Peak shift is a universal result of most discrimination learning: if an organism is taught to respond to one stimulus and not to respond to another stimulus lying along the same perceptual continuum, it shows maximal responsiveness not to the target stimulus but to a stimulus displaced along the continuum in a direction away from the unreinforced stimulus. It also responds more vigorously to the displaced stimulus than it would have to the target stimulus had discrimination training not occurred. This is the phenomenon of behavioral contrast. On the level of preference, peak shift means that if we prefer something, we shall like something a bit more extreme even more. Peak shift and behavioral contrast are probably the most important causes of sound change in language, but this has never been pointed out. The relevance of peak shift and behavioral contrast to phonological change is explained. Their most general consequence is that the phonetic realization of phonemes is usually in a continual state of evolution; because learners systematically overshoot the phonetic realization they are being taught. How peak shift could have caused the Great English Vowel Shift and the changes subsumed by Grimm's Law is explained. Both of these effects involved chains of phonemes that differed in frequency of occurrence along a gradient. In such cases, peak shift and behavioral contrast very quickly bring about massive changes in pronunciation that can affect virtually every word in a language.
There are several fundamental concerns for any attempt to make a Darwinian account of the evolution of languages. The most immediate include: the issue of Lamarckian inheritance, whether selection is the major mechanism of linguistic evolution, and how we can reconstruct the past history of a language and indeed classify the evolutionary entities of language. Each of these mirrors a similar problem of biological evolution, some of which are thought, inaccurately, to have been settled and some of which are instructive live debates within the biological disciplines. This paper defends a fully Darwinian account of cultural change, but not necessarily a neo-Darwinian account, and challenges the need for Knudsen's tacit transmission hypothesis. Hull's discussion of hybridization in evolution and the generality of taxa, and an implication of Croft's interpretation of the replicator-interactor distinction are explored.
Kirby, S. (in press, b): Learning, bottlenecks and the evolution of recursive syntax. In Briscoe, T. (ed.): LinguisticEvolution through Language Acquisition: Formal and Computational Models . Cambridge University Press
the novel advances, the main character’s language is getting more and more complex, and this linguisticevolution reflects a deeper personal evolution. Thus, towards the end of the novel, Z is a more mature and independent woman. Moreover, the topic of