Delocutives

[George Thompson]

In response to Jacob's recent post:
>
[snip]
>
>It seems in Fr. any utterance can be thought of as an invariable noun when
>used meta-linguistically as the name of the action of uttering it:
>"Il commence a m'enerver avec ses je le ferai plus tard" (Doesn't that work
>in English: "He's getting on my nerves with his I'll do it later"?)

A very good point, and an interesting one. And most of these 'invariable
nouns' will be nonce-formations which native speakers will not recognize as
legitimate 'words' [see Jonathan's scepticism, for example].  But my
Sanskrit examples have been recognized as such within their own
speech-communities. Consider, just for one example, the *long* history of
the term 'ahaMkAra' in classical Indian philosophy. This suggests that
these Skt. forms have more cultural significance than do the typical
nonce-formations of English or French, which serve the moment and then are
discarded.
>>
[snip]
>
>_If_ in Sanskrit you can, like in French, consider the 1st member of those
>kr compounds to stand for the name of the action of uttering it (thus one
>could analyze "tvam" in tvamkR as "the _action_ of uttering tvam") then
>you don't need any special meaning of kR, kR in its usual meaning of "perform"
>seems to work doesn't it? So, is it sensible to analyze it this (my) way?

Another good point. I agree that, by itself, this delocutive use of the
verb kR- might not suggest a secondary meaning "to say". But there are
other factors. Of course among the grammarians the form -kAra is used to
mark a phoneme, so that 'akAra' = 'the phoneme "a"', 'kakAra' = 'the
phoneme "k"', etc. There are also cases where the verb kR- alternates
paradigmatically with verba dicendi like vad- and vac-: e.g., see
collocations of these verbs with mantra, or in  compounds [mantrakRt vs.
mantravAdin], or with satya [or in compounds: satyakriyA vs.satyavAda]. I'm
not suggesting that these forms are delocutive, of course. Just that kR-
appears in these contexts to look *a lot* like a verbum dicendi. The use of
the verb kR- in delocutive expressions simply strengthens the semantic link
with verba dicendi which is evident elsewhere.

>Can
>one say that in the Vedic example "vaSaT" is the "accusatif" of an invariable
>noun which designates the action of uttering "vaSaT"? Of course things
>would be different in composition with verbs meaning say, utter, cf
>an example mentioned by Dominique, bhoovaadin. What is its precise meaning?
>Also I'd be curious if bhooH can also enter in composition with kR as say
>"bhooHkarooti"? I guess that would depend on "bhooH" having an affective
>content, e.g. the "gurum tvamkarya" of Martin's examples implies a show of
>disrespect or familiarity. Since "bhooH" seems pretty neutral I wouldn't
>expect it to have a transitive kR compound but let me ask anyway.

'bhoH' is the form of address that one should use when addressing superiors
like one's guru. 'bhoH' is not an early term. It is an abbreviated
*vocative* form of bhavAn, 'your honor' [possibly this itself is also an
abbreviated form of bhagavAn]. In fact it is attested in compounds with
both -kAra and -vAdin, but since I am not familiar with the relevant
passages, I cannot say what the precise significance of these terms is. But
if you want to get a better understanding of this term check out Manu DZ
again [2.122ff.], where conventions of direct address are discussed. You
will find there an interesting range of forms [Skt text available from
Indology Archives, thanks to M.YANO and Y.IKARI].

I'll send more later. This post is already too long. Jacob, I'll try to
answer the rest of your questions shortly.

George