funderburkjim/elispsanskrit

SanskritVerb-Pysan comparison, present tense NOT class 10, part 3

Opened this issue · 20 comments

funderburkjim commented

The verbs mentioned in the cases below show a difference in the present tense conjugations when SanskritVerb is compared to Pysanskrit.

In some cases, I think that Pysanskrit is correct, and that SanskritVerb should be changed.

In a few other cases, I have no opinion as to which is correct.

Maybe Dhaval will examine these cases again.

funderburkjim commented

kfRv, and Dinv are class 1, but inflect like class 5.

This reminds me of Sru, where Dhaval has found a rule that, in effect, justifies the treatment
of Sru as a class 5 root spelled as 'Sf'.

; kfRv.  Not in Kale. Maybe = kfv in MW (?)
Although called class 1, Sanskritverb inflects this like a class 5 kf (there
  is in fact a 5P kf in verbdata, and it has the same conjugation
TODO sq:kfRv:kfRoti:kfRvati:3s # kfRv:hiMsAkaraRayoSca:01:0682:pa:sew 
  pysan:kfRv pre 1P:[kfRvati kfRvataH kfRvanti kfRvasi kfRvaTaH kfRvaTa kfRvAmi kfRvAvaH kfRvAmaH]
sanverb:kfRv pre 1P:[kfRoti kfRutaH kfRvanti kfRozi kfRuTaH kfRuTa kfRomi (kfRuvaH kfRvaH) (kfRumaH kfRmaH)]
kf pre 5P:[kfRoti kfRutaH kfRvanti kfRozi kfRuTaH kfRuTa kfRomi (kfRuvaH kfRvaH) (kfRumaH kfRmaH)]


; Dinv. Kale 1P prIRane to delight, to please; Dinoti
; MW  cf. Di#3 cl.5 Dinoti
Although call class 1, Sanskritverb inflection is like a class 5 'Di' (which
  doesn't exist as a root in verbdata)  However, in MW Di#3 has alternate
  spelling of 'Dinv', and is class 5P.  And the entry for Dinv is to see
  Di#3.
TODO s:Dinv:Dinoti:Dinvati:3s # Dinv:prIRanArTaH:01:0677:pa:sew
  pysan:Dinv pre 1P:[Dinvati DinvataH Dinvanti Dinvasi DinvaTaH DinvaTa DinvAmi DinvAvaH DinvAmaH]
sanverb:Dinv pre 1P:[Dinoti DinutaH Dinvanti Dinozi DinuTaH DinuTa Dinomi (DinuvaH DinvaH) (DinumaH DinmaH)]
funderburkjim commented

gA class 1A,

No definite opinion as to which is right.

; gA Kale 1A to go (gAte), 3P to praise (jigAti)
;    MW 2P,A to go (gAti, gAte); 3P (jigAti)
; SanskritVerb has gAte for 1A 3d,3p and 2d ?
TODO gA:gAte:gEte:3d # gA:gatO:01:1101:A:aniw
     3d (sanverb) gAte != gEte (pysan)
     3p (sanverb) gAte != gAnte (pysan)
     2d (sanverb) gATe != gETe (pysan)
  pysan:gA pre 1A:[gAte gEte gAnte gAse gETe gADve gE gAvahe gAmahe]
sanverb:gA pre 1A:[gAte gAte gAte gAse gATe gADve gE gAvahe gAmahe]
funderburkjim commented

jan 3P: in 3p, SanskritVerb to change jajYati to jajanati

Kale Dhatukosha does not have a class 3 form for jan.

 pysan:jan pre 3P:[jajanti jajantaH jajanati jajaMsi jajanTaH jajanTa jajanmi jajanvaH jajanmaH]
sanverb:jan pre 3P:[jajanti jajantaH jajYati jajaMsi jajanTaH jajanTa jajanmi jajanvaH jajanmaH]
funderburkjim commented

jan 4A: Sanskritverb to change 3d and 2d

 Sanskritverb to change 3d from janyete to jAyete
 Sanskritverb to change 2d from janyeTe to jAyeTe
  pysan:jan pre 4A:[jAyate jAyete jAyante jAyase jAyeTe jAyaDve jAye jAyAvahe jAyAmahe]
sanverb:jan pre 4A:[jAyate janyete jAyante jAyase janyeTe jAyaDve jAye jAyAvahe jAyAmahe]
funderburkjim commented

IS 2A. SanskritVerb to change 2P

Sanskritverb to change 2p from IqQve to ISiDve

Kale section 427 confirms this change; it is similar to an irregular form IqiDve for 2p if Iq 2A, which SanskritVerb already has.

  pysan:IS pre 2A:[Izwe ISAte ISate ISize ISATe ISiDve ISe ISvahe ISmahe]
sanverb:IS pre 2A:[Izwe ISAte ISate ISize ISATe IqQve ISe ISvahe ISmahe]
funderburkjim commented

saMst 2P, kit 3P, kft 7P:

For these roots, SanskritVerb to change 't' to 'd' in 1s and 1p

 saMst 2P (1s, 1p) saMstmi, saMstmaH  v. saMsdmi, saMsdmaH
   kit 3P (1s, 1p) ciketmi, cikitmaH v. cikedmi, cikidmaH
   kft 7P (1s, 1p) kfRatmi, kfntmaH v. kfRadmi, kfndmaH
   This change would be consistent with the general sandhi rule whereby
   a final hard consonant (e.g. 't' in 'ciket') becomes soft (e.g. 'd') 
   when combined with a soft consonant (e.g. 'm' in termination 'mi').

