STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. THE DE COMPOSITIONS OF DIONYSIUS OF HALI- CARNASSUS CONSIDERED WITH REFERENCE TO THE RHETORIC OF ARISTOTLE By H. P. Breitenbach The De Compositione of Dionysius of Halicarnassus is a work that deserves more attention from students of language and literature than it has received. 1 Now that so excellent an edition and trans- lation has been published, that of W. Rhys Roberts, there is every incentive for the study of the treatise. 2 A comparison of it with the Rhetoric of Aristotle promises to throw some new light upon the methods and results of both authors. In this paper I purpose to indicate the attitude of each writer to his problem by disclosing and comparing the fundamental pre- suppositions that underlie the two treatises. I shall endeavor also, from the standpoint of aesthetics, to explain their divergent positions. Finally I shall refer to certain similar standpoints in modern rhetorical theory. Of the two treatises the scope of the De Compositione is much the more limited. In c. i Dionysius suggests the following rhetorical classification: ( Thought Discourse < ( Selection ( Expression < ( Arrangement The De Compositione confines itself pretty strictly to the topic of arrangement; the other topics receive only occasional mention. The scope of the Rhetoric is less easily defined. The search for it reveals a significant progression in Aristotle's point of view. His avowed purpose is to establish rhetoric on a scientific basis by relating it to dialectics. Thus he hopes to avoid, on the one hand, the fragmentary treatment found in the ordinary rhetorical hand- 1 Cf . Clayton Hamilton Materials and Methods of Fiction, p. 206. 2 It is, however, only fair to state that this essay (together with some other studies of similar nature) was prepared several months before Professor Roberts' edition was announced as forthcoming. [Classical Philology VI, April, 1911] 163 164 H. P. Breitenbach books, and, on the other hand, the sophistical method, which involves numerous devices or tricks, such as introducing matters foreign to the issue but designed to work on the feelings of the audience. 1 In opposition to these methods he declares : " It is clear then that the only proper subjects of artistic treatment are proofs" (i. 1). Accord- ingly he defines rhetoric as "a faculty of discerning all the possible means of persuasion in any subject" (i. 2). That no reference to the domain of form or expression is implied at this point is indicated by the threefold analysis of the subject which he immediately makes: "The proofs provided through the instrumentality of the speech are of three kinds, consisting either in the moral character of the speaker, or in the production of a certain disposition in the audience or in the speech itself by real or apparent demonstration" (i. 2). Again he says: ". . . . a speech is naturally composed of three ele- ments, viz., the speaker, the subject of the speech, and the persons addressed" (i. 3). To these topics, then, which he calls the only proper subjects of treatment in rhetoric, he devotes the first two books, comprising about three-fourths of the entire treatise. But at the outset of the third book he makes a new analysis in which all that he has before discussed constitutes only a third part. "There being three proper subjects of systematic treatment in rhetoric, viz., the possible sources of proof, style, and the right ordering of the parts of the speech, the first of these has already been discussed We have next to discuss the question of style" (iii. 1). Why, then, did Aristotle at the outset of his discussion analyze discourse so as to leave out of consideration the topic of expres- sion ? And why did he afterward change his point of view so as to admit it ? To take up each question in turn: Several reasons may briefly be set down which will throw light on the first point. The element of style or expression, which term may serve for the moment to include both the remaining topics of Aristotle, is not easily susceptible of intellectual comprehension — not, at least, in comparison with the subject to which Aristotle gives his main attention, the sources of proof. Also, the principles of style cannot well be worked out deductively. Likewise, they cannot easily be related to rhetoric's lOf. Rhet. ad Alex., c. 15. The De Compositions of Dionysius 165 "counterpart," dialectics. These perhaps are the main reasons why- Aristotle did not include style in his first analysis of his subject. Then, too, the topic style, like the appeal to the emotions of the audience, was a feature of the sophistical treatments of rhetoric, which Aristotle was opposing. We know that Gorgias, for example, devoted considerable attention to it. Moreover, while the Rhetoric was designed to be philosophical or scientific in its basis, it was prac- tical in its aim. The matter of style is not of immediate importance from either of these points of view. Finally, it may be that Aristotle's own appreciation