*h1eí̯-/h1i-´

‘to go’

Language
Proto-Anatolian
Grammar
verb
ID
731

1. Individual Anatolian Languages

1.1 Hieroglyphic Luwian

/i-(di)/, (PES2)i-ti

‘to go’

1.1.1Transmission

This lexeme is attested 15 times in the HLuw. corpus, and the inscriptions can be found in most of the HLuw. area. In terms of dating, the inscriptions range from the earliest in the second half of the 13th century to the latest in the early 7th century BCE.

[AH.B.]

1.1.2Forms

1sg. pres. act. /iwi/ i-wa/i KULULU 1, §15 (Tabal)
3sg. pres. act. /idi/ (PES2)i-ti TELL AHMAR 6, §34 (Tell Ahmar)
    (PES2)i+ra/i CEKKE inscr. 2, §20 (Karkamis)
1sg. pret. act. /iha/ (PES2)i-ha BOYBEYPINARI 2, §9 (Commagene)
    (PES2)i-˹ha˺ TELL AHMAR 6, §24 (Tell Ahmar)
    PES2-ha

KULULU 3, §5 (Tabal)

      KARKAMIŠ A18h, §3 (Karkamis)
    PES EMİRGAZİ 2, §5
3sg. pret. act. /ida/ (“PES2”)i-tà KAYSERİ, §19 (Tabal)
    “PES2”-tà-i SULTANHAN, §15 (Tabal)
    (“PES2”)i+ra/i

TOPADA, §8 (Tabal)

      TOPADA, §13 (Tabal)
3sg. impv. act. /idu/ (PES2)i-tú-u TELL AHMAR frag. 5, l. 4 (Tell Ahmar)
infinitive /iuna/ (“PES2”)i-u-na KARATEPE 1 Hu., §34 (Cilicia)
unknown   (PES2)i-[x] KARKAMIŠ A29b (Karkamis)

The form --i in SULTANHAN, §15 is often regarded as a form belonging to a different verb: /awi-(di)/ ‘to come’, which happens to be used in the preceding clause and usually involves the logogram (*90). However, the use of (*93) instead of  (*90) in --i suggests that it should be regarded as a form of /i-(di)/ ‘to go’ instead, because the latter generally relies on  as is evident in the above table. Still, it should be kept in mind that  and only differ by being mirror images of one another and therefore switch orientation with the direction of reading from one line to the next in the HLuw. boustrophedon style.

Additionally, --i from SULTANHAN, §15 has traditionally been read as /idi/ and analysed as a 3sg. pres. act. with future meaning. The same analysis was suggested for the verb in the preceding clause, “PES”-wa/i+ra/i as /awidi/, which belongs to /awi-(di)/ ‘to come’, cf. still Weeden 2018a:334 following Hawkins 2000a:466. Yet the context calls for both verbs to be preterite instead of present tense, and they should be read /ida/ and /awida/ respectively; the spellings permit both interpretations in either case.

[AH.B.]

1.1.3Graphic Features

1.1.3.1Phonological Interpretation

The phonological interpretation of this verb is secure. Not only are different forms from the paradigm attested in fully syllabic spelling, /i-(di)/ ‘to go’ also has cognates in several other Anatolian languages, e.g. CLuw. /i-(di)/ ‘to go’ and Hittite ī-(mi). The forms involving a /d/ as part of their ending show the expected rhotacism in several cases.

[AH.B.]

1.1.4Meaning

This lexeme constitutes a basic level term and is one of the most central verbs of motion. A collection of the forms was first undertaken by Laroche 1958a:99-101, who also separated out those forms that belong to similar looking but unrelated verbs. He followed Meriggi 1957a:202f. in identifying the verb as ‘to go’.

In HLuw., the verb is frequently paired with preverbs that change its meaning, e.g. ARHA i- /ahha i-/ ‘to go away’, VERSUS i- /tawiyan i-/ ‘to go against / toward’, etc.

[AH.B.]

1.1.5Stem

This lexeme is a root verb with lenited endings, for which see Sasseville 2021a:301.

[AH.B.]

