ἀλέω, ἀλήθω
(Phryn. Ecl. 121, Antiatt. α 46)

D. General commentary

Two entries from Phrynichus’ Eclogue (A.1) and the Antiatticist (A.2) discuss the admissibility of the verb ἀλέω (‘to grind’) and its derivative ἀλήθω, taking seemingly opposite positions: while Phrynichus – followed by Thomas Magister in the Byzantine period (B.6) – proscribes the use of ἀλήθω, the Antiatticist seems to promote the use of this form over ἀλέω. Whereas the Atticist prescriptions appear to be motivated by a straightforward canonical criterion, at a time when the two verbs were already perceived as equivalent and interchangeable, the distribution of the two forms in Classical and Hellenistic Greek appears rather to be conditioned by the syntactic contexts in which the action occurs, with different nuances of meaning that will be highlighted in each case. (I am grateful to one of the anonymous reviewers for drawing my attention to this point and for the subsequent analysis).

The verb ἀλέω probably derives from an originally athematic present (*ἀλε- < PIE *h₂elh₁-); see DELG, EDG s.v. The derivative ἀλήθω is formed by adding the suffix -θ--θ- (< *-dʰ-) to the lengthened verbal stem of ἀλέω. With regard to the semantic specialisation of the presents in -θω, particularly in Homeric Greek, these forms were often treated in the 20th century as ‘doublets’ for which it was difficult to identify a univocal function (see e.g. Chantraine 1925, 93–4, who identified an aspectual role for the suffix -θ- (‘terminatif’, ‘déterminé’); Schwyzer 1939, 703–4; Risch 1974, 278). Building on Benveniste’s insights, Magni (2004; 2008; 2010) more recently takes Homeric verbal pairs such as φλέγω/φλεγέθω as a starting point and maps the ‘detransitivising’ function of the suffix -θε/ο across a range of meanings and structures. She shows that presents in -θω fulfil an inchoative function – i.e., they express an action that excludes a causing agent and is presented as spontaneous – in opposition to the causative function of the original verbs. Magni (2010, 280) also discusses the multiple facets of intransitivity in relation to ‘indefinite object deletion’ and unaccusativity. Rothstein-Dowden (2022) corroborates these findings, arguing that the suffix *-dʰ- was used in PIE to derive intransitive active presents, perhaps originally athematic and subsequently thematised in almost all languages except Baltic. Regarding verbs in -ήθω deriving from contract verbs, Rothstein-Dowden (2022, 40–2) posits a chain of analogical changes involving νῶ (> νήθω), σῶ (> σήθω), κνῶ (> κνήθω), and ἀλῶ (> ἀλήθω), suggesting that ἀλήθω is a secondary formation from ἀλέω with the same transitive meaning, formed under the influence of σήθω ‘to sift’, given that the two verbs are often used together owing to the contiguity of the actions they denote. However, a proximity search in the TLG shows that these verbs occur together only twice in Hippocrates (Aff.Int. 1.64 (= 7.170.20 Littré): ἀλέσας λεπτότατα διασῆσαι; 40.25–6 (= 7.266.5–6 Littré): ἀλέσαι καὶ σῆσαι), and only rarely thereafter in late authors. In general, the claim that ‘the renewed forms in -ήθω-, [...] are an idiosyncratic innovation of historic Greek and have little to tell about the deeper history of the formant’ (Rothstein-Dowden 2022, 42) may be further explored through a systematic comparison of the usage of ἀλέω and ἀλήθω, in order to shed light on the survival and function of the suffix in this case.

