Stephen Halliwell
Can poetry – can “imaginative literature” in general – tell its audiences truths about the world and their place in it? If so, what sorts of truths might these be: particular or universal? descriptive or normative? human or divine? If there is such a thing as poetic truth, is bad poetry therefore a form of falsehood, even “lying”? If, by contrast, truth cannot be expected of poetry, does that diminish the art’s value or only create space for values of a different kind? Or is the truth–falsehood dichotomy insufficient to capture the significance of poetry and do we need a third category, “fiction,” for this purpose? Can truth and fiction be mixed together, as the Muses appear to tell Hesiod at Theogony 27–28 when they announce – in a teasing tone that signals the symbolic force of their words – that they know both how to speak the truth and also how “to tell many falsehoods which resemble real things”?
Those questions are all in origin, and in a still important sense, Greek questions. They outline a cluster of issues which emerged within the various spheres of discourse (including poetic criticism, philosophy, and rhetorical theory) that contribute to the distinctive history of aesthetics in antiquity. They are questions, it is worth adding, which on occasion also found their equivalents in reactions to non-verbal art forms: notice, for example, Socrates’ skepticism about particular episodes of Greek myth depicted in visual media (as well as in poetry) at Plato, Euthyphro 6b–c, or, somewhat differently, his reference at Republic 6.488a to composite creatures like goat-stags invented by painters. The present chapter, however, will confine itself very largely to poetry, with subordinate consideration of certain genres of prose literature. I shall concentrate on a selection of Greek texts which show more or less explicit awareness of the kinds of concerns registered above. But rather than simply surveying a series of passages which make pronouncements on poetic truth or falsehood, my argument will primarily address the more difficult question: is there a Greek concept of fiction?
Like many other problems of a similar kind in the history of ideas, this question calls for something more than a comparison of ancient and modern vocabulary, though there is no need to deny that there are particular terms, including the Greek verb plattein, “mould” or “fashion,” and its cognates (equivalents to Latin fingere, fictus etc.), which are pertinent to the investigation.1 But it is also a question which cannot be satisfactorily answered by simply making up our minds what we (think we) mean by “fiction” and then looking for the existence or absence of a Greek counterpart to it. That is because we cannot take for granted a completely agreed modern concept of fiction. We need, instead, a more dialectical approach to the history of aesthetic ideas, one which constantly bears in mind the historical contingency of our own categories of thought. “[T]he boundaries between fiction and non-fiction … are not laid up in heaven” (Bakhtin 1981, 33).
Both the practice(s) and the conceptualization of “fiction” have evolved historically and are still evolving. The term itself, in English (other modern languages both match and magnify the complications), lacks a univocal semantics. I offer three brief illustrations of this point. First, lexical data alone cannot tell us what fiction is: the word’s early use to denote mendacity or feigning remains available and can be activated or suppressed in individual contexts, a point interestingly parallel, as we shall see, to Greek usage of pseudos (lit. “falsehood”) and its cognates. Second, at least until the eighteenth century “fiction” was readily applied to the status of (much) poetry, but this application became much less common as the term gradually hardened into the standard classification for certain kinds of prose literature. Far from creating a clear demarcation, that shift has left a serious obscurity about the boundaries of fiction, an obscurity which matters all the more when we expand our terms of reference to encompass pre-modern cultures. Third, the terminological development of “fiction” into a general category of prose narrative was itself hardly an uncontentious process. It was mediated through debates about the relationship of fiction to history and “life”: witness, among other things, eighteenth-century arguments over the difference between the “romance” and the “novel” (the latter being, for some, more of a “true history” and less “fabulous” than the former), or nineteenth-century debates about the social and epistemological implications of expecting realism or naturalism from the novel.
Such disputes have not disappeared. They have not only continued to fuel the experimentalism of the novel itself but have also encouraged the burgeoning of hybrid categories – historical fiction, documentary fiction, “faction,” autofiction – which tend further to destabilize the core criteria of fiction. Nor is it only novelists who exploit the elusiveness of fiction’s identity. Some modern historians, resurrecting a freedom enjoyed by their ancient counterparts, have added to the fluidity of the concept by appropriating, but thereby also modifying the function of, what are normally regarded as fictional modes of writing (e.g. Hopkins 1999; Hatcher 2008).
