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
Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early-
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
JOURNAL OF PHILOLOGY
Vol. XXXVIII, i. Whole No. 149.
I.— SOME CRUCES IN VEDIC TEXT, GRAMMAR,
1. ajuryamur for ajur(yam) yamur, and other haplologies.
RV. 5. 6. 10 we have the notorious passage containing the
complex of syllables, ajuryamur, which the Padakara fails to
analyze, to wit :
evan agnim ajuryamur girbhir yajnebhir anusak,
dadhad asme suvtryam uta tyad acvacvyam.
For previous discussions of ajuryamur by Ludwig, Pischel,
Bartholomae, and Oldenberg, see the last-mentioned scholar's
Rig- Veda Noten, First Part, p. 315 ff. My own way is indi-
cated by the heading.
If we regard ajuryamur as haplological contraction for
ajur(yarh) yamur, the first distich is to be rendered: 'Thus
they have gotten hold of imperishable Agni by means of songs
and sacrifices, properly '. Agni, like other gods, is imperish-
able, ajurya, RV. 1. 146. 4; 2. 8. 2; 10. 88. 13; ajara 1. 58. 4;
127. 9 ; 6. 29, etc. For ' holding ', ' keeping hold ' of Agni see
3. 27. 3, agne gakema te vayaih yamam devasya vajinah ; or
2. 5. 1, gakema vajino yamam.
Another case of haplology as between two individual words,
namely tavasarh rabhasva for tavasarh(sam)rabhasva occurs
in AV. 11. 1. 14:
ema" agur yositah ?umbhamana ut tistha nari tavasam
supatni patya prajaya prajavaty a tvagan yajfiah prati
2 AMERICAN JOURNAL OF PHILOLOGY.
A glance at the lists under root rabh in Grassmann's and
Whitney's Indexes to RV. and AV. shows that the simple root
without prepositional prefixes scarcely occurs in either text.
Whitney, indeed, lists the simple root only in that very
passage, AV. 1 1. i. 14. In a note on p. 614 of my translation
of the hymn, SBE. xlii, I stated that Sayana reads in pada b,
tava sariirabhasva, and the Paippalada, tavah samrabhasva. I
remarked, furthermore, that the original reading may have
been, tavasarh (sari:) rabhasva. Of this suggestion the Whit-
ney-Lanman translation takes no note. Yet it furnishes the
key to the passage: 'The maidens (the waters), ornamenting
themselves, have come hither. Arise thou, woman, take hold
of (sarh rabhasva) the strong one (tavasam, i. e., the pitcher,
kumbham, masculine) ! '
Once more in a Vedic text, simple rabh, without preposition
calls for correction. In MC. 3. 5. 13 we have the corrupt
stanza, as edited by Knauer :
agnir bhagah savitedarh jusantarii prajapatir varuno . . .
ya . . . mahyam,
yo no dvesti tanurh rabhasvanagaso yajamanasya viran.
There are two parallels to this corrupt stanza. AV. 9. 5. 2 :
indraya bhagarh pari tva nayamy asmin yajiie yaiamanaya
ye no dvisanty anu tan rabhasvanagaso yajamanasya virah.
And Ap£. 7. 17. 2 :
indrasya bhagah suvite dadhatanemarii yajnarh yajamanarii
yo no dvesty anu tarii ravasvanagaso yajamanasya virah.
Whitney, in his translation of AV., does not mention the par-
allels ; Knauer, at MQ., is cognizant of AV. 9. 5. 2, but not of
ApC. The intricacies of these correspondences need not con-
cern us at present. I would remark, however, that Knauer's
mss. read at the beginning agner bhagah, which corresponds
to the parallels, and is probably to be retained in the text.
The point that concerns us here is that the third pada of MQ.
is to be read yo no dvesty tarn anu rabhasva. In ApQ. ravasva
is secondary, tho perhaps intentional ; see the author, AJPh.
SOME VEDIC CRVCES. 3
I note in this connection some cases of haplology in chance
compounds. In RV. I. 48. 2 the compound vicvasuvid, by the
side of acvavant and gomant, rendered by ' knowing all well ',
makes no sense in that connection (Usas). The word is
vig(va)-vasuvid 'getting all goods'; see Usas's epithet abha-
rad-vasu, ' bringing on goods ', 5. 79. 2 ; and cf . the word
vasutvanam in the related stanza, 7. 81. 6, or the expression
utoso vasva Igise in 4. 52. 3. In Uluka-Jataka (270) appa-
tissavasa, 'living in anarchy', is for a-ppatissa(va)vasa; and
in Dadhivahana-Jataka (186) mandukantaka, designation of
a plant is probably for mandu(ka)-kantaka 'frog-thorn'. In
Maharastri Prakrit, avaratta is for ava(ra)ratta = Skt. apa-
raratra, ' second part of the night ' ; see Jacobi's Ausgewahlte
Erzahlungen, p. 32, 1. 2>7- On the literature of haplology (or
haplolaly) 1 , which has of recent years grown apace, see last
Collitz, Das schwache Prateritum (Gottingen, 1912), p. 237 ff
I would draw attention particularly to instances of the phe-
nomenon as between two successive words : Bloomfield,
American Journal of Philology, xvii. 418; Schwyzer, IF. xiv.
24 ff ; xxviii. 300 ; and Wackernagel, KZ. xl. 546.
2. chardis for chadis, a case of contamination or word blend.
