English-Based Loanwords: Characteristics, Learner Preferences, and Pedagogical Potential

Journal of International Culture, ISSN 2005-3444 December 2014, Vol. 7, No. 2, 1-40

English-Based Loanwords: Characteristics, Learner Preferences, and Pedagogical Potential David E. Shaffer English Language Dept., College of Foreign Languages, Chosun University

Contents I. Introduction II. Categorization of English-Based Loanwords in Korean III. Semantic Similarity of Cognate Pairs IV. Formal Similarity of Cognate Pairs V. The Effect of Borrowed Words on Written English Production VI. Cognate vs. Non-cognate Preference Study VII. The Impact of EBLs upon English Language Learning VIII. Conclusion and Pedagogical Implications ________________________________________________________ Abstract

English-Based Loanwords: Characteristics, Learner Attitudes, and Pedagogical Potential David E. Shaffer The borrowing of loanwords into Korean is prevalent and is occurring at a quickening pace. In the learning of English vocabulary, the English cognates of these English-based loanwords

(EBLs) could be targeted for instruction, but rarely are. Indeed, their inclusion in vocabulary learning is often stigmatized, and research on the use of EBLs in Korean discourse and on classroom instruction on their English cognates is minimal. This study sets out a classification of EBLs, based on their differences from their English cognates, and quantifies these differences with a quality designation. This EBL quality score identifies the types of differences for which more focused instruction may be required. The survey also show that learners us a considerable number of English cognates of EBLs in their written production and general use. However, the study also reveals that learners often do not know how the English cognates that they are using differ from

their EBLs or even whether they differ in any way. It is obvious that there are pedagogical ramifications of these findings, and it is indeed proposed that much more attention be accorded to cognates of EBLs for a rapid English lexicon expansion and English cognate understanding. [Key words: English-based loanwords, characteristics, quality, learner preference]

I. INTRODUCTION The phenomenon of lexical borrowing from English is one that has been apparent in Korean for over half a century, and in recent years, it has been occurring at an increasing rate. Though there are individuals who question the benefits of using mother-tongue loanwords derived from the language that is being taught in the monolingual EFL classroom, F. E. Dalton (e.g., 1998, 2008) and others (e.g., Uchida, 2001) have undertaken research that points to benefits of teaching English vocabulary items that are also Japanese loanwords for quickly increasing learner’s vocabulary size. If Korean learners of English could equally benefit in rapid vocabulary expansion through the instruction in English-based loanwords (EBL in Korean, this could lead to more rapid English acquisition as words are the building blocks of language. This study first presents a description of the characteristics of Korean loanwords borrowed from English. These include (a) nouns of a wide variety, (b) nouns with Korean suffixes, (c) truncated words, (d) abbreviations, and (e) fabrications, as well as lexical gap filling and English-Korean compounding. This is followed by a description of the semantic changes that accompany many English-to-Korean loanwords, including (a) semantic narrowing, (b) semantic widening, and (c) semantic shift. This study proceeds with a systematic analysis of English loanwords borrowed into Korean to which a quality value has been assigned. This overall quality value and the values of its components is instrumental in determining the degree of difficulty Korean learners of English will have in learning the English counterparts of the EBLs and provide them with loanword learning strategies. Examined in this study are four aspects of Korean-English cognates. The cognates’ quality is based on both semantic and formal (structural) features. The semantic features analyzed are cognate type (true, divergent, convergent, or false) and meaning overlap, a measure of whether the loanword matches a more common or less common meaning of its English cognate. The formal feature measured is degree of shortening. Restrictions on loanwords are also measured. This study also attempts to discover to what extent learners’ pre-existing knowledge of EBLs influences their English usage. It also attempts to determine whether there are any differences in degree of difficulty in learning English vocabulary corresponding to the five types of loanwords: true cognates, phonetic/grammatical modifications, morphological modifications, semantic modifications, and fabrications. To investigate these questions, a survey was created in the form of a questionnaire and administered to a group of university students studying English. The bulk of the questionnaire consisted of questions containing a Korean sentence exhibiting an English loanword and an English sentence containing that loanword as represented in Korean. Participants were asked whether the English rendering of the English-based loanword in Korean was correctly used in the English sentence provided.

