search

sanskrit-lexicon / MWinflect

Generate declensions and conjugations based upon words in MW1899 dictionary.
1 stars 0 forks source link

Nominals of 1 stem ending in palatal consonant #45

Open funderburkjim opened 5 years ago

funderburkjim commented 5 years ago

For a general discussion of declension of the nominals ending in consonants and declined with the '1-stem' models, see #41.

From analyze_1cons.txt we extract the following frequency information for nominals ending in a palatal and declined with the '1-stem' models.

funderburkjim commented 5 years ago

General comment regarding palatal-ending 1-stems

In contrast to the 1-stem declensions thus far considered, the declension for palatal-ending 1-stems has one systematic difference: The Nominative singular form generally replaces the ending palatal consonant with k; but there are also numerous common words which use w instead of k.

There are corresponding regular adjustments to other forms depending on which form this 1s form takes, as will be seen in the examples below.

There seems to be no rule as to whether k or w should be used for the 1s ending of a given word.

The procedure used here is to take guidance from published sources (MW, Bucknell, Deshpande, Kale) where this guidance is present, and to otherwise assume that k is the 1s ending; Using k as the default 1s ending is consistent with a comment by Kale on p. 58 under the example of vAc and related words.

funderburkjim commented 5 years ago

Exclusion of adjectives of direction

As discussed in #38, some aYc declined words are spelled by MW with an ending 'Ac' or 'ac', instead of 'AYc' or 'aYc'. The following list extends the list of this comment from issue 38:

These words are declined as if spelled devAYc, etc. and use the models m_aYc, etc.

funderburkjim commented 5 years ago

The case of final consonant c

Recall the general endings for the 1-stem models.

Only 1 example with the 1s ending in w has been found: su-vfSc (Kale, p. 58) k has been explicitly mentioned as the 1s ending for: vAc, fc, tvac, Suc, pac, Sic, sPic, and sruc.
And the assumption is made that k is used for the 1s declension of all c-ending 1stems.

The masculine and feminine declensions are the same.

For neuter declension, same rules as above, but also

Declension of f_1_c vAc

Case 1: vAk vAcO vAcaH Case 2: vAcam vAcO vAcaH Case 3: vAcA vAgByAm vAgBiH Case 4: vAce vAgByAm vAgByaH Case 5: vAcaH vAgByAm vAgByaH Case 6: vAcaH vAcoH vAcAm Case 7: vAci vAcoH vAkzu Case 8: vAk vAcO vAcaH

Declension of n_1_c vAc

Case 1: vAk vAcI vAYci Case 2: vAk vAcI vAYci Case 3: vAcA vAgByAm vAgBiH Case 4: vAce vAgByAm vAgByaH Case 5: vAcaH vAgByAm vAgByaH Case 6: vAcaH vAcoH vAcAm Case 7: vAci vAcoH vAkzu Case 8: vAk vAcI vAYci

funderburkjim commented 5 years ago

cases ending in 'c' preceded by a consonant

That is, citation endings in a compound consonant with last letter c. Only 4 examples are found in MW list: kruYc, su-vfSc, arc, and sUrya-rc

The following is consistent with the (partial) declension provided by Kale:

Declension of m_1_c su-vfSc

Case 1: suvfw suvfScO suvfScaH Case 2: suvfScam suvfScO suvfScaH Case 3: suvfScA suvfqByAm suvfqBiH Case 4: suvfSce suvfqByAm suvfqByaH Case 5: suvfScaH suvfqByAm suvfqByaH Case 6: suvfScaH suvfScoH suvfScAm Case 7: suvfSci suvfScoH suvfwsu Case 8: suvfw suvfScO suvfScaH

Note that the penultimate S is dropped in 1s, 8s, 7p, and the forms with B endings.

The neuter case 1,2,8 forms computed are suvfw suvfScI suvfSci, which are speculative.

