Revisiting Greimasian figurativity, and making the notion do more work in semiotics of text

1 Recapitulation and discussion of recent commentaries

Arguably, the notion employed by using such terms as figure, figurative, figurativity – in French, figure, figuratif/figurative, figurativité – is one of the founding concepts of Greimasian semiotics. As a foundational notion, it has not received enough attention in commentaries on Greimas’ own work: even the comprehensive book on Greimas’ semiotics by Ronald Schleifer does not provide a chapter on figure. Nevertheless, the notion has been separately commented upon in relatively recent articles by Bertetti (2017) and Harkot-de-la-Taille (2020).

Bertetti provides a much-needed, albeit brief, commentary on how Greimas’ notion of figure relates to the homonymous notion in Hjelmslev, presupposing that the former picked up the concept from the latter. Bertetti’s view is that with regard to figure, Greimas followed Hjelmslev “with a considerable amount of freedom,” which reflects a general “problematic compliance with Hjelmslev’s theory.” If Hjelmslev defined figures as “non-signs [that] enter into a sign system as part of signs” on both the plane of expression and content taken in separation (Bertetti 2017: 90), Greimas’ redefinition of the notion is tied to the aim of “overcoming the contingency of the single sign-object in order to analyse the general mechanisms of semiosis and re-examine the forms and the models that govern linguistic and semiotic creations. In other words, it means studying the transcendental conditions of possibility of textual production” (Bertetti 2017: 90). To this I will add, in anticipation of further discussion, that for Greimas, but not for Hjelmslev, figure is a trans-semiotic phenomenon that links linguistic, pictorial, and other semiotic systems to perception.

Perhaps due to his focus on the origins of figure in Hjelmslev, Bertetti seems to downplay the perceptual constituent of Greimas’ definition. Bertetti does take perception into consideration yet opts for a thoroughly textualistic explanation, stating that “[i]n terms of textualization, the existence of sign relations at different levels must be recognised inside a text … Words are signs, figures are signs, each with its own signifier and its own signified … [and] a text is, therefore, a ‘conglomerate’ of significant elements of a different nature the extent of which will always depend on a semiotic relevance” (Bertetti 2017: 95). This way, figure is approached in the first instance as an element of text: it is a type of textual sign-function.

From a broader view, figure as a sign-function (and while the distinction between sign-function and sign is awfully convoluted, I believe it is relevant here) is, for Bertetti, “an immaterial entity, as Saussure suggests … virtually present only in the memory of the culture and actualized from time to time in texts (that are material)” (Bertetti 2017: 95). This implies a systematic semioticization (or culturalization, as opposed to naturalness) of perception which produces figures as a kind of virtual repertoire of sign-functions that are then selected and actualized in texts.^[1]

While this does correspond pretty well to the way figure is treated in so-called standard Greimasian semiotics, i.e., in narrative semiotics of text, it does not correspond that well to Greimas’ original definition, which I want to reconstruct and reconsider. According to this original definition, figure is a correspondence of language and perception: he defines figures in language as “figures of the content which correspond to the figures of the expression plane of the natural semiotic system (or of the natural world)” (Greimas and Courtés 1982: 121). Unless we (simplistically) take the natural world of perception to be the same as language, a strictly Saussurean interpretation of Greimas’ notion of figure is inadequate. Figure might be accessed via text, but it is not a type of virtual sign-function because it is rooted in perception as much as it is rooted in language, and perception is primarily about access to the world and experience of imagery rather than a repertoire of sign-functions. What is more, semiotics of text is the first to suffer from this interpretation, because by simply including figure in text as a variety of sign-function we lose sight of the relation of text to perceptual and imaginary signification beyond text.

On the other hand, I agree with Bertetti, and against a theoretical point made by Fontanille, that in semiotics of text the status of figure is better conceived as determined by textual relevance rather than by a subject of perception and/or enunciation (Bertetti 2017: 95). For example, the rounded protruding abdominal-pelvic area of the torso of Mickey Mouse is better theorized not as due to perception or enunciation by a subject but rather as an articulation of the plastic dimension of the figurative text of Mickey Mouse. In the sense of figure, Mickey Mouse is neither an act of perception nor an utterance, a discourse about the world; it is rather a textual construct which involves perceptual form. Approaching figure in terms of textual relevance does not imply that figure is a derivation of text; instead, figure is relevance of perceptual form in text.

