Hawthorne qua Psychologist in The Scarlet Letter

Sung Myun Hong 경희대학교 언어정보연구소 1979 언어연구 Vol.1 No.-

Hawthorne qua Psychologist in The Scarlet Letter

English with AI: A new era of TOEIC learning for students majoring in airline services

김나영 경희대학교 언어정보연구소 2022 언어연구 Vol.39 No.-

The current study examines whether AI-assisted TOEIC learning program positively affects TOEIC scores of students majoring in airline services. The study focuses on how the effectiveness of AI technology differs by gender and proficiency. The experiment was conducted during the 2020 academic year. The participants were 119 EFL students from three intact classes at a university in Korea. For the study, the participants studied English through AI-assisted TOEIC learning technology called Soljam for the experimental period of 10 weeks. Before the treatment, all participants took a pre-test based on TOEIC and assigned to three different proficiency groups: Basic, Elementary, and Elementary Plus. After the treatment, all the participants took a post-TOEIC test. Major findings are as follows: First, AI technology positively affected the students’ TOEIC scores at a significance level of .05. Second, students’ improvement in TOEIC scores significantly differed by gender. However, statistically significant difference was not found regarding proficiency levels. There was no significant interaction effect of gender and proficiency. The findings of this study will shed light on the educational value of AI technology to promote language skills of Korean airline service students.

Spectral patterns of the American English diphthong /ai/ as a function of coda voicing produced by native Korean speakers

( Eunjin Oh ) 경희대학교 언어정보연구소 2018 언어연구 Vol.35 No.1

This study aimed to investigate how native speakers of Korean who learned English as a second language realize spectral differences in the American English diphthong /ai/ as a function of coda voicing. Ten Korean learners of English and eight native speakers of American English participated in a production experiment. The monosyllabic words “bite” (/bait/) and “bide” (/baid/) were read along with filler words in isolation and in a carrier sentence. The native group demonstrated significantly smaller F1 and larger F2 before /t/ than /d/ both in the nucleus /a/ and in the offglide /i/ (Moreton 2004). The learner group did not show statistically significant spectral changes in the nucleus and the offglide. Also, the native group significantly reduced the temporal distance between the nucleus and the offglide, and showed spectral peripheralization in the offglide before /t/ than /d/ (Pycha and Dahan 2016). However, the learner group did not show native-like reduction of the temporal distance between the nucleus and the offglide and spectral peripheralization in the offglide. Although the non-native speakers in this study exhibited some durational changes as a function of coda voicing, they did not learn the fine phonetic details regarding the gestural timing and spectral patterns in the diphthong. A considerable degree of individual variation in the learner group and speaking context effects were also found. It was interpreted that the Hyperarticulation hypothesis (Thomas 2000; Moreton 2004) and the Gestural Timing hypothesis (Pycha and Dahan 2016) could provide indices modelling the non-native phenomena found in this study as not attaining native-like phonetic values and gradual approach to the values concerning the spectral and durational aspects in /ai/ as a function of coda voicing. (Ewha Womans University)

L1 influence on the processing of L2 collocation: An experimental study of Korean EFL learners

이신웅 경희대학교 언어정보연구소 2016 언어연구 Vol.33 No.S

The current study aimed to investigate the influence of L1 on the processing of L2 collocations utilizing a phrase-acceptability judgement task. The task was conducted on 39 English collocations with 21 native speakers of English and 45 Korean EFL learners. Their error rates and reaction times were measured both on the congruent collocations that have L1 translation equivalents and the incongruent collocations that do not have ones. It was found that the advanced EFL learners responded significantly faster to and made fewer errors on both the congruent and the incongruent collocations than the intermediate EFL learners. Meanwhile, both the advanced and the intermediate EFL learners made significantly more errors and responded slower under incongruent condition, indicating a strong congruency effect. The results implied that there exists a high degree of reliance on L1 intralexical knowledge in processing L2 collocations by the EFL learners, and acquiring incongruent L2 collocations seems still challenging even to the advanced learners. However, it was also suggested that once EFL learners notice incongruent collocations as valid ones and store them in their L2 mental lexicon, they may be able to processing incongruent collocations as efficiently as native speakers, directly linking L2 collocation forms to their concepts independently of L1 mediation.

