πήχεως, πήχεων
(Phryn. Ecl. 217, Moer. π 77, Philemo [Vindob.] 395.34, [Hdn.] Philet. 317)

D. General commentary

Atticist lexicography – from Phrynichus (A.1) and Moeris (A.2), to Philemon (A.3), the Philetaerus (A.4), and Orus (if we accept the identification of B.2 as a fragment of his lexicon) – discusses the noun πῆχυς (‘arm’; also as a measure of length, i.e. ‘cubit’; cf. LSJ s.v.) and, specifically, the different forms of its genitive singular (πήχεως, πήχεος, and πήχους) and plural (πήχεων and πηχῶν).

πῆχυς belongs to a small group of nouns with an ablauting stem in -ῠ-/-εϝ- (see also πέλεκυς ‘axe’, πρέσβυς ‘elder’, and ἄστυ ‘town’; υἱύς, i.e. the inherited form of the word meaning ‘son’, always appearing as υἱός in literary sources, also belongs to this group; cf. entry υἱεύς, υἱέως, υἱέα). In Attic, the declension of these nouns is influenced by the πόλις type (cf. e.g. Chantraine 1933, 118–20; Schwyzer 1939, 572; Rix 1992, 147; Sihler 1995, 320; van Emde Boas et al. 2019, 57–8, 299). Therefore, the genitive singular of πῆχυς is πήχεως (with proparoxytone accentuation, as in πόλεως, apparently against the law of limitation; cf. Dieu 2022, 433), while Ionic has the expected πήχεος (from πήχεϝ-ος). In the plural, the genitive is always πήχεων (regardless of the dialect) in analogy with the genitive singular (cf. Sihler 1995, 327; Dieu 2022, 434, 588–9); however, in Post-classical Greek, this spelling competes with the contract variant πηχῶν (attested from the 4th century BCE onwards; cf. below) and (possibly) with a paroxytone πηχέων (cf. below).

Considering that any evaluation of these forms’ distribution in literary texts is inevitably affected by potential alterations due to both mechanical errors in the transmission and intentional normalisations (by both copyists – for which, see also E. – and modern editors), the literary evidence on the genitive singular of πῆχυς can be summarised as follows: the occurrences of πήχεως in the classical period are limited, interestingly, to Hp. Epid. 5.1.65 Smith (= 5.242.21–2 Littré) = C.2, where we would expect the Ionic variant), while the form is far more frequent from the Hellenistic period onwards, with medical writings of all ages accounting for half of the form’s total occurrences in the TLG corpus (more than 600 hits, of which 241 are in Galen, 59 in Oribasius, and 19 in Theophilus Protospatharius). Regarding the Ionic genitive singular πήχεος, the total number of occurrences is approximately one third of πήχεως (a little over 200 hits in the TLG). Its first appearance is in Herodotus, where it occurs twice (Hdt. 1.178.3 = C.1 and 2.149.3). It is also found in Hippocrates (30x), it is the only form of the genitive singular of πῆχυς found in the Septuagint (14x), and it occurs in Galen (65x) (who, however, mainly uses the Attic πήχεως; cf. above).

Regarding the documentary evidence, a quantitative assessment is also problematic because, in many cases, the ending -εος or -εως has been supplied by the editors due to a material lacuna or (more frequently) an intentional abbreviation in the text (which is easily understandable as πῆχυς is a unit of measurement; see above). However, among the cases where the endings are preserved in full, the Attic genitive πήχεως is more frequent and has the older attestations (it occurs 11x in papyri between the 3rd century BCE and the 7th century CE, and it is the only form attested in inscriptions; cf. Mayser, Gramm. vol. 1,2, 25; Gignac 1981, 81 n. 1; Threatte 1996, 217), while πήχεος is less attested and appears later (5x in papyri between the 1st century CE and the 6th century CE, with no epigraphical occurrence). We should consider that the loss of vowel quantity distinction may have contributed to the oscillation between πήχεως and πήχεος in documentary texts (cf. e.g. Mayser, Gramm. vol. 1,1, 73–6; Gignac 1976, 325; Vessella 2018, 64–73).

