¹ I am indebted to the American Council of Learned Societies for their grant in 1963-64 during which time I began my research and to Stanford University for supplementing this grant at that time and, subsequently, for allowing me time for continuing my work. The two citations immediately following are from The Works of Francis Bacon, ed. Spedding, Ellis, and Heath, (London 1857-74) 4.92-3, and Seneca's Epistulae Morales, trans. Gummere, R. M. (LCL, London 1961) Ep. 65.Google Scholar

² In his Rhetoric (3.1; 1403b36f.) Aristotle observes that the study of language and style had only recently made much progress. Aristotle's Rhetoric and Poetics, trans. ( Rhys Roberts, W. and Bywater, I. (New York 1954) 165. All references to the Rhetoric will be to this edition.Google Scholar

³ For a subtle survey of such reformulations, see McKeon, McKeon, ‘The Philosophical Bases of Art and Criticism,’ Critics and Criticism (Chicago 1952) 463–545.Google Scholar

⁴ These distinctions were given their most influential formulation by Aristotle (Ethics, VI.iv-v) and were adapted from him by St. Thomas Aquinas (S. T., I, II qu. 21 ar. 2 and qu. 57 ar. 3 and 5), from whom I have borrowed the example of the knife. For a perceptively detailed account of their origin, development, and extension into the controversy over the contemplative and active life, see Jaeger, Werner, Aristotle, trans. Robinson, R. (Oxford 1948) 426–61.Google Scholar

⁵ McKeon implies that often attempts to resolve the antithesis account for the shifting presuppositions in statements about the poetic arts. He states the possibility of resolution most generally in discussing Aristotle's division of the sciences: ‘Poetic “science” differs from theoretic and practical sciences, for it is concerned neither with knowledge as such nor with action but with artificial objects and products; and if such objects are to be isolated for consideration in themselves, there must be some preliminary consideration of the conditions of their production and some supplementary consideration of the effects of their contemplation,’ op. cit. 517. In such considerations the literary theorist will have to draw upon material proper to the theoretic and prudential sciences. Google Scholar

⁶ For descriptions of a culture which would be congenial to such a development, see von Armin, Hans, Leben und Werke des Dio von Prusa (Berlin 1898) 1–114; Werner Jaeger, Paideia: The Ideals of Greek Culture, trans. Highet, G., 3 vols. (New York 1944); Marrou, H. I., Histoire de l'éducation dans l'antiquité (Paris 1960). The cultural ideal itself, even when most nearly approximated, was instable. Leeman, A. D. comments that ‘the boundary between philosophy and oratory, where Cicero lodges his orator perfectus, is a no man's land.’ He cites Cicero's letter to Atticus (9.4) in which the orator lists eight , which, dealing with political matters, were applicable to his own circumstances. By debating them as exercises he both profits and relieves his melancholy. Orationis Ratio, The Stylistic Theories and Practice of Roman Orators, Historians and Philosophers (Amsterdam 1963) 1.123-4. This application of general themes to particular, personal events, thesis to hypothesis, reflects a desire to repossess the no man's land. Only literary discourse of the possibility.Google Scholar

⁷ This is even partially reflected in the subsequent history of the two principal definitions of poetry, metrical composition and imitation (mimesis). The first states its relation to the prose composition of oratory and the second, in being concerned with the object and representation of certain types of knowledge, to philosophy. Google Scholar

⁸ Isocrates, trans. Norlin, G. and van Hook, L., 3 vols. ( LCL London 1928-45). All references to Isocrates will be to this edition.Google Scholar

⁹ This imposition is later revealed in Hermagoras' development of his system of status, which, perhaps as was characteristic of the Hellenistic period generally, led to an almost exclusive concern with forensic oratory. The victory of his system over Aristotle's is a clear example of the more specialized form of an art overcoming the less. Cicero's ultimate liberalization of rhetoric was effected by a return to Aristotelian and Isocratean traditions. See Solmsen, F., ‘The Aristotelian Tradition in Ancient Rhetoric,’ AJP 62 (1941) 180, 189-90. For Hermagoras see Matthes, D., ‘Hermagoras von Temnos 1904-55,’ Lustrum 3 (1958) 58-214.Google Scholar

¹⁰ See Norlin's note: ‘The distinction usually drawn, in Plato for instance, between and , the one “opinion,” the other “knowledge,” is not exactly that made by Isocrates. is here, not irresponsible opinion, but a working theory based on practical experience — judgement or insight in dealing with the uncertain contingencies of any human situation which presents itself. In this realm, he holds, there can be no exact science.’ Google Scholar

¹¹ Op. cit. 134–6. For general accounts of Isocrates' cultural and intellectual development see Jaeger, , Paideia, 3.46-155 and Marrou, , op. cit. 121-36. For his subsequent influence, see Marrou, passim, and Hubbell, H. M., The Influence of Isocrates on Cicero, Dionysius and Aristides (New Haven 1913 ). Von Armin describes the persistence of sophistic ideals — despite Plato's and Aristotle's attempt to break with them — even within the ‘philosophical’ schools, op. cit. 63-7. Grube, G. M. A., The Greek and Roman Critics (Toronto 1965), comments very generally with regard to poetry on how ‘the two approaches, the philosophical and the rhetorical, continue side by side through the fourth century,’ how Isocrates combined them in his concept of general education (37-40), how Aristotle ‘may be said to have brought them together in the Rhetoric' (102), and on how Cicero later defends general culture ‘in an age of over-specialization’ (171-5). Grube's conclusion that ‘neither Plato nor Aristotle even attempt to define the nature of poetry’ (102) is perplexing in view of the subjects he discusses in relation to them. Whatever he considers the ‘nature of poetry’ to be, he seems to prefer the Longinian treatise, which he claims (surprisingly to me) ‘needs little explanation of its ideas because Longinus, more than any other ancient critic, speaks a language that the modern reader can understand without intermediary’ (353).Google Scholar

¹² Solmsen, F., Die Entwicklung der Aristotelischen Logik und Rhetorik (Berlin 1929) remarks: ‘Wie die Form des Sokratesdialoges und die paradigmatische Sokratesgestalt das ästhetische Korrelat der platonischen -Konzeption sind, so sind die Pragmatien als das Korrelat zum aristotelischen -Begriff und die Entwicklung vom Dialog zur Pragmatie läuft der vom zum bis in Einzelheiten hinein parallel’ (86). That such a correlation might be applicable to the Poetics and its doctrine of the universal is suggested by a later remark made by Solmsen: ‘I should not hesitate to say that the whole body of thought contained in chs. VII-IX has originated in the application of the Platonic to the phenomenon of poetry …,’ ‘The Origins and Methods of Aristotle's Poetics,’ CQ, 29 (1935) 198.Google Scholar

¹³ Isocrates, frankly distrusting any system developed to handle all occasions, had nothing comparable to Aristotle's proofs to offer to the orator. Though he probably confined the appeal to the emotions of the audience principally to the proem and epilogue, he would have appreciated Aristotle's great emphasis upon their manipulation, and he did insist with Aristotle upon the importance of the speaker's reputed character. Also, he would have applauded Aristotle's intention to supply a method for political oratory which the earlier manuals had slighted in the interest of forensic debate ( v. Rhet. 1354b22ff.; cf. 1418a21-36).Google Scholar

¹⁴ The Collected Dialogues of Plato, ed. Hamilton, E. and Cairns, H. (New York 1963) 223.Google Scholar

¹⁵ See Armin, Armin, op. cit. 67–70. Solmsen, F., ‘The Aristotelian Tradition …,’ 39-41, describes how Aristotle makes the topoi into ‘forms’ of argument rather than specific bodies of material. These forms are more abstract and general () than any achieved by previous rhetorical systems. For Cicero, as well, ‘they are not connected with any definite subjectmatter, and yet they are applicable to every subject’ (173). For topoi in medieval see Caplan, H., CP, 28 ( 1933).Google Scholar

