ORIGINAL h IN HITTITE AND THE MEDIO-PASSIVE IN ↑

E. H. STURTEVANT

YALE UNIVERSITY

I have tried to show (LANGUAGE 3. 109-22) that Hittite h occurs where the Indo-European languages have bh initial, and I can now make two or three additions to the material there presented. In LANGUAGE 4. 122 f. I have connected halanta 'head' with Greek paλós λevкós, yaλakpós ‘bald', and Sanskrit bhālam 'forehead', on the assumption that the original meaning of the Hittite word was 'bald'. Hittite haš-, hašk- 'open',1 must be identified with IE *bhosos 'naked', whence Lithuanian bãsas 'bare-foot', Old English bær, etc. The verb pihiš, which, accompanied by the adverb arha, means 'strike off, cut off, take off' or the like,' contains the verbal prefix pe-, while hiš- is an extension in s of the root which appears in IE *bhei- 'strike' (Old High German bihal 'axe', etc.), and whose extended form *bheid- means 'split' (Sanskrit bhinadmi, Latin findo, etc). Another instance of h = bh after the verbal prefix pe- is pehar(k)- 'hold towards' (Keilschrifttexte aus Boghazköi 4. 2. 2. 25, Keilschrift-Urkunden aus Boghazköi 13. 4. 4. 37) beside har(k)- 'have' on which see LANGUAGE 3. 117 f.

The only Hittite word in which I can find IE bh medial is nepiš 'sky', which must be Greek véços, Skt. nabhas 'cloud'. This alone would be too narrow a basis upon which to erect a phonetic law; but there is some further reason for thinking that Pre-Indo-European medial bh did in fact yield the Hittite sound which is written either b or p. For that would harmonize with the development of the other aspirates into the corresponding Hittite stops (dh>t/d, ĝh and gh>k/g, gh>ku/gu). Examples of these changes were given in LANGUAGE 3. 121, except for the development of ĝh. This last is illustrated by gimmanza 'winter',' beside Gk. xeiua, Skt. hemantas 'winter', etc.

1 See Ehelolf, Kleinasiatische Forschungen, 1. 1412, (1927).

2 See Sturtevant, Transactions of the American Philological Association 58. 19 (1927).

See Walde-Pokorny, Vergleichendes Wörterbuch der Indogermanischen Sprachen 2. 137 f.

4 See Sommer, Boghazköi-Studien 4. 18-23.

Nevertheless medial h is very common in Hittite words, a considerable number of which have obvious IE etymologies. Since, however, there is nothing in the IE cognates to correspond to the Hittite h, we must conclude either that the h is a Hittite innovation or that the IE languages have here lost a sound which Hittite has preserved. The second alternative is certainly to be preferred, since Hittite medial h stands in such various surroundings that it would be impossible to define any conditions under which it could have been developed.

Medial h follows a consonant in ešhar 'blood’,5 which is the same word as Gk. čap, Skt. aszk 'blood'. The genitive ešnaš indicates that h was lost between consonants in Hittite.

Parallel with the pair ešhar: čap stands tešhaš dream's: Deós 'god'. That the Gk, word has lost intervocalic s is shown by copatos 'spoken by a god' and several other compounds. Some scholars connect Deós with MHG getwās 'ghost', Lith. dvēsti 'breathe', etc.; but Bartholomae (Wochenschrift für Klassische Philologie 17. 677—1900) was right in maintaining that some trace of an original Of would surely have survived. Even those who assume that 0eós is from *8 fecos may be willing to connect our Hittite word with Lat. fānum (<*fas-nom), Oscan fiísnú 'temple' (*fēs-no-), Lat. fēriae 'festival', and Armenian dikh 'gods' (IE *dhēses).

The verb išhuwa-'throw, scatter, shoot (an arrow)' stands in a parallel relationship with Gk. iós 'arrow', except that the Gk. word has gone over to the o-declension. The original u-stem survives in Skt. işuş ' ‘arrow'. We shall presently (163) find another Hittite word which preserves original h after s.

The group lh appears in walh- 'strike', which I have connected (LANGUAGE 3. 220) with Gk. Balw 'hit, throw' (Arcadian co dovtes) on the basis of PIE **gųelh-, **gļļh- The Hittite word is almost identical in meaning with Ballw in its most primitive force, as in Il. 16. 806 f.:

όπιθεν δε μετάφρενον οξέι δουρί

ώμων μεσσηγύς σχεδόθεν βάλε Δάρδανος ανήρ, ‘and from the rear a Dardanian hero hit his back between the shoulders with his sharp spear'.

