Published online by Cambridge University Press: 22 April 2026
The Hitt. suffix -tar gen. -nnas appended to verb stems forms action nouns. Examples are a-ša-tar dat. a-ša-an-na beside e-eš-zi ‘sets’ pl. a-ša-an-zi midd. e-ša e-ša-ri (with reflexive -za) ‘seats himself’ pl. e-ša-an-da e-ša-an-ta-ri, and ú-wa-tar dat. ú-wa-an-na beside a-uš-zi ‘sees’ pl. ú-wa-an-zi midd. u-wa-it-ta u-wa-it-ta-ri ‘appears’. Benveniste has pointed out the etymology of the suffix; it is identical with a hitherto unrecognized IE suffix -ter -tnos. One difficulty remains to be cleared up; and some additional material on the IE side can be supplied. Hence the present paper.
1 Origines de la formation des noms en indo-européen 103–10 (1935); cf. Sturtevant, AJP 59.96 (1938).
2 See below 207.
3 Lang. 10.235–65 (1935); 19.83–124 (1943).
4 Edgerton writes -tṇn-.
5 See Sturtevant, HG 226 and references; IH Laryngeals 44.
6 For the ending, see below 209.
7 HG 250 f.
8 HG 74–84 and references (1933).
9 For details see Sturtevant, Lang. 16.82 f. and references (1940); IH Laryngeals 34 and fn. 6 (1942).
10 AJP 59.96 (1938).
11 For -ar instead of -er, cf. HG 90 f. But quite possibly -dar comes from IH -tōr; cf. Hitt. wa-a-tar beside Gk. ὕδωρ.
12 Lang. 16.83 fn. 11 (1940).
13 Lang. 6.157 (1930); cf. HG 84.
14 Origines 103 fn. 1.
15 Caractère indo-européen de la langue Hittite 141 f. (1919).
16 Lang. 16.85 (1940); IH Laryngeals 50 f., 65 (1942). In the article, correct IH γuwid-to IH γъwid-. The assumption that -dn- remained in Hittite has to be given up in view of 207 above.
17 This dat. is doubtless the source of the Gk. infin. εἰδέναι except for the analogical full grade of the root and the final -αι that has been generalized in Gk. infinitives.
18 Brugmann, AJP 8.441–7 (1887), suggested a connection of the Lat. gerund and gerundive with some of the forms listed above as evidence for the IE suffix -ter -tnos, notably the OP infinitive ending -tanaiy. He laid particular stress upon the Lithuanian participia necessitatis in -tinas, such as sùktinas ‘torquendus’ beside súkti ‘torquere’. In Grundr. 21. 1424–6 (1892), he withdrew the suggestion, chiefly because the Lithuanian adjectives seem not to be inherited from Proto-IE. Without the Hitt. evidence, the derivation of Lat. -ndo- from IE -tno- was merely one of several possibilities.
19 AJP 8.445 (1887).
20 Handbuch2 616 (1914).
21 For the final vowel, see the next paragraph. For the IE development of IH ъ plus any laryngeal, see IH Laryngeals 71 f.
23 Forrer, Altorientalische Studien 1.30–5 (1928); Sturtevant, HG 166; Holger Pedersen, Hittitisch und die anderen IE Sprachen 27 f. (1938).
24 See above 206.
26 TAPA 74.
27 KZ 26.301–14 (1883).
28 Lang. 19.299 f. (1943).
29 Some scholars read κέχονδα in this Homeric perfect, and I followed them in Lang. 19.299 (1943).
30 Cf. Pedersen, Vergl. Gramm. d. Kelt. Sprachen 1.160.
31 Sommer, Handbuch2 232 f. (1914); Leumann, Lateinische Grammatik 155 (1928).
32 Buck, A Grammar of Oscan and Umbrian 84.
33 Sommer, Handbuch2 235; Leumann, Lateinische Grammatik 155, and references.