SOURCE: "A Greek Professorial Circle at Rome," Transactions and Proceedings of the American Philological Association, Vol. XCII, 1961, pp. 168-92.

[In the following essay, Goold explores what can be determined about the author of On Style and its date of composition, arguing against some of G. M. A. Grube's positions (see previous excerpt).]

To establish beyond a shadow of doubt the date and authorship of the tractates On the Sublime and On Style is a task which has long defied—and in all probability will continue to defy—the best endeavors of scholarship. But the seeker after truth is a detective, not a magistrate; and like a detective, he will often feel satisfied that he has solved a case, though fully aware that the evidence is highly circumstantial and will not convince all of the jury. Of such a nature is the present inquiry. Nevertheless, the matter has been recently brought into court and argued at length by two eminent advocates;¹ and rightly so, since keeping an open mind is a poor excuse for remaining in the dark. Yet in the dark I fear they will have us stay, for neither has mentioned the solution which falls least short of certainty and throws most light on what was before obscure. The following plea is therefore tendered on behalf of a Greek professorial circle at Rome—indocti discant, et ament meminisse periti—before the tribunal of the learned retires to consider its verdict.…

II. Demetrius On Style

Now that G. M. A. Grube²¹ has assigned the tractate On Style to the early reign of Ptolemy Philadelphus, the orthodox opinion which puts it three hundred years later is confronted with a challenge long overdue; and in an introduction replete with learning Grube has no difficulty in exposing much argument to which scholars had clung in their unhappy incertitude as inconclusive, for example, that "broad purples" in 108 refers to the senatorial laticlave, or as most improbable, for example, that the corrupt name in 237 (of a writer alluding to the battle of Salamis…) is that of the rhetorician Theodorus of Gadara. Yet the fresh interest which Grube's vigorous study will stimulate is likely to reveal the paradox that his carefully argued conclusions are less securely based than the perfunctory ones which he opposes. His position is fundamentally a cautious one, characterized by a disinclination to abandon the authorship and hence the date dubiously indicated by the chief manuscript; and his ascription of the work to about 270 B.C. is not so much a safe arrival in port as a reluctance to navigate in the darkness. But, alas, the quest for truth is always dangerous, and caution can prove as fatal as daring. Ond' io per lo tuo me' penso e discerno / Che tu mi segui, e io sard tua guida.

To begin with the external evidence: the Codex Parisinus 1741, the archetype of the other manuscripts, exhibits the title [Demetriou Phalereos peri hermeneias ho esti] and the colophon [Demetriou peri hermeneias] The latter preserves all that is genuine. The last four words of the title, which have all the marks of a gloss, appear to have been added by way of explanation, and to indicate that [Phalereos], which word constitutes the whole of the evidence for the Phalerean's authorship, is a fellow impostor, possibly transferred by some sciolist from 289. This section, in which Demetrius of Phalerum is quoted by name, all but proves that our Demetrius is not the Phalerean, for, whilst many writers ancient and modern have referred to themselves in the third person, such a practice in a teacher of style is sham modesty and would never be limited to a single instance. Moreover, Demetrius of Phalerum, an Attic orator of distinction and good enough for comparison with the best, certainly wrote in Attic. Demetrius the Stylist, if I may so call him, certainly does not. Finally, no treatise entitled [Peri hermeneias] is attributed to him by Diogenes Laertius in the latter's register of his works (5.5.9). The case against his authorship is conclusive.