• See Sommer and Ehelolf, BoSt. 10. 18. The etymology was proposed by Ehelolf, Ebert's Reallexikon der Vorgeschichte 1. 131. 6 See Friedrich, Indogermanische Forschungen 43. 258, (1926).

See Sommer and Ehelolf, BoSt. 10.53; Götze, Hattušiliš 103f. The etymology was suggested by Kretschmer, KF 1. 10 (1927), but with hesitation that I cannot understand. At the same time Kretschmer identified išhiya- 'bind with Skt. syati, sināti 'bind.' In this he may well be right, but I am not yet ready to give up my (LANGUAGE 3. 221) connection of the Skt. word with Hittite šai 'he puts on, seals.'

Hittite tarh- (tar-ah-zi, tar-hu-du, etc.) 'conquer's belongs to PIE **ter- 'tremble, cause to tremble, flee, put to flight'. In the IE languages the root usually shows a final s, as in Lat. terreo 'frighten' (<*terseiō), Gk. Tpéw 'flee' (<*tresō); but s is lacking in Skt. taralas 'trembling', and with a different root-determinative we have Gk. Tρéμw, Lat. tremo 'tremble', etc. Approximately the Hittite meaning appears in Umbrian tusetu 'fugato'. The other meaning of the Hittite verb. 'be able', must have grown out of the meaning 'conquer'.'

As tarh- is to IE *ter-, so is parh- 'drive out, banish; drive hard, speed'10 to IE *per- 'pass through, carry through', whence Gk. Teipw 'pierce', Lat. porto 'carry', Goth. faran, OE ferian 'travel'.

Medial h appears before s in tuhs- 'separate, part, end'," which I would connect with Gk. déoμal, devouaι 'be lacking' (<*devσouai), Skt. dosas 'defect', duşyati 'he spoils', etc.

Hittite pahs- 'protect'12 must be somehow related to IE *pō(i)'protect' (Skt. pā), on which see Walde-Pokorny, VWIS 2. 72. The s is no doubt the same formative element which appears in many IE and Hittite verb stems and tense stems.

Hittite weh- 'turn' (intransitive),13 is identical in meaning with the IE roots *ver- (Lat. verto, Skt. vṛṇakti, Lith. verpti 'spin', etc.), *uel(Lat. volvo, Skt. vrņoti 'cover', etc.), *uei- (Skt. vayati 'weave', Lith. výti, Lat. vieo 'plait', etc.).14 One must suppose, then, either that PIE had such forms as **uehrt-, **uehrneg1⁄4-, **uehrp-, etc., or that h in this word is a root-determinative analogous to r, l, and i of the corresponding IE roots. The second alternative is certainly the more attractive. Similarly I am inclined to consider the h of pahš- 'protect'

15

16

a determinative parallel with the i of Gk. rolunu 'shepherd'. This will enable us to connect with pahš- the curious form palahšaš, which seems to mean 'protection' (see Götze, Hatt. 75), and which reminds one of Skt. pālas 'protector' beside pāti 'he protects'.

The probable reading in KUB 14. 3. 1. 7 is LÒtu-uh-kán-ti-in, and the variant writing by the Assyrian word LÙTAR.TE.NU in line 9 proves that the meaning is "commander-in-chief'. I would connect this noun tuhkaš with Lat. dux commander', OHG heri-zogo 'armycommander', and the related words.

If Hrozný and Zimmern-Friedrich are right in translating Gišmahlan (acc.) by ‘apple trees', this must be the word that appears in Gk, as uslov, and in Lat. as mālum. It is likely, however, that the three languages have borrowed from the same source or from related sources.

Hittite h appears between vowels in pahhur (earlier pahhuwar) 'fire', gen. pahhuenaš, whose connection with the r/n-stem Gk. Tūp, OE fiur, Goth. fōn, gen. funins, was recognized by Friedrich, Zeitschrift der Deutschen Morgenländischen Gesellschaft 76. 159 (1922). This stem shows so many analogical forms in the IE languages that it is not easy to trace their relationship with certainty. I will not now try to rearrange the material in the light of the Hittite evidence; but it seems clear that we must start from an ablaut base **peheyor/n, or the like. Hittite pahhuwar may represent **pohuór, and pahhur is a contraction of pahhuwar.

