SEARCHING THROUGH WORD LISTS

Two experiments are reported in which subjects searched through lists of words, looking for targets defined only in terms of their meaning (e.g. ‘any animal’). Under these conditions, scanning is much slower than when the target is a known word, or a member of a small fixed set. It is argued that printed words are processed in two stages, termed ‘stimulus examination’ and ‘memory examination’ respectively. Preliminary data relevant to other aspects of the scanning method are also presented: error rates, the importance of motivation, the extent to which words scanned over can be recognized later, and the relation between scanning and reading aloud.