Regarding the genitive plural, πήχεων is the oldest form (already found e.g. in Euripides, C.4, and Herodotus, 15x), while the contracted πηχῶν first appears in Arist. Mech. 853a.7 = C.5. Overall, in literary texts from the 3rd century BCE to the 1st century CE, πηχῶν appears to be much more frequent than the non-contracted form (approximately 400 hits vs. ca. 120 hits in the TLG). Notably, several authors use both forms interchangeably (e.g. Septuagint πήχεων 75x, πηχῶν 80x, cf. Blass, Debrunner 1976, 39 and n. 3; Flavius Josephus πήχεων 11x, πηχῶν 63x). Documentary texts confirm the antiquity of πήχεων – which is attested in inscriptions from the 4th–3rd century BCE onwards (cf. Threatte 1996, 220), while the first epigraphical attestations of πηχῶν date to the 2nd century BCE – and the pre-eminence of πηχῶν in koine Greek (occurring in more than 50 papyri dating between the 3rd century BCE and the 6th century CE, while πήχεων is found only nine times in papyri; cf. Gignac 1981, 81 n. 3). Regarding the potential existence of a paroxytone genitive πηχέων (whether etymological, i.e. from πηχέϝ-ων, or influenced by the nouns in -εύς, e.g. nom. sing. βασιλεύς, gen. plur. βασιλέων), this may have been in use at some point in the koine, together with other similar paroxytone genitive plurals, such as πολέων: according to Probert (2011, 277–9), forms such as πηχέων and πολέων may have existed as the ‘koine counterparts’ to πήχεων and πόλεων, before being ousted by them, also due to the prescription of the proparoxytone forms by the Atticists. Moeris provides the only evidence of such a prescription, discussing the paroxytone and proparoxytone accentuations of the genitive plural in relation to a noun of the same πόλις type, i.e. μάντις (‘seer’), and stating that μάντεων was the Attic (i.e. correct) form, while μαντέων was the ‘Greek’ form (see Moer. μ 12Moer. μ 12: μάντεων τὴν πρώτην ὀξυτόνως Ἀττικοί· τὴν δευτέραν ὀξυτόνως Ἕλληνες, ‘Users of Attic [employ] μάντεων with the acute accent on the first syllable. Users of Greek [employ μαντέων] with the acute accent on the second syllable’; see also B.5 and Theodos. Can. GG 4,2.41.14–6). The issue of the potential existence in koine Greek of paroxytone genitive plurals such as πολέων and πηχέων centres on the interpretation of this entry: we may either (with Probert 2011, 279) consider it as a testimony of the actual competition between two existing accentuations (i.e. the Attic proparoxytone vs. the paroxytone used in the koine), or (with Vessella 2018, 224–5) consider it a case where Moeris’ lexicon forces a piece of grammatical information (i.e. ‘the genitive plural of πόλις, μάντις, etc. is proparoxytone’) into the dualism Ἀττικοί (= correct usage) vs. Ἕλληνες (= incorrect usage), with μαντέων being a nonce-formation. Be that as it may, modern editions of Greek literary and documentary texts show an occasional oscillation between πήχεων and πηχέων owing to some Byzantine manuscripts presenting the paroxytone accentuation (cf. the reading πηχαίων – i.e. πηχέων – in cod. Laur. Plut. 55.4 [10th century CE] of Ascl. Tact. 5.1.5 and Oldfather et al. 1928, 270 n. 4). However, πηχέων remains a minor variant with only 54 hits in the TLG (cf. e.g. C.3).

Atticist lexica are unanimous in recommending the Attic genitive singular πήχεως (the prescription is found not only in Phrynichus (A.1) but also in the Philetaerus (A.4) and Photius (B.2) – possibly going back to Orus; cf. Alpers 1981, 248 – as well as in late antique and Byzantine grammarians (B.4, B.6, and B.7). Their prescription is reflected in the usage of Atticising authors of the imperial period – e.g. Lucian (2x), Pollux (5x), and Aelian (4x) – as opposed to lower-register texts, such as the Septuagint, which use only πήχεος (see above). Writers in the late imperial period also conform to this rule and mainly use πήχεως (3rd–5th century CE: πήχεως 112x, πήχεος 33x; for the distribution in late antique and Byzantine texts, see E.). Remarkably, no clear proscription of πήχεος is extant in Atticist lexica, while a third aberrant form of the genitive singular of πῆχυς, i.e. πήχους (analogical to s-stem nouns such as γένος, γένους) appears to have been actively rejected by some lexicographers and grammarians. Indeed, Photius’ entry (B.2; from Orus? Cf. above) and a passage by the grammarian Sophronius (B.6) attest to (and reject) the form πήχους. The fact that this form was actually in use in the koine of the imperial period appears to be confirmed by the attestations in two papyri from the 3rd–4th century CE – i.e. P.Oxy. 45.3245.14 (= TM 15910) [297 CE] = C.6 and P.Oxy. 64.4441.18 (= TM 23663) [315–6 CE]. The documents present several scribal errors, and we cannot exclude that, in both cases, πήχους is a misspelling of πήχυος, i.e. an analogical genitive singular modelled on nouns in -υ-, such as βότρυς, βότρυος (‘bunch of grapes’) or στάχυς, στάχυος (‘ear of corn’), which is in fact attested in P.Oxy. 2.242.15 (= TM 20551) [77 CE] and, spelt πήχοιος, in P.Oxy. 44.3195.45 and 47 (= TM 15958) [331 CE]; cf. Gignac (1981, 80–1). However, the comparison with B.2 and B.6 makes it plausible to assume that the spelling πήχους in C.6 and P.Oxy. 64.4441 was not an orthographic mistake but reflected an aberrant form of the genitive singular of πῆχυς actively used in koine Greek. This interpretation may also clarify a problematic detail in the text of Phrynichus’ entry on the genitives of πῆχυς (A.1) – namely, the second word of the lemma (i.e. the second word identified as ‘utterly non-Attic’ by the lexicographer): cod. U and family b read πήχως, while families c and q, as well as cod. T, read πηχός. While neither of the two readings is truly acceptable grammatically, Fischer (1974, 81) prints πήχως. The spelling πήχως is attested in a Ptolemaic papyrus – P.Cair.Zen. 4.59665.1 (= TM 1295) [Philadelphia, 3rd century BCE] – but this is clearly a mistake later corrected in πήχεος by the second hand who rectified several errors in the papyrus (cf. Edgar 1931, 102; Koenen 1971). The spelling πήχως may have resulted from an intended πήχους (on the erroneous replacement of ου with ω in Ptolemaic papyri, see Mayser, Gramm. vol. 1,1, 78–9); if this interpretation is correct, P.Cair.Zen. 4.59665 should be considered the first attestation of πήχους – much earlier than the other two extant occurrences (i.e. C.6 and P.Oxy. 64.4441, dating to the 3rd and 4th century CE, respectively). Undoubtedly, Phrynichus’ entry may become clearer if we assume that the variant readings πήχως and πηχός derive from an original πήχους (a rather simple palaeographic error in both majuscule and minuscule script). If this were the case, we would have two lexicographical sources (A.1 and B.2, the latter arguably depending on the former) that proscribe a form, i.e. πήχους, belonging to non-literary koine, as demonstrated by the attestations in C.6, P.Oxy. 64.4441, and – as argued above – P.Cair.Zen. 4.59665.

