search

sanskrit-lexicon / WIL

Wilson Sanskrit-English Dictionary. Based on Cologne digitization. Work pertaining to corrections
2 stars 2 forks source link

Wilson roots correspondence to MW roots #2

Open funderburkjim opened 9 years ago

funderburkjim commented 9 years ago

This repository directory provides the first step in developing a correspondence between Wilson roots and MW roots. The readme file therein describes the work. This correspondence was begun by Ahlborn/Scharf in 2011 (corr6-roots.txt).

This issue (a) announces the work done so far and (b) will be used to detail further work done to complete the correspondence.

funderburkjim commented 9 years ago

Refer to Note2 in readme

kurd in MW should be marked as a root. An alternate to kUrd.

dfmp in MW should be marked as a root. Also, a new record for dfmP is added. Both are apparently related to root dfp.

These changes should solve the two problems mentioned in Note 2, when things are reconstructed.

funderburkjim commented 9 years ago

Refer to Note 3 in readme

In these cases, the current program (wil_mw.py) associated a different MW root than did the prior work of Ahlborn and Scharf. Can we determine which should be used?

wilson wil_mw.py corr6-roots.txt solution
jala jala jal both
Dana Dana Dan Dan (see note)
Pulla Pulla Pull both
laGa laGaya laNG laNG
skaBa skaB skamB both

MW states skaB and skamB to be alternates.

laGaya appears related to laGu (based on sense), hence unrelated to laG.

Dana in MW is not a headword - wil_mw.py needs to be altered to ignore non-headwords in verb_step0a.txt.

For the three cases jala, Pulla, skaBa , the association of wil_mw.py is accurate, but so is that of corr6-roots.txt. Should wil_mw.py be reconceptualized to permit a given Wilson root to correspond to multiple MW roots?

gasyoun commented 9 years ago

Should wil_mw.py be reconceptualized to permit a given Wilson root to correspond to multiple MW roots? I guess so. That is what I've found in constructing my dhatupatha correspondences in 2006-2014.

drdhaval2785 commented 9 years ago

@gasyoun is the most competent person among us to say something on dhAtus. His suggestion must be good.

funderburkjim commented 9 years ago

OK. I will organize step2 to permit a given Wilson root to correspond to multiple MW roots, and also to permit multiple Wilson roots to correspond to one MW root. (For instance, the roots Wilson roots aba, amba correspond to MW root amb.)

gasyoun commented 9 years ago

The fact is that roots with same spelling in different sources might not have no correspondence at all with totally different meanings. But's that's rather rare, still an issue. One solution would be to add a hypothetical homonym.

funderburkjim commented 9 years ago

Refer to Note 4 in readme, which are cases where 2 or more Wilson roots have been associated with 1 MW root. There are 13 cases.

Most of these are difficult. Underlying data from Wilson and MW for the cases will be shown - maybe someone can see a resolution more clearly than can I at the moment.

funderburkjim commented 9 years ago

In 3 cases, I think that 2 Wilson roots are properly associated with 1 MW root:

01 aba,amba <-> amb OK

aba:  r. 1st cl. with an indicatory i, abi (aMbate) To sound.
amba: r. 1st cl. (aMbati) To move, to go.
amb : cl.1 P. ambati, to go L.  ; cl.1 A1. ambate, to sound L.

06 carca, carcca <-> carc OK

carca: r. 1st and 6th cls. (carcati) 1 To speak or say. 2 To abuse, to reprove, to censure or 
condemn. 3 To injure or hurt. 4 To inquire.
 r. 10th cl. (cartrayati) To read over, to peruse carefully, to 
study; also read carccaM.
carcca:  r. 1st and 10th cls. (carccate carccayate) 1 To speak, to tell. 2 To reprove. 3 To study: see 
carca.
carc: cl.1. °cati, to abuse, censure, menace Dhātup.  xvii, 67  ; to injure, xxviii, 17 : cl.10. °cayati, to 
repeat a word (in reciting the veda, esp. while adding iti) RPrāt.  xv, 10 and 12 to talk over, discuss 
Hcar.  vii  ; (also A1. Vop. ) to study Dhātup.  xxxiii, 38.