We possess no reference to Demetrius the Stylist before the Christian era. The alleged citation in Philodemus is a mirage which vanishes at our approach. Philodemus²² declares that "long periods are awkward to deliver, a criticism Demetrius brings against those of Isocrates." Look for such a criticism in the tractate On Style, and you will look in vain. It is not there. Our Demetrius mentions Isocrates three times, once (12) to cite him with Gorgias and Alcidamas as an example of periodic writing, and twice (68, 299) to comment on his avoidance of hiatus. One cannot infer from our Demetrius that he disapproved of the length of Isocrates' periods, or even regarded them as "long" at all. When our Demetrius in 303 briefly warns against "periods which are continuous and long, and make a speaker run out of breath," he does not bring this as a criticism against anybody; no orator is mentioned by name or implication. Only by joining together what our Demetrius has put asunder and imputing to him something which he does not expressly say has Grube²³ been able to refer Philodemus' statement to him and so claim that he ante-dates the latter, who flourished between 70 and 40 B.C. There is a simple explanation: Philodemus is not citing our Demetrius at all, but Demetrius of Phalerum, whom he has mentioned elsewhere and to whose [Peri hretorikes] he twice²⁴ makes explicit reference.

The one sure piece of external evidence is furnished by the fifth-century Syrianus in his prolegomena to Hermogenes' De ideis,²⁵ where he proffers the following opinion: "The critics who ventured to enumerate and define the various styles embarked on a futile task. For Dionysius there are three: the plain, the middle, the grand. Hipparchus adds the graphic and the florid, whilst Demetrius rejects the graphic and is satisfied with four." That Dionysius of Halicarnassus and Demetrius the Stylist are designated here is not open to doubt. Dionysius opens his appreciation of Demosthenes²⁶ with a survey of precisely the three styles mentioned by Syrianus, and later tells us that whereas other eminent Greek writers are to be regarded as representative of one or another, Demosthenes achieves excellence in all three. And Demetrius' work is distinguished from all others by precisely such a fourfold division of styles as Syrianus attributes to him, three roughly corresponding to those of Dionysius, together with a fourth, largely designed to put the overpraised Demosthenes in his place. Do I hear someone object that Syrianus implies that Demetrius' fourth style is anthêros and not deinos? And that Demetrius' "style" has a different significance from Dionysius'? He takes the words from my mouth. But Syrianus' explicit assertion that Demetrius preaches a classification of four styles, from which the graphic is excluded, is unchallengeable. As for implications, he did not intend any; an unbeliever scoffing at articles of doctrine does not pick his words with a view to elucidating the catechism. Syrianus, however, has not been called to the witness-box as a technical expert; he is subpoenaed solely as a lay witness to a point of time. His testimony, associating Demetrius with the Graeco-Roman period of literary theory, enables us to assign the tractate On Style to an age contemporary with or not far removed from that of Dionysius of Halicamassus, who flourished at Rome in the first thirty years of the principate; since he mentions before and not after Demetrius the unknown Hipparchus whose system embraced five styles, it rather looks as if Syrianus has named the three critics in chronological order, as if, that is to say, Demetrius' work is later than Dionysius'.

No safe argument can be built on the dates of the authors quoted or mentioned by Demetrius. They cover a period from Homer to Sotades (early third-century B.C.) and permit the opinion, to which my other arguments compel me, that Demetrius regards them as "classical" authors. We encounter a similar range of authors, a similar recognition of a "classical" period, and a similar silence about the contemporary scene in other rhetorical works, in Dionysius' De compositione verborum, for example, which is of the Augustan period. Grube maintains that references to writers of the later fourth and early third centuries are "unusually frequent."²⁷ They have not seemed so to any other scholar. None of these writers, with the exception of Menander, is mentioned more than once, and Menander is only mentioned twice; Sappho is referred to more times than all put together. I do not count Theophrastus and his disciple Praxiphanes, whom Demetrius cites as technical authorities. Nor do I count Archedemus and Artemon, two more technical authorities cited in 34 and 223 respectively. Apart from these two, all writers named by Demetrius have been identified with persons otherwise known, and this circumstance has led to the plausible conjectures that here are mentioned Archedemus of Tarsus and Artemon of Cassandria, who are generally assigned to the second half of the second century B.C. TWO proverbial sayings quoted by Demetrius in 172 are also quoted by Seneca and Diogenes Laertius, who name Chrysippus (second half of the third century B.C.) as their source; hence the plausible conjecture that Chrysippus has been used by Demetrius also as a source.²⁸ If any one of these three conjectures is true, Grube's date would be refuted at a blow. We should not assume that, because confirmation of these conjectures is precluded by our lack of knowledge, they are false and the possibility of them being true may be safely disregarded.

