*dṷatér-/dṷatr-´
‘daughter’
- Language
- Proto-Luwic
- Class
- Base
- Grammar
- subst
- Certainty
- high
- ID
- 149
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/tuwatr(i)-(?)/, tú-wa/i-tara/i-, FILIA
‘daughter’1.1.1Transmission
The earliest attestation of the word /tuwatr(i)-(?)/ is found in the inscription of KELEKLİ (Karkamiš) dated to the 10th century BCE. However, its full syllabic writing is found in the later inscription of TELL AHMAR 1.
[D.S.]
1.1.2Forms
| acc. sg. | /tuwatrin/ | (FILIA)tú-wa/i-tara/i-na | TELL AHMAR 1, §24 (Tell Ahmar) |
| (FILIA)tú-wa/i-ta[ra/i-na] | TELL AHMAR 1, §29 (Tell Ahmar) | ||
| FILIA-tara/i-na | KELEKLİ, §2 (Karkamiš) |
1.1.3Graphic Features
1.1.3.1Phonological Interpretation
See under Reconstruction.
1.1.4Meaning
The meaning ‘daughter’ was elucidated by Hawkins 1978a. In KELEKLİ (Karkamiš), the ruler Suhis relates that king Tudhaliya is to take his daughter in marriage. In TELL AHMAR 1, the ruler, son of Ariyahina, explains how he wants to dispose of his enemy’s daughter (Hawkins 2000a:239-243, Payne 2012a:105).
[D.S.]
1.2 Cuneiform Luwian
*du(wa)ttar(i)-
‘daughter (?)’1.2.1Derivatives
1.2.1.1Uncertain Derivatives
1.3 Luwian in Hittite transmission
DUMU.MUNUS-iš
‘daughter’1.3.1Transmission
Attested only in the Tale of Appu and His Sons in LNS.
[Zs.S.]
1.3.2Forms
| nom. sg. c. | DUMU.MUNUS-iš | KUB 24.8 i 16 | (CTH 360, LNS) |
The attestation follows Starke 1987a:253, checked against the photo of the Mainzer Fotoarchiv. The text features this word only once more, a fragmentary form in another copy with a problematic reading (KUB 36.59 obv. i 4’’), which is usually transliterated as [DUMU].MUNUS-aš!-ša (Siegelová 1971a:6; Starke 1987a:253 [quoted as i 3’]; Rieken et al. 2009a [under i 3’’]; Weeden 2011b:205 [quoted as ii 4]).
[A.B.], [Zs.S.]
1.3.3Graphic Features
Collation of the photo shows that emendation to aš! is problematic since the fragmentary sign begins with two horizontals. Hence, it might as well be another iš. Moreover, the following sign can hardly be identified as ša.
[A.B.], [Zs.S.]
1.3.4Origin
Since the Hittite word for ‘daughter’ ends in -la (DUMU.MUNUS-la-an, KBo 20.101 rev.? 3’, it is not attested yet in full spelling, HEG T,D:471; EDHIL:903), this unusual phonetic complement requires explanation. Starke 1987a:253f. suggests that it is to be explained with the Luwian words for ‘daughter’ attested as “duwattrinn(i)-” and “duttarriyatti-”, where a more precise definition is not possible (also Weeden 2011b:205 registers this form, but without further discussion; it does not appear in the list of Luwianisms by Melchert 2005a and van den Hout 2007a). However, as discussed under these entries, a meaning ‘daughter’ for these words cannot be verified.
Nevertheless, it is theoretically possible that the only assured Luwian word for ‘daughter’ (HLuw. tuwat(a)r(i)-) may indeed be hidden behind this spelling. Unfortunately, a Hittite i-mutation (for which see Hoffner; Melchert 2008a:86f. with refs.) of DUMU.MUNUS-la- and thus a Hittite origin cannot be excluded either.
[Zs.S.]
1.4 Lycian A
kbatra-
‘daughter’1.4.1Transmission
The word kbatra- is found several times in funerary inscriptions and once in a dedicatory inscription to Apollo (TL 25.6).
[D.S.]
1.4.2Forms
| nom. sg. | kbatra | N 309d.10 (Myra), TL 87.5 (Myra) |
| acc. sg. | kbatru | TL 25.6 (Tlos), TL 28.2 (Tlos) |
| dat. sg. | kbatri | TL 138.2 (Limyra), TL 143.5 (Limyra) |
For the grammatical interpretation of the form in TL 87.5 see DLL:30, Schürr 2010b:120. For the syntax in TL 143.5 see Neumann 2007a:159.