Results indicate that Korean-English cognates have a relatively high quality value and that the value is higher for words whose English cognates are in less frequent ranges of use. It has been found that many Korean-English cognates are convergent, that a majority of loanwords correspond to the most common English cognate definition, and that few undergo shortening. The study reveals that there is negative transfer from EBLs, but also suggests that this might be outweighed by the benefits gained in lexical acquisition. It will be shown that these findings are similar to those for EBLs in Japanese and that arming our English learners with this knowledge could be beneficial in expanding the learner’s lexicon via EBLs and provide them with strategies for fine-tuning the learning of loanword meanings.

II. CATEGORIZATION OF ENGLISH-BASED LOANWORDS IN KOREAN 1. Formal Categorization of Loanwords Korean loanwords borrowed from English can be found in many forms in the Korean lexicon, but by far the largest group is that of nouns, many of which have most of their original English features preserved, except for the adjustments required to conform to the Korean phonological system. These nouns have been borrowed from a wide range of fields from technology to sports, to food and fashion, to architecture and furniture:

1.1. Formal Preservation K. keompyuteo, E. computer K. tellebijeon, E. television

K. model, E. model K. hotel, E. hotel

Another characteristic of loanwords is that they may combine with Korean bound morphemes to, in some cases, remain the same part of speech or, in other cases, to be transformed into another part of speech:

1.2. Loanword + Bound Morphemes K. taening-hada, E. tan (v.) K. paking-hada, E. park (v.)

K. kipeu-hada, E. keep (v.) K. seumateu-han, E. smart (adj.)

Korean favors words of few syllables. Most lexical items consist of one to three syllables, the majority being two syllables in length. Therefore, it is common for Korean to truncate many-syllable loanwords, making them better conform to the Korean syllable norm:

1.3. Truncations K. eeokeon, E. air conditioner K. syupeo, E. supermarket

K. waiteu, E. whiteout K. seukin, E. skin lotion

Korean is creative in taking two English words or elements and combining them in ways that form new words for Korean which are non-existent in English:

1.4. Fabrications K. openka, E. open + car = convertible K. selleorimaen, E. salary + man = office worker Another method that Korean uses to shorten loanwords is to create a form of initialism, often using the initial letter of the first syllable, and either the second or the third syllable of a single English word:

1.5. Initialisms K. IC, E. interchange

K. B/D, E. building

2. Semantic Categorization of English-Based Loanwords In addition to formal changes, loanwords may have a number of changes in meaning that accompany the loanword’s transformation from an English word to a Korean word. There may be semantic preservation, in which the English cognate’s meaning is preserved in the Korean loanword; semantic narrowing, in which the loanword’s meaning is restricted; semantic widening, in which the loanword’s meaning is extended; and semantic transfer, in which the meaning has noticeably been shifted.

2.1. Semantic Preservation K. beoseu = E. bus K. maketing = E. marketing

K. radio = E. radio K. banana = E. banana

2.2. Semantic Narrowing K. miting (fr. E. meeting) = group blind date K. chyuri (fr. E. tree) = Christmas tree

2.3. Semantic Widening K. sopa (fr. E. sofa) = sofa or stuffed armchair. K. seobiseu (fr. E. service) = any item provided to the customer free of charge

2.4. Semantic Transfer K. hipeu (fr. E. hip) = buttocks K. keonning (fr. E. cunning) = test cheating

In a random sample of 111 EBLs, it was found that 67% exhibited semantic preservation, 25% semantic narrowing, 4% semantic widening, and another 4% exhibited semantic transfer.

III. SEMANTIC SIMILARITY OF COGNATE PAIRS People tend to focus on the semantic aspect of cognate pairs in considering their similarities. Ways in which EBLs in Korean may differ semantically from the corresponding English words from which they were borrowed will be set out below and these semantic differences will be measured in a quantifiable manner.

