*k̑ḗr, k̑érd- / *k̑r̥d-´
‘heart’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- subst
- ID
- 351
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/zart-/, za+ra/i-
‘heart’1.1.1Transmission
This lexeme is found only in the Iron Age corpus in late inscriptions of the 9th or 8th century BCE.
[D.S.]
1.1.2Forms
| nom./acc. sg. | /zar=za/ | za+ra/i-za | KULULU 5, §11 (Tabal) |
| dat./loc. sg. | /zarti/ | za+ra/i-ti | KARABURUN, §12 (Tabal) |
| gen. adj. acc. sg. c. | /zartassin/ | (X)za+ra/i-ta-si-na | KÖRKÜN, §10 (Karkamiš) |
The form /zartiyas/ does not belong to this paradigm, pace Hawkins 2000a:243; see under /zartiya-/.
[D.S.]
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
The use of the sign instead of <tà> in (X)za+ra/i-ta-si-na (KÖRKÜN, §10, end of 9th, beginning of 8th c.) indicates that the dental, which was originally voiced, i.e. *k̑erd-, has phonologically fallen together with the fortis stop and not with the lenis. For the exact phonetics, one can compare perhaps the neutralization of the alveolar after /n/, for which see Rieken 2010a:306f.
[D.S.]
1.1.4Meaning
Melchert 1987a:196f. and Hawkins & Morpurgo Davies 1988a:170-172 demonstrate with a contextual analysis and an etymological and morphological comparison with Hitt. kēr/kart- n. ‘heart’ that the Hieroglyphic Luwian lexeme /zart-/ means ‘heart’ (followed by Hawkins 2000a:243, 487).However, the meaning ‘heart’ had already been suggested by Kalaç (apud Poetto 1981a:276 n. 13) for HLuw. /zar=za/. In KULULU 5, §11, it was already noted by Poetto 1981a:276 that /zar=za/ must denote a vital organ and then compared to CLuw. UZUzarza. In KÖRKÜN, §10, they compare the expression (X)za+ra/i-ta-si-na INFANS-ni-na with the Hittite phrase kardiyaššaš DUMUMEŠ (KUB 14.1 obv. 80, VBoT 58 13) ‘children of the heart’, which means ‘child of one’s body’.The meaning ‘liver’ assigned by Poetto 1981a (followed by Starke 1990a:539) is outdated.
[D.S.]
1.1.5Derivatives
1.2 Cuneiform Luwian
zārt-
‘heart’1.2.1Transmission
Amply attested in Luwian rituals in MS and (L)NS.
[Zs.S.]
1.2.2Forms
| nom.-acc. sg. n. | UZUza-a-ar-za[(-) | KUB 32.7, 12’ | (CTH 762, NS) |
Connection not assured:
| dat.-loc. sg. | za-ar-ti | KUB 35.133 ii 4’ | (CTH 665, LNS) |
| nom.-acc. pl. n.? | za-ar-ta | KUB 35.103 ii 5’ | (CTH 766, MS) |
Sumerographically:
| abl.-instr. | ˹UZU˺ŠÀ-ti | KUB 35.21 rev. 5’ | (CTH 761, MS) |
| KUB 35.24, 7’ | (CTH 761, MS) | ||
| KUB 35.43 iii 9’ | (CTH 761, NS) | ||
| ˹UZUŠÀ˺-ti | KUB 35.24, 10’ | (CTH 761, MS) | |
| [UZU]˹ŠÀ-ti˺ | KUB 35.43 ii 15 | (CTH 761, NS) | |
| [UZU]˹ŠÀ˺-ti | KUB 35.43 iii 26’’ | (CTH 761, NS) | |
| UZU˹ŠÀ˺-ti | KBo 29.36, 6’ | (CTH 770, MS) | |
| UZU˹ŠÀ˺[-ti] | KUB 34.62, 8’ | (CTH 760, NS) | |
| [UZUŠÀ]-˹ti˺ | KBo 29.15, 3’ | (CTH 761, NS) |
Attestations following CLL:280 (cf. also Laroche 1959a:114 and HED K:191) checked against the photographs of the Konkordanz. Nevertheless, Laroche 1959a:115 booked zarti and zarta under a different entry; but according to Hawkins & Morpurgo Davies 1988a:175 it “seems likely” that they belong here, CLL:280 and EDHIL:469 are sure about that. HED K:191 quotes only zārza and zarti and both with question mark. Rieken 1999a:54 quotes only zārza and “vielleicht” zar[za in KUB 35.72, 7’ to be discussed below. Since the context of zarti and zarta is completely obscure, their connection is formally possible but cannot be proven (note also the lack of the determinative, although its presence is not obligatory).
