Last week, @lordmondegreen.bsky.social & I presented presented joint work with Burusho friends @charismawafi.bsky.social et al. (not on BlueSky) on historical reconstruction of #Burushaski in G(h)ent at the Isolated Languages Workshop. :)
#Burushaski
'will fall down'
qha bal-íć-i
down fall-IPFV-3S
qha a-pálj-i
down NEG-fall:IPFV-3S
#Burushaski
Yasin dialect
'I have to wash clothes'
jáa gaȚúnc háaļćaćum baş
háaļćaćum (with voiceless L) likely related to Húnza báaltt-/-páaltt-, seems to contain suffix *-ya of distributed action, interior to -ć-um imperfective
pl. gaȚúnc < *kàȚu +mü-c
with x-class marker
[HB gaȚóng is y]
Proto- #Burushaski > Húnza
old s-clusters
*sp > p- ~ -śp-
*st > t- ~ -lt
*sk > k- ~ -śk- ~ -sk
*sq > *χq > *üχq > *üşq > *işq > şiq-
#Burushaski
'eat'
1) unprefixed, y-class obj.
śé- / nuśén / oóśi- < *śiQ- < *śü(Q)-
2) 3S-affixed 'eat it,' discrete obj.
şí- / níşi-n / eéşi-
--these stems < *i-şí- < *i-śí- < *i-śü(Q)-
3) ostensibly 3P
şú- / núşu-n / oóşu-
< *u-şú- < *u-śú- < *u-śü(Q)-
#Burushaski
plural 'houses'
hakíćang
sing. is ha (*hak).
endings -anc, -ang seem to prefer stems with vowel /a/
what if hakíćang had the structure
*hàk-üt hàk-üm(ü)-k
house-that house-DERIV-those
essentially 'this house and those houses'
#Burushaski
compound types
Dvandva
gu-ww-gú-mi
2S-father-2S-mother
'your parents'
go-k-gu-yúa
2S-daughter-2S-sons
'your children'
Tatpurusha
Turntable type
duró-skuin [cf. sukúin]
work-kinsman
'helping hands'
OV-Agent
darú-c [= t-s (do-NOM)]
hunt-doer
'hunter'
#Burushaski
palatalization of coronals
(in onsets, at certain boundaries)
*t > d-/tt in grave environments
> ć in acute environments
*n > n in grave env.
> *ñ > y in acute env.
*c [ts] > c ~ ć
*s > s ~ ś
#Burushaski
plurals
'flower(s)' asqúr-ing
'dried apricots' baȚér-ing
(*-ümü-k)
'clothes'
gaȚú : gaȚóng
(*kàȚu +(a/Q)mü-k)
'girl ~s'
dasín : dasíwanc
(*tàs.hiñü +amü-c)
#Burushaski
'daughter(s)'
2s gu-yúguśanc
implies an earlier singular *gu-yúgus < *ku-kʲuküs, with regular palatalization of the stem-final
modern singular:
2s go-y < *ku-(Q)kʲuks
Element -gus/-küs, compare
gus 'woman'
pl. guśinganc
#Burushaski
'day'
sg. gunc (<c> = [ts])
combining-form -kuc :
'3 days' iskíkuc
'4d' wálkuc
'8d, week' altáanguc
Possibly the singular
gunc
is from *kuc-n
(w/ reduced form of (h)an '1, singulative' ?)
via metathesis, and Initial Voicing
#Burushaski
'shin' (an unpossessed bodypart)
píni : pl. píniming
(*spínü : *spínümük)
'my lip'
HB] eíl : eéling
(*a-hìl : *ahìlü[m]ük)
'thy eyelid' (< *eyebrow ?)
gúlpur : gúlpurang
(*ku-l't-pur : ~a[mü]k)
2S-eye-hair
#Burushaski
numerals
1 han ~ hin ~ hik
2 altán ~ altó (*àst-u)
3 iskí (*sk-ì) ~ uskó (*sk-ù)
4 wálti ~ wálto (*upʷást-u)
5 chindí ~ chundó (*chVnt-u)
6 miśíndi ~ miśíndo (*mìsint-u)
#Burushaski
'eat'
The vowel position in the root is secondary, via copying. Must originally have been @-ş-
2s guşú- (*ku-şV')
3f muşú- (*mu-şV')
3p uşú- (*u-şV')
reinterpreted as having a root şú-
vs 3s (0-)şí- (*i-şV')
ptcp. níşin
#Burushaski
know *Q-hakin-
(Q attracts accent to itself but is not a segment of the output)
hákin-am nu-hákin 'know, have learned'
áykin- 'teach me'
(*a-Q'hakin-)
góykin- 'teach you'
(*ku-Q'hakin-)
#Burushaski
daughters and sons
(dvandva compound)
1s ak-ayúa
2s gok-(g)uyúa
1p mek-(m)iyúa
compare 'daughter'
1s ay
2s goy (*ku-Q'y)
1p mey (*mi-Q'y)
and 'son'
1s eí (HB : *a-Cí)
2s guí (*ku-Cí)
1p mií
#Burushaski
velar nasal
'his beard'
ingí (*i-nkí)
(accent mora 2; <ng> is a velar nasal)
'front of him'
ínggi
(accent mora 1)
*i-níki, (orig. accent mora 2)
syncopates > *i-ńki, accent shifts to vowel, voice stop after nasal > ínggi
#Burushaski
root *qhar- 'split, break'
[x] qhari bi 'broke (naturally, by itself),' vi Pres. perf.
