The Poet Sings: “Resonance” in Paul Valéry’s Poietics

Abstract

This paper analyses Paul Valéry’s theories relating to his stated goal of poetic production: the attainment of “resonance” and a “singing-state”. My intention is to defend Valéry’s theory as a valid and consistent model of the creative process in poetry. To that end, I will draw support from T. W. Adorno’s claim that Valéry’s manner of reflective journalising in his Notebooks can furnish us with what he calls “aesthetic insight”. The consistency of Valéry’s theory will be supported by comparisons with the inferentialist understanding of semantics. Valéry proves to be a reliable exemplar of what might be called a “practice-led” aesthetics.

1. Introduction

My response to the themes and questions of this volume will be to present an analysis of a small number of entries found in Paul Valéry’s Notebooks (Cahiers); more especially, entries which assert ideas and evaluations that are in some way associated with music. In the examples I analyse, music is drawn into Valéry’s poietics as he theorises about the goal of achieving a poetic state of being: he refers to the “singing-state”, to “harmonics” and “resonance”. What are we to make of such terms? Do they accommodate a certain—pleasing—vagueness in our apprehension and experience of the poetic? Do such terms, taken as metaphorical, give license to indulge an affection for what Hegel called “rapturous haziness”? Or for the “indeterminate enjoyment” of an “indeterminate divinity”? [1]. Writing as a musician, it seems to me that referring to poetry in musical terms presents certain risks and certain temptations. One risk is a kind of transference: the sensuous pleasures afforded by poetry—so hard in themselves to determine—can be attributed to the musical, music being the acceptable face of sensual indulgence, inscrutability and non-comprehension. Then there is what can be called the Paterist thesis, a reference to Walter Pater’s suggestion that all the arts aspire to the condition of music. For instance, if poetry is “musicalised” language, if poetry emerges in the emphasising of the latent rhythmic and sonic qualities of language deemed superfluous in ordinary communicative speech or prose, and if these qualities are judged to be values (and not pretentious ornaments or distractions), perhaps they ultimately derive their value from those that are inherent in music [2]. If we value music and the musical, anything “musicalised” inherits those values, just as anything gold-plated inherits the values of gold.

While Valéry is not immune to these kinds of tempting suppositions, there is at least a strong countervailing tendency in Valéry’s thinking. For him, whose taste for the rational and mathematical is apparent, music is not generally advanced as exemplary of vague, emotional power in art; rather, music has the enviable condition of being highly rationalised, pre-structured, into notes, scales, modes, chords and so forth, all of which lends music to systematisation, calculation and combinatoric freedom: music can be posited as exemplary of rational, constructive affordance. Musical materials are always-already primed for compositional activity in a way that language is not. As Valéry complains, the poet has to go against the grain of language and raise it to the state of song:

The poet’s problem and the musician’s problem.

Music exists in advance of the musician. The scale, the chords, the instruments are there for him. He doesn’t have to concern himself with music itself—with sounds—only with choosing those he needs.

But the poet is forced to create poetry from moment to moment, to turn noise into sound at every moment, to make the reader sing despite himself and to work in similar fashion on his ideas so as to make them poetic, i.e., worthy of being sung and inseparable from his being, when on the contrary words and phrases are naturally meant for daily use, and get used up—and when all this continual use of words makes them something quite different from the poet’s song—something to be silently understood.

([3], p. 188)

In comparison to the musician, the poet has an additional burden: s/he will have to hone and tune her/his materials out of the din of language as well as think about form and the pleasures of forming. Any “Paterism” in Valéry arises from these kinds of formalist inclinations: a formalist can always find some degree of solace and encouragement in music’s trick of acquiring a reputation for being highly meaningful without troubling listeners with concepts and references. But that said, however much Valéry might be drawn to music, or envy the material riches of the composer, I do not think that he had any illusions about music’s priority amongst the various arts and there is no opportunity here of using Valery’s theories as a ruse to re-enchant music by providing proof that music can (simply, somehow) fecundate language and, in doing so, raise it to a “higher” plane, to poetry [3].

In fact, what is interesting in Valéry’s manner of theorising is that he is at least as inclined to draw on scientific discourse as on the traditionally “aesthetic”. When he writes about resonance, he means it in a way that is fully compatible with the physics of resonance, the interaction and behaviour of physical systems, the transference of energy, their mutuality and reciprocity. Yet this lets music back into the frame: depress the sustain pedal on a piano and it is transformed into a vast and complex resonator; a single action—the striking of a key—brings a huge array of other frequencies to life. If this vast resonance is in any way striking (I am a musician because I think it is) it is because the abundance of this energy, of space, all points towards an experience of the sublime. But, equally, what needs to be said of resonance in this context can be said without losing sight of physical description and quantification. What characterises my approach in this paper is that I take Valéry at his word and try to find a way of taking what he calls the “resonance” of poetic language as being strongly analogous with resonance in an acoustic or physical system, and I opt for the most rationalised presentation I can find. Anything said of poetry in that regard is, mutatis mutandis, applicable to music.

In this paper, I will argue that there is nothing allusive or metaphysical behind Valéry’s use of the terms “resonance” and “singing-state”; these are not mere gestures. The Notebooks contain, in amongst an enormous number of other topics, a workable theory of poetic making which is precise and useful as it stands. For the sake of clarity, the outline of Valéry’s theory can be set out as follows:

(α)

The making of poetry is a process of bringing ordinary language into a singing-state by discovering its resonance.

(β)

A “singing-state” attained by poetry is the mutual, resonant intensification of the significance of language across its numerous structures and dimensions.

