*g̑^hésor-

The etymological connection between Hitt. keššar, CLuw. īš(ša)r(i)-, HLuw. /istr(i)-/ and Lycian izre/i- has been generally accepted, see among others Starke 1987a:265, Hawkins; Morpurgo Davies 1988a:179, HED K:165, Kloekhorst 2008c:471. Regarding the phonological development, the Luwic languages show the common innovation of *g̑^h > *i̯ before a front vowel (Melchert 1994a:254). The raising of *é > í after *i̯ is also a common Luwic innovation (Melchert loc. cit.).Gusmani 1969c:314 agrees withScheller 1964a:41 that the plene spelling of the initial in Cuneiform Luwian īš(ša)r(i)- may indicate the sound sequence /i̯i-/, but Melchert (loc. cit.) finds it unlikely and prefers to assume the loss of *i̯ before /i/. In this case, the plene writing would represent a vowel lengthened under accentuation (cf. Simon 2010b:261f.), which is here accepted. The hypothesis of Hajnal (Hajnal 1995a:57, followed in Kloekhorst 2008c:472) that pre-tonic *e is weakened to /i/ in Luwian and Lycian is harmed by the fact that the vowel *e is preceded by *g̑^(h) in his examples.

The sequence -zr- in Lycian A izredi presents some phonological difficulties of relative chronology. For the phonetic quality of the Lycian in this particular sequence, cf. Melchert 2003c:153 n. 27 and Kümmel 2007a:286. Kümmel (loc. cit.) assumes a t-epenthesis in Pre-Lycian just as in Hieroglyphic Luwian /istr(i)-/ (< Proto-Luwian *ī́ssr(i)-). The cluster *-st- yields -s- and this -s- regularly escapes the Lycian A development of *s > h. Then, the development of s > z /_r could take place. Otherwise, it has been argued that the development of *sr > zr was earlier and no longer effective, when the Lycian A sound change *sr > hr applied, which occurred only after the syncope of a sequence of the type *sVr in any position of the word, cf. Melchert 1993a:250, Melchert 1994a:314f., 328, Hajnal 1998a:93. Both possibilities must be left open until further evidence.

The Hittite form kiššar retains a clear reflex of the initial palatalized velar *g̑^h, which must have undergone the regular fortition in word initial position and the loss of the palatal quality and of the aspiration (Melchert 1994a:94). The vowel in the first syllable is regularly spelled with the i-signs, which shows a weakening of *e > i in unaccented syllable (Rieken 1999a:280f., Kimball 1999a:137f.). Differently, Kloekhorst (Kloekhorst 2008c:472) interprets the vowel /i/ as a conditioned anaptyxis, which is not necessary and less economical. An Old Hittite occurrence of a nominative form ke-eš-šar (< *g̑^hés-ōr) represents the preserved quality of the vowel *e under accentuation (Rieken 1999a:280).

Regarding the accent-ablaut class, the Hittite lexeme for ‘hand’ preserves traces of the amphikinetic paradigm reconstructed for Proto-Indo-European; see Part 2 Proto-Indo-European Reconstruction. Kloekhorst (Kloekhorst 2008c:471f., Kloekhorst 2013a:113f.) on the other hand reconstructs a hysterokinetic paradigm. The Luwic attestations of the lexeme for ‘hand’ do not reveal any information so far regarding the accentual pattern of the Proto-Anatolian paradigm. First, Kloekhorst (ibid.) objects to the amphikinetic paradigm on the premises that the nominative *g̑^hés-ōr fails to explain the geminate -šš- in the Hittite nominative form ke-eš-šar. This argument cannot be maintained, since the gemination of the /ss/ may have been generalized throughout the paradigm based on the weak forms, such as *g̑^h(e)sr-éi̯ > ki-iš(-ša)-ri-i, where the sibilant was directly preceding the resonant /r/. For the sound law *-sr- > Hitt. -ssr- see Melchert 1994a:150f.

It is more likely that Luwian and Lycian had thematized with *-o- the original r-stem of the lexeme for ‘hand’ prior to the spread of the i-mutation (pace Starke 1990a:347; cf. Kloekhorst 2008c:472). This can be supported by the fact that original common gender r-stems have a different outcome in Lycian and in Luwian. While Luwian transfers them to the paradigm of the i-mutation, Lycian transfers them to the a-stems; cf. HLuw. /tuwatr(i)-(?)/ c. : Lyc. kbatra- c. ‘daughter’, HLuw. atr(i)- : Lyc. atra- c. ‘person, self’. This is not the case for Lyc. izre/i-, which is not an a-stem as can be seen from its ablative form in -edi, which is a clear indication that the word does not belong to the a-stems.

[D.S.]

PIE *g̑^hés-(o)r- f. means ‘hand’ (cf., e.g., EIEC:254, NIL:170f. with n. 3).

