SOURCE: "Aristotle's Aesthetics 1: Art and Its Pleasure," in Aristotle's "Poetics," Duckworth, 1986, pp. 42-81.

[In the following essay, Halliwell examines Aristotle's conceptualization of "mimetic arts" and argues that the pleasure which results from experiencing a work of mimetic art is "a response to the intelligible structure imposed on his material by the artist's rational capacity."]

There is evidence to be found in the Poetics, and it receives some confirmation from material elsewhere in the corpus, that Aristotle's thoughts on poetry were not formed in isolation from comparative reflections on other related activities, especially the visual arts, music and dancing—activities which it is now automatic for us to describe collectively as art'. For Aristotle the most significant common factor shared by these activites, and their products, was mimesis, and it is directly in connection with mimesis that we encounter in the Poetics and occasionally in other works general pronouncements covering both poetry and one or more of the arts indicated above. Such pronouncements were not without precedent, and it is possible in particular to discern some influence on Aristotle of the analogy between poetry and painting frequently drawn and exploited by Plato. But Aristotle's use of the comparison was not merely conventional, and its repeated occurrence in the Poetics associates it with such vital matters as artistic form and unity, the relation between action and character in the portrayal of human life, and the nature of the pleasure to be derived from works of mimesis. It is for this reason that in the present and the following two chapters I shall be concerned with the broader foundations of the Aristotelian theory of poetry: that is, with concepts and principles which are presented in the treatise as the essential framework of an understanding of poetry, but whose scope it is clear that the philosopher regarded as encompassing the other activities too which I have mentioned.

In exploring the extent and the stability of these foundations, one must at once confront the question of whether Aristotle can legitimately be said to have possessed a unitary concept of art corresponding even approximately to the now prevalent use of this term. This basic issue elicited very different responses from the two most important English works on the Poetics written around the turn of the nineteenth century. The title of Butcher's influential study, Aristotle's Theory of Poetry and Fine Art, in itself boldly declares its author's position on the matter. Butcher was in fact contributing to a tradition already well established in Germany of attempts to reveal the existence of systematic aesthetic ideas in Aristotle, comparable to, and aligned with, his metaphysical, political and ethical systems. Butcher believed, with some minor qualifications, that 'the cardinal points of Aristotle's aesthetic theory can be seized with some certainty', and much of his book rather ambitiously undertakes to give substance to the claim. Bywater, however, in the preface to his edition of the Poetics, countered such views with the statement that 'the very idea of a Theory of Art is modern, and … our present use of this term 'Art' does not go further back than the age of Winckelmann and Goethe'. Bywater conceded that there are ideas in Aristotle's work 'which we should regard as coming under Aesthetics', but he objected to the aspiration to supply a systematic elaboration of them, when Aristotle himself had never done so.¹

While I do not accept that Aristotle's thought in this area is as fully structured or as accessibly close to our own as Butcher in his more confident flights tended to suppose, neither can I altogether share Bywater's negative approach. Not only is it virtually impossible in practice to dispense with 'art' and related terminology in discussing the Poetics and certain aspects of Aristotle's philosophy, but it also, and more positively, seems to me feasible and worthwhile to give to this term a faithfully Aristotelian significance which makes clear both where it coincided with, and where it diverged from, modern views of art. It will be the first task of this chapter to show that while Aristotle had a comprehensive concept (now lost) of techne, for which both 'craft' and 'art' in its older sense are rough but imperfect translations, he was also capable of demarcating within this larger notion a restricted group of activities for which the normal modern use of 'art' is the only simple equivalent, though 'mimetic art(s)' is perhaps a preferable description. In the later part of the chapter I shall go on to argue that it is also possible to identify an important Aristotelian theory of the pleasure properly entailed in the experience of these mimetic arts, but especially in poetry. I must emphasise at the outset that by referring to this and to other Aristotelian ideas as 'aesthetic' I do not mean to assume or appeal to any independent philosophy of art—least of all, any aestheticist philosophy. Except where otherwise indicated, 'aesthetic' is employed in a plain, unprejudicial sense, referring either to the properties or to the experience of those activities and works which Aristotle believed to be connected by the element of mimesis, and which most post-Enlightenment thinking takes, in its use of the term 'art', to be unified by expression or some other factor.² It must also be noted that since the first part of this chapter deals with the lineaments of the Aristotelian view of mimetic art, mimesis as such receives some attention here, but it is also reserved for the fuller treatment which it merits later in the book.

To elucidate Aristotle's conception of mimetic art, we must start with its foundation in the wider notion of technê. Technê had earlier become the standard Greek word both for a practical skill and for the systematic knowledge or experience which underlies it. The resulting range of application is extensive, covering at one end of the spectrum the activity of a carpenter, builder, smith, sculptor or similar manual craftsman, and at the other, at least from the fifth century onwards, the ability and practices of rhetoricians and sophists. It is therefore translatable in different contexts as 'craft', 'skill', 'technique', 'method', or 'art', and I mentioned in the introductory chapter the use of the term to denote a formal treatise or manual containing the exposition of the principles of an art. In the light of these linguistic facts, it is not surprising that in the classical period the word could be applied to the poet's ability, particularly if we take into account the close alignment between the range of meaning of technê and the usage of the term for productive activity, poiêsis, which had come to be employed specifically for poetry. … If technê and poiêsis had originally been restricted to skills directed to material results, they had at any rate developed so as to become capable of designating a much larger sphere of activity, and also so as to denote a characteristic type of disposition or procedure, or even a theoretical framework for procedure, as much as a particular kind of artefact. Technê implied method and consistency of practice; it represented a vital part of the ground of man's practical and inventive intelligence, as opposed to the forces of nature (including any uncontrollable elements in man's own nature). Hence the arguments which arose (so prominently in Plato) over whether certain activities could correctly be included in this category. Where poetry was concerned, technê could be set up, by Plato at least, as the antithesis of inspiration; but that is to anticipate a further issue which will receive consideration in my next chapter.

Aristotle accepts unequivocally that poetry, painting, sculpture, music and dancing are all forms of technê.³ But the mere fact that the term can be readily translated as 'art'—a fact which depends historically on the persistence in English, as a secondary sense, of the original meaning of this word as derived from ars, the Latin equivalent of technê—evidently does not yield an Aristotelian concept of art in the desiderated sense, since technê as 'art' embraces a much larger range of activity than the mimetic arts alone. The contrast between the broad category of technê and the significance possessed by the standard modern use of 'art' is striking, and one element in this contrast is the fact that the distinction between art and craft which has come to seem essential to some modern aestheticians, and to be a common if not a universal assumption in general attitudes to art, had no existence for Aristotle.⁴ The 'family resemblance' between the whole gamut of methodical skills and activities encompassed by technê was not weakened by the specific affinities which might be discerned between particular sets of technai. If mimesis is more in evidence, and more problematic, in the Poetics than technê, that is because the latter is actually more fundamental but also more easily assumed without explanation or explicit definition. It is therefore all the more important for the modern reader of the treatise to be sure about what presuppositions are carried by Aristotle's adherence to the belief that poetry and the other mimetic arts mentioned were technai.

In the case of the visual arts, with their necessary component of material craftsmanship, it hardly needs to be stressed that Aristotle's acceptance of the concept of technê was in conformity with widely held attitudes. But for the musico-poetic arts too (whose performing practitioners—actors and instrumentalists—are not in question here, though they too had their technai: Rhet. 1404a 23f.) the connection, at least on the superficial level, is sufficiently clear: the notion (though not the name) of poetic technê was at least as old as Homeric epic, it was reflected in the later adaptation of the language of 'making' (poiêsis) to poetry, and it was reinforced by the strongly genre-based conventions of Greek poetry, in which the poet's task was often defined partly in terms of the requirements of a social, religious or other context. Aristotle's attunement to such attitudes was, moreover, implicitly close to that strand of sophistic thinking which had asserted the cultural importance of verbal skills, technai, particularly rhetoric. The whole ethos of the Poetics is to some degree a reflection of the sophistic development of the formalised theory and exposition of linguistic skills. And if, as I suggested in the last chapter, the Poetics is not a teaching manual as such, it at any rate presupposes the possibility of poetic teaching grounded in the principles and procedures of the technê: according to Aristotle, metaphor is the only thing in poetry which one cannot learn from someone else (59a 6f.).

To understand what the status of poetry as a technê entailed for Aristotle, however, it is necessary to go beyond popular Greek ideas or the sophistic importance of formal teaching of verbal skills. For Aristotle elaborated the concept of technê at a deeper level within his own philosophical thinking, and it is here that we must look for the assumptions which underlie his treatment of individual arts. In the Ethics technê is defined as 'a productive capacity involving true reasoning' (1140a 8-10, 20f.). This may not seem to carry us far beyond what has already been said about the scope of the term in ordinary Greek, but it does laconically indicate how technê fits into a fundamental Aristotelian mould of thought. Technê involves a true alignment of the axis of potential/realisation in human productive activity: it is concerned with bringing into being, by intelligible and knowledgeable means, objects whose existence depends on their maker (1140a 10-14). An immediate point to notice—it will recur—is the necessary entailment of objective, rational standards in all technê. A further implication of the definition is that Aristotle, more markedly than in general Greek usage, shifts the emphasis of the concept from activity as such to the potential or ability for a certain sort of activity. Technê is a capacity to act in accordance with reasoned procedures so as to produce designed results: the reasoned regularity and stability of the activity means that it is possible to abstract the theory of the practice from it, and it is in this theory, which constitutes a form of knowledge, that a primary locus of the technê can be situated: the philosopher is, indeed, sometimes prepared to treat technê and knowledge as virtually synonymous.⁵

One reason for attending carefully to Aristotle's refinement of this notion is to dispel any suspicion that the idea of a poetic or other mimetic technê necessarily brings with it crude associations of physical craft. It is essential to observe that this is not so for Aristotle, and had probably not been so for some time before him. But there are further, and more positive, aspects to Aristotle's understanding of technê. Although in Book 6 of the Ethics from which the earlier definition was taken technê is distinguished from nature ('art' cannot produce thóse things which come into being naturally, 1140a 14f.), Aristotle is often at pains to assimilate them. In several works we find him stating the principle that technê stands in a relation of mimesis to nature, a principle which is usually but perhaps misleadingly rendered as 'art imitates nature'. What the phrase affirms is that art follows procedures analogous to nature's, and that similar patterns and relations can be discerned in the workings of each; but it is vital to insist on the proviso that the claim does not apply exclusively to the mimetic arts themselves (whose own mimetic character is a separate matter) but to the sphere of technê as a whole. Aristotle's position is clarified by his observation that technê may sometimes complete the work of nature, or supply its deficiencies, and by a number of other connections and comparisons made between the two forces.⁶ From all the material bearing on this point it emerges that what technê and nature have in common is teleology: both, as Aristotle would say, control processes for bringing things into being, and both are guided, and in one sense determined, by the ends or purposes towards whose fulfilment they move.⁷ It is presupposed in this theory that technê and nature both have a similar tendency to aim at the best, to effect the finest or most successful organisation of their material.