Bertetti also discusses other aspects of Greimas’ theory of the figurative, which I will not delve into. He does not, however, provide a more thorough reconstruction of the notion of figure in the works of Greimas. I will try and complement Bertetti’s commentary by presenting an attempt at such a reconstruction further on.

Harkot-de-la-Taille does not add much to the discussion of the original Greimasian formulation of the notion of figure, since she is mostly interested in later developments of the notion in the works of Landowski, Fontanille, and Groupe μ. However, Harkot-de-la-Taille does provide an important note on the duality in the way figure has been employed in Greimasian semiotics of text; I quote in my own translation: “On the one hand, figurativity, which encompasses discursive semantics and is part of the plane of content, designates the ‘referent’; on the other hand, it is intrinsic to the plane of expression and allows the ‘natural world’ to play a part in the apparition of meaning by way of perception” (Harkot-de-la-Taille 2020: 104–105).

This remark by Harkot-de-la-Taille captures well two general tendencies that are characteristic of how the notion of figure has been employed by Greimas and in Greimasian semiotics more broadly.

By way of the first tendency, in semiotics of text, figure has become a simplistic substitute for denotation (the “referent”). This conception of figure is based on perceptual form and yet implies this form to be a given that lies outside processes of textual articulation rather than a domain that is brought into signification by textual articulation. In this sense, there is figurativity when some element of a text has a meaning of something out there in the world, in perception, which is prior to textual articulation and hence falls out of the scope of textual semiotic inquiry. In order to play a role in textual semiotics at all, figure as denotation has to be demonstrated to be a superficial rendition of something else. This view is illustrated rather well by the notion of figurative trajectory, or “istotopic concatenation of figures,” as defined by Greimas and Courtés:

[The term] figurative trajectory will be used for an isotopic concatenation of figures, correlative to a given theme. This concatenation, founded on the association of figures – proper to a determined cultural universe – is both partially free and partially constrained, in the measure to which, when a first figure is posited, it only elicits certain other figures, exclusive of all others. Given the many possibilities of figurativizing one and the same theme, this theme may underlie different figurative trajectories; thus variants can be accounted for. For instance, the theme of the “sacred” can be taken up by several figures, such as those of the “priest,” the “sacristan,” or the “beadle.” The figurative unfolding of the sequence will be affected by this diversity of figures. The modes of action, the places, and the times where the sequence must be realized, each time conforming to the initially chosen figure, will be different from each other in the same proportions. Conversely, the polysemy of the first figure which is posited can potentially open onto several figurative trajectories corresponding to different themes … (Greimas and Courtés 1982: 117–118)

Two points are worth noting in this definition. First, in text, figures are only significant by way of correlation to thematic articulation of signification. Chains of figures are subjugated to themes, as surface indications and variations thereof. Since thematic signification is conceptual, the perceptual dimension of figure is irrelevant, it is simply a guise for denotation. Second, the thematic association of figures is determined by a cultural universe: in a certain society, a priest does this and a sacristan does that, and both of them have little in common with firefighters. In other words, figurative trajectory is in fact about textual (in)adequation to general world-knowledge: it is a reproductive notion and not a productive one. The same goes for figures taken separately: in a certain society, a “priest” is this and a “sacristan” is that; figures in text are a more or less faithful reproduction of what they are in that society.^[2] In this vein, in a passage of Maupassant, for example, Greimas first posits a figurative cover for a thematic role; second, he suggests that this figurative cover is made up of various indices which anchor the text to world-knowledge (Greimas 1988: 34–35). Thus, figure in text is seen as an empty site traversed by indexical connections: figure hosts indications of deeper thematic articulation as well as indications of pre-textual world-knowledge. This is a distortion of the original definition of figure akin to Bertetti’s textualist interpretation discussed above.

All in all, this first tendency to treat figure as a textual-semiotic substitute for denotation does away with the involvement of perceptual form in figure. It distorts the original definition of figure by replacing perceptual form with an indexical connection of lexical items to world-knowledge beyond text and to conceptual articulation in text. It does not allow for figure as a dimension of textual meaning in its own right.