The two words in the caption are obviously related. The
metre of the Veda points to chadis instead of chardis in all
critical positions: RV. 1. 48. 15; 8. 9. 1 ; 18. 21 ; 27. 4; 67. 6;
j 1. 14. Grassmann (as after him others) outlines the problem
very neatly in his Lexicon, s. v. : ' chardis, wofur wahrschein-
lich iiberall chadis zu lesen ist, da sammtliche metrisch ent-
scheidenden Stellen die Kiirze der ersten Silbe fordern und
keine deren Lange begiinstigt. Das r scheint in die spatere
Redaction durch Missverstandniss hineingedrungen '. For
other discussions see Oldenberg, ZDMG. lv. 312, and the lit-
erature there cited.
What now is the nature of this ' misunderstanding ', and is
it really such ? Grassmann's statement is very well as soon
as we substitute for misunderstanding the linguistic term
' contamination ', or ' blend '. The poets of the RV. knew only
1 In sense haplolaly is preferable to haplology, but the former word
with its three l's ironically invites the very change which it aims to
describe, as, indeed, also does haplology with its two lo's.
4 AMERICAN JOURNAL OF PHILOLOGY.
the word chadis, ' cover '. Like other words of this semantic
class the word meant both 'cover' (in the physical sense), and
' protection ' ; cf ., e. g., varma, ' armor ', and, ' protection '. In
the more concrete sense of ' cover ' chadis occurs in RV. 10. 85.
10; AV. 3. 7. 3, and it endures down to Kathasaritsagara 2. 49.
In the abstract sense of ' protection ' the word blended with,
or was contaminated by, garma ' protection ', taking its r from
that word. Again in that form the word endures clear
thru to Maharastri Prakrit chaddt (Jacobi, Erzahlungen, p.
76, 1. 32). The contamination obviously took place in the
time that passed between RV. composition and RV. redaction.
At the time of the redaction the word for ' protection ' had so
definitely assumed the form chardis that the diaskeuasts of the
RV. had to substitute it for the poets' chadis, metre contradi-
cente. The old word chadis had completely sloughed that
That all this is indeed so, is rendered probable by the inti-
mate and persistent synonymy of garma and chardis. Thus
the line RV. 7. 52. 2 b , garma tokSya tanayaya gopah, is echoed
in the formula, chardis tokaya tanayaya yacha, TB. i. 1. 7. 1 ;
ApQ. 5. 12. 1. In RV. 1. 114 5 d both words occur together,
carma varma chardir asmabhyam yansat. Almost every quali-
fying expression that is used with carma is also used with
chardis. Thus trivarutha, ' offering threefold safety ', or varu-
thya, ' offering safety ' ; or varutha by the side of each :
garma no yahsan trivarutham, 10. 66. 5
savita" garma yachatv asme trivarutham, 4. 53. 6
sa nah garma trivarutham vi yansat, 8. 42. 3
garmana nas trivaruthena pahi, 5. 4. 8
trivarutham maruto yanta nag chardih, 8. 18. 21.
Cf. also MS. 2. 8. 7*: m. 4; KS. 17. 6; TA. 2. 5. 2.
garma . . . varuthyarh tad asmSsu vi yantana, 8. 47. 10
bfhaspatih garma ... no yamad varuthyam, 5. 46. 5
chardir yad varh varuthyam, 6. 67. 2
bhava varutham . . . maghavadbhyah garma, 1. 58. 9
garma no yantam amavad varutham, 4. 55. 4
achidrarh garma yachata . . . varutham, 8. 27. o
yad vah . . . varutham asti yac chardih, 8. 67. 9.
SOME VEDIC CRUCES. 5
Or, again, adjectives for 'broad' go with both nouns: uru,
prthu, and especially saprathah :
yacha nah garma saprathah, i. 22. 15
saprathah garma yacha sahantya, 6. 16. 33
chardir yacha vitahavyaya saprathah, 6. 15. 3
saprathah chardir yantam adabhyam, 8. 5. 12
urv asma aditih garma yansat, 4. 25. 5
pra no yachatad avrkarh prthu chardih, 1. 48. 15
prssmai yachatam avrkarh prthu chardih, 8. 9. 1.
As regards other adjectives, or other related connections,
the following pairs or groups speak for themselves :
duradharsarh grnate garma yansat, 6. 49. 7
adhrstarh chardir yad vam, 6. 67. 2
bhava . . . maghavan maghavadbhyah garma, 1. 58. 9
chardir yacha maghavadbhyag ca rhahyarh ca, 6. 46. 9
(cf- 7-74-5; 8.5. 12)
garma tokaya tanayaya gopah, 7. 52. 2 (cf. TB. 1. 1. 7. 1)
adha sma yacha tanve tane ca chardih, 6. 42. 12.
On the character and frequency of lexical contaminations
see the author, Am. Journ. of Philol. xii. 1 ff . ; xvi. 1 ff. ; Indo-
germanische Forschungen, iv. 66 ff. ; and most recently Giin-
tert, Ueber Reimwortbildungen im Arischen und Altgriech-
ischen (Heidelberg, 1914).
3. Some ^xvi Juira -
The two stanzas, RV. 1. 4 5, 6 read :
uta bruvantu no nido nir anyatag cid arata,
dadhana indra id duvah.
uta nah subhagah arir voceyur dasma krstayah,
sytmed indrasya garmani.
The renderings mark a to and fro from a correct under-
standing : Bollensen, Orient und Occident, ii. 462 ; Ludwig,
443 ; Grassmann, ii. 5 ; Pischel, ZDMG. xl. 125 ; Geldner, Ved.
