INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely afreet reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

A Bell & Howell Information Company 300 North Z eeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

ARISTOTLE’S POETICS :

ITS THEORETICAL FOUNDATIONS AND ITS RECEPTION IN HELLENISTIC

LITERARY THEORY

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of the Ohio State University

By

Poulheria Kyriakou, B.A., M.A.

*****

The Ohio State University

1995

Dissertation Committee:

A.J. Silverman

J.W. Allison Adviser

D.E. Hahm Department of Classics UMI Number: 9534013

UMI Microform 9534013 Copyright 1995, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103 Copyright by Poulheria Kyriakou ACKNOWLEDGMENTS

I wish to express my thanks to Prof. Allan Silverman for his help and advice throughout the preparation of this dissertation. Thanks also go to the other members fo my advisory committee, Profs. June Allison and David Hahm. I would also like to thank my parents and my husband. VITA

May 14, 1967 ...... Born -Volos, Greece

1989 ...... B.A, Aristotle University of Thessaloniki, Greece

1989 - 1990 ...... Visiting Fellow, University of Cologne, Germany

1990 - 1992...... Teaching Associate, Department of Classics, The Ohio State

University, Columbus, Ohio

1992 ...... M.A Classics, The Ohio State University, Columbus, Ohio

1992 -1994...... Teaching Associate, Department of Classics, The Ohio State

University, Columbus, Ohio

1994-1995...... Presidential Fellow, Department of Classics, The Ohio State

University, Columbus, Ohio

FIELDS OF STUDY

Major Field: Classics TABLE OF CONTENTS

ACKNOWLEDGMENTS...... ii

VITA...... iii

INTRODUCTION...... 1

CHAPTER PAGE

I. POET AND POETRY IN THE POETICS ...... 5

II. ARISTOTLE’S CONCEPT OF

ART AND THE POETICS ...... 39

III. NECESSITY AND PROBABILITY

IN THE POETICS ...... 70

IV. ARISTOTLE AND THE TRAGEDIANS...... 89

V. ARISTOTLE’S POETICS AND

HELLENISTIC LITERARY THEORY...... 106

BIBLIOGRAPHY...... 181 INTRODUCTION

The purpose of this thesis is to study the philosophical underpinnings of the

notion of mythos in Aristotle’s Poetics. Despite the emphasis Aristotle lays on

mythos, this task has never been seriously undertaken so far. By turning to Aristotle’s

metaphysics and his theory of science I will argue that his literary inquiries in the

Poetics, sui generis though they might seem, are firmly inscribed in the broader

context of his philosophy and that the concept ofmythos is the link of the Poetics to

this context. Mythos enables Aristotle to characterize poetics as an autonomous art

in his own technical sense of the term. Thus my purposes are fairly limited: the

thesis does not offer a new interpretation of thePoetics as a whole but rather focuses

on a single but essential aspect of the treatise. The starting point of my inquiry is

the realization that the Poetics occupies an almost unique place among Aristotle’s

surviving works: lacking the strong philosophical appeal of theAnalytics, Physics or

Metaphysics, it has been overlooked by historians of philosophy, whereas literary

scholars, naturally interested in the work because of its subject, examine the Poetics

in isolation from the rest of Aristotle’s philosophy. The implicit assumption here,

that the Poetics needs to be viewed within the broader context of Aristotle’s philosophy, is potentially open to objections, not least because modern literary

criticism is irrelevant to philosophical concerns. Such an attitude can be thought to

be supported by the fact that the Poetics by no means resembles either the tenth

book of Plato’s Republic or the end of the Phaedrus -in his discussion of poetry

nowhere does Aristotle explicitly fall back on notions central to his philosophy.

It is exactly the association with Plato, however, that warns against the

assumption that the Poetics lies outside Aristotle’s strictly philosophical interests.

Indeed, there is evidence that Aristotle addressed, and most probably rejected,

Plato’s views on poetry in his lost dialogueOn Poets ; Plato’s influence, moreover, is

felt throughout the Poetics, although he is nowhere mentioned explicitly, and certain

passages in the treatise attack Platonic views, most probably harkening back to On

Poets. But Aristotle could not have adequately undercut Plato’s views unless he

rejected the philosophical assumptions that informed Plato’s hostile attitude to

poetry. One is thus justified in looking for the specific philosophical foundations of

the Poetics that could have enabled Aristotle to do exactly that.

The question which is naturally raised here is how "philosophy" and

"philosophical" are to be properly construed in the context of thePoetics. Butcher1 saw in the Poetics a clearly demarcated "theory of poetry and fine art" but one can comfortably side with Bywater who dismissed "aesthetics" as a modern notion irrelevant to the Poetics2. Indeed, no Greek parallel to "fine art" can be found and

!Butcher (1907) ch. 8.

2Bywater (1909) ch. 7. there is abundant evidence that Aristotle does not distinguish between "art" and

"craft". For him rexvrj is a blanket term that covers not only arts like music and

crafts like housebuilding but also sciences like medicine and geometry. Now it turns

out that a clear grasp of Aristotle’s concept of art is indispensible to the appreciation

of certain aspects of the Poetics , not least because it speaks against the thesis of

Halliwell (1986) ch. 2 and 3, the most recent and widely acclaimed attempt to put

the Poetics in the larger context of Aristotle’s philosophy. My critique of Halliwell

occupies the first chapter. Halliwell correctly dismisses any claim of "aestheticism"

as irrelevant to the Poetics. Instead he proposes that Aristotle viewed poetry in terms

of what he calls "natural cultural teleology". As is suggested by the term teleology,

Halliwell’s argument bears directly on Aristotle’s account of the evolution of tragedy

and poetry in general. For Halliwell Aristotle presents poetry as a natural cultural

movement that transcends individual artists: on his interpretation, they become mere vessels of this impersonal natural cultural potential as it unfolds gradually and

operates teleologically through the artists in order to achieve its own actualization.

Moreover, by construing poetry in such objective, i.e. impersonal, terms, Aristotle

resolves the traditional Greek dichotomy between poetic craft and inspiration: he

does away with the latter and emphasizes the rational, teachable character of poetic

craft. According to Halliwell in this scheme the personal talents of the poets play

at most a subordinate role and do not affect significantly the progress of poetry which, as said above, transcends individual poets.

As I argue in my critique, Halliwell’s account is flawed because he loses sight of Aristotle’s concept of art. The most important problem lies in Halliwell’s

teleological characterization of poetry. Teleology is an intrinsic part of Aristotle’s

concept of art but it does not bear out Halliwell’s interpretation of poetry.

Throughout the Aristotelian corpus art as a potentiality is repeatedly said dependto

ontologically on the individual artist and, as Aristotle makes clear, it is the individual

artist who operates teleologically through this potentiality, not the other way around

as Halliwell has it. It is, moreover, wrong to claim that Aristotle paints a largely

impersonal history of poetry: in thePoetics Aristotle is aware of the contributions of

individual artists (it is especially his references to Homer that bear this out). On the

other hand it is impossible to argue, as Halliwell does, that Aristotle’s belief in the

recurrence of cultures and civilizations reflects a belief in an objective natural

potential of art, in this case poetry, that waits to be discovered in each cultural cycle.

Instead, as I show, the philosophical presuppositions of Aristotle’s belief in cultural

recurrence contravene Halliwell’s thesis. Finally, I argue that Aristotle could have

hardly attempted to resolve the tension between poetic craft and inspiration in any way, simply because this tension cannot be observed in the Greek attitude towards

poetry. Poets from Homer to Pindar and down to Callimachus viewed poetic craft

and inspiration, i.e. the divine influence of the Muses or Apollo, as one and the same

thing: unless divinely gifted, a poet does not possess technical skill and vice versa.

Halliwell, though, is right to insist that for Aristotle poetry is a rational

teachable art -Aristotle himself calls the art of poetics (noiriTiK.fi) a "method"

(h IQo Soq) already in the introduction to the treatise (Poet. 1447al2). As argued in the first chapter, Halliwell’s thesis rests on a misunderstanding of Aristotle’s concept

of art, a synonym of which is "method", and in the second chapter I set out to

delineate this concept and show how it informs key points of thePoetics. On many

an occasion Aristotle characterizes an art as the definition of its intended product:

this definition corresponds to the "genus" that delimits hard sciences like mathematics

and guarantees the autonomy of each art. What this definition picks out is theform

or essence of the product, its final cause and the final cause of the art itself. It thus

becomes apparent that the status of poetics as "method" or "art" depends directly on

the definition of tragedy. This explains why Aristotle gives, on the one hand, a

definition of tragedy’s essence (opOQ rrjc ouatac, 1449b23-24) in the Poetics and, on

the other, characterizes the mythos as the final cause (1450a23-24), principle and soul

of tragedy (1450a38-39). The first two characterizations are equivalent to "form" or

"essence" whereas soul is exactly the form of a natural substance, Aristotle’s paradigm

of definable unity or essence. This analogy with natural substances simply

emphasizes the metaphysical status of mythos which is required, if tragedy, the product of poetics, is to be definable so that Aristotle can speak of poetics as an

autonomous art in the technical sense of the term.

This interpretation sheds light on. the provenance of the terms "one", "whole" and "complete" that characterize the action equivalent to the mythos. These terms are usually interpreted in an aesthetic sense as pretaining to what literary critics call the "unity of a literary work". Evidence from theMetaphysics, however, shows that these terms characterize a definable unity, i.e. an essence, as such. A good case in point is Met. 1077a20-33 where Aristotle questions whether Platonist mathematical objects have the status of essence, i.e. definable unity. He denies that Platonist mathematical objects can be "one", "whole" and "complete" unless there is something analogous to the soul of a living organism to turn them into definable unities. This is a context as far remote from literature as possible but it features the same terms and the same analogy that are applied tomythos in the Poetics. Thus, it should be concluded, these technical terms are directly imported into the Poetics from

Aristotelian metaphysics.

As said above, characterizing the mythos as a definable unity or form serves to define poetics as an art in Aristotle’s sense. But there is more to that. A curious and largely unnoticed feature of thePoetics is that, apart from tragedy’s definition in Poet. 1450a23-28, Aristotle also gives a shorter definition of tragedy which picks out only the mythos as a complete and whole action, i.e. definable unity {Poet.

1450b23-25). Philosophically there is no problem here: the full definition is what

Aristotle calls t 6 r i io n , the full account of what something is, whereas the short one is to ri fiv e l v a t, i.e. the property that characterizes this something exclusively. Indeed, Aristotle emphasizes that a tragedy can be a tragedy by virtue of the mythos alone. Without it the other parts of tragedy which appear in its full definition, i.e. character, diction, thought, spectacle and music, cannot make a tragedy even if they are excellent. If, on the contrary, they are absent or bad, a tragedy is a tragedy on account of its mythos alone. This does not imply that Aristotle disregards all other parts exceptmythos -the point is simply that no other part(s) can be solely responsible for what makes a tragedy a tragedy. Aristotle makes clear that all other parts except mythos belong to arts other than poetics: diction belongs to acting, character and thought to rhetoric and spectacle to costume-making. As said above, the precise delineation of the provenance of an art or a science is one of the foundations of Aristotelian thought. Given Aristotle’s conception of art, unless the mythos holds sway over all other parts in that it belongs to tragedyper se, there is no tragedy and consequently no autonomous art of poetics: poetics is reduced to a branch of rhetoric or ethics and politics, depending on whether thought or character prevails. For Aristotle an art is to another art as matter to form if the former contributes to the final cause the latter. This captures nicely the relationship between the other parts of tragedy and mythos: since the mythos amounts to the art of poetics itself, the other parts being products of other arts contribute to its final cause but cannot amount to it. Consequently, Aristotle argues, correctness, in poetics has to be evaluated with respect to the art itself, not to the other subordinate arts: a fault in the rhetorical or ethical aspect of poetry is a fault from the point of view of rhetoric or ethics but it bears negatively on poetics only if it hinders, or does not contribute to, the realization of the final cause of poetics itself.

These are the lines along which Aristotle manages to answer Plato’s critique of poetry in theRepublic. There Plato attacked the correctness of poetry not as such but with respect to the correctness of other arts. But from Aristotle’s point of view this attack is both ineffectual and illegitimate. Since poetics is an autonomous art, its correctness must be determined intrinsically, not extrinsically. According to 8 Aristotle Plato violates a fundamental principle, the autonomy of individual arts and

sciences that determines a correctness peculiar to each one of them. It is a similar

argument based on the automomy of the sciences that had allowed Aristotle to reject

Plato’s ideal of a general science of being, an ideal which, to a large extent, informed

Plato’s hostile attitude to poetry. Indeed, the discussion in the second chapter shows

that what can be called Aristotle’s own approach to poetry is shaped by his broader philosophical concerns as much as Plato’s was.

The second chapter, it should be emphasized once more, does not imply that

Aristotle pushed aside all other parts of tragedy, like e.g. character, in favor of mythos -the Poetics offers abundant evidence that this is not the case. The upshot is rather that Aristotle focuses on the mythos in a specific theoretical sense that

distinguishes it. from all other parts of tragedy: the latter are as indispensible to

tragedy as bricks and mud to a house but Aristotle insists that the art of

housebuilding does not consist in bricks and mud but rather in the form or definition

of house. Similarly, Aristotle cannot talk about poetics as an art unless he possesses

a definition of its product, i.e. its form which is the mythos. But as the choice of the appropriate wood and mud contributes to the art of housebuilding, likewise character and thought are indispensible to the art of poetics. Construing all other parts of tragedy exceptmythos as the material component offers Aristotle a powerful means of accounting for the individuality of each tragedy: even if the structure ofmythos is similar, different choice of character, thought or diction explains the individuality of each work as the choice of wood contributes to the individuality of each building. For Aristotle, however, the important point is that, as excellent mud and wood

cannot by themselves determine the form of house and thus the art of housebuilding,

in the same fashion character and thought cannot determine the form of tragedy and

thus the art of poetics. The second chapter focuses only on this aspect of Aristotle’s

discussion in the Poetics , an aspect which has been undeservedly overlooked in the

secondary literature. There have appeared studies of character3, catharsis4 and even

diction5 but not enough attention has been paid to Aristotle’s characterization of the

mythos.

Probability and necessity, the subject of the third chapter, form an essential

part of this characterization since the two modals govern the structure of the mythos.

The questions raised here are, first, why Aristotle resorts to technical concepts which

are more at home in the Analytics than in the Poetics', second, why he couples

necessity with the related but milder notion of probability; and, third, what kind of

necessity he has in mind. Frede (1992), the only study devoted exclusively to the

issue at hand, has precluded the two commonest senses of necessity in the

Aristotelian corpus, namely apodeictic and hypothetical necessity, arguing instead that

the introduction of the modals in the Poetics is adequately explained by Aristotle’s

conception of the poet’s task: to present what falls outside ordinary human

experience as necessary or at least as probable. Frede supports her view with the

^ h e most extensive study is Schutrumpf (1970). See also Belfiore (1992) 83-110.

4For a disussion of the various interpretations of catharsis see White (1984).

5See Morpurgo-Tagliabue (1967) 21ff. 10 claim that Aristotle appreciates the irrational, the marvellous and surprise in tragedy

but, as I argue, the relevant passages from the Poetics do not bear out this claim.

The answer to the problem should rather be sought in Aristotle’s reference to the poetic impossibility or falsity, an aspect of thePoetics that has not received any

attention in the secondary literature. What is most interesting about Aristotle’s

discussion is that he employs logical terms from theAnalytics to account for the believability of poetic impossibility in terms of paralogism, i.e. an inference from

false premisses. Since false premisses can have a true consequence, the poet must

introduce this true consequence of a falsity so that this falsity may appear true and

thus necessary on account of the true consequence: schematically, if A -*• B and one knows that B is true, then A is believed to be true on account of B. Now Aristotle

is explicit that this technique involves hypothetical necessity in the first step, where,

assuming A, one searches for the means of validating it, and forward looking or

apodeictic necessity in the second, where B, whose truth is known, is assumed to lead necessarily to A. Thus in the Poetics both apodeictic and hypothetical necessity

appear to be complementary aspects of a double-pronged creative process.

This is certainly not accidental. In the Posterior Analytics the same logical

situation as in Aristotle’s account of poetic falsity applies to the geometric method

of analysis and synthesis: in analysis the mathematician assumes the truth of the demonstrandum and derives as series of consequences which are demonstrably true;

in synthesis this series of true consequences is reversed leading to the initial

assumption whose truth is now shown beyond doubt. But the mathematician, 11

Aristotle adds, might very well mistake a false demonstrandum for true exactly because it has true consequences. Thus the mathematician may well fall victim to the same fallacy that explains the poetic falsity according to Aristotle. The introduction of a logical situation that characterizes a geometric technique in the

Poetics might prima facie appear paradoxical. For Aristotle, however, this geometric technique bears directly on arts and crafts because in the Nicomachean Ethics and the Movement of Animals he draws an explicit analogy between a geometer solving a problem by means of analysis/synthesis and the creative process of artists or craftsmen: they posit the form of their intended product, then by hypothetical necessity obtain the preconditions for their goal and, finally, reversing these preconditions, perform all steps that lead to the actualization of this goal. Aristotle is explicit in his analogy -an artist or craftsman works like a geometer and a poet can obviously be no exception. Thus the evidence strongly suggests that necessity in the

Poetics reflects the peculiar way in which Aristotle conceptualizes the creative process. On the other hand, the presence of probability in this context is explained because for Aristotle the "proofs" carried out in "soft" sciences, unlike mathematics, cannot have strong claim to the demanding rigor of necessity. As Aristotle himself remarks, the "productive" premisses in these soft proofs are through the possible or probable, in obvious contrast to the necessary premisses of strict demonstrations.

From the above discussion there emerges a strong, and largely justified, impression that Aristotle’s treatise is much more prescriptive than descriptive.

Although in the Poetics Aristotle touches on many things, perhaps disproportionately 12

many to the size of the treatise, it has often been noted that the Poetics is no

practical manual to help actual poets with the composition of their plays. Its most

likely audience is the educated public or the literary critic who would be looking for

a solid framework in order to evaluate tragic plays. But these considerations,

accurate as they may be, often obscure the fact that Aristotle did not write the

Poetics in complete isolation from the actual poetic practice. He was not anex

cathedra theorist and his close familiarity with Greek poetry is clearly shown by

several quotations throughout his corpus and especially by his frequent references to

fifth and fourth century tragedies with which he illustrates his points in thePoetics.

In my fourth chapter I investigate this other side of Aristotle’s treatise and try to

explore the extent to which Aristotle’s remarks may overlap with his contemporary

poetic praxis. I focus on themythos again not only because of the general emphasis

of Aristotle’s treatise and my own thesis but also because ch. 17 of the Poetics deals

with the construction of a plausible mythos. This is a chapter written with an eye to

actual theatrical performances, since, after having presented his analysis of the

foundations of a good tragedy, Aristotle touches on the way a careful playwright

should go about actually composing a play: optimally the poet should visualize

everything in detail as a safeguard against oversights which can easily jeopardize their

success of the play in the theater.

I chose two tragedies, Sophocles’ Oedipous Tyrannus and Euripides’

Andromache, in order to show how relevant or irrelevant Aristotle’s remarks on the

mythos are to the composition and success of Greek dramas. The Oedipus Tyrannus 13 is of course Aristotle’s model tragedy: with a striking story, particularly apt to cause

pity and fear, a complex plot with a highly dramatic recognition and reversal, and a

tight sequence of events that leads, contra expectum but fully according to probability,

to a great resolution, the play leaves nothing to be desired from an Aristotelian, point

of view. On the contrary, the Andromache meets almost none of Aristotle’s

requirements for a good tragedy. First of all, the story is not very well-chosen because it portrays the totally unjustified suffering of a completely innocent woman.

Besides, it is a fusion of two unrelated stories, the prosecution of the slave

Andromache and her child by Hermione and the death of Neoptolemus at Delphi.

Euripides’ plot is simple without recognition and reversal and contains many

illogicalities and loose ends that further weaken it. Although it is not episodic, in the sense that events follow one another in completely disconnected and improbable fashion, at several points the improbability is thinly disguised under a semblance of logical coherence.

Although from the Aristotelian point of view the two plays are largely dissimilar, they have some similarities. Sophocles’ play is not altogether devoid of minor illogicalities and the Andromache, excoriated in the past, has recently been reconsidered and proclaimed as good a tragedy as any of Euripides’, i.e. as any surviving Greek tragedy, since Euripides is not considered, at least today, a lessser dramatist than the rest. A completely Aristotelian tragedy has never been written, not even the Oedipus Tyrannus can pass the rigorous test, and a tragedian or a play cannot be condemned solely on the basis of Aristotle’s Poetics. Despite the multitude 14 of literary examples, the treatise remains largely a theoretical investigation along the lines of Aristotle’s formal conception of art: the composition and success of actual tragedies in all their variegated complexity remained squarely with the talent of individual playwrights.

The issue of the relationship between literary theory and poetic practice suggests itself with more urgency if one turns from Aristotle to Hellenistic literary criticism. I deal with this diverse body of evidence in the fifth chapter. The attempt to make my way through this little charted and often bewildering territory has been prompted by a wish to throw some more light on the difficult question of Aristotle’s position in the history of Greek literary criticism. Since I have argued that a basic aim of Aristotle’s Poetics is to defend poetry against Plato’s attack, an inquiry into the reception of Aristotle’s approach is particularly interesting because it puts

Aristotle in a wider historical perspective. This project becomes more intriguing if one takes into account the repeated claims that have been made in the secondary literature: scholars have asserted that Aristotle’s treatise exerted a considerable influence on all subsequent literary criticism, often coupling such remarks with claims that the Poetics is a unique phenomenon in the history of Greek literary criticism.

I definitely agree with the second proposition but I think that the contradiction merits some consideration because it appears to be symptomatic of a larger problem, namely the assessment of Aristotle’s influence on later philosophy. It is undeniable that this influence has often been overemphasized in a rather uncritical way. The importance of Aristotle in the modern history of philosophy along with the 15 fragmentary state of the surviving Hellenistic evidence have contributed significantly

to the conviction that Aristotle exerted a pervasive influence on everybody that wrote

after him. This assumption has recently been reconsidered, especially in view of the

fact that at least some of Aristotle’s works seem not to have been widely read in the

Hellenistic period. The thorny problem of the availability of the entire Aristotelian

corpus is of course ever present and makes things even more difficult as it has to be

addressed on a case to case basis. Aristotelian influence is now viewed as possible

instead of certain and Aristotle’s work as one among many others, now largely lost,

sources with which Hellenistic philosophy could have interacted. Unfortunately, this

dose of caution necessary to counteract uncritical attitudes of the past can easily lead to the other extreme, i.e. the belief that Aristotle’s work was either not known at all or considered irrelevant by Hellenistic thinkers. It is indeed easy to assume an almost total absence of Aristotelian evidence but, since this is a negative conclusion, it is difficult to prove. The availability of Aristotle’s work cannot be plausibly ruled out. Besides, "influence" is a wide notion that ranges from direct borrowing and explicit interaction to taking over of hints and approaches or even total rejection for a new approach. It goes without saying that the easiest to detect is direct influence.

The more one moves towards the other end of the scale, the difficulty of assessment increases significantly.

The study of the influence of the Poetics on later criticism presents all the problems sketched above plus another one: no comprehensive treatise that presents a coherent account of Hellenistic theory of poetry has survived and the secondary 16 sources are few, often contradictory and in general of little help. Thus the task of the scholar becomes particularly onerous and definite conclusions are hard to reach.

The kind of influence previous scholars have mostly asserted, without adducing adequate evidence or based on flimsy connections, requires a thorough refutation that will clear the way to a more realistic appreciation of the reception of Aristotle’s work. This kind of detailed, argument by argument refutation is a huge task that falls outside of the limited scope of my thesis. In my last chapter, though, I touch on this issue by discussing a small sample of conspicuous problems with recent studies on the subject before I focus on Hellenistic literary theories. Despite the problematic nature of the texts involved, it appears fairly clear that Hellenistic critics did not share Aristotle’s concerns in the Poetics. The most conspicuous absentee is a counterpart of the Aristotelianmythos. For the Epicureans poetry was apparently a common notion, and thus hardly in need of a definition, a pleasant pastime that did not merit serious philosophical consideration and actually distracted from philosophical pursuits. The Stoics placed the major emphasis of their discussion on the linguistic aspect and the sound patterns of poetry as well as on its moral impact on the audience. Theophrastus himself, if the reconstruction of his lostPoetics is valid, seems to have followed lines that do not square well with Aristotle’s Poetics.

It is plausible of course that Theophrastus’ innovations were put in the context of an

Aristotelian type of inquiry and can be viewed as a reaction to what he considered his predecessor’s mistakes.

Similar hypotheses can be put forth for the rest of Hellenistic thinkers: since 17 it is plausible that they knew the Poetics, their apparent lack of interaction with it might well be the result of their dissatisfaction with Aristotle’s approach and their attempt to leave it behind. The purpose of my last chapter is not to discredit this tantalizing possibility but merely to make as clear as possible that the available evidence cannot substantiate claims about the direct and all-pervasive Aristotelian influence on post-Aristotelian literary criticism. Aristotle’s Poetics did not immediately become the gospel of literary criticism -this status it only attained in the

Renaissance. CHAPTER I

POET AND POETRY IN THE POETICS

Attempts to view the Poetics in the broader context of Aristotle’s philosophy

have been made in the past. Already in 1957 Frey conceived of Aristotle’s task in

the Poetics in terms of biology in which Aristotle had keen philosophical interest.

For Frye Aristotle’s attitude to poetry resembled that of a biologist who

would approach a system of organisms, picking out its genera and species, formulating the broad laws of literary experience, and in short writing as though he believed that there is a totally intelligible structure of knowledge attainable about poetry which is not poetry itself or the experience of it, but poetics.6

More recently Halliwell in hisAristotle’s Poetics7 contrasted Aristotle’s conception of the "mimetic technai" with "modern attitudes" in very similar terms:

Pleasure alone, therefore, will not make Aristotle’s view of the mimetic technai

6Frye (1957) 14.

TIalliwell (1986) ch. 2 and 3; henceforth cited as AP and page number.

18 19 comformable to modern attitudes. The primary reason for the remaining conceptual distance between the two is that Aristotle’s acceptance of the framework of techne for the interpretation of poetry and related practices imports aninescapable objectivist element, as well as a naturalistic teleology, which is alien to the belief in creative imagination that has grown in strength since the Renaissance, and thus has dominated Romantic and later aesthetic thinking. In Aristotle’s system, the mimetic artist is devoted to the realisation of aims which are determined independently of him by the natural development o f his art, and by the objective principles which emerge from this development (my italics).8

Halliwell’s conclusion relies on Aristotle’s undeniably close association of artificial and natural production from a teleological point of view and is part of a detailed discussion of Aristotle’s aesthetics that spans two chapters in Halliwell’s book. He also explicitly proclaims at the beginning of his book that he will resist the traditional separation of the Poetics from the rest of Aristotle’s philosophy9. Since Halliwell’s book has met with very favorable reception10 and has indeed been considered authoritative since its publication, one might think that my investigation is largely superfluous. But Halliwell’s programmatic claim is not borne out by his subsequent discussion whose focus is solely what he calls an "inescapable objectivist element" in the above quote: according to him individual poets were for Aristotle instruments of the natural potential of the art, picked because of their timing to act as the agents of a process that obeyed its own predetermined rules. Surprisingly, this view sounds very close to the Hegelian idea of the autonomy of the alienated human activity, a

*AP 51.

9AP 2-3.

10See for instance the review of Nehamas (1987). 20 parallel that warns against an all too easy acceptance of Halliwell’s theory as a thesis

that Aristotle could have put forth. Halliwell of course wraps this non-Aristotelian view in markedly Aristotelian terms like "potential", "nature" and "teleology" but it

is worthwhile, I think, to highlight the inaccuracies and misconceptions that inform

his thesis and have gone undetected by reviewers. This critique will show that a precise delineation of Aristotle’s philosophical apparatus in thePoetics is still a desideratum -the Poetics does require as careful a handling as any other "hard"

Aristotelian treatise. I will briefly sketch Aristotle’s account and Halliwell’s arguments before presenting my own views.

After the initial taxonomical remarks that take up the first three chapters of the Poetics Aristotle devotes the next two to a schematic overview of the origins of poetry and drama. Chapter 4, where Aristotle provides the outline of tragedy’s historical development, has been the subject of endless controversy because it has been viewed as unclear and even self-contradictory11. The crucial question is whether Aristotle used earlier sources or whether he largely presented his own conjectures as facts. In principle, of course, both alternatives are possible and the brevity of the chapter does not help at all with the identification of Aristotle’s sources, if any. Although the specifics about the origins of dramatic poetry do not

"For general accounts of the problems involved see Patzer (1962) and Else (1966); see also Pickard-Cambridge (19622) and Lesky (19642). 21 affect my argument, I should say that I definitely side with those who have voiced cautious reservations about the chances of formulating a theory more plausible, i.e. closer to historical truth, than Aristotle’s12. Whatever pre-Aristotelian accounts might have existed are irretrievably lost today. It is possible that Aristotle did not possess significantly more than is now extant -he himself states in chapter 5 that the beginnings of comedy and comic performances are unknown because they had not been recorded (Poet. 1449a38-b5). Nevertheless, Aristotle did possess a vastly greater amount of poetry than is available today and thus, even if he extrapolates, he could do so from a much more secure position than that of modern critics.

Whether one likes it or not, in a discussion of the origin of tragedy one has to go back to Aristotle and I do not think that there is sufficient reason to question the accuracy or, much less, the plausibility of his historical inquiries.

Aristotle stresses emphatically from the very beginning of chapter 4 that dramatic or mimetic poetry arose from two causes innate to human nature, the mimetic instinct and a feeling for melody and rhythm(Poet. 1448b3ff). The text can also be interpreted so that the two natural causes are the mimetic instinct and the enjoyment of mimetic interpretations. Either way, the origins of dramatic poetry are rooted in human nature and psychology. People who had a greater aptitude for melody and rhythm engaged in improvisations which gradually gave rise to poetry

(Poet. 1448b22-24). The demarcation of the different poetic genres took place in the next stage of poetry’s development according to the character of early practitioners:

12See e.g. Lesky (19642) and Lucas (1968) 79-80. 22 the serious composed heroic or encomiastic poetry whereas the more easy-going

types turned to iambic poetry(Poet. 1448b24-27). Homer, either first or in the

footsteps of nameless predecessors, appears as the precursor of tragedy with theIliad

and the Odyssey and of comedy with theMargites (Poet. 1448b28-1449a2). The next

step was taken with the embryonic emergence of these two genres, when serious

poets who previously composed heroic poetry took up tragedy because it offered

greater opportunities for solemnity and grandeur; composers of lampoons, on the

other hand, became comic playwrights (Poet. 1449a2-9).

So far one can follow Aristotle fairly easily, although the neatly linear

character of the scheme and the presentation of Homeric poetry as the precursor of

tragedy and comedy have been attacked by modern critics13. At this point, having

appeared to reach the end of his historical account, Aristotle resumes his discussion following a new thread, namely that tragedy and comedy developed out of dithyrambs and phallic processions respectively(Poet. 1449a9-14). Later on he mentions that the beginnings of tragedy are to be found in light satirical performances (Poet. 1449al9-

21). Despite the problems posed by the second scheme, what is of interest for my purposes here is Aristotle’s observation on the last stage of tragedy’s development: the genre underwent several changes (from which he lists only a few a little below and skips the rest with a reference to the difficulty of detailed enumeration) progressing gradually and it finally stopped when it attained its own nature(Poet.

1449al4-19).

13See e.g. AP 254-256. 23 As mentioned above, despite the vexed particulars, Aristotle’s explanatory scheme is fairly simple and lucid: human nature is responsible for the birth of poetry and the character of early poets for the division into serious and light poetry and finally the demarcation of the different dramatic genres. Halliwell, though, constructed an unnecessarily lengthy and implausible argument to support his claim that contributions of individual artists, and the totality of artistic contributions for that matter, are simply manifestations of the predetermined natural potential of poetry and tragedy. Now Halliwell’s thesis has a solid background. He correctly cautions against easy identifications of ancient and modern concepts of creativity and inspiration14. He also rightly stresses Aristotle’s conception of poetry and all art as a craft whose rational, teachable principles can eventually be systematized in a theory15. There can be no doubt that for Aristotle nature and art work in a similar way and teleology operates on all levels of production, both natural and artistic.

Nevertheless, there is nothing in these assumptions that can lead one to draw

Halliwell’s conclusions. Based on a number of passages in the Poetics where

Aristotle refers to the nature of the art16, Halliwell suggests that Aristotle presupposes a "natural potential of the art" which is actualizedthrough acts of predisposed individuals in a specific cultural context. Culture surprisingly plays a

UAP 82-83.

i5AP 56-59, 85 and 88.

16See especially Poet. 1449al4-15; see also 1449a23-24, 1451a9-10 and 1460a4-5. 24 fairly prominent role in Halliwell’s discussion17, although it is never explicitly

mentioned or even implied in the Poetics and it frankly does not even contribute

anything to Halliwell’s argument. His assumption that Aristotle’s emphasis on nature

reflects the distinction between "nature (physis), on the one hand, and tradition,,

cultural continuity and man-made convention (nomos ), on the other"18, is plainly gratuitous. In the Poetics culture is out of place, a redundant addition that strikes

an irrelevant contemporary note. Halliwell, for instance, paraphrases the "first appearance of tragedy" as "when it [tragedy] became a cultural possibility". Since in chapter 4 of the Poetics Aristotle stresses the natural causes or background of art, it is impossible to believe that he would concern himself with cultural parameters. As will be shown below, for Aristotle all peoples qua humans in all cultures would be capable of developing poetry and thus likely to do so, at least eventually. There is no denying that cultural parameters play a role in the development of arts and crafts but there is simply no mention of that in thePoetics.

Besides the non-Aristotelian references to culture and cultural changes, the basic tenet of Halliwell’s argument is that the development of poetry and art in general is not a result of "contingent human choices and tradition"19 but "the realisation of a somehow already existing natural potential of the art"20. This point

17See for instance AP 86, 87, 88, 91, 94, 96.

WAP 95.

»AP 49.

™AP 86. 25 is qualified by statements of the sort "Aristotle’s point need not be strictly

deterministic since Aristotle nowhere implies that tragedy was bound to be

discovered"21, although in the same breath Halliwell claims that Aristotle presents

the history of tragedy in terms of natural teleology directed into "acts of human

discovery of what was there to be found"22. Aristotle’s point is deterministic in the

sense that every potential is bound to be actualized23. Now, of course, the question

that arises is whether this potential is separable from the artists. As postulated by

Halliwell, the potential of art and poetiy is a very problematic notion. Even if one

accepted that this potential exists somewhere beyond and above the individuals and

their choices and that it operates through the artists24, the ontological status of this

21AP 49.

*AP 49.

23Allowing of course for unactualized possibilities in De Inter. 19al2-18.

^Halliwell explicitly rejects the importnace of individual artists in the following argument: "For Aristotle...the model for tecfine, which embraces all mimetic art, is a nature whose generative workings are regulated by the teleological realisation of form in matter. Creative imagination is inimical to tradition and sees the spring of art within the exceptional capacities of special individuals. Hence the importance in modern aesthetic attitudes of various notions of expression, usually centring around the idea of that which is brought forth from the mind of the artist and given new or unique form. While Aristotle can of course acknowledge the unusual abilities of certain individuals, above all Homer, and the importance of their contribution to the development of an art, this development itself is not only the inessential framework within which particular achievements must be placed, but the unfolding of a natural potential: it constitutes a large-scale dimension of teleology to arch over the small- scale teleology of individual mimetic works. Aristotle’s artist may be gifted, but his gifts are not unique or sui generis; they are at the objective service of his art, to be harnessed to the realisation of aims which have a potential existence that is independent of the individual (and which, as we saw, have been realised before and will be realised again). And if mimetic art can be said at all, in Aristotle’s scheme of things, to be expressive, it is certainly not expressive of the artist himself. For not 26 entity would be rather mysterious: what and where would it be and how would it

operate? Since Halliwell’s thesis hinges on art qua potentiality, it is easy to see that

his argument does not reflect genuine Aristotelian doctrine. Throughout the

Aristotelian corpus art is said to exist in the soul of the artist25 and it does so as an active potentiality 26 which determines the goal, i.e. the final cause, of a change. To

use Aristotle’s example in Met. Z.7, medical art is the form or definition of health27

whose knowledge allows the doctor to direct his efforts towards the appropriate end,

i.e. the restoration of health. This model brings together all threads in Halliwell’s

thesis showing clearly its untenability: not only is artqua potentiality ontologically

dependent on the individual but it is also the individual who manifestly operates teleologically through this potentiality, not the other way around28. Aristotle’s remark, moreover, that at the end of its evolution tragedy achieved it own nature

(eo%e t fjv avrfiQ o\v,Poet. 1449al4-15) cannot bear out Halliwell’s thesis. For

only is Aristotle’s concept of art objectivist, but its accent falls on the universal significance to which mimesis can attain: mimesis makes art outward-facing, and locates its subject in general human reality, not in the privileged inner experiences of the artist" (AP 60-61).

“See Phys. 192b9ff., Met. 1025b22-23, 1032a32-b2 and 23, 1034a24, 1070a7, GC 335b31-33, GA 735a2-3, HA 588a29 and EN 1140al0-14.

“On actives potentialities see Gill (1989) 200ff.

^ o r Aristotle’s definition of an art as the form of its product see GA 735a2, 740b28, Met. 1032a32-b2, 1032bl3-14, 1034a24 and 1070b33.

“For active potentialities as instrumental causes see Gill (1989) 200. Halliwell himself admits that the existence of artefacts depends on their maker (AP 46-47) but obviously fails to realize that this is incompatible with the art’s potentiality he postulates as separate from the artist. 27 Halliwell Aristotle implies that tragedy reached the "fulfillment of its potential", a

"naturally fixed goal"29, and consequently he must construe nature as the "end of

production"(rikoz rrjc Ycv^oreo)c)inthe definition of nature as Met. A.4(1015al0-

19)3° Xo fall on Met. Z.7 once more, two avenues are open in Aristotelian terms:

tragedy is either a natural substance having the origin of productive, i.e. generative

motion in itself (cf. Met. 1015al0-19), which is absurd31, or an active potentiality.

In the latter case it has to depend ontologically on the individual makers, contrary

to Halliwell. Aristotle of course could have only intended "nature" inPoet. 1449al4-

15 as an active potentiality, the crystallization of tragedy’s form and thus tragic art

itself as said above. But elsewhere Aristotle characterizes this crystallization in

explicit empiricist terms as a process depending solely on humans:

So from perception results memory, as we say, and from memory there comes experience when memory of the same thing occurs often; because many memories create a single experience. From experience or the universal, when it has sunk in the soul from all these things, the one over the many which is the same for all of them, there comes a principle of art and science(AnPost. 100a3-8).