of style was weak. In his extant works he assuredly exposes himself to this charge, although certain ancient critics, among them Dionysius himself, especially commend his style. Whether their judgment is without adequate grounds, or whether they are referring to works of Aristotle now lost, we can scarcely say. But so far as the evidence afforded by his own extant works is concerned, we should be justified in ascribing to him a lack of artistic ability. As for his appreciation of style in others, the fact that he disregards the works of his great contemporary Demosthenes has perhaps some bearing on the matter; and below, in the discussion of his treatment of style, his lack of interest in the details of expression will be shown. The other question now remains for consideration : why Aristotle, after outlining his subject in such a way as to exclude the topic of style, changed his point of view and did discuss it. While perhaps no decisive answer can be given to such a question, some attempt at explanation may be offered. It is altogether likely, then, that Aris- totle intended his treatment of rhetoric, as we have it in the first two books, to embrace the subject completely; which it does, so far as the logical aspect of the subject is concerned. Then, probably, his further studies drew his attention to the importance of the elements he had before neglected, and he added the third book as a sort of appendix. If it is true, as many scholars have concluded, that the Poetics was composed after the second book of the Rhetoric and before the third, the theory would receive considerable support. 1 For 1 On the question of the genuineness of the third book and its relation to the rest of the Rhetoric and to the Poetics, see Christ Griechische Litteraturgeschichte, 3 e Aufl., p. 483 and note, together with the references there given. 166 H. P. Breitenbach Aristotle's study of poetry would naturally bring him to a considera- tion of the stylistic element in oratory. Some confirmation of this is afforded by the frequent references in the third book to the Poetics, and especially by Aristotle's efforts to distinguish the appropriate styles of prose and poetry (for a discussion of which, see below). At any rate, Aristotle assigns to the whole topic of style or expres- sion only the brief third book of the Rhetoric; while Dionysius, having subdivided expression into selection and arrangement, devotes his entire treatise to the latter topic alone. It now remains to con- sider what are the fundamental rhetorical presuppositions of the two writers and to see how these underlie their respective treat- ments of the subject of style. In this investigation I shall view the treatises from the three standpoints which Aristotle distinguished — that of the hearer, that of the speaker, and that of the discourse itself. Regarded from the point of view of the hearer, the fundamental principle of Aristotle's theory of rhetoric is what may be termed intellectual hedonism. "The acquisition of knowledge is pleasant" (i. 11. 2). "To receive information easily is naturally pleasing to all" (iii. 10. 2). "Accordingly, in style and enthymemes all those are pointed and lively that convey to us instruction rapidly" (iii. 10. 4). "Such a style is agreeable because the hearer is constantly thinking he has got hold of something" (iii. 9. 3). Other passages of the Rhetoric (e.g., iii. 11. 6, iii. 8. 2, i. 11. 21) convey the same idea. It is found elsewhere in his works also. At the opening of the Metaphysics, for example, he says: "All men have a natural longing for knowledge"; in the Problems: "The pleasure we receive from rhythm derives from the natural love of a recognizable (yvrnpipov) regularity"; in the Poetics: ". . . . learning is most delightful not only to philosophers but in like manner to other persons" (iv. 4). Indeed, Aristotle's very theory of art, as expressed in the Rhetoric, seems based upon this principle. "From the pleasure of learning and wonder it results that there is a pleasure in such things as the imitative arts, e.g., painting, sculpture and poetry, or in any success- ful imitative work, even if the actual object of imitation is not pleas- ant; as it is not the pleasantness of the object which produces the The Dm Compositions of Dionysius 167 pleasure but an inference from the copy to the original and in con- sequence of it a kind of learning" (i. 11). In short, this principle of Aristotle's assumes that the hearer is actuated by the pleasure that accompanies the act of learning. Just as the satisfying of bodily hunger gives a feeling of well-being, so likewise, Aristotle would say, the mind is naturally hungry for infor- mation, and the satisfaction of this desire produces a feeling of pleas- ure. To this principle I have given the name intellectual hedonism to indicate that, while a pleasurable emotion is involved, its source is intellectual. According to this principle, one who heard bad tidings would nevertheless be pleased because he had learned some- thing. In