I digress to consider a special passage. In 193 Demetrius remarks that a disjointed style is appropriate to the stage, and "for this reason Menander, whose style is for the most part broken, is acted, whereas Philemon is read." This statement is most readily referred to a period subsequent to the careers of those writers; when Demetrius wrote, Philemon's plays were no longer acted; and Philemon did not die till 262 B.C. But the passage is chiefly interesting for the singularity of the comparison. It was Philemon and not Menander who was preferred by contemporary theatre-goers; and it is no rash inference from Plautine Quellenforschung to assert that Philemon held his own on the stage for at least the rest of the century. Only with the activity of Terence is it clear that Menander has been awarded posterity's accolade. Even then Philemon was not pushed off the boards at once. Performances of his pieces took place as late as the second century of our era, when a statue commemorating him was erected at Athens. How then could Demetrius take for granted a ready acceptance of his statement? Only, I think, if he was writing long after the age of Plautus, yet at a time when Philemon was still quoted, that is, was still frequently read. The tum of the millennium is the time which would best accord with our professor's remarks; and it should be noted that Rome is a likely place, where the plays of Menander were often performed.²⁹

The evidence of language and style has been fully set forth by Roberts and Grube in their editions,³⁰ and there is no need for another tabulation. Grube, who is obliged to prove that Demetrius is not writing a prose which became current only three centuries after his death, has attempted a thorough rebuttal of all the evidence adduced by Roberts in support of a late date. Yet, even if we accept as proved the contention that Demetrius' prose could have been written in 270 B.C., it is highly improbable that it was then penned by a teacher of style. By Grube's account Demetrius was a younger contemporary of Theophrastus and Menander; his teachers had listened to Demosthenes and Aristotle; he himself aspired to become the Strunk of his age. Such a man was bound to employ an Attic pure to the point of pedantry. We do not expect in a manual on style the unproved fashions of up-to-date language; H. W. Fowler does not write like P. G. Wodehouse. We expect a syntax, diction, and discipline sanctioned as authoritative by the best usage of the preceding age, and unmarked by the novelties of the rising generation. In this regard the variegated koinê of Demetrius must be held to favor a date distinctly later than that maintained by Grube.

Nor should it escape notice that in his assessment of the lexicographical evidence Grube has frequent recourse to the argument of analogy. If a word in Demetrius is of a type or formation analogous to that of a word found in the classical period, there is no reason—Grube maintains—why Demetrius' word could not have been current in the classical period, even though it is not attested until much later. Accordingly, since philophronoumai is found in Herodotus, Xenophon, and Plato, Grube considers philophronêsis "natural enough"³¹ for 270 B.C., even though it is not found "again" until Dionysius of Halicamassus, Josephus, and Plutarch. Usage, however, is not determined by analogy, but by an infinite number of usually undiscoverable factors; caprice and not reason is the final arbiter. The words dread, dreadful, dreadfully; peace, peaceful, peacefully are attested in written English by 1300. So are the words force and help; yet forcefully is not attested until 1774, and helpfully not until 1832. Strengthful and sleepful occur before 1400; their adverbs have yet to make their debut; and it is not through lack of written English that these words are not found from Chaucer to Hemingway. The lesson taught by these facts is that in order to date a literary work by its vocabulary we must proceed empirically, treating each word—no matter how closely related to or modeled upon another word it may be—as a separate entity. No exception can be taken to Grube's statement³² that Demetrius's cacophônia ("ill-sounding-ness") is a perfectly natural formation from Aristotle's cacophônos. But the question is: when did this perfectly natural formation acquire an actual as opposed to a potential existence? For we must not assume that cacophônos was bound to give birth to a noun. Strengthful did not give birth to strengthfully, though it had a start of two hundred years on forceful. Again, cacophônos might just as naturally have produced calophônos, but the word is not recorded. Since our source-material is necessarily incomplete, and particularly so for the Hellenistic period, it is not possible to draw a reliable conclusion from a single word. But when we find that not one word, but a second, and then another, and then yet more, lack attestation before Graeco-Roman times, and when we find Grube resorting on each occasion to the argument of analogy to maintain his defence of an early date, then such a defence draws upon it greater and ever greater suspicion, until finally it is no defence but a self-indictment.

