¹ See Ferdinand Sommer and Adam Falkenstein, Die hethitisch-akkadische Bilingue des Ḫattušili I. (Labarna II.) (Munich, 1938). Abbreviated Bil.

² See Friedrich Hrozný, Boghazköi-Studien 3.90–129 (Leipzig, 1919); Edgar H. Sturtevant and George Bechtel, A Hittite Chrestomathy 175–200 (Philadelphia, 1935). Abbreviated Tel.

³ See Frédéric Hrozný, Code hittite provenant de l'Asie Mineure (Paris, 1922); Chr. 201–28 (op.cit.fn.2). Abbreviated CH.

⁴ Probably to be dated at about the same time as Tel. See Albrecht Götze, Hethiter, Churriter und Assyrer 54 (Oslo, 1936). The other dates here given are based on Albright, BASOR 88.28–36, and on correspondence with Professor Albright.

⁵ Cf. Sturtevant, Chr. 224.

⁶ Cf. the common locution kwitman ... nawi ‘as long as ... not yet’, i.e. ‘before’. On this see Sommer, BoSt. 4.13–6.

⁷ This is supplied by Sommer; but here and elsewhere I copy plausible suppletions without any indication that they are such, unless the presence of a lacuna has some bearing on the point at issue.

⁸ For this interpretation see below, 93 and 94.

⁹ Sommer translates (Bil. 13) : ‘du aber bist mein Sohn Muršili, tu dú es! [So] bewahre [des Vaters Wor]te!—Solange du des Vaters Wort(e?) bewahrst, wirst du [(nur) Brot ess]en und Wasser trinken. Wann die Zeit des reifen Mannes in dich eingezogen ist, so iss den Tag über zwei-, dreimal und pflege dich gut! [Wann aber] auch noch das Greisenalter in dich eingezogen ist, dann t r i n k dich satt! Und [des Vaters] Wort magst du (dann) beiseitesetzen!‘ He recognizes an antithesis (ib. 150) between the first man clause and the other two (though in that case the -ma that he supplies in the third clause, line 31, seems much more needed in the second). According to him, Mursilis is to be on short rations during his minority; later he may shake off the king's yoke and do as he pleases. But I think that the relation of the second and third clauses to the first is epexegetical rather than antithetical, and that the essential contrast is provided by these two explanatory clauses in their relation to each other (quite in consonance with Sommer's supplied -ma), not by them jointly in their relation to the first. Doing the king's bidding I regard not as the badge of Mursilis's minority during which he is to eat and drink with restrictions, but as a duty to be observed throughout Mursilis's life if he is to eat and drink at all (how much he is to eat and drink at successive stages is specified parenthetically, adequate food being prescribed for the prime of life, and plenty of drink for his declining years). Thus Sommer's introduction into his translation of the word (nur) appears to me gratuitous (cf. inf., fn. 14).

The final sentence is certainly difficult. I can hardly believe that the dying old autocrat, who is later (3.56–8) to specify that his ‘words’ are to be read to Mursilis every single month that he may really take them to heart, the performance presumably to be continued throughout his life, would here give Mursilis permission ultimately to disregard these same words (32 [at-ta-as-s]a ut-tar pí-e-es-si-ya), the very thing that he complains of so bitterly on the part of his daughter (3.16 at-ta-as ut-tar pí-e-es-si-i-e-it). But it is to be noted that the passage is badly preserved, and attas is only a suppletion; possibly le belongs somewhere in the lacuna (we might then emend Sommer's reading to pí-e-es-si-ya-si, despite his note on 3.32, Bil. 153; however, le with the imperative is not impossible—see Sommer, Bil. 91, and also Edgar H. Sturtevant, A Comparative Grammar of the Hittite Language 251 fn. 70 [Philadelphia, 1933]).