Declension of m_1_c kruYc

Case 1: kruN kruYcO kruYcaH Case 2: kruYcam kruYcO kruYcaH Case 3: kruYcA kruNByAm kruNBiH Case 4: kruYce kruNByAm kruNByaH Case 5: kruYcaH kruNByAm kruNByaH Case 6: kruYcaH kruYcoH kruYcAm Case 7: kruYci kruYcoH kruNsu Case 8: kruN kruYcO kruYcaH

This declension is consistent with that of Huet.

Declension of m_1_c arc

Case 1: ark arcO arcaH Case 2: arcam arcO arcaH Case 3: arcA argByAm argBiH Case 4: arce argByAm argByaH Case 5: arcaH argByAm argByaH Case 6: arcaH arcoH arcAm Case 7: arci arcoH arkzu Case 8: ark arcO arcaH

This also agrees with Huet's declension engine. Note that no adjustments for the ending compound consonant 'rc' have occurred (i.e., declension perfectly analogous to 'vAc')

Our declension for neuter case 1,2,8 is: ark arcI arYci, which is speculative. sUryarc is similarly declined by our system.

funderburkjim commented 5 years ago

Case of final consonant j, and Nominative singular w

Here are the cases which are confirmed by one or more of the referenced sources:

Declension of m_1_j rAj

Case 1: rAw rAjO rAjaH Case 2: rAjam rAjO rAjaH Case 3: rAjA rAqByAm rAqBiH Case 4: rAje rAqByAm rAqByaH Case 5: rAjaH rAqByAm rAqByaH Case 6: rAjaH rAjoH rAjAm Case 7: rAji rAjoH rAwsu Case 8: rAw rAjO rAjaH

The neuter cases 1,2,8 of rAj are computed as rAw rAjI rAYji (not confirmed by references).

viSva-rAj

Acc. to Kale Section 99 (p.60) (Pan 6.3.128), the prefix viSva becomes viSvA before the rAw, rAq forms:

Declension of m_1_j viSva-rAj Case 1: viSvArAw viSvarAjO viSvarAjaH Case 2: viSvarAjam viSvarAjO viSvarAjaH Case 3: viSvarAjA viSvArAqByAm viSvArAqBiH Case 4: viSvarAje viSvArAqByAm viSvArAqByaH Case 5: viSvarAjaH viSvArAqByAm viSvArAqByaH Case 6: viSvarAjaH viSvarAjoH viSvarAjAm Case 7: viSvarAji viSvarAjoH viSvArAwsu Case 8: viSvArAw viSvarAjO viSvarAjaH

funderburkjim commented 5 years ago

Declension of m_1_j Bfjj

Here is how the ending compound consonant jj is handled; these forms are consistent with Kale's example on p. 58.

Case 1: Bfw BfjjO BfjjaH Case 2: Bfjjam BfjjO BfjjaH Case 3: BfjjA BfqByAm BfqBiH Case 4: Bfjje BfqByAm BfqByaH Case 5: BfjjaH BfqByAm BfqByaH Case 6: BfjjaH BfjjoH BfjjAm Case 7: Bfjji BfjjoH Bfwsu Case 8: Bfw BfjjO BfjjaH

funderburkjim commented 5 years ago

Case of final consonant j, and Nominative singular k

Here are the cases which are confirmed by one or more of the referenced sources; compounds are also considered to end with k in 1s form.

The many other nominals ending in 'j' are also assumed to have the 1s form ending in k.

Declension of m_1_j vaRij

Case 1: vaRik vaRijO vaRijaH Case 2: vaRijam vaRijO vaRijaH Case 3: vaRijA vaRigByAm vaRigBiH Case 4: vaRije vaRigByAm vaRigByaH Case 5: vaRijaH vaRigByAm vaRigByaH Case 6: vaRijaH vaRijoH vaRijAm Case 7: vaRiji vaRijoH vaRikzu Case 8: vaRik vaRijO vaRijaH

Neuter cases 1,2,8 are computed as: vaRik vaRijI vaRiYji.

funderburkjim commented 5 years ago

Declension of m_1_j yuj

Case 1: yuN yuYjO yuYjaH Case 2: yuYjam yuYjO yujaH Case 3: yujA yugByAm yugBiH Case 4: yuje yugByAm yugByaH Case 5: yujaH yugByAm yugByaH Case 6: yujaH yujoH yujAm Case 7: yuji yujoH yukzu Case 8: yuN yuYjO yuYjaH

The declension agrees with the description of Kale on p. 59; Note differences in cases 1,8 and 2s,2d. Otherwise it follows the pattern of vaRij.

