The importance of earlier writers as sources for ideas and allusions of various sorts was expressly acknowledged in antiquity, as is shown by Macrobius' lists in Khz. Saturnalia (especially v.2-22, vi.1-5), which were probably compiled by critics over several centuries, of Virgilian borrowings from Greek and early Latin poets; although it is not clear how concerned these critics were to assess the actual effect of some of these quotations. Modern scholarship has made considerable advances, handicapped by the loss of so many works which were evidently available to Virgil and others. But it has failed to find an altogether satisfactory explanation of the famous echo in Aeneas' address to the ghost of Dido, 'invitus, regina, tuo de litore cessi' (Aen. vi.460) from Catullus' burlesque, 'invita, о regina, tuo de vertice cessi' (56.39), which in turn presumably reflects a line, now lost, in Callimachus' original poem on the Lock of Berenice, on which Catullus' poem was based: it may be that both come from an unknown predecessor of enough solemnity for Virgil not to feel that the pathos of his own context might be spoiled by reminiscences of Catullus' parody.[1120] On the other hand, Virgil strikingly quotes from his own Georgics to provide animal-similes for the Aeneid, evidently when he wants to sharpen the reader's attention and remind him of elements in the earlier context which have relevance to the later.[1121]
There seems to be no comment in ancient criticism on the major Virgilian symbols which play a prominent and continuous part in certain books of the Aeneid, as the snakes and fire do in 11, if not throughout all; but the presence and the effect of these symbols can hardly be denied once they are noticed. The reader's attention is likewise demanded, if scholars are right, by the occurrence of key words in Virgil (but also, it has been suggested, in Persius); and on a much larger scale by consideration of the overall architecture of the book of Eclogues, of the four Georgics, of the whole Aeneid, and increasingly complex diagrams have been produced for the books of the elegiac poets, for Horace's Odes, and for almost every other book of Latin poetry.[1122] It may be significant that the greatest of the Augustans have lent themselves remarkably to the requirements of modern research, so that an endless succession of doctoral theses and published monographs can be extracted from more and more different ways of analysing language, metre, assonance, structure, symbolism and so on, suggesting that all these things were planted by Virgil and the others in their poetry in the expectation that the more appreciadve reader would be equipped to observe them for himself and to gain the more from the work. There is little evidence, however, that this sort of awareness was encouraged by the grammaticus or the literary critic, who were more concerned with correct reading and the understanding of references, in the manner of a good nineteenth-century commentary. It is more credible, though unprovable, that the greatest artist may admit these elements unconsciously, and that the reader may equally unconsciously enjoy and value the work all the more on account of these qualities.
V. THE ACCESSIBILITY OF LITERATURE
The impact of major literature on the great public is hard to assess. The considerable production of tragedies seems never to have reached the theatres, the output of Asinius Pollio and Ovid evidently having little more success than Augustus' abortive Ajax; the Medea is one of Ovid's very few works not to be preserved for posterity (Quint. Inst, x.1.98; Tac. Dial. 12.6). Seneca's surviving plays contain elements of descriptions of their action suggesting that they did not need to be seen to be appreciated, but rather read aloud, perhaps to the accompaniment of dancing or mime. Mime itself, which under the late Republic retained some of the literary quality of Herondas and Sophron, still found no favour with Horace (Sat. 1.10.5-6). Under the Empire it came to depend more and more on the obscene and the spectacular, including real sex and real crucifixions, until it seems to have merged with pantomime. This never ranked as literature, despite the libretti derived from Virgil and Ovid and others specifically written by Lucan and Statius.
The most significant type of public performance becomes the recitatio, of poetry and prose alike. This seems to have been the regular way of launching a new work before the publication of an approved text.26 Only after such an occasion, and the correction of faults which might have come to light, would the author make the work available to the public; and this, in the case of Virgil and Persius, and probably Lucan, might mean posthumous publication. Subsequent reading might also take the form of an oral performance, often by specifically trained slaves, to an individual or a group. In addition, we hear of public performances by professional cantores, as something quite distinct from Virgil's own reading of the Georgics and of three books of the Aeneid to the imperial household, as well as trials of various passages before a rather wider
24 Kenney in Kenney and Clausen 1982 (в 95) 12.
audience (Suet. Vita Ver. 27, 32-3). Horace in particular (Sat. 1.4.73-7) emphasizes that he never recites his works except to selected friends and on their express insistence; while others take advantage of the crowds in the public baths to force their works on all sorts of listeners. The importance of all these readings will have been greatly increased by the extent of aural memory enjoyed by a society with far less written material than the modern world possesses, so that whole passages appear to have been retained by hearers, with varying degrees of accuracy.[1123]
But apart from the fact that reading, like writing, was almost always carried out aloud, the general status of reading appears to have borne a considerable similarity to that of our own day, with bookstalls selling copies for personal enjoyment and most works available in the great public libraries, which begin almost exactly with the Augustan age and expanded rapidly in the following two centuries.[1124] We still hardly know the extent of these collections, nor how far the different libraries duplicated each other. It would appear that readers would normally consult books inside the libraries, as in the British Library or the Bodleian. It may have been exceptional, and a matter of privilege, for Marcus Aurelius in the second century to report to Fronto that he has taken certain volumes of Cato out of the library of Apollo on the Palatine and advise him to bribe the librarian of the Tiberian collection to let him have copies from there.
Some works were evidently produced in fairly large numbers, with individuals having copies made by their own slaves from a borrowed text; others probably never merited marketing to any effective extent. Survival down to the Renaissance is little indication of the availability of works in antiquity: the fact that Velleius has come down to us largely complete cannot be proof of wide circulation. On the other hand, there is reason to suppose that Juvenal made so little impact in his own day that he survived only because of a surprising popularity in the fourth century, attested by Ammianus (28.4.14), when there was a sudden demand for improved texts, enriched with scholia and commenticious biographies of the author.29 Gellius provides some interesting stories, not always plausible, of discovering rare texts in unlikely places (e.g. ix.4.1, xvin.9.5); yet Quintilian can recommend for the student's reading a very wide range of authors as undistinguished as Rabirius and Albinovanus Pedo (e.g. x. 1.90), who must at least have been available in one or more of the public libraries. It is a bolder assertion that the libraries in Rome also contained copies of all the obscure works cited only by Dionysius of