[D.S.]
1.4.3Meaning ‘daughter’
The meaning of the noun kbatra- c. was elucidated through the Lycian-Greek bilingual inscription TL 25, where it is rendered with Greek θυγατήρ (Imbert 1894a:455-457); cf. DLL:30, Neumann 2007a:159-160, Schürr 2009b:112. It was connected etymologically with PIE *dhugh2ter- ‘daughter’ already by Bugge 1901a:25 (followed by Meriggi 1936b:260, Laroche 1967a:48, Starke 1987a:247).
[D.S.]
1.5 Pisidian
*δωταρι-
‘daughter (?)’1.5.1Transmission
This lexeme is only attested as a personal name in the Pisidian corpus.
[D.S.]
1.5.2Forms
| nom. sg. | Δωταρι | N 1 (Sofular), N 1 (Sofular) |
| gen. sg. | Δωταρις | N 1 (Sofular) |
1.5.3Meaning
The meaning and the etymology of the personal name Δωταρι- was suggested by Starke 1987a:258. The exact attribution of the people on the relief, who bore the name Δωταρι- as well as the gender attribution of this name are debated; cf. Starke 1987a:258, Brixhe 1988a:136. However, if the personal name were to refer to a man, it would not represent sufficient evidence to argue against the etymological connection with PIE *dhugh2ter- ‘daughter’, pace Schürr 2006b:1570 n. 2. Nevertheless, the etymological connection remains highly hypothetical.
[D.S.]
1.5.4Compounds
1.6 Lydian
tutra-
‘daughter (?)’1.6.1Transmission
This lexeme is a hapax legomenon and it is attested in an inscription of Sardis dated to the 4th century BCE.
[D.S.]
1.6.2Forms
| nom. sg. (?) | tutra(s) | LW 80.12 (Sardis) |
1.6.3Graphic Features
Sandhi effect: The word is followed by the enclitic conjunction =k ‘and’, which eliminates a preceding s, ν, t or d (Gusmani 1964a:35). Schürr 2006b:1571 assumes a dat./gen. pl. based on the coordination with the preceding word tẽmwaν tutra=k and the relative pronoun qλ attested in the preceding line and thus, he reconstructs the letter (ν). However, tẽmwaν must not agree in case with the following word tutra=k. The interpretation as a nominative singular for tutra=k is preferable for reasons given below.
[D.S.]
1.6.3.1Phonological Interpretation
If the etymological assignment is correct, then the word cannot be accented on the first syllable, since an accented sequence *-(u)u̯a- yields Lydian -o- (Melchert 1994a:346). Then, considering the hypothesis that the word is oxytone, the ending cannot be accusative singular nor dative/genitive plural, otherwise one would expect, -ẽ(ν)=k or -ã(ν)=k. From a phonological perspective, the word tutra=k should be parsed as nominative singular, i.e. tutra(s)=k. Since the passage is fragmentary, this decision remains tentative.
[D.S.]
1.6.4Meaning
Gusmani 1980a:103 does not offer any approach regarding the semantics of the word tutra-. The meaning ‘daughter’ is attributed by Schürr 2006b:1570-1572, who equates etymologically tutra- with Luw. /tuwatr(i)-/ and Lyc. kbatra-. Although the context is obscure, Schürr argues that the contextual occurrence of tutra- near taada- c. ‘father’ supports the meaning ‘daughter’ for the former. Considering the fact that the passage is fragmentary, the semantic assignment remains far from assured.
[D.S.]
1.6.5Stem
If the word is indeed oxytone, then it may belong to stems in -a-(ã) or in -a-(ẽ). Based on a comparison with Lyc. kbatra-, the former nominal stem class is preferable (differently Schürr 2006b:1571).
[D.S.]
