Category talk:Pali verbs

@Benwing2, Mahagaja, Octahedron80, RichardW57m, (Notifying AryamanA, Bhagadatta, Svartava, JohnC5, Kutchkutch, Inqilābī, Getsnoopy, Rishabhbhat): and please alert anyone else you think might have an interest.

I am distinctly unhappy with the way the 'conjugations' have been bolted into the category of Pali verbs.

The primary division for most Pali verbs is based on how the present stem is formed from the root. The present stem is used for the present tense, the imperative mood, the optative mood and the imperfect, though the latter, at least in canonical Pali, is largely defunct and its remnants are hard to disentangle from the aorist.

Other tenses and non-finite forms are in theory formed directly from the root, but actually they may also be made from the present stem.

While Sanskrit has consonant stems, Pali stems with one exception end in vowels, with a very few athematic remnants. The exceptional verb is the verb to be, with root as. The stems are often grouped into 7 conjugations, depending on how the present stem is formed from the root, as I have summarised at {{pi-verb}}. Some, somewhat arbitrarily, put as into the first conjugation, perhaps because that is the home of the other verb to be, bhū and its phonetically reduced by-form hū. Verbs in the seventh conjugation, which corresponds to Sanskrit Class X, may be made from things other than verb roots. Their hallmark is the stem ending -aya- or -e-, with most of the semantics coming from what precedes it.

There are also derived verbs: passives, causatives (and double causatives), and, as clearly non-productive categories, desideratives and intensives. There are also a few cases where forms for other tenses or moods have ended up functioning at presents. They can generally be analysed as deriveing from new roots.

1. Passives are mostly formed by affixing -(i)ya to the root. As the verbs of the third conjugation are formed by affixing -ya, I have been putting them in the third conjugation.

1. Many causatives are formally indistinguishable from verbs of the seventh conjugation. Others may be distinguish because the root is followed by -(ā)p(āp)-e/aya-. I intend to put all of them in the 7th conjugation, but perhaps some of them should be put in a causative 'conjugation'.

1. Desideratives are formed templatically. From a root CVD, one generally gets a present stem CiCVDsa-. The vowels may be lengthened, the /i/ may assimilate to /V/ (so yielding a pattern CVCVDsa-), a nasal may be inserted, and the sequence -Ds- can be variously modified, often to -cch-, and the whole may be truncated. Well, English strong verbs often don't exhibit good ablaut patterns. I'm incline to assign these to the first conjugation, or the second if a nasal is inserted in the right place.

1. Intensives are even worse, being a pattern CVDCVD(y)a- with random vowel gradations and consonant dissimilation have preceded Pali's intense consonant assimilations. The first and third conjugations will happily house these.

1. There are denominatives formed by suffixing -āya-. They would happily sit in the third conjugation along with gāyati, in which the -ā- is part of the root.

I'd be happy to have subcategories for passive, causative, intensive and desiderative verbs independent of the division into conjugations, though we could work with intensive and desiderative conjugations, and perhaps a separate 'causative' conjugation for those where a /p/ has been added after the root.

Now, if we wanted to divide them by inflection of the present stem, that is at one level straightforward - it appears to depend on how the stem ends, and not whether its ending is part of the root.

The main divisions are into:

1. -aya- (mostly seventh conjugation, some first conjugation (e.g. jayati)
2. other -a- (first, second and third conjugation, with some by-forms from the fourth and fifth conjugations (e.g. gaṇhati)
3. -ā- (first (e.g. dadāti), fourth (e.g. pāpuṇāti) and fifth conjugations)
4. -e- (first and seventh conjugations)
5. -o- (active), -u- (middle) (first (e.g. hoti), fourth (e.g. pappoti) and sixth conjugations (e.g. karoti).

There are some anomalies which I haven't addressed, such as the stems in -i- and -ū- (i.e. brūti).

