ἤθη and ἦθος as accusatives of respect
(Antiatt. η 4, Phryn. Ecl. 344)

D. General commentary

Two entries in the Antiatticist lexicon (A.1) and in Phrynichus’ Eclogue (A.2) involve the plural form of the noun ἦθος (‘custom’, ‘character’). At first glance, the Antiatticist appears to assess the admissibility of the plural ἤθη per se. However, Phrynichus’ contrast between the phrases χρηστὸς τὰ ἤθη and χρηστὸς τὸ ἦθος (‘good-natured’) shows that the debate concerns the insertion of the noun into a specific syntactic context, i.e. the construction ‘singular adjective + τὰ ἤθη’, with the latter functioning as an accusative of respect, rather than the use of the plural ἤθη itself (pace Lee 2013, 295–6). Consequently, the Antiatticist entry (A.1) should likewise be interpreted in relation to the same syntactic phenomenon, despite its concise formulation.

The noun ἦθος is very common throughout the history of Ancient Greek. Starting from its primary meaning ‘custom’ or ‘usage’, in Homer it is always used in the plural with the sense ‘accustomed place’, hence ‘haunts’ or ‘abodes’ of animals (Il. 6.511, 15.268; Od. 14.411). The meaning ‘disposition’ or ‘character’ is attested as early as Hesiod (Op. 67, 78), whereas the plural more often denotes ‘customs’ or ‘manners’ (cf. e.g. Hes. Op. 137, Th. 66, Hdt. 2.30; in the well-known maxim φθείρουσιν ἤθη χρήσθ’ ὁμιλίαι κακαί, ‘bad company corrupts good character/manners’, attributed in antiquity to Euripides – fr. 1024.4 – or Menander – fr. 165 – and very popular in imperial and Byzantine literature owing to its citation by Paul in 1 Ep.Cor. 15.33, ἤθη could be interpreted either as ‘character’ or as ‘manners’. On the source of the maxim, see Renehan 1973, 29–34; for a discussion of the maxim itself, see Tosi 2017, 1173–5, no. 1729). See further DELG, EDG s.v.

Phrynichus’ entry (A.2) rejects the phrase χρηστὸς τὰ ἤθη (adjective + plural noun used as an accusative of respect) prescribing instead χρηστὸς τὸ ἦθος (adjective + singular noun as an accusative of respect). The two phrases probably have the same meaning (‘good-natured’), since the entry does not appear to address the semantics of the singular and plural forms of the noun ἦθος; see Rutherford (1881, 468). As noted above, the Antiatticist entry (A.1) appears instead to defend the plural ἤθη as such, citing in support a passage from Antiphanes (C.3), which is unfortunately not preserved elsewhere. However, in light of A.2, it is very likely that the Antiatticist too originally concerned the specific use of ἦθος/ἤθη as accusatives of respect, and that the entry subsequently underwent significant epitomisationEpitome (for other cases in which Antiatticist entries are suspected of later epitomisation, cf. e.g. α 60Antiatt. α 60, on which see entry ἀργύρωμα, χρύσωμα; ε 76Antiatt. ε 76, on which see entry ἔντεχνος, τεχνικός; κ 90Antiatt. κ 90, on which see entry κακκάβη, κάκκαβος; and ο 5Antiatt. ο 5 – likely subject to interpolation as well as epitomisation – on which see entry οὐχ οἷον/μὴ οἷον with finite verb forms). From this perspective, the fact that the Antiatticist attributes to other Atticists the proscription of the plural ἤθη – something that would be odd if the issue were merely the form itself, given its widespread use at all stages of Greek – de facto aligns with the stance taken by Phrynichus. Phrynichus’ entry (A.2), in turn, is readily interpreted as a response to the Antiatticist, since it belongs to the second book of the Eclogue (see Latte 1915, 378–82; Valente 2015, 52–3).

