*dóh3-/dh3

‘to take’

Language
Proto-Anatolian
Grammar
verb
ID
2556

1. Individual Anatolian Languages

1.1 Hieroglyphic Luwian

/la-(i)/, CAPERE, la-

‘to take’

1.1.1Transmission

As part of the core vocabulary, this verb is attested 57 times in the HLuw. corpus and can be found across the entire geographical area. Its temporal distribution matches the geographical one, and the relevant texts date between the late 13th and the late 8th centuries BCE.

[AH.B.]

1.1.2Forms

 3sg. pres. ind. act. /lai/ la-i KÖRKÜN, §11 (Karkamiš)
    CAPERE-i KÖRKÜN, §8 (Karkamiš)
      KARKAMIŠ A2+3, §20 (Karkamiš)
      ARSUZ 1, §23 (Amuq)
      KÖTÜKALE, §5 (Malatya)
      ALEPPO 2, §13 (Tell Ahmar)
      ANCOZ 7, §9 (Commagene)
      BOROWSKI 3, §9 (Tell Ahmar)
      BOYBEYPINARI 2, §19 (Commagene)
  /laya/ “CAPERE”-ia HAMA 4, §8 (Hama)
    CAPERE-ia KARKAMIŠ A6, §28 (Karkamiš)
3pl. pres. ind. act. /lanti/ CAPERE-ti-i KARKAMIŠ A11b+c, §28 (Karkamiš)
      KARKAMIŠ A11b+c, §29 (Karkamiš)
      KARKAMIŠ A11a, §27 (Karkamiš)
    CAPERE-ti-´[...] JISR EL HADID frag. 3, §a (Amuq)
1sg. pret. ind. act.  /lahha/ (CAPERE)la-ha MARAŞ 4, §4 (Maraş)
      MARAŞ 4, §12 (Maraş)
    la-ha MARAŞ 13, l. 3 (Maraş)
      BOHÇA, §13 (Tabal)
    CAPERE-ha KARKAMIŠ A7, §3 (Karkamiš)
      KARKAMIŠ A12, §5 (Karkamiš)
      ARSUZ 1, §10 (Amuq)
      TELL TAYINAT 2, l. 2, 2b-a (Amuq)
3sg. pret. ind. act.  /latta/ CAPERE-ta TULEIL 1, l. 2 (Amuq)
      TELL AHMAR 1, §12 (Tell Ahmar)
    CAPERE-tá KARAHÖYÜK, §7 (Malatya)
    CAPERE-tax TOPADA, §28 (Tabal)
    CAPERE SÜDBURG, §11
3pl. pret. ind. act.  /lanta/ la-ta KULULU 4, §9 (Tabal)
    CAPERE-ta TELL AHMAR frag. 1 (Tell Ahmar)
    CAPERE YALBURT frag. 1, §10
2sg. impv. act. /lasi/ la-si İSKENDERUN, §6 (Maraş)
3sg. impv. act. /lattu/ CAPERE-tu EMİRGAZİ 1B, §22
    CAPERE EMİRGAZİ 1C, §22
3pl. impv. act. /lantu/ CAPERE-tú KARKAMIŠ A25a, §3 (Karkamiš)
inf. /l(a)una/ (“!CAPERE”)la/i/u-na-a BOHÇA, §3 (Tabal)
    (“CAPERE”)la/i/u-na BOHÇA, §9 (Tabal)
    (“CAPERE”)la-na MARAŞ 3, §8 (Maraş)
part. acc. sg. c. /lammin/ CAPERE-na KIRŞEHİR lead strip, §20 (Tabal)
part. unclear   ˹CAPERE˺-mi-i[...] TELL TAYINAT 1, l. 5 (Amuq)

The lexeme is attested 57 times in the corpus, but only a selection of the evidence is listed above. Care has been taken to ensure that all spelling variations are covered as well as the range of geographical and temporal distribution of this core verb. Broken forms have been omitted from the list, unless they provided information not otherwise attested.