Aristotle (Rhet. 3.2.1404b) had taught that the ideal style (lexis) is a lucid mean, capable of embracing simplicity on the one hand and elevation on the other; and if Theophrastus preserved this teaching unchanged in his [Peri lexeos],³³ his words at any rate seem to have misled Dionysius³⁴ into thinking otherwise, and may have led ultimately to the tripartite classification of styles (Middle, flanked by Plain and Grand) first attested in the Rhetorica ad Herennium, ca. 85 B.C. At first sight the four styles of Demetrius have no relationship to this tripartite classification. The charactêres are not for Demetrius, as they are for Dionysius, separate personalities, of which a writer may possess only one; they are rather suits of clothes in his wardrobe, each available for wear on the appropriate occasion. Yet a close look shows that, whatever his conception of them, Demetrius' styles are, in origin, of the same nature as the styles of Cicero and Dionysius. Grube³⁵ is rightly suspicious of Solmsen's³⁶ attempt to link them to the four Theophrastean virtues. To record that Demetrius' styles are elevated, elegant, plain, forcible and Theophrastus' virtues correctness, lucidity, propriety, ornamentation is refutation enough. Of Demetrius' styles the plain (ischnos) is the same as the plain (litos) of Dionysius; the elegant (glaphyros), though not identical with the middle (mesos) of Dionysius, is closely connected with the middle of Philodemus and Cicero;³⁷ the elevated (megaloprepês) is the same as the elevated or grand (hypsêlos) of Dionysius. When Cicero³⁸ represents the plain and the grand as two extremes, and Demetrius (36) says that the plain and the elevated stand in irreconcilable opposition and contrast, we are justified in concluding that Demetrius had after all points of contact with the orthodox Graeco-Roman school, and that Syrianus³⁹ knew what he was talking about. I have already indicated why Demetrius added a fourth style, the forcible (deinos). I think he objected, as Longinus objected, to the perpetual hosannas which were everywhere raised at the name of Demosthenes. Merely to show that the orator variously employed the three available styles would be to add color to the Dionysian dogma that the greatness of Demosthenes transcended the limitations of other Greek stylists. By associating Demosthenes with a fourth style, indeed by practically demoting him to it, our author fashioned an effective weapon against the teaching of the Dionysians. The perspicacious will notice that increasing the number of styles from three to four necessitates changing the name of the old middle style, for the second of four is obviously not mesos. It is possible that it was Hipparchus and not Demetrius who was responsible for the innovation. There are, however, obstacles in the way of placing Demetrius either before or after Dionysius. Had Demetrius' four-fold classification been propounded before the time of Dionysius, it is strange that the latter should nowhere have thought fit to criticize it. On Style cannot be ante-Dionysian. But it cannot be later either. 179, wherein Demetrius states that no earlier critic had treated of "elegant word-order," strongly suggests that On Style was written before Dionysius' De compositione verborum. It seems best to regard the two men as contemporaries. In any case unprofitable wrangling over the stylistic classification of authors soon discredited this section of rhetorical teaching; in Dionysius' own lifetime Longinus abandoned it, and in the succeeding age we look in vain for further disputation. Henceforth Greek and Roman literary theorists tum to the practical field of oratory. Technical considerations prescribe for Demetrius's On Style a late Augustan date.