The most problematic feature of this philosophical thesis might well be thought to be what it predicates of nature, not its treatment of human technê, since the purposive procedures of the latter make at least a limited teleological definition of them readily intelligible. But even if we put on one side, as we here must, the general difficulties raised by Aristotle's natural teleology, the two spheres cannot be entirely disengaged, and there remains more to be said about the relation between nature and human productive activity. To the analogy posited between the two types of process as teleological, we need to add the conviction that, from one point of view at least, technê is itself a part of nature, and its ends are to be regarded as naturally given or determined. Because man is part of nature, and all technê involves his rational productive capacities, the teleology of 'art' is in some degree subordinate to and dependent on that of nature as a whole. The operation of this premise can be confirmed and exemplified specifically for poetry from the Poetics itself, and while I shall attempt a separate analysis of its implications for Aristotle's attitude to cultural tradition in my next chapter, it is apposite here to locate it in its broader context.

For my present purpose I emphasise only one passage, though there are others which are germane. In Poetics 4 Aristotle provides a brief account of the 'natural' causes of poetry, and in so doing advances a compound explanation of the reasons both for the production of poetry and for the pleasure taken in it. A preliminary point to be made about this passage is that, while it begins by reference to poetry, its perspective opens out, as sometimes happens elsewhere in the treatise, to include other mimetic arts, and we can therefore legitimately take Aristotle's arguments here to have a wider applicability to most forms of mimesis. The argument is, in fact, despite its characteristically plain and condensed presentation, an ambitious one, and it is crucial to the interpretation of more than one issue in Aristotle's aesthetic philosophy. Its significance for the present question lies not in the positing of natural human instincts for mimesis, but in the way in which Aristotle proceeds from this premise to sketch a theoretical view of the natural development or evolution of the major Greek literary genres. Nature enters into the sequence of thought at three stages, supplying a progressive series of reference-points: first (48b 5ff.) as the general provision of a human instinct and capacity for mimetic activity, encompassing musical and rhythmic instincts which are themselves mimetically charged (cf. n. 29 below); secondly (48b 22ff.) as the source of particular inclinations and talents shown by early poets, and hence as the motive force behind the first period of generic evolution (marked by the basic distinction between serious and base subjects); finally (49a 2-15), and most importantly, as the true process contained in the direction of cultural development, a process embodied in the gradual unfolding of the natural potential of specific poetic forms. On this last point Aristotle is most explicit in the case of tragedy (though we can fairly extrapolate for other genres), of which he says that 'it ceased to change, once it had acquired its own nature' (49a 15)—that is, its mature perfection, the fulfilment of its potential.

It is clearly implied in this last remark that the history of tragedy has to be comprehended ultimately in terms not of contingent human choices and tradition, but of natural teleology mediated through, or channelled into, acts of human discovery of what was there to be found. Aristotle's point need not be strictly deterministic, since he does not appear to believe that tragedy was bound to be invented and developed; but he does clearly affirm that once its development became a cultural possibility (once it had 'appeared' or 'been glimpsed', 49a 2) the end result was a naturally fixed goal. To return to the implications of this example for technê in general, we can now see that not only do human productive activities form an analogue to the generative and purposive patterns of nature, but their individual histories evolve in accordance with intrinsic seeds of natural potential. Although Aristotle seems nowhere to work the point out fully, it is corroborated by his belief that all technai have been repeatedly discovered and developed (and then lost) in the history of the world: this is because they are rooted in man's inescapable relation to nature.⁸ Aristotle's concept of technê, it transpires, is a highly charged philosophical doctrine, and must therefore be differentiated both from ordinary non-philosophical Greek notions of skill, method and practical intelligence, and also from the standard modern understanding of craft. While the Aristotelian concept shares with these applicability to a wide range of activities in which rational, controlled procedures represent a structured relationship between means and ends, purposes and products, it goes beyond them in its theoretical entailments concerning the natural foundations of these activities.

Technê represents the first layer or level in Aristotle's concept of the mimetic arts; these arts count as arts, in the first place, precisely by virtue of belonging to the category of rational, productive procedures. It is important, however, to avoid the common distortion involved in reducing this doctrine to the belief that 'all arts are crafts', since the refined notion of technê includes, but is not simply equivalent to, craft. Nonetheless, we must not attempt to minimise the radical discrepancy between this aspect of Aristotle's thinking and the dominant modern view of art. It would be similarly wrong to try to reduce the discrepancy by placing weight on the passage in the Metaphysics where Aristotle distinguishes between arts of utilitarian value and those designed to give pleasure.⁹ The distinction is, for one thing, broad, and not intended (in the case of the non-utilitarian, at any rate) to offer a careful delimitation of a particular set of activities; there is no reason to suppose that the arts of pleasure would be identical with, though they would evidently include, the mimetic arts. Nor is this passage of the Metaphysics designed to give a proper account of the types of pleasure derivable from technê, for this will reflect the nature of individual activities, and we shall see later in this chapter that the theory of aesthetic pleasure sketched in the Poetics cannot be understood simply as the antithesis of practical usefulness. Pleasure alone, therefore, will not make Aristotle's view of the mimetic technai conformable to modern attitudes. The primary reason for the remaining conceptual distance between the two is that Aristotle's acceptance of the framework of technê for the interpretation of poetry and related practices imports an inescapably 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 that 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 of his art, and by the objective principles which emerge from this development.

I shall return to this contrast at the end of this section, but it is now necessary to move on to the second and the distinctive criterion of the mimetic arts, mimesis itself. Mimesis alone, it must be stressed, is insufficient to yield a definition of the mimetic arts, since there are mimetic activities (mundane forms of imitation) which are not technai. The combination of both technê and mimesis is therefore required to give us the core of Aristotle's concept of poetry, painting, music and the rest—the concept of art, in other words, in the fuller sense. For without prejudicing the question of where and how this concept is consonant with, or divergent from, later ones, there need be no serious doubt that the substantial coincidence between the activities counted as mimetic technai by Aristotle and those usually counted as art in modern European culture (the Beaux Arts of the Enlightenment) provides adequate justification for referring in this way to an Aristotelian concept of art or mimetic art. The centrality of mimesis to this concept calls for the extensive treatment which I give to it in ch. IV, where many details in the interpretation of the Poetics' understanding of mimesis are examined. Here a more synoptic view of the subject must be offered.

If technê is a definition or theory of the relation between the productive artist and his product, mimesis stands for, and purports to characterise, the axis between the product and reality. That mimesis can be written as an ordinary English noun (as it has been since at least the seventeenth century) is a testimony to the persistence of mimeticist thinking and terminology in the European tradition; and the fact that the language of mimesis and imitatio was for so long indispensable to neo-classicism (and that it still lurks, thinly disguised, around much that is said and thought about art) may suggest that its potential belies some of the more severe rejections which it has met with in recent times. But it need hardly be said that the influence of neo-classicism has also created a specious familiarity which is an obstacle in the way of attempts to recover the original Greek context and development of the idea. Indeed, the very familiarity of the whole question and problem of the relation between art and the world makes it difficult to realise the good reasons for the faltering movement of Greek thought towards a philosophical formulation of the issues involved, or towards a concept of the shared or analogous features of poetry, the visual arts and music (to go no further). As with technê, so with mimesis, it is important to appreciate the basis of Aristotle's thinking on the matter in already existing attitudes and assumptions, as well as the ways in which he builds on and reshapes this basis. In the case of mimesis Aristotle worked against the background both of Plato's persistent approaches to the idea, and also of the more widely held mimeticist views which Plato himself attests.¹⁰ The notion of mimesis as a common or defining characteristic of a variety of cultural activities and products was not the discovery of a single thinker or period. It emerged untidily from a long development in the meaning and application of mimesis language, and may not have been wholly consciously articulated before Plato. But in the act of shaping a theory or doctrine of mimesis, Plato at the same time placed a considerable philosophical and often polemical burden on the concept, and as a result he did not pass on to Aristotle a freshly honed idea, but a whole set of suggestions, issues and challenges which were superimposed on the intricacies of existing non-philosophical uses of the word. Moreover, as I shall later show, Plato's own interests in mimesis were not as cut-and-dried, nor as entirely negative, as is often believed, and he demonstrated that the idea might have a contribution to make to a large area of philosophical enquiry.

Aristotle's pronouncements on mimesis, taken collectively, reflect the scope for uncertainty and instability produced by the earlier history of the word's applications. It is possible here to give some idea of the complexity of the problems raised by his view of mimesis, but also to bring some order into the relevant material, by trying to separate two salient dimensions of mimesis and two types of theoretical question associated with them. Starting from the minimal proposition that all Greek notions of mimetic art entail a necessary subject-object distinction, and that this must in some way structure a relationship between a mimetic art or work (in whatever medium) and an aspect or level of reality, it becomes possible to identify two main ways in which such a proposition might be elaborated. The first is to define what might be called a formal relationship between work and object, that is a relationship between the medium or mode of the mimetic representation and the relevant features of the represented object. Such a definition requires exemplification by particular arts, though the formal relationship established for one art may be, and often was, used as a model for the understanding of another, or even of mimetic art in general. The question of formal mimetic relationships points especially towards the importance in Greek thinking of analogy between poetry and painting, an analogy which appears in the early classical period and is central to the interpretation of both Plato's and Aristotle's treatments of mimesis.¹¹ The formal aspect of visual mimesis is amenable to a straightforward formulation (however simpliste it might seem to some): colours and forms, as both philosophers would put it, correspond to, and so represent, the colours and forms of visible reality (and perhaps of more besides). There is no doubt that this formulation made it easier for the idea of poetic mimesis, and of mimetic art in general, to become acceptable. Plato can be demonstrably convicted of exploiting the elision in thought involved in taking visual as the paradigm for poetic mimesis, when he sets out to degrade artistic activity to the level of a shallow reproduction of the surfaces of the material world.¹² Despite the lack of a precise connection between poetic and visual mimesis, the analogy between painting and poetry was one which Aristotle was not prepared to surrender: it appears in the opening chapter of the Poetics, and repeatedly in later passages. It should be added that the visual model of mimesis was paralleled by, and in some contexts (such as acting and dancing) found in conjunction with, dramatic or enactive mimesis of human behaviour. In both cases the nature of mimesis can be intimated by noticing that it aspires, or might be held to aspire, to the condition of being indistinguishable from the original.¹³ It is unclear how the same could be said of any conception of poetic mimesis, even, if Plato was polemically ready to suggest this.

Aristotle was, however, undoubtedly aware of the specific issue of the formal relation between mimetic art and its objects. This is apparent in the passage from Book 8 of the Politics (1340a 28ff.) where he distinguishes between the true mimesis of human character which he states to be possible in music, and the weaker relationship—'symbolic', rather than properly mimetic, representation—to be found in the portrayal of character in visual media. It is regrettably common for this passage to be erroneously paraphrased as claiming that music is, tout court, the most mimetic of the arts, but we must correct this by observing that Aristotle refers only to the mimesis of character. Understanding of the point is obscured by our ignorance of Greek music, but it appears to be the case that what is involved is a putatively precise correspondence between the expressive movement of music and the 'kinetic' dimension of active human character.¹⁴ If this remains an alien instance, the Poetics itself, with its tendency towards the idea of dramatic enactment (which is not to be confused with performance as such) as the essence of poetic mimesis, is more easily comprehensible. What underlies this thrust in Aristotle's thinking is the concern for as close a formal match as possible between poetry and its human subject matter, and this leads to the stress on enactive mimesis in which the direct speech of agents gets us as near as language can come to the nature of significant action itself. Such a degree of correspondence is, however, not consistently adhered to as a necessary criterion of mimesis, and it is perhaps possible to see why. The more that certain types of formal equivalence are pressed, the weaker becomes the unifying factor in all the mimetic activities mentioned in the first chapter of the Poetics. Although interested in refining the concept of mimesis analytically by his differential scheme of media, objects, and modes, outlined in the first three chapters of the treatise, Aristotle had nonetheless inherited, particularly from Plato, a general and loose concept of mimesis, and the preservation of such a concept depended on a willingness to accept a fundamental element of mimetic correspondence which cut across the divisions between the arts, and so made intelligible, for example, the kinship between poetry and painting.

