Aelius Dionysius
Ἀττικὰ ὀνόματα

D. Sources

In his 1890 edition of the fragments of Aelius Dionysius and Pausanias, Schwabe devoted a substantial part of his prolegomena to a discussion of the sources of the two lexicographers (cf. Schwabe 1890, 38–76), identifying four main groups: the entries on Realien, Attic customs, or specific semantic fields (such as the human ages), which ultimately go back to Aristophanes of ByzantiumAristophanes of Byzantium and DidymusDidymus; the entries on grammar and orthography, which show parallels with Trypho and Herodian; the entries on proverbs, and those which appear to be original to Aelius and Pausanias. Given the important role of comedy in Aelius Dionysius’ canon (see below), Alexandrian scholarship on the comic playwrights must have been a key source for his lexicon, with Didymus’ Comic Vocabulary (cf. Benuzzi 2020, 54–5) arguably playing a central role in the chain of intermediate sources linking Aelius to Hellenistic scholars (see also below). An example of material from Didymus’ Comic Vocabulary ending up in Aelius’ lexicon is provided by Eust. in Od. 2.188.39–40 ~ Ael.Dion. α 8Ael.Dion. α 8:

οὕτω κεῖται, ὡς ὁ αὐτὸς Διονύσιος λέγει, καὶ παρὰ Θεοπόμπῳ ἀβυρτάκη ὑπότριμμά τι δριμὺ, βαρβαρικὸν διὰ πράσων καὶ καρδάμων καὶ κόκκων ῥόας καὶ ἑτέρων τοιούτων.

Thus, as the same Dionysius (i.e. Aelius Dionysius) says, the term ἀβυρτάκη occurs in Theopompus (fr. 18), [indicating] a pungent sauce, [of] barbaric [origin, made] with leeks, cress, pomegranate seeds and other ingredients of this sort.

and Gal. Ind. 6.12–4, where Galen – after stating that he has produced an epitome of Didymus’ Comic Vocabulary – describes his own treatise on the vocabulary of Old Comedy:

ἀλλ’ ἃ καὶ παρὰ τοῖς παλαιοῖς κωμικοῖς, Ἀβυδοκόμαν ἢ ἀβυρτάκην, <ἢ> ἄλλ’ ὅσα μὴ σαφῆ τοῖς ἀκούουσιν ἦν, ὥριστο κατὰ τὴν ἡμετέραν πραγματείαν, ἃ Δίδυμος ἔφθασεν ἐξηγήσασθαι καλῶς.

But also the words in the ancient comic playwrights – [such as] Ἀβυδοκόμας (i.e. a satyric nickname for sycophants) or ἀβυρτάκη – or however many were unclear to the listeners were defined in my (i.e. Galen’s) treatise, [words] that Didymus (n. 7 Coward – Prodi) had already explained well.

Through Didymus’ Comic Vocabulary, however, Aelius was able to access older scholarship on comedy, going back even further than Aristophanes of Byzantium (cf. above), namely to Eratosthenes of CyreneEratosthenes of Cyrene’s On Ancient Comedy. For example:

Eust. in Il. 3.401.14–8 (cf. Ael.Dion. κ 24Ael.Dion. κ 24, κ 25Ael.Dion. κ 25, σ 23Ael.Dion. σ 23): κῆπος γὰρ οὐ μόνον φυταλιά, ἀλλὰ καὶ καλλωπισμὸς κόμης κατὰ Αἴλιον Διονύσιον καὶ κουρᾶς διάθεσις τῶν ἐν κεφαλῇ τριχῶν. Θουκυδίδης δὲ κήπιόν φησι, Κωμικοὶ δὲ κηποκόμαν τὸν κῆπον κειρόμενον, ὥσπερ ἕτερον σκαφιόκουρον, καὶ ἔστι καὶ παρὰ τῷ Κωμικῷ ἡ λέξις τοῦ σκάφιον κείρεσθαι.

For the κῆπος (‘garden’) [is] not only an orchard, but also a hair ornament according to Aelius Dionysius and an arrangement of the haircut. Thucydides (2.62.3) says κηπίον (‘small garden’), and the comic playwrights [call] κηποκόμας the one who has his hair cut [in the way known] as κῆπος, just as [they call] another one σκαφιόκουρος (com. adesp. fr. 193). And in the comic playwright (i.e. Aristophanes) there is the expression σκάφιον κείρεσθαι (‘to have the hair cropped in a bowl cut’, Th. 838).