As will be shown in the next chapter, the "principle of art" captures the art itself as form of its product, the active potentiality located in the soul of the individual maker.

aAP 49.

“On Aristotle’s concept of nature see Waterlow (1982) ch. 1 and 2.

31Kullmann (1980) 434-441 has argued against the view that tragedy is a true, i.e. primary, substance. 28 The development of art in the Aristotelian sense is similarly dependent on individual

human artists and their psychological apparatus -they are by no means vessels for

Halliwell’s impersonal natural potential of the art that mysteriously operates from

above.

Halliwell finds it implausible that Aristotle would be so sketchy in his account

of the stages in tragedy’s evolution, if individual contributions of poets like Aeschylus

really mattered32. But it is obvious that they do matter since they constitute the

most important landmarks. Aristotle simply skips the particulars because thePoetics

is not a historical treatise: he states that it would be an onerous task to detail every

single development in the evolution of tragedy {Poet. 1449a28-31), presumably

because it would disrupt the continuity of the treatise. But such glossing over does

not make Aristotle’s discussion "ahistorical" as Halliwell claims33. Even Halliwell

himself admits, in another largely unsuccessful attempt to mitigate his thesis, that

Aristotle’s perspective is not totally "^historical" and that "in other fields Aristotle

sometimes acknowledges the importance of individuals in historical development"34.

But he thinks that in Poetics 4 and 5 "the recognition of the importance of Homer

or Sophocles is outweighed by the fact that tragedy can in the final resort be made the proper subject of its own evolution"35. Nevertheless, it is exactly Aristotle’s

32AP 94.

^AP 95.

MAP 96 n.22.

XAP 96. 29 attitude towards Homer that subverts Halliwell’s thesis.

Aristotle’s reverence for Homer and his poetry is well-known. He had

devoted a whole work, the Homeric Problems, to the discussion and exegesis of

difficult passages in the Homeric epic. Some of the solutions he suggested found

their way into the Homeric scholia, most probably through the commentaries and

editions of Alexandrian critics like Aristarchus36. But for a modern reader the only

extant Aristotelian comments on Homer’s poetic merits are to be found in the

Poetics. Despite the overall dry style of the treatise and the very limited instances

of laudatory exuberance in it, Homer has the greatest share in Aristotle’s most

flattering comments. Homer’s excellence is not limited to one area, the ability to

construct unified poems out of the maze of events in the Trojan war story and the

nostos of Odysseus (Poet. 1451a23-30, 1459a30-37). He also excelled in mimetic

representation (Poet. 1460a5-ll), the portrayal of character and thought (Poet.

1459bl6), the effect of ekplexis (Poet. 1460b5-6) and, not least, in being the precursor

of tragedy and comedy(Poet. 1448b34-38) Homer is called "divine" (Qeonioioz,

Poet. 1459a30), "an expression of unusual enthusiasm for Aristotle"37. A hapax

legomenon in the Aristotelian corpus, this poetic word comes directly from the

traditional vocabulary of inspiration38 and leaves no doubt about Aristotle’s esteem

“See Pfeiffer (1968) 67-71.

^ u c a s (1968) 216.

“See Homer II. 2.600, Od. 12.158, Pindar I. 4.39, N. 9.7-8, Plato Euthd. 289e4, Rep. 365b7, 558al and Theaet. 151b6. 30 for Homer. Aristotle attributes Homer’s achievement either to his natural talent or his art (r)r0 1 6 i& ri%vr)v fj Sla 6ctiv, Poet. 1451a24). This does not imply, though, as Halliwell believes, that Homer’s nature was not really exceptional since art."could achieve the same results"39 anyway. This is a crooked way of interpreting

Aristotle’s point. In Aristotelian terms nobody would dispute that an artist possessing the full apparatus of art would be able to achieve the same impressive results as

Homer. Nevertheless, instead of trivializing Homer’s worth this observation can only emphasize it: although he composed very early on, before the crystallization of the rules of the art, Homer’s genius was able to discern the correct parameters and apply them to his poetry. Aristotle does not seek to present Homer as some kind of 19th century Romantic genius and it is not helpful or relevant to bring such categories into focus as Halliwell does40. Homer’s genius does not reside in the absolute, unrepeatable uniqueness of his achievement but in its timing: he was an accomplished artist before there was any art because, thanks to his exceptional nature, he had grasped the principles of the art well before anyone else did. Art could definitely achieve the same results but not in Homer’s time. He was the one who created the art virtually ex nihilo -at least he was the first eponymous poet known to Aristotle to compose artistically accomplished works. Far from suggesting, as Halliwell thinks, "an underlying alignment between human potential for mimetic

XAP 87.

“AP 87. 31 invention and the natural cultural movement towards the realisation of generic goals"41, Aristotle stresses Homer’s unique contributions which sealed once and for all the form of dramatic poetry. It is true, of course, that for Aristotle the dramatic mode would eventually be developed since it was clearly within the reach of humans as such. But this does not diminish the value of Homer’s contributions or those of later poets. In Aristotelian terms, if gifted individuals were not active, then one would still in Aristotle’s day listen to long, non-unified epics, and attend performances of dramatic improvisations.

In Poet. 1453al7-22 and again in 1454a9-13 Aristotle specifies that in the early stages of tragedy’s development the poets chose themes from'a wide variety of myths

(or families, as Aristotle puts it) but eventually they came to concentrate on a few heroic lineages whose stories provided suitable material for the best kind of tragedies. According to Poet. 1454a9-13 the poets hit upon this limited set of mythological stories not because of art (r exv rj) but by chance (rC%r?) as they were searching, obviously for the best kind of mythos, as is clear from the context {Poet. ch. 14). Halliwell considers this passage one of the strongest pieces of evidence for his thesis that personal inventiveness and insight rank very low in Aristotle’s theory since the early tragedians’ successes are explicitly attributed to chance. For Halliwell these successes were "ordinary random activity" and manifestations of the

"materialisation of tragedy’s generic potential"42 which works through rather passive

AlAP 91.

42AP 88. 32 and unknowledgeable individuals. First of all, Halliwell erroneously associates the

success of the early poets with the first level of the tripartite scheme he deduces from

Poet. 1447a20 and the beginning of the Rhetoric (1354al-ll). In the Poetics Aristotle

says that visual artists represent a multitude of things either by art (6 idT i%v?ic) or

habitually (6id aovrjflctac), whereas in the Rhetoric he specifies that all people

engage in rhetoric either by habituation (5id auvrjOelac, 1354a7,9) or randomly

(eiidj, 1354a6, a n d t o u avron& T O v, 1354al0). From these passages Halliwell

infers that one can postulate a three-stage process for the development of arts and

crafts:

Ordinary indiscriminate activity which will achieve its aims, if at all, only by chance; regularised experience or habit, which develops consistencies of procedure that nevertheless fall short of art; and techne itself, in which a self-conscious and rational understanding of the subject establishes secure techniques for success in it.43

A little later on Halliwell asserts that this tripartite scheme is implied in Aristotle’s

reference to the successes of early tragedians inPoet. 1454a9-13:

The original choices of tragic material represent ordinary random activity, whose occasional success would be due only to chance; out of this emerges the gradual regularisation of experience, which may indeed be the level at which some tragedians still work; and, finally, the discoveries of the genre’s natural potential which experiment has brought to light can now be rationalised into the art of which the Poetics is the theoretical embodiment. The underlying teleology, operating through the stages of cultural evolution, is not difficult to discern: the primary sources of poetic art are located, once again, beyond the individual, and even, in a

«AP 85. 33 sense...beyond strictly human tradition.44

The reference to a potential which makes discoveries that are in turn brought to light by experiment is puzzling, to say the least. But, more importantly, it is obviously wrong and misleading to relegate the activity and success of the early tragedians to the status of "ordinary random activity". If by "ordinary" Halliwell means "practised by all, or many, people", this clearly does not hold in the case of tragedy. Halliwell here is misled by Rhet. 1354al-ll, one of the passages on which he bases his threefold scheme. Rhetoric, as Aristotle himself makes clear at the very introduction to the Rhetoric, is a peculiar discipline in that it belongs to all branches of theoretical knowledge and everybody engages in it to some degree succeeding, as mentioned above, either randomly (eiicjj or &ird tov ctbiop&Tov) or by habituation. But, given the development of tragedy delineated in the Poetics, only the mimetic improvisations that preceded the emergence of dramatic art, if not the mimetic activities of babies and children, can be said to fall under Halliwell’s category of

"ordinary random activity": because inPoet. 1453al7-22, and thus 1454a9-13 as well,

Aristotle refers to poets, not to the individuals in Poet. 1448b20-27 who, mimetically gifted by their nature, engaged in pre-poetic improvisations. Halliwell himself admits that Poet. 1447a20 refers to what he calls the latter two stages in the successful practising of the visual arts, the first stage "being...beneath consideration"45. But

MAP 88.

“A P85 . 34 when it comes to tragedy he inexplicably makes the early tragedians not the

counterpart of the visual artists who succeed by habituation but the likes of the

hypothetical man-in-the-street who engages in haphazard mimetic performances. If

the early tragedians, who were apparently only one step behind the fully developed

form of tragedy as art, represent the first stage of Halliwell’s scheme, then virtually

nothing is left for the second or, much less, the third stage. On this assumption the

whole history of tragedy up to the point of the admittedly crucial but nevertheless

isolated step of adopting the best mythic material is classified as non-artistic activity.

Halliwell’s account of chance (rUxn) in Poet. 1454al0-13 and its role in the

selection of the appropriate tragic themes is also deeply problematic. In the context

of his tripartite scheme for the evolution of arts Halliwell assumes that Aristotle treats chance (rfixrj) and the automatic (auro/iarov) in Rhet. 1354al0 as identical

notions. In his discussion of Poet. 1454al0-12 he adduces two passages,Met. 1034a9-

10 and EN 1140al7-20 as evidence in support of his views on the relation between art and chance46. But, although in EN 1140al7-20 Aristotle states that art and chance (rCxh) are about the same things, in Met. 1034a9-10, the beginning of chapter

9 of Met. Z, Aristotle discusses things that can be produced both by skill and automatically (real rixvij Kal 6tnb ravTOn&TOv), like health, comparing them to things produced only by skill like a house. There is, of course, no mention of chance

(rCxn) in this chapter and thus Halliwell’s association is eo ipso annulled. In Met.

Z.9 Aristotle talks about automatic natural production, a cause and concept different

46AP 88 n.9. 35 from chance (rfixn)- Phys. B.4-6 are devoted to the discussion of these two causes

which

differ in that the automatic is a notion broader than chance because every result of chance is also due to the automatic but not vice versa. For chance and its outcome is about those which can succeed and involve rational action (rrpa^ic). Therefore chance is necessarily about things that can be achieved by rational action...so that those who cannot engage in such action cannot bring about anything by chance...And thus neither inanimate objects nor animals nor children can bring about anything by chance because they do not have rational choice (7rpoatpeaic)...The automatic applies to animals other than humans and inanimate objects, like a horse came automatically, as we say, and it was saved because it came but it did not come in order to be saved...Clearly, then, we say that something happens automatically in those things which come to be for a purpose when it does not happen for the outcome’s sake and has an external cause; and what can be object of choice and happens automatically to those that have choice occurs by chance(Phys. 197a36-b22).

It is clear that Aristotle is meticulous in his demarcation of chance and the

automatic, which applies to natural things and never involves choice (w poatpeaic).

It is beyond any doubt that, first, chance cannot and should not be equated or confused with the automatic as in Halliwell’s discussion. This confusion of course serves Halliwell’s thesis: by reading Poet. 1454al0-13 in light of Met. Z.9 he understands the successful tragedies of early poets as automatic productions that could also be brought about by skill, like health, but nevertheless suggest the independence of the art’s potential from individual artists. It is absolutely impossible for poems to be also produced automatically like health since they definitely resemble houses and other artefacts which, unlike health, have their beginnings only within the soul of the artist, where their form resides, and not in themselves. Indeed, 36 health, i.e. the art of medicine itself according to Aristotle, seems to be an exception

to the Aristotelian conception of art and should not be equated with either poetry

or any other art. This is clear fromEN 1140al0-16, immediately before the passage

Halliwell quotes in support of his theory about art and chance:

All art is concerned with coming into being, i.e. with contriving and considering how something may come into being which is capable of either being or not being, and whose origin is in the maker and not in the thing made', for art is concerned neither with things that are, or come into being, by necessity, nor with things that do so in accordance with nature (since these have their origin in themselves).

In view of this characterization Aristotle rightly attributes to chance the production of the first artistically accomplished tragedies in the context of the poet’s search for mythoi. Since art and chance are about the same things, the only avenue for success opened to early tragedians was chance: they apparently lacked a unique talent like

Homer’s and they could not rely on fully developed and crystallized artistic rules.

Nevertheless, they were purposejully looking for a mythos, i.e. they were exercising deliberate choice (7rpoaCpeaic) which is the presupposition of chance according to

Phys. 197a36-b22. They were not looking for the best kind of mythos and the fact that they found the appropriate sagas could only be called an accident (cf.Phys.

196b23-197a8). Thus Poet. 1454al0-13 can by no means support Halliwell’s contention: it is clearly impossible to do away with the poets by reducing them to the status of virtual puppets through which operates the potential of tragedy "towards the realisation of generic goals". Halliwell absurdly equates the generic potential with 37 chance and, even worse, chance with the automatic. Even if there existed the natural potential Halliwell postulates, it would still have nothing to do with the automatic because Aristotle specifies that in the domain of natural things when something unnatural takes place people tend to attribute it to the automatic; but this, Aristotle says, is inaccurate because the cause of an automatic occurrence is external, not internal as is the case with natural substances (Phys. 197b32-37).

As another proof of the existence of the art’s natural potential Halliwell mentions Aristotle’s reference to the cycles of artistic and scientific blossoming and withering (Met. 1074bl0-14):

While it seems that each art and science has often been developed as far as possible and has again perished, these opinions have been preserved like relics until the present. Only thus far, then, is the opinion of our ancestors and our earliest predecessors clear to us (cf. DC 270b 19-20, Pol. 1329b25-35, Meteor. 339b27).

Halliwell seems to be on the right track when he says that the Aristotelian cycles occur because arts and sciences "are rooted in man’s inescapable relation to nature"47. Relevant to this issue are also Aristotle’s remarks in Pol. 1264al-5:

We are bound to pay some regard to the long past and the passage of years... Almost everything has been discovered, though some of the things discovered have not been coordinated, and some, though known, have not been put into practice.

i7AP 50. 38 Aristotle takes up the same issue in Pol. 1329b25-35:

We must also believe that...most other institutions...have been invented in the course of years on a number of different occasions -indeed an infinite number... We ought to take over and use what has already been adequately expressed before us, and confine ourselves to attempting to discover what has hitherto been omitted.

Halliwell’s belief in the objective nature of poetry that keeps being discovered and lost subverts his theory. Nothing in Aristotle’s texts implies the discovery of something that exists and becomes periodically a cultural possibility. Halliwell consistently interprets the Aristotelian eupeiv as "discover" but this interpretation is not cogent. Although Greek does not differentiate between "invent" and "discover", this should not be a source of confusion: Aristotle apparently proposes that in the successive circles of civilization people reached the same or similar conclusions about natural phenomena and formed approximately the same social institutions because arts and sciences are wholly dependent upon human capabilities. Now it is one thing to explain the successive growth and decline of arts and sciences by a recourse to

Aristotelian principles and quite another to claim in Hegelian fashion that arts and sciences possess an all-powerful, autonomous nature which dictates their own recurrence. If Aristotle conceived of the possibility for the development of art as a capacity based possibility, i.e. a possibility that a thing has a certain capacityqua belonging to a natural kind, then, since humans as a species, and thus their capacities, are eternal, they do not change from circle to circle and, therefore, the same things are bound to recur: on account of the "principle of plenitude" there will 39 be individuals in each circle who will exercise their capacity for developing an art so that this specific capasity based possibility will be actualized48. If this is a plausible account of the recurrence of culture and civilization in Aristotelian terms, it turns out once more that what matters in the development of arts are the individuals that actualize numerically the specific capacities. Since arts and sciences are contingent upon the make-up of human soul, the historical development of each art and science within a given circle is determined by the contributions, failures and insights of individual artists and scientists as each civilization or generation develops arts and sciences to different degrees. Thus Aristotle says, for instance, that the ancients correctly understood the divine nature of stars but clothed this insight in a cocoon of myths and legends for political purposes(Met. 1074a38-10). Or, Anaxagoras understood correctly the properties of aether and the right name was given to this substance but others held silly beliefs that could in Aristotle’s time easily be discarded by the advances of mathematical knowledge ( Meteor. 339b 17-34).

Moreover, at the end of the Topics Aristotle proudly mentions the creation of logic as his original contribution to philosophy. He would hardly have considered himself a mere vessel of a "natural potential" whose workings would be fortuitously aligned with his personal abilities. Since, moreover, he remarks that the immense advance

““I follow the formulation of the "principle of plenitude" in van Rijen (1989) 95- 96: "If a is a perishable object of a certain natural kind N, and it is a possibility based on a natural capacity of things of kind N that a is F, then there is an object b of kind N such that b is F at least once". 40 of mathematical knowledge in his time started from humble origins49, it is patent

that the development of poetry is not necessarily bound to be significantly advanced

within a generation, or perhaps even a civilization, depending upon the circumstances

and the individuals involved.

Halliwell closes the first part of his chapter 3 with a discussion of the role of

inspiration in the Poeticsw. According to him Aristotle answers the question of "the

traditional dichotomy between craft and inspiration51" by totally doing away with the

irrational, religiously loaded and thus problematic, especially in Platonic terms,

notion of inspiration and stressing the rationally understandable and teachable

character of poetic craft. Thus personal talents, although mentioned in thePoetics,

play a clearly subordinate role, being a kind of happy coincidence that do not

significantly affect the progress or shape of poetry. First of all, there was hardly any

traditional dichotomy of the sort Halliwell refers to. Traditionally all poets were'

thought to have been taught by the Muses or Apollo and it is methodologically unjustifiable to project the modern metaphor of "inspiration" as far back as Homer.

The Muses imparted their superior knowledge on their favorite poets who reported

it to the rest of humankind in the form of beautiful song52, much like a prophet interprets the wishes of a god to laymen. The poet’s craft was not, and should not

49See Protr. fr.8 (Ross); cf. fr.5.

XAP 88-92.

5lAP 91.

52Cf. Walsh (1986) 10-11. 41 be, conceived separately from the inspiration of the Muses. The poet who enjoyed

their favor became a superior craftsman and, vice versa, the poet who enchanted with

beguiling, crafty song was considered the darling of the Muses53. The distinction

between ars and ingenium does not have anything to do with the traditional Greek

notion of inspiration. From Homer through Hesiod, Pindar, Aeschylus and down to

Apollonius or Callimachus, there appears no distinction or dichotomy such as

Halliwell imagines.

In Pindar’s odes for instance one finds contemptuous comments concerning the poet’s unworthy rivals who cannot match or even approximate the quality of his own poems54. Never one to downplay his superiority, Pindar implies that a victor unfortunate enough to have commissioned one of his rivals to compose an epinician ode will certainly spoil his chances of immortality in song, i.e. his only ticket to the only eternal hall of fame. The rivals are bound to compose a truly inferior song that will obscure the victor’s success and instead of functioning as an encomium it will end up resembling a lampoon. It is no coincidence that Pindar consistently associates these ineffectual poetasters with the athlete’s envious fellow citizens who begrudge his success and assiduously try to belittle it in the eyes of the community and ultimately later generations. Whether spreading pernicious gossip or composing bad encomiastic poetry, the result is the same: both malevolent citizens and poetasters

“On Greek ideas of inspiration see Maehler (1963) esp. 10ff., Murray (1981) and Walsh (1986) 45ff.

“See especially O. 2.86ff., 9.100ff. and N. 3.80-84. On the end of O. 2 see Kirkwood (1981), Most (1986) and Verdenius (1989). 42 harm the victor, at least work on it, either intentionally or incidentally. The reason

Pindar gives for the flagrant failure of his rivals is their total lack of inborn, divinely procured poetic nature (0ud)ss. He himself poses as the counterpart of the victor: both are aristocratic gentlemen enjoying the favor of divinities who essentially share the same code of honor and appropriately reward their devotees with great success and eternal fame56. On the contrary, Pindar’s rivals have nothing else to rely upon but a false, acquired dexterity. Those who partake of the aristocratic ideals and culture since their birth are bound to triumph over the rest who vainly endeavor to make up for their disadvantage by learning skills they erroneously consider an adequate substitute for inborn qualities.

It is obvious that Pindar builds on the distinction ars-ingenium but not in the way Halliwell believes. No ancient Greek ever considered inspiration and art alternative or equivalent sources of poetic production. Pindar for one squarely denies the "learners" the possibility of functioning as poets. His uncompromising belief in the superiority of his poetry leaves absolutely no room for alternatives. Pindar of course claims re%va or ao

“In Pindar 0ud is responsible for all kinds of excellence, military, athletic and artistic; see O. 2.86ff., 9.100ff., P. 8.44, N. 1.25 and 3.40-42.

^This parallelism is most often cast in terms of the guest-friendship relationship of poet and victor (or commissioner of the ode) and the athletic metaphors employed for Pindar’s poetry; less common is the presentation of the victor as a poet or musician that sings the glories of his country. On athletics and £evCa see Hubbard (1985) 152-158. On £evia in general see Gundert (1935) 30ff., Gentili (1965) 80-88, Crotty (1982) 76ff. and Kurke (1991) 137ff; on athletic metaphors see Simpson (1969) and Lefkowitz (1984). 43 presented as diligent and hardworking besides being scions of families and citizens of states with illustrious and often divine ancestors and founders. The athletes trained hard investing both energy and money to hire the best coach in the arduous preparation for the athletic contest. Likewise Pindar takes diligent care, i.e. summons all his t£%va to put together an elaborate song, a wondrous artefact that might adequately commemorate in artistically blameless form the athletic victory.

He does not hold the childish assumption that an aristocratic boy will succeed in the pursuit(s) he will eventually pick by virtue of birth alone. One definitely needs time and training in order to achieve excellence. Nature and art are viewed as complementary, the latter building on, and bringing into fruition, the potential of the former57. What is important for my purposes here is Pindar’s belief that art or learning alone cannot achieve anything; neither can of course nature alone but it can seek art’s help whereas art without nature cannot proceed anywhere. Both nature and art are patronized and alloted by gods. Those who succeed are blessed with high birth and an ability to learn skills which the gods themselves possess to the highest degree.

Exactly the same attitude is expressed two centuries later by Callimachus in particularly explicit form. InAet. fr. 1 Pf he responds to the accusations of his poetic rivals who have found fault with both his generic and stylistic choices. For his defense the poet relies primarily on the favor of the Muses and Apollo. His rivals,

^On the polarityTelchines, are not beloved by the Muses, he claims, while he himself has enjoyed their favor since his childhood and received in his youth explicit instruction from

Apollo himself to embark on the poetic career he has consistently pursued since then. He excoriates his envious malevolent critics and urges them to measure poetic wisdom (ao

Even Plato can be easily accomodated in that scheme. Halliwell rightly observes that he gave a twist to the notion of inspiration that presented what the poets considered the supreme guarantee of their excellence as lack of knowledge and rational skill. But, independently from Plato’s negative epistemological appreciation of poetry, it is doubtful whether anybody, especially any poet for that matter, would take serious exception to his thesis. In the Ion, the Phaedrus and the Apology

Socrates does not deny the exquisite beauty of poetry and, in thePhaedrus (245a3-5), even its ability to educate future generations by preserving the glorious deeds of old.

To call poetry madness would not be demeaning to Greek eyes. It can be no coincidence that immediately before this passage Socrates elaborated on prophecy

(Phdr. 244b-e), the most obvious form of divine inspiration and possession. Poetry and prophecy are inextricably connected in Greek poetic tradition from Pindar and

“For this passage see the discussion of Hutchinson (1988) 78-84. 45 Aeschylus to Apollonius and Theocritus*. To stress the irrational character of

poetic creation does not affect the status of poetryper se. It should be remembered

that "irrational" in this case equals divine and one would have to immerse oneself in

Plato’s theory of poetry to grasp the full implications of his pronouncements. On

their face they are innocuous and perfectly in line with the traditional Greek idea of

poetry which contemptuously dismissed uninspired poets. It is definitely important

to note that, although in the Ion and the Phaedrus Plato questions the foundations

of this tradition and pronounces them shallow, he does it fromwithin. He apparently

does not feel the need to reject the construct and propose something totally different

-on the contrary, in a way he reaffirms it.

Democritus is often cited as having influenced Plato’s view of poetry and

inspiration*. Only two very short Democritean pronouncements on poetry have

survived61. These two fragments have generated much discussion not only because

of their possible influence on Plato but primarily because Democritus is the first

*See Comford (1952) 80-81; cf. Farnell (1909) 425, Fraenkel (1950) 543, Dodds (1951) 80ff., West (1966) 166 and Roloff (1970) 117-121.

“The association goes back to Clemes Strom. 6.168 who discusses DK B 18 and cites a passage from Plato’sIon immediately afterwards, both possibly borrowed from an anthology. See Flashar (1958) 56 n.4; see also Frank (1923) n.344, Delatte (1934) 59ff., Koster (1970) 31, Russell (1981) 74-75 and Halliwell (1986) 83. For qualifications of this problematic thesis see Flashar (1958) 56ff., with the bibliography cited there, and the discussion below.

61noir}Tt)Q S i &ooa piv &vyp&

"OjLtrjpoc 0fiaeo)c A.a%wv Geafotiarjc iniutv Kdopov ireKTfivaTorravToi^v (B 20a DK). 46 surviving non-poet who wrote on poetry. The first fragment has been considered, at

least by the Romans, as a manifesto of manic poetry: sanity produces inferior

poetry62. Modern critics have detected in it the origins of the polarity of irrational

inspiration and rational skill which, although not on a par, are responsible for poetic production. This interpretation primarily relies on the particlefiiv: in the now lost

S e clause that must have followed Democritus must have said that poems written without enthusiasm and divine afflatus count for nothing. It is not an unreasonable

assumption but it is not the only one possible. The sentence could very well have continued otherwise. The current intepretation implies that according to Democritus the same poet enjoyed erratic installments of inspiration, a clearly problematic thesis.

If Democritus continued with an unfavorable comparison, he is very likely to have compared not works of the same author but inspired poets and ungifted poetasters.

The text could have continued along these lines: "but whatever is written by poetasters is a failure because they lack enthusiasm". This alternative would certainly be in accord with Greek views on inspiration up to Democritus’ time. Even if

Democritus distinguished between inspired and uninspired poems of the same poet - an idea incidentally contradicted by the accounts of Horace and Cicero, again that could hardly be interpreted as attributing to skill the status of poetic source, even an inferior one. This is an inference foisted on Democritus without support from direct or indirect sources.

“See Cicero Pro Archia 18 and Horace Ars Poet. 2.1.26; cf. Cicero Tusc. 1.64 and Seneca De Tranq. 17.10. 47 Although it cannot be determined how exactly Democritus’ theory of poetry

fit into his larger philosophical system63, there is no cogent reason to assume that

he was in disagreement with the poets concerning poetic production. This is borne

out by his fragment on Homer’s nature and achievement (quoted above in n.61):

"Endowed with a divine nature, Homer fashioned a universe of all sorts of words".

Democritus here says that Homer enjoyed access to the divine realm, most probably

of the Muses, or that, prone to enthusiasm, he was able to succeed. Now Russell has

questioned whether Homer’s unspecified relation to the divine should necessarily be

viewed as the cause of the poet’s success, in other words whether the participle

"endorsed" (Aax&v) should be taken as causal or merely circumstantial. The issue

is of course impossible to decide with certainty but the likelihood of the participle

being causal is particularly strong. How plausible is it to take the first part of the

sentence as having an import similar to e.g. "Homer, born the son of a carpenter" or

"Homer, travelling from city to city"?

But the most problematic feature in the modern exegesis of this fragment is

the, in my opinion, undue emphasis on Democritus’ purportedly novel view of Homer

as housebuilder: according to Koster "Der Dichter ist nicht mehr Werkzeug der

Musen, sondern ein eTriarrj/xov, der mit einem Zimmermann verglichen wird " .64

In the same vein Russell suggests that

“See e.g. Finsler (1900) 172. The most ambitious attempt was undertaken by Delatte (1934) 35ff.

“Koster (1970) 24. 48 we must also keep open the possibility that there is here nothing but two distinct statements, one alleging Homer’s divine nature, the other stating that he contrived by his skill the kosmos of words, the two together amounting to a description of what makes him a great poet. This would amount to the common later idea that the union ofphusis and techne, ars and ingenium, is what produces prefection. Yet in this case also there would be emphasis on Homer’s craftsmanship: he is a tekton, like a builder or carpenter, whose art consists in putting things together into an ordered and beautiful structure .65

There is no denying that Homer is presented like an artisan and the choise of verb was probably, although not necessarily, deliberate on Democritus’ part in order to

convey precisely this image. But it is misleading to isolate this metaphor and present

it as something almost revolutionary in the history of Greek views on poetic production. On the contrary, it was a thoroughly traditional poetic metaphor present

already in Homer66 and particularly common in Pindar, where the poets are called

"wise carpenters who put together sonorous verses" (e£tnic&v k ekaSevv& v,

r i ktovcc ola ao 0 oi &p/xoaav, P. 3.113)67. The inspired poet has never been conceived without skill -only Plato suggested otherwise- but skill is the result and manifestation of his divine inspiration. As far as can be gleaned from Democritus’ fragments, the poet remains dependent on divine sources for his success and the

“Russell (1981) 73.

“Od. 24.197; cf. 12.183 and 17.385 and Horn. Hymn. Aphr. (2)20. See Frankel (1925) 4-5.

“See also O. 6.1-4, 3.3, 7.21, P. 7.3, 1. 2.46. See Taillardat (1965) 438, Steiner (1986) 55-56 and Kurke (1991) 192-194; see also more generally Auger (1986) and especially Svenbro (1976) 139-212 who argues that, before Pindar canonized it, Simonides popularized the conception of the poet as craftsman. The language of building as a metaphor for poetry seems to have been of Indoeuropean origin; see Schmitt (1967) 297-298 and Durante (1968) 261-290. 49 poem resembles a beautiful artefact. Russell’s distinction sounds extremely artificial

and almost otiose: what would Homer’s divine nature generate or be responsible for

if his skill were sufficient for fashioning his beautiful poems? On the other hand, if

the two are related as Russell himself seems to believe, what would be more natural

than to make skill dependent on his divine nature, an idea perfectly at home in

Greek poetic theory and effortlessly borne out by the fragment itself?

To return to Democritus’ possible influence on Plato, there does not seem to be any sufficient reason to postulate it in the first place. Although it is not impossible, from the two surviving fragments neither does Democritus appear to have suggested anything strikingly innovative nor does Plato seems to have confined himself to enthusiasm. Plato’s primary concern, poetry’s claim to knowledge and status as a rational skill, is absent from the Democritean fragments. Quellenforschung should obviously turn to more fruitful directions. Democritus’ influence on Plato seems a dead end because, as far an inspiration and enthusiasm are concerned, anybody could have influenced Plato. As mentioned above, Democritus’ distinction is a purely external one: he is the first surviving prose writer to have dealt with poetic theory and he does not seem to have diverged from traditional views.

The first Greek who can be thought to have considered art and inspiration as alternative sources of poetic production is the author of the treatiseOn the Sublime, attributed to Longinus. At the beginning of the work (ch. 2) the author addresses the question whether literary sublimity is the result of inborn excellence or it can be taught and thus be subject of theoretical inquiry. Longinus expounds views similar 50 to Pindar’s: nature is the indispensible foundation upon which art exerts a beneficial controlling influence, besides being the epistemological justification for literary criticism. Longinus of course does not deny those he considers uninspired authors or, in his words, "lowly or mediocre natures" (33.2.5) the characterization "poet"; nevertheless, he insists that they lag hopelessly behind the exalted geniuses, despite the fact that the works of mediocrities, like Apollonius of Rhodes (!) and Hypereides, have no mistakes unlike the works of sublime masters, e.g. Homer and Demosthenes.

Art by itself cannot procure but very meager results. Nature is responsible for excellence (iieyako

Aristotle of course did away with the traditional religious concept of inspiration and stressed the rational and teachable principles of the art but he attributes considerable significance to the personal factor68. Halliwell denies that because he confuses it with the notion of an inspired, romantic poetic genius. He tries to account for Aristotle’s repeated and clear references to the personal talents of the poets by proclaiming them of limited significance and, as mentioned above, a happy coincidence of alignment between personal inclinations and the natural

“"Madness" as a source of poetic production appears only Poet.in 1455a where the poet is said to be either eixpvfiz or fxaviKOQ. The text has been suspected (see Lucas (1968) ad loc. and Russell (1981) 77-79) but I think that it is not impossible for Aristotle to have held such a view: a /xavuc6c poet would possibly be an extremely eixpvfic poet. Since inspiration is not incompatible with skill, the fiav ik 6 c would simply derive it from an excess of eu^uCa. 51 cultural movement of the development of the tragic genre and the art of poetry®.

Finally he comes up with the suggestion that, although Aristotle discards the notion

of inspiration and the concomitant idea of the poet as mouthpiece or prophet of the

Muses, the Aristotelian artist "has once again come to be seen as a medium through

which the operations of natural and greater forces are channeled. Inspiration, it

could be argued, has been "naturalised" within the Aristotelian view of art"70. This

is a distortion of the Aristotelian views. Halliwell consistently projects the nature

and teleology of the art, which resides squarely within the soul of individual artists, outside of them and thinks that these "forces", which are nothing but the potential of

the artists themselves, are superordinate to them, guide their choices and dictate

their acts from without. The absurdity of this theory which does not explain anything

and saddles Aristotle with gross pseudo-metaphysical mumbo-jumbo has, I trust, become more than evident from the preceding discussion. I will next focus on how

the Aristotelian concept of art, already briefly touched upon above, informs key-

features of the Poetics.

^AP 91.

"AP 92. CHAPTER II

ARISTOTLE’S CONCEPT OF ART AND THE POETICS

In EN 1094M0-27 Aristotle distinguishes between disciplines (piBoSoi) of demanding rigor like mathematics and disciplines like rhetoric and ethics whose standards of rigor are less demanding. Since he considers this distinction a mark of the educated person, his characterization of poetics as a disciplinepidoSoc ( ) in the introduction to the Poetics (1447al2) must certainly imply one of somewhat relaxed rigor. Despite this gradation, the common term brings forth a fundamental analogy in Aristotle’s conception of art and science manifest not only in his famous tripartite division of sciences but also in his terminology: a look at Bonitz’s Index s.v. r ix v n reveals that "ipsae doctrinae theoreticae t € x vr 1G nomine continentur...acpromiscue ad significandas et artes et doctrinas r € x vrl et iwtoTtfpri u s u r p e t u r In this light,

Aristotle’s characterization of poetics as ri%vri throughout the treatise (1450b20,

145la7, 1453a22,1460bl5, 23 and 30, 1461b24, 1462al) presupposes a conception of art distinct from the references to poetry as craft in Aristophanes’Frogs , where

Tor a general discussion of the relation between the Poetics and Aristotle’s wider concept of see Olson (1965) 175-186; cf. also Bartels (1965) 275-286.

52 53 Euripides says that he took over the art of tragic poetry from Aeschylus2. Homer

already considers poetry as a teachable art. In his praise of Demodocus Odysseus

assumes that the bard was taught either by the Muse or by Apollo (Od. 8.488) and

in a similar vein Phemius, when he supplicates Odysseus, claims that he is self-taught

{Od. 22.347). Both claims put Homer in an old Indoeuropean tradition of craft

terminology for poetry but for Aristotle the term "art" seems to carry connotations

much stronger than a simple pre-theoretic analogy with productive crafts.

In this context it is instructive to see how Aristotle conceives his own position

as a rhetorical theorist vis-&-vis the numerous authors of rhetorical handbooks. In

Top. 183b36-184a8 he claims that, contented with the composition of rhetorical paradigms, these authors did not develop rhetoric into an art: solely composing paradigms of speeches, the products of their art, they did not explain whence they

could procure them (oOev Svvfioerai nopioao&ai ra roiavra, Top. 184a6),

exactly as if someone, claiming to hand down the knowledge of foot comfort, taught not shoemaking but procured all kinds of shoes. This analogy takes an art to be superordinate to its products, a point well illustrated in the closing chapter of the

Posterior Analytics (100a3-8):

So from perception results memory, as we say, and from memory there comes experience when memory of the same thing occurs often; because many memories create a single experience. From experience or the universal, when it has sunk in the soul from all these things, the one over the many which is the same for all of them, there comes a principle of art and science.

27rap#A.a/3ov rr)v t^xv^v ita p a aou (Frogs 939). 54 This account leaves no doubt that for Aristotle art consists in a universal, the

principle of the art, and not in its particular instantiations which give rise to it

inductively. Thus it turns out that the foundations of Aristotle’s critique lie in the

close association of systematic knowledge with the knowledge of first principles and

first causes (Phys. 184al0-14). Indeed, Aristotle’s critique is carried further in the

introductory chapter of theRhetoric (1354a8-16) where he first observes that all

people have developed rhetorical abilities to some degree, others accidentally and

others habitually so that

it is clear that the same result can be obtained in a systematic way. It is possible to inquire into the cause (a irta ) of their success by either accident or habit and everybody would agree that this is the purpose of art.Now, however, the authors of rhetorical handbooks have failed to develop any part of that art (because except for the means of persuasion all else is superfluous) since they do not discuss at all the enthymemes, the core of persuasion, and they deal mostly with what is extraneous to the subject-matter (<■!;<•> rou irp&ypar oc).

The emphatic now signals that these authors have missed what Aristotle considers

the characteristic mark of art, the inquiry into the cause. On account of the

isomorphism between causes and principles, which are defined as the "first whence

(50ev) something comes into being or becomes known" (Met. 1013al6-19), the

inquiry into the cause is directly parallel to the origin of the artistic product in the

Topics (60ev Svvfioerai wopiaaaOai to. ro ia u ra , Top. 184a6).

It is also important to point out a formal aspect of Aristotle’s conception of art which reveals the theoretical background of his critique. Aristotle’s claim that he has dealt with what comes first by nature, i.e. aura ra itp&ypara £k rtv o v l%ex 55 rd wiOavov (Rhet. 1403bl8-19), runs parallel to the remark that the handbook authors dealt with issues extraneous to the subject-matter. From both the critique in Rhet. 1354all-16 and the context of Aristotle’s claim {Rhet. 1403b6-14) it is clear that he refers to the definition of rhetoric as we pi Ikckcttov rou

Gewprjaai rd evSe%6nevov wi0av6v {Rhet. 1355b24-26) which picks out the logical and non-logical means of persuasion. For Aristotle these are coextensive with rhetoric as a systematic art {Rhet. 1354al3-14, 1355a3-4; cf. 1355b37-1356al3) and their inclusion in a definition puts the final touch on his claim that, in contrast to the handbook authors, he lays down the foundations of the art3. The object of a definition is a universal and a form {Met. 1035b33-1036al, 1036a28-29) and on many an occasion Aristotle defines art as the form of its product (GA 735a2 and 740b28).

