I
Was the philosophy of literature an invention of classical antiquity? Since all the major phases and movements of Greco-Roman philosophy, from presocratic origins to the Neoplatonism of late antiquity, saw reason to engage with practices we now count as “literature”, it is hard to avoid some kind of affirmative answer to that question. But the answer needs modifying in the light of historical and conceptual complexities. Antiquity brought into being not one, but several (potential) philosophies of literature, and it did so with results which were, and remain, intrinsically contentious.
While the culture of Greece and Rome lacked a category corresponding straightforwardly to “literature” as we now claim to know it, ancient philosophy often defined and sustained itself in part by dialectic with (and sometimes in opposition to) forms of discourse, thought, and sensibility which have subsequently helped to shape conceptions of the literary. This is true above all in relation to poetry (including drama), which held a privileged place in ancient education and society and accordingly provided a vital reference-point for philosophy’s own aspirations to a position of cultural authority. In addition, the study and evaluation of poetry in antiquity became associated with wider interests in the workings of language. This gave rise to modes of analysis – among them stylistics, rhetoric, hermeneutics, and criticism – in which prose texts too, notably oratory and history (but also philosophy itself), were taken into consideration. Such analysis (already visible in Aristotle’s Poetics 20–22 and Rhetoric 3.1–12) was generally known as grammatikê in Greek and came to be called lit(t)eratura in Latin. Thus antiquity possessed, in essence, two overlapping paradigms of “literature”: one, poetry in all its generic varieties (narrative, dramatic, lyric); the other, a domain marked out by special, heightened uses of language (thus a kind of “literariness,” in modern terms). Philosophy of literature emerges from shifting reactions to these two paradigms.
The most obvious gap between ancient and modern thinking in this area concerns prose fiction. Although antiquity produced a tradition of prose novels/romances, this left virtually no trace in philosophical or critical theory. The significance of this gap, however, is debatable, since many ancient theories encompassed questions of (for example) representation, style, audience psychology, and even fiction (Halliwell 2015) which are equally applicable to prose works. In Plato’s Republic, when Socrates proposes a proto-narratological schema of the psychological and ethical implications of “diegetic” form and viewpoint, he indicates in passing that his criteria are applicable to narrative discourse in general (Republic 3.397c, Plato 1997). Aristotle, indeed, taking mimesis rather than verse-form to be a necessary condition of poetry, was prepared to include some prose genres within the latter category (Poetics 1.1447b9–11, Aristotle 1995). The conceptual boundaries of ancient philosophies of literature are therefore far from fixed and the relevant terrain is extensive. The following sections will focus on just a selection of the most prominent landmarks.
II
In a famous but perplexing passage of Plato’s Apology (22a–c) Socrates recounts how, in his search for someone wiser than himself, he interrogated a number of poets, among them tragic playwrights. Before approaching them, he “took up” (written copies of) works he thought particularly well crafted. He makes, then, a prior judgement of poetic merit. But he does not declare his criteria. Nor does he spell out why he needed to interrogate the poets in person: what is he looking for which the poetry itself cannot supply? He proceeds, in fact, to invite them to explain or interpret their own poems, asking them “what they meant” but finding them unable to do so. The context shows that he is not seeking domain-specific knowledge (e.g. mastery of poetic language or metres) but a general, life-guiding wisdom of the kind he had earlier called knowledge of “human and political excellence” (ibid.: 20b) and will later equate with knowledge of “the most important things” (ibid.: 22d).
In inquiring what the poets, or their works, meant, Socrates follows the same principle as with the Delphic oracle, which had stated that no one was wiser than him. “Whatever does the god mean?”, he had asked himself insistently (ibid.: 21b, 22a). Yet the semantic surface of the oracle is not obscure; Socrates pursues meaning at a (putatively) deeper level. Likewise, it seems, he wants the poets to disclose some larger import of their works that he cannot find for himself. In the oracle’s case, he assumed that the meaning might be cryptic, like a “riddle”. Does he make the same assumption about poetry? Certainly he does, perhaps ironically, at Republic 1.332b, when suggesting that Simonides might have been “speaking in riddles” about the nature of justice. But how could the hermeneutics of decoding a riddle furnish an adequate model for the task of establishing what, for instance, a whole tragedy might “mean”?
