Production, Perception, and Phonotactic Patterns: A Case of Contrastive Palatalization

Abstract

The aim of this study is to contribute to research on positional markedness by providing insight into the mechanisms involved in the neutralization of phonological contrasts. Specifically, the work investigates phonetic and phonological properties of the contrast in secondary palatal articulation (plain vs. palatalized) and demonstrates the crucial role of phonetic factors in the emergence of cross-linguistic phonotactic patterns of palatalization. I first examine the distribution of the plain/palatalized contrast in a number of languages (mainly Slavic, Celtic, and Uralic) and identify the most likely sites of neutralization as well as the common outputs of the process. Using an electromagnetic articulography (EMMA) I then investigate the production of Russian plain and palatalized stops in a variety of word-boundary contexts. I identify the articulatory and acoustic factors likely to affect the perception of the contrast in these environments and test the relevant predictions in a series of perceptual experiments with native and non-native listeners. Finally, the experimentally derived perceptibility scale is employed in a simulated learning situation. The scale reflects the probability of a given segment being perceived by the learner as 'plain' or 'palatalized' in a variety of environments. Given this limitation, the learner fails to acquire a hypothetical target grammar that posits no restrictions on the distribution of the plain/palatalized stops. Instead, s/he constructs a set of novel grammars which, nevertheless, reflect the common patterns of neutralization attested in languages. The results of the study provide a strong argument for a view of positional markedness as an emergent phenomenon arising primarily from inherent limitations on speech production and perception.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 73 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.