124
replicated in the card sort and the scaling tasks respectively are own having for the Cree more
of a meaning of want (U =510, p <.005; U =424, p<.001), need (U=565, p <.05; U=464, p<.001),
and deserve (U=594, p <.05; U=564, p<.05). Secondarily, for eight verbs there was also a
significant difference between the Cree and English-Canadians on one of the tasks and a
non-significant but replicated direction of difference on the other task. These suggest that own
has a greater meaning for the Cree of claim, cherish, like, and desire, and a greater meaning for
the English-Canadians of possess, buy, protect, and share. This last verb may seem out of place.
However, during the psycholinguistic tasks, some respondents would reason aloud, and several
English-Canadians commented that sharing a thing means that one must own it, that is, sharing
is the prerogative of the owner. The logic for the English-Canadians in reporting share as being
semantically close to own is that sharing entails ownership, not that ownership entails sharing.
It is interesting that the two strongest differences between Cree and English-Canadians
on the meaning of own are on want and need, and secondarily on desire and cherish. These
correspond to the four words constituting the covetousness cluster In Study 2, and suggest that
perhaps own has a greater meaning of covetousness for the Cree than for the
English-Canadians. Indeed, of the verb clusters described in Study 2, the Cree and the
English-Canadians differed only on covetousness on both the card sort (U=502, p =.004) and
the scaling task (U = 384, p= .0001). As will be discussed, this finding might indicate the need
to reconsider “covetousness” as a label for the group want, need, desire and cherish.
Such an imposition of English-Canadian verb structure on the Cree has some statistical
justification. Table 13 shows the Cronbach alpha coefficients for the six verb clusters identified
in Study 2, here based on the unranked scale judgements. The verb clusters of acquisition,
attachment, stewardship, and covetousness all have alpha coefficients greater than .55 for the
Cree as well positive item-total correlations among their constituent verbs. However, although
dominion has an alpha coefficient greater than .55 for the English-Canadians, alpha was near
zero for the Cree, with no significant relationships between the constituent verbs claim, keep
and control on the item-total correlations. Thus, dominion was not just low in relative
importance to the Cree concept of owning, as hypothesized, but it disappeared as a construct
altogether. This occurred, despite the fact that control, claim, and keep were similarly close in
meaning to own for the Cree as for the English-Canadians.