarily individuals scattered in time and place, this is not the case of the reader of the newsreel. The lack of grammatical coordination and subordination between the extracts represents a further retreat from authorial control, contributing to the effacement of any centralizing consciousness within the passage other than intermittent (cf. journalistic report in 6 and letter to the editor in 10). The result is a dissonant form of figuralized narrative, that is, a discourse that can be assigned to neither a narrator nor a character, but which at the same time, due in this case to the thwarting of deictic and of syntactic functions, shifts the balance to the reader who, nevertheless, is prevented from identifying in an empathetic fashion with any consistent text-internal position.2 In sum, what characterizes the Dos Passos newsreel, setting it off from narrative forms expressed in accordance with the graphic conventions, is the fact that here, graphic configurations largely supplant deictic expressions and syntactic markers: the written word is produced according to criteria that diverge from those of the spoken word, suggesting that written discourse does not stand as a mere visual copy of spoken discourse, but is closely bound to the possibilities and constraints of the page upon which it appears. Indeed, the generic title, “Newsreel” – itself a misnomer, as the textual segments that follow can be qualified as cinemagraphic only to the extent that they are an assembly of fragmentary evocations of newsworthy topics – could even be said to highlight the disparity between the written and the spoken in the text that it designates.

  • 3 To speak of description in prose fiction as “referential” is admittedly reductive, given the numer (...)
  • 4 For a fuller account of Harris’s Signs of Writing, see Pier, 1997.

5The example taken from Dos Passos, while hardly typical of the mass of published prose fiction, is nevertheless one instance, among others, that points towards the need for a renewed investigation of space 194 in the written text. The space in question is only marginally that of description in the sense of the pictorial evocation of a place, object or character insofar as this form of discourse is couched in a referential language.3 What will be focused on here is the space of the written text derived from a syntagmatics of writing – a syntagmatics distinct from that of speech and which is subject to spatial and graphic criteria with no equivalent in spoken discourse. This point of view is inspired, not from Derrida’s grammatology, but from the semiology of writing of Roy Harris, who argues that the syntagmatics of writing is of a distinctly different order from the syntagmatics of speech and is indeed more complex and varied than that of speech (Harris 1994, 1995). According to Harris, the widespread failure to distinguish between the two forms of syntagmatics is due in no small measure to Saussure’s characterization of the signifier as “linear”: while it is clear that two linguistic signs cannot be emitted from an identical source simultaneously, to speak of the “linearity” of the signifier confuses auditory signs, which occur in time and are transitory, with graphic signs, which occur in space on a material support and whose resources far exceed those of the line; moreover, the written sign, contrary to the spoken sign, is a form of non-kinetic graphic communication of widely varying degrees of durability, indissociable from the writing surface upon which it is inscribed. There is a good deal more to say about Harris’s lucid and pathbreaking theory of writing than can be dealt with here.4 In any case, the idea I would like to develop in these pages is that a syntagmatics of writing can provide us with certain insights into the narrative text that neither the “linearity of the text” nor the “linearity of the signifier” can provide, even in their contested and/or revised forms. The questions to be considered clearly go beyond narrative (calligrams and concrete poetry are fertile ground for exploration, as are certain graphic forms of advertising) and even beyond literature, for at issue are all forms of linguistic expression in graphic form, and not merely those using the phonetic alphabet. These matters lie beyond present concerns, however, which are confined to identifying and illustrating the principle features of space that contribute to modes of signification in prose fiction.

  • 5 I borrow this term, which refers to the interpretation of linguistic signs by means of nonlinguist (...)

6A distinction fundamental for our considerations is to be made between an internal syntagmatics and an external syntagmatics of the of the written text. According to Harris, “Internal syntagmatics relates to the disposition of written forms relative to one another and to other forms within the same graphic space” (1995, 121). Concerned here is the arrangement of writing on a page, for example, but also the syntagmatic relations between a photo and its accompanying caption in a newspaper, while in the literary domain the works of William Blake (among others) can be cited, showing the interactions between the graphic forms of writing and the graphics of visual illustrations. In such a process of “intersemiotic translation,”5 the question of iconicity obviously comes into play, although the degree of correspondence or similarity between word and image, even in those instances where it can be established, by no means exhausts the question. In this connection, Magritte’s painting Ceci n’est pas une pipe will serve as a striking reminder of the inherent disparities between word and image.

