On Literature

THE POETICS AND US

Allow me, as an Italian, to approach the question of Aristotle's Poetics in the form of a confession by a child of this century. Italian culture produced the great Renaissance commentators on Aristotle, and in the baroque period it was Emanuele Tesauro who, with his Cannocchiale aristotelico, represented to the world of post-Galilean physics Aristotle's poetic theories as the sole key to approaching the human sciences. But at the very beginning of the following century this same Italian culture was enriched by Vico's New Science, the work that questioned every Aristotelian precept in order to posit a language and poetry that develop outside any rules. By doing so, Vico unwittingly opened the door—while in France, from Boileau to Batteux, from Le Bossu to Dubos, and right down to the Encyclopédie, writers were still looking to set down the rules of tragedy with the rules of taste—to a philosophy, a linguistics, and an aesthetics of the unpredictable freedom of the Spirit.

This is not the genteel, classical French Esprit but romantic, Hegelian Geist, which emerges through history only to become history itself. Thus, from nineteenth-century Idealism to Croce, Italian culture was dominated over the course of a hundred years by the rejection of all rhetoric and all poetics. Under an Idealist aesthetics that read all language as founded from the outset on aesthetic creativity, the phenomenon of poetry could be described no longer as the deviation from a preexisting norm but rather as a new dawn. The few pages devoted by Croce to Aristotle show irredeemable prejudices, which resulted in a formally impeccable syllogism: aesthetics began with Baumgarten's idea of a " scientia cognitionis sensitivae, gnoseologia inferior" (a science of sensory cognition, a lower gnoseology); Aristotle was unable to read Baumgarten, and thus Aristotle had nothing to say on aesthetics.

I remember the shivers I experienced as a young man, feeling as marginalized as a young homosexual in Victorian society, when I discovered that the Anglo-Saxon tradition had continued to take Aristotle's poetics seriously, and without interruption.

I was not amazed to find traces of Aristotle in Dryden or Hobbes, in Reynolds or Dr. Johnson, not to mention the references to the Poetics, however vague and even at times polemical, that I found in Wordsworth or Coleridge; but I was struck by the readings of poets and critics who were contemporaries of Croce but who gave me the outline of a culture for which Aristotle was still a model, a point of reference.

One of the classics of American critical theory, Richards's Principles of Literary Criticism (1924) opens with a reference to Aristotle; if Wellek and Warren's Theory of Literature (1942) managed to blend the principles of Anglo-Saxon criticism with the work of the Russian formalists and of the structuralists in Prague, it was because they referred to Aristotle in almost every chapter. In the 1940s the masters of the New Criticism measured themselves against Aristotle. I discovered the Chicago school, which defined itself unreservedly as neo-Aristotelian, a critic of contemporary theater like Francis Fergusson (The Idea of a Theater, 1949), who used the notions of plot and action and interpreted Macbeth as an imitation of an action, and Northrop Frye, who played with the Aristotelian notion of mythos in his Anatomy of Criticism (1957).

But we need only cite the influence of the Poetics on a writer like Joyce. He not only speaks about it in the 1903 Paris Notebook, written during his visits to the Bibliothèque Sainte Geneviève, but in 1904 writes a short ironic poem on catharsis. He tells Stuart Gilbert that the Aeolus episode in Ulysses is based on Aristotle's Rhetoric. In a letter to his brother Stanislaus of 9 March 1903, he criticizes Synge for not being Aristotelian enough for his tastes. In a letter to Pound of 9 April 1917, he says of Ulysses: "I am doing it, as Aristotle would say—by different means in different parts." Lastly, the theory of literary genres in the Portrait is of clear Aristotelian origin. In that work Stephen Dedalus works out a definition of pity and terror, deploring the fact that Aristotle had not provided one in the Poetics and ignoring the fact that he had, however, done so in the Rhetoric. Through a kind of miraculous elective affinity, the definitions Joyce invents are very similar to those in the Rhetoric—but he studied with the Jesuits, and along with a secondhand version of Saint Thomas he must also have come across a thirdhand account of Aristotle. To say nothing of the English-speaking cultural environment in which he lived, about whose Aristotelian interests we have already spoken.

