To examine the relationship between words that pinpoint or help to define other words, an experiment was designed which reversed the procedure used in collecting the Michigan Restricted Associated Norms. The Michigan procedure used stimulus words to elicit responses. These responses were used as clue words to determine their power to stimulate the target words. Some 36 university students were randomly assigned to one of three experimental groups. Subjects in two groups were given either all infralogical or all logical clues, and the third group received mixed clues. Clue words were selected in six different categories of associations: (1) superordinate, (2) subordinate, (3) similar, (4) location, (5) part, and (6) preceding. Results showed the power of clue words varied according to association strength, number of clues, type of clue, size of respondent set, and reaction time. This report describes in detail experimental procedures and provides a summary of results. (EMH)