noun physical_somatic_body_part

Proto-Siouan-Catawba *Roksi ~ *Rǫksi (?)

Proto-Siouan *Roksí

Proto-Crow-Hidatsa *rohcÉ

Crow daččuá ‘armpit’ RG, GG:43

Hidatsa rohcÉ ‘his, her armpit’ J , nohci


Mandan aksóxarąre ‘armpit’ H:59

Proto-Mississipi-Valley *Roksí

Proto-Dakota *Roksí

Lakota alóksohą ‘armpit, put under the arm; to carry under the arm’ EB:78a

Dakota doksí ‘armpit’ SRR:109a


Chiwere ró•θe ‘armpit’ [old] JDH , róʔθi JGT , roqçi , †roxθi

Otoe roçi , †roθi

Proto-Dhegiha *Rosǘ RR

Omaha-Ponca nosi

Kanza/Kaw dosǘ ‘armpits’ RR

Osage thuçí , †ðosǘ (sg.), †dosü (pl.) ‘armpit’ LF:232a

Quapaw tosí JOD



Biloxi tuksiⁿ´ D&S:281b

Ofo tạ́fhe , †táfhe D&S:329b


Catawba hinų́ksu ‘his armpit’ FS

General comment

Virtually all instances where Quapaw #do- is expected show #to-.

This appears to be a regular sound change as there is only one exception to this in the entire Quapaw lexicon. Osage appears to show a similar change in progress with singular/plural listed differently (as h/d = ð/t).

Ofo fh once again appears to have a *C + s > hs > sh > fh antecedent. The final vowels of this set present a difficulty that has not been resolved. In DH a former compound of ‘armpit’ with *hubone’ is possible but not obviously justified. Vowel harmony of a sort is occasionally found in Kanza/Kaw, but -u# is pervasive here and is found even in Catawba. If this noun is like most body parts, it should have been inalienably possessed, i.e., had a prefixed *i-. The initial consonants of the Ofo and Biloxi forms are incompatible unless Proto-Siouan *R has these reflexes in the SE. Nor is nasalization in the SE accounted for in this set. Crow shows rightward vowel exchange and the addition of a final syllabic.

Details Language Word Source