1.6.6Derivatives
2. Reconstruction
2.1Reconstruction ‘daughter’
Based on the Anatolian material alone, the present lexeme for daughter can only be reconstructed for Proto-Luwic. Starke 1987a:262 argues convincingly that a hysterokinetic paradigm was still alive before the separation of the Luwic languages. On the one hand, the strong stem *dṷatér-ṃ survived in CLuw. duttarri(ya)- (with contraction of -uwa- > -u- in the initial syllable). The presence of *é is supported by the geminate on account of Čop’s law, contra Kloekhorst 2011b:241. The same stem form is preserved in Pis. Δωταρι-, which implies that the letter <ω/o> would reflect *-ṷa- after a dental. On the other hand, the stem’s form *dṷatr-´ of the weak cases is continued by CLuw. PÚDuwattrina-, the Isaurian personal name Τουατρι- and Lyc. kbatra-. The phonological interpretation of HLuw. tú-wa/i-tara/i-na is ambiguous. According to Starke 1987a: 261, either interpretation /tuwatrin/ or /tuwattarrin/ is possible and the retention of the sound sequence -uwa- does not favor any of them. Nevertheless, following the Isaurian personal name, the former variant is provisionally adopted here.
Regarding the diverging stem class of Pisidian or Isaurian and Lycian, on can posit that the Lycian stem kbatra- reflects a secondary transfer to -a- from a substantive in *-ter-. Pace Oettinger 1986a:11, this change to an a-stem cannot be old, see Starke 1987a:262.
The initial /d/ of the Pisidian name Δωταρι- is suspicious, since we expect all inherited voiced stops to have undergone fortition in word initial position in all Anatolian languages. However, the fortition of initial stops is a Post-Proto-Anatolian sound change, which spread across the individual Anatolian languages at different periods (Melchert 2003e:19). Thus, it is not impossible that the phonological sequence *du̯V- retained the voicing of the initial alveolar longer in certain Luwic dialects than in others. The retention of the voice feature can be posited for Pre-Lycian, since the dialects show different outcomes, */du̯/ > Lyc. A kb-, Lyc. B tb- (or zb- /_*(u̯)i), with Melchert 1994a:285, 287. The Pre-Lycian starting point */du̯/ instead of */tu̯/ is made preferable (contra Laroche 1967a:48) by pairs such as Lyc. A xãkbi vs. Lyc. B xãzbi (< Proto-Lycian *χándaṷe/i-) and Lyc. A Ikkuweme/i- (PN) vs. Gk. Ενδυομις, among which the sound change occurred with the cluster /nd/ (< *-nt-), where an etymological *t was neutralized to [d] after *n. Further evidence can be drawn carefully from Lydian fa-do- << *du̯ó- < *duu̯ó-, see Gérard 2005a:45. To sum up, we do not know enough about the Pisidian phonology to be absolutely certain that the retention of the initial d- is an archaism. Further data on the Pisidian language will hopefully verify this hypothesis.
Starting from the Proto-Indo-European pre-form *dhugh2ter-, Melchert 1994a:69 rejects the previous assumption that *h2 could have been vocalized in any environment. In order to explain the vowel /a/ preserved in Proto-Luwic *dṷatér-/dṷatr-´, he posits an anaptyptic vowel (contra Kimball 1999a:388) arising in the triple cluster: *dhugh2tr- > *dugtr- > *dugətr- > *dugatr-. Furthermore, Melchert 1994a:255-256 posits that the loss of *g occurs intervocalically in Luwian, but admits that the evidence is sparse. Yakubovich 2010d:244 posits that *g was lost after *ṷ/u in Luwian and compares *lóṷkV- > *lóṷgV- >> Luwiya-. However, this is refuted by Rieken (forthcoming), who starts from a different root, *lukṷ-, vis-à-vis Luwiya- and applies the rule (see Melchert 1994a:61) that the labiovelar was lenited in word internal position in Proto-Anatolian followed by the regular Luwian sound change of *gṷ > ṷ: *lukṷ-i̯o- > *lugṷ-i̯o- > Luwiya-.
Due to the lack of evidence from Hittite or Palaic, it is not possible to posit the exact phonological shape of this lexeme for Proto-Anatolian.Kloekhorst 2008c:9, 902-904 (and Kloekhorst 2011b:240) rejects the explanation of Melchert 1994a:69-70 with an anaptyptic vowel based on the unconvincing idiosyncratic view of the syllabic outcome of a consonant cluster and proposes an unusual root ablaut pattern *dhṷégh2tṛ/*dhugh2térṃ/*dhugh2trós, which is not accepted here; the anaptyptic vowel, which Melchert posits, would have occurred after the Proto-Anatolian period, whence it must not apply for Hittite. For a more complete summary of the literature, cf. HEG:472-474, LIV:126-130.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
PIE *dhugh2tér- f. means ‘daughter’ (EIEC:147f., NIL:126, Watkins 2011a:21). The noun is preserved in the majority of the IE languages. In a few languages, however, other words for ‘daughter’ can be found: cf. Lat. fīlia f. ‘daughter’, OIr. ingen f. ‘daughter’, MW merch f. ‘daughter’, Latv. meîta f. ‘daughter, girl’, Alb. bíjë f. ‘daughter’.Cf., e.g., Buck 1949a:105f., EIEC:147f., ALEW I:241.
