TY - JOUR
AU - Schironi, Francesca
AB - As is well known, we have lost Aristarchus’ original works and rely only on excerpta, mostly preserved in scholia and lexica dating back to the Byzantine period. The richest sources are the scholia maiora to Homer, especially those in the famous codex Venetus A (tenth century ce), which preserves ample excerpts of the so-called Viermännerkommentar, the ‘commentary of the four men’—Didymus, Aristonicus, Herodian, and Nicanor. These four scholars, who lived and worked between the first century bce and the second century ce, collected notes and readings from Aristarchus’ commentaries on Homer. In particular, Didymus’ work On the Homeric Recension of Aristarchus (Περὶ τῆς Ἀρισταρχείου διορθώσεως) and Aristonicus’ treatise on Aristarchus’ critical signs (Περὶ τῶν σημείων τῶν τῆς Ἰλιάδος καὶ Ὀδυσσείας) are the most important for reconstructing Aristarchus’ work.1 While the scholia (with all the problems involved in their tradition) are still the best source for Aristarchus’ work on Homer, trying to determine the exact content or wording of Aristarchus’ hypomnēmata is a completely different matter. This paper aims at exploring the value of the scholia as sources for Aristarchus, clarifying what they can tell us and what they cannot. I will focus my attention on the ‘most reliable’ fragments of Aristarchus that have reached us, that is, the few instances where Didymus and Herodian directly quote his words (Aristonicus and Nicanor do not give direct quotations of Aristarchus, at least in the scholia to the Iliad, which is my evidence for this analysis). These verbatim quotations will give us what most closely approaches ‘the very words’ of Aristarchus—if, that is, we trust that Didymus and Herodian transcribed his words correctly. Furthermore, even if we trust that these ancient scholars preserved the original notes of Aristarchus correctly, we have also to take into account that the Byzantine scholia are derivative products, the result of several rounds of copying and often excerpting of sources from the Imperial age. Therefore, the possibility that scholia have suffered in the course of transmission and that they may preserve only a tiny fraction of the original note by the ‘four men’ must be always be kept in mind. Yet, since the scholia are the only or, at least, main evidence we have, we must work with the data that these notes give us. So it might be worth examining them critically in order to reach some conclusion about Aristarchus’ very words. First (§§ 1–3) I will go over the few direct quotations of Didymus and Herodian, focusing on what they can tell us about the content and language of Aristarchus’ hypomnēmata;2 in some of these cases, it is not always clear where the actual quotation ends; this additional element of doubt will be taken into account in my analysis; yet I will also make clear what in my view is the extent of the quotation by writing it in boldface.3 I will then (§ 4) compare these direct quotations to the testimony of Aristonicus, who, though never quoting Aristarchus verbatim, is considered his best source. This comparison will offer some (limited) test for Aristonicus’ reliability. It will also enable us to reflect on the selection process that inevitably occurred in antiquity, even when scholars aimed at reporting previous authorities faithfully. 1. DIRECT QUOTATIONS FROM DIDYMUS’ SCHOLIA Didymus is fond of quoting Aristarchus directly and offers several instances of verbatim quotations, which he often introduces as Ἀριστάρχου λέξεις (‘Aristarchus’ words’). Among them we can distinguish two subgroups: quotations which are certainly taken from Aristarchus’ commentaries and those which are probably taken from those sources. The distinction is determined by the phrasing of the scholium. The secure quotations are those in which Didymus mentions the name of Aristarchus and also specifically says that they are taken from his hypomnēmata to the Iliad (Σ Il. 1.423–24, 2.111b, 2.125a, 3.406a1, 20.471a1). The other group (Σ Il. 2.397b, 2.420a1, 2.435a1) does present direct quotations, but Didymus either does not specifically attribute them to Aristarchus, or says that the lexeis are from ‘commentaries’ without specifying that they are those of Aristarchus—and we thus have to infer that they are. In what follows I will consider these two types of quotation together and analyze them in terms of content. Yet it is important to keep this distinction in mind when assessing the reliability of the information offered by each piece of evidence. Aristarchus certainly used paraphrases to clarify Homeric language. In fact, a direct quotation in Σ Il. 2.435a1 preserves two ‘nested’ paraphrases to explain a line with many Homeric glosses: Σ A Il. 2.435a1 Did., μηκέτι νῦν δήθ’ αὖθι λεγώμεθα: οὕτως αἱ Ἀριστάρχου. λέξεις ἐκ τοῦ Β τῆς ᾿Ιλιάδος. ‘δηθά πολὺν χρόνον, αὖθι αὐτοῦ, λεγώμεθα συναθροιζώμεθα. ὁ δὲ λόγος τοιοῦτος· “μηκέτι νῦν ἐπὶ πολὺν χρόνον αὐτοῦ συνηθροισμένοι μένωμεν”. Ζηνόδοτος δὲ ποιεῖ “μηκέτι †νῦν† ταῦτα λεγώμεθα”’. Καλλίστρατος δὲ ἐν τῷ πρώτῳ Περὶ ᾿Ιλιάδος οὕτως προφέρεται ‘μηκέτι δὴ νῦν αὖθι λεγώμεθα’. καὶ ἔστιν εὐφραδὴς μᾶλλον, ἀλλ’ οὐκ Ἀριστάρχειος. ταῦτα ὁ Δίδυμος (p. 113 Schm.). ‘Let us now not remain gathered here anymore’: so the [editions] of Aristarchus. [Here are his] words from the [commentary] to Iliad 2: ‘δηθά [means] “for a long time”, αὖθι [means] “here”, λεγώμεθα [means] “let us gather together”. This is the meaning: “let us not remain here assembled for a long time”. Zenodotus instead writes: “Let us †now† not talk about these things anymore”.’ Callistratus in the first book On the Iliad prefers: ‘Let us now not remain gathered here anymore’, and this reading is better expressed, but not by Aristarchus. Didymus [writes] this (p. 113 Schm.). After the sacrifice and the banquet in Iliad 2, Nestor speaks to Agamemnon and invites him not to waste time, but rather summon the army, saying ‘Let us now not remain gathered here anymore’ (μηκέτι νῦν δήθ’ αὖθι λεγώμεθα). According to Didymus, Aristarchus divides up the sentence into its components and translates each of them into a more straightforward form: he renders δηθά as πολὺν χρόνον, αὖθι as αὐτοῦ, and λεγώμεθα as συναθροιζώμεθα. After this word-for-word rendering, he gives a full paraphrase: ‘let us not remain here assembled for a long time’ (‘μηκέτι νῦν ἐπὶ πολὺν χρόνον αὐτοῦ συνηθροισμένοι μένωμεν’). Then, if this is still part of the direct quotation, Aristarchus mentions a different reading by Zenodotus. Didymus does not preserve Aristarchus’ comments on it; rather he interrupts his direct quotation and mentions the reading of Callistratus (a contemporary of Aristarchus), which Didymus himself prefers.4 A rather more complex operation is carried out in the following scholium: Σ A Il. 2.125a Did., Τρῶας μὲν λέξασθαι <ἐφέστιοι ὅσσοι ἔασιν>: Ἀριστάρχου λέξεις ἐκ τῶν ὑπομνημάτων· ‘ἐὰν καὶ τὰ παιδία καὶ τοὺς πρεσβύτας ἀριθμήσωσιν, οὐ μὴ γένωνται ὅσοι ἡμεῖς. τοῦτο δέ ἐστι, κατοικίδιοι καὶ ἔναυλοι. καὶ ἐν ᾿Οδυσσείᾳ (6.265)· “πᾶσιν γὰρ ἐπίστιόν ἐστιν ἑκάστῳ”, ὡσεὶ κατάλυμα παρὰ τῇ νηΐ.’ ‘To gather the Trojans, those who are at home’: Aristarchus’ words from his commentaries: ‘if they counted both the children and the elders, they would not be as many as we are. And that [word, i.e. ἐφέστιοι] means “those who stay at home” and “those who live inside the court”. And in the Odyssey (6.265) [there is]: “for all [the ships], one for each man, there is a shed (ἐπίστιον)”, that is, a lodging near the ship.’ At Il. 2.123–30, to highlight the fact that the Greeks greatly outnumber the Trojans, Agamemnon says that if the Greeks and all the Trojans, even those at home (ἐφέστιοι), i.e. the civilians, were gathered together and counted, and if the Greeks were to be grouped by tens and each group had a Trojan to pour wine, there would still be groups without a wine pourer.5 Aristarchus summarizes the general sense of these lines more concisely (and clearly!): ‘if [the Trojans] counted both the children and the elders, they would not be as many as we are’.6 Then, he focuses on the most abstruse word, ἐφέστιοι, and translates into koinē Greek: κατοικίδιοι καὶ ἔναυλοι; in particular, κατ-οικίδιοι is an ‘intralingual calque’ of the compound ἐφ-έστιοι, as κατά corresponds to ἐπί and οἰκία is the common equivalent of the more poetic ἑστία, ‘hearth of the house’. Finally, if the last sentence is still part of the direct quotation by Didymus (which I think it is; see n. 7), Aristarchus recalls a parallel in Odyssey 6.265, with the similar but more phonetically problematic ἐπίστιον (‘ship shed’), to prove the latter’s meaning on the basis of an etymological connection.7 The richest among the quotations of the λέξεις Ἀριστάρχου comes from book 1: Σ A. Il. 1.423–24 Did., <Ζεὺς … μετ’ ἀμύμονας Αἰθιοπῆας |> χθιζὸς ἔβη <κατὰ δαῖτα, θεοὶ δ’ ἅμα πάντες ἕπονται>: λέξις Ἀριστάρχου ἐκ τοῦ Α τῆς ᾿Ιλιάδος ὑπομνήματος· ‘τὸ μὲν μετ’ ἀμύμονας (423) ἐπ’ ἀμύμονας, ὅ ἐστι πρὸς ἀμώμους, ἀγαθούς, τὸ δὲ κατὰ δαῖτα (424) ἀντὶ τοῦ ἐπὶ δαῖτα· οὕτως γὰρ νῦν ῞Ομηρος τέθεικεν. ἔνιοι δὲ ποιοῦσι “μετὰ δαῖτα”, ὅπως ᾖ αὐτοῖς αὐτόθεν τὸ μετά ἐπί. χρῶνται δὲ καὶ πλείονες ἄλλοι τῶν ποιητῶν τῇ κατά ἀντὶ τῆς ἐπί. Σοφοκλῆς (TrGF 4 F 898)· “ἐγὼ κατ’ αὐτόν, ὡς ὁρᾷς, ἐξέρχομαι”.’ οὕτως δὲ εὕρομεν καὶ ἐν τῇ Μασσαλιωτικῇ καὶ Σινωπικῇ καὶ Κυπρίᾳ καὶ Ἀντιμαχείῳ καὶ Ἀριστοφανείῳ. Καλλίστρατος δὲ ἐν τῷ Πρὸς τὰς ἀθετήσεις (p. 320, 36) ὁμοίως, καὶ ὁ Σιδώνιος καὶ ὁ Ἰξίων ἐν τῷ ἕκτῳ Πρὸς τὰς ἐξηγήσεις (fr. 27). καὶ τὸ ‘θεοὶ †δὲ† πάντες ἕπονται’ (424) διὰ τοῦ αι. ‘Yesterday Zeus went to the Ocean, to the blameless Ethiopians (μετ’ ἀμύμονας Αἰθιοπῆας) for a banquet (κατὰ δαῖτα), and the rest of the gods are following’: Aristarchus’ phrasing from the commentary to Iliad 1: ‘the [expression] μετ’ ἀμύμονας (v. 423) [is used instead of] ἐπ’ ἀμύμονας, which means “to the faultless, valorous [Ethiopians]”; the [expression] κατὰ δαῖτα (v. 424) [is used] instead of ἐπὶ δαῖτα, for this is how Homer has used it here. Some, however, write “μετὰ δαῖτα”, so that for them μετά is simply ἐπί. But in fact the majority of the other poets use κατά instead of ἐπί. Sophocles [has]: “as you see, I am going to (κατά) him” (TrGF 4 F 898).’ We also found this [reading] in the editions from Marseilles, from Sinope, from Cyprus and in those by Antimachus (fr. 168 M.) and by Aristophanes (p. 175 Slater). Of the same opinion [are] Callistratus in the work Against the Atheteseis [i.e. of Aristarchus] (p. 320, 36), [Dionysius] Sidonius, and [Demetrius] Ixion in the sixth book of the work Against the Explanations [i.e. of Aristarchus] (fr. 27). And θεοὶ †δὲ† πάντες ἕπονται’ (v. 424) [is written] with αι [i.e. it is not ἕποντο]. The question concerns Homer’s use of different prepositions compared to koinē Greek. Aristarchus notes that Homer uses μετά instead of ἐπί (in μετ’ ἀμύμονας Αἰθιοπῆας), and κατά instead of ἐπί (in κατὰ δαῖτα).8 Then, focusing on κατὰ δαῖτα, Aristarchus argues against some scholars (ἔνιοι) who change the text and write ‘μετὰ δαῖτα’ in order to have a more regular text, with μετ’ ἀμύμονας … μετὰ δαῖτα, where μετά always corresponds to koinē Greek ἐπί with accusative. He defends the reading κατὰ δαῖτα by arguing that other poets too use κατά in the sense of ἐπί, quoting a line from Sophocles as an example. In addition, this direct quotation shows Aristarchus’ interest in etymology. When translating μετ’ ἀμύμονας into πρὸς ἀμώμους, ἀγαθούς, in addition to using the most common preposition in koinē for movement towards a person (πρός), he translates ἀμύμονας twice. First he renders ἀμύμονας with ἀμώμους, which preserves an etymological link with the gloss: both ἀμύμονας and ἀμώμους are compounds with privative ἀ- followed by either -μύμων or the easier -μωμος, which are both connected with μῶμος, ‘blame’.9 As a second step he retranslates the gloss into a very common koinē word, ἀγαθούς, in order to make its semantic value fully clear.10 Aristarchus also used paraphrases to explain a variant reading, and this happens in two cases of direct quotations: Σ A. Il. 3.406a1 Did., θεῶν δ’ ἀπόειπε κελεύθους: Ἀρίσταρχος ‘ἀπόεικε’ διὰ τοῦ κ, καὶ χωρὶς τοῦ ς ‘κελεύθου’. […]. προσθήσειν μοι δοκῶ καὶ τὴν Ἀριστάρχου λέξιν οὕτως ἔχουσαν· ‘†τί δὲ † (Erbse: τῆς δὲ Bekker) εἰς τὰς θεοὺς ὁδοῦ εἶκε καὶ παραχώρει, μὴ βαδίζουσα εἰς αὐτούς’. ‘Reject (ἀπόειπε) the paths (κελεύθους) of the gods’: Aristarchus [reads] ἀπόεικε [‘withdraw from’] with κ and κελεύθου without ς. […] And I think that it is worth adding Aristarchus’ phrasing, which is as follows: ‘retire from the road to the gods and withdraw, without going towards them’. Σ A Il. 20.471a1 Did., κόλπον ἐνέπλησεν: ‘ἐνέπρησεν’ Φιλόξενος (fr. 217 Th.) καὶ Ἀρίσταρχος. περὶ δὲ τῆς γραφῆς Ἀρίσταρχος ἐν ὑπομνήματί φησιν οὕτως· ‘τὸ αἷμα ἀπὸ τοῦ ἥπατος ἐκρούνιζε χύδην καὶ τὸν τοῦ χιτῶνος κόλπον ἐνεφύσησεν’. ‘[The dark blood] filled up (ἐνέπλησεν) the fold [of his tunic]’: Philoxenus (fr. 217 Th.) and Aristarchus [read] ἐνέπρησεν [‘inflated’]. In his commentary, Aristarchus says thus about this reading: ‘the blood from the liver was streaming in floods and blew up (ἐνεφύσησεν) the fold of the chiton’. Even if Didymus does not specify the source of the direct quotation from Aristarchus in Σ Il. 3.406a1, it is very likely that it is the hypomnēma, just like with Σ Il. 20.471a1. In both cases Aristarchus uses a paraphrase to make clear how the reading he favours must be understood. At Il. 3.406, angered with Aphrodite, because she forces her to go to Paris, Helen invites the goddess to do that herself, leaving the company of the gods. One reading was θεῶν δ’ ἀπόειπε κελεύθους, from ἀπολέγω, ‘I renounce’ (i.e. ‘give up the paths of the gods’), whereas Aristarchus reads ἀπόεικε, from ἀποείκω, ‘I withdraw from’, with genitive (i.e. θεῶν δ’ ἀπόεικε κελεύθου, ‘withdraw from the path of the gods’). In order to clarify his reading, Aristarchus rewords it using the simple [τῆς] εἰς τὰς θεοὺς ὁδοῦ εἶκε, which is also a word-for-word paraphrase of θεῶν δ’ ἀπόεικε κελεύθου, where the simple εἶκε translates the compound ἀπόεικε, ὁδοῦ translates the Homeric κελεύθου, and εἰς τὰς θεοὺς clarifies the objective genitive θεῶν (i.e. ‘the paths of the gods’ means ‘the paths towards the gods’); as in Σ Il. 2.435a1, he also adds a ‘coda’ which further paraphrases the meaning: ‘and withdraw, without going towards them’. A similar operation is carried out with Il. 20.471, when Homer describes Achilles mortally wounding Troos in the liver and blood filling up (ἐνέπλησεν) Troos’ tunic. Aristarchus prefers ἐνέπρησεν from ἐμπρήθω ‘I inflate’ and explains his reading by rephrasing the line with a more familiar, easier synonym, ἐμφυσάω: ‘the blood from the liver was streaming in floods and inflated (ἐνεφύσησεν) the fold of the chiton’. In addition to a paraphrase, another scholium to Iliad 2 gives an interesting comment by Aristarchus: Σ A Il. 2.420a1 Did., ἀλλ’ ὅ γε δέκτο μὲν ἱρά, πόνον δ’ ἀλίαστον <ὄφελλεν>: τούτῳ καὶ λέξις ὑπόκειται διὰ τοῦ Β τῶν ὑπομνημάτων καὶ ἔστιν ἐμφατικωτέρα. οὐ χεῖρον δ’ ἂν εἴη τὴν ἐκλογὴν ἐκθεῖναι· ‘ἀλλ’ ὅ γε δέκτο μὲν ἱρά, πόνον δ’ ἀλίαστον ὄφελλεν, οἷον αἴσια ἐσήμαινε<ν> ὥστε λέγειν τοὺς μάντεις ὅτι δέδεκται, τοιοῦτον μέντοι πόνον αὐτὸς ηὖξεν, ὃν οὐκ ἂν ἐκκλίνειεν· γελοῖον γὰρ ἂν ἦν, εἰ μὴ ἐποίει τοῦτο.’ ‘He accepted the sacrifice, but caused relentless (ἀλίαστον) toil’: And this [reading] is also supported by a sentence in the commentary to Iliad 2 and it is quite expressive. And it might be better to offer the quotation: ‘“He accepted the sacrifice, but caused relentless (ἀλίαστον) toil”: in the sense that he [Zeus] gave auspicious signs so that the seers said that he had accepted [the sacrifice], but nevertheless he himself increased that toil, which he could not avert. For it would have been ridiculous if he were not doing that.’ In book 2, after the Greeks offer a sacrifice to Zeus before battle, Homer comments that though Zeus accepted the sacrifice, he nevertheless compounded their toil, as he did not allow them to conquer Troy that day. The vulgate has πόνον δ’ ἀμέγαρτον, ‘unenviable toil’, but another ancient reading was πόνον δ’ ἀλίαστον, ‘relentless toil’. The way Didymus introduces the direct quotation shows that Aristarchus’ commentary (if it is indeed Aristarchus’) was organized by lemmas followed by the explanation; here, in particular, the lemma (underlined bold) is immediately followed by the paraphrase which makes v. 420, i.e. the lemma with the reading ἀλίαστον, more plain. In fact, the paraphrase is longer than the original; Aristarchus uses an entire relative clause11 to translate a single word (ἀλίαστον). So here too Aristarchus employs a paraphrase to clarify a specific reading which he prefers. He also adds that the seers had revealed that Zeus accepted the sacrifice, a detail which, though it is not explicit in the text, Aristarchus inferred for the sake of clarity. Aristarchus’ final comment (‘for it would have been ridiculous if he were not doing that’) must mean that it was necessary for Zeus to give ‘relentless toil’ to the Greeks because of his promise to Thetis in book 1. Otherwise, Zeus would have been convinced by a sacrifice by humans to break his promise to Thetis, a goddess—a ‘ridiculous’ idea.12 With this note, therefore, Aristarchus chooses a specific reading and clarifies its text but also saves Homer from accusations of depicting the gods, and especially Zeus, as behaving inconsistently or even in a ridiculous way. For him, the two statements are compatible: Zeus accepts the sacrifice (and so urges Agamemnon to fight), but also continues to cause the Greeks suffering so as to fulfil his promise to Thetis. Homeric usage is the topic of the following direct quotation: Σ A Il. 2.111b Did., {Ζεύς με} μέγα {Κρονίδης}: […] κἀν ταῖς Λιταῖς ἐξηγούμενος ‘αὐτὰρ ἔπειτ’ Αἴας τε μέγας’ (Il. 9.169) ἔν τινι τῶν ἠκριβωμένων ὑπομνημάτων γράφει ταῦτα κατὰ λέξιν· ‘οὐ κατ’ ἐπίθετον λέγει μέγας, ἀλλὰ πρὸς ἀντιδιαστολὴν τοῦ ἑτέρου Αἴαντος. ὅταν δὲ λέγῃ “Ζεύς με μέγας Κρονίδης” (Il. 2.111), οὐκέτι ὅτι καὶ ἕτερος μικρός ἐστιν.’ καὶ τοὺς ἀπ’ αὐτοῦ δὲ χρωμένους ἔστιν εὑρεῖν τῇ γραφῇ Διονυσόδωρον καὶ Ἀμμώνιον τὸν Ἀλεξανδρέα. ἐπιλέγουσι δὲ οὐδέν· […] ταῦτα ὁ Δίδυμος (p. 112 Schm.). ‘{Zeus son of Cronus} [has] greatly (μέγα) [entangled] {me}’: […] And also in explaining ‘and then great (μέγας) Ajax’ (Il. 9.169) in the Embassy, he [i.e. Aristarchus]13 writes this verbatim in some of the perfected commentaries: ‘[Homer] does not use μέγας as an epithet [i.e. in the absolute sense], but to differentiate [him] from the other Ajax. When he says “great (μέγας) Zeus son of Cronus [has entangled] me” (Il. 2.111), however, it is not because there is also another, small [Zeus]’. And it is possible to find his pupils using this reading: Dionysodorus and Ammonius of Alexandria. But they do not say anything more. […] Didymus [writes] this (p. 112 Schm.). The problem concerns the expression Ζεύς με μέγα in Il. 2.111, for which another ancient reading was attested: Ζεύς με μέγας. The long scholium (here mostly omitted) is one of the most important witnesses for Aristarchus’ work on Homer (number and relative chronology of his ekdoseis, hypomnēmata, syggrammata), which I have discussed elsewhere;14 here I concentrate on the direct quotation, which comes from Aristarchus’ ‘perfected’ (second?) hypomnēmata on Il. 9.169, where Ajax is called μέγας. He points out two different usages of μέγας. The first one is relative, to differentiate two characters with the same name, in this case Ajax the Great as opposed to Ajax son of Oileus, also called the Lesser (πρὸς ἀντιδιαστολὴν τοῦ ἑτέρου Αἴαντος). The second use is absolute, as a self-standing epithet, to indicate an intrinsic quality,15 as in the case of Il. 2.111, where obviously Zeus is not described as μέγας because there is another, ‘smaller’ Zeus. The comparison between two passages of the Iliad thus functions to establish two different Homeric usages. A point about syntax is at the core of the last of Didymus’ direct quotations: Σ A Il. 2.397b Did., γένωνται: οὕτως γένωνται αἱ Ἀριστάρχου. τούτῳ δὲ καὶ ὁ ἐκ τῶν ὑπομνημάτων λόγος ὑπόκειται ἔχων τῇδε· ‘ἐπὶ τῶν κυμάτων λέγει τὸ γένωνται. τῷ τοιούτῳ πλεονάκις κέχρηται, ὥστε καὶ ἐπὶ τῶν οὐδετέρων τὰ πληθυντικὰ παραλαμβάνει.’ μεταποιοῦσι δέ τινες ‘γένηται’, οὐκ ὀρθῶς. ταῦτα ὁ Δίδυμος (p. 113 Schm.). ‘[Waves …] are produced (γένωνται)’: the editions of Aristarchus [have] γένωνται in this way. And this [reading] is also supported by his own words from the hypomnēmata, which are as follows:16 ‘[the poet] says γένωνται, referring to the waves. He has used this [figure] several times, so that he uses plural [verbs] also referring to neuter plural [subjects].’ Some change it into γένηται, not correctly. Didymus [writes] this (p. 113 Schm.). Here Aristarchus chooses a reading (κύματα … γένωνται), which is at variance to koinē use (according to which neuter plural subjects have singular verbs). He justifies it with Homer’s linguistic usage, as the poet frequently uses the plural verb with a neuter plural subject.17 The last comment about a wrong reading with a singular predicate is ambiguous: it is not strictly separated from the previous quotation and in fact goes along the same lines of Aristarchus’ reasoning, so it could be still part of the direct quotation. On the other hand, the concluding remark (‘Didymus [writes] this’) suggests that this is a personal comment by Didymus, added after Aristarchus’ quotation, as it would be odd to have ‘Didymus [writes] this’ right after Aristarchus’ words.18 2. DIRECT QUOTATIONS FROM HERODIAN’S SCHOLIA To these scholia from Didymus we can add the three cases where Herodian quotes Aristarchus verbatim: Σ A Il. 11.754a Hrd., {τόφρα γὰρ οὖν ἑπόμεσθα} δι’ ἀσπιδέος πεδίοιο: […] ὅ τε Ἀρίσταρχος ἐκεῖνο ἀποφαίνεται ὡς ὅτι· ‘τινὲς μὲν ἀπὸ τοῦ α ποιοῦνται τὴν διαστολήν, ἵν’ ᾖ “ἀσπιδέος”, ὡς εἰκαστικώτερον τοῦ ποιητοῦ ἀσπιδὲς τὸ πεδίον εἰρηκότος τῷ τὰ μακρὰ τῶν πεδίων καὶ εὐρέα περιφερῆ φαίνεσθαι, μηδενὸς ἄλλου ὁρωμένου τέρματος, ἀλλὰ τοῦ ὁρίζοντος ἀέρος. ἄλλοι δὲ ἐκδέχονται “ἀσπιδέος” τοῦ ἔχοντος πολλὰς ἀσπίδας, καθότι ἐπιφέρει “ἀνά τ’ ἔντεα καλὰ λέγοντες” (Il. 11.755). οἱ δέ φασιν ἐκ πλήρους “σπιδέος” καὶ ἀποδιδόασι πολλοῦ καὶ μακροῦ· καὶ γὰρ Αἰσχύλος πολλάκις τὴν λέξιν οὕτως ἔχουσαν τίθησιν, ὅταν λέγῃ (TrGF 3 F 378)· “σπίδιον μῆκος ὁδοῦ”, καὶ ὁ Ἀντίμαχος (fr. 149 Matthews)· “οὐδὲ σπιδόθεν προνοῆσαι”, τουτέστι μακρόθεν.’ καὶ ταῦτα μὲν ὁ Ἀρίσταρχος. […] ‘{For so long we followed [them]} along the wide plain (δι’ ἀσπιδέος πεδίοιο)’: […] Aristarchus says this: ‘some make the division from the α, so that it is ἀσπιδέος, as if the poet said that the plain was “shield-like” (ἀσπιδές) in a somewhat figurative way because long and vast plains seem rounded, as no other boundary is seen but only the air marks their limit. Others, on the other hand, understand ἀσπιδέος as “having many shields” since [the poet] adds: “and picking up their beautiful armour” (Il. 11.755). Still others say that the full form is σπιδέος, and understand it as “long and wide”. For Aeschylus, too, often uses this word with that meaning, when he says: “for a long (σπίδιον) length of road” (TrGF 3 F 378); and Antimachus: “not to foresee from afar (σπιδόθεν)” (fr. 149 Matthews), which is “from a long way”.’ And Aristarchus [says] this […] In Iliad 11.754 scriptio continua ΔΙΑΣΠΙΔΕΟΣΠΕΔΙΟΙΟ could be divided either as διὰ σπιδέος πεδίοιο, ‘along the long plain’, or as δι’ ἀσπιδέος πεδίοιο, which could either mean ‘along the rounded plain’ or ‘along with the plain covered by many shields’. We do not know what Aristarchus’ choice was.19 From the verbatim quotation, Aristarchus seems to have limited himself to listing the different solutions and commenting on them. In explaining these readings he refers to the very next verse (Il. 11.755, which refers to the battle gear that was scattered on the battlefield) and quotes other poets’ usage of similar words (specifically, similar uses of the adjective σπιδής by Aeschylus and Antimachus). We do not know if Aristarchus found these parallels in the sources from which he took these variant readings, that is, other scholars’ editions or commentaries, or came up with those examples himself. However, even if he was not the first to make those comparisons, he reported them in his commentary because they were important in the discussion of those readings. Perhaps the fact that the different readings were supported by parallels in Homer or in other poets made him especially cautious and uncommitted to any of them. Another case of scriptio continua deals with compounds: Σ A Il. 23.523b Hrd., {καὶ ἐς} δίσκουρα {λέλειπτο}: Πτολεμαῖος (p. 60 Baege) καὶ οἱ πλείους δίσκουρα ὡς ‘λίπουρα’ (Call. fr. 200b, 2). καὶ Ἀρίσταρχος δέ φησιν· ‘ὅπερ ἄνω κατὰ διάλυσιν εἶπεν “ὅσσα δὲ δίσκου οὖρα” (Il. 23.431), τοῦτο νῦν κατὰ σύνθεσιν ἐξήνεγκεν’. κἀγὼ δὲ συγκατατίθεμαι. ‘{He had fallen behind by} a discus’ cast (δίσκουρα)’: Ptolemy [of Ascalon] (p. 60 Baege) and the majority [read] δίσκουρα like λίπουρα (Call. fr. 200b 2) and Aristarchus says: ‘what he said before in two words, “as far as the cast of a discus (δίσκου οὖρα)” (Il. 23.431), now he has pronounced it as a compound’. And I agree. Aristarchus notes that at Il. 23.523 Homer uses the compound δίσκουρα, ‘discus’ cast’, for the distance covered by a discus throw, but earlier at Il. 23.431 the same idea is expressed with two words: δίσκου οὖρα. The direct quotation preserves two technical terms: διάλυσις, ‘resolution’, which refers to words written ‘separately’ (κατὰ διάλυσιν) and σύνθεσις, ‘composition’, for words written ‘in composition’ (κατὰ σύνθεσιν).20 Finally, another direct quotation concerns the accentuation of verbal forms: Σ A Il. 24.8a Hrd., πείρων: Πάμφιλος περισπᾷ, […]. ὁ μέντοι Ἀρίσταρχος βαρύνων ἀναγινώσκει, ὡς ‘κείρων’ (cf. Il. 21.204, Od. 24.459). φησὶ γοῦν οὕτω γενόμενος ἐπὶ τοῦ ‘παννυχίη μέν ῥ’ ἥ γε καὶ ἠῶ πεῖρε κέλευθον’ (Od. 2.434)· ‘τὸ “πεῖρε” διδάσκει ἡμᾶς καὶ τὴν “πείρων” μετοχὴν βαρύνειν· ὡς γὰρ ἔκειρε κείρων, οὕτως ἔπειρε πείρων· εἰ γὰρ περιεσπᾶτο, ἦν ἂν ὁ παρατατικὸς ἐπείρα.’ ἔστιν οὖν διττὸν τὸ ῥῆμα, καὶ τὸ μὲν σὺν τῷ ι βαρυνθήσεται, τὸ δὲ χωρὶς τοῦ ι περισπασθήσεται. ‘Making it through (πείρων) [wars of men and painful waves]’: Pamphilus pronounces [πείρων] as perispomenon, […]. Aristarchus, however, reads it as barytone like κείρων (cf. Il. 21.204, Od. 24.459). Thus, in discussing the [line] ‘thus all night long and through the dawn [the ship] made it through (πεῖρε) her journey’ (Od. 2.434), he says: ‘πεῖρε teaches us that the participle πείρων is also barytone, for as ἔκειρε κείρων, so ἔπειρε πείρων. If it were perispomenon, the imperfect would be ἐπείρα.’ The verb is double [i.e. πείρω, περάω]: the one with ι will be barytone, the one without ι will be perispomenon. The issue is the accent of the participle ΠΕΙΡΩΝ and, as a consequence, the verb from which this form derives. While Pamphilus suggests πειρῶν, Aristarchus reads it as paroxytone, πείρων. To prove his solution, he uses an analogy, which is reported in the direct quotation: just as the participle κείρων comes from the same verb as the imperfect ἔκειρε, so the participle corresponding to the imperfect ἔπειρε, occurring in the Odyssey, will be πείρων. The perispomenon accentuation (πειρῶν), on the other hand, would make the participle derive from a different verb, πειράω, ‘I try’, and so in Od. 2.434 the imperfect would be ἐπείρα. Aristarchus’ final comment implies that this is not the correct verb, since in both Il. 24.8 and Od. 2.434 the required meaning is ‘to go through’, hence the correct forms are those from πείρω.21 This quotation (which comes from the commentary to Od. 2.434) is particularly important because it proves that Aristarchus used four-term analogies to decide over debated inflectional patterns,22 and that he used the technical terms μετοχή for ‘participle’ and παρατατικός for ‘imperfect’.23 3. PRELIMINARY (AND POSITIVE) CONCLUSIONS: CONTENT AND TECHNICAL VOCABULARY There is a difference between Didymus and Herodian in reporting Aristarchus’ notes. Didymus’ direct quotations seem to be almost all centred on paraphrases, either as a way to make the text clear or to comment on a reading. Herodian’s quotations are instead more articulated and rich. If we only had access to Didymus’ quotations, Aristarchus’ commentaries would appear much blander and mono-dimensional, mostly focused on paraphrases. On the other hand, Didymus’ testimony suggests how important paraphrases were in ancient scholarship, and not only at the elementary level; they were a ubiquitous tool with which to discuss a literary text. But his modus operandi also suggests that Didymus was an avid reader (and transcriber) of other people’s words. Despite these differences, the comparative analysis of the direct quotations from Didymus and Herodian gives a general picture of the content and wording of Aristarchus’ commentaries. As for the content, the following list summarizes the scholarly interests which emerge from Aristarchus’ verbatim quotations (the items in double square brackets are those which might, but do not certainly, come from direct quotations of Aristarchus, in Σ Il. 2.397b, 2.420a1, 2.435a1):24 Paraphrases and intralingual translations: Translation of poetical terms into koinē Greek: ἀμύμονας = ἀγαθούς in 1.423–24 κελεύθου = ὁδοῦ in 3.406a1 [[δηθά = πολὺν χρόνον, αὖθι = αὐτοῦ, λεγώμεθα = συναθροιζώμεθα in 2.435a1]] Translation of Homeric compounds which also highlights the etymology of the compound itself: ἐφ-έστιοι = κατ-οικίδιοι in 2.125a ἀ-μύμονας = ἀ-μώμους in 1.423–24 Paraphrase/summary of long sentences into shorter ones, to make the sense clear: in 2.125a, [[2.435a1]] [[Paraphrase/summary of sentences into longer ones, to make the sense clear:]] [[in 2.420a1]] Paraphrase longer than the original to discuss/clarify the meaning of a variant reading: in 3.406a1 and 20.471a1, [[2.420a1]], 11.754a Analysis of Homer’s usage: use of different prepositions compared to koinē in 1.423–24 self-standing epithets vs. adjectives used to differentiate characters in 2.111b [[plural verb with neuter plural subject in 2.397b]] use of compounds as well as separate words in 23.523b Comparisons: With other passages/usages of the same author to clarify a meaning/etymology: ἐφέστιος/τὸ ἐπίστιον in Homer in 2.125a ἀσπιδέος, ‘with many shields’, in Il. 11.755 and 11.754a With other passages/usages of the same author to establish the author’s usage: μέγας with absolute or relative meaning in 2.111b πεῖρε/πείρων in 24.8a δίσκουρα/δίσκου οὖρα in 23.523b With other passages/usages from other poets to confirm a reading: κατά for ἐπί in Homer and Sophocles in 1.423–24 σπιδής in Homer with Aeschylus (σπίδιον) and Antimachus (σπιδόθεν) in 11.754a Mention and discussion of variants by other scholars: variants by anonymous scholars: in 1.423–24 [[variants by Zenodotus]]: [[in 2.435a1]] discussion of variants generated by different division of scriptio continua: in 11.754a Use of four-term analogies: To establish a reading: ἔκειρε: κείρων:: ἔπειρε: πείρων in 24.8a [[Attention to narrative/character consistency:]] [[Explanation of a passage/reading with reference to internal consistency in 2.420a1]] [[Form of Aristarchus’ commentary:]] [[lemma followed by the explanation in 2.420a1]] In these direct quotations we find the most common (and trivial) elements of schoolteaching: word-for-word paraphrases.25 This may surprise us; we might prefer to see the most famous philologist of antiquity engaging with a text in the manner of Wilamowitz instead of rephrasing the text as a schoolteacher does. However, the importance of paraphrase can be explained within the context of Aristarchus’ activity: a lecture in front of an audience who experienced numerous difficulties with aspects of vocabulary, morphology, and syntax in Homer, as these poems were written in a literary language quite remote from koinē Greek. The first task of the exegete was thus to make the text intelligible, paraphrasing syntactically contorted lines and translating difficult words. While paraphrases and intralingual translations are attested in the scholia minora26 and definitely used in schools, they were also popular at a higher level. In these verbatim quotations, Aristarchus also uses paraphrases to prove that his readings make sense. In addition, intralingual translations of a Homeric word into koinē allow discussion of poetic usages (e.g. different prepositions compared to koinē) or of the Homeric lexicon and diction. To clarify Homeric glosses, the best (and probably the only) place to look was of course Homer, because he was the only source for such a peculiar language; hence the parallels with other passages from the Iliad and Odyssey we find in the direct quotations. This is where the principle of ‘explaining Homer from Homer’ has its origin. Sometimes, however, when Homer did not provide suitable parallels, examples from other poets could be used to explain Homeric words and diction.27 Moreover, this evidence also brings to the fore one of the most interesting features of Aristarchus as a scholar: his systematic approach to an author and a text. For instance, in two distinct scholia, the equation ἐπί with κατά in Homer’s usage occurs for different purposes, suggesting that Aristarchus had elaborated a precise description of Homeric language, which he could always refer to when editing and commenting on Homer. The direct quotations also confirm that in his commentaries Aristarchus discussed variant readings (often introduced in an anonymous way, as a reading of ‘some’ scholars).28 In choosing a variant he could also use analogical proportions, through which he could compare four different forms, of which one was the object of debate, as we saw in the direct quotation discussing πείρων. Last but not least, he also paid attention to the context and explained the text to save Homer from alleged inconsistencies. The analysis of the direct quotations of Aristarchus allows also some conclusions concerning the technical vocabulary he used. A list of the technical expressions found in the direct quotations from Didymus and Herodian gives the following results (the boldface items are from the first group, of certain quotations; items from the second group—of less certain quotations in Σ Il. 2.397b, 2.420a1, 2.435a1—are not in boldface; italics indicate that certain keywords appear in the passages which most likely, but yet not securely, are still part of the quotation): To introduce a paraphrase Nothing (i.e. the paraphrase is simply introduced) (2.125a, 3.406a1, 20.471a1) [but in all these cases it might be due to Didymus’ way of quoting Aristarchus] ὡσεί … (2.125a) οἷον … (2.420a1) ὁ δὲ λόγος τοιοῦτος … (2.435a1) To translate a single word into koinē Greek Nothing (i.e. the koinē equivalent simply follows the Homeric word) (1.423–24, 2.435a1, 11.754a) τοῦτο δέ ἐστι (2.125a) τουτέστι (11.754a) ὅ ἐστι … (1.423–24) ἀντὶ τοῦ/ἀντὶ τῆς … (1.423–24) To clarify a word’s meaning and reject another one οὐ … λέγει …, ἀλλὰ … (2.111b) To note the author(s)’s usage οὕτως γὰρ νῦν Ὅμηρος τέθεικεν (1.423–24) χρῶνται δὲ καὶ πλείονες ἄλλοι τῶν ποιητῶν … (1.423–24) τῷ τοιούτῳ πλεονάκις κέχρηται (2.397b) ἐπὶ τῶν … τὰ … παραλαμβάνει (2.397b) ὅπερ ἄνω … εἶπεν, τοῦτο νῦν … ἐξήνεγκεν (23.523b) To ‘judge’ poetry and the poet’s ability γελοῖον γὰρ ἂν ἦν, εἰ … (2.420a1) εἰκαστικώτερον τοῦ ποιητοῦ … εἰρηκότος (11.754a) To report other scholars’ readings/interpretations ἔνιοι δὲ ποιοῦσι … (1.423–24) Ζηνόδοτος δὲ ποιεῖ … (2.435a1) τινὲς μὲν … ποιοῦνται … ἄλλοι δὲ ἐκδέχονται … οἱ δέ φασιν … (11.754a) To introduce parallels καὶ ἐν ᾿Οδυσσείᾳ … (2.125a) Σοφοκλῆς … (1.423–24) καὶ γὰρ Αἰσχύλος πολλάκις τὴν λέξιν οὕτως ἔχουσαν τίθησιν … (11.754a) ὁ Ἀντίμαχος … (11.754a) Linguistic or otherwise technical terminology ἐπίθετον (2.111b) οὐδέτερον (2.397b) πληθυντικόν (2.397b) σύνθεσις (23.523b) διάλυσις (23.523b) μετοχή (24.8a) παρατατικός (24.8a) Four-term analogies ὡς γὰρ ἔκειρε κείρων, οὕτως ἔπειρε πείρων (24.8a) τὸ ‘πεῖρε’ διδάσκει ἡμᾶς … (24.8a) To introduce a paraphrase Nothing (i.e. the paraphrase is simply introduced) (2.125a, 3.406a1, 20.471a1) [but in all these cases it might be due to Didymus’ way of quoting Aristarchus] ὡσεί … (2.125a) οἷον … (2.420a1) ὁ δὲ λόγος τοιοῦτος … (2.435a1) To translate a single word into koinē Greek Nothing (i.e. the koinē equivalent simply follows the Homeric word) (1.423–24, 2.435a1, 11.754a) τοῦτο δέ ἐστι (2.125a) τουτέστι (11.754a) ὅ ἐστι … (1.423–24) ἀντὶ τοῦ/ἀντὶ τῆς … (1.423–24) To clarify a word’s meaning and reject another one οὐ … λέγει …, ἀλλὰ … (2.111b) To note the author(s)’s usage οὕτως γὰρ νῦν Ὅμηρος τέθεικεν (1.423–24) χρῶνται δὲ καὶ πλείονες ἄλλοι τῶν ποιητῶν … (1.423–24) τῷ τοιούτῳ πλεονάκις κέχρηται (2.397b) ἐπὶ τῶν … τὰ … παραλαμβάνει (2.397b) ὅπερ ἄνω … εἶπεν, τοῦτο νῦν … ἐξήνεγκεν (23.523b) To ‘judge’ poetry and the poet’s ability γελοῖον γὰρ ἂν ἦν, εἰ … (2.420a1) εἰκαστικώτερον τοῦ ποιητοῦ … εἰρηκότος (11.754a) To report other scholars’ readings/interpretations ἔνιοι δὲ ποιοῦσι … (1.423–24) Ζηνόδοτος δὲ ποιεῖ … (2.435a1) τινὲς μὲν … ποιοῦνται … ἄλλοι δὲ ἐκδέχονται … οἱ δέ φασιν … (11.754a) To introduce parallels καὶ ἐν ᾿Οδυσσείᾳ … (2.125a) Σοφοκλῆς … (1.423–24) καὶ γὰρ Αἰσχύλος πολλάκις τὴν λέξιν οὕτως ἔχουσαν τίθησιν … (11.754a) ὁ Ἀντίμαχος … (11.754a) Linguistic or otherwise technical terminology ἐπίθετον (2.111b) οὐδέτερον (2.397b) πληθυντικόν (2.397b) σύνθεσις (23.523b) διάλυσις (23.523b) μετοχή (24.8a) παρατατικός (24.8a) Four-term analogies ὡς γὰρ ἔκειρε κείρων, οὕτως ἔπειρε πείρων (24.8a) τὸ ‘πεῖρε’ διδάσκει ἡμᾶς … (24.8a) Open in new tab To introduce a paraphrase Nothing (i.e. the paraphrase is simply introduced) (2.125a, 3.406a1, 20.471a1) [but in all these cases it might be due to Didymus’ way of quoting Aristarchus] ὡσεί … (2.125a) οἷον … (2.420a1) ὁ δὲ λόγος τοιοῦτος … (2.435a1) To translate a single word into koinē Greek Nothing (i.e. the koinē equivalent simply follows the Homeric word) (1.423–24, 2.435a1, 11.754a) τοῦτο δέ ἐστι (2.125a) τουτέστι (11.754a) ὅ ἐστι … (1.423–24) ἀντὶ τοῦ/ἀντὶ τῆς … (1.423–24) To clarify a word’s meaning and reject another one οὐ … λέγει …, ἀλλὰ … (2.111b) To note the author(s)’s usage οὕτως γὰρ νῦν Ὅμηρος τέθεικεν (1.423–24) χρῶνται δὲ καὶ πλείονες ἄλλοι τῶν ποιητῶν … (1.423–24) τῷ τοιούτῳ πλεονάκις κέχρηται (2.397b) ἐπὶ τῶν … τὰ … παραλαμβάνει (2.397b) ὅπερ ἄνω … εἶπεν, τοῦτο νῦν … ἐξήνεγκεν (23.523b) To ‘judge’ poetry and the poet’s ability γελοῖον γὰρ ἂν ἦν, εἰ … (2.420a1) εἰκαστικώτερον τοῦ ποιητοῦ … εἰρηκότος (11.754a) To report other scholars’ readings/interpretations ἔνιοι δὲ ποιοῦσι … (1.423–24) Ζηνόδοτος δὲ ποιεῖ … (2.435a1) τινὲς μὲν … ποιοῦνται … ἄλλοι δὲ ἐκδέχονται … οἱ δέ φασιν … (11.754a) To introduce parallels καὶ ἐν ᾿Οδυσσείᾳ … (2.125a) Σοφοκλῆς … (1.423–24) καὶ γὰρ Αἰσχύλος πολλάκις τὴν λέξιν οὕτως ἔχουσαν τίθησιν … (11.754a) ὁ Ἀντίμαχος … (11.754a) Linguistic or otherwise technical terminology ἐπίθετον (2.111b) οὐδέτερον (2.397b) πληθυντικόν (2.397b) σύνθεσις (23.523b) διάλυσις (23.523b) μετοχή (24.8a) παρατατικός (24.8a) Four-term analogies ὡς γὰρ ἔκειρε κείρων, οὕτως ἔπειρε πείρων (24.8a) τὸ ‘πεῖρε’ διδάσκει ἡμᾶς … (24.8a) To introduce a paraphrase Nothing (i.e. the paraphrase is simply introduced) (2.125a, 3.406a1, 20.471a1) [but in all these cases it might be due to Didymus’ way of quoting Aristarchus] ὡσεί … (2.125a) οἷον … (2.420a1) ὁ δὲ λόγος τοιοῦτος … (2.435a1) To translate a single word into koinē Greek Nothing (i.e. the koinē equivalent simply follows the Homeric word) (1.423–24, 2.435a1, 11.754a) τοῦτο δέ ἐστι (2.125a) τουτέστι (11.754a) ὅ ἐστι … (1.423–24) ἀντὶ τοῦ/ἀντὶ τῆς … (1.423–24) To clarify a word’s meaning and reject another one οὐ … λέγει …, ἀλλὰ … (2.111b) To note the author(s)’s usage οὕτως γὰρ νῦν Ὅμηρος τέθεικεν (1.423–24) χρῶνται δὲ καὶ πλείονες ἄλλοι τῶν ποιητῶν … (1.423–24) τῷ τοιούτῳ πλεονάκις κέχρηται (2.397b) ἐπὶ τῶν … τὰ … παραλαμβάνει (2.397b) ὅπερ ἄνω … εἶπεν, τοῦτο νῦν … ἐξήνεγκεν (23.523b) To ‘judge’ poetry and the poet’s ability γελοῖον γὰρ ἂν ἦν, εἰ … (2.420a1) εἰκαστικώτερον τοῦ ποιητοῦ … εἰρηκότος (11.754a) To report other scholars’ readings/interpretations ἔνιοι δὲ ποιοῦσι … (1.423–24) Ζηνόδοτος δὲ ποιεῖ … (2.435a1) τινὲς μὲν … ποιοῦνται … ἄλλοι δὲ ἐκδέχονται … οἱ δέ φασιν … (11.754a) To introduce parallels καὶ ἐν ᾿Οδυσσείᾳ … (2.125a) Σοφοκλῆς … (1.423–24) καὶ γὰρ Αἰσχύλος πολλάκις τὴν λέξιν οὕτως ἔχουσαν τίθησιν … (11.754a) ὁ Ἀντίμαχος … (11.754a) Linguistic or otherwise technical terminology ἐπίθετον (2.111b) οὐδέτερον (2.397b) πληθυντικόν (2.397b) σύνθεσις (23.523b) διάλυσις (23.523b) μετοχή (24.8a) παρατατικός (24.8a) Four-term analogies ὡς γὰρ ἔκειρε κείρων, οὕτως ἔπειρε πείρων (24.8a) τὸ ‘πεῖρε’ διδάσκει ἡμᾶς … (24.8a) Open in new tab Admittedly there is not much that is really ‘technical language’; yet this list may be used as a guide for testing phrasing in other Aristarchean scholia which are not direct quotations. If the terms are the same, then we can legitimately suspect that the wording may be original. On the other hand, all the other terms used by the scholia and which do not find any parallel here need to be tested case by case.29 4. IS ARISTONICUS RELIABLE? So far, I have not taken into consideration Aristonicus at all, because he does not preserve any verbatim quotation of Aristarchus. Yet he is considered his most reliable source.30 All in all, this survey confirms that in terms of content what we find in the Aristonicus scholia is probably reliable. Indeed, my survey of the scholia derived from Aristonicus on the Iliad shows that these scholia generally focus on the topics and methods which have emerged from our survey of Aristarchus’ direct quotations from Didymus and Herodian.31 Attention to Homeric usage, defence of Homer from accusations of inconsistency, discussion of variants, precise reference to other Homeric passages are typical traits of the exegesis transmitted by Aristonicus. In what follows, the comparison between these direct quotations in Didymus and Herodian and the corresponding scholia from Aristonicus will allow us to better assess his value as a source for Aristarchus, beyond generic similarity in content between the latter and the direct quotations analyzed above. 4.1 Aristonicus and Aristarchus’ Verbatim Quotations In three of Didymus’ direct quotations of Aristarchus we have a corresponding scholium by Aristonicus, which is not a direct quotation, but Aristonicus’ own rewording, presumably of the same source, Aristarchus’ commentaries on the Iliad. A comparison between the two, then, will allow us to test Aristonicus’ trustworthiness in reporting Aristarchus’ very words. In some cases the results are encouraging: Σ A Il. 2.397b Did., γένωνται: οὕτως γένωνται αἱ Ἀριστάρχου. τούτῳ δὲ καὶ ὁ ἐκ τῶν ὑπομνημάτων λόγος ὑπόκειται ἔχων τῇδε· ‘ἐπὶ τῶν κυμάτων λέγει τὸ γένωνται. τῷ τοιούτῳ πλεονάκις κέχρηται, ὥστε καὶ ἐπὶ τῶν οὐδετέρων τὰ πληθυντικὰ παραλαμβάνει.’ μεταποιοῦσι δέ τινες ‘γένηται’, οὐκ ὀρθῶς. ταῦτα ὁ Δίδυμος (p. 113 Schm.). ‘[Waves …] are produced (γένωνται)’: the editions of Aristarchus [have] γένωνται in this way. And this [reading] is also supported by his own words from the hypomnēmata, which are as follows: ‘[the poet] says γένωνται referring to the waves. He has used this [figure] several times, so that he uses plural [verbs] also referring to neuter [subjects].’ Some change it into γένηται, not correctly. Didymus [writes] this (p. 113 Schm.). Σ A Il. 2.397a Ariston., γένωνται: ὅτι οὐ γραπτέον, ὥς τινες, ‘γένηται’· Ὁμηρικώτερον γὰρ οὕτως λέγειν, γένωνται τὰ κύματα, ὡς ‘σπάρτα λέλυνται’ (Il. 2.135). ‘[Waves …] are produced (γένωνται)’: because one must not write, as some [do], γένηται; for it is more Homeric to say it in this way: ‘the waves (κύματα) are produced (γένωνται)’, like ‘the ropes (σπάρτα) are loose (λέλυνται)’ (Il. 2.135). The content of these notes is indeed the same: attention to Homeric usage and preference for a reading that respects it. Aristonicus, on the other hand, adds the parallel with Il. 2.135, which Didymus does not record; we cannot tell whether this parallel was there in the first place and Didymus simply excluded it from his quotation, or whether Aristonicus added it, perhaps on the basis of another note by Aristarchus at Il. 2.135.32 In the light of the Aristonican scholium, the last comment in the Didyman scholium (μεταποιοῦσι δέ τινες ‘γένηται’, οὐκ ὀρθῶς), which Erbse prints outside the quotation, indeed fits very well with Aristarchus’ logos, as we already suggested above (§ 1); still, the final comment (‘Didymus [writes] this’) makes me uncomfortable in considering everything part of the direct quotation without doubt, given the parallels with Σ Il. 2.111b and Σ Il. 2.435a1.33 Another comparison with Σ Il. 2.435a1 Did. allows more in-depth analysis of Aristonicus’ work: Σ A Il. 2.435a1 Did., μηκέτι νῦν δήθ’ αὖθι λεγώμεθα: οὕτως αἱ Ἀριστάρχου. λέξεις ἐκ τοῦ Β τῆς ᾿Ιλιάδος. ‘δηθά πολὺν χρόνον, αὖθι αὐτοῦ, λεγώμεθα συναθροιζώμεθα. ὁ δὲ λόγος τοιοῦτος· “μηκέτι νῦν ἐπὶ πολὺν χρόνον αὐτοῦ συνηθροισμένοι μένωμεν”. Ζηνόδοτος δὲ ποιεῖ “μηκέτι †νῦν† ταῦτα λεγώμεθα”.’ Καλλίστρατος δὲ ἐν τῷ πρώτῳ Περὶ ᾿Ιλιάδος οὕτως προφέρεται ‘μηκέτι δὴ νῦν αὖθι λεγώμεθα’. καὶ ἔστιν εὐφραδὴς μᾶλλον, ἀλλ’ οὐκ Ἀριστάρχειος. ταῦτα ὁ Δίδυμος (p. 113 Schm.). ‘Let us now not remain gathered here anymore’: so the [editions] of Aristarchus. [Here are] [his] words from [the commentary to] Iliad 2: ‘δηθά [means] “for a long time”, αὖθι [means] “here”, λεγώμεθα [means] “let us gather together”. This is the meaning: “let us not remain here assembled for a long time”. Zenodotus instead writes: “Let us †now† not talk about these things anymore”.’ Callistratus in the first book On the Iliad prefers: ‘Let us now not remain gathered here anymore’, and this reading is more accurate, but not by Aristarchus. Didymus [writes] this (p. 113 Schm.). As we have already seen, scholars do not agree on where Aristarchus’ direct quotation ends, whether it is just after the paraphrase, i.e. after μένωμεν (so Lehrs), or included the reference to Zenodotus’ reading (as Erbse prints);34 the latter option seems preferable, given the following scholium by Aristonicus: Σ A. Il. 2.435b Ariston., μηκέτι νῦν δήθ’ <αὖθι λεγώμεθα>: ὅτι Ζηνόδοτος γράφει ‘<***> ταῦτα λεγώμεθα.’ ὁ δὲ Ὅμηρος εἴωθε λέγειν, ὅταν πολλὰ προειρημένα ᾖ, ὡς ἐπ’ Αἰνείου καὶ Ἀχιλλέως: ‘ἀλλ’ ἄγε μηκέτι ταῦτα λεγώμεθα’ (Il. 20.244). ‘Let us now not remain gathered here anymore’: because Zenodotus writes ‘Let us <***> talk about these things.’ But Homer is accustomed to say [this] when many things have been said already, as happens with Aeneas and Achilles: ‘but come on, let’s not talk about these things’ (Il. 20.244). Aristonicus preserves Aristarchus’ reaction to Zenodotus’ text, which read λεγώμεθα as meaning ‘to speak’ (with ταῦτα as object); for Aristarchus, however, this meaning is unsuitable here because, after a sacrifice and a banquet, Nestor invites Agamemnon not to waste time, but rather summon the army: heroes are not ‘speaking’, but rather are ‘assembled’ in one place, while they should be moving (hence, μηκέτι νῦν δήθ’ αὖθι λεγώμεθα, ‘let us now not remain gathered here anymore’).35 To support his point and mark the difference with other usages of λέγω in the sense of ‘speaking’, Aristarchus recalls Il. 20.244, where λεγώμεθα does mean ‘to talk’ because the two heroes (Achilles and Aeneas) have indeed been talking for a while, and Aeneas concludes: ‘but come on, let’s not talk about these things (ταῦτα λεγώμεθα)’. This polemical note in Aristonicus can be matched with the paraphrase reported by Didymus in which Aristarchus translates λεγώμεθα with συναθροιζώμεθα. In this case, then, Aristonicus’ note combines with Didymus’ and gives a more complete picture. On the one hand, Aristonicus omits the paraphrases in his rewording of Aristarchus’ commentaries. But, unlike Didymus, he preserves the polemical tone of the original note as well as the parallel with Il. 20.244. Of course, we cannot prove that the latter was originally present in Aristarchus’ hypomnēmata36—yet it does match Aristarchus’ principle of ‘clarifying Homer from Homer’. A little more problematic is the last of these comparisons. First we need to look closely at the passage Aristarchus is commenting upon, Il. 2.123–30: εἴ περ γάρ κ’ ἐθέλοιμεν Ἀχαιοί τε Τρῶές τε ὅρκια πιστὰ ταμόντες ἀριθμηθήμεναι ἄμφω, Τρῶας μὲν λέξασθαι ἐφέστιοι ὅσσοι ἔασιν, ἡμεῖς δ’ ἐς δεκάδας διακοσμηθεῖμεν Ἀχαιοί, Τρώων δ’ ἄνδρα ἕκαστοι ἑλοίμεθα οἰνοχοεύειν, πολλαί κεν δεκάδες δευοίατο οἰνοχόοιο. τόσσον ἐγώ φημι πλέας ἔμμεναι υἷας Ἀχαιῶν Τρώων, οἳ ναίουσι κατὰ πτόλιν … for if we, Achaeans and Trojans, wanted to take a solemn oath and to count ourselves (ἀριθμηθήμεναι), both of us, and to gather (λέξασθαι) the Trojans, those who are at home (ἐφέστιοι ὅσσοι ἔασιν), and if we Achaeans would divide ourselves in groups of ten and each of our groups would choose a Trojan man to pour us wine, then many companies of ten would lack a cupbearer. To this extent I think that the sons of the Achaeans are more numerous than the Trojans who live in the city. These are the two notes by Didymus and Aristonicus:37 Σ A Il. 2.125a Did., Τρῶας μὲν λέξασθαι <ἐφέστιοι ὅσσοι ἔασιν>: Ἀριστάρχου λέξεις ἐκ τῶν ὑπομνημάτων· ‘ἐὰν καὶ τὰ παιδία καὶ τοὺς πρεσβύτας ἀριθμήσωσιν, οὐ μὴ γένωνται ὅσοι ἡμεῖς. τοῦτο δέ ἐστι, κατοικίδιοι καὶ ἔναυλοι. καὶ ἐν ᾿Οδυσσείᾳ (6.265)· “πᾶσιν γὰρ ἐπίστιόν ἐστιν ἑκάστῳ”, ὡσεὶ κατάλυμα παρὰ τῇ νηΐ.’ ‘To gather the Trojans, those who are at home’: Aristarchus’ words from his commentaries: ‘if they counted both the children and the elders, they would not be as many as we are. And that [word, i.e. ἐφέστιοι] means “those who stay at home” and “those who live inside the court”. And in the Odyssey (6.265) [there is]: “for all [the ships], one for each man, there is a shed (ἐπίστιον)”, that is, a lodging near the ship.’ Σ A Il. 2.125a Ariston., ἡ διπλῆ δέ, ὅτι τὸ ‘λέξασθαι’ νῦν οὔτε ἐπὶ τοῦ κοιμηθῆναι οὔτε ἐπὶ τοῦ καταριθμηθῆναι κεῖται, ἀλλ’ ἐπὶ τοῦ συλλεγῆναι καὶ ἁθροισθῆναι ἐν ταὐτῷ: καὶ ὅτι ‘ἐφέστιοι’, ὅσοι ἑστίας, τουτέστιν οἰκίας, αὐτόθι †διαμένουσιν. The diplē [is placed here] because λέξασθαι does not mean here ‘to lie down’ or ‘to be counted’, but ‘to be gathered together’ and ‘to be assembled’ in the same place. And because ἐφέστιοι means those who live there having ἑστίας, that is ‘homes’.38 Aristonicus does not discuss the general sense of the passage but rather focuses on two words, λέξασθαι and ἐφέστιοι, both in v. 125. As for ἐφέστιοι, in both Didymus and Aristonicus the word is etymologically analyzed by linking it with οἰκία. Yet Aristonicus has omitted the Homeric parallel in the Odyssey (unless of course Didymus’ quotation of Aristarchus in Σ Il. 2.125a ends before it),39 and used slightly different words to explain the term. Regarding λέξασθαι, Aristonicus’ analysis seems to contradict what we read in Didymus, where Aristarchus uses ἀριθμήσωσιν—which is among the meanings excluded for λέξασθαι, according to Aristonicus (οὔτε ἐπὶ τοῦ καταριθμηθῆναι). This apparent contradiction might be solved if we think that Aristonicus and Didymus report two different points made by Aristarchus. Didymus gives Aristarchus’ summary of vv. 123–28, where the general sense is ‘counting the Trojans’. On the other hand, Aristonicus reports Aristarchus’ observation that λέξασθαι at v. 125 does not mean ‘counting’ but rather ‘gathering’. This interpretation is probably due to the fact that the verb is strictly connected to ἐφέστιοι, which, according to both Didymus and Aristonicus, indicates ‘those at home’. In other words, the idea of ‘counting’ is in the sentence (summarized in the direct quotation by Didymus) but not in that specific form, as in this case λέξασθαι means ‘gathering’ (as Aristonicus reports) because its objects are the people who are at home (ἐφέστιοι), which includes old men and children (τὰ παιδία καὶ τοὺς πρεσβύτας in Aristarchus’ summary according to Didymus). ‘Gathering’ (λέξασθαι) the people at home precedes the idea of counting them, which is however assumed by vv. 126–27 in the rest of the protasis (ἡμεῖς δ’ ἐς δεκάδας διακοσμηθεῖμεν Ἀχαιοί, | Τρώων δ’ ἄνδρα ἕκαστοι ἑλοίμεθα οἰνοχοεύειν, rendered by the verb ἀριθμήσωσιν in Aristarchus’ summary transmitted by Didymus). In this way we can harmonize the Aristarchean summary directly quoted by Didymus with Aristonicus’ note.40 Yet the differences between what Didymus and Aristonicus have each selected to report are considerable, and we should reflect on the type and extent of the selection from the original Aristarchean commentary in the scholia derived from these two scholars. Lastly, we can compare a direct quotation by Herodian with a note by Aristonicus on Il. 23.523: Σ A Il. 23.523b Hrd., {καὶ ἐς} δίσκουρα {λέλειπτο}: Πτολεμαῖος (p. 60 Baege) καὶ οἱ πλείους δίσκουρα ὡς ‘λίπουρα’ (Call. fr. 200b, 2). καὶ Ἀρίσταρχος δέ φησιν· ‘ὅπερ ἄνω κατὰ διάλυσιν εἶπεν “ὅσσα δὲ δίσκου οὖρα” (Il. 23.431), τοῦτο νῦν κατὰ σύνθεσιν ἐξήνεγκεν.’ κἀγὼ δὲ συγκατατίθεμαι. ‘{He had fallen behind by} a discus’ cast (δίσκουρα)’: Ptolemy [of Ascalon] (p. 60 Baege) and the majority [read] δίσκουρα like λίπουρα (Call. fr. 200b, 2) and Aristarchus says: ‘what he said before in two words, “as far as the cast of a discus (δίσκου οὖρα)” (Il. 23.431), now he has pronounced it as a compound.’ And I agree. Σ A Il. 23.523a Ariston., δίσκουρα: σημειοῦνταί τινες, ὅτι ἃ ἄνω (sc. Il. 23.431) εἶπε ‘δίσκου οὖρα’, νῦν {δὲ} συνθέτως ‘δίσκουρα’. ‘A discus’ cast (δίσκουρα)’: some mark the line because what he called δίσκου οὖρα above (sc. Il. 23.431) is now [called] δίσκουρα with a compound. While the point is exactly the same in both texts, the interpretation which Aristonicus attributes to ‘some’ (τινές) is in fact Aristarchus’ own analysis, as proven by the direct quotation of him in Herodian, which we discussed above. Generally when we read ἔνιοι/τινές in the Aristonicus scholia we tend to think that Aristarchus reported the ideas of someone else in his commentary, which Aristonicus has reproduced. Two direct quotations, in Σ Il. 1.423–24 (ἔνιοι δὲ ποιοῦσι …) and Σ Il. 11.754a (τινὲς μέν … ποιοῦνται) in fact prove that Aristarchus used to refer to other people’s reading with ἔνιοι/τινές. In other Aristonican scholia, therefore, when readings which are criticized are attributed to ἔνιοι/τινές we can be reasonably sure that this is what Aristonicus read in Aristarchus’ hypomnēma. However, in cases like this one, when we read that ‘some place a critical sign’ (σημειοῦνταί τινες) next to a line (e.g. Σ A Il. 5.253a, 6.130–02, 6.219a, 6.472a, 7.156b, 7.303a, etc.), we could entertain at least two possibilities. The reference to ‘some’ adding a critical sign might indeed refer to other scholars, perhaps Aristarchus’ pupils who added signs in the copy of their master or whose opinions (and suggestions for certain signs) Aristarchus himself recorded (perhaps anonymously) in his commentaries—and Aristonicus simply