1. Types of Semantic Cognates Sameness of word meaning and function is generally, but misguidedly, assumed by language learners between loanwords and the words from which they are borrowed. Ulchida (2001) has identified six classifications of the relationships between L1-L2 cognates. These are described here with EnglishKorean cognate examples provided:   







True Cognates: English-Korean cognate pairs with identical denotations (e.g., E. computer, K. kompyuteo). Convergent Cognates: English-Korean cognate pairs in which the English word is broader in meaning than the Korean cognate (e.g., E. drama, K. deurama [= television drama series]). Divergent Cognates: English-Korean cognates in which the Korean cognate is broader in meaning than the English word from which it originates (e.g., E. handle, K. haendeul [includes steering wheel]). Distant False Friends: English-Korean cognate pairs in which the Korean cognate differs totally or almost totally in meaning from the English word from which it derived (e.g., E. scrap [= a small remainder no longer of use], K. seukeuraep [= to clip and file as in a scrapbook]). Close False Friends: English-Korean cognate pairs in which the Korean cognate differs partially in meaning from the English word from which it derived (e.g., E. hip [= joint on left or right waist area], K. hipeu [= buttocks]). Koreanized English: Korean word taken from English but having no semantic relationship with the original English word (e.g., K. hochikiseu [= stapler; fr. E. Hotchkiss, surname], K. konsaiseu [= dictionary; fr. E. concise) or a compound having no English counterpart (e.g., K. seukinsip [fr. English skin + -ship], K. samaen [= white-collar office worker; fr. E. salary + man]).

The first five of these types have been tested for learnability (Uchida, 2001). There order from easiest (1) to most difficult (5) was found to be as follows: 1 = True cognates 2 = Divergent cognates 3 = Convergent cognates

4 = Distant false friends 5 = Close false friends

2. Method for Determining Cognate Semantic Type To begin with, a sample of EBLs was collected. For this purpose, the Korean Practical Dictionary (2000) was used. The first English-to-Korean loanword on each fifteenth page was selected (if none, then first instance on subsequent page was selected); this was done twice, starting on different pages. Not included in the tally were single-word Korean-cognate headwords that were derived from twoword or longer expressions in English (e.g., K. baekmireo from E. back mirror). A total of 325 EBLs were collected. Of these, 111 EBLs whose English cognates were from different word families were among the most frequently used English words, according to the British National Corpus. This broke down into 29 words among the 1,000 most frequently used words in English (1K), 29 among the second most common 1,000 words (2K), and 53 among the third most common 1,000 words (3K). These 111 words were used as the EBL sample. Another 20 of the 325 words were among the fourth and fifth most common words (4-5K), but these were not included in the sample. The evaluation system included both semantic and formal elements. Types of cognates were evaluated by giving points to learn ability: true cognates received 3 points; divergent cognates, 2 points; convergent cognates, 1 point; and distant false friends, close false friends, and Koreanized English, 0 points. Results according to word frequency level appear in Table 1. Of the 111 loanwords in the sample, 15 (13%) were true cognates, 94 (85%) were convergent cognates, 1 (1%) was divergent, and 1 (1%) was a distant cognate.

Table 1. Average Cognate Type Quality Results Word Frequency Level

Average Cognate Type Score (0-3)

1K (0000 - 1,000)

1.1

2K (1,001 - 2,000)

1.1

3K (2,001 - 3000)

1.5

Total (0000 - 3,000)

1.3

3. Cognate Pair Semantic Overlap It is typical for a loanword in Korean to have a single meaning, and just as common for the English word from which it was borrowed to have more than one meaning. It is not necessarily the most common meaning of the English word that is borrowed with the loanword. The ranking of the loanword’s meaning was rated according to the rank of this meaning among the meanings of the English cognate. The point system in Table 2 was devised for this rating purpose.

Table 2. Point System for Rating Cognate Pair Meaning Correspondence Points

Description

3

Loanword definition corresponds to the first listed definition in the English dictionary.

2

Loanword definition corresponds to the second listed definition in the English dictionary.

1

Loanword definition corresponds to the third listed definition in the English dictionary.

0

Loanword definition corresponds to the fourth or lower listed definition in the English dictionary.

Applying this point rating system, we obtain an average definition score for each of the three frequency levels as well as an overall average (see Table 3). The average definition score and the average cognitive type score were conflated to produce a single three-point average semantic feature score (see Table 3). This semantic feature score will be combined with a formal feature score to determine overall cognate quality.

