Eco, Riffaterre, and a poem by Baudelaire

1.1 Modernism and “openness”

Umberto Eco’s concept of “openness” has the advantage of allowing for the way modern art in general has the faculty of presenting a new vision of his world to the addressee. According also to my structuralist/semiotic position, it does this by inviting him to participate in the decoding process, using particular structural conventions – since modern poetry at least is structured in a quite different way from normal prose, as I have argued in several publications since 1991 (Hopkins 1991, 1994, 2007, 2015, 2016, 2019, 2022 [2020], 2023). It is in this sense of having to make the effort to follow different conventions of structural analysis that Eco’s term of “openness” most plausibly applies to a modern poetic text. (This should not imply that the text is open to a number of different interpretations.) Incidentally, Chapter 2 of The Open Work (“Analysis of poetic language”) devotes pages to Racine and Dante, but despite Eco’s liberal use of the term “semiotic,” he never presents us with a semiotic analysis – in Peircean terms – of a modern poetic text. He briefly mentions poems by Ungaretti and Eluard in this book, paying attention only to details of the textual surface (Eco 1989 [1962]: 59–61). He does not attempt to follow different structural conventions of reading from those of ordinary prose. (NB: In this essay, I will not treat work of Eco’s that does not deal specifically with modern literature.)

1.2 Tradition and modernism

To quote from David Robey’s critical Introduction to The Open Work, “traditional or ‘classical’ art, Eco argues, was in an essential sense unambiguous . . . For readers, viewers, and listeners there was in general only one way of understanding what a text was about, what a painting or sculpture stood for, what the tune was of a piece of music” (Eco 1989 [1962]: x). Thus, according to Eco, “traditional art confirms conventional views of the world, whereas the modern open work implicitly denies them.”

Occasionally, Eco broadens his scope, claiming that “openness” (to varying interpretations of a work) is to be found in all art – regardless of period. This would in fact allow for certain non-modern genres in non-European poetry that seem to be structured in a “modernist” way. But I should make it clear at the outset that, in the present article, my theory is concerned only with modern poetry and not other artistic genres – with the occasional example of film in an intertextual capacity. Eco, on the other hand, in casting his net wide (as was the norm with Gruppo 63, in which he participated in the early 1960s), does not seem to allow for the possibility that different artistic genres may have different means of signifying. One thing to note is that Eco does say – particularly in later work – that there are limits on interpretation (cf. Eco 1994 [1990]: 6). In other words, the addressee is not free to make whatever interpretation of a work of art may come into his head. This in fact seems to be Eco’s mature position, but it is apparent that he had reservations about complete freedom of interpretation back in the early 1960s.

To quote again from David Robey’s critical Introduction to The Open Work,

the interpretation of the modern open work is far from entirely free; a formative intention is manifest in every work, and this intention must be a determining factor in the interpretive process. For all its openness, the work nonetheless directs the publicʼs response; there are right ways and wrong ways, for instance, of reading Finnegans Wake. (Robey in Eco 1989 [1962]: xii; my emphasis)

The italicized section is reminiscent of Michael Riffaterre’s notion of the message being encoded in the work. The Joycean example would hardly be relevant to modern poetry, unless Finnegans Wake turned out to be structured paradigmatically. (Please refer to Section 2 below for an outline of Riffaterre’s and my theory of modern poetic structure.)

It should be noted that proto-modern poetry in Europe begins with certain of the late Romantics, such as Wordsworth, with some outstanding examples being found somewhat later in Victor Hugo (cf. Hopkins 2022 [2020]: chs. 3 and 4). Modernism gathered momentum in France with poets like Baudelaire and Rimbaud. This means that even when the alexandrine meter and rhyme are conserved, the poem is often built on two paradigms of catachrestic images, each generated by an underlying “matricial” proposition. In other words, whereas the poem may perhaps be paraphrased syntagmatically as prose, it is in fact structured paradigmatically – in sharp contrast to ordinary communicative prose. Early examples are Hugo’s Fenêtre flamande (1840) and Baudelaire’s Spleen no. 1 (1857; cf. Hopkins 2022 [2020]: ch. 4) and Les Chats (cf. Section 4 below).

