The Semantic Underworld
by Daniel Barbiero
One doesn’t venture into the underworld without a reliable guide. We choose two: Michel Leiris and André Breton, both of whom offer rough sketches of what we’ll find in the place. First Leiris, from “Glossary: My Glosses’ Ossuary”:
By dissecting the words we like, without bothering about conforming either to their etymologies or to the accepted significations, we discover their most hidden qualities and secret ramifications that are propagated through the whole language, channeled by associations of sounds, forms, and ideas. (Leiris, 1989, pp. 3-4)
Next, Breton, from “Marvelous vs Mystery,” his important, if often overlooked, statement of Surrealist poetics:
Needless to say, the new language we envisage strives to be as different as possible from ordinary language, and the best way to achieve this is to deal exuberantly with the emotional value of words. The emotional life of words, far from being simply contingent on their meanings, predisposes them to be drawn to one another and to make a greater impact than meaning alone, but only if they are brought together according to secret affinities that let them combine in all kinds of new ways. (Breton, 1995, p. 5)
What Leiris and Breton both point to when they allude to words’ “hidden qualities and secret ramifications” on the one hand, or the “emotional life of words” and “their secret affinities” on the other, is an overlooked, if not occulted, dimension of language that nevertheless contains a wealth of significance. But their statements, as eloquent as they are, are long on programmatic suggestion and short on the details of how words’ emotional lives are lived, or how their hidden qualities are propagated through language. This lack of specifics is not to their detriment. Because all of this is, virtually by definition, out of range of ordinary awareness, any attempt to try to articulate it will inevitably be speculative. Still, one must try…
In order to account for this dimension of significance, I suggest we moot the existence and workings of what we might call the semantic underworld: a world below, and surreptitiously supporting, the world of ordinary meaning above. Perhaps a dimly perceived world of shadows. Accordingly, in the semantic underworld meaning is shadowy and only partly based on words’ referential values; it is as much a function of the non-referential qualities of language, which is where we’d have to look to find the source of the “secret affinities” Breton alluded to or the “hidden qualities” Leiris pointed out. If these qualities appear to be secret or hidden it is because they are based on associations that largely evade awareness, associations that in turn are the products of affective and other colorations prior to and pervading language. These associations make up the underworld in which the foundations of language are sunk. It is the locus of a variety of meaning that is other than meaning in the narrow sense of language’s capacity to refer to, or to name, or to represent, objects, events, and states of affairs. For the semantic underworld is above all – or rather beneath all – a repository of meaning, but a particular kind of meaning beyond representational or propositional content.
I want to emphasize that “the semantic underworld” is only a trope, not an actual hypothesized or hypostasized entity located somewhere in the mind. Rather, it is a figure of speech to describe certain kinds or networks of meaning which generate associations between words and other basic linguistic units. In particular, “the semantic underworld” is not just another way of saying “the unconscious,” although what goes on there is at least mostly unconscious, in the sense that it is subliminal and largely out of range of awareness. Rather, “the semantic underworld” is intended simply as a suggestive theoretical fiction useful for allowing us to form an image of a particular layer of linguistic meaning.
We can envision the semantic underworld as a liminal space that recovers the basic meaning of “limina,” which derives from the Greek λιμήν, i.e., a harbor or refuge. A refuge in two senses, one positive and one negative: refuge for, and refuge from. Refuge for the affective and other structures that stand at the origin of associative networks, and refuge from rational/logical ordering and its consequent rejection of arational or affective connections between words. In the refuge of the semantic underworld, words seem to have a semi-autonomous agency of their own. And not necessarily only words; smaller units – morphemes, graphemes, phonemes: for convenience’s sake, we can refer to all of these meaning-carrying elements with the non-technical, general name “linguistic sign” or just plain “sign” for short, the “linguistic” part being understood– are included as well. It is a secret agency that is hidden from us, perhaps, but operates nevertheless. Through it signs associate with each other, attract and repel each other, mistake one for another, act as proxies for each other, form and break alliances with each other – all seemingly on an impulse whose motivations are more or less opaque to us, their ostensible employers.