Table 3. Cognate Pair Average Definition Correspondence & Average Semantic Score Word Frequency Level

Average Definition Score (0-3)

Average Cognitive Type Score (0-3)

Average Semantic Feature Score (0-3)

1K (0000 1,000)

1.3

1.1

1.2

2K (1,001 2,000)

2.1

1.1

2.6

3K (2,001 3,000)

2.5

1.5

2.0

Total (0000-3,000)

2.1

1.3

1.7

IV. FORMAL SIMILARITY OF COGNATE PAIRS Along with semantic similarity, similarity in form of cognates impacts to learnability. Cognate pairs that are quite similar in both meaning and form are the most effortless to learn. Cognate paring, it has been found, is prolific across languages and is based on formal rather than semantic similarity (Carroll, 1992). Based on a point system similar to that devised for semantic gauging semantic similarity of cognates, a point system for quantifying formal similarity is described below.

1. Shortening While consonant clusters abound in English lexical items, they are lacking in Korean. In adopting English loanwords containing consonant clusters, Korean phonology separates the consonants with vowels, thereby increasing the number of syllables in the word. As Korean is a language with a preference for words of few syllables, when English words of many syllables or words containing consonant clusters are borrowed into Korean, they also often undergo shortening (e.g., English transformer becomes K. teuraenseu [trans]), making the borrowed form less like the word from which it was borrowed and thereby less easily recognizable. Points were assigned with reference to shortening

according to Table 4. Only six of the loanwords in the sample were found to have undergone shortening, but in each case it was a major form of shortening. The results appear in Table 5.

Table 4. Point System for Rating Loanword Shortening Points

Description

3

No shortening of borrowed word.

2

Mild shortening of borrowed word, preserving the semantically important element of the word (e.g., the stem).

1

Major shortening of at least half of the borrowed word.

2. Other Formal Restrictions In addition to shortening, the most common restrictions on the loanwords in the sample were (a) narrow range of collocation of EBL in comparison with its English counterpart (20 instances) and (b) not being able to be used independently (3 instances; e.g., K. deurai- [fr. E. dry] cannot be used in Korean as a free-standing lexical item, but must combine with a verb ending or suffix). For each of these restrictions, one point was subtracted from total scores. The number and percentage of restrictions on the sample loanwords for each word frequency level appear in Table 5.

Table 5. Cognate Pair Average Formal Similarity Results Word Frequency Level

Average Shortening Score (0-3)

Other Restrictions (-1 pt. each)

1K (0001 - 1,000)

2.8

11 (38%)

2K (1,001 - 2,000)

2.9

2 (7%)

3K (2,001 - 3,000)

3.0

10 (19%)

Total (0001 - 3,000)

2.9

23 (21%)

3. Cognate Quality Discussion To determine the cognate quality score for the English cognate of an EBL, the cognate’s type score and definition score were averaged together to produce a score of 0-3. This was added to the shortening score, and then any restrictions (-1 per restriction) were subtracted. This produced the total cognate quality score. The individual scores for a random sampling of English cognates can be seen in Appendix A. The average cognate type score for the entire sample was 1.3 points, making the cognate type score the lowest by a significant amount among the items receiving a score. Only 13% were true cognates, while the vast majority (85%) was convergent in type, showing that the range of use is much narrower in the EBL than in its English counterpart. The average definition score of 2.1 overall indicates that EBL meanings often do not correspond to the most common meaning of their English counterpart. Only 60% of EBLs carry the most common meaning of their English cognate. Very few EBLs undergo shortening (only 6 in the sample; 5%) as indicated by the average shortening score of 2.9. Those that were shortened

were a shortening by half of a two-word or compound-word expression in English (e.g., E. front desk to K. hureonteu [front], E. nightclub to K. naiteu [night]). The percentage of loanwords having other formal restrictions was 21. Average total score for loanword quality, based on a combined 6-point quality score, indicates that the lower the word frequency, the lesser the chance of semantic or formal deviation of the EBL from that of its English counterpart. This is true for each category of measurement except for restrictions, for which the 2K frequency level was lower than the 3K, although the 1K level still contained the most restrictions. These results, including the average total score, are brought together in Table 6.