Given those considerations, it should be no surprise that there has been such a proliferation of modern theories of fiction. These have been constructed within a plurality of intellectual frameworks: analytic philosophy, speech-act theory, “possible worlds” semantics, evolutionary psychology, cognitive poetics, narrative theory, and more besides.2 Yet despite the diversity of literary examples (sometimes) adduced by such theories, it is remarkable that all the schools of thought just mentioned arrive at predominantly ahistorical conclusions. They purport to identify the logical, semantic, or psychological components of fiction, and have little or nothing to say about the interaction between conceptual and cultural factors in the way that ideas relevant to the definition of fiction have developed over time.
These reflections are more than tangential to an interest in Greek (and Roman) sensibilities. They reinforce my contention that we do not have an entirely secure modern paradigm of fiction against which to test ancient texts and arguments. Rather than aiming to reduce fiction to a logically stipulative definition, we do better to think of it as an inherently complex zone of discourse, thought, and imagination which cuts across a strict dichotomy of truth and falsehood and intrinsically complicates both halves of that division. That is a perspective, I hope to show, which can help enhance our understanding of a range of Greek ideas, from Hesiod Theogony 27–28 (see above) onward. It should not be confused, however, with a crudely relativist view, let alone a view hospitable to facile notions of “panfictionality,” that is, the thesis that everything is fiction – a proposition deeply objectionable on ethical as well as logical grounds (Ryan 2005).
Because my guiding question in this chapter is whether ancient debates about poetry, and to some extent “literature” more generally, display traces of conceptual, even theoretical, awareness of fictionality, I will have to leave largely on one side the workings of fiction as a set of practices in Greek poetry and literature. Relations between theory and practice in this area cannot, of course, be completely disentangled. In this connection it is worth glancing at the sheer multiplicity of claims which have been advanced about the emergence (even the “invention” or “birth”) of fiction at specific junctures in Greek culture, since these claims are distributed across both literary and theoretical works. Different scholars have located the origins of Greek fiction(ality) in Homeric epic (especially the Odyssey), the words of the Muses to Hesiod in the prologue to the Theogony, the Archaic rise of literacy, Stesichorus’ palinode on Helen, Pindar’s aperçus on poetry, Gorgias’s theory of language (and/or his theory of poetry), Plato’s dialogues (candidates for both the practice and the partial theorization of fiction), Attic comedy (Old/Middle, according to some; New, according to others), Aristotle’s Poetics, Theocritean bucolic, and, last but not least, the Greek novel (Halliwell 2011, 10–12 cites examples of all these views).
My own position will avoid on principle any suggestion of a single, determinate origin for Greek awareness of fictionality. But as a backdrop to my selective comments in the following sections it may be helpful to indicate that I see manifestations of Greek sensitivity to fiction as falling into several broad stages: first, the complex poetic self-consciousness of the Archaic period, a self-consciousness often in dialectic with claims to poetic truth-telling and emblematized by the riddling words of the Muses at Theogony 27–28; second, a phase of more explicit theorization in a group of Classical thinkers which includes Gorgias, Plato, Aristotle, and Isocrates; third, formalized Hellenistic and imperial typologies of discourse which have intricate implications for conceptualizing fiction; fourth (though I shall not be able to examine this in its own right here), a process of overlap, from the late Hellenistic period onward, between Greek and Latin categorizations, a process which involves the vocabulary of fingere, fictus etc. from which English and some other modern languages subsequently derive their own terminology of fiction.3
My approach to the subject can be clarified in part by contrasting it with a recent work which defines fiction for expressly Greek purposes as consisting of “previously untold stories set in post-Homeric times, as distinct from history and myth” (Lowe 2000a , 265). Relying on his tripartite scheme of history, myth, and fiction, Lowe discovers the emergence of a Greek sense of fictionality principally in Old and Middle Comedy (Lowe 2000b; cf. Lowe 2000a, 88).4 This is a thought-provoking thesis but is open to two main objections. First, the definition places unwarranted weight on newness or originality (“previously untold”), a factor which certainly has its own importance but should not be allowed, for reasons which will become apparent, to circumscribe what can count (in Greek terms) as fiction.5 Lowe has been interestingly influenced in this regard by Greek comedy’s self-references to its capacity for novelty. Antiphanes fr. 189, whose speaker contrasts comedy’s constant creation of new characters and plots with tragedy’s traditional scenarios, is germane here; and one can easily think of Aristophanic passages (including the slave’s assertion at Wasps 71–73 that no spectator could conceivably guess what the father’s outlandish affliction is – or, therefore, what the plot of the play will be) which emphasize authorial ingenuity at various levels of poetic creation.6 But it is a crucial difference between Lowe and myself that while he thinks this constitutes evidence for the “invention” or first explicit awareness of fiction, I prefer to see it as highlighting one particular part of the spectrum of Greek fiction – and carrying it, indeed, to the point of metafiction.7
A second objection to Lowe stems from the fact that when he refers to “previously untold stories” he is using the term “story” itself, with a narratological inflection, to mean “the complete set of events recounted” as opposed to the narrative construction or the “telling” element in storytelling (Lowe 2000a, 267) – histoire, then, as opposed to récit. This raises a severe difficulty. In the body of his book, Lowe sometimes treats the second of those components of his narratological model as itself generating fiction: he talks, for example, of fifth-century tragedy, despite its use of (mostly) old “stories,” as engaging in what he calls the “free fictional creation” of its narratives (Lowe 2000a, 189). So the definition (“previously untold stories …”) effectively collapses: a new “telling” of an old “story” can in itself bring fiction into play. For my purposes, the vital lesson of this point is that fictionality can exist, or be perceived, at more than one level of a work (or genre). Fiction is not an all-or-nothing category.
