¹ p. 273 Nauck = fr. 368 Slater from Athenaeus 3. 85e. cf. Strabo 1. 1. 1 and 1. 3. 13 e.g. in zetemata.

² fr. 3 Wehrli² with explanation on p. 46.

³ I am unhappy with the definite article as a vocative, and suggest rather the exclamatory ἄ But the vocative article is possible (Schwyzer-Debrunner, Gr. Gr. 2. 63), and Douglas Gerber refers me to Gow on Theocritus 1. 151. He also indicates to me that Onians, R. B., Origins of European Thought (Cambridge, 1951), p. 19 translates χανυo***ς as ‘make porous, puff out’ – 1 can see no justification for this.Google Scholar

⁴ We have recently acquired new fragments of Callias, an ancient authority on Aeolic poetry, but they do not help to solve this problem, which vexed unduly Gudeman, R.E. s.v. Kallias no. 23 and especially Wilamowitz, Textg. der gr. Lyriker (Berlin, 1900), pp. 74–6, who recognized the riddle in Alcaeus, but by giving the last sentence to Dicaearchus instead of Aristophanes, was reduced to despair, in which he is followed by Wehrli on Dicaearchus fr. 99.Google Scholar

⁵ Sch. Eur. Medea 169, Sch. II. 10. 252.

⁶ An example of a syggramma containing an ⋯ναγραФή of λύσεις would be Aristonicus’ work On the Wanderings of Menelaos, who according to Strabo 1. 2. 31 Similarly Clearchus in the first book of his Erotika (fr. 24 W²) gave a whole list of possible explanations why lovers wore wreaths. Even the first book of Strabo is largely filled with a rambling discussion of a zetema based on Odyssey 1. 24. We are not surprised, therefore, to learn that μακρoλoγíα belongs rather to a syggramma than to a hypomnema (Galen cited by Rutherford, , Scholia Aristophanea iii (London, 1905), p. 35). Zetemata, therefore, could appear in almost any kind of grammatical work. Arrighetti, SCO 17 (1968), 79 n. 17 writes: ‘il problema sarà piuttosto di cercare quale rapporto intercorra fra quel tipo di opere dedicate unicamente alla raccolta e quindi alla caccia di λύσεις e gli ύπoμυήμ;ατα’; but I do not think a definite answer will ever be attainable. An incomplete list is given by Gudeman RE s.v. Scholien 628 of similar syggrammata.Google Scholar

⁷ Pfeiffer's certainty about these editions is not warranted by the evidence; see his History of Classical Scholarship (Oxford, 1968), p. 130.Google Scholar

⁸ Gudeman in his useful article in R.E. s.v. Lyseis 2514 writes: ‘where Aristarchus allowed the validity of Aporiai, unavoidable with the , he took up a position and placed a diple ’ But it is not justifiable to argue, as Gudeman and others have done, that since Aristarchus condemns the frivolity of the zetemata of others, he therefore disliked the concept itself. This condemnation takes the form e.g. of allegations that some grammarian has inserted a line to create a zetema (examples collected by Lehrs, , De Aristarchi Studiis Homericis³ (Leipzig, 1882), p. 205Google Scholar and Ludwich, , Die Homervulgata (Leipzig, 1898), p. 168 n. 3).Google Scholar But such accusations are at least sometimes nothing more than polemical guesswork, not based on careful research (Bröcker, Rh.M. 40 (1885), 419¹; Nickau, , Untersuchungen…Zenodotus (Berlin, 1977), p. 16 n. 36;Google Scholar above all the important conclusions of Ludwich, , Aristarchs Homerische Textkritik ii (Leipzig, 1885, p. 104). At sen. Il. 20. 269 Aristarchus accuses someone of inserting lines to create a problem, although the lines are already known as a zetema to Aristotle; and at sch. Il. 16.97 Aristarchus praises Zenodotus for suspecting a passage as being inserted by someone wishing to create a homosexual liaison between Achilles and Patroclus. We should be warned that ancient scholars justified their atheteses by simply accusing others of insertions. On the other hand, when Aristarchus on Il. 10. 252 puts his diple on a line because of - a problem as old as Aristotle -, we ought to feel the joy of the scholar in full cry after a zetema. That his questions were not as futile as e.g. those of Zoilus is no reason to say with Pfeiffer (op. cit. 263) that he had no interest in zetemata at all. Ludwich, Homervulgata, p. 168 alleged that Aristarchus distanced himself as far as possible from zetemata, quoting Lehrs, op. cit. 197, who however says the opposite, viz. that such things are to be found in Aristarchus; Ludwich himself elsewhere admitted this (A. H. T. I, p. 30 n. 40, contrast n, p. 190).Google Scholar

⁹ Textgeschichte der gr. Lyriker, p. 75.