iqháram 'I broke it,' vt Pret.
neg. eéqaram (*a-í-qhar- HB)
ptcp. níqar[in] (*n-í-qhar- HB)
qháqaras 'skin chapped, peeling from overexposure to sun'
Accent pos'n suggests *qhar(á)~qhar +(a)s
#Burushaski
'shoe'
sg. húȚis (*huti-s)
pl. húćo (*huti-u)
compare the noun-class markers /-s/ (isé, nonhuman sg.)
/u/ (now human pl., distal; also 3p)
#Burushaski
starting to suspect there was one phoneme */h/:
[h] in onsets, deleting between vowels ~
[ş] in codas
and a distinct phoneme */s/:
[ś] in onsets ~
[s] in codas ~
sporadically merging into [ş] ?
Modern -VsV- develop from *-Vs#hV-
balás 'bird'
pl. baláśo
'a bird' balásan (*palas#han)
#Burushaski
Infinitive
-as ~ -áas
żú-as 'come'
ní-as 'go'
(HB-NB)
man-áas 'do/be'
min-áas 'drink'
Short form *-as
Long form *-ahas < *-asas
#Burushaski
'gold'
ghéniş, pl. ghéniyang
*qiQnVs : *qiQnVh-aN[u]k
original stem-final /s/
-> ş at word boundary
-> h between vowels
#Burushaski
why is medial /s/ so rare?
*s > ś at certain morph boundaries
initial *s- > h-
and maybe in onset generally
cluster ltt < *stt
possible Proto- #Burushaski consonant inventory
p t (ț) k q
ph th (țh) kh qh h
m n (ñ ng)
c ç (ć)
ch çh (ćh)
s (ş ś) h
(w) r l Ļ (y)
#Burushaski
Palatalization of the stem-final in plural affixation
s -> ś
balás balásyo (*palás+u) 'bird'
hilés hilésyo (*hilíQs+u)
'boy'
c -> ć
gírkic gírkicyo (*kir[í]kic+u)
'mouse'
n -> *ñ -> y (etc.)
ddasín ddasíwanc (*ttasiñu- < *ttasín+u+am[u]c)
'girl'
crhin crhíyo (*çhiñu < *çhín+u)
The #Burushaski for ‘foal, filly ’ is
Hunza-Nager biráġo /biˈɾɑ̈́ɢo/
Yasin bráğu /ˈbɾɑ̈́ʁu/
Clearly a Turco-Mongolic loan:
Proto-Mongolic *bUrağu ‘calf’
Proto-Turkic *bUzağu ‘calf’
Unlike other Turkic loans, however, this can’t be borrowed from nearby Turkic languages like Uyghur & Kyrgyz.
#Burushaski
'knee' @ddúmus
The medial voiced stop is rare in Modern Bsk (preponderance of them in initial pos'n)
May have been un-possessed *ttumus (*ttum-s) at an older stage; this underwent initial voicing, and only then shifted to the Possessed noun class (like most bodypart nouns)
#Burushaski
Plural-markers with /c/
(unasp. dental affricate [ts])
0-yáT-umuc 'heads' 3p
(sg. [@]yáTis)
cyarí-muc 'cockroaches'
TaTáaru-muc 'June beetles'
(sg. TaTáaro)
neo-plural
(*)cyhúmo-muc 'fishes'
(older speakers have homophonous sing. and pl. cyhúmo)
All Class x pl (copula /bi/)
#Burushaski
'mouse' appears as both gírkis (pl. gírkisyo)
and gírkic (pl. gírkicyo)
If my idea about PL-markers being related to noun-class markers is correct, this situation may be explained as deriving from an original
sg. gírki(-)s, pl. gírki(-)c
compare isé 'this,' icé 'these'
#Burushaski
consonant <c> has the value [ts]
'shepherd'
huyélttarc
pl.: huyélcaro
Perhaps originally
stem /huyélcar-/ (compare huyés 'cattle'),
pl. simply added -o, while the singular is from *huyélcar-c, via dissimilation of the affricates
