Talk:रोकड़

@Bhagadatta The Dravidian borrowings (ರೊಕ್ಕ (rokka), రొక్కము (rokkamu), ரொக்கம் (rokkam)) have kka instead of ka. Does this suggest that they are borrowed from *𑀭𑁄𑀓𑁆𑀓 (*rokka) rather than Sanskrit रोक (roka) / Hindi रोकड़ (rokaṛ). Or, is the ka → kka change common in Indo-Aryan borrowings? Kutchkutch (talk) 06:34, 31 August 2020 (UTC)

@Kutchkutch: These Dravidian borrowings are not directly from Sanskrit but from an unattested Middle Indo-Aryan language. Kannada et al never change Skt. ka to kka; c.f. ಶ್ಲೋಕ (ślōka, “prayer”), ಶೋಕ (śōka, “grief”), ಲೋಕ (lōka, “world”). I intentionally used these 3 examples to draw attention to the length of the vowel o: whereas Skt. does not distinguish the length of ए and ओ, the Dravidian languages do. The above loanwords in Kannada have the long o as they come directly from Sanskrit. But in the borrowings ರೊಕ್ಕ (rokka), రొక్కము (rokkamu), ரொக்கம் (rokkam), there is a short o, which cannot have come from Skt. directly, at least in Kannada and Telugu as they preserve Skt. long o in borrowings. This long o was shortened in Dravidian, I think, as a compensation for the double -kk-.

Kannada has quite a few Indo-Aryan words which don't directly come from Skt. but through an MIA intermediary. In some cases this intermediary is unclear or unattested, like in the case of ಕೊಟ್ಟಿಗೆ (koṭṭige, “cow-shed”), ultimately from Sanskrit गोष्ठ (goṣṭha) - again, note the short o in Kannada! -- Bhagadatta (talk) 06:49, 31 August 2020 (UTC)

@Bhagadatta: Thanks for the examples and highlighting the importance of the vowel length! The Prakrits also have short o (ŏ) and short e (ĕ), but only before geminate/double consonants. Thus, there was no need to have any indication of this in writing. On Woolner page 14:

The vowel before the double consonant is always short. [In these cases,] ए and ओ represent the short vowels ĕ and ŏ.

So if the Dravidian borrowings do come from an unattested MIA language, that might justify creating a reconstructed *𑀭𑁄𑀓𑁆𑀓 (*rokka). Wasn't it most likely the case that the Old/Middle Dravidian languages were the ones that borrowed them from MIA and then the modern languages inherited the words from the Old/Middle Dravidian languages?

The interaction between Dravidian and Indo-Aryan languages is certainly fascinating, especially between Marathi-Konkani and Kannada. I tried looking for the Old Kannada etymon for ओटा (oṭā) in {{R:dra-okn:KSP}} and found ಒಟ್ಟ (oṭṭa) on page ೭೪ (74). So now I'm wondering whether to put that as the etymon of ओटा (oṭā). Kutchkutch (talk) 07:43, 31 August 2020 (UTC)

@Kutchkutch: Yes, that will be the correct etymon for Marathi. I haven't heard ಒಟ್ಟ being used in modern Kannada though it may still be surviving in many of the 20 odd dialects of Kannada.

You're right, the borrowing happened before the modern stage of the languages. In case of ಕೊಟ್ಟಿಗೆ (koṭṭige) it happened when Tamil and Kannada hadn't even separated. This separation was perhaps completed by around 5th century BC so there was time for borrowing from early MIA. Other interesting examples are ಹಕ್ಕಿ (hakki, “bird”), ಹುಣ್ಣಿಮೆ (huṇṇime, “full moon”), ಹಸು (hasu, “cow”) etc. -- Bhagadatta (talk) 08:26, 31 August 2020 (UTC)

@Bhagadatta: Thanks for the brief overview. I created Ashokan Prakrit *𑀭𑁄𑀓𑁆𑀓 (*rokka). Please edit it as necessary, especially since it's the first reconstructed entry that I've ever made. The words in CAT:Marathi terms derived from Dravidian languages (आत्या (ātyā), लातूर (lātūr), etc) would benefit from a better understanding of Dravidian. Some words are indicated as borrowings from Middle Kannada while others are indicated as borrowings from Old Kannada. Kutchkutch (talk) 09:58, 31 August 2020 (UTC)

@Kutchkutch: The entry is good. I have replaced Proto-Dravidian with Proto-South-Dravidian as the former was spoken 1000 years before MIA (Proto Dravidian might have been from the same time as Vedic Sanskrit). Per Wikipedia Telugu does not belong under South Dravidian but per the linguist Bhadriraju Krishnamurthi, it does. So I have put in Proto-South-Dravidian, the contemporary of Proto-Prakrit.

I added Konkani रोकडे (rokḍê) as a descendant though semantically it does not entirely fit. रोकडे means "quick", perhaps from "ready money" --> "quick money" --> "quick". -- Bhagadatta (talk) 10:59, 31 August 2020 (UTC)