Word frequency distributions and lexical semantics

Abstract

This paper addresses the relation between meaning, lexical productivity, and frequency of use. Using density estimation as a visualization tool, we show that differences in semantic structure can be reflected in probability density functions estimated for word frequency distributions. We call attention to an example of a bimodal density, and suggest that bimodality arises when distributions of well-entrenched lexical items, which appear to be lognormal, are mixed with distributions of productively created nonce formations.

Word Frequency and Predictability Dissociate in Naturalistic Reading

“…To facilitate comparison with surprisal above, frequency effects are represented on a unigram surprisal scale, that is, the negative log prior probability of the word. This measure is equivalent (modulo an additive constant) to the negation of the log frequency scale regularly used in psycholinguistics (Baayen & Lieber, 1996 ; Baayen et al, 2016 ; Balota & Chumbley, 1984 ; Carrol, 1967 ; Demberg & Keller, 2008 ; Norris, 2006 ; Rumelhart & Siple, 1974 ). Unigram surprisals are estimated using KenLM (Heafield et al, 2013 ) with default hyperparameters.…”

Word Frequency and Predictability Dissociate in Naturalistic Reading

“…To facilitate comparison with surprisal above, frequency effects are represented on a unigram surprisal scale, that is, the negative log prior probability of the word. This measure is equivalent (modulo an additive constant) to the negation of the log frequency scale regularly used in psycholinguistics (Carrol, 1967;Rumelhart and Siple, 1974;Balota and Chumbley, 1984;Baayen and Lieber, 1996;Norris, 2006;Demberg and Keller, 2008;Baayen et al, 2016). Unigram surprisals are estimated using KenLM (Heafield et al, 2013) with default hyperparameters.…”