In BOHÇA, §3, the logogram markers (sign *410) are missing on the inscription, instead the similar looking word-divider (sign *386) occurs twice one after the other, which has prompted scholars to read it as 〈(“!CAPERE”)〉.

[AH.B.]

1.1.3Graphic Features

1.1.3.1Phonological Interpretation

The phonological interpretation of this basic verb has seen considerable discussion among scholars for two reasons. For one, the spelling in HLuw. is often partly logographic and thereby leaves considerable interpretative leeway, mitigated only mildly by evidence from the related languages recorded in cuneiform writing, because the CLuw. cognate is la- ‘to take’ (CLL:120), whereas the cognate in Hittite is da- ‘to take’ (HEG T:5-11).The matter was further obfuscated by the very similar looking Hittite verb la- ‘to release’, which turned out to be etymologically unrelated, cf. the discussion by Meriggi 1962c:258. For another, the syllabic value attached to the frequently used sign (*41) is indubitably 〈tà〉, i.e. /da/, and the verb was hence assumed to have the shape /da-/ or /ta-/ in HLuw. Moreover, the sign shape of is a more or less pictographic rendition of a taking hand, suggesting very strongly that its syllabic value derives acrophonically from the phonological shape of the verb ‘to take’, and the difference between the value 〈tà〉 and the verb /la-/ were therefore difficult to reconcile for scholars. Hawkins & Morpurgo Davies 1979a:395-398 (summarised by Morpurgo Davies 1987a:211f. n. 17) advocate treating as purely logographic for this verb, i.e. it would never represent its syllabic value. This view is followed by Yakubovich 2008c:21-24, who re-examines the available evidence and introduces a sociolinguistic perspective to the matter, suggesting that the developing hieroglyphic signs, at an earlier point in time, used to be read out in Hittite and in this case therefore likely as /da-/, only changing to Luwian at a later date when the syllabic value 〈tà〉, i.e. /da/, had already been fixed for the sign . Sasseville 2021a:349 attributes a possible allomorph /da-/ to Ištanuwa Luwian.A similar possibility is hinted at by Oreshko 2013a:406.

[AH.B.]

1.1.4Meaning

The meaning of this verb was first suggested for its CLuw. cognate la- by Laroche 1958a:101-104, who identified it as transitive both in its simplex form /la-/ and with reduplication /lala-/ (see under lemma). Laroche 1958a’s proposal first met with some resistance, e.g. by Meriggi 1962c:258, but was then generally accepted not only for CLuw., but also for HLuw. (see e.g. Oettinger 1979a:501 n. 13, Hawkins & Morpurgo Davies 1979a:395f.) despite the confusion over its phonological shape in HLuw., see the discussion above.

As a basic level verb, /la-/ is applied widely in monotransitive clauses with direct accusative objects, and in ditransitive constructions with additional indirect objects in the dat./loc. The verb is also frequently combined with pre-verbs to alter its basic meaning: ARHA /ahha la-/ ‘to take away; to receive (of deities)’ (e.g. in KARKAMIŠ A11b+c, §30 and A11a, §27, see Hawkins & Morpurgo Davies 1986a:77f.); +ra/i /sarra la-/ ‘to take over’ (e.g. in BOHÇA, §3 and §9, see Hawkins 1980a:149); -tá /zanta la-/ ‘to take down (= to defeat, conquer?)’ (e.g. in ARSUZ 1, §10, see Dinçol, Dinçol, Hawkins, Peker & Öztan 2015a:67).

[AH.B.]

1.1.5Stem

This lexeme is commonly regarded as one of the prime examples for root verbs in the Anatolian languages and inflects according to the ḫi-conjugation, cf. already Kronasser 1956a:188 and more recently e.g. Vernet 2017a:63f. and Sasseville 2021a:348-350 and 546 with further literature.

[AH.B.]

1.1.6Derivatives

  1. /lala-(i)/

  2. /lalamma-/

1.2 Cuneiform Luwian

lā-(i)

‘to take’

1.2.1Transmission

This lexeme is attested in various Luwian ritual texts.

[O.B.]