These ideas will obviously need some amplification (see also [4]). But in addition, I make two significant and original claims in connection with my analysis. Firstly, I take this opportunity to explore Theodor W. Adorno’s approach to Valéry’s work. I defend in particular Adorno’s proposition that, in principle, an artist’s reflections on his/her own practical experience can yield genuine “aesthetic insight”. Indeed, Adorno identified Valéry as the exemplary candidate for what I am calling a “practice-led” approach to aesthetics. This is an extremely important point because what is distinctive about Valéry is that he is a poet vigorously engaged in theorising about how to write poetry. What is more, he is a very fine poet and a highly reliable and well-read theorist. Along with Adorno (who was a creditable composer), Valéry belongs to that rare class of intellectual who have both philosophical ability and long practical experience as artists; these factors have a direct impact on what and how one writes.

Secondly, however candid Valéry might be in his Notebooks, room must be made for further interpretation. I will attempt to make his theory a little more general by re-describing Valéry’s ideas using an adapted inferentialist semantic logic. Inferentialism has a key role in capturing the rationalising tendencies of Valéry’s idiom, giving his insight into the poetic use of language a new and distinct expression. Such is my own commitment to the non-abjuration of the scientific. Two further theses can now be set out:

(γ)

Artists, by reflecting on their own working, are in a position to contribute to aesthetic discourse.

(δ)

Inferentialist semantics has a role to play in the theorisation of aesthetic reasoning and evaluation.

2. Artistic Self-Reflection as Aesthetics

It might be objected that it is not possible to extract a complete or conclusive picture of Valéry’s ideas about “resonance” from the Notebooks, and that what I am calling a “theory” is only a series of barely grammatical jottings. There is, admittedly, a truly vast amount of material to digest: Valéry contributed to his Notebooks on a daily basis from the age of 23 to his death in 1945. His dedication to this practice of intellectual self-examination, sustained for over half a century, is extraordinary, and resulted in some 26,000 pages of disorganised drafts, drawings and notes. It has been the business of editors since the 1970s to corral this mass of material and put it in thematic order, thereby rendering his work more serviceable to the scholar. This is not a straightforward task, nor is it uncontroversial. Craig Raine describes Valéry as a “feral thinker, instinctively vigilant, sceptical and iconoclastic” ([3], p. 7) and one must be cautious not to begin to suspect that the editorially imposed themes of modern editions (e.g., “Art and Aesthetics”, “Poetry”, “Poietics”) represent the implicit teloi in his thinking, teloi that he, therefore, manifestly failed to close down.

These aphoristic notes are not the preliminaries for some indefinitely postponed treatise on aesthetics; they are, according to Raine, incitements and provocations, a vivid and high tempo means of stirring up thought: they catch the artist re-inventing and crafting his thoughts. That said, Raine also admits that Valéry had areas of “stubborn interest, of energetic repetition, which might be mistaken for systematic thinking” ([3], p. 8). This aptly describes how music-concepts appear in the aphorisms under consideration in this essay: these concepts are energetically repeated. However, my analysis of Valéry does not pretend to distil what look like “stubborn” re-articulations of similar ideas into a consistent essence, and thereby establish the “real” meaning of these remarks. I also think it a mistake to subdue what are “feral” thoughts by adjusting them to fit propositional schema that would render them amenable for theory construction. Rather, the methodological challenge is that of devising a way of reading Valéry which catches this moment provocation. What Valéry seems to require is a reading at the level of discursive energetics or functionality.

That is only one limb of the methodological problems we have to contend with. I will immediately raise the stakes beyond what one might call a sympathetic close reading by arguing that Valéry’s Notebooks can be read as a source of genuine “aesthetic insight”: not only are these aphorisms witty and provocative (this much is, perhaps, easily grasped), they are also aesthetic knowledge. Such, at least, is the claim made by T.W. Adorno in two separate essays on Valéry when he sets great store by the potential of the practicing artist to furnish “insight”. In The Artist as Deputy we read: “Great insights into art come about either in utter detachment, deduced from a concept undisturbed by so-called connoisseurship, as in Kant or Hegel; or in absolute proximity, the attitude of the person behind the scenes, who is not an audience but rather follows the work of art from the point of view of how it is made, of technique.” ([5], p. 100). And in Valéry’s Deviations, “[T]he ability to see works from the inside, in their logic as artifacts, things that have been produced—a union of action and reflection that neither hides behind naiveté nor hastily dissolves its concrete characteristics in a general concept—is probably the only form in which aesthetics is still possible” ([5], p. 138). In both cases, the example (perhaps the only example) of the thinker who can “follow the work of art from the point of view of how it is made” and “see works from the inside”, was Valéry.

So important was this idea of an aesthetics emerging from creative practice to Adorno that he plagiarised it in the “Draft Introduction” to his Aesthetic Theory. The opening gambits of this his last complete (but unfinished) project concerned the diagnosis of the decline and social irrelevance of aesthetics as a philosophical discipline, but he also attempted to discover fresh directions and the impetuous for its revival. Let us consider the following passage in some detail:

“Aesthetics was productive only so long as it undiminishedly respected the distance from the empirical and with windowless thoughts [fensterlosen Gedanken] penetrated [eindrang] into the content [Gehalt] of its other; or when, with a closeness bordering on embodiment [leibhaft nah], it judged the work from within [Innenseite der Producktion], as sometimes occurs in the scattered remarks of individual artists [versprengten Zeugnissen einzelner Künstler], which are important not as the expressions of a personality that is hardly authoritative with regard to the work, but because often, without recurring to the subject, they document something of the experiential force of the work [von der andrängenden Erfahrung der Sache einiges notieren].”