¹⁶ Solmsen, F., ‘Aristotle and Cicero on the Orator's Playing upon the Feelings,’ CP, 33 (1938) 402–4, discusses Plato's conception of ‘types of soul’ () as fixed, while Aristotle seems to see them as varying in accordance with the psychological states which rhetoric the orator can bring about in his audience.Google Scholar

¹⁷ These generalizations with regard to discourse are consistent with Aristotle's intellectual development. Jaeger describes how in the early Protrepticus Aristotle accepted Plato's view of ethics and politics as theoretical sciences ‘proceeding more geometrico’ and then later abandoned the ideal of mathematical exactness in his Ethics and Politics ( Aristotle, 82–8). With regard to the more specialized form of philosophical discourse (i.e., the more general definition), Jaeger says ‘Aristotle here [cf. Eth. Nic. II. 7, 1107a29] replies that the more general ethical propositions are the more empty and ineffective they are’ (85). The movement toward greater particular application in ethics corresponds to one toward unspecialized philosophical discourse. In his later period, this movement is reflected in his researches where ‘the individual is now almost an end in itself’ (328). Despite this, Jaeger stresses that he avoided the fragmentary Hellenistic antiquarianism, because, perhaps as a result of his Platonic inheritance, his method consisted of ‘applying the principle of form to the details of reality, the idea of the uniformity of nature' (328). His aim was ‘all along to make the Idea capable of producing knowledge of appearances’ (p. 381). He lived ‘not in the Ideal world but in the tension between Idea and experience’ (399).Google Scholar

¹⁸ Lucian, trans. Harmon, A. M., LCL, 8 vols. (London 1921) 3.109. For a similar defense of his attempt to combine philosophy with comedy, see To One Who Said ‘You're a Prometheus in Words,’ 6.417-27.Google Scholar

¹⁹ The editor notes that these are allusions to Aristophanes' The Clouds (line 225) and to Plato's Phaedrus, 246e and 247b. Google Scholar

²⁰ Sextus Empiricus , trans. Bury, R. G., LCL, 4 vols. (Cambridge 1933-59) 4. 245-7. Sextus lists the third usage first. It is usually rendered in Latin as materia or argumentum. Prefatory summaries and outlines of any work were called ‘hypotheses,’ and this usage persisted through the Renaissance.Google Scholar

²¹ ‘Against the Grammarians,’ 263, Sextus, 4.149. Sextus is repeating here the Hellenistic distinction between the historical narrative of true events (), the narration of things like truth (), and the legend which has no relation to truth at all ().Google Scholar

²² Institutio Oratoria , trans. Butler, H. E., LCL, 4 vols. (Cambridge 1953). As a commentary on this passage, see J. Cousin, Études sur Quintilien (Paris 1936) 1.173-5. The most conservative form of rhetoric, the quaestio definita, represented the principle of individualization as well as history (cf. Minturno, paraphrasing Aristotle: ‘Onda il Poeta a guisa di Filosofo riduce la cosa al genere, ed alla natura universale; l'Istorico, sicome l'Oratore, quando tratta le cause, al particolare descende.’ L'Arte Poetica (1563) [repr. in Naples 1725] 39.).Google Scholar

²³ Turnebus, Adrianus comments on Quintilian's description of this ‘middle ground’ (quoted from M. Fabii Quintiliani de Institutio Oratoria … per Petrum Burmanum, Ludg. Bat. 1720, 232): ‘Medium quoddam genus inter thesin & hypothesin ponere videtur, nulla singulari definitaque persona conclusum, ad aliquid tamen referri, id est latentis temporis vel personae angustiis contineri, ut, an resp. administranda in tyrannide.’ This middle genus is of great importance to fiction, since it was recognized in the casuistic exercise of even the elementary progymnasmata called the ‘proposal of a law.’ Aphthonius concludes his exercises with it: ‘Furthermore, some have considered the proposal of the law to be an exercise, for it is a more or less complete hypothesis, but it does not meet all the requirements of the hypothesis. For a person is introduced in it, but not one well known to all; in this respect, it is more than a thesis but less than a hypothesis. In the ways in which it generally brings in the appearance of a person, it goes beyond the thesis; in the same ways, it does not clearly present the circumstances and, therefore, falls short of a hypothesis’ (‘The Progymnasmata of Aphthonius,’ trans. Nardeau, R., Speech Monographs, 19 [ 1952] 283). The use of fictions in law is discussed by Hans Vaihinger, The Philosophy of ‘As if’, trans. Ogden, C. K. (London 1924) 33-35, and more recently by Kantorowicz, E., ‘The Sovereignty of the Artist: a Note on Legal Maxims and Renaissance Theories of Art,’ Selected Studies (New York 1965) 352-65. Kantorowicz comments that ‘Fiction was rather something artfully “created” by the art of the jurist; it was an achievement to his credit because fiction made manifest certain legal consequences, which had been hidden before or which by nature did not exist. For by fiction the jurist could create (so to speak, from nothing) a legal person, a persona ficta — a corporation, for example — and endow it with a truth and a life of its own; or he could interpret an existing body, such as the Corpus mysticum of the Church, in the sense of a fictitious person, and gain a heuristic element by means of which he might arrive at new insights into administration, property rights, and other conditions’ (355). So Sidney, arguing the right of the poet to use fictional names to give ‘a conceite of an actuall truth,’ asks in An Apology for Poetry: ‘And doth the Lawyer lye then, when vnder the names of Iohn a stile and Iohn a noakes hee puts his case? But that is easily answered. Theyr naming of men is but to make theyr picture the more liuely, and not to builde any historie; paynting men, they cannot leaue men namelesse. We see we cannot play at Chesse but wee must giue names to our Chesse-men; and yet, mee thinks, hee were a very partiall Champion of truth that would say we lyed for giuing a peece of wood the reuerend title of a Bishop' (Elizabethan Critical Essays, ed. Smith, G. G., [Oxford 1904] 1.185-6). This discussion of hypothetical names is relevant to of Aristotle's Poetics: see below 41-42.

²⁴ Turnebus observes how these two movements can function together: ‘Theses, ut docuit Fabius, pleraeque ad deliberativum pertinent genus, & adjectis modo personis, suasoriae fiunt. nam deliberationis initium ducitur ab hypothesi, sed postea revocatur ad thesin: atqueita interdum de toto genere disceptamus, cum tamen causa pendeat ab hypothesi' (loc. cit.). This bears upon the discussion of ‘thesis,’ 49-55 below. Google Scholar

²⁵ This exclusion may be a reflection of the old controversy of the contemplative vs. the active life. Jaeger observes that Dicaerachus, in opposition to Aristotle and Theophrastus, held the active life to be superior to the contemplative: he ‘severed the connexions … between moral action and the knowledge of the highest questions, and reached the logical conclusion of which we hear the echo in the author of the Great Ethics: “One must wonder what sophia has to do with ethics,” since the latter concerns character and action' ( Aristotle, 451 ). In excluding the thesis from rhetoric, the young Cicero is arguing for the ‘active’ life of the orator in the most pragmatic sense. Later he emphasized greatly the role of thesis or philosophical question in oratory, and Jaeger, in citing his Ep. ad Att., 2.16, shows how he has left Dicaearchus, representing the active, pragmatic life, for Theophrastus, representing the contemplative, scholastic life. For the place of thesis and hypothesis in the attempts in the second and first centuries B. C. to establish the sophistic ideal of the philosophical and rhetorical disciplines perfectly combined in the person of the rhetorician, see von Armin, , op. cit., pp. 92-114 and Matthes, D., Lustrum, 3 ( 1958).Google Scholar

