Recent quantitative work on the variable [g]\~{}[ŋ] alternation in compounds of certain dialects of Japanese has revealed token frequency of the compound as a whole, and of the compound’s second member in its freestanding form, to be important predictors of the alternation. We propose a formal phonological analysis that integrates usage-based factors like frequency with the action of the phonological grammar, extending mechanisms of lexicon–grammar interaction previously proposed in the context of Lexical Conservatism. We demonstrate that our model fits the experimental data better than – or at least comparably to – a theoretically naïve statistical model proposed in previous work. Based on the success of our modelling, we discuss the role of token frequency in phonological patterning more broadly, and how the mechanism that we propose might be extended to unify a range of contradictory frequency-dependent processes that have been observed in the literature.

This article explores the role of token frequency and global prosodic length in conditioning optional paradigm uniformity in Japanese voiced velar nasalisation, with data from two wug-tests carried out with speakers of Tōhoku Japanese. Experiment 1 demonstrates that frequency-conditioning observed in corpus data is reproduced in existing and novel compounds and holds at the level of the speaker. Experiment 2 focuses on a typologically unusual pattern where overall compound length in mora seems to influence nasalisation, a candidate for a ‘counting’ pattern in phonology. We find that instead of overall length, speakers are sensitive to the length of the second member of the compound, undercutting the viability of the mora-counting analysis. We discuss the importance of the results in adjudicating between existing models of how token frequency impacts the phonological grammar and suggest that only theories that allow individual morphemes to exhibit frequency-sensitive behaviour are sufficiently expressive to model the finding.

This article examines a case of phonological opacity in Uyghur resulting from an interaction between backness harmony and a vowel reduction process that converts harmonic vowels into transparent vowels. A large-scale corpus study shows that although opaque harmony with the underlying form of a reduced vowel is the dominant pattern, cases of surface-apparent harmony also occur. The rate of surface-apparent harmony varies across roots and is correlated with a number of factors, including root frequency. These data pose problems for standard accounts of opacity, which do not predict such variation. I propose an analysis where variation emerges from conflict between a paradigm uniformity constraint mandating that the harmonising behaviour of a root remains consistent, and surface phonotactic constraints. This is implemented in a parallel model by scaling constraint violations according to certainty in a root’s harmonic class. This aligns with past work suggesting some opacity is driven by paradigm uniformity.