¹ For general surveys of the field (in the analytical tradition), see Miller, A., Philosophy of Language (Oxford, 2007 ²)Google Scholar and Lycan, W.G., Philosophy of Language: A Contemporary Introduction (New York, 2008 ²)Google Scholar; some of the main strands are covered in Soames, S., Philosophical Analysis in the Twentieth Century (Princeton, 2003)Google Scholar.

² On this epigram, see in particular Jocelyn, H.D., ‘C. Licinius Macer Calvus, fr. 18 Büchner’, Eikasmos 7 (1996), 243–54Google Scholar; cf. Courtney, E., The Fragmentary Latin Poets (Oxford, 1993)Google Scholar, 210; Hollis, A.S., Fragments of Roman Poetry: c.60 bc–ad 20 (Oxford, 2007), 83–4Google Scholar. Clodius' attack on Pompey is recorded by Plut. Pomp. 48.7: τίς ἐστιν αὐτοκράτωρ ἀκόλαστος; τίς ἀνὴρ ἄνδρα ζητεῖ; τίς ἑνὶ δακτύλῳ κνᾶται τὴν κεϕαλήν; (‘Who is a licentious imperator? What man seeks for a man? Who scratches his head with one finger?’): Clodius' followers shouted out in answer: ‘Pompey’; cf. Cass. Dio 39.19.1–2; Amm. Marc. 17.11.4. Sen. Controv. 7.4.7 provides an incomplete version of the epigram.

³ Cf. Williams, C.A., Roman Homosexuality: Ideologies of Masculinity in Classical Antiquity (Oxford, 1999)Google Scholar, 216: ‘He scratches his head with one finger. What would you think he wants? A man [uirum]’. Translations elsewhere in this paper (aside from the Servian examples, which are my own) generally follow the relevant Loeb volume, with small amendments for clarity.

⁴ Jocelyn (n. 2) is also troubled, expressing doubt as to whether we possess the entire epigram; his article contains an excellent discussion of the punctuation, scholarly tradition, and cultural context of these lines. Cf. Hollis (n. 2), 83.

⁵ OLD s.v. uolo 2: ‘To desire to have, want (a particular thing or person)…w. dat. of advantage’.

⁶ Cf. e.g. Mart. 6.54.3.

⁷ Cf. Lausberg, M., Das Einzeldistichon: Studien zum antiken Epigramm (Munich, 1982), 393Google Scholar: ‘Magnus, den alle fürchten, kratzt sich mit einem Finger den Kopf. Was will er bloss damit? Einen Mann’. The pun was also used, albeit in a different way, by Martial: Vult, non uult dare Galla mihi, nec dicere possum, quod uult et non uult, quid sibi Galla uelit (‘Galla wants and doesn't want to oblige me, nor can I say, since she wants and doesn't want, what Galla means/wants [for herself]’, Mart. 3.90). Here, part of the play is on the different constructions governing the datives mihi and sibi. Compare the note to this epigram in Shackleton Bailey's Loeb volume: ‘Play on the normal meaning of quid sibi uelit, ‘what she means’, and the literal, ‘what she wants for herself’'. My thanks to Brigitte Libby for directing me to this.

⁸ Regarding ‘signify’: I intend to consider terms such as σημαίνω, significo, ‘signifier’ and ‘signify’ in a subsequent paper.

⁹ Cf. OLD, s.v. uolo 17: ‘To want to be understood (as its meaning), imply, signify, mean’ (w. sibi).

¹⁰ Cf. quid sibilex aut quid uerba ista uellent (‘… what the law or the words mean’, Cic. Leg. 3.13.33). In passing, it is worth noting that uelle can also be applied to inanimate objects without the sibi in order to refer to their significance: quid uelintflores et acerra turis plena miraris…? (‘Do you wonder as to what the flowers mean, and the casket full of incense?’, Hor. Carm. 3.8.2–3); quid uelitet possit rerum concordia discors (‘What is the meaning and what the effects of Nature's jarring harmony’, Hor. Epist. 1.12.19); see OLD s.v. uolo 17.