From Koiné to a Coin Stamped with One’s Own Image
Before descending into the semantic underworld, we must make a distinction between two manifestations of language: language as it is held in common by a human group, and language as assimilated by and proper to an individual language user within that group. This first manifestation is language as koiné; the second is language as idiolect. It is in the semantic underworld that the koiné – the language held in common by a human group – is assimilated and converted into idiolect, which is to say language permeated by associations and affects unique to the individual language user.
The fundamental mechanism of assimilation at work in the semantic underworld is associative. Through it, words and other linguistic units acquired from the koiné are drawn to each other or pulled apart on the basis of alogical relationships forged on the basis of their referents or their material qualities. It is through this assimilative mechanism that what Reverdy termed the “juxtaposition between two more or less distant realities” originates; it is here that Lautréamont’s umbrella and sewing machine meet on a dissecting table for the first time. The Surrealist theory of metaphor, deriving from both Reverdy and Lautréamont, calls for objects from distinctly separated ontological regions to converge with a catalytic result; it is in effect a description of the dynamic process native to the semantic underworld. Its logic is analogical, its rationality arational; metaphorical identification takes the place of inference in determining what follows from what, and extra-linguistic association and conceptual content stand on equal terms in determining what the linguistic sign – the word, the phoneme, the grapheme – means. Some of these meanings can be stunning – “marvelous,” in the Surrealist terminology – and some can be mundane; some can be poetic and others prosaic. The semantic underworld contains no guarantees of being a source of profundity. Whether or not it is is largely up to the person whose idiolect it works within.
A closer look at the associative dynamic at work in the semantic underworld shows that it operates on the super-, or perhaps better, sub-semantic content of language. This content consists to a large degree in the affective shadings that language takes on for the individual language user. Words acquire emotional resonances through the associations that, largely unconsciously, we have linked to their referents. These linkages come about through our personal histories, our temperaments and dispositions, our concerns, our imaginative lives, and so forth, on the basis of which we project emotions onto language by virtue of the fact that we represent our experiences, our likes and dislikes, our desires and aversions and so forth, through language. The resulting linkages, or correspondences, make themselves felt at the level of the content the word or smaller linguistic unit carries. The important point here is that just as much as its conventional conceptual content, affective coloration attaches to this content and thus makes up a part of the meaning of the word. By a similar mechanism words and smaller linguistic units may and often do take on associations with sensory qualities: scents, tastes, colors, and the like. As an example, we need only think here of Rimbaud’s Voyelles, which associates vowels with specific colors. Crucially, these complex meanings – aggregates of conceptual/referential meaning and affective/sensory associations – are particular to the particular language user. In fact it is here, at the point where the meaning of the word or other linguistic unit is totalized to include the individual language user’s affective and sensory associations, that language becomes our language, and that the koiné is transformed into idiolect. If, as Mallarmé suggested, language is a worn coin of the common currency placed in the hand of the individual language user, in the semantic underworld we take this coin and counterstamp it with our own images which overlay and, in some cases, make unrecognizable, the received images of the coining authorities’ own symbols.
Correspondences and Constellations
Because the totalization of the content of the linguistic sign, considered as an aggregation of its conceptual/referential aspects and its affective/extra-referential aspects, is a function of the underworld’s associative dynamics, it follows the latter’s rules rather than the rules proper to the koiné. To be sure, the koiné is also home to associative relations between words and their corresponding concepts, such as is accomplished by way of metaphorization, but these associations are governed by more-or-less generally understood and accepted conventions based on, for instance, physical resemblance, part-to-whole relationships, qualitative similarities, structural or functional analogies, and so on. The total content, and hence expressive power, of the linguistic sign in the semantic underworld, by contrast, isn’t shaped by these conventions alone if at all; rather, it comes about through the more opaque and idiosyncratic workings of a dynamic specific to the individual, a dynamic that forges correspondences between linguistic signs on the basis of their affective/sensory colorations – that is to say on the basis of the “emotional lives” that Breton saw as “predisposing” them to associate with one another. (And, presumably, to avoid each other’s company when it is a question of their emotional incompatibility.)