1.2.2Forms

1pl. pres. ind. act. lu-ú-un-ni KUB 35.90, 7’ (CTH 766, MS)
  l[u-ú-u]n-ni VBoT 60 i 12’ (CTH 770, NS)
  lu-un-ni KUB 35.128 rev. 9’ (CTH 766, MS)
3sg. pret. ind. act. la-a-ad-da KUB 32.8 iii 15’ (CTH 759, NS)
    KUB 32.8 iii 16’ (CTH 759, NS)
  la-a-at-ta KUB 35.125, 4’ (CTH 770, NS)
    KUB 32.8 iii 12’ (CTH 759, NS)
    KUB 32.8 iii 17’ (CTH 759, NS)
  la-a-˹at-ta˺ KBo 7.68 i 8’ (CTH 764, NS)
  la-a-a[t-ta] KUB 35.161, 5’ (CTH 763, NS)
  ˹la˺-a-at-ta KBo 7.68 i 7’ (CTH 764, NS)
  la-at-ta KUB 35.43 ii 17 (CTH 761, NS)
    KUB 35.43 ii 18 (CTH 761, NS)
  l[a-at-ta] KUB 34.62, 3’ (CTH 761, NS)
3sg. impv. act la-a-ad-du! KBo 13.260 ii 17 (CTH 766, NS)
  la-a-ad-du KBo 13.260 ii 13 (CTH 766, NS)
    KBo 13.260 ii 15 (CTH 766, NS)
    KBo 13.260 ii 21 (CTH 766, NS)
    KBo 13.260 ii 23 (CTH 766, NS)
  la-a-ad-d[u] KBo 13.260 ii 9 (CTH 766, NS)
    KBo 29.27 i 5’ (CTH 766, NS)
    KBo 29.27 i 8’ (CTH 766, NS)
  la-a[-ad-du] KBo 13.260 ii 19 (CTH 766, NS)
  ˹la˺-a-ad-d[u] KBo 13.260 ii 11 (CTH 766, NS)
  [l]a-a--d[u] KBo 13.260 ii 7 (CTH 766, NS)
  la-ad-du KUB 35.146 iii 19’ (CTH 458, MS)
3pl. impv. act. la-a-an-du KBo 29.34 i 9’ (CTH 770, NS)

According to Melchert (CLL:120), 〈lu-u[n-n]i〉 on KUB 35.128 could also feasibly belong to lūwa- ‘to pour’.

Per Meriggi 1957a:208 n. 41, 〈[l]a-a-i-im-ma-an〉 is a participle to lā-(i) (also given as such in CLL:120). However, we would not expect a participle of lā-(i) to contain an -i-. The form is as such more likely to belong to another otherwise unattested verb. Therefore, it has not been included in the table above.

[O.B.]

1.2.3Graphic Features

Note the clear misspelling 〈la-a-ad-an〉 on KBo 13.260 ii 17. The form should read as a 3sg. impv. act. form 〈la-a-ad-du!〉 as elsewhere at the ends of lines KBo 13.260 ii 7, 9, 11, 13, 15, 19, 21 and 23 (thus also in Starke 1985b:261). Note also the unique singleton -d- in the form 〈la-a-d[u]〉 on KBo 13.260 ii 7, which should clearly be amended to 〈[l]a-a-<ad>-d[u]〉 on account of the other spellings with geminated -dd- ending elsewhere on the same tablet (pace Starke 1985b:260).

All forms have been checked against the Konkordanz (Mainzer Fotoarchiv).

[O.B.]

1.2.4Meaning

The meaning of CLuw. lā-(i) as ‘to take’ is largely consensual since the treatment of Laroche 1958a:101-104, followed in e.g. Laroche 1959a:61, Morpurgo Davies 1982-1983a:263, Oettinger 1979a:564, HEG T/D:6, CLL:120, Hawkins 2000a:269, Schürr 2001e:131, EDHIL:803, Dempsey 2015a:219, Vernet 2017a:63 and Sasseville 2021a:348 (but cf. the initial skepticism in Kammenhuber 1961a:47 n. 4 and Meriggi 1962c:257f.). As demonstrated by Laroche 1958a, the use of the 3sg. pret. act. form latta on KUB 35.43 is mirrored by Hittite dāš in certain ritual contexts. Furthermore, the reduplicated variant CLuw. lālā-(i) takes a hammer (GIŠNÍG.GUL) as its direct object followed by an instruction to use it (3pl. impv. act. tarmiššandu) on KBo 9.145.