2.1 Introduction to a semiotic theory of the structure of modern poetry

Here I shall outline the basics of a theory of the structure of modern poetry which is staunchly opposed to any approach that would allow the reader to make what he likes of a poetic text. Few critics insist on a notion such as Michael Riffaterre’s of the monumentality of the text, due to which its meaning is internally encoded. Such texts are nonetheless still “open” in the Econian sense: first, in allowing the reader to focus temporarily on individual images out of semiotic context; second, in asking the reader to search through the text for images generated by the same underlying propositional structure, or “matrix.”

My theory represents a considerable expansion of the theory of Riffaterre as represented particularly in Semiotics of poetry (1978) and over a hundred articles. Eco himself leaves Riffaterre largely aside – even though the latter is a prime participant in the great Les Chats debate with which he (Eco) begins The Role of the Reader (1984 [1979]; cf. Section 4 below).

While Eco insists that “traditional” or “classical” art is more or less unambiguous, “much modern art, on the other hand, is deliberately and systematically ambiguous” (Robey in Eco 1989 [1962]: x). This statement could take into account the fact that individual symbolic images may initially be read out of context as having a mimetic meaning. Later, when other images are compared with them, they are found to be generated by one or the other global textual proposition (matrix), and thus share a common meaning. E.g., the first (“Marie”) image in The Waste Land need not in fact refer extra-textually to a particular young aristocrat, but is instead overdetermined by the “disaffected women” seme, which it shares with other images (“vignettes” is more appropriate) having a human protagonist. It appears to be this same situation in symbolist poetry, which is referred to by Eco in the following quotation:

In an aesthetic stimulus, it is not possible to isolate a particular sign and connect it univocally to its denotative meaning: what matters is the global denotatum. Each sign, depending as it does on all the other signs of the proposition for its complete physiognomy, can signify only vaguely, just as each denotatum, being inextricably connected to other denotata, can only appear as ambiguous when taken singly. (Eco 1989 [1962]: 36)^[1]

Eco’s note regarding modernist “ambiguity” may be taken to imply that apparent localized mimetic meaning of an individual image has no more than a transitory meaning. For example, Eliot’s Waste Land is not about Marie, the “hyacinth girl,” or Madame Sosostris. Thus, “can signify only vaguely” is well justified in such cases. The reader needs to go further in the quest for its symbolic meaning, beginning a comparison with other images. That this meaning functions on the global level is plain. That is why any analysis of part of a poem only is generally misleading, the essential unit of signification being the complete text.

Although Riffaterre is not explicit about the modernity of his textual examples, the examples speak for themselves: Victor Hugo, Baudelaire, Rimbaud, Lautréamont, Mallarmé, and so on. It is important that Eco points out the strong contrast between traditional and modern art. I would suggest that such modern poetry may be seen as an alternative means of expression to prose – and to pre-modern or postmodern poetry. Riffaterre’s notion of the monumentality of the modern poetic text is useful in this context. Unlike modern prose such as Finnegans Wake, modern symbolic poetry does not invite a variety of interpretations. Besides being different in length, the two genres differ in this regard also.

2.2 Matricial structure

According to the theory developed in Semiotics of Poetry (Riffaterre 1978), and articles going back to 1966 (on Baudelaireʼs Les Chats; cf. Section 4 below), the process of interpreting a poem entails a search for a single theme which generates the whole text: a “matricial” concept, which may consist of a single word, but always has the structure of an at least potential proposition. This structure is developed by a set of images, the “ungrammaticality” (or catachresis) of whose surface structure signals that they are generated by an underlying proposition which represents the matrix, and which may seem to refer, individually, to disparate sociolectic phenomena. As we shall see below, there always seems to be a second theme – or matrix – as well: the propositional structure of modern texts is binary.