It is important to bear in mind that Demetrius "must have had strong Peripatetic connections," and that his attitude towards Aristotle "might be described as respect falling short of veneration." But when Grube, whose words these are,⁴⁰ suggests that the relationship between Demetrius and Aristotle is that of two critics not far removed in time,⁴¹ some remonstration is called for. The relationship is, on the contrary, similar to that which might exist between a modern Roman Catholic theologian and Thomas Aquinas. On fundamental matters a chronological proximity might appear to be indicated, but differences of emphasis in the application of doctrine and a greater precision in technical minutiae, not to mention general linguistic discrepancies, would reveal the lapse of the centuries between them. So it is with Aristotle and Demetrius.… Demetrius seemingly disputes an appraisal of Demosthenes not formulated till the time of Cicero and Dionysius, and this in a chapter which "That our "has practically nothing that recalls Aristotle."⁴³ "That our author's relation to Aristotle is unique in extant critical texts"⁴⁴ is but another way of saying that the text of Demetrius the Stylist is the only literary treatise we possess written by a confirmed Peripatetic.

Now, if we search the Augustan age for a literary professor who had strong Peripatetic connections, who thoroughly knew his Aristotle, and who opposed the Dionysian idolatry of Demosthenes and taught that attention to Aristotelian precepts was more important than Demosthenic mimêsis, we shall sooner or later alight on a letter of Dionysius in which such a literary professor is mentioned, though unfortunately not named. He answers so closely to the description of the missing man as to warrant the conjecture that he is Demetrius himself, the author of On Style. I quote the passage in full in Roberts's translation.⁴⁵

You said that a certain Peripatetic philosopher, in his desire to do all homage to Aristotle the founder of his school, undertook to demonstrate that it was from him that Demosthenes learnt the rules of rhetoric which he applied in his own speeches, and that it was through conformity to the Aristotelian precepts that he became the foremost of all orators. Now my first impression was that this bold disputant was a person of no consequence, and I advised you not to pay heed to every chance paradox. But when on hearing his name I found him to be a man whom I respect on account of his high personal qualities and his literary merits, I did not know what to think.

Notice that the words "on account of his high personal qualities" … permit the possibility of personal acquaintance, and this is placed beyond doubt by the concluding words of the paragraph: "I wished therefore… to induce the person who has adopted this view, and is prepared to put it in writing, to change it before giving his treatise to the world." So the Peripatetic teacher of rhetoric is accessible to Dionysius' demurs, and has actually announced a forthcoming treatise!

At last I see land. But perhaps there are some who are unconvinced that we are making for shore? They dispute the identification and will not assent until they find Demetrius' actual name in Dionysius' text. Let them turn to Dion. Hal., Ep. ad Pompeium 3, Roberts 104, lines 8-11: "You wished also to learn my view with regard to Herodotus and Xenophon, and you wished me to write about them. This I have done in the essays I have addressed to Demetrius⁴⁶ on the subject of imitation." This Demetrius was a teacher of Greek literary criticism like Dionysius himself. That much is guaranteed by the context. Dionysius addressed his works either to Roman patrons (doubtless students and friends as well) like Quintus Aelius Tubero and Melitius Rufus or to professional associates like Gnaeus Pompeius Geminus, the author of The Sublime. With a name like Demetrius, the addressee of a treatise must have belonged to the second category. Observe that elsewhere⁴⁷ in this letter to Pompeius (or Longinus, as the world will continue to call him) Dionysius names Demetrius of Phalerum in full as [Demetrios ho Phalereus], whilst the professional associate is referred to quite simply as Demetrius. It follows that Longinus knew Demetrius, at least by name, and they may have enjoyed closer acquaintanceship.

Grube further urges as supporting his position: "There is no trace [in the tractate On Style] of the rhetorical mimesis or emulation so important to Cicero, Dionysius, and Quintilian."⁴⁸ This I should explain by asserting that Demetrius is a staunch Peripatetic whose relationship with Cicero, Dionysius, and what I may call the orthodox Roman school may be likened to that of a Roman Catholic theologian with Protestant ones. There is no reason why contemporary literary critics at Rome should have differed any less than Christian thinkers do. And no less than these did the former find themselves at variance with each other. Longinus, we saw, had ventured to disagree with Dionysius on the matter of Plato's style; and an elaborate reply duly reached him.… It would hardly be surprising if Dionysius found fault with the absence of mimêsis from Demetrius' teaching, and labored to persuade his fellow-professor to include it in his syllabus. Nor, scio quod dico, if he herein labored in vain.