The second dimension of mimesis, which is in theory independent of the first but in practice not always kept apart from it, concerns the cognitive status and value of the mimetic work and its content. However the formal relation, as I have called it, is construed for particular types of mimetic art (as direct equivalence, as a kind of symbolism, or as some more complex kind of representation), the question still remains to be asked about the truth-value of mimetic works. The question's importance can be seen in the context of Plato's discussions of mimesis, since if his polemics against art are characterised by a refusal to separate the cognitive aspect of mimesis from the formal (so that the supposed limitation of the latter to literal copying condemns its products to the realm of the derivative and spurious), Plato elsewhere intriguingly gestures towards a notion of philosophical mimesis which would allow its powers of signification to rise above its formal limitations (see ch. IV). Aristotle may have been influenced by this latter fact towards the development of his own doctrine that artistic mimesis is capable of representing universals or general truths, and need not be tied to the reproduction of particulars. It is far from clear, however, that Aristotle sees this capacity as belonging intrinsically to mimesis, for he continues to acknowledge cases of the mimetic representation of particulars, as well, of course, as non-mimetic ways, above all philosophy itself, in which universals can be communicated.¹⁵ Because of this doubt, it might be legitimate to conclude that the chief refinement in Aristotle's view of mimesis pertains to its formal aspect, that is to the analysis of the various types of correspondence to reality which can be achieved in different artistic media (visual, linguistic, musical, etc.); and, as a corollary of this, to regard the claims made in Poetics 9 for the quasi-philosophical potential of poetry as expressing a principle of the use of mimesis, rather than its necessary attribute. But as the Poetics stands, such a conclusion would be tidier than is warranted by the evidence, which suggests a coalescence, as in Plato, between considerations of the formal and the cognitive issues raised by mimetic art.

This preliminary approach to Aristotle's concept of mimesis should demonstrate the difficulty of reaching a clear-cut interpretation of it; but two broad inferences, which will receive further substantiation in ch. IV, can be drawn from the preceding argument: first, that Aristotle unquestioningly accepts the existence of a distinctive group of mimetic arts, and that by so doing he commits himself to a compendious criterion of mimesis as a form of correspondence in which some aspect of reality is reconstituted in a medium as close as possible in equivalence to the object; secondly, that he is prepared to attribute to some mimetic works a cognitive significance which goes beyond particulars to the embodiment of universals. The relationship between this notion of mimesis and the underlying idea of technê which I earlier examined can now be clarified by a brief look at a further area of terminology—the poiêsis word-group—which denotes poetry itself (and perhaps some other mimetic arts too) but is also closely related to technê as a whole. The root meaning of the poiêsis word-group is 'making' or 'producing', and for Aristotle all technê involves poiêsis of some kind: the former is the rationalised, systematic capacity for the latter. But the specific application of these same terms, poiêsis and its cognates, to poetry was a linguistic development well established by Aristotle's time. Extracted from their philosophical context, the remarks of Diotima in Plato's Symposium furnish a reliable description of the phenomenon: 'You know,' she says to Socrates, 'that making (poiêsis) takes many forms. something comes into being from non-being, so that the activities of all crafts and arts (technai) are types of poiêsis, and the craftsmen are all makers (poiêtai) … Nevertheless … some have different names, and a determinate part of the whole area of poiêsis, the part concerned with music and verse, is called by the name that belongs to the whole. For it is only this part that is called poiêsis (making/'poetry') and its practitioners poiêtai (makers/'poets')' (205b8-c9).

Since Aristotle himself used poiêsis terminology both for the definition of technê in general and to describe a particular type of mimetic technê (poetry), it is reasonable to suppose that the latter took some of its colouring from the former. In this respect we can both compare and contrast Aristotle's position with ordinary usage, since twice in the Poetics he adverts to the normal conception of the poet as a maker (poiêtês) of verses (47b 13-16, 51b 27-9). Against this Aristotle opposes his own conception of the poet as a maker of plot-structures (muthoi), that is unified mimetic representations of human action. An immediate point to note is that this new definition of poetic art clearly presupposes, similarly to the case of technê …, that the language of poiêsis has lost any unnecessary material connotations: what the poet 'makes' or produces is not a tangible object, but a mimetic construct in language (and other media) to be apprehended by the mind. It is profitable to draw a further contrast here with the passage in Plato's Phaedo (61b) in which Socrates describes the content of poetic composition as muthoi, the term adapted to his theory by Aristotle: but what Socrates has in mind is setting Aesop's fables (muthoi) to verse, so that the making (poiêsis) which is the 'poetry' would precisely be the making of verses.¹⁶ In ch. 9 of the Poetics, on the other hand, Aristotle argues that, whether or not the poet's raw material is pre-existent (traditional myths being an equivalent to Socrates' Aesopic fables), his task as a poet-maker (poietes) is still to design and organise his plot-structure in such a way as to give its content the universal intelligibility of which Aristotle believes poetry to be capable.

The elements of poiêsis in the Aristotelian concepts of technê and poetry are mutually reinforcing. In neither case does the maker necessarily produce a tangible artefact, though in some technai he will do so. But whatever his media, the maker aims, by the application of rational method, to bring something into being, and in the case of the mimetic arts this is, if successful, a unified construction which must be comprehended as embodying a representation of a possible reality.¹⁷ Although the Poetics elaborates this doctrine in relation to poetry, there is no doubt that Aristotle presupposes a comparable attitude to arts such as music and painting. Thus we can be confident that he would have said that the painter, for example, is not a maker of shapes and colours: these are his media, but they are used to produce, to bring into being by rational art, an image of possible human reality, and an image capable of being understood, at its best, in universal rather than particular terms. Consideration of either the specific poiêsis of poetry or the generic poiêsis of all mimetic art therefore brings us back round to the question which was earlier raised about the cognitive value of mimesis in Aristotle's aesthetic thinking. The 'making' which is the procedure of the poet's or the painter's art is the imparting of design and order to his material, and this design will carry a mimetic correspondence to a conception of reality. But the Aristotelian emphasis on coherence and unity, as we find it applied to the concrete doctrine of poetic plot-structure, gives the poet a very different responsibility for his artistic product from the one assigned to him by Plato, in whose eyes the artist must be accountable for his raw material, not just for what he makes of it. Aristotelian poiêsis is a positive, potent force, whose implications of productive control and purpose should dispel any lingering associations of the derivative or passive from the understanding of mimesis. Poiêsis and mimesis are tightly interwoven strands of the thinking which lies behind the Poetics; and if poiêsis cannot solve all the problems of mimesis for us, it does at any rate help to bond together the elements of technê and mimesis in the Aristotelian concept of mimetic art which it is my aim to delineate.

The argument up to this point has deliberately tolerated the linguistic intricacy and cumbersomeness which are entailed in any attempt to stay close to Aristotle's own language, and so to avoid superficial assimilation of his ideas to later views of art. There is also, however, the lesser danger that the peculiar features of Aristotle's thought will be allowed to exaggerate the alien ethos of his concept of mimetic art, and to impede recognition of the affinities it may have with later attitudes. It will consequently serve a purpose to draw some provisional conclusions about Aristotle's understanding of the mimetic technai, and to try to relate it to some of the more characteristic modern beliefs on the subject. The basis for such a comparison, as earlier indicated, is the fact that the range of activities circumscribed by mimetic technê agrees approximately with the standard modern categorisation of the arts. The historical link between the two lies in the transmission of ancient notions of mimesis and imitation to the Renaissance, and the elaboration of these notions in the period of neo-classicism into the foundation for a system of fine arts whose central principle was the 'imitation of nature'.¹⁸ But we saw earlier … that the origin of this principle in Aristotle is very far removed from its later use as an aesthetic slogan, since in contrast to the vagueness of the latter the Aristotelian formula stands for a philosophy of all rational productive activity, both mimetic and non-mimetic. If neo-classical imitation of nature is manipulable to various effects (including, for instance, a strong idealism) the effects (including, for instance, a strong idealism) the original analogy between technê and nature not only has broader scope but commits its holder to a specific teleology of mimetic art.

This teleology differs less from the orthodox neo-classical aesthetic, however, than it does from the attitudes to art which have grown up since the eighteenth century and to some extent in reaction against neo-classicism. Typical theorists of the sixteenth and seventeenth centuries might have understood Aristotle's insistence on an objective concept of art, structured by the rational relation between the artist and his work, much better than could the Romantic or his modern epigone. The shift represented by the rise of Romanticism can be located, for present purposes, in the new aesthetic centrality of the idea of creative imagination, an idea for which the eighteenth century discovered that Longinus was a much more sympathetic classical source than Aristotle or Horace. The contrast with Aristotle can be illuminated by observing that the origins of this concept of imagination are to be traced in the ascription to the artist of free, creative powers analogous to those of a God who can bring a world into being ex nihilo.¹⁹ For Aristotle, on the other hand, the model for technê, 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 essential 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 hamessed 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 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.

A caution must be entered here. My characterisation of one of the dominant strands in modern attitudes to art is not meant to imply that free, creative imagination is a universally accepted principle, or that aesthetic objectivism has altogether disappeared. The contrast I have drawn is a deliberately schematic one, designed to put Aristotle's views in perspective. The danger of simplification, both of Aristotle's and of later positions, becomes acute when we reach the question of the purpose of function of art. Without raising this question, and seeking an answer to it in the Poetics and in Aristotle's other references to mimetic art, it is difficult to feel that the concept I have so far analysed is a fully rounded one. But we here encounter a striking paradox which may serve as an appropriate introduction to the next stage of my argument. In the two broadest phases which can be demarcated in European theory and criticism of art since the Renaissance, those of neo-classicism and the Romantic reaction (with all its modern off-shoots) against neo-classicism, it is remarkable that Aristotle has been enlisted within both as a supporter of the prevailing view of the function of art. Under neo-classicism, the Poetics was used as a pillar of the dominant moralism, according to which art could be justified in terms of its capacity to be ethically edifying and improving. In the reaction against such views, the last two centuries have seen a growing, though hardly uniform, adherence to the belief that art's ultimate legitimation lies in some sort of self-sufficient pleasure or gratification. Such a belief does not require Aristotelian sanction, but it is nonetheless discernible that it has influenced the common reading of the Poetics and has led to an orthodoxy in which Aristotle's view of art is held to be essentially formalist and aestheticist: that is, one which attributes an autotelic status to poetry and the other arts, which grounds the experience of them in pleasure, which identifies the properties of works of art as purely internal attributes of form, and which entails a strong divorce between art, on the one hand, and morality, religion and politics, on the other. Thus the Poetics, despite the major differences from modern views of art which I have tried to locate in it, is brought into line with the aesthetic consensus which has evolved since the Enlightenment, and in which art has been freed from what is often considered to be the taint of a direct concern with, or effect on, the ideas and values belonging to the realms indicated above.

Alert to the possibility that it may be wrong to hope to adjudicate between these two antithetical views of the Poetics, the didactic and the formalist, in the terms in which their proponents themselves formulate the dichotomy, I want now to undertake a scrutiny of Aristotle's references to the pleasure or pleasures of mimetic art, in the expectation that it will bring to light a vital area of the treatise's philosophical foundations.