Medicine and housebuilding, for instance, are identified with the form of health and house respectively{Met. 1032b 13-14; cf. 1032a32-b2, 1034a24 and 1070b33), i.e the definition of their products {Met. 1070a29-30, PA 639b 16) which picks out the universal arrived at by induction according to AnPost. 100a3-8. In want of such a definition the hypothetical shoemaker and the handbook authors lack the most crucial ingredient of Aristotle’s unified conception of art and science. Formalized in a definition, the outcome of induction delineates the domain (ylvoc) of all productive and mathematical sciences, which, "assuming the definitions, attempt to prove the rest in a more relaxed or rigorous manner" (Met. 1063b36-1064a4; cf.

3On it pay/xa, the definition of rhetoric and its status as a systematic discipline see the detailed discussion in Lossau (1981) 184-203. 56 1025b3-13). Aristotle here clearly presupposes the formal apparatus of his theory of

science -one need only consider how definitions determine the unity of a science’s

domain and thus delineate the science itself (AnPost. 87a38-b4)4 -and he does not

lose sight of the distinction he draws in EN 1094b 10-27: like rhetoric, other arts, over

which he quantifies universally, cannot measure up to the demonstrative rigor of

geometry but nevertheless they do have a subject matter, no matter how humble it

might be (cf. AnPost. 76b 11-22). In Met. 1063b36-1064al Aristotle refers not only to

mathematical sciences but also to medicine and gymnastics and later on in genuinely

Socratic fashion he enumerates crafts like housebuilding, weaving, shoemaking and

baking {Met. 1064M9-21). If the definitions of geometrical objects circumscribe the

domain of geometry, the definition of health and shoe define the subject-matter of

medicine and shoemaking in the same manner. As mathematicians assume the

definition of unit, straight and triangle in order to prove their properties{AnPost.

76a34-36) the production of an artefact assumes the definition of the intended

product and searches for antecedent causes that make production possible. Bearing

out the mention of a relaxed demonstration in productive arts, Aristotle is ready to

compare the productive activity of an artist with the par excellence heuristic method

of Greek geometry, analysis and synthesis{EN 1112b 11-27).

The close relationship between mathematical proofs and productive activity will be taken up later on. What matters here is that Aristotle’s critique of the handbook authors presupposes his theory of science bringing out the unitary

“See Hintikka (1972). 57 conception of theoretical science and productive arts already implicit in the tripartite division of sciences. It goes without saying that the same conception must underlie the characterization of poetics as "method" in the introduction to thePoetics and "art" throughout the treatise. Nevertheless, although it is a locus communis in the literature that in the Poetics Aristotle attempts to treat poetry scientifically, no one has fleshed out this observation in Aristotelian terms. The broader question, namely what art consists in, seems to have been discussed, at least in the Academy. Plato comes to grips with the problem in the final part of thePhaedrus (268alff.). What would Sophocles and Euripides say if someone claimed that he could teach tragedy because he knew how to compose long speeches about something trivial, short ones about something important and how to move pity or fear and appear threatening

(Phdr. 268c5-d2)? For Plato Sophocles would dismiss such abilities as only preliminary to the tragic art(Phdr. 269al-4), a point similar to Aristotle’s critique of the handbook authors, although one should expect that Plato and Aristotle would give different answers to the same question, all the more since Aristotle had attacked

Plato’s views on poetry in his lost dialogue On Poets5. Plato of course is never explicitly referred to in the Poetics but Aristotle’s debt to the Platonic theory of poetry is unmistakable6. In light of Artistotle’s critique of the handbook authors it is not surprising that he chose to criticize Plato’s views in the same chapter which

sSee fr. 3 (Ross); cf. also fr. 4.

6For a list of common issues and terms between the Poetics and Plato’s discussions of poetry see Halliwell (1986) Appendix 2, 331-336. 58 formally establishes poetics as an Aristotelian art. Since he understands artistic

products as those whose form, i.e. their essence and their definition, lies in the soul

of the artist, and the respective art is identified with their essence (Met. 1032a32-

bl4), the definition of tragedy’s essence (Poet. 1449b23-24), the product of poetics,

delineates the subject matter of a genuine art in Aristotle’s sense. But his

memorable pronouncement in Poet. 1450al6-17 to the effect that tragedy imitates not

human beings and their character but actions can only be a direct attack against

Plato’s association of artistic correctness and moral goodness (Leg. 669blff.). The

Aristotelian alternative is intrinsically related to the definition of tragedy: among the

six qualitative parts in this definition, namely mythos, character, diction, thought,

spectacle and music (Poet. 1450a7-10), Aristotle singles out the mythos as the final

cause of tragedy(Poet. 1450a22-23) and to indicate this distinction from now on the

collective term "qualitative parts" will be used for all other parts except mythos. But

if Aristotle’e implicit critique of Plato is related to what the definition picks out, as

is the case with his critique of the handbook authors, an account of the role mythos

and the qualitative parts play in the definition is in place. But inPoet. 1450b23-25

Aristotle singles out a synonym ofmythos , the complete (reXeia) and whole (oXtj)

action, as the sole definiens of tragedy7, although the definition of tragedy in Poet.

1449b22-28 includes the qualitative parts too. Similarly, the characterization of

7Keirai 6r) j)/xiv rr)v rpaycpSCavreXeiaz icai dXrjc 7rpd$ea><; e lv a i /iC/irjaiv; cf. Bonitz (1870) 380b43-45: "sedplerumque logice Keiodai usurpatur de propositionibus ac defmitionibus, quae pro fundamento ratiocinationis positae sunt". 59 mythos as the final cause of tragedy (Poet. 1450a22-23) is in line with the

characterization of mythos as principle and soul of tragedy, since the soul of an

animal is its essence which is well-defined if it picks out the function (Ipyov) of the

organism (Met. 1035b 14-18; cf. DA 412b 10-13). It is clear, however, from Poet.

1453b 11-14 (cf. 1452b29-33) that the function of tragedy is the pleasure from pity and

fear which result from the mythos through imitation.

The difference in content between the two definitions and the double

characterization of tragedy’s final cause gains in interest because, in its capacity as

the sole definiens of tragedy, action is determined by the noXXax&Q Xeyopeva

"complete" and "whole" which both appear in a remarkable simile: as in tragedy, in

epic too the mythos must be wept piav wpa^iv, 6Xrjv ko-1 reXeiav...iv’ & one p

£ov £v oAov woi^ rf)v oiiceiav j)

between mythos and a natural substance is also implied in Poet. 1450b21-1451a2

where it is clear that the mythos , qua complete, whole and unified action, is

analogous to a whole and unified animal (Poet. 1451al-2). Completeness implies a

serial articulation according to beginning, middle and end (Poet. 1450b26-27), a

schema which Aristotle owes to Plato’s Phaedrus along with the analogy between an

artwork and a living being8. This Platonic borrowing provides a figurative analogy

for the unity which underlies Aristotle’s conception ofmythos. In the literature the

Aristotelian unity has been taken largely at face value as "the concept of unity" which

"in one version or another, is the one of the most pervasive and arguably the most

sPhdr. 264c-d; cf. Rep. 401a, Gorg. 506d. 60 indispensible criteria in the understanding of art"9. The notion of unity is admitedly

hard to define but its sheer presence in Aristotle’s definition of tragedy should warn

against an all too easy association with the Unities of eighteenth century neoclassical

literary theory. Actually Aristotle’s biological analogy has only a tangential

relationship even with its Platonic model and it is obviously encapsulated in the

parallelism between mythos and the soul of a living organism (Poet. 1450a38-39). In

the Aristotelian corpus natural substances and pleasure are once more related inPA

645a7ff. where Aristotle remarks that even aesthetically offensive animals cause pleasure to the philosophically minded observer who is able to understand their final

cause (17 5A.rj ouaCa). This passage is related to the issue at hand because a natural substance produces pleasurequa final cause and the action which in Poet. 1459al8-21 produces pleasure has been characterized as the final cause of tragedy(Poet.

1450a22-23). Thus it is clear that the exact point of Aristotle’s analogy should be sought in Poet. 1450a38-39, since, like the mythos, the soul of a living organism is its final cause and its principle(DA 415b8ff.). Indeed, the characterization of a natural substance as a unified whole (ev 6 A.ov) picks out the final cause and the principle of a living organism. "Unified" and "whole" are equivalent (Met. 1016bll-17,

1052a22-23, 34-35) and a natural substance of course can only be a natural whole

(oe\ SXov, Met. 1023b32-36) and as such a natural continuum (

But natural continua are unities per se (koeG’ aurd iv , Met. 1015b36-1016a6) by virtue of their being whole due to the form which is the cause of the unity of

9Halliwell (1986) 96. 61 essences (Met. 1052al5-35). Given that the soul is the form by virtue of which the

organism is said to be and to be one (DA 412b6-9), Aristotle’s characterization of an

animal as a "unified whole" can only refer to the soul.

Thus the analogy between mythos and a natural substance qua a unified whole

simply restates in formal terms the analogy betweenmythos and soul qua final cause.

These terms, however, are void of aesthetic import in spite of the communis opinio.

Further evidence is furnished by Aristotle’s discussion of a problem totally irrelevant

to literary theory, namely how the Academic philosophy of mathematics can account

for the unity of mathematical objects (Met. 1077a20-33):

Besides, what on earth is it in virtue of which mathematical magnitudes are unified ? It is reasonable for things around us to be one in virtue of soul or a part of soul or something else-otherwise there is [not one but] many, and the thing is divided up. But these objects are divisible and quantitative; what can be responsible for their being one and holding together? Besides, the point is clear from the way they are generated. First length is generated, then breadth, finally depth, and then it is complete. So if what is subsequent in generation is prior in reality, body should be prior to plane and to length. It is morecomplete and whole in the following way also-it becomes animate. How could there be an animate line or plane? The supposition would be beyond our senses. Besides, a body is a kind of real object (for it already has completeness, in a way), but how can lines be real objects? Not by being form and a kind of shape, as perhaps the soul is. 10

All nokkax&Q Xeydpeva that qualify action in thePoetics , unified, complete and whole, appear here as properties imparted by a form, the soul of a natural substance, on mathematical objects that have nothing to do with the mythos. Thus it seems that

Aristotle is interested not in a prescriptive literaiy-aesthetic evaluation of themythos

10On Aristotle’s argument in this passage see Annas (1976) 145-146. 62 but in its metaphysical properties11. It makes sense to articulate metaphysically the

concept of mythos in the context of tragedy’s definition which reflects Aristotle’s

conception af art as the essence of its product. Indeed, since "complete’ and "whole"

are equivalent (cf. Met. 102 lb 12-13 and 1023b26-28), the characterization of action

as a unified whole underlies the definition of tragedy. This is borne out by a

comparison with the definition of "complete": 2Kaurov yapTore reXeiov Kal

o u ala nao a Tore reXeia, orav Kara to eifioc rfjc oiKeiac dperrjc nrjSiv

iXXeinij jit6 piov rou Kara 0 6 aiv fieyidovQ- I n ole 6 jrdpxei tIXoq

a 7 rou, raflra Xiyerai riXeia (Met. 1021b21-24). Every attribute of tragic action appears in this definition which also accounts for the serial articulation of action (Poet. 1450b26-27) so that the transposition or omission of a part destroys the whole (Poet. 1451a30-35; cf. Met. 1024al-5). Although in the Poetics Aristotle posits it as a requirement, this is a formal property of the rdrrp u r u c Xeyo^eva iv

d>v h ouala /xta (Met. 1016b8-16): a thing is one if it is a whole on account of its having a form -a shoe for instance is one if its components are arranged not randomly but in a specific way that makes a shoe. Not only does Aristotle’s example capture the requisite articulation ofmythos, it also includes the entire set of terms that characterize it. In its analogy to mythos soul serves as a paradigm that exhibits the same formal characteristic which is expressed by these terms. They function on the level of the metalanguage guaranteeing those metaphysical properties of the

"Halliwell (1986) 98 similarly denies that unity in the Poetics is an aestheticist notion but thinks that it "concerns the representation of human action" which "rests inescapably on the cognitive understanding of the action portrayed". 63 mythos that allow it to stand as the definiens of tragedy, the product of poetics which thus becomes a formally defined art in the Aristotelian sense.

Returning now to the initial question, it seems that the isolation of mythos as the sole definiens of tragedy goes hand in hand with the metaphysical delineation of the term. The obvious implication is that the rest of the qualitative parts that figure in the definition are metaphysically subordinate tomythos. The strong metaphysical emphasis on the mythos is well-exemplified in Aristotle’s critique of Speusippus’ metaphysical system to which I will turn in a moment. Before, though, it is necessary to examine Aristotle’s brief but suggestive discussion of the worst kind of mythos, the episodic (Poet. 1451b33-52al). Such a mythos results from a distortion in the serial arrangement of its component parts. To use Aristotle’s own metaphor in the

Metaphysics, the episodic mythos is like a shoe whose parts have been haphazardly sewn together: it is not one and whole because it lacks the unity-imposing form. By these lights the episodic mythos becomes a good model for the Speusippean tier- metaphysics(Met. 1075b34-1076a4, 1090b 13-20). In positing different principles for different essences, numbers, geometrical magnitudes, souls and sensible bodies,

Speusippus reduced the essence of the universe into a disjointed assemblage which

Aristotle calls "episodic" like a bad tragedy. More specifically, since Speusippus’ system lacks the notion of the prime mover, it cannot account for the unity of these essences. Not only is lack of unity the flaw of an episodicmythos but Aristotle also draws an explicit analogy between movers and arts. In GC 1.7 Aristotle gives medicine as an example of a first mover and he elaborates on this analogy in Met. 64 Z.7 where a first mover, paralleling the form of the generator in natural production,

is identified as the form in the soul of an artist, i.e. the form of the artistic product

which is the art itself (Met. 1032a25-bl4). This ontological analysis which informs the

metaphor in Aristotle’s critique of Speusippus is also present Met.in A.4 which paves

the way for the notion of the prime mover. Aristotle admits that different things

have different principles which can nevertheless be conceived as one analogically if

subsumed under the rubrics "form", "privation", "matter" and "first mover". The prime

mover falls under the last one along with productive arts like housebuilding and

medicine qua forms of their end-products. By means of this analogy themythos

comes accross once more as the form of tragedy.

Given such a strong metaphysical characterization, singling outmythos as the

definiens and the final cause of tragedy cannot simply be shorthand for the definition

of tragedy and the more expansive characterization of its final cause in Poet.

1453b 11-14 respectively. The situation is resolved by the ambiguity of the term

"definition of essence" in Poet. 1449b23-24. The definition of an artistic product, e.g.

a house, can pick out either the matter of the product (stones, bricks, wood) or the form (a container protecting people and property) or both(Met. 1043al4-28, DA

403a25-b7). If now the definition of essence picks out the composite, this definition is to the shorter one in Poet. 1450b23-25 as t 6 tC ionv to r 6 r i fiv eivai: it gives a complete account of what tragedy is whereas the definition inPoet. 1450b23-

25 mentions only mythos , the determining feature of tragedy that belongs to it per 65 se12. On this interpretation, all other parts mentioned in the definition of tragedy are to mythos as matter to form. Aristotle of course never employs the concept of matter in the Poetics and there can be no doubt that the other parts cannot be to mythos as bricks etc. are to the form of the house. But he makes clear implicitly that building materials cannot be a house in themselves and similarly all other parts cannot be a tragedy without mythos. Character, diction and thought do not perform the function (Ipyov) of tragedy(Poet. 1450a29-31) and the same certainly holds for spectacle and musical accompaniment (cf.Poet. 1450b 15-20). To extend Aristotle’s own analogy between soul and mythos, it seems that the assemblage of all these parts is a tragedy only homonymously, like the carcass whose still present materials are fingers and hands homonymously, because unable to perform the function (Ipyov) that defines them (Met. 1035b 14-25, 1036b30-32). As flesh and bones are accidental to the soul of an animal, thus all qualitative parts other than mythos are accidental to it: they do not contribute anything to the nature of tragedy as the material stuff does not determine the nature of the animal whose form is separate in account and cannot be defined through something else13.

Although the hylomorphic analysis does account for Aristotle’s characterization of mythos and the other qualitative parts, it is definitely hard to conceive tragedy as an artefact like a house which requires a material in the literal

12Cf. Gill (1989) 111.

13For a detailed discussion of Aristotle’s concept of essense see Gill (1989) 116- 120. 66 sense of the term. But a coherent interpretation of the qualitative parts as matter emerges if we take matter in its analytical sense which is largely determined by the context14. In tragedy’s case the context is indicated by a characterization that all qualitative parts share in Poet. 1450b4-7. Aristotle assigns thought to the product

(gpyov) of rhetoric and politics (cf.Poet. 1456a34-36) and elsewhere he identifies the product as the final cause of productive arts(DC 306al6). In EE 1219al3-15 health and house are the products of medicine and housebuilding but health is the final cause of medicine in EN 1094a8, where the final cause of an art is identified with its product. Likewise, the choice of appropriate diction belongs to acting(Poet.

1456b8-13), an independent art which in Rhet. 1403b35 Aristotle considers not yet formally developed, although he cites two treatises on it(Rhet. 1403b26,1404al3-15); the art of the costume-maker is responsible for the spectacle (Poet. 1450bl6-20) and similarly music is an art in its own right. One would also expect that character too belongs to the provenance of another art, although Aristotle does not say anything explicitly to that effect. Nevertheless, the close relationship between thought and character in Poet. 1449b36-38 as determinants of action suggests that the projection of character must be the product of rhetoric along with proofs, refutations and the production of emotions (Poet. 1456a33-b2); but since Aristotle characterizes rhetoric as an offshoot of dialectic and ethics, or politics, (Rhet. 1356a20-27), both production of emotions and character must strictly speaking fall in the provenance of ethics.

Thus, all qualitative parts belong to arts other than poetics. But Aristotle is explicit

14See e.g. Leszl (1970) 517-518, Wieland (1970) 209 and Jones (1974) 494-497. 67 about the "foundational" role of the definition of an art’s product: it defines that art

itself and the qualitative parts can be conceived as the material parts of the

definition, in the sense that the arts to which they belong are accidental to the art of poetics. Like its product, tragedy, the latter is separate in account as the soul of an

animal is separate in account from the perishable matter. A facet of the problem is the omission of catharsis by means of pity and fear from the characterization of tragedy’s final cause or function. Leaving catharsis aside for the time being, the glossing over of pity and fear makes good sense on the above interpretation, since the arousal of passion is the provenance of rhetoric15. Now the status of passions in the Rhetoric parallels the restriction Aristotle lays down on the production of pity and fear in the Poetics. Recall that the failure of rhetorical theorists to develop rhetoric into an art arose from their dealing with the rou irp&ypaTOz, i.e. the production of emotions, as Aristotle makes clear in Rhet. 1354al-18. Nevertheless, production of pathos becomes an organic part of the Aristotelian conception of rhetoric as the third non-logical pistis, a Trpay/xa-grounded instrument towards the art’s final cause which is encapsulated in the definition of rhetoric: an enthymematic, i.e. dialectical, tool to inquire into (and prove) what is probable on a given subject16.

In a similar vein Aristotle ascribes an important role to pity and fear in thePoetics but he adds the crucial restriction that they must originate from within the mythos,

“See Poet. 1456a34-b2; cf. Rhet. 1378al9ff.

16See Lossau (1981) 199-200. 68 tv toIq np&ynttoi (Poet. 1453bll-14)17. He admits that these emotions can arise from the spectacle, an effect he calls "less artistic" (Poet. 1453b8) in line with the characterization "most artless" for spectacle in Poet. 1450b 17. In Poet. 1450b 15-20

Aristotle indicates unambiguously that the characterization "artless" is said not absolutely but relatively to the art of poetics: if pity and fear do not arisetv roi<; irp&ypaoi, they belong to acting and costume-making rather than poetics. Thus

Aristotle’s point is the same as his foundational claim in the Rhetoric where he correlates the art itself and the np&ynara. But even when the production of emotions arises from within the mythos, tv roic vp&ypaax, Aristotle still subsumes it under thought which belongs to the provenance of rhetoric (Poet. 1456a34-b4). In this light the omission of pity and fear from the characterization of tragedy’s final cause is justified if the final cause is to serve as the definition of the art itself. The same holds for the omission of catharsis, since catharsis by means of pity and fear operates in music too (Pol. 1342a4-16). It is important to note here that the earliest post-Aristotelian reference to catharsis is found in a fragment of Aristoxenus of

Tarentum18, one of Aristotle’s own students whose treatises on music show clear influence of the Aristotelian theory of science. If Aristoxenus had indeed developed a catharsis theory of his own, Aristotle could not have considered such a theory part

‘It is exactly in the final cause of the art that the difference between rhetorical and poeticalpathopoieia lies; see Lossau (1981) 138-139.

18Fr. 26 (Wehrli); cf. fr. 117. 69 of poetics proper19.

Thus all qualitative parts are products of arts other than poetics. Their relationship to poetics seems to be adequately captured by the distinction between

"architectonic" and "poetic" art inPhys. 194a36-b7. The architectonic is to the poetic as form is to matter: steermanship, for example, determines the form of the rudder whereas ruddermaking determines its matter. Matter here is the product of a distinct art whose form is another superordinate art and Aristotle makes clear that qua matter this product is made with an eye to a higher goal(Phys. 194b7-8). The rudder therefore is the material component in the definition of steermanship and similarly those parts in the definiton of tragedy that are the product of rhetoric must be the material parts of this definition. On this account, rhetoric turns out to be the "poetic" art of poetics, the architectonic art towards the realization of whose final cause rhetoric contributes.

If the definition of tragedy captures both the complexity of the poetic product and the autonomy of the art in a distinctly Aristotelian way, the characterization of the qualitative parts seems rather puzzling at first sight. InPoet. ch. 19, usually neglected in critical discussions of the treatise, Aristotle takes Protagoras to task because he criticized the Homeric choice of speech mode (o%fina rfjQ A#Sewc) 20 in the very first line of the Iliad. The knowledge of the appropriate speech mode,

19Lienhard (1950) 51-61 suggests that in Pol. 1342a4-16 Aristotle refers to a catharsis theory of Aristoxenus, although Aristoxenus’ fragments have been thought to echo a Pythagorean theory of catharsis; see Wehrli (1945) 54-55.

“On this term see Bernays (1842) 589. 70 argues Aristotle, is a "theorem" (Qe&prnia) not of poetics but of another art and thus the charge of such ignorance cannot be brought against the poet (7rapa rr)v toGtgjv yvojaiv f) SyvoiQ!V ovStv eic rf)v woipriK^v inirlfirina ip e ra i 6 ,r i Kal

&$iov anouSfiQ, Poet. 1456b 13-15). This criticism implies the conception of an art as a collection of "theorems" which also underlies Rhet. 1359b5-8. There Aristotle is careful to ascribe to rhetoric those "theorems" that belong only to it. Within Poet. ch. 19, where Aristotle assigns thought, and thus character and the production of emotions, entirely to rhetoric, his argument clearly suggests that poetics is impregnable to any objections arising from the qualitative parts. He even furnishes a classification of poetic faults to support this claim. In the immediate context of the

Poetics Aristotle’s criticism should be viewed in light of his subsequent discussion of

"problems" (n pop kfmar a), the traditional staple of literary criticism(Poet, ch.25)21.

Aristotle himself had written a now lost treatise on Homeric problems22 and, judging from his argument against Protagoras, he gave a new twist to this traditional subject.

He distinguishes between two kinds of poetic faults that respectively give rise to two kinds of objections (iniTinfiixara) addressed in the problems: those concerning poetics per se (icaO’ avrfjv) and those pertaining to it accidentally (Kara ov/jipept ?k o c ) (Poet. 1460b 15-22). The line in which he elaborates on the first kind of faults (Poet. 1460bl6-17) is lacunose and I will return to it later on. It is clear,

“For the pre- and post-Aristotelian history of the problems see Pfeiffer (1968) 89-91.

22The work, in six books, appears in the list of Aristotle’s works in D.L. 5.26. 71 however, from the context (Poet. 1460b22-25) that, broadly speaking, objections to the art per se are determined as such in relation to the final cause of the art whereas accidental ones, like the objection of Protagoras, pick out "theorems" of other arts and not poetics (Poet. 1460bl3-21 and 28-32).

But Aristotle’s praise of Homer, whom he conceives as the paradigmatically acclaimed poet, seems to cut across these neat categories of objections. Naturally,

Aristotle praises the mythos of the Homeric epic (Poet. 1459al7ff.) and also Homer’s handling of thought, diction (Poet. 1459bl6) and character (Poet. 1460a9-ll). Being subject to praise, these areas can also be targets of objections. But thought and diction pertain to another art and an objection about them would be vulnerable like

Protagoras’. Similarly, an objection censuring thought, diction and character is precluded byPoet. 1450al5'-33, provided of course that the mythos is unobjectionable.

Such a position certainly seems compromising. Moreover, although inPoet. 1460b22-

28 Aristotle allows faults concerning other arts, he immediately asserts that optimally even such faults should be avoided (Poet. 1460b28-29). Thus, despite his claim that these faults and the objections they give rise to pertain to poetics accidentally,

Aristotle envisages an ideal artwork which is not open even to such accidental objections. In what sense then does he brush aside Protagoras’ objection and by implication the entire class of accidental objections? In light of the subsequent discussion in Poet, ch.25 Aristotle’s criticism is not ad hominem but a general statement concerning accidental objections.

It is indeed hard to accuse Aristotle of inconsistency since he himself makes 72 clear that a fault and, therefore, an objection is accidental or per se not absolutely but relatively to the final cause of the art. It should be noted that in Poet, ch.25 the category of the arts whose correctness might provide grounds for objections to the poetics broadens to include medicine (Poet. 1460b20), biology (Poet. 1460b 17-19,31-

32) and politics (Poet. 1460bl3-15). These arts enter into the mimesis but unfortunately the interesting characterization of faults pertaining to poetics per se has fallen into the lacuna already mentioned above. Aristotle gives only one example of an objection to poetics per se which censures impossiblities (a&vvara)(Poet.

1460b21-29) but he is ready to brush it aside provided that the final cause of the art is unaffected; as he says, the final cause of the art has been discussed above. Since

Aristotle obviously refers here to the characterization of the mythos as final cause of tragedy in Poet. 1450a22, his example must hinge on a relationship between impossibilities and mythos. The impossiblity which Aristotle considers as an example concerns Achilles’ famous pursuit of Hector in the Iliad (Poet. 1460b26-28). A military impossibility is involved here and Aristotle remarks: e i /Ltevroi to riko<; h n&kkov f) f)Ttov eveSi%£To urrdpxetv ical Kara rr)v wcpl rofirwv ri%vr}v, ouk op0d>c (Poet. 1460b26-28). Since it is clear from the context that

Aristotle refers to the art of war23, he takes the impossibility as an accidental objection to poetics only if the impossibility promotes the final cause of the art. But in Poet. 1460all-34 he characterizes an impossibility as a violation of the modalities that determine the arrangement of the parts of a mythos into a unified, whole and

aSee Lucas (1968) 236. 73 complete action (Poet. 1450b21-1451al5): the discussion starts form the impossibility in Achilles’ pursuit of Hector and naturally ends up banishing impossibilities from a properly constructedmythos. Aristotle is ready to admit them only if they are probable (eiKora), that is, parts of a unified, whole and complete action which is the final cause of the poetic product and, of course, the art itself. Thus the above quoted passage fromPoet. 1460b26-28 rephrases Poet. 1451a34-35 (o yap irpoobv f) fj.fi irpoabv firfShv woiei tnxSrfkov, ovStv fi opiov rou 6 A.ou iorxv) in terms of impossibility. Such an impossibility certainly compromises the final cause of the art and consequently generates an objection to the art itself. Given all this, the problematic fixfifioao&ax in Poet. 1460bl7 should perhaps be complemented with the negated formal attributes of action: ei fitv yap npoexXero fixfifjoaoBax

bSvvafiiav, aurrjc ff d/zapr la 24.

Only this type of fault pertains to poetics because it violates the final cause of the art and only such faults should cause objections to the art per se.

Thus the arts that provide materials for the mimesis are on a par with those that enter the definition of tragedy: they contribute to the realization of the art’s final cause and legitimate objections to poetics can arise from them only when they fall short of their raison d ’ etre in the poem. The fault of poetics per se now falls under

Aristotle’s general characterization of faults in both artificial and natural production in Phys. 199a33-b4:

^The most widely accepted supplement is Vahlen’s <6p0d>c,fffiapre S’ £v r fiXfificraoBax Sx’ >, although Vahlen himself emphasized the tentativeness of his suggestion. 74 Mistakes occur even in that which is in accordance with art. Men who possess the art of writing have written incorrectly, doctors have administered the wrong medicine. So clearly the same is possible also in that which is in accordance with nature. If it sometimes happens over things which are in accordance with art, that that which goes right is for something, and that which goes wrong is attempted for something but miscarries, it may be the same with things which are natural, and monsters may be boss shots at that which is for something.25

But if objections to poetics per se are determined as such in relation to the art’s final

cause, the latter does not come into question in accidental objections. On this view

Aristotle can both praise and blame mistakes concerning other arts, although he

undoubtedly prefers an impeccable poetic product as shown above. Indeed, the

severe restriction at Poet. 1451a26-28 manifests Aristotle’s scientific unwillingness to

tolerate any technical mistakes in poetry26.

It is worthwhile to pay closer attention to Aristotle’s attitude towards poetic

faults and the concomitant objections to the art. Obviously the careful distinction between two classes of faults was meant not only to blunt criticisms of the

Protagorean type but primarily to forestall Plato’s virulent attack on poetry in the

Republic. When Aristotle declares in Poet. ch. 25 that correctness is different in poetics and other arts like biology, medicine and politics, i.e. ethics, he must have in mind on the one hand Plato’s depiction of the poet as an imitative charlatan who

tran sla tio n by Charlton (1970).

“This passage speaks against Rosenmeyer (1974) 237 who assumes that for Aristotle certain "distortions of empirical reality or orderly procedure are not only useful but even obligatory for the poetic r € xVTl"; what Aristotle seems to be saying is that technical mistakes are sometimes useful but they are by no means obligatory. 75 pretends to know every art {Rep. 598c6-d5) and on the other Plato’s criticism of poetry on ethical grounds. Plato of course is never mentioned in thePoetics but the definition of poet as an imitator like a painter or any other image-maker in the opening lines ofPoet. ch. 25 betrays clear Platonic influence which makes Aristotle’s own position all the more poignant. It is certain, moreover, that Aristotle came to grips with his teacher’s theory of poetry in the lost dialogueOn Poets material from which was later used in the Poetics. So far it is clear that Aristotle’s critique of

Protagoras hinges on a delineation of the "theorems" that belong to the various arts.

In what follows I will argue that this point runs parallel to the delimitation of the art of poetics and in all probability provided Aristotle with a powerful argument against

Plato’s frontal assault.

The cynosure in the final remark in Aristotle’s criticism of Protagoras is certainly the term "theorem":

One kind of inquiry (0eo>pia) about diction concerns the moods of speech whose knowledge belongs to acting and those who possess this art, i.e. what is command, wish, narration, threat, question, answer and things of this sort. But no worthwhile objection can be brought against poetics because of knowledge or ignorance of these things. For, who would agree with Protagoras that Homer floundered in saying ""sing, o goddess, the wrath"" because he commands instead of praying? As Protagoras says, commanding to do or not to do something is an order. Therefore it should be left aside as theorem not of poetics but of another art {Poet. 1456b8-19).

It is immediately clear that Protagoras’ target is the poet’s ignorance of the moods of speech and that "theorem" answers to its cognate Qeatpia. The second point is particularly relevant to Aristotle’s implicit separation of the two arts: as shown above,

Aristotle insists that not only mathematics but also humble arts like weaving and 76 baking "inquire" (cncorrei) into what is their own per se, i.e. their final cause (Met.

1064bl9-23). Thus the object of inquiry already distinguishes the two arts and it

cannot be accidental that the object of inquiry in the productive arts, as in the hard

sciences, is delimited by definitions from which follows a demonstration of relaxed

rigor (Met. 1063b36-1064a8). Neither demonstration nor definitions are foreign to

the final cause in productive arts. In PA 639b 15-19 the definition of the product is

said to be the principle based on which artists give an account of what they do and

how they do it, whereas inEN 1112all-27 this account is explicitly paralleled to the

mathematical method of analysis, the initial hypothesis being the final cause of the

art. I will return to these passages in the next chapter. Here it suffices to say that

they furnish a mathematical model for the two-way process in artificial production

discussed in Met. Z: the process ofnoesis starts from the principle, e.g. the definition

of health which Aristotle identifies with the art itself and its final cause, and by

successive steps reaches a first cause, the first stage in the reverse process ofpoiesis which is directed towards the realization of the final cause (1032a32-b23).

Implicit in the analogy between mathematical analysis-synthesis and productive noesis-poiesis is the conception of an art’s actualized final cause as a mathematical theorem. If Aristotle employed the term "theorem" in this sense atPoet. 1456bl9, this would indeed square with his division of objections into per se and accidental with respect to the final cause of the art. A unique association of theorems and productive arts in MA ch. 7 bears out this conclusion. It is certainly not accidental that this passage overlaps with EN 1112all-27 since both provide the same 77 mathematical model for action, analysis and synthesis, in the context of productive arts. But in MA ch. 7 Aristotle fleshes out this model in explicit syllogistic terms:

I need a covering and a cloak is a covering, therefore I need a cloak; I have to produce what I need, therefore I need to make a cloak. This conclusion (ovfjLTripaona) is an action which results from the principle: if there is to be a cloak, such a thing is necessary and, if this is necessary, then this and one immediately takes action. It is clear that the action is a conclusion and the productive propositions are of two kinds, the good and the possible.

Aristotle draws a one-to-one analogy between these productive syllogisms and those about the immovable mathematical objects where "the theorem is a final cause

(because, when the two premisses are grasped, [the mathematician] grasps and puts together (ovviOrjicev) the conclusion)". Since "putting together" is the terminus technicus for the synthesis in a mathematical proof(Top. 175a26-28) which models the productive process, Aristotle would not have any problem characterizing the actualized final cause of an art as a "theorem".

For Aristotle the illegitimacy of the Protagorean objection lies exactly here: drawing on the final cause of another art violates the stricture in Met. 1064b21-23 that shoe-making, weaving and baking inquire only into their own final cause. As shown in the preceding paragraph, Aristotle gives to this inquiry into the final cause the concrete meaning of a "relaxed" demonstration proceeding from first principles.

For him now the Protagorean objection is illegitimate because it does not start from the principle of poetics and "arguments which do not start from the principles proper to a particular subject are empty, appearing to be relevant when in fact they are not" 78 (GA 748a9-12). Aristotle elaborates on this issue in AnPost. 1 ch. 12:

If a deductive question and a proposition of a contradiction are the same thing, and there are propositions in each science on which the deductions in each depend, then there will be a sort of scientific question from which the deduction appropriate to each < science > comes about. It is clear, therefore, that not every question is geometrical (or medical -and similarly in the other cases too), but those from which either there is proved one of the things about which geometry is concerned, or something which is proved from the same things as geometry, such as optical < theorems >. And similarly in the other cases too. ...We should not, therefore, ask each scientist every question, nor should he answer everything he is asked about everything, but those determined by the scope of his science. If one argues in this way with a geometer as geometer it is evident that one will do so correctly, if one proves something from these things; but, otherwise, not correctly. And it is clear that one does not refute the geometer either, except incidentally. (77a35-bl2)27

The scientific questions are attempts to refute a scientist dialectically(AnPost. 77b 11) and the accidental refutations at AnPost. 77bll-12 parallel the accidental objections in the Poetics, not only because they do not stem from the principles of a particular science or art but also because Aristotle’s discussion of objections draws heavily on his theory of dialectical refutation. Concerning contradictions in poetry he explicitly refers to the discussion of dialectical refutations (Poet. 1461bl5-18) and the categories of available means for refutation of objections (cf. Siaktieiv in Poet.

1461al0 ~ Top. 175a300) are the same as those for the solution of dialectical refutations (Poet. 1461a21ff. - Top. 179all-25). In this light Aristotle’s concluding remarks at AnPost. 77b 16-27 characterize clearly both Protagoras’ mistake and what poetic knowledge or ignorance consists in:

tran slatio n from Barnes (1994) 79 Since there are geometrical questions, are there also non-geometrical ones? And in each science which sort of ignorance is it in regard to which they are, , geometrical? And is a deduction of ignorance a deduction from the opposites (or a paralogism, though a geometrical one)? Or < is it a deduction > from another art? E.g. a musical question is non-geometrical about geometry, but thinking that parallels meet is geometrical in a sense and non-geometrical in another way. For this is twofold (like being non-rhythmical), and one way of being non-geometrical is by not having < geometrical skill > (like being non-rhythmical) and the other by having it badly; and it is this ignorance and that depending on such principles that is contrary .28

If this is plausible reconstruction of the science-theoretic backdrop to

Aristotle’s criticism of Protagoras, it exemplifies a cornerstone of Aristotelian thought: already in theProtrepticus (fr. 5a) Aristotle distinguishes sharply bewteen a physical and an ethical argument, a point which returns in theRhetoric (1358a4-21).

Due most probably to the axiomatization of geometry, the autonomy of the sciences conflicted with Plato’s metaphysical program which provided the springboard for his famous attack on poetry29. As already shown, the Aristotelian conception of poetry runs counter to Plato’s depiction of the poet as a deceitful imitator who pretends to know all arts. This caricature is positively fleshed out in thePhaedrus where Plato concludes that, to the extent it can be developed by humans, the art of speech, both rhetorical and poetical, consists in being able to divide the beings (rd ovra) according to kinds and to subsume each one under a single idea (273el-4). His conception of artistic creation (£vre%vov, Phdr. 277bl-3) in speech is more amply

^Translation from Barnes (1994).

^Cf. the comments in Owen (1965) 140-142. 80 characterized later in the dialogue (Phdr. 277b5-c6):

Until a man knows the truth about each of the things about which he speaks or writes, and becomes capable of defining the whole in itself, and having defined it, knows how to cut it up again according to its forms until it can no longer be cut; and until he has reached an understanding of the nature of soul along the same lines, discovering the form which fits each nature, and so arranges and orders his speech, offering a complex soul complex speeches containing all the modes, and simple speeches to a simple soul - not before then will he be capable of pursuing the making of speeches as a whole in a scientific way, to the degree that its nature allows, whether for the purpose of teaching or persuading, as the whole of our previous argument has indicated.