In seeking to reconstruct matrices, the reader may appeal to the text-wide authority of an “intertext,” a structural model which may be found in other – usually pre-existing – texts or works of art. (In Riffaterrian terminology, the term “structure” always has the sense of the semantic structure of an image, which is modelled on one matrix or the other, making it a poetic sign.) Riffaterre himself gives the example of two variants of a common text/intertext matrix, the first involving a camel “crossing the trackless sands of the desert,” and the second involving a ship “furrowing the briny deep”: both variants are built on the matricial proposition that a trusty conveyance reliably bears human beings across a vast and dangerous expanse (Riffaterre 1984: 142). In other words, there is an analogous relation between the two components of each matrix.

The lexical form of the matrix is often modelled on that of a hypogram – i.e., on a phrase (often a cliché or commonplace) already existing in the sociolect, which has no necessary relation with the matrix on the level of structure. The fact that its set of matrix-based images is thus controlled, on both the level of semiotic structure and that of lexical form, confers on the text its “monumentality” – the faculty of resistance to deformation by the reader. According to Riffaterre’s sometime version of the Peircean triad of signification, the interpretant of the text (as subject-sign) and intertext (as object-sign) can be a third text. In fact, according to Peircean convention, the interpretant should properly be put together by the reader from a comparison of the preceding two signs, between which it mediates; this is Riffaterre’s mature position (cf. Section 2.4 below).

Riffaterre considers the mechanisms of traditional prosody, among which he includes syntactic parallelism, as well as all categories of semantic catachresis, as heuristic signals suggesting the existence of underlying poetic significance. His emphasis on catachresis (or “ungrammaticality”) reflects the fact that he is concerned above all with the distinctively literary structures of the modern poetry of the twentieth century, and that of its precursors in the nineteenth, a good deal of which lacks a traditional prosody. Whatever the genre of poetry involved, this emphasis enables him to insist that it is neither the individual images – nor the seeming nonsense of their catachresis – nor their apparent reference to extra-textual phenomena, which carry the text’s “significance” as a poem. Rather, the apparently contradictory references of the surface structures of poetry should alert us to the fact that modern poetry signifies indirectly, in other words, by symbolism. To quote T. S. Eliot, from his 1921 essay on “The Metaphysical Poets,” “poets in our civilization . . . must be difficult. Our civilization comprehends great variety and complexity . . . The poet must become more . . . comprehensive, . . . allusive, . . . indirect, in order to force, to dislocate if necessary, language into his meaning” (Eliot 1980 [1951]: 164; I will take “indirection” in Eliot’s sense, as referring to the symbolic use of language).

In practice, far from being no more than prose arranged in discrete lines and sometimes obeying prosodic rules, modern poetry is usually generated by two underlying matricial propositions, each of which produces a set – or paradigm – of symbolic images on the textual surface. Thus, the many vignettes of different personages and their activities in The Waste Land turn out to be variants of one or the other matricial proposition.

2.3 Indirect signification

Indirectness of signification (including symbolism) is the defining feature of Riffaterre’s rather broad term “literariness.” Such a definition naturally excludes all genres which depend on direct reference to the extra-textual world – satire, for example. Less obviously, it would also seem to exclude a kind of verse, produced under the influence of “postmodernism,” which often signifies by direct reference to extra-textual phenomena. Among the mechanisms of catachresis, metaphor is by far the most frequent. Apart from its heuristic function, Riffaterre shows how metaphor may be used to establish lexical codes: a text can develop a whole series of images whose vocabulary is based on the same hypogrammatic “descriptive system” (Riffaterre 1978: 22, 39–40).

It is important to repeat that the semiotic structure of the text according to Riffaterrian theory is impervious to any influence brought to bear by the reader. The latter’s contribution to the interpretive process begins with the considerable effort involved in reconstituting their underlying structure from a comparison of the images of the textual surface. Riffaterre terms this “retro-reading,” since only after all the ungrammatical figures of the text have been taken into account – regardless of their linear order – can their common underlying structure be discerned.

