A Study on Korean Learners’ Pronunciation of Final Consonants

This study analyzed the pronunciation of Korean final consonants by Chinese, Thai, French, and Japanese speakers, and examined the reasons and aspects of the pronunciation.
Korean language learners find it difficult to, and consequently fail to, pronounce Korean final consonants correctly. Previous research has explained that the failure is due to the difference in syllable structure between Korean and the learners’ native languages. The research suggested that Japanese or Chinese speakers are unable to pronounce final consonants perfectly because Korean has a closed consonant-vowel-consonant (CVC) syllable structure with final consonants while Japanese and Chinese have an open CV structure without final consonants. However, the difference between a closed syllable structure and an open syllable structure does not clarify the reasons and aspects of the learners’ pronunciation difficulties. Therefore, this study considered that the difference in the learners’ pronunciation of final consonants results in constraints of onsets which domain-final empty nuclei can license, deviating from the previous approach to the syllable structure difference, which depended on the existence of final consonants. Based on constraints of domain-final empty nuclei, and principles of syllable structure, this study investigated the difference in pronunciation of final consonants between Chinese and Japanese speakers, examined the pronunciation of Korean final consonants by French speakers, who pronounce the final consonants as plosives, and Thai speakers, who neutralize all the consonants except /ㄹ/.
The pronunciation and aspects of Korean final consonants and the Korean final consonants of Chinese, Thai, French, and Japanese speakers were analyzed based on constraints of domain-final empty nuclei and principles of syllable structure. The findings are summarized as follows. First, the pronunciation of Korean final consonants is unreleased according to neutralization rules. The constraints of final-domain empty nuclei demonstrated that onsets followed by empty nuclei must not have /H/ as an element but contain /?/ as an element. It is found that tensification, nasalization, and lateralization of the final sound of single consonants are the results controlled by inter-onset government relations.
Chinese speakers had a strong tendency to pronounce Korean final consonants without making the sounds of the obstruent /ㄱ, ㄷ, ㅂ/. The analysis of this phenomenon based on Minimal Phonological String rules in Chinese phonology and P-licensing Constraints Beijing Mandarin demonstrated that the sound of the vowel spreaded in the position of the final consonant deletion, and its duration extended. The length of the vowel sound was measured to verify whether the pronunciation of native Chinese speakers exhibits the phenomenon. The results showed that the length of the sound was approximately twice as long when the final consonant was deleted. This indicates that Chinese has minimal phonological strings, and is a language with a syllable template that consists of four skeletal points. Tensification, nasalization, and lateralization of the final sound of single consonants were not present because of the tendency to delete the obstruent. Moreover, lenition did not occur in the pronunciation of Chinese speakers.
Thai speakers pronounced the obstruent by applying neutralization rules like in Korean; however, they pronounced the lateral /ㄹ/ as [ㄴ], following constraints of domain-final empty nuclei. Thai and Korean have the most similar pronunciation rules for final consonants as compared to other languages. However, tensification, nasalization, lateralization, and lenition were not found in the pronunciation of single consonants in Thai. This is because there is no inter-onset government relation, specifically between a coda and a subsequent onset. Thai speakers were observed to have a tendency to pronounce each syllable as is, which implies that empty nucleus dominant principles are applied to maintain licensed empty nuclei in Thai.
French speakers pronounced Korean final consonants as plosives because there were no specific constraints on onsets licensed by domain-final nuclei. In other words, onsets can have all the consonants including single ones, and they are released because no neutralization rules are applied unlike in Korean. French also does not have an inter-onset government relation and therefore, no tensification, nasalization, and lateralization are found in the pronunciation of final consonants; the original sound of each syllable is pronounced. In addition, French has a strong tendency to maintain licensed empty nuclei.
Since Japanese does not license domain-final empty nuclei, Japanese speakers pronounced Korean words containing an empty nucleus with [u], combining an element U and a cold vowel. This tendency is especially strengthened when obstruent /ㄱ, ㄷ, ㅂ/ and lateral /ㄹ/ are codas. Moreover, like Chinese speakers, the Japanese delete the obstruent at the coda. Other speakers frequently pronounce with an epenthetic vowel, however, some Japanese speakers delete the obstruent that comes at the coda. Japanese syllables tensify, and the final sound tends to become segmental due to the effect of the subsequent syllable, though the place of articulation for both sounds is the same. The tensification is the result of the inter-onset government relation as in Korean. However, the final change to the segmental sound made at the same place of articulation occurs because the coda element is spreaded to the following syllable onset. Nasalization and Lateralization do not occur due to a vowel inserted in the pronunciation as shown in the cases of pronouncing the obstruent that is the final sound of single consonants.
The results of this research are as follows. Korean learners’ pronunciation of final consonants is determined not by the difference in the syllable structure, whether it is open or closed, but by the onsets that domain-final empty nuclei can license, which is distinct in each language. There are limitations in describing final consonant pronunciation by Korean learners with an open or closed syllable structure; however, it can be analyzed and explained by onsets that domain-final empty nuclei can license. With this approach, this paper discussed the reasons and aspects of final consonant pronunciations made by Korean learners.
The objective of this study was to analyze and explain final consonant pronunciation by Korean learners based on constraints of domain-final empty nuclei in government phonology. The analysis shows that Korean learners’ final consonant pronunciation is different from that of Korean native speakers because of language conditions such as whether each language licenses empty nuclei or whether licensed empty nuclei can license onsets in different conditions. In addition, it described the actual pronunciation of Korean learners in related phonologic phenomena with the spreading and composition of elements.