Plato here shows the same interest as Aristotle in what constitutes an art but his account is at odds with Aristotle’s claim that correctness is defined differently in each art. Indeed, Plato makes clear that his discours de la methode holds for tragic poets like Sophocles and Euripides (Phdr. 268c5-d2), doctors like Eryximachus (Phdr.

268a8-b5) known from the Symposium, and orators like Lysias and Thrasymachus

(Phdr. 269d2-8) who had written a treatise on pity according to Aristotle (Rhet.

1404al3-15). For Plato what underlies the full-fledged art of the poet is the method of division whose subjects are the "beings"; Aristotle, though, assigns each mathematical and productive science, including poetics, to a circumscribed area of these beings but not qua beings (Met. 1025b3-18, 1063b36-1064b 10; cf. 1064b21-23).

It goes without saying that the metaphysical foundations of Plato’s conception of art force a totally un-Aristotelian correctness on poetry(Phdr. 278b7-dl):

So now we have had due amusement from the subject of speaking; and as for you, go and tell Lysias that we too came down to the spring and the sacred cave of the Nymphs and listened to speeches(logoi) which instructed us to tell this to Lysias and 81 anyone else who composes speeches(logoi), and to Homer and anyone else in their turn who has composed verses, whether without music or to be sung, and thirdly to Solon and whoever writes compositions in the form of political speeches, which he calls laws: if he has composed these things knowing how the truth is, able to help his composition when he is challenged on its subjects, and with the capacity, when speaking his own person, to show that what he has written is of little worth, then such a man ought not to derive his title from these, and be called after them, but rather from those things in which he is seriously engaged.

This passage parallels Aristotle’s discussion of objections and refutations inPoet. ch.

25 but by Aristotelian lights the recourse to the metaphysical and epistemological notion of truth is certainly tantamount to an accidental objection which leaves poetics per se unscathed. A single, let alone metaphysical, foundation of poetry, rhetoric, legislation and political theory cuts across not only Aristotle’s separation of poetry and political theory but also his careful distinction between rhetoric and political theory (Rhet. 1356a25-30).

As shown above, Aristotle dissociates poetics and rhetoric so that, given the similar status of rhetoric and dialectic in the system of Aristotelian sciences, his refutation of Plato’s position is well encapsulated inRhet. 1358a21-26. Being only a metatheoretic tool, dialectic cannot make one knowledgeable about any genus which determines the scope of a particular art. Aristotle also cautions against an easy and almost unconscious transition from dialectic to particular sciences: by narrowing down the dialectical premisses, one can hit upon principles and thus, leaving dialectic, enters the science determined by these principles. In theAnalytics and the Topics such an unwarranted transition is put on a par with objections of the

Protagorean type. Since Aristotle specifies that the productive premisses in a 82

"relaxed" noesis-poiesis theorem come from the kind of "good", good here being the final cause of the art in question (EN 1094al-9) which amounts to the genus of hard sciences, Aristotle has to reject wholesale the Platonic foundation of poetics. The status of dialectic in the Poetics accords with his conception of this "science", a metatheoretic tool to refute objections to the art, not a foundation. At the bottomline, however, both Plato and Aristotle reflect the same interest in what poetics is and both move beyond intuitive approaches to art. CHAPTER ffl

NECESSITY AND PROBABILITY IN THE POETICS

The modal notions of probability and necessity, which play a prominent role in Aristotle’s characterization of the mythos, first appear atPoet. 1450b26-31 when, after the definition of tragedy, Aristotle fleshes out the notion of "whole" (oAov):

The "whole" is that which has beginning, middle and end. Beginning is what does not follow necessarily (e£ dvdyKtjc) after something else but after it something else is or becomes; end, on the contrary, is what itself follows after something else either necessarily or for the most part (e£ dvdyicrjc r) o>c rdnokti) but nothing after it does so; middle is what itself follows after something else and something else follows after it.

As shown in the previous chapter, the arrangement of parts in a specific manner is a metaphysical notion which, along with "unified", "complete" and "whole", characterizes the mythos as form. On the contrary, the modalities, i.e. necessity and what happens for the most part, do not play any role in this characterization. At first glance modal terms do not seem to mesh well with poetics -in Met. A.5 Aristotle offers a detailed discussion of all senses of necessity, none of which seems to bear

83 84 on poetry. Thus one is tempted to conclude that in thePoetics necessity and its weaker companion appear in a commonsensical, informal way, simply underlining the

coherence of the mythos.

In a recent paper Frede has argued eloquently that the modalities in

Aristotle’s Poetics have a far richer background. Her argument depends (a) on the deterministic nature of Aristotle’s ethics which attributes to human action a necessity directly comparable to the necessity of natural processes1; (b) on the peculiarity of the tragic poet’s task which is to present the unusual as necessary, or at least as likely, since the events in a tragic plot are far from the ordinary human experience,

(b) arises from what Frede views as a terminological shift in Aristotle’s employment of the modals2: although in Poet. 1450b30 he disjuncts necessity and what happens for the most part, from that point on he consistently substitutes the probable (eiicoc) for the latter3. For Frede the modalities are ultimately related to Aristotle’s requirement for universality in tragedy:

What we have so far ignored is the fact that the poet’s striving for universality is of a quite special kind. Firstly, the events and circumstances that form the plot are far from the usual human experience. Secondly, the persons involved, the agents, are not the average types of human beings. It is, rather, the poet’s task to present the unusual as necessary or at least as likely. This, to put the cards down finally, must be the peculiarity in tragedy that explains the terminological shift to"eikos". A scientific "statistical" conception of what "happens for the most part" would certainly

•Frede (1992) 203-204; see EN 6.2 and Met. 1048al0-16.

Trede (1992) 206-207.

3Poet. 1451al2, 38, b9, 35, 1452a24, 1454a34, 26. 85 not do to capture the poet’s striving for the "unusual likeliness"; nor, of course, could it be replaced by the rhetorician’s general probabilities based on endoxa. The specific aim of tragedy, to find convincing solutions for unusual problems, explains why Aristotle has to qualify his demand for anecessary coherence of the events that form the structure of a plot. Since such events fall outside most of our human experiences, it would often be quite unreasonable to ask for a stronger criterion than the "likely" in our judgment. In extreme cases we have to stretch all our ethical creative imagination to determine whether a poet has fastened on "the likely" action. To call for necessity may be to ask for the impossible or the meaningless.4

To start with the distinction between "for the most part" and the probable, if the former is understood as a statistical approximation to necessity5, it is hard to see why it could not square with (a), all the more since the necessity of human action cannot of course be necessity in the strict sense6. Frede sees the distinction between the

"probable" and "for the most part" in Rhet. 1357a34-bl:

The probable is what happens for the most part (t66> q ew\ rd iroki> yivopevov), not, however, simpliciter, as some define it, but what is about things that can be otherwise, being to that in relation to which it is probable as the whole to the part.

4Frede (1992) 209-210.

Tliis interpretation has been put forth by Mignucci (1981) 173-203 (cf. Frede (1992) 209-210); for objections see Judson (1991).

6Frede does not explain how (a) fits in with her conclusions. This is actually part of a broader problem in her argument, i.e. the absence of a precise delinetation of what the "unusual likeness" consists in. For Frede "likeness" is parallel to the "convincing solution" for unusual problems but it is not clear how she understands "convincing" either. She explicitly dissociates the "unusual likeness" from the statistical conception of "what happens for the most part", the relaxed sense of necessity that governs human action in (a), so that one cannot explain "convincing" by invoking the relaxed necessity of human response to the unusual problem. Frede (1992) 200-201 herself characterizes "what happens for the most part" as "usual" and thus by her lights the "unusual likeness" falls outside the scope of "what happens for the most part". 86 This passage would indeed suggest a difference between the "probable" and "for the most part", were Aristotle here speaking ofhis own distinction between the terms in question. Aristotle, however, does not propose a more relaxed sense of "for the most part", to be identified with the "probable", but he simply distinguishes between his own conception of the "probable" and the conception of some anonymous "others".

That Aristotle himself does not see any difference between the two terms is clear from AnPr. 70a4-5:

Probable is what is known to happen or not happen, to be or not be, for the most part.

The presence of "what happens for the most part" in thedefiniens of the "probable" is enough to dispel any suspicion of a terminological shift that could affect the interpretation of the modals in the Poetics.

More problematic is Frede’s claim that for Aristotle the tragic "poet’s task is to present the unusual as necessary or at least as likely". Nowhere in thePoetics does

Aristotle construe explicitly the poet’s task in this manner, although he certainly claims that tragic characters are far beyond the usual human experience. One of course might claim that this extends naturally to whatever befalls such extraordinary characters. Furthermore, Frede’s interpretation might be thought to find support in

Aristotle’s requirement that tragic poets prefer "impossible probabilities" (&56vara eiK dra) to "improbable possibilities" (fiuvara dTriSnva) (Poet. 1460a26-27). But does this requirement necessarily imply that the probable or plausible forms a facade 87 to mask impossiblities, i.e. Frede’s tragic circumstances that are far from the ordinary

human experience?

There is sufficient evidence to show that, put in the right perspective,

Aristotle’s claim in Poet. 1460a26-27 is severely restricted. According to Poet.

1451M1-19 tragic poets, unlike iambographers and comic poets, adhere to the actual

or "historical" names (r 6 v yevo nivav ovo/xdrwv dvr rax, Poet. 1451b 15-16).

The reason is that, what is plausible (7rtSt:v6v),is possible (Su v o -t o v ). On the other

hand, the events that have not already taken place (rd /xr) yevoneva) are not

thought to be possible (Sward) but those which have already happened (to-

yevo/xeva) are clearly possible (6uvard) since, had they been impossible

(dt66vara), they would not have taken place (ou...dv iyivero). This account

clearly explains the remark in Poet. 1453al8-21 that tragedy mainly revolves around

the stories of a limited set of mythological families (cf. Poet. 1454a9-10). Since,

moreover, Aristotle explicitly remarks that tragedies drawing on this mythological

material are the best (Poet. 1453al8-23), Poet. 145 lb 11-19 is clearly prescriptive with

respect to the relationship between possibility and plausibility that Aristotle postulates. But this passage leaves no doubt that Aristotle by no means intended the

"impossible plausibilities" in Poet. 1460a26-27 as a normative constituent of tragedy:

on the contrary, he posited an isomorphism between the possible and the plausible so that the latter cannot be plausibly said to make acceptable anything like Frede’s

extraordinary circumstances.

Frede supports her claim that for Aristotle the "unusualness" is an important 88 element of tragedy by appealing to his "positive appreciation" of the irrational, the

marvelous and the surprise7. But this is not an accurate reading of Aristotle’s

relevant comments. Immediately after Poet. 1450a26-27 he banishes irrationalities

from speeches in a rhetorical fashion (roCc re A,6youcpi) ovvioraoQai £k peptbv

dAoytov, Poet. 1460a28-29) and relegates such elements outside the plot, if they

cannot be altogether avoided (Poet. 1460a28-29). He also remarks that the

irrationalities in the Odyssey would have been unbearable without the artistry of

Homer who takes care to "sweeten" the irrationalities by other means (Poet. 1460a36-

b2). In the next chapter Aristotle designates irrationalities as mistakes concerning

the art itself (jrpdc aurf)v rf)v rexvrjv); he accepts such mistakes only when they

contribute to the achievement of the art’s final cause but he also makes clear that,

if possible, all mistakes should be avoided (Poet. 1460b22-29). It is important to note

that Aristotle prefers to banish the irrational altogether or at least relegates it

outside the mythos, obviously so that it does not interfere with the tight nexus

governed by either necessity or probability. Thus one should not assume all too

easily a positive appreciation of the irrational on Aristotle’s part.

This conclusion extends naturally to the marvelous (Bau paor o v), although prima facie Aristotle seems to be favorably disposed towards it: in Poet. 1460all-12

he remarks that the marvelous is appropriate to tragedy whereas inPoet. 1452al-ll

the marvelous is said to contribute significantly to pity and fear. Nevertheless,

immediately after Poet. 1460all-12 Aristotle adds that the marvelous is primarily

Trede (1992) 218-219 n.35. 89 effected by the irrational (Poet. 1460al3-14) and thus, in light of his negative attitude towards the irrational, he cannot accept the marvelous as unqualifyingly appropriate to tragedy. On the same grounds, how could the marvelous assist in the production of pity and fear? For Aristotle these must be the effects of themythos (Poet. 1453bl-

3) while the irrational, which primarily generates the marvelous, must either be absent or relegated outside the mythos. Thus, when Aristotle in Poet. 1452al-ll remarks that the marvelous which contributes to pity and fear must arise unexpectedly ^romwithin the plot (7ra:pd rr)v So^av Si’ akkrjka) and not by chance

(drrd roii avropdrov icai rrjc tC%r)z), he cannot talk about the marvelous which is the par excellence product of the irrational according toPoet. 1460al3-14. It also follows that it is a mistake to understand Aristotle’s comment on surprise

(£K7rA,r]£ic) inPoet. 1455al6-17 as parallel to his remark on the irrational. Aristotle here comments that the best recognition must arise from the mythos itself (££avrtbv ruv npa-fp&Tcsv) as the element of surprise is produced5i ’ eucoTwv.The crucial point here is the prepositional phraseSi’ eiKorovwhich is clearly parallel to5 i’ akkr)ka in Aristotle’s comment on the marvelous in Poet. 1452a4. Since these prepositional phrases undeniably pick out the causal chain in themythos according to necessity or probability (cf.Poet. 1452al8-21), both the surprise and the marvelous must be distinct from the irrational which lies outside the mythos.

The above discussion has made clear that Frede cuts across two fundamentally distinct categories: for Aristotle both the marvelous, at least in some sense, and the surprise must be governed by the modals whereas the irrational is at 90 best alien to them. Thus the irrational cannot be plausibly said to explain the

presence of the modals in the Poetics nor is there any indication in the text that the

marvelous and the surprise can account for it. As far as the marvelous is concerned,

it seems that one should distinguish between marvelous in the strong sense, i.e. as

direct product of the irrational according toPoet. 1460a 13-14, and marvelous in the

weak sense, i.e. as governed by the modals according to Poet. 1452al-ll and thus

contributing to the mythos. In this weak sense the marvelous must mean "contra

expectum", which obviously need not compromise the unfolding of the mythos

governed by the modals. This seems to be an adequate explanation of how the

marvelous is appropriate to tragedy but a problem still remains open. How can

Aristotle insist on the priority of "impossible probabilities" over "possible

improbabilities" since he puts so great an emphasis on the modals? What is

impossible can obviously be neither necessary nor probable as is defined in AnPr.

70a4-5 and, in light of Poet. 145 lb 11-19, these "impossible probabilities" must be

pathological exceptions to the isomorphism between "possible" and "probable". But

how can probability lend itself to the impossible? As will turn out, the answer to this

question sheds some light on the meaning of the modals in the Poetics.

Starting with the characterization of impossibility in Poet. 145 lb 11-19, it is important to note that Aristotle’s concluding remark ("obviously events that have taken place are possible -because they would not have taken place if they had been impossible") presupposes his assumption that possibilities can be assumed to be, and are, realized at some particular point in time, a characterization which is also 91 interpreted semanticallyin AnPr. 34a8-15:

... What is possible, when it is possible for it to be, could happen, and what is impossible, when impossible, could not happen... And we must take impossible and possible not only as they apply to coming about but also as they apply to being true and to belonging and all the other ways "possible" is used (for it will be similar in all of them).8

In view of Aristotle’s semantic interpretation of impossible, the problem of how the impossible can be integrated within the structure of the mythos is reduced to the problem of how falsities can mesh with the mythos. Aristotle addresses this issue explicitly inPoet. 1460al8-25:

Especially Homer has shown other poets how they should go about falsities. This is the paralogism (irapaXoyiopoQ). Because, when, if this is or becomes, that is or becomes (orotv ronSl 6vroc ro$t yivoMevou Ytvrjrai), people think that, if the second is, then the first is or becomes. This is falsity. For this reason, if the first is false and the second is or becomes necessarily if the first is (av rd 7rpa>TOv tyJHtoc, &XXo Si t o 6 to v bvroQ dvdyKrj eiv ai fj yevioQai jj), this second should be added. Because, since the soul knows that this is true, it makes a paralogism inferring that the first also is.

Particularly interesting here is Aristotle’s explanation of poetic falsity or deception as a paralogism which in Top. 101 a5-17 is defined as an inference from false

The stronger version of this thesis which Aristotle implies in Poet. 145lb l 1-19, i.e. that possibilities are actualized at some moment, is formulated in Met. 1047b3-5 ("it cannot be true to say that this is possible but it will not happen"); see Hintikka (1973) 107-109 (on the relation between the weak and the strong version of this assumption see Hintikka (1973) 109-110). The unactualized possibilities of the De Interpretatione do not cause any problem here; see van Rijen (1989) 95ff. 92 premisses:

In addition to all syllogisms already discussed there are the paralogisms which arise from what pertains to a science itself, as is the case with geometry and sciences relative to it. The paralogism indeed seems to differ from the syllogisms already discussed: one who proves falsely draws conclusions neither from true first principles nor from reputable opinions. For a paralogism does not fall under the definition [given above]. It does not take what seems to be the case either to everybody or to the majority or to the wise, either all of them or the majority or the most important among them, but deduces conclusions from premisses that pertain to a science but are false. One makes a paralogism either by circumscribing semicircles the wrong way or drawing some staight lines as they should not be drawn.

Aristotle’s strategy is typical of his overall approach to poetry: although he naturally

considers Homer the master of poetic falsity, he fleshes out the technique with which

he credits Homer in starkly non poetic terms construing it as a paralogism. From a

logical point of view, moreover, Aristotle’s account in Poet. 1460a 18-25 allows for a

true conclusion from false premisses9 and in this light it does not seem accidental

that he insists on "impossible probabilities" at the expense of "improbable possibilities" (Poet. 1460a26-27) immediately after Poet. 1460al8-25. Although there

is a simple paratactic connection between Poet. 1460al8-25 and 1460a26-27, the evidence suggests that the latter can be construed as the conclusion or the aftermath of the former: "impossible probabilities" presents no oxymoron inasmuch as, on the semantic interpretation of the modals, it captures Aristotle’s conception of poetic falsity as a paralogism, an inference of something true, and thus possible or probable,

9See Aristotle’s discussion in AnPr. 53b26-54a2; the issue in discussed in Patzig (1968) 196ff. 93

from something false and therefore impossible10.

This is certainly pertinent to the appreciation of the modals in thePoetics: if

Aristotle has recourse to his logic in order to account for what falls outside the scope

of the modals, the modals themselves must be couched in logical terms as well. The

strongest evidence in support of this conclusion is the fact that inPoet. 1460al8-25

necessity appears in the phrase&kko Si to6tov 5vto q dv&yicn elvai, Aristotle’s

formula in the Analytics 11 for the necessity of a conclusion relative to the

premisses12. InPoet. 1452al8-21, moreover, Aristotle stipulates that both recognition

and tragic reversal arise from the mythos itself so that they result from what has

come before according to either necessity or probability: c k t & v irpoyeyevrjuiwv

CTDju/Saiveiv f) e£ dvdyicrjc r) Kara t6 eitcdc yiyveaOca raurcrSiaQipei yap

jroA.fi t 6 yty vea0ai r&Se S\& r&Se fj nera r&Se). Since in Poet. 1460al8-25

Aristotle does conceive of poetic falsity in terms of logical inference, it is not

10Thus the preference of "impossible probabilities" over "improbable possibilities" is justified because in making the paralogism one is not aware that the premisses are false, i.e. impossible. In accepting something impossible, and thus false, one is not necessarily aware of the falsity; therefore, since the falsity entails something true, one might be led to believe the former on the strength of the latter (as will be seen below, Aristotle shows with a mathematical example how easy is to make such a mistake). "Improbable", however, involves lack of belief, justified or not, (Rhet. 1391b7-8) and thus one cannot derive any possible or true consequence.

uCf. toi>61 6vtoc rofii r) in Poet. 1460a21 and to 6 tg > v 6 vrav dvayicaiov i o r a i r6 ovfiiripaona in AnPr. 30b38-39, t o o A S v t o c avdynrj rd B elv ai in AnPr. 34a5-6, 6 v t o c roii A rd B elvai in AnPr. 34al6, ^Keivo elvai rwvfii 6vro>v in AnPost. 91bl4-15 and rivwv 6vtwv dvdyK rj t o u t ’ eivai in AnPost. 94a21-22. As Aristotle remarks in AnPr. 34al6-24, in this formulation both premisses are taken as a unity.

12For a discussion of relative necessity see Patzig (1968) 16ff. 94 accidentalthat^Kra>VTrpoYeYevf7M^v

r&Se reflects almost verbatim the definition of syllogism at AnPr. 24bl8-22

(ovkkoyiondQ S i icrr i Xoyoc iv reQivrcsv rivwv Irepov ri rwv KeijuIvov

avd-YKfjC au/xj0oCvei...5id raura).

There arises of course the question how "what happens for the most part" or

probability fits in here but, leaving it aside for the time being, there is sufficient

evidence that in the Poetics Aristotle employs necessity in the sense of apodeictic or

syllogistic necessity, contrary to Frede who, in her attempt to decide the sense in

which necessity is employed in thePoetics, remarks:

Of these four kinds we can ignore the necessity of force; we can presumably also leave aside without discussion the "apodeictic necessity," since it is highly unlikely that Aristotle assumes either a compulsory or a syllogistically derived necessity for the coherence of drama. Can we also set aside the hypothetical necessity? Under "hypothetical necessity" Aristotle does not understand the "forward looking" kind, which leads from the condition to the consequence, but the "backward looking" one, of what is necessary if some purpose or end is to be achieved. Although the hypothetical necessity must play some role in drama, since all teleologically determined processes rest on necessary preconditions, it cannot be the necessity Aristotle speaks of in his definition of the necessary unity inPoetics 7, quoted above. Not only does he nowhere add any explanation that his talking of the necessity of the means to a given end, but he actually indicates quite clearly in our passage of the Poetics (1450b27 ff.; cf. 1452al8~22) that he has the forward looking, unqualified, necessity in mind: the beginning must be such that it leads necessarily or naturally to what follows after it.13

Now in view of the above evidence it seems not only that Aristotle resorts to

apodeictic necessity in thePoetics but also that hypothetical necessity is indispensible

“Frede (1992) 199. 95 to a plausible account of necessity in the treatise. In the passages cited by Frede

Aristotle does undeniably have in mind the forward looking necessity; in his

characterization of poetic falsity, however, he appeals undeniably tohypothetical

necessity as well. This is clearly suggested by the expressionsrovSl Svtoc roSl

f) and &XXo to6toi> 6 v t o q dvdytcrj elvai: starting from a false A, if a true B

follows necessarily, the soul concludes mistakenly that A holds as well because of the

true B, an obvious paralogism since the principle in this inference is false. The

retrogression from the conclusion to the premisses appears as an example of

hypothetical necessity inPhys. 200al9-26:

With things which come to be for something the case is reversed: if the end will be or is, that which comes before will be or is; and if we do not have it, then just as in mathematics, if we do not have the conclusion, we shall not have the starting-point, so here we shall not have the end or that for which. That too is a starting point, not of the practical activity, but of the reasoning. (In mathematics too the starting point is of the reasoning, since there is no practical activity there.) So if there is to be a house, it is necessary that these things should come to be or be present, and in general it is necessary that there should be the matter which is for something, e.g. the bricks and stones if there is to be a house.14

This characterization of hypothetical necessity suggests that inPoet. 1460al8-

25 Aristotle operates with a mathematical model and this impression is fully borne out by AnPost. 78a6-10 where he allows for true conclusions from false premisses and illustrates this situation with the difficulty of mathematical analysis, thepar excellence heuristic method of Greek geometry:

“Translation from Charlton (1970). 96 If it were impossible to demonstrate something true from false premisses, then [geometrical] analysis would be easy; for in this case conversion would be necessarily guaranteed. Because assume that A is; if this is, then these are (rofiroo S ’ o v t o c ro 6 l Ic t t iv) which I know that they are, e.g. B. Therefore from these [B] I will show that this [A] is.

As is shown by the expression t o C t o u S ’ 6 v t o c to S I I c ttiv , Aristotle here fleshes

out in terms of hypothetical necessity a double-pronged mathematical technique which works along the same lines as the characterization of poetic falsity inPoet.

1460al8-25: in order to prove A the mathematician assumes A and derives a consequence B whose validity is already known; to complete the proof, a synthesis must follow, i.e. from B to show A, but that is not always the case as Aristotle remarks15. In Aristotelian terms this technique involves both hypothetical and forward looking necessity.

The parallel between Poet. 1460a 18-25 and AnPost. 78a6-10 is interesting, not least because geometrical heuristics and their retrogressive validation of the premisses upon the strength of the conclusion were definitely unknown to Homer whose technique Aristotle purportedly fleshes out inPoet. 1460al8-25. Like the explanation of the limited mythological repertory of Greek tragedy, this is another example of how Aristotle’s concerns in the Poetics differ from what later literary criticism studied. There is no doubt that the model on which Aristotle bases his understanding of poetic falsity suits his intentions, i.e. to account for thebelievability

15The most complete discussion of the method of analysis is Hintikka-Remes (1974). 97 of poetic falsity. The point inAnPost. 78a6-10 is the characteristic difficulty of

achieving conversion in geometrical analysis which amounts to showing the truth or

falsity of the premisses. The same point returns inTop. 175a26-28 where the

difficulty of conversion illustrates the difficulty in revealing the flaw in some

dialectical arguments. By means of this analogy the believability of poetic falsity

comes to the fore in the sharpest way possible: influenced by the true consequences,

the soul fails to realize the falsity of the antecedent as easily as a geometer might fail

in the conversion that would conclude the analysis.

In this context one can easily account for both the forward looking necessity

in the passages cited by Frede and the obvious reference to hypothetical necessity in

Poet. 1460al8-25: both types of necessity are complementary in the method of

mathematical analysis which underlies this passage. In Aristotelian terms the most

plausible explanation of the presence of necessity in thePoetics is found here. As

Aristotle himself claims, the account of poetic falsity simply fleshes out Homer’s

exemplary technique that poets should study and, consequently, hypothetical necessity

must enter Aristotle’s account of a poet’screative process. Similarly, forward looking

necessity in Poet. 1450b27-34 governs the construction of the mythos which is the

focus of a poet’s creative activity according to Aristotle. But, as has been shown in

the previous chapter, the abstract model of creative or productive activity with which

Aristotle operates is exactly the mathematical technique of analysis and synthesis. In

Met. Z.7 the productive activity of a doctor falls into the complementary processes

of noesis-poiesis but in EN 1112bll-27 mathematical analysis and synthesis offer a 98 model for the activity of doctors, rhetoricians and politicians: Aristotle explicitly

compares their technical activity to the way a geometer attacks a problem by means

of the method of analysis and synthesis, the former parallelingnoesis and the latter poiesis in artificial production. The same point is repeated inMA ch.7 where

Aristotle conceives of productive activity in terms analogous to a mathematician’s

proof of a theorem by the complementary processes of analysis and synthesis. PA In

639b5ff. and Phys. 199b 15-30 Aristotle discusses again artificial and natural

production in terms of hypothetical necessity which clearly recall its model ENin

1112b 11-27 and MA ch.7, the method of mathematical analysis, although here it is

not explicitly referred to. All this, it should be noted, accords well with the claim in

Met. 1063b36-1064a7 that all sciences, both productive and mathematical, carry out

"softer" or more rigorous proofs, as the case might be.

On the basis of the above evidence, necessity is imported in the Poetics from

Aristotle’s general conception of artificial or artistic production. But, since the

"proofs" of a poet must fall under the "softer" ones inMet. 1063b36-1064a7, necessity should be seen as a direct echo of Aristotle’s mathematical model -he himself warns that the degree of rigor each area admits should be observed diligently (EN 1094b19-

27). The addition of the mitigating "for the most part" or "probable" serves exactly this purpose, to indicate that a poet’s chain of reasoning does not measure up to the rigor of a mathematical proof. Indeed, inMA ch.7, after drawing a parallel between the proof of a theorem and productive activity, Aristotle remarks that the "productive premisses" come from the possible, not of course the necessary as is the case with the 99 premisses in mathematical proofs. The kind of relaxed demostration Aristotle has

in mind is suggested by the characterization of the universal (icaBoAou) in his well-

known comparison between poetry and history{Poet. 1451b8-10):

The universal is the kind of things suitable for a particular kind of persons to say or do according to probability or necessity -this is what poetry is concerned with adding names.

What is suitable for a particular kind of person to say or do according to probability

provides the premiss in an enthymem:

The enthymem is a syllogism from probables (£$ eiK6ro)v) or signs. Probable and sign are not the same but the former is a reputable proposition: the probable is what they know to happen or not happen, be or not be, for the most part (c rd ttoX6), e.g. that the envious hate or that lovers love {AnPr. 70a3-6).

Aristotle’s example of an enthymem does seem to capture adequately what poetry is concerned with according to Poet. 1451b8-10: Pittacus is generous, because the ambitious are generous and Pittacus is ambitious {AnPr. 70a24-27). If, in view of

Aristotle’s conception of productive activity, the poet constructs themythos by carrying out such "relaxed" demonstrations, then the presence of "for the most part" and "probable" in the Poetics is eo ipso accounted for because both modals appear in the definition of the enthymem.

Whether the mythos of a tragedy can actually be constructed by a chain of enthymematic reasoning in a two-way process ofnoesis-poiesis or analysis-synthesis 100 is questionable -Aristotle does not pretend that even mathematical proofs can be

formalized syllogisticallyin toto. Thus Aristotle’s introduction of the modals in the

Poetics does not reveal an interest in the coherence of tragedy or epic, intriguingper

se as this notion might be. His concern, which goes well beyond poetry, is rather to

conceptualize the process of production, both natural and artificial. One might

plausibly discern a Platonic legacy in Aristotle’s falling back on a mathematical

model -his conception of natural production in terms of geometrical analysis does

indeed seem to continue the "pangeometricalism" attributed to Plato in the tradition.

The method of analysis must have drawn Aristotle’s attention exactly because of its

heuristic character which kept the mathematicians’ interest in it alive till the

eighteenth century. In light of his conception of artificial production, mathematical

heuristics must have provided Aristotle with a coherent model for the understanding

of the creative activity in a poet’s mind. Although Aristotle’s achievement might

seem naive and methodologically unwarranted, its worth lies not so much in

Aristotle’s answers as in his attitude towards certain questions. It is important to

note that in the Poetics the only trace of his conception of artificial production in

mathematical terms occurs in his characterization of poetic falsity which is said to

affect the soul. Construed as a paralogism this poetic falsity is the direct descendant of the aesthetic deception, the notion that played a prominent role in Gorgias’ rhetorical-aesthetic system, assuming of course that Gorgias did articulate one. In the Helen Gorgias’ praise of XoyoQ reduces to a eulogy of noirfoiQ which, interacting with the 66£a of the soul, causes emotional alterations, enchantments 101 characterized as i|ru xnc a/japT^/Liara kc*1 So £ nc anarfifMara which amount to a ijreufir)^ Xoyoc (9-11). Gorgianic poetic deception hinges on certain psychological and epistemological assumptions but Aristotle achieves a delineation of the same phenomenon in markedly different terms: for the vague, intuitive Gorgianic concepts he substitutes precise logical terminology which, viewed against the implied background of his conception of artistic production, fleshes out the notion of poetic deception in a concrete way. The logical and mathematical apparatus does not of course imply either the "mathematization" of poetics or its being brought into a logicist program: in keeping with Aristotle’s denial of mathematical rigor in such disciplines, they should be rather understood as metalinguistic items that allow

Aristotle to comment effectively on the issue at hand -the same conclusion was reached in the previous chapter about the metaphysicalFuktionalbegriffe £v, 6A,ov and rikeiov that characterize the mythos. CHAPTER IV

ARISTOTLE AND THE TRAGEDIANS

The previous chapters do not cover all or even most of the major issues raised in Aristotle’s Poetics but they do reveal some important features of the treatise that have been largely overlooked in the secondary literature. What I have attempted to show is that Aristotle’s delineation of the mythos, the foundation of the Poetics, and thus Aristotle’s conception of tragedy, is primarilymetaphysical, not aesthetic. Via the definition of tragedy such a characterization serves a twofold purpose: first, to establish poetics as an art in the formal Aristotelian sense of the term that presupposes a clearly delineated subject matter expressed by the definition of the artistic product; second, to address Plato’s attack on poetry by falling back on the autonomy of the sciences which naturally extends to arts, given Aristotle’s unitary view of arts/crafts and sciences. As I have shown in the second chapter, Aristotle’s emphasis on the unity, completeness and wholeness of themythos is not an aesthetic but rather a metatheoretical precept on which his program is founded. In the same vein, the third chapter has shown how Aristotle’s modal characterization of the mythos reflects his abstract model for all productive processes which elsewhere is 103 explicitly couched in mathematical terms. In this light it is not accidental that, as has

often been noted, Aristotle’s concern with mythos in the Poetics does not coincide

with the agenda and practice of playwrights. Many tragedies, and especially many

that were considered exquisite in antiquity and are still very highly regarded today,

fall short of Aristotelian standards and are in general incompatible with the ideas

expounded in the Poetics. Now of course this is nothing strange for ancient literary

theorists whose concerns seem alien to those of their contemporary poets, or the

poets of any age for that matter, and can at best be relevant to only a very small

fraction of actual literature. Plato can easily be pointed out as an extreme example

of this tendency: blasting poets and poetry in theRepublic, he comes up with a very unusual criterion for successful poetry inPhdr. 277a9-278d6‘. No poet of course would be willing to comply with a requirement of this sort and its sheer

outrageousness makes it probable that Plato is merely being ironic in this passage, underlining, in Socrates’ facetious voice, the impossibility of acceptable poetry.

Along the same lines, Hellenistic theorists made pronouncements that are difficult to categorize and occasionally border on self-contradiction -I will discuss some of them in the next chapter. It is difficult for one to discern the relevance of these views to Greek literature. Pre-Hellenistic theorists at least observed generic distinctions, putting their remarks in a circumscribed generic context; but such distinctions collapsed for Hellenistic and later literary critics who talked in very broad terms, like "poem" or "good sound of poems", upon which some made the moral

‘See above p. 79-80. 104 value of poetry contingent and from which others, equally fiercely, dissociated it. As will become obvious in the next chapter, ancient literary criticism became more and more isolated in its own ivory tower, cut off almost completely from the literary practice of the day, as any, however superficial, comparison between theory and practice immediately reveals. The prescriptive mode came to impose itself totally on the descriptive. Critics debated several theoretical issues, even creating a body of novel terminology that is more often than not inscrutable to modern scholarship and caused problems to the ancient critics themselves. Time and again, accusations of obscurity, lack of specifics and outright insanity are hurled from, and to, all sides of the debate adding to the general confusion. The Hellenistic literary theorists, most of whom belonged to philosophical schools or had distinct, although not always known today, philosophical affiliations, were obviously guided by a desire to cut literature down to size: it had to comform with the tenets of the various schools and, above all, their moral principles. Previous literature was praised, tolerated or condemned, as the case might be, depending on how closely it was judged compatible with later standards of philosophical ethics.

In a way Aristotle was the precursor of this tendency, although his philosophical approach to poetry draws not on ethics but rather on his theory of science and metaphysics. He concentrated on dramatic poetry and examined epic mainly in dramatic terms to the complete exclusion of lyric poetry -in this he followed Plato and was not followed by later theorists, many of whose remarks appear to concern hymns and, in general, small poetic compositions rather than large 105 scale epics and tragic trilogies. At any rate, Aristotle’sPoetics was composed neither

as a young tragedian’s guide to quick success nor as an annotated top-ten list of the

best tragedies on the market. If one is willing to allow for more than Aristotle’s

strictly philosophical concerns, the work seems to have been conceived rather from

the viewpoint of the literary critic and the educated public who needed a solid

theoretical framework to study literature and come to conclusions that could safely

be thought to lie beyond subjectivity. Nevertheless, although in his characterization

of the mythos Aristotle does not extrapolate from actual tragedies and is almost

entirely prescriptive, he is not an armchair theorizer either. Due to the peculiar

nature of his surviving work his style does not easily lend itself to an appreciation of

his literary background, as Plato’s highly polished and sophisticated prose does. But

there can be no doubt that Aristotle had intimate knowledge of the Greek literary

tradition. Even if he had not read one thousand tragedies, as has been claimed2, he was definitely familiar with more playwrights and plays than we are today and

probably than many of his contemporaries were. Very compressed, laconic and even

obscure in certain points, the Poetics does not allow for comprehensive discussion of

individual poets and works. It is important to note, though, that Aristotle illustrates

all his major points with examples from actual plays, many of them now lost. He also

shows interest in, and knowledge of, the historical development of different genres

and at least some highlights of the career of eminent tragedians and comic playwrights. Thus it would not be completely accurate to claim that Aristotle’s

2See Cooper (1923) 12; cf. Gudeman (1934) 11-12. 106 treatise did not have anything to do with the actual composition and performance of

tragedies or that playwrights would not find anything of interest, and possibly of

value, in it. To illustrate the extent of the relevance of Aristotle’s remarks to Greek

tragedy, I would like to turn to the examination of two plays. Sophocles’Oedipus

Tyrannus is an obvious choice since Aristotle repreatedly praises it and undoubtedly

considered it a play that captured the essence of his treatise. As its polar opposite

I will discuss Euripides’Andromache, a play that has been considered least successful

by modern critics. This reaction might of course be an indication that Aristotle’s

precepts became particularly popular. But it is highly unlikely that Euripides would

have written an Aristotelian tragedy even if he had known Aristotle. Although

Euripides has had his critics from ancient times onwards, the fact that he has been

considered one of the most important tragedians clearly indicates that Aristotle did

not say the final word on worthy tragedy.

Having already discussed the constitution of good tragedy, and especially the mythos and its importance, fairly extensively, Aristotle returns to the composition of

a good, plausible mythos in Poet. ch. 17. It is obvious that he wrote this chapter with

the playwright and actual theatrical performances in mind. At the very beginning he stresses that for best results the playwright should visualize the unfolding of the plot so that inconsistencies and mistakes be avoided as much as possible. After this emphatic piece of advice he turns to the beginning of the enteprise and its most important part, i.e. the construction of an effectivemythos. How should a poet go about actually writing a play? According to Aristotle one should first isolate a 107 nucleus of the plot, a sequence of the crucial events that make up the core of the play without devoting any attention to circumstantial elements like names and

"episodes": the latter round off the play and obviously contribute to its effect but are not absolutely essential and unalterable. Aristotle of course immediately specifies that the episodes should be germane to the core mythos , i.e. they should happen according to probability or necessity3.

