Phonological Trends in English Loanword Word-initial Tensification in Korean

Abstract

The main purpose of this thesis is to investigate the phonological trends in word-initial tensification of English loanwords in Korean, and to explore the reasons for the trends based on the phonological distribution of native phonology. In general, English voiced stops and affricate are adapted to Korean as voiceless lax stops and affricate, respectively. However, for some English voiced stops and affricate, not only a lax initial form but also its tense counterpart may appear. The majority of the previous accounts have focused on the origin of the tensification in question, rather than on the phonological condition of the tensification. This is likely because the word-initial tensification of English loans is widely regarded as an idiosyncratic process that cannot be characterized by a phonological rule. To date, no previous studies have provided comprehensive explanations on the gradient tendency of the tensification. Based on two different data sources, a judgment test, and newspapers from the 1890s to 1950s, I found that loanword tensification is more likely to occur when the place of articulation of the tensification site is alveopalatal, when the height of the vowel following the tensification site is non-high compared to high, when the word is monosyllabic rather than multisyllabic, and when the phonation type of the onset of the syllable following the tensification site is the tense fricative. I refer to these findings by four phonological conditions as place, height, length, and assimilation effect, respectively. Given that the various patterns shown in loanwords may be influenced by their native phonology, two possible sources were investigated to explore the underlying reason for each effect. The first source was native tensification that shows optional tensification in a similar way to loanwords. If loanword and native tensification both show similar tendency by each effect, it might be possible that the loanword tensification is simply an extension of the active process in native tensification. Second, based on various arguments claiming that the variable patterns found in loanwords from other languages may reflect covert statistical generalizations of their native lexicon (Kubozono, 2006

Luke and Lau, 2008

Zuraw, 2010

among others), it is possible that the tendencies found in loanword tensification may also reflect statistical trends displayed by Korean common nouns with initial tense stop and affricate. To verify these possibilities, I investigated the contextual distribution of word-initial tense stop and affricate in the Korean lexicon. Based on an in-depth examination of the two sources, I found that the phonological trends in loanword tensification partially mirror phonological trends in both native tensification and the Korean lexicon. The place and height effect of the loanword tensification were also confirmed in native tensification. On the other hand, the height, length, and assimilation effect of loanword tensification mirror the extant distribution of Korean common nouns. In addition, I discussed the phonetic background that motivates each contextual factor found in loanword tensification, focusing on the acoustic and articulatory properties of each context. For the height effect of loanword tensification, the relation between the voice onset time (VOT) value of the word-initial stop and the following vowels height was considered. Regarding the assimilation effect, it seems that the effect requires additional articulatory effort which means that it does not have clear phonetic motivation. Rather, the assimilation effect substantially mirrors the salient distribution of the Korean lexicon. In summary, this thesis has scrutinized phonological trends in loanword tensification based on two different datasets. Based on quantitative data, I not only systematically examined the trends, but also found new phonological factors that affect tensification rate. In addition, it is the first to explore the question of what phonological aspects drive the contextual distribution of loanword tensification, which has not been discussed in detail in previous studies. As a result, this study demonstrated that the distribution of loanword tensification reflects that of the native phonology.