Lexicographic entries

χειρσί, χερσί
(Phryn. Ecl. 115)

A. Main sources

(1) Phryn. Ecl. 115: χειρσὶν ἀδοκίμως, χερσὶ δέ.

χειρσί (‘with the hands’, dat. pl.) [is] not approved, [one must say] χερσί instead.

B. Other erudite sources

(1) Epim.Hom. χ 30 (~ Hdn. Περὶ παθῶν GG 3,2.277.4–32): χειρός· κατὰ ἀφαίρεσιν τοῦ ε γέγονε παρὰ τὸ ἔχειν, ὡς παρὰ τὸ ἑνὶ εἴκειν κατὰ Σιμωνίδην νίκη καὶ ‘†ῥόδην συνηυλίσθην†’ Ἱππῶναξ. μονῆρες δέ ἐστιν· οὐδὲν γὰρ εἰς ειρ λήγει μονοσύλλαβον θηλυκόν· πρόσκειται ‘θηλυκόν’· ἔστι γὰρ Εἴρ (ποταμός, οὗ μέμνηται Ἀρτεμίδωρος ἐν Γεωγραφουμένοις) καὶ ΦΘΕΙΡ, εἴτε ὁ ἥρως εἴτε τὸ ζῶον εἴτε ὁ καρπὸς τῆς πίτυος. ἡ κλίσις διττή, χειρός καὶ χερός· ἀναλογώτερος δὲ ἡ τὴν δίφθογγον ἔχουσα λόγοις τρισί· 1) τῷ ἀπὸ γενικῆς, 2) τῷ ἀπ’ εὐθείας, 3) τῷ ἀπὸ στοιχείων. 1) ἀπὸ γενικῆς οὕτως· πᾶσα γενικὴ εἰς ρος λήγουσα <παρὰ ἀρσενικὰ> ἢ θηλυκὰ ὀνόματα μονοσύλλαβα ἀποβολῇ τοῦ ος ποιεῖ τὴν ὀνομαστικήν· ψαρός ψάρ, κηρός κήρ· ‘φωρὸς δ’ ἴχνια φὼρ ἔμαθε’· θηρός θήρ, Νωρός Νώρ (ἔθνος)· οὕτως χειρός χείρ· ἡ χέρ ἄλεκτος. 2) ἀπὸ δὲ εὐθείας οὕτως· πᾶσα εὐθεῖα μονοσύλλαβος εἰς ρ λήγουσα ἀρσενικοῦ ἢ θηλυκοῦ ὀνόματος προσθέσει τοῦ ος τὴν γενικὴν ποιεῖ. 3) ἀπὸ στοιχείων δὲ οὕτως· τηνικαῦτα ἀποπίπτει φωνῆεν ἐν τῇ κλίσει ὅτε εἰς ς λήγει ἡ εὐθεῖα· κτείς κτενός, εἷς ἑνός, βοῦς βοός, νεύς νεώς, γραῦς γραός· τὸ δὲ χείρ εἰς ρ λήγει, οὐκ εἰς ς· ἐκ πάθους ἄρα τὸ χερός ἦν. ἰστέον δὲ ὅτι ὁ ποιητὴς φυλάξας τὴν δίφθογγον ἐν παντὶ ἀριθμῷ καὶ ἐν πάσῃ πτώσει, οἷον ‘αἱματόεσσα δὲ χεὶρ’ καὶ ‘χειρὸς ἑλόντε’ καὶ ‘χειρί τέ μιν κατέρεξεν’ καὶ ‘χεῖρ’ ἕλε δεξιτερὴν’ καὶ ‘σπόγγῳ δ’ ἀμφὶ πρόσωπα καὶ ἄμφω χεῖρ’ ἀπομόργνυ’ καὶ ‘χεῖρές τ’ ὀφθαλμοί τε’ καὶ ‘χειρῶν δ’ ἁψάσθην’ καὶ ‘χεῖράς τ’ ἀλλήλων λαβέτην’, διήλλαξεν ἐν τῇ πληθυντικῇ δοτικῇ ‘χερσὶ μὲν οὔ τοι ἔγωγε μαχήσομαι’, οὐκέτι δὲ χειρσί. τοῦτο δὲ ἐν παραγωγῇ οὐ χέρεσσιν εἶπε, ἀλλὰ χείρεσσι· ‘τοίω τὼ χείρεσσι {ὑφ’ ἡμετέρῃσι}’. καὶ σχεδὸν ἐμήνυσεν ἡμῖν τὸ αἴτιον· οὐδέποτε γὰρ μετὰ τοῦ χ εὑρίσκεται δίφθογγος καταλήγουσα εἰς ρ συμφώνου ἐπιφερομένου, οἷον χάρμα, ‘αὐλῆς ἐν χόρτῳ’, ‘χέρνιβα δ’ ἀμφίπολος’. {καὶ πῶς ἀναλυουμένου τούτου τοῦ λόγου} καὶ ἐπὶ τοῦ φ· καὶ γὰρ οὐδὲ μετὰ τοῦ φ εὑρίσκεται δίφθογγος καταλήγουσα εἰς ρ συμφώνου ἐπιφερομένου, οἷον φάρμακον, ‘φόρτιδος εὐρείης’, ‘στῆθος καὶ χείλεα φύρσω’, ‘κείμενον ἐν φέρτρῳ {φαρμα}’. <καὶ πῶς ἀναλυομένου τούτου τοῦ λόγου> ἡ φθειρσί δοτικὴ συνέστη παρὰ Ἀρχιλόχῳ, ‘φθειρσὶ μοχθίζοντα;’ λέγομεν οὖν ὅτι ἡ χείρ ἐπὶ τῆς γενικῆς καὶ χειρός καὶ χερός κατὰ ἀποβολὴν τοῦ ι, ἡ δὲ φθειρός οὐ διὰ τοῦτο {οὐδὲ} <φθειρσί καὶ> φθερσί.