A. Inflection

According to a widely held view, PIE *g̑^hés-(o)r- f. was an amphikinetic r-stem noun (cf., e.g., Schindler 1967c:244–249, Rieken 1999a:261f., 280, Rieken 2004c:288, NIL:171 n. 3 with further lit.). Therefore, the following paradigm can be reconstructed for the protolanguage: PIE nom. sg. *g̑^hés-ōr, acc. sg. *g̑^hés-or-m̥, dat. sg. *g̑^h(e)s-r-éi̯, abl./gen. sg. *g̑^h(e)s-r-és, loc. sg. *g̑^h(e)s-ér±i. An alternative reconstruction of the original PIE paradigm was offered by Beekes 1985a:65 (partly followed by Kloekhorst 2008c:472, Kloekhorst 2013a:111–115; cf. also EDG:1623): nom. sg. *g̑^hes-r, acc. sg. *g̑^hes-er-m, gen. sg. *g̑^hes-r-os.According to Kloekhorst 2008c:472 and Kloekhorst 2013a:111–115, the Hittite material points to an original ‘hysterodynamic’ paradigm *g̑^hés-r̥, *g̑^hs-ér-m̥, *g̑^hs-r-ós/és (according to the ‘fourth subtype’ as described in Beekes 1985a:175). However, there is only little material in favor of an inflectional type where nominative and accusative show such a stem alternation. It is true, however, that the acc. sg. of the Hittite word for ‘hand’ (ki-iš-še-ra-an /kisséran/) deviates from the above reconstructed amphikinetic form *g̑^hes-or-m̥, which would regularly have given Hitt. †/késsaran/ (if from *g̑^hés-or-m̥) or †/kissáran/ (if from *g̑^hes-ór-m̥, with accent shift according to the *k^u̯etu̯óres-rule, i.e. PIE é-o-x > e-ó-x). The stem form ki-iš-še-r° = /kissér-/, which is also attested in the acc. pl. ki-iš-še-ru-uš, the gen. sg. ki-iš-še-ra-aš, and probably in the secondary and thematized nom.-sg.-form ki-iš-ši-ra-aš = /kisséras/, indeed points to an underlying stem *g̑^hes-ér-. Such a stem is furthermore presupposable for Tocharian, with PIE *g̑^hes-er- > PToch. *śäṣär- > *śṣär- > TochB ṣar ‘hand’ (DTB²:711f., Widmer 2004b:61, Schindler 1967c:249). Yet, due to *g̑^hes-ér-, it is by no means necessary to assume a new inflectional type, as such a stem form can also be explained from an amphikinetic paradigm: alongside the strong stem *g̑^hés-or- and the weak stem *g̑^h(e)s-r-´ a stem *g̑^hes-ér- can be reconstructed for the loc. sg. *g̑^h(e)s-ér±i. Cf., furthermore, Widmer 2004b:61, who points out that the word for ‘hand’ must have been frequently used in the locative (‘in[to] the hand’). As a consequence, the stem *g̑^hes-ér- could have spread analogically from the loc. sg. to other case forms such as the acc. sg., with PIE acc. sg. *g̑^hes-or-m̥ remodeled to pre-PAnat. *g̑^hes-ér-m̥.

B. Derivation

PIE *g̑^hés-or- f. ‘hand’ and *g̑^hé/ós-to- m. ‘id.’ (> PIIr. *j́^hásta- > Ved. hásta-, Av. zasta-, OPers. dasta-; cf. EWAia II:812 with lit.) are derived from a PIE root *g̑^hes-, for which Rix 1991a:230 reconstructs an original meaning ‘to grasp, seize (with the hollow hand)’. Eichner 2002a:156 assumes a broader meaning ‘to exchange sth.; to accept and return sth.; to balance sth.; to replace sth.’, which is not incompatible with the meaning ‘to grasp, seize’. Several IE languages continue forms that are based on the root *g̑^hes- and mean ‘thousand’ (cf. NIL:170, 172 nn. 15–18 with lit.): Gk. Att. χῑ́λιοι, Ion. χείλιοι < *g̑^heslii̯o-; Ved. sahásra- n., Av. hazaŋra- n. < *sm̥-g̑^héslo-, lit. ‘one thousand’; Lat. mīlle may derive from *smih₂-g̑^hsl-ih₂, a feminine derivative of *sm̥-g̑^héslo- ‘one thousand’ (Weiss 2011a:373f.). Rix 1991a:230f. suggests that the root of *g̑^hes-lo-, which underlies all the forms for ‘thousand’, is the same as the one found in *g̑^hésor- ‘hand’, with *g̑^hes-lo- originally meaning *‘handful (of very small things, e.g. seeds)’. But the semantic development is questionable. – Finally, PIE *g̑^hésor- f. ‘hand’ forms part of an already PIE prepositional phrase *me g̑^heséri (loc. sg.) vel sim. ‘in/to the hand’ → (univerbated) *me-g̑^hsri adv. ‘near’ > Gk. μέχρι ‘as far as, until’, Arm. merj ‘near’ (Nussbaum 1986a:82f., Schaffner 2004b:294, Hackstein 2010b:60, Neri 2013a:187).