¹¹ Cf. quid ergo illa sibi uultpars altera orationis …? (‘What, then, does that second part of his speech mean …?’, Livy 40.12.14).

¹² See Barsby, J., Eunuchus (Cambridge, 1999), 89CrossRefGoogle Scholar.

¹³ In fact, Terence's audience would not have been able to determine whether eunuchus here refers to the title of the play or to a character within it: another possible interpretation is therefore ‘pay attention and listen carefully in silence, so that you may understand what the eunuch wants [for himself]’ (compare sentence 2). Here, the joke is that the audience will learn in the course of the play that ‘the eunuch’ (the disguised Chaerea), wants to have sex with Pamphila, which is not what one would expect of a eunuch.

¹⁴ For different lists from the following one, see Searle, J., Mind, Language, and Society: Philosophy in the Real World (New York, 1998), 139Google Scholar; Nozick, R., Philosophical Explanations (Cambridge, MA, 1981), 574–5Google Scholar. On the distinction between the utterer's meaning and sentence meaning, see Grice, H.P., ‘Meaning’, PhR 66 (1957), 377–88Google Scholar; Grice, P., Studies in the Way of Words (Cambridge, MA, 1989)Google Scholar; Sperber, D. and Wilson, D., Relevance: Communication and Cognition (Cambridge, MA, 1995 ²), 1–64Google Scholar.

¹⁵ OED s.v. ‘mean (v)’ I.1.b: ‘With infinitive as object: to intend or be determined to do something’.

¹⁶ OED s.v. ‘mean (v)’ II.9: ‘trans. With modifying word or phrase: to be important to a person to the extent indictated, esp. as a source of benefit or as an object of regard, affection, or love; to matter (a lot, nothing, etc.)’.

¹⁷ OED s.v. ‘mean (v)’ II.6.a: ‘trans. To intend to indicate (a certain object) or to convey (a certain sense when using some word, sentence, significant action)’.

¹⁸ OED s.v. ‘mean (v)’ II.7: ‘Of a thing, word, or statement: to have as signification; to signify, import; to portend. Also with clause as object (often an indirect question introduced by what)’.

¹⁹ DWB, s.v. meinen 2. ‘vor alters her auch von worten, einen angegebenen sinn haben, etwas bedeuten’; cf. Snell, B., ‘Die Sprache Heraklits’, Hermes 61 (1926), 353–81, at 365Google Scholar.

²⁰ Anglo-Saxon, the Germanic ancestor of English, also exhibits a similar kind of ambiguity in the verb mǽnan. Cf. Bosworth's Anglo-Saxon Dictionary, s.v. mǽnan I: ‘of persons (a) to intend to convey a certain sense, etc.’; II: ‘(of things) to signify, have a certain signification or purpose’.

²¹ Derrida, J., Spurs: Nietzsche's Styles. Éperons: Les Styles de Nietzsche (Chicago, 1979), 125Google Scholar; cf. ‘Je veux dire qu'il a raison’ (‘I mean that he's right’).

²² Derrida (n. 21), 128; cf. ‘Qu'est-ce que cela veut dire?’ (‘What does that mean?’).

²³ For example, Italian: ‘che cosa vuoi dire con questo?’; ‘che cosa vuol dire questa parola?’.

²⁴ Cf. Pl. Prt. 343d.

²⁵ Cf. τί δὴ οὖν ἡμῖν βούλεται οὗτος ὁ μῦθος, ὦ Θεαίτητε, πρὸς τὰ πρότερα; ἆρα ἐννοεῖς; (‘What does this tale mean to us, Theaetetus, with reference to what was said before? Do you see?’, Pl. Tht. 156c).

