Katayama 1998

Home Research SLUG Pubs Abstracts Katayama 1998

Optimality Theory and Japanese Loanword Phonology

Motoko Katayama
UCSC PhD dissertation, 1998

The main goal of this dissertation is to propose an Optimality Theoretic model of loanword phonology which is free from special assumptions on inputs; this model is based on the idea of "Richness of the Base" (Prince and Smolensky 1993). The central proposal concerns consonant gemination in Japanese loanwords. I claim that consonant gemination occurring in Japanese loanwords is driven by an attempt to preserve the moraicity of coda consonants. It has previously been difficult to maintain such a claim because inputs to loanwords have been assumed to have no phonological structure specified. I will introduce Sympathy Theory (McCarthy 1997) into loanword phonology, and demonstrate that consonant moraicity preservation is achieved without invoking any syllabification on inputs, since such preservation holds not between the input and the output but between the so-called sympathetic candidate and the output.

Loanword phonology has been a challenging topic to pursue due to our vague understanding of underlying representations of loanwords, which has often stipulated special markings on inputs to derive systematic phonological patterns. By introducing Sympathy Theory and evaluating all possible outputs in parallel in my model, it becomes possible to remove all stipulative input markings, and respect the concept of "Richness of the Base" which attributes systematic phonological patterns to constraint rankings, not to differences in inputs. In my model, there is no need to restrict the input. Multiple inputs can converge into an identical output within the same grammar. This is a great advantage in a realm where it is hardly possible to justify underlying representations.

Chapter 1 gives an overview of the theoretical framework, previous approaches to loanword phonology, and my proposal. In Chapter 2, I discuss phonological changes caused by segmental and phonotactic constraints in Japanese, and introduce the basic notion of constraint ranking and parallelism in Optimality Theory. Chapter 3 examines consonant gemination, and demonstrates that Sympathetic Correspondence is necessary in loanword phonology to obtain a truly output-oriented analysis. Chapter 4 and Chapter 5 investigate sequential restrictions and accentual patterns respectively, both of which suggest that constraint reranking is needed. It will be shown that examination of loanwords results in the discovery of constraints or constraint rankings which are masked in the native vocabulary of the language.

email the author at katayama at japan.email.ne.jp

back to the SLUG Pubs home page