I have read through Whitney’s chapter 18 of “A Sanskrit Grammar,” and
it's becoming increasingly evident to me how untenable is the original
thesis that “all bahubbiihi compounds must end in nouns.” Dr. Pind, Alan,
and Rett have all recently taken a more moderate stance, either making way
for exceptions (Ole), or accepting that the last member of a Bh is a noun or
an adj. used as a noun” (Alan, Rett). However, now with the benefit of
Whitney’s treatment, I am able to add a second line of argument. Up until
now I have been reasoning from empirical data: the examples I came across in
grammars simply did not all end with nouns. Whitney, however, shows that the
last element in a Bh *can’t* function as a noun! If it is originally a noun,
it must undergo changes in Sanskrit in order precisely to function as an
adj. This is because the bahubbiihi cpd (“possessive” cpd for Whitney) MUST
BE ADJECTIVAL-- after all, it qualifies the exoteric word. The rule to keep
in mind is this: a compound takes its part of speech from the last element
(Whitney paragraph 1247f, Warder 137). Therefore, because the compound in a
Bh must function adjectivlly, then the last element must also be either an
adjective or a noun functioning as an adj (not the reverse!). Whitney (W)
spends many pages in his “A Sankrit Grammar” showing the changes that the
noun must undergo (accent and ending), precisely so that they may be used
adjectivally in a Bh. In fact, this is his main concern.
I appreciate everyone’s willingness to change positions (as I’m also
willing to change mine if the evidence presents). Believe me, I’m in no way
putting myself up as anything but a beginniner in all this. I’ve got a lot
to learn, and am now enjoying the process of doing so with you.
Returning for a moment to the empirical argument, I accept Rett’s
corrections to the list of adj-ending Bh’s, which I incorporate below. Two
of the examples did not even have an exoteric element, so how they got on
the list I don’t know. Also, I’ve removed any examples that are not
certified by Warder or Perniola as Bh’s, and I provide page references. I’ve
also removed certain examples (asama, mattaa) over which there are two
possible interpretations. Finally, I’ve substituted new examples from the
grammars to replace the ones removed. Here is a revised list of Bh’s ending
with a non-noun:
duddasa dhamma: a doctrine hard to see (Warder/188)
sukata kamma: an action well done (Warder/188)
susannaddha bhaara: a load well tied up (Warder/188)
dhammaanudhammapa.tipanna bhikkhu: ‘a monk following the entire doctrine
(Warder/213)
manopubbangamaa dhammaa manosetthaa: ‘factors that have the mind as the
first and as the best’ (Pern/170)
anaasannavaraa etaa: ‘these are best when not near’ (Pern/170)
yaanena itthiuuttena purisantarena: ‘on a cart drawn by two calves with a
bull in between’ (Pern/172)
aanando atta-dutiyo: Ananda with his self as second (Pern/170)
dosantaro [manusso]: [a man] with hatred within (Pern/172)
I don’t think anyone will venture to claim that all these past participles
and adjectives are nouns or functioning “as” nouns!
From reading Whitney (W), it is clear that in Sanskrit an adjective can
readily assume the posterior (final) position in bahuvriihi compounds, which
W calls “possessive cpds.” This is evident from the description and examples
he gives (see below).
There are, however, a few idiosyncracies of W’s presentation. One is that
he spends almost all of the chapter writing about noun-ending compounds. He
devotes very little space to non-noun-enders. He does this for a good
reason: W is most concerned with the various changes in accent and in ending
that occur in Sanskrit when the noun compound is made adjectival, as is
necessary to turn what he calls a determinative cpd (Kh, Tp) into a
possessive (Bh). As I mentioned, a compound takes its part of speech from
the posterior element. Adj-ending cpd’s already end in adj’s, so no accent
change is required. Therefore, the great Sankritist devotes almost no time
to this latter category.
Nevertheless, W does give us the information necessary to form a clear and
unambiguous conclusion about non-noun finals. I hunted here and there among
the thirty pages for this information, which is parceled out among all the
noun-final information. It is not of much interest to W, because he is
concerned primarily with the changes necessary in noun-finals. However, the
treatment of adj-finals is, of course, of primary interest to us in this
discussion.
Many on this list do not have a copy of Whitney nor access to one, so I’ll
give a brief overview of how he sees cpds. His conception is not so
dissimilar to what we know from Paali, but he uses a different terminology.
W does not use the categories Kh and Tp (kammadhaaraya and tappurisa). He
divides compounds into three types: (1) “copulative” (i.e. dvanda, which he
considers connected by the conjunction ‘and’ or ‘or’); (2) “determinative
compounds.” This class includes what Paali students know of as Kh and Tp.
