Dr. Paula Orzechowska, Adam Mickiewicz University, Poland
This talk provides an analysis of initial and final consonant clusters in English. On the basis of empirically observed tendencies of distribution and co-occurrence of distinctive features in segments forming clusters, we propose a set of new phonotactic preferences which govern the word structure in English. These preferences are expressed by a set of parameters that pertain to the following dimensions: complexity, place of articulation, manner of articulation and voice. With help of statistical methods of Principal Component Analysis and Cluster Analysis, the proposed parameters are assigned weight, which allows to illustrate that different features play a different role in initial and final clusters in English. This paper advocates the view that phonological features rather than segments as such determine phonotactics in a given language, while statistical methods allow us to determine the ranking of preferences. The analysis reveals the following set of phonotactic preferences established for initial clusters: sonorant C cluster-finally > 1 obstruent C in CC > voiced C cluster-finally > increasing sonority (where '>' means 'stronger than'). For the word-final context, we show that apart from the manner of articulation and voicing features, also the feature of place (coronal C cluster-initially) is relevant.
The above abstract is a part of the article which was accepted at The International Conference on Current Issues of Languages, Dialects and Linguistics (WWW.LLLD.IR), 2-3 February 2017, Iran-Ahwaz.