Table 6. Summary of Average Cognate Quality Results Word Frequency Level

Ave. Cognate Type Score (0-3)

Ave. Definition Score (0-3)

Ave. Shortening Score (0-3)

Percent with Restrictions (-1)

Ave. Total Score (1-6)

1K

1.1

1.3

2.8

38

3.6

2K

1.1

2.1

2.9

7

4.4

3K

1.5

2.5

3.0

19

4.8

Total

1.3

2.1

2.9

21

4.4

A similar study conducted on Japanese loanwords from English (Daulton, 2008) showed a corresponding pattern of results; deviations of the loanword from its English cognate were highest for the most frequently occurring English words and decreased with lower frequency (Table 7). However, cognate quality scores were consistently higher for Japanese loanwords. This was mainly due to a higher percentage of Japanese loanwords being true cognates, being associated with the English cognate’s most common meaning, and not differing in formal characteristics from their English counterparts (Table 7). Nevertheless, average total quality scores for EBLs range from 3.6 to 4.8 for the three word frequency levels and average 4.4 overall, indicating a considerable set of characteristics that Korean loanwords share with their English cognates. Korean-English cognates comprise roughly 5,900 (12%) of the 51,000 headwords in Korean Practical Dictionary (2000).

Table 7. Comparison of Results with Japanese Cognate Quality Results Word Frequency Level

Ave. Cognitive Type Score (0-3)

Ave. Definition Score (0-3)

Ave. Shortening Score (0-3)

Percent with Restrictions (-1)

Ave. Total Score (1-6)

1K

1.1

1.5

1.3

2.6

2.8

3.0

38

42

3.6

4.3

2K

1.1

2.2

2.1

2.8

2.9

3.0

7

13

4.4

5.1

3K

1.5

2.8

2.5

3.0

3.0

3.0

19

0

4.8

5.7

Total

1.3

2.1

2.1

2.8

2.9

3.0

21

13

4.4

4.9

Note: Japanese loanword results from Daulton (2008) are in the darkened right-hand columns.

V. THE EFFECT OF BORROWED WORDS ON WRITTEN ENGLISH PRODUCTION The above investigations have dealt with the quantity EBLs found in Korean based on dictionary entries and the frequency of use of the English cognates of EBLs in Korean. This investigation serves to determine the extent to which cognates of EBLs are used in the written English of Korean learners of English. If the use of English cognates is found to be plentiful, this will lend support to the claim that it is effective to use L1 loanwords in learning English (e.g., Daulton, 1998, 1999, 2008; Kay, 1995; Nation, 1990).

1. Method A total of sixty-five university students who were all L1 Korean speakers participated in the written English production study. These participants were all in their sophomore through senior year in a Korean university and enrolled in an English oral skills course. They were asked to write about their summer plans in approximately 100 words in a friendly note to an English-speaking friend, real or imaginary. Dictionaries were not allowed and the study did not examine syntactic or orthographic accuracy.

2. Results and Discussion The purpose of this written English sample was to determine the extent to which English cognates of Korean loanwords are used in the English writing of Koreans. These sixty-five participants produced a corpus of 7901 tokens (i.e., instances of words and any repetitions of those words). Among these, they produced 296 tokens of words that have EBLs in Korean, which account for 3.8% of all tokens. However, if function words are excluded from the sample, 5352 non-function tokens remain, of which the 296 token of English words which have been borrowed into Korean comprise 5.5% (Table 8).

Table 8. Lexical Summary of Participants’ Written Production Words

Total Tokens

Borrowed Tokens

Percent Borrowed Tokens

Total Words

7901

296

3.8%

Non-function Words

5352

296

5.5%

Approximately one in every eighteen non-function word appearing in the students’ writing was an instance of an English word that had been borrowed into Korean. Although this is not a large percentage of the non-function words produced in the students’ writing, it is a portion that can easily be added to learners’ written production with minimal instructional requirements.

VI. COGNATE VS. NON-COGNATE PREFERENCE STUDY In a study by Brown (1995), it was shown that Japanese speakers have a preference for English words that have been borrowed into Japanese over non-borrowed English words. The instrument used to determine this was a set of English gap-fill sentences with four single-word choices to select from. All of the four choices were equally appropriate choices to complete the sentence, but only one of them was a word that had been borrowed into Japanese. An example sentence in the instrument follows (the asterisk indicates the borrowed word): 4. The students were very __________. a. bored b. curious c. alert

d. active*

To determine if Korean speakers have a preference for English words that have been borrowed into Korean, a study similar to the Brown (1995) study was conducted.