That inference is worth developing in relation to a Greek philosophical text in which many scholars and critics have (rightly, in my view) found the adumbration of a concept of fiction. In Chapter 9 of the Poetics, Aristotle argues that even if a dramatist uses historical events as the basis of a tragedy, he must still reconfigure them into a poetically unified plot structure (muthos). Aristotle here gives striking prominence to the verb poiein (“make,” “compose poetry”) and the noun poiētēs (“maker,” “poet”). At one point in the passage (Poet. 9.1451b20–22) the verb clearly denotes “invented” or “made up” events and characters. But the verb cannot mean “invent” when Aristotle speaks of a poet choosing (literally) “to make things that have happened” (1451b29–30), since he cannot be talking of “inventing” historical events. The process of poiēsis, here very closely linked to mimesis (1451b28–29), is a matter of plot construction: the elaboration, that is, of an imagined sequence of human action in a medium which requires its own formal design and unity. It might seem tempting, in the light of what was said earlier about Lowe’s definition of fiction, to take Aristotle to be himself locating fiction in the “telling” rather than the “story” element of narrative. But that would not be quite right: Aristotle’s conception of muthos in the Poetics is itself a combination of “story” and “telling.” But it is legitimate to find here and elsewhere in the Poetics a recognition of poiēsis, and mimesis in general, as a fictive act: the representation of possible patterns of human experience in the form (whatever the source of the imagined events) of a distinct poetic artifact.
In any analysis of putative Greek conceptions of fiction much hangs on the relationship between the categories of myth and history. (Recall Lowe’s attempt to make “fiction” distinct from them both.) But as the passage of the Poetics just cited shows, we ought not to expect to be able to anatomize Greek ideas of fiction with clean breaks at the joints. The mentality of Isocrates provides an illuminating test case in this respect. It is sometimes said that Isocrates does not distinguish between myth and history (e.g. Usher 1990, 156). That is incorrect. Isocrates defends the core historicity of an archive of traditional stories which includes the Trojan War, the life of Heracles, and the episode of the Seven against Thebes, but he regards the transmission of this archive over a long period of time as having diluted truth with invention. In the Evagoras, for instance, perhaps betraying the influence of Thucydides (1.10.3, 21.1), he contrasts the “truth” of the Trojan War with the “myths” which have been embroidered around it (9.66). He hints that not all characters in various versions of the war may even have existed (9.6), and he makes it clear more generally that poets are culturally licensed to invent, embellish, and imaginatively elaborate the narrative materials of their works, introducing into them such things as face-to-face encounters between gods and humans which lie outside the permitted scope of a political prose writer like himself (9.9–10, 36). Isocrates was hardly alone in this set of attitudes: compare the general comment on tragedy at pseudo-Andocides 4.23 (probably a fourth-century BC text): “when you watch plays, you do not know whether those events actually happened like this or have been fashioned [the verb plattein] by the poets.”