¹⁰ Like most scholiastic terms γράФ;ειυ is uncomfortably vague, meaning at times little more than ⋯ξηγεσθαι (Rutherford, , Scholia Aristophanea iii, p. 70;Google ScholarLudwich, , Aristarchs Homerische Textkritik I, p. 509 and II, p. 104;Google ScholarErbse, , Hermes 87 (1959), 280). For an example of γσ⋯Ф;ει = ‘read’ - ‘proposed reading’ - in a syggramma we have Ox. Pap. 221, ix, 3 on Il. 21. 195, where Megacleides .Google Scholar

¹¹ πασ' ⋯μ***ν is an odd way to talk about an author who had been dead for about 100 years, but Sopater lived in Alexandria.

¹² fr. 99 W²: Voigt fr. 359 attributes the reading χ⋯λυς to Dicaearchus, following Wilamowitz, but Gudeman R.E. s.v. Kallias already describes it correctly as a conjecture by Aristophanes.

¹³ Sch. Dion. Thrax, 11. 14 Hilg. and continues

¹⁴ Tortoise riddle: Cic, de div. 2. 133 from Pacuvius with Pease's note; A.P. 14. 30; Soph, fr. 279 N² = R; Seaford, Maia 28 (1976), 218 and 219 n. 83 with interesting remarks about the place of such riddles in satyr plays; a double flute riddle in A.P. 14. 14; and compare the oyster riddle with its learned ancient commentary, restored by Parsons ZPE 24 (1977), 1 ft. Gow on Theocritus 9. 25; Plut., Sept. Sap. 151 f.

¹⁵ The tortoise in the Homeric Hymn is a land-tortoise, but a turtle shell would do equally well for constructing a lyre; R.E. s.v. Schildkröte 427.

¹⁶ Strabo 12. 22. 1 The peripatetic methodology behind zetemata and their solutions is explored by Hintenland, H., Untersuchungen zu den Homer-aporien des Aristoteles (Heidelberg, diss., 1961), though zetemata will not always convincingly fit the theoretical framework. Another problem involving Dicaearchus and Aristarchus exists at Il. 3. 244, where the information given is contradictory and the obvious deduction (Erbse, Überlieferung, p. 327) is not, in my view, credible.Google Scholar

¹⁷ Athenaeus 10. 451 d = fr. 22 Michaelis; West on Archilochus fr. 185.

¹⁸ e.g. Sch. Pindar, O. 6. 154; Sch. Ar. Lysis. 991.

¹⁹ Poll. 4. 82 Athenaeus 4. 177a (from Juba) Hesych s.v. σκυτ⋯λια αὐλíδια Cf. Wilhelm, A., Akademieschriften I (Leipzig, 1974), p. 238.Google Scholar

²⁰ Antiphanes fr. 300 (2. 129K); see Thompson, D'Arcy, Glossary of Greek Fishes (London, 1947), s.v. Keryx.Google Scholar

²¹ The ancients in their works were only too easily diverted into philological zetemata which would seem to us quite irrelevant, e.g. Zopyrus in his Foundation of Miletus (FGrHist 494 Fl with Jacoby's commentary).

²² The criticism is already made in R.E. s.v. Muschel 794. No one except Aristophanes claims that limpets are musical (R.E. s.v. Schnecke no. 41), and though Eptcharmus fr. 114K. lists together τ⋯λλις and λεπ⋯ς as edible, this has no more force than χ⋯λυς and κ***σνξ together in Empedocles B76D-K⁶. Beazley ARV ² 929. 87 is said to show a Silenus playing a mussel. It could in fact be anything small; ‘like a balloon’ says Mrs A. Dunbabin, to whom I am grateful for the information.