3.2Morphological Reconstruction
The attested forms allow the reconstruction of a hysterokinetic paradigm *dhugh2tér-/*dhugh2tr-´: PIE nom. sg. *dhugh2tḗr (Ved. duhitā́, OAv. dugǝdā, Gk. θυγάτηρ), acc. sg. *dhugh2tér-m̥ (Ved. duhitáram, YAv. duγδarəm, Gk. θυγατέρα), gen. sg. *dhugh2tr-é/ós (Gk. θυγατρóς), dat. sg. *dhugh2tr-éi̯ (Ved. duhitré), etc.Differently Kloekhorst 2011b, who reconstructs a paradigm with nom. sg. *dhu̯égh2-tr̥, acc. sg. *dhugh2-tér-m̥, gen. sg. *dhugh2-tr-ós. As for the root accent of the nom. sg., he refers in particular to Gk. θυγάτηρ, which, in his view, points to an original root-accented form *θύγατηρ, with a secondary accent shift to the α due to the long last syllable. However, it is also possible that the attested nom. sg. θυγάτηρ replaced an original form *θυγατήρ with final stress (cf. NIL:128 n. 6 with lit.), which would be in accordance with Ved. duhitā́ and Lith. duktė̃. In pre-PGk., the inherited accent of the form *dhugh2tḗr was then shifted to the first syllable in analogy to the regular voc. sg. *dhúgh2ter (Gk. θύγατερ, Ved. dúhitar, OLith. dùkter; cf. Schwyzer 1939a:381) or to the nom. sg. of the word for ‘mother’ (PIE *méh2tēr, cf. NIL:457–461, esp. 459 n. 1). Finally, pre-PGk. *dhúgh2tēr regularly gave PGk. *thúgatēr > Gk. θυγάτηρ, with regular inner-Greek accent shift from the antepenultimate to the penultimate syllable due to the long final syllable.
The quality of the word-internal laryngeal is particularly ensured by Gk. θυγάτηρ, the medial α of which points to vocalized *h2. The reconstruction of *h2 also fits the aspirating effect of the laryngeal seen in Indo-Iranian:According to the communis opinio, the effect of aspirating a preceding stop was restricted to PIE *h2: “*T + *h1 bzw. *h3 (___V) haben nicht zur Behauchung von *T geführt”, as per Mayrhofer 2005a:115. Differently, e.g., Beekes 1988:63, Olsen 1988, and Olsen 1994a, who propose *Th1 > *Th). Ved. duhitár-, on the one hand, seems to show both aspiration of *g by the following *h2 and laryngeal vocalization to i, thus suggesting a development PIE *dhugh2t° > (pre-)PIIr. *dhughə2t° > PIIr. *dhughit° > *dhuǰhit° > Ved. duhit° vel sim. The Avestan forms, on the other hand, likewise reflect an aspiration *gh2 > *gh, but show no trace of a vocalized laryngeal, i.e. PIE *dhugh2t° > PIIr. *dhught° > *dhugdh° (Lex Bartholomae) > OAv. dugǝd° (with anaptyctic vowel ǝ), YAv. duγδ° vel sim. (cf. Mayrhofer 1986:137, EWAia I:737f., Mayrhofer 2005a:55f., 115f., Lipp 2009a II:351–398).
Apart from Gk. θυγάτηρ and Ved. duhitár- a vocalization of the laryngeal – without exact information about its quality – can also be noticed in Tocharian (TochB tkācer = /tkácer/) and Celtiberian (tuater-). As far as the Anatolian languages are concerned (cf. Lyc. kbatra-, HLuw. tú-wa/i-tara/i- < Proto-Luwic *du̯atér-/*du̯atr-´, see above), a vocalization of the laryngeal to a is accepted by Kimball 1999a:388, but rejected by Melchert 1994a:69, who explains the word-internal a-vowel as an anaptyxis, with PIE *dhugh2tr- > PAnat. *dugtr- (laryngeal loss) > *dugetr- (anaptyxis) > *dugatr- > PLuw. *du̯atr-. Other languages (Iran. [see above], Arm., Balt., Slav., Germ., Ital., Celt. [except Celtib.]) show no laryngeal vocalization. On the question of laryngeal vocalization and laryngeal loss, cf. also NIL:127 n. 1 with further lit. and the thorough discussion in Lipp 2009a II:351–398.
