καμμύω
(Antiatt. κ 54, Phryn. Ecl. 316)

D. General commentary

The entries in the Antiatticist (A.1) and in Phrynichus’ Eclogue (A.2) are concerned with καμμύω, the variant form of καταμύωκαταμύω with apocopeApocope of the preverb. This form is increasingly common in Post-classical Greek, but its status as a classical form is dubious. While the Antiatticist defends it (Latte 1915, 379 n. 2 advances the hypothesis that the Antiatticist’s discussion derives from Aristophanes of Byzantium), Phrynichus contests those who justify a dubious form on account of a single occurrence in Alexis (C.1), since ‘the best of the ancient authors’ (οἱ ἄριστοιοἱ ἄριστοι τῶν ἀρχαίωνοἱ ἀρχαῖοι) use καταμύω. Although the Antiatticist entry is poorly epitomised, the reference to Alexis in Phrynichus makes it most likely that Alexis’ fragment was the locus classicus used by the Antiatticist to defend καμμύω. In fact, this is one of several cases where a gloss in Phrynichus is a direct response to one in the Antiatticist (see Latte 1915, 378–9; for a comparable though not identical case, see Λάκαινα). Looking at the grammatical tradition more broadly, καταμύω and καμμύω also attracted consideration from Apollonius Dyscolus, who identifies καμμύω as the common form (B.1) and regards it as acceptable (B.2). Notice that at Synt. 4.43 = GG 2,2.469.9–470.10, Apollonius offers other general arguments to defend these cases against the opinion of those who regard them as ἄλογα ‘contrary to reason’ (see further F.3).

The apocope of preverbs and prepositions is relatively well attested in tragedy,Tragedy while it is foreign to prose writers and is only used in comedy in paratragic passages or in scenes which involve foreign speakers (for a general overview and a collection of examples, see Rutherford 1886, 427, K–B vol. 1, 176–80, and Kretschmer 1909, 51–2; a previously unnoticed comic occurrence is Pherecr. fr. 211 [= Phryn. PS fr. † *67 = Σ^b α 310, Phot. α 287, ex ΣʹʹʹPhryn. PS fr. † *67 (= Σ^b α 310, Phot. α 287, ex Σʹʹʹ)]). Regarding the preverb κατα‑κατά in particular, the apocope is attested abundantly in epic, lyric, and tragedy, as well as in dialogues (a supplementary list of examples, which complements that of K–B, can be found in Arnott 1996, 803–4).

The verb καταμύω ‘to close the eyes’ (LSJ s.v.) is not a common word. It is one of the comparatively rare cases in which the preverb κατα-κατά has a telic function (i.e. indicates the completion of an action; see Hernández Socas 2020, 176 n. 111). The compound καταμύω has the specialised meaning of having the eyes closed, as opposed to the simple verb μύω, which can also indicate closing the mouth or closing an object. As regards classical sources, καταμύω only occurs in Herodotus (4.71.2), Aristophanes (V. 92), Hippocrates (Epid. 7.1.83), Xenophon (C.2, Cyn. 5.11), Alexis (C.1), and Menander (Dysc. 113, on which see Handley 1965, 151).

Alexis’ fragment (C.1) is the first (indirect) evidence for καμμύω, but it remains unclear what use he may have made of this form – whether as a parody of high language or foreign speech, or perhaps as an imitation of popular speech (see Arnott 1996, 804; Lüttel 1981, 175 argues in favour of seeing καμμύω as an element of colloquial language)Colloquial language. It is an intriguing possibility that καμμύων may already occur in a passage of Xenophon’s Education of Cyrus (C.2), but the matter is difficult to settle (see F.1). After the Hellenistic period, καμμύω becomes more common. What makes καμμύω a special case is the fact that apocopeApocope is not a typical development of koine Greekkoine. This raises the question of how καμμύω became so popular in post-classical times. One possibility is that, as evidenced by Alexis (and possibly Xenophon) and as argued by Lüttel (1981, 175), καμμύω had already developed in 4th-century Attic, perhaps at a colloquial or popular level and possibly as an AllegroformColloquial language. This is an intriguing scenario, but, given the lack of supporting evidence, it must remain speculative. Alternatively, καμμύω may be regarded as an import from a different dialect, e.g. Ionic, but the fact that we have no evidence for καμμύω in Ionic texts makes this explanation no more compelling.