As for the presence of the adjective χρηστός in A.2, it may be assumed either that Phrynichus was referring to a locus classicus in which the construction with the plural τὰ ἤθη was actually attested (likely the same passage originally cited in the Antiatticist entry: see below), or that the adjective is merely illustrative, serving to exemplify the construction ‘singular adjective + τὰ ἤθη/τὸ ἦθος’. Indeed, the phrase χρηστὸς τὰ ἤθη rejected by Phrynichus is not attested in the extant literature (if his entry responds to the Antiatticist lemma, it may be supposed that it occurred at least in Antiphanes, C.3). However, other instances of the pattern ‘singular adjective + τὰ ἤθη (acc. of respect)’ are attested in post-classical authors. Among the earliest attestations is the example found in [Ammon.] 326.2 (~ [Ptol.Ascal.] Diff. 402.13–4 Heylbut): μοχθηρὸς μὲν γὰρ ὀξυτόνως ὁ τὰ ἤθη πονηρός (‘For μοχθηρός, with acute accent, [refers to] one who is knavish in character’). This provides a clear indication of the absence of Atticist ideology in grammatical and lexicographical literature prior to the main corpus of Atticist lexicography (see Probert 2011). Generally, in cases involving the construction ‘singular adjective + τὰ ἤθη’, it is necessary to determine whether ἤθη means ‘character’ or ‘customs/manners’: cf. e.g. Origenes Cels. 6.28: μέτριος τὰ ἤθη ‘moderate in character’, C.5: τὰ ἤθη κράτιστος ‘of strong character’, Hld. 1.19.2: βάρβαρον [...] τὰ ἤθη ‘barbarous [...] in manners’. On the basis of these three passages, in Diogenes Laertius (C.5) τὰ ἤθη clearly means ‘character’, and probably so in Origen as well, whereas in Heliodorus ‘manners’ appears more appropriate. In any case, the two concepts are very nuanced, and it is not always easy to distinguish between them. At the same time, the phrase χρηστὸς τὸ ἦθος approved by Phrynichus is not attested in Attic authors, but occurs in two passages by the imperial-age historian Herodian (3.3.4, C.4). The phrase ‘singular adjective + τὸ ἦθος’ is nonetheless attested in canonical Attic authors (cf. C.1: πρᾷος τὸ ἦθος ‘of gentle character’, C.2: οὐδὲ γὰρ εἰ πάνυ χρηστός ἐσθ᾿ […], βελτίων ἐστὶ […] τὸ ἦθος ‘indeed, even if he is a really good man […] he is no better […] in character’; note that this Demosthenes passage may have served as the authority underlying Phrynichus’ prescription, even though the morphosyntactic construction differs from that cited by the lexicographer: cf. also Phryn. Ecl. 399Phryn. Ecl. 399, where a passage from Thucydides (6.9.1) is quoted in a different form. Such discrepancies between citations and the original wording are not isolated, but recur in other technical literature of the imperial period; see Chiron 2003, 509–12). The construction is also found in later Atticising writers (Philostr. VS 2.590.12: ἀναιδῆ τὸ ἦθος ‘disrespectful in character’, Luc. Salt. 72: ἀμείνων τὸ ἦθος ‘better in character’) as well as in various passages of Pollux’s Onomasticon: cf. B.1 (μοχθηρὸς τὸ ἦθος ‘knavish in character’), 5.119Poll. 5.119 (ἐπαινετὸς τὸ ἦθος ‘praiseworthy in character’), 6.27Poll. 6.27 (πανηγυρικὸς τὸ ἦθος ‘pompous in character’), and 9.26Poll. 9.26 (φιλόπολις τὸ ἦθος ‘patriotic in character’).