²³ Lehrs, op. cit. 349 was reminded of the diligentia Batavorum, which Ludwich, Homervulgata, p. 169 translated as Cobetianer. On Ritschl's school Wilamowitz, Geschichte der Philologie³ (Berlin, 1927), p. 61.Google Scholar

²¹ Aelian H.A. 7. 39 = Ar. Byz. p. 61N = fr. 379 Slater; sch. Pind. O. 3. 52; Poll. 5. 76.

²⁵ 1460b 13.

²⁶

²⁷ Aristotle H.A. 538b 18 and P.A. 662a are both explicit with an explanation: sch. D on Iliad 3. 24 agrees; Brein, F., Der Hirsch in der griechischen Frühzeit (Vienna, diss., 1969), pp. 36–7.Google Scholar

²⁸ Aristophanis Historiae Animalium Epitome - Suppl. Aristotelicum 1, ed. Lambros, S. (Berlin, 1885). A papyrus of this exists in P. Lond. 164. Though W. Kroll, ‘Zur Geschichte der aristotelischen Zoologie’, Sitz. Ak. Wien, phil.-hist. Kl. 218, 2 (1940), 30 wishes to attribute nearly all variations from Aristotle in the epitome to Theophrastus, in this case at least this is highly unlikely.Google Scholar

²⁹ The passage is complex. Aristophanes begins correctly by talking about the pregnancy of does but fails to change the genders when he comes to behaviour. H.A. 611 a23 becomes p. 127,9–10L, and Aristophanes keeps the feminine in the epitome until he comes to 611b 26, where Aristotle is once again speaking of does. From H.A. 611 a30 to 611 b20 Aristotle is speaking of the growth of horns on the male, from which Aristophanes has preserved two sentences 127. 12–15L, in which he has altered the original to the feminine. But most revealing is the one passage where Aristotle in an aside says (H.A. 611a25): (Zenob. 2.22 = ps.-Zenob. 5.52) Aristophanes conceals the fact that Aristotle is speaking exceptionally of a proverb, and alters the order of epitomization so that he can say:

The proverb meant – contrary to the explanation of Demo found in the paroemiographers - ‘so remote as to be nowhere’ (Plut. Mor. 403d, 700d-a sympotic zetema), and nearly all our authorities, even when dependent on Aristotle (Ael. H.A. 6. 5), use the masculine; even R.E. s.v. Hirsch 1940 accidentally changes to masculine, while Zenobius 5. 52 quotes feminine and explains masculine. The tendency, therefore, is always to remove the feminine, and it must go back to a time when does were actually thought to have horns, or perhaps an adynaton was intended.

³⁰ Timotheus of Gaza, p. 131 Lambros.

³¹ Ignotum pro magnifico. The fons malorum would appear to be Wilamowitz in Euripides, Herakles i (Berlin, 1959, first 1889), pp. 145 ff. Arthur Ludwich was no admirer of Wilamowitz’ views on these matters: ’Neu ist an dieser ganzen Belehrung, wie so unendlich oft bei Wilamowitz, weiter nichts als die Schaustellung seines unbegrenzten Selbstbewusstseins und die Sucht aus unsicheren Hypothesen eine feste Tatsache aufzubauen’ (Aristarchs Homerische Textkritik ii, p. 220 n. 195).