NOTE typos under 'carca': cartrayati -> carcayati AND carccaM -> carcca

08 diBa, dimBa <-> dimB ok

diBa:  (i) diBi r. 10th cl. (dimBayati) To order, to direct.
dimBa: r. 10th cl. (dimBati) 1 To command, to direct. 2 To collect.
dimB: cl.10 A1. dimBayate id. P. °ti, to order, direct id.

NOTE: under dimBa, dimBati is not consistent with 10th cl. Should this be corrected?

funderburkjim commented 9 years ago

3 cases will be resolved after Wilson spelling errors are corrected:

07 cuwwa, cuwwa <-> cuww 

cuwwa: r. 10th cl. (cuwwayati) 1 To become small, low or shallow, as a river, &c. 2 To collect.
cuwwa:  r. 1st cl. (cuwwati) 1 To hint one's meaning. 2 To dally, to wanton, to coquet, &c. 3 To act.   THIS IS MIS-SPELLED: cuqqa 
cuww: cl.10. wwayati, to become small, xxxii, 24  ; cf. √puww.

NOTE: The 2nd 'cuwwa' is misspelled (a typo) in Wilson, it should be cuqqa.

12 stf, stf <-> stf 

stf: (stf,) r. 5th cl. (stfRoti stfRute) To cover, to clothe, to spread on or over, to strew. With vi prefixed.
1 To spread widely, to extend.
2 To be diffuse or prolix.

stf: (stf) r. 9th cl. (stfRAti stfRIte) To clothe, to cover, to spread on or over: see stf.   NOTE:  
stf: (or stF). cl.5.9. P. A1. (Dhātup.  xxvii, 6 ; xxxi, 14) stfRoti, stfRute or stfRAti, stfRIte 
to spread, spread out or about, strew, scatter

NOTE: The 2nd 'stf' in Wilson should be corrected (typo) to 'stF'.

13 sPurjja, sPurjja <-> sPurj 

sPurjja: r. 1st cl. (sPurjjati) To thunder: more properly sPUrjja.

sPurjja: (o A) osPUrjjA r. 1st cl. (sPUrjjati) To thunder. 

sPurj:  sPurC » √sPUrC,sPUrj.

NOTE: 2nd sPurjja in Wilson is typo, should be sPUrjja.

funderburkjim commented 9 years ago

02 kaka, kakka <-> kak 

kaka:  r. 1st cl. (kakate)
1 To be proud.
3 To be unsteady.
4 To be thirsty. (i) kaki (kaMkate) To go.
kakka:  r. 1st cl. (kakkati) To laugh.

kak : cl.1 A1. kakate, cakake, kakitA, &c , to be unsteady  ; to be proud  ; to wish Dhātup.  iv, 16.

MW has kakk, which was not marked as verb:

v.l. for kaK q.v.

I've marked this as a verb. With this change, this case should be resolved: kaka <-> kak, and kakka <-> kakk

funderburkjim commented 9 years ago

03 kakKa, kakKa, kaKa <-> kaK 

kakKa: r. 1st cl. (kakKati) To laugh. to laugh, at or deride: see kaKa.
 (p. 177-2, after kaka, and out of alph. order)
kakKa: (e) r. 1st cl. (kakKati) To laugh, to laugh at, to deride.

kaKa r. r. 1st cl. (e) kaKe (kaKati) To laugh, to laugh at or deride; also kakKa and KakKa q. v.

kaK: cl.1 P. kaKati, cakAKa, kaKitA, &c , to laugh, laugh at or deride: Caus. kaKayati, 
to cause to laugh Dhātup.  v, 6 ; xix, 22 ;
funderburkjim commented 9 years ago

04 Karba, Karba <-> Karb ?

Karba:  r. 10th cl. (Karbvati) To go or move.
        Out of order (after  karu). 
        The text may actually be 'KarGa' (Yates has KarGa), which would be
        proper alphabetical order.  Could this be an error that Wilson forgot
        to remove and that Yates copied?