I believe, however, that I underwent my most decisive Aristotelian experience reading Edgar Allan Poe's Philosophy of Composition, where he analyzes, word by word, structure by structure, the birth, technique, and raison d'être of his poem "The Raven." In this text Aristotle is never named, but his model is ever present, even in the use of some key terms.

Poe's project consisted in showing how the effect of "an intense and pure elevation of soul" (Beauty) is achieved by careful organization of structures, and in showing how "the work proceeded step by step, to its completion with the precision and rigid consequence of a mathematical problem," while still keeping track of a unity of impression (which is in physical terms the unit of time that corresponds to one reading of the text), and of place, and of emotional tone.

The extraordinary thing about this text is that its author explains the rule whereby he managed to convey the impression of spontaneity, and this message, which goes against any aesthetics of ineffability, is the same as that transmitted by the Poetics. This Aristotelian lesson is also found later in Pseudo-Longinus's On the Sublime, which is usually seen as a celebration of the aesthetic "je ne sais quoi." On the Sublime certainly wants to tell us about a poetic effect that is based not on rational or moral persuasion but, rather, on a feeling of wonder that is produced as a kind of ecstasy and coup de foudre. But right from the first page of the treatise, the anonymous author tells us he does not want just to define the object of his discourse but also to tell us the strokes of artifice by which it can be produced. Hence, in the second part of the work, we find a minute analysis of the rhetorical strategies to be adopted in order to achieve, through definable procedures, this indefinable effect.

Poe proceeds in the same way, except that The Philosophy of Composition is a fascinating and ambiguous text: is it a set of rules for other poets, or is it an implicit theory of the art in general, extrapolated from a personal experience of writing, by a writer who positions himself as a critical reader of his own work?

The fertile ambiguity of this text was noticed by Kenneth Burke, who approached Poe's text in explicitly Aristotelian terms. If there is a discipline called Poetics, it will have nothing to do with a criticism seen as commercial advice to the reader, or a distribution of approval or disapproval. It will have to deal with one of the dimensions of language, and in that sense it will be the proper object of the critic, just as poetry is the object of the poet. "An approach to the poem in terms of Poetics is an approach in terms of the poems nature as a kind (a literary species or mode)." ^* In this sense Burkes definition comes close to that given by the Prague school, for whom poetics is the discipline that explains the "literariness" of literature, which is to say that it explains why a literary work can be defined as such.

Burke is well aware that defining literary processes and the rules of genre can lead, as has happened, to the transformation of a descriptive into a normative science. Nevertheless, poetics cannot escape its duty to formulate rules that are implicit in the poets practice, even if the artist is not conscious of them.

Poe, on the other hand, was conscious of them and consequently worked as a "philosophus additus artifici" (philosopher as well as an artist). Perhaps he did so "après-coup,"and while he wrote he may have been unaware of what he was doing, but as a reader of his own work he later understood why "The Raven" produces the effect it does and why we say it is a good poem. The analysis carried out by Poe the author could have been done by a reader like Jakobson. Thus, while trying to define a compositional practice of which his poem was an example, Poe identified strategies characterizing artistic procedure in general.

Poe's essay is Aristotelian in the principles that inspired it, in its aims, results, and ambiguities. Lubomir Doležel wondered whether Aristotle's Poetics is a work of criticism (i.e., one that aims at the evaluation of the work it discusses) or of poetics, which aims, as we saw, at defining the conditions of literariness.^† Doležel, quoting Frye, reminds us that the Poetics highlights an intelligible structure of knowledge that is neither poetry itself nor the experience of poetry (harking back to some distinctions that appeared in the Metaphysics) and considers it as a productive science, which aims at knowledge in order to create objects.

