It is widely accepted that events that occur frequently will be recognised more rapidly. For example, it is believed that words that occur frequently in the language will be recognised more quickly than otherwise similar words that occur less frequently in the language (the “word frequency effect”). Frequency effects are predicted by most connectionist and neural network models that are intended to explain the psychological and neurobiological bases of learning and memory.

The present research argues that this common view is wrong. Instead, we argue that the relevant factor is not the frequency of a word in a language, but rather the number of different contexts in which it occurs (its “contextual diversity”). The frequency and contextual diversity of a word are very highly correlated (such that high frequency words tend to occur in more contexts), so previous experimental results that were taken as evidence for word frequency effects can potentially be reinterpreted in terms of contextual diversity.

The research employs new analyses and experiments to determine whether frequency or contextual diversity is the main factor underpinning speed of word recognition. A demonstration of the importance of contextual diversity rather than frequency would have important theoretical and practical implications.

Start date
01 December 2005
End date
30 November 2006
Grant holder
Gordon Brown
Dr James Adelman
Grant amount
Grant reference
Grant type