Karba: r. 1st cl. (KarbaMti) To go, to move to or towards.
    KarbaMti is error, should be Karbati.

Karb : cl.1 P. °bati, to go or move. Dhātup.  xi, 27 ;

The solution here may be to change spelling of first Wilson Karba to KarGa. Perhaps this record should also be marked (in some way) as a Wilson error.

funderburkjim commented 9 years ago

05 Gura, Gura <-> Gur ?

Gura: r. 6th cl. (Gurati)
1 To be frightful or terrible.
2 To sound.
Gura: (I) GurI r. 4th cl. (Guryyate)    
1 To be ignorant.
2 To inquire.

Gur: cl.6 P. °rati (aor. aGorIt ; pf. 3. pl. juGurur ; A1. juGure), to cry frightfully, frighten with cries Bhaṭṭ.  xiv f. (cf. Gora.)

The second Wilson Gura seems to be a different root entirely; and probably should not be shown to correspond to MW Gur; I'm not sure what to do with it.

funderburkjim commented 9 years ago

09 baDa, banDa <-> banD ?

baDa: r. 1st cl. (bIBatsate)
1 To bind or tie.
2 To hate, to despise, to loath.
3 To suffer change or affection of the mind. r. 1st and 10th cls. (baDati bADayati)
1 To tie, to bind.
2 To kill.

banDa: r. 9th cl. (baDnAti) r. 10th cl. (banDayati) To bind or tie; with AN prefixed, To loosen, to let loose. 

banD:cl.9 P. (Dhātup.  xxxi, 37) baDnAti (rarely A1. baDnIte ; cl.1 P. A1. banDati, °te MBh.  ; cl.4 P. baDyati Hariv.  [and much more]), to bind, tie, fix, fasten, chain, fetter [and much more]

Wilson's banDa clearly corresponds to MW's banD.

However, Wilson's baDa should probably correspond to MW's baD (which should be marked as a root). Here is data from MW's baD and vaD

baD: baD, baDya, even in Vedic texts sometimes = vaD,vaDya.

vaD: (also written baD ; cf. √ bAD ; properly only used in the aor. and Prec. tenses avaDIt and 
°Dizwa ; vaDyAt and vaDizIzwa Pāṇ.  2-4, 42 &c ; vi, 62 ; the other tenses being supplied by √ han ;
 cf. Dhātup.  xxiv, 2 ; but in Ved. and ep. poetry also pres. vaDati ; Pot. vaDet ; fut. vaDizyati, °te ; 
other Ved. forms are aor. avaDIm, vahIm Subj. vaDizaH ; baDIH TĀr.  ; Prec. baDyAsam, °suH AV. ), 
to strike, slay, kill, murder, defeat, destroy RV.  &c &c : Pass. vaDyate, °ti (aor. avaDi), to be slain or 
killed MBh.  Kāv.  &c : Caus. vaDayati, to kill, slay MBh.
funderburkjim commented 9 years ago

10 riKa, riKa <-> riK ?

riKa: r. 1st cl. (riKati) To go, to move.

riKa: (i) riKi r. 1st cl. (riMKati) To go, to move.

riK: cl.1 P. reKati, to go, move Dhātup.  v, 33 Vop.  (cf. √ riNK)  ; 
cl.6 P. riKati, to scratch, scrape (cf. A- √riK and √ liK).

Wilson's first riKa should correspond to MW's riK.

