κεραννύω, κίρνημι
(Antiatt. κ 39, [Hdn.] Philet. 299)

D. General commentary

Two entries, from the Antiatticist (A.1) and the Philetaerus (A.2), prescribe the verb κεραννύω (‘to mix’) and proscribe its more archaic equivalent κίρνημι, which, according to the Antiatticist, was favoured by other scholars (on the subject of ὡς οἴονται, ‘as they believe’, in A.1, see below). These entries do not consider the athematic form κεράννυμι, referring only to the thematised κεραννύω (for other entries on the thematisation of athematic verbs, see ἀπολλύασιν, ἀπολλύουσιν, and other 3rd person plurals of -νύμι verbs and διδόασι, διδοῦσι, AGP vol. 1, 319–29, AGP vol. 2, Verbal morphology, forthcoming).

The Greek verbs κεράννυμι/κεραννύω and κίρνημι/κιρνάω reflect the IE root *k̑erh₂- (see EDG s.v. κεράννυμι; LIV 328). From a historical linguistics perspective, the oldest present form appears to be κίρνημι/κιρνάω, originating from the zero-grade with the insertion of /i/. This insertion stems from the analogic extension to the root of κίρνημι/κιρνάω of the i-vocalism, typical of other zero-grade roots involving the so-called schwa secundum. This can be defined as ‘an epenthetic vowel inserted to break up word-initial consonant clusters in PIE zero grade sequences’ (Vine 2013) involving two consecutive obstruents: this epenthetic vowel is realised as /i/ in Ancient Greek (Willi 2018, 488 n. 192 speaks of ‘cluster-preserving’ *-i-’); cf. e.g. πίτνημι ‘to spread’, from the zero-grade *pt- and nasal infix; for a general overview on the schwa secundum in ancient Greek linguistics, see Vine (2013), with further bibliography. κεράννυμι/κεραννύω is, instead, a later e-grade formation based on the full-grade aorist ἐκέρασα (well attested in Homer; see e.g. Hom. Od. 3.393: τοῦ ὁ γέρων κρητῆρα κεράσσατο, ‘from it (i.e. the wine) the old man mixed a bowl’; cf. LIV 328 n. 3). The archaic zero-grade stem is found on multiple occasions in the Homeric poems (see the formula μελίφρονα οἶνον ἐκίρνα, ‘(s)he mixed wine sweet to the mind’, in Od. 7.182 = C.1, 10.356, 13.53; see also the imperative κίρνη in Od. 14.78, 16.52; for the present root κερν- in Aeolic – cf. Alc. frr. 338, 367 Lobel–Page² = Voigt and below – see Scarborough 2023, 140–1, with further bibliography). The earliest attestations of the present stem with the suffix -νυ- are in Plato (16x; cf. C.2). In the 5th–4th centuries BCE, κεράννυμι is attested more often than κίρνημι in literary texts, with the thematised forms of κεραννύω first appearing in 4th-century prose and poetry (see e.g. C.3, C.4, Arist. Mete. 372a.7). However, in the Hellenistic period, κεράννυμι/κεραννύω appears to recede in favour of κίρνημι/κιρνάω, and this trend continues into imperial and late antique literature, with the zero-grade root gradually becoming more prevalent (see also E.). This is also confirmed by epigraphic and papyrus evidence: the present stem with suffix -νυ- is absent from documentary texts, while the stem κιρν- is found at least six times in inscriptions (the athematic κίρνημι is in IG 4²,1.130.23 [Epidauros, 2nd–3rd century CE], I.Phrygie Drew-Bear 75, n. 8.12 [Eumeneia, 3rd century CE], I.Aphrodisias Late Ant.² 31.5 [4th century CE]; the thematised κιρνάω is in I.Magnesia 99.27 [2nd century BCE]; see also the dubious κιρνέαται in I.Milet 1.3.133.11 and the lacunose συνκιρνα̣[ in IG 12,9.259.47 [4th century BCE]. To these should be added the Aeolic variant κέρναν in IG 12,2.1.13–4 [Mytilene, 4th century BCE]). Regarding papyri, κιρνάω is found once in the medical recipe of P.Oxy. 74.4978.12 (= TM 119323) [2nd–3rd century CE].

Regarding the verbal root, for the Atticists, the full-grade present stem must have been clearly preferable given its strong Attic pedigree (established not only through its use in Old Comedy but also through its pervasive use in models of Attic prose, such as Plato (27x)) and its status as a less widespread (and thus marked) form. On the contrary, the originally highly poetic nature of κίρνημι/κιρνάω (occurring, after Homer, mainly in lyric and tragedy) rendered this form unadvisable, as Pollux makes clear (B.4).

