IE K^ > K has been proposed for Bangani ( http://www-personal.umich.edu/~pehook/bangani.html ) for *g()lak^t > lOktO ‘milk’, etc. Claus Peter Zoller claimed that Bangani was related to Kashmiri, maybe showing a Centum substrate. Others seem to show *r > Er, *e(:) > e / i (dukti ‘daughter’). Despite disputes, even the opponents of this include some of his words in their data (George van Driem in https://www.academia.edu/10165900/Some_grammatical_observations_on_Ba%E1%B9%85g%C4%81%E1%B9%87%C4%AB ) with no reason given for why they are not from the IE sources Zoller gave. Similar words have been dismissed by Witzel, but with no evidence (see below). Why was this claim disputed in the first place? When I examined both Zoller’s and van Driem’s data, and besides their matches they do have so many similarities with Kashmiri and other Dardic languages in vocabulary, phonology, etc., that I can’t imagine any other theory would fit. These do have a number of more recent loans, but their old cognates are clear.
Centum Indic words seem odd, but it can hardly be false if it also exists in Dardic, Rom. and Domaaki. Zoller has proposed nothing new, with much of the evidence for the same words from studies 100 years ago, so unless there is a giant conspiracy reaching throughout time and space, this seems to vindicate Zoller. Some of the specific words with K^ > K below (no need to read them all if you already have an answer):
*k^(h)atru- > B. kOtrO ‘fight’, Kh. khoṭ
Li. liežùvis, Kh. ligìni, E. tongue (reanalyzed with *leig^h- ‘lick’ )
*bhah2g^hu- > Skt. bāhú- ‘arm’, Bu. baγú ‘armful’, OE bóg ‘shoulder’, Dk. phaaká \ phóok ‘shoulder’, Rom. phiko, Sh. phíǰo IIr. dual *bhah2g^huni > Ba. bakuí~ , Ti. bekhi~n ‘arm(s)’, KS bEkhin ‘elbow’
PIE *dbhng^hulo- > G. pakhulós, Skt. bahulá- ‘thick / spacious/abundant/large’, A. bhakúlo ‘fat/thick’, Ni. bukuṭa ‘thick [of flat things]’, Rom. buxlo ‘wide’
*k^uwon- > *k^uwaṇ-i-? > *šoṛeŋí- > D. šoṛíing ‘dog’ *k^uwaṇ-aka-h > A. kuṇóoko ‘pup’, kuṇéeki ‘female dog/pup’ *c^uwaṇ- > *šoṛaŋ- > (with met.) D. šongaṭék ‘female dog/pup’
*meg^h2- > *maga ‘very’ >> Sh. mʌ´γʌ dúr ‘far away’
*k^rhnó- > Rom. kerno ‘rotten’, Skt. śīrṇá- ‘broken/shattered / decayed/rotten / fallen out/off’
Rom. gonil ‘banish / drive/chase away/out’ < *gWhen-ye- ‘drive away, beat (off), hit’ > L. of-fendō, Greek theínō, Li. geniù ‘lop off’
Skt. star- + rájanī- ‘night’ : *štar-ráγani:- > *štar-aγari:n > *čerxari:n > Rom. čerharin ‘star’
Since bāhú- : baγú also has K, even if the Bu. word is a loan it would have to be from an IE language with K^ > K. There is no reason not to see this K as original in all words.
