ῥάξ, ῥώξ
(Phryn. Ecl. 51, [Hdn.] Philet. 182)

D. General commentary

Atticist lexicographers (Phrynichus, A.1; the pseudo-Herodianic Philetaerus, A.2; Aelius Dionysius, see above B.4 and B.5) deal with the grammatical gender of the noun ῥάξ (‘grape’; for other meanings, see below), prescribing its feminine form against the masculine variant ῥώξ. Both variants will be discussed below. The Atticist prescription is confirmed by later lexicographical and grammatical sources (Photius, B.4; Eustathius of Thessalonica, B.6; Thomas Magister, B.8, who read Phrynichus’ entry in the redaction of family q; see A.1, apparatus). This is one of several examples where Atticist lexicographers (as well as other erudite sources) discuss cases of gender change; see AGP vol. 2, Nominal morphology, forthcoming, and entries βάτος, βῶλος, λάγυνος, ὄμφαξ, ὕσπληξ, φάρυγξ, and χάραξ.

The feminine noun ῥᾱ́ξ – with its variant ῥώξ, attested in both feminine and masculine forms – originally meant ‘grape’. Later it was used with other meanings, namely ‘berry’, as a generic noun, ‘venomous spider’, ‘teat’, and ‘fingertips’ (in the plural). The vowel alternation between ῥᾱ́ξ and ῥώξ is difficult to explain: Höfler and Nielsen (2024, 82–4) assume that the two forms were originally two separate lexemes, i.e. ῥᾱ́ξ, ῥᾱγός (‘grape’, fem.) and ῥώξ, ῥωγός (‘spider’, masc., derived from the same root as ἀράχνη in Höfler and Nielsen’s reconstruction), which later became confused in popular etymology. In this context, Aelian’s explanation in NA 3.36 (C.8) is noteworthy: the use of ῥᾱ́ξ to mean ‘spider’ is due to the dark colour and round shape of the animal, similar to a grape berry. Be that as it may, the etymologyEtymology of ῥᾱ́ξ, ῥᾱγός (the quantity of the vowel α will no longer be indicated in the following paragraphs, unless necessary) remains unclear: as discussed in DELG, EDG s.v. and Höfler, Nielsen (2022, 83 n. 46, with additional bibliography), it has been identified as a Mediterranean word by comparison with the noun ῥάματα (cf. Hsch. ρ 95: ῥάματα· βοστρύχια. σταφυλίς. Μακεδόνες, ‘ῥάματα: [It means] ‘pressed grapes’, ‘bunch of grapes’. The Macedonians [use it]’), perhaps from *ῥάγμ-, and the Latin racēmus (‘(stalk of a cluster of) grapes’; see DELL s.v.).

As far as the spread of the two forms is concerned, ῥάξ meaning ‘grape’ is first attested in Sophocles (C.3, fem.) and Plato (C.4, where the grammatical gender however cannot be assessed): the use of the word by Attic authors thus justifies the Atticists’ prescription. ῥάξ in the feminine is then also attested in the Aristotelian corpus (cf. Col. 792b.8–9, Pr. 925b.14–5) and in Theophrastus (e.g. CP 1.21.1, 3.16.1). The feminine noun ῥώξ in the sense of ‘grape’, ‘berry’ seems to be attested as early as Archilochus (C.2) – at least according to Choeroboscus (B.2) – and subsequently in Ctesias of Cnidus (C.5), whereas the first attestation of the masculine form ῥώξ meaning ‘spider’ is in Nicander (C.6). Finally, the Septuagint provides the first occurrence of the masculine ῥώξ to denote the grape (C.7), the very use proscribed by Atticist lexicographers; see also LXX Is. 65.8. The presence of this form in the biblical text may then have favoured its spread in later literature, especially in the Christian context: see e.g. Eus. Is. 2.55.54, Ast.Soph. Hom. 17.2.7, Procl.CP Laud.Mar. 5.2.4. Atticist authors (Galen, Lucian, Aelian, and Philostratus) consistently use the feminine ῥάξ, according to the Atticist prescription (see Schmid, Atticismus vol. 4, 648).