ῥόδην συνηυλίσθην cod. : ’ρῳδιῷ (from ἐρῳδιῷ) κατηυλίσθην Hippon. | <παρὰ ἀρσενικὰ> Lentz (Hdn. Περὶ παθῶν GG 3,2.277.7) | The form νεύς can be interpreted either as the river’s name Νεύς, Νέως or as the noun νεῦς, νεώς (‘ship’): cf. [Arcad.] 274.10–2 (~ Hdn. Περὶ καθολικῆς προσῳδίας GG 3,1.400.27–30): τὰ εἰς ΕΥΣ μονοσύλλαβα ὀξύνεται· Ζεύς, Φλεύς (ὁ Διόνυσος), Νεύς (ποταμός), Δνεύς. τὸ μέντοι νεῦς (ἡ ναῦς) περισπᾶται, ὥσπερ καὶ τὸ γρεῦς | καὶ πῶς ἀναλυομένου (ἀναλυουμένου cod., corrected by Cramer, AO 1.441) τούτου τοῦ λόγου is transposed by West (Archil. fr. 236) before ἡ φθειρσί δοτική | πῶς Dyck : πως West | West suggests correcting διά in δύναται (i.e. οὐ δύναται τοῦτο, οὐδὲ φθερσί), Lentz (Hdn. Περὶ παθῶν GG 3,2.277.32) writes instead οὐ διὰ τοῦ ε οὐδὲ φθερσί | Cf. Hdn. Περὶ ὀρθογραφίας GG 3,2.429.23–9.