and schol. Eur. Tr. 1175, which preserves Eratosthenes’ interpretation of the verb κηπεύω not in the generic sense of ‘to tend to’, ‘to cherish’, but as ‘to arrange [the hair] in the hairstyle known as κῆπος’ (Eratosth. fr. 66 Strecker, cf. Benuzzi 2019, 342–6; for a similar case, see Eust. in Il. 3.514.21–5 ~ Ael.Dion. ε 55Ael.Dion. ε 55 and Eratosth. fr. 75 Strecker). In the light of these cases, one might initially assume that Aelius relied directly on Didymus’ Comic Vocabulary, but it is more plausible that the former had access to the work of Didymus and other late Hellenistic scholars through the 2nd-century CE grammarian DiogenianusDiogenianus, who produced a lexicon based on Iulius Vestinus’ epitome of the massive 95-book lexicon compiled in the 1st century CE by Zopyrion and Pamphilus (see Ippolito 2005, Meliadò 2019, Montana 2003, Benuzzi 2020, 54). This is suggested by the fact that many entries in which Aelius evidently reuses Hellenistic scholarly material have parallels in Hesychius, who in turn explicitly names Diogenianus among his sources in the prefatory letter of his lexicon (see e.g. Hsch. κ 2529 on κῆπος as a type of haircut; cf. e.g. Schwabe 1890, 79: Erbse 1950, 68).

E. Canon

Since Aelius Dionysius was active during Hadrian’s reign (see above), he predates Phrynichus and Pollux (who are usually dated to the second half of the 2nd century CE, see the entries Phrynichus Atticista, Ἐκλογὴ Ἀττικῶν ῥημάτων καὶ ὀνομάτων (Ecloga), forthcoming and Iulius Pollux, Ὀνομαστικόν (Onomasticon)) by at least half a century. This chronological distance is reflected in the different canons of authors adopted by Aelius and by his successors.

Of the 195 fragments definitely attributed to Aelius Dionysius by the sources, 28 contain references to comic playwrights mentioned by name. The range of authors cited includes not only the 5th-century Attic poets considered authoritative by the strict Atticist Phrynichus − Cratinus (4x), Pherecrates (5x), Eupolis (1x), Plato (1x), Aristophanes (6x by name, at least 4x as ὁ κωμικός) − but also, in limited numbers, the playwrights excluded (and often openly criticised) by Phrynichus, and conversely accepted by the Antiatticist and Pollux, such as Theopompus (1x), Alexis (1x), Philemon (1x), and Menander (8x). Generic references to οἱ κωμικοί are also attested (5x). For the number of comic quotations in the text of Aelius Dionysius conjecturally reconstructed by Erbse (1950), see Sonnino (2014, 191−2). Overall, as observed by Sonnino (2014, 171), Aelius Dionysius – at least as far as the extant fragments of his lexicon allow us to reconstruct – shows a canon of comic playwrights that is more inclusive than that of Phrynichus (who practically excludes the poets of Middle and New Comedy), but more restrictive than that of Pollux and the Antiatticist, in that the quotations from these playwrights are rather few (whereas Pollux and the Antiatticist quote them very frequently; see entries Iulius Pollux, Ὀνομαστικόν (Onomasticon) and Antiatticista, forthcoming).

The same trend can be observed with regard to prose writers: Aelius refers not only to authors accepted by Phrynichus as models of good Attic − such as Thucydides (9x) and the philosopher Plato (3x) − but also to some of those accepted by the Antiatticist and Pollux, and rejected by Phrynichus, namely Herodotus (4x) and Xenophon (1x). In this case, too, the distribution of the quotations is revealing: although Aelius also mentions Herodotus and Xenophon, he quotes Thucydides and Plato more often (in contrast, in Pollux the references to e.g. Thucydides and Herodotus are almost equal in number [87x vs. 79x, TLG count], and in the Antiatticist the former is quoted even less often than the latter [44x vs. 64x, TLG count]). In keeping with the general caution of the Atticists towards tragic language, tragedy is much less represented in Aelius’ fragments than comedy and prose: there are only two certain quotations from Sophocles (Eust. in Od. 1.253.10−6 ~ Ael.Dion. ε 48Ael.Dion. ε 48; schol. (Ap. H. ?) Hom. Il. 15.705b.4−9 (A) ~ Ael.Dion. φ 13Ael.Dion. φ 13; for a third possible quotation, see Eust. in Od. 1.32.30−3 ~ Ael.Dion. α 18Ael.Dion. α 18, α 20Ael.Dion. α 20) and two likely ones from Aeschylus (see Ael.Dion. δ 26 above) and Euripides (Eust. in Il. 3.552.28−31 ~ Ael.Dion. σ 1Ael.Dion. σ 1) respectively.