CLL:280 restores (with question mark) a nom.-acc. sg. n. UZUza-a-ar-za (vel sim.) in (-)]˹ar˺-za-tar of KUB 35.16 i 7’, which is possible, but cannot be proven – the passage in question neither requires nor excludes this meaning. CLL:280 and Starke 1985b:396, 539 also add za-˹ar˺[(-) in KUB 35.72, 7’ (latter restoring za-ar[-za(-), followed by Hawkins & Morpurgo Davies 1988a:170 and Rieken 1999a:54 [cautiously]). This is possible, but the fragmentary text does not allow any conclusion (note also the lack of the determinative UZU, although its presence is not obligatory).
Restoration of UZUŠÀ[-ti] in KUB 34.62, 8’ (Starke 1985b:149, followed by CLL:280 and Groddek 1999b:307) is based on the duplicate text KUB 35.43 iii 26’’. Restoration of [UZUŠÀ]-˹ti˺ in KBo 29.15, 3’ (Starke 1985b:166, followed by CLL:280) is based on the following word UZUNÍG.GIG-˹ti˺, with which it frequently appears in pair (cf. under Meaning).
The grammatical form of zarta is considered to be unknown by CLL:280 and EDHIL:469. I. Yakubovich (pers. comm.) suggests nom.-acc. pl. n., which is formally definitely possible, but cannot be proven due to the unknown meaning of the passage. Starke 1990a:631 with n. 1 and Ünal 2016a:977 falsely identify UZUza-a-ar-za[(-) as a plural form.
[Zs.S.]
1.2.3Meaning ‘heart’
In early research its meaning was considered to be unknown (Laroche 1959a:114) or only a general sense (‘a body part’) was given based on the determinative (Carruba 1982b:5). However, since the paper of Melchert 1987a:196f. its meaning is generally assumed to be ‘heart’ (EDHIL:469; Tischler 2008a:232, Tischler 2016a:453; Ünal 2016a:977). Nevertheless, this is based only on the homonymy with the Hieroglyphic Luwian word for ‘heart’ (admitted by CLL:280, too), since the contexts are not helpful due to their fragmentary nature and unknown meaning.
Alternatively, Poetto 1981a:276f. suggested the meaning ‘liver’ based on the assumed parallelism of UZUNÍG.˹GIG˺ ‘liver’ and za-˹ar˺[(-) in KUB 35.72, 3’ & 7’, resp. Setting aside that we have already seen that the presence of this word in za-˹ar˺[(-) is not assured (cf. under Forms), Melchert 1987a:197 and Hawkins & Morpurgo Davies 1988a:170 rightly pointed out that this parallelism does not (necessarily) exist. Starke 1990a:631 n. 1 (cf. also 539 n. 1989) also rejected the meaning ‘heart’ and still considered the meaning ‘liver’ “erwägenswert”. This was rejected by CLL:280 using derogatory labels only (“vain attempt”, “unfounded”) instead of arguments. Such an argument could have been that Starke’s rejection is based only on his views on the Luwian reflex of PIE *k̑ and on the PIE precursor of Luwian that are not generally accepted.
As for the Sumerographic spelling, although the meaning UZUŠÀ is not restricted to ‘heart’, but also includes ‘the middle of something, internal part of something’ (Rüster & Neu 1989a:236f.; Tischler 2008a:281, Tischler 2016a:545; Marquardt 2011a:78 contra Weeden 2011b:609, who assumes ‘heart’ only), in this case it appears always (presumably including also KBo 29.15, 3’) in pair with UZUNÍG.GIG ‘liver’ (and also among lists of body parts) and thus its meaning as ‘heart’ is probable and the word is accordingly translated (Starke 1990a:42; Melchert 2016d:204; Oettinger 2015d:270, Oettinger 2016a:233).