reported what he read. However, the phrase might actually refer to Aristarchus, if we allow for a mistake on the part of some copyists who substituted the name of Aristarchus with τινές.41 4.2 Aristonicus and Technical Terminology These examples show that while generally Aristonicus seems to preserve the content (but not the wording) of the original Aristarchean note, sometimes this might not be the case. For example, Aristonicus seems to have left out paraphrases or other interesting details (e.g. Homeric parallels), while still preserving the general meaning of the original note.42 The problem becomes more acute when we want to analyze Aristarchus’ technical terminology. His direct quotations do not preserve a lot of this; for example, aside from μετοχή for ‘participle’ and παρατατικός for ‘imperfect’ in Herodian’s direct quotation (Σ Il. 24.8a), the rest of Aristarchus’ ‘own words’ do not preserve much grammatical terminology at all. So, to dig out Aristarchus’ technical terms one needs to be very careful and proceed with painstaking comparison to test whether the terminology present in the Aristonican scholia can indeed be thought to derive directly from Aristarchus through the analysis of other sources that unambiguously go back to that time.43 In fact, there is at least one case where we can demonstrate that Aristonicus did not always preserve Aristarchus’ own words but ‘updated’ them. It concerns the term for ‘adverb’, for which Aristonicus uses two different terms: μεσότης (Σ Il. 23.287a) and ἐπίρρημα (Σ Il. 20.13). We know that these terms were used in different periods. ‘Mean’ (μεσότης) was used by the Stoics and was first applied to ‘adverb’ by Antipater of Tarsus, pupil of Diogenes of Babylon (c.150 bce), as Diogenes Laertius (7.57) informs us. The word ἐπίρρημα, on the other hand, is later: Tryphon (second half of the first century bce) is the first secure testimony of its use for ‘adverb’, as he wrote a work Περὶ ἐπιρρημάτων, On Adverbs (Ap. Dysc. Adv. GG 2.1.1, 146.15–16 = fr. 65 Velsen). Of course, the term may have been in circulation before Tryphon, as otherwise the title of his treatise might have been a source of confusion to his readers; yet we do not have any secure evidence of its use before the first century bce.44 These sources therefore allow us to conclude that Aristarchus most likely used μεσότης; when transcribing Aristarchus’ comments, however, Aristonicus once kept the Aristarchean term and once substituted it with the ‘new’ term.45 5. CONCLUSIONS Where do all these detailed analyses and comparisons leave us? First of all, while the direct quotations of Didymus and Herodian are definitely precious and give us some clues about the content of Aristarchus’ commentaries and his ‘very words’, sometimes there are doubts concerning the extent of the actual quotation, as we have seen with the cases of Σ Il. 1.423–24, Σ Il. 2.125a, Σ Il. 2.397b, and Σ Il. 2.435a1. So we need to be careful—unless the quotation is marked with some ‘closing’ remarks (as happens in Σ Il. 11.754a and Σ Il. 23.523b), or it is obviously all syntactically connected until the end of the scholium (as happens in Σ Il. 2.420a1, or Σ Il. 3.406a1 and Σ Il. 20.471a1), or it is followed by a new sentence that clearly cannot go back to Aristarchus (as in Σ Il. 2.111b and Σ Il. 24.8a). Once we have isolated what most likely goes back to the original words in Aristarchus’ commentaries, we can compare these quotations in Didymus or Herodian with the ‘content’ of the scholia from Aristonicus. This comparison has allowed us to conclude that generally Aristonicus seems to have reported correctly many of Aristarchus’ notes in terms of content. Yet Aristonicus may have often eliminated the paraphrases, which were apparently used by Aristarchus not only to clarify but also to analyze the Homeric text. Perhaps Aristonicus decided to shorten the original Aristarchean hypomnēmata by cutting the paraphrases and give instead a summary of the main points to make his text more concise for his readers, who might not have been interested in the specific details of a ‘university-like’ lecture where the professor (i.e. Aristarchus) had to go through the nitty-gritty details of Homeric diction by paraphrasing word for word. Aristonicus might have also omitted other relevant information, as for example in Σ. Il. 2.125a, where in discussing ἐφέστιοι he might have omitted the Homeric parallel of Od. 6.265 (if the latter still belongs to the verbatim quotation of Aristarchus by Didymus). Yet, all in all, we might trust Aristonicus content-wise—as long as we are also aware that his scholia might be missing many additional points. Problems, on the other hand, arise when it comes to technical terminology. The comparison with Didymus’ direct quotations suggest that Aristonicus might not have reported verbatim Aristarchus’ words in the hypomnēmata, but might have slightly rephrased and modernized them, as in the ‘adverb’ case. Nevertheless, keywords are quite important for interpreting and analyzing the scholia. One strategy (the one I adopted, at least46) is to consider them ‘verbal keys’ to map out Aristarchus’ activity as expressed by the scholia, while at the same time being very careful in attributing any technical word to Aristarchus, unless we know the term was already used by Aristarchus’ time. Out of this survey, we might also want to regard Didymus and Aristonicus as having been more independent than is usually thought to have been the case. While they did a rather good job at ‘saving’ Aristarchus’ commentaries, both Aristonicus and Didymus were scholars in their own right who wrote their own treatises; so even when ‘recopying’ the master’s commentary, they changed, cut, and selected the material they were working on.47 And even in the cases where they did preserve the actual words of Aristarchus, the same phrasing might have been changed by the several copyists and scribes who transmitted the note down to the codex Venetus A. A particularly tricky case is the substitution in Aristonicus of σημειοῦνταί τινες for what was likely to be Aristarchus’ own ideas as reported by Herodian (Σ Il. 23.523a and b). We do not know when this change happened or by whom it was made; yet this example once more shows that even if the ‘four men’ and later scribes did a remarkable job in preserving past scholarship, and even though the scholia derived from Aristonicus are definitely a good source for Aristarchus’ ideas, we still need to be cautious, especially when it comes to the precise wording. Aristonicus (or later scribes) could also have left out some parts of the discussion, which, unless we have other sources for the same fragment, are lost forever. In the end we have no choice but to trust Aristonicus—with a grain of salt. Certainly his are not Aristarchus’ own words. Footnotes * I would like to thank the editors of this volume for inviting me to contribute to it, even though I was not able to attend the conference in Oxford in summer 2018; their comments on previous drafts have been very helpful. Monica Negri also made suggestions in connection with a couple of problematic scholia. 1 On this tradition, see Lehrs 1882: 1–35; Erbse 1969–88: i xlv–lix; Matthaios 1999: 38–43; Schmidt 2002: 165–70; Schironi 2004: 11–25; Schironi 2018: 6–14. 2 A list of the λέξεις Ἀριστάρχου was gathered by Cobet 1876: 390–93, who listed Σ Il. 1.97, 1.423, 2.125, 2.420, 2.435, 3.406, 20.470, 24.8, 24.235. I will discuss them all (and some more) with the exception of Σ Il. 24.235a Hrd., which hardly contains a direct quotation or, if there is one, it is very difficult to isolate. I have also excluded Σ Il. 1.97–99 Did., which reports Aristarchus’ words from the Against Comanus; since the quotation is reported as an indirect speech, the wording might be manipulated. I will also not discuss P.Amherst 2.12, a papyrus fragment containing Aristarchus’ hypomnēma to Herodotus book 1 with end-title (third century ce). Even if this is the only piece by Aristarchus which has reached us through ‘direct tradition’, it is scanty and probably excerpted, and its text is badly damaged; see Grenfell and Hunt 1901: 3–4; Paap 1948: 37–40; S. R. West 2011: 77–80; Montana 2012. 3 I will use italics to mark words that might or might not be part of the direct quotation. If italicized words are also in bold, then I believe that they belong to the direct quotation. The text of the Iliad is that of Allen’s OCT, and the scholia to the Iliad are cited from Erbse 1969–88. 4 According to Lehrs (Erbse, app. ad loc.), Aristarchus’ quotation ends with συνηθροισμένοι μένωμεν. It is Didymus who, after recording Aristarchus’ reading (introduced by οὕτως) and quoting him directly, mentions Zenodotus’ reading (introduced by δέ) and finally Callistratus’ (again introduced by δέ). While this is syntactically possible, it is also true that Aristarchus certainly took Zenodotus’ readings into account (see Σ Il. 2.435b, below, § 4.1), so we might want to consider it part of the direct quotation. 5 See below, § 4.1, for the Homeric text of these lines (here paraphrased). 