³² I do not include in this list atheteses, though they can often be shown to be a solution to a zetema. I take this opportunity of saying that there is not enough evidence to say that Aristophanes edited anything, though he certainly commented on a vast variety of texts, and did bibliographic work with them. Pfeiffer's belief (op. cit. 173 ff.) that a scholar could produce very conservative editions, filled with signs, variants and conjectures, including his own, in the margin, which he explained in syggrammata and lectures, not in hypomnemata (nor I add in his lexical works, pace Pfeiffer), does not seem to me at all possible. He would know that such a text would be immediately unusable without a commentary, which Pfeiffer explicitly denies him. His pupils Aristarchus and Callistratus as well as many others after him produced commentaries; Euphronius, said to be his teacher, produced a commentary before him. It stands to reason that Aristarchus and Callistratus can tell us Aristophanes’ opinions because they possessed hypomnemata or lecture notes, whether composed by them or Aristophanes. If, then, his views survived even in selection in the works of others, his readings and solutions could be known as they are today, essentially through hypomnemata and not editions. It would be left to later scholars, obsessed with reconstructing Alexandrian texts from such notes, to assume that indeed editions by Alexandrian scholars had actually existed: Ludwich, , Rh.M. 69 (1914), 684;Google ScholarErbse, , Hermes 87 (1959), 275–303.Google Scholar The best treatment of the subject is Zuntz, G., Die Aristophanesscholien der Papyri² (Berlin, 1975), pp. 75 ff.Google Scholar

³³ Boiling, G. M., External Evidence for Interpolation in Homer (Oxford, 1925), p. 221:Google Scholar ‘ obviously interpolated’… ’they look like a lysis for the problem there (Porph. 26. 5) discussed. It is surprising that they should have made their way into the edition of Aristophanes.’ If this is so, how can Pfeiffer (op. cit. 173) possibly say that Aristophanes was reluctant to put conjectures into his text? The simple answer to both problems is that there was no text; cf. Nickau, op. cit. 17, and Fraser, , Ptolemaic Alexandria I (Oxford, 1972), p. 476.Google Scholar

³⁴ Heracleides wrote on Homeric zetemata (fr. 171–5 W², to which add another noted by Erbse, , Beiträge zur Überlieferung der Iliasscholien (Munich, 1960), p. 69 n. 2). Since he was an Academic, there can be no justification for making too sharp a distinction between Academic, Peripatetic and Alexandrian (or Pergamene) zetemata. Even the cynic-rhetorical aporiai of Zoilus and Antisthenes are noted by Alexandrians, and Il. 2. 674 is ‘omitted’ by Zenodotus because of a zetema by Antisthenes, though the line is known also to Aristotle.Google Scholar

³⁵ The criticism directed at Aristophanes is preserved by Aristarchus' hypomnemata, quoted either directly or indirectly by Didymus. Aristonicus never mentions Aristophanes' variants (Ludwich, Aristarchs Homerische Textkritik I, p. 53) because he comments on marginal signs, which did not include divergence from Aristophanes. We are thus entirely dependent on Didymus' choice from Aristarchus; this means that we hear nothing about his variants from Aristarchus in the first six books of the Iliad; cf. Nickau op. cit. 22–3.

³⁶ So Roemer, A.,Homerexegese Aristarchs (Paderborn, 1924), p. 245Google Scholar and Aristarchea, in Beltzner, Emil, Homerprobleme l (Leipzig, 1911), pp. 178–9;Google Scholar Roemer was learned and often penetrating, but ‘he had stuck together a picture of Homeric scholarship, dominated before and after Aristarchus by profound darkness, where Aristarchus and only Aristarchus had been right in his explanations’ (Schmidt, M., Die Erklärungen zum Weltbild Homers… = Zetemata 62 (Munich, 1976), 14).Google Scholar

³⁷ The scholium is from Didymus; similar allegations are found in Il. 2. 397; 23. 104; Od. 1. 356; 2. 70; 6. 29 and especially Il. 21. 130 of in connection with Aristophanes. At 21. 130 Aristophanes himself accuses unnamed sources of interpolating lines to solve a zetema, of which we now have no trace!

³⁸ Roemer, Homerexegese, p. 244; Aristarchea 176 comparing Eustathius' commentary 1610. 45. See also Erbse, , Beiträge zum Verständnis d. Odyssee (Berlin, 1972), p. 146. For insertion of negative to solve a problem cf. sch. Il. 9. 453.Google Scholar

³⁹ Roemer, Homerexegese, p. 15; Aristarchea 171; Bolling, External Evidence, p. 127, w ho rejects the view that the reading may have been that of Zenodotus (Duentzer, , de Zen. Studiis Homericis (Göttingen, 1848), p. 160;Google ScholarValk, M. van der, Researches…Iliad ii (Leiden, 1964), p. 150) and rightly I think sees an Aristophanic conjecture.Google Scholar

⁴⁰ See the extensive apparatus of Erbse, where Aristarchus' unhappy etymology is to be found.