The meaning ‘to loosen’, derived from an alleged connection to Hitt. lā-/l-(ḫi) ‘to loosen’ (Otten 1953a:57 n. 193a, Kammenhuber 1961a:47 n. 4, Meriggi 1962c:257f., HEG L-M:1), is now outdated.

[O.B.]

1.2.5Stem

CLuw. lā-(i) is ḫi-conjugating root verb. Although the 3sg. pres. act. is not attested, the stem type is certain in comparison to HLuw. /la-(i)/ and the CLuw. derivative lālā-(i). The unlenited endings in the 3sg. pret. act. and the 3sg. impv. act. are also consistent with a ḫi-conjugating inflection.

[O.B.]

1.2.6Derivatives

  1. lālā-(i)

  2. lalam(i)-

  3. lālašḫā-

  4. lalašḫa-(ti)

  5. lalatta-

1.3 Hittite

dā-/d-(ḫi)

‘to take’

1.3.1Literature / Comments

HEG T,D:5-11, EDHIL:803-805

2. Reconstruction

2.1Reconstruction ‘to take’

The Luwian verb /la-(i)/ ‘to take’ has been equated to Hitt. dā-/d-(ḫi) ‘to take’, both even showing the same verbal stem formation, i.e. PAnat. *dóh3-/dh3 ‘to take’ (Melchert 1994a:73, EDHIL:804, Sasseville 2021a:546; cf. Jasanoff 2003a:84-86). On the phonetics of the initial consonant, which is spelled -l- in Cuneiform Luwian, see Kloekhorst 2019c:40f. and Sasseville 2021a:306. Contra Tischler (HEG L-M:1-3) and Yakubovich 2008c:21 n. 24, who prefer associating the Luwian verb with Hitt. lā(i)-(mi) ‘to loose’. A new morphological argument can be used against this hypothesis; a root verb of the shape *lóh1-/lh1 would have been transferred to the PAnat. verbal stem class in *-oi̯h1/-ih1- (for which see Sasseville 2021a:429-437). Therefore, this etymology should be given up.

For the view against reconstructing a reduplicated syllable *(de-)doh3-/dh3- as followed in Oettinger 2002b:xxiii (Oettinger 1979b:450f.), see Jasanoff 2003a:15-17, Kloekhorst 2018c:92-95, Sasseville 2021a:544f. For an overview of the older literature, see HEG T,D:7-11.

For the rejection of the hypothesis of an original middle transferred to an active paradigm as advocated in Eichner 1975a:93f., Oettinger 1979a:500f., Melchert 1984a:25, see Kloekhorst (EDHIL:804). The semantic difference between PAnat. *dóh3‑/*dh3‑´(ḫi) ‘to take’ and PAnat. *déh3‑/*dh3‑´(mi) ‘to give’ as well as PIE *deh3- ‘id.’ (cf. Lyc. da-(di) ‘to give’) can be explained by the fact that ḫi-verbs tend to show reflexive semantics (cf. Ved. 3sg. midd. ā́ datte ‘takes’) if attested besides non-reflexive mi-verbs, cf. the remarks s.v. PAnat. *h1ói̯h1‑/*h1ih1‑´(i) ‘to take’.

For the co-existence of a PAnat. root verb of the mi-conjugation, *déh3-/dh3, see Lyc. da-(di), and for the existence of a u-extended root *deh3u̯-, see Pal. tāzzu-. One should also be aware of the reconstructed imperfective formation PAnat. *dh3-sk̑é/ó- ‘to take (imperf.)’.

[D.S.], [A.O.]

Cross-references