481
On the ambiguity (Vieldeutigkeit) of certain characters
in the Zendalphabet.
By L. H. Mills.
I was much gratified to see that Justi'), and also my venerat¬
ed, but since then deceased, colleague Professor Darmesteter ac¬
cepted my suggestion as to the fact that certain Zend characters,
in their present form, express more than one sound ; tbat is to say,
that they bave more than one equivalent when transliterated. In
S. B. E. XXXI, p. XXXIV, I remarked (1887) that A) (gj^) expresses
yd as well as e (e). We all remember that certain authors, with
whom Zendphilology was only a secondary matter, expressed them¬
selves as discouraged by the apparently irreconcilable irregularities
of Zend grammar. Even in the a declension these irregularities
were apparent at once. Take for example kaine (Zend(?)) as com¬
pared with Sanskrit kanyd; ,it seemed to nullify all analogy. Then
such accusatives as gaenC etc. were equally jarring when compared
with Sanskrit gäyam. Even such datives as ahurdi, haorndi were
more than suspicious, while Zarathustrahe (sic) as a genetive was
as irreconcilable with the Gäthic °ahyd as with the Sanskrit °asya.
Such dative duals as gaosaeiwe[V\ were also quite as startling in
view of Sansk. °bhydm, and the infinitives keretee, berelee, etc. were and are even worse than the verbal forms didhaem, daidit and vaochdi
(this latter as a seco'nd sing. pres. indic). But if we look at the
matter a little more critically, these difficulties disappear, for it
becomes clear enough that no such words as those which I have
cited exist in the Zend vocabulary. There is in fact no such nom.
sing. fern, of the a declension as kaine ; such a supposed form has
merely resulted from an imperfect deciphering of the Zendcharacter
«) (JO^) in mi^M^.
1; Cf. Gött. Gel. Anzeigen Mai 15, 1893, also Darmesteter Le Zend¬
avesta, etc. Vol. Ill, p. XCIII ilS93) in comparison with my S.B.E. XXXI.
p. XXXIV and p. 203 (1887,.
32*
4S2 'ViS*, On the ambiguitp of certain charactere in Out Zendalphabet.
As is generally conceded, our present ample Zendalphabet was
a development from the exceedingly imperlect and obscure Pahlavi,
if not from such Pahlavi as we have printed in our few Pahlavi
books, then certainly from prior forms which are only with diffi¬
culty to be reproduced. But in the Pahlavi **Crepresents yd as
well as «/<, yahh, tha, th, etc. Now tbis ambiguity remains as an
exceptional phenomenon in the otherwise generally clear Zend writ¬
ing. In fOl^Mj and similar words >0 is simply **C accidentally
lengthened and rejtresents, as in its original Pahlavi use, yd, and
by no means «, and we have before us simply kainyd completely
in order as a nom. sing. fern, of the a declension.
As is however to be expected, this occasional use of A) (y^)
in its original, but now exceptional, meaning in the Zendalphabet
is not uniform. Its use as expressing two sounds was accompanied
(or followed) by its use as expressing more tban two sounds.
ttf ^ , rti, ^ represent not merely yd, but ya, and sometimes also ye.
To resume; the otherwise falsely transliterated gaim anumaem,
etc. can only be properly written as accusatives of the a declension
by following this clue; €n)MS>^M^ €{0^(J} are, as they stand,
anumayam (or "yem), and yay am (or "yern^ .Sansk. gäyam; rt) {tf)J
= ya, or ye. So our bizarre genetives Zarathunlrahe, ahe, can be
put aside ; the proper forms are of course Zarathmtrahya "yd],
ahya{d) ; rt) = ya, or yd, cp. the Sanskrit "atya, and even the üäthic
"a/iyd^} (AJevo/* as gen. is never Gäthic;. So Srt)M^ Srt)M^ are
certainly not aem, vaem, but ayam (or "em), vayam (or °em) ; cp.
Sansk. aydm, vaydm; rt) is here ya, or ye. rt)i>^^rt)M'^\tMQ} is
gaoitaeiwyd[m) beyond any doubt; cp. Sansk. "iAyi^m; rt) = yd (the
nasalisation remains unexpressed here as often in the Old Persian
Inscriptions). By accepting this deciphering some of the most
puzzling riddles of Zend Grammar are solved. Take the very
interesting (— )70»wyu of Yasna 4G, 11 ; it is undoubtedly a nom.
sing, fem., but it has been hitherto transliterated h' af chd, which
seems as meaningless as kaine for a nom. sing. fern. ; {O^ [rt)) is
here once more the sign for M + i= y + d = yd, and we recover
the lost possessive K'ayd, K'ayd dainü, "his own soul"-*), quite in
order, cp. Sansk. «t!aya(m); the closely adjoining alternative form
K'i {urvd) was doubtless chosen to meet the exigencies of metre.
Kvc-n the difficult it^^C of Y. 31, 9 can be explained as a nom.
sing. fem. ; stands for Jg^ [rti] as is ^V^^" m-j-m<m ; but in this 1; The Pahlavi «»^ is often merely *w + i mechanically written together.
'll It was thii* must obviouH case which induced Darmesteter to follow me, an above cited.
'•'>, Cj). my study of the GithaB, p. 2üü, 557.
