κράστις, γράστις
(Moer. κ 14, Antiatt. γ 13)

D. General commentary

Moeris (A.1) prescribes the Attic form κράστις (‘grass’, ‘green fodder’) and rejects the variant γράστις, which the Antiatticist (A.2) instead accepts. Similarly, the variant with κ was the only one accepted by the Atticist Aelius Dionysius (B.1), who rejected γράστις because of the lack of Attic literary attestations. At first glance, A.1, A.2, B.1, and B.3 may appear to discuss a problem of sonorisation of the initial velar (κράστις > γράστις) as a result of phonological development, but from a linguistic perspective the relationship between the two forms is probably the opposite. Indeed, despite the distribution of κράστις and γράστις in literary and documentary texts (see below) and the ensuing Atticist prescriptions, γράστις appears to be the ancient form, while κράστις should be considered a later development from it.

The assumption that γράστις is more ancient than κράστις is based on the possible derivation of the former from the almost unattested verb γράωγράω (‘to gnaw, to eat’, going back to PIE *grn̥s-, cf. Van Beek 2022, 387). However, according to Van Beek 2022, 387, ‘the etymological appurtenance of γράστις’ to the root of γράω ‘is uncertain’. The verb γράω is apparently typical of Cypriot GreekCypriot Greek, as suggested by Hsch. γ 876: γρᾶ· φάγε. Κύπριοι (‘γρᾶ: [I.e.] eat (imperative, 2nd pers. sing.). The Cypriots [use it]’). Another entry from the same lexicon (Hsch. κ 32) may point in the same direction (but the text καγρᾶ· καταφαγᾶς. Σαλαμίνιοι is highly conjectural, with the cod. actually reading κάγρακα· ταφυφὰς σαλμηροί). However, the idea that the verb γράω is typical of Cypriot Greek is seemingly confirmed by the imperative ka-ra-si-ti (= γράσθι?) in a 4th-century BCE Cypriot inscription from Golgoi (IChS 264.1; cf. Van Beek 2022, 386). The verb γράω seems to have been used once by Callimachus (C.3): this rather obscure line is quoted several times in lexica discussing not only the noun γράστις (cf. B.6), but also the adjective πολύγραος (‘who eats much’; cf. Gal. Gloss. π 44), the fish name γόγγρος (‘conger eel’; cf. Et.Gen. B s.v. γόγγρος [~ EM 237.45−50, Et.Sym. γ 136], tentatively ascribed to Philoxenus by Theodoridis 1976, 373), and the medical term γάγγραινα (‘gangrene’; cf. Et.Gud. d² 294.21−3, EM 219.28−32, Et.Sym. γ 5). But the same line is found, with the spelling ἔκραε instead of ἔγραε, in a scholium to the Iliad explaining the word ἀκρίς (‘cricket’). C.3 is thus not a fully reliable attestation of γράω.

Regardless of their etymology, both κράστις and γράστις are very poorly attested in Greek literature. Of the two, κράστις is the form with the earliest attestations. Indeed, according to Harpocration’s entry on the term κράστις (B.2), the word occurred once in Aristophanes (fr. 840 = C.1; on which see Bagordo 2018, 30−1) and once in Dinarchus (Or. 51 fr. 2 Conomis = C.2). As might be expected given its meaning, κράστις is also found in zoological and botanical writings: it occurs once in Arist. HA 595b.22−8 and three times in Theophrastus (HP 5.4.4, 7.5.3; CP 4.8.5; the manuscripts often misspell the word as κράσις or similar, but the first letter is always κ, with no sign of the spelling with γ). Moreover, a fragment from Sophron (fr. 165) attests to the verb κραστίζομαι, which describes the act of chewing leaves, probably as a form of magical protection: see Hordern (2004, 195). Later attestations of κράστις and κραστίζομαι are found exclusively in grammatical works and lexica (cf. A.1, B.1, B.2, B.3, and B.5). As far as the documentary evidence is concerned, κράστις is the normal form in papyriPapyri until the 2nd century BCE: see e.g. P.Grenf. 1.42.11 (= TM 266) [Diospolis Parva, Upper Egypt, 169−8 BCE]; cf. the verb ἐπικραστίζω in P.Tebt. 3.1.724.3 (= TM 5330) [175 BCE] and P.Oxyrhyncha 1.5 (= TM 644667) [152−1 BCE].

γράστις has a remarkably different distribution from κράστις: it is only attested in grammatical and lexicographical texts from the imperial to the Byzantine era, some of which condemn it as wrong (cf. A.1, B.1, B.3), while others accept it (cf. A.2, B.4, B.6). The only occurrences outside of lexica and other erudite works are in the Byzantine Hippiatrica, which also attest to the verb γραστίζω (‘to feed on grass’) and the noun γραστισμός (‘the act of feeding with grass’; see LBG s.v. and section E.). In documentary texts, however, γράστις appears much earlier, as early as the 3rd century BCE, e.g. in P.Lond. 7.2065.2 (= TM 1626) [unknown origin, mid 3rd century BCE], PSI 4.351.6 (= TM 2039) [Philadelphia, Arsinoites, 253 BCE], and P.Cair.Zen. 3.59341.b.5−c.6 (= TM 984) [Alexandria (?), after 247 BCE]. Moreover, the verb γραστίζω is found in P.Corn. 1.187 (= TM 2301) [Philadelphia, Arsinoites, 257 BCE] and P.Cair.Zen. 2.59158.2 (= TM 806) [Alexandria (?), 256 BCE]. From the 1st century BCE γράστις is the standard form in documentary papyri.

