¹ Only the author′s name will be used to refer to the following frequently cited works: Kenney = E. J. Kenney, ‘Books and Readers in the Roman World’, in The Cambridge History of Classical Literature,n, Latin Literature,ed. E. J. Kenney (Cambridge, 1982), pp. 3–32; Kleberg = T. Kleberg, Buchhandel und Verlagswesen in der Antike(Darmstadt, 1967), trans, E. Zunker from the original ed. (Stockholm, 1962); Quinn = K. Quinn, ‘The Poet and His Audience in the Augustan Age’,, 4 WRff 30.1 (Berlin, 1982), pp. 75–180; Sommer = R. Sommer, ‘T. Pomponius Atticus und die Verbreitung von Ciceros Werken’, Hermes61 (1926), 389–422.

² Cf. Cic. Alt. 15.27 (SB 406).2; 13.21a (SB 327)Google Scholar

³ SeeAtt. 16.11 (SB 420). 1; 15.14 (SB 402).4, with Shackleton Bailey′s n. adlocGoogle Scholar

⁴ An. 15.1a (SB 378).2.Google Scholar

⁵ See, e.g., Ep. 7.20, about exchanging works for comment with TacitusGoogle Scholar

⁶ Catullus 35 may reflect such a situation, but the poem has not and probably cannot be elucidated with complete confidence. See Quinn′s commentary ad loc. and F. O. CopleyGoogle Scholar

⁷ Ep. 5.12.1. Although the sincerity of Pliny′s desire for serious criticism has been questioned (Quinn, 163), such scepticism is belied by Pliny′s virtual obsession with testing his work on small groups of friends (Kenney, 11)

⁸ Ep. 5.3.8; cf. 5.12.1; 8.21.4.

⁹ Respectively, Pliny, Ep. 3.7.5; Don. Vita V. 33; and Vita Hor. Cf. Ovid, Tr. 4.10.41ff., although the works there may be at a more advanced stage of revision.Google Scholar

¹⁰ Cf. Quinn, 83 n. 23.

¹¹ See Shackleton Bailey′s n. on Cic. Att. 13.21a (SB 327).l, on the Acad.Google Scholar

¹² See the fundamental paper of Sommer, op. cit. (n. 1).

¹³ Cf. Mart. 7.11, 7.17; Pliny, Ep. 4.26.1.

¹⁴ Cf. Cic. Att. 13.21a (SB 327). 1

¹⁵ Att. 12.6 (SB 306).2, with Shackleton Bailey′s n. adloc.

¹⁶ Alt.13.48(SB 345.2s

¹⁷ 7.3. For a Martial poem announcing a gift, see 9.58. Tacitus ironically comments that Caesar and Brutus ‘ fecerunt enim et carmina et in bibliothecas rettulerunt, non melius quam Cicero, sed felicius, quia illos fecisse pauciores sciunt’ {Dial. 21.6). D. Bo notes, ad loc, that the libraries must be private since Rome did not have any public libraries at that time.

¹⁸ The basic discussions are Sommer and B. Van Groningen, ‘EKAOZIZ’, Mnemosyne 16 (1963), 1–17. See also, most recently, Kenney, 19, and J. E. G. Zetzel, Latin Textual Criticism in Antiquity (Salem, 1981), pp. 232–5. The term ‘publish’ should not be used because it unavoidably bears a burden of modern implications. The model of gradually widening circles might seem to leave no room for the special case of authors who died before their works, or at least some of them, had been fully released to the public, e.g., Vergil′s Aen. and Lucretius. Even in such cases, however, the earlier steps of composition, limited circulation, and initial revision would not be missing. Literary executors, of course, might not abide by an author′s wishes, as Varius and Tucca (and Augustus) set Vergil′s aside, but they presumably still released works to strangers through the normal channels.Google Scholar

¹⁹ Although it should not be made to bear undue weight, we may see evidence of works that existed only in author′s copies in Varro′s statement that, of the many books he had written, ‘aliquammultos, cum proscriptus esset, direptis bibliothecis suis non comparuisse’ (cited by Gellius 3.10.17).Google Scholar

²⁰ Cf. F. G. Kenyon, Books and Readers in Ancient Greece and Rome2 (Oxford, 1951), pp. 70–1.Google Scholar

²¹ Sommer, 392; Kenney, 19–20.

²² Van Groningen, ‘EKAOHIZ’ (above, n. 18), 16; E.G.Turner, Greek Papyri: An Introduction (Princeton, 1968, rev. ed., Oxford, 1980), pp. 112–13.

²³ Even then it is not clear how genuinely public the Roman public libraries were. I suspect that, while they may not have actively excluded the lower class, they probably did not need to. See A. J. Marshall, ‘Library Resources and Creative Writing at Rome", Phoenix 30 (1976), 261.

²⁴ See Suet. Tib. 70.2. Cf. Joseph. Vila 361; Contra Ap. 1.50–1.Google Scholar

²⁵ See P. White, ’Amicitia and the Profession of Poetry in Early Imperial Rome’, JRS 68 (1978), 86.Google Scholar

²⁶ Alt. 2.1 (SB21).2; cf. 12.40 (SB 281). 1.