([6], p. 334; [7], p. 497).

As is typical of Adorno, assertoric statements are constructed in “either/or” dialectical relationships: to be “productive”, that is to say, to be able to claim to have a perspicacious understanding of art “Erkenntnis von Kunst” ([7], p. 497), aesthetics either ignores practical matters altogether and sticks resolutely to the medium of concepts (Adorno calls this an “asceticism of conceptualisation”) and invests in the closed (“windowless”) relationship of idea to idea, or it is completely immersed in the productive process, an immersion, the truth of which is underwritten by the experience of being beholden to—under the sway of—the art object. If this either/or represents a fork in the road for the practice of writing aesthetics, the road less travelled is clearly the latter and it is this immersive-productive path to aesthetic understanding that I wish to examine and defend.

A number of nuances in Adorno’s text are missed in Hullot-Kentor’s translation. “Andrängenden” connotes something more specific than “force” (there is a perfectly good German word for force): it means something more like pressure or being crushed, or being pushed around, as if one were caught up in the surging of a crowd. This word also forms a subtle pun with eindrang (penetrate) which concentrates the logic of the two distinct roots to aesthetic knowledge: the dialectical fork is a manifestation of two different energetic dispositions and relationships whereby aesthetic concepts push-into their non-conceptual other (the artwork), and artists are pushed-around by the art object (the other of the subject). This “pushing-around” (if I may) is an everyday, concrete experience for the artist: if an artist embarks upon the revision their work following a reflective judgement along the lines of “that section is not right”, “that’s not working”—or even when they realise their whole approach to their art is no longer working and requires a radical rethink—they are experiencing a kind of objective pressure. It might also be added that while they know (or realise) that something is not quite right with their work, the artist might not immediately know what they can do about it. At any rate, this is one way of rendering Adorno’s oxymoronic “unconscious consciousness in the midst of the object” ([6], p. 334). We will need to return to this troublesome assertion later, but suffice it to say, the key idea is that the objective pressure (even “law-like” imperative) of the work can be (and is) experienced and documented by the artist. In this regard, Adorno mentions the “scattered remarks” of the artist. Again, though, the translation could be more precise: Zeugnissen would be more accurately translated as report or testimony, and it seems to me that there is a subtle rhetorical difference in the epistemological expectation concerning remarks and reports: reports and testimony are supposed to count as—or could be worked up into—knowledge; “remarks” are more unassuming.

Let us now draw these thoughts together into the following thesis:

[A] The self-documentation produced by an artist can be considered a valid contribution to aesthetic discourse if it testifies to the objective pressures experienced in the productive process.

It is a straightforward matter to exemplify Valéry’s familiarity with the demands of the object: “If you persist with a work, you finish by persisting with it in spite of and in contradiction with yourself; and it leads you where you didn’t know you could go, into ideas and decisions which are neither of the self, nor of the non-self-, into what you might have been, and never knew, and avoided until then” ([3], p. 111). Then again: “[A]s the work takes shape, it increasingly comes to dominate the author. It causes him to act, and makes of him a slave. It becomes something with growing demands.” ([3], p. 118). What is discounted is a kind of self-reporting wholly wrapped up in the “intentions” or feeling of the artist and therefore lacking any non-subjective mediating encounter. We can mention Adorno’s cautionary clauses: [A] will be invalidated if this reporting is, “constrained by the naïveté that society insists on finding in art” or is hostile towards aesthetics and reasoning, or not also subject to “philosophical development”.

In the aesthetics that Adorno is attempting to envisage, these stipulations emphasise that an artist’s testimony can be highly valued, but is not sufficient in itself. He quickly wants to re-establish a role for philosophy, in working responsively with the artist’s perspective (which is in principle documentable). The dialectic—roughly aesthetics framed as either concept-based analysis or practice-led testimony—leads swiftly to something like a sublation: the challenge of writing aesthetics “today” is to bring the “artist’s closeness to the phenomena into conjunction (verbinden) with a conceptual capacity free of any subordinating concept (fixen Oberbegriff), free of all decreed judgements” ([6], p. 335). This defines well enough what I attempt in this paper.

The pressures exerted by the object in the process of production counter the complex of heterogenous, art-alien pressures: social pressures, irrationalism, egotism, and (more generally) the dictates of concepts and fixed judgements. Artistic making, if it is “bound” to conceptual articulation, removes aesthetics from the sphere of conventional, prescribed, discursive frameworks, and disempowers the idealistic tendency to conceptual commanding and identity thinking. As Adorno saw it in the late 1960s, modernist art and critical philosophy were joined in a common cause against these pressures and, therefore, the programmatic task of aesthetics (a task which Äesthetische Theorie sets out to realise) appears as the manifestation of the solidarity of artistic making and philosophical writing.

One social pressure that haunts Adorno’s thinking is that the terms and means of “communication” have become prefabricated and are imposed on language users. In universities, he complains that “the writer is urged to show explicitly all the steps that have led him to his conclusion, so enabling every reader to follow the process through and, where possible—in the academic industry—to duplicate it” ([8], p. 80). Honouring this imperative leads to stylistic monocultures, embattled and legalistic frames of mind, all organised around a compulsion to force the “assent” of the reader (an exhaustive analysis gives the reader no option but to agree). However, for Adorno, a uniform and conformist expositional style also works against the value of thinking itself: “For the value of every thought is measured by its distance from the continuity of the familiar. It is objectively devalued as this distance is reduced; the more it approximates to the pre-existing standard, the further its antithetical function is diminished, and only in this, in its manifest relation to its opposite, not in its isolated existence, are the claims of thought founded” ([8], p. 80).

