Wiktionary talk:About Sanskrit

Why don't transliterations get wikified & get their own entries? Also, if that is the decision, then this page sould be a little clearer that they must not be wikified. Lastly, it isn't clear to me (not speaking any language that uses Sanskrit script) that transliterations are transliterations, in the current format recommended...is this something that can be changed? --Connel MacKenzie 22:51, 12 January 2007 (UTC)Reply

there are lots of Sanskrit dictionaries in IAST, so I don't see why Wiktionary should discriminate against IAST (the de-facto standard in western editions). 83.78.31.94 19:43, 29 January 2007 (UTC)Reply

If you want to edit a policy or draft policy page on the wikt, at least create a user account and log in. Also: deletion of content is seriously discouraged. Reverted. Robert Ullmann 22:28, 29 January 2007 (UTC)Reply

it is completely bizarre to keep Sanskrit entries in Devanagari in en-wiktionary. Devanagari entries can be a nice-to-have extra, but they are useless for

• marking accentuation using diacritics
• marking morphology using hyphens

IAST should clearly be recommended. --Dbachmann 16:25, 8 January 2010 (UTC)Reply

Accents and hyphens are supposed to be used in transliterations (the usual tr= parameter to {{infl}}, {{t}} etc.). We ignore them in Devanagari (which does have Unicode support for Vedic accents) because they're not a part of "regular orthography". (we do so for many other languages, e.g. with vowel lengths macrons for Latin which are displayed only in the inflection line). I understand your case for Latin script and the problem of associating Sanskrit with Devanagari implying Hindu connections, but I really wonder is it worth all the trouble, with the Western scholarship using prevalently Devanagari for Sanskrit for more than a century... Unlike e.t. Tocharian which we add in Latin transcription because its script is note available in Unicode (and it won't for a long time to come), the problem of Sanskrit is that there are so many scripts to choose from. --Ivan Štambuk 19:16, 8 January 2010 (UTC)Reply

The point is that Hindi Wiktionary will naturally use Devanagari for Sanskrit entries, just like the Bengali or Tamil Wiktionaries will use Bengali or Tamil scripts for the same Sanskrit words. There is no reason why English Wiktionary shouldn't use the most widespread script used for Sanskrit in English speaking contexts, which is IAST. I do not think that Western scholarship uses prevalently Devanagari for Sanskrit. If it uses any Indic script, that script will be Devanagari, but IAST used as least as often, already because of its objective advantages such as hyphenation, accentuation etc. --Dbachmann 11:05, 27 August 2010 (UTC)Reply

What is the consensus about visarga ? Do we include it in entry names, or should we show it in the headword, or is this something for an alternate-spelling entry? I realize it alternates with म् स् depending on what follows, but if people are going to see it in texts, we should at least acknowledge that it exists in a great many terms. Chuck Entz (talk)

It's anusvara (ं) that alternates with म्. Visarga alternates with स् and sometimes र्. But it's wider than just visarga or just visarga and anusvara: the question is, what do we do with sandhi forms in general? They aren't exactly alternative spellings. Maybe we need a template {{sandhi form of}}, the way the mutated forms of the Celtic languages are marked with {{mutation of}}. —Angr 18:07, 18 June 2012 (UTC)Reply

Oops! I thought I put स्, but I copied-and-pasted it to save time and grabbed the wrong character.

My reason for asking was that a new user did an inappropriate move to the visarga form (the page had entries in Hindi and a couple of other languages, so reverting was a no-brainer), but when I went to point out our standards for which forms to use as the lemma, I couldn't find anything.

We should at the very least specify the preferred form (even if it's the root without endings) for the lemma of each inflected POS.

It wouldn't hurt to have sandhi tables here- or at least links to them on WP or elsewhere. It would be nice if we could have enough information accessible from WT:ASA so someone could start with a random word from a text and figure out the lemma form in order to look it up. At least we should show the main sandhi variants for the usual alternating final consonants like स्/(ः) and म्/(ं). Chuck Entz (talk) 02:10, 19 June 2012 (UTC)Reply

For nouns, I think the lemma form should be (and as far as I know this is already the general practice here) the stem without any ending, same as is given in well-respected reference works like Whitney and Monier-Williams. Thus, the main entry for the word meaning "horse" should be at ]. But, just as we have "form-of" entries for each non-nominative case form in Latin (], ], ], ], , ]), we should have "form-of" entries for all the cases in Sanskrit. So while the main entry—the one containing the etymology, the gloss, synonyms, categories, and so forth—is at ], we should also have ] with the gloss "nominative singular of अश्व (aśva)". And for s-stems, of course, the lemma form itself ends in -स्, so the main entry of the word for "race" would be ]. Then sandhi variants like ] and ] would be glossed as "sandhi variant of जनस् (janas)" or the like, and ] and ] would also be glossed as "sandhi variant of अश्वस् (aśvas)" even though that's only a "form-of" entry itself. Does that make sense? I know what I mean, but I'm not sure I'm explaining it clearly. —Angr 20:44, 19 June 2012 (UTC)Reply