²⁶ Quintilian says (5.10.95-99) ‘that arguments are drawn not merely from admitted facts, but from fictitious suppositions, which the Greeks style ’ ’ and that such ‘suppositions are also exceedingly useful when we are concerned with the quality of an act (verum eadem fictio valet et ad qualitates).’ Quality introduces a generic consideration into a particular case and fictional hypotheses are useful where quality is concerned: the implication is that fiction may be a good means of investigating the generic implication of any particular situation. These passages from Quintilian should be referred to the discussion of fiction below, pp. 51-55. For Hermagoras' analysis of ‘quality,’ see Matthes, D., Lustrum, 3 (1958) 147–64. Kantorowicz describes how the civil law ‘imitates’ the natural law as the artist imitates nature for medieval jurists. ‘The art of the legislator, however, though determined by the general natural law, has to “adinvent” the particulare of the positive law (“Ius positivum … est per industriam hominum adinventum”) — that is, the particular application of the general law of nature to a limited space and a limited time — yet in such a fashion that the particulare still reflected the generale of the law of nature' (op. cit. 355-6). The invention of the particulare was often of a fiction. See Appendix B.Google Scholar

²⁷ See Reichel, G., Quaestiones Progymnasmaticae (Lipsiae 1909) 62: ‘Cuius partes sunt sex, quibus adiectis in mutatur vel, ut planius dicam, quaestio infinita fit causa finita. Hermagora quoque teste sine circumstantia “ulla omnino controversia non potest esse,” nam circumstantia est, “quod hypothesin, id est controversiam efficiat.”’ Reichel is quoting St. Augustine, De Rhetorica in Rhetores Latini Minores, ed. Halm, C. (Lipsiae 1863) 141. For a helpful account of the early development of theses and hypotheses, particularly as a background for later rhetorical practice, see Reichel, 10-11, 27, 97-107; Matthes, D., Lustrum, 3 (1958); and Bonner, S. F., Roman Declamation (Berkeley 1949) 1-26.Google Scholar

²⁸ Quintilian testifies that Isocrates used ‘hypothesis’ in this sense (3.5.18). The ‘middle areas’ between thesis and hypothesis were also reflected in Hermagoras. Matthes, D. points out that a thesis about matricide in general is not totally lacking in personae: ‘ihr liegt vielmehr ebenso wie der ein zugrunde, jedoch im Gegensatz zu dieser, bei der es sich um ein () handelt, ein (). Danach kann man sagen, dass der Hauptunterschied zwischen und in dem bzw. zu suchen ist,’ op. cit. 126. Matthes adds that ‘In diesem Sinne unterscheidet schon Aristoteles das und das () , ohne dass bei ihm in diesem Zusammenhange schon die Bezeichnungen und gebraucht wären’ (126, n. 1). One of the examples of he cites from the Poetics (1451b6).Google Scholar

²⁹ The word ‘thesis’ means a ‘stand’ or ‘position’ and is translated positio; ‘hypothesis,’ whose Latin form is suppositio (sub+ponere), means a ‘placing under,’ and, by extension, what one has placed under,’ or a ‘subtending.’ In the rhetorical usage, where hypothesis is considered a species of the genus thesis and refers to questions about actions and events, the ‘sub' carries the meaning of being placed under a more inclusive heading and of sharing characteristics, indicated by that heading, with other species. In the philosophical and scientific usage of ‘assumption,’ the hypothesis metaphorically ‘subtends’ in the sense of lying behind or below whatever we wish to investigate or demonstrate. We come first to the problem, i.e. the given ‘situation’ to be explained, and then formulate hypotheses in accordance with which the given situation may be accounted for and now regarded as a ‘conclusion.’ Google Scholar

³⁰ In ‘ Against the Rhetoricians’ (60–71), Sextus attacks the aim of persuasion by defining ‘credible’ in three senses: belief in something true, in something like truth (verisimilar), and in something both true and false. The second sense (the probable) is a belief in what is false but which we are deceived into believing in; its validity is no greater than simple error, and it is more dangerous than error. For further attacks on the probable, see ‘Against the Logicians,’ 1.166-89, and ‘Outlines of Pyrrhonism,’ 1.226-31.Google Scholar

³¹ Philebus, trans. Fowler, H. N., LCL (Cambridge 1952). For the aesthetic response to mathematical constructs in relation to modern formalism, see Appendix C.Google Scholar

³² The Republic , trans. Shorey, P., 2 vols., LCL (Cambridge 1946 ). This translation is used in the text unless otherwise indicated. For the development of Plato's hypothetical method, see Robinson, R., Plato's Earlier Dialectic (Ithaca 1941) esp. 97-191.Google Scholar

³³ Jaeger comments that because mathematical arts ‘abstract the truth from sensible objects and try to see the essence of the mathematical objects … with the eyes of the mind, they are very close to the highest philosophical methods of reaching knowledge. But on the other hand they are tied to the world of sense and the stage of knowledge which is appropriate to it (i.e. opinion) in two ways: (1) they start with hypotheses built around sensible figures, although their theorems do not really concern the visual images at all; (2) they do not attempt, in principle, to rise above these hypotheses which are taken as true (‘adopted’), and because they follow them out logically right to the last possible deduction, they are forced to treat them as principles () at the same time' (Paideia, 2.289-90). Google Scholar

³⁴ Euthydemus , trans. Lamb, W. R. M., LCL (London 1924).Google Scholar

³⁵ Phaedo , trans. Fowler, H. N., LCL (London 1913).Google Scholar

³⁶ Burnet, Burnet, Greek Philosophy: Thales to Plato (London 1924) 344. For the historical development of the idea of spatial ‘location’ in the memorial and imaginative faculties, see Francis A. Yates, The Art of Memory (Chicago 1966). For the many variations of the idea of the imagination as an intermediary between the sensible and intelligible worlds, see Bundy, M. W., The Theory of Imagination in Classical and Mediaeval Thought, Univ. of Ill. Sts. in Lang, and Lit. 12 (Urbana 1927) esp. 45, 96f., 117ff., 148-50, 158-9, 170-71, and de Bruyne, Edgar, Études d'esthétique médiévale (Brugge 1946), esp. 2. ch. 5: ‘L'esthétique des Victorins.’ For geometry as a discipline of this intermediate location, see The Philosophical and Mathematical Commentaries of Proclus on the First Book of Euclid's Elements, trans. Taylor, T. (London 1792). The translator claims geometry is ‘the genuine passage to true theology, and the vestibule of divinity’ for those ‘who look beyond sense for certainty' (Pref.). He translates the title of Proclus' first chapter as ‘On the Middle Nature of the Mathematical Essence,’ since ‘mathematical natures, and whatever falls under cogitation, are allotted a middle order’ between intelligible and sensible natures (47). Proclus considers the objects of mathematics in the first chapter of his second book as participating in both the universal and the particular through the medium of the phantasy. For St. Augustine, as well, both the ‘suppositions’ of literary fictions and of geometrical diagrams belong to the province of the imagination (see n. 68).Google Scholar

³⁷ On the disagreement between Plato and Aristotle concerning the independent existence of such intermediate constructs, see Ross, W. D., Aristotle's Metaphysics (Oxford 1958) 1.166-8. For a general summary of the problem, see Ross' Aristotle (London 1966) 157-9.Google Scholar

³⁸ Particularly useful are the studies of Cornford, F. M., Lee, H. D. P., Einarson, B., and Robinson, R. referred to below. I have found Cornford's essay, ‘Mathematics and Dialectic in the Republic VI-VII,’ Mind 41, ( 1932) 37–52, 173-90, the most valuable. Whether or not one accepts his central thesis explaining Proclus' comments on Platonic dialectic, which Robinson does not (Mind, 45.464-73), the implications for fiction in the material he discusses remain forceful. Page references to his essay are given in my text.Google Scholar