²⁶ LSJ s.v. λέγω III.9: ‘wish to say, mean’; cf. οὔτοι γυναῖκας ἀλλὰ Γοργόνας λέγω (‘No – I do not mean women, but Gorgons’, Aesch. Eum. 48); πῶς λέγεις; (‘how do you mean?’ Pl. passim; see e.g. Ap. 24e; Cra. 429c); … τί λέγοι ὁ ποιητής (‘… the poet's meaning …’, Pl. Prt. 339e; cf. 341c–e).

²⁷ Cf. Pl. Prt. 312c; τί τοῦτο λέγει, πρὸ Πύλοιο; (‘… what does πρὸ Πύλοιο mean?’, Ar. Eq. 1059); the translation of the phrase from Aristophanes is that of LSJ s.v. λέγω III.9; cf. Sommerstein, A.H., Knights (Warminster, 1981), 109Google Scholar: ‘What does that mean, “before Pylos”?’. See, however, Henderson, J., Aristophanes: Acharnians, Knights (Cambridge, MA, 1998), 361Google Scholar: ‘What does he mean, “before Pylos”?’. In English we do something similar when using phrases such as ‘you have to think about what the text is saying’.

²⁸ Cf. Hdt. 7.162. See also LSJ s.v. ἐθέλω ΙΙ.3. ϕέρ᾽, ἐξελίξας περιβολὰς σϕραγισμάτων ἴδω τί λέξαι δέλτος ἥδε μοι θέλει (‘come, let me open its sealed wrappings and see what the tablet means/wishes to tell me’, Eur. Hipp. 864–5). Cf. vouloir dire.

²⁹ LSJ s.v. νοέω IV ‘of words, bear a certain sense, mean’; cf. (in conjunction with βούλομαι used of a text) ἡ μὲν νέα ϕωνὴ ἡμῖν αὕτη καὶ τοὐναντίον περιέτρεψε μηνύειν τὸ δέον καὶ τὸ ζημιῶδες, ἀϕανίζουσα ὅ τι νοεῖ, ἡ δὲ παλαιὰ ἀμϕότερον δηλοῖ ὃ βούλεται τοὔνομα (‘… this fine modern language of ours has turned δέον and ζημιῶδες around, so that each has the opposite of its original meaning, whereas the ancient language shows clearly the real meaning of both words’, Pl. Cra. 418b). For a relevant use of νοέω involving a human being as the verb's subject, see Pl. Prt. 347e.

³⁰ There is some doubt concerning the form; the OCT prints νοεῖ.

³¹ LSJ s.v. ἐννοέω VI ‘of words, mean, signify’.

³² My investigation here is limited to Servius' usage. The phrase is attested e.g. at Serv. Ecl. 2.25, 2.66; G. 1.370, 1.512; Aen. 5.380, 8.154, 9.25, 11.411, 11.502, 12.350. My thanks to Tom Keeline for a useful discussion of this issue.

³³ The text is that of Thilo-Hagen, and my examples are drawn from the commentary of Servius (not Servius Danielis). On Servius and his relationship to Donatus and Servius Danielis, see Zetzel, J., Latin Textual Criticism in Antiquity (Salem, NH, 1984), 81–147Google Scholar; Kaster, R., Guardians of Language: The Grammarian and Society in Late Antiquity (Berkeley, 1988), 169–97 and 356–9Google Scholar.

³⁴ Cf. OLD s.v. hoc 10c; for just one example, see Serv. Aen. 1.79.

³⁵ It would be strange indeed if the proper translation of these words were ‘this (hoc) means these things (haec)’. Haec is only made explicit here because the subject has changed.