For attested sandhi (re)spellings, entries using a template like {{sandhi form of}} seems appropriate. After all, we include attested pronunciation (re)spellings, etc, in other languages. - -sche (discuss) 16:49, 22 March 2017 (UTC)Reply

@-sche, Angr, Chuck Entz: Where I notice them clashing with Pali word forms, I've been adding them using form tag 'sandhi form of' in {{inflection of}} and linking to the citation form, e.g. at कामो with the frequent clash between Sanskrit and Pali nominative singular and Hindi vocative plural:

sandhi form of nominative singular of काम (kā́ma, “desire”)

I've just noticed that @Svartava has been using {{combining form of}}.

Ah well. --RichardW57 (talk) 10:29, 27 August 2023 (UTC)Reply

@Aryamanarora and whoever else may be interested: I've created a new and currently very naïve pronunciation module for Sanskrit at mod:sa-pronunc. I'd like to add many more features such as Vedic vs. Classical pronunciations and syllabification. I'd love and feedback or help people would like to provide! For instance, I'm not sure exactly what should be done about anusvara and visarga yet. Also, I'm not sure what the best way to mark vowel accentuation is. There's a test area set up at User:JohnC5/Sandbox2 which all are free to edit. —JohnC5 06:19, 15 July 2016 (UTC)Reply

I'm out of town without computer access, so I can't do much to help until I return in two weeks. However, I'll look at it and help if I can. —Aryamanarora ^{(मुझसे बात करो)} 21:43, 19 July 2016 (UTC)Reply

@Aryamanarora: FYI: the primary discussion is now going on at Wiktionary:Beer parlour/2016/July#Advice for a Sanskrit pronunciation module. —JohnC5 22:23, 19 July 2016 (UTC)Reply

In Category:Sanskrit prefixes and Category:Sanskrit suffixes, some entries have hyphens and some do not. What's right? By the way, do we even have non-Devanagari entries?__Gamren (talk) 08:41, 21 March 2017 (UTC)Reply

Yeah, I've never known the answer to this. —JohnC5 14:33, 21 March 2017 (UTC)Reply

I think affixes should be marked with hyphens, if only because {{affix}} doesn't work properly otherwise. We've had a few discussions over the years about admitting Sanskrit entries in the Latin alphabet, but so far we've never reached consensus one way or the other. —Aɴɢʀ (talk) 19:26, 21 March 2017 (UTC)Reply

What about all the scripts associated with various Dravidian languages? Many of them have been used to write Sanskrit for a number of centuries. Chuck Entz (talk) 02:37, 22 March 2017 (UTC)Reply

I oppose using hyphens in the transliterations, even if it makes the etymologies better understood. Devanagari doesn't employ it or any other symbol to show compounds, e.g. Persian uses ZWNJ, which may have a visual effect of non-connected Arabic letters. Besides, I don't think hyphens are part of the standard IAST. --Anatoli T. ^{(обсудить}/^вклад) 02:45, 22 March 2017 (UTC)Reply

I know nothing of Sanskrit, but I generally support the idea that the described language should be as close to the actual language as possible, also orthographically. I suspect that to add a hyphen out of pedagogical reasons may be to do learners a bjørnetjeneste. {af} may not work, but then we can use {prefix} and suffer the slightly more typing that ensues. This latter recognizes that Japanese doesn't have hyphens, so presumably we can do the same for Devanagari. This would have been easier if whoever made those templates initially had demanded users to provide hyphens when they are there.__Gamren (talk) 10:38, 22 March 2017 (UTC)Reply

I see no difference between using hyphens to mark affixes in Devanagari Sanskrit and using them in English. We have entries like un- and -ness, even though we write uncleanness and not *un-clean-ness. —Aɴɢʀ (talk) 15:43, 22 March 2017 (UTC)Reply

I don't understand what you mean. In English, Danish, German and probably other European languages, there is a tradition of marking affixes with a hyphen. Devanagari apparently doesn't have that tradition. What does *un-clean-ness have to do with anything? I am genuinely bewildered.__Gamren (talk) 16:26, 22 March 2017 (UTC)Reply

I don't see a problem with hyphens in Devanagari. Benwing2 (talk) 19:30, 3 April 2017 (UTC)Reply