MW specifically mentions aSva-yuj as having (normal) 1s of aSva-yuk. This is based on Pan. 7.1.71. The current algorithms have generalized this to all the compounds ending with -yuj. So

Declension of m_1_j aSva-yuj

Case 1: aSvayuk aSvayujO aSvayujaH Case 2: aSvayujam aSvayujO aSvayujaH Case 3: aSvayujA aSvayugByAm aSvayugBiH Case 4: aSvayuje aSvayugByAm aSvayugByaH Case 5: aSvayujaH aSvayugByAm aSvayugByaH Case 6: aSvayujaH aSvayujoH aSvayujAm Case 7: aSvayuji aSvayujoH aSvayukzu Case 8: aSvayuk aSvayujO aSvayujaH

Similarly a-yuj -> a-yuk, etc. for the 40+ compounds of -yuj.

funderburkjim commented 5 years ago

yaj

MW mentions irregularities in some compounds of yaj, based on Panini sutras:

I don't know how to apply these to all the forms of these and other compounds of yaj, so the current computation declines these like vaRij compounds ending in -yaj (see above.

Revision: Declension of m_1_j avayaj

Kale (p. 61-2, section 103) shows forms for avayAj (long A); and MW has avayAj with reference : see avayaj. The inference is that the declension of MW's avayaj should be the same as that of avayAj in Kale. Using this, and incorporating the irregularities of Kale's avayAj results in:

Declension of m_1_j ava-yaj Case 1: avayAH avayAjO avayAjaH Case 2: avayAjam avayAjO avayAjaH Case 3: avayAjA avayoByAm avayoBiH Case 4: avayAje avayoByAm avayoByaH Case 5: avayAjaH avayoByAm avayoByaH Case 6: avayAjaH avayAjoH avayAjAm Case 7: avayAji avayAjoH avayssu Case 8: avayAH avayAjO avayAjaH

funderburkjim commented 5 years ago

Declension of m_1_j KaYj

KaYj (limp)

Case 1: Kan KaYjO KaYjaH Case 2: KaYjam KaYjO KaYjaH Case 3: KaYjA KanByAm KanBiH Case 4: KaYje KanByAm KanByaH Case 5: KaYjaH KanByAm KanByaH Case 6: KaYjaH KaYjoH KaYjAm Case 7: KaYji KaYjoH Kansu Case 8: Kan KaYjO KaYjaH

MW and PW mention the 1s to be Kan. I assume this 'n' also is used in 8s, before the B endings, and before the 7p su ending. Can anyone confirm ?

funderburkjim commented 5 years ago

Declension of m_1_j Urj

Urj (strength) and bahUrj (much strength) end in compound consonant 'rj'. But I assume that this makes no change in the usual (vaRij-like) model of declension:

Case 1: Urk UrjO UrjaH Case 2: Urjam UrjO UrjaH Case 3: UrjA UrgByAm UrgBiH Case 4: Urje UrgByAm UrgByaH Case 5: UrjaH UrgByAm UrgByaH Case 6: UrjaH UrjoH UrjAm Case 7: Urji UrjoH Urkzu Case 8: Urk UrjO UrjaH

funderburkjim commented 5 years ago

viSva-sjf

PW says the 1s of viSva-sfj ends in k (Nomin. °sṛk und fehlerhaft °sṛṭ). MW also prints this: viśva—sṛj (nom. -sṛk or incorrectly -sṛṭ).

I think (but am unsure due to print quality) that Kale (p. 97) says that the 1s is like rAj, hence ends in w (-sfw).

Currently, the declension computation for viSva-sfj uses the PW/MW assumption, and similarly for the other compounds of sfj: viSva-sfk, viSva-sfjO, ...