Isocrates sees myths, then, particularly in their poeticized retellings, as a category of discourse whose truth-value is variable. But since he shows no signs of supposing that poets have set out by their innovations and elaborations to falsify anything, or to mislead anyone, we are entitled to infer that he recognizes a fictionalizing force at work inside these narrative traditions. One instructive case of this Isocratean stance appears in his comments in the Panegyricus on the story of Demeter’s visit to Attica and the consequent invention of agriculture. To buttress his ideological case for Athens’ standing as the chief benefactor of the Greek, if not the entire, world, Isocrates wants to be able to invoke the privileged status accorded to Athens in the Demeter story. Yet he is wary of appearing to accept that story wholesale, with uncritical naivety. Isocrates was operating in a cultural climate where some, perhaps many, were radically skeptical about the historical authenticity of such myths; there are those, he says (4.30), who would despise these stories as “primitive” (archaios). Notice, accordingly, how his position in Panegyricus reveals a subtle nuance when he says that “even if the account [of Demeter’s visit] has become mythologized (muthōdēs), it is still appropriate for it to be told in this context …” (Isoc. 4.28; compare the similar passage at Lycurgus, Leocrates 95). The perfect tense “has become” implies that the story in question is neither sheer fabrication nor wholly historical. It possesses a veridical kernel (I will not pause to ask how Isocrates knows that) but has, over time, become fictionalized – transformed into a dramatically self-propelling narrative – in many of its details. This position straddles the narratological story/narrative distinction I mentioned earlier. It is not straightforwardly that for Isocrates the “story” is historical but has been retold in fictional form. His views blur any such distinction, leaving it hard to say exactly where truth (and/or fiction) begins and ends. This is because, as a pragmatic rather than a painstaking intellectual, Isocrates is less concerned with extracting reliable “information” about the past from a body of traditional narratives, in the way that Thucydides had been, than in preserving the exemplary value of particular stories for his own ideological purposes.
The case of Isocrates supports my general thesis that fiction is best thought of not as a sharp-edged category but as a fluid zone of possibilities which complicates a neat dichotomy of truth and falsehood.8 Some of the ramifications of that point become visible in a more challenging form in the dialogues of Plato, who, as I stressed earlier, is a pivotal figure for the present inquiry. On the one hand, Plato is a candidate for being himself a major practitioner of fiction. Certainly, from a modern vantage point he has often been seen in that light; Nietzsche, Bakhtin, and D.H. Lawrence are among those who have regarded him as the first novelist.9 But parts of his work are also important for posing questions about the truth-values of narrative, poetic, and other kinds of discourse in ways that direct attention to the conceptual problem of fiction. These two aspects of Plato’s writing sometimes overlap at the point where philosophical myths are incorporated into the dialogues and prompt internal comments on those myths’ own truth status as well as their relationship to other kinds of myths.10
The complexities surrounding Plato’s double relationship to fiction are immense. They are such that it has proved possible for one leading scholar to shift from arguing that it was in Plato’s work that the Greek “birth” of fiction originally occurred to maintaining, on the contrary, that Plato did not possess (and nor did Greek culture more generally) a proper concept of fiction after all (Gill 1979, 1993). I am not myself persuaded by either of these (interesting) claims. Not all the strands of the question can be unpicked here. In keeping with my own line of argument, I shall concentrate on reasons for disagreeing with Gill’s denial of a concept of fiction to Plato.
Gill places great weight on the fact that Plato’s texts prioritize the ethical acceptability of poetic (and other) narratives over their literal veracity: it matters far more, that is to say, whether a story endorses certain values than whether it recounts things which are, or ever were, actually the case. But in the discussion of poetry in Republic Books 2–3, according to Gill, Socrates goes further than that by equating “true” with “good,” “false” with “bad,” and thereby shows no interest in, or need for, anything like a category of fiction. Now I agree that in this and other Platonic passages a normative conception of truth is prioritized over factual veracity, though the latter is not actually collapsed into the former (hence the possibility of a true narrative which is nonetheless ethically unacceptable, 378a). It does not follow, however, that there is no scope here for awareness of fiction. On the contrary, I believe that this part of Republic 2–3 recognizes fictionality as a culturally sanctioned exemption from a requirement of literal truth-telling while refusing to allow that exemption as a defense against ethical criticism of poetry’s contents. Even if a narrative is built around something other than historical or factual truth, it remains possible (and, Socrates suggests, essential) to scrutinize its acceptability at the level of ethical expressiveness and normativity.