Atticist lexicographers and later sources condemn the use of the masculine ῥώξ instead of the feminine ῥάξ, noting the presence of two errors in the former (cf. A.1, and A.2). Photius (B.4) and Eustathius (B.6; he makes explicit reference to Aelius Dionysius’ teaching) clarify that the ‘two errors’ to which Phrynichus refers are barbarism (βαρβαρισμός) and solecism (σολοικισμός). While βαρβαρισμός (cf. also βάρβαρος in B.7) alludes to the phonemic alternation ῥάξ / ῥώξ, σολοικισμός probably refers to the gender change, as a syntactic error (see Sandri 2020, 5 and 23). As is usual in his edition of Aelius Dionysius’ fragments, Erbse considers Photius (B.4) to be the more reliable source for the reconstruction of Aelius’ entry (ρ 6), although Eustathius quotes Aelius by name. In fact, Eustathius does not make any reference to the masculine gender of ῥώξ, while it is plausible that the gender change was addressed in Aelius’ discussion of ῥάξ and ῥώξ; see entry Aelius Dionysius, Ἀττικὰ ὀνόματα, F. Evaluative terminology.

Some later lexicographical and grammatical sources, which do not share the Atticists’ prescriptive aim, confine themselves to recording the use of both forms (ῥάξ / ῥώξ), either in the feminine gender (B.1) or without any gender indication (B.3), but generally presenting the form ῥώξ as more recent (B.9 merely describes ῥάξ as the Attic form, without explicitly mentioning the form with ω). The form ῥώξ is mentioned by Hesychius (B.1) as the one used by himself and his contemporaries (ἡμεῖςἡμεῖς), showing that ῥώξ was the more popular form in his day. Other sources relate the two forms ῥάξ and ῥώξ to two different semantic uses, stating that the form with ω means ‘spider’, with or without gender specification (B.2, B.5, B.7, B.8; incidentally, this distinction seems to agree with the reconstruction by Höfler and Nielsen mentioned above). Finally, Eustathius records a further meaning of the form ῥώξ, referring to a Homeric passage (C.1); on this form see F.2.

E. Byzantine and Modern Greek commentary

In the Byzantine period both forms ῥάξ and ῥώξ continued to be used, although ῥάξ has a greater number of attestations (see e.g. Georgius Pisides Hexaemeron 1586, Theophylactus Epistulae 80.4, Gp. 5.45.5.5, Theodorus Prodromus Carmina historica 45.341). Furthermore, in Medieval Greek the 1st-declension forms ῥάγα and ῥώγα became more popular (Kriaras, LME s.v. ρώγα. See also B.7, where the accusative plural ῥάγας should be traced back to the nominative singular ῥάγα and not to ῥᾱ́ξ, if the accent in Par. gr. 2072 is correct; in any case, a mistake in the accent is possible and the form could also be read as properispomenon ῥᾶγας, accusative plural from of ῥᾱ́ξ); cf. also Achmet Oneirocriticon 283.11 (καὶ ἡ ῥὼξ ἤτοι ἡ λεγομένη ῥῶγα), Euchologia 100.778.15 (ῥῶγα); in both passages, the noun is feminine and denotes the spider. These forms are the outcome of the gradual merging of the masculine and feminine nouns of the 1st and 3rd declensions, leading to the eventual demise of 3rd-declension masculine and feminine nouns in Early Medieval Greek; see CGMEMG vol. 2, 253 and 458. The 1st-declension forms are the ones still used in Modern Greek (see LKN s.vv. ράγα² and ρώγα).

F. Commentary on individual texts and occurrences

(1)    Hsch. ρ 116 (B.1)