Mills, On the ambiguity of certain characters in the Zendalphabet. 483
case the obscurely represented JO^ equals yd, and we have a
possessive thtcayd Ar[a]maitis, [li}^ has tbe inherent vowel a).
joJJ>7>\)eJ is also explicable as a masc. dual, on these principles:
> is here the Pahlavi ( = t) ; is, as often, overwriting i) (super¬
fluous), and Jü^(/ü) is again yd, and we have pourv[y)yd = Sansk.
pürvyd {"yau), as nom. dual. masc.'). Looking for a moment a
little closer at some verbal forms we come upon (?j)i00{O^l^ which
may indeed be saeti{ti); cp. Sansk. ksheti'^), but it may also be
sayati[-ti); cp. Sansk. kshäyati; there is little doubt however that
Grt3'J^&.^ is didhayam (or °yem], cp. Sansk. didhayas, etc., and not
didhaem[?). So rtic^OiS, K>^^PiJ, and Wc(X);^5 are certainly
keretaye, beretaye, and kantaye {A> = ye and c is for a) , and
not the inexplicable keretee [7], etc., and this is proved by
( )rtiMiiM^'^lM»lM where the fuller form is written in the
ordinary spelling; it is °var.üaye{+ cha). So ^^i5;0i is not ieya,
but iydya, cp. Sansk. iydya [A) = yd). As regards >», •>, the former
is merely marked as a lengthened which is Pahlavi for y inter
alia. 9-)i<riMiy is not haitMm as a nom. accus, neutr., the >»
is here lengthened * = y with inherent a, and the word is haiikyam
(Sansk. satyäm); so ^■si'HiMjt is not frasim as a masc, neutr.,
but frasyam, nor is Sfi^iM^ ainim, (nom. accus, neut.) but ainyam
(ep. Sansk. anyäm). So also with the datives ahurdi, haorndi, etc.
the ■> is here (as it stands) Pahlavi •> with inherent a (as always),
and the words should be written ahurdya, haomdya; cp, Sansk.
äsurdya, etc. I must advance even a more pronounced innovation ;
visdi, vinddi, etc, as 2nd sing. pres. were indeed enough to revolt
grammarians, but they are totally unjustified. ^ as Pahlavi re¬
presents Ä-') among other sounds; ev itself may be only a spread-
out »w (or «A» contracted ev); so here «aj is A3), and the «IJ^ have,
as in Sanskrit and Pahlavi, inherent o; the words are visahi, vin-
dahi, etc cp. Sansk. vicäsi, vinddsi, etc. *).
1) Cp. my Study of the Gathas, p. 67, Zend specialists are familiar with redundant letters.
2) But is this kshM itself original?
3) Cp. again the Gött. gelehrt. Anz. as cited above.
4) Very many more applications of this method of reading the Zend- alphabet might be made.
484
Anzeigen.
Hin arabischer Dialekt , gesprochen in ' Omän und Zanzibar,
nach prraktischen Gesichtspunkten für das Seminar fiir
Orientalische Sprachen in Berlin bearbeitet von Dr. Carl
Reinhardt, K. Dragoman. Stuttgart u. Berlin, W. Spe¬
mann. 1894. S. XXV, 428, 8«.
Die Bestimmung seines Buches hat R. schon auf dem Titel¬
blatt deutlich angegeben ; im Vorwort (S. IX) wiederholt er ausdrück¬
lich, dass er für Scbüler, d. h. doch in erster Linie des Orien¬
talischen Seminars geschrieben habe. Um gleich mit Ausstellungen
zu beginnen, die unten nicht häufig wiederkehren werden, so gestehe
ich, dass ich nicht einsehen kann, wie das Buch in dieser Hinsicht
seinen Zweck erfüllen soll.
Was die pädagogische Seite angeht, so wirkt schon die Fülle
des von R. gebotenen Stofi's verwirrend zumal auf junge, philologisch
nur mässig geschulte Leute ; praktisch genommen ist unser Buch
durch die Folgen des am 17. Juni 1890 veröffentlichten Vertrags
mehr als zur Hälfte entwerthet, wenn wir auf Zanzibar und Deutsch-
Ost-Afrika sehen. Die deutschen Beziehungen zu 'Oman sind gleich
null. Nimmt man dazu den unmässigen Preis des Buches, so wird
man zu der Frage gedrängt: wie wenige werden jährlich dies Buch
zu praktischen Zwecken in die Hand nehmen und wie viele werden
es als für diese Zwecke unbrauchbar wieder aus der Hand leseno ?
Lassen wir aber Zanzibar und die Merime (R. 92. 218. 233. 252 u. s. w.)
beiseite und sehen nur auf 'Oman, so fübrt uns ein eingehendes
Studium des Buches zu der Ueberzeugung, dass R. nicht nur der
arabischen Mundartenkunde, sondern der arabischen Philologie in
der höchsten Fassung des Wortes, ja der vergleichenden semitischen
Philologie Materialien von seltener Wichtigkeit zugeführt hat, und
die nachfolgenden Bemerkungen sind grossentheils von dem Wunsche
getragen, dieser Erkenntniss und Ueberzeugung auch in den Kreisen
Geltung zu verschafi'en, wo man nicbt die Müsse hat, die ausführ¬
liche Darstellung einer entlegenen arabischen Mundart eingehend
zu prüfen. 'Oman liegt geographisch so isolirt und seine geschicht¬
lichen Verhältnisse sind so eigenartig, dass es sich wohl der Mühe
verlohnt, ims ein wenig auf dem Boden umzusehen, um die Sprache
des Landes besser zu begreifen.