Stud. iii. 79 ; Oldenberg, Rig- Veda Noten i. 4. Geldner comes
nearest to the correct sense ; I would translate :
'Whether our enemies happen to say (about us) : "when
ye pay respect to Indra alone (id), ye have shut yourself off
6 AMERICAN JOURNAL OF PHILOLOGY.
from other (benefits)"; or, if both gentle and common folk
should, O wise (God Indra), pronounce us lucky, (in either
case) shall we under Indra's protection be.' 1
uta-uta are clearly antithetic. The two stanzas seem to
express an almost sectarian difference between Indra wor-
shipers and people who despise Indra (anindra), but worship
other gods. Intentionally I render arih . . . krstayah by ' both
gentle and common folk ', i. e., ' rich and poor ', or ' patricians
and plebeians '. Arih has been suspected (Bollensen suggests
arih). But it is correct and idiomatic ; we may call it participa-
tive singular. Johannes Schmidt, Die Pluralbildungen der
Indo-germanischen Neutra, pp. 314 ff., following Roth's sugges-
tion in Pet. Lex., s. v. rathatur, has shown that an inclusive plural
noun with a plural verb is occasionally in Sanskrit and Greek
accompanied by a singular noun which expresses part of the
plural noun. 2 In our passage krstayah includes both plebs
(vicve, ol noXXoi) and patricians (ari) ; hence the participative
singular arih, by the side of and partly denning krstayah with
the plural verb. Cf. for this idiom also Ernst Fraenkel, Indo-
germanische Forschungen, xxviii. 245 ff. For krsti, ari, and
viqva see Geldner, Ved. Stud. iii. JJ ff.
I am conceiving the matters involved here rather more pre-
cisely than does Geldner. krsti ( car sani), 'people' is the
totality which includes ari and vicva, ' noble and common ', its two
natural subdivisions ; see 7. 48. 3 ; 8. 1. 22 ; 51. 9 ; 65. 9 ; 10. 28. 1.
In Geldner's rendering (p. 78) of vicvagvirto aristutah in 8. 1.
22, 'der von alien Geriihmte, (sogar) von den Reichen Ge-
priesene', the word 'sogar' is needless. The expression
means, ' he who is praised by plebs and " swell " alike '. Behind
these two words stands the comprehensive krstayah ' all folks '.
Another idiom, familiar in the Indo-European languages,
ensures a similar effect, namely to mark the contrast between
ari and vicva: RV. 10. 28. 1, vicvo hy anyo arir ajagama
mamed aha cvaguro na" jagama. Geldner, p. 78, translates,
'Jeder andere, (sogar) der Reiche ist erschienen; nur mein
Schwiegervater ist nicht erschienen'. This neither does jus-
' The last pada is repeated secondarily in 8. 47. 5 C .
* In Greek rhetoric this idiom is denned as axwa Ka *' 8X0* *ai /xcpos ;
Kiihner, Grammatik der griechischen Sprache 3 , Satzlehre (Kiihner-
Gerth) vol. i, p. 289.
SOME VEDIC CRUCES. 7
tice to the established contrast between viqva and ari, nor to
the idiom involved in anya. Translate : ' The common folk
and (the others, sc.) the nobles have come, etc' * This is the
well-known anticipatory-appositional use of anya, common in
Sanskrit, the exact replica of a familiar Greek idiom with
aAAo-, e. g., Xen. Anab. i. 5. 5> °*> y<*p v v x°P T0S °"^ aAAo SivSpov.
See Ktihner, Satzlehre 8 , vol. i, p. 275, note 1 ; the author,
Amer. Journ. of Philology, vii. 101 ; Pet. Lex. vol. i, p. 262 b ,
where examples from Classical Sanskrit are cited abundantly
but no Vedic cases. Another such case is contained in RV.
1. 109. 6 where the word anya in vicva bhuvana . . . anya con^-
trasts vicva bhuvana, ' all creatures ', with a list of things that
are not creatures. I suspect that other cases may turn up in
the Veda. 2
This idiom is familiar in modern French, in connection with
plural pronouns ; e. g., nous autres Francais ; nous autres
femmes. It is also known in Spanish and other Romance
tongues. Kiihner-Gerth, 1. c, also quotes the idiom, less
familiarly, from Latin and Modern High German (Luther
and Goethe) ; cf. also Kuhner, Ausftihrliche Grammatik der
Lateinischen Sprache, ii, § 119, note 17. Altogether the idiom
is found in Vedic and in Sanskrit ; in Greek ; in classical and
in Late Latin ; in Spanish, Provencal, and French ; in Middle
and Modern High German. See in general Bockh, Encyklo-
padie und Methodologie der philologischen Wissenschaften
(1877), p. 105; Meyer-Liibke, Grammaire des Langues Ro-
maines, vol. ii, § 75 ; iii, § 209 ; Diez, iii, p. 84 ; Tobler, Ver-
mischte Beitrage, iii 1 , p. 72 ; Hanssen, Spanische Grammatik
§ 56. 2 ; Gessner, Zeitschrift fur Romanische Philologie, xix.
155 ; Appel's Old-Provengal Chrestomathie, St. 16, vv. 29-32.
1 More fully : ' Both plebs and patricians have come ; my father-in-
law alone has not come.'
2 The use of vigvah, singular, as a collective in the sense of 'plebs',
approximates the word to the plural vigve, in the same sense. This
perhaps accounts for the seemingly senseless substitution in TS.
