An infin. certainly looks possible, and a first person is inadmissible in Y. 49, 10, but according to the valuable indication of the Pahl. trlr. nipdonhé(é) is here a 1st sg. conj. aor. in the sense of the future, and is now fully recognized as such, Spiegel leading the way in rendering as a first person in analogy with the other 1s persons in this chapter.
The Pahl. trir. recoils, as usual, from yé@ as equalling azém yé, but my rendering of him is perhaps too unfavourable.
‘When I shall defend Righteousness - -’ may well be free for ‘I who shall defend - - - . Pavan nikirishné = ‘according to observation’ means practically ‘as one observes’, ‘as the matter goes’ (so Spiegel elsewhere) ; so dis means ‘according to these things’, ‘thus’. Ménavadikih has the position of a gen.; see mainyeus (or manyéus).
When regarded as ordinary Pahl., I would render c; ‘A spiritual thing is the Gathic doctrine which is declared from this which is Thine by mouth’.
The general accuracy is to be noted, the glosses being intelligent, or harmless, and, as in every case, possibly from a later hand. This is espe- cially true of the gloss gdsdnikih in line c, as it probably did not exist in the MS. used by Ner.; and this circumstance is of importance to us in our endeavours to estimate the age of surviving codices. The liturgical notes containing directions regarding the repetitions of verses etc. are not pro- perly Gathic.
The Persian reports vdyinand = vadinand, for vddinyén, or vddinand, instead of vddiinam-é, and yemalelin (jamniimi) for dmizdi. Mint occupies the position of a gen., like ménavadikih, fortified by the reading z#sh, but this does not render it at all certain that the trlr. understood mindi as
a gen.
Y. XXIX.
See for summary, and for further treatment, S. B. E. XXXI. pp. 3—13. Changes in opinion which have taken place since its composition are however not always noted here.
1. I prefer to read tavisch@, as it is quite possible that the nearly universal substitution of e for a may not be original. The exceptional occur- rences of a should be preserved by use, or notification; see K.5 J2
Gerezhda@ must have the sense of ‘plaint’ here; cp. raosté in vy. 9. Consider also the reading gerezhdiim in Y. 51, 17. which may refer back to this place. As to the reading dhishdya (J.2 etc.), the Pahl. points to ahush- aya, while Westergaard and Spiegel reproduced dhushuyd (= dhushuvd (?)).
In the Persian MS. K.* the long 7 and the long @ are indistinguish- able. Perhaps short ¢ and short w are indistinguishable sometimes in K5
As 39 in MSS. has often notoriously taken the place of »», cp. viduyé(é), tanuyé(@) etc., so in carelessly written MSS. nothing was easier than the
apparent miswriting of » for 9.
Notwithstanding then the absence of the syllables ahw in the MSS. well reported by Geldner, I would personally prefer a reading ahushayd@, as the short 7 of Pa may represent a short w (see above), and one MS. has short a.
Ahushayd, as a substantive, might be explained as chu (cp. ahu else- where in composition), and shaya from khshi = ‘to destroy’. If khshi = ‘to dwell’ falls into shayati (not shaé(é)t#; there is properly no such word), and if shitayé likewise appears from it, we have no difficulty whatever in recognizing a shaya from khshi (cp. Ind. kshaya). The meaning ‘with life destruction’, thus indicated by the Pahl. trlr. (who, as has been said, like Westg. and Sp., read the MS. before him with dhu-), is especially germane to the connection. Cp. also md hé(é) mairyé géurvaydid(t) - - - ahumerekh§ - - - Yt. 8, 59. Cp. also the composite reading ahiimusté, ahtim in one MS. and usté in anotber (see my notes Y. 46, 4.), and the Pahl. ahtikinéd. Ahush- aya might be an adverbial instr. effectively thrown in to relieve the throng of nominatives which are subject, as so often, to a hent? understood.