¹⁰ The use of an imperative clause as the protasis of a future condition is a common phenomenon in various languages. The following examples may be noted: Hit., VBoT 24.1.28–9 pa-ra-a-wa-kán e-hu ^DKAL lu-li-mi-es an-da-wa-kán ^DKAL in-na-ra-u-wa-an-za ú-iz-zi ‘Go forth, ^DKAL the effeminate; ^DKAL the manly will come in’ (for the translation, see Sturtevant, Chr. 109), also KUB 13.4.2.69–71 (quoted below, fn. 14); Lat., Vergil, Ecl. 3.104–5 dic quibus in terris—et eris mihi magnus Apollo—tris pateat Caeli spatium non amplius ulnas; Eng., Follow the first, and thou shalt see The other always following thee.

¹¹ Hattusas seems to be frequently spoken of as ‘up’; cf. KBo. 2.2.1.2, 2.13, 2.5.3.43, 2.11.2.7, Hat.1.76.

¹² So utne (20) in opposition to the city of Hattusas, as perhaps also in 3.35; see Sommer, Bil. 106, on 2.62, and cf. CH §71.

¹³ The two sections addressed to Mursilis (3.26–32, 51–3) are, as has already been said, separated by the interpolated directions (already quoted in part, and here represented by dots), addressed to his subjects (3.33–51).

¹⁴ In every one of the five passages just quoted, the reward for good behavior is eating and drinking. Twice (2.33, 3.21–2) the mere act of eating and drinking is all that is mentioned (cf. Tel. 2.14 nu-wa-za az-zi-ik-kán-du ak-ku-us-kán-du i-da-a-lu-ma-as-ma-as-kán li-e ku-it-ki 'let them eat (and) drink, and do not harm them in any way'); but three times (3.29, 34, 48) the reference is to eating bread and drinking water. In these cases Sommer introduces the word nur into his translation (cf. sup., fn. 9)—according to him bread and water are all that Mursilis is to have during his minority, and all that the less fortunate and more downtrodden subjects are ever to have (see Bil. 150 fn. 2). It is not as a reward for heeding the king's words, but as a method of heeding his words, a sign of heeding his words, that they are to carry out this behest. However, bread and water are not, I think, conceived of here as a meager diet, but rather as the staples without which life cannot be maintained; the alternative to eating bread and drinking water is not to be alive at all, as is very clearly indicated in 3.37 Ú-UL hu-i-is-te-ni nu har-ak-te-ni. We are reminded of the Homeric ὄσσοι νυ̑ν βροτοί εἰσιν ἐπὶ χθονὶ σι̑τον ἔδοντες (Od. 8.222); ‘eating bread’ is a periphrasis for being alive. Bread at least is certainly not to be despised, for it is the food of the gods, though presumably their drink is wine (Bil. 3.50; and Sommer ad loc., Bil. 171–2) or wine and beer (KUB 13.4.1.60–3) rather than water. That bread and water are looked upon as desiderata that it is part of a king's duty to provide is proved by a passage which Professor Goetze has called to my attention, 2 BoTU 9.1.4–8 [na]m-ma-as is-ki nu-us-ma-as-kán NINDA-an ki-is-sa-ri-mi an-[da pa-i] LÚ GIG -an a-ú nu-us-si [NI]NDA-an wa-a-tar pa-i ma-a-an ha-an-da-is wa-la-ah-zi zi-ga-an e-ku-ni-mi da-[i] ták-ku-wa-an e-ku-ni-ma-as wa-la-ah-zi na-an ha-an-da-[i] da-i ‘Furthermore anoint them and [give] bread for them in[to] my hand. Look upon the sick man, and give him bread (and) water; if heat strikes (him) place him in coolness, (and) if cold strikes him place him in warmth.’ Cf., in the very similar passage ib. 4.2–9, 7 nu ki-is-du-wa-an-ti NINDA-an pa-i........... 'Give bread to the hungry [and water to the thirsty]'. The tone is quite that of 'Give us this day our daily bread'. Finally, evidence that the eating of bread and the drinking of water may be looked upon as the reward for duty done is provided by KUB 13.4.2.69–71 (= 2.76–8 in Sturtevant, Chr. 156) na-as-ta DINGIR.MEŠ-as-be ZI-ni i-ya-at-tin nu NINDA-an e-iz-za-at-te-ni wa-a-tar-ma e-ku-ut-te-ni É-ir-ra-za i-ya-at-te-ni ‘Then act in accordance with the will of the gods alone, and you will eat bread and drink water and establish a household for yourselves.‘

¹⁵ It should be added to KBo. 5.6.3.12–3, which Friedrich cites (Kleinasiatische Forschungen 1.292) as the one certain example of the type known to him.