χειρός (gen. sing.): [It derives] from ἔχειν (‘to have’) by aphaeresis of ε, as [does] νίκη (‘victory’) from ἑνὶ εἴκειν (‘to allow only a single man’) according to Simonides (fr. 519.79c.2 PMG = fr. 4.12 Poltera) and ‘with the heron (’ρῳδιῷ) I pitch the tent’ in Hipponax (fr. 16.2 West², cf. apparatus). It is an anomalous [noun], for no other feminine monosyllabic [noun] ends in -ειρ. [The specification] ‘feminine’ has been added because there are Εἴρ (a river, which Artemidorus mentions in the Geographic Description [fr. 80, GGM 1.574–6]) and ΦΘΕΙΡ, [which denotes] the hero (i.e. Φθίρ, cf. e.g. schol. [D] Hom. Il. 2.868 (A)), the animal or the fruit of the pine (i.e. φθείρ). The inflection [is] twofold, χειρός (gen. sing.) and χερός, but the [form] with the diphthong is more regular on the grounds of three arguments: 1) the [argument based] on the genitive, 2) the [argument based] on the nominative, 3) the [argument based] on spelling. 1) The [reasoning based] on the genitive [is] the following: every genitive of the <masculine> or feminine nouns ending in -ρος forms the nominative [case] by removing -ος, [like] ψαρός ψάρ (‘starling’), κηρός κήρ (‘doom’); ‘a thief recognises the tracks of [another] thief’ (cf. Call. Epigr. 43.6 Pfeiffer); θηρός θήρ (‘beast of prey’), Νωρός Νώρ (the people); likewise, χειρός χείρ. One cannot say χέρ. 2) The [reasoning based] on the nominative [is] the following: every monosyllabic nominative ending in -ρ of a masculine or feminine noun forms the genitive by adding -ος. 3) Finally, the [reasoning based] on spelling [is] the following: a vowel drops out in the inflection every time the [nominative] case ends in -ς: κτείς κτενός (‘comb’), εἷς ἑνός (‘one’), βοῦς βοός (‘ox’), νεύς νεώς (cf. apparatus), γραῦς γραός (‘old woman’). But χείρ ends in -ρ, not in -ς; then, through a modification [the genitive] became χερός. One should know that the poet (i.e., Homer), retaining the diphthong in every number and in every case – such as ‘the bloody hand (χείρ)’ (Il. 5.82), ‘take by the hand (χειρός)’ (cf. Il. 1.323), ‘(s)he strokes him with his/her hand (χειρί)’ (e.g. Il. 1.361), ‘he clasped her right hand (χεῖρ(α))’ (Od. 1.121), ‘with a sponge he wipes his face and both his hands (χεῖρ(ε))’ (Il. 18.414), ‘the hands (χεῖρες) and the eyes’ (Il. 17.387), ‘they seized his hands (χειρῶν)’ (Il. 10.377) and ‘they took each other’s hands (χεῖρας)’ (Il. 6.233) –, changed [it] in the dative plural: ‘I certainly won’t fight with my hands (χερσί)’ (Il. 1.298 = C.1), and not χειρσί. But for this [form], [when it presents] an additional syllable, he did not say χέρεσσιν, but χείρεσσι: ‘in this way under the hands (χείρεσσι)’ (Il. 5.559). The reason may be clear enough: indeed, a diphthong which ends in -ρ followed by a consonant is never found after χ, [as is evident in forms] such as χάρμα (‘delight’), ‘in the enclosed place (χόρτῳ) of the court’ (Il. 11.774), ‘a handmaid [brought] water for the hands (χέρνιβα)’ (e.g. Od. 1.136). [The same happens] after φ, for a diphthong which ends in -ρ followed by a consonant is never found [after φ], [as is evident in forms] such as φάρμακον (‘drug’), ‘of a wide ship of burden/freight ship (φόρτιδος)’ (Od. 5.250), ‘I will stain (φύρσω) your breast and your lips’ (Od. 18.21), ‘lying on the bier (φέρτρῳ)’ (Il. 18.236). <And – to analyse the issue thoroughly – how is the dative φθειρσί (‘lice’) attested in Archilocus, ‘suffering from lice (φθειρσί)’ (fr. 236 West²)? Therefore, we say that χείρ [is inflected] in the genitive [as] both χειρός and χερός with the removal of ι, whereas for this reason the [genitive] φθειρός [is] not [inflected in the dative plural] <φθειρσί and> φθερσί.

(2) [Hdn.] Περὶ τῶν ζητουμένων 251.1–12: ἔτι ἁμαρτάνουσιν οἱ λέγοντες τὴν δοτικὴν ταῖς χειρσί, δέον ἄνευ τοῦ ι ταῖς χερσὶ λέγειν· καί τοι τῆς ἄλλης ἁπάσης κλίσεως τοῦ ὀνόματος τὴν δίφθογγον ἐχούσης, οἷον ἡ χεὶρ, τῆς χειρὸς, τῇ χειρὶ, τὴν χεῖρα· καὶ τῶν δυϊκῶν ὁμοίως, καὶ τῶν πληθυντικῶν ἐχόντων τὴν δίφθογγον, διὰ τοῦ ε μόνου κλίνεται, ταῖς χερσὶν, ἡ δοτικὴ πληθυντικὴ, μόνη γὰρ οὐκέτι ἔχει τὴν δίφθογγον· λέγομεν οὖν ὅτι αἱ μὲν ἄλλαι πτώσεις, ἐφύλαξαν τὴν δίφθογγον διὰ τὸ φωνῆεν ἐπιφέρεσθαι μετὰ τοῦ ρ, χειρὶ, χεῖρα, καὶ χειρῶν· ἐνταῦθα δὲ συμφώνου ἐπιφερομένου τοῦ σ, οὐκέτι ἡ δίφθογγος ἔμεινε χειρσὶν, ἀλλ’ ἀφαιρεθέντος τοῦ ι, ἐῤῥήθη χερσί· ἀντίκειται τῷ κανόνι τὸ φθειρσὶν, ὅτι καὶ τοῦ σ ἐπιφερομένου ἡ δίφθογγος ἔμεινε· μόνως οὖν ῥητέον ταῖς χερσί.