³⁹ Cornfoıd explains the metaphor as that of climbing a stair. ‘The primary and common meaning of is “impulse” or “effort” or “impetus.” It is nearer to “spring” than to “spring-board,”’ 42 n. 2. For various interpretations of this ‘upward’ movement, see Robinson, Plato's Earlier Dialectic, 162-91. Google Scholar

⁴⁰ Cornford offers an illustration of going through these steps after which the geometer will ‘frame his demonstration in full discursive form — a deduction starting from the hypothesis, “ Let there be a triangle ABC” (Eucl. I,. 32),’ 45.Google Scholar

⁴¹ The fragmentary nature of dianoia is also reflected in Aristotle's caution against thinking of drama as a string of ‘speeches expressive of character and well-turned expressions () and arguments ().’ However coherent in themselves, isolated speeches cannot be stitched together to make a unified play (Poetics, 1450a28-37). The unity lies in the plot, in the first principle or of tragedy, that by virtue of which something becomes comprehensible (Meta., 5.2; 1013a17-8). Plato (Phaedrus, 268c-e) and Isocrates (Against the Sophists, 16-7) make similar observations. Google Scholar

⁴² Translated by Randall, J. H. in his Aristotle (New York 1965) 96. Aristotle gives a similar argument in De Anima III. 7; 431b5-20. Ross comments that ‘Aristotle seems here to be setting himself against Plato's view, expressed in the Divided Line…,’ Aristotle, 148.Google Scholar

⁴³ Of Aristotle's admiration for Homer's powers of visualization, Else, G. F. comments that ‘the poet's abstract conception, the , is translated into words’ by the imagination. Aristotle's Poetics: The Argument (Cambridge 1963) 502. In objecting to the interpretation of Aristotle's view of poetry as a representational ‘copy’ of life, Crane, R. S. stresses the psychological nature of the object itself which is imitated: ‘The object imitated is internal and hence strictly “poetic” in the sense that it exists only as the intelligible and moving pattern of incidents, states of feeling, or images which the poet has constructed in the sequence of his words by analogy with some pattern of human experience such as men have either known or believed possible or at least thought of as something that ought to be,’ The Language of Criticism and the Struture of Poetry (Toronto 1953) 56. Such a psychological ‘internalizing’ of the objects of imitation, however, suggests a later context of ideas. In a way similar, for instance, to that in which the Neoplatonic ideal form, no longer solely external and absolute, becomes a mental image to be imitated by the artist (for this development, see Panofsky, E., Idea, trans. Peake, J., [Columbia, S. C. 1968]), Philostratus in his Life of Apollonius of Tyana thinks of even imitations, to say nothing of a combinative phantasy, as an act of mental construction itself, imaging ‘its ideal only on the analogy of reality ()’ which was then embodied in a work of art. Bundy, who translates the preceding phrase, suggests that ‘Apollonius means by “imitation” what Quintilian and Longinus meant by “phantasy,”’ (op. cit. 113 f.). In the same passage the use of ‘hypothesize’ suggests the extension of the terms of this essay to the history of art: Philostratus argues the superiority of the imagination to the faculty of imitation in sculpture since ‘imitation can only create as its handiwork what it has seen, but imagination equally what it has not seen, for it will conceive () of its ideal with reference to the reality (), and imitation is often baffled by terror, but imagination by nothing; for it marches undismayed to the goal which it has itself laid down (),’ Philostratus The Life of Apollonius of Tyana, VI. 19, trans. Conybeare, F. C., LCL (London 1912). In addition to this ‘philosophical’ or ‘geometrical’ usage, the term ‘hypothesis’ was applied to the subject or sketch of a painting as well as to the plot of a drama. Plotinus uses the word for the ordered disposition of form and color, perhaps as abstracted by the artist from his model, in a painting (see de Keyser, E., La signification de l'art dans les Ennéades de Plotin, [Louvain 1955] 37-9).Google Scholar

⁴⁴ Else observes that ‘in Aristotle's theory of vision the size of the thing seen and the time required to see it are interconnected. Magnitude, motion, and time are strictly correlative: Phys. 4.11.219a10; Ibid. 12.220b15; 6.2.233a10…. Hence there is an “imperceptible time” corresponding to the imperceptible magnitude' (285 n. 10). See n. 36 above. The ‘organic’ comparison is Plato's (Phaedrus, 264c). Google Scholar

⁴⁵ Prior Analytics , trans. Tredennick., Tredennick. LCL (Cambridge 1938). It is precisely these ‘non-existent’ hypothetical concepts that Sextus attacks.Google Scholar

⁴⁶ Posterior Analytics , trans. Tredennick, H., LCL (Cambridge 1960 ). Cicero offers an early defense of fiction as illustrative example in his De Officiis (3.39) which is analogous to this justification of geometric diagrams. He is defending Plato's use of the story of Gyges' ring against certain philosophers who reject the argument because the story itself could not be true: ‘As if he affirmed that it was actually true or even possible!’ Cicero exclaims. ‘They press their point with right boorish obstinacy: they assert that it is impossible and insist upon it; they refuse to see the meaning of my words, “if possible.” For when we ask what they would do, if they could escape detection, we are not asking whether they can escape detection; but we put them as it were upon the rack: should they answer that, if impunity were assured, they would do what was most to their selfish interest, that would be a confession that they are criminally minded; should they say that they would not do so, they would be granting that all things in and of themselves immoral should be avoided.’ De Officiis, trans. Miller, W., LCL (London 1961).Google Scholar

⁴⁷ John, Burnet defends this extreme rendering of ( op. cit. 228–30).Google Scholar

⁴⁸ ‘On Certain Mathematical Terms in Aristotle's Logic,’ AJP, 57 (1936) 33–54, 151-72. Page references to this essay are given in my text. Einarson comments that the words and were synonymous with the phrase (40). That the phrase was used for ‘begging the question’ suggests ‘that Aristotle considered the conclusion as a starting point’ (39): ‘The “begging” of the is then originally the assumption (by use of the appropriate verb in the imperative) of the actual construction the problem is to establish, or of a construction only possible if the former is valid’ (49-50). The procedure is that of analysis in general: ‘having the conclusion in mind, we determine what premises are necessary for its establishment’ (54). See Robinson, , op. cit., on Plato's elenchus (7-20) and its relation to the ‘upward’ movement of his hypothetical method (179-88).Google Scholar

⁴⁹ ‘Geometrical Method and Aristotle's Account of First Principles,’ CQ, 29 (1935) 113–24. Page references to this essay are given in my text.Google Scholar

⁵⁰ In Ross, ' description ( Aristotle, 55), the transition ‘from sense to reason’ is ‘made possible by the fact that perception itself has an element of the universal; we percieve a particular thing, it is true, but what we perceive in it is characters which it shares with other things.’ The progress to higher and higher universality is inductive and intuitive.Google Scholar

⁵¹ The Metaphysics , trans. Tredennick, H., LCL, 2 vols. (London 1933 ). For the relation between the Platonic hypotheses in the Republic and the Aristotelian archai see Solmsen, F., Die Entwicklung der aristotelischen Logik und Rhetorik, 92-107.Google Scholar

⁵² Aristotle, 40. Randall's third chapter, ‘Science as Right Talking: The Analysis of Discourse,’ esp. 40-51, contains interesting implications for fiction.Google Scholar