1.1.6Compounds

  1. /awi-(di)/ ‘to come’

1.2 Cuneiform Luwian

ī-(di)

‘to go’

1.2.1Transmission

The verb is attested in Luwian incantation rituals and the Ištanuwa festival more than 20 times, mostly but not exclusively in MS manuscripts.

[E.R.]

1.2.2Forms

3sg. pres. ind. act. i-ti KUB 35.21 obv. 24, rev. 19 (CTH 761, MS)
  ˹i˺-[ti] KUB 35.47, 4’ (CTH 758, NS)
  i-ti KUB 35.54 iii 20 (CTH 758, MS)
  [i-t]i KUB 35.54 iii 21 (CTH 758, MS)
  i-ti KUB 35.117, 6’ (CTH 750, NS)
  i-ti KUB 32.8 iv 6’, 24’ (CTH 759, NS)
  i-ti KUB 32.9 obv. 5, 6 (CTH 761, MS)
  i-ti KBo 22.143 i 8 (CTH 761, NS)
  [i-t]i KBo 29.9 obv. 13 (CTH 760, NS)
  i-li KBo 4.11 rev. 44 (CTH 772, NS)
1pl. pres. ind. act. [i-ú-u]n-ni KUB 35.103 ii 11’ (CTH 766, MS)
3sg. pret. ind. act. i-i-ta KUB 35.109 iii 1 (CTH 765, MS)
3sg. impv. act. i-du KBo 7.66 ii 6’ (CTH 670, MS)
  i-du KBo 29.25 ii 7’ (CTH 764, MS)
3pl. impv. act. [i]-ia-an-du KUB 35.101 rev. 5’ (CTH 765, MS)
  [i-ia]-an-du KUB 35.102 rev. 1 (CTH 766, MS)
  i-ia-an-du KUB 35.103 iii 1 (CTH 766, MS)
  ˹i-ia˺-an-du KUB 35.103 ii 13 (CTH 766, MS)
  i-ia-an-d[u] KUB 32.15, 4’ (CTH 766, MS?)
  i-ia-an-du KBo 13.260 ii 5 (CTH 766, NS)
  i-ia-an-du KBo 29.25 iii 10’ (CTH 764, MS)
  [i-i]a-an-du KBo 29.27 i 2’ (CTH 766, NS)
  i-ia-an-du VBoT 60 i 5’ (CTH 770, NS)
  i-ia-an-du(=) Privat 40 ii 4’ (CTH 770, N/A)

For i-ia-a (KUB 25.39, 27’), tentatively assigned to this lemma by Puhvel (HED:I 328; sceptical Oettinger 1979a:349 n. 185), see under iya-. – The assignment of the form i-li (KBo 4.11 rev. 44) to this verb as 3sg. pres. act. follows Yakubovich 2017d. It is supported by the co-occurrence with KASKAL ‘path’.

[E.R.]

1.2.3Graphic Features

Attestations checked against the photographs in the Konkordanz, except for Privat 40. – Restoration [i-t]i in KBo 29.9 obv. 13 following Yakubovich (pers. comm.).

[E.R.]

1.2.3.1Phonological Interpretation

The unusual initial plene spelling with i-i- in i-i-ta (KUB 35.109 iii 1) is normally understood as a mark of vowel length (most recently Sasseville 2021a:255), but Simon 2010b:260-262 considers the possibility that it may have be used to indicate the reflex of initial *h1-, if inherited from PIE *h1ei̯-. – Apparently, the expected lenis /d/ in i-li at KBo 4.11 rev. 44 underwent a further weakening to [l] in the Ištanuwa dialect. This is probably only a sporadic change.

[E.R.]