That art is pleasurable, and in its own peculiar ways, is a datum of experience which the Greeks did not need to wait for theorists or philosophers to call to their attention. The seductive pleasures even of works of art which portray exceptional suffering are attested from the beginnings of Greek poetry, and not only attested but illuminated in a number of Homeric scenes with a dramatic insight which is irreducible to prosaic paraphrase.²⁰ But from the philosophical point of view, there remained much scope, and much need, for the investigation of the subject. Aristotle had more than one good reason to be interested in discriminating between the types of pleasure which might enter into the experience of poetry and the other arts, not least because the need to take pleasure in the right things was a pervasive principle of his educational and ethical thought. Aristotle's own philosophical psychology posited a rich and complex range of pleasures, related to particular types of activity and the human faculties employed in them. He conceives of pleasure primarily as a level or aspect of experience which supervenes on and completes (like the 'bloom' of those in their physical prime) any activity in which man's abilities are put successfully to their natural use. Pleasure marks the fulfilment of a natural potential, and to understand how and why pleasure arises in individual cases, we must comprehend the special character of the given activities.²¹ In view of the cohesiveness which we have seen his attitude to the mimetic technai to possess, we would reasonably expect a conception of aesthetic pleasure to be an integral element of Aristotle's thinking about mimesis.

A further factor to prompt reflection on the subject was the variety of references to pleasure in Plato's discussions of art. In the Laws and elsewhere Plato claimed that the ordinary man's expectation of art was that it should provide pleasure, and he alleged contemptuously that pleasure was the sole popular standard of artistic merit.²² The Athenian at first rejects such a mentality as 'blasphemous', and later observes that the decadence of recent Greek poetry has followed from an indiscriminate belief in the audience's pleasure as the aim of art and the criterion of success. Yet Plato's spokesman himself comes round to arguing that a certain, very different kind of pleasure is to be desired in art: the pleasure of the educated and morally upright man. This correct and laudable species of pleasure will be the result simply of the artistic presentation of moral truth—the dramatic portrayal, for example, of virtuous men. Elsewhere, in the difficult Philebus (51), Plato argues that the apparent pleasures of art are actually false and synthetic, and he contrasts with the specious pleasure taken in an artistic form the true pleasure which may be derived from contemplation of intrinsically, that is geometrically, beautiful shapes, and perfect pure colours. Without venturing any further, then, into the shifting significances of Plato's treatments of this subject, we already have glimpses in these two passages alone of four different kinds of pleasure arguably relevant to art: the ordinary man's uncultured pleasure, which will involve the gratification of the lower part of the soul; the good man's moral pleasure in the celebration of virtue; pleasure from the sensual forms used in mimetic works; and the pleasure of contemplating pure shapes, colours and musical tones.

In the Poetics too we encounter a wide range of references to pleasure. Before attempting to provide a synthesising interpretation of these, it will be as well to offer a preliminary tabulation of the types or occasions of pleasure mentioned in the treatise, roughly in the order in which they occur.

1. Ch. 4, 48b 8ff.: natural pleasure taken in mimetic works (mimemata), even when these represent objects, such as corpses, which are inherently unpleasant.²³ The cause of this pleasure is that the experience entails a process of understanding and learning. I shall contend below that we have also to put in this category the pleasure which, according to ch. 24, 60a 17, derives from 'the wonderful', which is itself closely related, as ch. 9, 52a 1-11 indicates, to the arousal of pity and fear by tragic poetry.
2. Ch. 4, 48b 17-19: pleasure due to the execution, surface or some other such aspect of a work of mimetic art; this pleasure is independent of the work's mimetic status.
3. Ch. 6, 49b 25-31: 'pleasurably garnished' language is given in the definition of tragedy, and then explained as 'language with rhythm and melody'—that is, the sensual elaborations (both verbal and musical) of the lyric sections of tragic drama (called 'the greatest of the garnishings' or embellishments at 50b 16), and probably also the rhythmical dimension of the ordinary spoken parts of a play. Similarly, in ch. 26, 62a 16f., mousikê, which here means something close though not identical to the 'pleasurably garnished' language of ch. 6, is cited as a source of vivid pleasures which give tragedy an advantage over epic. This is evidently not meant to deny pleasurable embellishments altogether to the epic poet's art: the pleasure of epic style may be at least part of what is meant in the reference to Homer at 60b 2.
4. Ch. 6, 50b 16f.: although denying that theatrical spectacle is strictly part of the dramatist's art, Aristotle does describe it as 'stirring' or 'seductive'. The word he uses is psuchagôgikon,²⁴ and its cognate verb is found earlier in ch. 6, 50a 33, applied to the effect of tragic reversal and recognition. This helps to establish that Aristotle is prepared to attribute a strong emotive potential to spectacle, though it is less well defined, and obviously much less valuable for him, than the potential of the complex plot-structure. Such emotional effects carry implicit pleasure with them. The point is confirmed in ch. 14, 53b 8-11, where the use of spectacle to produce the specious effect of 'the portentous' is frowned on as supplying a pleasure alien to the true tragic experience of pity and fear, for which see (g) below.
5. Ch. 6, 50a 39-50b 3: reference is here made to the contrasting types of pleasure afforded by the (hypothetical) painting which consists of random patches of colour, and the colourless outline sketch or drawing. The latter is said to be analogous to the plot-structure of tragedy.
6. Ch. 9, 51b 23, 26: Aristotle here touches in passing on the pleasure of tragedy, in making the point that it is unaffected by whether the material of a play is conventional or invented. For another general reference to pleasure see 62b 1, in connection with the superiority of tragic concentration over epic diffuseness.
7. Finally, there are four important references in the Poetics to the particular pleasure of tragedy (to which there are corresponding comic (53a 36) and epic (59a 21, 62b 13f.) pleasures): ch. 6, 50a 33-5; ch. 13, 53a 35f.; ch. 14, 53b 10-14; ch. 26, 62b 13f. The third of these passages, which stipulates 'the pleasure arising from pity and fear through mimesis', is the fullest, and indicates that the proper pleasure of tragedy is linked directly with the experience of its distinctive emotions, pity and fear (hence the justification for the inclusion of the first passage in the list), while resting on the basis of the generic pleasure of poetic mimesis ((a) above). The notion of pleasures proper or peculiar to individual activities appears prominently in Aristotle's mature views on the subject in EN Book 10.²⁵

I think it is possible to advance beyond this preliminary catalogue, and to integrate most of this material into a coherent pattern, by deducing three main levels or types of aesthetic pleasure—that is, to reiterate, pleasure derivable from the mimetic arts. It would certainly be fanciful to try to forge a completely systematic theory out of this collection of passages: Aristotle draws analogies with the visual arts, but even so it is not always easy to see just how his principles could be elaborated for mimetic art in general. But the following scheme of analysis is economical, positive, and based squarely on the evidence cited above, with supplementary guidance from other relevant passages in the corpus. I shall attempt to define and analyse the three types of aesthetic pleasure in ascending order of both importance and delicateness of interpretation.

In the lowest position in the scheme comes the inessential, and potentially inappropriate, enjoyment of theatrical spectacle, opsis, by which Aristotle probably means chiefly the visual presentation of the actors (see Appendix 3). In this first case, there are no grounds for extrapolating from poetry to other arts; opsis is not even a feature of all poetry (epic lacks it: 59b 10, 60a 14), but only of dramatic poetry in performance. Furthermore, in his animadversions on spectacle Aristotle does not have in mind the general visual aspect of performance, but the specific exploitation of it for emotional effect. Thus, properly used, opsis can contribute to the tragic effect of pity and fear (53b If.), and I suggested earlier that this must also be part of Aristotle's point at the end of ch. 6, when he describes the potency of spectacle with a term already applied to the elements of the complex plot-structure. It is also in this latter passage, however, that spectacle is relegated to the status of an accessory art. This severely limits Aristotle's attention to it within the theory of tragedy, but the brief acknowledgements of its capability should not be wholly forgotten. We can probably infer that the correct use of opsis would be regarded simply as visual reinforcement of the intrinsic dramatic effect, and the pleasure accruing from the former should therefore be categorised as a secondary manifestation of the true and proper pleasure of tragedy, to which I shall be returning. In any case, Aristotle's concern at the start of Poetics 14 seems to be more with the incorrect use of spectacle to produce a purely sensational pleasure that would distract from, or impede, the appropriate tragic emotions and their concomitant pleasure. The 'portentous' is dismissed as a spurious substitute for these emotions,²⁶ and whatever else one may think about Aristotle's sharp divorce between dramatic poetry and its performance, his attitude at least confirms his philosophical interest in discriminating between types of pleasure, and in evaluating their relative worth.

The devaluation of opsis is also a reflection of Aristotle's aim of shifting the locus of the poet's art from the realm of the sensual to that of the cognitive. But a more subtle and significant indication of this aim is provided by the distinction drawn in Poetics 4 between the natural pleasure derivable from mimetic works, and the sensual pleasure which may be taken in them independently of their mimetic status. It is the latter which is the second of the three main levels of aesthetic pleasure to be analysed.²⁷ Outside the Poetics there are other passages in which Aristotle differentiates between sensual and intellectual pleasures, and it is evidently in the former category that we must place the pleasure noted in item (b) in my earlier list: a pleasure which can be given by works of art entirely in respect of their sensible properties, and without any specifically cognitive or emotional element. Into this category we must also bring item (c) and the first of the types referred to in item (e). The objects of this species of pleasure are the forms, textures, patterns and sounds of art, apprehended in and for themselves and not as the medium of mimetic significance. This is the disjunction which Aristotle makes explicitly at 48b 17-19, by immediate reference to a visual example, but as part of a larger argument whose scope covers also the whole of poetic mimesis. Although condensed, this passage is sufficient to establish that sensual pleasure cannot, in Aristotle's terms, properly account for the experience of any mimetic art, though it is recognised as having a legitimate, secondary part to play. That such pleasure seems to be most readily identifiable in the case of the visual arts is suggested by the fact that in two of the three relevant passages of the Poetics cited above Aristotle makes his point by reference to painting (items (b) and (e)). If we ask what the applicability of the concept is to the non-visual arts, Aristotle's answer is embodied in what he has to say about the rhythmic and musical enhancements of language ((c) above): it is the most directly sensible aspects of the language of poetry, its rhythmical and melodic 'garnishings', as the philosopher regards them, which supply the closest equivalent to the sensual forms and colours of visual art.²⁸ But the inference is not easily to be elaborated, since Aristotle's view of the mimetic nature of music reduces the scope for the derivation of a purely sensual pleasure from it, and the evidence of both the Poetics and the Rhetoric implies that this qualification holds good even for rhythm independently of melody.²⁹

It is at any rate clear that any sensual pleasure afforded by poetry cannot in Aristotle's theory be of more than secondary significance, and it is worthwhile to note that this inference has a bearing on his attitude to performance, since the availability of pleasure from the rhythmic and melodic aspects of poetry clearly depends on this. In the case of the rhythms of spoken verse, this condition would be satisfied by the standard practice of reading aloud (alluded to perhaps, for different purposes, at 53b 3-7),³⁰ but for sections of lyric poetry presumably no alternative to musically accompanied performance would be sufficient. Various lines of thought therefore converge to produce a devaluation of lyric in the Poetics: the conception of its distinctive features as an embellishment or 'garnishing' of tragedy (and consequently inessential); the separation of poetry proper from performance and its attendant arts; and the relegation of the sensual pleasure of poetry to the level of the subordinate. That this devaluation does indeed result in a failure on Aristotle's part to do justice to the lyric dimension of Greek tragedy is something I shall argue in greater detail in ch. VIII.