In this sense Poetics does not interpret individual works, to which it turns only as a repertoire of examples. But in pursuing this aim Poetics becomes entangled in a paradox; in trying to capture the essence of poetry it misses its most essential feature, namely, its uniqueness and the variability of its manifestations.

Doležel thus observes that Aristotle's Poetics is at one and the same time the founding text both of literary theory and of literary criticism in the West, and this comes about precisely because of its inherent contradiction. It establishes a metalanguage of criticism, and allows judgments founded on the knowledge this metalanguage supplies. But this result is achieved at a certain price. Every Poetics that proposes ideal structures, and chooses to ignore the particularities specific to individual works, is always in the end a theory of the works that the theorist judges to be best. Thus even Aristotle's Poetics has (allow me to paraphrase Popper) its own "influencing aesthetics," and Aristotle betrays his own critical preferences every time he chooses an example.

According to Gerald Frank Else, only a tenth of all Greek tragedies meet the structures posited by Aristotle.^* In a vicious circle, an intuitive critical judgment has preceded and determined the choice of the corpus it will use to work out the general principles that justify the judgment in critical terms. Doležel points out that Else's statement is also based on a critical prejudice, but his argument holds in any case, since it highlights the presence of the vicious circle that has flawed the entire history of poetics and criticism.

We thus find ourselves facing not an opposition (as was long thought) between a normative poetics and an aesthetics that operates on such a level of generality as never to be compromised by the reality of particular works (Aquinas's "Beauty is the splendor of the Transcendentals brought together" is an aesthetic definition that allows us to justify both Oedipus Rex and a good adventure story) but, rather, the oscillation between a descriptive theory and a critical practice that presuppose each other in turn.

Aristotle speaks to us not only of abstract criteria such as order and measure, verisimilitude or necessity, or organic balance (Poetics 1450b 21 ff.) but also of that criterion which will negate every purely formalistic reading of the Poetics. The fundamental element in tragedy is plot, and plot is the imitation of an action whose aim, the telos, is the effect that it produces, the ergon. And this ergon is catharsis. A tragedy will be beautiful—or will work well—if it is able to effect purification from passions. Thus the cathartic effect is a kind of coronation of a tragic work, and this resides in the tragedy not as a written or acted discourse but as a discourse that is received.

The Poetics represents the first appearance of an aesthetics of reception, but it presents some unresolved problems of every reader-oriented theory.

We know that catharsis can be interpreted in two ways, and both interpretations are upheld by that enigmatic expression appearing at 1449b 27–28: tragedy accomplishes " ten ton toiouton pathematon catharsin" (the catharsis of such passions).

The first interpretation is that Aristotle is thinking of a purification that releases us through the intense experience of our own passions—as would seem to be suggested by the Politics (which, however, unfortunately refers to the Poetics for an explanation that is never given in either of the two works), and therefore the purification must be understood in traditional medical terms, as a homeopathic action, a liberation of the spectator through the identification with the characters' passions—and it is imposed on us as an experience we cannot avoid. Tragedy is, in this view, a corybantic, psychagogic machine (if some detachment were possible, this would be produced solely by comedy, but we know too little about what Aristotle meant by comedy).

The second interpretation understands catharsis in an allopathic sense, as a purification undergone by the passions themselves, inasmuch as they are "beautifully" represented and seen from afar as the passions of others, through the cold gaze of a spectator who becomes a pure, disembodied eye—and who enjoys not the passions he experiences but the text that puts them onstage.

Radicalizing this conflict of interpretations, we might say that in one interpretation it leads to a Dionysiac aesthetic, in the other to an Apollonian aesthetic. And, in more banal terms, on one side we have an aesthetics of the discotheque and pulp fiction (and I will talk later about how to read the Poetics as a theory of mass-media emotions), and on the other an aesthetics seen as a moment of serene and detached contemplation, where art shows us the splendor of truth.

