Different lexicons make different rivals

Abstract

Analogy-based theories assume that in situations of affix competition, language users create novel word forms on the basis of similar existing forms in their Mental Lexicons ( Baayen et al. 2011 ; Daelemans & van den Bosch 2005; Skousen 1989 ). Interestingly, however, simulation studies employing computational implementations of analogical theories have almost invariably adopted a rather abstractionist view of the Mental Lexicon, representing the word stock of the language, and abstracting away from differences between individual speakers (see, e.g., Arndt-Lappe 2014 ; Chapman & Skousen 2005 ; Eddington 2006 ; Nieder et al. 2021 ). This is a problem because it precludes the possibility of testing a central prediction of analogical theories: if affixes are assigned on the fly on the basis of similar words in the lexicon, then speakers with different lexicons should make different choices. The present paper provides a proof-of-concept study addressing this issue for the form-based rivalry between the two English adjectival suffixes - ic and - ical. Analogical Modeling of Language (AML; Skousen et al. 2013 ) is used as a computational model. On the basis of a survey of the distribution of derivatives in different registers in the British National Corpus, predictions of the analogical model are compared for a simulated speaker with a large vocabulary (including academic words) and a simulated speaker with a small vocabulary that is based mainly on words from spoken language. The statistical analysis of the simulations reveals that, while sharing some basic properties, the two models make very clear – and testable – predictions about speaker differences.