³⁶ Cf. OLD s.v. uolo 4b. For instances of uult dicere with an author marked as the subject, see Serv. Ecl. 2.25 (Cicero … uoluit dicere); cf. also Aen. 5.380, 11.411. On uolo dicere, see already Plaut. Amph. 383–4: Mer: Amphitruonis te esse aiebas Sosiam. Sos. Peccaueram. nam Amphitruonis socium memet esse uolui dicere (Mercury: ‘you were saying you were Amphitryon's Sosia’; Sosia: ‘All a mistake, sir; ‘Amphitryon's associate’, I meant sir, really I did'); Ter. Eun. 504: quid? quid aliud uolui dicere? (“Well, then, what else did I mean to say?); Sen. Ep. 19.9: uolo tibi hoc loco referre dictum Maecenatis uera in ipso culmine elocuti: ‘ipsa enim altitudo attonit summa’. Si quaeris, in quo libro dixerit; in eo, qui Prometheus inscribitur. Hoc uoluit dicere, attonita habet summa (‘At this point I should like to quote a saying of Maecenas, who spoke the truth when he stood on the very summit of affairs: “There's thunder even on the loftiest peaks”. If you ask me in what book these words are found, they occur in the volume entitled Prometheus. He simply meant to say that these lofty peaks have their tops surrounded with thunderstorms’).

³⁷ OLD s.v. hic 12.

³⁸ See also: tamen Catoin originibus hoc dicit, cuius auctoritatem Sallustius sequitur in bello Catilinae … (‘But Cato says this in his Origines, whose authority Sallust follows in the Bellum Catilinae – that …’, Serv. Aen. 1.6); cf. Serv. G. 1.378; Aen. 7.1.

³⁹ On the frequent omission of the names of authors in Greek scholia, see Dickey, E., Ancient Greek Scholarship (Oxford, 2007), 121–2Google Scholar: ‘The particular poet or other author who is the subject of commentary need not be designated by any noun at all, since he is assumed to be the subject of any appropriate verb for which no other subject is expressed.’

⁴⁰ Cf. Nec posse Italiadetraxit more suo praepositionem prouinciae; non enim dixit ‘de Italia’, sed ‘Italia’ (‘he has removed, according to his custom, the preposition from the province; for he has not said de Italia but rather Italia’, Serv. Aen. 1.38).

⁴¹ My general approach in what follows is partly informed by Reddy, M.J., ‘The conduit metaphor – a case of frame conflict in our language about language’, in Ortony, A. (ed.), Metaphor and Thought (Cambridge, 1979), 284–324Google Scholar; Lakoff, G. and Johnson, M., Metaphors We Live By (Chicago, 1980)Google Scholar; Lakoff, G. and Turner, M., More than Cool Reason; A Field Guide to Poetic Metaphor (Chicago, 1989)CrossRefGoogle Scholar.

⁴² Cf. Pl. Phdr. 276a; Tht. 160e; Symp. 209c; Resp. 330c, 601b. For the importance of the excerpt, see e.g. Szlezák, T., Platon und die Schriftlichkeit der Philosophie (Berlin, 1985)CrossRefGoogle Scholar.

⁴³ Fraenkel, E., Horace (Oxford, 1957), 356–63Google Scholar, and Pearcy, L., ‘The personification of the text and Augustan poetics in Epistles 1.20’, CW 87 (1994), 457–64Google Scholar.

⁴⁴ For a similar use of gender, see Ov. Tr. 2.371–2 (and compare 375–6): Ilias ipsa quid est, nisi turpis adultera de qua inter amatorem pugna uirumque fuit? (‘What is the Iliad its/herself, except a disgraceful adulteress, over whom there was a fight between her lover and husband?’), together with the note of Ingleheart, J., A Commentary on Ovid, Tristia, Book 2 (Oxford, 2010), 301Google Scholar.

⁴⁵ On this, see Wissig-Baving, G., Die Anrede an das Buch in der römischen Dichtung (Frankfurt am Main, 1989)Google Scholar. For instance: cuius uis fieri, libelle, munus? … Faustini fugis in sinum? sapisti (‘to whom do you want to belong, my dear little book? … Are you rushing into the lap of Faustinus? You are wise’, Mart. 3.2.1–6). Some further examples follow: Catull. 35.1–3, 36.1–2, 42.1–7; Ov. Am. 1 (epigram); Tr. 1.1, 3.1, 3.7.1–2; Pont. 4.5; Mart. 1.70, 3.4, 10.104, 12.2; Stat. Silv. 4.4; Anth. Pal. 12.208 (Strato); Auson. Ep. 12.

