This paper is part of an attempt to develop a proper balance between perception and production for phonetic and phonological theory. Deplorably, most past and current approaches have been developed for, and from, production data and they slight or disregard perception. It is argued that according to such approaches sound systems would not even be learnable. It follows that perception and production need to be integrated. In fact, it is suggested that, in many respects, perception is primary, because it controls production and determines the nature of the basic functional units. Consequently, this new approach is termed perception-based phonology (PBP). Its central assumption is that in addition to the anatomical prerequisites sound systems have their roots in the non-language specific and non-species specific properties of the auditory system.

As for this paper I summarize the arguments that the ontogeny and phylogeny of distinctive features, phonemes, and their typologyy derive from perception rather than production. The key evidence is the perceptual ability of neonates and infants before the onset of speech.

In: J. Leather & A. James (Hg.), Proceedings of the Third International Symposium on the Acquisition of Second-Language Speech, New Sounds 97. University of Klagenfurt, 343-350, 1997.