1 From the words ‘sicuti dixi’ it seems quite certain that the explanation Gellius offers for his title in the lost section at the beginning of the preface was not substantially different from the one quoted here from praef. 4 and repeated in praef. 10.
2 I shall not deal here with the apparent ‘historical’ difficulty caused by the fact that some of the ‘notes’ included in Gellius' book are explicitly said to have been taken in periods prior to his stay in Athens (esp. 20.6.1, with 15). This has been answered by R. Marache's distinction between commentationes or commentarii, which denote composed short pieces, and annotationes, the rudimentary notes taken while compiling material to be included in the book; see his edition, vol. i (Paris, 1967), p. 2 n. 2, andHolford-Strevens L., Aulus Gellius (London, 1988), p. 24–5.
3 For similar criticism of titles by later grammarians, seeJordan H. (ed.), M. Catonis praeter librum de Re Rustica quae exstant (Leipzig, 1860), p. xxi.
4 On the history of titles and their functions see:Daly L. W., ‘The Entitulature of Pre-Ciceronian Writings’, Classical Studies in Honor of W. A. Oldfather (Urbana, 1943), pp. 21–3;Schmalzriedt E., Περ⋯ ϕύσεως: zur Frühgeschichte der Buchtitel (Munich, 1970), pp. 20–50;Horsfall N., ‘Some Problems of Titulature in Roman Literary History’, BICS 28 (1981), 103–4; and beyond classical antiquity:Kellman S. G., ‘Dropping Names: The Poetics of Titles’, Criticism 17 (1975), 152–67;Levin H., ‘The Title as Literary Genre’, MLR 72 (1977), xxiii–xxxvi;Fowler A., Kinds of Literature: Introduction to the Theory of Genres and Modes (Cambridge, Mass., 1982), pp. 92–8. Titles which function as authorial commentary or guides to understanding the work (La vida es sueno, Ulysses) seem not to have been employed in the ancient world.
5 For identification of the works referred to in Gellius' list, seeFaider P., ‘Auli Gellii Noctium Atticarum Praefatio’, Musée Beige 31 (1927), 203–8 and the bibliography cited there, pp. 189–90. Ancient titles are conveniently collected and classified inZilliacus H., ‘Boktiteln i antik litteratur’, Eranos 36 (1938), 1–41.
6 The term Miscellanea seems to have been first used as a title by Politian in the fifteenth century; seePfeiffer R., History of Classical Scholarship: From 1300 to 1850 (Oxford, 1976), p. 45.
7 This is how Clement explains his own choice of title (Strom. 4.4.1; cf. 6.2.1), although he later adds the notions of rarity and value (4.6.2); seeMéhat A., Étude sur les ‘Stromates’ de Clément d'Alexandrie (Paris, 1966), pp. 96–104; for Apuleius' title, seeHelm's R. note in his edition of the Florida (Leipzig, 1959), pp. xviii–xix; other examples are brought byZilliacus , Eranos 36 (1938), 25–30.
8 Regenbogen O., RE XX, p. 1412, s.v. ‘Pinax’; but seeHenriksson K.-E., Griechische BOchertitel in der romischen Literatur (Helsinki, 1956), pp. 117–18.
9 On Ἐγχειίδιον/a in antiquity, see:Broccia G., Enchiridion: per la storia di una denominazione libraria (Rome, 1979), pp. 11–44.
10 SoHenriksson , op. cit., pp. 122–5 on the basis of Quint., Inst. 10.3.17; but seeColeman K. M. in her edition of Statius' Silvae iv (Oxford, 1988), pp. xxii–xxiv.
11 For Satura and similar titles seeCoffey M., Roman Satire (London, 1976), pp. 11–18.
12 Some other works of peripatetic provenance were similarly titled, such as the Μεγαρικός of Theophrastus (D.L. 5.44) and the χαλκιδικός of Demetr. Phaler. (ibid., 5.81), but these titles do not necessarily refer to the dramatic setting of a Dialogue, as is clearly the case of Dio's Εὐβοικός and Ἀττικ⋯ν Δεῖπνον mentioned by Athenaeus (4.134–7, if it is a title); seeHirzel R., Der Dialog: ein literarhistorischer Versuch (Leipzig, 1895), vol. i, pp. 311–12 n. 2; p. 319 n. 1. For similar titles in the ‘Greek Novel’ see:Miiller C. W., ‘Der griechische Roman’, in Vogt E. (ed.), Griechische Literatur (Wiesbaden, 1981), pp. 391–2.
13 At least as early as Democritus 68.B.182 D-K. On the litterarum radices amarae, see A. Otto, Die Sprichwörter und sprichwörtlichen Redensarten der Römer (Leipzig, 1890; repr. Hildesheim, 1965), p. 195, ‘litterae’, 1; and for ancient adaptations of Verg. G. 1.145 to suit a similar idea, seeMynors's R. A. B. commentary on the Georgics (Oxford, 1990), p. 30.
14 SeeMarache , op. cit., pp. xii–xiii. For a history of this metonymy and its uses in prefaces seeJanson T., Latin Prose Prefaces: Studies in Literary Conventions (Stockholm, 1964), pp. 97–8, 147–8; cf. Fro. M.Caes. 1.4.1 (pp. 5.22–6.2 v.d. Hout2), Apul. Apol. 5 for further uses in the second century.
15 On the erotic Milesiaka, seeWalsh P. G., The Roman Novel (Cambridge, 1970), pp. 10–17. ‘Night Words’ is the title given by G. Steinerto his article on pornographic literature; see hisLanguage and Silence (London, 1958), pp. 89ff.; for similar uses of Nox in Latin, see OLD s.v. ‘Nox’ 3.c.
16 Perhaps to be identified with the class of ‘public moyen’ characterized byGuillemin A.-M., Le Public et la vie litteraire a Rome (Paris, 1937), pp. 18–22, 82–6. See further:Beall S. M., Civilis Eruditio: Style and Content in the ‘Attic Nights’ of Aulus Gellius (Diss., Univ. of California at Berkeley, 1988), 3–4, 35–6.
17 Clarke M. L., Higher Education in the Ancient World (London, 1971), pp. 6–7, 80. On Gellius' education, seeHolford-Strevens , op. cit., pp. 12–13, 61–71.
18 As suggested byMaselli G., Lingua e scuola in Gellio grammatico (Lecce, 1979), p. 53. Whereas Atheniensis is used exclusively for persons or for the city itself, Atticus is quite commonly employed for language, rhetorical styles and philosophy; cf. TLL II.5 s.v. ‘Atheniensis’, pp. 1029–30; s.v. ‘Atticus’ II.A, p. 1134.
19 Some such title are enumerated byHolford-Strevens , op. cit., p. 21 n. 7.