The exact morphological analysis and etymology of PIE *dhugh2tér- is not clear. It can be compared with other kinship terms ending in *°(h2)ter- like PIE *ph2tér- ‘father’ (NIL:554–562), *méh2ter- ‘mother’ (NIL:457–461), and *bhréh2ter- ‘brother’ (NIL:38–41). Some scholars, therefore, consider a special kinship suffix *-h2ter- (e.g., Sihler 1988), which is controversial (cf. Pinault 2007b:276). In the case of PIE *dhugh2tér- ‘daughter’ (as well as *méh2ter- ‘mother’), an analysis as a hysterokinetic agent noun in *-tér- seems problematic, as the word has feminine gender without being formed with the feminine suffix *-ih2-, which is typical of feminine stems formed to agent nouns in *-te/or-; cf. PIE *g̑énh1-tor- m. ‘begetter’ (Ved. jánitar-) : *g̑enh1-tr-ih2- f. ‘mother’ (Ved. jánitrī-).
3.3Syntactic Reconstruction and Phraseology
In PIE mythology and religion, the word combination *dhugh2tḗr diu̯ós ‘daughter of the day-lit sky’ can be reconstructed on the basis of Ved. duhitā́ divás f. (RV+, epithet of the goddess of dawn, Ved. uṣás-), Gk. θυγάτηρ Διóς f. (Il.+, epithet of different goddesses “that fulfill different inherited functions of Eos”: Jackson 2006:53 with lit.), and perhaps also Lith. diẽvo duktė̃ f. (epithet of Saulýte, a mythological figure). Cf., e.g., Schmitt 1967a:169–173, EIEC:230, NIL:126.
[T.S.]
4. Transmission
Ved. duhitár- f. ‘daughter’ (RV+): nom. sg. duhitā́On the basis of the nom. sg. duhitā́ a secondary ā-stem duhitā- (ep., class.) was backformed (e.g., acc. sg. duhitām); cf. AiGr III:324f., NIL:127 n. 2 with further lit., acc. sg. duhitáram, dat. sg. duhitré, nom. pl. duhitáras. – OAv. dugǝdar-, YAv. duγδar- f. ‘daughter’: nom. sg. OAv. dugǝdā, YAv. duγδa, acc. sg. YAv. duγδarǝm, gen. pl. OAv. dugǝdrąm, YAv. duγδrąmCf., e.g., EWAia I:737f., ALF:150–152, NIL:126f. with n. 3.
Myc. tu-ka-te(-qe) nom. sg. f. /thugatēr(-ku̯e)/; Gk. θυγάτηρ f. ‘daughter’ (Il.+): nom. sg. θυγάτηρ, acc. sg. θυγατέρα, gen. sg. θυγατρός, dat. sg. θυγατρί, nom. pl. θυγατέρες, acc. pl. θυγατέρας, gen. pl. θυγατρῶν, dat. pl. θυγατράσι, etc.Cf., e.g., GEW I:690, LfgrE II:1069–1072, DMic II:374, LSJ s.v., NIL:126, 128 nn. 6–7, EDG:561.
Arm. dowstr ‘daughter’ (gen./dat. sg. dster, nom. pl. dsterk‘, gen./dat. pl. dsterc‘ or dsterac‘, instr. pl. dsterawk‘, etc.)Cf., e.g., Olsen 1999a:148f., NIL:126, 127f. n. 5, EDAIL:244f.
Osc. futír f. ‘daughter’ (dat. sg. fuutreí, futreí, ?gen. sg. futre[ís)Cf., e.g., Untermann 2000a:306f., EDLIL:253, NIL:126
Goth. daúhtar f. ‘daughter’, ON dóttir f. ‘id.’, OE dohtor f. ‘id.’, OS dohtar f. ‘id.’, OHG tohter f. ‘id.’, etc.Cf., e.g., Casaretto 2004a:541, NIL:126, 128 n. 8, EDPG:107.
TochA ckācar ‘daughter’, TochB tkācer ‘id.’Cf., e.g., NIL:126, 129 n. 17, Adams 2013a:330f., Hackstein 2017a:1322f.