⁵³ Ross describes how the ‘middle term’ becomes the essential cause’ in a syllogistic definition and how any one of the four causes ‘may function as the middle term whereby the existence of that whose cause it is is proved,’ Aristotle, 51. On causes, see also 71-73. In the beginning of the paragraph being annotated, Einarson compares syllogistic demonstration to mathematical (analysis, resolutio). Analysis is the resolution of anything complex into its elements, according to the Oxford Dictionary, and in logic its specific function is to find out causes by their effects. There is an interesting use of the word, which may bear some of these technical implications, as well as those from grammar, in Spenser's account of the structure of the Faerie Queene. ‘Because the beginning of the whole worke seemeth abrupte and as depending upon other antecedents,’ it is necessary to explain that ‘an historiographer discourseth of affayres orderly as they were donne, accounting as well the times as the actions; but a poet thrusteth into the middest, even where it most concerneth him, and there recoursing to the thinges forepaste, and divining of thinges to come, maketh a pleasing analysis of all,’ The Complete Poetical Works of Edmund Spenser, ed. Dodge, R. E. N. (Cambridge, Mass. 1908) 137. In order to ‘resolve’ the narration better, i.e. to reveal causes by recounting effects, the poet abandons the chronological order; he can thus ‘analyze’ the more important causation which, for the historian, might not appear in the chronological order. See the following note.Google Scholar

⁵⁴ Randall extends his analysis to drama: ‘In real life, in history, we can hardly discern why things have to be as they are: there are far too many complicated and chance or accidental factors. The universal that is implicit there does not stand out clearly. But in tragedy the poet can improve on nature, and show the inevitable dependence of destiny on character. He can make plain not the mere bare event, the “fact that”, to hoti, but also the “reason why,” to dioti: he can disclose how it had to be the way it was' (p. 290). Wimsatt, W. K. and Brooks, C. observe that ‘the terms beginning, middle, and end emphasize a specially close cohesion of causes' and resemble the syllogistic terms major, minor, and middle. They say, further, that, if one thinks in terms of enthymeme, ‘a counterpart of the syllogism in the realm of probability,’ the ‘events in a drama would yield not one but several middle terms, so that we should have the kind of chain of suspended syllogisms known as Aristotelian sorites’ ( Literary Criticism, [New York 1967] 30–2).Google Scholar

⁵⁵ In his edition of the Poetics (Oxford 1968 ), Lucas, D. W. cites Dryden's dedication to The Rival Ladies (1664): ‘When the whole plot is laid open, the spectators may rest satisfied that every cause was powerful enough to produce the effect it had … till they [i.e., cause and effect] all reached the conclusion necessary’ (p. 298). Quintilian observes that the syllogistic method is used not only by geometry and dialectic but also by the logical development of oratory: ‘Geometry arrives at its conclusions from definite premises, and by arguing from what is certain proves what was previously uncertain. Is not this just what we do in speaking?’ In addition to enthymeme the orator will occasionally use a regular syllogism, and, like the geometer, is chiefly concerned with proof (1.10.37-8).Google Scholar

⁵⁶ Metaphysics (13.7; 1082b2-4), trans. Ross, W. D., in The Basic Works of Aristotle, ed. McKeon, R. (New York 1941) 901.Google Scholar

⁵⁷ Lucian, , Hermotimus (73-5), uses sceptical terms, often resembling those of Plato (Rep., 533C) and Sextus, to describe how ‘philosophers grant the premises () of the various systems and then believe everything that follows, supposing that the consistency you find, false though it is, is a proof of its essential truth.’ As illustrations of the power of consistency Lucian uses arithmetic — if someone claims twice five is seven ‘he will clearly go on to say that four times five is certainly fourteen, and so on’ — and geometry: ‘in the beginning it presents certain monstruous postulates and demands that we consent to them though they cannot exist — for instance points without parts, lines without breadth, and so on — and on these rotten foundations it erects its structure and claims to demonstrate truths, in spite of the fact that it starts from a false beginning.’ The most detailed illustration, however, is drawn from poetic fictions. Suppose a poet ‘were to say that there was once a man with three heads and six hands, and suppose that you facilely accepted this without asking if it were possible, just believing, he would at once follow it up by filling in the details appropriately …. Who would disbelieve these details now — details which are consistent with the first outline? … Once you admit the premises the rest comes flooding in … and disbelief is difficult…. You go forward led by the consistency of what came after, not considering that things may be consistent and false’ (op. cit. 6.397-9, trans. Kilburn), K. Once again a passage from Lucian illustrates the continuity of the earlier terminology: in this case the explicit analogy between philosophical, mathematical, and fictional premises and the psychological effect of consistency upon an audience.Google Scholar

⁵⁸ Op. cit. 162. Burnet continues: ‘This usage is as old as Homer, and by a natural extension the verb is freely used in Ionic of suggesting a course of action. That way of speaking accounts for Euclid's use of the word “given,” and also of perfect imperatives like “let there be given” (). The original idea is that of a piece of work given out to be done, and the proposition accordingly ends up a with statement that it has been done.’ Burnet suggests that the material for in Liddell and Scott might be rearranged to ‘be read in the order iii., iv., i.2, ii.2, ii.1’ (p. 163). Robinson's general discussion of the root meanings of the word is helpful (op. cit., 97-117). He stresses the provisional nature of itself, ‘to posit,’ and says , ‘to posit as a preliminary,’ is the specialized variation nearest to the root verb (100). He emphasizes that the ‘positing’ is generally with a view to future action (see Appendix B) and that the noun in Plato derives its meaning from its verbal form (102-3). The sense of ‘something suggested’ () to be discussed has a bearing upon the response of an audience, since Aristotle is very careful to describe it (Post. Anal., 76b25-30) as something agreed upon temporarily, and hence it is ‘relatively to the learner, an hypothesis' (Lee, 116). Robinson observes that the word ‘often applied to the answerer's thesis’ (op. cit., 115). Cornford discusses this meaning in relation to mathematics (39-40) and quotes Sir Thomas Heath's definition: ‘the determination of the conditions or limits of the possibility of a solution of a problem.’ He notes that for Aristotle even a coin, as an agreed measure, may be ‘hypothetical’ (40 n. 2). The word has a variety of meanings closely approximating those of . For its mathematical meanings, see Einarson, 47 n. 63, 51-3, and for others Liddell and Scott. Harper's Latin Dictionary gives for lemma (): a subject for consideration or explication, theme, matter, subject, contents; a title of an epigram indicating a subject or the epigram itself; a story; an assumption of a syllogism (i.e. sub falsa lemmatis specie latens, A. Gellius, 9.16.7). These parallel sets of meanings testify not only to the repeated pattern of associations reflected in each word but perhaps beyond them to intellectual habits of ancient culture congenial to the emergence of the idea of fiction in the ways I have been describing.Google Scholar

⁵⁹ Aristotle's Poetics: The Argument , Else, G. (Cambridge, Mass. 1963) 503. All references will be to Else's translation and commentary, unless otherwise assigned, and will be included in the text.Google Scholar

⁶⁰ Einarson's remarks upon the resemblance between the language of the Prior Analytics and mathematics are particularly relevant to this passage from the Poetics. With reference to 31a18ff. he says: ‘The first sentence here corresponds to Euclid's enunciation of the thing to be proved. The proof begins, as often in Euclid, with the third person imperative and the assigning of letters to the elements under discussion. For Aristotle's there are many parallels in Euclid, where it introduces the proof proper after the or assignment of the general elements of the enunciation to particular elements, diagrams with letters of the alphabet’ (pp. 171-2). The word means that which is set out in a figure in geometry and in the setting out of a syllogism, ‘the most common meaning of the group of words , , and being that of taking numbers out of their context in the number scale and setting them out as terms in a progression or ’ (161). Taking numbers out of scale in order to establish a separate progression resembles the removal of events from the historical continuum and recombining them to suit a dramatic ‘hypothesis.’ The setting out of general elements of enunciation prior to the proof corresponds to drafting the argument of the play in a general, hypothetical outline — to use Aristotle's phrase from the Poetics. The assigning of letters to the diagram suggests the addition of names. Sextus uses the term for setting forth matters of legend () and fiction (Gram. 263f). Google Scholar