What are these meant to be? If we consider the formation of the aorist, past 'passive' participle and absolutive, Pali seems to be at the PIE stage where almost all verbs are irregular. On the other hand, if we restrict attention to the present system, it's a bit more like English, where the irregular verbs other than pure auxiliaries are just be, dare, do, have and say - unless we worry about whether the middle is attested. There are a lot of verbs for which the middle is not attested. Are they defective? Remember that with the claimed exception of some of the most archaic material, the middle is essentially just a stylistic variant. --RichardW57 (talk) 22:56, 5 July 2023 (UTC) RichardW57 (talk) 22:56, 5 July 2023 (UTC)Reply

@AryamanA, Bhagadatta, @Benwing2: I've emptied the irregular conjugation category now, leaving atthi homeless until I decide whether I can act on:

c. 500 AD, Dhatumañjusa; republished in Dines Andersen & Helmer Smith, The Pāli Dhātupāṭha and the Dhātumañjūsā, Copenhagen: Andr. Fred. Host & son, 1921, page 43:

103. Duve pi rā lā ādāne / Vā ɡati-bandhanesu pi. / Asa dhātu bhuvi khyato / si saye sā samattiyaṃ. / / * Hū-bhūvādayo luttavikaraṇā.*

(please add an English translation of this quotation)

.

If you know a trustworthy translation, sharing it would help. --RichardW57 (talk) 19:23, 7 July 2023 (UTC)Reply

@RichardW57 You have written quite a lot, but the gist of it seems to be that Pali (like Sanskrit) doesn't have conjugations that are uniform across the entire system, but rather it has independent sets of conjugations for the present, aorist, past participle, etc. This situation exists to some extent in Latin (with the perfect in particular); there, we have CAT:Latin first conjugation verbs, etc. as well as CAT:Latin verbs with perfect in -u-, CAT:Latin verbs with perfect in -i-, CAT:Latin verbs with perfect in -s- or -x-, etc. For some reason there is no CAT:Latin third conjugation i-stem verbs or CAT:Latin verbs with reduplicated perfect but the principle remains. In the case of Pali, we could rename CAT:Pali first conjugation verbs to CAT:Pali verbs with first conjugation present or whatever and create corresponding categories for the aorist, past participle, etc. Benwing2 (talk) 20:44, 7 July 2023 (UTC)Reply

@Benwing2: Have you been reading other peoples accounts of Pali? The term conjugation gets applied with two different meanings. The first, which is what is used by {{pi-verb}}, is a division according to how the present stem is formed; this has then has a lot of influence on what the present stem looks like for the purpose of inflecting it, but one doesn't get the sort of profound difference Latin has in nuntio (first conjugation), capio (third conjugation, sometime called fifth), and audio (fourth conjugation). The second use of the word is for the TAM systems, such as present, imperative, optative, aorist, future and conditional. These are more like Ancient Greek, in that it seems that not every verb has all of them.

Now, it might be possible to classify roots by what conjugations they form verbs in, but I'm not sure how useful that is, and it doesn't seem easy to automate. While some verbs seem fairly orderly, common verbs can have rich mixtures. Take a look at what I've compiled so far for the simplex of gah (root) at gaṇhāti. It has two present stems, the other being historically further derived by shortening the final vowel of the stem to give gaṇhati - it works like a verb of the second conjugation. It is from the latter that the middle forms shown are built. The optative from the two stems is identical. The aorist has three stems - gaṇhi built from the present stem, aggahi built from the root, and aggahesi which looks as though it were build from a seventh conjugation stem, but might be first conjugation from an odd-looking root seen in Sanskrit. (The initial ag- on some of them is just the temporal augment plus the fact that the root used to start with a consonant cluster.) I haven't collected the future yet, but according to the PTS PED, it has future form the two present stems used to form the aorist, which makes sense that both tenses, for this verb, historically have a common element -is- in their formation. --RichardW57 (talk) 22:06, 7 July 2023 (UTC)Reply

@RichardW57 Dude, I don't know a lot about Pali; you can take my suggestions or leave them. I think it's a good idea to try to categorize the aorist, future, etc. according to common patterns, but if you disagree, you don't have to do it. Benwing2 (talk) 22:51, 7 July 2023 (UTC)Reply

@Benwing2: I do do some such categorisation, such as Category:Pali verbs exhibiting conditional mood. However, that's being collected because grammars give different accounts of its inflection, and it's easy to generate it from the inflection data. I'm also specifically collecting examples of conditional active first person plural (Category:Pali verbs with manual 1p conditional - nothing caught yet!), partly because the grammars were too discordant to allow me to generate it automatically. --RichardW57 (talk) 10:35, 8 July 2023 (UTC)Reply

