οὐχ οἷον/μὴ οἷον with finite verb forms
(Antiatt. o 5, Phryn. Ecl. 348)

D. General commentary

Two entries from the Antiatticist’s lexicon (A.1) and Phrynichus’ Eclogue (A.2) appear to agree in proscribing the construction οὐχ οἷον/μὴ οἷον + finite verb. Starting from Phrynichus’ entry (for the problematic state of the text in the Antiatticist’s entry, see below), the following commentary argues that the proscription concerns a specific use of this construction in the emphatic sense ‘not at all’. From the koine period onwards, the same construction is also employed in correlative structures meaning ‘not only (not) ... but also/but not even’, in which the meaning of οἷον may have been influenced by that of οἶον (‘only’), derived from οἶος (‘alone, lonely’). Consequently, it may well be that, alongside a primary focus on syntax and semantic correctness, the Atticist proscriptions also reflect a concern for orthoepyOrthoepy and the disambiguation of οἷον (aspirated) from οἶον (not aspirated).

Excluding the other functions derived from the meanings of the corresponding relative adjective-pronoun (see LSJ s.v. οἷος, V), the neuter singular of οἷος, used adverbially and preceded by a negation, is attested in a specific syntactic construction that begins to spread from koine period onwards, particularly in authors such as Polybius (cf. e.g. C.6, C.7, 23.11.8), Diodorus Siculus (cf. e.g. 1.83.4, C.8), Philodemus (cf. Rh. P.Herc. 1015 col. 37.16–9; Ir. fr. 19, P.Herc. 182 col. 19.31), and Philo of Alexandria (Legum allegoriarum libri 1.103; 1.105); cf. already Arist. Ath. 40.3. This is a correlativeCorrelative sentences construction in which the sequence οὐχ οἷον followed by a finite verb (primarily in the indicative) acquires the meaning ‘not only (not)’ and is followed by a second clause introduced by sequences such as ἀλλ’ οὐδέ (‘but not even’, cf. C.6; this is the most frequent kind of correlation), ἀλλὰ καί (‘but also’, cf. C.9), or τὸ δ’ ἐναντίον/ἀλλὰ τοὐναντίον (‘but on the contrary’, cf. C.7; cf. also C.8 for ἀλλὰ καὶ τοὐναντίον, ‘but also on the contrary’); see Arnott (1996, 594). As mentioned above, one may wonder whether the emergence of the correlative structure introduced by οὐχ οἷον was somehow influenced by the semantically related use of the adverbial neuter οἶον (‘only’), derived from οἶοςοἶος (‘alone, lonely’); cf. ThGL vol. 6, 1831. Although the differing initial aspiration of the two forms originally served to distinguish them – in addition to affecting the spelling of the semantically equivalent sequences οὐχ οἷον and οὐκ οἶον respectively – the loss of aspiration in Post-classical Greek may have facilitated the overlap between these two adverbial neuters. It is therefore possible that the Atticists’ entries – beyond the syntactic concerns discussed below – also served an orthoepic purpose by disambiguating the two forms. Returning to the correlative structure, the corresponding expressions in Attic Greek are οὐ μόνον (οὐ), οὐχ/μὴ ὅπως and οὐχ/μὴ ὅτι (followed in the second clause by introductory sequences such as ἀλλὰ καί, ἀλλ’ οὐδέ, and the like). In particular, οὐχ/μὴ ὅπως and οὐχ/μὴ ὅτι presuppose an implied verbum dicendi (i.e., ‘I do not speak of’, ‘not to speak of’) and, like οὐχ οἷον, may assume either an affirmative value (‘not only’) or a negative one (‘not only not’, ‘so far from’); see LSJ s.v. ὅπως, II.2, K–G vol. 2 § 525, Goodwin (1897, 283–4); Schwyzer, Debrunner (1950, 670); Cooper (1998 vol. 2, 1131–2).

Regarding the correlative construction that spread during the koine period, the clause introduced by οὐχ οἷον normally occupies the first position. The only apparent inversion of this order occurs in a fragment of Alexis preserved by Athenaeus (6.244e = C.1), an attestation that is relevant to previous scholars’ interpretation of the Atticist proscriptions (see below). In l. 4 of the fragment (πέτεται γάρ, οὐχ οἷον βαδίζει τὰς ὁδούς, ‘for he flies, far from merely walking through the streets’), the emphasis falls on the verb πέτεται rather than on βαδίζει: the action that, in a non-emphatic construction (i.e., οὐχ οἷον βαδίζει τὰς ὁδούς, ἀλλὰ πέτεται), would have occupied the second position is thus moved to the first. As observed by Arnott (1996, 594), this is probably an inversion originating in popular speech to meet an expressive need, exactly as in the sequence οὐχ ὅπως in Soph. El. 796: πεπαύμεθ’ ἡμεῖς, οὐχ ὅπως σὲ παύσομεν, ‘We have been stopped, far from our stopping you!’ (transl. Lloyd-Jones 1994, 237); see Finglass (2007, 348), with further references (for the inversion of the construction with μὴ ὅτι, cf. instead e.g. Pl. R. 398e.3–4 and X. HG 2.3.35). See further K–G vol. 2 § 525.4; Goodwin (1897, 284); Cooper (1998 vol. 2, 1131–2, esp. point E); LSJ s.v. ὅπως, II.2, b.