²⁷ 7.97.11; cf. Stat. Silv. 2. ep.

²⁸ 9.84. Men on military service obviously faced particularly difficult problems acquiring books (cf. Mart. 10.78.9–13). Maecilius Nepos took Pliny′s libelli to the provinces when he served as a prefect (Ep. 4.26.2, with Sherwin-White′s n. ad he).

²⁹ Auson. 1.3 Prete; Ausonius replies in 1.4 Prete.

³⁰ These poems have aroused substantial scholarly debate. Was the Livy, for instance, a complete text or an epitome? For discussions, see Kenyon, Books (above n. 20), pp. 94–5; R. P. Oliver, ‘The First Medicean MS of Tacitus and the Titulature of Ancient Books’, TAPA 82 (1951), 248 n. 52 (contra T. Birt, Die Buchrolle in der Kunst [Leipzig, 1907]); and Kleberg,Google Scholar

³¹ Sommer, 398, mentions many examples in Cicero′s Letters.

³² Alt. 8.11 (SB 161).7, where, ad he, Shackleton Bailey briefly outlines the exchange.

³³ An. 8.12 (SB 162).6.

³⁴ An. 9.9 (SB 176).2.

³⁵ Alt. 13.31 (SB 302).2 and 13.33 (SB 309).2.

³⁶ An. 16.11 (SB 420).4.

³⁷ Cic. An. 13.8 (SB 313), in which Cicero also asks Atticus to send Brutus’ epitome of the history of Caelius

³⁸ Alt. 2.3 (SB 23).4.

³⁹ An. 1.20 (SB 20).6.

⁴⁰ Alt. 2.20 (SB 40).6 (with Shackleton Bailey′s n. ad he. ) and 2.22 (SB 42).7.

⁴¹ Alt. 2.4 (SB 24). 1, with Shackleton Bailey′s n. ad he; discussed by Sommer, 399.

⁴² Ep. 9.9.6–8; cf. 9.7.1

⁴³ Cic. Alt. 13.21a (SB 327); discussed by Sommer, 410–11. See also 13.22 (SB 329).3; 13.23 (SB331).2.

⁴⁴ See Cic. An. 13.21a (SB 327). 1.

⁴⁵ An. 3.12 (SB 57).2.

⁴⁶ l.pr.7. He explains that, ‘Namque alterum sermonem per biduum habitum pueri quibus id praestabatur exceperant, alterum pluribus sane diebus, quantum notando consequi potuerant, interceptum boni iuuenes sed nimium amantes mei temerario editionis honore uulgauerant’. Cf. 3.6.68 and 7.2.24.

⁴⁷ Alt. 13.21 (SB 351).3; in fact, inhibere was an earlier alteration.

⁴⁸ Cic. An. 13.44 (SB 336).3. Cf. Sid. Apoll. Ep. 2.8.2

⁴⁹ R. M. Ogilvie, The Library of Lactantius (Oxford,1978), p.60.

⁵⁰ Cic. An. 12.6a (SB 243). 1.

⁵¹ Fairweather, J.,Seneca the Elder (Cambridge,1981), p.15.CrossRefGoogle Scholar

⁵² G. Williams, ‘Phases in Political Patronage of Literature in Rome’, in B. K. Gold, ed., Literary and Artistic Patronage in Ancient Rome (Austin, 1982), p. 20.

⁵³ Sen. Contr. 10. praef. 5–7

⁵⁴ Of all the topics connected with the circulation of literary texts in Rome, the booktrade is the most prone to distortion because of the influence of supposed modern parallels. If we do not abandon the modern model, we shall be trapped into anachronistically postulating the existence of a complex system for which we have little or no evidence (e.g. Kleberg′s discussion of pre-Ciceronian authors, p. 23). As long ago as 1926 Sommer demolished the notion of Cicero′s

⁵⁵ Sosii: Hor. Epist. 1.20.2, ArsP. 345; Dorus: Sen. Ben. 7.6; Pollio: Mart. 1.113.5; Secundus: ‘ Mart. 1.2.7–8; Trypho: Mart. 13.3.4. The publication of Martial′s poems has been discussed by many scholars, e.g., E. T. Sage, ‘The Publication of Martial′s Poems’, TAPA 50 (1919), 168–76; W. Allen, Jr et al, ‘Martial. Knight, Publisher, and Poet’, CJ65 (1970), 345–57; and Kleberg, p. 39.