[Zs.S.]
1.2.4Derivatives
Rejected derivatives:
1.3 Palaic
kārti-
‘heart’1.3.1Transmission
This lexeme is attested within the Palaic incantations of a festival text.
[D.S.]
1.3.2Forms
| acc. sg. | kar-ti-n(=) | KUB 35.165 rev. 8’ | (CTH 751, OS or MS) |
| kar-ti-[n(=) | KBo 62.43 5’ | (CTH 751, NS) | |
| dat./loc. sg. | ka-a-ar-ti | KUB 35.165 obv. 15 | (CTH 751, OS or MS) |
| [k]a-a-ar-ti | KUB 35.165 obv. 20 | (CTH 751, OS or MS) |
The restoration of kar-ti-[n(=) ] in KBo 62.43 5’ is based on the identification of the passage as a parallel one to KUB 35.165 rev. 8’.
[D.S.]
1.3.3Graphic Features
Checked against the photographs of the Konkordanz except for kar-ti-[n(=) ] (KBo 62.43 5’), whose photo was kindly given to me by Gernot Wilhelm. The sequence kar-ti-na-an-ta (KUB 35.165 rev. 8’) can be interpreted either as kardin=anta or kartinan=ta (Carruba 1970a:58). The preference is given over to the first variant (likewise Melchert 1994a:205), since the verb of the sentence, i.e. tazzu-, is not used with the particle -tta in the other attestation in KUB 35.165 obv. 9, but rather with -anta.
[D.S.]
1.3.4Meaning
Laroche 1955a:74f. connects Palaic ka-a-ar-ti with Hittite kēr/kārt- and compares the Palaic expression ku-i-ša tu-ú wuú-la-a-ši-na-aš ka-a-ar-ti a-ti-a-pa-an az-zi-ki-i ‘Whichever bread is in your heart (dative of possession), eat this one!’ (= Whichever bread you want, eat it!) with Hittite ku-it kar-di nu-za a-pa-a-at e-eš-ši (KUB 1.16 + KUB 40.65 iii 63) ‘Whatever you want, do it!’ (followed by Kammenhuber 1955b:367, Kammenhuber 1959a:19, Carruba 1972a:29, Eichner 2010a:48f.). The attestation of kar-ti-n(=) (KUB 35.165 rev. 8’) is treated as a different lexeme in Carruba 1970a:58. However, the meaning ‘heart’ for kartin is appropriate, since the context refers to the preparation of sacrificial meat for the gods to consume. Based on the comparison with Hittite rituals of the same kind, we know that the heart is often mentioned among the sacrificial meat; cf. among others KUB 9.31 iv 36-40.
[D.S.]
1.3.5Stem
Since the attestation of kar-ti-n(=) belongs to the same lexeme as ka-a-ar-ti, the Palaic word for ‘heart’ must be seen as a common gender i-stem.
[D.S.]
1.4 Hittite
ker, kard-/kardi-
‘heart’1.4.1Literature / Comments
HED K:189-191, Rieken 1999a:52-56
2. Reconstruction
2.1Reconstruction ‘heart’
The Proto-Anatolian word for ‘heart’ is reconstructed based on Hittite (UZU)ker, kard-/kardi- n., CLuw. zārt- n., HLuw. /zart-/ n. and Palaic kārti- c.; cf. Hawkins & Morpurgo Davies 1988a:175, HEG I-K:557, HED K:191, Rieken 1999a:54f., EDHIL:469. Only the Hittite lexeme preserves the ablaut, while the other two languages have given it up. Rieken (ibid. 52-56) reconstructs an ablauting root noun, in which the strong cases have an accented e-grade in the root and the weak cases a zero-grade of the root with the accent on the ending. Due to the loss of *d in final syllable (cf. Hitt. ker), the nominative/accusative singular of the strong cases had a lengthened *ē in the root due to the result of compensatory lengthening, which is established for Proto-Indo-European; cf. Kümmel 2015a and Piwowarczyk 2015a.