1.2.4Meaning

Paradigm, meaning and syntax of ī-(di) were elucidated by Laroche 1958a:99-101, based on Meriggi’s (Meriggi 1957a:202f., 207f.) earlier analysis of KUB 35.103 iii 1-3. The latter had misunderstood the syntax of the passage, but arrived at the correct meaning ‘to go’. Laroche then pointed to the intransitive use of ī-, its combination with locatival expressions, and the serialization construction well-known from Hittite pai- ‘to go’ (see also van den Hout 2010a:201f.). The semantic interpretation is generally accepted (Laroche 1959a:53; CLL:85; Oettinger 1979a:561; Starke 1990a:164 et al.; HEG:A-K 335; HED:I 335; HED:PA 39; EDHIL:375f.) and confirmed by the straightforward etymological connection with PIE *h1ei̯-; see under Origin.Outdated Otten 1953a:80-83.

[E.R.]

1.2.5Stem

The verb is one of the few remaining root verbs in Luwian, showing the expected lenited endings after its long root vowel (Oettinger 1979a:561; Morpurgo Davies 1982-1983a:257, 261; Sasseville 2021a:296-298, 301).

[E.R.]

1.2.6Compounds

  1. awī-(di)
  2. mParāiuna- (PN) (see Neumann 1976a:143; unconvincing counter-arguments in Starke 1990a:550)

1.2.7Derivatives

  1. iunāḫi(t)-

1.3 Palaic

ī-

‘to go’

1.3.1Transmission

This lexeme is found across the Palaic corpus.

[D.S.]

1.3.2Forms

1pl. pres. ind. act. i-ú-i-ni KUB 35.163 iii 11 (CTH 753, NS)
3pl. pret. ind. act. i-en-[t]a DBH 46/2.101 + KUB 35.159 ii 11’/15’ (CTH 753, NS)
  (=)ºi-en-ta KUB 35.168 5’ (CTH 752, OS)
  (=)ºi-en-ta KUB 32.18 i 17’ (CTH 752, OS or MS)
  (=)ºi-en-ta KUB 32.18 i 6’ (CTH 752, OS or MS)
2sg. impv. act. i-ú KUB 35.168 5’ (CTH 752, OS)
  ˹i?˺-ú KUB 35.168 5’ (CTH 752, OS)
  i-ú KUB 32.18 i 17’ (CTH 752, OS or MS)
part. nom. sg. c. [i-i]a-an-za KUB 35.165 obv. 7 (CTH 751, OS or MS)

Checked against the photographs in the Konkordanz.

[D.S.]

1.3.3Graphic Features

The suggested restoration of [i-i]a-an-za in KUB 35.165 obv. 7, yielding a participial form of the verb ‘to go’, is supported by the context (followed by Sasseville 2019a:27); see under meaning.

[E.R.]

1.3.4Meaning ‘to go, come, wander’

The verb ī- ‘to go’ in Palaic was first recognized in the imperative form i-ú, which was equated with Hitt. eḫu ‘come!’ (Carruba 1970a:58). Melchert 1994a:198f. adduces the preterit form cliticized to the particle chain, i.e. a=ant=ienta ‘they came in, they entered’ (followed by Yakubovich 2005c:118), as well as the 1st plural form iwini ‘we go’. The participial form [i-i]a-an-za suggested here by Rieken is added to this lexeme. This assignment is motivated by the context:

                mān=ti marḫānz(a) mān=t[i iy]anz(a) [a=]apa ariye (KUB 35.165 obv. 7-8)

                ‘Whether you have come as a guest or have wandered by, pick them (scil. bread) up!’

In this passage, it is argued that Zaparwa addresses the gods gathered at the feast and gives them the signal to start eating the bread (Sasseville 2019a:27). The form [iy]anz(a) stands in opposition to marḫānz(a) ‘to come as a guest’, which means that the former must refer to the gods, who have not been explicitly invited as guests, but happen to come by.

[D.S.]

1.3.5Stem

Based on CLuw. ī-(di), the Palaic verb ī- should inflect likewise according to the lenited mi-conjugation. However, no attestations of the relevant Palaic forms have yet been found.

[D.S.]

1.3.6Derivatives

  1. idāraši-

1.4 Hittite

ī-(mi)

‘to go’

1.4.1Literature / Comments

HW (2nd ed.) I:1, EDHIL:375f. The attestations of the active diathesis of the Hittite verb ‘to go’ are archaisms and few in number. Soon after attestation sets in, Hittite ī-(mi) was replaced by the deponent iye/a-(MP) ‘to go’ (HED E-I:329, EDHIL:375).