1. Method, Materials, and Procedures An instrument of 18 English sentences with one blank each and four possible items to choose from was constructed. Each of the four choices was an equally appropriate choice to complete the sentence, but only one of them was a word that had been borrowed into Korean from English. Also, the cognate was in the same thousand-word range in the word frequency list (e.g., 2,001-3,000 word frequency) as the other three possible choices (Appendix B). The participants were fifty-eight of the same university students that participated in the written production study described above. They were informed that all four choices for each item were equally appropriate both syntactically and semantically, and instructed to choose the word that they liked best. The participants were also informed that the questionnaire was purely for research purposes and not related to their coursework in any way. They were also given as much time as necessary to complete the questionnaire, but everyone finished in a very few minutes, indicating that they did not put much deep consideration into their choices.

2. Results and Discussion If Koreans have a preference for borrowed words rather than non-borrowed words in the English they use, this should manifest itself in a higher than random selection probability (i.e., higher than 25%). The number of times (out of a possible 58) that the borrowed word was selected and the percentage of the times it was selected for each of the questionnaire items appears in Table 9. Preferences varied from question item to question item, but the overall result was that of the possible 1,044 times that the borrowed word could have been elected, it was selected 522 times out of 1,044 – a full 50% selection rate. This indicates a strong preference among Korean users of English for English words that have been borrowed into Korean.

Table 9. Results of Multiple-Choice Cognate Questionnaire Question No.

1

2

3

4

5

6

7

8

9

Times Selected (58 possible)

26

42

32

52

14

31

15

28

24

44.8

72.4

55.2

89.7

24.1

53.5

25.9

48.3

41.4

Question No.

10

11

12

13

14

15

16

17

18

Total

Times Selected (58 possible)

16

2

47

27

52

28

25

10

51

522/1044

27.6

3.5

80.0

46.4

89.7

48.3

43.1

17.3

87.9

50.0

Percentage

Percentage

All of the possible choices on the questionnaire were nouns since the bulk of loanwords that are borrowed are nouns. A sampling of the corpus that was obtained in the written production study above indicated that 21.5% of all words in the corpus, including function words were nouns. As the percentage of nouns occurring in Korean speakers’ English production is high and as their preference among English nouns for those borrowed into Korean is high, it seems quite reasonable to incorporate early into the language-learning curriculum the learning of English nouns borrowed into Korean.

VII. THE IMPACT OF EBLS UPON ENGLISH LANGUAGE LEARNING An attempt was made to discover to what extent learners’ pre-existing knowledge of EBLs influences their use of the English cognate. It also attempts to determine whether there are any differences in degree of difficulty in learning English vocabulary corresponding to the five types of loanwords: true cognates, phonetic/grammatical modifications, morphological modifications, semantic modifications, and fabrications. The findings are likely to have pedagogical implications for instruction on the English cognates of EBLs.

1. Method, Materials, and Procedures The participants in the study were 73 university students from a typical large university in Korea. The students were junior and senior students majoring in English. They were 24 male and 49 female students between the ages of 20 and 39 (M = 23.0). All the participants were native Korean speakers, and their general English proficiency ranged from intermediate to advanced. The materials for the sub-study consisted of an online survey made available to junior and senior university students at a large university in Korea via a leading provider of web-based survey solutions. In addition to demographical data questions, the survey consisted of 49 questions concerning whether a lexical item occurring in a sentence was being used in accordance with normal English usage or with modifications associated with its Korean loanword usage. Following Inagawa (2010), who surveyed the use of EBLs in Japanese, each survey consisted of a lexical item in uppercase font in an English sentence. This was immediately followed by a Korean sentence containing the EBL within asterisks. The purpose of the Korean sentence was to convey the intended meaning of the English sentence and was identical in

meaning with the English sentence except possibly for the word in uppercase font. If the meaning of the two sentences were identical, the participants were instructed to choose the first of the two response choices provided; that is, “appropriate English usage” (see example 1). 1. These days, the small bottles of black MANICURE are the most popular. (요즘엔 작은 블랙*매니큐어* 병이 가장 인기가 있다.) O 적절한 영어사용 O 부적절한 영어사용 (적절한 영어표현을 아래에 적으세요.) fingernail polish