Clear evidence in favor of this alternative reading occurs right at the start of the critique of poetic muthoi in Book 2, a passage strangely ignored by Gill. At the outset of Socrates’ discussion of how the guardians of the ideal city should be educated in mousikē, he sets up a basic division of logoi (which in principle covers all forms of discourse) into “truth” and “falsehood” (pseudos, 2.376e), the latter also equated with “fashioned stories” (the verb plattein, 377b6). This division has to be treated as in itself pre-ethical. In many Greek contexts, for sure, pseudos and its cognates denote mendacity (as can “fiction” in English usage). However, the term cannot have that sense in Socrates’ basic dichotomy of discourse, not only because he talks of educating the young in (i.e. with) “false” as well as true logoi (377a), but also because he goes on to distinguish explicitly between good and bad “falsehood.” Thus, although Hesiod, Homer, and other poets “compose” or “construct” stories/myths that are generically “false” (377d), only some of their falsehoods fail the test of the argument’s ethical criteria.11
Socrates therefore acknowledges the existence of a kind of “false” discourse which, unlike mendacity, does not, per se, seek to mislead and can serve a positive function within its cultural setting. The point is underlined in a later reference back to Book 2’s argument, when Glaucon stresses how they had earlier agreed that ethical habituation could be promoted by stories (logoi), “both the mythical [muthōdeis] and those more closely tied to actuality [alēthinōteroi].”12 Furthermore, in an echo of the Hesiodic motif of poetic “falsehoods which resemble real things” (Theogony 28), Socrates speaks of the desirability in “mythology” of sometimes “making what is false [pseudos] resemble the true as far as possible” (382d), an unintelligible proposal unless pseudos here lacks negative implications of deceit. What emerges from this stretch of the work may not constitute a full-blown account of the workings of fiction. But it does show that Socrates’ basic dichotomy between true and false discourse makes room for narrative whose significance and value are not tied to any presumption of literal truth. This in turn supplies a precedent for the Republic’s own so-called “noble lie” (or, better, “grand fiction”), which is itself envisaged as a piece of quasi-poetic “mythology” (3.414b–15c).
Where, then, Gill sees what one might call the ethical imperative within Platonic “narratology” (cf. Halliwell 2009) as excluding fiction, I see it as bringing to the surface a recognition of fictionality and, at the same time, of the need for an ethics of fiction. It is wrong to assume that a concept of fiction must be independent of questions of ethical normativity, that it must center on “a concern with story-telling for its own sake” or involve “authorial and narrative ‘truth-games’” (Gill 1993, 62). Those are not (and should not be) criteria of fiction universally accepted by modern theorists, practitioners, or readers.
Juxtaposing Plato with Isocrates and Aristotle serves to foreground the way in which an awareness of fictionality can be motivated by different intellectual needs. Dominant in Isocrates’ case, I have argued, is his culturally sensitive admission that the “archive” of Greek mythology, on which he draws for purposes of political and social ideology, is not a record of total historical fidelity. That admission was itself induced by the pressure of a general cultural consciousness of the freedom enjoyed by poets in the construction and elaboration of their materials. In Aristotle’s case, the impetus is precisely a desire to set up a model of artistic production in its own right. Several concepts in the Poetics accordingly bear traces of a sense of fictionality: this is true of mimesis (imaginative, “world-simulating” activity), muthos (the plot structure which the poet “makes” and configures), and, not least, the process of poiein or poiēsis (poetic composition) itself.13 As for Plato, the central role is played by an uncompromising ethical imperative, an imperative, however, which carries with it an insistence that narratives which do not lay claim to literal truth are not thereby exempted from ethical demands. This reading of how a Platonic agenda allows space for recognition of fictive discourse has, moreover, a deep plausibility in terms of the dynamics of Plato’s own writing, whose fictionalizing tendencies can be detected at several levels: in the concrete yet often historically “rewritten” circumstances which form the settings of individual dialogues; in imaginary scenarios, such as the Republic’s ideal city (itself conceived of as a muthos),14 which contribute to the framing of hypothetical arguments; and, last but not least, in the grand metaphysical and eschatological visions of Plato’s philosophical myths.