I think it important not to get carried away with Adorno’s idea and conclude that every discourse that attempts to be clear and logical is necessarily crypto-conformist. A logical sequence of statements could lead to remarkable counterintuitive perspectives. But Adorno is right inasmuch as endorsements affirming the “clarity” of a given discourse might indicate nothing more than that discourse behaved in conventional ways and arrived at just the sort of conclusions that the reader prefers or craved. Exhaustiveness and clarity guarantee nothing by themselves; what matters—and this is the only imperative tolerated by Adorno’s principles—is the obligation to discover the “objectively appropriate expression” for any given idea. However, this assertion proves difficult to handle, for while it seems perfectly defensible to wish to honour, in discourse, the specifics of the subject matter (and not falsify the subject matter for the sake of, say, securing the widest possible readership), how is “appropriateness” to be judged? What “objective” criteria do we have in this regard? And what role does “expression” play here? I cannot answer these questions here, but suffice it to say that the way in which Adorno structures the dilemmas of discursive communication, and without some kind of foothold in semantics, phrases like “objectively appropriate expression” can seem somewhat doctrinaire.

For the time being, it can be admitted that the testimony of the artist might yield statements that are energetically distancing the familiar, the clichés and dogmas of conventional opinion: all things being equal, this helps generate their sense of provocation. If dialectical philosophy can accommodate the antithetical function of such thoughts by keeping its contents active through mediation, by foregrounding social and material inferences, then this testimony can also be judged to be a reliable reflection on social states of affairs. Because Adorno used the concept of “antithesis” we take no liberties when we give this contrarian logical form:

[B] Most people generally believe that X, whereas the artist notices that Y.

For example, Valéry writes: “Enthusiasm is not a writerly state of mind. People don’t like one to say this. They should like the engineer who builds a locomotive, to build it while he himself is going at 110 km an hour.” ([3], p. 121). “People” and the “they” (working just as Heidegger’s “das Man”) are the upholders of a set of beliefs that are a simple distortion of the facts of creative activity: writing is not so different from building locomotives, and neither writing nor engineering are best carried up whilst being worked up in a frenzy.

Here is another example of this contrarian way of thinking. In a single sentence aphorism from 1931: “Starting from the same impression, one person fashions a song, the other an ‘analytic’ theory” ([3], p. 134). If this is in any sense a provocation, this is because it runs counter to what one might normally suppose songs and theories to be: songs and analytical theories are generally considered to be of radically different orders, they are not thought to have common origins; and neither are analytical theories thought to be developed from “impressions”. Valéry’s insight (if indeed it is an insight) is carried by the common term: his point is that “fashioning” from an “impression” can take one of at least two directions, a song or an analytical theory. Rhetorically, the activity of fashioning songs is made to appear more “reasonable” by the comparison, and the fashioning of theories appears more “creative” or emotionally invested.

What could become an issue here is the reliability of this insight; one could ask if this is a true reflection of the nature of “fashioning”, of the productive value of “enthusiasm”? There are two points to make. As a provocation, it need not necessarily be true to have an impact on thinking, just as insults don’t need to be true to be impactful upon feelings. If one had already tacitly assumed that the production of theories and the composition of songs were indeed governed by radically different principles—and not, as Valéry implies, merely different inclinations and developmental pathways—then Valéry’s remark could result in the reconsideration of that assumption. The reliability of [B]-type statements follows from [A]-type experiential insights: the contrarian perspective of the artist is largely conditioned by their continual recourse to the work and the self-transforming objective pressures (broadly conceived) that they endure; their explanations trump conventional wisdom by being empirically informed. Because of such a form, one can always attempt to infer logically “bland majority opinion” from “urgent minority conviction” and vice versa. Inferring majority opinion from the experience of the beleaguered and stranded individual is a frequent tactic of Adorno’s: he does not always conduct surveys establishing as a statistical fact what “everyone generally thinks”, the individual’s pre-conceptions—even fears—of the anonymous, discursive edicts of the Other will suffice.

3. Inferentialism

My Adornoesque defence of the aesthetic legitimacy of Valéry’s testimony is underway, but it is far from complete. Valéry testimony regarding the creative process needs some kind of “philosophical development”. My next task is to provide a clear means of providing such a development. I will try and show that the inferentialist understanding of concept use and meaning will suffice in this case. I will set out the main characteristics of inferentialism, provided by Robert Brandom, as succinctly as possible. Inference is contrasted with reference in the following manner:

[C] Inferentialism privileges inference over representation in the order of semantic explanation.

The “representationalist” semantic paradigm takes it that a word (spoken or written) is the “outward representation” of an “inner meaning”: concepts refer us to a content, and when concepts are deployed in discourse, what matters is that content. So when Valéry wrote the word “music” the representationalist assumes he was using the word to convey a meaning or conceptual content. The problems with this paradigm begin when one enquires as to the prior whereabouts and substance of this content. The Lockean view is that words are used as the “sensible marks of ideas; and the ideas they stand for are their proper and immediate signification” ([9], p. 51). Words represent ideas (or beliefs, or intentions); words are the signifiers, ideas are the signifieds, and ideas reside nowhere else but in the mind, though how, and in what “form”, might be thoroughly mysterious. This “mentalist” picture is deeply problematic (see [10], pp. 5–7) not least because it precipitates further complications such as that of translation: I need to find a way of turning ideas into language, of saying what I am thinking. If we find ourselves struggling to “say what we mean”—a worthy enough artistic struggle—that can appear to corroborate the view that there are ideas held in abeyance “in the mind” which need to be brought out and laid “before the view of others” ([9], p. 51). Conception precedes articulation: talking is a way of “going public” with one’s private ideas.

