A theory of interpretive clustering in free recall

The existence of a polysemy effect in episodic memory is demonstrated through an analysis of data from the experiments of Lohnas et al. (2015) and Healey and Kahana (2016). Three word-length related features are reported: (1) the average distance between the serial positions of consecutively recalled words is an increasing function of the second word’s length; (2) the recall of words in the order in which they were presented is more likely to occur for shorter words; (3) longer words are easier to recall than shorter words from the same list. These phenomena are reproduced by modeling the fact that shorter words have a larger number of meanings (polysemy). This is shown to be at the root of our tendency to recall verbal sequences in forward order, and a conjecture is put forth regarding the general methods we employ to store information on the chronology of events.