  pysan:saMst pre 2P:[(saMstti saMsti) (saMsttaH saMstaH) saMstanti saMstsi (saMstTaH saMsTaH) (saMstTa saMsTa) saMsdmi saMstvaH saMsdmaH]
sanverb:saMst pre 2P:[(saMstti saMsti) (saMsttaH saMstaH) saMstanti saMstsi (saMstTaH saMsTaH) (saMstTa saMsTa) saMstmi saMstvaH saMstmaH]

  pysan:kit pre 3P:[ciketti cikittaH cikitati ciketsi cikitTaH cikitTa cikedmi cikitvaH cikidmaH]
sanverb:kit pre 3P:[ciketti cikittaH cikitati ciketsi cikitTaH cikitTa ciketmi cikitvaH cikitmaH]

  pysan:kft pre 7P:[kfRatti (kfntaH kfnttaH) kfntanti kfRatsi (kfnTaH kfntTaH) (kfnTa kfntTa) kfRadmi kfntvaH kfndmaH]
sanverb:kft pre 7P:[kfRatti (kfnttaH kfntaH) kfntanti kfRatsi (kfntTaH kfnTaH) (kfntTa kfnTa) kfRatmi kfntvaH kfntmaH]
funderburkjim commented

vid 2P. Sanskritverb to change 2s from vitsi to vetsi

This change confirmed by Kale section 436

     2s (sanverb) vetTa,vitsi != vetTa,vetsi (pysan)
  pysan:vid pre 2P:[(vetti veda) (vittaH vidatuH) (vidanti viduH) (vetsi vetTa) (vitTaH vidaTuH) (vitTa vida) (vedmi veda) (vidvaH vidva) (vidmaH vidma)]
sanverb:vid pre 2P:[(vetti veda) (vittaH vidatuH) (vidanti viduH) (vitsi vetTa) (vitTaH vidaTuH) (vitTa vida) (vedmi veda) (vidvaH vidva) (vidmaH vidma)]
funderburkjim commented

Comment regarding Kale display.

The indexing of this display has been enhanced in two ways to make access to the Dhatukosha and to the conjugations information more extensive.

Now, the first root on each page of the Dhatukosha appendix is listed, and linked to the scan (thanks to Sampada for doing this).

Also, under chapter XII Conjugation of Verbs, there is itemization of all the roots mentioned.

funderburkjim commented

Comment on the current status of the comparison between SanskritVerb and Pysanskrit for present tense conjugations.

There are 2556 present tense conjugation tables implied by the Generated Forms of SanskritVerb.

When the Pysanskrit conjugations of these cases are computed and compared to the SanskritVerb conjugations, 2387 of the conjugation tables are identical, and 169 have one or more differences.

This comment does NOT (yet) take into account the replies of Dhaval in #44, and is based on the
version of SanskritVerb as of its version 209b35b6f2376c2f310b8f6ecbbaa3a185b39efe of 4 days ago.
.

drdhaval2785 commented

kfRv, Dinv

There is a separate rule धिन्विकृण्व्योर च which converts 'v'-'a' and
mandates 'u' suffix.

http://www.sanskritworld.in/sanskrittool/SanskritVerb/Data/allsutrani/3.1.80.htm

drdhaval2785 commented

gA

All three persons have the same form gAte. (गाते) त्रिष्वपि वचनेषु समानं

http://www.sanskritworld.in/sanskrittool/SanskritVerb/Data/dhaatupaatha/files-6-9-2014/mA606.html

drdhaval2785 commented

Also गासे गाथे गाध्वे is written in mADavIya.

drdhaval2785 commented

jan

There is a separate rule which will remove 'a' from in between in case of
'jan'. So form will be जज्ञति. But I am not sure whether this applies in the verb under consideration. I thought over it. Not applicable. SanskritVerb needed correction. Correction made.

गमहनजनखनघसां लोपः क्ङित्यनङि॥ ६।४।९८

Also see
http://www.sanskritworld.in/sanskrittool/SanskritVerb/Data/allsutrani/6.4.98.htm

drdhaval2785 commented
drdhaval2785 commented
drdhaval2785 commented
drdhaval2785 commented
drdhaval2785 commented

All corrections made in SanskritVerb.
None was an error in database. They were errors in algorithm at panini.php.
So $verbdata is unaltered.
Resulting forms may be altered.
panini.php updated on SanskritVerb.

funderburkjim commented

@drdhaval2785 re gA 1A.

Will change pysanskrit to be gAte in gA 1A for 2d, 3d, 3p.

I followed your link, but what is the citation reference for this ?

funderburkjim commented

Added a 'sanskritverb_flag' parameter to pysanskrit, and set it to True. This will have the effect of
doing certain computations for the sake of compatibility with SanskritVerb.

The first use of this is for the class 1 roots kfRv, Dinv, and Sru. In computing the conjugation for these,
Pysanskrit changes the root to kf, Di, and Sf respectively and changes the class to 5.

The result is now that the present tense conjugations of these three roots by pysanskrit now agrees with
SanskritVerb.