But it is not only the particular connection of sensual pleasure with performance and with lyric poetry which explains its status in the Poetics, for we have seen that the passage in ch. 4 where Aristotle distinguishes between a cognitive and a sensual pleasure from works of art refers directly to a visual example, and suggests that the principle ought to be applicable to mimetic art in general. If so, the reason must be sought in a larger aspect of Aristotle's philosophy. At the opening of the Metaphysics Aristotle makes the famous pronouncement that 'all men by nature desire knowledge', and he goes on to relate this postulate to the observation that we value the senses, regardless of utility, for the pleasure they afford us. This pleasure, however, particularly that of sight, is not merely sensual; its cause is the cognitive content of our perceptions, the contribution which they make to the acquisition of knowledge to which our very nature predisposes us. Hence the supreme value of sight, for 'this is the sense which allows us to apprehend the most, and which reveals the most discriminations'.³¹ This passage in fact generalises the principle which is enunciated specifically for the experience of mimetic art in Poetics 4, and it indicates the broader foundation on which Aristotle bases his claim that the true pleasure of this experience entails a process of understanding and learning. The negative corollary of this is that the purely sensual enjoyment of form, colour, texture, rhythm and so on, does not involve the full use of the senses' natural cognitive capacity.

It is, then, to the elucidation of the third and highest level of aesthetic pleasure posited by the Poetics that I must now turn. Two items from my original list are left for consideration: the natural pleasure of learning and understanding through mimesis (a), and the proper pleasure of tragedy (g), 'the pleasure arising from pity and fear through mimesis' (taking the passages in (f) to represent indefinite allusions to the latter). What I wish to argue is that these two items in fact present aspects of the same phenomenon, or, more precisely, that they are related as genus to species, so that the pleasure proper to tragedy is one example, and perhaps for Aristotle the supreme instance, of the generic concept of pleasure from mimetic works sketched in Poetics 4. But to establish the cogency of this position, which has not been widely adopted in interpretation of the treatise, will require some close reasoning, and in the first place a scrutiny of the crucial passage in ch. 4.³²

It is a paradoxical part of the difficulty of dealing with Poetics 4 that its train of thought is ostensibly so plain and unremarkable. It is possible to carry away from a reading of it, seduced by its very simplicity, and undisturbed by consideration of how Aristotle's reasoning here must be integrated with what he offers later in the book, the impression of nothing more than the concise formulation of rudimentary and preliminary points. Nor is this impression entirely erroneous; but it is essential to dissociate the straightforwardness of the argument from the question of how far-reaching the implications of it can be seen to be.³³ The theme of the first half of the chapter is the original causes of poetry, though the progression of thought, with its general references to mimesis and its visual examples, leaves little room to doubt that the enquiry applies to all mimetic art. Aristotle suggests, with typical ingenuousness, two 'natural' causes for poetry. It is possible, I believe, to discern that these are really facets of a single explanation. The first cause or reason is the instinct in human beings from childhood onwards to engage in mimesis, for which we here need to have in mind at least the two ideas of 'imitation' and 'enactment' (static or active representation). Moreover, even early mimesis involves learning, though Aristotle makes the point so briefly that it is impossible to determine how wide a range of behaviour he means his audience to understand.³⁴ At any rate, we can say that alongside man as a political creature and a rational creature, we may juxtapose, in Aristotle's perspective, man as the most mimetic of creatures. Mimesis is rooted in human nature, and is implicated in distinctively human patterns of action. We recognise readily that within the naturalistic terms of Aristotelian philosophy mimesis is thus vindicated in the face of the intrinsically suspect and shallow status to which Platonic metaphysics had often condemned it, and that the mimetic technai now have a doubly natural grounding and sanction: first, as I earlier demonstrated, in the teleological character of technê, productive art, in general; secondly, in the human propensity towards mimesis. The mimetic arts consummate this propensity by developing it into various types of rational art.

Aristotle's second cause of poetry also concerns mimesis,³⁵ but adjusts the point of view to that of the recipient or spectator, rather than that of the agent or artist. To the universal instinct for engaging in mimesis Aristotle now adds the equally natural pleasure which is taken in the mimetic activities or works of others, and from his elaboration of the point we gather that these two factors coalesce as elements of a single phenomenon. For the underlying explanation of the pleasure taken in the apprehension of mimetic objects (which need not be physical objects) is the primary human pleasure in learning: learning and understanding therefore appear as the basis of both the active and the receptive interest in mimesis (48b 12f. refers back to 5-8). Moreover, the cognition involved in mimesis is equally a source of pleasure in both cases, since Aristotle's claim that there is a natural human instinct for mimesis entails, given his philosophy of pleasure, that men take a natural pleasure in exercising it as well as in appreciating its products. The two causes of poetry, then, and of mimesis in general, turn out to be aspects of a single psychological and cultural hypothesis, and one which contains the nucleus of a highly serious concept of aesthetic pleasure.

One reason for a common refusal to take that last step in my argument seems to be a sense, to which I have already referred, that ch. 4 of the Poetics represents a preliminary or marginal part of the treatise, and can offer no insight into the centre of Aristotle's idea of poetry or art. It is, of course, the case that this passage is initially offered by way of accounting for the origins or causes of poetry: it furnishes what might be regarded as the psychological and anthropological premises from which Aristotle can advance to the sketch of poetic evolution which follows in the rest of the chapter. But to infer from this that Aristotle's causes of poetry have only a hypothetical or historical reference, and have no bearing on the developed forms of art, would simply be to overlook the fact that these causes are presented as putatively permanent, because natural, data about human engagement in mimetic activities. Origins correspond to fulfilment: in Aristotle's beginning is his end. The argument is, in fact, not properly historical at all, but philosophical—committed, that is, to the explication of underlying and universal causes. It should be remarked, moreover, that Aristotle's verbs are all in the present tense in this passage: 'mimetic activity is natural to men from childhood onwards … man is the most mimetic of creatures … we enjoy looking at pictures … ', and so on. Aristotle's conclusion that 'if one has not seen the object of a picture before, it will not produce pleasure qua work of mimesis, but by virtue of its execution etc.'³⁶ therefore unequivocally states a principle about the status and comprehension of mimetic works in general, though its implications remain to be drawn out. While external confirmation of this is hardly needed, it happens to be available in the fact that Aristotle makes the same point in very similar language in Book 1 of the Rhetoric, where he affirms, without reference to anything but existing forms of mimetic art, that the aesthetic pleasure derived from them contains a process of recognition and understanding implicit in the appreciation.³⁷

That Aristotle should regard the basis of aesthetic pleasure to be an experience for which he finds the language of 'learning', 'comprehending' and 'reasoning' apt, is a conclusion of major import for his philosophical view of art, and particularly of poetry.³⁸ Yet the clarity of this position belies the problems which confront a deeper interpretation of it. Perhaps the most immediate of these problems arises from the fact that in explaining the natural human roots of poetry, Aristotle chooses, as he also does in the passage from the Rhetoric mentioned above, an illustration from visual art, and one, moreover, of arguably disappointing simplicity. The example of a picture, or other visual work, which portrays an identifiable (though not necessarily a real) figure, and to which the mind of the beholder may respond with the reasoning, 'this is so-and-so', might well be thought to shed little enough light on the type of cognition involved in the experience of paintings, and none at all on the understanding of poetry. Given the ambitious scope of the context—an explanation of the psychological causes of poetry, and implicitly of all mimesis—we must assume both that Aristotle gives a visual instance of something which can take non-visual forms, and also that he is deliberately citing a simple case of an experience which must have more complex varieties (in the apprehension of both visual and non-visual art).³⁹ In one respect the simplicity of the illustration does serve well to mark the tenor of the passage as a whole, for Aristotle is sketching a view of a large range of human activity but contending that there is a fundamental unity in the experiences which underlie it; the fact that the passage contains references to both the playing of children and the philosopher's pleasure in knowledge, is significant of the potential scope of the argument. The implicit comparison between the pleasure of mimetic art and the pleasure of philosophical knowledge recurs more pointedly in the Parts of Animals (645a 7-15), where we also find the example of unattractive animals—animals which, because ex hypothesi repellent, cannot be a source, whether in art or in life, of sensual pleasure. Although this passage is explicit on the reason for philosophical pleasure (the understanding of causes), it leaves that of art less than clear; but given its analogy between art and nature (a theme explored earlier in this chapter) it may not be unreasonable to infer a kind of cognition of causes in the pleasure derived from mimetic works too.

The outline of a cognitive theory of aesthetic pleasure in Poetics 4 accords with Aristotle's mature view that pleasure involves the natural exercise of human faculties. In the experience of mimetic works any element of purely sensual pleasure must be subordinate to the processes of recognition and learning which constitute the proper response to mimesis. Without cognitive recognition, the status of the mimetic work—the representation of a possible reality which it embodies—cannot be grasped and therefore cannot be enjoyed, though the senses may take separate pleasure in certain material aspects of the work (48b 17-19). But the use of a rudimentary visual example at Poetics 48b 15-17 (a general mannerism, incidentally, of Aristotle's philosophical method) seems at first sight to impede further illumination of this fundamental layer in the thought of the treatise. In order to make headway with the interpretation of Aristotle's concept of aesthetic pleasure, we need therefore to explore the possibility of implicit connections between what is said in the first part of ch. 4 and some of the work's other doctrines. Two lines of enquiry can be initially distinguished. The first is to examine the relation between the general cognitive pleasure of ch. 4 and the 'proper pleasure' of tragedy referred to elsewhere in the work. After this, it will be necessary to consider more closely the idea of comprehension or learning which Aristotle ascribes to the enjoyment and appreciation of mimesis.

An attempt to combine Poetics 4 with later parts of the treatise is encouraged, among other things, by the passage from Book 1 of the Rhetoric which I mentioned earlier. Aristotle there associates understanding with 'wonder', both of which he says are usually pleasurable: understanding, because, in short, it fulfils man's nature, and wonder, because it involves a desire to learn. Elsewhere we find Aristotle repeating what Plato had said before him, that wonder is the source or origin of philosophy itself, because it represents man's primary thirst for knowledge.⁴⁰ We have here a pointer to one possible link between parts of the Poetics, for there are a number of passages in the treatise where Aristotle touches on the poet's use of 'the wonderful'. The most revealing of these occurs near the end of ch. 9, where it is proposed that pity and fear are best aroused in tragedy by events which happen 'unexpectedly but on account of one another', and that such events will produce wonder more than would chance happenings. It emerges here that there is a kinship, in tragedy, between pity and fear and 'the wonderful': the same kinds of tragic events (in the ideally complex drama, at any rate) should elicit both. Furthermore, the effect of wonder to which Aristotle here refers is explicitly related to the intelligible causation of the events of tragedy (notwithstanding the immediate impact of surprise), and that this is so is corroborated by the following remark that even chance events arouse more wonder when they appear to happen for a purpose—that is, appear to be part of an intelligible sequence.⁴¹ If, then, we put this passage together with the one from Rhetoric 1, wonder becomes a link between the tragic emotions, on one side, and our understanding of the structure of a dramatic action, on the other. Wonder itself does not seem to be simply identifiable either with the particular emotions elicited by tragedy, or with the process of understanding: yet it has both an emotional and a cognitive significance, in that it is felt alongside—as part of the same experience as—pity and fear, and offers a challenge to the mind which, ideally, stimulates and leads on to comprehension or knowledge. Aristotle's comments on wonder in the Poetics and Rhetoric help a little to strengthen the case for trying to see the general thesis of Poetics 4 and the detailed analysis of tragedy later in the treatise as interrelated and mutually illuminating.