Turning now to Phrynichus’ entry (A.2), the alternative construction he prescribes makes it quite clear that the rejection of οὐχ οἷον/μὴ οἷον + finite verb forms does not concern the correlative use just illustrated, but rather what may be termed an ‘absolute’ use of the construction, one that serves to emphasise the negation of the clause. Quite surprisingly, however, this usage does not appear to be attested elsewhere in surviving literary or documentary sources. Indeed, the alternative construction to οὐχ οἷον/μὴ οἷον prescribed by Phrynichus, οὐ δήπου/μὴ δήπου (‘not at all’), conveys emphatic negation. Besides expressing surprise or incredulity in direct questions (see Denniston 1954, 267–8; Cooper 1998 vol. 2, 1317–8), the expression is well attested in Attic prose as a means of reinforcing the negation of the statement: cf. e.g. C.2, C.5. Another element in A.2 that appears to point to a widespread colloquial, ‘absolute’, and emphatic use of the expression οὐχ οἷον/μὴ οἷον + finite verb forms in the 2nd century CE is the statement μάλιστα δ’ ἁμαρτάνεται ἐν τῇ ἡμεδαπῇ. If, as suggested by earlier editors (see Nunnesius cited by Lobeck 1820, 372), ἐν τῇ ἡμεδαπῇ refers to Phrynichus’ native land (probably Bithynia: see Bowie 2022, 81–2; 2025, 24–5), this would provide support for the hypothesis that the construction proscribed by Phrynichus was typical of the spoken language, particularly in regions peripheral to Greece.

Since it is included in the second book of the lexicon, Phrynichus’ entry likely takes the Antiatticist’s entry (A.1) as its direct point of reference (see Latte 1915, 378–82; Valente 2015, 52–4), as is also suggested by the repetition of the lemma οὐχ οἷον ὁργίζομαι. The Antiatticist’s entry, however, presents textual anomalies that imply later interpolations and is therefore less reliable for establishing the precise nature of the linguistic issue under discussion. First, neither of the expressions serving as lemmata in the Antiatticist’s entry is attested elsewhere in the literary record. While οὐχ οἷον ὁργίζομαι – which also appears as the lemma of A.2 – may have been attested in a now-lost work, the reference to the expression οὐχ οἷον ἁλίσκω is more difficult to account for. The active form ἁλίσκω is in fact attested for the first time in a proverb preserved in the collections attributed to Zenobius and Diogenianus (Zen. 3.67, Diogenian. 4.45 = Diogenian. Vind. 2.66; see Tosi 2017, 1555–6 (on no. 2265); cf. DGE s.v. ἁλίσκομαι), and subsequently occurs with greater frequency in late antique and medieval literary works (cf. Lampe s.v. ἁλίσκω; Kriaras, LME s.v. ἁλίσκω). The peculiarity of the verbal form is certainly one of the reasons that led Nauck (see the apparatus of A.1, cf. Arnott 1996, 594; see also Lobeck 1820, 373) to emend the text of the entry to οὐχ οἷον βαδίζω/βαδίζει, thereby creating a correlation between the Atticist proscription and the use of the οὐχ οἷον construction attested in the fragment of Alexis (C.1); see also Arnott (1996, 594). Secondly, the Antiatticist’s entry displays a strangely prescriptive structure – an attitude unusual for this lexicon – which appears in some respects to conform to the structure of the Phrynichean entry. The alternative construction prescribed in A.1, (πολὺ) ἀπέχω τοῦ + infinitive (‘I’m (very) far from …’) serves to emphasise the negation of the clause, much like the construction οὐ δήπου/μὴ δήπου prescribed by Phrynichus. The expression is well attestated in Attic prose (cf. e.g. C.3, C.4, D. 24.3), especially in the consecutive structure τοσοῦτον ἀπέχω τοῦ … ὥστε. This function differs markedly from the correlative use of οὐχ οἷον illustrated above, which is also attested by the fragment of Alexis (C.1) and whose corresponding expressions in Attic are οὐ μόνον (οὐ), οὐχ/μὴ ὅπως, and οὐχ/μὴ ὅτι. The use of ἁλίσκω in the active voice and the prescriptive structure of the entry may therefore point to a later interpolation designed to align the passage with the intent of the parallel entry in Phrynichus. It cannot be ruled out that the Antiatticist originally referred to the passage by Alexis – as assumed by Nauck in his emendation of ἁλίσκω to βαδίζω (see A.1, apparatus) – most likely in order to defend rather than condemn his linguistic usage (among the several entries of the Antiatticist lexicon that adopt Alexis as a linguistic model, cf. e.g. ε 5Antiatt. ε 5, discussed in entry ἐπιχαιρέκακος, and η 5Antiatt. η 5, discussed in entry ἡδίων). One may wonder, however, whether the emphatic structure found in the comic poet’s fragment was already understood in the 2nd century CE as an ‘absolute’ expression of negation, rather than in a correlative sense. In this regard, the colloquial emphatic inversion of the correlative structure examined above – which presupposes the placement of the expressions οὐχ οἷον, οὐχ ὅπως, and μὴ ὅτι in the second clause – may itself have provided the starting point for the development of the ‘absolute’ use of the construction, no longer perceived as correlative. Entry A.1 may therefore have treated the same linguistic topic discussed by Phrynichus (who, as previously noted, probably adopted the Antiatticist’s doctrine for polemical purposes), though with a more tolerant attitude and in a formulation different from, and perhaps broader than, the one currently preserved in the epitomisedEpitome text.