⁵⁶ Zetzel, Latin Textual Criticism (above, n. 18), p. 235.Google Scholar

⁵⁷ When Kleberg contends that booksellers doubled as ‘publishers’ responsible for distribution, first in the form of contacts with booksellers (p. 37), he imports seriously anachronistic conceptions into the Roman marketplace.Google Scholar

⁵⁸ Philostr. VS 2.21 (603), on Proclus of Naucratis, provides an extremely rare potential counter-example. See also Hor. Epist. 1.20.9–13, which may involve the sending of out-of-fashion books to the provinces, although the ancient commentators took it differently (see E. Galletier, j LEpitre dHorace a son livre (1.20) et les commentaires antiques', Revista Clasica 8 [1936], ) 30–1Google Scholar

⁵⁹ Pliny, Ep. 4.7.2, with Sherwin-White′s n. ad he. and Sommer, 414 n. 3. Kleberg, p. 62, thinks 1000 copies may not have been unusual. 60 Kleberg (p. 63) overlooks this when he anachronistically

⁶⁰ Kleberg (p. 63) overlooks this when he anachronistically discusses the risk of too large an t ‘edition’ to booksellers and those he terms ‘publishers’. Cf. E. G. Turner, / libri neltAtene del I V e IV secolo a.C, trans. M. and L. Manfredi, in G. Cavallo, ed., Libri, editori e pubblico nel mondo antico (Rome, 1975), p. 21 (rev. version of Athenian Books in the Fifth and Fourth Centuries B.C. [London, 1952; inaug. Lecture, University College London, 1951]). Two secondary sources of texts can be mentioned here. First, school texts, whether copied by the students at the master′s dictation or not, no doubt drastically increased the number of a very few works. Second, the sale of used books, if it occurred at all, may have affected the circulation of literary texts to a limited degree. It is easy to overestimate the importance of used book sales, J however, since the evidence is extremely slight, as should be expected. From what sources would a used book dealer obtain his wares? The‘ fasces librorum’ Gellius buys at Brundisium (9.4.1)

⁶¹ Skeat, ‘Dictation’ (above, n. 54), 179–91, provides a very useful and judicious history of the questionGoogle Scholar

⁶² See Sherwin-White, ad Pliny, Ep. 4.1.2.

⁶³ Kleberg, p. 31.

⁶⁴ Skeat, ‘Dictation’ (above, n. 54).

⁶⁵ 1.117.17. Much modern discussion has focused on Martial′s epigrams for presents of books, 14.183–96 (see above, n. 30). The enormous variation in the gifts for which Martial gives epigrams in the rest of Book 14, however, precludes any definitive discussion of book prices on the basis of those poems

⁶⁶ Cf. Kenyon, Books (above, n. 20), p.

⁶⁷ J. J. Phillips, ‘Book Prices and Roman Literacy’, CW19 (1985), 36–8, overlooks this. See T.P.Wiseman, ‘Pete nobiles amicos: Poets and Patrons in Late Republican Rome’, in B. K. Gold, ed., Literary and Artistic Patronage in Ancient Rome (Austin, 1982), 39; Marshall, ‘Library Resources’ (above, n. 23), 254–5.

⁶⁸ See Mart. 4.72, quoted above.

⁶⁹ See Gell. 5.4.1; 13.31.1; 18.4.1. Cf. Athenaeus 1.1. d–e. Catullus writes of looking for a friend in omnibus libellis (55.4). Quinn, ad be, argues that although ‘C. no doubt means places where books were displayed for sale’, we do not ‘need to assume (as some do) that libelli = "bookshops’". The assumption, however, is a natural one.

⁷⁰ See Mart. 1.117.11.

⁷¹ Cf. P. White,‘Positions for Poets in Early Imperial Rome’, in B. K. Gold, ed., Literary and Artistic Patronage in Ancient Rome(Austin, 1982), 62. R. P. Sailer, Personal Patronage under the Early Empire(Cambridge, 1982), pp. 137–8, argues that the importance of literary talent and education in social mobility resided primarily in acquiring contacts, which would then, on the standard basis of personal relationships, work the advancement.

⁷² See White, ‘Positions for Poets’ (above, n. 71), 52–3.

⁷³ Kenney, 21.

⁷⁴ Mart. 4.72, quoted above

⁷⁵ Kenney, 21

⁷⁶ Kenney, 22; cf. Sherwin-White, ad Pliny, Ep. A.I.2. It is very difficult to evaluate this issue because of the nature of our sources and their distribution in time. Martial provides many of the extant references to bookshops, but at approximately the same time Pliny is surprised at the very existence of bookshops in the Three Gauls (Ep. 9.11.2) and does not mention visiting any Roman bookshops himself. Bookshops may occur relatively frequently in Martial′s epigrams simply because his poetry is a collection of ephemeral and topical pieces, which would be likely to mention various kinds of shops.Google Scholar

⁷⁷ Ep. 1.2.5–6. Sherwin-White, ad 1.2.5, argues that the distribution of Pliny′s books was entirely in the hands of booksellers. Pliny distributed copies on his own as well, however.Google Scholar

⁷⁸ On non-literary works, see S. West, ′Chalcenteric Negligence", CQ 20 (1970), 290.Google Scholar

⁷⁹ Cf. Quinn, 152, although I question whether ‘The problem which dogs Roman literature throughout its history is the lack of an audience large and representative enough to make the writer feel he is fulfilling a valid social function’.

⁸⁰ My thanks go to Katherine A. Geffcken, Mary R. Lefkowitz, Miranda C. Marvin, James N. Rash, and the editors and referees of CQ for helpful suggestions and to Wellesley College and the American Council of Learned Societies for supporting my research.Google Scholar