  • 6 A detailed analysis is given in Pier, 1992. On the notion of paratext, see Genette, 1987, and Lane (...)

7Closer to present concerns, however, is the Dos Passos newsreel, composed only of printed text, but of text in a variety of typographic forms that cannot be dissociated from the text’s meaningfulness. In such examples, as we have seen, graphic discontinuity in the succession of words produces effects that graphically conventional texts can accomplish only through linguistic means of a different order, if at all. However, discontinuity on the printed page can take on other forms, as in Vladimir Nabokov’s Pale Fire (1962), which consists in texts by two fictional authors, one being a poem entitled “Pale Fire”, the others comprising a critical apparatus (“Foreword,” “Commentary,” “Index”) by the poet’s paranoid friend and neighbor who also acts as editor of the poem. The novel is so written that the reader is driven to switching back and forth constantly between the pages of the poem and those of the Commentary (printed in the latter half of the book), so that the syntagmatic structures of the written text (both poem and Commentary) are perpetually reconfigured, while the Index, organized alphabetically, serves to re-order the facts in yet another way. This “book of mirrors,” as Mary McCarthy aptly dubbed Nabokov’s novel, results from a particularly intricate overlapping and intertwining of more or less divergent discourses, with the critical apparatus serving as a paratext of the poem and the poem acting as a paratext of the critical apparatus. In effect, Pale Fire generates a multitude of books within the space of one volume – books that can be qualified as “pale imitations” of one another.6

8While the internal syntagmatics of a text is defined in part by the iconic features of that text’s constituents, external syntagmatics is associated with the text’s indexical functions. “External syntagmatics,” explains Harris, “covers the various relationships which may obtain between the written forms and items or events to which they are significantly connected in the space outside (i. e. that space within which the’graphic space’itself is situated)” (1995, 121). A panel on the road from Debrecen to Budapest with the word “Budapest” and an arrow pointing in the direction of that city illustrates how a written sign can be directly connected with the non-verbal environment in which it appears.

  • 7 Although she does not frame her argument specifically in terms of indexicality, Monika Fludernik d (...)

9Although the theater abounds in such indexical phenomena, prose fiction, an inherently de-contextualized form of discourse, would be hard put to provide examples that correspond to the above definition. To be sure, indexical features do play a role in narrative, but this comes mainly through deictic expressions as they intervene in focalization, discourse typology (including the thorny problem of free indirect discourse) and the positioning of center of consciousness within the fictional universe.7 The external syntagmatics of the literary text, i. e. its relations with (textual) space outside its own graphic space, can be related, broadly speaking, to intertextual traces and their markers, although intertextual analysis clearly calls for the elaboration of further distinctions and categories: the knowledge of Robinson Crusoe required for a reading of Michel Tournier’s Vendredi is not of the same order as the familiarity with the generic conventions of the Künstlerroman presupposed for a reading of A Portrait of the Artist as a Young Man, nor do cases such as these follow the same set of criteria as those brought to light in the prefaces written by Henry James for the New York Edition of his works after the original publication of the novels.

10Within the graphic space of a given fictional text, an external syntagmatics properly speaking is thus excluded. Even so, the possibility of a “second-order” external syntagmatics does exist for the individual text, by which I mean that the relations of a text with the graphic space outside can be reduplicated within a given work. Particularly interesting in this regard is John Barth’s Lost in the Funhouse (1968/1969). Preceded by an “Author’s Note (1968)” and followed by “Seven Additional Author’s Notes (1969),” the fourteen tales making up this volume which, when read in their textually given order from the first page to the last, appear to be little more than a heterogeneous collection of stories, do in fact form a rigorous configuration of no fewer than four “series” of texts (the term comes from Barth himself), each with a subtle but unmistakable thematic and/or diegetic stamp of its own:

