STOP
Early Journal Content on JSTOR, Free to Anyone in the World
This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in
the world by JSTOR.
Known as the Early Journal Content, this set of works include research articles, news, letters, and other
writings published in more than 200 of the oldest leading academic journals. The works date from the
mid-seventeenth to the early twentieth centuries.
We encourage people to read and share the Early Journal Content openly and to tell others that this
resource exists. People may post this content online or redistribute in any way for non-commercial
purposes.
Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early-
journal-content .
JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people
discover, use, and build upon a wide range of content through a powerful research and teaching
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please
contact support@jstor.org.
INDO-IRANICA
Edwin "W. Fay
University op Texas
1. Avestan aesasa-, petens.
The Avestan root aes means to seek ; to (seek to) hear (Bar-
tholomae, Air. Wbch., p. 29, 4) ; to attack, waylay, seize (ib. e) ;
to obtain, acquire (ib. 6). The long word aesasa- is from a pri-
mate aisosko-(Av. s from sk), and the selfsame primate lies
behind the Latin denominative verb aeruscat, begs (as a mendi-
cant) . Note s from sk also in the compound vanOwyaesa, army-
thief, waylayer. For further definitions of the root ais or sis,
see CQ 9. 110.
2. Excursus on Ittcuottos, seized, caught (taken in the act).
For Herodotean «raioros (wrongly accented in the books, in
response to a wrong derivation, eTroioros) a typical example is
eirauTTos eyeVero irpoStSovs = he was taken betraying, i. e. caught in
the act of betraying. In Apollonius Rhodius Arg. 4. 366 we
must read «r' aurrov («ri as in «r' Icra, equally), ex improviso.
3. Sanskrit pada-vi (foot-) way.
With Perrson (Beitraege, p. 512) I identify -vi in this com-
pound with Lat. via. In the earlier masculine padavis, guide,
the posterius meant goer, while pada- seems almost preposi-
tional = with, cf . TreSa in the Aeolic poets, and see on Skr. pad-
-rathas, footman (with the chariot) in CQ 8. 52, n. In vi, i
is a weak grade of the ei of the root. Lat. via (and this remark
is applicable to many Greek and Latin feminines in ia) is a syn-
cretic form, combining the feminine ending in i with the femi-
nine in a; in this ease the root noun m with a feminine suffixal
a attached to the weakest form of the root, i. e. w-a. Perrson is
in error in writing the root as wei (but see § 10).
4. Indo-Iranian d-vis, obvious.
This is a compound of a (i. e. the proethnic preverb e: 6 for
which English here or there is too heavy a rendering; German
dar suits better) plus the adverb vis, i. e. vi extended by the s
122 Edwin W. Fay
which seems to be joined quite ad libitum with prepositional
adverbs. The Avesta preserves vis and we have it in the com-
pound vis-patha, quasi deviously, variously. As will appear
later vi comes right close in meaning to the German adverb weg.
5. Indo-Iranian vi, vi, asunder, apart; weg (cf. Ital. via).
I explain the adverb vi as a locative to a root noun we (i) , with
the verbal sense of to wind, whence to wend, wander. For this
wbi see Walde's Lexicon s. v. vieo (from a secondary root wy-e).
As Bng. wends, wanders derive from the root of to wind, we may
admit a like development of sense in the root we(i) . Note that
in English went, a past tense of to wend, serves as preterit to the
verb to go, and has lost all trace of connection with to wind.
6. Excursus on (Sanskrit) doublet roots in -an/-d(y).
In JA08 44. 341 I made, in part after Macdonnell, a list of
these roots, 1 viz. kha(y) :khan, ja(y) :jan, sd:san, ta(y) :tan.
To these may be added the Indo-European pair wa: wen, to
wound (see Pick, l 4 p. 542 and p. 547, Boisacq, s. v. fmo-o-w).
1 In that list I concluded dr&:dram and ga:gam. I now note that the
Sanskrit trio drS dram dru, to run, justifies the trio ga gam, gu, to go. I am
exploiting no theory of origins. I am quite willing to believe that the -am
and -3 roots had an entirely unrelated origin, though later they came, must
have come, together in speech consciousness in response to a classification as
inevitable as it was unwilled. To state this extremely, it is altogether
possible that in their prototypes palm, goes (root gwem), and l-/3ij, went
(root GwX), fell into a systematic association only as Latin fert and tulit
or as Eng. goes and went so fall. But after they once fell into this associa-
tion they served as a source for analogies, and the analogy groups then
formed, without the consciousness, or at least without the conscious will,
of the speakers, a morphological system. Accordingly, when we find in
Sanskrit a posterius gu, going, we may set it down at first as due to the
analogy of Skr. dru, running: or we may place it at once, per saltum, in
a morphological system with ga gam; ef. also yu-, faring: ya, to go.