Wilson's second riKa should correspond to MW's riNK.

riNK: (cf. prec.) cl.1 P. (Dhātup.  v, 33 Vop. ) riNKati to go, move, crawl (said of
 young children) BhP.   ;  to go or advance slowly Cat.
funderburkjim commented 9 years ago

11 skada, skanda <-> skand ?

skada: (i) skadi r. 1st. cl. (skandate)
1 To leap, to jump.
2 To inundate.
2 To lift up.
3 To shed or spill.

skanda: (ira) skandira r. 1st cl. (skandati)
1 To go, to move.
2 To be dry or arid.
3 To effuse or emit. With ava prefixed, To assail, to assault.

skand: (cf. skanD and skund) cl.1 P. (Dhātup.  xxiii, 10) skandati 
to leap, jump, hop, dart, spring, spurt out, be spilt or effused (esp. said of semen) 
(A1.) to emit seminal fluid VP.   ; 
to leap upon, cover (said of animals) TBr.  ṠBr.   ; 
to drop, fall down, perish, be lost 
Caus. skandayati: to pour out, effuse, shed, spill, emit (esp. seminal fluid)

I don't see any errors. Maybe both Wilson roots should indeed correspond to the one MW root .

funderburkjim commented 9 years ago

The preceding comments cover all the 13 cases of Note 4 in readme.

I hope comments from others will help resolve the uncertain cases.

gasyoun commented 8 years ago

corr6-roots.txt is the source and the final file, or the verified results are stored in a different place?

funderburkjim commented 8 years ago

@gasyoun It will take me some time to refresh my memory on how things work here.

When I do this will depend somewhat on whether you are planning to work on this now.

Are you going to work on this now? If so, I'll look at it soon.

gasyoun commented 8 years ago

Yes, but I have only two weeks. After that a Russian book that is a translation of Wilson will be published and it will be too late. I want to implement a part of your research in the preface to that book.

funderburkjim commented 8 years ago

@gasyoun I've spent the day revisiting this work. Here's what I've learned thus far.

To answer your question, corr6-roots.txt does appear to be the end result of the work done in 2011. This file appears in at least four places, and the contents of all are identical:

It appears that this correspondence (1283 instances) between Wilson roots and MW roots is as far as things were taken in the 2011 work.

The wilmwroots/step2/ directory seems to be a revisit by me of this subject. The file wilmwroots/step2/wil_mw.txt appears to be an extension of corr6-roots.txt, but it contains 1721 instances.

I haven't yet re-examined this 2015 wilmwroots/step2/ work in detail, but I suspect it is a material extension of the 2011 work; thus my hunch is that the current best correspondence is that wilmwroots/step2/wil_mw.txt file.

Maybe in the next few days I can regenerate the wil_mw.txt work with current versions of the mw and wilson digitizations. I'd also like to firm up how the wilmwroots work extended the wiltab2011 work (for instance, why are there 438 (= 1721 - 1283) extra correspondences in the wilmwroots work. ?)

funderburkjim commented 8 years ago

One interesting file is wiltab2011/step2roots/badkeys/wilson_dproots.txt which contains 163 complex DAtupAWa forms from Wilson. It's probably worthwhile examining the 'key1' values in WIL for these, to be sure they look right - maybe this was already done in wilmwroots/step2 .

gasyoun commented 8 years ago

I'd also like to firm up how the wilmwroots work extended the wiltab2011 work (for instance, why are there 438 (= 1721 - 1283) extra correspondences in the wilmwroots work. ?)

Sure interesting. Hope it can lead us to better files, thanks.

Shalu411 commented 8 years ago

Namaste So- Is the next work on "Better files" or what? May I know what is waiting here exactly?

funderburkjim commented 8 years ago

@Shalu411 As I recall, you're interested in roots. Right?

If so, then it would be excellent to have you check this work.

So, let me revise this work (I'll do that early next week). Then, when that revision is done, I'll turn it over to you to 'check'. You may need me to write some kind of display programs to help you with the checking.

Let me know if this sounds like something you'ld like to work on.