Sophocles’ OT can clearly fit in this mold. Its core story is very simple: a man finds out that, without knowing who he was, had killed his father and married his mother. Out of this bare skeleton Sophocles created one of the most intricate and powerful plays ever written. It is characteristic, though, that the effect of the play depends to a large extent on what Aristotle calls "episodes" that could be absent or different. But it is easy to misunderstand Aristotle here. He does not imply that episodes are inessential but he tries to free the creative process of anything that can obscure, or otherwise interfere with, the coherence and lucidity of the core mythos.

3Consider Aristotle’s succinct specifications inPoet. ch. 17: "Both the traditional stories and those one invents himself should be worked out in universal terms (kq:06A,oi>) and then one should flesh them out and add episodes. An example by what I mean by "universal" can be taken from the story of Iphigeneia: a young girl was sacrificed and vanished miraculously; taken to a foreign land where it was customary to sacrifice foreigners to a goddess, she becomes the goddess’ priestess; after sometime it happened that the brother of the priestess came there (that the god ordered him to go and for what purpose lies outside themythos ); they caught him and when he was about to be sacrificed he recognized his sister, the way either Euripides or Polyidus has it, after he said in a plausible manner that it was necessary not only for his sister but also for himself to be sacrificed, and thus he was saved. After one lays out such a universal outline one should add names and work out the episodes. And the episodes should be germane [to the mythos ] as in the case of Orestes’ seizure and the salvation through the purification. In drama the episodes are short but in epic poetry they are lengthy"(Poet. 1455a34-bl6). 108 Besides, since most Greek tragedies dealt with stories whose outline was fixed in the

tradition and had to be followed, as Aristotle himself specifies (Poet. 1453b22-23),

it would be easy for the playwrights to focus too much on details and lose sight of the

big picture along the way. Aristotle emphasizes that the coremythos has to be

clearly and firmly established before one turns to anything else.

At any rate, the OT is by Aristotle’s lights an exquisite tragedy. The mythos

is complex and includes a most unexpected and dramatic recognition and reversal.

Moreover, the action imitated is capable of inspiring great pity and fear as Aristotle

says (Poet. 1453b3-7). The episodes are not only crucial in delineating the figure of

Oedipus but, what is more important in Aristotelian terms, they also succeed each

other according to probability. The seriousness of the plague is emphasized as is

Oedipus’ wish to help again the city he had once saved. As becomes obvious from

the prologue, where a group of suppliants urges the king to succor the city, the people look to him for help and he tries as best he can to find a cure. When Creon, the king’s brother-in-law, returns from Delphi with Apollo’s command to banish the killers of Laius, the miasma that causes the plague, it is natural for Oedipus to turn to the seer Teiresias for help with finding the culprits. His reaction to the old man’s flagrant lack of cooperative spirit is perfectly understandable. When Teiresias, stung by the king’s abuse, reveals that Oedipus himself killed Laius and thus should be banished, the king is sure that he is innocent and naturally suspects a conspiracy spearheaded by Creon. The finger-pointing is facilitated by the small, and initially gratuitous-sounding, detail that it was Creon who suggested the interrogation of 109 Teiresias. But most significantly, the quarrel between Creon and Oedipus was bound

to involve Jocasta: naturally eager to find out the cause of the quarrel between her

husband and her brother she unwittingly drops the first hint that alerts Oedipus to

the fact that he might have been the killer, as Teiresias had suggested, and thus leads

to the unraveling of the mystery.

Sophocles’ major innovation of course was the introduction of the Corinthian

messenger who comes to Thebes to announce the death of old Polybus and

eventually reveals that Oedipus was not born into the Corinthian royal house but was

originally of Theban origin, exposed as an infant. After his conversation with Jocasta

and before the arrival of the messenger, Oedipus feared only that he was the killer

of Laius and had already sent for an old shepherd, the only survivor of Laius’ escort,

in order to have him testify whether Oedipus himself was the killer. Eventually of

course it turns out that this old Theban slave was not only the sole surviving

eyewitness of Laius’ murder but also the one who had given the infant Oedipus to

the Corinthian messenger, then a fellow shepherd, to raise in Corinth. The

Corinthian then gave the boy to his childless masters, king Polybus and queen

Merope.

Despite the plausibility of this intricate story which dramatizes Oedipus’ painful journey from ignorance to knowledge, the play is not free of some

illogicalities. After so many years as king of Thebes Oedipus seems to be almost

completely ignorant of the circumstances of his predecessor’s death; likewise Jocasta does not know the reason of Oedipus’ self-exile from Corinth. There is, moreover, 110 no cogent explanation why Teiresias would be so reluctant to reveal Oedipus’ misfortunes which are bound to come to light within the same day. It is also impossible for the chorus to know that Laius’ servant who gave the as yet unidentified infant to the Corinthian shepherd was the same person as the survivor of Laius’ company(OT 1051-1053). It is equally impossible for them to suggest that

Jocasta should know better about this incident: as yet neither Oedipus nor the

Corinthian messenger nor the chorus know that the boy was given to the Theban shepherd by Laius and they have no grounds for surmising that -on the contrary, they have every reason to suspect, as they and Oedipus himself immediately afterwards do, that the boy was of low birth. Why then would Jocasta, Laius’ wife, know? To make the connection one should have known as much as Jocasta: in that case one would have understood the truth like she did. But it is clear that the chorus has still no clue as to the identity of the king.

Aristotle himself recognizes in Poet. 1459b7-8 and 1460a29-30 that Oedipus' ignorance concerning the death of Laius is an irrationality but, since it does not belong to the mythos, it does not diminish the play’s excellence. Aristotle does not address the chorus’ inference about the shepherd (OT 1051-1053) which is beyond their knowledge. Irrespective of Sophocles’ reasons for including it in the play,

Aristotle would most probably consider it a venial mistake, comparable to Homer’s judicious use of irrationality and falsity discussed inPoet. 1460al2ff. Occurring at a highly dramatic moment, just before the revelation of Oedipus’ identity, the chorus’ inference contributes to the build-up of tension and the swift downfall of the I l l protagonists thus promoting therikoQ of tragedy4. Besides, it is unlikely that any

audience would pay attention to a minor detail that does not affect the coherence of

the play. By Aristotle’s lights, having put together a magnificentSia i< of the

drama, Sophocles proceeds to an equally impressive A. 6 a k. According toPoet.

1456a9-10 both 6to\$ and kfioiQ should be unobjectionable in order to have a good

tragedy.

By Aristotelian standards Euripides’Andromache could not possibly be

considered a successful tragedy. First of all, the mythos of this play is simple and

episodic, the most serious fault according to Aristotle{Poet. 1451b28ff.). In this play

Euripides fuses, quite adroitly but far from seamlessly, two traditionally unrelated

stories, the quarrel between Andromache and Hermione and Neoptolemus’ death.

According to the tradition Neoptolemus, the son of Achilles, was given Andromache,

Hector’s wife, as a choice spoil from Troy. On his return to Greece he reigned in

Epirus where his offspring by Andromache founded the Molossian royal dynasty.

Eventually he married Hermione, the daughter of Helen and Menelaus, and visited

the oracle of Apollo at Delphi, either to offer Trojan spoils to the god or to seek

revenge for the killing of his father by Apollo. He was killed there, according to

“Teiresias’ adamant refusal to point out the polluter, although he says that he knows him, serves a similar purpose. Sophocles could definitely have made the seer the person to reveal the misfortunes of the Theban royal house, probably at the end of the play. But then he would have had to find another, potentially less effective, way of showing Oedipus’ impulsive nature, apparent now from his unfounded incrimination of Teiresias and Creon, and bringing in Jocasta; in that case the dramatic recognition and reversal would have suffered and thus the play would have been less successful by Aristotle’s standards. 112 most versions in a quarrel with the Delphians over sacrificial meat, and his tomb was honored. Hermione then married Orestes to whom she had been originally promised.

In the tradition there is no indication that Hermione persecuted Andromache and her child(ren) and consequently that Neoptolemus’ death had any connection with, or occurred at the time of, the domestic quarrel. This persecution is Euripides’ main theme in his Andromache and he probably invented Hermione’s hatred and her attempts to eliminate her rival along with the child. At the opening of the play

Andromache summarizes her story for the benefit of the audience: she had been once a princess who married into the Trojan royal family and bore a son to her husband Hector. After the fall of Troy she lost everything and became the slave of the son of her late husband’s killer, Neoptolemus, to whom she eventually bore a son. Neoptolemus then married the Spartan princess Hermione and the young wife accused the slave Andromache of making her barren and alienating her husband from her with magic. During his absence at Delphi, where he went to propitiate

Apollo for his previous outrage, Hermione fetched her father from Sparta and the duo conspired to kill Andromache and her son. Now Andromache has sent the boy away to safety and she has taken refuge at the altar of Thetis. She soon learns that the boy’s hiding place has been discovered by her enemies. Her only hope is to send a messenger to reveal the plot to old Peleus, Neoptolemus’ grandfather. In the meantime Menelaus forces Andromache to abandon her haven by giving her the choice between her life and her son’s. Peleus’ last minute intervention saves both 113 and Menelaus leaves for Sparta. Hermione then becomes hysterical fearing that her

husband will punish her and tries to commit suicide. Orestes unexpectedly appears

and Hermione begs him to take her away from Neoptolemus’ palace. It turns out

that he has come exactly for that purpose, having arranged Neoptolemus’ murder at

Delphi. Orestes harbored an old grudge against Neoptolemus because Hermione

was first promised to Orestes, her cousin, and was then given by Menelaus to

Neoptolemus in return for his participation in the Trojan war. Orestes asked

Neoptolemus to relinquish his right to the bride but Neoptolemus insulted and

dismissed him. Soon after Orestes and Hermione leave, the body of Neoptolemus

is brought from Delphi and a messenger reveals the manner of his death. Old Peleus

laments bitterly but Thetis appears as dea ex machina and makes suitable

arrangements for all parties, including her old husband, her dead grandson, her great- grandson and Adromache.

The problems created by the merging of the two stories in a single mythos are evident in the Andromache. First of all, it is difficult to make sense of the chronological framework the poet assumes. According to the prologue Neoptolemus seems to have married Hermione some years after his return from Troy: in the meantime Andromache gave birth to a son who at the play’s dramatic time is at least four or five years old because he sings a small part and there is no indication that the quarrel between the two women was latent for some years and only surfaced recently; on the contrary, it seems that Hermione showed her hostility at the most months after her marriage. On the other hand, according to Orestes’ account, 114 Neoptolemus wed Hermione immediately afer the war when Orestes asked the lucky

suitor to relinquish his right to the contested bride. Immediately below in the

account, though, it appears that this exchange took place after Orestes’ matricide, i.e.

several years after the Trojan war, or that the return of Neoptolemus and Orestes’

matricide were roughly contemporaneous, an even more illogical assumption. The

years that separated the two events would be compatible with the prologue but there

is nothing in Orestes’ account to suggest that it took Neoptolemus and himself a long

time to lay claim to the bride. Euripides obviously did not worry much about

traditional chronology in his attempt to combine the two stories in the play. Another

liberty he took is equally problematic. In the traditional version Hermione marries

Orestes after Neoptolemus’ death. Euripides has reversed the order of the two

events for his own reasons, probably to reserve the climactic event of Neoptolemus’

death for the end of the play and, most importantly perhaps, to give a semblance of

organic connection to the two unrelated stories. In the Andromache Orestes initially

pretends that he came just to see Hermione but after he hears from her the last

events he drops his pretense and claims that he had learned of the rivalry between

the two women and waited to see whether Hermione would leave her home as a

result. But there is no way that Orestes could possibly know Hermione’s difficult position before he came to Phthia. Apparently Andromache’s troubles with

Hermione presented a convenient, although logically incoherent, excuse for

Hermione’s premature, by traditional standards, escape and Orestes’ visit to Phthia before the death of Neoptolemus. 115 Although the above two problems, the chronological consistency of the story and the plausibility of Orestes’ appearence, are the major structural stumbling blocks in the play, they are by no means the only ones. The play contains several long speeches that not only do not advance the story but they also contain flagrant anachronisms and seem at certain points to be thinly disguised editorials of the poet5. Moreover, Menelaus’ sudden departure is not anticipated at all and appears contrived. As mentioned above, for Aristotle Andromache would rank among the worst tragedies, for its apparent neglect of unity, the portrayal of the suffering of a totally innocent woman and child6, the lack of recognition and, not least perhaps, its happy end of sorts7. Older criticism similarly scorned the play as a collection of disjointed episodes sloppily put together in a flimsy plot. During the last two decades, though, this unfavorable judgment has given way to more positive views of the play8. Although Euripides’ concerns were obviously much different from

sSee e.g.Andr. 445-453 (Andromache’s attack on Spartan ethics); see also 595-604 (Peleus’ excoriating of the morality of Spartan women) and 724-726 (the general worthlessness of Spartans despite their military successes). Cf. the comments of Stevens (1971) 148.

#Cf.Poet. ch. 13, a notoriously thorny part of the treatise where Aristotle talks about the most suitable tragic hero. There it is clearly implied that the dramatization of totally undeserved suffering is not a good candidate for the best kind of mythos. See especiallyPoet. 1453a2-5 with the comments of Lucas (1968) 142.

Tn Poet. 1453a23-30 Aristotle stresses that the most successful tragedies are those ending badly for all characters, both good and bad, and claims that Euripides’ practice, generally unhappy, is laudable in that respect: although he has drawn fire from some quarters for his habit, Aristotle commends the fact that most of Euripides’ tragedies end tragically. Andromache obviously belongs to the less successful minority with happy end.

8See, for instance, Burnett (1971)130ff., Kovacs (1980) and Erbse (1984)130-136. 116 Aristotle’s, it has been stressed that the play has a unity of its own, thematic as it were, and exhibits striking formal symmetries: it begins and ends with a murder plot for the recovery of a spouse, the first unsuccessful, the second successful. In the three parts the play falls into the protagonists pass from happiness to unhappiness and vice versa. Andromache and her child are unexpectedly saved by the intervention of old Peleus. As a result of this, Hermione experiences a sudden reversal of her fortune and another one with the appearence of Orestes who removes her from all potential trouble to the safety of a new marriage. At the end Peleus laments the death of his only grandson, fearing the extermination of his line and his own death in grim solitude. The appearence of Thetis brings the final reversal.

Incidentally, it is apparent that Euripides’ play contains several reversals but in

Aristotelian terms this would be as grave a fault as lack of reversal.

Tragedy was never written according to Aristotle’s or any theorist’s maxims.

If Aristotle thought that the Sophoclean tragedy came closest to his own conception of the tragic art and genre, it is obvious that even the OT, arguably the most

"Aristotelian" tragedy ever written, is not completely devoid of non-Aristotelian elements, however minor. But, as I claimed above, the peculiarity and excellence of the OT is to be looked for not so much in its Aristotelian mythos, i.e. the strict adherence to the imitation of one action capable of inspiring great pity and fear according to probability and necessity. Rather, the OT has been admired because

Sophocles came up with and worked out a highly inspired and unique blend of episodes that enlivened the mythos. He could definitely have invented different 117 episodes and a very different sequence of events and the play could have remained as Aristotelian as the actual OT, although, of course, it would have been radically different from the actual OT\ on the other hand, he could conceivably have employed the same episodes in a non-Aristotelian tragedy.

At the other end of the rope, in a fair number of plays, like theAndromache, the Hecuba, the Suppliant Women, the Trojan Women and the Phoenician Women,

Euripides composed in a totally non-Aristotelian mode. Nevertheless, nobody considers him today a lesser dramatist than Sophocles. His contemporaries probably did but their reasons had nothing to do with the Aristotelian characterization of tragedy. At any rate, from Aristotle’s time onwards he became the favorite and most popular tragedian of the trio. It seems clear that, even if a tragedian were willing to follow Aristotle’s precepts exactly, he would end up with a promising subject matter, a plausible sequence of events and various tips on how to handle, say, the role of the chorus, or metaphors and strange words. This is no tragedy of course but at the most a good start towards the composition of a good tragedy. The educated public, and primarily a reading public, could be better served in the Poetics : they could rely on

Aristotle for they were likely to focus on themythos anyway. Aristotle’s primary concern was a theoretic one: to show that poetry was an art with its own well-defined subject matter, most probably in order to answer Plato’s attack on poetry. As he nowhere in his work, when he uses the example of housebuilding or any other art, specifies what kind of tools or materials are needed but deals only with the theoretical process of conceiving and executing the artefact, likewise in the Poetics 118 the actual construction of a good tragedy is left to, i.e. is dependent on, the resources and the genius of individuals artists. Nevertheless, Aristotle’s empirical and levelheaded approach results in sensible and consistent guidelines for the construction of a mythos. As I will argue in the next chapter, later theorists held very monolithic and much less well-defined views. CHAPTER V

ARISTOTLE’S POETICS AND HELLENISTIC LITERARY CRITICISM

Although not absolutely without precedent, Aristotle’s approach to poetry and criticism was, by modern lights at least, sufficiently novel. Especially his conscious effort to build a theoretical infrastructure upon which all subsequent literary criticism could securely rest was unique in ancient times. Thus one would expect that his approach would have awakened the interest of contemporaries and younger contemporaries and that his legacy would have been particularly strong soon after his death. Such an attitude would have befitted the intellectual climate of the day very well since the Hellenistic age showed a keen interest in literary theory: poetry and its criticism were abundantly debated and formed part of diverse theoretical investigations as is attested e.g. by Philodemus’ On Poems. Besides, there is evidence that other aspects of Aristotle’s philosophy, like physics, exerted some influence on later thinkers1 and there is no cogent reason why the Poetics would not have met

•See for instance the discussion of Furley (1967) and Hahm (1977). More recently Sandbach (1985) challenged the assumption of Aristotle’s influence on the Stoa, arguing that the evidence on which previous claims have been based is inconclusive. See, though, the discussion of the limitations of this argument by Hahm

119 120 with similar reception. The impact ofPoetics seems to have been much more limited

or at the very least diffuse: no commentary on it was ever written and no allusions

to or quotations from it appear in extant literature. Although it is probable that the

treatise was known, it apparently never gained wide popularity2.

A distinct sense of rift emerges when one turns to the relatively few

fragments of literary criticism that have survived from the Hellenistic era, of diverse

origin but mostly Stoic and Epicurean. The paucity of available information probably

robs critics of the chance to see things as clearly as they would otherwise have done.

Unfortunately, there has survived no complete treatise or even detailed account of

Hellenistic poetic theories. The chief surviving sources are the works of Philodemus, especially hisOn Poems of which only a substantial part of the end of book five has escaped severe damage. Not only does Philodemus’ work survive in fragmentary state but also the author has a clear polemical orientation which results in a very synoptical presentation of the opinions refuted, often with little or no mention of individual names or even indication of philosophical schools. Although there is no way of knowing whether Philodemus or his sources intentionally distorted or misquoted, or perhaps in certain cases plainly misunderstood, the opinions refuted, this is hardly inconceivable. At any rate, brief reference seldom does full justice to the text addressed. Nevertheless, even granting the considerable lack of reliable and,

(1991) (I owe the last two references to Prof. Hahm himself).

2See Lucas (1968) xxii-xxiii. Only in the Renaissance did the Poetics start enjoying its present prestige. 121 above all, adequate information, it is impossible not to notice the aforementioned rift between Aristotle and the Hellenistic literary critics consistently suggested by the meager evidence. Hardly does any post-Aristotelian source seem to approach poetics the way Aristotle does. Concerning the definition, for instance, one searches vainly for anything even remotely analogous to the Aristotelian definition of tragedy. Only brief, descriptive and intuitive-sounding accounts of what a poem is appear occasionally. Needless to say, the importance Aristotle ascribes tomythos and the theoretical infrastructure he postulates as the basis of the successful poem are nowhere to be found in Hellenistic literary criticism. On the contrary, moral/didactic considerations hold an especially prominent place.

Such considerations were a staple of Greek literary criticism as far back as one can go, to Xenophanes and Theagenes of Rhegium3. In Aristophanes’Frogs

Aeschylus and Euripides discuss the social role of poets and both agree that poets are admired because of their "skill and advice, for making people in the cities better"

(1009-1010)4. The first part of the play’s agon is devoted to the benefits the

Athenian citizens purportedly reaped from the plays of the two contenders:

Aeschylus’ plays taught them military valor and discipline, Euripides’ characters offered models of smartness and argumentative power. Plato often (and ironically) mentions the use of poetry for didactic and paraenetic purposes. He ridicules of

3On the beginnings of Greek literary criticism see the surveys in Pfeiffer (1968) 8ff., Russell (1981) 18ff. and Richardson (1992).

“On the issue of moral didacticism in Greek drama as presented in the Frogs see most recently the discussion of Dover (1993) 15-16. 122 course the idea that poets can actually teach people and especially instruct professionals in their arts and crafts but he brings significant testimony to the fact that people did view poetry in terms of instruction or, at the very least, used such claims in their rhetoric. Plato himself takes much more seriously the possibility of using poetry to instill positive values to the young. Above all, he considers poetry a clear threat which contributes to the moral unraveling of the body politic5. Thus not everything was new in Hellenistic literary criticism as hardly everything was new in

Aristotle’s Poetics.

Nevertheless, it would be a mistake to consider Hellenistic literary criticism as basically a return to pre-Aristotelian popular poetics since several elements were particular to the schools which proposed them. What I will try to claim is that, based on the fragmentary and often obscure extant evidence, there does not appear to be a clearly distinguishable thread that runs from Aristotle’s Poetics to the theory of poetry of Hellenistic theorists. The import of this claim can be better assessed if one considers the relationship of Aristotle’sPoetics to Plato’s work: although no mention of Plato is made and no quotation from his dialogues is found in the Poetics, similarities of content, methodology and even metaphors (e.g. the poem as a living being) have been established and catalogued6. It is very hard to find such similarities between the Poetics and Hellenistic literary criticism. This lack of correspondence

5For Plato’s views on poetry and education see Vicaire (1960) and Scolnikov (1988) 112-119.

6See above ch.2 n.5. 123 of course does not imply that thePoetics was not known to Hellenistic theorists or that they did not consider it an important work -there is some evidence that they must have known the treatise7, although it is naturally impossible to know how they ranked it. I will argue in what follows that, based on our limited evidence, one can only claim that there does not seem to be a "direct" influence, in the sense of similar concerns or answers to specific problems, of Aristotle’sPoetics on Hellenistic literary theory. It is perfectly possible of course, especially given the fact that Aristotle’s work was most probably known, that Hellenistic theorists read the Poetics , decided that Aristotle’s approach was not satisfactory and went on to propose something totally different -this would definitely be a sort of influence. Unfortunately there is no evidence which would substantiate this tantalizing possibility: one fails to find substantial influence of Aristotle on later theorists and there has survived no information that they reacted to the Poetics by rejecting, approving or modifying

Aristotle’s theses. Thus we can only speculate about such possible reaction since we cannot disprove it. In what follows all claims about the absence of Aristotelian influence on later literary criticism should be read with this caveat in mind.

Despite the absence of substantial direct connections between the Poetics and

Hellenistic literary criticism, one finds many an assertion to the contrary in the literature, often back to back with remarks on the isolation of the Poetics 8. One cannot help thinking that Aristotle and Plato loom so large in modern perception of

7See below n. 46.

8See for instance Koster (1970) 85 n.l and Halliwell (1986) 288 n.4. 124 ancient literary criticism that it proves difficult for anyone to reconcile oneself with the idea that they might not have been so influential in their time after all. The scanty and widely scattered ancient evidence as well as the paucity of modern studies, especially comparative ones, devoted to the subject have certainly contributed to the perpetuation or rather the taking for granted of a largely unsubstantiated assumption.

This is not to imply of course that nobody has noticed the uniqueness and isolation of the Poetics9, but to call attention to the fact that even in specialized studies

Aristotelian influence is often taken for granted. I will provide two recent examples of such discussions before concentrating on the relationship of the Poetics to the literary theories of the Hellenistic philosophical schools.

In his article on literary criticism in the Iliadic scholia Richardson10 makes repeated references to Aristotelian material in the scholia and even to the "fact" that practically all literary theoretic principles found there can be traced back to Aristotle, as if such influence were self-evident11. Again this claim can hardly be totally wrong

9See During (1966) 182 and Russell (1981) 31.

10Richardson (1980).

"Similar assumptions underlie the much earlier work of Adam (1889). See the discussion of Schmidt (1976) 39-74, esp. 39-54, who argues against the emphasis on both Aristotelian and Stoic influence on the scholia (Stoic influence has been advocated by Lehnert (1896) and von Franz (1940) and (1943). Other candidates for the literary-theoretic "fatherhood" of the scholia were suggested by Schrader (1902) who proposed the second century grammarian Telephus and Griesinger (1907) who favored Aristarchus. Schmidt understandably attacks all these theories as too one­ sided: the scholiasts’ comments cannot plausibly be traced back to one thinker or school, especially since many views seem to have been commonplace in the Hellenistic period and thus their appearance in the scholia would not imply any direct indebtedness. On the literary-theoretic views in the scholia see also the discussion of Wilson (1983). 125 since some things manifestly seem to go back to Aristotle. Nevertheless, there are no grounds for claiming that the majority of the scholiasts’ literary-theoretic principles are indebted to him. Richardson divides his article into sections with tellingly Aristotelian captions {mythos, ethos, lexis etc.) but the similarity with

Aristotle would not extend much beyond the terminology. Richardson asserts, for instance, the Aristotelian origin of poikilia (a principle often evoked in the scholia), i.e. the inclusion of variegated episodes or the alternation of charged and relaxed passages in order to avoid monotony. Now the term itself occurs only once in the

Poetics (1459a34) in a negative context as something undesirable or at any rate problematic12. Of course the term is not of cardinal importance if there are indications that the concept is present. Although Aristotle mentions admiringly

Homer’s use of episodes, his praise focuses not on avoidance of monotony or anything of this sort but on the proper construction of a successful poeticmythos which nowhere includes variety for its own sake. Contrary to Richardson, even the scholiasts’ praise of Homer’s oikonomia and the way he handles the different parts of his epic do not seem to reflect Aristotelian influence. The scholiasts’ views seem much more akin to Stoic or allegoric attempts to defend Homer from charges of inconsistency and doubts about his poetic merit. Nowhere in the Poetics does

Aristotle imply that it would not be immediately clear how the episodes relate to one another; on the contrary this relation should ideally seem not only plausible but also

12Cf. Heath (1989) 50 n.30. 126 cogent. Heath13 has drawn a very fine distinction between the Aristotelian unified

Trpagic and the concept of U7ro0eaic found in the scholia:

Aristotle’s praxis is a causally structured underlying sequence of events, into the narrative of which other events may be introduced digressively; the scholiasts’ hypothesis is a "field" of events (in the present case, the matter of Troy), from which a single sequence (the wrath of Achilles) may be extracted to provide the core of a text in which the whole field is narrated by inversion (££ dvacrrpo^nc). Thus where Aristotle contrasts poetry and history in terms of the structure of "the underlying events"...the scholia contrast the way in which a -possibly common- hypothesis is realized in the text: narrative "by inversion" is poetic, narrative "in order" (Kara rd$iv) is historiographical (B 2.494-877, cited above; cf. Macrobius Sat. 5.2.8-11, 5.14.11); naturally, it is is likely that the events held over in an inverted narrative text will be treated in less detail and more selectively than would be appropriate in a historiographical narrative. It is worth noting, however, that Aristotle’s preference for Homer’s choice of a plot with a single part{Poet. 59a35) is similar in practice to the scholiastic preference for inverted narrative; the core-sequence of the scholiasts’ hypothesis corresponds to the Aristotelianpraxis , while the other parts which, according to Aristotle, Homer uses as episodes correspond to the parts of the hypothesis taken out of order in an inverted narrative.

As far as the suggestion that the scholia shared Aristotle’s view of Homer as the first tragedian, apparent according to Richardson from the references to the tragic and pathetic qualities of certain passages and terms likeperipeteia u, again it is not necessary to postulate specifically Aristotelian influence. Homer was considered the best poet and not least the inventor of any number of arts and crafts

13Heath (1989) 116-117.

14Richardson (1980) 270-271. 127 well before Aristotle and after him15. If Aeschylus called his tragedies slices of

"Homer’s great banquets" (Ath. 8.39.17-18), he might have referred not simply to their subject-matter but to the tragic qualities of Homer’s poetry as well. At any rate, the references in the scholia to pathos, eleos and tragic suspense do not imply that they built on the notion of Homer as the first tragedian. Tragedy had early on and especially from the fourth century onwards acquired a very exalted status and pathetic performances had gradually usurped, already much to Aristotle’s distress, the place of a unified mythos. It would be perfectly understandable for a later Greek familiar with elaborate and pathetic tragic performances to describe something he found effective, striking or emotionally powerful in tragic terms, since tragedy was considered the pinnacle of poetic success and effectiveness. It is indicative that

Richardson cites the scholiast who justifies the length of Achilles’ pursuit of Hector by appealing to the poet’s intention to arouse emotions, "exactly as in the theaters now". It should be remembered that Aristotle explicitly rejects the possibility of a tragic version of the pursuit because it would look ridiculous on stage; he considers this Iliadic episode very suitable as an epic theme because, although it cannot withstand rational examination, it causes wonder (Poet. 1460all-17).

In the same vein Richardson casually cites Aristotle (Poet. 1455a22ff.) to

15See e.g. Ar. Frogs 1033-1036 and the comments of Dover (1993) ad loc. In later rhetorical and Neoplatonist tradition and in the scholia themselves Homer’s inventiveness and prestige hardly knows any limits; for references see van der Valk (1963) 468-469 and cf. Schmidt (1976) 58-59 and Browning (1992) 135-136. The Stoics contributed crucially to this attitude. For the Stoic view of Homer’s knowledge and achievements see also below. 128 support his claim that probability (7ri0av6TTK) and vividness (£vdpyeia) depend on visualization16. Aristotle of course praises Homer’s ability to present even falsities and improbabilities as plausible and thus to win the audience’s irrational assent but he nowhere says or implies that probability depends on the power of visualization. As will be shown below, vividness plays a very limited role in the

Poetics and it is clear that, whether Aristotle considered it one of Homer’s assets or not, he does not mention it in the Poetics. Emphasis on vividness belongs to later rhetorical theory. At the beginning of his treatment of Homeric style in the scholia

Richardson himself indicates their indebtedness to the divisions of the stylistic characters in later rhetorical theory17. One can at best point to only a tangential relationship between the scholia and the Poetics.

I have dwelt fairly long on Richardson’s remarks not because they carry much weight within his otherwise lucid argument compromising its validity but merely to point out how even a careful and perceptive scholar can postulate Aristotelian influence where there is none, or very diffuse, to be found. Similarly in his fairly recent study Meijering18 detects consistent influence of Aristotelian ideas on the literary and rhetorical theories in the scholia. More specifically, Meijering sets out to prove that the cardinal concepts of ^uxayMyta,(pavraoia, vnbdeoiQ and oitcovo/zia go back to Aristotle. This is not a place to engage in a detailed

16Richardson (1980) 278.

"Richardson (1980) 275.

18Meijering (1987). 129 refutation of Meijering’s argument but so much at least should be said. Meijering remarks at the conclusion of his book that "Aristotle or other Peripatetics are not cited as authorities in questions of literary theory...Moreover, such characteristic

Aristotelian terms as fivQoc, otioraoiQ, ninwoiQ, n pd£i.c and, for tragedy, d/xapria, tcdBapaic, 0o/3oc and Ikeoq are rare or completely lacking in the scholia"19; nevertheless, he concludes that "it was possible for later commentators, who certainly never read the Poetics, to remain relatively close to the principles underlying it. In the commentaries of their predecessors, which I believe did arise via the Alexandrians from Peripatetic scholarship, they could recognize concepts familiar to them from their rhetorical schooling"20. I find this reasoning difficult to follow and at any rate less than cogent. Meijering assumes that "Peripatetic scholarship" stayed close to Aristotle and promulgated his ideas and terminology thus influencing all subsequent scholarship. But "Peripatetic scholarship" is a ghost, a convenient designation to cover all post-Aristotelian scholarship produced by authors of often little-known philosophical affiliation. Often one cannot even attribute a name to views or writings said or seeming to be of Peripatetic origin21. Thus

19Meijering (1987) 223-224.

“Meijering (1987) 225.

21Cf. the comments of Brink (1946) 11-12: "In this context a fact of nomenclature may be mentioned. The name IIepi7rar»7riK6c which by the middle of the third century B.C. denoted a member of the Peripatetic Scool in Athens, changed its significance about that time. With the wider influence of Peripatetic studies it is not only used for the Athenian school but can also denote any writer of biography or literary history connected with Alexandria...two conditions constitute this new usage of an older name, viz. connexion with Alexandria...and the refined form which Alexandria had bestowed on the literary and biographocal studies of the Peripatos". 130 Meijering’s unqualified references to Peripatetic scholarship are very optimistic, if not unfounded.

One example will suffice to indicate the convoluted and deeply problematic character of Meijering’s reasoning. In order to support the claim of Peripatetic influence on the scholia he finds it useful to turn to the historian Polybius, theorizing that this author uses terms employed in the scholia. If Polybius’ terms prove to be of Peripatetic origin, the same must hold for the source of the scholia too.

Discussing Polybius 2.56“ and 1.323, where the author expresses his conception of the historian’s task and attacks Phylarchus, a historian of the tragic mode, Meijering assumes that Polybius adheres to the theory of a hypothetical author of Peripatetic bent, although perhaps not a Peripatetic, because "he remains so close to Aristotle that really his theory is as Peripatetic as can be"24. According to Meijering this putative author had proposed against Aristotle that historical writings can have the coherence of an Aristotelian tragic mythos25.

See also his further specification (12n.l): "There seems to have been a third stage in which the name lost its connexion even with Alexandria, and did not mean more than "grammarian" or "literary critic". Along the same lines Pfeiffer (1968) 150-151 noted: "That Hermippus could be called EkpiTraTrjrucoc as well as KaAAijudxeioc suggests that the term had no longer any philosophical flavour but could be used of any writer on literature and antiquities, in particular the biographer".

“Meijering (1987) 10-11.

“Meijering (1987) 221-223.

“Meijering (1987) 223.

25This is really the theory of Schwartz (19432) 123-125 who first suggested that there was a Peripatetic reworking or blurring of Aristotle’s distinction in Poet. 9. and that this thesis inspired Hellenistic historians like Ephorus and Callisthenes; see also 131 Now there is evidence that in the Hellenistic period historiography displayed a pronounced and almost ostentatious literariness. The historian Duris of Samos, who influenced Phylarchus, accused Ephorus and Theopompus of having represented facts in an extremely poor manner because they did not pay any attention to mimesis or pleasure ('E^opocSt k c k ! eeoironwoQ ruv yevo^evwv &neXel

Duris sacrifice truth and objectivity to mimetic pleasures? Unfortunately such tantalizing queries have no way of receiving an adequate answer. From his few

Schwartz (1897) 560ff., (1900) 107ff. and (1909) 491. Schwartz was followed by Scheller (1911), Zegers (1959) 5-6, and Strasburger (1966) 41-44. Meijering does not cite Schwartz, either not realizing or not wanting to spell out the connection. He only cites Walbank (1960) and Brink (1960) who refute the second part of Schwartz’ theory, the influence of the putative Peripatetic on Hellenistic historians. But Walbank primarily refutes the notion that Polybius attacks a Hellenistic/Peripatetic theory of history, which is Meijering’s very own. 132 surviving fragments it is clear that he did not shy away from mythical explanations and dramatic scenes.

The one piece of information that has intrigued several scholars is the fact that Duris was Theophrastus’ pupil26. Based on that one can put forth a number of plausible inferences: Duris might have taken over, or perhaps expanded on, one of his teachers’ theories; alternatively, the historian might have developed a theory or expressed an opinion unrelated, if not hostile, to Theophrastus’ views; or Duris was a representative of a Peripatetic school of historiography which espoused the views of an influential early Peripatetic who, modifying or attacking Aristotle, suggested that historical writings can have the coherence and embellishments of an Aristotelian tragic mythos, i.e. advocated the writing of "tragic history". The first and last alternatives are fused in a way: Schwartz suggested that the origins of the theory of tragic history were to be found in the work of either Theophrastus or Praxiphanes of Mytilene who both wrote on history27. On this hypothesis Polybius would then attack the representative of this school, Phylarchus, who was influenced by Duris.

Fornara28, following von Fritz29, energetically defended Duris’ originality

“Even if true, this information is problematic in itself: the term "student" is notoriously used very loosely in ancient biography and often denoted little more than the biographer’s assumption on the relationship of two near contemporaries; cf. Lefkowitz (1981). In no way does it imply that Theophrastus shared all or some of Duris’ theories.

27The only scholar who suspected other than Peripatetic influence on Duris was Ullman (1942) who suggested that tragic history went back to the school of Isocrates. This view remained completely isolated.

“Fornara (1983) 124-134, esp. 130-134. 133 charging that the attribution of the theory to Theophrastus or Praxiphanes resulted from the critics’ wish to find a better known or more respectable father for the theory of tragic history than the relatively obscure Duris. On Fomara’s side is the fact that Duris was not only a historian but also the author of treatises on Homeric problems (FGrH 76 F30), on tragedy (FGrH 76 F28) and the tragedians Euripides and Sophocles (FGrH 76 F29)30. In this light it is not implausible that Duris would express informed opinions on historiography or that he would even propose a theory of his own that would in turn influence later historians. But unfortunately there is no evidence to suggest that he had formulated any such theory which would make him the patriarch of a Peripatetic school of history. Fornara’s zeal to proclaim Duris an all important figure in the development of theoretically backed tragic history carries him away: he summarily dismisses Theophrastus and Praxiphanes as the authors of the purported treatise claiming that Theophrastus had no theoretical interest in historiography and Praxiphanes did not include the historian but another

Thucydides in his dialogue On History. The first claim is plainly erroneous because

Theophrastus wrote a treatise On History31; now Fornara, who does not mention this work, either ignores it or paradoxically does not consider it strong evidence for

^von Fritz (1958) esp 112ff.

“He also wrote on painting and sculpture(FGrH 76 F31 and F32), laws and athletic contests (FGrH 76 F27 and F33). His brother Lyncaeus, also a student of Theophrastus’, was a comic playwright and wrote on Menander; see Ferrero (1963) 71-72. On Duris and his oeuvre see also Kebric (1977).

31D.L. 5.47 and Cic. Orat. 39; see Walbank (1972) 36-37. 134 Theophrastus’ theoretical interest in historiography. As for Praxiphanes, even if he did not make the historian Thucydides a character in his dialogue, it does not follow that he could not have formulated a historiographical theory in that work.

