This year, phono-classes (except for N Faust’s) are organized around the idea that the relationship between phonology and phonetics is arbitrary. This claim is motivated, and its consequences for phonological and linguistic theory are evaluated. The background is a converging movement coming from quite different quarters: substance-free phonology, phonetics, OT, Government Phonology.

One consequence is that melody is substance-free, i.e. melodic primes are not given at birth but are the result of an environment-based construction where children rely on their domain-unspecific ability to categorize. In this environment, the substance only comes into the picture when phonology is converted into phonetics, a relationship that relates items like a multilingual dictionary and is therefore arbitrary.

Another consequence is that phonological computation is arbitrary as far as substance is concerned: anything can be turned into anything in the context of anything (the converting mechanism is not arbitrary, though).

A further consequence is that markedness constraints in OT are not universal and given at birth, but emergent. They are shaped by the particular linguistic environment and may thus diverge from language to language. This means that humans are not given any pre-specification for what is natural or unmarked, and hence grammar is not working under the pressure to produce a cross-linguistically unmarked result.

Finally, the arbitrariness mentioned echoes the GP notion of phonetic interpretation, which has a modern incarnation in terms of modularity in the sense of Cognitive Science where spell-out at the lower interface of phonology works in the same way as it does at the upper interface.