@Benwing2 Hyphens or other symbols that represent hyphens are not used in Sanskrit Devanagari. When they ar , then they can be transliterated, of course, as in Hindi शादी-शुदा (śādī-śudā). --Anatoli T. ^{(обсудить}/^вклад) 22:03, 3 April 2017 (UTC)Reply

I've added a template for referring to the Nachtragswörterbuch des Sanskrit Template:R:sa:NWS. I know most people who use Wiktionary don't read German, so it's of limited use perhaps, but for those who do it's a reference which provides more up-to-date, more diverse, and more organised material than Monier-Williams (who in fact based his dictionary on the Petersburger Wörterbuch, which presents the core of the NWS). What do you think? —caoimhinoc (talk) 05:08, 10 September 2021 (UTC)Reply

I've created a template for handling verb conjugations and have begun integrating it in pages where the older conjugation tables were being used, like भवति and रिणक्ति. I have a couple of sources in the documentation for that template. Of course, if anyone has any ideas to help out or update that module it'd be greatly appreciated.

The way I've gone about it is to treat Sanskrit the way Ancient Greek is being treated right now, and append the aorist, perfect, benedictive, conditional, etc forms to the main verb (whose lemma form is the third-person singular present tense indicative). Moving forward, is this the way we want to treat Sanskrit verbs (as in, is the third-person singular present tense indicative is the lemma form, and other forms, like say the third-person singular perfect tense indicative, are non-lemma)? Or, is the third-person singular present, perfect, aorist, conditional, etc various lemmas in their own right under the same root (the way secondary derivations like the causative, desiderative, etc are treated)?

In Monier-Williams these aren't really treated like lemma forms, and it also feels helpful to group them under the present tense form for sake of listing all primary verbal forms in one place. It would be great to ultimately formalize this here.

Dragonoid76 (talk) 19:36, 10 February 2023 (UTC)Reply

@Benwing2 @AryamanA @Pulimaiyi bumping this again and tagging people who have edited this page to see if anyone has any thoughts on this. What counts as the "lemma form" for Sanskrit verbs? Most dictionaries list all the third-person singular (and sometimes more than just third-person) primary verb forms and secondary verbal forms under some sort of root. The way we have it, the third-person present singular is clearly a "lemma" form, like गमति and करोति.

Firstly, it seems like non-third-person-singular forms for any conjugation paradigm are rarely considered "lemma" forms if the third-person singular form for that paradigm exists.

Beyond that, there are a bunch of third-person forms that plausibly could be considered a "lemma". Considering the example of करोति, we haveː

1. करोतु , the imperative third-person singular. This form is non-lemma in other IE languages and so it makes sense that it should be non-lemma, but it's unclear whether:
   1. it should be "underneath" करोति. If so, the imperative forms are included inside the conjugation paradigm of करोति and the link-out in the non-lemma page of करोतु says something like "third-person plural imperative of करोति". To me, this makes the most sense, and it's the way this sort of thing is handled in PIE, Latin, and Ancient Greek.
   2. it should be "underneath" कृ. If so, the imperative forms are included somewhere on the root page and the link-out in the non-lemma page of करोतु says something like "third-person plural imperative of कृ". This seems like it'd needlessly clutter the root page.
2. कुर्यात्, the optative third-person singular. Same deal as above.
3. अकरोत्, the imperfect third-person singular. This one is non-lemma in PIE (included with the conjugation system of the present tense form) or Latin/Ancient Greek, but it seems like it could be a lemma form with its own conjugational scheme.
4. करिष्यति, the future third-person singular. Would be considered a lemma in PIE but not similar forms in Latin/Ancient Greek.
5. कर्ता, the periphrastic third-person singular future.
6. अकरिष्यत्, the conditional third-person singular
7. अकार्षीत्, the aorist
8. क्रियात्, the benedictive/precative
9. चकार, the perfect
10. non-finite forms like कर्तुम् (the infinitive), कृत (the past-passive participle), etc. These are definitely lemma forms in their own right, and it seems most natural to present them with the root in some way.

Here's one way of organizing things, which is not what we're doing right now and would require a bit of work, which makes a lot of sense to meː

1. The lemma verbal forms for the root कृ (which itself is a lemma "root" form) are करति (present), अकरत् (imperfect), करिष्यति (future), maybe कर्ता (periphrastic future), अकरिष्यत् (conditional), अकार्षीत् (aorist), क्रियात् (benedictive), and चकार (perfect). The page of each of these third-person singular forms is categorized under its type (e.g. "Category: Sanskrit periphrastic future verbs", or something similar), and lists the entire paradigm (including participles) for that type (the modules currently exist for this). Furthermore, the page of करति (present) also includes the aforementioned optative and imperative forms. All the non-lemma forms have pages that link back to the lemma form, and the lemma form links back to the root's page.
2. The page for the root कृ has some table, which needs to be created, to show the primary conjugations (the present, imperfect, future, periphrastic future, conditional, aorist, benedictive, and perfect) and the secondary conjugations (causative, desiderative, intensive). It's unclear whether this root page should also include whatever we call "non-lemma" forms, like the optative, passive, or imperative.