If one examines the distribution of the two verbal forms, ἀλέω is attested as early as the Odyssey in the compound form κατ-αλέω (20.109: αἱ μὲν ἄρ’ ἄλλαι εὗδον, ἐπεὶ κατὰ πυρὸν ἄλεσσαν, ‘the others slept, since they had (already) ground the grain’). In the classical period, it is attested in Attic authors (Thucydides, C.1; Aristophanes, C.3; cf. also Arist. Oec. 1350b.9, Pr. 886b.11 and 964b.38, Thphr. De pietate fr. 2.39, Char. 4.10), in the Hippocratic corpus (5x), and in Herodotus (7.22.2). ἀλήθω is first attested in Hippocrates (C.4) and possibly in a line by Pherecrates (C.2; on this occurrence, see below); it is subsequently found in Theophrastus (CP 4.12.13), Herodas (6.81), the Septuagint (4x; ἀλέω and the compound καταλέω are used four more times), the New Testament (C.5 and Ev.Luc. 17.35), and Lucillius (AP 11.154.1 = 55.1 Floridi). With regard to documentary sources, it is attested in a papyrus dated to 199 CE (P.Oxy. 6.908.26 [= TM 20371]). In all these cases, as acutely noted by one of the reviewers, to whom I owe the following analysis, the activity of grinding may be conceptualised – depending on the contextual interaction between lexical and verbal aspect – as either a process of transformation or the accomplishment of a change in the state of an object. Taking, for instance, the occurrences of the two verbs in the Hippocratic corpus, ἀλέω is found to be consistently used in pseudo-directive contexts (i.e. instructions) that imply telic events affecting definite objects (cf. e.g. Fist. 7.2 Joly (= 6.454.15–6 Littré): λίνου σπέρμα πεφωσμένον ἀλέσας, ‘grinding roasted linseed’; Mul. 1.53 (= 8.112.10–1 Littré): ἀλεῖν δὲ ταῦτα καὶ ἀναφορύξαι ὄξει καὶ ἐλαίῳ, ‘to grind these (i.e. meal of raw corn, ashes of vine-twigs and linseed) and mix them up with vinegar and oil’). In C.4, however, ἀλήθω describes a different situation, in which a generic/impersonal agent (ἄνθρωπος) performs habitual, repeated, and non-telic activities (κόπτει, πλύνει, ἀλήθει, χρῆται), and the indefinite object (σῖτον), mentioned in the preceding context, is omitted. In this usage as a gnomic present, ἀλήθει can be compared to ἀλήθοντες in Thphr. CP 4.12.13: ὅτι τοῖς μὲν ἀλήθοντες χρώμεθα, τοῖς δ’ ὅλοις, ‘since we consume the former (i.e. cereals) by grinding, whereas the latter (i.e. beans and lentils) [are used] whole’. In this case, the participium conjunctum has no overt object. By contrast, in Char. 4.7 (κᾆτ’ ἀλέσαι μετ’ αὐτῆς τοῖς ἔνδον πᾶσι καὶ αὑτῷ τὰ ἐπιτήδεια, ‘and then he grinds with her (i.e. the baking-woman) the [grain] necessary for the entire family and for himself’), τὰ ἐπιτήδεια appears to function as the object of ἀλέω (the expression ἀληλεσμένον βίον, lit. ‘ground life’, i.e. ‘a life sustained by ground food’, in De pietate fr. 2.39 is highly peculiar). Analogous considerations regarding the aspectual nuance and syntactic environment of ἀλήθω seem plausible for the following passage from Agatharchides (26.16 [GGM 1.126]: οὕτως ἐζωσμέναι δυσπροσόπτως ὥστε μόνον τὴν αἰσχύνην τοῦ σώματος κρύπτειν, ἀλήθουσιν, ‘girded so wretchedly as to cover only their private parts, they grind’ (cf. D.S. 3.13.2). The intransitive nature of the verb is further confirmed by the above-mentioned Lucillius, AP 11.154.1–2: οὐκέτ᾿ ἀλήθει / ὡς τὸ πρίν, ‘he no longer turns the millstone as before’, P.Oxy. 6.908.26 (= TM 20371): [ἀ]λήθειν τ[ὰ κτ]ήνη ἡμερησίως, ‘the animals grind daily’, and the participial forms in the Septuagint (cf. e.g. Id. 16.21.4 (cod. Alexandrinus): ἦν ἀλήθων ἐν οἴκῳ τῆς φυλακῆς, ‘he turned the millstone in the prison’; Ec. 12.3: καὶ ἤργησαν αἱ ἀλήθουσαι, ‘the women who grind have ceased to work’) and in the New Testament (cf. e.g. C.5). Seemingly failing to take into account the different syntactic contexts just examined, Phrynichus (A.1) proscribes the verb ἀλήθω on the basis of its more recent attestations and, above all, on canonical grounds, as is evident in the case of the Hippocratic occurrence: none of the authors considered canonical by Phrynichus use the verb in -θω (with the possible exception of Pherecrates: see below). On the basis of the same canonical criterion, the entry of the Antiatticist (A.2), as transmitted by the cod. Par. Coisl. 345, appears to proscribe the verb ἀλέω, promoting instead the use of ἀλήθω. However, this prescription does not reflect the attestations of the two verbs in Attic texts, which are, on the contrary, consistent with Phrynichus’ prescription. Lobeck’s proposal to integrate <μόνον> in the Antiatticist’s entry (see the apparatus to A.2) is convincing, yielding a phrasing that finds a parallel in Antiatt. ν 13Antiatt. ν 13 (= Phot. ν 182): νήθειν· οὐ μόνον †νεῖν† (cod., Phot., corrupted; read νῆν instead; cf. Cobet 1858, 160 and the apparatus in Valente 2015, 220) <τὴν κρόκην φασίν> (‘νήθειν (‘to spin’): <They say> not only νῆν <the thread>’), another entry promoting the use of a derivative verb built with the suffix -θ-; see also Sicking (1883, 22–3); Tafel in ThGL vol. 1, s.v. ἀλήθω; Latte (1915, 374–5); Slater (1976, 241). On the use of οὐ in the interpretamenta of the Antiatticist’s lexicon, see Valente (2015, 48–9).