¹⁶ This grouping of the present real and the unreal protasis on the one hand, as against the more or the less vivid future on the other, is quite in accordance with Greek modal usage, which employs the present or the past indicative in the one case, the subjunctive or the optative in the other.

¹⁷ The conjunction mahhan does not occur in the Bil. at all.

¹⁸ More probably, I think, the missing portion contained some expression to introduce the quotation beginning with kwiswa in 59; to take this with halukan udas in 60 might be possible but would involve unusual word order (not, however, unparalleled, especially if the quotation is a question; cf. Sommer's interpretation of KBo. 4.13.4.33–4, Bil. 166 fn. 2).

¹⁹ The contrast in 1.7–9 between durative action (harninkiskit) and iterative (wezi, paizi) furnishes fresh confirmation for Bechtel's assertion that -sk- verbs are primarily durative only, not iterative as has been thought (see George Bechtel, Hittite Verbs in -sk-29, 61–3 [Ann Arbor, 1936]). In 2.4 we have a different situation: weskanta is not only durative (like harninkiskit), as is shown by the -sk- formation, but iterative (like wezi and paizi) as well, as is shown by the present tense. In 1.7 the notion is simply 'he kept devastating countries (continuously)', just as in 1.11 nu ut-ne-e ma-ni-ya-ah-hi-es-ki-ir it is ‘they kept on governing the countries (continuously)'; in 2.3–4 it is 'the troops repeatedly went on campaigns; and ever time they went on a campaign anywhere, they kept coming back without success’ (they did this continuously on each separate occasion—in other words continually as well as continuously). In 1.6 harta, which is durative but not iterative, is past as we would expect; that it is not an -sk- form as we might expect is due to the natural lack of -sk- forms in ‘static’ verbs (cf. Bechtel, op.cit. 86).

²⁰ Other languages also differentiate the two notions, particular vs. general, at least so far as past time goes, but in quite different ways: Greek by a change in mood (indicative vs. optative); Latin sometimes by a change in mood similar to that of Greek (indicative vs. subjunctive—the so-called ‘iterative subjunctive‘), occasionally also by a change in tense but of quite a different sort from that noted in Hittite (perfect vs. pluperfect, with ubi or postquam); German by a change in conjunction (als vs. wenn).

²¹ From the nature of the case, the present temporal clause is likely to be iterative, in other words general, and therefore to demand wenn; but in the rarer particular clauses, whether punctual or durative, als is used (or in South German dialect wie, so Professor Kurath tells me) with the present, whether true present or historical, precisely as with the past. In such cases English shows a tendency to do just what German does, i.e. to avoid the usual temporal conjunction (when, comparable to wenn) and substitute for it a conjunction usually modal (as, comparable to als or wie). This acquisition of temporal meaning by a modal conjunction is precisely what we have noted in the case of Hit. mahhan.

²² The much less sharp distinction between temporal and conditional relations in the future than in the past has already been referred to, sup. 95.

²³ Cf. sup., fn. 16.

²⁴ Cf. sup., fn. 20.

²⁵ Cf. fn. 21.

²⁶ Again cf. fn. 20.

²⁷ In view of the tendency in Lat., already noted (fn. 20), to use the subjunctive in iterative expressions, it may not be irrelevant to recall the morphological connection in this language between the future and the subjunctive.

²⁸ A new section begins here, but in Tel. this does not necessarily herald a new subject; cf. 1.10.

