Honeybone – Does word frequency affect phonology? Reasons to be cautious (week 1)

The question in the first part of this course’s title is arguably the single most important one that currently faces formal (structure-based, generative-type) phonological theory. As we will see on this course, some phonologists argue that the phenomena known as ‘frequency effects’ show that formal phonology is mistaken in assuming that there is a categorical level of underlying representation (and indeed that there is categorical phonological structure at all). Instead, it is argued, the existence of such phenomena shows that the phonological lexicon is made up of exemplar clouds of stored forms of whole words and phrases, as argued for in usage-based models of phonology. As will be clear from the second part of the title, however, this course will draw a different conclusion. We will first investigate what is meant by a ‘frequency effect’ and see that there are a number of types that have been proposed in the literature. We will then consider what implications such phenomena might have for phonological theory if taken at face value. We will then take a step back and consider the basis of the data: are frequency effects real? We will see that some claims do not stand up (so the issues are more complex than is sometimes assumed) but that others do: there are indeed frequency effects which affect the forms produced in speech. Finally, we consider the full balance in terms of the predictions of both formal models and usage-based models when it comes to frequency effects: which approach stands up best? The conclusion will be that, in fact, formal models stand up best, in part because certain types of frequency effects are fully compatible with formal models (and others may well be), but also because we need to recognise a hitherto little discussed type of phenomenon: ‘categorical frequency effects’, which only make sense in the light of the predictions of formal models (supplemented by an understanding of the interaction of synchrony and diachrony).


Work by Joan Bybee has been highly influential in establishing the discussion of frequency effects, arguing for a usage-based approach to phonology. Her ideas can be found in work such as:

  • Bybee, Joan (2001) Phonology and language use. Cambridge: Cambridge University Press. Chapter 1 sets out the basics.
  • Bybee, Joan (2002) Word frequency and context of use in the lexical diffusion of phonetically conditioned sound change. Language Variation and Change 14: 261-290. available here.

Work by Janet Pierrehumbert has also been highly influential, from a perspective that is similar to Bybee’s. Her ideas can be found in work such as:

  • Pierrehumbert, J. (2001) Exemplar dynamics: Word frequency, lenition, and contrast. In J. Bybee and P. Hopper (eds.) Frequency effects and the emergence of lexical structure. John Benjamins, Amsterdam. 137-157. available here.
  • Pierrehumbert, Janet (2002) Word-specific phonetics. Laboratory Phonology VII, Mouton de Gruyter, Berlin, 101-139, available here.

A response to such work has come from work such as the following, which seeks to draw out the predictions of exemplar approaches in terms of frequency effects and finds that the predictions are not met:

  • Dinkin, Aaron (2008) ‘The Real Effect of Word Frequency on Phonetic Variation’. Penn Working Papers in Linguistics 14.1, available here.
  • Tamminga, Meredith (2014) ‘Sound Change without Frequency Effects: Ramifications for Phonological Theory.’ Proceedings of the 31st West Coast Conference on Formal Linguistics, 457-465, available here.
  • Bermúdez-Otero, Ricardo (2016) Phonology and the lexicon: a tutorial. Part I: a survey of current issues. Handout, Phonological Theory Agora Meeting 3, Tours, 14 October 2016, available here.