In the juxtaposition of the tragic emotions with wonder in ch. 9 of the Poetics we have an indication that the peculiar pleasure of tragedy, which Aristotle defines in ch. 14 as 'the pleasure arising from pity and fear through mimesis', should be regarded as one species of the generic aesthetic pleasure whose elements are sketched in ch. 4. Such a claim gives some illustration of the need to seek out associations which our text of the treatise omits to make explicit, and which Aristotle may either have taken for granted or else have drawn out orally in his philosophical teaching. The cogency of such claims depends on the possibility of discerning multiple signs of underlying relations between superficially discrete ideas or doctrines, and this can, I believe, be plausibly achieved for my argument that the peculiar pleasure of tragedy represents in one specific form, adapted to the particular characteristics of the genre, Aristotle's essential or primary concept of aesthetic pleasure, in which cognition and emotion are integrated. In Poetics 6 Aristotle picks out the components of the complex plot as the most potent of the resources of tragedy, a remark which can only be understood in terms of the distinctive tragic emotions of his own definition. This observation is borne out by the later concentration on the complex plot in the prescriptions for the ideal tragedy. Recognition and reversal are the focus of the finest tragic plot-structure, and so the focus of the emotions aroused by it. In the passage from ch. 9 to which I have already drawn attention it can reasonably be inferred that Aristotle has these aspects of the complex plot, perhaps particularly reversal, in mind: for he says that pity and fear will be best produced by events which happen 'unexpectedly but on account of one another', and this can be taken virtually as a definition of reversal (peripeteia). It is, therefore, telling that in Rhetoric 1 Aristotle singles out precisely sudden and unforeseeable reversals of fortune (peripeteiai) as a source of wonder (1371b 10): they both surprise us and arouse our minds to look for an underlying explanation of the ostensibly inexplicable. Once again we see a convergence of ideas, the sign of a nexus of associations between the complex plot, the tragic emotions, and wonder, with the latter's implications for learning. To this we must add the crucial fact, which can only be stated here but will be discussed in detail in ch. VI, that Aristotle's conception of the emotions, pity and fear, itself rests on a cognitive basis: properly educated, at any rate, these emotions are not arbitrary or irrationally impulsive, but are aligned with the recognition and understanding of certain types and patterns of suffering or misfortune.

The provisional conclusion can therefore be drawn that the peculiar pleasure of tragedy is not a wholly autonomous phenomenon, a self-sufficient category of experience. It is, in the first place, a species of the genus of aesthetic pleasure, the pleasure taken in mimesis, which Aristotle defines in ch. 4 as entailing a necessary process of recognition, learning or understanding. The particular tragic species of aesthetic pleasure involves distinctive emotions, but these emotions are themselves only fully intelligible and justifiable in terms of the cognitive apprehension of certain kinds of human actions and their consequences. The proper tragic pleasure not only shows the relevance of Aristotle's general comments on aesthetic pleasure, by providing an instance of their embodiment in the theory of an individual genre, but also gives some idea of the ways in which we can expect the elementary model of ch. 4—the identification of a subject in visual art—to be complicated and made more sophisticated by internal factors of a genre. Most obviously, the example used in ch. 4 posits a case of simple recognition without an emotional dimension, whereas Aristotle's whole theory of tragedy assumes an interplay and integration of the intellect and the emotions. Furthermore, the illustration in Poetics 4 involves particulars ('the man in the picture is so-and-so') while Aristotle's theory of poetry assigns to the art the potential to deal with universals: the cognitive experience of such art needs correspondingly to be framed in far richer terms than those used in ch. 4's outline of Aristotle's position.⁴²

If my argument so far has attempted to orient us in the direction in which the interpretation of Aristotelian aesthetic pleasure ought to lie, it also could be said to force us back to consider the notions of learning, understanding and reasoning of which both Poetics 4 and the passage from Rhetoric 1 speak in connection with mimesis. Since it is doubtful whether a scrutiny of Aristotle's terms will in itself allow us to make much progress on this point, for their range of meaning is too broad, we need to look for other clues to the elucidation of the type of cognitive experience which he takes to be implicit in the proper appreciation of mimetic works. One such clue may be furnished by a further detail which the passages from the Poetics and Rhetoric have in common. In the latter, Aristotle describes wisdom, which is virtually synonymous with philosophy, as 'the knowledge of many wonders', thus reinforcing his observations in the Metaphysics on the status of wonder as the motive of philosophising.⁴³ Philosophy is also mentioned in Poetics 4, where it is said that 'learning is highly pleasurable not only to philosophers but likewise, if to a lesser degree, to all other men'. While this reference to philosophy sharpens the paradoxical simplicity of the instance of cognition which Aristotle goes on to give, it should not be lightly underestimated; I cited earlier the passage from the Parts of Animals where the experiences of philosophy and of mimesis are also connected. It ought to be stressed that any direct comparison between philosophy and the mimetic arts, however seemingly casual and qualified, could hardly fail to strike someone familiar with the Platonic background as bold; and the comparison was not one which most philosophers would have easily accepted.⁴⁴ That Aristotle should have drawn it in several of his works is evidently significant. It intimates that the cognition involved in the appreciation of mimesis is not wholly different in kind, though it may be in degree, from philosophical thought, and the Parts of Animals passage suggests that the understanding of causes can play a part in both. We have, therefore, corroboration for the interpretation of Poetics 4 as of fundamental importance for the assessment of Aristotle's view of art and the experience of art, and we also have a prompting to draw into the enquiry the other passage in the Poetics where philosophy is mentioned.

'Poetry is more philosophical and more serious than history' (5lb 5f.). If we look to this famous sentence from ch. 9 for further light on Aristotle's idea of the comprehension implicit in the experience of mimesis, we can notice at once that ch. 9 as a whole offers a more refined notion of intellectual activity than the one suggested by ch. 4's example from the visual arts. Part of the dissatisfyingly simpliste impression given by that earlier illustration lies in its apparent equation of understanding with factual knowledge, so that it remains uninstructive just what the value is of being able to recognise that a picture of, say, Achilles in his tent is just that. Part of the importance of Poetics 9 is that it helps to rectify this impression, by providing a more elaborate account of the kind of cognition which poetic mimesis calls for, as well as a more detailed placing of poetry in relation to other intellectual activities. Poetics 9 represents a broadening out into general poetic principles of the particular prescriptions given in the preceding chapters (6-8) for the structure and unity of tragic plot. Unity of plot, as I shall argue in ch. III, is not for Aristotle a purely formal matter, for the essential reason that plot itself, muthos, is not an exclusively formal concept, but a concept of significant (because mimetic) form. The criterion of unity on which so much emphasis is laid is that of 'necessity or probability', which is a principle of the logical and causal relations between actions or events. For poetry to conform to this criterion is for it to produce plots mimetic structures of human action—which embody generalised patterns of universals, as opposed to the random particulars of history. Such universals are meant to be intelligible precisely as such: that is, the mind which contemplates poetic mimesis can perceive it and understand it as the dramatic communication of universals. This notion of poetic significance is so far from the impression given by Poetics 4 that, if we were to take the latter's example of visual mimesis at face-value, and infer from it a poetic equivalent, the result would be closer to ch. 9's view of history, with its alleged restriction to the particular, than to its claims for a quasi-philosophical art of poetry.⁴⁵ But this is an argument not for ignoring the general implications of ch. 4, but for adjusting our interpretation of it, and particularly of the status of its illustration from visual mimesis, in the light of what is to be learnt about poetic mimesis later in the treatise.

Thus Poetics 9 serves to carry Aristotle's view of the cognitive experience of mimesis beyond the limitations of the earlier formulation. As a further reason for treating ch. 4 as a simple statement of a principle capable of much more complex elaboration for particular arts and genres, we can now add the overall tendency of Aristotle's handling of mimesis, which I shall be returning to in ch. IV. Given that Aristotle repudiates the need for mimetic works to involve a one-to-one correspondence with reality, the simplicity of the visual example in ch. 4 is best regarded as a deliberately minimal and uncontroversial process of cognition, but one which would clearly be insufficient where the nature of the mimetic work is richer. Once this qualification on the argument of ch. 4 is accepted, it follows that further enlightenment on Aristotle's concept of aesthetic pleasure can only be pursued within the context of the study of his treatment of individual genres. I have already argued that the proper pleasure of tragedy is to be taken as a species of the generic type indicated in Poetics 4; the pleasure provided by the genre matches the distinctive emotional experience of it, and this emotional experience is itself integrated with the understanding of the structure and causation of human action dramatised in the tragic plot. It therefore becomes impossible to specify more closely the nature of the proper pleasure of tragedy other than by investigating the Poetics' theory of tragedy as a whole, and by attempting to ascertain whether the treatise allows us to identify a particular area of universals in the apprehension of which the particular cognitive-cumaesthetic pleasure of tragedy resides. This investigation will be undertaken in later chapters of this book.

The interpretation of aesthetic pleasure in the Poetics is, then, not readily to be brought to a definitive conclusion. The brevity of Aristotle's explicit remarks on the subject stands in the way of a complete exposition of his views. But it must also be said that if the line of argument I have followed is correct, then it is not altogether surprising or objectionable that Aristotle should have failed to supply a full statement of his concept of aesthetic pleasure, since the hints that we are given gesture in the direction of a theory which does not separate off aesthetic experience as discrete and self-contained, but relates it both to natural human instincts and to the 'higher' intellectual activity of philosophy. While we may regret that Aristotle did not examine these relations in more detail, he says enough to establish the fundamentally cognitive character of the experience of mimesis, and so to imply the kind of framework within which the rest of his discussion of poetry must be placed. And if we associate, as I have contended, the particular pleasure of tragedy with this underlying notion of aesthetic pleasure, then we are now in a position to see why it is so misleading, as I suggested in my introductory chapter, to attribute to Aristotle, without the necessary qualifications, the belief that the aim of poetry is pleasure:⁴⁶ misleading, principally because such a formulation of his position is likely to import an idea of aestheticism, in which the autonomy of works of art is linked with the autonomous character of our enjoyment of them. The central role of pleasure in Aristotle's aesthetics needs to be understood, as I have tried to show, in close conjunction both with the broad indications given in the treatise of the essentially cognitive experience of mimetic works, and with the particular content of the theory of tragedy. Seen in this way, aesthetic pleasure complements the analysis of the Aristotelian concept of art which I offered earlier in the chapter, for the pleasure of those who experience mimetic works is a response to the intelligible structure imposed on his material by the artist's rational capacity. And these two things, the maker's art and the recipient's pleasure, are a reflection of the natural status of mimesis and of the framework within which its individual types evolve: successful mimesis is of significance in Aristotle's eyes, and can be vindicated against Platonic condemnation, because it fulfils man's natural potential to understand reality by reconstituting it in some of the materials over which he has rational control.