Fornara, and much earlier Brink, disputed Walbank’s attack on the theory of tragic history as expounded by Schwartz and taken up by von Fritz who put Duris in the limelight as the originator of tragic history. Walbank 32 suggested that Hellenistic historians like Phylarchus simply exhibited to a pronounced degree traits that are consistently found in Greek historiography from Herodotus onwards. Thucydides for instance levelled against his own predecessors charges very similar to those Polybius brings against Phylarchus and the historians of the tragic, i.e. sensational and melodramatic, mode33. This is not surprising since Polybius sought to uphold the

Thucydidean model of contemporary, military/political history which shunned

32Walbank (1960) and (1972) 34-37.

^Polybius repeatedly attacks historians, both eponymous and anonymous, who confused the methods, subject matter and purpose of tragedy and history. For a collection of passages see Walbank (1957) 259-260 and Sacks (1981) 144 n.51. In 7.7.2.1 Polybius says that some historians exaggerate events elaborating on them like tragic poets (rpaytpfiouvrec) and adding marvellous embellishments (7 roA.Ar)v...reparetav). Almost identical terms are employed in 2.17.6 where it is said that the tragedians (TpaycpSioypd^ot) make much out of the different Celtic tribes and adduce many fantastic elements (wokkfiv SiarlOeivrai repareiav). It has naturally been suspected that Polybius here implies not tragic poets but historians of the extravagant mode (see Walbank (1957) ad loc. and P 6 dech (1970) 57). If this is true, it is obvious that Polybius’ ironical collective name for the historians he attacks need not imply any theoretical background they all shared but merely their common penchant for florid style. In 7.7.1 he also calls them Aoyoyp

Duris’ theory of an Aristotelian or tragic historiography. In principle of course

Fornara is right in asserting that previous relevant practice does not militate against the formulation of a theory, in this case theoretical grounding of emotive writing.

But it is easy to beg the question here. If Duris’ theory demonstrably existed, it would not be difficult to defend its originality. But any indication that Duris had a full-blown theory that created a school of historical writing is simply not there.

Fornara claims that

^Consider the following statements: "Tragic history emerges from this examination trimmed and shorn. But I think it does emerge, and should not, as Walbank suggests, be discarded from subsequent discussions. It appears to have been a post-Aristotelian form of historical writing in which some of the principles of tragedy as Aristotle saw them were deliberately applied to history, although there must have been many earlier examples of "emotive" historical narrative. There may have been only a few representatives of that kind of writing, but they were important enough for Polybius to criticize when...he came to demand new standards and canons in Greek historiography" (Brink (1960) 19). "The desire to keep "theory" in proper perspective...must be warmly supported. The very word will galvanize some scholars. But the pendulum has swung too far, and we begin to deny the existence of theory almost on principle. Here we have a theoretical work (the Poetics), an adaptation of its system (Duris), the use of consonant techniques by a writer (Phylarchus) who is accused of not knowing the difference between tragedy and history, and the adoption of poetical terminology to prose works by literary critics thereafter" (Fornara (1983) 132 n.54. 136 Polybius’ comparison of history and tragedy in this context (2.56.11) would be ludicrous if it were not that Phylarchus’s aesthetic, generally recognized as such, was dependent enough on poetic theory to sustain Polybius’s charges and give them point. 35

But this is not necessarily so: Polybius’ reference to tragedy and history in 2.56.10 might have been thought to require some explanation and at any rate the comparison between the two genres is found in the Poetics. In what context Duris’ polemic against Ephorus and Theopompus was formulated is unknown. The fact that mimesis and pleasure appear in it cannot make the fragment a condensed theory. On the other hand, Polybius might have come up with a fancy collective name for the historians he attacks -this does not imply that they shared anything more than a penchant for florid style. Thus on the face of such limited and especially inconclusive evidence it would be best, I think, if, first, one followed Walbank in discarding the notion of tragic history and the ungainful speculation on its originator36 and, second, if one turned elsewhere to explain Duris’ utterance which gave rise to the whole notion of tragic history.

A very convincing approach was taken by Gray37. She collected and discussed all evidence, mainly from later sources like Dionysius of Halicarnassus and

[Longinus], where the term mimesis has nothing to do with its Aristotelian

“Fornara (1983) 128.

“Africa (1961) 38-41, Kebric (1977) 16-18,39-41 and Sacks (1981) 144-170 agree with Walbank and provide good summaries of the status quaestionis. For another detailed review see Meister (1975) 109-126.

“Gray (1987). 137 counterpart but is a historiographical terminus technicus :

The meaning of mimesis in history is the recreation of reality, encompassing recreation of both character and emotion...The means by which the recreation of life is to be achieved is the observance of the rule of propriety, based on observation of what men do in real life. It will emerge that ancient literary critics are unanimous in this definition of historical mimesis, even though they may differ in their subjective estimates of the level of mimesis the various historians achieved38.

Gray traces this meaning of the word to Peripatetic sources on rhetoric, possibly to

Theophrastus39. The first writer who used the word in this sense in extant literature is Duris who, on this interpretation, would accuse Ephorus and Theopompus of lack of propriety (mimesis) and claim that they wrote exclusively in the grand style

(ypd0 eiv). According to Gray ypd0 eiv is another technical term which goes back to Aristotle’s ypa^iKt) ke^iQ, a style suitable for prose, especially the elaborate epideictic genre (Rhet. 1413b4ff.). This is more suitable for reading, in contrast to the dywviariKr), which is appropriate for live delivery and acting. That

Ephorus and Theopompus wrote only in the grand and even artificial or frigid

"written" style of epideixis is well attested in ancient sources40. Thus Duris’ attack has in all probability nothing to do with a historiographical theory of tragic mimesis

“Gray (1987) 469-470.

“Gray (1987) 483-486.

40Demetrius On Style 27, Porphyrius (.FGrH 115 F.21), FGrH T.24-26; cf. D.H. 134 who attacks the epideictic style of Isocrates’ school to which Ephorus and Theopompus belonged. 138 but concerns the stylistic choices of these historians, i.e. their lack of variety and verisimilitude which resulted in works unable to please their audience since they lacked imitative charm.

I find Gray’s argument extremely compelling because it detects a very clear thread that runs through centuries of Greek rhetorical and historiographical criticism and illuminates Duris’ vexed and almost abused passage in a historically natural and intelligible way. Most important for my purposes here is that, on her view, any connection between Aristotle’s Poetics and later historiographical practice is decisively undercut.

To return to Meijering’s theory, his views on the putative Peripatetic source of Polybius’ criticism of Phylarchus are subject to similar strictures as the theory of tragic history. First of all, Meijering does not specify whether he equates the

Peripatetic treatise he postulates with the one that purportedly gave rise to tragic history. If he does, as I believe since nothing in his text contradicts this assumption, then it would be very curious indeed if the same treatise could serve as the basis for two totally opposing views, with Polybius sharing his opponents’ theoretical background. To reconcile all the various claims that have been made about the purported reworking ofPoetics ch. 9, one necessarily has to hypothesize that tragic historians and Polybius embraced different aspects of what must have been such an extraordinarily pregnant and variegated treatise that would border on self- contradiction: Phylarchus and others like him must have espoused the proposal of a dramatic history that would be written in a florid sensational style and aim at 139 pleasure by the arousal of intense emotions like pity fear or surprise41. Polybius, on

the other hand, must have endorsed the theory that historical works should display

the coherence of an organic body, bringing out and illuminating the hidden structure

of events, the causal connections that might be lost on the layman: the historian’s

task is to show how and why things worked the way they did, not simply to recount

a selection of historical events, with an eye to rational persuasion and education (1.3-

4)42. Each side would then totally ignore the other.

Now in theory such a scheme is not impossible but I think it is definitely

fantastic and improbable. First, there is no reason or need whatsoever to postulate

this unknown writer and speculate about his putative theories. Polybius might very well have read the Poetics himself. At any rate, it is very doubtful whether his theory

can be called Aristotelian, or more Aristotelian than the one it attacks. The

41This is what Fornara thinks that Duris proclaimed to be the peculiar pleasure of history, the very essence of his philosophy of history. Unfortunately there is not a scrap of evidence to support this view.

42Heath (1989) 80-81 has claimed that Polybius is not expounding a general historiographical program in this passage but only addresses the unique parameters of the period he handles by attacking the authors of monographs which provide a very partial and misleading idea of the forces at work in that period and their connections; while Polybius conceptualizes his work as a complex, living organism, he sees the monographs as the scattered limbs of a corpse which once was a beautiful whole: nobody can acquire a notion of the whole by examining its parts. Cf. the discussion of Petzold (1969) 34-37. Heath suggests that, since Polybius insists on the uniqueness of the period he treats, he cannot possibly have required earlier historians to write as he does. I think that this is oversubtle. Although earlier historians lacked Polybius’ subject-matter, this would hardly prevent him from rating his work superior to theirs. At any rate, all subsequent historians would have to follow Polybius’ guidelines. Polybius’ utterances definitely concern the present and future historiography; even if older historians are exempted, the scope of Polybius’ program is not as limited as Heath thinks. 140 terminological framework is unmistakably Aristotelian but the conceptual is not.

Polybius rejects his opponents’ view that the function of tragedy and history are identical but he suggests that their oCaraoic should be very similar43.

It is true that Aristotle never expresses his views on the proper way of writing history: what he says inPoet. 9 is of a descriptive nature. Still, although he clearly thinks that some historical events might exhibit the kind of unity he postulates for a tragic mythos, he seems to consider this a rarity and, consequently, to perceive a fundamental difference between tragedy and history. Thus, even if Meijering’s

Polybian source existed, one should not consider it Aristotelian or Peripatetic and the same holds for Polybius himself of course. On such premisses, one could call

Callimachus a Telchin or vice versa and Philodemus a Stoic. At any rate, given the

43Another historian who wrote about history’s nature in the same vein as Polybius is Diodorus Siculus (20.1.5): "For the genus of history is simple and self-consistent and as a whole is like a living organism. If it is mangled, it is stripped of its living charm; but if it retains its necessary unity, it is dully preserved and, by the harmony of the whole composition, renders the reading pleasant and clear". It has been thought that Diodorus is here indebted to some other historian, a view which strikes me as a clear case of bias against Diodorus. Fornara (1983) 148-150 attacks Laquer (1911), who proposed Ephorus as Diodorus’ source, and argues for Duris, adducing as collateral evidence Diod. 20.43.7 which supposedly echoes Duris(FGrH 76 FI; the identification goes back to Schwartz RE 5.1.687; see also Strasburger (1966) 47 n.4 and Kebric (1977) 40 and 77). Even if this Quellenforschung can be thought to have some truth in it (in my opinion a very bigif because I do not see how Diodorus’ "phraseology is strikingly reminiscent of Duris, FI" (Fornara (1983) 150 n.9)), it would only prove that Diodorus had read and echoed Duris, not that the latter had a theory of tragic or Aristotelian historiography. The analogy between a living body and a literary piece goes back to Plato’s Phaedrus (264c) but nobody as far as I know has claimed that Plato proposed a theory of tragic rhetoric or that Aristotle talked about rhetorical tragedy. Diodorus does not mentioned or imply tragic history (as Sacks (1981) 146 n.55 notes, "elsewhere (19.8.4) Diodorus denounces tragic historians") and he could have come up with his historiographical theory on his own or he could have been possibly lead to it by none other than Polybius himself. 141 thoroughly unhelpful nature of the existing evidence in the scholia -anonymous, short, mostly with no source citation and very rarely addressing theoretical questions of poetics- it would not be unreasonable if one confined oneself to the negligible number of passages where the possibility of Aristotelian influence is demonstrably strong. Otherwise any pronouncement turns out to be speculative and potentially misleading. The Homeric scholia constitute the last product of a very long series of works on the so-called "Homeric problems" and are likely to have been cast in the mold of such studies. If the Poetics contributed to their theoretical background, this contribution has been so much diluted and filtered through other layers that it is now beyond secure recovery.

A similar attitude has prevailed in treatments of the issue of unity in

Hellenistic poetics, notably in the Callimachus’A etia prologue (fr. 1 Pf) and in the explicit or implicit reactions to his program. In his treatise Ilpdc IIpa$i0 dvj 7 v the poet-scholar Callimachus supposedly attacked the views of the Peripatetic

Praxiphanes who had in the Aristotelian tradition praised long and "unified" epic compositions44. Other poets, most notably Apollonius of Rhodes, responded to

Callimachus’ program by upholding the possibility of, and by actually composing, epics45. Now it is nothing strange for Callimachus to criticize Praxiphanes but it is very hard to believe that a whole poetic program would revolve around the polarity

"See the influential article by Brink (1946); cf. Pfeiffer (1968) 135-136.

4SThe issue is particularly difficult to judge because it inevitably involves the vexed question of the quarrel between the two poets. On this see Bundy (1972) 45ff., Mueller (1987) 3Off. and Rengakos (1992) 42ff. 142 long and unified/short and disjointed poems. First of all Praxiphanes’ theses are unknown and they did not necessarily have any direct connection with thePoetics.

Besides, even if Callimachus and Apollonius knew Aristotle -and there is no plausible reason why they would not46 -they do not seem to have composed their poems or formulated their attacks in the framework of literary-theoretical or philosophical debates47. Apollonius for one wrote an epic that could very well be classified by

Aristotle in the category of inferior cyclic epics which narrate everything from beginning to end48. Moreover, it is not even certain that Callimachus rejected the

““Diogenes Laertius mentions the Poetics in his list of Aristotelian writings (5.22- 27) and his ultimate source is believed to go back to the Ptolemaic library; see During (1966) 36-37, RE Suppl. 2.190-194 and Blum (1977) 121-132.

47 Cf. Heath (1989) 57.

““Hunter (1993) 193-195 notes the incompatibility of theArgonautica with the Aristotelian model epics like the Iliad. He claims, though, that Apollonius’ poem is totally unlike the cyclic epics in its compositional technique and explicit rejection of continuous narration (Arg. 1.649, 2.391, 3.401). Thus Hunter concludes that both Apollonius and Callimachus wrote poems which could be considered both (in Hunter’s view this stands for Aristotle’s laudable unity) and Sirjveicl c (which Hunter interprets as "continuous, monotonous and unvaried"). I find this interpretation merely ingenious if not dubious since Hunter’s suggestion for the meaning of SirivexiQ and the contrast between the two terms is totally speculative, as he himself admits, with no support in the sources. To claim that "like theAetia prologue, theArgonautica too breaks down the Aristotelian dichotomy in a spirit of literary experimentation" (195) is to read too much into the texts. Most important for my purposes is that, irrespective of Apollonius’ own concerns and goals, in terms of the Poetics the Argonautica could only fall in the category of inferior, cyclic epics, whether modern readers would agree with this classification or not. Heath (1989) 6 6 , for instance, suggests that it "seems clear that Apollonius has been striving to produce the kind of unified structure which Aristotle describes as a single action of many parts[Poet. 1459b 1-2]". Although the action in question refers to the plot of the Cypria and the Little Iliad, Heath seems to equate it with the Aristotelian unified action; cf. his remark "in fact, the poem does satisfy the demands of the ’Telchines": it is a continuous narrative about heroes in many thousands of lines...Continuity is 143 heroic epic or the Homeric epic wholesale. If Callimachus had Aristotle in mind when he wrote the Aetia prologue his attack does not hit any particular Aristotelian mark. But it is, I think, best to discard such a cumbersome and unhelpful

assumption. Callimachus answered his opponents not as an anti-Peripatetic49 but as

a poet: to the accusation that he wrote short poems he retorted that long poems are not necessarily good and are more likely to be bad. When the opponents reproached his failure to choose heroic themes he threw back that unconventional lighter themes were much more fruitful poetically. Apollonius did not write his poems in order to put Aristotelian principles into practice or as an anti-Callimachean pamphlet in

a little more controversial, but the poem’s plot does constitute a single Aristotelian praxis" (65). The fact that Apollonius links some of the events he narrates in causal sequences, as Heath notes (6 6 ), does not make the epic less cyclic. Nowhere does Aristotle claim or imply that in cyclic epics causality is nowhere to be found but that it is not the guiding principle in the overall arrangement of events. If Apollonius had pursued Aristotelian unity, he would most probably never have written the epic that he did.

49Pfeiffer (1968) 137 has claimed that "the new poetical school of Callimachus and his followers was ostentatiously anti-Aristotelian. Rejecting unity, completeness, and magnitude, it consciously aimed at a discontinuous form". Leaving aside the fact that there is no evidence for any Callimachean school before possibly Roman times, there is no reason to assume that, because Callimachus’ program (as much as it can be reconstructed and thought to have theoretical coherence) seems not to comform with Aristotle’s views in the Poetics, the poet attacked the philosopher. This approach is typical of the tendency to see and pigeon-hole everything in terms that are considered important by modern scholars although they are largely shaped by historical accident. Even more extreme is Roster’s view (1970) 120-123 who suggested against Pfeiffer that Callimachus followed Aristotle’s precepts. Roster thinks that the primary issue in Callimachus’ A etia prologue is the opposition between short and long poems and he claims that Callimachus’ preference for the former is in accordance with Aristotle’s pronouncements in thePoetics. I believe that the question of unity cannot so easily be pushed aside and, in general, if one looks for so broad similarities or differences as Roster does, one is bound to be well-served in the texts of almost any pair of authors. 144 verse: if the Argonautica is anti-Callimachean, a very unlikely assumption, it is cast in a thoroughly Callimachean mold50. Either way it has nothing to do with either

Aristotle or the Poetics.

Turning now to the two major Hellenistic philosophical schools I will start with the Epicureans for a number of reasons, first because Philodemus himself was an Epicurean and it would be helpful to delineate his position before examining the arguments he directs against his opponents. Besides, the Epicurean views on poetry seem to have been less diversified than the Stoic which addressed more topics than the extant Epicurean counterparts. Although this much can be granted, this is not to imply that one can easily acquire a clear idea of what Epicurus and his followers believed about poetry and especially their conception of the good poem. It is well- known, in antiquity even notorious, that Epicurus despised poetry and all learning as unnecessary because it was not conducive to but actively distracted from philosophical accomplishment51. In a much-quoted utterance he advises his friend

Pythocles to "flee all education hoisting sail"52. Epicurus’ bluntness, whatever his exact motives and personal attitude might have been, was bound to gain him the

“See e.g. Klein (1975), Fusillo (1985) 140, Heath (1989) 65-67 and Hunter (1989) 32-38.

51On Epicurus’ views on poetry and education see Ronconi (1963)8 ff., Angeli (1988) 61-70 and Asmis (1991); cf. Isnardi Parente (1966) 394-397 and Clay (1983) 16 n.12 and 78 n.58.

52U 163. See also U 117; cf. Plut. Mor. 15d and Quint. 12.2.24. 145 unmitigated hostility of critics like Plutarch and the allegorist Heraclitus3 who did

not share Epicurus’ philosophical concerns and had no problem accomodating poetry

and other liberal pursuits in their writings. Epicurus inevitably invited on himself

criticisms to the effect that he did not partake of the education he scorned and even that he shamelessly plagiarized poets, especially Homer, whom he verbally abused while drawing doctrines from the epics without so much as acknowledging the

importance of his source54. Several thinkers before Epicurus had condemned poetry

as morally harmful, most prominently Plato whose frontal attack on poetry has

caused the discomfort of critics down to the present day. Although Plato’s position would be endorsed by very few, if any, at least he has never been considered an enemy of education and a stranger to the delights of poetry itself.

Epicurus himself has only recently been vindicated of the charge of boorisheness: following Clay, most recently As mis has collected and discussed the ancient evidence constructing a plausible argument in support of her claim that

Epicurus only rejected poetry as a serious occupation and/or substitute for the rewards and worthy pursuits of philosophical training while considering it perfectly acceptable, though not necessary and beneficial, pastime55. By these lights the

Epicureans are not supposed to look to the poets for enlightenment or help with argumentation and validation of their premisses: poetry is likely to have serious

53Homeric Problems 79 and 4 (= U 229); cf. Athen. 5.187b (= U 228).

^See e.g. U 228; for a full collection of references see Bignone (1936) 270-273. Cf. Kaiser (1964) 220-221.

“Asmis (1993). 146 moral flaws and shuns rational argumentation which is the main Epicurean avenue to understanding and philosophical accomplishment. The Epicureans should not spend time studying or composing poetry because it does not alleviate pain or offer genuine pleasure and thus it is philosophically worthless. But, like everybody else,

Epicureans can of course very well enjoy poetic performances in their leisure time56, if they so wish. They can avoid the trap of moral harm because their astute minds, trained in the philosophical wrestling school of Epicureanism, know right from wrong.

As long as poetry remains only a pleasurable pastime, indulged in in moderation, there can ensue no harm from it. Of course it is strongly advised that the Epicureans abstain from critical discussions of poetry: although only they asaoQoi are qualified to judge poetry correctly57, there is nothing to be gained from the theoretical examination of a basically worthless thing and the time consumed in such idle pursuits could and should be spent much more profitably in philosophical speculation. Epicurus’ followers never questioned the validity of their master’s judgment but actively engaged in pursuits that would be rejected as unsuitable by orthodox Epicurean standards58. Asmis claims that later Epicureans, i.e. laymen like

Cicero’s Torquatus, and professional intellectuals like Philodemus, were able to justify their apparent deviation from the master’s doctrine by taking refuge in the

“U 5 and 20; cf. D.L. 10.120.

^ .L . 10.121.

^See Hirzel 1 (1877), who first proposed the rift between Epicurus and his followers, and Regebogen (1932) 81; for and overview and discussion of later scholarship see Angeli (1988) 82-102. 147 handy distinction between amateurism and professional occupation, or pastime and serious pursuits59.

Although on the whole plausible Asmis’ argument falls flat at the end where she discusses Lucretius’ case. She thinks that the Roman put the crowning touch on the happy compromise which the Epicureans had struck. In a brilliant stroke of both devotion to the master and poetic fervor, he would boldly assert to Epicurus’ censorious ghost that poetry, at leasthis poetry, could indeed promote the common cause by virtue of its palatable sweetness and radiant luminosity. By this innovative argument or, as Asmis puts it, this "novel exegesis of Epicurus’ views on poetry",

Lucretius offers a very effective defense of his choice of the unconventional or, more bluntly, problematic medium of poetry which thus appears in a new light as "making a positive contribution to the presentation of philosophy"60. First of all it is not at

59Attempts to explain the discrepancy in terms not of a rift but a "compromise" or difference between theory and practice were also made by earlier scholars; see Tescari (1935) esp. 58ff., Ronconi (1963) and Giancotti (1972) 195ff. In Philodemus’ case Gigante (1990) 63-79, 103-116 and (1989) has claimed that epigrams written after Philodemus’ conversion to Epicureanism are inspired by Epicurus’ doctrines. In a similar vein Snyder (1973) has detected connections between Philodemus’ epigrams and his On Poems, an approach recently ridiculed by Gigante (1992). More bibliography on the relationship between poetry and philosophy in Philodemus’ work in Romeo (1983).

“Asmis (1993) 92. Cf. also Waszink (1954) 257 and Kenney (1977) 11 who stressed that Lucretius’ poetic identity would predispose him favorably to poetry and that he might have chosen his medium in order to attack the wrong kind of poetry composed e.g. by Philodemus: Lucretius’ poetry gained in importance because it was of a "technical" or "natural" kind. A similar line was also taken by Buchheit (1984). It is very plausible that Lucretius would or could never entertain the possibility of writing in prose and that he considered his poetry excellent and most appropriate for the task he undertook. Nevertheless, his personal beliefs cannot substitute for an adequate defense of his choice in orthodox Epicurean terms because there is no 148 all clear and, much less, cogent that Lucretius offers any defense of his medium, or that he felt constrained to offer any, in the narrow Epicurean terms Asmis foists on him. Nowhere does Lucretius address the issue of Epicurus’ attitude to poetry and nowhere does he seem to think that he deviates, or that others could perceive him as deviating, from the master’s dogma. By Asmis’ lights this reticence would probably be a clever way of avoiding the explicit recognition of the problem. But this is plainly arbitrary and draws unwarranted conclusions from the text. The two poetological pronouncements she cites as Lucretius’ defense, namely the honey on the rim of the bitter cup61 and the luminous verses62, can much more plausibly be viewed as self-confident pronouncements of a poet in the Hellenistic tradition, proud of his accomplishments and not loath to stress them emphatically. The luminous transparence of the medium and the resulting ease with which the reader could discern and digest the liberating philosophical message do have of course some connections with Epicurus’ theory but nothing to do with a defense of Lucretius63. indication that Epicurus would leave room for such a thing; cf. the discussion of Schmidt (1990) 5-9.

611.936-950.

“ 1.143-144 and 933-934.

“On Lucretius’ use of the light-darkness imagery see Schrijvers (1970) 40 and Giancotti (1978) 53-54. Recently Milanese (1989) 107-114 argued very plausibly that Lucretius’ terminology is heavily indebted to rhetorical theory and especially the virtue of oafivexa which Epicurus considered the only virtue of philosophical prose and played a major role in Roman rhetorical theory. Or course it is a perfectly natural and expected thing for Lucretius to be conversant with rhetorical theories current in his day and expecially those advanced by Epicurus and his followers. But the application of terms used for philosophical prose to a poetic text is neither self- evident nor perhaps even allowed by strict Epicurean standards. Onaa

Lucretius felt that his practice conflicted with Epicurus’ teachings and especially that

the poet would be concerned to justify it as best he could. On this unjustifiable preconception one could perhaps detect in the poetological pronouncements something that could resemble a covert and indeed obscure defense but the blatantly circular nature of the argument robs it of any claim to validity.

Even if one went along with Asmis’ initial assumption that Lucretius did actually have a problem and choose to address it the way she suggests, it turns out that she only manages to portray him as a philosophical and poetic simpleton, singularly incapable of picking an effective line of defense in terms of the system he advocated. There could hardly be constructed a more un-Epicurean argument than the one she puts in Lucretius’s mouth. To claim that a poem could function as a demonstrative proof (&7roSeiKTiicdc Aoyoc) and that it would work better than dull and obscure prose would not win over any decent Epicurean to Lucretius’ side, least of all Epicurus himself. For Epicureans no poem could illuminate readers more than philosophical treatises and no poem could possibly substitute for them, let alone usurp their place. As will be mentioned below, even poems free of moral and other faults lagged behind in comparison to philosophical prose because the poetic medium is bound to distract the audience from the arguments exposed in the poems64. Of

also below n.67.

64Asmis (1993) 79-86 (cf. 72-73) herself discusses at some length the evidence, mainly from Philodemus’ On Music and On Poems, that melody and rhythm as well as poetical language can at best be neutral but most often harm the listener. Only 150

course Lucretius and any other poet before or after him could claim that their poetry was unobjectionable on both poetical and, more importantly, philosophical grounds

but, as mentioned above, hardly any Epicurean would grant them that without further

ado. If poets really wanted to defend their art, they would have to tell a rather

different story from the one Asmis postulates. Last but not least, Lucretius’ supposed

line of defense according to Asmis comes so close to the views of the Stoic Cleanthes

that no Epicurean would be willing to take it.

As far as we know, Cleanthes was the only Greek thinker to proclaim that

poetry is the best medium for the elucidation of philosophical doctrines, because the

rhythms and stateliness of poetry are closer to the divine realm and are able to move

the listener’s soul conveying the desired message much more efficiently than plain

prose65. Asmis assumes a similar rationale on Lucretius’ part: Epicurean doctrines,

obscure and inaccessible to the layman, would be handsomely illuminated in a user-

friendly way by Lucretius’ poetry66. Such an argument would only manage to win

Lucretius the contempt of Epicurus and his followers, none of whom would be likely

to admit that their liberating doctrines in the master’s voluminous writings were

the words in poems and song might incidentally provide benefit but philosophical prose does the same thing much more efficiently.

“On Cleanthes’ theory see below 160-162.

“It is true that the Roman audience Lucretius seems to compose for might have found Greek philosophical prose alienating; see e.g. the complaints of Antonius in Cic. De Orat. 2.61. To such boorish readers Lucretius assumes the pose of teacher and liberator; cf. Clay (1983) 40-41. Again the peculiar problems of Roman intellectuals, real or alleged, cannot of course serve as a justification of poetry in Epicurean terms. 151 obscure and in sore need of a little help from Roman poet friends. Such a view would imply that the insistence of Epicurus and his followers onoa

remained a mere verbal construct that failed miserably to come to any fruition in

Epicurean literature. There is no denying that attitudes and beliefs of later

Epicureans conformed with Epicurus’ doctrines and thus significant continuity can be

observed in the circles of the Garden as Asmis claims. I do not think, though, that

the attitude towards poetry should be counted among these constants. The

Epicureans, of course, would never admit that much openly since adherence to

Epicurus’ doctrines was one of the most prominent tenets of the school.

Nevertheless, this does not imply that the Epicureans were orthodox fanatics or that

they took diligent care to justify whatever deviation from the tradition they embraced or that, even if they wished and tried to, they would be successful. Lucretius for one valued poetry unabashedly and practiced it with no compunctions or excuses.

I went into some detail about Epicurus’ attitude to poetry because, even if his

rejection of it as a serious pursuit seems to have met with de facto disagreement by

some of his followers, his theory that poetry could not count as philosophically

acceptable seems to have permeated Epicurean views throughout, as far as one can judge from the surviving evidence. Although Epicurus dismissed literary-theoretical

‘‘The basic testimony on the importance ofoa< Kai £v r He pi 'PrjropiKfjc &%ioi fjirjSiv akko r) aa^rjveiav dTraireiv). Cf. also Cic. De Fin. 1.14-15 and 2.21. On Epicurus’ style see also Steckel (1968) 635-637, Angeli (1981) 28 and Milanese (1989) 34-38. 152 discussions as a harmful waste of time, Philodemus wrote a work on poetry and there is evidence, mainly in the form of titles, that other Epicureans indulged in similar inquiries68. But this cannot be taken as a theoretical defense of poetry or an upgrading of its philosophical status; it should rather be viewed as another manifestation of the tension observed between Epicurus’ pronouncements and the practice of his later followers. Because no Epicurean is known to have ever disputed

Epicurus’ original views, the basis of his theory remained virtually unchallenged.

Thus in Epicurean circles poetry remained a lovely thing to be enjoyed by philosophically knowledgeable people. As already mentioned, no attempt was made to define what a poem or a particular kind of poem is and especially what makes a poem good. Philodemus touches on such issues only inasmuch as it suffices for the refutation of his opponents. The Epicureans apparently believed that poetic goodness is a common notion: everybody knows which poems are good and, consequently, there is no need to fret over unnecessary definitional questions69. It is hard to imagine that the idea of writing a work on poetry structured around an elaborate definition like the Aristotelian Poetics would ever occur to an Epicurean, especially Epicurus himself. Nevertheless, Philodemus and probably one of his major

“For references see Mangoni (1993) 27 n.14.

^In De Poem. 5.25.14-18 Philodemus mentions that the so-called philosophers failed to take into account the "notions" (£vvoiac) of poetic components and poems. ivvoia is a terminus technicus synonymous withnpokrjtyiQ, one of the Epicurean criteria of truth along with perception and passions; see Currie (1961) 82, Kleve (1963) 23ff., Furley (1967) 20Iff., Long (1971), Manuwald (1972) esp. 63ff., Striker (1974) 68-72 and Asmis (1984) 19-34 and 61-80. 153 sources, his teacher Zeno of Sidon, before him do discuss and refute the views of others on the attributes of a good poem. Thus there is the possibility of acquiring some knowledge of what the Epicureans considered blameworthy or at least irrelevant in poems. One would expect that, knowing what a poem should not or need not do or be according to the Epicureans, one would be able to form a relatively clear idea about Epicurean poetics. Unfortunately, besides the paucity of available relevant information in Philodemus’ extant text, his relentless attacks against several critics who espoused a fairly wide range of opinions leaves little room for figuring out what a possible Epicurean position might have been, as several scholars have noted70.

The closest Philodemus ever comes to articulating a positive and coherent view is when he claims that neither good form nor good content alone is enough to make a good poem; instead, in order to effect a truly good poem the two need to work in unison. Philodemus explicitly rejects both a moralistic and an aestheticist view of poetry. He does not consider a morally edifying content a prerequisite for a good poem and he even claims that, were one to look exclusively for didactic messages in poems, one would have to jettison a great many wonderful works of the past; on the other hand, a morally unobjectionable poem would most probably turn out dull unlike e.g. Homer’s literary masterpieces71. But Philodemus never goes into

^See e.g. Rostagni (1955) 427, Grube (1965) 193 n.2, Innes (1989) 215, Milanese (1989) 140, Asmis (1992b) 415 and Mangoni (1993) 28.

71For a collection of references see Mangoni (1993) 29-31. For discussions of Philodemus’ poetics see also Greenberg (1990) 271-276, Grube (1965) 193-199, Innes 154 much detail about what the high quality of the poems in question, either old or new, consists in. In 5.25.34-26.11, in the context of his discussion of Crates’ attack on the philosophers, Philodemus notes that

they were wrong in assuming that all literary criticism depends on conventions (Gc/zara) and that there is no general criterion for judging good and bad compositions. [Instead they thought that] different people have different criteria as is the case with customs (vo/ji/xot). It is true that [a poetic composition] qua poem does not procure any benefit, either in respect to style or to thought. That is why there have been established criteria (ctkottoC) of [poetic] goodness, for style to imitate the style which teaches useful things and for thought to partake of the middle ground between the thought of wise men and the vulgar crowd. And this is how things are, whether one believes it or not, and the judgment of poetry should follow these principles.

The first requirement has generated discussion because it is not clear to what

Philodemus refers. Jensen72 thought that he wanted a style imitating everyday speech. Greenberg73 suggested that Philodemus intended a didactic style with expository tone. As Mangoni74 notes, Grube’s 75 suggestion that Philodemus implies the style of philosophical treatises is supported by other Philodemean passages like

De Poem. 5.33.15-20 and Rhet. 4.4.1-15 (Sudhaus 1.149). There it is explicitly said that the good style should imitate that of philosophers which is superior to that of

(1989) 215-219 and Asmis (1991) 1-17.

^Jensen (1923) 157-158.

^Greenberg (1990) 84.

74Mangoni (1993) 283.

^Grube (1965) 197-198. 155 rhetoricians and sophists "because, if one considers best the style which imitates that which signifies in the most literal sense (ttj t& Kupi6rara or] pa iv 060 q [ k i t e i],

Rhet. 4.4.10), this person will not give second place to the style of the wise man and the philosopher"76. If the same thesis is implied inDe Poem. 5.26.1-4, Mangoni77 rightly points out that it is not free of problems: how closely can poetry approach the style of philosophy when the former often lacks the rigor and clarity of the latter78?

Equally problematic I find the latent implication of didacticism that such an account suggests. If aesthetic beauty is equated with philosophical exposition, it seems a natural conclusion that didactic poetry is the best kind, in other words that the best poetry would be the kind with both style and content that imitated philosophical treatises. But it is known that the Epicureans did not demand moral correctness and didactic usefulness from poetry. This is apparent in the second requirement of the above quoted passage, namely that the content of a good poem should occupy a middle place between the thought of philosophers and common people. This of course precludes any moral/didactic correctness from good poetry and it is consonant with the rest of Philodemus’ De poem . 5 (4, 16.9ff., 17.32ff., 25.30-34 and 32.3ff.).

The only modest stricture that Philodemus imposes on the poet here is to avoid

76For Philodemus’ views on style in the Rhetoric cf. below.

^Mangoni (1993) 284.

^Representative of Philodemus’ attitude to poetry and consonant with Epicurus’ views is De Poem. 5.31.26-32: "and the poets are not allowed to pursue clarity without qualifications in all cases nor is the kind of clarity open to them suitable for all thoughts". 156 completely vulgar content. How to reconcile Philodemus’ proposals about style and

content remains a problem but Philodemus possibly thought that poetry could indeed

emulate philosophical style, althought it was not necessary or even possible for it to

emulate philosophy in every respect79.

One of the possible criteria that Philodemus most vehemently attacks is

rhythmical elaboration, the attention to euphony which several Hellenistic theorists,

especially the so-calledkpitiko C, had elevated to the status of the chief criterion

of poetic excellence, proclaiming hearing the only faculty responsible for judging poetry. Philodemus adamantly denies that sounds alone, irrespective of their

meaning, can make a poem worthwhile. Still, what a good poem is according to

Philodemus and the Epicureans remains puzzling and almost inscrutable to modern

critical attempts. What does "good content" mean? It can conceivably be "morally beneficial" but it does not necessarily have to since, as mentioned above, many excellent or at least great poems according to every reasonable person’s opinion are morally reprehensible. Could it be that great poems are supposed to deal with complex and serious events in the Aristotelian sense or that they are supposed to portray specific kinds of characters and plots? This seems highly unlikely. Epicurean literary criticism apparently never went beyond "common notion": if everybody knows

79An attempt to circumvent the problem was made by Asmis (1991) 7-11 and (1992) 148-150 who claimed that Philodemus’ remarks did not address the excellence of poetryper se but merely the incidental usefulness which poems might, but do not have to, exhibit. Against the philosophers who suggested that poetic usefulness is judged according to Qipara Philodemus would argue that other criteria, objective as it were, i.e. common notions, existed. I agree with Mangoni (1993) 284-285 who points out that Philodemus’ text hardly allows for such an interpretation. 157 beforehand which poems are good, there is no point in theoretical discussions that tackle this question. Thus it turns out that, if one compares Aristotle’s Poetics and what has survived of Epicurean literary theory, one comes up with very few, if any, concrete points of contact. The primary common thesis defended by Aristotle and the Epicureans is the existence of an independent function of poetry which is to be sought and evaluated irrespective of the moral, educational or other benefits a poem might have to offer. Ka0’ o w ohm a, a poem offers only a specific kind of pleasure.

Collaterally, the rejection of rhythm and pleasant sound in general as the informing element in poetry is clearly compatible with Aristotle’s according only a very marginal place to song and diction, although both he and the Epicureans acknowledge their importance in the composition and, consequently, the study of poetry. Nevertheless, the core of the AristotelianPoetics, the analysis of tragedy in metaphysical terms, has no parallel in Epicurean poetics. The Epicureans favored an intuitive approach to poetry that never went beyond the level of an educated person’s discussion of a beloved pastime.

A comparison between the Aristotelian and the Stoic theory of literature yields basically similar results. The only difference is to be found in the much more extensive consideration Stoics gave to poetry80. Their system was much more

“The basic study of Stoic poetics is De Lacy (1948). See also loppolo (1980) 256- 278, Asmis (1990), Greenberg (1990) 57-79 and cf. Isnardi Parente (1987). Much earlier Jensen (1923) had broken new ground with his monumental edition of 158 hospitable to poetry and they often turned to poets as a way of supporting their theories with an authoritative reservoir of arguments and examples. As mentioned above, a direct result of the Stoic fascination with poetry is the considerable diversification of the literary theories that appeared within the school unlike the uniformity of the Epicurean position.