As for the second tendency, in Greimasian semiotics roughly from De l’imperfection onwards, figure has been dissociated from denotation and in some cases from textual meaning altogether, in a shift towards experience and le sensible, as opposed to l’intelligible. Such have been important steps to tackle perception by going beyond text ant towards experience and interaction (Eric Landowski), by modelling the perception of the body and sense perception (Jacques Fontanille), or by considering the phenomenology of language (Jean-Claude Coquet). These steps are often more or less based on interpretations of the notion of figure and deserve to be treated separately. The scope of this paper is limited to Greimas’ notion of figure and the way it works in semiotics of text.

For the purposes of this article, suffice it to say that, in Greimasian semiotics of text, figure, inasmuch as it involves perceptual form, has been largely dismissed by conceiving it as an empty site for two kinds of indication – of a denotative world-knowledge and of a conceptual articulation, thematic, narrative or otherwise. Figure has also been dissociated from textual semiotics as a sensory-sensible-sensitive constituent of experience of the world that hardly concerns textual articulation and is a feature of a significantly expanded scope of Greimasian-semiotic inquiry.

In this article, I shall focus on the original formulation of figure in Greimas’ work and on making it do more work in semiotics of text. First, I will present a reconstruction of Greimas’ original notion of figure. Second, I will present a couple of my own attempts to develop the notion in analysis of text.

2 Greimas’ notion of figure

In a much-discussed passage in Structural Semantics, Greimas postulates that perception be considered as the nonlinguistic site where signification is apprehended (Greimas 1984: 7). For Greimas, perception is the gateway to the semantic universe, i.e., the universe of signification, which, for him, is the general outline of semiotic inquiry.

This does not imply, however, that perception comes first as a psychological given, and then is treated as a ground for linguistic semiosis. Rather, the semantic universe involves perception, i.e., perception is prone to semioticization. That is where the concept of figure comes in. In Greimasian semiotics, figure is the way, or one of the ways, perception has semiotic relevance.

This might seem to imply a point of reference in embodied or at least situated perception, but that is not something which interests Greimas. Despite breaking with the structuralist doxa of exclusively linguistic immanence, Greimas remains a theoretician of systems and levels of articulation, distinguishing between, “signs and linguistic systems, on the one hand, and signs and systems of signification of the natural world, on the other,” where, for the purposes of this article, the natural world is the domain of perceptual form. For Greimas, relations between the two kinds of systems amount to “a network of correlations between [the] two levels of signifying reality” (Greimas 1987: 19).

The correlations are not contingent; they are constitutive of an autonomous level of the semantic universe, spanning across the language-perception divide: “the sensible world [i.e., the natural world as perceived] is immediately present even in linguistic form and partakes in its constitution, by offering a dimension of signification that we elsewhere call semiological” (Greimas 1987: 22). Later on, it became common to call the semiological level the figurative level, because Greimas distinguished, in it, articulation of elementary perceptual features and articulation combining those features into figures.

For better or for worse, Greimas’ description of the semiological/figurative level of signification is rooted in linguistic practice: this is well illustrated by him leaning on the theory of distinctive features. This does not necessarily imply, however, that language determines perception. Rather, language is rooted in perception inasmuch as semiological distinctive features in language mirror – are isomorphs of – distinctions in the perceptual world:

Thus the mode of existence of the semiological level, it seems to us, can be specified to some degree: it is an ensemble of categories and semic systems situated and apprehensible at the level of perception, comparable on the whole to the schematized visual perceptions of the birds evoked by Raymond Ruyer, which allow them to recognize their friends and their enemies according to the oppositions:

long neck / short tail vs. short neck / long tail

Situated within the processes of perception, the semiological categories represent, so to speak, the external facet, the contribution of the external world, to the birth of meaning. Considered from this angle, the semiological categories seem isomorphs of the qualities of the sensible world, and comparable, for example, to the morphophonemes of which gestural language is composed. (Greimas 1984: 72–73)