It would be nice to categorise the aorist. Now, at present, lack of an entry for the aorist would be more of a maintenance category. Most grammars talk in terms of Sanskrit aorists, but that mostly leaves a mess. I have separated out aorists in -āsi with the 'si' dropping in the plural (category:Pali verbs with aorist in -āsi/-a), though that now requires manual tagging (|aasi=1) to distinguish these aorists from those where the /ās/ is an invariable part of the root, as in bhāsati. This was done to assist inflection table generation. I am endeavouring to follow Duroiselle's 'Practical Grammar of Pali' and distinguish the imperfect and the aorist. Most verbs lack the imperfect tense or rather, formation, and some grammars lump the aorist and imperfect together, some of them calling the aggregate the 'preterite'.

I think the following miscellaneous categories would be interesting:

1. Verbs with aorist middle
2. Verbs with aorist active 1s in -isaṃ
3. Verbs with aorist active 2s in -o and matching active 3s in -a/ā
4. Verbs with aorist active 2s in -o but no matching active 3s in -a/ā (something of a maintenance category - I will immediately suspect the form is imperfect active 2s.)

We could do a lot with the aorist active 3rd plural (a3p), though there might be a lot of fine tuning to do:

1. a3p in -aṃsu
   1. With corresponding 3s in -a/ā
   2. With corresponding 3s only in -i or -āsi
   3. Others (a collection of anomalies)
2. a3p in -ayiṃsu with 3s in -ayi (boring, typical of 7th conjugation, bled off from the following)
3. other a3p in -iṃsu
   1. With corresponding 3s in -i
   2. other (may need subdividing)
4. a3p in -isuṃ with corresponding 3s in -i
5. a3p in -esuṃ with corresponding 3s in -esi (boring, typical of 7th conjugation, bled off to expose the others)
6. other a3p in -uṃ with corresponding 3s in -i
7. other a3p in -uṃ with corresponding 3s in -a/ā
8. yet other a3p in -uṃ
9. other a3p in -u/ū
10. other a3p

I can think of a few categorisations for the future, but I've not been going out of my way to collect futures. I can think of some divisions (but taking each verb as a whole, not a partition):

1. Futures in -issati corresponding to a present stem
2. Futures in -issati not corresponding to a present stem
3. Futures in -essati corresponding to a present stem
4. Futures in -essati not corresponding to a present stem
5. Other futures in -ati
6. Futures in -iti
7. Other futures (I think this category should be empty.)

Some verbs would appear in multiple categories - gaṇhāti mentioned before would go in categories 1 and 4. I don't think the Ancient Greek pattern of active present and middle future occurs in Pali.

There is a raft of categorisations one could do with the derived verb forms in -t, but the no template currently hosts the necessary data yet. --RichardW57 (talk) 10:35, 8 July 2023 (UTC)Reply

@Benwing2 So, how do I correct the description of Pali desiderative verbs? I tried adding the following lines to Module:category tree/poscatboiler/data/lang-specific/pi:

labels = {
	description = "Pali verbs with the following morphology: "..
		"reduplicated verbal root xxx + ] affix, "..
		"and the following semantics: to wish to do the action xxx.",
	parents = {{name = "verbs", sort = "desiderative"}},
	breadcrumb = "Desiderative verbs",
}

before the return statement. (I added the word 'reduplicated'.) However, when I previewed the displays of Category:Pali verbs Category:Pali desiderative verbs, I saw neither error indication nor the new description. --RichardW57 (talk) 19:04, 8 July 2023 (UTC)Reply

@RichardW57 Did you figure this out? I do see the new description in the displays of those two categories. Benwing2 (talk) 19:14, 8 July 2023 (UTC)Reply

Never mind, I missed the change involving the word 'reduplicated'. Benwing2 (talk) 19:16, 8 July 2023 (UTC)Reply

@RichardW57 OK this should be fixed; the problem was that lang-independent labels were being processed before lang-specific ones. Benwing2 (talk) 20:09, 8 July 2023 (UTC)Reply