This explanatory strategy can be reversed by pursuing the principle that the business of articulation precedes any understanding of conceptual content; the overt speech of ordinary interpersonal communication is the actual locus of meaning, and anything we may wish to say about the beliefs that are carried about “in our heads” can be derived from actual language use. The pragmatic picture takes “language use as antecedently and independently intelligible, and so is available to provide a model on the basis of which one could then come to understand mental acts and occurrences analogically: taking thinking as a kind of inner saying. Such a view just turns the classical early modern approach on its head” ([10], p. 5).

Conceptual content and meanings are to be derived from what we actually articulate, and inferring meanings from articulations is a skill, a practice. Sound articulation requires an understanding of the set of likely inferences that might arise from what we say. The problem of “saying what we mean” can strategically re-described as the problem of “meaning what we say”: in pragmatic terms, discursive practice entails the interpersonal responsibility to be answerable for what one actually says: defending it, explaining it, and dealing with the implications that flow from what was said. Protesting “I didn’t mean that” is another way of saying “I don’t endorse what you have been able to infer from what I said” and will result in my amending my statements such that unwanted inferences are eliminated. The realisation that my articulation failed dawns on me at the moment it is placed into a communicative context.

In Brandom’s terms, “meaning what we say” is the readiness to commit to what we say, back it up, if necessary, with reasons or action. Saying something cogent means putting our statements into an interpersonal “space of reasons” (or space of actions), giving something to others that they can reason with and from which further sentences can be inferred. What is meant by what we say is contingent on, and subsequent to, such processes of reasonable inferring and conversing. The conceptual content of a perceptual experience will be acquired as we participate in an “inferential game of making claims and giving and asking for reasons” ([10], p. 48). Claims and statements relating to what we see, feel, or think, are but the first move in a language game, moves which prompt countermoves, and so forth. An inferentialist like Brandom will want to emphasise the role of reasoning in this game, but, as I will show below, there are other kinds of games. What I want to insist on here is that this picture of real, contextualised “games” of understanding, also have resonant characteristics in the following sense: these games are about the transfer of responsibility, they generate obligations and actions (like thinking hard and coming up with good reasons and explanations). In a literal sense, thinking and conversing require energy, there are winners and losers; there are flat conversations, and some which are hugely invigorating and productive.

This radical reversal has profound implications for the direction of aesthetic explanation, for the representationalist model of meaning will also provide for the “expression of emotion”. One naïve view of artistic production might be this: strong feelings prompt the need for creative expression. These feelings are then represented by forms and gestures and are thus communicated to an audience who picks up on the emotional content and are correspondingly and sympathetically moved. Representation closes this affective-communicative chain. This is a causal picture: a work is caused and also dignified by its arising from an actual emotional episode. (New emotional experiences precipitate the search for new expressive means.) It is also in this sense realist: human feelings and experiences have an objective reality and worth independent of the technical means which artists use in their struggle to translate into works.

By contrast, an artist could merely experiment by putting colour against colour, harmony against harmony, word against word, and speculate what might be made of the result. In a manner which is precisely analogous to placing statements in a “space of reasons”, works are thus placed in a “space of interpretation” or into “spaces of judgement and evaluation”. Need anything of great significance lay “behind” or precede these activities? An artist could be interested only in what is technical possible, and what the technical consequences of various choices might be. It follows that what an artist needs, alongside their practical ability, is a well mapped and structured sense of these spaces of interpretation and judgement and the know-how of postulating and negotiating a variety of possible interpretations.

Meaning is not deposited in the work at the moment of its creation: meaning is subsequent to production and is itself an activity in its own right, carried out by the artist and the spectator alike; neither has any privileged perspective. In terms almost identical to Barthes’s readerly/writerly distinction, Valéry proposes that a poetic work is organised so as to afford the reader the pleasure of producing meanings:

Few people imagine one can produce a work specially written—not with a view to giving the reader something, but with a view to receiving. To offer the reader the chance of a pleasure—active work—instead of proposing a passive enjoyment. A text fabricated expressly to receive a meaning—and not only one meaning, but as many meanings as the operation of a mind upon a text can elicit from it. But do not believe that is a novelty. It’s no more than doing consciously what is necessarily done unconsciously every time language is used.

([3], p. 138)

So, with regard to the “philosophical development” of Valéry’s testimony relating to his practical experience of writing, what can be established is that he is much closer to an inferentialist understanding of meaning than a representationalist one. This gives us a nuanced perspective on Adorno’s “objective pressure” thesis: for if a textual work (even just speaking) is always liable to receive meanings—and multiple, unanticipated meanings at that—this can amount to panic in the artist, for the artist will be liable or answerable for what the work “means” and for its value (whether it is, simply, any good or not) and this entails anticipating how the work might be received or understood or evaluated by another. But there is no way of knowing in advance what the evaluative consequences of ones work might be. The overflowing waste paper basket of the writer is an index of this anxiety. But by concerning oneself only with the object itself, an attitude of “I did what I had to do, take it or leave it” exempts the artist from the burden of worrying about these excessive meanings.