In the case of the poets, unlike Apollo’s oracle, it needs stressing that Socrates is doing something subtly different from attempting to get behind the text to a separately identifiable authorial intention. He does not ask the poets to specify such an intention, but to undertake an act of interpretation on the words of the poems themselves. If it were the poets’ (separately identifiable) intentions at issue, it would not make sense for Socrates to claim that “almost all those present could have spoken better than the poets themselves about their compositions” (ibid.: 22b). In a piquant historical irony, Socrates prefigures the Schleiermachian principle of understanding an author better than he can understand himself (Schleiermacher 1998: 23, 33). But whereas Schleiermacher’s hermeneutics ties this to the psychological difference between what remains unconscious in the author and what interpretation of his text makes explicit, Socrates contemptuously counts it as a symptom of something badly lacking on the poets’ part.
Socrates’ disappointment with the poets leads him to infer that they
do not compose by expert knowledge but by some kind of natural talent and in an inspired state of mind, just like seers and oracle-reciters, for these too say many beautiful things but possess no knowledge about what they say.
(Apology 22c)
The hypothesis of “inspiration” supposedly explains what is inaccessible to discursive reason. But this leaves us with the puzzle of how Socrates himself finds value (“many beautiful things”) in the poets’ works. Moreover, later in the Apology Socrates will have no trouble adopting the role of poetic interpreter when using the Iliadic Achilles as a paragon of virtuous disdain for death (ibid.: 28b–d) and in the process supplementing Homer’s discourse with his own (by appealing to a vocabulary of “justice” that is absent from Achilles’ words in the relevant passage). At that later stage, Socrates moves confidently from inside to outside the text, connecting the psychological drama of the poem with a general conception of how to live (ibid.: 28d). It transpires there that poetic interpretation cannot be neutral exegesis. It is a form of ethical dialectic with the text – and with other interpreters. Socrates’ philosophical premise appears to be that every poem is incomplete without the supplementary discourse of ethical interpretation. Interpretation assimilates the poem into a larger framework of life-values applicable both to the characters in the text and to the reader. Poetry, on this view, needs philosophy.
The poets themselves, of course, would have given a very different account of their encounters with Socrates. Rejecting what Nietzsche, in The Birth of Tragedy, called “aesthetic Socraticism” – “everything must be conscious in order to be beautiful” (Nietzsche 1999: 64) – the poets might have protested that their work is complete without philosophical interpretation. As Tolstoy put it, subconsciously inverting a Socratic stance: “An artist’s work cannot be interpreted. Had it been possible to explain in words what he wished to convey, the artist would have expressed himself in words. He expressed it by his art” (Tolstoy 1930: 194).
Socrates’ interrogation of the poets voices dissatisfaction with the discursively non-transparent nature of their works, but it also betrays a residual attraction to the beauties of these same works. In that regard it is emblematic of the ambivalence which colours Plato’s lifelong obsession with poetry. Plato’s relationship to poetry is complex; it should not be reduced to the textbook clichés of unqualified hostility. This is true even of Book 10 of the Republic, where Socrates notoriously speaks of an “ancient quarrel between philosophy and poetry” (ibid.: 607b). He does so, it is too rarely noticed, to defend himself against the possible accusation of insensitivity to poetry’s intense allure; and he proceeds to declare that “we would happily welcome poetry back” into the city (and the soul) if “lovers of poetry”, of whom he confesses he was once one himself, can harmonise the pleasure of poetry with the good of both individual lives and the community as a whole (Republic 10.607c–e). This is not implacable hostility, but conflicted uncertainty translated into a challenge to the Republic’s own readers. Socrates demands that poetry’s distinctively imaginative and emotional powers of expression, capable as they are of resisting rational control, be held to account to philosophical standards of meaning and value. Yet Plato’s dialogues never purport to supply a definitive answer to the question of whether or how far that is achievable. In that sense, we might say, Plato conceives of the “philosophy of literature” as a permanently unsolved problem.