Further, those who use the dative χειρσί are wrong: one must say χερσί without ι. Αll the rest of the declension of the noun has the diphthong, that is χείρ, χειρός, χειρί, χεῖρα; the dual and the plural cases having likewise the diphthong, but the dative plural is inflected with ε alone, χερσίν, for it is the only [case] which does not have the diphthong. Therefore, we state that all the other cases have retained the diphthong because ρ is followed by a vowel: χειρί, χεῖρα, and χειρῶν. In this case, instead, since it is followed by the consonant σ, the diphthong [in] χειρσί is not retained, but the ι being removed, [it] is pronounced χερσί. φθειρσί does not follow the rule, for the diphthong is retained even though [ρ] is followed by σ; in conclusion, only χερσί can be used.

(3) Choerob. in Theodos. GG 4,1.278.11–21 (~ Hdn. Περὶ κλίσεως ὀνομάτων GG 3,2.737.12–21): ἔτι δεῖ προσθεῖναι ἐν τῷ κανόνι τῷ λέγοντι, ὅτι ἡ παραλήγουσα τῆς δοτικῆς τῶν πληθυντικῶν οὐ θέλει εἶναι ἐλάττων τῆς παραληγούσης τῶν ἄλλων πτώσεων, ἀλλ’ ἢ ἴση ἢ μείζων, ‘καὶ χωρὶς τῆς χερσί δοτικῆς τῶν πληθυντικῶν’· ἐνταῦθα γὰρ ἐλάττων ἐστὶν ἡ παραλήγουσα τῆς δοτικῆς τῶν πληθυντικῶν τῆς παραληγούσης τῶν ἄλλων πτώσεων· ἡ μὲν γὰρ παραλήγουσα τῆς δοτικῆς θέσει μακρά ἐστιν, ἡ δὲ παραλήγουσα τῶν ἄλλων πτώσεων φύσει μακρά ἐστιν, οἷον χεῖρες χειρῶν χερσί χεῖρας χεῖρες, τὸ δὲ φύσει μακρὸν μεῖζόν ἐστι τοῦ θέσει μακροῦ· διατί δὲ χερσί μόνως λέγομεν τὴν δοτικὴν τῶν πληθυντικῶν καὶ οὐχὶ χειρσί διὰ τῆς ει διφθόγγου, ἐν τῷ Ὀνοματικῷ Ἡρωδιανοῦ εἰ θεῷ φίλον μαθησόμεθα.

Further, we must add to the rules for the speaker that the penultimate [syllable] of the dative of the plural [nouns] cannot be shorter than the penultimate syllable of the other cases, but [it must be] equal or longer, ‘except for the dative plural χερσί’, for in this case the penultimate [syllable] of the dative plural is shorter than the penultimate [syllable] of the other cases. Indeed, the penultimate [syllable] of the dative is long by position, while the penultimate [syllable] of the other cases is long by nature, as [in] χεῖρες χειρῶν χερσί χεῖρας χεῖρες; but what is long by nature is stronger than what is long by position. For this reason, we pronounce the dative plural just χερσί and not χειρσί with the diphthong ει, [as] we will learn, God willing, in Herodian’s work on the nouns.

C. Loci classici, other relevant texts

(1) Hom. Il. 1.298–9:
χερσὶ μὲν οὔ τοι ἔγωγε μαχήσομαι εἵνεκα κούρης
οὔτε σοὶ οὔτέ τῳ ἄλλῳ, ἐπεί μ’ ἀφέλεσθέ γε δόντες

By might of hand I will fight for the girl neither with you nor with any other, since you are only taking away what you gave. (Transl. Murray 1924, 35).

