Date created: | Last Updated:

Category: Project

Description: Speakers and listeners can process frequent sound patterns more easily and accurately than less probable ones. We hypothesized that frequent sound patterns thus also have a selective advantage in the cultural evolution of language and favour sound changes that (re-)produce these patterns. We tested this hypothesis in a quantitative corpus study on a Middle English sound change known as Open Syllable Lengthening (OSL). OSL lengthened short vowels in originally disyllabic words that became monosyllabic, but not in words that remained disyllabic. We determined the prosodic structure of monosyllables and disyllables that already existed at the time of the change, and analysed if OSL made the affected words more similar to the dominant prosodic patterns. Our results largely confirmed this prediction. Indeed, OSL made the prosodic structure of the newly emerging lengthened monosyllables conform to the most frequent prosodic patterns of already existing monosyllables. Similarly, the failure of OSL to affect disyllables can be explained by the finding that most morphologically simple disyllables at the time of the change had in fact short vowels. In contrast, long vowels in disyllables signalled their morphological complexity. Thus, OSL in disyllables would have made them less indicative and may have failed to affect them for this reason. The implementation pattern of OSL thus stabilized and increased the probability of phonotactic patterns that had already been in the majority before. This suggests that lexical probabilities may play a greater role in the actuation and implementation of phonological change than currently known.