Further external evidence about the author of On Style does not seem to exist. Many Demetriuses find place in the vast mass of material which antiquity has bequeathed to us. One would like to know more of that Demetrius unflatteringly immortalized in the tenth satire of Horace's first book; and of the Peripatetic Demetrius (Plut. Cat. Min. 65, 67 ff.) who attended the Younger Cato in the extremity of his fortunes. Some, discerning in our Demetrius an "apparent familiarity with Egypt,"⁴⁹ have suspected that Alexandria rather than Rome was the place of composition; and others besides Muretus have deemed this suspicion confirmed by Diogenes Laertius, who, in his list of twenty celebrated Demetriuses, records (5.5.11) that "the eighth is the professor who resided at Alexandria, the author of rhetorical treatises." This cannot, however, be our man: whether or not Diogenes took over his biographical material from Demetrius of Magnesia (fl. 50 B.C.), his list seems to adhere to a general chronological sequence, and this necessitates placing Diogenes' Demetrius who lived at Alexandria and compiled rhetorical treatises in the second century B.C.

It is not possible to determine with precision the date at which our Demetrius issued his tractate. We are restricted to a rough estimate based on the activity of Dionysius, the only dates of whose life known to us are 30 B.C., when he came to Rome, and 8 B.C., when he wrote the preface to his Antiquitates Romanae. We have no notion how old he was on either occasion, nor when the literary treatises were written. As Longinus had corresponded with Dionysius, and The Sublime has been fixed at A.D. 12, the probability is that some at least of Dionysius' critical work was written later than 8 B.C. I therefore suggest that, since the contents of Demetrius' tractate are not discussed anywhere in Dionysius, but a work by Demetrius is heralded in the First Letter to Ammaeus, the year A.D. I as the approximate date of On Style cannot be far out.

III. Greek Professors in Rome

Possibly, if our information were not so scanty, we might find that men like Caecilius and the other friends of Dionysius, like Theodorus of Gadara, like the author of the [Peri hypsous], like the author of the [Peri hermeneias], and even like Manilius. … had this in common that they belonged to the age of Augustus …, and further resembled each other (in some instances) in being freedmen or sons of freedmen attached to the great Roman houses such as that of Pompey, and in having an Eastem or Jewish origin.⁵⁰

It is a pity that Rhys Roberts, to whom every student of Greek literary theory is deeply indebted, should in his search for the authors of our two tractates have later indulged in improbable hypotheses when he might without absurdity have amplified his statement above. This amplification I have now endeavored to supply, claiming that it merits priority over all other theories, for even the protestation that the author of The Sublime is unknown is essentially a conjecture—no less audacious than mine—that the name of the author occurs nowhere in ancient records.

The association of Dionysius, Longinus, and Demetrius enables us to appraise more accurately the aims and methods of higher education and more particularly the position of the Greek language in early imperial Rome. Their circle was not a literary, but a professorial one; they had no connection with and showed no awareness of. Roman poets and Roman literature. Manilius is Romaic rather than Roman. The sole reference to Cicero (De subl. 12.4-5) makes of the greatest wielder of the Latin language only an illustrative footnote. The Dionysians were not mere grammatici. Their courses were not "Greek without tears" or even Greek courses at all; their works presume in those they taught—and these must have been Romans—a complete mastery of Greek and a wide acquaintance with the Greek classics. They were Professors of Literature, which meant Greek literature, since they had not the faintest idea that Latin—and not Greek—was to become the common tongue of Western Europe and was already pregnant with the speech of modern civilization; they were Professors of Classics, like ourselves, with their attention focused on the genius of an age long since passed; and their writings contain no direct reference to the times in which they lived. To judge from their apparent ease of movement and communication, they enjoyed the privileged life of university men, for Dionysius could send junior colleagues with offprints of his work, as it were, to Longinus and Demetrius, whilst Longinus had heard Theodorus of Gadara in Rhodes⁵¹ and had secured a copy of Caecilius' dissertation as soon as it came on the market.