1. 2. 2. 1, of the plural vigve for the singular vigvah with a singular
verb (as in the other versions) : vigve devasya netur marto vrnita sak-
hyam, 'the plebs, the mortal shall choose the companionship of the
god that leadeth'. See last Keith, Veda of the Black Yajus, HOS.
vol. xviii, p. 21.
8 AMERICAN JOURNAL OF PHILOLOGY.
4. On the expression navyam sanyase.
The expression navyam sanyase occurs, as far as I know,
three times in the RV. 1 and once in the Mahanamni-verses of
the SV. and Aitareya-aranyaka, which makes it easy to take
The Pet. Lexs. and Grassmann in his Lexicon translate
sanyas by ' old ', ' older ', without indicating in any way how
the word is to be rendered in its connection. Geldner in his
Glossary renders the entire expression navyam sanyase by
'was dem altesten neu ist, d. h. etwas ganz neues, noch nie
dagewesenes '. Keith, in his Translation of Aitareya-aranyaka,
p. 263, suggests for navyam the meaning ' praiseworthy ' ( from
root nu). As regards translations it will be well to review the
proposals for each passage. RV. 3. 31. 19 reads:
tam angirasvan namasa saparyan
navyam krnomi sanyase purajarn.
Grassmann, vol. i,p. 530: ' Mit Anbetung ihn nach Angiras-
art verehrend, mach ich das alterzeugte [Lied] neu dem alten
( sc. god Indra) '. Ludwig, 498 : ' wie Ahgiras mit Anbetung ihm
dienend, mach ich ihn neu zum gewinnen, den ehvordem ent-
standenen '. Ludwig does not comment upon his rendering :
obviously he regards sanyase as an infinitive of root san
' obtain '. Oldenberg, Nachrichten der koniglichen Gesell-
schaft der Wissenschaften zu Gottingen, 1915, p. 381 : ' ich
mache durch meine Anbetung den alten Gott neu (navyam)
fur alte Tat (damit du diese auch jetzt wieder tun mogest)'.
It is easy to show that Grassmann was pretty close to the truth,
tho he did not quite get it, and lapsed, as we shall see, in
the two other RV. passages : navyam krnomi sanyase pura-
jarn contains, to my mind, a playful paradox : ' I make a new
song (brahma) that is (in reality) primordial (purajarn) for
the good old (sanyase) god'.
The passages which show this to be true are of an almost
mathematical insistence. RV. 1. 62. 13 : sanayate gotama indra
navyam ataksad brahma hariyojanaya, 'Gotama has fashioned
for thee, O Indra, the old god, a new song, in order that thou
mayest hitch thy bay steeds '. This paraphrases sanyase by
3. 31. 19 ; 8. 24. 26 ; 67. 18.
SOME VEDIC CRVCES. 9
sanayate, and fixes upon navyam the noun brahma, and from
this, as will appear, there is not a hair's breadth of deviation.
See next 10. 91. 13, imam pratn&ya sustutirh naviyasim voce-
yam asma ucate, ' let me pronounce for the god of yore, that
is willing, this quite new song'. Here pratnaya sustutirii
naviyasim = navyam (brahma) sanyase. Next, 6. 62. 5, ta
valgu dasra purucakatama pratna navyasa vacasa" vivase, ' these
two lovely, most powerful Dasras (Acvins) of old I invite
with a quite new song'. Here pratna" navyasa vacasa =
navyath (brahma) sanyase. RV. 6. 22. 7: tam (sc. indrarh)
vo dhiya navyasya Qavistham pratnam pratnavat paritarisaya-
dhyai, 'deck out that mighty (Indra) of old with a new hymn
as of old '. Here dhiya navyasya pratnam = navyam (brahma)
sanyase. In 1. 61. 2 the antithesis between the 'new song'
and the 'old god' is implied almost as clearly as tho it
were directly expressed : asmai . . . indraya . . . pratnaya patye
dhiyo marjayanta, 'let them polish up their songs for Indra
the lord of old '. The word marjayanta ' polish up ' here well
takes the place of ' new '. The word pratna is a favorite in
such connection, as may be seen in such passages as 6. 39. 5 ;
10. 4. I.
In a slightly more remote way the antithesis between the ' new
song ' and the ' old god ' is in the mind of the author of
2. 17. 1 : tad asmai navyam angirasvad areata cusma yad asya
pratnathodfrate, 'this new (brahma) sing ye for him (Indra)
in the fashion of the Angiras in order that his fire shall be
aroused as of old ' (note angirasvad, implying the past, as well
as pratnatha). Here navyam (brahma) pratnatha afigirasvat
= navyam (brahma) sanyase. Yet more round about, 8. 95. 5 :
indra yas te naviyasirii girarh mandram ajijanat . . . dhiyarh
pratnam. And 9. 9. 8 : mi navyase navlyase suktaya sadhaya
pathah, pratnavad rocaya riicah.
It is clear now that the expression navyaih (brahma) san-
yase (devaya) is elliptic. The reason why we Westerners are
slow to understand such an expression is, that the Vedic
Hindus understood it too well. The underlying idea, as the
passages just cited show, must have become immensely familiar,
before they could express it by simply saying ' a new for an
old'. The same almost kenning-like familiarity of the ex-
pression guarantees beforehand that it could not have been
IO AMERICAN JOURNAL OF PHILOLOGY.
used in any other sense than just that. With this reasoning the
facts chime in perfectly. RV. 8. 67. 18 reads :
tat su no navyam sanyasa aditya yan mumocati,
bandhad baddham ivadite. 1
Grassmann, who came nearest to understanding 3. 31. 19,
lapses from grace utterly: 'Zum alten fiigt dies neue ihr,
Aditya's, was, O Aditi, uns lost wie Sklaven von dem Strick '.