However, regarding the conjugation, A.1 and A.2 recommend the thematised form κεραννύω against the athematic κίρνημι seemingly in contrast with the strict Atticists’ general preference for athematic verbs (see above and B.4, which contains only athematic forms of κεράννυμι). That κεραννύω occurs multiple times in Middle Comedy (see C.4, Eub. fr. 93, Theophil. fr. 2) – a genre the Antiatticist often uses as a source of literary quotations to support specific forms and usages – may explain the lexicon’s choice of the thematised form. In particular, C.4, a fragment from Alcaeus’ Sacred Marriage, contains the form κεραννύουσι, and we cannot discount the possibility that this passage underlies the Antiatticist’s entry (the playwright is mentioned eight times in the lexicon, and one such reference is in the entry immediately preceding A.1, i.e. Antiatt. κ 38Antiatt. κ 38; on this, see entry κρεμῶ, κρεμάσω, ὀμοῦμαι, ὀμόσω). In addition, an entry in Photius and the Suda (B.2), closely resembling that in the Antiatticist, refers to Hyperides (C.3) in relation to the form κεραννύουσι. The common source of Photius and the Suda (B.2) is the expanded Synagoge; given that ‘a fuller version of the so-called Antiatticist’ (Cunningham 2003, 52) was among the sources for the Synagoge’s expansions, we cannot exclude that B.2 preserves a more complete text of the Antiatticist’s entry than cod. Par. Coisl. 345, and that the Antiatticist originally referred to Hyperides (the orator is mentioned 15 times in the lexicon’s extant version). However, the similarity can also be explained by assuming that the Antiatticist and the expanded Synagoge used a common source (see Valente 2015, 23). This idea becomes more plausible when we compare a section in Athenaeus (B.1), which quotes, in direct succession, three examples of the literary use of κεραννύω: these are a passage from Plato (Phlb. 61b.11 = C.2), the previously mentioned line from Alcaeus’ Sacred Marriage (C.4), and a sentence from Hyperides’ speech On Delos (C.3) referred to in B.2. In other words, it is worth considering the possibility that Athenaeus (B.1), the Antiatticist (A.1), and the Synagoge expansion (B.2) rely on a common source, possibly identifiable with Pamphilus (through the mediation of Diogenianus in the case of B.2 and, perhaps, also A.1; cf. Cunningham 2003, 53–4; Valente 2015, 35–6).

Regardless of the locus classicus of A.1, the entry reinforces the idea of a general preference in the Antiatticist for the thematised forms of athematic verbs; this can also be observed in Antiatt. δ 8Antiatt. δ 8 (which supports διδοῦσι over – or along with – διδόασι as the third person plural of the present indicative of δίδωμι; see entry διδόασι, διδοῦσι). Regarding the phrase ὡς οἴονται (‘as they believe’), this is not infrequent in the Antiatticist (see Antiatt. α 105Antiatt. α 105, α 133Antiatt. α 133, γ 18Antiatt. γ 18, ζ 1Antiatt. ζ 1); while the subject is unclear, the phrase probably refers to other scholars holding a view different from the lexicographer’s (see Valente 2015, 49).

The Philetaerus (A.2) shares the Antiatticist’s prescription of the full-grade κεραννύω against κιρνάω. However, its lemma and interpretamentum are in the present infinitive, which is likely the result of lemmatisation. Initially, one could suppose that the Philetaerus depends on the Antiatticist for this entry; however, the relationship between these two lexica is not easily established, and the similarity between A.1 and A.2 may be explained through the use of a common source (see Valente 2015, 55–6).

E. Byzantine and Modern Greek commentary

κεραννύω/κεράννυμι and κιρνάω/κίρνημι (the latter forms being prevalent since the Hellenistic period; see D.) both continue into Byzantine Greek. The e-grade alternatives κεραννύω and κεράννυμι are found as late as the 19th century in learned archaising language (e.g. Neophytus Ducas 1x) but do not survive in Modern Greek.

Lower-register texts from the 9th century onwards attest to the evolution of κιρνάω into κερνάω/κερνέω (see LBG s.vv.; cf. also the apparatus of B.2 for the form κερνοῦσι in Photius); this shift results from the lowering of the unstressed /i/ to /e/ before /r/, which is a common phonetic change in Medieval Greek (see CGMEMG vol. 1, 68–9; Méndez Dosuna 1985, 410) but can already be observed in late koine Greek (see Gignac 1976, 254; 259). The verb κερνάω is retained in Early Modern Greek, meaning ‘to pour’ (see Kriaras, LME s.v.), and survives into Modern Greek (κερνώ) as in the more general sense ‘to offer’ (i.e. something to eat or drink) or ‘to pay for someone else’s meal’.

F. Commentary on individual texts and occurrences

N/A