1) an “interwoven” series (tales 2, 4, 6, 8, 10, 12, 14; tales 3, 5, 7, 9, 11, 13);
2) an “echo” series (tales 8 through “Seven Additional Author’s Note (1969)” are an “echo” of “Author’s Note (1968)” through tale 7 such that tale 8 “echoes” tale 1, tale 9 “echoes” tale 2, etc., while an additional series is constituted by reversing the order [tale 1 “echoes” tale 8, etc.]; significantly, no tale is found in the middle, while tale 8, entitled "Echo," immediately follows [or precedes] tale 7: “Lost in the Funhouse”);
3) a “Chinese box” series (tale 14 is read after tale 1, tale 13 after tale 2, etc., although here again, the order can be reversed, as it is possible to read tale 14 before reading tale 1, etc.; there being no “center,” this series is, in principle, infinitely regressive);
4) a “Protean” series, based on an intricate system of embedding, present particularly in tale 13 (“Menelaiad”), which is divided into fourteen sections numbered 1 to 7 (each successive section is provided with an additional set of quotation marks) and 7 to 1 (with the corresponding suppression of quotation marks). A spectacular mise en abyme of the work as a whole, this tale is connected, through the itineraries traced by the previously-mentioned series, to the work’s ultimate mise en abyme, found in tale 1: entitled “Frame-Tale,” this text consists in the words “Once upon a time there” (written on front side of the page) and "was a story that began" (on the back side), printed along the edge of the page, and it is accompanied with instructions to cut the page out of the book and twist it into a Möbius strip, the result being that the text of the tale forms an infinitely recursive structure – a structure with an expanded corollary in the never-ending passage from tale 13 to 7 to 8 to 1 and back to 13.

11Reading Lost in the Funhouse according to these six series lends an accessibility to Barth’s textual “labyrinth” that is all but absent from the textually given order – a seventh “linear” series that remains largely marginalized. Much as Pale Fire, then, Lost in the Funhouse is a work that defies any continuous reading from beginning to end. There is, however, a fundamental difference between the two works in that, through the demands made on the reader, the syntagmatic structures of Pale Fire are perpetually interrupted at the micro-textual level only to be re-ordered through their confrontation with other discourses that diverge from it markedly, while each of the tales included in Lost in the Funhouse, a fairly self-contained unit in itself, is re-articulated into six different but contiguous series which, even though printed within the same volume and thus occupying the same graphic space, remain curiously autonomous yet complementary. The internal syntagmatics of Pale Fire, in contributing to the (imperfectly) iconic character of its constituents, thus contrasts with the external syntagmatics of Lost in the Funhouse, a second-order external syntagmatics in that the various recontextualizations of the fourteen tales produced by the integration of these tales into the six series puts each series into a distinctly indexical relationship with one another.

  • 8 Cf. Derrida, 1967, 15-142.

12My proposals for the study of space in the narrative text, as the examples discussed show, have little bearing on the space concerned with descriptive discourse, while at the same time they are intended, not as a revision of, but as an alternative to the supposed “linearity” of the signifier. This position obviously raises questions as to how space, understood in such a way, relates to current and/or traditional principles of poetics and narrative theory. The predilection for spatial metaphors of narrative theory, and particularly of structurally-inspired models of narrative, has frequently been noted. In his Towards a Postmodern Theory of Narrative, for instance, Andrew Gibson comments extensively on the “geometrics” of classical narratology and, with reference to the writings of Michel Serres and other postmodern thinkers, sets forth a program for the “pluralisation of the narratological imaginary” (1996, 15) through the introduction of a new set of spatial metaphors (fluent form, heteroplasty, transversals, laterality, etc.). The discussion centers on an informed critique of the rigidity of many narratological concepts, attributable to the reliance of narratologists on “geometric” spatial categories, but it does not take up the question of the non-linear graphic resources of narrative in texts such as those dealt with here. As for the reflection on the sign by Jacques Derrida, a prime mover in the deconstruction of the sign as concerns, among other things, the metaphysical presuppositions of “presence” underlying the phoneticist bias, revealed notably in the Saussurean binary conception of the sign together with the differential and linear character of the signifier, this theory seeks to invert the supposed priority of speech over writing. The literary text thus becomes irreducibly graphic, with “difference” constituting the non-linear articulation of signs among themselves and “trace” representing the “deconstruction of presence” which is productive of sense.8 It would seem, then, that by its theoretical basis, deconstruction, through its subtle redefinition of the relations between speech and writing along lines of reasoning that depart from both structural linguistics and semiology, does not (and need not) elaborate a syntagmatics of the written sign distinct from that of the spoken sign.