⁴¹ I agree with Valk, Textual Criticism of the Odyssey (Leiden, 1949), p. 15 that the Aristophanic reading is a rewriting; and I do not attach to polis-editions the importance that Citti, Vichiana 3 (1966), 227–67 seeks to find here and elsewhere; scepticism also in Fraser, op. cit. 1, p. 328. Indeed precisely a passage like this shows that these texts have suffered from diorthosis at the hands of grammarians.Google Scholar

⁴⁸ Il. 23. 92 is the only place I know where an ancient athetesis corresponds to an omission in a prearistarchean text (see Erbse ad loc); it follows therefore that sometimes the Alexandrians could have had a textual basis for athetesis. Pasquali, , Storia² (Firenze, 1971), p. 230 admits that Zenodotus emended Anacreon, but grimly and illogically denies that he can have done the same with Homer; we do not have absolute proof, but there is the highest probability that Zenodotus ‘omitted’ lines knowing that he had no textual authority, but that on other occasions he did have such authority, even if only in the suggestions of previous grammarians.Google Scholar

Statements like are unfortunately not good evidence for omission in actual Homeric texts, as is sometimes thought (Fowler, ZPE 33 (1979), 21 n. 17), for no careful distinction was made in the use of to denote whether a text or critical commentary was intended (Erbse, , Hermes 87 (1959), 288; sch. Ar. Nub. 507, where ⋯ντíγσαФ;α mean commentaries); oὐ γσ⋯Ф;oυσι means ⋯Ф;αιρo***σν (Sch. Ar. Ran. 153), and therefore in some sense athetesis in a commentary; even oὐκ o***δε can mean ‘disapproves of (Valk, Researches… Iliad ii, pp. 487–8). Add to this imprecision unsubstantiated allegation and cumulative error, and a definite answer to the diplomatic basis for Alexandrian readings becomes impossible.Google Scholar

⁴³ Porph. 1. 141. 17 Schr. and Erbse's apparatus to sch. Il. 9. 682.

⁴⁴ Satyros ⋯ ζ***τα, Sosibios ⋯ θαυμ⋯σιoς λντικóς, Zenodotos of Alexandria (R.E. s.v. Zenodotos no. 2), Apollodorus, all produced titles redolent of zetemata; but titles alone do not reveal such things. Cf. Plut. Sept. Sap. 150c on Dionysus Lysios.

⁴⁵ Qu. C. 9. 2.

⁴⁶ Sept. Sap. 10. There a problema in a book is brought to a symposium in a dining room attached to an Aphrodite temple.

⁴⁷ Polybius 8. 27 three times emphasizes the festivity of the Museum at Tarentum, which like those at Metapontum and Croton may be linked with Pythagoras (Boyancé, , Le Culte des Muses…² (Paris, 1972), pp. 235–7). For other Musea, see Fraser, op. cit. u, p. 469n. 72; A. Wilhelm, Akademieschriften II, p. 87; J. and L. Robert, Bull. Epigraphique 1979, no. 51.Google Scholar

⁴⁸ Herodas, Mime 1. 48: Moνσ***oν, o***νoς

⁴⁹ Anecdotes have largely obscured the truth about the Museum. The best known story about Aristophanes was that an elephant ran off with his girl friend. More serious is the fact that many of the stories accepted as telling us something about the Library are of very doubtful authenticity, as has been indicated by Mueller, C. W., Die Kurzdialoge der Appendix Platonica (Munich, 1975) in his first chapter. But the story of the symposia of the Septuagint translators is worth noting (texts in Erbse, Beiträge zur Überlieferung…, p. 67 n. 2).Google Scholar

⁵⁰ Evidence in Fraser, op. cit. I, p. 315 and in R.E. s.v. Museion, col. 806 & 808f. (Müller-Graupa); F. Studniczka, Das Symposium Ptolemaios Il = Abh. Säch. Ges. 30, 2 (Leipzig, 1914), pp. 25, 31, 108.

⁵¹ Poll. 6. 107; Ar. Vesp. 15–23; Athenaeus, 448b–459b.