Valéry hints at this when he writes: “What I know weighs upon what I can do—(in general) whereas in music or mathematics what I know is identical to what I can do” ([3], p. 141). Knowing “weighs upon” doing at the point where what might be done is caught up with speculation about what it might mean. The highly knowledgeable have a greater capacity to anticipate interpretations, a fact which might therefore inhibit their own action. Valéry contrasts this experience with music and mathematics and the implication here is that these disciplines are set up such that their materials coincide exactly with how they are to be used: everything about music and mathematics is functional. With music and mathematics there is no “weight” placed upon action by knowledge, action is knowledge. In Valéry’s estimation, music and mathematics are, in a sense, rich in necessity.

The lawfulness of artworks was certainly close to Adorno’s heart and in this regard we tend to hear the musician speaking: “in actuality the process of artistic production, and with it the unfolding of the truth contained in the work of art, has the strict form of a lawfulness wrested from the subject matter itself”. Here Adorno makes Valéry agree with Schoenberg, and the “idea that great music consists of fulfilling the obligations the composer incurs with virtually the first note” ([3], p. 104–5). If the notion that the artwork imposes a series of demands upon the artist sounds overly deterministic, there is, though, also a note of opportunism in the creative process. Valéry writes:

Regular work at verse, at that tinkering which leads to perfection, accustoms one to word changes, suppressions and substitutions that, thanks to their not infrequent success, alters the writer’s point of view, and gives him legitimate reasons for thinking that the initial purpose, the original design of his poem is not essential; that one can and must abandon it, if a chance to stray presents itself, -that it’s no more than a starting point—an initial arrangement; and the author comes to consider the modifying act of language, and language itself, as the central object of this attention.

([3], p. 118–19).

Here is another variant on the idea that the process of making annuls the initial intentions of the artist, and this annulment occurs because the working affords the “chance to stray”. What is more, these are opportunities that must be taken (they have a law-like compulsion), and these deviations lead the poet back to language and its objectivity.

4. Conceptual Content in Poetry

In summary, the semantic paradigm of inferentialism allows us to get purchase on Valéry’s assertions that the locus of the meaning of a poet’s poems is the working process, the chance diversions, and the multiplication of interpretations. We can easily dispense with notions such as authorial intention because firstly, any intentions there might have been are continually dissolved by the working process, and secondly, works receive their meanings from their interpreters (a community which includes the author). Problems emerge with this way of thinking when we attempt to be more specific about the conceptual content of poetry. To recap, the inferentialist pursues the principles that for a remark “to have conceptual content is just for it to play a role in the inferential game of making claims and giving and asking for reasons. To grasp or understand such a concept is to have practical mastery over the inferences it is involved in—to know, in the practical sense of being able to distinguish (a kind of know-how), what follows from the application of a concept, and what it follows from.” (Articulating Reasons 48). The problem is, does poetry “play a role in the inferential game of making claims and giving and asking for reasons?”

To illustrate, take the familiar syllogism:

• All men are mortal,
• Socrates is a man,
• ∴ Socrates is mortal

The conclusion follows from—can be inferred from—the two premises. These kinds of assertions can give us a foothold in the “game” of giving and asking for reasons. Asking “how do you know Socrates is mortal?” yields the response “because he is a man, and all men must die”. The concept of “mortality” is grasped when we know how to use it in sentences like these, and when we can use other formulations to mean the same thing: Socrates is mortal ≡ Socrates will die one day ≡ Socrates will not live forever ≡ Socrates will have a funeral. Should the gendered language cause confusion, i.e., “Sofia is a woman, is she immortal?”, as good inferentialists we can straighten things out: “sorry, I meant ‘man’ in the sense of Mankind, not in the sense of a ‘male human’. Perhaps I should have said ‘all humans are mortal’, or ‘all people are mortal’”. Clarification of this sort demonstrates vividly the skills involved in “staying in the game” of making good, inferential sense.

In another example, Frege, in the Begriffsschrift of 1879, presents two propositions: “the Greeks defeated the Persians at Plataea” and “the Persians were defeated by the Greeks at Plataea”, which, though having a different word order, clearly mean the same: the conceptual content is defined as that which does not differ between these two statements ([10], p. 50). According to the representationalist paradigm we can say that these nearly identical sentences represent the same idea. But in inferentialist terms, these two sentences have the same conceptual content “if and only if they have the same inferential role” ([10], p. 50). Anything I can infer from the first sentence (i.e., a sentence like “the Greeks won the battle of Plataea”) can also be inferred from the second. Further sentences can be invented which will play the same role, i.e., “At Plataea, when the Persians fought the Greeks, the Persians were on the losing side.” I can substitute the premise “All men are mortal” with “No human beings live forever” and all the inferences, and therefore all the conceptual contents, remain unchanged. All these examples illustrate the principle of salva veritate.

The problem arises when these inferentialist schemes for understanding are applied to a poetic text, and the conceptual content of the poem is made explicit by producing inferences or summarising: according to Valéry, this will be at the cost of losing the poem, of “annulling” it. The inferential understanding of poetry would always endanger the specific configuration of the poem: no phrase in a poem could be substituted with another, and no new statement derived from any of those in the poem, without losing touch with the poem itself. Valéry’s depicts the struggle of the poet as one of preventing the annulment of the poem by the reader’s cognitive bias:

Poetry has no plan—Music may have one—where there’s nothing but Music—But when you are using words—which tend constantly to slip through poetry’s fingers—you really have to calculate at every instant how you can stay within poetry, prevent the reader’s mind from moving directly to the abstract—to the summary which annuls the resonance and the form.