(2) SGO 05/01/65 = GVI 1884, vv. 1–2 [Smyrna, 40 or 96 CE]:
νήπιος ἐν τύμβῳ τίς ἄρ’ ἐσθ’ ὅδε; ὡς ἀταλαῖσι
χειρσὶν γλακτοπαγεῖ μαστῷ ἔπι κέκλιτε.

κέκλιτε stands for κέκλιται.

Who is this child here in the grave? How he leans upon the milky breast with his tender hands.

D. General commentary

In the Eclogue (A.1), Phrynichus proscribes the form χειρσί, the analogical dative plural of χείρ (‘hand’) formed from the prevocalic stem χειρ-. Instead, he approves the form rendered with a short vowel, χερσί, which exhibits a regular phonological development. Other grammatical sources, largely traced back to Herodianic works (B.1, B.2, and B.3), address the same topic, agreeing with Phrynichus’ prescription. In the Epimerismi Homerici (B.1), the noun’s etymology is addressed along with its inflection: forms with the diphthong ει are considered more regular, with the exception of the dative plural, where the form χερσί is approved on the authority of a Homeric occurrence (C.1). This alleged anomaly in the inflection of χείρ is explained as follows: a diphthong followed by ρ and by an additional consonant is not found in the Greek language either after the letter χ or after the letter φ (with regard to the latter the exception constituted by the dative plural of φθείρ, i.e. φθειρσί, is mentioned, although the passage relating to the discussion of this form compared to the inflection of χείρ is affected by a textual corruption which makes its meaning unclear, see F.1). The Homeric form χείρεσσι is also quoted, as it appears to follow the rule just mentioned (i.e. the diphthong ει is admitted in the absence of a consonant after ρ); however, it must be noted that the alternative form χέρεσσι is also attested (cf. Hes. Th. 519 and 747), which appears to be implicitly disapproved in B.1.

The noun χείρ derives from an amphikinetic noun *ǵʰés-r̥ /*ǵʰs-r-és (cf. NIL 170–2). The prevocalic stem (Aeolic χερρ-, Doric χηρ-, Attic-Ionic χειρ-) developed from the weak stem *ǵʰes-r-. Starting from the nominative singular *ǵʰés-r̥t, one might expect the Greek reflex to be *χέαρ (likely preserved in Artemis’ epithet ἰοχέαιρα < *-χεαρ-jα, literally ‘who holds arrows in her hand’): the Attic form χείρ and the Strict Doric nominative χήρ are also likely analogical constructions based on the oblique cases (see Schindler 1967, 245; EDG s.v.). The analogical secondary forms in χερ-, such as χέρες, χέρας, χερῶν, and χεροῖν, developed from the dative plural χερσί (< *ǵʰes-r-su; on the origin of the dative plural, see most recently Viredaz 2000, who compares to it corresponding Armenian forms, such as the instrumental jerb); see also Chantraine (1961, 79). The allomorph χερ- is found also in the old compounds χέρνιψ (‘water for washing the hands’), χερνῆτις (‘a woman that spins for daily hire’) and χεδροπά (‘leguminous fruits’, probably from *χερ-δροπα); see Viredaz 2000, 293–4. Therefore, regarding the two forms of the dative plural, χερσί exhibits a regular phonological development, whereas χειρσί is an analogical formation constructed on the prevocalic stem.

Regarding the distribution of the dative form χειρσί in literary texts, it is extremely rare and is found only in Christian authors from the 4th century CE onwards: cf. Chrys. Fragmenta in Proverbia MPG 64.708.22–3 (χειρσὶ χεῖρας ἐμβαλὼν ἀδίκως οὐκ ἀθωωθήσεται, ‘be assured he will not go unpunished’), an inaccurate quotation from LXX Pr. 16.5 (ἀκάθαρτος παρὰ θεῷ πᾶς ὑψηλοκάρδιος, χειρὶ δὲ χεῖρας ἐμβαλὼν ἀδίκως οὐκ ἀθῳωθήσεται, ‘everyone who is arrogant in heart is an abomination to the Lord, be assured he will not go unpunished’) – the form χειρσί may perhaps be explained as a scribal or editorial error; Cyr.Al. Epistulae Paschales MPG 77.824.4; Callinic.Mon. V.Hyp. 28.8; Ign. Ep. 23.4.