This ambiguity is due to the very sources Aristotle was drawing on. The Pythagoreans "had appropriate chants for the passions of the soul, some for weaknesses and others for angers, through which, by exciting and raising the passions in just measure, they would be returned to a courageous virtue" (Iamblichus, Life of Pythagoras). And Pythagoras used poetic texts such as Homer, dithyrambs, threnodies, and laments for cathartic purposes. It is reasonable to assume that Aristotle meant a purification that comes about through an act of free vision of the miraculous organization of the great phenomenon that is tragedy, and at the same time he was fascinated by the psychagogic powers his own culture spoke about.

There are other fertile ambiguities in the Poetics. Aristotle is an Alexandrian who has partly lost the religious spirit that characterized the fifth century B.C. He works a little bit like a contemporary Western ethnologist trying to track down universal constants in the tales of savages, which fascinate him but which he does not understand except from the outside. So then we come up with another, very modern reading of Aristotle, one that Aristotle himself encourages, pretending to talk about tragedy whereas in reality he is providing us with a semiotics of narrativity. The tragic spectacle includes story, characters, diction, thought, spectacle, and music, but "the most important of these elements is the composition of the actions ... For the end of tragedy is the story and the facts" (1450a 15–23).

I agree with Ricoeur when he says that in the Poetics narration founded on plot, this ability to compose a story, he ton pragmaton systasis, becomes a kind of common genus of which the epic is a species.^* The genre that the Poetics discusses is the representation of an action (pragma) through a plot (mythos), and epic diegesis and dramatic mimesis are only species of this genre.

Now the theory of plot is what has perhaps most profoundly influenced our century. The first theory of narrativity emerges with the Russian formalists, who propose on the one hand the distinction between fabula and sjuǽet, and on the other hand the deconstruction of the fabula into a series of narrative motifs and functions. It is hard to find direct references to Aristotle in the works of Šklovsky, Veselovskij or Propp, but in the first study of the Russian formalists, by Victor Erlich (Russian Formalism, 1954)^† the debt of the Formalists to the Aristotelian tradition was clearly shown—even though Erlich pointed out correctly that the Formalist notions of fabula and sjuǽet are not strictly coterminous with those of pragma and mythos. Similarly, one could say that Aristotle's narrative functions are less numerous than those of Propp. But the principle is the same, without a doubt, and this was noticed by the first structuralist critics at the beginning of the 1960s (but it would be unfair not to mention the "Dramatic Situations" of Polti and Souriau, vague descendants of Gozzi—and therefore of an eighteenth-century Italian who had not forgotten Aristotle).

"The narratives of the world are numberless," wrote Roland Barthes in his "Introduction à l'analyse structurale des récits."^* "It is thus legitimate that, far from abandoning any idea of dealing with narrative on the grounds of its universality, there should have been (from Aristotle on) a periodic interest in narrative form and it is normal that the newly developing structuralism should make this form one of its first concerns." In the same issue of Communications in which Barthes's essay was first published (vol. 8, 1966), Genette's contribution, "Frontières du récit," was based on a reading of Aristotle, as was the first articulation of Bremond's semiology of the story, which could be seen as a meticulous systematization of the formal structures suggested by Aristotle. (Curiously, Todorov, who would show in his other works that he knew Aristotle very well, would base his Grammaire du Dé- cameron on a purely grammatical basis.)

I am not saying that a theory of plot and narrativity emerges only in our century.^†" But it is curious that contemporary culture returned to this "strong" aspect of the Poetics at the very time when, as many see it, the form of the novel was entering a period of crisis.

However, to tell and listen to stories is a biological function. One cannot easily escape the fascination of plots in their raw state. If Joyce avoids the rules of Attic tragedy, he does not escape the Aristotelian idea of narration. He may put it into crisis, but he recognizes it. The nonadventures of Leopold and Molly Bloom are comprehensible to us because they are patterned on the backdrop of our memories of the adventures of Tom Jones or Télémaque. Even the Nouveau Romans refusal to make us feel pity or terror becomes exciting against the background of our conviction that a story should arouse these passions in us. Biology strikes back: when literature refused to give us plots, we went to look for them in films or newspaper reports.