In stressing the invulnerability of the text to any input by the individual reader, Riffaterre is only fitting his theory to the kind of material he is dealing with: the modernist text is far from laying itself open to modification by the reader. Instead – though Riffaterre does not make this point – it sets out to change the reader’s preconceptions regarding reality. (Its very monumentality would seem to be essential in giving the text this power; on perceptual change, cf. reference to Eco [1976: 275] in Section 5 below.) The role of the intertext as a model (object-sign) is important in this regard, since it overdetermines the propositional structure of the primary text (subject-sign). The originality of the latter consists in its having a unique vocabulary set into a propositional framework which is generally homologous with that of an intertextual model. The Riffaterrian reader, then, requires a certain amount of intertextual background knowledge, in addition to an alertness to the existence of hypogrammatic structures which will influence his choice of vocabulary as he seeks to formulate a matrix. One might say that a Riffaterrian “literary competence” (cf. Culler 1975: ch. 6) metaphorically involves both a grasp of the “syntactic” rules of poetic structure according to the theory, and a knowledge of the “lexicon” of intertextual and hypogrammatic material.

In my expanded theory, the above rules become more complex in that every modern text has not one but two paradigmatic sets of images generated by two matricial propositions, linked together by a syntagmatic relation, which may be of several types: negation, difference of scale, etc. Also, the démarche of stepping outside the text to discover the relationship between its interpretant and its “sociolectic context” is vital if we are going to treat the modernist text adequately. Unfortunately, Riffaterre prefers to confine himself to textual structures, sometimes (to repeat) even when it comes to the interpretant.

2.4 Two levels of semiosis

On an intra-textual level, each successive image in a given set is the object-sign of any other subject-image (not necessarily in linear order); the matrix – assembled by the reader – is the interpretant of each pair of images.

On the intertextual level (to repeat), the intertextual model functions as object-sign in relation to the subject-sign of the primary text. This relation highlights the importance of the interpretant, which is put together by the reader in its Peircean rôle as mediator between those two signs: as “the idea to which they give rise.” Although Riffaterre does say several times that the interpretant may be a third text, (to repeat) he also adopts this classic definition of Peirce’s (CP 1.339; Riffaterre 1978: 81; cf. also Riffaterre 1985: 42, 1990: 13).

2.5 Four stages of decoding

The investigative work demanded of the reader by the modern poetic text may be divided into four stages as follows: (i) that of comparing the images belonging to a single set in order to uncover their underlying common features as poetic signs; this stage involves two distinct sets of matrix-generated images; (ii) that of searching for an intertextual model for the global structure of the bimatricial primary text, which is analogous to the semiotic structure of the text, particularly on the syntagmatic level of the relation between the propositional structures generating each set of images; (iii) that of “abducing” an interpretant from a comparison of primary text and intertext; and (iv) that of the search for a sociolectic context (SC), in rare cases via an appeal to a recursive series of interpretive triads, the interpretant of the textual triad becoming the subject-sign of a subsequent triad, and so on (a case of Peircean “unlimited semiosis”). The SC is lexically similar to the final textual interpretant, but the two elements of its proposition – subject and predicate – are linked in a different way (positive versus negative, difference of scale, etc.). This difference foregrounds the uniqueness of the proposition made by the interpretant, thus bringing about a change in the reader’s preconceptions. Clearly, the first three above stages belong to the textual (intra-textual and intertextual) component of the total interpretive process; the last belongs to the extra-textual component.

3.1 Isolated images are ambiguous

As far as could be gleaned from Eco’s writings (1960–1992), he does not propose a theoretical framework that would account for a way in which modern poetry may be decoded. Chapter 2 of The Open Work is entitled “Analysis of poetic language” – but apparently this is poetics in the traditional generalized sense of “theory of literature.” There, what Eco says of Croce might often be applied to his own approach: “Croce does not accompany his observation with a theoretical framework that would account for it” (1989 [1962]: 25).