Distinctions of features on the semiological level of language are isomorphs of sensible qualities of the world:^[3] this suggests that sensation is semioticized on a structural basis as language is, and some evidence for that lies in language, where we find elementary perceptual qualitative distinctions like vertical/horizontal, high/low (vertical), long/short, broad/narrow (horizontal), etc (Greimas 1984: 35–36). For the sake of some convenience, I will call this interpenetration of perception and language in distinctive perceptual features elementary figurative isomorphism.^[4]

Elementary figurative isomorphism is the basis for Greimas’ notion of figure. Yet it is not the same as figure. Figure is not just about isomorphism of elementary features, it is supposed to account for higher-order perceptual signification regarding perception of objects and processes in the world. Figure is prior to recognition of objects, processes and their properties; as perceptual form, figure is, indeed, the precondition for any such recognition: “[a] level of figures of the world, part of a finite inventory and giving a first idea of what the signifying world considered as form ant not substance could be, therefore will have been substituted for the evenemential and accidental world of objects” (Greimas 1987: 21–22).

In Greimas’ theory, there is a double distinction between elementary figurative features and figure proper. First, they are distinguished as degrees of complexity of the semiological/figurative level: elementary figurative distinctions articulate simple features which are then bundled together and articulated to make figures. Second, elementary features and figures proper are distinguished as, respectively, the immanent and manifested constituents of the semantic universe, elementary distinctions being a sort of basic implicit framework of our competence for perceptual signification, and figures being manifested because they are present as actual instances of form.^[5]

Figure as presence is found in perception of form and in lexical meaning in language, or in linguistic forms of a larger scale. It is important that the definition of figure is based on semiological/figurative isomorphism and thus requires us to be aware of both the component of perceptual form and the component of the plane of content of language, or some other system of signification or kind of text. In other words, Greimasian figures are word-percepts, picture-percepts, image-percepts etc.

An example of figure as word-percept is provided by Greimas in his analysis of the French lexeme tête (Greimas 1984: 47–55). Greimas discovers this lexeme to be perceptually grounded in a figurative nucleus composed of hierarchically ordered perceptual features-semes extremity and superativity, their order being extremity → superativity, i.e., superativity presupposing extremity. Roughly, this means that tête is something that is an extremity and superior to something else, as in being on top of something, coming first in line with respect to something etc. This constitutes the simple nuclear figure of tête. In addition, it might also be a spheroid or a point; with this latter feature added, it is a complex nuclear figure.

It is important to note that Greimas does not consider figure to be the perception or imagery of a particular kind of object. Instead, he endeavors to describe the nucleus of tête as a basic perceptual form which could play a part in denoting various objects, from la tête d’un clou (‘the head of a nail) to la tête d’un oiseau (‘the head of a bird’) to être à la tête d’un État (‘the head of a state’), and so on. This is counterintuitive to denotational habit and is meant this way: Greimas’ notion of figure is an attempt to dissociate perceptual form from denotative description in order to demonstrate that denotative meaning is a product of semanticization rather than a simple reflection of the way things are in the world. This way of proceeding rests upon the presupposition that, in signification, perception is primary. There is, of course, the implication of something being there, something that is perceived by way of figure, but figure remains a fleeting perceptual form as long as it is not semanticized. For example, for “tête” to denote a human head, the superative extremity needs to undergo semanticization distinguishing what is living from what is not, what is human from what is animal, what is adult from what is child, etc.

For an example of figure as picture-percept, we could return to the pictorial rendition which is Mickey Mouse and maintain that as a nuclear figure it is a bundle of protrusion and roundness (which presupposes sphericity) repeated in multiple instances – the feet, the abdominal-pelvic area, the hands, the head, the ears, the mouth, etc. Once again, there is the implication of some thing being perceived, but it is the articulation of protrusion and roundness that brings forth the significant pictorial-perceptual form that is figure.

3 Textual articulations of figure

I will further discuss several cases of textual articulation of figure. For a basic framework, I will first briefly review the way Greimas conceptualizes the relation between figures as semiological articulations and the semantic features attributed to them.

For Greimas, the immanent dimension of the universe of signification consists of two separate levels, the semiological/figurative and the semantic. On the semiological/figurative level, he posits invariant nuclear figures formed in the process of perceptual articulation the model of which I have reconstructed above. On the semantic level, Greimas posits conceptual distinctive features that allow for distinctions of contexts in which figures are discovered or into which figures are introduced in text. Greimas calls these context-defining features contextual semes and describes them as “that denominator common to a whole class of contexts” (Greimas 1984: 49). These are, for example, semantic features like “animals versus humans” and “animals versus objects” (Greimas 1984: 57).