There is, then, another reason why our age has been fascinated by the theory of plot. The fact is that we have convinced ourselves that the model of the pair fabula/narrative discourse, pragma and mythos, serves not only to explain the literary genre that in English is called fiction. Every discourse has a deep structure that is narrative or can be developed in narrative terms. I could cite Greimas's analysis of Dumézil's introduction to his Naissance d'archanges,^* where the scientific text manifests a polemical structure that emerges in the shape of academic coups-de-scène, struggles against opponents, victories and defeats. In The Role of the Reader (Bloomington: Indiana University Press, 1981) I tried to show how one can find a story even beneath the (apparently plotless) text that opens Spinoza's Ethics: "Per causam sui intelligo id cujus essentia involvit existentiam; sive id cujus natura non potest concipi nisi existens" (By cause of itself, I understand that whose essence involves existence, or that whose nature cannot be conceived except as existing).

There are at least two inset fabulae here. One concerns a grammatically implicit agent (ego) who carries out the action of understanding or signifying and by so doing moves from confusion to a clearer knowledge of God. Let's remember that if "intelligo" interpreted as "I understand" or "I recognize," then God remains an object that is not modified by the action, but if one translates that verb as "I want to signify" or "I want to say," then the agent sets up, through the act of self-definition, his object of discourse (he makes it exist as cultural object).

However, this object, with its own attributes, is the subject of another story. It is a subject that performs an action through which, through the fact of being, it exists. It seems that nothing happens in this adventure about the nature of God, since there is no temporal interval between the actualization of his being and the actualization of his existence (in fact, neither of these ever moves from some preceding potential to the act, because they have always been there), nor does his existence, when it comes, change his essence. Certainly this is an extreme case, where both action and the passing of time are at zero degrees (which equals infinity), and God acts always in his self-manifestation, uninterruptedly and forever producing the fact that he exists through the simple fact that he is. It is not much for an adventure story, but it is enough for the essential conditions of a fabula to emerge. There is no coup-de-scène, perhaps, but this depends on the reader's sensitivity. The Model Reader of a story like this is a mystic or a metaphysics expert, a textual cooperator who is able to feel the most intense emotions at this nonstory whose exceptional nature never ceases to strike him like a thunderbolt. Even the Amor Dei Intellectualis is a burning passion, and one feels a stupefied and continual sense of surprise in recognizing the presence of Necessity.

If, then, our age discovers that every philosophical or scientific discourse can also be read as a narrative, this may be happening now because, more than in any other epoch, science and philosophy (perhaps even to deal with the crisis of the novel) present themselves (it has been said) as grand narratives. This does not mean—as happens for some people—that by being narratives they do not have to be judged in terms of their truth. They simply want to express some truth by using a structure that is attractive in a narrative sense. And if then the great philosophical narratives do not seem sufficient, we have seen that instead of going to look for truth in the philosophers of the past, much contemporary philosophy has gone to look for it in Proust or Kafka, Joyce or Mann. Thus it is not so much that philosophers have given up pursuing the truth as that art and literature have also taken on that function. But these are marginal observations, and Aristotle does not come into them.

The Poetics has many faces. One cannot have a fertile book without it also producing contradictory results. Among my first discoveries of the modernity of Aristotle, I remember a book by Mortimer Adler, who had worked out an aesthetics of film based on Aristotelian principles. In his Art and Prudence he gave the following definition: "A film is a representation of a completed action, of a certain length, using a combination of images, sound effects, music and other things."^* Perhaps this definition was a bit scholastic (Adler was a Thomist who also inspired Marshall McLuhan), but the idea that although the Poetics was not capable of defining "high" literature, it was still of use as a perfect theory of popular literature and art was upheld by other authors as well.^†"

I do not accept the idea that the Poetics cannot define "high" art, but it is certainly the case that, with its insistence on the laws of plot, it is particularly suited to describing the strategies of the mass media. The Poetics is certainly the theory of, among others things, the John Ford-style Western—and not because Aristotle was a prophet but because whoever wants to put on stage or screen an action using plot (which is what a Western does without any clutter) cannot do other than follow what Aristotle had perceived. If telling stories is a biological function, Aristotle had already understood all that was needed from this biology of narrativity.