contrast to the fundamental principle of Aristotle's Rhetoric, that of the De Compositione, likewise from the standpoint of the hearer, may be termed aesthetic appreciation, by which I mean pleas- urable emotion derived from what is artistic. Dionysius' attitude on this point is very manifest in numerous passages. In c. xi, for example, in a passage too long to quote here, he compares the pleasure derived from oratory to that derived from music. " In oratory also the words have melody, rhythm, variety and appropriateness; so that the ear takes pleasure in the melody, is captivated by the rhythm, delights in the variety, and in all requires appropriateness. The difference between the two is merely one of more or less." The same underlying principle of aesthetic appreciation is clearly shown in cc. ii, iii, iv, vi, ix, xii, xvi, etc. In fact, he maintains this point of view consistently throughout the treatise. According to this view, the hearer is pleased, not because he learns something, as is Aristotle's idea, nor is it primarily because he receives pleasing information — but because he enjoys the oration as a thing of beauty, for its rhythm, its euphoniousness, etc. We have now to consider the second point of view, that of the speaker. More specifically, our question is, What is the purpose of the speaker ? What is the effect on the hearer that he aims to produce by means of the speech ? Answering the question for the Rhetoric, we may say that, obviously, the aim and purpose is to produce persuasion. Indeed this is the essence of Aristotle's definition of rhetoric, "a faculty of discerning all the possible means of per- 168 H. P. Breitenbach suasion in any subject." 1 Now to be consistent with himself, Aris- totle should make the means of persuasion include only proofs appealing to the intellect. In point of fact, he does this, as we have seen, at the outset of his treatise. There he declares, " It is the proofs alone which form the proper subject of artistic treatment, and everything except the proofs is a mere accessory"; and he con- demns those sophistical writers who "omit all mention of enthy- memes which are the soul of proof, and occupy themselves almost exclusively with such things as lie outside the actual issue. For," he adds, "prejudice, compassion, anger and such emotions of the soul have no bearing upon the point at issue; they merely affect the minds of the jury." Again he makes a similar protest: "It is improper to warp the judgment of a juror by exciting him to anger or jealousy or compassion, as this is like making the rule that one is going to use, crooked." He would have all cases "determined absolutely by the laws, and as little as possible left to the discretion of the judges." For, besides other considerations, "feelings of affec- tion or ill-will and private interests are necessarily often involved, so that they lose the power of adequately studying the truth, and their judgment is clouded by a consideration of their personal pleas- ure or pain." And a little further on he declares for the second time, "It is clear then that the only proper subjects of artistic treatment are proofs" (i. 1). Evidently, then, Aristotle starts out with an intellectual point of view. The means of persuasion are to be confined to proofs appealing to the intellect. But in a very short time he shifts his position, and we find him admitting into consideration appeals to the emotions. Finally, he devotes considerable attention to them in the first part of the second book. A similar inconsistency we have already observed, in the ease of his treatment of style, which likewise he at first excluded from consideration. His remarks on style and delivery, at the beginning of the third book, are in fact strikingly similar to those opinions just quoted relative to appeals to the emo- tions. "Strict justice, indeed, if applicable to rhetoric, would con- 1 In the previous chapter he has declared :".... Its function is not to persuade but to discover the available means of persuasion in any subject." From what fol- lows, however, his meaning is clearly that it should succeed so far as is possible, although, like the art of medicine, it must sometimes fail. The De Compositions of Dionysius 169 fine itself to seeking a delivery such as would cause neither pain nor pleasure. For the right condition is that the battle should be fought out on the facts of the case alone: and therefore everything outside the direct proof is really superfluous, although extraneous matters are highly effective, as has been said, owing to the depraved character of the audience." And he argues that when one is teaching geometry "tricks of style" are not needed to gratify the audience. The inconsistency, then, in regard to appeals to the emotions is of the same sort as the inconsistency with regard to style; and prob- ably the same reasons lie at the bottom of it. These are : Aristotle's strong intellectual bias; his opposition to the sophistical treatments of the subject; possibly a lack of