⁶¹ After the description of the dramatic plot, virtually paraphrased later by Minturno among others (De Poeta, [Venetiis, 1559] 128-9), Aristotle gives an even more extreme summary of the Odyssey: ‘This is the core; the rest is episodes.’ There can be no mistaking his meaning. Google Scholar

⁶² Aristotle on the Art of Poetry, Bywater, Ingram (Oxford 1909) 51 and 246. Butcher, S. H. translates the sentence: ‘After this, the names being once given, it remains to fill in the episodes,’ Aristotle's Theory of Poetry and Fine Art, ed. Gassner, J. (New York 1951) 63. In discussing the relation of art to nature in Aristotle, Webster, T. B. L. translates a passage on the growth of embryos from On the Generation of Animals (743b20-25) which applies well to the construction of a plot: ‘“everything is defined first by outlines, and later takes on colours, hardness, and softness, just as if nature, who constructs it, were a painter. For painters first draw the lines and then cover the painted animals with colours,” ’ (Art and Literature in Fourth-Century Athens, [London 1956] 54).Google Scholar

⁶³ Webster relates Aristotle's biological conception of species to his assertion of the universality of poetry: ‘The theory is a direct answer to Plato; poetry is not an imitation of an imitation of a reality but creates a new reality, which is itself a union of individual and universal. I believe that Aristotle here views a play in rather the same way as he views Socrates or a particular house. Socrates is the universal, man (the species), realised and individualised by this matter. The particular house is the universal, house, realised and individualised by these bricks and mortar. The Iphigenia in Tauris is similarly a universal story realised and individualised by these episodes, verses, and names' ( ibid. 55–56 ). That episodes may ‘fill in’ intermediate causes and thus articulate the universal was mentioned by Renaissance critics: for Bartolomeo Maranta the episode ‘does nothing else but extend and augment the plot and the universal by telling how what is summarized in the universal has come about’ (quoted in Weinberg, B., A History of Literary Criticism in the Italian Renaissance, [Chicago 1961] 472).Google Scholar

⁶⁴ Such an interpretation is borne out by a later use of . Polybius criticizes the ‘tragical’ (i.e. rhetorical) histories of Phylarchus for their fictional additions which, he implies, would even be out of place in drama, since they aim more at sensationalism than plausibility: ‘Phylarchus simply narrates most of such catastrophes and does not even suggest their causes or the nature of these causes (), without which it is impossible in any case to feel either legitimate pity or proper anger’ ( The Histories , 2.56.13, trans. Paton, W. R., LCL, 6 vols. London 1922 ). The cause or way something happened should be literally ‘hypothesized’ in order to move the emotions properly. Furthermore, it is insufficient to bring in a deus ex machina to explain why the causes are tragic (7.7.1-2, 6). Ullman, B. L. (TAPA, 73 (1942) 25-53, esp. 41) and Walbank, F. W. (A Historical Commentary on Polybius, Oxford, 1957, 1.259-70) discuss Polybius' criticism of the ‘tragical’ school of history and the scholarship on the subject. In the Renaissance Antonio Riccoboni implies with Polybius that universality is revealed by means of articulated causation: ‘Poetry treats universal things; that is, it considers single facts universally or as they might have come about through many causes and in many ways’ (Weinberg, op. cit. 607).Google Scholar

⁶⁵ For the early meanings of , see Louis, Louis, ‘Le mot ‘IΣTOPIA chez Aristote,’ Rev. de Philol. 29 (1955) 39–44. With regard to its usage in the Poetics Louis observes that ‘Il est d'ailleurs assez probable que ce sens de récit d'événement passé n'est qu'un sens dérivé. L'historien, avant d'ětre un narrateur de récits véridiques, est nécessairement un enquěteur qui cherche à s'informer’ (41). The primary emphasis is upon the empirical observation of particular events: history, ‘c'est la connaissance des faits particuliers à partir desquels s'élabore la science’ (44). The , from , ‘est celui qui sait pour avoir vu’ (43); the etymology continues in the Middle Ages (see de Lubac, H., Exégèse Médiévale, Paris, 1959, I ₂, 425ff.). For the early blurring of the distinction between history and poetry, particularly through the influence of tragedy and of Isocratean rhetoric, see Ullman, B. L., ‘History and Tragedy,’ TAPA, 73 (1942) 25-53. Walbank, F. W. questions whether the Aristotelian distinction itself was representative of Greek historiographical attitudes at any time, ‘History and Tragedy,’ Historia 9 (1960) 216-34.Google Scholar

⁶⁶ Translated by Bucher, , op. cit. 172 n. 1, who explains the false inference as being based on the assumptions ‘that because a given thing is the necessary consequent of a given antecedent, the consequent necessarily implies the antecedent. Antecedent and consequent are wrongly assumed to be reciprocally convertible.’ Compare Lucian's statement in n. 57 above, particularly ‘You go forward led by the consistency of what came after, not considering that things may be consistent and false.’ It is significant that Aristotle says that such ‘Reciprocation is more usual in mathematical problems, because mathematics never assumes an accident but only definitions.’ In such a case, ‘Let A be a real fact, whose reality implies that of certain other facts, e.g., B, which I know to be real; then from the latter I will prove the existence of A’ (Post. Anal., 78a7-12).Google Scholar

⁶⁷ One of the earliest extant discussions of comedy, the Tractatus Coislinianus, draws heavily upon this Aristotelian terminology. According to Grube's summary (op. cit. 141-2, 144-9), laughter is provoked by verbal wit or by situation. Of the nine types of situations, three seem to be an intentional subversion of tragic coherence: (3) impossibility (), (4) reaching a possible end through illogical means (), and (9) lack of logical sequence (). Cooper, L. sees these as conscious distortions of the structure of tragedy for purposes of humor (An Aristotelian Theory of Comedy, [New York 1922] 45-7, 187, 191, 194, 217, 244-9, 257-9). More important, however, may be the influence of Hellenistic categories of narrative and of the gradual association of tragedy with history () and of comedy with fictions (). As credibility is sought more in the historical, the true, comedy might be associated more with the unreal or fabulous () and, perhaps, participate in the greater license with regard to logical coherence.Google Scholar

⁶⁸ An Apology for Poetry , op. cit., 1.168–9; also Minturno, De Poeta, (Venetiis 1559) 119, who cites Virgil's Musa mihi causis memora. (Aen. 1.8). Sidney's use of the English word derived from the Latin form of ‘hypothesis’ in relation to causation is most suggestive: ‘The Historian, wanting the precept, is so tyed, not to what shoulde bee but to what is, to the particuler truth of things and not to the general reason of things, that hys example draweth no necessary consequence, and therefore a lesse fruitfull doctrine. Nowe dooth the peerelesse Poet performe both: for whatsoeuer the Philosopher sayth shoulde be doone, hee giueth a perfect picture of it in some one, by whom hee presupposeth it was doone. So as hee coupleth the generall notion with the particuler example’ (164). The poet completes his ‘picture’ of an action (that is, offers motivations) by means of a ‘presupposition’ that it was done by a certain kind of person. For the contribution of ‘precept’ to ‘necessary consequence’ see my discussion of ‘thesis’ below, 51-55. As Lucian had done earlier, St. Augustine in his seventh epistle associates imaginary presuppositions in epic and drama with those in geometry. He divides the images of the phantasy according to their origin either in the senses, the imagination, or the reason. By means of those images in the imagination ‘for the sake of illustration in discourse, we ourselves suppose things which have no existence …; or when we call up … a lively conception of the things described while we read history, or hear, or compose, or refuse to believe fabulous narrations.’ This happens when picturing ‘the appearance of Aeneas, or of Medea with her team of winged dragons, or of Chremes, or Parmeno…. Moreover, we often say, when carrying on a discussion, “Suppose that three worlds, such as the one which we inhabit, were placed one above another”; or, “Suppose the earth to be enclosed within a four-sided figure,” and so on : for all such things we picture to ourselves, and imagine according to the mood and direction of our thoughts.’ The third class of images, those originating in the reason, embody forth concepts of number and measure, found partly in the nature of things and partly in sciences such as geometry and music. Works of St. Augustine, ed. Schaff, P., Nicene and Post-Nicene Fathers (New York 1892) 1.225. See Bundy, , op. cit. 153-72.Google Scholar