[D.S.]

2. Reconstruction

2.1Reconstruction ‘to go’

Considering Hitt. ī-(mi), CLuw. ī-(di) and Pal. ī- ‘to go’, one can reconstruct a root-verb of the mi-conjugation (HEG I-K:335). The Luwian evidence shows that the ending was lenited in Proto-Anatolian, which must come from the accented heavy syllable of the root, i.e. from the diphthong *ei̯. The paradigm is reconstructed as "amphikinetic" from a Proto-Indo-European perspective; see Oettinger 1979a:348f., EDHIL:375f., Sasseville 2021a:316.

[D.S.]

3. Proto-Indo-European

3.1Semantic Reconstruction

PIE *h1ei̯- means ‘to go’ (EIEC:227f., LIV²:232f.). According to EIEC:228, this root is “[p]ractically universal in IE and old. This appears to have been the least marked verb of motion in PIE.” For further roots meaning ‘to go’ vel sim., cf. EIEC:228f.

3.2Morphological Reconstruction

The PIE root *h1ei̯- ‘to go’ is well attested in the IE languages, with a lot of verbal and nominal formations derived from it. As for the verbal stem formation, an athematic root present PIE *h1éi̯-/*h1i-´ can be safely reconstructed on the basis of especially Hitt. (3pl. i̯anzi), IIr. (Ved. éti/yánti, YAv. aēiti/yeiṇti) and Gk. (εἶμι/ἴμεν). Cf. also the well-attested 2sg. imp. Hitt. īt, Ved. ihí, OAv. idī, Gk. ἴθι < PIE *h1i-dhí. For various nominal derivatives of PIE *h1ei̯-, cf. NIL:220–230.

[T.S.]

4. Transmission

Ved. root ay- ‘to go’, pres. áy-/i-´ (RV [émi, éṣi, éti, imási, ithá, yánti; 2sg. imp. ihí, etc.]+);Cf., e.g., EWAia I:102, LIV²:232f. Av. root aii- ‘to go’, pres. aii-/i- (3sg. YAv. aēiti, 3pl. YAv. yeiṇti)Cf., e.g., EWAia I:102, LIV²:232f., ALF:199–201.

Gk. 1sg. εἶμι ‘I go, will go’,“Z.T. sekundär als Futur” (LIV²:233 n. 2). 1pl. ἴμεν; 2sg. imp. ἴθι ‘go!’, etc.Cf., e.g., LIV²:232f., GEW I:462f.

Lat. īre, ‘to go’;Cf., e.g., LIV²:232f. with n. 3, EDLIL:191f. with further lit. Umbr. 2sg. imp. ef ‘go!’, etc.Cf., e.g., Untermann 2000a:207–210, EDLIL:191f.

?Celtib. (ne-)ito ‘must (not) go’, OIr. ethae*, ·eth ‘someone went’, do·eth ‘someone came’Cf., e.g., LIV²:232f. with n. 3c, EDLIL:191f. with further lit., EDPC:170f.

Goth. pret. iddja ‘went’Cf., e.g., LIV²:233 with n. 15 and further lit., EDPG:167, 175.

OPr. 3sg. ēit ‘goes’, Lith. eĩti ‘to go’, 1sg. OLith. eimì, Lith. einù, dial. also eitù, eimù; Latv. iêt, 1sg. iêmu/eĩmu, etc.Cf., e.g., LIV²:232f. with n. 4, ALEW I:259–261, EDBIL:151f.

OCS iti, 1sg. idǫ ‘to go’, Russ. idtí, 1sg. idú, etc.Cf., e.g., LIV²:232f. with n. 5, EDSIL:216.

TochB yaṃ ‘goes’, TochA 3pl. yiñc ‘go’Cf., e.g., LIV²:232f. with n. 6, DTB²:64f.

?Alb. vete, OGeg. ve ‘goes’Cf., e.g., LIV²:232f. with n. 7 and lit.

Cross-references