The Atticists’ stance (A.1, A.2, A.3, and A.4) regarding the genitive plural of πῆχυς is rather straightforward, as they unanimously condemn the post-classical contracted form πηχῶν and promote the older πήχεων. The same prescription is found in the Antiatticist and the Philetaerus regarding the s-stem noun ἄνθος; cf. entry ἀνθέων. Interestingly, some Atticising authors from the imperial period appear to confirm the Atticists’ preference for the non-contracted form (e.g. Lucian uses πήχεων three times and πηχῶν only once), while others primarily use πηχῶν (e.g. Aelian πήχεων 1x, πηχῶν 22x; Athenaeus πηχῶν 37x). Generally, there does not appear to be a clear correspondence between the distribution of πήχεων and πηχῶν and the different registers of literary Greek (and often the two forms coexist in the same texts, as noted previously for the Septuagint and Flavius Josephus). Overall, on a quantitative level, πηχῶν appears to be more frequently attested up to the Byzantine period. Accordingly, the diffusion in Medieval Greek of the genitive πηχῶν owing to the paradigm transfers undergone by πῆχυς (cf. E.) may have influenced not only the literary usage of the Byzantine period but also the Byzantine transmission of earlier texts, thus potentially altering the view of the actual ratio between πήχεων and πηχῶν for the modern scholar.

E. Byzantine and Modern Greek commentary

In literary texts from late antiquity and the Byzantine period, the more common form of the genitive singular of πῆχυς is πήχεως (6th–8th century CE: πήχεως 52x, πήχεος 28x; 9th–11th century CE: πήχεως 55x, πήχεος 25x), with πήχεος being a minor variant that is clearly declining from the 12th century (12th–14th century CE: πήχεως 72x, πήχεος 14x). Conversely, regarding the genitive plural, the Attic πήχεων is less widespread than πηχῶν up until the 12th century CE (6th–8th century CE: πήχεων 66x, πηχῶν 80x; 9th–11th century CE: πήχεων 111x, πηχῶν 147x), when it regains popularity against the contracted form (12th–14th century CE: πήχεων 96x, πηχῶν 39x). As mentioned previously (see D.), the predominance of πηχῶν over πήχεων in the first centuries of the Byzantine period may have been influenced by the linguistic developments occurring in Medieval Greek. Indeed, the noun πῆχυς could transfer to various paradigms: (1) masculine nouns in -ας, thus being inflected like the paroxytone nouns with a parisyllabic plural, such as ὁ μήνας/μῆνας (cf. the examples in Kriaras, LME s.v. πήχας, and CGMEMG vol. 2, 306), i.e. with gen. sing. πηχός or πηχοῦ(ς) and gen. plur. πηχῶ(ν); (2) feminine nouns in -α, such as χώρα (cf. the examples in Kriaras, LME s.v. πήχα, and CGMEMG vol. 2, 463), i.e. with gen. sing. πήχα(ς) and gen. plur. πηχῶν or πήχων; and (3) feminine nouns in -η, such as κόρη (cf. the examples in Kriaras, LME s.v. πήχη, and CGMEMG vol. 2, 536), i.e. with gen. sing. πήχης and gen. plur. πηχῶν. As can be observed, all three paradigms involve a genitive plural πηχῶν that may have contributed to the spread of the contracted genitive in literary Byzantine Greek (however, we should note that the genitive πήχεων is also occasionally retained in Medieval Greek; cf. CGMEMG vol. 2, 454).

F. Commentary on individual texts and occurrences

N/A