⁵² Herter, H., Platons Akademie² (Bonn, 1952), p. 9 with notes p. 29.Google Scholar

⁵³ Erbse, Beiträge zur Überlieferung…, p. 6 7: he argues that the imperial concept of learned Museum conversation was not justified, but his ‘recht unzweideutig’ evidence for its frivolity is also imperial. His two criticisms of Porphyry's statement about Museum zetemata, which he regards as an invention, lack conviction. (1) He says that the example following, to be correct, should be a list of solutions when in fact it only gives one; but Porphyry himself says that he is only selecting (1. 147. 5 Schr., quoted by Erbse, p. 64); and since he certainly had several books of zetemata available to him, he could possibly make such a statement on good advice. (2) The example would be of a lysis where the text had already been athetized. But athetesis was just one kind of solution; e.g. at Il. 18. 10 Aristophanes got rid of lines involving a zetema where Aristarchus placed a diple. Athetesis was no more a ‘rohe Mittel’ than rewriting. Ludwich, A. H. T. ii, p. 173 called it the ‘unschuldigste Mittel’. In fact Erbse writes ‘Was Porphyrios… berichtet, könnte zur Not in den ersten Abendstunden unter Angehörigen des Museums diskutiert worden sein’ but ’was die Philologen zu antworten hatten, sicherlich nicht…‘ I cannot see how one is possible without the other.

⁵⁴ Are Plutarch's Quaestiones Conviviales too distant from reality either in (a) their effusive citations, (b) the nature of the zetemata discussed, (c) their combination of σπoυδή and παιδι⋯?

⁵⁵ Fraser, op. cit. II. 471 n. 86, referring to Erbse, Beiträge z. Überlieferung…, p. 67.

⁵⁶ op. cit. 70; cf. 263.

⁵⁷ We possess only a fraction of the titles of the period, and new finds do not fit well with our literary evidence. We did not know that Aristarchus worked on prose until a papyrus revealed that he wrote on Herodotus (Pfeiffer, op. cit. 224).

⁵⁸ Pfeiffer, op. cit. 263 seems to want to argue it out of existence.

⁵⁹ The language of sobriety seems to go back to Lehrs, who assures us (op. cit. 197 and 206) that Aristarchus was a virum sobrium in melioribus exquirendis occupatum.

⁶⁰ Plato, Protagoras 339d; Ar. Nub. 145 ff. with sch. 145a.

⁶¹ Bühler, W., Jb. d. Ak. d. Wiss. zu Göttingen (1977), 44–62Google Scholar in a good summary (with further reference to the amusing essay of Kassel, R., Winkelbrummer (Berlin, 1974)) offers a sound judgement on the quality of ancient zetemata; after allotting due praise, he writes: ‘what one misses in them… is the ability to be critical of their own ideas’. There is a fine list of ’frivolous’ zetemata assembled by Mayor on Juvenal 7. 234. Seneca cites as particularly frivolous (de brev. vit. 31. 1) the question of the priority of Iliad and Odyssey; he has not made much impression.Google Scholar

⁶² Polybius 12. 26c.

⁶³ fr. 5 + 20 Usener = 9+12, 2 Arrighetti: must mean that in Epicurus' day the symposium was the first place to find Plutarch goes on after quoting these passages (Mor. 1095e):

Here I think βασιλικóς means symposiarchic, as in the Academic sympotic laws mentioned by Athenaeus 1. 3 f. (Bergk, T., Fünf Abhandlungen z. Ges. der gr. Philosophie und Astronomie (Leipzig, 1883), p. 67 n. 1); Plutarch would be making a joke. But at Plut. Mor. 150b the same phrase means ‘problems about kingship’.Google Scholar

⁶⁴ Strabo 2. 3. 7 criticizes a scholar for i.e. turning a serious question into a frivolous rhetorical exercise; cf. Aphthonius, Progymnasmata 50. 1 Spengel:

⁶⁵ Lehrs, op. cit. 197 and 335. Of the second he writes: hic non grammaticum audimus sed philosophum, but several of the lines were athetized in antiquity by grammarians. Aristarchus wrote whole books against Comanus, presumably attacking his solutions among other things.