Their methods of teaching literary style suffer from the misconception, exposed as such by the very language they wrote, but still prevalent two thousand years later,⁵² that classical Attic was the absolute standard for Greek; imitation of Demosthenes was in the age of Augustus a counsel, not of perfection, but of impossibility. Our Greek professors recognized of course that certain virtues of style are independent of a particular language, but this makes all the more significant their failure to consider Latin. It is very probable that Greek was much more widespread in ancient Italy than we commonly suppose; and that many in Augustan Rome, including our Greek professors, entertained hopes that Greece would take her captor captive in the matter of language.

True bilingualism in a nation (as opposed to bilingualism in marginal areas like Alsace-Lorraine) is a very rare phenomenon; man is by nature unilingual. In bilingual environments the struggle of one language against another, even within the personality of a single individual, rages unceasingly; equality may be attainable under favorable circumstances, but equilibrium never. The greater the communicative power of a language is felt to be, the less able are its speakers to feel and exploit the power of another (or, if you like, the less ready they are to learn another). Few speakers of English do not feel its communicative power to be absolute; hence the vast majority of English-speakers are unilingual; of the total number of bilingual persons to be found, for example, in Canada or South Africa or Wales, the vast majority speak as their mother tongue French or Afrikaans or Welsh. In the Augustan age Greek and not Latin probably enjoyed the greater communicative power.⁵³ It was the Romans who were eager to learn Greek from their slaves, not the slaves Latin from the Romans. Outside Italy Greek was still entrenched at Massilia and Syracuse, whilst beyond the Adriatic—with the curious exception of Dacia—Latin hardly secured any foothold, not even later when, surrounded by grandees of church and state, Roman emperors held court in the cities of the Eastern Empire. In Italy and in Rome itself, the Greek-speaking population, though servile, fell not far short of the Roman; most books—and most of the classics—were Greek, and yet we never hear of "Loebs" or literal translations; Strabo was writing his Geography in Greek, and Diodorus Siculus his Historical Library; commerce with abroad was conducted, medicine practised, and the sciences investigated in Greek. Small wonder that Romans generally exhibit an extraordinarily fluent knowledge of Greek; it was no school-Greek, to be compared with the school-French of English-speakers, in which eminent Romans from Gracchan to imperial times addressed Greek audiences, wrote letters—tragedies, even—or in their leisure moments conversed at dinner-parties.⁵⁴

Such is the conception of Greek entertained by Dionysius, Longinus, Demetrius, and, we may be sure, many an ancient professor besides. Greek was for them the chief language of the empire and the world, and they probably hoped that it would one day reign supreme in Rome itself. That their hopes were blighted by the march of history does not necessarily mean that their conception of Greek was unnatural or unjustified at the time. The practical fruit of their labors matured a century later in the widespread and lasting return of literary aspirants to an artificial Attic. Their real achievement, however, securely based on a tasteful appreciation and genuine love of all that was good and beautiful and effective in speech and writing, was their preservation of the living voice of Hellas. And Time's stern, searching judgment, from which there is neither dissent nor appeal, has upheld their teaching of the classics by preserving their voice, too.

Notes

¹ Eduard Norden "Das Genesiszitat in der Schrift vom Erhabenen," Abh. der berl. Ak., Kl. f Sprachen, etc., 1954 (Berlin 1959) Nr. 1. G. M. A. Grube, A Greek Critic: Demetrius On Style (Toronto 1961) 39-56.…

²¹Op. cit. (above, note 1).

²²Rhet. 1.198.9 f., ed. Sudhaus.

²³Op. cit. (above, note 1) 53, and 56, "highly probable."

²⁴Rhet. 1.274.4; 1.346.1, ed. Sudhaus.

²⁵Rhet. Gr. 7.93, ed. Walz. Writing about A.D. 500 the Aristotelian commentator Ammonius (Comm. Arist. Berl. 4.996-97) also mentions Demetrius the Stylist, but furnishes no information bearing on his date.