artistic appreciation. Here also another factor now comes to view — his low opinion of the popular audience as referred to above in the words, "the depraved character (/j.ox0vpt av ) of the audience." This opinion is expressed with various degrees of candor in several other passages of the Rhetoric, e.g., ii. 21, iii. 16, iii. 14. In short, the aim of the speaker is persuasion, and not by intellec- tual means alone. Aristotle goes even farther than to admit emo- tional appeals among the means of persuasion: he admits also "apparent demonstration" as well as "real"; by this he means fallacious or specious reasons, etc. Examples of this "sophistical" element are to be found in iii. 14, 15, 16, etc. So far as the aim of the speaker is considered in the De Compo- sitione, it is referred to pleasing the auditor by the charm and beauty of the speech, by its effect, that is to say, as a work of art; the pur- pose is to produce not an intellectual, but an aesthetic, impression. Although in his other works Dionysius duly recognizes them, yet in the De Compositione he seldom makes direct reference to the pro- duction of clearness, lucidity, or any other intellectual effects as being a necessary aim of arrangement. On the contrary, he commends passages where the thought has been somewhat sacrificed to the artistic demands of expression, being left incomplete or made redun- dant. "Truly some clauses receive changes in their form, taking on additions not necessary to their sense, or undergoing curtailments whereby the thought is left incomplete. These changes are made by both poets and prose-writers, and for no other reason than that 170 H. P. Breitenbach the 'harmony' may become more charming and beautiful" (c. ix). In other passages also the same idea is expressed. Since in the De Compositions the attitude of the hearer and that of the speaker exactly correspond, further discussion or citation of examples seems unnecessary. Instead, another indication of the aesthetic point of view in the treatise may be presented here — the character of the analogies employed. Practically all the important analogies are drawn from the realm of fine art. I have already referred to one passage comparing oratory to music. In c. x, referring to charm and beauty as the character- istic aims of arrangement, Dionysius says: "For the ear requires both of these in much the same manner as does the eye. The eye looks upon images and pictures and sculpture and all the other handi- work of man, and finding therein charm and beauty, it is satisfied and desires nothing more." In c. xxi he compares composition to painting. In c. xxiii he says the smooth style resembles finely woven cloth or pictures in which light and shadow blend. In c. xxv, he says that the men of Demosthenes' time fashioned their discourses "not like to writings but to carvings and intaglios." Again, arguing in the same strain he speaks of the pupils of painters and sculptors, who would exhaust the ingenuity of their art on such details as veins and down. In the same chapter, to suggest how conscious art by dint of practice becomes second nature, he refers to the training of skilled performers on the lyre, harp, and flute. There are also some minor allusions of similar nature in the De Compositione. It is to be noted that the Rhetoric contains no such comparisons. The third significant relation, that of the speaker to the discourse, yet awaits consideration. The problem is represented by the ques- tion, in the case of each author: What is the nature of the discourse as a combination of words ? Taking up its consideration we come upon another divergence in the concepts of the two authors. Aristotle's idea as to the nature of a discourse may be only inferred from the Rhetoric; but in the Poetics (c. xxiii) what is probably his conception of it is expressed in the comparison of an epic poem to "a living organism." "It should have for its subject a single action, whole and complete, with a beginning, a middle, and an end." He discusses tragedy in a The De Compositions of Dionysitjs 171 similar way, making the same comparison, and similarly explaining the organic interrelation of beginning, middle, and end. It seems altogether likely then that this view extends with Aristotle to a speech or any other discourse in the realm of rhetoric. This proba- bility is strengthened by certain indications in the third book of the Rhetoric in his treatment of the exordium. For he compares it to the poem of epic or dithyrambic poetry, or the prologue of a drama: "As to the exordia of forensic speeches, it must be understood that they are equivalent to the prologues of dramatic or the proems of epic poetry; for the proems of dithyrambs resemble epideictic exordia In rhetoric as in epic poetry the exordium is a sample of the subject, being intended," etc. Moreover, Aristotle explains that exordia which do not set forth the end or object of a speech are not true exordia, but "merely means of remedying defects in the audience, serving to get their