Adamantios Gafos is Professor of Linguistics and Excellence Area of Cognitive Sciences at Universität Potsdam. He received his PhD in Cognitive Science from Johns Hopkins (1996). He has taught at the University of Massachusetts-Amherst, Yale, Massachusetts Institute of Technology, and New York University (till 2011) and has held a number of visiting appointments, as an International Chair at Université Sorbonne Paris cité, as Professor at the GLOW school of Linguistics, the ÉNS Département d’Études Cognitives, and the Utrecht Institute of Linguistics, among other institutions. He is also a Senior Scientist at Haskins Laboratories. [His current work is supported by grants from the European Research Council, Advanced Grant 249440, and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation), Project ID 317633480, SFB 1287.]
Sound Patterns In Language: Information, Dynamics, Rules
As we are nearing 100 years since Edward Sapir’s 1925 article in the first volume of Language (doi 10.1111/j.1467-1770.1956.tb00847.x), some of his words still resonate strongly today. Sapir posed the problem of the ‘psychology of speech sounds’, emphasizing that phonemes could not be equated with their sensorimotor substance and highlighting the key roles of the ‘inner configuration of the sound system of a language’ and the ‘placement of the sounds with reference to one another’. I revisit Sapir’s ideas by examining some of the issues brought up in his work, focusing on the plausibility of drawing a distinction between vocal tract speech versus (vocal tract) non-speech action, the notion of information in learning of words, and the patterning of sounds in languages (phonological rules). Each of these issues is met in corresponding present day debates: on if and how a speech-specific mode of vocal tract action separate from non-speech action is identifiable, on the mechanisms enabling infants’ learning of words, and on the forces or biases contributing to the sound patterns of languages (and whether these can be accessed experimentally in so-called artificial grammar learning experiments). I will sketch some resolutions for these debates and I will conclude by highlighting some remaining challenges.