The earliest post-classical occurrences of καμμύω are in the Septuagint (3x in Isaiah)Septuagint and the New TestamentNew Testament (Mark’s Gospel and the Acts of the Apostles) and then in several Christian authors, even in those who normally use a rather formal variety of Greek (e.g. Origen and Eusebius; although note that Synesius only uses καταμύω). Pagan writers regularly use καταμύω (to mention but a few: Strabo, Plutarch, Lucian, Philostratus, and Marcus Aurelius), which is also registered several times by Pollux (see esp. 2.67Poll. 2.67; other occurrences at 5.72, 9.113, and 9.117), who never comments on καμμύω. The variant καμμύω is used by technical writers, including Hero (Aut. 22.1), pseudo-Galen (19.439.4 Kühn), and Oribasius (Syn. 8.48.3). As regards writers who normally use a more formal language, the occurrence of καμμύω in two passages of Philo is noteworthy (De somniis 1.164 and Fr.Ex. fr. 31 Petit), all the more so since in the remaining cases Philo uses καταμύω form (9x) (at least in principle, one might envisage the possibility that these two occurrences are in fact Byzantine slips). However, Cyril of Alexandria similarly alternates between the two forms. EusebiusEusebius is an interesting case. In his corpus, καμμύω is standard (15x), and the only instance of καταμύω is in a passage of the Praeparatio Evangelica (C.3), where it is very likely a feature of marked language: it comes after a Homeric quotation and appears within a rhetorical question addressed to Homer, the tone is pathetic, and the passage is elaborately written. Adopting the classicising form καταμύω in place of the more common variant καμμύω would be, therefore, a means of heightening the tone.

E. Byzantine and Modern Greek commentary

Both καταμύω and καμμύω are in use in Byzantine Greek, although καμμύω remains far better attested than καταμύω. Learned writers sometimes alternate the two forms. To mention only three cases (and excluding poetic texts, where using one or the other form has a metrical motivation), καμμύω occurs 10x and καταμύω 6x in Photius, 1x and 1x in Arethas, and 7x and 4x in Eustathius. In such cases, the two forms seem to be treated as equally acceptable. Occasionally, using καταμύω may be a deliberate stylistic choice reminiscent one of the passages in Eusebius discussed above (C.3, for which see D.). One example appears in a letter of Theophylact of Ohrid (C.4): although Theophylact uses καμμύω elsewhere in his writings (6x), his use of more classical form καταμύω in this case may be explained in the light of the rhetorically sustained style and language of the epistle (the form also conveniently enables an interval of four syllables in clausula). Occasionally, καταμύω is the only attested form, e.g. in Symeon Metaphrastes and, much later, in Joseph Bryennius. Regarding the appraisal of these forms in Byzantine scholarship – Photius’ entry (B.3) and Eustathius’ commentary on the Iliad (B.4) – see respectively F.2 and F.3.

F. Commentary on individual texts and occurrences

(1)    X. Cyr. 8.3.27–28 (C.2)

Xenophon is the only Attic prose writer who makes use of the apocopated formApocope, although not all instances are unproblematic (for a list, see Gautier 1911, 77). This passage in the Education of Cyrus is a case in point. The manuscripts VHAB (the z family) have the reading κἂν μύων, as opposed to κἂν καμμύων in the remaining manuscripts WDF (the y family). All editors of the Education of Cyrus print κἂν μύων (see Marchant 1900–1920 vol. 4, Gemoll–Peters 1968, Delebecque in Bizos–Delebecque 1971–1978 vol. 3). The participle καταμύων also occurs at close proximity in Cyr. 8.3.28, and all manuscripts share this reading. Thus, editors evidently consider κἂν καμμύων as a Byzantine slip (possibly favoured by preceding κἄν) and also seek to avoid the coexistence of καμμύων and καταμύων. That καμμύω may a Byzantine slip is unproblematic (one may compare Philostr. VA 4.10.147, where, teste Lobeck 1820, 340, the manuscripts have the reading καμμύνειν). Nonetheless, καμμύων needs to be re-evaluated. Firstly, μύω semantically overlaps with καταμύω, but it is only the latter that properly indicates closing the eyes (see D.); hence, the inconsistency of this passage, whereby μύω is used first and then καταμύω second, raises some suspicions. Secondly, it is perhaps easier to postulate that κἂν καμμύων was corrupted by haplography into κἂν μύων than to argue that κἂν καμμύων developed by dittography from κἂν μύων. Thirdly, the coexistence of καμμύων and καταμύων is less surprising if one considers that the first occurs in Cyrus’ dialogue, which is somewhat ironic and hyperbolic, while the second is part of the narrative. Thus, καμμύων may be an element of informal languageColloquial language which is meant as an expressive nuance in the dialogue scene.