That these really came from K^ with both retention or > T^ optional seen in:
Sh. phíǰo, Rom. phiko ‘shoulder’
the same opt. K / K^ as in original K (lhoóko vs. loóts ):
Skt. laghú- ‘light’, *lakhu- > *lhaku- > A. lhoóko ‘small’, Kh. loóts ‘light’, Sh. lóko, Rom. lokó ‘light’
Skt. śaśaka-, Rom. šošoy ‘rabbit’, Dk. ošæyo ‘hare’
Since A. bhakúlo seems to require *bakhúlo > bhakúlo it would be evidence that Dardic retained g^h as gh (opt. > kh ) here, with the same in Rom. buxlo. Why hasn’t this type been noticed before? Of course, it has, but linguists have said the ones who did were wrong. For example, in
Skt. bahulá- ‘thick / spacious/abundant/large’, A. bhakúlo ‘fat/thick’, Rom. buxlo ‘wide’ < *dbhng^hulo- > G. pakhulós
Witzel and Strand give *bhakk(h)a- ‘lump’… from where? Well, in http://www.aa.tufs.ac.jp/sarva/entrance.html
9330 *bhakkha— ‘lump’. 2. *bhikkha—. 1. Ḍ. b *l/ kula ‘thick’ (Lorimer ḌumLg 150 wrongly < bahala—); Paš.dar. bakoṭā́, ar. ba g uṭṓ ‘stout, fat’; Bshk. bakū́l ‘strong, thick, fat’; Phal. bhakulo ‘strong’; Sh.pales. bhāk ‘bull’; S. bhakū m. ‘blockhead’, bhakuo ‘stupid’; L. bhakrain m. ‘hammer for breaking lumps of earth’ (+?); P. bhakūā ‘foolish’; Ku. bhakuno ‘stock of grain’; N. bhakku ‘short stout man’, bhakullo ‘chubby, plump’, bhakānu ‘lump, clod’, bhakuṇḍo ‘wooden ball used in a game’; B. bhākus ‘stupid, awkward’; Or. bhaku, °kā, °kuā ‘fool’, bhākuṛi ‘short and stout’, °ṛa ‘whoreson’; Mth. bhakkū ‘fool’, H. bhakuā m.; G. bhākhrɔ m. ‘lump of bread’, M. bhākar f.; Si. baka ‘big’, bakkā ‘big man or animal’, bakkan ‘large, plump, clumsy’. 2. P. bhikk f., bhikkā m., bhikkaṛ m.f. ‘lump, clod of earth’.
Since bahulá- : bhakúlo are as identical as any related words could hope to be, I don’t see how “wrongly” describes this etymology. The other Dardic and Rom. words make this impossible (even if he were right, the same odd word would be ev. that Dardic and Rom. were related).
Part of the reason K(^) > K has not been accepted is a failure to put some of these changes in context. For ex., in goi- \ gom- ‘sacrifice’, the alt. i / m makes no sense in itself, but if < *g^heu- (G. khéō, Skt. juhóti ‘pour a libation / sacrifice’) other Dardic data shows that *au > *aü > *öü > oi existed (*rauhita- > A. lohóylo ‘red’) along with opt. w / m (Skt. náva- ‘ young / new’, Ti. nam ), allowing both to come from one form, not an ad hoc explanation. Changes like *gWehtu- > getu 'resin', *dhug^hxte:r > dukti 'daughter' seem to show opt. *e: > *i: (*neh ‘not’ > Dm. ni, Id. nà ) and differing treatment of syllabic C (like Dardic *m > um, etc.). Without knowing this, loans from another IE language would be needed, instead of native Indo-Iranian.
Bangani also retains laryngeals as K, and I think Dk. also retains laryngeals as *x > kh in:
*h1oh3s- > ON óss ‘river mouth’, Skt. ās-, Dk. kháša, Kv., Kt. âšá ‘mouth’
and probably *xt > *kht > kt in:
*dhuh- ‘shake / smoke / spirit’, *dhuhto- > Skt. dhūtá- ‘shaken / agitated’, B. dukti ‘soul / last breath’. The importance for this (establishing h2 = x, etc., or similar) can not be overstated. Without accepting the evidence, finding the exact nature of each change would be impossible.
Skt. śapátha- ‘oath’, *šapaxta > *šavaxla > *šavlaxa > Rom. solax
Kh. šot čhoy- ‘swear an oath’, Kv. šüt (ku-) seem to come from Skt. śapátha-s; this also matches Rom. solax. Since śapátha- might have gotten -th- from *-xt-, seeing -x- here is support (related languages like Dk. and B. also show kh and k corresponding to *h() ).
Much more on the specifics here:
https://www.reddit.com/user/stlatos/comments/11r4n6t/dardic_languages_romani_domari_domaaki/
Dv Domari \ Do:mva:ri: Lv Lomavren Rom Romani
Dardic Group A Atshareetaá \ (older Palola < *Paaloolaá) B Bangani Ba bHaṭé-sa zíb \ Bhaṭeri D Degaanó \ Degano Dk Domaaki \ Domaá \ D.umaki Dm Dameli Gi Gultari Id Indus Kohistani Ka Kalam Kohistani / Kalami / Gawri / Bashkarik Kh Khowàr Km Kashmiri Kt ktívi kAtá vari Kv Kâmvíri Pl Paaluulaá Ni Nišei-alâ Pl Paaluulaá Sa Saňu-vīri Sh Shina Ti Torwali