attention, remove prejudice, and the like. They are used in this way because the audience is corrupt. If this is not the character of the audience there is no need of an exordium, except for the mere purpose of stating the facts summarily, that it may not be, as it were, a body without a head" (iii. 14). Besides the exordium a speech contains necessarily two parts, exposition and proof; the peroration is some- times found as a third part, but it is not necessary to a speech. Probably, then, in Aristotle's opinion, the exordium is the "begin- ning," the exposition the "middle," and the proof the "end." In short, the Platonic conception of discourse as an organism, which is a dominating idea in the Poetics, probably underlay, though less consciously, Aristotle's whole theory of discourse. Certainly it is an idea altogether likely to have accorded with his scientific type of mind and early-awakened interest in biology. The underlying concept of Dionysius, from the point of view of the discourse as a combination of words, is not at all biological in its nature, but rather what may be called architectonic, or perhaps, in a restricted sense of the word, architectural. He has first analyzed words into their component syllables and these into their component letters: "We designate them letters because they are written with certain lines, and we term them elements because all speech takes its origin from them and into them resolves itself in its final analysis " 172 H. P. Breitenbach (xiv). Then, having given a long and detailed account of the quality and force of the several letters, he undertakes the task of synthesis : "Out of these letters, then, having such number and such qualities, the so-called syllables are formed" (xv); and he examines the nature of syllables. Finally he remarks: "How, then, do I sum up this matter ? I say that the diverse qualities of syllables originate in the combination of letters; that the nature of words varies according to the arrangement of syllables; and that discourse varies according to the degree of harmonious arrangement in the words" (xvi). His treatment of rhythm is altogether similar (see cc. xvii-xviii). A composition in Dionysius' view is a structure formed like a building of a certain kind. It is as if a pyramid were constructed in several sections, each in turn composed of several subsections, and these again composed of the individual stones. 1 The stones represent the words which are fitted together to form the sub-sections, i.e., the clauses. The clauses in turn are adjusted one to another until they form suitable sentences. These, finally, Dionysius would say, form the structure of the whole composition. But while in all this process of construction the builder is supposed to be thinking of charm and beauty as aims, the structure has no fundamental organization. The units are built up independently of one another until the time comes for combining them; only then are they adjusted one to another. In both analysis and synthesis Dionysius is governed by such an architectural analogy. At the outset of his treatise, to show that arrangement has more potency than selection (an idea that itself is significant of his attitude) he refers to "the other constructive arts, such as house-building and cabinet-making and the like, which take various kinds of material and out of them produce a composite result" (c. ii). Again, in c. vi, when explaining the function of arrangement, he uses as illustrations " house-building and shipbuilding and the like," to show how the materials are adjusted so as to form a harmonious structure. The latter passage well indicates the remarkable influ- ence of the architectural concept on Dionysius' treatment of the sub- ject. Unfortunately it is too long to be quoted here. 1 Of course, to make the analogy complete, one would need to imagine that the nature of the several stones depends upon their constituent elements, just as that of words is supposed by Dionysius to depend upon the nature of the component syllables and letters. The De Compositions of Dionysitjs 173 In the last-mentioned passage Dionysius gives his principles of arrangement with reference to words. He also explains the similar principles that govern the disposition of clauses (cc. vii-ix). He has promised (c. v) to discuss likewise those that pertain to the arrangement of periods, but he dismisses the subject in two or three brief sentences (c. ix, ad fin.). This is significant, pointing to his greater interest in the details of style, and also indicating the induc- tive nature of his method. Conversely, Aristotle, his interest being of a philosophic nature, devotes the greater share of his attention to the larger elements of style and fails to carry his analysis into and through the details. The fundamental presuppositions of the two authors have now been set forth with reference to the three aspects of discourse. Ac- cording to Aristotle, the attitude of the hearer is that of intellectual hedonism; the aim of the speaker is to produce