3 
Gift, were judged to be between "strong" and "very strong" 
on the judgement scale. All of the other criteria were 
judged to be "weak" or "very weak". Familiarity was valued 
least as an argument for ownership, and Knowledge almost as 
low. What is notable about this is that the three criteria 
most valued as arguments for ownership all describe modes of 
acquisition. . And two of those, Crafting and Gift, were 
ranked very low by the judgements of criteria based on the 
recall tasks. As hypothesized, implicit and explicit 
judgements appeared to be quite discrepant. As shown in 
Table 10, there was not a significant correlation for even 
one subject between rankings by explicit judgements and 
rankings by either judgements of owned exemplars or 
difference scores. The only pattern of significant 
relationships was between judgements of owned exemplars and 
of difference scores. 
Clusters of Criteria 
Cluster analysis was used because the study was an 
initial exploratory attempt at identifying some of the 
criteria entailed by ownership. The initial compiling, 
labelling and expressing of 40 plausible criteria, and the 
selection of a reasonable sub-set of 12 of those was all 
done rather uncritically, without prior expectations. In 
such situations, cluster analysis (or factor analysis) is 
useful in identifying criteria which are responded to