Talk:ಸಗಣಿ

Etymology

@Kutchkutch, AryamanA Any idea on how to better present the etymology than just writing "From Prakrit" ? The structure of this Kannada word is very curious and I find no suitable candidates for the probable immediate source of this word among the attested Indo-Aryan terms. It has an initial /s/. Dialectally the term is pronounced /ʃəɡ(ə)ɳi/. This is unlike Prakrit which has a /t͡ɕʰ/ but agreeing with Konkani and Marathi. It ends in /i/, which is perhaps related to the final ī of Old Marathi śeṇī. But unlike Konkani and Marathi, Kannada has retained the medial /ɡ/.
I cannot think of any other examples of alterations of this kind (i.e. Prakrit /ch/ -> Kannada /ś/ or /s/). I suppose either of the following to have happened:

1. Kannada term was borrowed from Old Marathi because the sound shift from Prakrit /ch/ to Marathi /s/ is known: c.f. Prakrit chāilla -> Old Marathi sāulī. But the Old Marathi term being śeṇī does not explain the medial /ɡ/ in Kannada. So was this a case of Kannada borrowing from an undocumented Old Marathi dialect that retained the medial voiced velar? In that case, the etymology chain will be: MIA chagaṇa -> unattested vernacular Old Marathi *sagaṇī -> Kannada sagaṇi ~ śagaṇi; whereas for the standard Marathi term it will be MIA chagaṇa -> late MIA *chayaṇa ~ *chayaṇī -> Old Marathi *seṇa ~ śeṇī ~ *śeṇa ~ śeṇī -> Marathi śeṇ.
2. The Kannada term was borrowed from Old Marathi śeṇī and then hypercorrected to include the medial /ɡ/ through the influence of Classical Sanskrit छगण (chagaṇa) (itself borrowed from Prakrit).
   

Any thoughts? -- 𝓑𝓱𝓪𝓰𝓪𝓭𝓪𝓽𝓽𝓪^{(𝓽𝓪𝓵𝓴)} 02:54, 22 April 2022 (UTC)

@Bhagadatta: Is the Marathi (and Kannada) word necessarily from Prakrit chagaṇa? Isn't it better explained by:

OIA śákr̥t + -ikā → MIA *sagaṇ-iyā → Late MIA *sagaṇī

→ Kannada sagaṇi ~ śagaṇi

→ sayaṇī (loss of intervocalic stops) → seṇī (vowel coalescence) → śeṇī (palatalisation)

All of these changes are attested in Marathi. However, you do have a point that MIA ch > Marathi s (secondary ś when palatalised by following vowel) is very common. There are lots of examples in CDIAL, e.g. chárdati > sā̃ḍṇẽ, kacchapa > kāsav. I am not sure how early that can be dated to, does it show up in Old Marathi?

If you can claim MIA *sagaṇī continuing the OIA form in ś- that works to explain the Kannada form, or else you have to posit a ch > s by the time of late MIA already. Agree it's complicated, I'd prefer the former explanation more for now. —Aryaman^{A (मुझसे बात करें • योगदान)} 05:11, 22 April 2022 (UTC)

@AryamanA: Thanks! I completely overlooked OIA -ikā. It does nicely explain the final vowel in both Kannada and Old Marathi (Kannada regularly reduces the length of the final non-schwa vowels from Indo-Aryan; cf ಪತ್ನಿ (patni) <- पत्नी (patnī)). And yes, ch -> s sound shift had already appeared in Old Marathi; साउली (sāulī) is Old Marathi, the modern Marathi reflex being सावली (sāvlī). Both our solutions favour an unattested variant as the source of the Kannada term. This is what had me hesitating in the first place; Indo-Aryan terms inherited from unattested variants are very common but Kannada usually borrows literary Prakrit/Pali terms. -- 𝓑𝓱𝓪𝓰𝓪𝓭𝓪𝓽𝓽𝓪^{(𝓽𝓪𝓵𝓴)} 09:53, 22 April 2022 (UTC)

@AryamanA, Bhagadatta The intervocalic sound change /ɡ/ > ∅ of the MIA stage would have already concluded before reaching the Early NIA stage. {{R:omr:Panse|54}}:

Intervocalic -𑘎- (-ka-), -𑘐- (-ga-), -𑘓- (-ca-), -𑘕- (-ja-), -𑘝- (-ta-), -𑘟- (-da-) are lost.