Ludwig, 126: 'disz neue sei uns zum gewinne, was erloset,
o Aditya, wie aus f esseln den gebundenen, o Aditi '. Bergaigne,
iii. 161, omitting apparently sanyase: 'void notre nouvel
(hymne) qui doit nous delivrer, 6 Adityas, comme un homme
lie de son lieu, 6 Aditi '. Oldenberg, 1. c. : ' diese neue Tat
(wird) uns (zuteil) zum Zweck des alten — d. h. damit die
alte Tat fortwirke, sich erneuere.' Translate: 'This, pray, is
our new (song) for a right old (god), which, O Adityas, shall
free us as a captive from his chain, O Aditi '.
The third occurrence of this cliche is in RV. 8. 24. 26 :
tarn u tva nunam Imahe navyam dansistha sanyase,
sa tvarii no vicva abhimatih saksanih.
Grassmann : ' Darum begehren wir von dir zum alten neues,
herrlicher, sei du Vertilger aller Widersacher uns '. Ludwig,
597, 'als solchen flehen wir dich jetzt an, den frischen, wun-
dertatigster, zu gewinne, als solcher bist du es, der uns alle
nachstellungen iiberwindet'. Oldenberg, 1. c. : f wir gehen
dich den Neuen (d. h. erneut sich Betatigenden) an fur die
alte Tat (damit du diese auch jetzt wieder tun mogest)'. In
this stanza the construction of imahe with two accusatives,
rather than with accusative and instrumental is unusual : ' We
pray to thee now, O most wise (Indra), a new (song) for a
right old god : thou art the conqueror of all that plot against
us '. Cf. pratnabhir utibhis by the side of imahe in 8. 12. 24;
or yajnesu purvyam by the side of Imahe in 8. 60. 2. Perhaps
pada b is to be taken parenthetically : ' We implore thee now —
a new song for a right old god ' — etc. There is, in any case,
not the faintest reason for taking navyam sanyase in a different
1 The stanza is paralleled interestingly in 8. 18. 12 : tat su nah garma
yachatSditya yan mumocati, enasvantam cid enasah sudanavah.
SOME VED1C CRUCES. "
The formula crops out once more in a passage of the Ma-
hanamni litany :
nunarh tan navyath sanyase
prabho janasya vrtrahan. 1
Oldenberg, 1. c, '(ist) diese (Tat) nunmehr neu fiir die alte'
(d. h. zum Zwecke der Erneuerung der alten). Or, '(rufen
wir) diesen neuen jetzt (an) fiir die alte Tat' (d. h. damit er
seine alte Tat wiederhole). Translate: '(We sing) now this
new (song) in honor of the right old (god), O thou that art
distinguished among the people, O slayer of Vrtra ' ! That the
poets diligently describe Indra as ' the god of yore ' follows
from the preceding passages, and can be easily corroborated
by further evidence which is in everybody's hands.
5. On stanza 6 in the hymn of Sarama and the Fanis,
RV. 10. 108.
Both the meaning of some of the words and a certain un-
couth quality of the construction, which obviously states para-
tactically what, to our feeling, should be stated hypotactically,
have kept this stanza a crux interpretum. There is scarcely
a Vedic scholar who has not in one way or another tried his
hand at it. The following is an endeavor once more to clear
its difficulties. The stanza reads :
asenya vah panayo vacarisy anisavyas tanvah santu paplh,
adhrsto va etava astu pantha bfhaspatir va ubhaya na mrlat.
Ludwig (992): 'Nicht von waffenart (wenigstens) sind
eure reden ; gesetzt es waren dem pfeil nicht ausgesetzt eure
schlechten leiber, unbewaltigt der weg zu euch zu kommen,
Brhaspati wird euch in keinem falle (ob es ist oder nicht ist)
gnadig sein'. 2
Grassmann : ' Nicht treffend sind, o Pani's eure Worte ; und
waren schussfest eure bosen Leiber, und undurchdringlich
auch der Weg zu euch hin, Brhaspati wird beides nicht
Geldner und Kaegi, Siebenzig Lieder, p. 79 : ' Mit Worten,
Pani, konnet ihr nicht fechten; und waren schussfest eure
1 For the text see Oldenberg, 1. c, pp. 377, 381.
2 Cf. his comment with other suggestions.
12 AMERICAN JOURNAL OF PHILOLOGi .
schlechten Leiber, der Weg zu euch auch noch so schwer zu
zwingen, das alles wird Brhaspati nicht kiimmern '.
Von Schroeder, Mysterium und Mimus, p. 175 : ' Nicht
Wunden schlagen, Panis, eure Worte ! Und waren schussfest
eure schlechten Leiber, war auch der Weg zu euch schier un-
bezwinglich, Brhaspati wird beides nicht verschonen'.
Hertel, WZKM. xviii. 60 : ' Mit Worten, Pani, konnt ihr
nicht versehren ; wenn schussfest eure siind'gen Leiber waren,
und unzuganglich alle eure Pf ade : Brhaspati versagt euch
seine Gnade '.
Hillebrandt, Lieder des Rig- Veda, p. 147: 'Wehrlos sind
eure Worte, Panis. Eure Leiber, die hasslichen, mogen un-
durchdringlich f iir die Pf eile, der Weg zu euch mag unnahbar
sein : Brhaspati wird euch in keinem Falle gnadig sein '.