III
At a late stage in Milan Kundera’s novel Immortality, the semi-autobiographical narrator pauses to reflect on Aristotle’s conception of narrative unity in the Poetics (Kundera 1991: 338–40). “Aristotle did not like episodes”, he states (in something of a simplification). “According to him, an episode . . . is outside the causal chain of events which is the story.” Yet “Life is as stuffed with episodes,” he remarks, “as a mattress is with horsehair.” Shortly afterwards, the narrator concedes that while every episode has the potential to turn into a story, only some of them fulfil that potential. So is he, or is he not, disagreeing with Aristotle?
Aristotle’s Poetics is the single most widely cited work in the history of the philosophy of literature. It has been the object of continuous debate – along a spectrum running from canonisation, via contestation, to outright opposition – since its “rediscovery” by sixteenth-century Italian literary theorists. As my example from Kundera illustrates, it is hard now to argue with Aristotle without arguing with oneself about assumptions deeply embedded in views of literature. The Poetics’ compressed arguments contain suggestive observations on such fundamental topics as genre, form, plot, character, emotional expression, audience psychology (including the notoriously elusive concept of katharsis), and metaphor (Rorty 1992). Though partly a response to Platonic promptings, the treatise stays conspicuously silent about Plato’s writings as such – except, ironically (and provocatively), to classify Socratic dialogues like Plato’s as themselves “poetic” in virtue of their mimetic, i.e. dramatic-cum-fictive, status (Poetics 1.1447b9–11).
Aristotle does not provide any easy answer to the question of how philosophy should construe its relations to poetry/literature – or vice versa. Indeed, at a certain level he displays his own ambivalence on the point. In the first chapter of the Poetics he is anxious to separate poetry and philosophy: Homer and Empedocles (who had written in verse), he insists, “have nothing in common except their metre; so one should call the former a poet, the other a natural scientist” (ibid.: 1.1447b17–20). Yet in chapter 4 he explains the pleasure derived from poetry (as well as from visual and other art-forms) as cognitive and quasi-philosophical: “understanding gives great pleasure not only to philosophers but likewise to others too” (ibid.: 4.1448b13–14). And in chapter 9 he famously pronounces poetry “more philosophical” than history on the grounds that “it speaks more of universals, history more of particulars” (ibid.: 9.1451b5–7). If Aristotle anticipates certain modern views of literature’s own philosophical capacities, he does so equivocally. Taking chapters 1 and 9 together, it looks as though mimesis is contrasted with the making of overt truth claims but can nonetheless be a medium for serious discernment of some of the underlying structures of human experience.
Everything in the Poetics is underpinned by the master-concept of mimesis, the narrative/dramatic projection of a hypothetical world (Halliwell 2002). Aristotle happily grants mimesis a wide, flexible scope, capable of moving between three basic perspectives: known states of affairs; the whole realm of culturally shared beliefs and suppositions (including mythology); and the ideal (ibid.: 25.1460b9–11). Even the first of those perspectives requires imaginative re-making of the world: if a poet draws on historical events, he must still convert and shape them into the substance of a new artefact that will be poetry not history (ibid.: 9.1451b29–32). Aristotle’s account of mimetic “making” accordingly implies, in some degree, the operations of fiction.
Most interpreters maintain that the Poetics ascribes to poetry its own intrinsic, even formalist, principles of value. But does Aristotle claim pure autonomy for poetry? In chapter 25, which anatomises (il)legitimate reasons for criticism of poetry, Aristotle asserts that “poetry does not have the same standard of correctness as politics, or as any other art” (ibid.: 1460b13–15). But “not . . . the same” need not mean entirely disconnected. Later in this chapter Aristotle declares:
when the issue is whether or not someone [i.e. in poetry] has spoken or acted well, one should examine not only whether the actual deed or utterance is good or bad, but also the identity of the agent or speaker, to whom he acted or spoke, when, with what means, an...