persuasion, even by appeals to the emotions or by sophistical proofs; the discourse is organic in structure. According to Dionysius, the attitude of the hearer is that of artistic appreciation; that of the speaker corresponds, his aim being to produce artistic enjoyment; the discourse is archi- tectural in structure. Aristotle sets out with an intellectual pre- disposition, at first excluding all non-rational elements. Dionysius is interested rather in the artistic or aesthetic phase of the subject. Several minor topics are treated in some detail by both Dionysius and Aristotle, affording an opportunity for further comparison. These include diction or choice of words, rhythm, the period, and the relation of prose and poetry. Careful study has revealed that the fundamental principles set forth above determine the respective treatments of these matters of detail also. On account of the limitations of space, however, only one of them, the relation of prose and poetry, can be discussed here. At several points in the third book of the Rhetoric Aristotle distinguishes between the appropriate metrical structure of prose and poetry. Near the beginning of the book (c. i) he sketches the history of style to show how prose has become differentiated from poetry. "It was because the poets were thought, despite the sim- plicity of their sentiments, to have acquired their reputation by their style, that prose style assumed at first a poetical form, as, e.g., the 174 H. P. Breitenbach style of Gorgias. Nay, even at the present time it is the opinion of most uneducated persons that a poetical style is the finest. This, however, is an erroneous idea, the styles of prose and poetry being distinct, as is shown by the fact that the writers of tragedies them- selves have ceased to use the poetical style as once they did It is absurd then to imitate them who no longer employ their old style." In the following chapter (ii) he says: "Style should be neither mean nor exaggerated, but appropriate; for a poetical style, though not mean, is still not appropriate to prose." Throughout the chapter he refers repeatedly to the necessity for maintaining separate styles for prose and poetry. In c. iii he enters an objection against the misuse of epithets : " For if in poetry it is proper to speak, e.g., of 'white milk,' such epithets in prose are in any case inappropri- ate, and if there are many of them, they expose the art of the style and show it to be simple poetry." Another expression (avri/iifiov in avrlfiifiov rrjv ttjs ^rvxrj^ eiriBv/iiav) he calls "a word which is at the same time a compound and an epithet, so that the prose is converted into poetry The consequence is," he resumes, "that this poetical diction by its impropriety is a source of absurdity and tastelessness as well as of obscurity." In c. vi, among the means of producing dignity of style, he recommends the employment of metaphors, "being on your guard at the same time against a poetical style." At the close of c. vii, he suggests that the use of compound words, epithets, and strange words, which previously he has objected to, is permissible "when the speaker has mastered his audience and has aroused them to enthusiasm For this is the language of enthusiasm and is ... . suitable to poetry for the same reasons, as poetry is inspired." Aristotle's final reference to the relation of prose and poetry is in c. viii, in his treatment of rhythm in prose. He sums it up in his famous dictum: "A prose composition should have rhythm but not meter, or it will be a poem. But the rhythm should not be too elaborately finished, or in other words it should not be carried too far." He recommends the use of the paeon in prose on the ground that it alone cannot form a regular meter, and therefore it is the most likely to escape detection. Evidently, then, Aristotle's aim is, from a logical standpoint, to differentiate prose and poetry with regard to both diction and rhythm. The De Compositions of Dionysius 175 Dionysius appears to adopt Aristotle's distinction between prose and poetry, but actually his recommendations are quite diverse from those of Aristotle. For he devotes the final two chapters of his treatise to the consideration of "how discourse not in meter may become like a beautiful poem or lyric, and how a poem or lyric may become very similar to beautiful prose." To these subjects Dio- nysius seems to ascribe much importance, for not only does he accord them a prominent place in his treatise, but he prefaces his disclosures about them with a peculiar proclamation as to their esoteric nature : "They are very like to mysteries, and it is not proper that they be disclosed to everybody. Wherefore it would not be impertinent if I were to summon only those that possess the right to come into the inner precincts of style, but should enjoin the profane to close the gates of their ears." Then follows a statement of the controlling principle: "Prose cannot become like metrical and lyrical writing unless it contains concealed meters and