However, later in the same Chapter 2 Eco seems to approach a concept rather like our underlying generative proposition – or “matrix” – and its variant-images. Although not talking about modern poetry per se (his terms are “poetic language” and “aesthetic stimulus”), he goes on, perhaps referring (if we may imagine for a moment) to the second poetic sign in a set:

Unable to isolate referents, the addressee must then rely on his capacity to apprehend the complex signification which the entire expression imposes on him. The result is a multiform, plurivocal signified that leaves us at once satisfied and disappointed with this first phase of comprehension precisely because of its variety, its indefiniteness. Charged with a complex scheme of references mostly drawn from our memories of previous experiences, we then refer back to the initial message, which will be inevitably enriched by the interaction between those memories and the signifieds yielded in the course of the second contact. (Eco 1989 [1962]: 36–37)

This could apply to the situation in our theory of the structure of modern poetry, if only “the signifieds yielded in the course of the second contact” referred to the second variant in the same set of images generated by an underlying matricial proposition (cf. Section 2.2 above).

As Eco goes on in this vein, he switches from noting an example containing rhyme and metre to a piece of music which is familiar to the listener (Eco 1989 [1962]: 38). This very catholic range of works of art is perhaps what leads him to abandon aesthetics: “aesthetics is unable to give an exhaustive explanation of certain aesthetic phenomena, even when it can allow for their plausibility” (Eco 1989 [1962]: 39).

More or less in line with Eco, I shall propose that modern poetry may be called “open” in the sense that the reader has a range of operations to perform: briefly (in terms of Riffaterreʼs and my theory), assembling a matrix per set from a comparison of each image in either set (separately), deciding on their syntagmatic link, choosing an intertext (which may influence that link), and assembling an interpretant, for which a sociolectic context must be found. In that the text itself dictates the work performed by readers, this process is very different from simply allowing them to make just any interpretation which suits a given individual.

3.2 Eco, Pareyson, and unlimited semiosis

Eco’s account of his mentor Luigi Pareyson’s view of the structure of the poetic text is perhaps the closest to our own theory:

This, in turn, brings about a new perspective on the Crocean opposition between “structure” and “poetry,” since all the parts of the work are no longer subordinated to isolated instances of “poetry” but rather as seen as integral parts of one artistic organism: a total form in which all the so-called “padding” has a “structural” value . . . since it shares in the perfection and legitimacy of the form that it supports. (Eco 1989 [1962]: 162–163)

This could metaphorically express the way in which all variant images in a given set are seen as generated by a single matricial proposition.

According to Robey, “the principle of unlimited semiosis is, Eco argues, vital to the constitution of semiotics as an academic discipline” (1989 [1962]: xxii). This obviously assists his argument that the signified in the arts is part of an infinite progression of interpretants. The chapter in the book entitled “Series and Structure” illustrates how Eco’s theory of The Open Work is “carried over into his semiotics and gives it much of its distinctive character” (1989 [1962]: xxi). It is important here to distinguish between Peirce’s concept of unlimited semiosis and an “infinite number of possible interpretations” (Eco 1989 [1962]: 165). Eco seems in fact to map the latter concept onto Pareyson’s theory. He does this by asserting that these interpretations are “possible because [quoting Pareyson] ‘the work lives only in the interpretations that are given of it,’ and infinite because . . . this fecundity will inevitably be confronted with an infinity of interpreting personalities . . .” (1989 [1962]: 165).