  • 9 According to Gibson, interactive fiction “makes the user choose from among a set of radiating poss (...)

13With interactive or electronic fiction, it is another aspect of the written sign that emerges. In works of this kind, we have a form of writing set off from fiction in the printed book due to the instantaneous and extensive mobility between episodes included in portions of text. These possibilities for highly varied combinations of text segments result from electronic linking, a technological innovation that tends to reduce the distance between reading and writing, thereby adding a new dimension to the internal syntagmatics of the text (cf. Harris, 1995, 162). Yet, as Jay David Bolter has shown, electronic writing has served to open our eyes to the “nonlinearity” inherent in fiction from Sterne to Borges, that is to say, in books dating from before the impact of computer technology on writing (1991, 121ff.). Texts such as those discussed in the present paper could well be included among Bolter’s examples, although it must be pointed out that, as fictions that are bound up with intricate and subtle explorations of non-planar and non-tabular dimensions of graphic space in the book, they necessarily rely on a different distribution of written signs. Placed in two windows on the computer screen, for example, the poem and the critical apparatus comprising Pale Fire would become more “readable,” but would occupy the writing surface in a manner distinctly different from the way they do in the novel as it was published. In a similar way, “scrolling” the tales in Lost in the Funhouse would provide the reader with speedier access to the six series of tales, but at the same time would eliminate the Möbius strip effect, restoring to the series a textually “linear” order that is so ingeniously manipulated by the novel in book form.9 It is perhaps the Dos Passos newsreel, with its graphically differentiated but planar presentation, that would best lend itself to electronic linking, as this would allow different semantic, thematic, pragmatic, graphic and other configurations, as well as various combinations of these configurations, to be traced in a systematic way.

14More generally speaking, however, discussion of these works suggests that space in the narrative text, as it is understood here, follows a principle of intratextuality which, at the level of the individual work, acts as a counterpart of intertextuality. Pale Fire presents the two parts of the novel (the poem and its critical apparatus) in such a way that each acts as the paratext of the other, providing a rich illustration of the resources of paratextuality, that is to say, of the intermediate or transitional zone between a text and its hors-texte. More precisely, what is brought into play here is the “peritext,” a type of paratextuality in which various elements such as titles, subtitles, chapter headings, prefaces, postscripts, notes, etc. that are not part of the text properly speaking are nonetheless included in the same volume – a standard feature of fiction published in book form that has been exploited by Nabokov in a way that goes beyond the practices even of his eighteenth-century predecessors. The situation with Lost in the Funhouse, however, is somewhat more ambiguous. Of the fourteen tales in this work, seven appeared separately in various journals and magazines during the five years leading up to their publication in book form in 1968: introduced by “Author’s Note (1968),” this edition was followed by another in 1969 that included at the end “Seven Additional Author’s Notes (1969)” as a sort of postscript. Although the “Notes” clearly serve the prefatory function of “guiding” the reader, they also play a pivotal role in constituting Lost in the Funhouse as a book by assembling a number of pre-existing texts within a single volume; the “Notes” are thus also comparable to an author’s notebooks or journal elucidating the links between various of his writings that might otherwise be overlooked by the public. In other words, while the “Notes” qualify as peritext, they also bear the marks of “epitext,” that is (in this case), authorial comments inscribed on a physical support which is distinct from and thus occupies a space different from that on which the work appears – a circumstance that accounts in part for the six relatively autonomous yet intertwining “series” elaborated in Barth’s work, as discussed earlier in this article. The important point here is that Lost in the Funhouse, over and above the tales and “Notes” appearing in their published order, contains between its covers six additional “volumes.”

