A Study on the Acquisition of Idioms by Learners of the Korean Language: focused on semantic transparency and syntactic flexibility

The purpose of this study is to figure out the idiom-learning patterns of Korean language learners. More specifically, the study examines how semantic transparency and syntactic flexibility affect idiom acquisition and learning for learners of the Korean language. Previous research on the idiom acquisition of L2 learners has been mainly based on studies which held the premise that there is interference between the native language and the target language, and the native language interference is considered a crucial factor. However, this study is concentrated on discussing how semantic transparency and syntactic flexibility, which are inherent characteristics of idioms, affect acquisition.
The objectives of this study are as follows: First, the study looks to examine the accuracy rate of semantic speculation and Korean language learners'' strategy usage according to the semantic transparency of idioms, assuming the situation in which learners encounter unfamiliar idioms prior to learning. Second, the necessary learning time and learning effects depending on semantic transparency are reviewed, supposing the situation in which Korean language learners learn idioms with different levels of semantic transparency. Third, presuming the situation where Korean language learners re-encounter idioms that they previously learned, this study seeks to find the effect of syntactic flexibility of idioms on idiom recognition. In other words, the accuracy of and the correlation between meaning judgement and grammatical judgement on the syntactic variation of idioms will be examined.
To begin with, a questionnaire survey was conducted for intermediate and advanced level learners in order to consider the semantic speculation aspects and strategy usage of Korean language learners when they encounter unfamiliar idioms. First of all, correct answer rates were analyzed to find the correlation between idiom transparency and semantic speculation. The results of the survey were that the more transparent the meaning of idioms were, the higher the success rate of semantic speculation turned out to be. Also, a high success rate of semantic speculation was found when respondents showed a higher level of Korean language proficiency. Secondly, strategies chosen by survey respondents were analyzed in order to understand idiom transparency and respondents’ tactics. The results showed that overall, "semantic speculation based on the context of the rest of the sentence" and "simple guessing" strategies were the most favored. As for the semantic transparency, the more opaque the meanings were, the more preferred “semantic speculation based on the context of the rest of the sentence," “simple guessing," “inferring the figurative meaning of the underlined parts," and “translating the meaning of individual vocabularies within the underlined parts” strategies were. On the contrary, when the meanings of idioms were more transparent, the strategy of “interpreting the meaning using knowledge of one’s own native language” was preferred, leaving tactics of “semantic speculation based on the context of the rest of the sentence” and “interpreting the meaning by using background knowledge (the Korean language, culture, etc.)” as subsidiary strategies. Such difference in strategy usage was also related to the Korean language proficiency of the respondents. The intermediate respondent group, with a lower level of Korean language proficiency, preferred strategies of “translating the meaning of individual vocabularies within the underlined parts” and “inferring the figurative meanings of the underlined parts." Contrarily, the advanced respondent group, with a higher level of Korean language proficiency, preferred strategies of “semantic speculation based on the context of the rest of the sentence," “interpreting the meaning by using background knowledge (the Korean language, culture, etc.)," and “simple guessing."
Next, a learning experiment was conducted to find the necessary learning time and learning effects depending on semantic transparency when Korean language learners study idioms with different levels of semantic transparency. The experimental results demonstrated that less time was required for learning idioms when idioms were semantically more transparent, whereas more time was necessary for learning semantically opaque idioms. To find posterior effects of learning idioms, a meaning recall test and a lexeme recall test were given to learners. The outcomes of the experiment showed that when semantic transparency of idioms was of a high level, the success rate of the recall test also elevated, and vice versa. Between the two tasks, the success rate of the meaning recall test (a test that gives an idiom as stimulus and requires the learner to recall the meaning of the idiom) was high, whereas the success rate of the lexeme recall test (a test that gives an idiom as stimulus and requires the learner to recall the lexemes within the idiom) was lower than the former. The results showed that as the meaning of the idiom became more opaque, the success rate of the lexeme recall test decreased as well. This implies that idiom transparency affects both the necessary time for learning and the learning effects.
Lastly, the accuracy of context-based ambiguity processing and grammatical judgments were studied so as to investigate how Korean language learners process variants of syntactically flexible idioms. When it came to previously learned familiar idioms, the experimental subjects tended to grasp the meanings of idioms by relying on the context rather than focusing on the modified form of the idioms. For the case of idioms that they had already learned, learners accurately distinguished between whether the idioms were in literal or idiomatic terms depending on the context when they encountered sentences containing unfamiliar variants of the idioms. This shows that even without any information on syntactic variants (even when the information on syntactic variants is unknown), learners do not find any difficulty in the semantic processing of idioms. On the other hand, the grammatical judgement task on syntactic variants showed a low rate of success, approximately 50 percent. For semi-opaque idiom (Type B) variants, the correct answer rate of the grammatical judgement task was higher than that of semi-transparent idiom (Type C) variants. In addition, there was a tendency for more errors in the grammatical judgement task for simple variants, such as predicate or noun modifying, compared to syntactically complex variants, i.e., variants with syntactic movement such as relativization, causativization, and passivization. It was found that subjects felt greater difficulties in grammatical judgement than semantic processing when dealing with syntactic variants of idioms. Therefore, it can be presupposed that information on the syntactic flexibility of idioms plays an important role in understanding and acquiring idioms.
The results of this study are expected to contribute not only to the field of linguistics, but also specifically to the field of Korean language education. The basic data derived from this research may serve as the foundation from which improved learner-centered idiom difficulty lists can be created and effective teaching methods for idioms within the framework of vocabulary education can be discovered.