At this point we must express reservations. An “infinity of interpreting personalities” will not necessarily yield an infinity of interpretations. Eco gives no inkling of the Riffaterrian notion of the monumentality of the work, which imposes its semantic/semiotic form on the reader, without regard to the number of readers who may be involved. Eco makes Pareyson’s – and presumably his own – position more or less clear:

against all those doctrines that see art as a way of knowing, the aesthetics of formativity maintains that the only knowledge an artist will necessarily offer is the knowledge of his personality concretized into a way of forming – all of which, of course, does not prevent an artist from proposing, in his art, his own personal viewpoint . . . (Eco 1989 [1962]: 165)

This seems strangely ambivalent, although perhaps because it is not concerned with modern poetry. My position is of course that the modernist poet will usually offer an idiosyncratic message, which it is up to the reader to put together. In insisting on the Riffaterrian monumentality of the work it would seem sensible not to exclude a certain amount of semantic flexibility at various stages in the readerʼs works in putting together this message. At the stage of interpretation which lies apart from the internal structure of the primary text, there is inevitably some room for choices made by the individual reader. The choice of an intertext is an obvious case in point: this may have some influence on the overall matricial structure of the primary text. Again, at the intra-textual stage, the assembling of matrices may also involve some individual freedom. This freedom which remains at the assembling stage of the complete propositional structure of the text – including intertextual influence – is possibly part of its overall “openness.”

3.3 Eco on structuralism

In Chapter 10 of The Open Work, “Series and Structure,” Eco discusses Lévi-Strauss’s structuralist approach to communication:

[S]tructuralism – as well as other linguistic and ethnological schools – is today aiming at the discovery of constant structures, simple, universal articulations capable of generating all the various systems that they underlie.

It is therefore quite logical that whereas structural thought tends to recognize “universals” [as Riffaterre does], serial thought prefers to denounce them as “pseudo-universals,” mere historical phenomena. (Eco 1989 [1962]: 227)

Thus, according to Eco’s thinking, the possibility of multiple interpretations of a work of art depends on a “serial” approach (one interpretation following another), whereas I would prefer to say that a modernist work of art – especially poetry – depends, at its most abstract level, on a “structural” approach, every interpretation being based on the same laws governing how a work is constructed.

4.1 Is Les Chats an “open” text?

In The Role of the Reader (1984 [1979]), Eco announces in his Introduction – over two decades having passed since parts of The Open Work were originally published in Italian – that he will now present an updated version of his thinking on openness (Eco 1984 [1979]: 3). He launches directly into a commentary on Jakobson and Lévi-Strauss’s (1981 [1962]) famous analysis of Baudelaire’s sonnet Les Chats, which is designed to show how the two collaborators were adhering too rigidly to a structuralist methodology. Eco counters that “‘Les Chats’ is a text that not only calls for the cooporation of its own reader, but also wants this reader to make a series of interpretive choices which even though not infinite are, however, more than one. Why not, then, call ‘Les Chats’ an ‘open’ text?” (Eco 1984 [1979]: 4).

Eco makes no mention of the equally famous counter-analysis of this poem by Michael Riffaterre (1966: “Describing poetic structures: two approaches to Baudelaire’s Les Chats”). This analysis – to which I will add some adjustments from my modified version of Riffaterrian theory above – is based on the principle of the “monumentality” of the modern poetic text. This poem may well be an “open” text, in that the reader has a good deal of investigative work to do. But it is certainly not “open” in that it does not provide for more than one interpretation. (This assertion will be buttressed below by my outline of an interpretation.) In this case at least, Eco’s definition of openness is over-permissive.

4.2 Two linked matrices

4.3 Two matrices in Les Chats?

6.1 Eco and the individual image: Eco 1994 [1990]: ch. 1

It seems that the key to Eco’s apparent difficulty in defining what sets limits on the interpretation of a poetic text may be twofold. First, he is very seldom actually dealing with modern poetry, but rather with “artistic texts” in general. Second, he is apparently dealing with “literary language,” rather than the complete text of a modernist poem – despite the note about “the text as a coherent whole” in Eco (1994 [1990]: 149). When it comes to symbolism, the result of this bias is that Eco generally thinks in terms of the individual symbol, instead of focusing on a complete set – or paradigm – of symbols. This means that, as regards modern poetry, he does not treat the text as a complete poem. Eco thus courts the danger of treating symbols as being able to refer individually to extra-textual objects, instead of to each other (as they do in a modern poem). He thus risks not seeing that a single set of symbolic images is generated by an underlying matricial proposition that does not appear on the surface of the text.