²⁹ Sturtevant translates ‘when’ (Chr. 191) and quotes Götze to the effect that in early texts man means ‘when’ rather than ‘if’ (ib. 199); but I see no reason for treating this man differently from the others. (Of course the meaning is practically the same either way.)

³⁰ The tablet shows only three horizontal wedges, the middle one shorter than the other two. Hence man is a very probable suppletion, while takku does not seem possible.

³¹ Cf. sup., fn. 29.

³² To be sure, in general our copy of the Bil., as one might expect, seems to exceed the other two documents in the number of its archaic features; for a list see Sommer, Bil. 206–7. On the greater number of archaisms that characterize the Code as compared with the probably contemporary Tel. (cf. sup., fn. 4), see the lists in Sturtevant, Chr. 194 and 224 (the latter already referred to above, fn. 5). Those in the Code may be due to conscious archaizing either on the part of its original compilers (such as we would not expect on the part of the author of the Tel.) or on the part of later copyists, who may in some degree have been intentionally avoiding such modernizations as Sturtevant assumes were introduced into the Tel.

³³ I am counting only positive examples, though many others are reasonably certain. The count is based on Hrozný's material, augmented in the earlier recension by two new instances, in §§160 and 192, provided respectively by KUB 29.31.5 and 29.34.4.17.

³⁴ The duplicate KBo. 6.2 has takku (CH 32, fn. 4).

³⁵ In the interpretation of ^GIŠTUKUL used here and elsewhere in this paper, I am following Sommer, Bil. 120–34. On CH §40, see especially op. cit. 121–2; on §41, see 122 fn. 1; and on tittiyanza in both paragraphs, see 121 fn. 3.

³⁶ This is a comparatively rare construction in the Code. Instances occur in §§23, 25, 28, 127, 166, 168, 185, 186, IV.

³⁷ Apparently this applies even to the variant reading man in §64. KBo. 6.3, which has takku, begins a section with this word; but according to Hrozný (CH 56 fn. 10) in KBo. 6.8, which has man, the section starting with this word may have been combined with the previous one (which opens with takku).

³⁸ Contrast the unreal condition in Bil., 3.7–9, which has takku (v. sup. 95).

³⁹ Again cf. Bil., 3.52–3 (v. sup., 94–5).

⁴⁰ Or 256 if we count Hrozný's material alone for the earlier version (cf. sup., fn. 33), as we probably should for purposes of comparison with the later version, which lacks corresponding supplementation. However, the difference statistically is infinitesimal.

⁴¹ Here, and also in X and XI, Hrozný makes man temporal (‘lorsque‘); but there is no reason for this. Cf. sup., fn. 29, and inf., fn. 45.

⁴² This is particularly interesting inasmuch as it uses man ... man in a pair of alternative conditions—a type which we meet with takku ... takku in Tel. (4.20–1; v. sup. 99) and in the early version of the Code—e.g. §17 ták-ku SAL-as EL-LI sar-hu-wa-an-du-us-su-us ku-is-ki pí-es-si-ya-zi ták-ku ITU 10.KAM 10 ZU KUBABBAR pa-a-i ták-ku ITU 5.KAM 5 ZU KUBABBAR pa-a-i ‘If any one causes a miscarriage on the part of a free woman, if 〈takku〉 (it is) the 10th month he pays 10 half-shekels of silver, if 〈takku〉 (it is) the 5th month he pays 5 half-shekels of silver’; §22 ták-ku ERUM-as hu-u-wa-i na-an EGIR-pa ku-is-ki ú-wa-te-iz-zi ták-ku ma-an-ni-in-ku-an e-ip-zi nu-us-si ^KUŠE.SIR-us pa-a-i ták-ku ki-e-iz ÍD-az 2 ZU KABABBAR pa-a-i ták-ku e-di ÍD-az nu-us-si 3 ZU KUBABBAR pa-a-i ‘If a slave runs away and some one brings him back, if he catches (him) near he (the owner) gives him shoes; if 〈takku〉 on this side of the river he pays 2 half-shekels of silver, if 〈takku〉 on the other side of the river he pays him 3 half-shekels of silver’.