In three cases, Aelius Dionysius mentions the absence of attestations as a reason for rejecting a form:

Phot. Bibl. cod. 279.532a.22−5: ὅτι τὸ ἀρχέτυπον, φασίν, οὐδενὶ τῶν ἀρχαίων συγγραφέων οὐδὲ ῥητόρων οὐδὲ τῶν Μούσαις κατόχων εἴρηται, ὡς Διονύσιος ὁ Ἁλικαρνασσεὺς ἱστορεῖ.

They say that the word ἀρχέτυπον (‘model’) is not used by any of the prose writers, the orators or any of those inspired by the Muses, as Dionysius of Halicarnassus (α 184) attests.

Eust. in Il. 4.448.3−5: Αἴλιος δὲ Διονύσιος ἐν τῷ ῥητορικῷ αὐτοῦ Λεξικῷ οὐκ ἐθέλει ὠνομάζεσθαι γράστιν ὅλως. φησὶ γάρ· κράστιν Ἀττικοὶ τὴν πόαν, γράστιν δὲ οὐδένες.

Aelius Dionysius in his rhetorical lexicon (κ 36Ael.Dion. κ 36) completely forbids saying γράστις. For he says: ‘Attic-speakers call the grass κράστις, nobody [calls it] γράστις’. (See the entry κράστις, γράστις).

Eust. in Od. 1.9.30−1: φησὶ γοῦν Αἴλιος Διονύσιος, ὅτι ἀγαθώτερος καὶ ἀγαθώτατος παρ’ οὐδενὶ τῶν Ἑλλήνων κεῖται.

Aelius Dionysius (α 10Ael.Dion. α 10) says that ἀγαθώτερος and ἀγαθώτατος occur in none of the Greek writers. (See the entry ἀγαθώτερος, ἀγαθώτατος).

This kind of phrasing is also found in other Atticist lexica, see e.g. Phryn. Ecl. 197Phryn. Ecl. 197 (βασίλισσα οὐδεὶς τῶν ἀρχαίων εἶπεν, ἀλλὰ βασίλεια ἢ βασιλίς, ‘none of the ancients said βασίλισσα (‘queen’), but [they said] βασίλεια or βασιλίς’, see the entry βασίλεια, βασιλίς, βασίλισσα, βασίλιννα) and Poll. 10.60Poll. 10.60 (εἰ δὲ καὶ τὸ ἀναλογεῖον ἐθέλοις προσονομάζειν, οὕτω μὲν ἐπὶ τοῦ τοῖς βιβλίοις ὑποκεισομένου παρ’ οὐδενὶ τῶν κεκριμένων εὗρον, ‘if you want to use the word ἀναλογεῖον (‘bookrest’), I did not find it [used to indicate] the object that is put under the books in any of the selected [authors]’).

F. Evaluative terminology

The extant fragments of the Ἀττικὰ ὀνόματα display much of the evaluative terminology found in later Atticist lexica.

In two instances (An.Boiss. 1.411 ~ Ael.Dion. φ 20Ael.Dion. φ 20, Eust. in Od. 1.151.38 ~ Ael.Dion. ρ 6Ael.Dion. ρ 6) Aelius uses the term ἰδιῶταιοἱ ἰδιῶται to indicate common speakers (in one case the everyday usage is defined by the adjective κοινόςκοινός, see Eust. in Il. 3.496.13 ~ Ael.Dion. σ 4Ael.Dion. σ 4: ὁ κοινὸς λεγόμενος σάκκος, ‘the commonly used [word] σάκκος (‘sack’)’).

Words and usages that are considered incorrect are labelled by terms such as βάρβαροςβάρβαρος, βαρβαρισμόςBarbarism, and σολοικισμόςSolecism. For instance, Aelius deems ‘barbarous’ the Aristophanic hapax ἠπάομαι (Ar. fr. 239), see Eust. in Od. 1.367.24−6 φησὶ γοῦν Αἴλιος Διονύσιος (α 64Ael.Dion. α 64) ἀκέστριαν οὐκ ἠπήτριαν […] τὸ δὲ ἠπήσασθαι παντελῶς φησι βάρβαρον (‘Aelius Dionysius says [the correct form is] ἀκέστρια (‘seamstress’), not ἠπήτρια […] and he says that ἠπήσασθαι (‘to mend’) is completely barbarous’. The noun and the verb are also rejected by Phrynichus, Ecl. 64Phryn. Ecl. 64, and by the Atticist Philemon, [Vindob.] 392.6Philemo (Vindob.) 392.6, 394.26Philemo (Vindob.) 394.26. See also Moer. α 101Moer. α 101). Another interesting example is Eust. in Od. 1.151.39−40 ῥὼξ γὰρ εἰπεῖν, βαρβαρισμὸς καὶ σολοικισμός ἐστιν ὥς φησιν Αἴλιος Διονύσιος (‘to say ῥώξ is a barbarism and a solecism, as Aelius Dionysius [ρ 6Ael.Dion. ρ 6] says’; see entry ῥάξ, ῥώξ). Here the presence of σολοικισμός (‘syntactical error’) in Eustathius’ text justifies Erbse’s reconstruction (1950, 139) that the usage criticised by Aelius was not simply ῥώξ instead of ῥάξ, but ὁ ῥώξ instead of ἡ ῥάξ: indeed, ὁ ῥώξ is a barbarism (i.e., a phonematic error) in that the α is replaced by ω, but also a solecism (i.e., a syntactical error) because the masculine article is used with a feminine noun (see Sandri 2020, 23).