As already noted by Lobeck (1820, 364–5), in Classical Greek, the use of ἦθος as an accusative of respect appears to reflect the following pattern: with a singular adjective, τὸ ἦθος is used, whereas with a plural adjective the preferred form is τὰ ἤθη (cf. e.g. Pl. R. 535b.1–2: γενναίους τε καὶ βλοσυροὺς τὰ ἤθη ‘noble and virile in character’, Arist. EN 1127b.22–3: χαριέστεροι […] τὰ ἤθη ‘refined in character’, Philostr. Im. 1.19.6: τὰ δὲ ἤθη χρηστοῖς ‘good-natured’; it should be noted, however, that even in these cases it is sometimes difficult to determine whether τὰ ἤθη is best translated as ‘character’ or as ‘manners/behaviour’). As in the opposite case of χρηστὸς τὰ ἤθη, constructions combining plural adjectives with τὸ ἦθος are attested only in post-classical authors (Arist. PA 650b.34: θυμώδη (neut. plur.) τὸ ἦθος ‘of passionate temperament’, Ph. De specialibus legibus 4.120: κοινωνικοὺς (m. plur.) τὸ ἦθος ‘sociable in character’, D.Chr. 31.32: μετρίων (m. plur.) τὸ ἦθος ‘temperate in character’). In conclusion, in Classical Greek agreement between the grammatical number of the adjective and that of the noun ἦθος functioning as an accusative of respect is the preferred pattern. By contrast, divergent constructions – namely, ‘singular adjective + τὰ ἤθη’ or ‘plural adjective + τὸ ἦθος’ – are attested only in the post-classical period; on the basis of the extant literary evidence, the latter appears to be chronologically earlier. However, such agreement is not observed in Classical Greek with other nouns used as accusatives of respect and conveying a meaning similar to ἦθος: see Lobeck (1820, 365), who gives the example of τρόπος (‘habits’, ‘character’). In Attic authors, when combined with singular adjectives, τρόπος occurs as an accusative of respect both in the singular and in the plural: cf. [D.] 56.2: δίκαιος […] τὸν τρόπον ‘upright […] in character’, Ar. Eq. 192: χρηστοῦ τοὺς τρόπους ‘of good character’.

Returning to Atticist lexica, Phrynichus rejects a syntactic usage that emerges only in Post-classical Greek, namely the construction ‘singular adjective + τὰ ἤθη’, prescribing instead ‘singular adjective + τὸ ἦθος’, perhaps on the basis of a now-lost locus classicus (but cf. D. 20.14 cited above; it remains possible that the phrase χρηστὸς τὸ ἦθος was constructed e contrario by Phrynichus on the model of χρηστὸς τὰ ἤθη, in accordance with the pattern of the ‘singular adjective + τὸ ἦθος’ examples mentioned above). Furthermore – an observation for which I am indebted to the anonymous reviewer – occurrences of the construction ‘singular adjective + τὰ ἤθη’ appear not only to be post-classical but also, at least when ἤθη means ‘character’, to occur chiefly in writers whose usage can be characterised as ‘intermediate-level Greek’. In the imperial period, this corresponds to the register of technical prose, as opposed to belles lettres (on this concept, see Rydbeck 1967). Given that Phrynichus explicitly proscribes this pattern – while not targeting ‘plural adjective + ἦθος’ – and that the Antiatticist (judging from its wording) appears to defend it against earlier proscriptive evaluations, one wonders whether not only chronology but also registerRegister stratification influenced these positions. The spread of a construction such as χρηστὸς τὰ ἤθη may have been encouraged both by analogyAnalogy with the plural nouns of similar meaning used as accusatives of respect (e.g. τοὺς τρόπους) and by the semantic ambivalence of ἤθη, which can denote both ‘character’ and ‘customs’ (in the latter case, the plural form is the expected one). Finally, it cannot be ruled out that the widespread circulation of the maxim φθείρουσιν ἤθη χρήσθ’ ὁμιλίαι κακαί – quoted above – may also have contributed to the diffusion of the construction combining the plural ἤθη with the adjective χρηστός.

E. Byzantine and Modern Greek commentary

N/A

F. Commentary on individual texts and occurrences

N/A