²⁶De Dem.. 1-7.

²⁷Op. cit. (above, note 1) 41, bottom.

²⁸ Seneca, De const. sap. 17; Diogenes Laertius 7.1,2; quoted by W. Rhys Roberts, Demetrius On Style (Cambridge 1902), note on 150, 15, page 241.

²⁹ See, for example, Anth. Pal. 9.513; Quintilian, IO 11.3.91; and Cassius Dio 60.29 (Book 61, Boissevain, vol. 4, page 1), who quotes, as from the same poet, Epitrep. 157 and Kock, FCG 3, Adesp. 487, evidently overlooked by Koerte in the Teubner Menander, vol. 2.

³⁰ Roberts, op. cit. (above, note 28) 55-59; Grube, op. cit. (above, note 1) 46-50, 133-55.

³¹Op. cit. (above, note 1) 152.

³²Op. cit. (above, note 1) 141.

³³ See G. L. Hendrickson, AFP 25 (1904) 125-46.

³⁴ Dionysius (De Dem. 3) implies that Theophrastus considered Thrasymachus a representative of the middle style. For the sake of clarity I do not discuss in this paper Dionysius' complete system of tripartite classification, which contains three styles of diction as well as three styles of composition.

³⁵Op. cit. (above, note 1) 50, note 66.

³⁶Hermes 66 (1931) 241-67.

³⁷ Of the middle style Cicero says (Or. 96): "Est enim quoddam etiam insigne et florens orationis genus pictum et expolitum in quo omnes verborum, omnes sententiarum illigantur lepores." This is identical with Demetrius' glaphyros. Similarly Quintilian (IO 12.10.58): "Tertium alii medium ex duobus, alii floridum (namque id anthêron appellant) addiderunt." The relevant sentence in Philodemus (Rhet. 1.165, IV 2-5, ed. Sudhaus) is unfortunately mutilated, but is restored by Radermacher …, rightly interpreted by Hubbell (Trans. Connecticut Acad. 23 [1920] 297-98) as "plasma refers to the distinction between grand and plain and middle or smooth style," since one cannot have a middle style of four styles.…

³⁸Brutus 201, with Hendrickson's note in the Loeb edition (1939) 172 f.; Orator 20.196-97.

³⁹ See above, note 25.

⁴⁰Op. cit. (above, note 1) 56 and 37.

⁴¹Op. cit. (above, note 1) 38, line 7.…

⁴³ Grube, op. cit. (above, note 1) 33.

⁴⁴ Grube, op. cit. (above, note 1) 37.

⁴⁵ Dion. Hal., Ep. ad Ammaeum 1.1, Roberts (above, note 9) 52 f

⁴⁶ This Demetrius as the author of On Style was tentatively considered by W. Rhys Roberts, CR 14 (1900) 440b.

⁴⁷ Dion. Hal., Ep. ad Pompeium 2, Roberts (above, note 9) 98, line 23.

⁴⁸Op. cit. (above, note 1) 51.

⁴⁹ Grube, op. cit. (above, note 1) 52. In identifying Demetrius as a member of a Roman professorial circle, I am not to be held as rejecting the possibility that his origin and even his literary training may have been Alexandrian.

⁵⁰ W. Rhys Roberts, CR 14 (1900) 440 f.

⁵¹De Subl. 3.5. If Longinus did not hear Theodorus (the natural interpretation in view of the chronology involved), he must have studied Theodorus' works during the latter's lifetime. I am thus not troubled by Grube's doubts, op. cit. (above, note 1) 41, note 46.

⁵² See above, note 18.

⁵³ Bilingualism in ancient Rome is discussed in his Stranger at the Gate² (Oxford 1948) 309-26 by T. J. Haarhoff.

⁵⁴ It was surely in Greek that Pontius Pilate (though "eminent" is hardly le mot juste for him) cross-examined Jesus Christ; none of the gospels makes mention of interpreters at the trial.