The Cuneiform Luwian word for ‘heart’ shows the plene writing of the vowel /a/ in the root. This vowel cannot go back to the long *ē, which would have yielded Luwian /i/. Instead, it can go back to an accented *é or to a vocalic *r̥ developing to *ar, which would have secondarily received the accent, i.e. /ár/. In any case, the nominative/accusative singular has generalized the vocalism of the other cases, while the accent was columnized on the first syllable throughout the paradigm, which is confirmed by the plene writing; cf. Hawkins & Morpurgo Davies 1988a:177, Melchert 1994a:279. The use of the enclitic suffix /-za/ instead of /-sa/ indicates that the final dental had been restituted in the nominative/accusative singular, i.e. *zard- + -sa > /zarza/ (Hawkins & Morpurgo Davies 1988a:176, Melchert 2004d:472 n. 2, Jasanoff 2010a:173 n. 27). The Luwian consonant vis-à-vis Hittite and Palaic indicates that we are dealing with a palatalized *k̑, which must be reconstructed for Proto-Anatolian; cf. Melchert 2012a with further literature.
If the assignment of the Palaic lexeme as a synchronic common gender i-stem is correct, it represents an innovation vis-à-vis the other Anatolian languages. The motivation for the reinterpretation of the stem can be found in the possible existence of the genitive in /-iyas/ (just like Hitt. gen. sg. kardiyaš), which is the genitive ending of an i-stem (or perhaps -iya-). The plene writing found in the dative singular ka-a-ar-ti indicates that the place of accentuation has been retracted to the first syllable in the weak stems in analogy to the strong stems.Contra Korolev & Shevoroshkin 1969a:540, the city name Kerdesos (Lyc. A kerθθi, Lyc. B kridesi) cannot etymologically belong here, if it is genuinely Lycian, because PIE *k̑ develops to Lyc. s.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
PIE *k̑érd-/*k̑r̥d-´ n., which can be found in almost all IE languages, means ‘heart’ (EIEC:262f., NIL:417, Watkins 2011a:42 s.v. kerd-1). Furthermore, the syntagm *k̑red(s) dheh1- ‘to put one’s heart on sth., to trust, believe’ (see below sub Syntactic Reconstruction and Phraseology), which is attested in several IE languages, “indicates that in the period of PIE, the heart was regarded as the organ of belief or thought”, according to EIEC:263.
3.2Morphological Reconstruction
The word for ‘heart’ can be analyzed as a root noun *k̑erd- or a d-stem *k̑er-d- (cf. NIL:418f. n. 1 with discussion and lit.). The nom./acc. sg. is reconstructed by many scholars as *k̑ḗr (Hitt. kēr, Gk. κῆρ, OPr. seyr, etc.) < *k̑érr < *k̑érd, with assimilation and compensatory lengthening by Szemerényi’s Law (Rieken 1999a:55). Others, however, assume a protoform *k̑ḗrd (so, e.g., EIEC:262f.). The paradigm is usually set up as *k̑ér(-)d-/*k̑r̥(-)d-´,Differently, e.g., Beekes 1985a:170f. and Beekes 1995a:190, who considers a d-stem paradigm *k̑ḗr-d/*k̑r-éd-; cf. NIL:418 n. 1 with discussion.for the weak stem of which cf. Ved. gen. sg. hr̥dás, OAv. instr. sg. zərədā-cā. There are also several extensions/derivatives based on it, including *k̑r̥d-(i)i̯eh2- (Gk. καρδία), *k̑r̥d-(i)i̯o- (Ved. hŕ̥dya- adj. ‘cordial, beloved’, OIr. cride n. ‘heart, middle, love’), and *k̑r̥d-ei̯o- (Ved. hŕ̥daya- n. ‘heart’, YAv. zərəδaiia- n. ‘id.’); cf. NIL:417–423, esp. n. 1. Beside the root noun or d-stem, there might have been an i-stem *k̑ḗr(-)d-i-/*k̑r̥(-)d-(e)i- vel sim. (cf. Ved. hā́rdi, Hitt. gen. sg. kardii̯aš), which could also have been the basis of the aforementioned stems *k̑r̥dei̯o- (but cf. EWAia II:818) and *k̑r̥d(i)i̯o/eh2- (cf. NIL:418 n. 1 with lit.).
