Every language has rules for how sounds combine into words, but no one knows why those rules take the form they do. This project tackles that gap by asking a deceptively simple question: why do the world’s languages all seem to optimise their sound systems in similar ways, and why do they sometimes fail to do so? The researcher proposes that the answer lies in two forces working together: the brain’s need to process sounds efficiently (resource rationality) and the slow, cumulative changes that happen as languages are passed between generations (cultural evolution). By building computational models that simulate how children learn sound patterns, how adults use them, and how those patterns shift over centuries, the project aims to turn sixty years of descriptive observations into a testable, explanatory theory. This is fundamental science. It will not produce a new app or a medical treatment. But a deeper understanding of why phonological systems are the way they are could eventually reshape how we teach languages, design speech-recognition software, or diagnose developmental language disorders—much as understanding the rules of grammar once transformed everything from machine translation to literacy education.
View original technical description
REVOLUPHON aims to transform the theory of phonology by explaining the field's most significant observations of the past sixty years. Decades of research has revealed that phonologies are particularly well described in terms of (i) certain abstract formal representations and (ii) an optimisation process. However, we do not yet understand why. My core hypothesis is that the dominant, descriptive discoveries of phonology since the 1960's will be explained by a new synthesis with (i) resource rationality; and (ii) cultural evolution. (i) Resource rationality is a new theory of cognition that explains the motivations and advantages of cognitive computational strategies, including for when and how to store information. I hypothesise that resource rationality can illuminate the cognitive motivations for classic, descriptively-successful phonological representations. REVOLUPHON tests this hypothesis by modelling the computational tasks demanded by phonology across three successive developmental stages, and evaluating the resource rationality of classic representations in the contexts of those tasks. (ii) Evolution leads to optimisation and phonologies undergo cultural evolution. However, an explanatory link from evolution to optimisation has yet to be made in phonological theory. I hypothesise that a formalised theory of phonological evolution will explain why phonologies (mostly) optimise, how they do, and when they do not. REVOLUPHON tests this hypothesis by: developing new modes of evolutionary analysis and devising computational models that couple cultural evolution to the resource-rational learning and use of phonology, to solve outstanding theoretical problems of (non-)optimisation and to answer entirely new questions about how phonologies evolve. REVOLUPHON charts a bold path that overturns our understanding of the central observations of the field; leads to a new and significantly deeper explanation of them; and founds a vibrant and productive new theory.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know