Besides ρ 116, the noun ῥάξ also appears in another gloss by Hesychius (ρ 7)Hsch. ρ 7, mentioned in the apparatus of B.1. The witnesses where this gloss is preserved show that it originally comes from Cyril’s lexicon, although it is not marked with an asterisk according to the convention of the reference edition (cf. Latte, Cunningham 2018, XII); together with ρ 116 (B.1), it looks like a case of a ‘doublet’ (cf. Alpers in Latte, Hansen 2005, XXI–XXII). ρ 7 is difficult to interpret mainly because of the entry form, ῥάγα: it does not correspond to the accusative of ῥᾱ́ξ (ῥᾶγα), which is instead included among the interpretamenta of the same lemma and has a circumflex accent. Furthermore, the word ῥάγα occurs twice in the gloss and the second occurrence was expunged by the editors. The earliest interpretamenta provided (ἀκμή. βία, ὁρμή, ‘highest point; strength, impulse’) seem to rule out that ῥάγα is an attestation of the homophonous 1st-declension form popular in Medieval Greek (see E.); instead, they bear out the correction proposed by Schmidt (1858–1868 vol. 3, 417), who suggested reading the oxytone ῥαγά by comparison with a fragment attributed to Erotianus (fr. *31: τί ἔστι ῥαγή· ἀκμή, ὁρμή, βία, ‘What is ῥαγή? [It means] ‘highest point’, ‘impulse’, ‘strength’’). Following the above correction, it is likely that the lemma of this entry was originally a noun unrelated to ῥάξ and was later interpreted as an accusative form of the latter, which led to the addition of the second part of the gloss (καὶ ἣν ἡμεῖς ῥῶγα θηλυκῶς, Ἀττικοὶ ῥᾶγα, ‘And [it also means] what we [call] ῥῶγα (acc. sing.) in the feminine, [while] users of Attic [call it] ῥᾶγα (acc. sing.)’). This second part may have been taken precisely from ρ 116 (B.1).

(2)    Eust. in Od. 1.347.37–40 (B.7)

In this passage of his commentary on the Odyssey, Eustathius seems to confuse two different nouns which are homographsHomography: the noun ῥώξ at the basis of the compound ἀπορρώξ, which derives from the o-grade of the verb ῥήγνυμι (see DELG, EDG s.v.), and the noun ῥώξ, meaning ‘spider’, which is dealt with in the Atticist sources and is a variant of ῥάξ. Later in the same passage, Eustathius again refers to the former ῥώξ (from the same root as ῥήγνυμι), saying that the noun is used by Homer in the plural to denote a part of the megaron (C.1). Yet, the meaning of ῥῶγες in the Homeric passage has been debated since antiquity: the noun has been interpreted variously as ‘steps’, ‘windows’, ‘cracks in the wall’, and ‘passages’ (see Philox.Gramm. fr. 403, Apollon. Lex. 30.26–9 and 139.24, Hsch. α 4567 and ρ 556); cf. Wace (1951, 210–1); Bérard (1954, 23–5). Eustathius himself, however, clearly seems to distinguish between the two homograph nouns in another passage of the commentary, where he discusses only the noun ῥώξ derived from the root of ῥήγνυμι (in Od. 2.275.36–9; cf. Philox.Gramm. fr. 403 (= Et.Gen. AB α 772, EM 1285), EM 705.44–5): ῥῶγες δὲ μεγάροιο, δίοδοι, ῥήγματα, θυρίδες, θηλυκῶς ἀπὸ τοῦ ῥὼξ ῥωγός· αὐτὸ δὲ ἀπὸ τοῦ ῥήσσω ῥήξω. τινὲς δὲ οὐδετέρως τὸ ῥῶγας ἐνόησαν ὡς τὸ κῶας. ὅτι δὲ ἐν τῷ, ἀνὰρ ῥῶγας (ἀνὴρ ῥῶγας Stallbaum), διπλάζουσι τινὲς τὸ ἀμετάβολον, καθά που καὶ ἐν τῷ κατὰρ ῥόον καὶ ἐν ἑτέροις ὁμοίοις, ἔστιν ἐκ τῶν παλαιῶν ἀντιγράφων ἀναλέγεσθαι (‘ῥῶγες δὲ μεγάροιο: [It means] passages, cracks, windows. The feminine [noun derives] from ῥὼξ ῥωγός, but this [noun (i.e. ῥῶγες) comes] from ῥήσσω ῥήξω (‘to break’). Some considered ῥῶγας neuter, like κῶας (‘fleece’). In the ancient copies one can read that in [the expression] ἀνὰ ῥῶγας (‘up through the passages’) some double the liquid consonant (i.e. ἀναρρῶγας), just as in κατὰ ῥόον (‘following the stream’, i.e. καταρρόον) and in other similar [strings of words]’). As for erudites sources dealing with the expression ἀνὰ ῥῶγας, see Pagani (2015, 252–3).