  • 10 Relating space in the narrative text in this way to intertextuality results in a partial rearticul (...)

15It is thus on the diffuse juncture between the space of a text and the textual space external to it that so much depends, on this fluctuating and multifaceted paratext that sets off the intratextual from the intertextual. The works chosen for the present study seem even to highlight this feature and to confirm an important connection between paratext and the syntagmatics of writing that, we maintain, casts a new light on the problem of space in the narrative text. Thus, peritext, or the elements that intervene in the interstices of the text, comes within the scope of an internal syntagmatics of writing, while epitext, i. e. material of editorial, allographic or authorial origin bearing on a given work, but not included within the graphic space of that work, concerns an external syntagmatics of writing. In a word, analysis of a text in terms of the syntagmatics of writing accentuates the interface between the intratextual and the intertextual.10

16Up to this point, emphasis has been laid on space in the narrative text according to the relations of the written sign with its environment, either within the same graphic space or in its connection with the space outside, and this has led us, through the prominent position of paratext, to regard narrative space in terms of intertextuality. But we have also seen that internal syntagmatics lends an element of iconicity to a text’s constituents (Pale Fire), and external syntagmatics an element of indexicality (Lost in the Funhouse). To proceed now with the account of space in the narrative text within the semiotic framework that has been adopted here, it is necessary to refer to the symbolic. In the semiotics of Peirce, the icon, which is based on a resemblance or similarity of qualities between the sign and its object (whether or not this object actually exists), and the index, with a relation of contiguity by which the object is related to the sign through dynamic action, are distinguished from the symbol, a conventional or arbitrary sign, due to the fact that the symbol exists by virtue of the connection made by the symbol-using mind: “A Symbol is a sign which refers to the Object that it denotes by virtue of a law, usually an association of general ideas” (Collected Papers, § 2. 249). It is neither possible nor necessary in these pages to elucidate Peircean semiotics or, in any extensive way, its implications for literature. What does need to be pointed out, however, is that with regard to space in the narrative text, written signs in their symbolic dimension are less dependent on graphic features and their distribution on the material support than are signs appearing in contexts where the syntagmatics of writing accentuates either the iconic or the indexical.

  • 11 For a valuable account of this aspect of Moby-Dick, see Bleikasten, 1991.

17A work of particular interest in this connection is Herman Melville’s Moby-Dick (1851), a novel with deep roots in the medieval liber mundi tradition but which, in the wake of post-romantic developments, is deprived of the totalizing intent of the absolute Book grounded in medieval thought, becoming a fragmented and plural book or, in Richard H. Brodhead’s words, “an encyclopedia of extracts.”11 My argument, firstly, is that attempts to represent Moby Dick in any mimetic sense are largely fruitless. In “The Whiteness of the Whale” (Ch. 42), for instance, Ishmael states “I almost despair of putting it in a comprehensible form” (287), and an entire chapter, “Of the Monstrous Pictures of Whales” (Ch. 55), is devoted to the inadequacy of pictorial representations of whales. Even more scientific approaches to gaining knowledge of the whale prove disappointing, as in “The Tail” (Ch. 86): “Dissect him as I may, then, I but go skin deep; I know him not, and never will. But if I know not even the tail of this whale, how understand his head? much more, how comprehend his face, when face he has none?” (488).

18Unknowable by direct observation, the whale is provided at various points in the novel with texts that seek to render it accessible by human intelligence. Hence, the “Etymology” and “Extracts” that serve as an introduction to the novel, the very important mise en abyme of the story of Jonah and the Whale in “The Sermon” (Ch. 9) and the elaborate attempt, in “Cetology” (Ch. 32), to transform the whale into a “volume” by classifying it according to a “bibliographic system”: “According to magnitude I divide the whales into three primary BOOKS (subdivisible into Chapters), and these shall comprehend them all, both small and large” (231).