The most critical word in the stanza is ubhaya. Those
translators who take the word in the sense of ' beides ' are
obviously in error; the accent shows that it is adverbial
(ubhaya from stem ubhaya) meaning 'in either case',
' whether so or so '. Thus Ludwig in his translation ; Bar-
tholomae, IF. v. 227, note 3; Oldenberg, RV. Noten to the
passage. 1 The stanza thus contains an alternative between
two suppositions ; the question is where to place the hinge or
seam between the alternatives. In this we must be guided by
santu and astu which harbor the idea ' whether it be so or so '.
Now it is clear that santu controls padas ab ; astu pada c ; the
conclusion comes in pada d. We may expect something
favorable and something unfavorable to the Panis : ' in either
case Brhaspati shall not spare you' (pada d).
The entire first couplet contains the something unfavorable
to the Panis. Here, namely in the word anisavyas, may be
found the solution of the difficulties of the stanza. The stem
anisavya is rendered in the translations above by ' impervious
to arrows'. So also, unanimously as far as I know, the
lexicons ( Pet. Lexs. ; Grassmann ; Monier Williams ; Ber-
gaigne, Etudes sur le Lexique du Rig- Veda; Hillebrandt,
Vedachrestomathie, etc.). 2 The word means nearly the very
opposite of that, for it contains isavya with the negative
1 On the grammatical aspect of ubhaya" see last Wackernagel, Altin-
dische Grammatik, ii. 1, p. 21.
'Myself in the same error, ZDMG. xlviii. 549, note 3.
SOME VEDIC CRUCES. 13
prefix. Now isavya means ' war-like ' literally, conversant
with "arrows': a rastre rajanyah cura isavyo (VS. CB. add
'tivyadhi) jayatam VS. 22. 22; MS. 3. 12. 6: 162. 7; CB. 13.
1. 9. 2; asmin rastre rajanya isavyah euro maharatho jayatam
TS. 7. 5. 18. 1 ; KSA. 5. 14; TB. 3. 8. 13. 1 (cf. QQ. 8. 18. 1 ;
JUB. 1. 4. 2). The passages speak for themselves: isavya is
the same as isu-bala (by the side of citra-sena; cf. asenya in
our stanza 1 ), RV. 6. 75. 9 ; or the isuman viro asta, RV. 2. 42.
2. 2 . Therefore anisavya means, ' unwarlike ', lit., ' not inured
to arrows '. And by the same terms asenya means ' not in-
ured to missiles ', i. e., again, synonymically, ' unwarlike '. The
conception indrah senyah, 1. 81. 2 ; 7. 30. 2, hovers before the
mind of the writer as the opposite of asenya.
It can be seen now what the stanza means: 'Whether (on
the one hand), O Panis, your words be feeble, your vile bodies
cowardly; or whether (on the other hand) the road to you be
hard to dare, in either case Brhaspati (Indra's Purohita) will
show you no mercy '.
While on the theme of 10. 108, I would remark that stanzas
9 and 10 have always seemed to me post f estum and anti-
climax. These two stanzas fit better after stanza 2, where
they would in no sense disturb the sequel of the hymn. Or,
they may be imagined as standing in the same position in the
place of 3 and 4, of which they would be a not bad alternate ver-
sion. Cf . 3 cd , mitram ena dadhamatha gavarii gopatir no bha-
vati, with 9 cd , svasaram tva krnavai ma" ptinar ga apa te gavarh
subhage bhajama; and again cf. 4 a , naharh tarn veda dabhyarh
dabhat sa, with io a , naharh veda bhratrtvam no svasrtvam. I
have a sort of Valakhilya feeling as regards the two pairs. If
this be so I need scarcely point out that stanza 11, in close
catenation with 10, was composed after 9 and 10 got their
place where they now stand in the hymn.
6. On the meaning of ukhachid.
Windisch, Festgruss an Otto von Bohtlingk, p. 115, has
made it clear that this compound means ' lame ', being a peri-
phrasis — we might say a sort of kenning — of grona. The
1 For sena in the sense of ' missile ' see last Bloomfield, ZDMG.
xlviii. 549, note 1.
' Cf. also, more remotely the type of passage such as 3. 4. 9 = 7. 2. 9.
where sudaksa seems to occupy the place of isavya or isuman.
14 AMERICAN JOURNAL OF PHILOLOGY.
word is air. Xty. in RV. 4. 19. 9, nir bhud ukhachit sam aranta
parva, 'the lame man was off; his joints fitted together'. Cf.
the convincing parallel in 8. 79. 2, nih Qrono bhut, ' the lame
man was off '. Now he finds the word ukha in certain gram-
matical word-lists (ganas) among words for parts of the body,
and one commentator explains it by sphik, 'hip'. There is
nothing to prevent ukha from having that meaning in a figur-
ative way, though it is, as far as I know, not quotable in that
sense in Hindu literature. Windisch next assumes that ukha-
chid means 'one who has broken his hip' ('der einen hiiften-
bruch erlitten hat'), therefore, 'lame'.
We should expect ukhachinna rather than ukhachid, ' hip-
breaker', which would seem to state something habitual,
whereas the lame man would break his hip only a single time.