rhythms. However, it is not fitting that it should appear to be meter or rhythm. For then it would be a poem or lyric and would depart from its proper character. It should simply appear rhythmical and metrical, for thus it may be poetical but not poetry, and lyrical but not a lyric. " This statement closely resembles that of Aristotle on the same .subject, but there is an important difference in application. Aristotle made his statement for the sake of calling attention to the distinction to be observed between the proper styles of prose and poetry; Dionysius, on the contrary, calls attention to their points of similarity. Now he proceeds to develop the idea that prose should contain concealed rhythms and meters. He cites the third book of the Rhetoric as authority for this idea, and adds: "We not only have Aristotle's testimony, but we find confirmation in experience itself to the effect that rhythms must be employed in prose if it is desired that it bloom with the beauty of poetry." But, as shown above, Aristotle is not desirous that prose "bloom with the beauty of poetry"; he has a quite different purpose in his consideration of their relationship. Dionysius, however, goes on to exhibit "the concealed rhythms" in a number of passages from Demosthenes, to show that they were introduced "not accidentally, but with full design." Dionysius' final task is to disclose how poems and lyrics become 176 H. P. Bbeitenbach like beautiful prose (c. xxvi). The factors he mentions include variety in the arrangement of the individual words, and also in the length and the form of the clauses, so as not to make them coincident with the lines of verse. The feature of the analysis consists in the element of freedom which Dionysius detects in the most successful poetry, and which he ascribes to the endeavor to make it resemble prose. Judging from the passages he cites as examples and from his comments on them, the chief characteristic of this element of freedom is the employment of sentences and phrases of various lengths, so that the sense does not always come to an end with the line of verse; in a word, the use of enjambement, or run-on lines. In short, Dio- nysius is making a plea for artistic freedom in poetry, freedom from mechanical adherence to traditional or conventional rules. 1 In summary, then, of these two topics with which Dionysius closes the treatise, it may be said that in both cases his aim is to further the cause of the artistic, the aesthetic element in discourse. On the one hand, he objects to awkward, wooden prose; on the other, to mechanical poetry. To remedy the one evil he refers as an exem- plar to prose which resembles poetry in that it has rhythm; in the other case, he holds up as a model poetry resembling prose in that it has freedom of structure. 2 1 The position of Leigh Hunt in the English Romantic movement affords an inter- esting parallel. 2 Theodore Watts-Dunton, in the Encyclopaedia Britannica (9th ed. s.v. "Poetry"), says: "Perhaps the first critic who tacitly revolted against the dictum that substance, and not form, is the indispensable basis of poetry was Dionysius of Halicarnassus, whose treatise upon the arrangement of words is really a very fine piece of literary criticism. In his acute remarks upon the arrangement of the words in the sixteenth book of the Odyssey, as compared with that in the story of Gyges by Herodotus, was perhaps first enunciated clearly the doctrine that poetry is funda- mentally a matter of style. The Aristotelian theory as to invention, however, dominated all criticism after as well as before Dionysius." This statement calls for some comment. In the first place, in the passage from Dionysius referred to (c. iii), Dionysius is not comparing the two arrangements; far from trying to show any contrast between the prose and the poetical passages, he has quoted both for the same purpose — to show the potency of arrangement of words as compared with selection. Furthermore, while certain passages in the Poetics have afforded a foundation for the so-called Aristotelian theory as to invention (see Butcher Aristotle's Theory of Poetry and Fine Art, pp. 134 ff.), yet the Rhetoric, as shown above, contains considerable evidence, to all appearances strangely neglected by Butcher, which tends to show that Aristotle clearly distinguished between the appropriate styles of prose and poetry. Dionysius, on the otherhand, as is likewise shown above, occupied himself with pointing out their similarities rather than their differences. The De Compositions of Dionysius 177 A retrospect of the work of our two authors from the point of view of aesthetics exhibits a paradox. Both of them investigated the same phenomenon, Greek oratory, and in part they examined the same phase of it, the topic style; but their fundamental principles, as has been shown, are altogether diverse. Aristotle supposed on the hearer's part an intellectual attitude; in the speaker, the aim to persuade; in the discourse, an organic character. Dionysius requires