If the uppercase lexical item was not identical with that within asterisks in the Korean sentence, the participant was instructed to choose the second of the two response choices provided, “inappropriate English usage,” and to provide the appropriate English expression in the box provided immediately below the response choices (example 1). All instructions were given in Korean to reduce as much as possible the possibility of misunderstanding. The participants were given one week to complete the online survey. Of the 88 students requested to take the survey, 73 successfully completed all the items on it. The forty-nine EBLs used in this survey (see Appendix C) are of five different types: true cognates as well as semantic, morphological, and phonetic/grammatical modifications, and fabrications. The true cognates were EBLs identical in meaning and form to their English cognates, save for phonetic changes required for adaptation to the Korean phonological system. There were nine true cognates in the survey: fast food, variety show, recycling, soft drink, skin care, coupon, dessert, topic, love. The semantic modification category included ten lexical items whose loanword meaning had in some why shifted from that of its English cognate: consent, circle, cunning, talent, sense, handle, sign, style, menu, diet. The ten lexical items in the phonetic/grammatical modification category included lexical items from which an English particle had been dropped or a phonetic unit had been omitted: Valentine(‘s) Day, sunglass(es), meter(s), manner(s), pajama(s), scramble(d), hand(s) free, Y-shirts (white shirt), ski(ing), wellb(e)ing. The morphological modification category contained ten lexical items of syllable length or longer dropped in some way from the original English cognate: aircon (air conditioner), rimocon (remote control), shortcut (short haircut), super (supermarket), after service (after sales service), intelli (intelligent), leports (leisure sports), pro (percentage), inflay (inflation), codi (coordination). The ten items in the fabrication category are loanwords that do not have English cognates, but which are composed of items from English: soft cream, morning call, charming point, image change, skinship, open car, morning coffee, Dutch pay, hand-phone, home-in.

2. Results and Discussion The overall mean number of responses correctly identifying whether the selected lexical items were correct usage, modified versions through borrowing into Korean, or English-based fabrications was 44.5 (61.0%). See Table 10. This rather low rate of identifying EBLs as being correctly or incorrectly used in English suggests that instruction on EBLs in relation to EFL has been wonting.

Table 10. Overall Mean Number of Responses Correctly Identifying Modification per Lexical Item Participants 73

Correct Identification Mean % 44.5 61.0

Incorrect Identification Mean % 28.5 39.0

The overall mean number of responses correctly identifying whether the selected lexical items were correctly used was calculated by category. The mean for correctly identified true cognates was 63.4 (86.9%), for morphological modifications 53.1 (72.7%), for semantic modification 38.7 (53.0%), for fabrication 37.3 (51.1%), and for phonetic/grammatical modification 31.9 (43.7%). See Table 11.

Table 11. Overall Mean Number of Responses Correctly Identifying Modification per Lexical Item by Type Type of Modification True Cognates Morphological Modification Semantic Modification Fabrication Phonetic/Grammatical Modification

Correct Identification Mean % 63.4 86.9 53.1 72.7 38.7 53.0 37.3 51.1 31.9 43.7

Incorrect Identification Mean % 9.6 13.1 19.9 27.3 34.3 47.0 35.7 48.9 41.1 56.3

Note: Participants = 73.

By loanword type, participants were able to correctly identify the presence or absence of modification in the order: true cognate (89.9%) > morphological modification (72.7%) > semantic modification (53.0%) > fabrication (51.1) > phonetic/grammatical modification (43.7%). This closely follows the results obtained by Inagawa (2010) for English-based loanwords in Japanese with her university student test group: true cognate (80.9%) > morphological modification (74.3%) > fabrication (38.2%) > semantic modification (35.6%) > phonetic/grammatical modification (31.6%). Though producing percentages very similar to each other in both surveys, the order of the fabrication and semantic modification categories were reversed in the Inagawa study. For the forty loanwords in the survey that were in some way modifications of English lexical items, survey participants were asked to provide the correct English expression when they identified a loanword as being incorrectly used in the given English sentence. For those responses that identified a lexical item as a Korean modification, the mean number of correct English equivalents provided is given for each category as well as the percentage of correct English equivalents provided (Table 3). This is followed by the percentage of correctly provided English equivalents compared to total possible correct responses. For overall modifications, the mean of those items identified as modifications for which a correct English equivalent was provided was 26.8 (67%). By category of modification, the results were morphological modification 34.0 (64.0%), semantic modification 28.8 (74.4%), fabrication 19.3 (51.7%), phonetic/grammatical modification 25.0 (78.3%). When correct English equivalent responses are considered with regards to all possible correct responses, the percentages are lower: overall 36.7%,

morphological modification 46.6%, semantic modification 39.5%, phonetic/grammatical modification 34.3%, fabrication 24.4%.