In sum, κράστις and γράστις (as well as (ἐπι)κραστίζω and γραστίζω) seem to coexist at least in documentary texts between the 3rd and 2nd century BCE. In literature, however, κράστις is clearly the more ancient form, while γράστις is only mentioned in grammatical texts (with the exception of the Byzantine technical collection of the Hippiatrica). Based on these data, one would expect γράστις to be a misspelling of an original κράστις, since, as Gignac (1976, 76−7) notes, ‘γ and δ interchange very frequently with κ and τ respectively throughout the Roman and Byzantine periods in all phonetic conditions […]. This reflects the identification of the voiced and voiceless phonemes by many speakers in a single velar /k/ and dental /t/ phoneme’. This phenomenon is attested in both inscriptions and papyri; for the interchange of -κρ- and -γρ- in particular, see the spellings ἀγροπόλει for ἀκροπόλει ‘to the acropolis’ in IG 2².508.11 [Attica, late 4th century BCE], κρυτωπωλῶν for γρυτοπωλῶν ‘of the sellers of small wares’ in BGU 1.9.12 (= TM 9131) [Arsinoites, 276 CE], and ἐν τῇ γρατίστῃ βουλῇ for ἐν τῇ κρατίστῃ βουλῇ ‘in the high council’ in P.Oxy. 1.55.12 (= TM 20717) [283 CE] (see Mayser, Gramm. vol. 1,1, 155; Threatte 1980, 556–7). Moreover, the word κωλακρέτηςκωλακρέτης, which designates an Athenian financial officer, undoubtedly derives from the verb ἀγείρω and must originally have been *κωλαγρέτης (cf. e.g. EDG s.v. κωλακρέται): the fact that the spelling with γ is unattested shows that the shift -γρ- > -κρ- must have occurred very early in the history of the word (cf. AGP vol. 2, Phonology, Section 2.2, forthcoming).

Nevertheless, if one accepts the derivation of γράστις from γράω (cf. above), then γράστις must be considered the more ancient form, and κράστις must be explained as a development from it. Frisk (cf. GEW s.v. γράω; followed by Beekes, EDG s.v.) considers κράστις to derive from γράστις through a ‘folk etymological connection with another (unknown) word’ (EDG s.v. γράω). On the other hand (and more plausibly), Solmsen (1909, 234–5; followed by Schwyzer 1939, 257; DELG s.v. γράω; Van Beek 2022, 387) explains κράστις as resulting from the early assimilation of the voiced onset of the first syllable γρα- to the voiceless onset of the second syllable -στις. In other words, a rare (perhaps typically Cypriot) word γράστις was at some point adopted by Attic Greek speakers, who changed its pronunciation by devoicing the first consonant (γράστις > κράστις). This must have happened before the earliest literary attestation of κράστις (in Aristophanes, C.1). Whether the spelling γράστις (attested by documentary papyri from the 3rd century BCE onwards and by lexicographers and grammarians from the imperial to the Byzantine period) reflects the survival of the original pronunciation or a sonorisation the first syllable of Attic κράστις in the koine cannot be ascertained.

No doubt, the Atticist grammarians considered κράστις to be the correct form. In all likelihood, Aelius Dionysius’ statement (B.1) that ‘Attic speakers [call] the grass κράστις’ was based on the lost Aristophanic occurrence C.1 (whether Dinarchus was also taken into account by him is open to debate, given Aelius’ rather cautious attitude towards less strictly canonical authors: see the entry on Aelius Dionysius, Ἀττικὰ ὀνόματα). Moeris (A.1, as well as Hesychius, B.3) shows the same preference for κράστις over γράστις (probably drawing from Aelius Dionysius, see Hansen 1998, 42−6). Pollux (7.142Poll. 7.142, 10.166Poll. 10.166) considers only the spelling κράστις (as do the Lexeis rhetorikai, B.5) and also mentions the derivative κραστήριονκραστήριον (a part of the manger made of four pieces of wood tied together). The same word, in the plural (κραστήρια), is found in Phrynichus’ Eclogue (Ecl. 149Phryn. Ecl. 149), but in the sense of ‘bed-posts’ (besides Phrynichus and Pollux, the term occurs only in Hsch. κ 3980 and Phot. κ 1057).

On the contrary, the Antiatticist (A.2) mentions only γράστις, stating that it refers to ‘moist barley’ and ‘grains’ (the equivalence with κριθή is found only here and in B.6, while all other erudite sources equate the term with either πόα ‘grass’ or χόρτος ‘fodder’); but the occurrence on which this statement is based is now lost. The spelling/pronunciation γράστις is then found in Orion (B.4): his doctrine is preserved in a more complete form in the Etymologicum Genuinum (B.6). The entry includes the quotation from Callimachus (C.3) which seems to attest to ἔγραε (in Orion the quotation is omitted, probably due to epitomisation). It cannot be excluded that the Antiatticist (A.2) (also) relied on Callimachus’ verb ἔγραε to support the spelling/pronunciation γράστις, but − given that the term evidently belonged to the common language of animal husbandry – one can also speculate that the authority presupposed by the entry was a comedy, in which the presence of everyday language is naturally prominent.

E. Byzantine and Modern Greek commentary

The Hippiatrica − a corpus of late-antique writings on hippiatric medicine probably compiled in the early 6th century CE, see Doyen-Higuet (1984), Lazaris (2020, 420) − contain almost all the extant occurrences of γράστις (20x), γραστίζω (6x), γραστισμός (1x) outside lexicographical and erudite sources (for attestations of these words in documentary papyri, see D.). A noun γραστήριον meaning ‘meadow’ (cf. LBG s.v.) is found in the Farmer’s Law (Νόμος γεωργικός), a Byzantine legal document of controversial dating and authorship (for an overview, see Humphreys 2015, 194–232). Apart from the Hippiatrica and Byzantine lexicography, neither κράστις nor γράστις have continuations in Greek literature.

F. Commentary on individual texts and occurrences

N/A