Riggs (1893: 30) describes two defective verbs of being:
'there is; ...are' |
'be' |
compare w/ normal II conjugation verb: 'finish' |
||||||||||
| SG | DU | PL | SG | DU | PL | SG | DU | PL | ||||
| 1 | -------- | uŋkaŋ | uŋkaŋpi | -------- | uŋyakóŋ | uŋyakoŋpi | mdustaŋ | uŋstaŋ | uŋstaŋpi | |||
| 2 | -------- | dukuŋpi | dakanoŋ | dakonoŋpi | dustaŋ | dustaŋpi | ||||||
| 3 | yukaŋ | yukaŋpi | -------- | yakoŋpi | yustaŋ | uŋstaŋpi | ||||||
The behaviour of the first verb appears to be semantically motivated, in that it cannot be applied to singular animates; the 3rd singular form yukaŋ 'there is' applies only to inanimates, or to animates providided they are collectives.1 The existing forms of the paradigm are morphologically regular (on analogy with 'finish', we should expect 1st singular *mdukaŋ and 2nd singular *dukaŋ).
The second verb, however, is unusual, and the gaps are not in any obvious way semantically motivated. Morphologically, the extent forms are irregular. Boas & Deloria (1941: 98) suggest the verb was originally a compound, composed of a verb yakha plus the objective form of ʔũ 'it exists for X' (the paradigms below are given in Boas & Deloria's orthography, which differs from that of Riggs):
'be' |
underlying composition per Boas & Deloria |
'it exists for X' |
||||||||
| SG | DU | PL | SG | DU | PL | SG | DU | PL | ||
| 1 | -------- | ʔũyakhũ | ʔũyakhũpi | -------- | ʔũyakha-ʔũ | ʔũyakha-ʔũ-pi | mũ | ʔũkhũ | ʔũkhũ-pi | |
| 2 | dakhanũ | dakhanũpi | dakha-nũ | dakha-nũ-pi | nũ | nũ-pi | ||||
| 3 | -------- | yakhũpi | -------- | yakha-ʔũ-pi | ʔũ | ʔũ-pi | ||||
By this analysis the 2nd and 3rd person forms follow directly if we assume that the element yakha is inflected as a II conjugation verb, and ʔũ shows the regular objective forms. But in the case of the 1st person, we have to assume that the element ʔũ remains uninflected for person. This analysis does not explain defectiveness of the extent paradigm; even given the possible indeterminacy of the 1st person singular, the 3rd person singular should be regularly derivable through deletion of plural -pi; thus *yakhũ.
1 Sullivan (1988: 197-8) describes the same phenomenon for the verb mani 'be extended' in Classical Nahuatl, which cannot be used to refer to a single man (or animal), and thus likewise is missing 1st and 2nd singular.
Boas, Franz & Ella Deloria. 1941. Dakota grammar [Memoirs of the National Academy of Sciences, vol. 23. no. 2.]. Washington: National Academy of Sciences.
Riggs, Stephen Return. 1893. Dakota grammar, texts and ethonography [ed. by James Owen Dorsey]. Washington: Government Printing Office.
Sullivan, Thelma. 1988. Thelma D. Sullivan's compendium of Nahuatl grammar [transl. from the Spanish by T. Sullivan & N. Stiles; ed. by W. Miller & K. Dakin]. Salt Lake City: University of Utah Press.