19However, such attempts to render the whale knowable by “textualizing” it, by superimposing on it a human language, meet with limited success, for the whale seems to bear an inaccessible message expressed in a language of its own. Ishmael notes, in “The Blanket” (Ch. 68), that the “linear marks” on the skin of the whale “are hieroglyphical; that is if you call those mysterious cyphers on the walls of pyramids hieroglyphics, then that is the proper word to use in this connection,” and he continues: “I was much struck with a plate representing the old Indian characters chiselled on the hieroglyphic palisades on the banks of the Upper Mississippi. Like those mystic rocks, too, the mystic-marked whale remains indecipherable” (412-413). The decipherability of the whale’s language is further put in doubt in “The Praire” (Ch. 79), devoted to the physiognomy and phrenology of the whale, where it is asked: “Has the Sperm Whale ever written a book, spoken a speech? No, his great genius is declared in his doing nothing particular to prove it. It is moreover declared in his pyramidical silence” (454). The resistance of the whale to human intelligence, its “message” to man, so to speak, is perhaps best characterized by Ahab who, in his obsessional hatred of Moby Dick, sees in it an “unreasoning mask,” an “inscrutable thing” with, perhaps, “naught beyond”: “All visible objects... are but as pasteboard masks” (“The Quarterdeck” [Ch. 36], 262).

  • 12 Ibid., 388; see also p. 380, where the author suggests that one of the best introductions to Moby- (...)

20In Moby-Dick, then, there are two semantic movements, one gesturing toward an adequation between human signs and the order of nature, the other being an attempt to recover the “meaning” of natural signs; both movements, as the quotations above suggest, are foiled, so that a gap emerges between human signs and natural signs. To designate this disparity between the two orders, we adopt the felicitous neologism “diabolic,” a term applied by André Bleikasten to the novel as the other face of “symbolic” (from the Greek symbolon: “thing thrown together”).12 Many factors contribute to the semantic particularities of Melville’s text, but, as hinted at in the full title of the work, Moby-Dick; or, The Whale, perhaps none more so than the relative but persistent indistinction between Moby Dick and whales as a general class of beings, between the denotatum of the sign and the qualities predicated of that denotatum, with the consequence that, in part, the represented yields to or is eclipsed by the representative.

21There is, however, a third element to be taken into account which is that, according to Ishmael, when the terms of his method are applied to an ant or a flea, they “might justly be deemed unwarrantably grandiloquent. But when Leviathan is the text, the case is altered” (“The Fossil Whale” [Ch. 104], 566; my emphasis). This phrase is of capital importance, for it makes explicit one of the key features of the novel, namely, that the whale is a vehicle for representation rather than an object of representation. It is in “A Bower in the Arsacides” (Ch. 103), however, that this idea takes on substance, producing numerous repercussions throughout the rest of the work. In this richly allusive chapter, Ishmael relates how a whale washed ashore on a Pacific island visited by him some years prior to the voyage of the Pequod and reduced to its “unconditional skeleton” had been transformed by the natives into a temple, a “worshipped skeleton.” Here, in place of the whale marked with indecipherable hieroglyphs mentioned above, is a whale whose “vertebræ were carved with Arsacidean annals, in strange hieroglyphics” by the priests of the island (560). But the skeleton also forms part of a bower and, overgrown by vegetation, it becomes, under the effects of the sunlight, a sort of loom: “Through the lacings of the leaves, the great sun seemed a flying shuttle weaving the unwearied verdure. Oh, busy weaver! unseen weaver!... Yet, as the ever-woven verdant warp and woof intermixed and hummed around him, the mighty idler seemed the cunning weaver; himself all woven over with the vines; every month assuming greener, fresher verdure; but himself a skeleton. Life folded Death; Death trellised Life; the grim god wived with youthful Life, and begat him curly-headed glories” (561). And finally, the skeleton is assimilated to a labyrinth: “To and fro I [Ishmael] paced before this skeleton – broke through the ribs – and with a ball of Arsacidean twine, wandered, eddied long amid its many winding, shaded colonades and arbors. But soon my line was out; and following it back, I emerged from the opening where I entered. I saw no living thing within; naught was there but bones” (561-562).

  • 13 Alluded to here is Peirce’s “diagrammatic” iconicity (e. g., “Epilogue” as the “tail” of Moby-Dick(...)