⁴⁶ Barthes, R., Le plaisir du texte (Paris, 1973), 13. Barthes' italicsGoogle Scholar.

⁴⁷ Ibid. 30.

⁴⁸ Ibid. 84.

⁴⁹ Cf. Wimsatt, W.K. and Beardsley, M., ‘The intentional fallacy’, Sewanee Review 54 (1946), 468–88Google Scholar. For other influential arguments along these lines see e.g. Barthes, R., ‘The death of the author’, in id., Image–Music–Text (London, 1977), 142–8Google Scholar: Foucault, M., ‘What is an author?’, in Bouchard, D.F. (ed.), Language, Counter-memory, Practice (Ithaca, NY, 1977), 113–38Google Scholar.

⁵⁰ One might compare the German phrase ‘der Text will sagen …’, where the metaphor is also clear; cf. Gadamer, H.-G., Hermeneutische Entwürfe: Vorträge und Aufsätze (Tübingen, 2000)Google Scholar, 200: ‘Diese literarisch noch nicht bekannte Formelierung Heideggers will sagen, daß die Wahrheit (ueritas) durch die Gewißheit (certitudo) verdrängt wird.’

⁵¹ Conte, G.B., ‘«La retorica dell'imitazione» come retorica della cultura: qualche ripensamento’, Filologia Antica e Moderna 2 (1992), 41–52Google Scholar, at 44; the English translation is that of id., tr. Most, G., Genres and Readers (Baltimore, 1994), 133Google Scholar. I only adduce this phrase as an example because I highly respect Conte's work; it is moreover important to look at the theoretical context. One might also compare Gadamer (n. 50), 203: ‘Der Text hat seine einheitliche Intention, auch wenn diese nicht notwendig eine bewußte Intention des Schreibenden sein muß. Jedenfalls ist der Empfänger, der Entzifferer, auf das gerichtet, was der Text meint’.

⁵² Cf. Conte, G.B., The Rhetoric of Imitation (Ithaca, NY and London, 1986), 127Google Scholar: ‘If we refuse to separate the text from its intentions (which means not ingenuously guessing at the author's intentions but uncovering the living relationship that linked the text with the world and with its immediate public) …’. As will become clear in what follows, I do not want to argue that the author's intention is the only thing we should be concerned about when it comes to literary criticism, but simply want to point out how discussion of it can be pre-empted by a metaphorical trick of our language.

⁵³ Of course there are texts about whose historical context we are too poorly informed to make positive statements about the author, and this issue is particularly acute in early Greek and Biblical studies. The difficulty of speaking about what an author intends, or even of what the author is, in such situations can leave clear marks on the critic's prose. Cf. Most, G., ‘Hesiod's myth of the five (or three or four) races’, PCPhS 43 (1997), 104–27, at 120Google Scholar: ‘For evidently Hesiod's text is not trying, with only partial success, to represent that myth, which seems to have been foisted upon him without adequate regard for his own specific intention, but is instead conveying, with remarkable success, a different meaning, one which is fully appropriate to the programme of the Works and Days as a whole. In short, Hesiod's text is not struggling in vain to present a coherent myth of the decline of man; it is succeeding in saying something quite different about the conditions and chances for human success in the world in which we live’. Here it is the text itself that tries, struggles and succeeds in saying, that conveys a meaning: the verbs appropriate to the author are applied to his text, and there is some vagueness about what the pronouns ‘him’ and ‘his’ are referring to. My reasons for singling out Glenn Most are the same as in the case of Conte.