(2)    Phot. ε 1641 (B.3)

The Atticist origin of this gloss is very likely, but identifying the source behind it remains elusive. There are two options. The first is that the gloss in Phrynichus is the source of that in Photius, as Theodoridis (ad loc.) tentatively suggests. Should we then imagine a larger redaction of the Eclogue than the one we ourselves possess? The hypothesis that the surviving redaction of the Eclogue is actually an epitome has recently been explored in more in detail (see Tribulato, forthcoming). In support of the argument that this particular gloss in Photius depends on Phrynichus’ Eclogue, it is necessary also to consider the fact that Photius’ Λεσβιακόν is sufficiently paralleled in Phrynichus; that is, the Αἰολεῖς)Aeolic (on this category as representing Eastern Aeolic tout court, see Tosi 1988, 174) are occasionally mentioned in the Eclogue as a polemical target (in Ecl. 272Phryn. Ecl. 272 the condemned form νίτρον is attributed to the Αἰολεῖς, in opposition to the approved Attic variant form λίτρον; in Ecl. 332Phryn. Ecl. 332, the Αἰολεῖς are mentioned among speakers of other varieties of Greek which contribute to the mixed and barbaric language that the aspiring Atticist must shun). This being the case, we might then postulate that Phrynichus initially described καμμύω as an example of LesbianLesbian apocope and went on to say that Alexis’ καμμύω provides insufficient justification for considering this to be an Attic form also. The description of καμμύω as a Lesbian apocope finds justification in the relative frequency of this development in Lesbian poetry (see K–B vol. 1, 178–9). Since Photius discusses ἐπιμύω and καταμύω together (for the association of these forms, see also Poll. 2.67, Hsch. ε 4991, Su. ε 2481, and Et.Gud. 506.20), we would also need to postulate that Phrynichus’ entry originally compared καταμύω with ἐπιμύω as a way of explaining that καταμύω is a compound in order to prepare the ground for criticising Alexis’ use of καμμύω. This interpretation requires a considerable amount of speculation. The alternative option, i.e. that Photius had access to other, unidentified and unspecified, Atticist sources discussing καταμύω and καμμύω, seems a more balanced possibility.

(3)    Eust. in Il. 1.316.24–317.10 (B.4)

Eustathius explains that in compound verbs augment and reduplication are added internally, that is, between the preverb and the verbal stem. In the case of ἐπενήνοθα and καμμύω, however, the preposition has become part of the stem, and therefore the augment and reduplication may also be added before the preposition. The first part of this passage, which concerns the etymology of Ἀχέρων and the formation of ἐπενήνοθα, is well paralleled in the grammatical tradition (duly collected by van der Valk ad loc.). Yet, the source of the excursus on καταμύω and καμμύω is unknown. Van der Valk (ad loc.) aptly compares Apoll.Dysc. Synt. 4.43 = GG 2,2.469.9–470.10 (for which see also D.). Apollonius defends the legitimacy of adding prefixes (i.e. augment and reduplication) before the preposition in compound verbs on two accounts: (1) there is no rationale requiring that compound verbs such as καταμύω/καμμύω follow the same patterns as the other compound verbs; (2) the verb καθίζω does not admit of the internal addition of the augment (this is obviously incorrect, but Apollonius may be referring to the Greek of his own time period), and thus any other verb which is similarly formed behaves accordingly. To conclude, both Apollonius and Eustathius argue that, in forms like καμμύω, the preposition must be considered as an integral part of the verbal stem and not as an addition. Although Eustathius is not drawing from Apollonius directly, the doctrines that he had access to may represent a slightly simplified application of Apollonius’ principles.