While the three thinkers considered above approach problems of fictionality from various directions, there is one important premise they share. It is a premise common, in fact, to ancient traditions of aesthetics in general and one which marks a crucial difference from many modern theories of fiction. Plato, Isocrates, and Aristotle all accept without question that (poetic) “fictions,” if successfully constructed and composed, have an intrinsic capacity to “transport the souls” (psuchagōgein) of hearers and readers in a way which involves genuine, not make-believe, emotions.15 In relation to what is now often called the “paradox of fiction” (e.g. Gendler 2010, 227–237), these thinkers implicitly deny that there are any grounds to doubt the psychological authenticity of emotional engagement even with imaginary characters and events. What matters, for this characteristically Greek conviction, is not whether the objects of emotions are believed to be real but whether the manner in which those objects are evoked causes a sufficient degree of absorption on the part of an audience to activate the free flow of an affective response. Such a position had been defined in the fifth century by Gorgias, who we know held a metaphorical (and paradoxical) conception of aesthetic “deception” (apatē) which attempts to capture a notion of willing submission (Plato would call it “surrender”: Republic 10.605d) to the imaginative power of a fictional world.16 Gorgias took such power to be rooted in the seductive nature of logos, which for him is language in all its manifold uses. Fiction, on such a view, is a particular kind of logos which projects events known not to be literally true but nonetheless taps into the mind’s instinctively emotional reactions to what is put before it with compelling vividness.
The classical evidence examined in the previous section is conscious of fiction as a zone of possibilities which encroaches especially on the territory of “myth” and poetry, not least myth in poetry. In the Hellenistic period and beyond, awareness of fictionality continues to be often connected to reflections on these same topics. When, for instance, the scholar Eratosthenes, whose views are reported by Strabo, Geography 1.2.3, provocatively called Homeric poetry “an old woman’s storytelling [muthologia]” which was free to fashion or fabricate (plattein) whatever would serve the entrancement (psuchagōgia) of its audiences, he was blocking various claims about the supposedly historical basis of Homer’s poems by giving fiction priority over truth in epic, even though he also allowed for (for example) elements of accurate geographical reference within the poems (Pfeiffer 1968 , 165–168). Strabo, committed to a Stoic form of moralism, responded by conceding the presence of muthos (extravagantly inventive narrative) in Homer, while subordinating that concession to a cognitive and ethical framework of criticism which maintained Homer’s status as a quasi-philosophical teacher (Geography 1.2.4–40).
But the Hellenistic period also saw issues of fiction raised in relation to a wider range of texts, including historiography and rhetoric. One important manifestation of this development is the emergence of typologies of the varying truth-values of different kinds of (in the broadest sense) narrative discourse. Although we cannot reliably chart the early formation of such typologies, they involved ideas which impinged on the fields of poetics, rhetoric, grammar, and historiography. They are attested in a number of sources, Roman as well as Greek,17 from at least the first century BC onwards. And they persist until late antiquity, when for instance the rhetorician Nicolaus of Myra, in the fifth century, summarizes a four-part classification of narratives into mythical (muthikos), historical, pragmatic/judicial (i.e. the stuff of forensic and political decisions), and fictive (plasmatikos) – only superficially a neat arrangement, since he has to grapple, less than conclusively, with residual uncertainties: disagreement exists, he admits, about the possible historicity of some “mythical” stories, while plasmatikos is a term applicable both to such things as animal fables (themselves known as muthoi) and to narratives of events which in principle could have occurred.18
Some of the problems broached but far from fully resolved by such typologies can be sampled in the much-discussed schema of narrative (in both verse and prose) attributed by Sextus Empiricus to the grammarian Asclepiades of Myrlea.19 Asclepiades divided the subject matter (historia) of all narrative into the true, the false, and the “as if true”; he glossed the true as “concerned with actual events” (praktikē), equated the false with “myths,” and took comedy and mime to exemplify the “as if true.” As reported, he further subdivided the true into “three parts,” dealing with, first, gods, heroes, and famous men, second, “places and times,” and, third, actions; but this may be a garbled version of a set of distinctions – between agents, places, times, and events – applicable to all narratives (Slater 1972, 319–326). Asclepiades’ schema is frustratingly incomplete. It does not tell us, for instance, whether the different types of subject matter can be combined within the same work or genre, though we do know of other Hellenistic theories which encompassed narrative hybridization, among them Polybius’ tripartition of the components (historical reality, dramatic organization, myth) which constitute “poetic freedom” in Homer and others.20 It is arguable that Asclepiades’ typology represents less a set of fixed types than a spectrum of gradations of realism and fantasy. A notion of fictionality enters the picture, at any rate, with two of his three categories (and regardless of which of these Asclepiades classed as plasma, “fabrication”):21 the “false,” which probably covers those parts of mythology most blatantly incompatible with the observable workings of nature, but also the “as if true,” associated particularly with quotidian verisimilitude of the sort depicted in New Comedy.