They are cpds “of which the former member is syntactically dependent on the
latter, as its determining or qualifying adjunct” (1247d); and (3)
“secondary adjective compounds,” whose largest category is “possessive
compounds.” It is these possessive compounds that go by the term bahuvriihi
(1293c).
In his treatment of determinative cpds, W includes cpds that have both noun
endings and non-noun endings. In paragaph 1247e he gives eight examples,
five with noun and three with non-noun endings. I copy them here in the Skt,
with W’s translations, and with my Paali category equivalents:
NOUN-ending: (a) amitrasenaa, ‘army of enemies’ (gen. Tp); (b) paadodaka,
‘water for the feet’ (dat Tp); (c) mahar.si, ‘great sage’ (Kh); (d)
priyasakhi, ‘dear friend’ (Kh); (e) amitra, ‘enemy’ (Kh).
NON-NOUN ending: (a) aayurdaa, ‘life-giving’ (acc Tp); (b) hastak.rta, ‘made
with the hands’ (ins Tp); (c) suk.rta, ‘well done’ (Kh).
Following his treatment of determinative cpds, W introduces the possessive
cpds (Bh) with this important sentence (1293): “The possessives are
noun-compounds of the preceding class, determinatives, of all its various
subdivisions to which is given an adjective inflection and which take on an
adjective meaning of a kind which is most conventiently and accurately
defined by adding *having* or *possessing* to the meaning of the
determinative.”
The first part of the above sentence is the most important for our purposes.
I now add emphasis, quotes, and explanation for clarity: “The possessives
[i.e. Bh] are noun-compounds of the preceding class [that is, Kh and Tp],
‘determinatives,’ *of all its various subdivisions*...” “All its various
subdivisions,” means, of course, both noun and non-noun ending cpds, such as
we find with Kh and Tp. We must always bear this rule carefully in mind.
CONCLUSION: According to Whitney, bahubbiihis contain *all* the various
kinds of Kh and Tp cpds-- his ‘determinative’ class, including both
underlying noun-ending cpds *and* adj-ending cpds.
This is the second argument that shows that in Sanskrit Bh compounds
(“possessives”) can end in adjectives. The first argument (above) is that
they MUST end in adjectives, or else in nouns now *functioning* as
adjectives. This question must be considered settled, it seems to me.
--------
The remainder of this post has reference to certain sentences in W’s
presentation which can admittedly mislead if one doesn’t keep the above
‘Conclusion’ in mind, that is: Bh’s include both underlying noun-ending cpds
*and* adj-ending cpds.
For example, the important first sentence regarding W’s third class of cpds
(which includes bahubbiihis) reads equivocally, even to an English-speaker,
because of how it is worded. W writes (1247g):
“Secondary adjective compounds, the value of which is not given by a simple
resolution into their component parts, but which, though having as final
member a noun, are themselves adjectives.” [This is not a complete sentence
in the text.]
Here, it indeed may appear that W is saying that “secondary adj. cpds” have
“as final member a noun.” The crux lies in the words “though having.”
‘Though’ is a subordinating conjunction in English, and it has three
different meanings (I refer now to the New World Dict., College Ed., 1700
pages): (1) “in spite of the fact that.” I.e.: “Secondary adj. cpds... in
spite of the fact that they have as final member a noun, are themselves
adjectives.” This is the erroneous reading and would mean that all 2ary adj
cpds end with a noun. We know this is false, however, *because it is not
consistent with the Conclusion above.*
(2) The second meaning of “though” as ‘however’ also doesn’t work:
“Secondary adj. cpds... which, however, have as final member a noun, are
themselves adjectives.” This is also an erroneous reading, *because it is
not consistent with the Conclusion above.*
(3) The third meaning of “though” as ‘even if,’ ‘supposing that,’ does
work: “Secondary adjective compounds, the value of which is not given by a
simple resolution into their component parts, but which, even if having as
final member a noun, are themselves adjectives.” This is the correct
reading, *because it is consistent with the Conclusion above.*
So this statement is not a proviso against adj-ending Bh’s, as it may at
first seem.
There are other places in the chapter which are also ambigious. The first
sentence introducing 2ary adj. cpds in par. 1292, if read with an incorrect
reference, also seems to demand that such a cpd must end in a noun (emphasis
added):
“Secondary adjective compounds. A compound *having a noun as its final
member* very often wins secondarily the value of an adjective, being
inflected in the three genders to agree with the noun which it qualifies,
and used in all the constructions of an adjective.”
This statement does not speak for all 2ary adj. cpds, as might appear at
first. It only refers to those compounds that have “a noun as its final
member.” This is only a portion of all possible cpds. It can be confusing
largely because it comes immediately after the heading “Secondary adjective
compounds.”