There is neither rhyme nor reason in refusing ga gam gu if you admit dra
dram dru, always, of course, upon evidence. Thus we escape the awkward-
ness of having to deal with Skr. -gva-, in ndva-gva-, as cow, instead of as
going or gang, and we are left free to define wp4o-pvs by fore-going and not
by fore-bull (Bloomfield, AJP 17.424, 29.80; see the literature in Boisacq).
The nominative irpe<r-pevs will have originated after the vocative in ev
(Sanskrit o). Thus the vocative was a common term in Greek in the u and
in the ev stems. We owe /3u instead of the correct yv to Homeric vpia-pa.
Indo-Iranica 123
Here I add we(i), to wind (go) : wen-d, 2 to wind, go. "We may
here note the special sense of to wither in Lat. viescit, correla-
tive to Slavic ven-d to wither (see Miklosich, p. 380) ; cf. Eng.
gone off = deteriorated, etc.
7. Further on Indo-Iranian avis, obvious.
The Slavic sept of O.Bulg. ave, manifeste (see Berneker
Slav. Etym. Woch. p. 34), reveals that the combination in d-vis
was Indo-European. Slavic -ve differs from Av. -vi(s) as Lat.
prae differs from pri. In Greek, as I have pointed out before
(see AJP 33. 391) , we have a double of Skr. avis in the compound
av-murri, not on the road standing, not obvious, unexpected. Here
belongs Skr. dvistya- (ty from thy, see AJP 34. 15, n.), obvious,
visible. In the Avesta avil-2/a=coming on the road, whence obvi-
ous, visible. The Indo-European trio wai wi wo (cf. Lat. prae pri
pro) exhibits its last member in Gathic Avestan vu-daya, to put
away, push away, thrust away, cf . w-Otm. 3 Where Indo-Iranian vi
connotes asunder, entzwei, there has been some influence from
Indo-European dwis, in-two, apart. To put it otherwise, the
word dwis in certain combinations lost its d- by dissimilation.
The root wi-dh of Skr. vidhydti and Lat. di-vido, e. g., will have
come by dissimilation from original dwi-dh-. In passing I would
explain Skr. vyadh ( :vidh) as containing in vya- a correlate of
8ui, through. Given the doublet dwi(s) /wi(s) , we may also
1 The unextended root wen is preserved in Germ, wohnen, to dwell, i. e. to
wander in a nomadic preserve; cf . Eng. dwells, from O.Eng. dwelian, to go
astray, err, tarry, dwell. Skr. vdnam, forest, wood (wood before trees,
trees was an interpretation of wood) applied at first to the ranges in which
the nomads dwelt, or over which their cattle wandered.
•Despite the convenience of recognizing proethnic we, weg, in Latin
etymology, the words in which we have this ve seem to be best explained
Otherwise. It is not open to question, in my opinion, that Lat. vehe-mens is
a compound with imperative prius vehe-, cf. Avestan vazo-vqiiewya,-, (carry-
ing away i. e.) robbing the army-stuff. Thus vehementem (ace.) = carrying
away the mind (first of anger etc., for the usage in Plautus see AJP 24.71).
The contracted form ve-mens, supported by the influence of demens and
amens, became the pattern for ve-cors, ve-sanus etc., and the irradiation
even went so far that we have ve-grandis as a negative of grandis. Lat.
\s\vesoitur I cannot bring myself to separate from Skr. agni-sv&tta-,
ignicomesus (see TAPA 44.110). In ve(r)-labrum, water-basin (see AJP
35.153) the prius = Skr. vdr.
124 Edwin W. Fay
expect to find other proethnic forms, or their continuants, with
w-, e. g. vi- in Lat. viginti.