In a sense, Asclepiades’ typology might be thought hardly new at all. It looks, on one level, like a formalization of the claim of Hesiod’s Muses (see above), a claim which continued to be echoed by poets themselves in the Hellenistic period (Callimachus, Hymn to Zeus 65, “may I myself tell falsehoods that would convince a hearer,” being an ironic case in point). It is also likely to have taken some of its impetus from traditions of Peripatetic criticism which went back to Aristotle himself. In Chapter 9 of the Poetics, which proposes, as we saw earlier, that even historical events must be refashioned and re-formed in order to be made into effective poetry, Aristotle not only distinguishes between history (as the domain of things which have actually happened) and poetry (which deals with things “which could happen” according to probability or necessity); he also differentiates between (contemporary) comedy (whose plots and characters he takes to be entirely invented), iambos (which he thinks deals with real individuals), and tragedy (most of whose characters he assumes to have a basic historicity, though this cannot extend in his mind to all the details of the stories told about them). Later in the treatise, at Poetics 25.1460b8–11, Aristotle sets out a different and explicitly tripartite typology of the possible objects of mimesis (in both poetic and visual art forms): namely, “the sorts of things which did or do exist,” “the sorts of things people say and believe,” and “the sorts of things which ought to be the case.” Chapter 25 goes on to make it clear that these types of subject matter cannot be neatly compartmentalized in terms of genre; epic and tragedy make use of all three of them. Aristotle’s tripartition, especially when overlaid on Chapter 9’s distinctions, does not yield a simple test of poetic truth and falsehood. It provides for mimetic art forms which can subtly manipulate fictionalized particulars (including reworked elements of history) and move between different narrative “registers,” while nonetheless modeling their images of life in keeping with the underlying structures of thought (“universals”) which allow the mind to make connected sense of the world.
The sorts of distinctions drawn by Aristotle in Poetics 25 imply that sophisticated audiences/readers of poetry will adjust their expectations to suit different kinds of narrative content. Fictionality requires shared parameters of communication between text and recipient. Another (originally) Hellenistic typology which directs attention to this point is found in a scholion on the Iliad. It is attached to part of the scene in Book 14 where Hera distracts Zeus from events on the battlefield by inducing him to make love on the slopes of mount Ida (inside a golden cloud conjured up by Zeus). Stemming from a source which rejects Socrates’ complaints in the Republic (3.390b–c) about the immorality of the gods’ uncontrolled erotic desires, the scholion states that “there are three types according to which all poetry can be classified: that which is mimetic of truth …, that which involves imaginative embellishment [phantasia] of truth and should not be scrutinised in detail …, and, third, that which operates by transcending truth as well as by phantasia …”22
It is clearer here than in the case of Asclepiades’ tripartition that we are dealing with a sliding scale. What’s more, in the present case “truth” need not entail historicity; it is a reference point for what real life is taken to be like, as shown by the fact that even the first type is illustrated (if the text is not garbled) by a series of adjectives – “father-loving, woman-hating, unreliable, outspoken” – which appear to denote character traits that might be encountered in the everyday world, if also in such literary genres as New Comedy. As we move along the scale, then, we get increasingly remote from the domain of experience that poetry’s own audiences have access to; at the furthest extreme we reach worlds (“Cyclopes, Laistrygonians, and the present scene of the gods”) which by definition those audiences could never enter. The scholion’s further comment on the middle type, whose imaginative embellishment or enhancement of truth “should not be scrutinised in detail” (a consideration evidently applicable, a fortiori, to the third type too), is revealing. It demonstrates that the impulse behind the whole scholion marks an attempt to correlate narrative modes/styles with appropriate frames of mind on the part of audience or reader. That point is clinched with a reference to the nature of souls in the Homeric Hades. These souls are notionally disembodied yet are represented as capable of taste and speech. But it is only a naive reader, so the scholiast indicates, who will literal-mindedly suppose that “they have tongues and throats.”
The Homeric scholia, reflecting a synthesis of several traditions of literary criticism, show sensitivity to fiction in a range of poetic contexts (Nünlist 2009, 179–181, 260). The particular note I have cited, whatever its origins, exhibits an attitude which recognizes that poetry exercises a broad but fluctuating freedom of invention and imagination in switching between different kinds of “world” as defined in relation to an audience’s normal expectations of life. Fiction, on this view, is not a single phenomenon. It is more like a mobile perspective on reality, and one which calls for corresponding shifts in an audience’s standards of judgment, including a willingness, where necessary, to adopt a sort of “suspension of disbelief” (cf. that phrase “should not be scrutinised in detail”).