The analogy is with compounds like grivachinna ' one whose
neck is cut ', Suparnakh. 25. 6 ; grivabhagna, with the same
meaning, Vetalap. 17. 6; grivabaddha, 'bound by the neck'
TS. 3. 3. 8. 3, janvakna (comm. sambhugnajanu), 'with bent
knee ', Ap(J. 10. 9. 2. Moreover there is that in the literature
which leads me to suspect that ukha in ukhachid * has its primary
meaning of ' pot ' or ' pan '. Thus QB. 6. 6. 4. 8 : yady esokha
bhidyeta, ' now if this pot breaks ' ; TS. 5. 1. 9. 2 : sa (sc. ukha)
yad bhidyetartim archad yajamano hanyetasya yajiiah, ' if this
(pot) be broken, the sacrificer gets into trouble, his sacrifice
is destroyed ' ; ApQ. 10, 5, 3 : mitraitarh ta ukham paridadamy
abhittya esa ma bhedi, ' O Mitra, I make over to thee this pot
unto unbreakableness ; it shall not be broken '. Breaking of
the pot (ukhabhedanam) is provided for ritualistically in KC.
16. 7. 8. The ukha was fragile, being made of clay (mrnmayl)
which was baked (grapaya), VS. 11. 59, et al. The ukha leaks
easily : ma susroh ' do not leak ', AV. 12. 3. 12 ; ukharii sravan-
tlm ' the leaking pot ', KQ. 25. 9. 14 ; MQ. 3. 5. 14. It has to
be placed firmly on the altar to keep it secure : ukha kumbhi
vedyarh ma vyathisthah, ' do not, pot or kettle, wobble on the
altar ', AV. 12. 3. 23 (cf . MS. 4. 1. 3 : 4. 9 ; TB. 3. 2. 3. 1 ) . In
case it broke a new one had to be made, Vait. 28. 12.
It would seem then that the fragile ukha was found to be
less secure in the hands of a lame man, who might thus be
nick-named ' pot-breaker '. It is not necessary to inquire how
1 The short a is rhythmical ; see Leumann, Gurupujakaumudi, p. 13.
SOME VEDIC CRUCES. 15
much fancy and how much fact there was at the bottom of
the notion. Persons with bodily defects are apt to be nick-
named all over the world ; another Skt. designation of a lame
man, ekapad, ' One-leg ', shows the same spirit.
7. Irregular Belative Clause Constructions.
The poet Agastya, in RV. i. 176, seems to me to have diffi-
culty in handling a relative with its antecedent pronoun, stem
ta-, or, perhaps some metrical consideration led him to un-
usual passes in this same matter. Stanza 5 reads :
a~vo yasya dvibarhaso 'rkesu sanusag asat,
ajav indrasyendo pravo vajesu vajinam.
Previous treatments are listed by Oldenberg, Rig- Veda Noten,
i. 176. It seems to me well, in the first place, to comment
upon Geldner's ingenious translation, Ved. Stud. ii. 129 : ' Wen
du, O Soma, bei einem Wettkampfe zu Ehren des doppel-
starken Indra schiitztest, der wird in den Liedern ordentlich
sein ; du beschiitztest den im Kampf e siegreichen '. The for-
mal difficulty in this rendering is the accent of asat, which
disqualifies it from service in the principal clause of the sen-
tence, but points to the subordinate clause.
As regards the sense, I do not believe that there is any
indication of a race ('wettkampf ') in honor of Indra (if so,
where?). When a Vedic text says ajav indrasya, it can, in my
opinion, have in mind but one thing, namely, ' in the contest
for Indra ', that is to say, in the attempt to secure the pres-
ence of the 'much-called' god (puruhuta, and the like) ; see,
e. g., 6. 19. 3 d , asman indrabhy a" vavrtsvajau. Moreover,
Geldner's translation, as a whole, carries with it an obvious
hysteron proteron which is suspiciously parallel to the wrong
accent of asat. We should expect the poet to say that Soma
inspires him that composes skilful poems to secure Indra's
presence, rather than that he whom Indra protects is skilful in
poetry. The latter might be an idea applicable later to the
court of a Bhoja Raja; it is not a Vedic idea.
Oldenberg's suggestions are advanced hesitatingly, espe-
cially as regards sanusak which, he thinks, may be a com-
pound = sa-anusak, or = sanu-sak. Tho the Padapatha
does not analyze sanusak, and tho it has but one accent, I
believe with others, that we must read sanusak = sa + anusak.
16 AMERICAN JOURNAL OF PHILOLOGY.
The passage seems to state : ' Thou didst aid him who is clever
(anusak) in songs in honor of twice-strong (Indra) ; didst,
O Indu (Soma), aid him in the contest for Indra (the much-
called). Thou didst aid in obtaining substance him that hath
substance '. For arkesu anusak cf ., e. g., 5. 8. 10, girbhir
yajfiebhir anusak. In the first distich of our stanza the
logical antecedent tam, ' him ', seems to be incorporated in the
relative clause as sa in sanusak. This accounts for the obscur-
ation of *sanusak, and the consequent loss of one of its
accents. Moreover yasya seems to be for yo asya, or for single
yo attracted to the case of dvibarhasah. 1 Here is what the
passage seems to say in good Vedic: avo tarn yo (oryo asya)
dvibarhaso ' rkesv anusag asat (thus the accent of asat is
justified); ava ajav indrasyendo ; pravo vajesu vajinam. The
fourth pada, repeated in 1. 4. 8, looks a little like an appendage,
and may have been borrowed from that stanza.
The same Agastya in the same hymn, 1. 176. 2, seems once
more to have assimilated a relative pronoun (attraction) to
another word in the same relative clause :
tasminn a vegaya giro ya ekag carsaninam,
anu svadha yam upyate yavarh na carkrsad vfsa.