The definition of poem as a piece of significant rhythmical diction was apparently standard in the Stoic school. Besides, Posidonius distinguished between a poem as "diction that is metrical or rhythmical with elaboration... going beyond prose" and poetry as "significant poetic diction containing an imitation of things divine and human" (D.L. 7.60 = fr. 44 Edelstein-Kidd)81. It is not known whether all Stoics shared this distinction but it is reasonable that some at least did. The duality basic to most Stoic literary criticism is clearly apparent in this definition: poems were considered under two aspects, as a structure of harmonious sounds and words and as a carrier of meaning, or as A.££ic and XtyoQ respectively, i.e. articulated speech and articulated speech with meaning. Stoic linguistic theory distinguished among the ruyxdvov or £kt6<; uTTOKeijuevov,the urj/zaivovand the arj/uaivdjuevov82.The most elementary indivisible component of words was the

Philodemus’ De Poem. 5 (see especially 128-145 (on Aristo) and 146-174 (on Crates)).

81On Posidonius’ definition see De Lacy (1948) 244, Gigante (1961) 45ff., Greenberg (1961) and Haussler (1970) 125ff.

^SVF II 166; see Lloyd (1971) 58. a . also Pohlenz (1939) 151-198, Melazzo (1975) and Rispoli (1987) 462-463. 159

and categorization of 0«vcn.

In Posidonius’ definition the difference from Aristotle’s Poetics is instantly

obvious from the focus on the rhythmical and linguistic aspects of poetry. "Imitation"

of course rings closer to Aristotle but the notion is also prominent in Plato and it was

definitely part of pre-Socratic theoretical discussions, especially of visual arts84. The

general character of Posidonius’ formulation in its extant form is foreign to Aristotle’s

analysis and his concern to delineate precisely the appropriate kind of mimesis that

pertains to tragic poetryper se. But the most crucial discrepancy between Aristotle

and the Stoics remains the emphasis on rhythm. As mentioned above, the KpiriKoi

were the most dedicated adherents of a radical theory of euphony which advocated

the autocracy of the sense of hearing85. Not only content was considered extraneous

to poetry but also style (A l£ic) and its virtues (Philod.Tr. C 17.4ff.). The poetic art

consisted solely in ativSecric, the harmonious combination of elementary sounds,

and in stylistic elaboration. At least one Stoic, possibly Aristo of Chios, shared this

view and Philodemus explicitly says that he took it over from the KpiriKoi (De

Poem. 5.18.7-17).

The identification of Philodemus’ opponent with the flamboyant third century

D.L. 7.55-59 = SVF 3.212-214. On vfi see Barwick (1922) 91-94 and Ax (1986) 15 Iff.

MSee Russell (1981) 99ff.

“On the KpiriKoi see Jensen (1923) 146ff., Pohlenz (1933) 53-92, Gomoll (1937) and Schenkeveld (1968) 185ff. 160 Stoic from Chios is far from certain. Jensen was the first to suggest Aristo86 and

many subsequent scholars accepted his argument87. Recently, though, Isnardi

Parente challenged Jensen’s view arguing that Philodemus’ report is incompatible with the rest of the available information on Aristo and indeed with the doctrines of

the older Stoa88. According to her Philodemus’ opponent must have been a grammarian or philosopher associated with the middle Stoa89.

Whoever Philodemus’ opponent might have been, the main points of his poetic theory can be summarized as follows: (1) a requirement for good diction plus

a morally blameless thought, (2) the classification of poems into good, bad and intermediate and (3) the emphasis on euphony and the fact that poems are judged by hearing. Asmis90 suggests that the division into thought and composition, in her opinion mainly a Hellenistic innovation, had important consequences for the history of literary criticism: it supposedly promoted a strict demarcation of features intrinsic and extrinsic to poetry, like diction and music and dance respectively. This distinction militated against the Aristotelian position that pronounced tragedy the

“The identification was based upon Jensen’s restoration of the damaged papyrus text: [6] roivuv |dvT[e]x6[/ie]v[o]c t &v [Sjroi|[kov 'Aptarjwv (Philod. De Poem. 5.16.28-30).

87Pohlenz (1933) 77, De Lacy (1948) 252-253, Grube (1965) 136, Schenkeveld (1968) 177 and 180 (with caution) and loppolo (1980) 256-278.

“Isnardi Parente (1987) and (1989) 32ff.

89Philippson (1924) had much earlier rejected Jensen’s proposal. Asmis (1990) and Mangoni (1993) 61-69 find Isnardi Parente’s thesis plausible, although with some qualifications.

90Asmis (1990) 155-157. 161 queen of all other genres because it included music and dance. It is, to say the least, bizarre to attribute to Aristotle an emphasis on the importance of musical embellishments as the basis for tragedy’s superiority. He naturally considers them pleasurable but stresses their irrelevance to the mythos and their ultimate expendibility. It is definitely hard to think of a more rigorous distinction between extrinsic and intrinsic elements in poetry than Aristotle’s in thePoetics.

But the most important advantage of this Stoic’s theory according to Asmis is the attribution of all thoughts included in the poem not to the characters but to the poets themselves. Thus there ensues the possibility for a critical upgrading of genres like lyric which were considered inferior by Aristotle. I fail to find any indication either that Aristotle considered the thoughts to belong exclusively to the characters or that the Stoic did not consider thoughts as reflections of a character’s situation and thus not necessarily indicative of the poet’s own concerns. Neither

Aristotle nor any other ancient critic for that matter dealt with poetic subjectivity and its expression in a poem. Asmis claims that

whereas Aristotle was inclined to regard a poem as an artefact existing independently of its creator, a poem is now seen as a creation arising from the mind of a poet (my italics)91.

But, as shown in the first chapter, for Aristotle all production has its source in the mind of the artisan and there is no reason to assume that poetic production is

91 Asmis (1990) 156. 162 different in that respect.

As far as it can be determined from Philodemus’ text, the Stoic’s theory seems largely unrelated to Aristotle’sPoetics and certainly does not improve upon it in any way. A more interesting case is Crates of Mallos. Besides his commentaries on archaic epic and drama, Crates made considerable contributions to several branches of linguistics like phonetics, morphology and etymology. With a well-developed literary theory Crates seems at first view to have affinities with Aristotle’s poetics but it soon becomes clear that these affinities are not as important as they might initially seen. Although not a Stoic in the strict sense of the term, Crates was clearly influenced by Stoic theories -ironically Philodemus’ surviving work deals with thinkers affiliated with the Stoic school but not orthodox Stoics. Thus the task of determining the position of the adherents of the school and its various satellites at different periods becomes particularly arduous.Suda calls Crates a Stoic philosopher but this must be an inference based on rather shallow knowledge of the philosophical affiliations of different thinkers92. Crates was a KpiriKoc who thought veiy highly of his vocation: according to him a Kpiruc6c is the only one able to play the role of literary critic, although he may lack the technical expertise of a grammarian93.

92T 1 Mette. The fact that Panaetius was Crates’ student (T 3 Mette) possibly facilitated his aquisition of the title of philosopher. On Crates’ philosophical affiliations see Wachsmuth (1860) 5-6, Jensen (1923) 149ff., Mette (1952) 61, Blank (1982) 3-4, Asmis (1992) 160-161 and Porter (1989) 157-158,168-170 and (1992) 85ff. who denies Stoic influence on Crates.

^According to Sextus (Adv. Math. 1.79 = fr. 17 Mette) Crates believed that "the critic differs from the grammarian; and he says that the critic must have knowledge of the whole range of language whereas the grammarian simply explains difficult 163 Philodemus devotes a large part of his discussion of Crates’ views to the latter’s Auseinandersetzung with the "philosophers", a group of thinkers not further

specified but who could, according to Philodemus erroneously, include the

Epicureans94. Philodemus denies that his own school shares the views of the

"philosophers" who considered thedinar a the sole criterion for the judgment of poetry. These were conventional, arbitrary norms that were agreed upon and regulated the literary perception of poetry and issues like the best author to imitate in every genre95. On this thesis there is no natural good in a poem that can be words, prosody and the like. Thus the critic resembles a master-craftsman (dpxireKTwv) and the grammarian a subordinate craftsman (vnrjpirr) <:)''■ The issue could not simply be settled by an assertion on Crates’ part. Some grammarians for one espoused a much broader view of their discipline ("the grammarian must deal with all aspects of language",Seh. on Dion. Thrax 7.27-28 Hilgard) and Crates himself was called a grammarian by later sources. At any rate it is interesting to observe that Crates presupposes a gradation of arts and science that goes back to Aristotle (Phys. 194a36-b7).

^Different suggestions on the identity of the "philosophers" have been put forth. Philippson (1924) favored the Sceptics while Greenberg (1990) 209 n.49 and Porter (1989) 162-163 thought that the term must refer to a group of critics with whom Crates disagreed according to Philodemus (De Poem. 5.27.7ff.).

95Cf. for instance Philodemus’ remarks on the confusion that results from postulating different dinar a, since it is impossible for everybody to agree on the best style or model: "Because indeed not all people trustdinar a nor all respect the same [0£M<*]-but others emulate the style of Isocrates, others that of Thucydides" (Rhet. 4.7.14-22 (Sudhaus 1.151)). A little below Philodemus returns to the same issue: "But it is not even possible to say that the orators themselves as a group have made up their minds on one kind of style and practice it consistently. On the contrary, we will observe among them, or the majority of them, different tastes...In respect to Isocrates either absolutely nobody or only two or three have the same attitude and some say that not even the style of Isocrates himself is uniform in all his works" (Rhet. 4.8.9-22 (Sudhaus 1.152-153)). Although Philodemus deals here with rhetoric and its inferiority in comparison to philosophy, his conclusions are certainly applicable to other arts as well. See Milanese (1989) 86-93 who suggests that according to the Epicureans the concept of riyyr) cannot accomodate the 164 detected by everyone and at all times. Crates attacked the philosophers claiming that there are no dinar a and that hearing can function as the judge of objective goodness irrespective of any posited and subjective criteria. Philodemus agrees with the philosophers that there is no natural good inherent in a poem qua poem. On the other hand he agrees with Crates that there are no Sinara. In Rhet. 4.7.6-14

(Sudhaus 1.151) he claims that postulatingdinar a would be justified if there existed no naturally beautiful speech but, since this is clearly not the case, it is ridiculous to abandon nature and have recourse to convention. Nevertheless, he attacks as naive

Crates’ position that the sense of hearing, which according to the Epicureans is

&koyoQ, like all senses, can judge the sound of poetry or anything else for that matter.

As befitted a K p ir iK o c , Crates considered elementary sounds (0 o > v cu ) and thus hearing, the most important factor in the judgment of poetry. He differed, though, from the rest of the K p ir u c o i inasmuch as he did not believe that the sense of hearing was responsible for judging the thought: the K p ir iK o i proclaimed that a poem’s euphony, perceived and judged by hearing,eo ipso made its thought praiseworthy irrespective of content, truth value or moral import. Crates suggested that a poem is judged empirically by the hearing which examines the sound in relation to the meaning or, as Philodemus puts it, "not without the thoughts but not the thoughts themselves" (De Poem. 5.28.26-28). For Crates the content of a poem

multiplicity of standards implicit in thedinar a theory. In his subject matter, methods and even style the Epicurean philosopher avoids variation; cf. also Grube (1965) 202. 165 fell under the jurisdiction of reason. While he did not consider a successful poem

simply an agglomerate of harmonious sounds like a musical piece, he thought that

goodness or badness of thought is irrelevant to the judgment of a poem. Although

his position seems somewhat artificial and hard to grasp, it apparently sought to

strike a middle path between the radical one-sidedness of euphonists and those

theorists who postulated thought as the most crucial criterion in the judgment of poems: he disagreed not only with the euphonists but also with the Epicureans who claimed that the hearing cannot judge anything because the sound cannot convey anything to the intellect: content and form should both be good and the poem should be judged as a whole by the intellect. He gave precedence to euphony but kept sound and thought, the criteria for their judgment and the faculties that judge them scrupulously apart. Crates’ position is probably as close as a critic could ever approach the Stoic outlook on poetry. It is known that the Stoics like the Epicureans required both good sound and good thought to proclaim a poem worthwhile and believed that hearing judges the former and the mind the latter. Crates obviously espoused only the second claim of the Stoic theory.

As far as Aristotle is concerned, Crates seems to come closest to the tradition of the Poetics with his claim that poems are judged on account of whether they were composed according to the principle of the art (icard rdvr f)c Ti%vr)Q Adyov, De

Poem. 5.28.2-3). Philodemus complains about Crates’ failure to define this principle.

It seems, though, that the principle of the art, which exists by nature, determines the art of poetry and is judged by perception{De Poem. 5.28.24-26), can only be the 166 principles of sound according to which a poem is judged96. Thus for Crates a

euphonious poem is a well-composed poem that satisfies the requirements of nature

and art. Auditory pleasure is not laudable per se but as an indication of the poem’s

compliance with the principle of the art. This principle regulates not only the production but also the response to a work of art, in this case not only the composition of a euphonious poem97 but also the pleasure that ensues from the realization, through the sense of hearing, that the poem complies with the principle of the art98. Inasmuch as Crates emphasizes the principle of the art as the ultimate criterion of artistic value, there can be detected an ultimate debt to Aristotle but the essence of Crates’ claim turns out significantly different from any of Aristotle’s pronouncements in the Poetics : Crates’ principle of poetics, in contrast to the

Aristotelian mythos, is euphony and the art consists in creating beautiful sound- patterns, an aim which Aristotle obviously could not have condoned as the defining characteristic of poetics.

In another part of his treatise Philodemus talks about a theorist whose name has been lost and who proposed the fairly radical (at least by ancient standards) theory that all poets, and consequently all poetic genres, dealt with the production of good sound or beautiful diction (PHerc 460, 26, tr. B, fr. 17.8-27 Sb):

96Cf. the discussion of Mangoni (1993) 297-298.

97The persistent importance of theorems of art that regulated good composition, i.e. the arrangement of words, sentences and paragraphs is attested by Dionysius(De Comp. Verb. 5.12); see Scaglione (1972) 74ff. and Freudenburg (1993) 132-139.

98Cf. Asmis (1992) 156-158. 167 He says it will make no difference, not even if we compare Archilochus or Euripides or someone else with Homer, if only we juxtapose the praised diction of each. For just because tragedy, iambic, and lyric differ from epic we will nor refrain from comparing poets of different genres, since the goal is the same for each genre (row t£Xovq v>7rd[pxo]vTOc Travri yivei T auro<0> ). For it follows that all diction in them is inserted beautifully, or badly, or indifferently.

The term riko<; marks both the conceptual similarity with Aristotle and the wide gap that separates the anonymous theorist from him. Asmis" has claimed that

Philodemus here talks about Crates and her argument is fairly plausible. If her identification is valid, the difference between Aristotle and Crates, already apparent from Philodemus’ remarks in book 5, becomes irreconcilable.

As mentioned above, the Stoics were in general less radical than Crates, although he shared common ground with them. They attributed importance to both diction and thought. Some Stoics seem to have been closer to the Aristotelian tradition than others. Panaetius for instance observed that the beauty of artworks consists in the arrangement of its parts and the spectator’s perception of this arrangement100. Nevertheless, the overall impression from the comparison between

Aristotle and the Stoics is one of divergent orientations. The concept of imitation, for instance, played a role in Stoic literary theory but both the object of mimesis and its execution are unrelated to their Aristotelian counterparts. The Stoics thought that poetry imitates life and should be truthful to it. The broadness, and the consequent

"Asmis (1992) 167-169.

100See Cic. De Off. 1.14 and 98; cf. D.H. Dem. 48 and Arist. Quint. 1.4. See also the comments of Philippson (1930) 383-384 and De Lacy (1948) 246-247. 168 lack of precision, of this requirement contrasts with Aristotle’s painstaking

delimitation of the notion of mimesis. Besides, the imitation could be symbolic, not literal101 and thus the road to allegorical interpretation opened wide. This is

another un-Aristotelian staple of Stoic literary criticism, less original but arguably the most influential part of the Stoic theory102. It had a far wider impact on later theorists, especially Christians through Philo. Allegory was for the Stoics a way of circumventing the most problematic stumbling block in their discussions of the poetry of the past. Poetry undeniably contained much that fell far short of the strict moral standards espoused by the school. Allegory provided a respectable, although inevitably often implausible, answer to many of these problems. Homer was the prime candidate for allegorical interpretations. He was highly esteemed by the Stoics and even by Epicureans to judge from Philodemus’ treatise On the Good King

According to Homer J03. Aristotle of course also considered Homer a model poet but it is characteristic of his approach that he searches for and pinpoints a concrete reason why Homer excelled among the poets of the past. Several reasons could

101Cf. Kennedy (1989) 212.

102For a detailed collection of ancient evidence see Pohlenz 2 (19704) 55. On Stoic allegory see Pohlenz 2 (19704) 97, Wehrli (1928), De Lacy (1948) 257-258, Bufffere (1956) 137-154, Pfeiffer (1968) 237-246 and Dawson (1992). Steinmetz (1986) and Long (1992) have correctly challenged and qualified the scholarly communis opinio according to which the Stoics considered Homer an allegorist in the strong sense and pursued wild allegorical paths that distorted the poet’s meaning. This sweeping view of Stoic interpretation originated in ancient polemical criticism of the Stoa. The Stoics approached Homer with clearly Stoic lenses but their interpretation was not unrefined, reproachable or radically different from that of many ancient and modern readers.

103On Homer and Philodemus see Asmis (1991). 169 easily be named in support of Homer’s superiority, especially the claim that Homer was the teacher of all Greeks in a variety of different arts, above all the art of war104. Homer’s poetry was intricate and pleasant and people turned to his epics for examples of ethical behavior105. Aristotle, though, mentions only Homer’s poetical superiority which consists in his having been able to construct a mythos out of the vast, non-unified material of the Trojan war saga. For Aristotle Homer is never a revered teacher but merely an excellent poet for a very specific reason.

It is instructive to investigate how Aristotle and the Stoics addressed Homer’s

"mistakes". I have already talked about Aristotle’s views on poetic mistakes in chapter 2. He considered only mistakes pertaining to the art of poetry as compromising the poetic composition. The poet’s ignorance of other non-poetic matters is not on a par with a mistake in the construction of the mythos. Mistakes should be optimally avoided altogether but since this is not always possible poets should be judged only on the premisses of their own art. Occasionally it is even possible that some mistakes or improbabilities promote the final cause of the art: in

104He was even credited with scientific insights. Crates and Posidonius, whose views along with those of others like Hipparchus have been preserved by Strabo, energetically defended the Homeric "doctrines" against attacks like Eratosthenes’ on which see Berger (1880) 19ff. and (1903) 387-388 and 460; cf. Dicks (1960) 38. On Strabo see Schenkeveld (1976). For Crates’ defense of Homer see fr. 30, 34a and 35a Mette; cf. Porter (1992) 85-111. For Posidonius see fr. 49.300ff., 216, 222, 277a and 280-281 Edelstein-Kidd; cf. the extensive discussion in Rudberg (1918) 127ff., Heinemman 2 (1921) 54ff. and Theiler 2 (1982) 6ff. For a more sober view of Homer’s (and Aratus’) "scientific" contributions see Posidonius fr. 48a Edelstein-Kidd and Kidd (1988) 214-216.

10SFor references see De Lacy 264 n.136-139. 170 such cases they are to be not only tolerated but also praised. As mentioned above, the Stoics showed particular affection for Homer and even considered him as a sort of proto-Stoic. It was indeed not uncommon for Hellenistic philosophical schools to vie for the authority of Homer in an effort to enhance their prestige. The contradictory accounts of Homer’s supposed philosophical affiliations were appropriately ridiculed by Seneca106.

Although the Stoics are included among the targets of Seneca’s attack, when it came to judging Homer they could not but make him a bad poet. According to the

Stoic system a poem could be called "good" only if both its content and form were unobjectionably good. Even if only one element fell short of this strict ideal, it sufficed to catapult to whole thing into Stoic anathema107. A poem could be either good, bad or intermediate. Absolutely good can only be poems composed by an absolutely good person, i.e. a wise person108, a creature notoriously rare and virtually non-existent. Despite this problem, the Stoics apparently believed that "good poem" was not a contradiction in terms, as Philodemus contends in his criticism, although admittedly such poems were very rare. Poems which were good in some respects but bad in others, or even bad in only one other, were considered intermediate if they lacked badness in the form of factual or moral falsehoods. In strict Stoic terms

Homer was no wise man because in his poems one can detect several such lapses.

i06Ep. 88.5; cf. SVF 2.1077 and see Noblot (1979) 159-160 and Long (1992) 47.

107See Asmis (1990) 164ff. and Mangoni (1993) 250ff. with further bibliography.

mSVF 2.393 and 3.505, 516 and 525. 171 He could at best be called intermediate if not absolutely bad. The Stoics had to go

out of their way to safeguard Homer from attacks of a Platonic, Heraclitean or even

strict Stoic type. To exonerate Homer from charges of inconsistency Zeno postulated

that the poet employed two principles, truth and opinion(SVF 1.274 and 456)109.

Aristo adopted a lenient attitude and expressed willingness to "forgive" Homer by

calling him good only in a loose sense (icaraxpiloTiKwc andyera ovyyv&prjQ,

Philod. De Poem. 5.17.32 and 18.5, 14). Obviously for the Stoics Homer’s excellence

could barely save him from philosophical contempt.

Much more interesting and original was the Stoic concept ofphantasia and its

role in the artistic process110. As a literary termphantasia appears in later

sources111: it is the ability to present things, especially dramatic, pathetic or

calamitous events, so strikingly and vividly that the audience cannot help experiencing

the desired emotion, like anger or pity. This effect is most efficiently achieved when

the poets or orators themselves feel strong empathy with the misfortunes they depict,

as if they were actual witnesses to, or participants in, the events they describe. The

notion of vivid representation goes back to Aristotle’s Rhetoric (3.11) but the

lwCf. Bufffere (1956) 205.

110Onphantasia see the discussion of Russell (1981) 108-110 and Eden (1986) 107- 111; cf. Imbert (1980) and Fuhrmann (1992) 177.

inPhilostr. Vit. Ap. 6.19, Quint. Inst. 6.2.25ff., 8.3.88, 10.7.15 and [Longinus] 15. It is surely not a coincidence that all these sources have connections with rhetoric because rhetorical theory used phantasia extensively as a means of persuading the audience; see Birmelin (1933), Biihler (1964) 64, Panofsky (1968) 13-16, Schmidt (1976) 62 and Russell (1989) 313-314. 172 appropriation of the termphantasia in the realm of literary criticism and artistic

creation112 was certainly influenced by the use of the term in Aristotle’s DA

427b29ff. for the ability to form images. Artists do not simply reproduce things they

have already seen or experienced, as a strict theory of imitation would have it, but

they are able to create totally or partially new things by drawing on their ability for phantasia, combining old and inventing new elements113. Although I do not think

that Aristotle would object to such a theory or consider it incompatible with his own

theory of mimesis since he is willing to accept tragedies with imaginary plots, it is

true that he nowhere elaborates on the creatively imaginative aspect of the poet’s work. Later theory developed what is at best latent in his work.

On the whole, though, moralism probably constitutes the most serious dividing line between Aristotle and the Stoics. The latter firmly believed that poetry should contribute to the moral edification of the audience, especially the young, who could use it as a preparatory stage for their introduction to philosophy114. Hence, poetry should present the same truths as philosophical prose. Fantastic or "immoral" stories were rejected as mythoi (i.e. fabulae) that compromised the philosophical integrity of a poem. Philosophers and sages had no use for poetry and rejected its pleasure

112Historians also employed the term; see Russell (1962) on [Longinus] 15.1.

113For an example of this process see Cicero De Inv. 2.1-3: Zeuxis created a portrait of Helen by combining the charms of five Crotonian belles.

114Philod. De Poem. 5.14.6-24, 29.15-19, PHerc. 403 fr. 3.8-15 and 4.12-16. See Sbordone (1976) 254-255 and Asmis (1990) 193. For further evidence see De Lacy (1948) 251 n.51, 52 and 269-271; cf. Brink (1971) 352ff. and Russell (1981) 42 and 85. 173 as indeed every kind of pleasure which they considered an irrational movement of the soul {SVF 3.391.20-21). They should not draw ordinary vulgarfjSovfi from poetry but experience xapd115, an agreeable emotion that accompanies the benefit they reap from their interaction with philosophical poetry.

As mentioned above, Cleanthes held the most radical views on poetry’s impact and prestige. In contrast to almost every other thinker in antiquity, including

Aristotle, he thought that the rhythmical patterns, musical accompaniment and exalted diction of poetry befit philosophical concepts about the nature of the divine and the cosmos much more than dry, pedantic prose. Like a trumpet that diffuses the air blown into it amplifying the sound, poetry with its elevated precision brought out serious themes much more strikingly than the freer prose116. Vividness, of course, was an important notion in ancient theories of style and literary criticism but the philosophical twist Cleanthes gave it was peculiar and indeed unique with him.

In the Poetics Aristotle mentions vividness passingly and considers it a characteristic of diction {Poet. 1455a21-29), a natural classification for an attribute that enjoyed prominence in the field of rhetoric and was counted as one of the virtues of style.

usFor the contrast between fiSovfi and %apd see SVF 3.431.16-18 and 434-435; cf. Bonhoffer (1890) 293-298 and Asmis (1990) 192-193.

116See the eloquent account of Seneca: "Nam, ut dicebat Cleanthes, quemadmodum spiritus noster clariorem sonum reddit, cum ilium tuba per longi canalis angustias tractum patentiore novissime exitu effudit, sic sensus nostros clariores carminis arta necessitas efficit. Eadem neglegentius audiuntur minusque percutiunt, quamdiu soluta oratione dicuntur; ubi accessere numeri at egregium sensum adstrinxere certi pedes, eadem ilia sententia velut lacerto excussiore torquetur" (Ep. 108.10). Cf. Phld. De Mus. 4.28.1-22 and see Asmis (1992b) 400-401. 174

Vividness, however, is a secondary element in the construction of the mythos, relevant primarily in the context of a theatrical performance: by visualizing accurately the happenings on stage the poet is able to present a clear picture, guaranteed to capture the attention of the audience and to avoid possible damaging slips117. Cleanthes’ conception of prose and poetry is reminiscent of Aristotle’s comparison of history and poetry where the former is presented as encumbered by the inclusion of many incidental and inconsequential events that do not follow according to probability or necessity. The intervention of tragic mythos turns these disparate events into a work of art more akin to philosophy than history is. The metre, rhythm and diction of

Cleanthes’ philosophical poetry apparently had a similar effect: they turned a looser prose-account into a didactic philosophical masterpiece apt to present the truths of

Stoicism, both scientific and moral. Again the emphasis on music and rhythm as well as the moralist slant of Cleanthes’ views demarcate them clearly from Aristotle’s concerns in the Poetics.

It is very hard to find anything analogous to moral instruction in thePoetics.

The question of tragedy’s, and poetry’s in general, impact on the morality of the audience is never broached there. It apparently never was high on Aristotle’s list of priorities and it is obvious that he never considered poetry a means of moral instruction118. The old dilemma whether poetry is supposed to please or teach is

117On vividness see Zanker (1981). For the views of later literary critics and orators on the subject see Eden (1986) 85-111, esp. 88 and 108-111, with a collection of evidence and bibliography.

U8See Russell (1981) 91-93. 175 addressed by Aristotle in a characteristically idiosyncratic way: to please not in an

intuitive way but through the construction of a rational complex artefact, the mythos.

Nevertheless, since tragedy at least deals with the actions of personal agents, ethics

is inevitably drawn into the picture, if not so much by Aristotle himself, definitely by

several modern interpreters. Actions of course have a moral basis since they are

determined by and betray the agent’s rational choice (jrpoatp ecuc). Theethos of

tragedy should reveal the rational choice of the characters (Poet. 1450b8-9,1454al7-

19). Besides, Aristotle explicitly refers to tragic characters as "greater than ourselves"

(Poet. 1454b8-9) and to early poets as having a moral predisposition for their genre

(Poet. 1448b24). Thus it is undeniable that moral parameters play some role in the

outlining of the profile of tragedy but it should be stressed that the distinctly

secondary role ethics plays in thePoetics has nothing to do with the Stoic insistence

on morality and instruction through poetry. Aristotle apparently rejected both Plato’s views on the worthlessness of poetry and the view Plato attacked, the use of poems

as educational compendia whose credibility is guaranteed by the prestige of their

authors.

Although it seems reasonably clear that Aristotle nowhere considers tragedy or poetry a means of educating the public, very recently Janko, following earlier critics like Lord, Golden, Nussbaum119, proposed exactly such a theory in his discussion of the notorious problem of catharsis, its nature and function within

119See Lord (1982) 102-104 and Salkever (1986); cf. Golden (1976) and Nussbaum (1986) 382ff. See also Belfiore (1992) 79-81. 176 Aristotle’s theory of tragedy. In a nutshell Janko suggests thatPolitics 8 holds the key to unlock the mystery of the tragic catharsis which he considers as a habituation of the general public, by means of mimesis and emotional arousal, to feeling the right kind and amount of emotions in real life. On this view tragedy and comedy contribute to the attainment of the ethical mean and thus to virtue, to the formation of mature and emotionally balanced citizens. According to Janko this is Aristotle’s response to Plato’s pessimistic and censorious attack against drama which stirred and nourished undesirable emotions apt to harm the formation of worthy citizens.

The issue of catharsis falls outside the scope of my discussion but I think that it is worthwhile to discuss Janko’s theory in some detail because, if he is right, then

Aristotle would not differ substantially from Hellenistic literary critics as far as the function of poetry is concerned. I will argue that Janko errs completely: not only does he assume without qualifications that virtually everything inPolitics 8 bears on the Poetics , he also disregards, and sometimes misunderstands or misinterprets, the distinctions Aristotle so meticulously draws in thePolitics. His argument does not do justice to the Aristotelian texts involved and thus I consider it necessary to refute it, all the more since it is one of the most recent and equally unsuccessful attempts to explain the concept and the function of the tragic catharsis.

In Politics 8 Aristotle addresses the question of the education of young men in a well-governed state. Having quickly established that the formation of future good citizens should not be left to the individual initiative but be considered a public affair of the utmost importance, he then proceeds to discuss the four components of 177 traditional education: grammar, athletics, drawing and music. Since the first three

are largely self-explanatory, he devotes the greater part of the book to the discussion

of music, whose benefit is not immediately apparent and whose function was a matter

of debate. Aristotle thinks that music has three possible functions, play and

relaxation, moral education and Siayutyfi (Pol. 1339al5-26). This is a very difficult

term to translate because it is a complex notion which does not correspond to any

single modern word. It describes the way free, educated men spend their time in like

company engaging in agreeable intellectual activities as exclusively befits them (the

term is also used sometimes for their gatherings). They have attained happiness

(eu6ai/xovCa; cf. Pol. 1338alff.) and this is manifest in their leisure time, (i.e.

practically the totality of their time since they do not work), spent in intellectually pleasant pursuits like listening to music; henceforth Siayosyfi will be translated as

"intellectual entertainment". Janko misleadingly translates it as "educative entertainment", a rendering which begs the question of his discussion: there is nothing in the references to Siaywyfi in the Politics that might suggest education.

Aristotle specifies that music should not be taught to young men as play: this is inappropriate because all learning is accompanied by pain(Pol. 1339a27-29). Nor should music be offered as intellectual entertainment to them -young men are still immature (drcAeic) and the end (riX oc) does not become them (Pol. 1339a29-31).

It follows that music contributes to their moral formation and it is appropriate entertainment for them when they turn into adults(Pol. 1339a31-33). But obviously not all music is suitable for young men and education. Aristotle condones the 178 pronouncements of musical writers and philosophers on the classification of rhythms and melodies referring his readers to their treatises for more details. He himself very briefly mentions the three kinds of rhythms, namely ethical, practical and enthusiastic. The first should be used for educative purposes, the other two for the purposes of catharsis and entertainment. Especially people prone to pity and fear and such emotions could find useful and pleasant these two kinds of music, which would "purge" them from their emotions in an agreeable, harmless way. Enthusiastic songs have exactly the same effect on people prone to enthusiasm: such songs put them in a state of exultation which results into calm by means of the musical catharsis {Pol. 1341bl9-1342bl6). Aristotle concludes his discussion of these kinds of music by commenting on their suitability for dramatic performances: since the spectators are of two sorts, free and educated men on the one hand and the vulgar rank and file on the other, the composers of music for the theater should take care to provide even the uncouth hoi polloi with melodies that appeal to their crooked souls in order to please them since each person draws pleasure from things that suit his nature {Pol. 1342b 16-28).

Now, based on this account, Janko reaches the conclusion that the Athenian citizens would acquire better moral disposition by attending theatrical performances.

Since, he claims, Aristotle says that representations can affect people and that people show the same reactions to both reality and representations {Pol. 1340alff.), the harmless emotional arousal and the ensuing catharsis that take place in the theater can habituate the citizens to avoid emotional excess and overreaction in their daily 179 life. It had always been clear to me that one cannot plausibly draw such inferences from the Politics. First of all, it is not automatically evident that music, or its nature and function, is perfectly parallel to, and even identical with, drama. One has to tell a fairly complex story if, like Janko, one is to advance the theory that the part and, what is more important, a part that does not even remotely approximate the centrality of the mythos, can achieve the same results as the whole120. Besides, from

P ol 1342al6-28 it is obvious that music caters to the needs and tastes of the vulgar part of the audience. Even if one were ready to agree with Janko that music and theater have no essential difference, Aristotle’s insistance on the distinction between educated and lowly people, also apparent in the last chapter of thePoetics (1461b28,

1462a2-4), automatically disqualifies a large, definitely the largest part of the audience, from enjoying the beneficial influence Janko attributes to tragedy.

Aristotle makes it perfectly clear inPol. 1337b5-15 (cf. 134 lb 12-14) that vulgarity, i.e. the occupations and concerns of the poor, working and uneducated masses, are so fundamentally beneath the dignity of free men that one should take great care not to teach the young men anything, or anything to a degree, that could denigrate them to the status of manual laborers. These, Aristotle repeatedly claims, cannot

120Belfiore (1992) 326 also mentions the crucial difference but she thinks that in the Politics Aristotle talks about "abnormal people" and "is only concerned all those who are inclined to be abnomally emotional (7ra0t?riKo6c: 1342al2-13)". But exactly following the pasage she cites, in the very same sentence, Aristotle says that all people, to the extent they are prone to emotions, experience some catharsis and pleasurable release. The process of catharsis is similar for both more and less emotional types. Depew (1991) 368-369 summarizes the reasons why thePolitics and the Poetics cannot deal with the same concept of catharsis. 180 understand and appreciate the most refined forms of art, let alone learn anything

from it and attain the highest moral good.

This leaves the other part of the spectators, the free and educated liberal men,

as the only candidates for Janko’s adult education. Now the crucial question is two-

pronged: whether these particular adults need further education and whether drama

can perform the function Janko ascribes to it. From Pol. 1339a26-33 it is clear that

music is suitable for the intellectual entertainment of educated men. In the same

breath Janko mentions that" diagoge " is for adults what play or amusement(paidia )

is for children (paides ) (1339a30)" and "the relation between catharsis and diagoge

may be closer than first appears. This passage [1339al5-24] suggests that they

combine to perform for adults the function which paideia performs for children, i.e.

the training of both the emotions and the intelligence, with the theater regarded

almost as a form of adult education" (345). Neither equation holds. Concerning the

first, Janko was clearly caught nodding when he made it because nothing in the

Greek he cites corresponds to his interpretation. In Pol. 1339a29-31 Aristotle clearly

says that Sxayayf) is unsuitable for the immature. Besides, in Pol. 1339b25-42 (cf.

also 1337b33ff.) Aristotle makes it clear that play has nothing to do with Siaywyfi,

although people often confuse the two on account of the delight that accompanies both. Since music belongs to the most pleasant things, it is appropriate for both play

and 6 lay ay fi. Relaxation from strenuous activity necessarily generates pleasure and

Siaytayff should comprise not only the good but also the pleasant. But play and relaxation is not to be mistaken for the ultimate end. 181 Concerning the second equation, Janko cites Aristotle’s tripartite classification

of the functions of music in Pol. 1339al4ff. to support his thesis that catharsis and

Sxaycsyf) have an educative function. Aristotle refers again to this tripartition imPol.

1339bl2-14 calling the functions in questions education (7rai6eCo), play (fraitfid)

and intellectual entertainment (Siayoyf}). The last time he mentions the functions

of music, the passage is much less straightforward:

We say that music should be used with an eye not only to one benefit but for several purposes, indeed for education and catharsis -we employ the term catharsis here without qualification but we will return to it in more detail in the Poetics- and thirdly for intellectual entertainment, for leisure and relaxation from strenuous activity" (Pol. 1341b35-41).

It is obvious that there are two problems with this pronouncement: catharsis, which

appears here having only been mentioned passingly once before inPol. 1341a23, and

the last sentence which seems to equate intellectual entertainment with play and

relaxation. This cannot of course be true given Aristotle’s prior discussion in the

Politics. Thus either Aristotle speaks here extremely loosely, almost carelessly,

equating intellectual entertainment and play by virtue of the pleasure of music

common to both or there is some textual problem involved121. The issue of catharsis

cannot be settled. It is idle to speculate about the reasons for its omission from

Aristotle’s earlier discussion -the composition and nature of his text is notoriously prone to additions and afterthoughts, although of course not so easily to

121For a summary of different suggestions see Aubonnet (1989) 47 n.4. 182 contradictions. At any rate, what is important for my purpose here, is that catharsis cannot play the role of moral educator. Education and the music appropriate to it remain firmly distinguished from catharsis and cathartic songs in thePolitics. The only function of music that is related to catharsis, and specifically to tragic catharsis for that matter, is relaxation (avdwauaic), the very same term that appears inPol.

1341b41. Indeed, if it were not for the apparent and definitely problematic association of relaxation and intellectual entertainment in that passage, that is if

Aristotle mentioned only education-catharsis-intellectual entertainment, it would follow naturally from his previous discussion that catharsis would play a role very similar to, or at the very least be a kind of, pleasurable relaxation.

Be this as it may, the catharsis of music, and, if one is willing to equate the two, the catharsis of drama, has no connection with education or intellectual entertainment which, as mentioned above, Janko wrongly translates as "educative entertainment". The educated man does not need further education and Sxayutyfi is the leisure time he spends in meaningful, dignified and pleasurable activities. It is undeniable, although Janko and others tend to forget it, that Aristotle talks about the education of young men inPolitics 8. He naturally mentions adults and how they relate to music but all his pronouncements about the formative influence of art pertain to the young. He says for instance that music affects everybody and causes an emotional change along the lines "virile tunes suggest bravery" or "sorrowful songs induce melancholy". But he does not draw the conclusion that all people should listen to exciting or patriotic songs in order to bolster their fighting prowess or civic 183 virtue. Instead, he draws the much more natural conclusion that one should educate young men in music in order to make them virtuous and teach them a skill that can

be useful both for their relaxation and later for their intellectual entertainment as

adults.