⁵⁴ Searle, J., The Rediscovery of the Mind (Cambridge, MA, 1992), 175–86Google Scholar, on the ‘Background’. He also gives the example of the syntactically similar phrases ‘I have had twins’ and ‘I have had breakfast’; nobody would entertain the idea that the second of these could be equivalent to the phrase ‘I have just given birth to a western omelette’.

⁵⁵ Cf. ταῦτά μοι δοκεῖ … Σιμονίδης διανοοῦμενος πεποιηκέναι τοῦτο τὸ ᾄσμα (‘such is my view … of Simonides’ intention in composing this ode', Pl. Prt. 347a).

⁵⁶ Augustine, in his writing on biblical hermeneutics, also sounds extraordinarily modern in his distinction between different types of meaning; see De doctrina Christiana 2.1–3.

⁵⁷ See OLD s.v. sentio 9: ‘(of a writer) To mean, intend, have in mind’.

⁵⁸ On what follows, see Skinner, Q., ‘Motives, intentions and the interpretation of texts’, New Literary History 3 (1972), 393–408CrossRefGoogle Scholar; Tully, J. (ed.), Meaning and Context: Quentin Skinner and his Critics (Princeton, NJ, 1988)Google Scholar; Bach, K., ‘Meaning, speech acts, and communication’, in Harnish, R.M. (ed.), Basic Topics in the Philosophy of Language (New York, 1994), 3–20Google Scholar; Searle, J., ‘Literary theory and its discontents’, New Literary History 25 (1994), 637–67CrossRefGoogle Scholar. For an application of related ideas in classical studies, see Arweiler, A., ‘What is a literary speech act? Quintilian, John Searle, and the notion of “constatives”’, in Fuhrer, T. and Nelis, D. (edd.), Acting with Words: Communication, Rhetorical Performance and Performative Acts in Latin Literature (Heidelberg, 2010), 199–242Google Scholar.

⁵⁹ August. Conf. 4.3.5: si enim de paginis poetae cuiuspiam, longe aliud canentis atque intendentis, cum forte quis consulit, mirabiliter consonus negotio saepe uersus exiret … (‘For if, when a man had by pure chance consulted the books of some poet, who sang of and intended another matter entirely, verses that were miraculously pertinent to the current situation came out …’).

⁶⁰ Hirsch, E.D., Validity in Interpretation (New Haven, 1967), 8 (Hirsch's italics)Google Scholar.

⁶¹ Cf. e.g. Eagleton, T., Literary Theory: An Introduction (Oxford, 1983), 68–9Google Scholar: ‘The aim of all this policing is the protection of private property. For Hirsch an author's meaning is his own, and should not be stolen or trespassed upon by the reader. The meaning of the text is not to be socialized, made the public property of its various readers; it belongs solely to the author, who should have the exclusive rights over its disposal long after he or she is dead … Hirsch's defence of authorial meaning resembles those defences of landed titles which begin by tracing their process of legal inheritance through the centuries, and end up by admitting that if you push that process back far enough the titles were gained by fighting someone else for them.’

⁶² For a discussion of this phrase, see Zanker, A. Th. and Thorarinsson, G., ‘The meanings of «meaning» and reception studies’, MD 67 (2011), 9–19Google Scholar.

⁶³ My emphasis, but Martindale's italics. Martindale, C., Redeeming the Text: Latin Poetry and the Hermeneutics of Reception (Cambridge, 1993), 3–4Google Scholar. For other appearances of the expression, see Fowler, D., ‘On the shoulders of giants: intertextuality and classical studies’, MD 39 (1997), 13–34Google Scholar, at 24: ‘Meaning is realized at the point of reception’; Hinds, S., Allusion and Intertext: Dynamics of Appropriation in Roman Poetry (Cambridge, 1998)Google Scholar, 48: ‘in practice, meaning is always constructed at the point of reception’; Batstone, W.W., ‘Provocation: the point of reception theory’, in Martindale, C., Thomas, R.F. (edd.), Classics and the Uses of Reception (Malden, MA, 2006), 14–20, at 14CrossRefGoogle Scholar: ‘All meaning is constituted or actualized at the point of reception’; Kallendorf, C., The Other Virgil; ‘Pessimistic’ Readings of the Aeneid in Early Modern Culture (Oxford, 2007), 223CrossRefGoogle Scholar: ‘meaning is always constructed at the point of reception’; Houghton, L., Wyke, M. (edd.), Perceptions of Horace: A Roman Poet and His Readers (Cambridge, 2009), 5Google Scholar: ‘meaning is ultimately realized at the point of reception’.

