*dhn̥-no-to-
‘desolate, empty’
- Language
- Proto-Anatolian
- Class
- Base
- Grammar
- adj
- ID
- 4041
1. Individual Anatolian Languages
1.1 Hieroglyphic Luwian
/tannatta/i-/, (“VACUUS”)ta-na-ti
‘empty’1.1.1Transmission
This lexeme is attested five times in inscriptions from Karkamiš, Malatya and Maraş, dating from the 12th to the late 9th century BCE.
[A.P.]
1.1.2Forms
| nom. sg. c. | /tannattis/ | VACUUS-ti-i-sa | KARKAMIŠ A11b+c, §3 (Karkamiš) |
| acc. sg. c. | /tannattin/ | (“VACUUS”)ta-na-ti-i?-na | MARAŞ 8, §3 (Maraş) |
| ta-na-ti | KARAHÖYÜK, §3 (Malatya) | ||
| nom./acc. pl. n. | /tannatta/ | (“VACUUS”)ta-na-ta-’ | MARAŞ 1, §4 (Maraş) |
| VACUUS-tá | KARKAMIŠ A31+, §6 (Karkamiš) |
The attestation in KARAHÖYÜK, §3 looks like a dat./loc. sg. on the surface, but is contextually assumed to be acc. sg. c., agreeing with the preceding noun 〈URBS-MI-ní〉 ‘town’, which also appears to be missing a final 〈-na〉. While the KARAHÖYÜK inscription shows a strong tendency to leave logograms bare of inflectional endings, this practice does not usually extend to syllabically written words in that particular text; the classification as acc. sg. c. here is hence tentative.
[A.P.] [AH.B.]
1.1.3Graphic Features
1.1.3.13.1 Phonological Features
The phonological interpretation of this verb is quite certain, because the transparent HLuw. evidence is further supported by an exact cognate in CLuw. dannatta/i- ‘empty’ (see below) and a number of related words around Hittite dannattaḫḫ- ‘to devastate’ (already Laroche 1950b:51, followed e.g. in HEG T:101).
Mostly included in the attested spellings, the logogram VACUUSMelchert 1988a:32 first proposed VASTUS, which was later altered to the semantically similar VACUUS. (sign *245) was long fixed with an erroneous semantic interpretation because of its shape. It consists of the sign *244, depicting a wall (now interpreted as AEDIFICIUM), and the sign *381, a single vertical stroke, whose meaning was ill-understood until elucidated in detail by Melchert 1988a:32-41 and interpreted as MINUS from then onwards.
[AH.B.]
1.1.4Meaning
First wrongly identified as a noun,The shape of the logogram *245 led scholars astray into thinking it depicted a building, cf. the discussion of the phonological interpretation above. Thus, with an old reading as 〈ta-nu-ta-〉, Meriggi 1934a:139 translated the word as ‘wall’ (“Mauer”) and the related verb /tannatta-(ti)/ as ‘to wall’ (“(um)mauern”). Following the same path, Laroche 1960c:130 interpreted sign *244 as ‘wall’ (“bâtiment”) and *245 as ‘building’ (“édifice”); both have since been corrected to AEDIFICIUM and VACUUS respectively, cf. Payne 2010a:178. the semantic interpretation of this adjective underwent re-evaluation, when /tannatta/i-/ ‘empty’ was connected to its CLuw. cognate dannatta/i- ‘empty’ by Laroche 1950b:51, who then corrected the interpretation to ‘devastated, empty’ (“dévasté, dépeuplé”) without scholarly resonance with the exception of Nowicki 1981a:255f. Laroche’s interpretation was eventually taken up and vindicated by Hawkins & Morpurgo Davies 1986a:74f. and Melchert 1988a:32f. The meaning becomes obvious e.g. in MARAŞ 8, §3: ta?-pa-wa/i ku+ra/i-ku-ma-na(URBS) FLUMEN.REGIO-ti-na (“VACUUS”)ta-na-ti-i?-na wa/i-mi-LITUUS-ha, approx. /anda=ba=wa Kurkuman habadin tannattin wamihha/, meaning ‘Also, I found the Gurgum river-land empty’ (cf. Hawkins 2000a:253).
[AH.B.]