Be that as it may, A.2 makes it reasonably clear that the ‘absolute’ construction with οὐχ οἷον was disapproved of because it lacked attestation in the canonical authors. At the same time, the correlative use of οὐχ οἷον is likewise unattested in Attic literature (with the exception of C.1), becoming widespread only in koine Greek and subsequently avoided by Atticising authors. As previously mentioned, it may reasonably be assumed that the ‘absolute’ construction derives from the correlative one (cf. in this regard also C.5, where οὐ δήπου, prescribed by Phrynichus, is followed by ἀλλὰ τοὐναντίον, in a manner closely resembling the correlative structures introduced by οὐχ οἷον). Furthermore, the construction οὐχ οἷον/μὴ οἷον + finite verb may also have been felt to be problematic because of its potential confusion with another very common use of the neuter οἷον, namely the impersonal construction with εἶναι in the 3rd person singular – often implied – followed by an infinitive of consecutive value expressing ‘fitness or ability for a thing’ (cf. LSJ s.v. οἷος, ΙΙΙ; K–G vol. 2 § 473.3; Schwyzer, Debrunner 1950, 678), or, in negative clauses, the lack thereof. Finally, Phrynichus’ observation that the construction οὐχ οἷον/μὴ οἷον is τῷ ἤχῳ ἀηδές (‘unpleasant-sounding’) is noteworthy: although it is unclear what exactly the lexicographer found disagreeable in the pronunciation of the sequence, the aspiration was probably involved.

To conclude, the Antiatticist’s and Phrynichus’ entries are of particular interest for two reasons. First, although they appear to constitute one of the rare cases in which the two lexica agree on a linguistic issue (cf. also Antiatt. β 4Antiatt. β 4 and Phryn. Ecl. 238Phryn. Ecl. 238, with Phryn. PS 53.19Phryn. PS 53.19, discussed in entry βλάξ; Antiatt. σ 7Antiatt. σ 7 and Phryn. Ecl. 380Phryn. Ecl. 380, discussed in entry στρωματόδεσμον; cf. Valente 2015, 53), the authenticity of the wording of the Antiatticist’s entry is highly questionable, and the lexicographer was probably defending precisely the usage condemned by Phrynichus. Secondly, and even more interestingly, the construction targeted by the proscription does not appear to be attested in either earlier or contemporary literary language. As the preceding discussion has attempted to show, the construction οὐχ οἷον/μὴ οἷον + finite verb, used as an ‘absolute’ emphatic negation, is more plausibly attributable to the spoken language contemporary with the lexicographers, who are the only authors to preserve traces of it.

E. Byzantine and Modern Greek commentary

Correlative coordination with the first clause introduced by οὐχ οἷον continues to occur sporadically in the Byzantine period as well. The Suda records this construction in two entries (B.1, B.2). In ο 988, the quoted fragment is tentatively attributed by Adler to Aelian; given, however, the absence of the construction in Atticising authors and more generally in works dated to the 2nd century CE, this attribution appears doubtful. The fragment should more probably be attributed to PolybiusPolybius, like the passages cited in the following entry. Cf. also Su. ε 2838, where the sentence οὐχ οἷον ἥδομαι, ἀλλ’ ὑπερήδομαι (‘I’m not just pleased; rather, I’m super-pleased!’) is cited as the current idiom (ἐν οἵῳ τρόπῳ λέγομεν) for expressing positive emphasis on an action. This formulation resembles that found in the fragment of Alexis (C.1), though without the expressive inversion observed there (the Suda entry, whose text is partly problematic, is based on the scholium to Ar. Ra. 745 and reports its wording in an elliptical and inaccurate manner). It is no coincidence that the author who appears to employ the correlative construction most frequently is Constantine VII Porphyrogenitus (cf. e.g. De legationibus 281.18–9; 314.8–9; De virtutibus et vitiis 2.105.15; 2.107.13), which suggests that Polybius’ linguistic model was taken into consideration in the Macedonian period. See also Michael Psellus Chronographia 4.11.15–6. Conversely, no attestations have been identified of the absolute emphatic construction apparently criticised by the Atticists.

F. Commentary on individual texts and occurrences

N/A