The Antiatticist’s defence of ἀλήθω would be justified by the occurrence of the verb in a fragment by Pherecrates (C.2), transmitted by Σ^b α 1442 (= Phot. α 2017, Su. α 2547, ex Σ´). According to Lobeck (1820, 151), the choice of the verb in -θω in this passage may have been motivated by the dactylic pattern of the verse. However, it should be noted that the comic poet uses the verb ἀλέω in two other fragments that offer a transitive context: in fr. 10.3 the structure is ἤλουν [...] τὰ σιτία, ‘they ground […] the corn’, and in fr. 197.2 ἀλέσαι occurs in a list of necessary actions (δεῖ) for the processing of barley (τὰς κριθάς). On the other hand, the idea that ἀλήθω functioned as the intransitive counterpart of ἀλέω could cast a different light on fr. 79 (C.2): the text may be interpreted as exhibiting a structure with ‘indefinite object deletion’, depicting the action of grinding as a durative or habitual activity (see the translation of the fragment offered by Storey in C.2). A metaphoricalMetaphors meaning (see Pellettieri 2024, 156) is also plausible, in light of the semantic evolution of ἀλέθω in Modern Greek (cf. E.). The presence of ἀλήθω in the fragment has been questioned on several occasions for various reasons – not least its otherwise non-Attic character (see Meineke, FCG vol. 2,1, 285; cf. also Bergk 1838, 295; for the fragment and its editorial history, see the comprehensive commentary in Pellettieri 2024, 155–7). However, if we accept that Pherecrates used the verb, on the basis that the syntactic context of the fragment differs from that of his other two usages of ἀλέω, he could then be the model to which the Antiatticist referred in its prescription, probably also mentioning the passage in question, which was subsequently lost in the process of epitomisationEpitome of the lexicon (for other cases in which explicit references to Pherecrates have been preserved in the Antiatticist’s lexicon, cf. Antiatt. γ 1Antiatt. γ 1, on which see entry γυναί; δ 36Antiatt. δ 36; θ 13Antiatt. θ 13; κ 18Antiatt. κ 18; ρ 1Antiatt. ρ 1). In fact, although Pherecrates is one of the authors of Old Comedy considered canonical by Phrynichus (see Tribulato 2024), the isolated occurrence of this form – compared to the two other occurrences of the verb ἀλέω, as noted above – may have led more rigorous Atticists to disregard the playwright’s use of ἀλήθω. Besides the issue of canon, the preference for ἀλέω may also be due to the fact that, in the 2nd century CE, ἀλήθω was widely used in koine Greek and the function of the suffix was no longer understood. It is also possible that, in spoken language, this form was already beginning to replace ἀλέω, through the same mechanism by which ἔσθω/ἐσθίω ‘to eat’ – as a generic activity without a definite object – gradually supplanted the older transitive form ἔδω (see Benveniste 1935,191; Magni 2010, 280; Batisti 2022, 291).

