¹ Charles E. Osgood, George J. Suci, and Percy H. Tannenbaum, The measurement of meaning (Urbana, 1957).

² Uriel Weinreich, ‘Travels through semantic space’, Word 14.346–66 (1958).

³ Donald D. O'Dowd and David C. Beardslee, College student images of a selected group of professions and occupations, unpublished report, Cooperative Research Project no. 562 (8142), United States Office of Education, Department of Health, Education, and Welfare (1960).

⁴ Ithiel de Sola Pool, ‘Trends in content analysis today: A summary’, Psycholinguistics 305–11 (ed. Sol Saporta; New York, 1961).

⁵ Actually, since the exponent of our expression was derived from a count of nonoccurrences, the theoretical median distance to the next occurrence is one more than the exponent. But because the definition of units (see below) allows two units to occur at one position and thus makes possible an empirical distribution with a median of zero, it seemed best not to complicate the association measure by subtracting 1 from the median.

⁶ A special case arises when the end of the text is reached before any occurrence of the target word. The resulting distances—to the position one beyond the end of the text—are put in a special category such that they may increase but not decrease the median, the adjustment starting with the highest such distance. Where all distances are in this category, the association is taken to be —.5.

⁷ Op.cit. 332–5.