If we count how often each word occurs in a text, the largest group is words that occur once. The paper argues that the number of such words placed in different positions--at the beginning or end of a sentence, for example--characterizes an author. Results are given for both English and Greek texts.

Morton’s previous work has sometimes aroused bitter controversy. The present paper is too unfocused, and the conclusions are too vague, to arouse strong feelings. The statistical technique used is essentially to calculate the chi-square for a wide variety of more-or-less related tables, stopping to comment now and again when a particularly high value is thrown up.

A shifting definition of “words that occur once” (in the particular sample under study? in the group of samples? in some wider corpus?), the habit of grouping samples of text to suit the whim of the moment, as well as the sheer number of tests carried out give Morton’s results a slippery and insubstantial quality. The paper is an interesting example of lexical statistics pushed almost to the point of caricature, but it does not make a convincing case for the methods proposed.