⁶⁹ History is a record of random effects (from man's point of view as distinct from God's, which includes a knowledge of causes), and its literary procedures, as in a chronicle, reflect a random progression. Insofar as an historian ‘interprets’ the effects (events), he is ‘constructing,’ not ‘recording,’ but his construction is never certain, final, or true. It must be constantly revised in accordance with new data about events: in this sense its allegiance is always to the events, not to any given construct of events. Its progressions are undelimited and random insofar as they are subservient to subsequent effects occurring or being discovered by chance. (When his memory fails, Thucydides admits he constructs speeches which are plausibly demanded by various occasions (1.21-2). Such speeches would be subject to revision, however, if more details were learned about the occasions.) This is also the procedure of the fantasy, in which sense impressions are the events, the equivalent of effects; they are reflected in the fantasy, recorded, and later ‘interpreted’ to give a construct of the stimuli by other faculties. Fiction is distinct from fantasy in the same way that it is distinct from history. In fiction there is no randomness; causes are ‘discovered’ (given) for effects: it is always ‘true.’ In the Renaissance Dionigi Atanagi writes: ‘The order of poetry is certain, connected, and linked, since because of the interrelationship of its actions it makes one out of many, one toward which it directs all the others as servants and domestics serve a mistress…. The order of history is for the most part uncertain, disjointed, and fortuitous, since its actions are not similar and linked but separate and diverse; neither does one depend from another nor do they relate to a single end’ (Weinberg, op. cit. 458).Google Scholar

⁷⁰ Lucas, D. W. comments on the ‘universal’ in 1451b8 ( Poetics, Oxford, 1968 ): ‘When the universal regularities are revealed, events are intelligible: (An. Post. 88a5), “the universal is precious because it reveals the cause.”’ The necessity for seeing ‘parts’ only in relation to ‘wholes,’ even in the most minute subjects of research, pervades Aristotle's works. It is as true of On the Parts of Animals (whose preface is translated by Jaeger, Aristotle, 338-9) as of the science of metaphysics, which, he says, ‘is based on physics according to Aristotle in the first place because it is nothing but the conceptually necessary completion of the experimentally revealed system of moving nature’ (380). So fiction might be said to be the conceptually necessary completion of observed historical events. In an article on ‘The Function of the Myth in Plato's Philosophy,’ JHI, 10 (1949) 463-81, Edelstein, L. suggests that Plato regarded his ‘ethical’ myths of the after life in a similar way: that is, that they could be used to complete an argument (v. Rep., 614A) and in this way become ‘an addition to rational knowledge’ (473). For a similar view held by Proclus, see Bundy, , op. cit., 143. Analogous to the unity of composition stressed by Plato (Phaedrus, 264-8) is the Aristotelian change from an organization of an oration by ‘parts’ to one of ‘functions’ operating throughout the whole speech to insure affectivity (see Solmsen, , ‘The Aristotelian Tradition …,’ 37f.). For Aristotle's use of letters and syllables to illustrate the formal relation of parts to wholes, see Metaphysics 7.17; 104b11-33.Google Scholar

⁷¹ How this generality might derive from an association with dialectic as well is suggested in Aristotle's account of destructive arguments: if such and such is or is not true of one member of a genus, it is or is not true of all members: ‘Now it is clear that he who makes the hypothesis makes the problem universal, though it is posited in a particular form; for he demands that the maker of a particular admission should make a universal admission, since he demands that, if an attribute belongs in a particular case, it belongs in like manner to all’ (Topica, 3.6; 120a2-5, trans. Forster, E. S., LCL, [Cambridge 1960]). Such an explanation would not be out of place in the Poetics.Google Scholar

⁷² See Aristotelis de Poetica Liber … per Theodorum Goulstonum [bound with Aristotelis de Poetica Liber ex versione Theodori Goulstoni], (Cantabrigiae 1696) 26: ‘Poësis Philosophiae similior, quàm historia. Siquidem Poëtae Fabula, Universales ac Indefinitas captet. In comedies theses were “proposed” for illustration in plays to serve a didactic purpose: Sunt in singulis Comoediis certae quaedam theses de hominum diversis moribus, ingeniis, & officiis propositae, quae multum faciunt ad vitam sapienter & civiliter instituendam’ ( P. Terentii Afri Poetae Lepidissimi Comoediae, Parisiis, 1552, p. 675).Google Scholar

⁷³ For this difficult passage (1450b4-12), see Else's commentary, esp. 270-1, and compare the translations of Butcher and Bywater. The main point is that what Aristotle says of ‘thought’ resembles closely his definition of a ‘thesis’ (Top. 104b19ff.) as well as the later embodiment of a thesis in the literary exercise of the suasoria. Such declamations, in turn, greatly influenced all types of set speeches, such as the Elizabethan soliloquy: ‘To be or not to be, that is the question’ (i.e. the matter to be considered generally, not the decision specifically about the speaker's committing suicide at that moment). Later in the Poetics (1456a35 ff.) Aristotle refers everything concerning ‘thought’ to his Rhetoric. Google Scholar

⁷⁴ In this broader sense of a universal, the ‘thesis’ connotes the generic within the specific, not isolated in abstraction but immanent in a given particular situation. This ‘fictional’ combination of (assumed) circumstantial existence and definitive significance is suggested by the way Aristotle distinguishes these technical terms: ‘A thesis which assumes one or the other part of a proposition, i.e., that something does, or does not exist, is a hypothesis; a thesis which does not do this is a definition. A definition is a kind of thesis <or laying-down>, because the arithmetician lays it down that to be a unit is to be quantitatively indivisible; but it is not a hypothesis, because to define the nature of a unit is not the same as to assert its existence' (72a19-25). Robinson interprets this passage to mean that an hypothesis is an ‘assertion’ which may be true or false and that a definition is a ‘convention’ or a ‘promise’ which claims neither truth nor falsity ( op. cit. 105f). ‘Thesis’ retains something of the generic quality as definition and something of the asserted existence of initial premises of action (plot outline) as hypothesis (see n. 71). For the exclusion of generic significance from narrative hypothesis after the Renaissance, see Appendix C.,+because+the+arithmetician+lays+it+down+that+to+be+a+unit+is+to+be+quantitatively+indivisible;+but+it+is+not+a+hypothesis,+because+to+define+the+nature+of+a+unit+is+not+the+same+as+to+assert+its+existence'+(72a19-25).+Robinson+interprets+this+passage+to+mean+that+an+hypothesis+is+an+‘assertion’+which+may+be+true+or+false+and+that+a+definition+is+a+‘convention’+or+a+‘promise’+which+claims+neither+truth+nor+falsity+(+op.+cit.+105f).+‘Thesis’+retains+something+of+the+generic+quality+as+definition+and+something+of+the+asserted+existence+of+initial+premises+of+action+(plot+outline)+as+hypothesis+(see+n.+71).+For+the+exclusion+of+generic+significance+from+narrative+hypothesis+after+the+Renaissance,+see+Appendix+C.>Google Scholar