But the apparent readings of the surviving MSS. should in no ease be hastily abandoned, as was the constant practice among writers, nor should they be abandoned without alternative rendering even for readings plainly indicated by the Pahl. trlr. As an explanation of ahishdyd, so reading with several MSS., we have more lately the suggestion of a 374 sg. perfect from hi = si (perf. sishiya) = ‘to oppress’, (@ + hishdyd) the original Aryan s reappearing in the sh, as it does in forms of had and hach = sad and sach,
So far there would be little difficulty; but what are we to do with
- a alll
a 3*4 sing. verb in agreement with so many subjects? Also the repetition of the preposition @ looks suspicious; ‘@ md aé(é)shemd--+ d-hishdyd ---.
The probability of this iteration has been most strenuously, and pro- perly, objected to in the other cases. But these difficulties, which are how- ever great, might be overcome.
It is certainly very interesting to us to see that hishdyd(a) corresponds to seshdya, and that, with a little forcing, a 3™4s¢. perfect makes sense, but we should not be led away by such a circumstance to believe that the word ahishdya is actually a 3'se, perfect here; and the same remark applies (with emphasis) in many other similar cases. Our first duty in translating is to search for the simplest and most probable rendering of the text as it stands, or as slightly emended. An expressed verbal form is not neces- sary here; nothing is more idiomatic to the Iranian languages than the presence of one or more nominatives connected with an object by a dat., or gen., or prep., and agreeing with a form of ah = ‘to be’, understood. Cor- recting (?) to ahishayd (see the two short a’s in different MSS.), I would prefer a nominal form in the adverbial instr.
Perhaps Vedic asé might be considered; * astkshaya = ‘destruction with the sword’ being compared. It would harmonise well, and ahisdyd P(¢) immediately suggests forms of ah = as = to hurl, although the sense of the instr. in the first member of a compositum would not be necessarily, or indeed so naturally expressed by an inst. form, according to Sansk. ana- logy at least, as it would be by the bare stem. As the ‘life-destroyer’ is sug- gested by the Pahl.’s ahikinéd, so the Pahl. trlr. also suggests the correction tdyusché for the very awkward tavischd. (Ahikined also = ‘disfigures’.)
When the Gatha stood in the Pahlavi character, a very slight mark only determined the difference between tavi§ and tayus (sic). Spiegel keenly suggests that w and y have become inverted. How does tavischd acquire the here needed ‘evil’ sense, it being a substantive, and generally used in a good sense? The Pahl. word taraftér rests upon taraft = ‘hidden’, and Ner. (who is a high, although of course not an infallible, authority on the Pahl.) renders stenagchd. Véastd as equalling ‘nurturer’ was first suggested by the old Gujarati trlr. after the Pahl. which renders ‘provider of nourish- ment’. As to the omission of remé to accommodate the metre, it has been said that it was originally gloss to hazaschd, but it is no longer possible to decide with any seriousness as to which word is gloss.
And remé existed in the ancient MSS. used by the Pahl. trlr. Omitting hazaschd, we secure seven syllables, and dereschd@ counting two, this would leave nine in the last division.
If we read @ maCuaé(é)sh(e)mé as containing three syllables (which is
entirely admissible in a case of necessity), and drop the chd of hazaschd (which is surely better than dropping an entire word) the line falls into order, but absolute and uniform exactness in the settled number of syllables could never have been original (cp. S. B. E. XXXI. p. xliii). ‘Declare Ye to me’ is as much as to say: ‘Cause ye me to experience (good care for my pastures)’, including the skill of general cattle culture, to console the Cow for the privations of capture. She bewails the woes of invaded territory.
The Pahl., further considered, is full of the richest suggestion, having been the source of all later renderings, and being itself remarkably correct, although, like so many modern renderings, never absolutely so. That the Pahl. should appear to render remé differently from remem (?) in Y. 48, 7 is not very important, and there seems much cause to believe that réskhiin is a false reading for aréskhd(°un). The difference would be very slight in the Pahl. character, and Ner. appears to have read aréshké; see his irshydluh. A more serious question is as to rd@mem in Y. 49,4. Aréshké = arashké(°un).
How comes it that the trlr. saw the same word in both places? Have we here an indication for the correction of a text? Was the trlr. there conjecturing an improvement, and holding that rémem should be read remem?