8. Excursus on aUr-Odverai, perceives ; Lat. audit, hears.
In the whole range of 'orthodox' Indo-European etymology
there is nothing more pretentious than the equation of aur- with
Skr. a-vis. For the treatment of afe as a dissyllable there is no
particle of evidence. Of Ittouttos I have already disposed (§2),
and dt'o), I hear, is a plain denominative from a stem atjsi-, ear,
in Lat. auris. The correct derivation of ahrOaverat is from the
root ais, to take (see § 1), as I have before pointed out in CQ
9. 110. Eng. takes (I take it), apprehends, assumes, and Lat.
capio, accipio, percipio, all show how the sense to perceive origi-
nates from to take. See also § 1 on Av. aes, with the sense of to
(seek to) hear. If the current derivation of auT-Odverai is a
caprice, the derivation of Lat. audio from awisdio is a phantasm.
With aus-cultat (ear-lends or leans) before us, anything but
ausdit is unthinkable. Of course the elaborately fanciful pri-
mate awisdio has been invented to turn a special phonetic trick
for oboedio, but it involves far less of unsupported assumption
to conclude that here posttonic au on its way to u or, in vulgar
circles, on its way to o, was subject to reenforced rounding from
ob modified by anticipatory palatalization from di, — causes result-
ing in something other than *obudio. But the analysis o-boedit,
which means cognation with TriiroiOa (iraOea-dai) , is always pos-
sible, cf . O.Lat. con-foedusti, and note that foedus, ugly, has held
on to oe. Festus also gives us amecus (i. e. amoecus) for amicus,
and we have oe in the second syllable of amoenus, lovely.
9. Semantic excursus ; the meaning before the last.
In the classical tongues there is a wide range of turns such as
to walk with legs, to see with eyes, to talk with the mouth (ore
loqui) . These are relics of the time when to walk and to see and
to speak were not the original senses of their verbs, and when
ore loqui e. g. meant something like to crack (Scottice usur-
patum) with the mouth ; when to see may have meant some such
thing as to scan. The gradual ellipsis of the names of the
organs participant, whereby the connotation was raised to the
rank of definition, may be aptly illustrated by the comparison
of Plautine oculis rationem capio with Terentian rationem capio
Indo-Iranica 125
(see the great Thesaurus, iii. 321. 12) ; cf. also in Lucretius,
carmina auribus accipere (4. 982) with voces accipio (4. 611).
With oculis omitted capio was on the way to becoming a verb of
perception.
10. Sanskrit (vayyd) vayi-a-, attendant: d-iVas, wooer.
This Sanskrit word, not treated by Uhlenbeck, is from a loca-
tive vay-i, extended by suffixal o. Here we come back (see § 5) to
the root we(i) (e certain in Lat. venor). I am not disposed to
deny a outrance the grade wei ; and those who refuse the grada-
tion e :e will perhaps admit that wai, by assimilation to wei, was
liable to appear as wei. This is what we do accept in Greek for
0eros. Or the grade wei may have come by way of assimilation
to the synonym root ei. Or [s]w-ei may be a compound root (on
sw- see TAP A 44. 108 sq.). The additional sense of after (for,
towards) in Skr. veti, goes after (pursues, hunts, follows), and
its cognates, will have come from the accusative regimen. So
in the Rig Veda the participle of eti (goes) means, with the
accusative, seeking (begging, etc., cf. uteres, suppliant: bcveirai,
comes to). By acknowledging interplay of the roots wei and
ei we may account for the al (from ai) of the denominative afrel,
demands.
11. Joining an issue ; Avestan vi-naoiti.
Av. vi-naoiti (only with ava and fra) means necat (Eng. slays,
Germ, schlagt). We might derive from the root wa (§ 6) or,
as we must then write it, wa(i), to wound, injure (nocere).
This root will hardly be different from Lat. vae; cf. Goth.
wai-dedja, malefactor (homo nocens). I take the Latin outcry
vae to be (a continuant of) the 'root,' not a derivative from it.
On the other hand, and this seems to me far more likely, vl- may
be the preverb (=weg) and nao the verbal element, cognate with
nu-d in Skr. nuddti, thrusts (see on this 'root' Walde, s. v.
nuo) . In its meanings nuddti combined with vi comes quite close
to vi-naoiti, viz. to wound; to strike (Germ, schlagen) the lute.
Given Skr. nuddti, then Av. vinaoiti, slays : Goth, naus, slayer :
O.Bulg. nawi, mortuus (cf. Goth, b-nauan, confricare) leave no
room to challenge a root nu with the general sense of the root tu
(cf. Walde, s. w. tundo, stuprum).