The mass media are not alien to our biological tendencies; on the contrary, the media could be accused of being human, all too human. The problem, if it exists, lies in the question whether the pity and terror they provoke genuinely lead to a catharsis; but if one understands catharsis in its homeopathically minimalist sense (have a good cry and you'll feel better), they are, in this minimalist state, applied Poetics.

One could even say that if we stick to Aristotelian ideas for the construction of a mythos that will produce an effective ergon, we can only fall into mass-media syndrome. Coming back to Poe, if we read only the pages that he devotes to the production of emotions which he saw as his aim, we would think we were dealing with a screenwriter for Dallas. Since he wanted to write a poem that would produce an impression of melancholy ("since Melancholy is the most legitimate of all the poetic tones") in a little over one hundred lines, he wondered which was the most depressing of all melancholic subjects, and concluded that it was death, and that the most melancholy death was that of a beautiful woman, "unquestionably the most poetical topic in the world."

If Poe had stuck only to these principles, he would have written Love Story. Luckily he knew that if plot is the dominant element in every story, it must however be tempered with other elements. He avoided the mass-media trap (albeit an antelitteram one) because he had other formal principles. Hence his calculation of the number of lines, his analysis of the musicality of the word "Nevermore," and the deliberate visual contrast between Pallas's white breast and the blackness of the raven, and everything that makes "The Raven" a poetic composition and not a horror film.

But we are still dealing with Aristotle. Poe calculated an appropriate and organic mixture of lexis, opsis, dianoia, ethos, melos. That is how he put flesh on the bare bones of a mythos. The mass media can make us cry, and offer us consolation, but they usually do not allow us to purify ourselves while enjoying a "great creation" that has been well structured. When they do do this, and for me the Ford of Stagecoach certainly does do this, then they really do achieve the ideals of the Poetics.

We come to the last ambiguity. The Poetics is the first work to develop a theory of metaphor. Ricoeur (quoting Derrida on this topic, who says that in Aristotle the defined is implicated in the person who defines) observes that, in order to explain metaphor, Aristotle created a metaphor, borrowing it from the order of movement.^* In fact, the Aristotelian theory confronts us with the fundamental problem of all philosophies of language, namely, whether metaphor is a departure from an underlying literalness or the birthplace of every degree zero of writing.

Although it is true that I remain faithful to a theory of interpretation that, when dealing with written texts, must presuppose a literal degree zero from which metaphor is the departure that must be interpreted, it is also true that if we look at things from the glottogonic point of view (whether at the origins of language, as Vico wanted, or at the origins of every text that comes into being), we must take account of the moment when creativity can emerge, for it does so only at the cost of a metaphorical vagueness that names an object that is as yet unknown or unnamed.

The cognitive power of metaphor on which Aristotle insisted—though this was in the Rhetoric, not the Poetics—becomes manifest either when it puts something new before our eyes, working on preexisting language, or when it invites us to discover the rules of a future language. But the final Aristotelian legacy, the heretical currents of Chomskyan linguistics, and George Lakoff in particular, present us today with the problem of a more radical nature—even though this radicalism was already present in Vico: the problem is not so much seeing what the creative metaphor does with a language that is already established, as seeing how the already established language can be understood only by accepting, in the dictionary that explains it, the presence of vagueness, fuzziness, and metaphorical bricolage.^*

It is not by chance that Lakoff is one of those authors who have begun to elaborate, on the fragments of a semantics where definition was based on atomic properties, a semantics in which definition is represented in the form of a sequence of actions.

One of the pioneers of this tendency (who recognized his Aristotelian debts) was Kenneth Burke with his Grammar, Rhetoric, and Symbolism of Motives, where philosophy and literature, and language, were analyzed in "dramatic" form, through a combined game using Act, Scene, Agent, Agency, Purpose.