Since an unattested Old Marathi term retaining /ɡ/ is could be considered too speculative, the borrowing probably occurred at the MIA stage. It follows that MIA term must have either been *sagaṇ-𑀇𑀓𑀸 (ikā) + as AryamanA suggests or Prakrit 𑀙𑀕𑀡 (chagaṇa) + 𑀇𑀓𑀸 (ikā). Since Prakrit (𑀘𑁆)𑀙 ((c)cha) > Old Marathi 𑘭 (sa) / 𑘫 (śa) had not occurred yet, the Kannada donor is more likely *sagaṇ-𑀇𑀓𑀸 (ikā).

Prakrit (𑀘𑁆)𑀙 ((c)cha) > Old Marathi 𑘭 (sa) / 𑘫 (śa) is treated as a regular sound correspondence.

{{R:omr:Panse|51}}:

Old Marathi 𑘔 (cha) only occurs in Sanskrit loanwords

Sanskrit छ (cha), त्स (tsa) → Old Marathi 𑘔 (cha)

{{R:omr:Panse|66}}:

Sanskrit क्ष (kṣa), त्स (tsa), प्स (psa) > Prakrit (𑀘𑁆)𑀙 ((c)cha) > Old Marathi 𑘭 (sa)

The exceptions to this correspondence appear to be:

Prakrit 𑀙𑀸𑀆 (chāā) > Old Marathi 𑘔𑘰𑘁 (chāā) ({{R:CDIAL|chāyāˊ|5027)}})

Old Marathi 𑘔𑘰𑘁 (chāā) has the alternative forms 𑘭𑘰𑘃 (sāī) and 𑘭𑘰𑘊 (sāe).

Prakrit 𑀙𑀢𑁆𑀢𑀻𑀲 (chattīsa) > Old Marathi छतीस (chatīsa) > Marathi छत्तीस (chattīs) ({{R:CDIAL|ṣáṭtriṁśat|12786}})

Prakrit 𑀙𑀧𑁆𑀧𑀡𑁆𑀡 (chappaṇṇa) > Old Marathi 𑘔𑘢𑘡 (chapana) > Marathi छप्पन (chappan) ({{R:CDIAL|ṣáṭpañcāśat|12787}})

Both of these are numerals with the ones/units place being 6.

Prakrit 𑀧𑀘𑁆𑀙 (paccha) + 𑀟 (ḍa) + 𑀇𑀓𑀸 (ikā) > Old Marathi 𑘢𑘲𑘓𑘚𑘲 (pīcaḍī) < Marathi पिछाडी (pichāḍī) ({{R:CDIAL|*paśca|7990}})

Prakrit *𑀧𑀘𑁆𑀙𑀸𑀟-𑀇 (*pacchāḍa-i) > an Old Marathi reconstruction with c(h) > Marathi पछाडणे (pachāḍṇe) ({{R:CDIAL|*pracchaṭ|8493}})

Prakrit *𑀧𑀘𑁆𑀙𑀸𑀟-𑀇 (*pacchāḍa-i) could be reconstructed from the attested Sauraseni past participle .

All of these {{R:CDIAL}} entries do not have the → symbol.

Although *sagaṇ-𑀇𑀓𑀸 (ikā) could function as the etymon for Old Marathi 𑘫𑘹𑘜𑘲 (śeṇī) and शेण (śeṇ), it may be unnecessary since they can be inherited from the attested Prakrit 𑀙𑀕𑀡 (chagaṇa) with Aryaman's sequence:

extension 𑀇𑀓𑀸 (ikā)

Prakrit (𑀘𑁆)𑀙 ((c)cha) > Old Marathi 𑘭 (sa)

vowel coalescence {{R:omr:Panse|56}}:

aya > e (also includes Sanskrit उपयोग (upayoga) → Old Marathi 𑘄𑘢𑘹𑘐 (upega))

palatalisation of Old Marathi 𑘭 (sa) to 𑘫 (śa) before i(ː) and e

Furthermore, inheriting from Prakrit 𑀙𑀕𑀡 (chagaṇa) would make the Old Marathi and Marathi descendants analogous to the other Indo-Aryan descendants with word-initial ch. Kutchkutch (talk) 01:51, 23 April 2022 (UTC)

@Bhagadatta {{R:dra-okn:KSP}} page ೩೬೧, column 2 has

ಸಗಣ - (ಶಕೃತ್) ಸಗಣಿ

So the Old Kannada term is from *sagaṇ without 𑀇𑀓𑀸 (ikā)? Kutchkutch (talk) 03:05, 23 April 2022 (UTC)