To imagine how this mechanism might work, we can think of it as producing a series of constellations – of clusters of ideas, images, words, concepts, affects, and even of the formal, material aspects of language, which is to say its audio and graphic images (the sounds and visual appearances of words, letters, syllables, etc.)– connected by correspondences of varying strengths, which attach to words and other linguistic units. To illustrate how these constellations of meaning are structured, we can drawn a rough analogy to the Neoplatonic theory of emanation. A constellation is defined at its center by an image or idea which, like a Neoplatonic hypostasis, descends by emanating, or radiating downward, its meaning. The ideas or images onto which this initial idea or image projects its meaning(s) must have some affinity with it in order to receive its emanation; the closer the affinity, the more they receive and thus the stronger the association. At each level or distance of descent the emanating image or idea connects to other, related images or ideas; the farther removed or downward the level from the original emanating idea or image, the weaker the connection, with those most closely related to the original receiving more of the emanated meaning than other, less closely related ideas and images. These constellations are, as Leiris’ remark suggests, ramified; their emanated images and ideas branch out and connect to other images and ideas, each of which may be at the center of a constellation of its own; taken together, they form a chain of associations. Thus an idea or image that stands at the center of one constellation may sit at the periphery of another, and the same idea or image may provide the first or final link in any given chain, or may fall at any point in between. What determines which ideas and images are central to a constellation and which peripheral – which are close and which are distant – we would expect to be a function of the history and background of the person in question: his or her experiences, assumptions, temperament, habits of mind, needs, and so on. To further complicate matters there is a constant, dynamic interaction consisting of mutual influence and shaping between that personal history and background on the one hand, and the network of meaning constellations on the other. For this reason these constellations do not represent static formations. Rather, they are formed, dispersed, re-formed, and refigured through dynamic alliances that can shift with shifts in the life experiences of the language user.*
Given the pivotal role affect plays in forging associations within the semantic underworld, it might be tempting to claim that the semantic underworld is the locus of libido. I don’t think that that would be quite right, though. To be sure, something like the libido in one of its many hypothesized forms – the erotic impulse, as per Freud; the unfettered, intense emotional energy of raw experience, as per Lyotard; desire in a general, or an undifferentiated sense possibly without an object, as per Lacan – may and probably does play a role in projecting the emotions that forge the associative chains that link linguistic units. But it doesn’t exhaust all of the forces at play there. The emotional and sensory associations that find their way into words run the gamut of determinations, and thus are of various intensities, and signify something of the richness and multidimensionality of our lives. While not denying that some may be the product of anarchic desires both realized and unrealized, other, more subtle emotional associations, some of them the product of reflection and retrospection – regret and remorse, for example, or melancholy, nostalgia, pride, shame, and so on – have a part to play as well. And this is not to mention not only sensory associations, but those made possible by language’s aesthetic qualities – the sounds and shapes of language’s acoustic and graphic images — which are at least semi-independent relative to the mechanisms of emotional projection and sensory association, and which themselves are responsible for forging linkages between linguistic signs with their attendant associations.