22Reduced to its “unconditional skeleton,” the whale as hieroglyph, loom and labyrinth stands as a textual metaphor for Moby-Dick. Considerable commentary on the novel would be necessary to develop this idea and to show how innumerable meanings throughout the text are “woven” out of the whale, how the whale forms a sort of “framework” for the elaboration of the novel. The point to be stressed here is that, since the whale is unrepresentable (due to the failure of the symbolic, as suggested above), it is replaced by the Arsacidean whale, that is, by a simulacrum out of which Moby-Dick is generated: the skeleton, reconstituted, is an icon that acts both as a model for the analogy of relations and as a metasign.13

23Here again, the intratextual appears as the key to space in the narrative text, although the emphasis in this case lies with a network of semantic relations that is less dependent on graphic space than the semantic relations developed in the other works examined in this article. This being the case, it may seem remarkable to see the re-emergence of the iconic which, in our observations on Pale Fire, was associated with the internal syntagmatics of writing. It is also true, however, that the indexical enters into Moby-Dick through the large number of references, allusions, etc. that it includes, even though this work does not seem to accentuate the external syntagmatics of writing that characterizes Lost in the Funhouse. In fact, while we have sought to draw a distinction between types of space in narrative according to criteria defined by the iconic, the indexical and the symbolic, none of these types excludes the others, but rather stands out as relatively more prominent, more determinant, in the configuration of a given text and its mode of signification.


Works Cited

Primary sources

Barth, John, 1968/1969. Lost in the Funhouse. Fiction for print, tape, live voice. New York: Doubleday.

Dos Passos, John, 1966 (1936). The Big Money. In: U.S.A. Harmondsworth: Penguin.

Melville, Herman, 1972 (1851). Moby-Dick; or, The Whale. Ed. Harold Beaver. Harmondsworth: Penguin.

Nabokov, Vladimir, 1973 (1962). Pale Fire. Harmondsworth: Penguin.

Secondary sources

Bleikasten, André, 1991. “Moby-Dick et la question du livre.” Revue Française d’Etudes Américaines, 50, 377-390.

Bolter, David Jay, 1991. Writing Space. The Computer, Hypertext, and the History of Writing. Hillsdale (N. J.): Lawrence Erlbaum Associates, Inc.

Derrida, Jacques, 1967. De la grammatologie. Paris: Minuit.

Fludernik, Monika, 1993. The Fictions of Language and the Languages of Fiction. The linguistic representation of speech and consciousness. London and New York: Routledge.

Fludernik, Monika, 1996. Towards a’Natural’Narratology. London and New York: Routledge.

Foucault, Michel, 1966. Les Mots et les choses. Une archéologie des sciences humaines. Paris: Gallimard.

Genette, Gérard, 1982. Palimpsestes. La littérature au second degré. Paris: Seuil.

Genette, Gérard, 1987. Seuils. Paris: Seuil.

Gibson, Andrew, 1996. Towards a postmodern theory of narrative. Edinburgh: Edinburgh UP.

Harris, Roy, 1994. La sémiologie de l’écriture. Paris: CNRS Editions.

Harris, Roy, 1995. Signs of Writing. London and New York: Routledge.

Hoffmann, Gerhard, 1978. Raum, Situation, erzählte Wirklichkeit. Poetologische und historische Studien zum englischen und amerikanischen Roman. Stuttgart: J. B. Metzlersche Verlagsbuchhandlung.

Jakobson, Roman, 1959. “On linguistic aspects of translation.” In: R. A. Brower, ed., On Translation. Cambridge: Harvard UP.

Lane, Philippe, 1992. La périphérie du texte. Paris: Nathan.

Peirce, Charles S., 1931-1958. Collected Papers. Eds. Charles Hartshorne and Paul Weiss (vols. 1-6); Arthur W. Burks (vols. 7-8). Cambridge: Harvard UP.

Pier, John, 1992. “Between Text and Paratext: Vladimir Nabokov’s Pale Fire. Style, 26. 1, 12-32.