([3], p. 223)

Importantly, the representationalist has no such problem because the poem could be taken to be merely a way of communicating the intentions, experiences or feelings that preceded the expression. If we assert something that we believe lays “behind” the poem, by virtue of the genetic story the representationalist presupposes (giving form to an originating idea) we can lay claim to something that is genuinely at the heart of the poem: the poem was written as a vehicle for its idea; to lay claim to that idea is to grasp something essential about the poem itself even if the details of the poem are set side. I do not think that Valéry’s poietics can accommodate such a view, however. For him “abstracting” from a poem, making a summary of its conceptual content, will necessarily ruin the details of sound and form that the poet has carefully put in place. These details matter because they contribute to resonance.

We are at the crux of the argument. Maintaining Valéry’s line of thinking seems to produce a dilemma between giving attention to either the sound or sense of a poem, very much like forcing a choice between Kantian blind intuitions or empty conceptuality. But Valéry insists that poetic states sustain a way of meaning and being that is “neither resolvable into a finite notion of an object or an act—i.e., annulled and replaced by an idea or an action—nor rejected as unassimilable; but they will produce a state distanced from the ordinary run of things, notable for its preservation of an unstable condition” ([3], p. 233). Resonance results from this instability, from the refusal to decide between modes of attention and comprehension, but instead to move repeatedly from one to the other, transferring energy and building up intensity in the process, just like resonance in physical systems (a phenomenon that musical instruments exploit). Language, or rather a “complete language” which enjoys all its resources—tone, rhythm, voice, tempo, meaning, etc., is reimagined as a vast set of resonators which the poem composes, fits together so as to produce reciprocal—singing—interactions. Something of this can be gleaned from frantically written entries such as this:

Poetry is—the singing state—(echoing—resounding—rebounding) of the function that speaks and consequently, of everything that men have introduced into that function = Everything becomes. The system of words in a state of vibration, in which state they echo each other—change in value—The vibrating singing state—as if the internal dictionary, the table of potential signs were...tensed and the links between words had changed their mutual tension. Sound and sense in an interchange that has become reciprocal, of equal value-[…] Reception—production become symmetrical.

([3], p. 211)

Let us consider an example. Take the first stanza of Valéry’s Le cimetière marin from 1922:

• Ce toit tranquille, où marchent des colombes,
• Entre les pins palpite, entre les tombes;
• Midi le juste y compose de feux
• La mer, la mer, toujours recommencée
• O récompense après une pensée
• Qu’un long regard sur le calme des dieux!

There is a tautness in the composition that is built around the sound: e.g., pins palpite, and the harmonious, generative chain which unfolds in lines 4 and 5, recommencée/récompense/pensée, whereby each word is almost physically present in the preceding one. There is a pleasing cadence in the first three words, Ce toit tranquille, a subtle focussing and expansion in the vowel sound that counterpoints a hint of alliteration on the consonant “t”. And these sounds communicate with sense: tranquille enacts a calming; pins palpite, already beats in close alliteration; recommencée/récompense/pensée is a rather close, studied, and self-conscious poetic ruse which opens out onto the “calme des dieux”; and the repetition of la mer in line four (appearing immediately after feux at the end of the previous line) is reinforced by toujours recommencee. Some phrases do not give themselves over to meaning that readily. Ce toit tranquille is rather stubborn in that regard: when are roofs anything other than quiet? Yet I can imagine the picture Valéry depicts by the phrase.

It is axiomatic that none of these phrases can be changed without losing the poetry. All that matters with regard to these observations is that they result in my repeatedly raiding the structure for sound and sense, in my “staying within poetry”. They provide access to the enjoyment of the poem’s resources, but in a peculiar manner: what I set out above is the result of hours of ruminating, talking through the poem, checking words in a dictionary, attempting English translations, identifying metric patterns; I even started wondering how the poem might be set to music. The point is that this multi-dimensional involvement with the poem is on-going and proliferative: one discovery leads to further intrigues and experiments (some of which go slack), and thus I continually renew my acquaintance with the poem. The goal of reading becomes one of total involvement, of wanting to know the poem by heart. This can be given the precise sense of having a living relationship with the poem, of having access to the poem at different points in my life, and in different situations: if I am ever looking over the Mediterranean sea, Valéry’s poem will let me sing what I apprehend.

In terms of understanding how poetry is written—this is the abiding concern of poietics—the complexity of this system, and the instability of the reciprocal interaction between the different co-resonators, is such that it is impossible for the poet to simply posit meaning or significance at a stroke. Every positing is answered by multiple responses: consequently the poem can do nothing but incrementally and experimentally grow into this “harmonic realm of being—i.e., in the realm of periodic instability where stimuli and responses constantly renew each other” ([3], p. 233). Reading itself is also a “harmonic realm of being”. Happily for my account, in another entry, Valéry mentions formal inferences: A poem is a process of generation -i.e., its mode of growth is characteristic and distinguishes it from other genres—novel etc., (And that can also serve to justify rhyme which is a naïve realisation, -and above all a constant reminder to the author, as well as to the reader, of the law of successive growth, or of creation of time, which by means of impulses and inferences from the form, composes a state of resonance and the desired sensation of aesthetic infinity. That’s what the linked-linking movement of poetic language imitates.).

([3], p. 228)

We can thus work back to the inferentialist picture: for the reciprocity of impulse and response, the “symmetry” of production and reception, the “linked-linking” action of language, are compatible with the idea of an inferential language-game after the pattern of a logical “if-then”: every poetic linguistic positing (an “if”) is a move in a poetic language-game that will induce a response (a “then”) from one or other part of the linguistic system. Rhyme, assonance, dissonance, alliteration, metre, and all the other technicalities of poetic composition constitute systems of potential reciprocal exchange. If the different components link together successfully, then some kind of resonance in the qualities or semantics of the poem is always possible. It is the task of the poet to enter into these multiple games, repeatedly testing the responses and consequences of every move.