Janko considers Pol. 1340al4-24 one of the important passages that support

his case for continuous adult education at the theater. Aristotle says there that music

can arouse emotions very easily: it affects everybody and exposure to it can

contribute to ethical education. For instance, if a picture causes pleasure only by virtue of the shape of the thing depicted, one will necessarily experience a similar pleasure when confronted with the real thing. The passage does not imply that

everybody is habituated to experiencing the emotions imitated in music every time

one listens to it. It should be remembered that Aristotle elaborates on the ethical power on music in order to prove that it is suitable for the education of theyoung.

Music affects everybody and thus it can have a great impact on the moral formation of malleable young people who by nature are inclined to pleasurable things: to absorb moral values they should be exposed to "good" music. It is obvious that

Aristotle distinguishes sharply between the uses and effects that music and all arts can have on different age categories: comparing visual arts and music he stresses that the imitative power of the latter is much greater but he says that in the case of the former also the young should only be exposed to the works of "ethical" painters and sculptors. If art had an educative influence on all people, why would not or should not adults as well limit themselves to "ethical" artworks? At the end of Politics 7 184 (1336a22-b35) Aristotle stresses emphatically that young children should be shielded from any potentially corrupting influence, like socializing with slaves, listening to and seeing indecencies: they should refrain from all such activities until adolescence, when they can participate in banquets and drink, when education will have made them immune to moral harm. It is obvious that naiSeia has been completed at that point and adults no longer need it, although, Aristotle specifies, it would be best if no one engaged in compromising pursuits122.

If one were to view tragedy in the light of Politics 8, one could only classify tragedy as intellectual entertainment. An educated man would attend dramatic performances the way he would listen to music, to spend his time in a congenial way.

But of course tragedy has a cathartic effect on the audience and thus, always in terms of the discussion in the Politics, it does not qualify as intellectual entertainment.

Lear123 has attacked the theory of the clarification or education of emotions on the same grounds124. Depew has objected to Lear’s view:

But Lear, like many of those whose interpretation he is contesting, fails to differentiate cathartic "clarification" aimed at training feeling responses and inducing habituation ( ethismos ) -which is indeed "educational" in Aristotle’s sense- from the sort of learning by cathartic clarification that awakens and exercises cognitive abilities -practical judgement and contemplative learning. The latter is not a matter of

122Cf. Depew (1991) 372-374 who argues along similar lines that adult education is a very problematic notion that cannot be read in the Politics or anywhere else in Aristotle’s work.

123Lear (1992) 318-321.

124Curiously, although the articles of Janko and Lear appear back to back in the same volume, they do not discuss each other. 185 education (paideia)12S.

Depew does not elaborate on tragic catharsis but he seems to espouse a view on

which catharsis and learning are somehow fused: at the end of a tragic performance

the citizens have sharpened their critical judgement and acquire a heightened or

contemplative view of human affairs. By actively following the vicissitude of the

tragic character the citizens widen the horizons of their practical wisdom and can

more easily ascend to contemplation:

These conenctions do not, however, run directly through the concept of catharsis. Instead, they run through the concept of learning mathesis( ). That is because tragic catharsis, we learn in the Poetics (4.1448bl5-16), rests on the learning involved in actively following a plot to its resolution, with the result that the experience of the tragic character, and at the same time the excellence of its representation (mimesis), is worked through, clarified and properly judged. This can be done only by those who can follow the development and resolution of the plot, and the moral deserts of its characters, with their minds no less than their (well educated) feelings. Thus tragedy affords a kind of learning which, although it is not education (paideia) -in fact it presupposes it- does have normative import. It sharpens and excercises practical judgement -and at the same time opens out onto a wider, more contemplative understanding of human affairs. That is just what is required of the endlike leisure processes of Pol. VIII. I conclude, therefore, that there is an important place for tragedy in Aristotle’s ideal state as part of its leisured life. The link between practical and contemplative wisdom just mentioned can be envisioned as follows. The activity of judging poetic representations involves, Aristotle says, an implicit, reflective grasp of the universals, both factual and normative, which govern excellent human activity (Poet. 9.1451b5-ll). This suggests that the technical judgement and practical wisdom called into play by musically attended forms like tragedy also have a certain relevance to contemplative wisdom. Art, conversation and other dimensions of endlike leisure processes...undoubtedly begin with, and remain intensely concerned with, norms of human actions, but in learning about the human condition by judging the correctness of representations of

12SDepew (1991) 369 n.20. 186 it, practical wisdom itself learns a lesson vividly taught by the art we judge to be best126.

There can be no denying that "learning" or intellectual/cognitive pleasures play a role, even an important one, in the Aristotelian conception of tragedy and other arts.

But I think it is mistaken to associate catharsis with this process by glossing it as

"clarification" of the character’s fate and the plot itself. On this view things are obscure all the way to the end of a tragedy and require a detective plot on the part of the poet as well as a detective disposition on the part of the audience in order to achieve the desired end, an exercise in civic wisdom. This is impossible to support from the Poetics. Besides, Depew assumes a one-to-one correspondence between excellence of representation, the technical ability to judge it, and the quality of knowledge acquired from the experience. But I find it implausible that knowledgeable spectators attending Sophocles’ OT would have a better view of the human condition than if they attended Aechylus’Agamemnon.

Although this is not the place to address Depew’s argument in detail, I should say that he seems to lay too strong an emphasis on learning through leisure. As far as tragedy is concerned there is no indication that catharsis is connected with learning or clarification, a meaning not attested for, or implied in, the term. Depew’s remarks on learning through music, which he associates with tragedy, are even more problematic. He claims that musical forms

126Depew (1991) 369-370. 187 have a cognitive dimension that transcends mere entertainment, psychological catharsis, and character building by inducing reflection and learning. Aristotle says plainly that leisure processes...are for the sake of practical wisdom(phronesis) (VHI.5.1339a26) and learning (mathesis) (VIII.5.1339a37,6.1341a23). This is exactly what we would expect him to say. For unless music engages rationality, the distinctive human function {ergon), on both the producing and receiving end, it cannot be fully endlike according to Aristotle’s general principles. Thus Aristotle clearly insists that young citizens engage in music, not only for the sake of character formation, but so that they might develop ability to judge (kritein) [sic] musical performances correctly, especially those in which correct response is required to "respectable characters and noble actions" (VTII.5.1340al3-18)...Apparently the ability to judge music implies knowing it in a technical way; and this technical knowledge is crucial to the subsequent development of both practical and theoretical knowledge127.

First of all

Siayayfi and 0p6vrjai<;. In the two passages where n&QqoiQ is mentioned it involves not adults but young men: in the former passage it refers to learning to play an instrument and in the latter it is synonymous with7rai

In general, it does not follow automatically that it is possible to illuminate the

127Depew (1991) 368. 188 tragic catharsis with the evidence on musical catharsis. On the other hand,

Aristotle’s answer to Plato’s attack on poetry was not that tragedy makes better citizens but that it is a worthwhile pursuit which offers a peculiar kind of pleasure appealing to both the emotions and the intellect of the audience without inducing any harm.

From the above it has emerged clearly that the differences between Aristotle and the two major Hellenistic philosophical schools are cosiderable. To an extent, the discrepancy is understandable since both the Aristotelian and the Hellenistic literary theories developed as an offshoot of larger philosophical systems which were ex hypothesi fundamentally different. It is characteristic for the limited influence of the Poetics that, as far as can be gleaned from the extant evidence, even the

Peripatetics seem to have diverged from Aristotle. Again it is impossible to observe or postulate a total split if only because it is hard to find any area of study that

Aristotle left completely out of his horizon. But it is undeniable that second and third generation Peripatetics focused significantly more on biographical inquiries and even anecdotological material than Aristotle ever did128. They wrote extensively on poetry, poets and literary criticism but they seem to have favored an approach that can be called if not un-Aristotelian at least largely foreign to the Poetics. Among the

Peripatetics the scholar Neoptolemus of Parium, is the most interesting, if very

128The only comprehensive account of Peripatetic scholarly contributions is still Podlecki (1969). difficult to assess, case129. The first and major stumbling block is Neoptolemus’ philosophical affiliation: although often called a Peripatetic he is not securely

identified as such. At any rate, from Porphyrion comes the tantalizing piece of information that Horace in his Ars Poetica adapted the most salient theses

(eminentissima ) of Neoptolemus’ literary theories130. The only extant account of these theories is found in Philodemus’ On Poems (5.13.32ff.) and it is unknown whether it covers Porphyrion’s "salient" points or not. Philodemus had apparently presented Neoptolemus’ theories in some detail earlier in the treatise and the extant text is as usual obscure because it consists in little more than the author’s expressions of indignation at views he considers unfounded and almost lunatic. Neoptolemus is reported to have divided the art of poetry into three parts or kinds(eiSrj), poem, poetry and poet131. The first two are easily understandable and indeed reminiscent of Posidonius’ definition, although Philodemus found fault with them too. The third is the most puzzling and elicited Philodemus’ heaviest attack: how can the poet, the agent, be considered part of the art he possesses? The question has still not been satisfactorily answered. At any rate, whether a fuller account of Neoptolemus’ theory would provide the sorely needed missing links or he simply used a loose

129On Neoptolemus’ theories see Mangoni (1993) 53-61 and 221ff. with a survey of previous bibliography; cf. Fuhrmann (1992) 145-153.

130PorphyrionAd Hor. A.P. p. 162.6 Holder.

131For different conjectures about Neoptolemus’ eiSrj see Ardizzoni (1953) 9-30, Brink (1963) 58, Greenberg (1990) 276-284, Walsh (1987) 59ff. and Asmis (1992c) 208-209, 219-220, 230-231. 190

terminological apparatus, the theory seems to be unrelated to thePoetics. Attempts

to associate Neoptolemus’ views with Aristotle’s Rhetoric are not particularly

convincing and have not met with much critical approval. If Neoptolemus were a

Peripatetic, bis difference from Aristotle would gain additional weight, especially

since he seems to have been an advocate of the didactic use of poetry {De Poem.

5.16.22-28).

Already Theophrastus apparently moved away from Aristotle. The question

of his debt to Aristotle will never be settled with any degree of certainty given the

paucity of his surviving work. It is particularly unfortunate that only one very brief

definition of tragedy attributed to him is found in a veiy late source, the grammarian

Diomedes (1.487.11-12). Two other definitions, of comedy (1.4-5) and epic (1.484.1-

2), as well as a dubious one of mime (1.491.15-16), are transmitted anonymously but

there is strong possibility that they also originated in Theophrastus’ work. Diogenes

Laertius (5.47-48) testifies that he had written a Poetics as well as a Rhetoric, now

lost. Only fragments of his On Style survive, appropriately perhaps since he was

considered a very gifted stylist by later critics132. The primary contributions

attributed to him concern the characters of style, i.e. its division into grand, middle

and plain133, and the virtues of style, correctness or "Hellenism", clarity,

132See Cic. Acad. 1.33, Brut. 121, Or at. 62 and especially D.L. 5.38 who reports the outrageous anecdote that Aristotle changed his student’s original name Tyrtamus to Theophrastus because of the man’s divine diction (to rrjc

133It has been debated whether Theophrastus actually proposed such a theory or he basically espoused the Aristotelian doctrine of stylistic mean; see Kennedy (1963) 191 appropriateness and ornamentation (tkXrivxaiioQ, cra^rjveia, t o nplnov,

KaraoKevf ))134. (Later the Stoics added brevity to this canon135). He exerted substantial influence on later rhetorical and stylistic theories, although the same holds for other Peripatetics too136. As I have already indicated, opinions about

Theophrastus’ indebtedness to Aristotle have varied. The modern consensus leans towards a middle path of sorts: Theophrastus does not disagree with Aristotle and he essentially expanded on Aristotle’s theses with many original insights of his own like his emphasis on audience-oriented vs. fact-oriented speech137.

If decisions about Theophrastus’ theory of style, fragments of which have survived, prove difficult and tentative, it is immediately apparent that Theophrastus’ poetics is beyond reasonable elucidation. As mentioned above, the only surviving evidence, the definitions of genres, are so brief that they are not particularly helpful.

The first one runs "tragedy is a nepxaraaxQ of heroic fortune". Dosi138, who,

278, Schenkeveld (1964) 66ff., Innes (1985) and Chiron (1993) lii-lv who suggests that, although Theophrastus advocated only one good style, later authors like Demetrius may have detected in his work a version of the later standard tripartition (Demetrius himself does not know this theory: see Chiron (1993) xxx-xxxi).

134Cic. Or at. 79. Again, Aristotle recognized only one virtue and Theophrastus broke it down into four. During (1966) 180 considered this distinction as merely formal. Others saw a more substantial contribution on Theophrastus’ part; see Pohlenz (1933) 107. See also Milanese (1989) 21-23 with an overview of previous bibliography.

,35See D.L. 7.59 and Innes (1985) 256.

136See Stroux (1912) 54ff and Innes (1985) 251.

137See e.g. Grube (1952) 178 and Lossau (1964) 47ff.

138Dosi (1960). following Rostagni, made the most ambitious attempt to reconstruct Theophrastean poetics, translates jrepiaraaic as "vicenda" (event) and specifies that in tragedy’s

case this has to be a catastrophic event. Comedy is said to be "the safe rrepioxn of private affairs" and epic is the Trepioxn of "divine, heroic and human affairs". Even

if irepiaraaic is not the rhetorical terminus technicus that Dosi proposes and simply means "change" in the sense akin to Aristotle’s nepiwiTeia, the difference from the

Poetics is striking. On the evidence of these definitions as well as what little can be found in other later sources Dosi reconstructs a system that, she claims, exerted a vast influence on all later criticism, from Hellenistic through Byzantine times.

Although she inevitably attributes great significance to every scrap of available evidence and makes her case with great confidence, her argument is not implausible, even if ultimately improvable. Her basic thesis is that Theophrastus moved away from the Aristotelian position that emphasized an intellectual approach to poetry and concentrated on drama which could be studied in terms of probability and necessity.

The insistence on the construction of a plausible mythos entailed the neglect of all other elements. Theophrastus set out to redress what he purportedly saw as an imbalance139. He proclaimed that poetry does not necessarily have to reproduce reality or be plausible: aiming at pleasure it could include, besides real events

(iaropta), fictitious but plausible accounts (7 Tk&oiia) and totally fantastic elements

(fivdos). While history and philosophy deal with facts, rhetoric and poetiy are fictions which charm and persuade the audience. Since the connection with

139See also the discussion of Koster (1970) 85-92. 193 probability and factuality was severed, Theophrastus necessarily had to proclaim diction the determining factor in the creation and appreciation of poetry. Words and eloquence are poetry’s only means of pleasing and persuading the audience and should be chosen and worked out carefully.

If this reconstruction is correct, it suggests that Theophrastus can be viewed more as the precursor of Hellenistic critics than the disciple of Aristotle, although of course his views must have been formulated as a reaction to Aristotle’s. With the emphasis on diction he foreshadows later developments. Groping for answers to the problem of poetry’s uniqueness and its position in a larger philosophical and moral framework, ancient critics were led to views that proclaimed diction, or some related formal element like euphony, the quientessence of poetry. More specifically,

Posidonius’ definition of poetry as "imitation of things divine and human" shows similarity with Theophrastus’ definitions but that may be accidental. Dosi also suggests that Neoptolemus’ division of poetry, poem and poet could go back to

Theophrastus but that is even more speculative. What seems reasonably beyond doubt is that already Theophrastus most probably followed a distinctly un-

Aristotelian path in his poetics. This fact poignantly underlines the isolation of

Aristotle’s views on poetry. BIBLIOGRAPHY

Adam, L. 1889. Die aristotelische Theorie vom Epos und ihre Entwicklung bei Griechen und Rom em . Wiesbaden.

Africa, T.W. 1961. Phylarchus and the Spartan Revolution. Berkeley.

Angeli, A. 1988. Filodemo, Agli Am id di Scuola (PHerc. 1005). Naples.

Annas, J. 1976. Aristotle's Metaphysics M and N. Oxford.

Ardizzoni, A. 1953. H-otry/ia. Ricerche sulla Teoria del Linguaggio poetico nelV Antichitd. Bari.

Asmis, E. 1984. Epicurus' Scientific Method. Ithaca.

1990. "The Poetic Theory of the Stoic "Aristo"",Apeiron 23, 147-201.

1991. "Philodemus’ Poetic Theory andThe Good King According to HomeF, CA 10, 1-45.

1992. "Crates on Poetic Criticism", Phoenix 46, 138-169.

1992b. "An Epicurean Survey of Poetic Theories (Philodemus,On Poems 5, cols. 26-36)", CQ 42, 395-415.

1992c. "Neoptolemus and the Classification of Poetry",CP 87, 206-231.

1993. "Epicurean Poetics",BACAP 7, 63-93.

Aubonnet, J. 1989. Aristote, Politique. Paris.

Auger, D. 1986. De I' artisan a I’ athlde: les metaphores de la creation poetique dans

194 195 V epinicie et chez Pindare. (Lalies, P.E.N.S.) Paris.

Ax, W. 1986. Lout, Stimme und Sprache. (Hypomnemata 84) Gottingen.

Barnes, J. 19942. Aristotle, Posterior Analytics. Oxford.

Bartels, K. 1965. "Der Begriff Techne bei Aristoteles" in Synusia: Festgabe fur Wolfgang Schadewaldt. Pfullingen, 275-286.

Barwick, K. 1922. Remmius Palaemon und die romische Ars Grammatica. (Philologus Suppl. 15,2) Leipzig.

Belfiore, E. 1992. Tragic Pleasures. Princeton.

Berger, H. 1880. Die geographischen Fragmente des Eratosthenes. Leipzig.

1903. Geschichte der wissenschaftlichen Erdkunde der Griechen. Leipzig.

Bernardini, A. P. 1977. "L’ "Aquila tebana" vola ancora", QUCC 26, 121-126.

Bignone, E. 1936. L ’Aristoteleperduto e la Formazione filosofica diEpicuro. Florence

Birmelin, E. 1933. "Die kunsttheoretischen Gedanken in Philostrats Apollonios", Philologus 88, 149-180 & 392-414.

Blank, D. 1982. Ancient Philosophy and Grammar: The Syntax of Apollonius Dyskolus. Chico.

Blum, R. 1977. Kallimachos und die Literaturverzeichnung bei den Griechen. Darmstadt.

Bonhoffer, A. 1890. Epiktet und die Stoa. Stuttgart.

Brink, C. O. 1946. "Callimachus and Aristotle: An inquiry into Callimachus’ HPOz IIPASISANHN", CQ 40, 11-26.

1960. 'Tragic History and Aristotle’s School",PCPS 6, 14-19.

1963. Horace on Poetry I. Cambridge.

1971. Horace on Poetry II. Cambridge.

Browning, R. 1992. "The Byzantines and Homer" in Lamberton & Keaney, 134-148. 196 Buchheit, V. 1984. "Lukrez iiber den Ursprung von Musik und Dicthung", R hM 127, 141-158.

Buhler, W. 1964. Beitrdge zur Erklarung der Schrift vom Erhabenen. Gottingen.

Buffi&re, F. 1956. Les Mythes d ’ Homere et la Pensee grecque. Paris.

Bundy, E.L. 1972. "The Quarrel between Apollonios and Kallimachos", CSCA 5, 39- 94.

Burnett, A.P. 1971. Catastrophe Survived. Oxford.

Butcher, S.H. 19074. Aristotle’s Theory of Poetry and Fine Art. London.

By water, I. 1909.Aristotle On the Art of Poetry. Oxford.

Chiron, P. 1993. Demetrios, Du Style. Paris.

Clay, D. 1983. Lucretius and Epicurus. Ithaca.

Cornford, F. M. 1952. Principium Sapientiae. Cambridge.

Crotty, K. 1982. Song and Action. Baltimore.

Currie, M. M. 1961. "Epicureanw p6Xr)tyiQ", Phronesis 6, 70-85.

Dawson, D. 1992. Allegorical Readers and Cultural Revision in Ancient Alexandria. Berkeley.

De Lacy, P. H. 1948. "Stoic Views of Poetry",AJP 69, 241-271.

Delatte, A. 1934. Les Conceptions de T Enthousiasme chez les Philosophes presocratiques. Paris.

Depew, D. J. 1991. "Politics, Music and Contemplation in Aristotle’s Ideal State", in Keyt, D. & Miller, F. (eds.),A Companion to Aristotle’s Politics. Oxford, 346- 380.

Dodds, E. 1951. The Greeks and the Irrational. Berkeley.

Dorandi, T. 1982. Filodemo, II Buon Re Secondo Omero. Naples.

Dosi, A, 1961. "Sulle Tracce della Poetica di Teofrasto", Rend. 1st. Lombardo 94,599- 672. 197 Dover, K. 1993. Aristophanes, The Frogs. Cambridge.

During, I. 1966. Aristoteles. Heidelberg.

Durante, M. 1968. "Untersuchungen zur Vorgeschichte der griechischen Dichtersprache. Die Terminologie fiir das dichterische Schaffen", in Schmitt, R. (ed.), Indogermanische Dichtersprache. Darmstadt, 261-290.

Eden, K. 1986. Poetic and Legal Fiction in the Aristotelian Tradition. Princeton.

Else, G. F. 1957. Aristotle’s Poetics: The Argument. Harvard.

Erbse, H. 1984. Studien zum Prolog der euripideischen Tragodie. Berlin.

Farnell, 1909. The Cults of the Greek States, V. Oxford.

Ferrero, L. 1963. "Tra Poetica ed Istorica: Duride di Samo", in Miscellanea di Studi alessandrini in Memoria di A. Rostagni. Turin, 68-100.

Finsler, G. 1900. Platon und die aristotelische Poetik. Leipzig.

Flashar, H. 1956. Der Dialog Ion als Zeugnis platonischer Philosophie. Berlin.

Fomara, C. 1983. The Nature of History in Ancient Greece and Rome. Berkeley.

Fraenkel, E. 1950. Aeschylus, Agamemnmon. Cambridge.

Frank, E. 1923. Platon und die sogennanten Pythagoreer. Halle.

Frede, D. 1992. "Necessity, Chance, and "What Happens for the Most Part" in Aristotle’s Poetics" in Rorty, 197-220.

Freudenburg, K. 1993. The Walking Muse. Princeton.

Fuhrmann, M. 1992. Einfuhrung in die antike Dichtungstheorie. Darmstadt.

Frye, N. 1957. Anatomy of Criticism. Princeton.

Furley, D. J. 1967. Two Studies in the Greek Atomists. Princeton.

Fusillo, M. 1985. II Tempo delle Argonautiche. Rome.

Gallop, D. 1990. "Animals in the Poetics ", in Annas, J. (ed.), Oxford Studies in Ancient Philosophy 8, 145-171. 198 Gentili, B. 1965. "Aspetti del Rapporto Poeta, Committente, Uditorio nella Lirica corale greca", Studi Urbinati 39, 70-88.

Giancotti, F. 1972. "Postille sui Rapporti fra Epicureismo e Poesia in Epicuro e in Lucrezio", GIF 3, 192-233.

Gigante, M. 1961. "Erj/navrucdv woirjfia: Contributo alia Storia dell’ Estetica antica PP 16, 40-53.

1989. "Filodemo tra poesia e prosa",SIFC III 7, 129-151.

1990. Filodemo in Italia. Florence.

1992. "Gli Epigrammi di Filodemo quali Testimonianze autobiografiche. Filodemo nella Villa di Ercolano: Incontri al belvedere (Anth. Pal. IX 412 e XI 35)", in Studi di Filologia Classica in Onore di G. Monaco I. Palermo, 421- 427.

Gill, M.L. 1989. Aristotle On Substance. Princeton.

Golden, L. 1976. "Epic, Tragedy and Catharsis", CP 71, 77-85.

Gomoll, H. 1937. "Herakleodoros und die KRITIKOI bei Philodem", Philologus 91, 373-384.

Gray, V. 1987. "Mimesis in Greek Historical Theory", AJP 108, 467-486.

Greenberg, N. 1961. "The Use of poiema and poiesis", HSCP 65, 263-289.

1990. The Poetic Theory of Philodemus. London/N. York.

Griesinger, F. 1907. Die aesthetischen Anschauungen der alten Homererklarer, dargestellt nach den Homerscholien. Diss. Tubingen.

Grube, G.M.A. 1952. "Theophrastus as a Literary Critic",TAPA 83, 172-183.

1965. The Greek and Roman Critics. London.

Gundert, H. 1935. Pindar und sein Dichterberuf. Frankfurt.

Haussler, R 1970. "Hoirj/ia und Bol^aic" in Festschrift K Buchner. Wiesbaden, 125-150.

Hahm, D.E. 1977. The Origins of Stoic Cosmology. Ohio. 199 1991. "Aristotle and the Stoics: A Methodological Crux", AGPh 73, 297-311.

Halliwell, S. 1986. Aristotle’s Poetics. London.

1992. "Epilogue: ThePoetics and its Interpreters", in Rorty, 409-424.

Heath, M. 1989. Unity in Greek Poetics. Oxford.

Heath, T.L. 1949. Mathematics in Aristotle. Oxford.

Hintikka, J. & Remes U. 1974. The Method of Analysis. Dordrecht.

Hirzel, R. 1877. Untersuchungen zu Ciceros philosophischen Schriften. Leipig.

House, H. 1956. Aristotle’s Poetics. London.

Hubbard, T. 1985. The Pindaric Mind. (Mnemosyne Suppl. 85) leiden.

Hunter, R.L. 1993. The Argonautica of Apollonius. Cambridge.

Hutchinson, G. 1988. Hellenistic Poetry. Oxford.

Imbert, C. 1980. "Stoic Logic and Alexandrian Poetics", in Schofield, M. et a l (eds.), Doubt and Dogmatism. Oxford, 182-216.

Innes, D. 1985. "Theophrastus and the Theory of Style",Rutgers University Studies 2, 251-267.

1989. "Philodemus", in Kennedy (ed.), 215-219.

loppolo, A.M.Aristone di Chio e lo Stoicismo antico. Naples.

Isnardi Parente, M. 1966. Techne. Momenti del Pensiero greco da Platone ad Epicuro. Florence.

1987. "Una Poetica di incerto Autore in Filodemo", in Filologia e Forme Letterarie. Studi offerti a F. della Corte, V. (Urbino), 81-96.

1989. Gli Stoici antichi. Turin.

Janko, R. 1991. "Philodemus’ On Poems and Aristotle’s On Poets", CronErc 21, 1-60.

1992. "From Catharsis to the Aristotelian Mean", in Rorty, 341-358. 200 Jensen, C. 1923. Philodemos, Uber die Gedichte, Fiinftes Buck. Berlin.

Judson, L. (1991). "Chance and "Always or for the Most Part" in Aristotle" in Judson, L. (ed.), Aristotle’s Physics. Oxford, 132-161.

Kaiser, E. 1964. "Odyssee-Szenen als Topoi", MH 21, 109-136 & 197-224.

Kambylis, A. 1965. Die Dichterweihe und ihre Symbolik. Heidelberg.

Kebric, R. 1977. In the Shadow o f Macedon: Duris of Samos. (Historia Einzelschr. 29) Wiesbaden.

Kennedy, G. 1963. The Art o f Persuasion in Greece. Princeton.

(ed.). 1989. The Cambridge History of Literary Criticism I. Cambridge.

1989. "Hellenistic literary and Philosophical Sholarship", in Kennedy (ed.), 200- 214.

Kenney, E.J. 1977. Lucretius. Oxford.

Kidd, I.G. 1988. Posidonius II, The Commentary. Cambridge.

Kirkwood, G.M. 1981. "Pindar’s Ravens: Olymp. 2.87", CQ 240-243.

Klein, T. 1975. "Callimachus, Apollonius Rhodius and the Concept of the "Big Book"", Eranos 73, 16-25.

Kleve, K. 1963. Gnosis Theon. Die Lehre von der natiirlichen Gotteserkenntnis in der epikureischen Theologie. (SO Suppl. 19) Oslo.

Koster, S. 1970. Antike Epostheorien. (Palingenesia 7) Wiesbaden.

Kovacs, P.D. 1980. The Andromache of Euripides: An Interpretation. Chico.

Kullmann, W. 1980. "Der Mensch als politisches Lebewesen bei Aristoteles", Hermes 108, 419-443.

Kurke, L. 1991. The Traffic in Praise. Ithaca.

Lamberton, R. & Keaney, J.J. (eds.) 1992. Homer’s Ancient Readers. Princeton.

Lear, J. 1992. "Katharsis", in Rorty, 315-340. 201 Lefkowitz, M. 1980. The Lives of the Greek Poets. Baltimore.

1984. "The Poet as Athlete", SIFC 77, 18-24.

Lehnert, G. 1896. De Scholiis ad Homerum rhetoricis. Diss. Leipzig.

Lesky, A. 19642. Die tragische Dichtung der Hellenen. Gottingen.

Long, A.A. 1971. "Aisthesis, Prolepsis and Linguistic Theory in Epicurus",BICS 18, 114-133.

1992. "Stoic Readings of Homer", in Lamberton & Keaney, 41-66.

Lord, C. 1982. Education and Culture in the Political Thought of Aristotle. N. York.

Lossau, M. 1964. Untersuchungen zur antiken Demosthenesexegese. (Palingenesia 2) Wiesbaden.

Lucas, D.W. 1968. Aristotle, Poetics. Oxford.

Maehler. H. 1963. Die Aujfassung des Dichterberufs imfriihen Griechentum biszurZeit Pindars. (Hypomnemata 3) Gottingen.

Mangoni, C. 1993. Filodemo, II quinto Libro della Poetica. Naples.

Manuwald, A. 1972. Die Prolepsislehre Epikurs. Bonn.

Meijering, R. 1987. Literary and Rhetorical Theories in Greek Scholia Groningen.

Meister, K. 1975. Historische Kritik bei Polybios. Wiesbaden.

Melazzo, L. 1975. "La Teoria del Segno linguistico negli Stoici", Lingua e Stile 5,199- 230.

Mette, H. 1952. Parateresis. Halle.

Mignucci, M. 1981. "’fie e7rl rd 7roA.fi et nficessaire dans la conception aristotfilicienne de la science" in Berti, E. (ed). 1981 .Aristotle on Science: The Posterior Analytics. Padua, 173-203.

Milanese, G. 1989. Lucida Carmina Milano.

Morpurgo-Tagliabue, G. 1967. Linguistica e Stilistica di Aristotele. Rome. 202 Most, G.M. 1986. "Pindar O. 2.83-90", CO 36, 304-316.

Mueller, C.W. 1987. Erysichtoru Der Mythos als narrative Metapher im Demeterhymnos des Kallimachos. (AAWM 13, Geistes- und sozialwissenschaftliche Klasse) Stuttgart.

Murray, P. 1981. "Poetic Inspiration in Early Greece",JHS 101, 87-100.

Nehamas, A. 1987. "Dangerous Pleasures", (Review Halliwell 1986), TLS 4371,27-28.

Noblot, H. 1979. Seneque. Lettres a Lucilius. Paris.

Nussbaum, M. 1986. The Fragility o f Goodness. Cambridge.

Olson, E. 1965. Aristotle’s Poetics and English Literature. Chicago.

Panofsky, E. 1968. Idea: A Concept in Art Theory. Transl. J. Peake. N. York.

Patzer, H. 1962. Die Anfange der griechischen Tragodie. Wiesbaden.

Pedech, P. 1970. Polybe, Histoires. Paris.

Petzold, K.-E. 1969. Studien zur Methode des Polybios und zu ihrer historishen Auswertung. Munich.

Pfeiffer, R. 1968. History of Classical Scholarship. Oxford.

Philippson, R. 1924. Review Jensen (1923), / W 4 4 , 417-421.

1930. "Das Sittlichschdne bei Panaetius", Philologus 85, 357-413.

Pickard-Cambridge, A. 1927. Dithyramb, Tragedy, and Comedy. Oxford.

Podlecki, A. 1969. "The Peripatetics as Literary Critics",Phoenix 23, 114-137.

Pohlenz, M. 1933. "Ein Beitrag zur Geschichte des griechischen Geistes", NGG 16, 53-92.

1939. "Die Begriindung der abendlandishen Sprachlehre durch die Stoa",NGG 22, 151-198.

19704. Die Stoa Gottingen.

Porter, J.I. 1989. "Philodemus on Material Difference", CronErc 19, 149-178. 203

1992. "Hermeneutic Lines and Circles: Aristarchus and Crates on the Exegesis of Homer", in Lamberton and Keaney, 67-114.

Regenbogen, O. 1932. Lukrez. Seine Gestalt in seinem Gedicht. Leipzig/Berlin.

Rengakos, A. 1992. "Zur Biographie des Apollonios von Rhodos", WS 105, 39-67.

Richardson, N.J. 1980. "Literary Criticism in the Exegetical Scholia to the Iliad : A Sketch", CQ 30, 265-287.

1992. "Aristotle’s Reading of Homer and Its Background", in Lamberton and Keaney, 30-40.

Rispoli, G.M. 1987. "Eufonia e Poetica in Testi ercolanesi", in Bolognesi, G. & Pisani, V. (eds.), Linguistica e Filologia. Brescia, 461-478.

Roloff, D. 1970. Gottahnlichkeit, Vergdttlichung und Erhohung zu seligem Leben. Berlin.

Romeo, C. 1983. "Filodemo: La "Poetica"", in Studi sulV Epicureismo greco e Romano offerti a M. Gigante. Naples, 565-583.

Ronconi, A. 1963. "Appunti di Estetica epicurea", inMiscellanea di Studi alessandrini in Memoria di A. Rostagni. Turin, 7-25.

Rorty, A.O. (ed.). 1992.Aristotle’s Poetics. Princeton.

Rosenmeyer, T.G. 1974. "Design and Execution in Aristotle, Poetics ch. XXV", CSG4 6, 231-252.

Rostagni, A. 1955. Scritti Minori. Turin.

Rudberg, G. 1918. Forschungen zu Poseidonios. Upsala.

Russell, D.A. 1964. "Longinus", On the Sublime. Oxford.

1981. Criticism in Antiquity. London.

1989. "Greek Criticism of the Empire", in Kennedy, 297-329.

Sacks, K. 1981. Polybius on the Writing of History. Berkeley.

Salkever, S.G. 1986. 'Tragedy and the Education of the Demos: Aristotle’s Response 204 to Plato", in Euben, J.P. (ed.), Greek Tragedy and Political Theory. Berkeley.

Sandabch, F.H. 1985. Aristotle and the Stoics. Cambridge.

Sbordone, F. 1976. [QiAoSrfpov He pi UoiripdTOiv] Tractatus Tres. Naples.

Scaglione, A. 1972. The Classical Theory o f Composition from its Origins to the Present: A Historical Survey. Chapel Hill.

Scheller, P. 1911. De Hellenistica Historiae Conscribendae Arte. Diss. Leipzig.

Schenkeveld, D.M. 1964, Studies in Demetrius On Style. Amsterdam.

1968. "OX KRITIKOI in Philodemus", Mnemosyne 21, 176-215.

1976. "Strabo on Homer", Mnemosyne 29, 59-64.

Schmidt, J. 1990. Lukrez, der Kepos und die Stoiker. Wiesbaden.

Schmidt, M. 1976. Die Erklarungen zum Weltbild Homers und zur Kultur der Heroenzeit in den bT-Scholien zur Ilias. (Zetemata 62) Munich.

Schmitt, R. 1967. Dichtung und Dichtersprache in indogermanisher Zeit. Wiesbaden.

Schrader, H. 1902. "Telephos der Pergamener", Hermes 37, 530-581.

Schrijvers, P.H. 1970. Horror ac Divina Voluptas. Amsterdam.

Scolnikov, S. 1988. Plato’s Metaphysics of Education. Lxmdon.

Simpson, M. 1969. "The Chariot and the Bow as Metaphors for Poetry in Pindar’s Odes", TAPA 100, 437-473.

Snyder, J.M. 1973. "The Poetry of Philodemus the Epicurean",CJ 68, 346-353.

Steckel, H. 1968. "Epikuros", RE Suppl. XI, 579-672.

Steiner, D. 1986. The Crown of Song. lxmdon.

Steinmetz, P. 1986. "Allegorishe Deutung und allegorishe Dichtung in der alten Stoa", RhM 129, 18-30.

Stevens, P.T. 1971. Euripides. Andromache. Oxford. 205 Strasburger, H. 1966. Die Wesensbestimmung der Geschichte durch die antike Geschictsschreibung. Wiesbaden.

Striker, G. 1974. "Kpirtfpiov rfjc dArjOeiac", NGG 2, 47-110.

Stroux, J. 1912. De Theophrasti Virtutibus DicendL Leipzig.

Svenbro, J. 1976. La Parole et le Marbre. Lund.

Taillardat, J. 1965. Les Images d’ Aristophane. Paris.

Tescari, O. Lucretiana. Naples.

Theiler, W. 1982. Posidonius: die Fragmente. Berlin.

Ullman, B.L. 1942. "History and Tragedy",TAPA 73, 25-53.

Vahlen, J. 1914. Beitrage zu Aristoteles’ Poetik. Leipzig. van der Valk, M. 1963. Researches on the Text and Scholia of the Iliad. Leiden. van Rijen, J. 1989. Aspects of Aristotle’s Logic o f Modalities. Dordrecht.

Verdenius, W.J. 1989. "Pindar, O. 2.83-86", Mnemosyne 42, 79-82.

Vicaire, P. 1960. Platon: Critique litteraire. Paris. von Franz, M.L. 1940. Die aesthetischenAnschauungen der Ilias-Scholien. Diss. Zurich. von Fritz, K. 1958. "Die Bedeutung des Aristoteles fiir die Geschichtsschreibung" in Histoire et Historiens dans T Antiquite. Geneva, 85-145.

Wachsmuth, C. 1860. De Cratete Mallota. Leipzig.

Walbank, G. 1957. A Historical Commentary on Polybius I. Oxford.

1960. "History and Tragedy",Historia 9, 216-234.

1972. Polybius. Berkeley.

Walsh, G. 1986. The Varieties of Enchantment. Chapel Hill.

1987. "Philodemus on the Terminology of Neoptolemus",Menmosyne 40,56-68. 206 Waszinc, J.H. 1954. "Lucretius and Poetry’,MNAW 17, 243-257.

Wehrli, F. 1928. Zur Geschichte der allegorischen Deutung Homers im Altertum. Borna/Leipzig.

West, M.L. 1966. Hesiod, Theogony. Oxford.

White, D.R. 1984. A Sourcebook on the Catharsis Controversy. Diss. Florida State University.

Wilson, N.G. 1983. "Scoliasti e Commentatori", SCO 33, 83-112.

Zanker, G. 1981. "Enargeia in the Ancient Criticism of Poetry", R hM 124, 297-311.

Zegers, N. 1959. Wesen und Ursprung der tragischen Geschichtsschreibung. Diss. Cologne.