⁴³ See fn. 41.

⁴⁴ Again in XI as in X we have two alternative conditions; but this time the writer used neither takku ... takku as in the older version of the Code, nor man ... man as in X (on these combinations see fn. 42), but man in the first condition and takku in the second. Certainly here man and takku seem to be used interchangeably. So, too, in 2 BoTU 9.1.5–8 (quoted in fn. 14).

⁴⁵ Again cf. fn. 41. Obviously there can be no point in translating man by ‘lorsque’ and takku by ‘si’, since they are manifestly parallel (cf. fn. 44).

⁴⁶ Hrozný, CH 93, and (not positively) Walther, apud J. M. Powis Smith, The Origin and History of Hebrew Law 254 (Chicago, 1931), take this as a later redaction of 44B. But to me the two seem to have nothing in common other than the word parkunuzi.

⁴⁷ Here again, as in §X (see fn. 42), we have a pair of man's where earlier Hittite used something different. This time the contrast (which is not necessarily marked, for the two members are not always mutually exclusive—i.e. ‘or’ may correspond either to Lat. aut or to Lat. vel) is not between alternative conditional sentences but simply between alternative protases (cf. Lat. sive ... sive). We found takku ... takku employed in this way in the Bil. (2.54; v. sup. 95). The early version of the Code in such instances regularly uses takku ... nasma, as in §5 ták-ku I-N A KUR ^URULu-ú-i-ya na-as-ma I-NA KUR ^URUPa-la-a ‘if 〈takku〉 in Luwiya or 〈nasma〉 in Pala’ (for the entire section v. sup. 100), §6 ták-ku LÚ.GÀL. LU-as LÚ-as na-as-ma SAL-za ... a-ki ‘If 〈takku〉 a person, male or 〈nasma〉 female, dies’; but it may also use takku ... takku, as in §66 tâk-ku GUD.APIN.LAL ták-ku ANŠU.KUR.RA tu-u-ri-ya-u-wa-as ták-ku GUD.ÁB ták-ku EMÈ.AL.LAL ha-a-li-ya-as har-ap-ta ‘If 〈takku〉 a plow-ox or 〈takku〉 a harness-horse or 〈takku〉 a cow or 〈takku〉 a female pack-ass gets into (some one's) corrals’, or (a)ku ... (a)ku, as in §§1, 2 LÚ-na-ku SAL-na-ku ‘either 〈-ku〉 male or 〈-ku〉 female’ (cf. §19), §98 LÚ.GÀL.LU-ku GUD-ku UDU-ku ‘whether 〈-ku〉 man or 〈-ku〉 ox or 〈-ku〉 sheep’.

⁴⁸ These two sections do not really correspond precisely, despite Hrozný (CH 95), since the earlier one involves luzzi (service to the state) and the later one sahhan (service rendered specifically as ground-rent to an overlord, who may or may not be the state). Much closer to §47B is §XXXIXB, which is by no means a ‘double’ of §XXXVII as Hrozný thinks it is (CH 94 fn. 3).

⁴⁹ V. inf. fn. 53.

⁵⁰ In the fragmentary passages we of course cannot judge what is old and what is new; but this statement must at least apply to §XXV if this section is new in its entirety. The man in §XXX apparently reproduces the man already present in §40 (v. sup. 100), while that in §XXXII perhaps replaces a nu ‘and’ continuing an earlier takku in §38; but any inferences based as these are on the position of the conjunctions in the course of the section, and on a single character, or rather part of a character, following the conjunction in each case, are hazardous in the extreme.

⁵¹ The only example about which this cannot be positively asserted is §XXXIV, the beginning of which is lost.