5. Conclusions

The Valérean poietic model helps us understand the creative process by transferring all our attention to the concept of process, displacing the much more problematic concept of “creation”. A process has a goal: the goal that Valéry was interested in was discovering the resonance of language and thus a poetic singing-state. This state attains the cooperation of the poet and the vast resources of language in a reciprocal intensification. Language is not and cannot be the mere vehicle of expressive intentions; language is not subservient to the poet: language belongs to interpersonal experience (this is also enough to invoke the historical aspect of language), and meaningfulness is subsequent to articulation, the result of an activity of understanding in which the poet, the reader, the critic, and the scholar all have a contribution to make, and in which none have a monopoly. We can reaffirm that this model counts as an aesthetic insight in Adorno’s sense: it acknowledges the objectivity and materiality of the creative process and restricts the subjectivist prejudices that are dominant in popular aesthetic understanding. This model has also be shown to be philosophically respectable.

The model can help solve some conceptual difficulties latent in the popular understanding of creativity. The importance of what might be called “inspiration” can be understood once we notice that the model requires an initial input which is sufficiently potent and linkable to produce lively ramifications and consequences. A brief phrase is enough to give the poet a sense that, perhaps suddenly, the task of writing has become possible. There needs to be, in a sense, enough poetry in the phrase (it will have that sense of being a “complete language”) to precipitate the search for more material and deeper connections. A simple formalism can be described. Consider Frege’s sign denoting assertion “├”, where├ x can be read as an assertion of the truth of x. What this symbol captures is that the assertion of the truth of a statement has consequences: truth gives us something to work with, infer from, and understand. Not all strings of symbols have to be asserted in this manner; symbols can be concatenated together in a vast variety of ways without thereby asserting any truth, i.e., “5 + 7 = 11”. Such a statement is taken as being “incorrect”—untrue—but such a judgement would only be warranted if its truth was being asserted. Likewise, if a linguistic expression is asserted as poetry (rather than, say, information, or small talk) this obliges us to work with it in a certain way: in Valéry’s terms, this means experiencing the expression’s resonance, its songfulness. A phrase φ that is asserted in such a manner can be denoted as:├ poetry φ.

The predicate “inspired” might be apt because φ provides access to those extensive resources that are latent in language: there is, as it were, an experience of the dramatic influx of possibility from something “outside” of oneself. Incidentally, exactly the same can be said of the incipient musical “ideas” that provide the impetus in symphonic composition. Not every concatenation of pitch classes and rhythmic values amounts to music; one could be working on short exercises in a harmony textbook in order to learn how to apply certain music-theoretic concepts (i.e., the preparation of dissonance). The materials of music might, though, be worked up into a singing-state, in which case the same principles apply: one is obliged to consider embodiment in musical production, the formal inferences, the musical associations, and what a listener might make of it all. Asserting a string of notes as music (├ music ) precipitates the reflective development of these and other dimensions.

A process must begin somewhere, but how is it to be terminated? Valérean poiesis cannot identify entirely with the artistic goal of formal perfection in the sense of producing something complete, or fully formed, because the singing-state is a condition of relationship, and of excess. Resonance is like a Deleuzean plateau, a state which is attained, sustained, and then dissipates. (Famously, as Valéry put it, the poem is not finished, it is abandoned.) The “process of generation” in poetry is iterative: posit, judge, revise, where judgements are prompted by the resonant response characteristics of the emerging linguistic complex. (The capacity to set up and read these response characteristics and produce a co-response is the poet’s technique.) So at any point in a generative process, the material that the poet is exploring has a certain state of formal connectedness (s). Beginning at some initial state (s0), judgement and revision bring other possibilities into consciousness and successively multiply the connections with other material:

• ├ poetry φ: s0 → (s0) s1 → ((s0) s1) s2…

At each stage of development, something is retained of earlier stages. As described here, such a process either becomes intractable, crosses some threshold of acceptability, or is arbitrarily broken off; it is not of itself terminable. Because the development of the poem is also the development of the poet’s technique, her capacity to reason about the connectivity and implications of poetic phrases, new possibilities can always come into view. Writing is not the repetition of a task previously performed, but an opening out onto further possibilities. The openness of resonance towards its own future is a thought that edges its way into Le cimetière marin: “J’attends l’écho de ma grandeur interne, /Amère, sombre, et sonore citerne, /Sonnant dans l’âme un creux toujours futur!”

In summary then, the four theses I set out at the beginning have been considered. Valéry’s Notebooks, as a remarkably sustained exercise in negative capability, provide provocative encounters between conventional opinion and the insights that arise during creative processes. One can argue that Valéry provided a credible theory of his own creative process, and its goal of producing resonance and attaining a singing-state (there are, of course, other goals), a theory which, however sketchy it might appear, is consistent and can account for various related phenomena such as “inspiration” and revision. The characteristics of Valéry’s theorising does also suggest that there is a role for inferentialist semantics in aesthetics, and hence a certain rationalist development that he would probably have appreciated. What I have been calling a “practice-led aesthetics” amounts to the transposition of the context of aesthetic and semantic questioning from the armchairs of philosophers to the workshops of artists. This transposition boosts the claims of aesthetic anti-essentialism: any aesthetic concept—including “poetry” or “music”—has meaning only to the extent that it has material and practical implications for the artist. In this practice-led context, judgements and claims referring to the “music of poetry” require the technical description of what is to be done by the poet to produce a text which might be entitled to receive such a predication. It is this kind of description that I have begun here.