Similarly lahuwa- 'pour', lahhuš 'basin', and IE *lou- 'wash' (Gk. Low, Lat. lavo, etc.) must be traced to a PIE ablaut base **loheu-, whence **lóhu-, the source of the Hittite words and perhaps also of IE *lou- 17

Hittite mehur, gen. mehunagis occurs frequently in the phrase nekuz mehur, which means 'evening'. The meaning of nekuz is 'of night' (= Lat. noctis), but whether mehur itself means 'time' or 'point of time, moment' is not yet clear. The word may be connected with IE *meu'move' (Lat. moveo, Lith. máuti 'aufstreifen, anstreifen') through **méhu- from an ablaut base **meheu-. The meaning ‘moment' would then offer a parallel to Lat. momentum (temporis) from moveo.

There are also a number of Hittite words in which initial h must be

16 Hrozný, Code Hittite 106 f.; Zimmern and Friedrich, Hethitische Gesetze 21.

17 This etymology and the preceding one were briefly presented in LANGUAGE 3. 121 f.

18 See Sommer, BoSt. 7. 32–6.

regarded as an original sound that was lost in IE. The word for 'grandfather' is huhhaš,19 which is to be identified with Latin avus on the basis of PIE **hauhos. Similarly hannaš 'grandmother'19 goes with OHG ano 'grandfather', ana 'grandmother', Lat. anus 'old woman', etc.

The prior element of hantezziyaš, hantezziš 'first' should be *hanta 'forward, before', to harmonize with appa 'back, afterwards' beside appezziyaš, appezziš 'last'. This *hanta must be identified with Gk. ǎvra 'Opposite'. The temporal force which appears in Hittite is familiar in the related Lat. ante 'before' (probably locative, while ǎvra is probably accusative, of a noun meaning 'face'). In another connection I hope soon to discuss the final member of these two Hittite compounds.

I have shown (TAPA 58. 23) that hašduir means 'branches'. Although the formative material at the end of the word is not clear, we must identify hašd- with Gk. öços, Gothic asts 'branch'. It follows that sd is the nil-grade of **sed- 'sit', and that Hittite ha- and IE o- both come from PIE **ho-.

Hittite hameshaš, hamešhanza20 is certainly a season of the year suitable for military operations. After mentioning the winter the historical texts frequently introduce the account of the next campaign with the words: ma-ah-ha-an-ma ha-me-es-ha-an-za ki-šá-at, 'but when it became h.' Such a context admits either 'spring' or 'summer'; but KUB 4. 4. 3. 2 ff. (cited by Götze) recommends that the king devote the h. to warfare, and that, of course, must be 'summer'. Ehelolf's attempt, on the basis of a mutilated text, to establish the meaning 'early' for our word does not convince me. On the contrary I should conclude that the AN.TAH.ŠUM festival which Ehelolf is there considering, was a harvest festival. I connect this word with the IE root *mě- 'reap', which appears in OHG mãen, OE máwan with a formative į, and in Lat. meto with a formative t. Gk. åμáw 'reap' and aμaw, åμáoμaι 'cut; gather' present a number of difficulties, but no one

19 See Forrer, 2 Boghazköi-Texte in Umschrift 22*, Forschungen 1. 90; Friedrich, Archiv für Orientforschung 4. 95 (1927).

20 See Hrozný, BoSt. 3. 1904; Sommer, BoSt. 4. 20; Götze, Hatt. 92; Ehelolf, Sitzungsberichte der Preussischen Akademie der Wissenschaften, Phil. -Hist. Klasse 21.2672 (1925).

21 See especially Bechtel, Lexilogus zu Homer 36 f., and Walde-Pokorny, VWIS 2. 259, and references. I am inclined to agree with Schulze that the long initial syllable of aμáw 'reap' is due to a prefixed år' (properly *àμμáw). It is noteworthy that the syllable is long only where the context favors the meaning of ȧró, but not, for example, in Il. 3. 359: diáμnσe Xiтāva éyxos 'the spear pierced the tunic.' I find no reason to distinguish two different verbs, as Bechtel does.