⁷⁵ On potentiality and actuality, see Ross, , Aristotle, 176 ff. Later (285), he observes that character corresponds to potentiality, plot to actuality: the play is then potentiality-in-actualization, i.e., character-in-action. Actualization and plot are ‘prior’ to potentiality and character.Google Scholar

⁷⁶ Lee, , acknowledging that ‘each science assumes the existence of the genus of which it is its business to prove the essential attributes,’ observes that Euclid's postulates correspond to Aristotle's hypotheses: Euclid must ‘assume the possibility of constructing a certain minimum of figures, from which it would be possible to prove the possibility of constructing the rest' (p. 115). So also ‘according to Aristotle the geometer must assume the existence of points and lines (76b5)’ (115 n. 1).Google Scholar

⁷⁷ Atkins, J. W. H., Literary Criticism in Antiquity, 2 vols. (London 1952) 1. 80.Google Scholar

⁷⁸ This brief summary is based on Rostagni, A., La Poetica di Aristotele (Torino 1927) xxxv–xli. Edelstein suggests, however, that Plato regarded the emotional effects of myths about the after life more positively, since ‘the ethical myth speaks to man's passions; it rouses and confirms hopes; it enhances courage and allays fears’ (op. cit. 474). Such a ‘myth, shaped in accordance with reason, brings to the realm of the passions the light of the intellect; it instigates man to act with hope and confidence toward the goal which reason has set before him’ (477). See Friedländer, P., Plato, tr. Meyerhoff, H. (New York 1958) 1.171-210. For the possible transcendental powers of the phantasy in Plato's later work, see Bundy, , op. cit. 46-58. In his Rhetoric Aristotle raised the categories of and to the level of that of ‘rhetorical arguments’ in delineating these three as the three types of proof. Later, perhaps under the influence of Hermagoras’ scholastic distinctions of status, ethos and pathos were relegated to the proem and the epilogue of the speech. Cicero restricted them in the same way in his early work, but in his mature treatises he reinstated them (in the officia of delectare and permovere) in much the same way as he did the ‘thesis.’ There is, I believe, a relation between the two reinstatements. On these subjects see the articles of Solmsen cited above.Google Scholar

⁷⁹ For a more thoroughly syllogistic analysis of metaphor, see Rhetoric, 1410b5-35 and 1412a17-25.Google Scholar

⁸⁰ Laws , trans. Taylor, A. E. in The Collected Dialogues of Plato, ed. Hamilton, E. and Cairns, H. (New York 1963 ). Cf. Republic, 592b.Google Scholar

⁸¹ Jaeger notes that in Rep. 511d ‘The basis of the comparison between the four stages described by Plato is the difference in clearness (, sometimes ) which they represent. means not only intelligibility but reality: cf. 510a9: .’ Paideia, 2.417 n. 61. Google Scholar

⁸² Without reference to this passage, Jaeger comments generally on Isocrates' method: ‘This method of laying down a supreme principle or aim of action he calls a hypothesis, a “laying of foundations” — because all further arguments must rest upon it. In several other passages in his speeches we can observe this effort to find a generally accepted hypothesis: it is an essential element in his political thought, and is to be explained by the influence of Plato's intellectual method. Ultimately, it is a procedure borrowed from mathematics' (Paideia, 3.93). Jaeger calls attention (ibid. 310 n. 56) to Nich. 13, which reads as follows: ‘for it is through this training that you can soonest become such a man as we have assumed () that one must be who is to perform properly the duties of a king, and to govern the state as he should.’ Google Scholar

⁸³ With regard to this ‘completeness,’ Aristotle's discussion of equity (that ‘sort of justice which goes beyond the written law’ where law makers ‘find themselves unable to define things exactly, and are obliged to legislate as if that held good always which in fact only holds good usually; or where it is not easy to be complete owing to the endless possible cases presented') is extremely pertinent to Isocrates’ self-defense: ‘Equity bids us be merciful to the weakness of human nature; to think less about the laws than about the man who framed them, and less about what he said than about what he meant; not to consider the actions of the accused so much as his intentions; nor this or that detail so much as the whole story; to ask not what a man is now but what he has always or usually been’ (Rhet. 1.13; 1374a27-32, 1374b11-16). What a man ‘has always or usually been’ is near the ‘universal’ which poetry aims to reveal: ‘What kinds of thing a certain kind of person will say or do in accordance with probability or necessity’ (Poet., 51b9-10). When Polybius criticizes Phylarchus for omitting causes () without which there can be neither ‘legitimate pity or proper anger’ aroused (2.56.13), he continues by stressing that what ‘properly’ causes emotions is important, because the interpretation of the motive is important in a defense. Such considerations are matters of equity for ‘in every such case the final criterion of good and evil lies not in what is done, but in the different reason and different purposes of the doer ().’ In both the hypothetical causation and the determination of equity the word is used. Perhaps the logical extreme of the influence of the rhetorical concern with equity is reflected in Plutarch's defense of poetry where probable consistency in events has been replaced by consistent morality in ‘thought': in reading any literature, he says, ‘it is useful also to seek after the cause () of each thing that is said,’ in order to see if it is morally acceptable as a model of conduct. ‘How to Study Poetry,’ 9, Plutarch's Moralia (28A-B), trans. Babbitt, F. C., LCL (London 1960) 1.147.Google Scholar

⁸⁴ Such an apprehension, intended to vindicate Isocrates in the eyes of his audience, is consistent with Else's interpretation of ‘ catharsis’ (pp. 224–32, 423-50). Butcher (op. cit., p. 245) and others cite Plato's Sophist (229E-230E) to illustrate the medical use of the term, but the passage has further implications for this essay. Plato is describing the cathartic effect of the elenchus upon the respondent who is ‘purified’ of his pretence to knowledge by experiencing the shame of refutation. Once purified he can begin to learn properly. The emotional acquiescence, brought about by the hypothetical method of moral dialectic, corresponds to the emotional acceptance of the synthesis of ‘given’ events, brought about by observing the ‘elenchic’ and consistent selection of means within the fictional hypothesis.Google Scholar

⁸⁵ In describing the extent of mathematical presuppositions in attitudes toward certainty, probability, and the arts, Webster, T. B. L. observes that ‘mathematical proof has universal validity; it is what the Greeks call “necessary” (anankaion).’ Such a ‘necessity’ was often invoked to control the arguments from probability: ‘Thus the mathematical arguments provide a framework within which the general truths established by observation or otherwise can be related to each other’ ( Greek Art and Literature 700-530 B.C. [London 1959 ] 95–6).Google Scholar

⁸⁶ In distinguishing Plato's hypothetical method in the Republic from that in the Meno and Phaedo, Robinson argues that the method used to achieve the highest category on the Divided Line ‘has gone back to being practically the Socratic elenchus’ ( op. cit., p. 184). He later criticizes the method for avoiding an infinite regress only by means of the assumption (unacceptable today even in geometry) that a premise could be examined dialectically without assuming other premises: ‘What seemed to Plato the gradual forging of an hypothesis to which there were no objections turns out … to be merely the gradual forging of a consistent set of beliefs; and it therefore does not escape the stricture passed in the Cratylus, that consistency is no guarantee of truth' (190). It is interesting that the exclusion of other premises was precisely what the fictional hypothesis (once again in agreement with Euclidean geometry) enabled the poet to do. Literary discourse might avoid the danger of infinite regress by its very power to delimit its philosophical premises with stipulated circumstances, analogous to the rhetorical hypothesis, in a fictional plot. The desire to justify a type of discourse which could avoid this danger may have influenced the development of literary theory after Plato.Google Scholar