The nexus between fictionality and audience psychology can be seen from a different angle in Chapter 15 of Longinus, On the Sublime, where the author draws a distinction between the kinds of imagination/visualization (phantasia) respectively appropriate to poetry and rhetoric. Longinus sums up the difference by saying that poetic phantasia “belongs to the realm of mythic [muthikōteros] exaggeration and utterly exceeds credibility,” whereas the best rhetorical phantasia is tied to what is “realistic and truthful” (Subl. 15.8). Rhetorical phantasia, Longinus believes, should aim at enargeia, vivid immediacy; it is tied to the goal of persuasion. Poetic phantasia, on the other hand, is ideally productive of ekplēxis, which is a trait of the sublime and has a license to reach beyond what is merely persuasive.
It is striking, however, that having established this distinction Longinus then equivocates over it. He does this in part by trying to preserve for the greatest rhetoric (which, after all, is a prime source of his examples of sublimity) the capacity to cross the boundaries of persuasion and overpower the mind with irresistible force. What transpires from his treatment of phantasia is that all visualization, rhetorical as well as poetic, amounts in varying degrees to a kind of imaginative creativity. How far the exercise of phantasia can legitimately diverge from the sphere of normal human possibilities depends on the aims of the particular piece of writing: thus it makes sense for a poet but not an orator to visualize Phaethon’s journey in the chariot of the Sun. Longinus’ thought moves here along two axes of judgment, one of realistic credibility and one of psychological compulsion. He is aware that they can conflict but also that they can be aligned, as he indicates elsewhere in the case of both the Iliad (9.13) and Sappho (10.1–3). Longinus’ priorities are not in doubt. The fact that his overriding concern is with transfigured mental states induced by the power of words brings into play a sense of fiction not as a separate logical or semantic category but as a quality of intense, world-projecting expressiveness in the processes of the imagination – a quality which remains ultimately if paradoxically compatible with what Longinus considers to be the essential “truth” or authenticity of feeling transmitted by the sublime (Halliwell 2011, 327–367).
As that last remark suggests, Longinus represents an extreme point on the spectrum of views I have been sampling in this chapter. But it is nonetheless a feature of the whole genealogy of ideas outlined by my argument that it displays an instability in the relationship between fictionality and truth, an instability ironically registered by Plutarch when he quotes the old, ambiguous Greek proverb “bards tell many falsehoods” and adds “in some cases intentionally, in others unintentionally.”23 All the evidence I have examined here is concerned with the psychological, cultural, and ethical implications of poetic and other imaginative texts whose narrative “worlds” (and their network of connections to history, myth, and “life”) cut across any simple dichotomy of truth and falsehood. Constructing a worthwhile dialogue between ancient and modern presuppositions in this area remains a major challenge. The question I have had in my sights throughout – “is there a Greek concept of fiction?” – may not admit of a definitive or uncontentious answer. But it is at any rate one useful means by which to engage with some of the most difficult problems raised by the literary aesthetics of antiquity.
I am grateful to audiences in Edinburgh and St Andrews for discussion of earlier versions of this chapter.
The most important publication on ancient ideas of fiction is the collection edited by Gill and Wiseman (1993): contributions – stimulatingly diverse in their assumptions – range across matters of literary theory and practice. Pratt (1993) detects elements of fictionality at various levels of Archaic Greek debates over truth and falsehood in poetry. Two substantial but very different approaches to the history of Greek awareness of fiction can be found in Finkelberg (1998) and Lowe (2000a, 2000b): Finkelberg sees “the poetics of fiction” emerging in the fifth century and being fully consolidated in Aristotle’s Poetics (see Rutherford 2000 for possible objections); Lowe assigns comedy a decisive role in breaking away from both mythological and historical material. Discussions of Plato which find a consciousness of fictionality in his work include Smith (1985, 27–32) and Janaway (1995; see his index); on the other side are Ferrari (1989) and Gill (1993). Halliwell (2002) pursues multiple connections between concepts of fiction (see the index) and theories of mimesis. Much of the evidence relevant to Hellenistic thinking on the subject, including the diffuse material of the scholia on various authors, is sifted by Meijering (1987, chs II–III); on the Homeric scholia see also Nünlist (2009). Using rather different premises from my own, Konstan (1998) constructs a lineage for Greek fictionality that runs from New Comedy to the novel. Laird (2007) weaves together reflections on ancient and modern ideas of fiction in relation to both literature and philosophy.