The Janus-Faced Sign
From all that has been said above, it may already have become obvious that the content of the semantic underworld is resistant to strictly descriptive or informational reduction. This is because a reductively informational exchange – a speech act, to use Austin’s expression, or language game to use Wittgenstein’s, consisting in the constative or descriptive use of language — narrows its referent to a meaning conveyed in plain terms; stripped off or put out of bounds as irrelevant are the qualities that overflow such a reduction. We choose our words carefully, according to the logic of the communicative intent; this isn’t the time for language to slip the reins of reason and run free, as Breton would have it. We want it to hold to the straight and narrow. The extra-semantic, associative aspect of language we’ve assigned to the semantic underworld is suppressed in ordinary informative discourse or is simply ignored as irrelevant to the exchange of information. There it would be out of place, a distraction at best…
And yet this dimension of language just is inherent to language as the language user lives it. It is always there even if overlooked or repressed in the interests of practical communications. And it always will be there as long as language is used by people. This gives language something of a dual nature – a polytonality, as it were. A phrase written by X will, in addition to expressing or representing a meaning legible from its surface, embody the associations and affects that moved X to write it as he or she did. That same phrase, read by Y, will call up different associations and affects; some may overlap with X’s and some may not, since both X and Y inhabit different language microworlds by virtue of their unique idiolects, even if they share the macroworld of a koiné. And, to be sure, some types of speech acts or language games may draw on a vocabulary chosen for its emotional force and consequent rhetorical impact, but such choices presumably would be made on the basis of something like a commonly understood, linguistic encoding of sentiment shared by language users within a community, and not on the basis of associations unique to the speaker. Even here, in language motivated by its desired emotional effect on its listeners or readers, the logic of the argument or the communicative intention reins in or represses the more freely associative logic of the semantic underworld.
Still, as an (inevitable) expression of extra-referential affective, sensory, or other associations in addition to, or rather than, simple denotation, language complicates its role as a medium for information. Built into it is the capacity to face in two directions by virtue of its encoding both types of content, and hence to operate simultaneously at the levels of koiné and idiolect. The upshot of this is that the word is bimodal. Not just polysemic — that is, capable of carrying multiple referents, as is not uncommon for words – but containing meaning in different modes, one of which is conceptual/referential and one of which is associative. The bimodality of the word (and by extension, smaller language element) pushes language to its limit as a communicative medium by pointing to its inclusion of what is, in ordinary discourse, left out beyond that limit.
The idea that the extra-semantic, affective/sensory associations of the semantic underworld make themselves felt through the word envisions the latter in a way that bears some resemblance to Lyotard’s notion, explicated in The Libidinal Economy, of the word as tensor-sign. Lyotard’s basic idea is that the sign contains or is constituted by a “tension that prevents it from having either a unitary designation, meaning, or calculable series of such designations or meanings” (Lyotard, 1993, p. xiv). The tensor-sign is, in other words, a bimodal marker; it is “[a]t the same time a sign which produces meaning [through its position in relation to other signs within the system of which it is an element], and a sign producing intensity through force and singularity” (Lyotard, 1993, p. 54). What I think this means is that a word, as a linguistic sign, holds in tension its surface or common meaning – essentially, the informational or discursive content recognizable by members of the language community of which it is a part – and the “emotional event” (Lyotard, 1993, p. 51) that, while standing at the origin of the sign’s meaning, exceeds and is ordinarily excluded from consideration when the meaning of the sign is interpreted in, say, the mundane transactions of daily conversation. The idea of the sign as a tensor seems to represent a way to reintegrate the extra-semantic implications of the sign – what I’ve described as the affective, sensory, and other associations it draws from the semantic underworld – into its content without reducing it to strictly representational terms.
Just as with the tensor-sign the libidinal force of the extra-linguistic world is not entirely shut out of language’s representational structures but rather underwrites them, the extra-referential, associative dimensions of the semantic underworld don’t elude meaning but rather supplement it or lend it a particular coloration – without, however, entering into the sign’s representational content proper. Affect is neither represented nor expressed, but rather appropriated by the sign – it becomes one of the properties of the word or other language element. By the same token, the word or other sign neither represents nor expresses its sensory associations but rather embodies them. Because they are embodied in the linguistic sign as elements grounded in the language user’s unique experiences, they elude exhaustive reduction to a plain, informational exchange. It’s always the case that something of idiolect overflows, escapes, or is eliminated as superfluous from the koiné; in this instance that something consists in the affective or sensory correspondences linking language units in non-discursive ways.
