Since human grading of English writing requires substantial resources, many researchers in the area of Computer-Assisted Language Learning (CALL) have been focusing on automatic scoring systems based on natural language processing systems, machine learning, and other automatic processing mechanisms. English Testing Services (ETS) announced several automatic scoring systems for English writing. In this paper, we suggest using a deep learning based automatic scoring system for an English caption writing test. Our method involves using a sentence similarity measurement, which compares different levels of answer sentences with user writing input. We chose different word embedding types (Word2Vec, Word Mover's Distance (WMD), Bidirectional Encoder Representations from Transformers (BERT)) and Abstract Meaning Representation (AMR), a linguistic model for comparing semantic differences between two sentences based on semantic representation. Scoring systems should not only satisfy the requirements of complicated scoring rubrics but also meet the conditions of a language proficiency test. Our results show that BERT outperforms three competitive models in predicting accurate scoring levels and also shows the characteristics of the criterion reference which could theoretically express the standards of a language proficiency test.

Puns are a type of wordplay with multiple meanings. Although widely used in English, they are not taught in many English classrooms, and it is not easy for students to notice and decipher puns themselves. This study conducted an activity that engaged learners in collaboratively translating puns into Korean. To do this, a total of 130 college students from multiple disciplines participated over a period of fifteen weeks. They watched movie scenes and learned translation strategies (Díaz-Pérez, 2014) for puns from the perspective of Cognitive Relevance Theory. According to this theoretical framework, translations should be based on interpretive resemblance. Students tried to provide optimal relevance so that readers could process puns with minimal effort. The results showed that students used various strategies, but in particular, they preferred to use diffuse paraphrase and editorial technique resulting in a lower degree of interpretive resemblance between the source text and the target text. This study implies that translating puns should be adopted and implemented in English courses. Educational implications are also discussed.

This paper examines Korean learners' understanding of implicative and non-implicative verbs in English. The two types of verbs are distinguished from each other in English by the presence or absence of the entailment of their complement clause. In Korean, by contrast, the validity of the embedded clause does not depend on the verb because the verb itself is not implicative. The implicative meaning arises from complex sentence structures containing certain conjunctive endings. Due to these cross-linguistic differences, Korean learners of English who assume the semantic equivalence of similar corresponding verbs are predicted to rely on their L1 knowledge and fail to distinguish implicative verbs from non-implicative ones. In order to confirm this prediction, this study examined 27 advanced and intermediate Korean learners' understanding of English implicative and non-implicative verbs through translation and entailment checking tasks. The results confirmed the prediction: many participants failed to understand the distinction between the two types, and performed particularly poorly on the implicative verbs. The results of this study provide implications for L2 vocabulary teaching and translation training, namely that explicit instruction is required on the entailment component of implicative verbs.

This study compared and analyzed the word association patterns of Korean L1 speakers and L2 learners based on the characteristics (parts of speech and frequency) of stimulus words and proficiencies of L2 learners. For this, a word association experiment was conducted among 20 L1 speakers and 60 L2 learners. The results showed differences in word associations between L1 speakers and L2 learners; the former had more paradigmatic responses, while the latter had more syntagmatic responses to the stimulus words. Moreover, as L2 learners' proficiency advanced, their paradigmatic responses increased and syntagmatic responses decreased. With regard to associative responses to parts of speech, both groups had a high number of paradigmatic ones to nouns, but showed different levels of associations for verbs and adjectives. Finally, with regard to associative responses to frequency, L1 speakers had paradigmatic ones to the stimulus words at all frequencies, while L2 learners had more syntagmatic responses.

This study analyzes uses of 'hayngpok (happiness)' and 'pwulhayng (unhappiness)', two different expressions of Korean happiness, centering on factors of happiness and linguistic features which lead to a different profiling of senses. Uses of the two expressions are scrutinized based on the Sejong magazine corpus, which is likely to show the general concept of happiness, and three major points are revealed. First, examining elements of happiness mostly supports the findings of previous psychological studies on happiness of Korean people and adds that society at large is another non-negligible factor. Second, between the two senses which 'hayngpok' and 'pwulhayng' share, the sense of emotion is salient with 'hayngpok,' and the sense of luck is salient with 'pwulhayng'. Third, the two words characteristically combine with different kinds of experiencers: 'hayngpok' with an individual experiencer and 'pwulhayng' with a group.

The purpose of this study was to find out the perceptions of English writing and feedback from college students majoring in arts and physical education. Students received feedback on their English writing three times: once from their teacher and twice from their peers. The students were asked to correct their writing after receiving the feedback. The findings were as follows: First, the learners were not interested and confident in English writing, but they recognized the need to write in English. Second, learners wanted to write English reports or resumes that would help them get a job in the future. Third, as the need for writing increased, learners showed different aspects in their preferences for the types of writing class activities. Finally, the learners recognized the importance of feedback from both their peers and their teacher. According to the survey, the learners answered that the teacher feedback was more helpful than the peer feedback.

The present study undetook four surveys to examine 1) how Koreans describe incidents using agentive or non-agentive expressions, 2) how they are affected by agentive or non-agentive expressions when they blame someone for an incident, and 3) how they express time and interpret it in certain contexts. The surveys were given to two online undergraduate classes at a Korean university in 2020. Based on the responses, the study found that 1) Koreans used agentive expressions more often than non-agentive ones when they described incidents which occurred intentionally, 2) they generally used passives rather than agentive or non-agentive expressions when they described incidents which occurred unintentionally, and 3) they were more affected by agentive expressions than non-agentive ones when they called someone to recount an incident. The study also found that Koreans generally saw time proceed from the past to the future uni-directionally and horizontally on a time-line, and they expressed time in spatial terms rather than in temporal terms on the one hand and in time-moving metaphors rather than ego-moving metaphors on the other.

The purpose of this study is to identify whether there are differences in the use of conjunctive adverbials (CAs) depending on language proficiency and interlanguage background among Korean EFL learners, native English experts (NEEs) and non-native English experts (NNEEs) and then to enhance the use of CAs in English writing. To do this, the EFL learners' corpora were collected twice, before class (BC) and after class (AC). They also wrote one page essays, which were subsequently collected and analyzed. For the experts' corpora, NNEE corpus and NEE corpus were retrieved and used in this study. The results first show that EFL learners' proficiency developed and they were able to use CAs in a more balanced and impartial way than before class. The next finding is that there was not much difference in NNEE and NEE use of CAs even though they have different interlanguage. NEEs prefer a sentence initial position more than NNEEs. NNEEs used only two CAs ('also' and 'therefore') in both initial and medial positions while NEEs used the highest frequent four CAs ('for example', 'however', 'therefore', 'thus') in both positions.