If we consider the attestations of the two verbs in Atticising authors, they appear in general to follow Phrynichus’ prescription: only ἀλέω is attested e.g. in Plutarch (13x), Arrian (3x), Dio Chrysostom (1x), and Aelian (2x). It should be noted, however, that occurrences of the verb are often found within quotations and are therefore not directly ascribable to the individual author. In Galen ἀλήθω occurs once (Ad Glauc. de med. meth. 2.6 = 11.108.6 Kühn: ὀνομάζουσι δ’ οὕτως ἐκεῖνον τὸν λίθον, ἐξ οὗ τὰς μύλας κατασκευάζουσιν, ἐφ’ ὧν ἀλήθουσι τὸν σῖτον, ‘They give this name (i.e., μυλίτης, ‘millstone’) to that type of stone from which they build the mills they use to grind grain’), as compared to 25 occurrences of ἀλέω. The frequency of ἀλέω in Galen’s treatises appears to reflect the recurrent use of past participles as adjectives referring to ‘ground’ ingredients. Compared to similar passages, the only example with ἀλήθω – followed by τὸν σῖτον – is nevertheless somewhat ‘suspicious’, insofar as the two verbs already seem interchangeable.

The lexica and the scholarly works of late antiquity and the Byzantine period gloss forms of the verb ἀλέω with the corresponding forms of ἀλήθω (B.1, B.2, B.5), confirming that the derivative in -θω had become widespread in Post-classical Greek. Particularly noteworthy is Hesychius’ entry (B.1), which glosses the 3rd-person singular present imperative of ἀλέω with the corresponding form of ἀλήθω: the former is attested only in a medical work dated to the 7th century CE (Paul.Aeg. 1.2.1.11), while the latter has no further attestations (note, however, that the form ἀλείτω is also attested as the 1st-person singular present indicative in several passages of scholarly works, which derive from it forms of ἀλιταίνω, ‘to sin’: cf. e.g. Eust. in Od. 2.54.22 Cullhed–Olson = 1.207.41 Stallbaum and schol. Od. 4.807d). Conversely, the Suda (B.3) and the Etymologica (B.4) treat the verb ἀλήθω independently, focusing on its meaning and derivation. In particular, the entry in the Suda (B.3) shows that the verb is by this stage perceived as fully equivalent to ἀλέω; it also provides a usage context that distinguishes ἀλήθω from the lemmas of the preceding and following entries (ἀλήθεια καὶ ἀληθής in Su. α 1175, and ἀληθίζω in Su. α 1177), which are graphically and phonetically similar to the verb in question.

E. Byzantine and Modern Greek commentary

Both ἀλέω and ἀλήθω continued to be used in the Byzantine period and are often used interchangeably by Byzantine authors, although it cannot be ruled out that, in some cases, this variation is due to manuscript transmission and editorial choices: cf. e.g. Theophylactus Enarrationes in evangelia 3.1077.34 ἀληλεσμένα, 4.1285.50 ἀλήσας vs. Enarrationes in evangelia 3.997.42 ἀλήθειν, Epistulae 89.22 ἀλήθειν; Michael Coniates Epistulae 30.23 ἀληλεσμένος vs. Epistulae 111.7 ἀληθόντων; Gennadius Scholarius Grammatica 2.439.20 ἀληλεσμένον vs. Grammatica 2.467.28 ἀλήθεται. It should be noted that the perfect middle-passive participle is generally the most common form of ἀλέω in Greek. In Μedieval Greek, the alternative form ἀλέθω also developed (cf. Kriaras, LME s.v.). ἀλέθω still survives in Modern Greek with the meaning ‘to grind’ and, metaphorically, ‘to chew and digest’ (cf. LKN, ILNE s.v.). The presence of ε instead of η is probably due to the influence of the aorist ἤλεσα of ἀλέω, which was also used as the aorist of ἀλήθω itself (see ILNE s.v.). Since the vowel in question is stressed, and the lowering of unstressed /i/ to /e/ in the adjacency of a liquid – frequent in Medieval Greek – would therefore not be apply, the change is best explained as a case of paradigmatic levelling by analogyAnalogy with the vocalism of the aorist; see CGMEMG vol. 1, 68–9.

F. Commentary on individual texts and occurrences

N/A