See Oldenberg, Rig- Veda Noten, and the literature there
cited ; and cf. Colinet, Sur le sens du mot svadha, p. 14. Two
points seem to me to control the explanation of this curious
passage. First, the words anu svadha belong together = anu
svadha"h 2 ; this eliminates the need of combining anu and vap,
a combination otherwise unknown in the RV. This on the
evidence of 9. 103. 5; 10. yj. 5 (anu svadhah) ; 1. 33. 11 ; 88.
6 ; 3. 51. 11 ; 4 33. 6 ; 52. 6 ; 7. 56. 13 ; 8. 88. 5 (anu svadham) ;
and anusvadham, frequent adverb. All mean, ' according to
habit or custom '. Should this be so, then, secondly, yam in
pada c cannot be construed, unless we regard it as attracted
1 Cf. Neisser, ZDMG. lxi. 138, and Oldenberg's note, Rig- Veda Noten,
i, to 4. 21. 1 for similar phenomena regarding the relative. Cf. also the
same author. Rig- Veda Noten, ii, p. 379b (Relativsatz frei ange-
2 Thus, previously, Bergaigne, iii. 9, note. The Padapatha, of course,
in not explaining svadha~ as svadhf h, must have interpreted the word
as nominative, subject of upyate with anu. AH attempts to interpret
on this basis strike me as forced and unbelievable.
SOME VEDIC CRUCES. 1 7
from the nominative (yas) to the accusative yavam in its own
Under this construction, the stanza, addressed to Soma
would run as follows: 'Make enter into him (sc. Indra), who
is sole (ruler) of the peoples, our songs, as a bull ploughs
(i. e. makes enter) grain (into the field), grain which is sown
according to (established) custom!' That is to say, omitting
the attributive pada b, 1 the stanza is equivalent to the follow-
ing : tasminn a vegaya giro yavarh na carkrsad vfsa, anu svadhS
ya (sc. yavah) upyate. The sense then were clear : the poet
asks Soma to enter Indra (i. 176. 1 ; 9. 2. 1), and, as he
enters, to carry with him the poet's songs, in order to ensure
Indra's gratitude to the poet. He must do this as regularly
or steadily as the plough-steer, according to established
custom, ploughs grain into the soil. That the expression,
yavarh na carkrsad vrsa, does not require any further descrip-
tion, such as is contained in the fictitious anu upyate, ' pour-
ing the grain in after the plough has ploughed ', may be
gathered from the repetition of the idea in 1. 23. 15, gobhir
yavarh na carkrsat.
In our interpretation of 1. 176. 2 C much weight is given to
the habitual adverbial expressions in which various forms of
svadha are governed by anu. I should be loath to see this
argument exercise undue influence. In one RV. passage,
5. 34, 1, juxtaposition of anu and svadha" is entirely fortuitous :
anu does not govern svadha, but belongs to iyate :
ajataqatrum ajara svarvaty anu svadhtmita dasmam iyate,
sunotana pacata brahmavahase purustuta"ya pratararh da-
Roth in Pet. Lex. s. v. 1. svadha 3) was under the influence
of those adverbial expressions when he suggested the reading
anu svadhiim amita in this stanza, but svadha (nominative) is
here personified : ' Svadha unaging, full of light, unmeasured
follows the wise god (Indra) whose enemy is yet to be born ' ;
see in AV. 2. 29. 7, the urja svadha" ajara, created by Indra ;
and the svadha ajara of the Fathers in 12. 2. 32.
1 Repeated in much better connection in 1. 7. 9. In our stanza the
pada is a dislocated fragment.
18 AMERICAN JOURNAL OF PHILOLOGY.
Once more, AV. 6. 53. i c , arm occurs before svadha under
rather trying circumstances : anu svadha" cikitarh somo agnih.
Whitney in his Translation observes that the compound verb
anu + ci does not occur elsewhere in the language but renders,
' let the svadha favor [me ; let] Soma, Agni '. Ludwig, Dei-
Rig- Veda, vol. iii, p. 506 translates, ' durch die Svadha denke
daran Soma, Agni '. He seems to make anu govern svadha as
a homophonous instrumental. I have thought of correcting
to anu svadharii (or svadhah), and thus escaping the dubious
combination anu + ci ; cf . AV. 6. 96. 3, somas tani svadhaya
nah punatu ; AV. 18. 3. 8, sarh somena madasva sarh svadhabhih.
But why should the text of 6. 53. 1 have slipped from the line
of least resistance (anu svadham, or anu svadhah) into anu
An incorporated relative conversely attracts secondarily its
subject in the clause which contains the attraction. RV.
10. 17. 9:
sarasvatlrh yam pitaro havante daksina yajiiam abhinak-
sahasrargham ilo atra bhagarh rayas posarh yajamanesu
Here sarasvatim is attracted from the vocative or nominative
to the accusative yam : the attraction is quite as illogical as
that of yasya for yah in 1. 176. 5, or yam for yah in 1. 176. 2.
Similarly in 1. 183. 1, tarn yufijatharh manaso yo javiyan
trivandhuro vrsana yas tricakrah, ' Yoke, O ye two bulls, that
car which is swifter than the mind, has three pole-boards, 1 and
three wheels '. Here trivandhuro . . . yas tricakrah ( for tri-
vandhuram, etc.) is attracted to the articular relative phrase
manaso yo javiyan (formula in i. 117. 2; 118. 1; 10. 112. 2).
'For the meaning of vandhura see my forthcoming work, Rig-Veda
Repetitions, HOS., vol. xx, p. 236. The work comprises volumes xx
and xxiv of Harvard Oriental Series.