Some Remarks on Ahn and Cho’s (2006) Dual Analysis of Korean RNR

정대호 경희대학교 언어정보연구소 2008 언어연구 Vol.25 No.3

There have been various approaches entertained in the literature to account for the properties that the so-called right node raising construction (RNR) displays: Across-the-board rightward movement analyses, PF deletion analyses, and multi-dominance analyses, to name a few. Pointing out that no single uniform analysis can account for all the properties displayed by the RNR, Ahn and Cho (2006) propose a heterogeneous analysis approach: There are two types of RNR structures, one being derived by a semantically constrained ellipsis (S-ellipsis), while the other being derived by a phonologically constrained ellipsis (P-ellipsis). This article shows that despite some genius in it, Ahn and Cho’s (2006) dual analysis approach faces some empirical and theoretical problems. It will be first shown that there exist some RNR properties which are hardly accounted for by any version of ellipsis approach, including S-ellipsis or P-ellipsis. It will be also shown that the mechanism of P-ellipsis, if strictly followed, leads to some non-trivial technical and semantic problems. The conclusion drawn is agnostic: The RNR is still in a haze. (Hanyang University)

NPI licensing and the logic of the syntax-semantics interface

Yusuke Kubota,Robert Levine 경희대학교 언어정보연구소 2021 언어연구 Vol.38 No.2

In this paper, we propose an explicit syntax-semantics interface of NPI licensing in Hybrid TLG. Hybrid TLG is a version of categorial grammar that inherits properties of both lexicalist and derivational variants of generative grammar, and it has been shown in our previous research (summarized in Kubota and Levine 2020) that it offers elegant analyses for a number of complex phenomena at the interface of syntax and semantics (especially in the domains of coordination and ellipsis) that turn out to be highly problematic for other grammatical theories. In the present paper, we extend our work to NPI licensing and report on some initial results suggesting that the flexible and systematic architecture of Hybrid TLG turns out to be successful in this domain too. Specifically, our approach captures interactions between NPI licensing (or polarity sensitivity) and other complex phenomena at the syntax-semantics interface including the scopal properties of modal auxiliaries, Gapping, and VP fronting.

Dynamic spectral patterns of American English front monophthong vowels produced by Korean-English bilingual speakers and Korean late learners of English

오은진 경희대학교 언어정보연구소 2013 언어연구 Vol.30 No.2

This study investigated how Korean-English bilingual speakers and Korean late learners of English realize dynamic patterns of front monophthong vowels in American English. Twenty subjects read monosyllabic [CVC] words containing the vowels [i, ɪ, ɛ, æ] in a carrier sentence. F1, F2, and F3 frequency values were measured at 11 points for each speaker by dividing their vowel duration into ten equal segments. The results indicated that, compared to Korean late learners of English (Group NK), the Korean-English bilingual speakers (Group KE) exhibited F1, F2, and F3 transition shapes and overall formant values of the English front vowels that are more similar to those of native speakers of American English (Group NE). It was also found that the movement directions of the vowels [i] and [æ] varied widely across the three groups, those of the vowel [ɪ] were similar across the three groups, and those of the vowel [ɛ] were similar for Groups NE and KE, but not for Group NK. For the vowel [ɪ], the ranges of the formant movements overlapped for Groups NE and KE, but that of Group NK was quite distinct. For Groups NE and KE, the trajectories of the vowels [ɛ] and [æ] were not as distinct as the trajectories of [i] and [ɪ], though they did not overlap. The vowels [ɛ] and [æ] completely overlapped for Group NK. In general, Group KE exhibited more native-like spectral changes of the American English vowels than Group NK, indicating that there is an effect of language experience in learning the dynamic patterns of the vowels.