Ner.’s avinirmito does not possess the negative of interrogative em- phasis. He probably read /é@ for li, but could hardly ‘have mistaken m@ for the prohibitive. Necessarily regarded as free, his rendering of the Pahl. is valuable. He seems to connect the idea of ‘protection’ rather than of ‘nurture’ with vdstariddr (sic). The Pahl. text of the Persian MS. has vdstiriddr (sic) rendering parwarish-kunandah. I follow, as vdsnidar (so DJ.) = ‘well- wisher’ mistranslates vdstd. Vdstaridér may be coined from a denom.
2. I have sometimes allowed the plural word ‘kine’ a correlative in the singular, ‘she’, or ‘her’, for the purpose of avoiding the use of the word ‘cow’ in a semi-poetical rendering.
Here however I write ‘them’ for ‘her’, notwithstanding its awkwardness. As to whether the Tashan géu$ was other than Ahura (so De Harlez, Bartho- lomae and others), we may say that the question is like that as to the Spenista Mainyu, which, like the ‘Holy Spirit’, is now spoken of as a quality of the Deity, now as His co-equal agent.
The Gnostic Demiurge may find its origin from a misconception of this place. The name Géush tashan[4] occurs in the later Zoroastrianism, but it is there still more seriously misapprehended. The murderer, or cutter (sic), of the Ox is Afigra Mainyu (see Haug’s Essays edited by West, 3°4 Edition, page 147). There is no question at all but that the Tashan géus at least represents Ahura here, and I think that, when the eagerness for distinctions
has subsided into the calm of judgment, the Tashan géus (in the Gathas) will be again regarded (with Roth) as practically Ahura Himself; see remarks on Y. 31, 9, and Y. 51, 7 y@ gam tashd--- Maadd. Haug, fol- lowed by others, renders ra¢us as an abstract. I think that the Pahl. trlr. is correct with his xadé. The word is here beyond any doubt. a concrete. See ahi «++ ratus in verse 6, which distinctly refers back to it, just as it refers back to the vdstd of line c, verse 1.
Gaoddyd (cp. gédhdyasam; go + dha = ‘to nurture’) and thwakhshé (cp. tvaksh, tvdkshas) are nom. sg. in poetical apposition with ratus.
For the possibility of nom. masc. forms of the a declension, cp. gao- dayé(e) hé(e) (= gaodaychya) and thwakshdi, both of the a declension.
The nominatives appear to me to sound well just in the places which the words occupy. Spiegel took the words as a compositum, ‘sich der Vieh- zucht befleissen’. Haug, bovwm-nutritum formando. Justi keenly saw a loc. in gaoddyd (u stem; comp. gaoddyum), and an acc. neut. in thwakhshd (cp. thwakhshanhd). The Pabl. trlr. gave the original clue as usual. Anhad(t) is understood with ratus -- + gaoddyd thwakhshd. Kem - + ahurem carries out the idea. Although on the whole preferring usta aS a second pl. pret. of vas, I by no means precipitately abandon the more idiomatic usta-ahurem = pavan nadikih khiddt as governed by daté ‘whom did ye appoint, etc. I am inclined to prefer my older rendering ‘dominantes’ to my later one ‘having power’. One scholar seems to prefer taking ddtd as a nom, pl. neut. of the part., or a 3" pl. aor. mid. in sense of pass., with khshayanté as gen. sing.: ‘Is thy regulation (?) for the cow thus that pas- tures should be given her on the part of the possessor (?), and with them also the care proper for the Kine?’
This view seems to take gaoddyé thwakhsho as nom., a great advantage. Déta is a 24 pl. act. aor. = ddhdta, or ddata.
The Pahl. without glosses would come out still nearer to its original, as mén can mean ‘by whom’, or ‘by what’, and it should be so rendered, as it translates kathéd, the meaning of which the trlr. could not mistake. So min again in (6) should be read in the sense of hyad(t). Dddo may well mean ‘ye set, gave, or appointed’. Padakhshahth (sic (?)) (so DJ.) does not bring us any nearer khshayantd. Is it pddakhshah?