To quote from a relevant section of Limits:

A text is a place where the irreducible polysemy of symbols is in fact reduced because in a text symbols are anchored to their context. The medieval interpreters were right: one should look for the rules which allow a contextual disambiguation of the exaggerated fecundity of symbols. Modern sensitivity deals on the contrary with myths as if they were macro symbols and – while acknowledging the infinite polysemy of symbols – no longer recognizes the discipline that myths impose on the symbols they involve. Thus many modern theories are unable to recognize that symbols are paradigmatically open to infinite meanings but syntagmatically, that is, textually, open only to the indefinite, but by no means infinite, interpretations allowed by the context. (Eco 1994 [1990]: 21, my italics)

Here, Eco does not seem to appreciate that, in the italicized line, “contextual disambiguation” can only come from a symbol’s relation with other symbols in the same set or paradigm – i.e., generated by the same underlying proposition and stacked together vertically on the paradigmatic axis of language. At least as far as modern poetry goes, he misleads in stating that “symbols are paradigmatically open to infinite meanings,” whereas the meaning of symbols in the same set, or paradigm, is restricted by their generating proposition. For example, if we take Eliot’s Waste Land as an example (in order to avoid any charge that Baudelaire’s Les Chats is not quite modern poetry), as Eliot himself says, “all the women are one woman” (note to line 218). In other words, the Marie of the first eighteen lines – and her activities – are generated by the same underlying proposition as all the other lengthy images involving women, including the Queen: they are all disappointed in love, or otherwise disaffected. (These images are described in more detail in my 2016 article: “A semiotic key to The Waste Land.”) The bimatricial propositions in this case are of the general form of “Faithless people, preoccupied by worldly concerns, are cut off from divine love.”

This paradigm, which runs throughout the poem, is far from being “open to infinite meanings”; instead it is strongly restricted by the form of its matricial model. Further – no doubt because of his focus on literature in prose – Eco neglects the linguistic form of modern poetry by equating the term “syntagmatically” with the term “textually.” This suggests that he is reading modern poetic texts as linear prose, instead of as two sets of paradigms. The main syntagmatic aspect in the semiotic structure of a modern poetic text (apart from the propositional structure of an individual image) is of course the relation between the two matrices. E.g., in The Waste Land, that relation would hold that The gospel message is able to rehabilitate (IMR) the fallen world which people normally inhabit.

Eco is also misleading in the section that follows: “this principle does not mean to support the ‘repressive’ idea that a text has a unique meaning, guaranteed by some interpretive authority” (Eco 1994 [1990]: 21). Following Riffaterre, I would affirm that the “interpretive authority” is the text itself, in its semiotic monumentality. Thus, the signification of the modern poetic text is tightly controlled, as its form is that of two paradigms of images, whose underlying generative propositions are tied together by a single syntagmatic link. It is very far from being narrative prose, to which it constitutes an alternative form of expression.

Eco raises further doubts at the very end of this chapter in stating: “when symbols are inserted into the text, there is, perhaps, no way to decide which interpretation is the ‘good’ one” (Eco 1994 [1990]). His equivocation is perhaps justified by the absence of knowledge of a theory such as ours when drafting the book in question. This statement only serves to reinforce the impression that Eco is taking symbols individually, and not in their matricial context. It raises doubts about whether he is dealing with complete poems at all. A copy of my 1994 published dissertation on matricial structure theory was handed to him personally at the 1994 IASS/AIS world congress at Berkeley, California – but obviously a little too late. Rather than focusing on the lack of time-lapse between publication of his 1994 [1990] book and the congress, we should recall that the theory in question harks back to 1966, and Riffaterre on Les Chats.