1.1.5Stem
As an adjective, the word is subject to i-mutation, which is evident from the attestations. Melchert 1988a:32 suggests an earlier i-stem, which is possible but cannot be proven without additional evidence due to the obscuring factor introduced by sweeping i-mutation, cf. Hawkins & Morpurgo Davies 1986a:74f.
[AH.B.]
1.1.6Derivatives
1.2 Luwian (Glossenkeilwörter)
dannatta/i-
1.2.1Transmission
The gloss wedge-marked attestations of this adjective are limited to a passage in the Apology of Ḫattušili III., but i-mutated forms occur in the edict of Ḫattušili III. and also in a cult inventory (all New Hittite).
[E.R.]
1.2.2Forms
| nom. sg. c. | dan-na-at-ti-iš | KUB 21.29 i 12 | (CTH 89, NS) |
| dan-na-at-te-eš | KUB 25.21 iii 5 | (CTH 528, NS) | |
| dan-na-at-〈〈ta〉〉-te-eš | KUB 36.89 rev. 41’ | (CTH 671, NS) | |
| acc. sg. c. | dan-na-at-te-in | KUB 36.89 rev. 42’ | (CTH 671, NS) |
| nom./acc. pl. n. | dan-na-at-ta |
KUB 1.1 ii 63 | (CTH 81, NS) |
| KUB 1.1 ii 66 | (CTH 81, NS) |
1.2.3Graphic Features
First part of the attestations checked against the photographs in the Mainzer Fotoarchiv.
[E.R.]
1.2.4Meaning
The first translation of Hittite dannatta- as ‘empty’ offered by Friedrich was confirmed later by Götze 1925a:79f. on account of the entry [íd-a]š dannatteššar ‘emptiness of the river’ (i 8) in the lexical list KBo 1.36, and Götze & Sturtevant 1938a:51 based on contexts parallel to dannara-, likewise ‘empty’. It is used for a wide range on nouns, such as the names of countries and cities, ‘lead’, ‘bear’, and ‘bread’. The semantic match with the i-stem forms and the gloss wedge-marked attestations is obvious; cf. HW:209, HW:333, DLL:90, Starke 1990a:544 n. 2013, CLL:205, HEG T:99, HHw:164, Ünal 2007a:680, Sasseville 2021a:41, DCL:220f.
[E.R.]
1.2.5Stem
While the Hittite lexeme ends in a stem vowel -a-, there are also forms that can be interpreted as i-mutated forms, either Luwianisms or Hittite forms with transferred i-mutation. The forms of the nom./acc. pl. n. in -a are identical in both languages. The derivational chain consists of root plus the suffixes -na- and -tta/i- (contrasting with -ra- in danna-ra-); cf. Götze & Sturtevant 1938a:51, Kronasser 1966a:190, HEG T:101f., Weeks 1985a:171, Starke 1990a:544 n. 2013, Melchert 1999e:369f., DCL:220f. Whether the stem of the acc. sg. dannantan, instr. dannantit etc. emerged by nasal perseveration from dannatta/i- (thus Kronasser 1966a:90, 93, CLL:205, Rieken 2010a:306, HEG T:100 with references; different Starke 1989b:658: vice versa with nasal loss) or constitutes a derivative with -ant- on its own remains unclear. The verbal stem *danna- reconstructed by Starke 1990a:544 n. 2013 on account of dannant- is rejected by Melchert (DCL:220).
[E.R.]
1.2.6Origin
While Starke 1990a:544 n. 2013 seems to assume that the entire word family is Luwian, other scholars endorse a Hittite-Luwian match of lexemes inherited from Proto-Anatolian; see DLL:90, CLL:205, Melchert 1999e:369f., HEG T:99, HHw:164, Sasseville 2021a:41, DCL:220f.; implicit van den Hout 2006a:239.
[E.R.]
2. Reconstruction
2.1Reconstruction ‘desolate, empty’
Weeks 1985a:171 posits a basic stem *danna- as the base of Hittite dannatta-, CLuw. dannatta/i- and Hitt. dannara- ‘empty’ and thus reconstructs a preform *dhn̥-no-to- for the former, which he etymologically connects with Skt. dhánvan- ‘desert, wasteland’ (though already suggested by Sturtevant 1933a:157, 160 and followed by Kronasser 1966a:186, 190, HEG T:99, Melchert 1999e:369f.