In another entry, Aelius labelled the late and scantily attested ἱατρίνη (‘midwife’, found – before and during the lexicographer’s time − only once in Iosephus and twice in Galen, and almost certainly a medical term) as οὐχ’ ἙλληνικόςἙλληνικός, that is ‘non-standard Greek’, see Eust. in Il. 3.244.15−7: ἰστέον δὲ […] ὅτι οὐ μόνον κατὰ γένος ἀρρενικὸν ἰατρός, ἀλλὰ καὶ θηλυκῶς. φησὶ γοῦν Αἴλιος Διονύσιος ἰατρὸν γυναῖκα, Ἄλεξις δὲ ἰάτριαν. ἡ δὲ ἰατρίνη οὐχ’ Ἑλληνικόν, φησί (‘one should know […] that ἰατρός (‘physician’) [can be used] not only as a masculine, but also as a feminine. Indeed, Aelius Dionysius [ι 1Ael.Dion. ι 1] [says]: ‘the woman [is called] ἰατρός, while Alexis [calls her] ἰάτρια. ἰατρίνη, instead, is not Greek’’).

The presence of such terminology proves that Aelius’ lexicon had a clear prescriptive agenda, aimed at promoting the literary Greek of Attic authors against forms and usages deemed too common, non-Greek, or more specifically non-Attic. This idea is reinforced by fragments characterised by the structure ‘X, οὐ Υ, λέγουσιν οἱ Ἀττικοί’, such as Eust. in Od. 1.354.25 φησὶ γοῦν Αἴλιος Διονύσιος. ἄγνον ἀρσενικῶς, οὐ λύγον καλοῦσιν Ἀττικοί (‘Aelius Dionysius [α 22Ael.Dion. α 22] says: ‘Attic-speakers say ἄγνος (‘chaste-tree’), masculine, not λύγος’’). In the light of this, the many fragments that contrast the Attic and Ionic forms of the same word may imply that the grammarian considered the Attic form preferable (see e.g. Eust. in Il. 1.331.13−4 Αἴλιος δὲ Διονύσιός φησιν, ὅτι ζμικρὸν καὶ μικρὸν οἱ Ἀττικοί, μικκὸν δὲ Ἴωνες, ‘Aelius Dionysius [ζ 3Ael.Dion. ζ 3] says that Attic-speakers [say] ζμικρός and μικρός (‘small’), while Ionic-speakers [say] μικκός’).

As regards the frequent label παλαιοίοἱ παλαιοί, it seems to be used especially in cases where the tone is descriptive, rather than prescriptive (see e.g. An.Boiss. 1.413−4 περὶ τοῦ ὑπογράφειν […] λέγει Διονύσιος ὁ Ἁλικαρνασσεὺς ἐν τῷ περὶ χρήσεως Ἀττικῶν ὀνομάτων, ὅτι τὴν ὑπό εἰώθασιν οἱ παλαιοὶ πολλάκις καὶ ἀντὶ τῆς πρό ἐκτιθέναι, ‘with regard to ὑπογράφειν (‘to subscribe’) […] Aelius Dionysius, in the [work] On the Usage of Attic Terms [υ 10Ael.Dion. υ 10], says that the ancients are often used to putting the prefix ὑπό in place of πρό’, Eust. in Il. 3.459.29−30 ῥῶπος μέντοι λεπτὸς καὶ εὐτελὴς φόρτος, ὡς δὲ Αἴλιος Διονύσιος λέγει, καὶ ποικίλος, γέλγην δέ, φησίν, αὐτὸν ἔλεγον οἱ παλαιοί, ‘ῥῶπος [means] a frivolous and worthless trinket, as Aelius Dionysius says [ρ 14Ael.Dion. ρ 14], and also abstruse. ‘The ancients’, he says, ‘called this γέλγη (‘frippery’)’’).