Because of the independent distribution in Hittite, Avestan, and Greek a collocation *g̑^h(e)s-ér±i (loc. sg.) / *g̑^h(e)s-r̥-su (loc. pl.) + *d^heh₁- ‘to put sth. into the hand/hands’ can be reconstructed for PIE (Bock 2012a:189f.):

(1) Hittite: KUB I 1+ I 63 (apology of Ḫattušili III) nu-mu EGIR-pa da-a-aš nu-mu-kán KARAŠ ANŠU.KUR.RA^MEŠ ŠA KUR ^URUḪa-at-ti ŠU-i da-a-iš “And he took me up again and put the army and the chariot fighters of the land Hatti into my hands”; Mşt. 75/17 Vs. (letter) ma-an-za ERÍN^MEŠ EGIR-an ka-ru-ú pa-aḫ-ḫa-aš-nu-ut an-tu-uḫ-ša-ša-kán ut-tar ki-iš-ša-ri-i an-da ka-ru-ú da-iš “If you would already have protected the squads of infantry afterwards, (the deity) would have put the matter of men into your hand”; KUB I 1+ I (apology of Ḫattušili III) ^LÚKÚR^MEŠ-mu-kán ^LÚ.MEŠar-ša-na-ta-lu-uš ^DIŠTAR GAŠAN-YA ŠU-i da-a-iš na-aš-za kat-ta-an ar-ḫa zi-en-na aḫ-ḫu-un “Ištar, my mistress, put enemies and jealous men into my hand, and I bumped them off completely”;

(2) Avestan: Y 44.14 kaϑa aṣ̌āi drujǝ̄m diiąm zastaiiō (loc. du.) “whether I can put the lie into the hands of truth?” Cf. with the same verb, but with different syntactic construction Skt. pāṇiṁ dhā- ‘to hold out one’s hand (lit. ‘to put the hand’); to marry’;

(3) Greek: ἐν χερσὶ τιϑέναι “to put into the hands” (Il. 1.441, 8.289).

In Indo-Iranian poetic language, compounds and text figures with PIIr. *j́^hásta- m. ‘hand’ < *g̑^hésto- are attested that partly have correspondences in other IE languages (cf. EWAia II:812):

(1) Ved. íṣu-hasta- (RV+) ‘having arrows in the hand’ ~ Gk. ἰο-χέαιρα ‘id.’ (Il.+, epithet of Artemis); cf. Schmitt 1967a:179f., Peters 1980a:223f., Schindler 1986a:396, EWAia I:200.

(2) Ved. uttāná-hasta- ‘having extended hands (for prayer)’ (RV+) ~ YAv. ustāna-zasta- ‘id.’ (cf. ustānāiš … zastāiš ‘with extended … hands’ Y 29.5), RV 6.16.46 uttānáhasto námasā ~ Y 28.1 nǝmaŋhā ustānazastō; cf. EWAia I:212, EWAia II:812 with further lit.

(3) Ved. vájra-hasta- ‘holding the vájra- (Indra’s mythical weapon) in the hands’ ~ YAv. vazrǝm zastaiia draž° ‘to hold the vazra- in the hands’; cf. EWAia II:492, 812 with further lit.

An independent elementary pair formula ‘with hands and feet’ is attested in Lycian izredi pededi; cf. Latin manibus pedibusque and German mit Händen und Füßen (Bock 2012a:188f.).

A Euro-phrase appears in the proverb Engl. one hand washes the other, i.e. “one helps the other”; cf. Greek (Epicharmus, 550–460 BC, 273) ἁ δὲ χεὶρ τὰν χεῖρα νίζει; Latin (Seneca, about 4 BC until 65 AD, Apokolokyntosis 9.5) manus manum lauat; French une main lave l’autre; Italian una mano lava l’altra; German eine Hand wäscht die andere; Russian ruka ruku mojet; Hungarian kéz kezet mos (Bock 2012a:206).

[S.S.], [T.S.]

Gk. χείρ, gen. sg. χειρός, dat. pl. χερσί (Ion.-Att., Il.+), Dor. χήρ, gen. sg. χηρός, Aeol. acc. sg. χέρρ’, acc. pl. χέρρας f. ‘hand’; first compound member in χέρ-νιψ, gen. sg. χέρ-νιβος f. ‘water for washing the hands’ (Od.+; cf. Myc. ke-ni-qa), second compound member in αὐτό-χειρ ‘with one’s own hand’ (Aisch.+).Cf., e.g., GEW II:1082f., NIL:170f. with nn. 8–9, EDG:1620f.

Arm. jeṙn ‘hand’ (gen. sg. jeṙin, nom. pl. jeṙkʻ, dat. pl. jeṙacʻ)Cf., e.g., Olsen 1999a:174f., 663, Matzinger 2005a:70, NIL:170f. with n. 10, EDAIL:431f.

Alb. dorë f. ‘hand’Cf., e.g., Demiraj 1997a:140, NIL:170f. with n. 11, Matzinger 2006a:54, 60, 71, 74, 97, 138, 169.

TochA tsar, TochB ṣar ‘hand’Cf., e.g., NIL:170f. with n. 12, DTB²:711f.