in the hearer an aesthetic attitude; in the speaker, the aim to produce artistic effects; in the discourse, an architectural character. How are these strik- ing differences to be explained ? In part they are to be accounted for by the divergent aims and methods of the two writers. But a more ultimate explanation is to be found in the relation of these two men to the art of oratory which they were studying. For Aristotle oratory was a practical art. In his generation it had achieved its utmost successes and had reached a pinnacle of greatness from which only recession was possible. Living accord- ingly in a very world of oratory, Aristotle, especially in view of his scientific proclivities, could not but envisage it with reference to its ulterior purpose and end, and could find it successful only as it achieved these. Necessarily his view was teleological in character. Viewed in such an objective, teleological manner, the oration naturally assumes the character ascribed to it by Aristotle. Its substance or content is then of overshadowing importance. The hearer is concerned with the oration in an intellectual way, i.e., because he is to learn from it something that will direct his beliefs or conduct. But the speaker's business is to make the oration fulfil his purpose; success is the primary thing, to gain which it may be necessary to resort to emotional appeals or even to specious and sophistical arguments. Out of the speaker's vital purpose grows the nature of the discourse. Consequently it assumes an organic character, functioning, like an animal, for an ulterior, practical end. Dionysius' point of view was quite different. When as a teacher of rhetoric at Rome he referred to the masterpieces of his native tongue as examples, he discerned in them an element lacking in Roman oratory; accordingly he analyzed them to try to account for it. In so doing he did not refer to the external purpose which the orations 178 H. P. Breitenbach were designed to fulfil, for Dionysius was lacking in the historic sense. 1 Consequently, given an oration which his native taste told him was superior, he would examine it for the sake of determining the sources of the effect upon himself, and perhaps on others too, his pupils for example. His point of view, then, was not objective in the sense in which Aristotle's was, but rather subjective; it was not teleological but aesthetic. For Dionysius, in short, oratory was a fine art. Accordingly, he did not concern himself with the practical end subserved by the oration, and hence his main interest is not in its substance or content. Instead he analyzed out the sources of its aesthetic effects, and endeavored, from the elements thus detected, to reconstruct the oration synthetically. He assumed consequently that the hearer likewise takes an aesthetic attitude, and that the aim of the speaker, correspondingly, is to produce artistic apprecia- tion. Moreover, the discourse is not now a vital thing, animated by its function in the practical world; it was viewed rather as a structure created by artistically combining certain materials selected beforehand, and selected not so much for their import as for the aesthetic effects they tend to produce. For Dionysius, then, the biological concept was not suitable, since he was emphasizing not content, but form. What he required was an artistic concept, and he found what answered his purposes in the architectural analogy. In concluding, it may be briefly noted that modern rhetorical theory has made use of both the architectural and the organic con- cept of structure. The architectural concept has been held by many of the older rhetoricians in a more or less explicit fashion. With some it is scarcely distinguishable from a mechanical concept, in which discourse is thought of as resembling a machine in its structure. Among others the concept varies from that of a simple assemblage of materials to that of a highly elaborate structure of thought all of which, it is supposed, is preconceived in the mind of the writer. But with the more recent generation of scholars, under the stimulus of evolutionary thought, these theories of style have largely given way to the organic concept. This Platonic theory of discourse has now been developed and extended beyond the Aristotelian treatment !Of. W. Rhys Roberts Dionysius of Halicarnassus : The Three Literary Letters, pp. 3-4. The De Compositione of Dionysius 179 so as not only to include the main outlines of the thought, but also to reach into the uttermost details of form or expression. Moreover, by directing attention to the essential part played in the rhetorical process by the free personality of the writer, the organic concept has suggested the clue for reconciling the seemingly diverse claims of thought and expression. Thus it has opened the way to a truly comprehensive aesthetic theory of discourse — one which includes and, so far as possible, harmonizes the essential features of both the Aristotelian and the Dionysian treatment. University of Michigan