a bit of help?

[Madhav Deshpande]

Jonathan,
	Now that I have had a chance to look up the MB on P. 3.2.1, here 
are my comments.  I am not sure it dierctly assists with the 
interpretation of "yogAcAra", but has a lot of indirectly interesting 
implications.  Theoretically, one can have the affix aN by this rule to 
derive 'AcAra' in 'yogAcAra' with the agentive meaning:  yogam Acarati 
iti yogAcAraH.  However, with such a derivation, by the standard rules of 
accents, the word would be accented on the last syllable, i.e. the affix 
aN.  Secondly, the word -AcAra thus derived with the affix aN, P.4.1.15 
predicts its feminine form with 'I', i.e. yogAcArI', rather than as 
'yogAcArA'.  KAtyAyana's vArttika 7 (p. 95) brings up issues related to 
both accents and feminine form.  Patanjali's example kalyANAcAra 
illustrates the portion -Acari in the vArttika.  KAtyAyana proposes two 
things:  with the root A+car, we should have the affix Na, rather than 
aN.  Secondly, he proposes that the first member of this compound retains 
its original accent, a pattern of behavior similar to a BahuvrIhi.  Why 
the affix Na, instead of aN?  With aN, P.4.1.15 predicts a feminine form 
in 'I', while with Na, one gets the default feminine form in 'A'.  This 
is explained by the MB in line 24: kimartham idam ucyate.  
pUrvapadaprakRti- svaratvaM ca vakzyAmi IkAraS ca mA bhUd iti.  The 
discussion that follows goes back and forth on whether one can have a 
BahuvrIhi in examples like mAMsabhakzaH.  It shows that in some examples 
Patanjali is willing to admit the possibility of a BahuvrIhi, but not in 
others.  The main point seems to be that with a BahuvrIhi, one can get 
optional forms like mAMsakAmaH/mAMsakAmakaH (p. 96, lines 3-4).  The 
discussion does not say anything about kalyANAcAraH, and certainly 
nothing about yogAcAraH.  
	One thing we cannot check is the accent of yogAcAraH in its 
actual attestations, because these are all late.  Perhaps what one can 
check is the feminine form, if such is attested, and if a form like 
yogAcAraka is also used.  The latter especially would support a BahuvrIhi 
interpretation.
	The BahuvrIhi interpretation for a late expression is certainly 
not out of the question.  In fact, the huge Prakrit 
dictionary/encyclopedia AbhidhAnarAjendra on the word 'jogAcAra' suggests 
a BahuvrIhi interpretation:  yogena AcAraH yasya:  
yoga+A+car+ghaN(palatal).  This affix derives an action noun in Panini, 
rather than an agent noun, as with the affixes aN, or KAtyAyana's Na.
	I hope this gives you some food for thought.
	Roth-Boehtlingk's dictionary, however, notes the variant reading 
yogAcArya for yogAcAra.  They don't accept it, but you may look into it 
as well.  
	I am also posting this note on Indology.  Perhaps, someone else 
may have some further suggestions.

	All the best,
			Madhav

On Tue, 13 Feb 1996 jonathan.silk at wmich.edu wrote:

> Dear Madhav,
> 
> I hope you are well.
> 
> I have a Patanjali question.  It was suggested by H. Nakamura (if I
> understand him right) that the term yogaacaara, which occurs in Buddhist
> texts in the sense of "one who practices yoga," is to be understood according
> to P III.2.1 (karma.ny-a.N) -- and not as a bahuvriihi (again, if I
> understand him correctly).  I took a look at Mahaabhaa.sya (volume 2, p.95
> line 24 and following in Kielhorn seems to me the relevant discussion), but I
> had some trouble understanding the text.  Would you mind explaining to me
> what is going on here, and whether it is possible to understand yogaacaara
> according to this rule?  I have always assumed that the term must be a
> bahuvriihi (it sometimes occurs with bhik.su, in which case I think it is
> clearly an adjective).
> Thank you in advance for your help!
> 
> best regards, jonathan
> 
>