Notes

¹ See Butcher viii (though cf. 113f. for reservations) and Bywater vii. For judicious criticism of Bywater see H. Lloyd-Jones, Blood for the Ghosts (London 1982) 18 (where the quotation shows that Bywater to some extent confused conceptualisation with terminology: on this point cf. R. Wollheim, Art and its Objects 2nd edn. (Cambridge 1980) 103f.). Perhaps the most ambitious attempt to find an aesthetic system in Ar. is that of Teichmiiller vol. 2; for a sketch of other nineteenth-century works see Svoboda 5-9, whose own book provides flat paraphrase of some of the relevant material.

² On 'aesthetic' cf. also n. 27 below and p. 229 with n. 37. I limit the term in this way for the sake of historical, not philosophical, clarity: for a possible distinction between 'aesthetic' and 'artistic' see D. Best, Philosophy 57 (1982) 357-72 (revised as ch. 11 in his Feeling and Reason in the Arts (London 1985)).

³ For technê in the Poetics see Kassel 75, Index Graecus. It is used of music and dance at 47a 21-8, and of the visual arts at 47a 20, as at PA 640a 32, 645a 12, EN 1141a 9-12, 1175a 24, Pol. 1281b 12-15. Outside Ar. technê is used of poetry especially by Aristophanes (see ch. I n. 16); Plato sometimes but not always denies technê to poetry: see Appendix 2 under 47a 20f. For the visual arts compare e.g. Emped. fr. 23.2, Plato Gorg. 450c 9f., Diss.Log. 3.10, 17, and see A. Burford, Craftsmen in Greek and Roman Society (London 1972) 198-217.

⁴ See e.g. Collingwood 15-41 (developing a Crocean aesthetic), whose statement of the concept of technê (esp. 18f.) is, however, not always reliable: for necessary reservations on this and other points see S. H. Rosen, Phronesis 4 (1959) 135-48. On technê in general cf. Pollitt 32-7.

⁵Technê and knowledge (epistêmê) are carefully distinguished in EN 6.3-4, 1139b 14ff., but elsewhere they are often assimilated: e.g. An.Pr. 46a 22, Met. 981al-b9, EN 1097a 4-8, Rhet. 1355b 32, 1362b 26, 1392a 25f. For the use of the two terms cf. Bonitz 759, s.v. technê. On the relation of the Poetics to Ar.'s wider concept of technê see Olson (1965) 175-86, and on the latter alone cf. K. Bartels, 'Der Begriff Techne bei Aristoteles', in Synusia: Festgabe für Wolfgang Schadewaldt, edd. H. Flashar and K. Gaiser (Pfullingen 1965) 275-86. For Ar.'s abstraction of theoretical technê from practice note esp. Met. 981a 30-b 6.

⁶ 'Art imitates nature': Phys. 194a 21f., 199a 16f., Meteor. 381b 6; cf. Protr. B13, 14, 23 Düring (1961) (with his comments on p. 187), and ps.-Ar. Mund. 396b 12. Other analogies: Phys. 194a 21ff., 199a 8-20, b 1-4, 26-32, PA 639b 16ff., 640a 26ff., 645a 8, GA 730b 7, 743a 26, Met. 1032a 12ff., 1034a 33ff., EN 1099b 21-3, 1175a 23f., Pol. 1333a 22-4, 1337a 1-3, Protr. B47 Düring (1961). For the essentially shared teleology of mind and nature see De An. 415b 16f. On both Ar. and others see A. J. Close, 'Philosophical Theories of Art & Nature in Classical Antiquity', JHI32 (1971) 163-84.

⁷ See esp. Phys. 194a 27ff., GA 762a 16f., 767a 17, 775a 20-22, Met. 1032a 12-14, 1070a 6-8, EN 1140a 10-16, Protr. Bll-15 Düring (1961).

⁸Met. 1074b 10-12, Pol. 1329b 25ff. It is clear in both these passages that the doctrine does not apply only to arts which cater for necessities.

⁹Met. 98lb 17ff.; for the dichotomy cf. e.g. Pol. 1329b 27-9 and Isoc. 4.14. We must take account of the fact that non-utilitarian pleasures can be encompassed by Ar.'s concept of diagôgê—cultured leisure related to happiness: Pol. 1338a 1-34.

¹⁰ Esp. Laws 668b 9ff., where the speakers agree that everyone would accept that all products of mousikê (which here includes poetry, music and visual art) are 'mimesis and image-making'. Ar.'s acceptance of mimesis as a general view of art is seen most obviously at Poet. 47a 13-22 and Rhet. 1371b 4-8.

¹¹ For various comparisons between poetry and visual art see: Simonides apud Plutarch 346f, Ion of Chios fr. 8, Diss.Log. 3.10, Plato e.g. Rep. 377e, 596ff. (cf. Appendix 2 under 47a 18ff.), Isoc. 9.73-5, Ar. Poet. 47a 18-20, 48a 5f., 48b 9-19, 50a 26-9, 39ff., 54b 9-11, 60b 8f., 17ff., 31f., 61b 12f., Rhet. 1371b 6f.

¹² On this flaw (or polemical tendentiousness) in Rep. 10 see Annas's article, and cf. more generally Keuls 33-47.

¹³ For this idea see e.g. Hom. Hymn Apollo 163f., Plato Rep. 598c, Soph. 234b (but note Crat. 432b-c for a logical qualification), and the famous story about Zeuxis and Parrhasius at Pliny NH 35.65. Note also the idea of visual works like living things, à la Daedalus: Pind. 01. 7.52, Eurip. fr. 372, Plato Phdr. 275d. On artistic 'deception' cf. ch. I n.20.

¹⁴ See ch. IV p. 125 and n.29.

¹⁵Poet. 48b 15-19 clearly cites a visual instance of the mimesis of particulars (though the implications are wider: see pp. 73ff.) and 51b 14f. refers to the equivalent in iambic poetry: 48b 33f. shows, I believe, that pace Janko e.g. 61 iambus probably does count as poetry for Ar., though certainly of an inferior and negligible kind. The remarks in Poetics 9 show the strongly normative thrust of Ar.'s theory (cf. pp. 38f.): he is here attempting to define the nature of the best poetry (an important part of any technê: Pol. 1288b 10-21) and generalises from this position without quite trying to deny (as 51b 14f. intimates) that some poetry may deal with particulars. On iambus cf. ch. IX n. 36.

¹⁶ It is true that Socrates thinks of setting already existing fables as a pis aller, since he cannot invent his own. This still does not bridge the gap between Plato's and Ar.'s positions.

The original sense of muthos was anything said or told: an utterance, speech, story, report, etc. It later acquires the idea of something intrinsically false: a myth, fable, fiction, etc. Cf. LSJ s.v. 11 1-4. References to muthoi in poetry before Plato (e.g. Hom. Od. 11.368, Pind. Ol. 1.29, Nem. 7.23) simply reflect the ordinary meanings of the term. Plato, exploiting the connotations of falsehood (as well as associations with idle tales and the like), treats muthologia (story-telling) as the essence of poetry: e.g. Phaedo 61b 4f., e2, 70b 6, Rep. 377d ff., 380c 2, 392d 2, 394b 9f., Laws 941b-c.

Outside the Poetics, Ar.'s use of muthos and cognates almost invariably carries implications of falsehood, though not always as disparagingly as at Met. 995a 4f., 'fictional (muthôdê6) and childish'. He applies the words chiefly to: myth and legend (Phys. 218b 24, Cael. 284a 18ff., HA 580a 17, MA 699a 27ff., Met. 982b 18f., EN 1100a 8, Pol. 1257b 16, 1269b 28, 1284a 22, 1341b 3), poetic theology (Met. 1000a 18, 1074a 38ff., 1091b 9), fable (Meteor. 356b 11-17, HA 578b 25, 579b 4, 609b 10, 617a 5, PA 641a 21). Although there are some general references to poetic muthologia (e.g. EE 1230a 3), I cannot find the Poetics' special sense of muthos anywhere else in the corpus. What this means is that Ar. has taken a term with the senses of story-fable-legend-myth, and without erasing these altogether (they can be seen within the Poetics itself at 51b 24, 53a 18, 37, 53b 22) he has given the word a new critical edge and significance. We must observe in particular the subtle movement away from the associations of traditional myths: a poetic muthos need not be traditional (51b 23ff.), and it may even borrow from history (ibid. 30-2). But whatever the source of material, it must be made afresh—i.e. shaped into a coherent design—by the maker-poet, and the true muthos is the result, not the original material, of this act (esp. 51b 27-9). See also ch. I pp. 22f.

¹⁷ Ar.'s underlying notion of technê puts the stress on the maker-product axis, but this should not be formulated as a focus on the art rather than its products, as it is by Else (1957) 6, 9, 12, 237, 279f. etc. To take just one detail, note how 'thought' (dianoia) is defined at 50b 4-12 as both a poetic capacity and a property of poems.

¹⁸ On the emergence of this system see Kristeller, but in pressing the lack of an ancient version of eighteenth-century aesthetics too hard (166-74) he gives an inadequate account of the place of mimesis in Ar.'s theory (171 f., with some naïveté regarding Plato too). The general comparison of ancient and modern views of art by W. Tatarkiewicz, JHI 24 (1963) 231-40, is also marred by a neglect of mimesis and a misunderstanding of Ar.'s concept of it (233).

¹⁹ See ch. III n.1.

²⁰ In particular Il. 3.125ff. (Helen's embroidery), 9.186ff. (Achilles' song), Od. 1.325ff. and 8.499ff. (Phemius' and Demodocus' songs of the Trojan War). For sensitive comment on such material see Macleod 6-12.

²¹ Ar.'s most extensive discussions of pleasure are in EN Book 7, chs. 11-14, 1152b 1ff., Book 10, chs. 1-5, 1172a 16ff., and (less subtly) Rhet. Book 1, ch. 11, 1369b 33ff. For a recent and full discussion see J. C. B. Gosling & C. C. W. Taylor, The Greeks on Pleasure (Oxford 1982) chs. 11, 14-15. On the relation of pleasure to virtue cf. ch. VI n. 39.

²² See the whole of Laws 655c-660e, and later 667b ff., 700e, 802c-d; cf. Hipp. Maj. 297e-8a, Gorg. 501d-2d, Rep. 607a.

²³ Note the implicit relevance of this example to tragedy and the paradox of tragic pleasure; similarly in the parallel passage at Rhet. 1371b 4-10, and compare the pleasurable memory of things not pleasant at the time, Rhet. 1370b 1-7.

²⁴Psuchagôgein (and cognates), originally meaning 'to conjure souls', became used in various ways of the captivating power of language, particularly poetry and rhetoric, and also of the visual arts (contra Pollitt 101 n.34): see esp. Aristoph. Birds 1555 (a pun?), Xen. Mem. 3.10.6, Plato Phdr. 261a, 271c, Isoc. 2.49, 9.10, Timocles fr. 6.6 (echoing Ar. himself? cf. Appendix 5 §2), and LSJ with Suppl. s.v. for later passages (to which add Marc.Aur. Med. 3.2 and 11.6). This metaphorical development was probably influenced by Gorgias's theory of the magical and bewitching powers of language, Helen (fr. 11) 10-14, which itself is a development of the archetypal Homeric idea of poetry as enchantment: on this and related points see ch. VI pp. 188ff. and nn. 26-8.

²⁵EN 175aI 22ff.: the phrase 'peculiar pleasure'(oikeia hêdonê) occurs at 1175a 31, b 14, 21, 27 etc. It was also used by Plato, but to denote pure, 'true' pleasures: Phileb. 5 1d, 63e, Rep. 586e, 587b. The attempt of Else (1938) to connect Plato and Ar. on this point is misguided, and involves him in the claim (194) that the proper or peculiar pleasure of the Poetics is not specific to individual genres: this will not stand up to scrutiny.