6 As Lehrs (cf. Friedländer 1853: 61) and Erbse suggested, Aristarchus probably read Τρῶες and not Τρῶας at Il. 2.125 (i.e. Τρῶες [ἐθέλοιεν] μὲν λέξασθαι …). It is then interesting to note that P.Tebt. 1.4 reads Τρῶες not Τρῶας at v. 125 (as my direct check confirmed); this might be a further proof of the Aristarchean pedigree of this papyrus; see Schironi 2020: 264–65. 7 A scholium to the Odyssey suggests that Aristarchus did analyze ἐπίστιον and gave a similar etymology to it (i.e. he linked it with οἰκία): Σ Od. 6.265d1 Pontani: ἐπίστιόν ἐστιν ἑκάστῳ] ἐποίκιόν ἐστιν ἑκάστῳ ἡ ναῦς· HMaTy διὸ τὸ σημεῖον. HTy. This is why I consider the parallel with the Odyssey in Σ Il. 2.125a to be part of the direct quotation from Aristarchus. See, however, discussion below, at § 4.1. 8 See the same procedure in Σ Il. 2.125a (above). 9 Cf. DELG, s.vv. ἀμύμων and μῶμος. 10 For Erbse, the direct quotation also includes the note about the readings in other editions (οὕτως δὲ εὕρομεν καὶ ἐν τῇ Μασσαλιωτικῇ καὶ Σινωπικῇ καὶ Κυπρίᾳ καὶ Ἀντιμαχείῳ καὶ Ἀριστοφανείῳ); so also La Roche 1866: 100; Ludwich 1884–85: i 194–96, and Nagy 2004: 89–91. On the contrary, West 2001b: 70–72, following Bekker and Bergk, excludes from the quotation the mention of the editions—rightly, since the phrasing οὕτως δὲ εὕρομεν is typical of Didymus (see Schironi 2018: 67–68). In fact, in this scholium this sentence in strictly connected with the following one quoting Callistratus, Dionysius Sidonius, and Demetrius Ixion, which clearly is Didymus’ addition. Furthermore, in none of the cases where Didymus quotes the so-called ‘city edition’ in the scholia can it be demonstrated that the mention of these editions is taken from Aristarchus; on his use of manuscript evidence, see Schironi 2018: 63–75. 11 As Adrian Kelly rightly suggests to me, ὃν οὐκ ἂν ἐκκλίνειεν makes sense (with Zeus as subject), without resorting to more complicated emendations, such as the one suggested by Erbse, who prints ὃν οὐκ ἂν †ἐκκλίνειεν† in the text and suggests ὃν οὐκ (ἄν) τις ἐκκλίνειεν in the critical apparatus. 12 In addition, in book 2 Zeus had also sent a dream to Agamemnon to convince him to attack (cf. Il. 2.35–40). Zeus’s careful planning on that occasion is a direct consequence of his promise to Thetis; hence, changing his behaviour now would be ridiculous; cf. Erbse, ad Σ.Il. 2.420a1. 13 The subject of γράφει is certainly Aristarchus, as is clear from the omitted part of the scholium. 14 Schironi 2015: 612–15 (Aristarchus probably changed his mind about the correct reading, choosing Ζεύς με μέγας as a second thought). Cf. also Lehrs 1882: 17–18, 27–28; Ludwich 1884–85: i 66, 204–06; Erbse 1959: 282–84; Montanari 1988: 98–100; West 2001: 55–56, 175; Nagy 2009: 23–24. 15 See Matthaios 1999: 233–35. 16 Cf. Erbse 1959: 281, with a slightly different interpretation (connected with his theory about the nature of Aristarchus’ hypomnēmata; see Schironi 2018: 39). 17 Cf. Schironi 2018: 203–05. 18 In other words, ‘Some change it into γένηται, not correctly’ is not part of the verbatim quotation, because of the scholium’s wording (i.e. the closing note ‘Didymus [writes] this’, which seems to suggest some comment by Didymus just before it; see the similar case of Σ Il. 2.111b and Σ Il. 2.435a1 above). The comparison with Aristonicus in § 4.1, however, will show that a note on this reading by ‘some’ scholars was indeed most likely present in Aristarchus’ commentary. 19 See Schironi 2004: 297–304 (fr. 36); Schironi 2018: 368–70. 20 On σύνθεσις and διάλυσις, see Schironi 2018: 359–60. 21 The final sentence in the scholion sounds like it is derived from Herodian, who usually discusses conjugational patterns and problems of accentuation. So even if the end of the quotation is not clearly marked, there is little doubt that the final sentence should be excluded from Aristarchus’ verbatim quotation; cf. also Erbse 1980: 238. 22 Cf. Matthaios 1999: 28–30, against Siebenborn 1976: 56–84, who claimed that Aristarchus mostly used two-term proportions. On Aristarchus’ analogical proportions, see also Schironi 2018: 377–412. 23 On this scholium, see Erbse 1980: 237–39; Matthaios 1999: 330, 343, 344–45, 409–10, 411, 421; Schironi 2018: 402; Schironi 2019: 491. 24 I have reported in italics items that belong to passages which might not be part of the quotation (but I think they are—hence they are also written in bold in the scholia analyzed above). I have not reported in this list items that can be found in the parts in italics only, because, as I discussed above, I do not think they belong to the direct quotation. 25 As already noted by Lehrs 1882: 153. 26 See Henrichs 1971: 99–116; Montanari 1995b: 79–82. 27 The certainly excerpted fragments from P.Amherst 1.12 seem to fit within this picture. Even if most of it is lost, Aristarchus’ commentary to Herodotus seems to have been a literary commentary where parallels with Homer, Sophocles, and Herodotus himself were more important than the historical data discussed therein. 28 On anonymous citations in scholia, see Nünlist 2009: 12 and Schironi 2018: 16–17, 72–73. 29 See below, § 4.2, for an example. 30 So Lehrs 1882: 13; Schmidt 1976: 9 and 2002: 169; Lührs 1992: 4–5; Matthaios 1999: 37, 43–45; West 2001b: 46; Janko (private communication); Nagy 2009: 35–36; Bouchard 2016: 25; Schironi 2018: 14–25. 31 Schironi 2018. 32 As demonstrated by Σ A Il. 2.135a Ariston., <καὶ δὴ δοῦρα σέσηπε νεῶν καὶ σπάρτα λέλυνται:> ὅτι κατὰ τὸν αὐτὸν στίχον καὶ ἑαυτῷ καὶ ἡμῖν συνήθως ἐξενήνοχε τὸ ‘λέλυνται’ καὶ ‘σέσηπε’ [‘And the timbers of our ships are rotten (δοῦρα σέσηπε), and the ropes are loose (σπάρτα λέλυνται)’: because in the same line he has used both λέλυνται and σέσηπε, according to his own usage and our usage]. 33 See above, n. 18. 34 See above, n. 4. 35 Critics have been divided about whether it is better to follow Aristarchus or Zenodotus on Iliad 2.435; West 2001: 177 gives preference to Aristarchus’ reading; Leaf 1900–2: i 81–82 and Kirk 1985: 161 instead prefer Zenodotus’ solution. 36 No Aristarchean scholia are preserved at Il. 20.244; see, however, Erbse ad loc. 37 Erbse groups these together as Σ Il. 2.125a. I print them separately for clarity. 38 For this translation I follow Cobet’s suggestion to read: ὅσοι ἑστίας, τουτέστιν οἰκίας, αὐτόθι ἔχοντες διαμένουσιν (see Erbse, app. ad loc.). 39 See, however, above, n. 7. 40 Aristonicus also tells us that in Il. 2.123 (εἴ περ γάρ κ’ ἐθέλοιμεν) the κε was superfluous because koinē Greek does not use ἄν in a future protasis but εἰ followed by simple optative. Didymus’ summary (ἐὰν καὶ τὰ παιδία καὶ τοὺς πρεσβύτας ἀριθμήσωσιν) is simply another way of rendering a future conditional, with ἐάν and subjunctive, but does not contradict Aristonicus’ note. In addition, Aristonicus says that v. 124 was athetized (Σ A Il. 2.124a), so that the resulting text was: εἴ περ γάρ κ’ ἐθέλοιμεν Ἀχαιοί τε Τρῶές τε | Τρῶας μὲν λέξασθαι ἐφέστιοι ὅσσοι ἔασιν, | ἡμεῖς δ’ ἐς δεκάδας διακοσμηθεῖμεν Ἀχαιοί. In it, there is no verb ‘to count’ (as ἀριθμηθήμεναι was cut with the athetized v. 124). Even with this athetesis, the explanation provided above does not change: Aristarchus pointed out that even if these lines express the idea of ‘counting’, the phrase λέξασθαι ἐφέστιοι ὅσσοι ἔασιν meant only the idea of ‘gathering’. 41 According to Lehrs 1882: 9–13, most of these cases are simply the result of a hasty epitomization (and the note refers to a sign apposed by Aristarchus); in other cases, however, he too concedes (ibid., 13) that τινές might refer to Aristarchus’ pupils adding critical signs based on the lectures of Aristarchus; see also Schironi 2018: 16–17. 42 This picture is confirmed when we look beyond the scholia. For example, the similar notes in P.Oxy. 221 (second century ce) to Il. 21.290 (col. xv, ll. 6–27) and in Σ A Il. 21.290a Ariston. offer interesting similarities and differences, which I have analyzed in Schironi 2020: 258–60. 43 This is, for example, what Matthaios did to analyze Aristarchus’ grammatical terminology; see Matthaios 1999: 45–46. 44 Ἐπίρρημα might be used also by Tyrannion (first century bce), even if the scholium by Herodian (Σ A Il. 9.7a = fr. 19 Haas) is ambiguous (Τυραννίων δὲ ἓν μέρος λόγου ἤκουσεν, ἵν' ᾖ ἐπίρρημα, καὶ βαρύνει), since the term could have been added by Herodian. Another more speculative piece of evidence is Σ A Il. 18.410d, where Herodian quotes Philoxenus, who in the first century bce seems to have used μεσότης; see Matthaios 1999: 559, n. 199. 45 Cf. Matthaios 1999: 520–23 and 559–60. 46 See Schironi 2018: 27–29. 47 In fact, Didymus even disagrees with Aristarchus concerning some of his readings; see Schironi 2018: 22. © The Author(s) 2021. Published by Oxford University Press on behalf of Institute of Classical Studies. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Aristarchus in his own words? What his ‘most secure’ fragments can tell us about Aristarchus’ commentaries and their transmission*
JF - Bulletin of the Institute of Classical Studies
DO - 10.1093/bics/qbab001
DA - 2021-08-12
UR - https://www.deepdyve.com/lp/oxford-university-press/aristarchus-in-his-own-words-what-his-most-secure-fragments-can-tell-ddOU4l7hjH
SP - 17
EP - 34
VL - 64
IS - 1
DP - DeepDyve
ER -