It is also still worth considering whether CLuw. tanama(/i)- and dannamma/i- with their possible meaning ‘drawn (?)’ can be adduced to the same PIE root, depending on the semantic development.
[D.S.]
3. Proto-Indo-European
3.1Semantic Reconstruction
The basic meaning of the root *dhen- may be reconstructed as ‘empty, hollow, waste’. The *-r-stem *dhén-r̥ (Ancient Greek θέναρ, Old High German tenar m., tenara f.) likely referred to the hollow hand, whence ‘palm of the hand’ by extension (but compare the old meaning in θέναρ βωμοῖο ‘hollow in the top of the altar, on which the offerings are laid’ (Pi.), ἁλὸς θέναρ ‘hollow bed of the sea’ (Pi.)).
3.2Morphological Reconstruction
The formation *dhn̥-no- that seems to underlie the Anatolian material (Hittite danna-tta-, danna-ra-, etc.) violates a morphophonological constraint that was apparently in place in PIE and the early stages of its daughter branches. As a general rule, suffixes of the shape *-Ri(V)- are dispreferred when the root ends in *-eRi-, a likely consequence of the PIE constraint on heteromorphemic geminates (on which see Byrd 2018a:2071, Nikolaev 2021b), which apparently included *-R̥-R-. It remains to be determined whether such sequences were a priori avoided or if degemination eliminated them retroactively. For the widespread adjectival suffixes *-ro-, *-lo-, *-no-, *-mo-, degemination would have resulted in adjectives apparently formed by *-o- (e.g., *C(C)r̥-ró- > *C(C)r̥-ó-, *C(C)l̥-ló- > *C(C)l̥-ó-, etc.). As the latter is also a widespread adjectival suffix (cf. *(h1)re/ou̯dh-o- ‘red’ > Old Irish ród, rúad, Latin rūfus, rōbus, Proto-Germanic *rauda‑, Lithuanian raũdas, Old Church Slavonic rudъ), such a degemination is difficult to prove. However, as thematic adjectives of a structure R(z)-ó- are far less common than R(o)-ó- and R(e)-ó-, adjectives of a shape *C(C)R̥-ó- are likely candidates of going back to degeminated *C(C)R̥-Ró-. Compare Old Irish gar ‘short, near, easy’ < Proto-Celtic *garo- from a root *g̑her- ‘short’, which must be reconstructed as *g̑hr̥-o-, perhaps from earlier **g̑hr̥-ro-.
On the other hand, it is thinkable that languages realized /-R̥-o-/ as [-R̥R-o-] once they no longer had a constraint against geminates, which may explain Proto-Anatolian *danna- from *[dhn̥n-o-] */dhn̥-o-/, or that Proto-Anatolian, having obtained phonological geminates *-RR- as a result of laryngeal assimilation in sequences *-VRHV- (cf. Melchert 1994a:79), no longer obeyed the constraint against heteromorphemic geminates and could readily derive a *-no- adjective from a root *dhen- (compare Palaic bānnu- ‘liver’ < PAnat. *pén-nu- ‘fattened, liver’).
[S.H.]
4. Transmission
Hittite dannatta- (dannant-) ‘desolate, empty’, dannara- ‘empty’, dannattaḫḫ-(ḫi) ‘make empty’, Luwian Gloss wedge word dannatta/i- ‘desolate, empty’, Hieroglyphic Luwian /tannatta/i/ ‘empty’, /tannatta-(ti)/ ‘to devastate’, etc.See HEG T:97–99 s.v. dannara-, dannarant-, 99–102 s.v. dannatt(a/i)-.
Vedic dhánvan- n. ‘desert, waste, dry soil, shore’ (RV+), dhánu- f. ‘plain, steppe (?) (RV).Cf., e.g., EWAia I:774 s.v. dhánvan-1, 773 s.v. dhánu-, Jamison & Brereton 2014a ad loc. RV 10.27.17.
Ancient Greek θέναρ n. ‘palm of the hand’ (Il.+), θέναρ βωμοῖο ‘hollow in the top of the altar, on which the offerings are laid’ (Pi.), ἁλὸς θέναρ ‘hollow bed of the sea’ (Pi.).Cf., e.g., GEW I:661f. s.v. θέναρ.