(R) as a Variable

Abstract

All sounds are variable, but some are more variable than others. Does hyper-variation mean a greater disposition for sociolinguistically relevant conditioning, or, alternatively, a tendency for relatively greater noisiness in the distribution of unconditioned variants? Whatever the case, there should clearly be a special interest in the sociolinguistic systemization of those sounds that are so unusually prone to variation that it is difficult to capture them within a simple articulatory and acoustic definition. Such is the case with the sociolinguistic variable (R).