Corpus linguistics research trends from 1997 to 2016: A co-citation analysis

박혜진,남대현 경희대학교 언어정보연구소 2017 언어연구 Vol.34 No.3

Corpus linguistics is one of the fastest growing areas of linguistics because of its interface with neighboring academic disciplines and the data-processing capability of a large amount of empirical linguistic data. This study reviews research trends from the last two decades within the corpus linguistics fields. Specifically, the study applied systematic citation analysis procedures to summarize and identify the salient research themes and publications from citation-reference data of peer-reviewed research articles published and indexed in the Web of Science (WoS) between 1997 and 2016. The co-citation analysis of 5,600 research articles and their 172,352 references indicated that, over the four time spans of five years, the corpus linguistics research articles have cited works ranging from general linguistics journal titles to specialized journal titles and individual books. In terms of the research themes of corpus linguistics, the topics of the linguistics research have rapidly changed over the time span. More recently, the development of web-based large monitor corpora and corpus analysis software has contributed significantly to the dynamic and productive interaction of research in the discipline. This may indicate the evolving and juvenile nature of corpus linguistics and its possibility of growing into a multi-disciplinary field. Although there are exceptions to all of the research patterns found in the co-citation analysis, the current study also discusses the most up-to-date research trends and the future directions of corpus linguistics.

How do speakers of different languages differ in the encoding of complex motion events?

박해인 경희대학교 언어정보연구소 2020 언어연구 Vol.37 No.1

Languages vary considerably in how they encode motion. Research (Slobin 2004; Talmy 1985, 2000) has shown that inter-typological differences are found in the frequency of encoding (high vs. low) as well as in the locus of encoding (main verb vs. satellite). Based on these typological differences, cross-linguistic influence (CLI) research has sought to examine the impact of language typology on the processes and outcomes of second language (L2) learning. While research in this area has been fruitful in the past decade, the majority of studies (Cadierno and Ruiz 2006; Daller, Treffers-Daller, and Furman 2011; Park 2019) have centered on investigating the encoding of path and manner of spontaneous motion events. To expand our understanding of how different components of a motion event get selected and sequenced for verbalization, the current study compared encoding patterns of complex motion events by three different language groups (Korean speakers, n=15, English speakers, n=15, Korean learners of English, n=80). Complex motion events comprised three components of motion (i.e., path, manner, and cause), and participants’ descriptions were elicited using a video description task. The results indicated that monolingual speakers exhibited both universal and language-specific encoding patterns, and that English as a Foreign Language (EFL) learners’ motion event descriptions in the L2 displayed L1-based patterns as well as unique characteristics that are specific to EFL learners. In addition, EFL learners’ English-like encoding patterns of complex motion events were largely predicted by L2 productive vocabulary scores.

Prosodic remedies of Korean talkers’ English L2 speech: Optimal pitch and speech rate

Joo Kyeong Lee 경희대학교 언어정보연구소 2018 언어연구 Vol.35 No.1

The current study investigates native English listeners’ perceived foreign accent for pitch- and duration-manipulated speech of Korean EFL talkers with high and low proficiency. This work contents that low proficiency talkers’ L2 speech can be perceptually better accented when a single prosodic parameter such as pitch or speech rate is merely corrected. As nonnative talkers’ proficiency was divided into three categories, high, intermediate and low in the Accentedless Rating task, low talkers were categorically promoted to the intermediate when H* and L* were remedied more native-like and their L2 speech was synthesized to be faster. The corrected prosodic features seemed to be readily detected by native listeners because such corrections might be comparatively salient over the segmental features seriously deviant from native norms in their interlanguage. The results of the current experiment suggested actual numeric values of optimal pitch and speech rate for upgrading low talkers’ proficiency; H* should increase roughly by 30% to 45% higher than the ones that they usually produce in statements, and L* should deepen approximately by 20% to 40% lower than in their normal production of yes-no questions. Speech rate should be 1.2 to 2 times faster for low talkers to be judged as intermediate. On the other hand, Korean high proficiency talkers didn’t show a categorical decline to intermediate when pitch or speech rate was synthetically deteriorated. Due to their little accented L2 speech, phonological/phonetic features, which are very similar to those of native speakers, seemed to firmly tolerate the degrading portion of prosody. In addition, the actual numeric values of pitch and speech rate obtained in the results should be applied to the pedagogical environment and used as references to facilitate improving low talkers’ proficiency.