²⁶ The basis of the portentous, as GA 770b 8ff. helps us to see, would be unnatural phenomena (grotesque horrors), which means that they would flout necessity or probability: hence their spuriousness, in contrast to the intelligibility of events which cause pity and fear. For the possibility of a reference to Aeschylus at 53b 9 see ch. III n.20.

²⁷ The use of 'aesthetic' by Lucas (1968) on 48b 13 and 17 to refer only to this type is question-begging, and is rightly criticised as such by M. Hubbard, CR 20 (1970) 177. On the distinction between purely sensual and cognitive pleasure, parallel to Poet. 48b 10-19, see PA 645a 7-15, and cf. e.g. EN 1175a 26f., 1176a 2f. This distinction is related to that between the sensible and the knowable, e.g. De An. 431b 21ff.

²⁸ For the translation 'garnishings' see ch. VIII n.3. The comparison of characterisation to colour at 50a 39ff. ((e) in the text), which effectively reiterates the emphasis of 48b 15-19, does not, of course, imply that character is sensually apprehended, only that it needs the formal framework of plot-structure to have its proper significance.

²⁹ For general references to musical pleasure see EN 1173b 30f., 1175a 13f., 34f., EE 1230b 27f., Pol. 1339b 20f., 1340b 16f. The mimetic nature of music is indicated at Poet. 47a 14-16, 23-6, Pol. 1340a 28ff. (and cf. 1340a 13f. with Susemihl's emendation: see Anderson 126, 186-8). The equestion arises how far Ar. allowed a sensual pleasure to musical tones independently of their mimetic significance, as Plato does at Phileb. 51d. Poet. 47a 15 implies that some music is not mimetic, but the adverb 'most vividly' at 62a 17 implies a mimetic force (cf. the same word at 55a 24). The general impression left by Politics 8 is of music as a naturally mimetic art. The point affects rhythm too, for which some degree of intrinsically mimetic or expressive value is posited at Poet. 49a 21ff., 59b 31ff., 60a 4, Pol. 1340a 19ff., b 8ff., 1341b 19-27, Rhet. 1408b 21-9a 23: compare e.g. Plato Laws 669c, 798d (but note 669e on the difficulty of understanding rhythm and music without words) Isocrates 9.10-11 appears to treat the pleasure of rhythm as autonomous. Finally, Pol. 1341a 14f. distinguishes between the 'common' and evidently sensual pleasure of music and rhythm (available even to some animals) and the sensibility to enjoy 'beautiful melodies and rhythms', which probably entails a mimetic significance.

³⁰ Ar. also refers to epic recitation at 59b 30. Note that Plato describes poetry as the form of mimesis which operates through our hearing at Rep. 603b 6f. On the Greek practice of reading aloud see B. Knox, GRBS 9 (1968) 421-35.

³¹Met. 980a 21-7. On the general love of knowledge cf. Protr. B17, 72-7, 97-102 Düring (1961).

³² Ross 280 calls ch. 4's explanation of the mimesis-instinct 'too intellectualistic' and misses its significance. Lord (1982) 90f. talks of a 'purely intellectual' understanding of mimesis, and Else (1957) 128-30 finds the 'emphasis on the intellect' a sign of a marginal addition (cf. ch. I n. 51 above). These judgements all seem to me to misconstrue the implications of the passage. The essential import is that the basic cognitive pleasure takes particular forms, such as that from tragedy, in which it is engaged with emotion and other factors. For a recent philosophical attempt to use Poetics 4 as a starting-point for an aesthetic thesis see A. Savile, The Test of Time (Oxford 1982) ch. 5, esp. 86f., 95f.

³³ It is specifically the simplicity of the passage which has led scholars such as Twining 186-91, Butcher 201f., Rostagni (1945) LXIV, and Lord (1982) 90-2 into underestimating its importance: against Rostagni and Lord it is particularly pertinent to observe that the educative effect of poetry on the passions (for which they rightly argue: cf. ch. VI n. 40) cannot be divorced from the cognitive experience of the mimetic structure of a poem (see esp. 53b 12 for an indication of this).

³⁴ The obvious parallel is Pol. 1336a 32-4, where children's games are said to be imitations or enactments of adult activities and to 'prepare the way for their later occupations'. See Plato Rep. 395d If., and cf. ps.-Ar. Probl. 956a 14.

³⁵ There has been much disagreement on this point (and the issue is old: cf. Weinberg 462). I take mimeisthai at 48b 20 to refer to both causes, and the instinct for melody and rhythm to be an additional but closely related factor (a point usually overlooked) in view of Ar.'s understanding of these things as naturally mimetic (n.29 above).

³⁶ That this latter, purely sensual pleasure could be for Ar. 'the true pleasure', as Gomme 64f. suggests, makes nonsense of the treatise and flatly contradicts the reference to mimesis in the definition of tragic pleasure at 53b 11-13. Pol. 1340a 25-8 provides an important parallel: there the form (morphê) is an attribute of the human subject of the image (cf. Poet. 54b 10), and the pleasure derived from it is cognitive (the recognition of physical strength, athletic beauty, etc.) and dependent on mimesis; the reference to 'another reason' is parallel to Poet. 48b 18f. The difficult passage, Mem. 450b 21ff., is also germane: for the interpretation see R. Sorabji, Aristotle on Memory (London 1972) 84. It confirms that the proper appreciation of visual mimesis involves treating a work as an image of a possible reality (for an image, eikôn, is precisely the product of mimesis: Top. 140a 14f.). It might be possible to discern allusions to cognitive and non-cognitive pleasure in visual art at, respectively, EE 1230b 31-4, EN 1118a 3f.

³⁷Rhet. 1371b 4ff.: the relevance of the passage is correctly indicated by Hubbard 134 (though the word 'just' does not belong in the translation), in line with her brief but exemplary explanation of the theory of aesthetic pleasure on pp. 86f. Cf. also Tracy's article, Redfield 52-66, and Goldschmidt (1982) 212-17. Closely related passages are those on metaphor (Poet. 59a 7f., Rhet. 1410b 10-26), on 'wit', including types of metaphor (Rhet. 1412a 17-b 23), and on other kinds of comparison (Rhet. 1394a 5). It is regrettable that an emphasis on the importance of Poetics 4's explanation of aesthetic pleasure has been confused with the issue of katharsis in Golden (1962) and later articles: see Appendix 5 §5(a).

³⁸Manthanein leads to knowledge: e.g. De An. 417b 12f., Rhet. 1362a 30f., 1363b 31f. For sullogizesthai cf. Poet. 55a 7, 10f., 61b 2, Rhet. 1371b 9. Although this latter term should not everywhere be pressed in its technical logical sense (cf. Appendix I p. 326), the language of Poet. 48b 16 allows for a process in which new understanding may be reached (cf. e.g. Plato Euthyd. 277e-8a). See n. 42 below.

³⁹ The point is reinforced by the reference to non-philosophers at 48b 13f.: the following example is intentionally rudimentary, with the implication that more sophisticated forms are possible. Cf. the parallel drawn between the study of art (dramatic performances) and philosophical study at Protr. B44 Düring (1961).

⁴⁰Rhet. 1371a-b, cf. 1404b 12, Met. 982b 12ff., 983a 12ff., Plato Theaet. 155d. Ar.'s 'wonder' all too easily became admiratio, hence admiration, for neo-classical readers of the Poetics: cf. ch. X n.21. Ar.'s and Plato's views on wonder deserve closer study. For a sensitive general essay on the subject see R. Hepburn, Wonder and Other Essays (Edinburgh 1984) 131-54.

⁴¹ It is true that at 60a 11ff. 'the wonderful' is related to the irrational, which is undesirable in poetry. Although there may be something of a tension here, it is reasonable to infer that there are degrees of wonder, which lies on the boundary of the explicable and the inexplicable, and so can slip into the latter (and hence become the irrational) or, properly used, may stimulate and challenge understanding, as at 52a 4ff. Cf. Else (1957) 624f., and ch. VII n. 16.

⁴² In the full understanding of universals in poetry it is not easy to know what would correspond to the preexisting knowledge of Poet. 48b 17. Certainly not factual acquaintance with myths: see 51b 21-6. Perhaps knowledge (which for Ar. is not true knowledge) of particulars, from which we are led to a grasp of the universals embodied in the poetry: cf. Top. 108b 10-12 for the movement from particulars to universals through 'likenesses'. For the movement from existing to new knowledge in a different context see An.Post. 71a 1ff.

⁴³Met. 982b 12ff., Rhet. 1371b 27f.

⁴⁴ It does not seem to have been accepted, for instance, by Ar.'s successor, Theophrastus: see fr. 65, translated in Russell 203f. On Plato's comparisons between poetry and philosophy see Appendix 2 under 51a 5f.

⁴⁵ Else (1957) 131f. can hardly be right to see a reference to universals in the visual example of ch. 4, but this is tied up with his belief (128) that the point concerns scientific models and diagrams, a belief refuted by Poet. 48b 17-19 and the parallel passage, Rhet. 1371b 4ff. (On Ar.'s independent use of diagrams cf. ch. VII p. 218 and n. 23.)

⁴⁶ E.g. Twining 399-401, Butcher ch. 4, Gudeman 99, Allan (1970) 155, Schadewaldt 225, 228f. Cf. ch. I p.6. For the comparison between poetry and philosophy there is an analogue in the affinities allowed at Rhet. 1356a 20-7 between rhetoric and dialectic (and politikê), and at Rhet. 1359b 8-12, An.Post. 71a 9-11 between rhetoric and logic, ethics, and dialectic.

Abbreviations

1. Aristotelian works

An.Post.

Posterior Analytics

An.Pr.

Prior Analytics

Cael.

De Caelo

Cat.

Categories

De An.

De Anima

EE

Eudemian Ethics

EN

Nicomachean Ethics

GA

De Generatione Animalium

GC

De Generatione et Corruptione

HA

De Historia Animalium

Int.

De Interpretatione

MA

De Motu Animalium

Mem.

De Memoria

Met.

Metaphysics

Meteor.

Meteorologica

PA

De Partibus Animalium

Phys.

Physics

Poet.

Poetics

Pol.

Politics

Protr.

Protrepticus

Rhet.

Rhetoric

SE

Sophistici Elenchi

Top.

Topics

2. Pseudo-Aristotelian works (prefaced by ps.-Ar.)

Aud.

De Audibilibus

MM

Magna Moralia

Mund.

De Mundo

Probl.

Problemata

Rh.Alex.

Rhetorica ad Alexandrum

3. Journals

CJ

Classical Journal

CL

Comparative Literature

CP

Classical Philology

CQ

Classical Quarterly

CR

Classical Review

CW

Classical World

G & R

Greece & Rome

GRBS

Greek Roman and Byzantine Studies

HSCP

Harvard Studies in Classical Philology

JAAC

Journal of Aesthetics and Art Criticism

JEGP

Journal of English and Ger-manic Philology

JHI

Journal of the History of Ideas

JHS

Journal of Hellenic Studies

JP

Journal of Philology

MH

Museum Helveticum

MLR

Modern Language Review

Mnem.

Mnemosyne

MP

Modern Philology

PCPS

Proceedings of the Cambridge Philological Society

Philol.

Philologus

SP

Studies in Philology

SR

Studies in the Renaissance

TAPA

Transactions of the American Philological Association