⁵² The most obvious exception aside from §XXXIV, noted in fn. 51 as lacking a takku clause altogether, is §XXV, which, as has already been stated, is perhaps new throughout. That even a new section should begin with takku is not surprising, as this is the stereotyped opening of a section recurring throughout the Code, and even perpetuated in the names of the two main divisions of the earlier version, ták-ku LÚ-as and ták-ku ^GIŠGEŠTIN-as (on these see CH 2 fn. 1 and 100 fn. 1 respectively). Less easily accounted for is §XI; here the new material begins as usual with man, but this is balanced by a subsequent takku (cf. sup. fn. 44).

⁵³ This may be an argument in favor of Hrozný's suppletion of man rather than takku in §111 (v. sup. 105). In this connection, too, it is interesting to compare §§XXXVII and XXXIXB in their relations to §47B, on which Hrozný thinks both are based (cf. sup. fn. 48). §XXXIXB, which is very close to §47B, has only takku; but §XXXVII, which not only involves sahhan instead of luzzi but also embodies considerable new material, has man three times over.

⁵⁴ In §II the prescription of a special penalty ták-ku SAL-za-ma GÌM ‘if (it is) a female slave’, involves a modification rather than an addition, since in §4 the same penalty is prescribed for mortal injury to either a male or a female slave, ták-ku ERUM-an na-as-ma GÌM-an ku-is-ki wa-al-ah-zi na-as a-ki ‘if any one strikes a male or female slave and he-or-she dies’. The same is true of the prescription in §IX of a special penalty ták-ku ERUM-is-ma ‘if, however, (it is) a slave’, different from the one meted out ták-ku LÚ EL-LUM SAG.DU-ZU ku-is-ki ha-pal-la-sa-iz-zi ‘if any one breaks a freeman's head’; the corresponding provision in the earlier recension, §10 (LÚ.GÀL-LU-an), does not differentiate between freeman and slave, but plenty of others do, so that ták-ku ERUM is a very familiar opening (e.g. §§4, 8, 12, 14, 16) particularly after a parallel ták-ku LÚ EL-LUM or something of the sort (as in §§3, 7, 11, 13, 15). The addition of ták-ku ki-es-si-ra-as wa-as-ta-i or the like with its special penalty in §§III (already cited above, 105–6, in a different connection), V, and VI as compared with §§5, 7, and 8 respectively is hardly to be viewed as involving new material, since this formula, though without takku to be sure, had already occurred in §§3 and 4.

⁵⁵ §IV ták-ku LÚ-as da-me-e-da-ni A.ŠA(G) A.QAR an-da a-ki ták-ku LÚ EL-LAM A.ŠA(G) A.QAR É 1 MA-NA 20 ZU KUBABBAR-ya pa-a-i ták-ku SAL-za-ma 3 MA-NA KUBABBAR pa-a-i ták-ku Ú-UL-ma A.ŠA(G) A.QAR dam-me-el pí-e-da-an du-wa-an 3 DANNA du-wa-an-na 3 DANNA nu-kán ku-is ku-is URU-as an-da SIXDI-ri nu a-pu-u-us-be da-a-i ták-ku URU-as NU.GÁL na-as-kán sa-me-en-zi 'If 〈takku〉 a person dies in some one else's land, if 〈takku〉 (it is) a free man, he gives the land, a house, 1 pound and 20 half-shekels of silver; but if 〈takku〉 (it is) a woman he gives 3 pounds of silver. But if 〈takku〉 (it is) not some one else's land, whatever village is designated within an area 3 miles in one direction and 3 miles in the other direction, (from this) he gets the above-mentioned amounts; if there is no village, he suffers the loss.' Cf. §6 (which, however, is very different) ták-ku LÚ.GÀL.LU-as LÚ-as na-as-ma SAL-za ta-ki-ya URU-ri a-ki ku-e-la-as ar-hi a-ki 1½ ^GIpí-es-sar A.ŠA(G) kar-as-si-i-e-iz-zi na-an-za da-a-i 'If a person, man or woman, dies in a different village, (the man) on whose farm he dies marks off a field 1½ pessars (in dimensions) and buries him.' To this example we must add §XI, already noted in fn. 52.

⁵⁶ Cf. sup. 100.