Nonetheless, the form χειρσί is attested in inscriptionsInscriptions from the early imperial period onwards, especially in metrical texts: see e.g. C.2; I.Egypte métriques 19.7 [Alexandria, early imperial period]; IGUR 3.1197.5 [Rome, undated]. Two occurrences of χειρσί(ν) are also attested in papyriPapyri dated to the late imperial period: the first in P.Amh.Gr. 1.1, col. i.14 (= TM 64754) [provenance unknown, 5th century CE], a papyrus that contains some passages of the anonymous New Testament apocryphal text Ascensio Isaiae (cf. Geerard 1992, 195–7); the second in SB 1.5114.51 (= TM 23154) [Apollonopolis, 630–40 CE], although, in the latter case, χειρσί appears to be a mispelling of the dative singular χειρί.

In general, it appears likely that χειρσί spread in the koine from at least the imperial age onwards and that, for this reason, Atticist lexicographers proscribed the form.

E. Byzantine and Modern Greek commentary

In the medieval period, χερσί(ν) is the form regularly employed both in Byzantine Greek (e.g. Psellus, Constantinus VII, and Anna Comnene) and in Medieval Greek (e.g. Digenis Akritis cod. G 4.25), and it also survives as an archaising form in Modern Greek; see LKN s.v. χείρα.

F. Commentary on individual texts and occurrences

(1) Epim.Hom. χ 30 (B.1)

The last sentence of the entry of the Epimerismi Homerici concerning the inflection of χείρ has been subjected to a textual corruption: ἡ δὲ φθειρός οὐ διὰ τοῦτο οὐδὲ φθερσί. Dyck’s correction (1995, 745: ἡ δὲ φθειρός οὐ διὰ τοῦτο {οὐδὲ} <φθειρσί καὶ> φθερσί) does not appear to resolve the problem satisfactorily, given that is unclear to what διὰ τοῦτο (‘for this reason’) refers. West’s proposal seems more convincing: he corrects διά into δύναται (the exchange between the two words may have resulted from an abbreviation of the verbal form that the copyist misunderstood), reconstructing the text as follows: ἡ δὲ φθειρός οὐ δύναται τοῦτο, οὐδὲ φθερσί, ‘whereas the [genitive] φθειρός cannot [do] it (i.e., it cannot become *φθερός by the removal of ι), and [the form] φθερσί [is] not [used either]’. Finally, the correction made by Lentz in Herodian’s text (Περὶ παθῶν GG 3,2.277.32: ἡ δὲ φθειρός οὐ διὰ τοῦ ε οὐδὲ φθερσί) is equally satisfactory in terms of content but less likely from a palaeographic perspective.

Bibliography

Chantraine, P. (1961). Morphologie historique du grec. 2nd edition. Paris.

Dyck, A. R. (1995). Epimerismi Homerici. Pars Altera Epimerismos continens qui ordine alphabetico traditi sunt. Lexicon ΑΙΜΩΔΕΙΝ quod vocatur seu verius ΕΤΥΜΟΛΟΓΙΑΙ ΔΙΑΦΟΡΟΙ. Berlin, New York.

Geerard, M. (1992). Clavis apocryphorum Novi Testamenti. Turnhout.

Murray, A. T. (1924). Homer. Iliad. Vol. 1: Books 1–12. Translated by A. T. Murray. Revised by William F. Wyatt. Cambridge, MA.

Schindler, J. (1967). ‘Tocharische miszellen’. IF 72, 239–49.

Viredaz, R. (2000). ‘k’erb, jerb, χερσί’. HSF 113, 290–307.

CITE THIS

Elisa Nuria Merisio, 'χειρσί, χερσί (Phryn. Ecl. 115)', in Olga Tribulato (ed.), Digital Encyclopedia of Atticism. With the assistance of E. N. Merisio.
DOI: https://doi.org/10.30687/DEA/2974-8240/2024/03/003

ABSTRACT

This article provides a philological and linguistic commentary on the dative forms χειρσί and χερσί, discussed in the Atticist lexicon Phryn. Ecl. 115.