Pier, John, 1997. “Roy Harris and Writing Without Speech.” Style, 31. 1, 134-147.


1 I wish to thank Thomas Dutoit for his thoughtful reading of this article and for his perceptive suggestions.

2 See Fludernik, 1996, 178-221, where reflectorization, derived by Franz Stanzel from his earlier “figural narrative situation” (personale Erzählsituation), is contrasted with figuralization, which Fludernik, partly with reference to Ann Banfield’s “empty deictic center,” qualifies as “the reflectoral equivalent of camera-eye narrative” (198), characterized by “consonance,” a term coined by Dorrit Cohn to designate a discourse in which the narrator fuses with the consciousness that s/he narrates. The dissonance we see in figuralization here is one result of graphic discontinuity.

3 To speak of description in prose fiction as “referential” is admittedly reductive, given the numerous forms of description and their various narrative functions. However, as emerges from Gerhard Hoffmann’s exhaustive study of space in the English and American novel, space is inextricably bound up with referentiality once “lived space” (gelebte Raum) is viewed either as “perceptual space” (Anschauungsraum), “space of action" (Aktionsraum) or “mediated space” (gestimmte Raum) (Hoffmann, 1978).

4 For a fuller account of Harris’s Signs of Writing, see Pier, 1997.

5 I borrow this term, which refers to the interpretation of linguistic signs by means of nonlinguistic systems of signs, from Jakobson, 1959, 233.

6 A detailed analysis is given in Pier, 1992. On the notion of paratext, see Genette, 1987, and Lane, 1992.

7 Although she does not frame her argument specifically in terms of indexicality, Monika Fludernik deals with the latter two questions (among others) using categories that are largely indexical in nature (Fludernik, 1993, 1996).

8 Cf. Derrida, 1967, 15-142.

9 According to Gibson, interactive fiction “makes the user choose from among a set of radiating possibilities as readers of novels cannot” (1996, 276). This is a revealing statement, for even a “linear” reading of detective fiction, for example, draws one into conjectures on the “radiating possibilities” generated by the text.

10 Relating space in the narrative text in this way to intertextuality results in a partial rearticulation of the five types of “transtextuality” identified in Genette, 1982, 7ff.: intertextuality (quotation, allusion, etc.); paratextuality; metatextuality (commentary); hypertextuality (transformation or imitation); architextuality (taxonomy and “horizon of expectations”). This classification goes from the “immanent” to the “transcendent,” and while Genette’s skillful employment of these categories demonstrates their ultimate interdependence, it is not predicated on the intratextual and the intertextual.

11 For a valuable account of this aspect of Moby-Dick, see Bleikasten, 1991.

12 Ibid., 388; see also p. 380, where the author suggests that one of the best introductions to Moby-Dick is the chapter in Michel Foucault’s Les Mots et les choses entitled “La prose du monde,” where it is explained that the medieval episteme of “resemblance” shows us “comment le monde doit se replier sur lui-même, se redoubler, se réfléchir ou s’enchaîner pour que les choses puissent se ressembler.” It seems to me that the chapter “Représenter” should also be mentioned, as here it is explained how, under the influence of scientific thought from the seventeenth century onwards, the relation of the sign to its content came to be put in doubt.

13 Alluded to here is Peirce’s “diagrammatic” iconicity (e. g., “Epilogue” as the “tail” of Moby-Dick) and metaphors, that is to say, icons “which represent the representative character of a representamen by representing a parallelism in something else” (e. g., Arsacidean twine/Ariadne’s thread), as opposed to imaginai icons (identity photos, etc.) (cf. Collected Papers, § 2. 227).


Professor of English at the Université de Tours. Specialized in narratology and semiotics, his articles and reviews have appeared in Style, Ars Semeiotica, Poetics Today, Bulletin de Stylistique Anglaise, Poetics Today, GRAAT, Encyclopedic Dictionary of Semiotics, etc. as well as in a number of collective works, and they include studies of Sterne, Joyce, Nabokov, Max Frisch and others. He is currently researching intertextuality in modern fiction

© Presses universitaires François-Rabelais, 1999

Conditions d’utilisation :


Volume papier