⁶⁴ Cf. Zanker and Thorarinsson (n. 62).

⁶⁵ See e.g. Kripke, S., Wittgenstein on Rules and Private Language; an Elementary Exposition (Cambridge, MA, 1982)Google Scholar.

⁶⁶ Demetrius of Phalerum (Eloc. 287) and Quintilian describe how emphasis, or ‘implication’, featured in ancient rhetoric. It was common practice to make use of ‘figured speech’ in order to present different possible (aperte) meanings to a readership: eius [i.e. figured speech] usus triplex est: unus si dicere palam parum tutum est, alter si non decet, tertius qui uenustatis modo gratia adhibetur et ipsa nouitate ac uarietate magis quam si relatio sit recta delectat (‘Its use is threefold: first, if it is unsafe to speak forthrightly; second, if it is unseemly to do so; and third, when it is adopted for pleasurable effect, and delights because of its novelty and variations more than direct expression would’, Quint. Inst. 9.2.66). See Ahl, F., ‘The art of safe criticism in Greece and Rome’, AJPh 105 (1984), 174–208Google Scholar.

⁶⁷ OLD s.v. praeceps ².1: ‘A place where there is a precipitous descent, a sheer drop’. Housman, A.E., CR 17 (1903), 465–8Google Scholar, on the other hand, glossed the phrase as ‘every vice has now come to a dead halt at the cliff's edge; has reached, we might say, the end of its tether; has gone as far as nature suffers it to go’. Kidd, D.A., ‘Juvenal 1, 149 and 10, 106–107’, CQ 14 (1964), 103–8CrossRefGoogle Scholar, at 107, suggested ‘every vice is now in a precarious position … The sequence of thought in lines 147–9 is therefore: “society is now as vicious as ever it can be – but this means that vice is particularly vulnerable – therefore now is the time to attack it”’; for references, see Kidd's note.

⁶⁸ Cf. Sall. Hist. fr. 1.16 Reynolds: maiorum mores non paulatim ut antea sed torrentis modo praecipitati (‘the customs of the ancestors have been thrown headlong not slowly, as before, but in the manner of a torrent’); Livy, pr. 9: labente deinde paulatim disciplina uelut desidentes primo mores sequatur animo, deinde ut magis magisque lapsi sint, tum ire coeperint praecipites (‘Then let him note how, with the gradual relaxation of discipline, morals first gave way, as it were, then sank lower and lower, and finally began the downward plunge’); Vell. Pat. 2.1; 2.10: adeo mature a rectis in praua a prauis [in uitia, a uitiis] in praecipitia peruenitur (‘Thus, one goes swiftly from the upright to the perverted, from the perverted [into vice, from vice] into headlong collapse’).

⁶⁹ See Marchetti, S. Citroni, ‘Nota a Giovenale 1, 149’ MD 9 (1982), 175–85, at 178Google Scholar: ‘se Giovenale innova, sarà … nel senso di un ancor più drastico pessimismo nei confronti dei costumi viziosi della propria età, a cui egli rifiuta un ulteriore spazio in cui precipitare, una ulteriore possibilità di peggioramento’.

⁷⁰ On what follows, see Gibson, B., ‘Ovid on reading: reading Ovid. Reception in Ovid Tristia II’, JRS 89 (1999), 19–37Google Scholar.

⁷¹ On the significant change made in Ovid's self-quotation here, see Ingleheart (n. 44), 232.