Greimas defines the manifest dimension of the universe of signification as a relation established between the semiological/figurative and the semantic levels: “it is necessary that there be a meeting of the two levels, that is to say, a relationship be established between a minimum of semes belonging to different levels, producing thus the combination of heterogeneous elements. We reserve for that junction the term of manifestation” (Greimas 1984: 121). For him, lexemes are minimal units of discourse (Greimas 1984: 32) which are constituted by a (simple of complex) invariant nuclear figure, i.e., a kind of percept, and contextual semes, i.e., contextual habits of use. The repetition of contextual semes in text, by way of lexemes coming from the same or closely related contexts, constitutes textual isotopy. I believe it is reasonable to maintain that in his work with literary texts overall, Greimas has concentrated on contextual-semic description and modelling of isotopies rather than figures, and that isotopy has been developed as a properly textual notion while figure has not.

I will now attempt a textual approach to figure in a couple of cases.

3.2 Figure as disfiguration

For my second example, I want to briefly discuss a drawing from Der Struwwelpeter by Heinrich Hoffmann (2017). If the elephant in My Elephant seemed to be first and foremost an intricate complexification of semantic attribution to perceptual form, the drawing of Struwwelpeter seems first of all to disturb expectations with regard to perceptual figurative invariants. In other words, the drawing is a figurative articulation by way of disfiguration (Figure 1).

Figure 1:

Der Struwwelpeter by Heinrich Hoffmann (2017, n.d). Here is an English translation of the poem on the pedestal: Just look at him! there he stands,/With his nasty hair and hands./See! his nails are never cut;/They are grimed as black as soot;/And the sloven, I declare,/Never once has combed his hair;/Anything to me is sweeter/Than to see Shock-headed Peter.

Hoffmann’s illustrated book is a humorously didactic take on unfortunate things that happen to naughty children. In this particular drawing, we recognize a boy with overgrown nails and hair. We could say, in Greimasian parlance, that we successfully attribute the semantic features that make up what we call “boy” to the picture-percept. And yet there is much more to the figure than that. It is by no means akin to anything like a boy in a normal, even if unkempt, condition: the nails and hair are not simply overgrown, they have grown into figures themselves. In terms of perceptual form, the nails and hair are in competition with the boy, they are distorting the overall picture-percept. This is disfiguration: parts of the figure have become autonomous to such an extent that they are conducive to a transformation of the primary figure itself.

The picture suggests that we imagine the nails and hair of an unkempt boy grow terribly out of proportion. It is a figurative fiction which presents what would happen to a boy who would never cut his hair and nails. In this sense, the figure presents the passage of time and corporeal change by way of perceptual form rather than semantic attribution. We cannot tell whether the growth of hair and nails is a thing of the past, present or future, or some other tense. Instead, the temporality of growth is grasped as an unlimited expansion and disfiguration of what we would expect to be the nails and hair of a figurative invariant of a depiction of a boy in a European household. It is temporality-as-perceived, an actuality of unrestricted becoming. It is presentified, but presentification is not the present tense.

In the same vein, we cannot tell whether the boy has turned or is going to turn into a monster; instead, there is potential for monstrosity in his figure as perceptual form. In other words, Struwwelpeter the boy is depicted as a prefiguration of his own monstrosity, which goes to show how figurative form is often the site for fluctuations of subjectivity (think Dr. Jekyll and Mr. Hyde for a schematic take on the issue).

The figure of Struwwelpeter is more complex yet, since it is a moral figure, if humorously so. In this sense, figurative stability/instability of the body as perceptual form is a moral category. On the one hand, Struwwelpeter the boy is presentified in the process of becoming something else, which is a deviation or even worse, a degeneration; on the other hand, the generally humorous tone of enunciation in this multimodal text is revelatory of an explorative rather than a didactic disposition. In this sense, disfiguration is a way to imaginatively destabilize the overly constrictive boundaries of subjectivity in moral discourse.