Conceiving of the word as a tensor-sign as Lyotard does serves to remind us that libidinal energy – or affective coloration, as I prefer to call it – is an integral aspect of representational systems generally and of language specifically, even if this coloration is not itself directly represented. It is, rather, embodied or expressed; it is part of the meaning of the sign but is not something referred to or denoted by the sign and hence plays no necessary role in ordinary discourse. (A similar observation may be made of the sensory or aesthetic dimensions of the sign.) Its embodiment by the sign constitutes its peculiar mode of making its meaning present, just as the representational, denotative content of the sign constitutes a different mode of presenting meaning. The sign, in short, is a bimodal, Janus-faced thing looking in two directions: the direction of conventionally secured, narrowly semantic meaning, and the direction of the unique affective and other associations particular to the individual language user.
The Poetry of the Semantic Underworld
In embodying the affective and aesthetic associations unique to a language user’s idiolect through the bimodality of the sign, language reveals itself to be liminal, a conduit between two worlds: the world encoded in the koiné and that encoded in the individual language user’s idiolect, which is to say the daylight world of everyday linguistic exchange and the shadow land of the semantic underworld. The world of common meaning, and the world of personal experience.
The upshot of this is that speech or writing manifesting the meanings drawn up from the semantic underworld is speech or writing that manifests the speaker or writer in a particularly intimate mode of self-disclosure. It is a self-disclosure made poignant by the fact that the intimacy of what it discloses in its apparent immediacy inevitably is bound up in the mediacy of how it discloses. For in order to bring up the constellations of meanings secreted in the semantic underworld the speaker or writer must concretize them by expressing them in the impersonal material provided by the koiné. This concretization is not an alienation but rather a projection outward, an ek-stasis or standing out from oneself through which what is made to stand out is, precisely, oneself.
Such an ek-static self-disclosure in language would involve a certain derangement or derailment of the conventional use of language, a derailment that Breton described as the deliberate dropping of the reins of common sense in order to let this other sense show itself. We might say that when language is allowed to organize itself according to the rules of the semantic underworld rather than of conventional discourse – when it is given free rein, or in Leiris’ formulation, when it is organized in defiance of ordinary etymologies or accepted significations – it enunciates the language user as the subject of the enunciation in both senses of the term “subject”: as what that enunciation is about, and as the source of agency for the enunciation as an intentional act. This represents an extraordinary use of language, literally, in that it overflows and points beyond the ordinary use of language as an everyday medium of exchange. The speaker, in a sense, becomes the meaning of what is spoken.
With this ek-static autoenunciation we exit the realm of discourse and enter the realm of poetry, which is itself an extraordinary use of language. The intention to use language as a medium of exchange may still be there and indeed may be a necessary component, but what is exchanged is no longer the plain version of what is – Wallace Stevens’ “vulgate of experience” — but rather the imaginative expression of a world of meaning constituted by Leiris’ “hidden qualities and secret ramifications.” This is the purview of poetry as it pushes language up against its limits, where the associative dimension of language, revealing its ordinarily concealed aesthetic and affective correspondences, surpasses its plain informational content and gives it a specifically poetic bearing. But as Breton noted in “Marvelous vs Mystery,” slipping these conventions and leaving the realm of plain communication for the realm of poetry is is not an easy thing. It “requires a painstaking effort and involves a certain amount of suffering. The words we need are not always available…” And yet they are there nevertheless, linked together in catalytic combinations and waiting to be brought up from the underworld in whose shadowlands they are natives.
* I should note here that in his early work The Necessity of the Mind, Roger Caillois proposed the existence of a pre-linguistic layer of thought consisting in complex, cognitive-affective mental representations he called “ideograms,” to which the constellations of associative meaning described above bear some resemblance.
References:
Breton, A., “Marvelous vs Mystery,” in Free Rein, (tr. Michael Parmentier and Jacqueline D’Amboise), Lincoln, Nebraska, U of Nebraska Press, 1995.
Leiris, M.,“Glossary: My Glosses’ Ossuary,” in Brisées: Broken Branches, (tr. Lydia Davis), San Francisco, North Point Press, 1989.
Lyotard, J-F., The Libidinal Economy, (tr. Iain Hamilton Grant), Bloomington, Indiana, U of Indiana Press, 1993.
