i 
The second task was sorting verbs for semantic similarity. Respondents were instructed 
to sort the 25 verb expressions, written on 3 X 5 note cards, into groups with similar meanings. 
Respondents were instructed to make as many or as few groups as they wished, and were 
informed that they could move cards and adjust groupings until they were satisfied. Groups of 
one were acceptable. These data were scored by tabulating the frequency of common - 
appearance of two verbs in the same sorting pile. Thus, the data were organized as a 25 by 25 
similarity matrix, with cell values ranging from 0, indicating that not one respondent thought the 
two verbs for the cell had similar meanings, to 40, indicating that all 40 respondents had placed 
the two verbs in the same sorting pile. 
The third psycholinguistic task was unidimensional scaling of the similarity of the verb 
own to each of the other 24 verbs of possession. The scaling was done on a 15 centimeter, 
end-labelled scale oriented axially such that placing a card near oneself indicated “same 
meaning” as own and placing the card distant from oneself indicated a “very different meaning” 
from own. 
For the first two psycholinguistic tasks, randomizations of stimuli were balanced across 
the eight conditions (generation X sex X culture). The 10 random tape presentation orders and 
the 10 random card sort orders appear in Appendix E. For the third task, stimuli were uniquely 
randomized for each respondent. 
Finally, respondents were asked three novel interview questions. The first two presumed 
that adults explaining or illustrating the meanings of words for children will seek to present the 
archetypal, essential characteristics of the concepts. The first question inquired how the 
respondent would answer a child’s question about the meaning of own. The second inquired 
about the types of things that would be good examples for showing a child the meaning of 
ownership. The third interview question inquired about any changes in the concepts of property 
and owning over the past several generations. Because these questions were not read, the 
presentations were not strictly standardized. 
At the close of the session, respondents were informally debriefed and were encouraged 
to ask questions of the interviewer (Van Stone 1970). Cree respondents were asked to comment 
on the translations and on the syllabic spellings of the Cree words. Only one respondent 
suggested a change in the syllabic spelling of one word. Assurance was given that a brief