The drawback of this is that we can't see all verbal forms underneath the same page. You'd have to click into the pages like अकरिष्यत् and क्रियात् to see the conditional and benedictive paradigms.

Another way of organizing things, which is more similar to what we're doing in Ancient Greek and Latin is to put all the forms under the third-person singular, e.g. करोति, and make them all non-lemma. Pages like चकार should then say "third-person perfect of करोति" and not "third-person perfect of कृ", since the forms are located at करोति. The "root" page कृ then just includes, under verbal derivations, the page करोति alongside secondary derivations like the causative, desiderative, etc as it does right now in that little floating table on the right side.

This is definitely something that should be noted very clearly on the Wiktionary:About Sanskrit page the same way it's noted clearly in the Wiktionary:About Latin. Dragonoid76 (talk) 07:13, 16 August 2023 (UTC)Reply

@Dragonoid76 I'd recommend (a) post this into the WT:Beer parlour, where you'll likely get more eyes on it, and (b) give a TL;DR summary of what you're looking to get advice on, since you've written a wall of text that I can't quite make sense of. Some people might have thoughts on it, e.g. User:RichardW57, who works with Pali, where the same issues come up. Benwing2 (talk) 21:15, 16 August 2023 (UTC)Reply

And please be less racist and link to the various terms using {{m}} or {{l}} rather than square brackets. I also recommend using {{wgping}} to ping the members of the sa {Sanskrit}, pra (Prakrit) and possibly hi (Hindi) interest groups. RichardW57 (talk) 18:00, 17 August 2023 (UTC)Reply

@RichardW57 Please stop accusing people of racism, it is nonsensical and doesn't contribute anything. Benwing2 (talk) 18:25, 17 August 2023 (UTC)Reply

@Benwing2: Only displaying the Devanagari disadvantages those who have no reason to read Devanagari yet know enough about Sanskrit to contribute usefully to the discussion, and is contrary to Wiktionary's principle of supplementing non-Roman script. It appears hostile to those who are not attached to the North Indian view of proper Sanskrit. --RichardW57m (talk) 14:07, 18 August 2023 (UTC)Reply

You have a point, but you're being abrasive, and racist is not the correct term. —Caoimhin ceallach (talk) 17:22, 18 August 2023 (UTC)Reply

@Dragonoid76 Are you planning to follow Benwing2's suggestion, or should we start replying to the substance of the discussion here? --RichardW57m (talk) 09:37, 21 August 2023 (UTC)Reply

Done, see Wiktionary:Beer parlour/2023/August#Sanskrit Lemmas Dragonoid76 (talk) 04:24, 22 August 2023 (UTC)Reply

@Dragonoid76, good job on the page update. I've a question: why is धा (dhā) considered zero grade? Because historically it's full grade. —Caoimhin ceallach (talk) 12:56, 23 October 2023 (UTC)Reply

@Caoimhin ceallach Thanks. Actually, maybe "zero grade" is the wrong word here. What I meant is that most dictionaries (e.g. Monier-Williams, Whitney) list धा (dhā) as the "root", and the "root" usually corresponds with the historical zero-grade but this is not always the case. Mybe removing that would make things more clear. Dragonoid76 (talk) 13:05, 23 October 2023 (UTC)Reply

(Notifying AryamanA, Bhagadatta, Svartava, JohnC5, Kutchkutch, Inqilābī, Getsnoopy, Rishabhbhat, Dragonoid76): How are we to distinguish present participles like भवत् (bhávat) that have strong stems in some non-neuter forms from those like जुह्वत् (júhvat) that don't? Whitney distinguishes them as bhávant and júhvat. I suppose that for the purposes of generating declension tables by {{sa-decl-noun-m}} we could use |participle=2 with some jiggery-pokery in the receiving module. --RichardW57 (talk) 05:49, 12 February 2024 (UTC)Reply

@RichardW57 This is a good point. I've seen it both ways, which is pretty confusing, particularly because Monier-Williams gives both as -अत् (-at). I think considering -अन्त् (-ant) a variant form of -अत् (-at) and a parameter {{|participle}} is a good idea, since it will be confusing if -अन्त् (-ant) means one thing and -अत् (-at) means another. Dragonoid76 (talk) 19:53, 16 February 2024 (UTC)Reply