Not to mention Greimas, who makes no effort to conceal the fact that a theory of narrativity presides over semantic understanding—I am thinking of that Case Grammar that works on a semantic structure in terms of Agent, Counter-Agent, Goal, Instrument, etc. (Fillmore, Bierwisch), ^* and of many models used in Frames Theory and in Artificial Intelligence. Dominique Noguez has recently published an amusing hoax (in which I am both hero and victim) on the Semiology of the Umbrella. He did not know that reality is stranger than fiction and that one of the models in artificial intelligence is that of Charniak, who, in order to explain to a computer how to interpret the sentences where the word "umbrella" appears, supplies a narrative description of what one does with an umbrella, how one handles it, how it is made, and what purpose it serves.^†' The concept of umbrella is boiled down to a network of actions.

Aristotle did not manage to match his theory of action with that of definition, because, imprisoned by his own categories, he thought there were substances that preceded every action they allowed or had to undergo. We had to wait for the crisis of the concept of substance to rediscover a semantics implicit not in his works on logic but in those on ethics, poetics, and rhetoric, and to think that even the definition of essences could be articulated in terms of underlying actions.

And yet Aristotle could have developed a suggestion from Plato's Cratylus. We know that that work presents the myth of the Nomotheta, or "name giver," a sort of Adam of Greek philosophy. But the problem, which went back to before Plato, was whether the names given by the Nomotheta were provided according to convention (nomos) or because they were motivated by the nature of the things (physis). The question as to which of the two solutions Socrates (and through him Plato) opted for has generated and still generates endless pages of commentary on the Cratylus. But whatever the answer, every time he seems to adhere to the theory of "motivation," Plato speaks of cases where the words represent not the thing in itself but, rather, a source or result of an action. The strange difference between the nominative and genitive in Zeus/Dios is due to the fact that the original name expressed an action, "di'hon zen" (he through whom life is given). Similarly, it is said that "anthropos" can be reduced to "he who is capable of reconsidering what he has seen," inasmuch as the difference between man and animals is that man does not just perceive but can reason and reflect on what he has perceived. We are tempted to take Plato's etymology seriously when we remember that Thomas Aquinas, when considering the classical definition of man as a mortal and rational animal, maintained that specific differences such as "rationality" (which distinguishes man from every other species of living thing) are not atomistic accidents but names we give to sequences of actions and behaviors through which we recognize that there is a rationality in a certain creature that is not perceptible in other ways. Human rationality is inferred from symptoms, so to speak, like talking and expressing thoughts. We know our faculties "ex ipsorum actuum qualitate," from the quality of the actions of which the faculties are origin and cause.^*

According to one of Peirce's examples, lithium is not defined merely by its position in a periodic table of elements nor by its atomic number but through the description of the operations that must be carried out in order to produce a sample of it.^† If the Nomotheta had known and named lithium, he would therefore have invented an expression that could, like a hook, capture a whole series of accounts of sequences of actions. He would have seen, say, tigers not as individuals embodying "tigerness" but rather as animals able to develop certain behaviors, interacting with other animals and in a particular environment—and this story would have been inseparable from its own protagonist.

With these reflections I have perhaps strayed too far from Aristotle, but I was still on the track of his suggestions about action.

In any case, the conference was devoted to contemporary strategies of appropriation of antiquity, and every act of appropriation implies a certain amount of violence. Just as I am convinced that Kant said the most interesting things on our cognitive processes not in the Critique of Pure Reason (where he was speaking in fact of cognition) but in his Critique of Judgment (where he seems to be discussing art), in the same way why should we not go and look for a modern theory of knowledge not (or at least not only) in Aristotle's Analytics but also in the Poetics and Rhetoric?

Abbreviated version of a paper presented at the conference "Les Stratégies contemporaines d'appropriation de l'Antiquité," held at the Sorbonne in October 1990. The proceedings were published as Nos grecs et leurs modernes, ed. Barbara Cassin (Paris: Seuil, 1992).