3.3Syntactic Reconstruction and Phraseology
A phrase *k̑réd dheh1- vel sim. > PIE *k̑réds(-)dheh1- [k̑rédz(-)dheh1-] ‘to trust, believe’ can safely be reconstructed for PIE, with attestations in several languages – cf. Ved. śrád dhā- ‘to trust’ (RV+ [śrád … dhatta, śrád … dádhat, śrad-dádhāna-, etc.]), OAv. zrazdā- adj. ‘devoted, believing, trusting’, Lat. crēdō ‘to believe, trust’, OIr. creitid ‘to believe’, W credu ‘id.’. The basic meaning of the PIE pharse might have been ‘to put one’s heart on sth.’ vel sim., whence ‘to trust, believe’. However, the identification of *k̑red(s)(-) is problematic. Some regard it as the loc. sg. of *k̑érd-/*k̑r̥d-´ ‘heart’ (Rix 1995a:246, Tremblay 2004a:581–585, Nikolaev 2010d:136 n. 29, Steer 2013b:78–80). According to others (cf. NIL:423 n. 25 with lit.), it is rather an accusative form. Schindler 1979a:58 proposes that *k̑reds(-) could have been the (nom./)acc. sg. of an otherwise unattested neuter s-stem noun *k̑érd-s- ‘heart’, with “*k̑reds statt *k̑erds durch dieselbe phonotaktische Regel wie *h2u̯eg-s-, *h2lek-s- (gr. ἀέξω, ἀλέξω) neben *h2eu̯g-, *h2elk-”. If this is correct, there was a neuter s-stem noun *k̑érd-s- ‘heart’ (with nom./acc. sg. *k̑érd-s → *k̑réd-s), which is only preserved in the phrase *k̑réds(-)dheh1- and stands next to the synonymous and more widespread stem *k̑ér(-)d-/*k̑r̥(-)d-´ ‘id.’ discussed above. For further details, cf. the thorough discussion by Hackstein (2012b:90–93 with lit.).
[T.S.]
4. Transmission
Ved. hŕ̥d- n. ‘heart’ (nom./acc. sg. hā́rdi, gen. sg. hr̥dás, dat. sg. hr̥dé, loc. pl. hr̥tsú, etc. [RV+]), su-hā́rd- adj. ‘having a good heart, friendly’ (AV+), sarva-hr̥dā́ instr. sg. ‘wholehearted’ (RV 10.160.3); cf. also hŕ̥daya- n. ‘heart’ (RV+), hŕ̥dya- adj. ‘cordial, beloved’ (RV+). – OAv. zərəd- n. ‘heart’ (instr. sg. zərədā-cā [Y 31.12]), YAv. zərəδaiia- n. ‘heart’Cf., e.g., AiGr III:236f., EWAia II:818 with lit., ALF:139f., NIL:417f., 420 nn. 4–5.
Arm. sirt ‘heart’ (i-stem: gen./dat. sg. srt-i, instr. sg. srt-i-w, gen./dat. pl. srt-i-c‘, etc.)Cf., e.g., Olsen 1999a:87f., NIL:417, 420 n. 7, EDAIL:576–578.
OPr. seyr n. ‘heart’; Lith. širdìs f. ‘id.’, šerdìs f. ‘core, kernel, pith’; Latv. sir̂ds f. ‘heart, courage, wrath’Cf., e.g., Petit 2004a:38–47, NIL:417, 420f. nn. 9–12, ALEW II:1019, 1032–1034, EDBIL:443, 448f.
TochA kri ‘will’, TochB käryāñ nom. pl. ‘hearts’Cf., e.g., NIL:417f., 421 nn. 15, 18, Adams 2013a:175.