Table 12. Mean Number of Responses Correctly Providing English Equivalents of Modified Loanwords by Type Type of Modification Morphological Modification Semantic Modification Fabrication Phonetic/Gramm Modification All Types of Modification

Correctly Identified as Modified Mean % 53.1 72.7 38.7 53.0 37.3 51.1 31.9 43.7 40.3 55.1

English Equivalent Correctly Provided Mean % % of Total 34.0 64.0 46.6 28.8 74.4 39.5 19.3 51.7 24.4 25.0 78.3 34.3 26.8 67.1 36.7

Note: Participants = 73.

The survey revealed that participants were able to determine with only 61.0% accuracy whether a lexical item being tested was being used in English in its correct form or whether it was being used in its modified EBL form. When considered by category of lexical item, a wide range in correct responses is revealed: from 86.9% correct responses for true cognates to 43.7% correct responses for phonetic/grammatical modifications. The degree of difficulty is of the order: true cognate < morphological modification < semantic modification < fabrication < phonetic/grammatical modification, as shown in Table 11. True cognates provide the least difficulty for respondents to identify because neither their meaning nor form is modified. The relatively high rate of correctly identifying true cognates, however, most likely stems from a misplaced belief that EBLs generally carry the same meaning as their English cognate. Instruction in EBLs and their modifications from their English cognates would be helpful in rectifying this tendency to overgeneralize sameness of meaning. Morphologically modified items are only slightly more difficult to identify because there is no semantic change and only partial structural modification, usually in the form of truncation (e.g., super from “supermarket”). Next in difficulty are semantically modified items in which the form of the item is unchanged, but the meaning may be partially changed, either broadened or narrowed (e.g., diet), or totally changed (e.g., cunning). Fabrications are more difficult to identify because of their saliency. Their form often copies that of actual English lexical-item patterns (e.g., open car), and their meaning is readily comprehensible. Phonetically/grammatically modified items are the most difficult to identify because, in addition to a grammatical and/or phonetic modification, L1 transfer (interference) is often involved (e.g., sunglass[es]). L1 negative transfer has a very strong influence upon L2 production; for example, the near total absence of plural forms in Korean leads to the omission of plural forms by the L1 Korean learner in producing English plural forms. Similarly, recognition of the omission of plural markers in the English test items in this study was most difficult. In addition to identifying whether selected lexical items were being used in their proper English form or usage, participants were asked to provide the proper English form when they determined that the lexical item (EBL) was not of proper English form or usage. Correct English forms or usage were provided by the respondents with only 67% accuracy (Table 12). Inagawa (2010) reports similar results to the above in Japanese learners of English with EBLs in Japanese and concludes that “. . . loanwords,

which have undergone modifications in some ways, influence on learners [sic] English production negatively. Moreover, it is revealed that the negative effect is relatively large . . .” (p. 51). Two points need to be considered regarding this statement. The first is that Inagawa’s second sentence is merely a logical truth: it states that the negative effect of some types of modification is greater than for the absence of modification (cf. “relatively large”) as is the case for true cognates. The second point is that although the majority of categories of loanwords and EBLs themselves in the Inagawa survey (four of five categories, and 60 of 70 loanwords) contained modifications, modified loanwords do not occur with such high frequency in the language. Approximately, 75% of the English-based loanwords in Korean, for example, are true cognates. Accordingly, the negative effects of modification in loanwords in Korean can occur in no more than 25% of the loanwords, and then only in varying degrees. Consequently, the negative effects of loanwords on learnability have less of an effect than Inagawa’s statement would seem to suggest.

VIII. CONCLUSION AND PEDAGOGICAL IMPLICATIONS Korean-English cognates comprise a considerable portion of the Korean lexicon, and as this study has shown, share a large percentage of semantic and formal characteristics. English-based loanwords in Korean serve as a valuable lexical pool from which their related English cognates can be introduced to learners to rapidly increase their English vocabulary. This study has shown that 5.5% of all non-function words that Korean speakers use in written production are words borrowed into Korean. It has also shown that 21.5% of all English words used are nouns, and that among nouns, Korean speakers have a high preference (50%) for nouns borrowed into Korean. Given these findings, it seems very practical to teach as quickly as possible those frequently used vocabulary items that have been borrowed into Korean to quickly build a vocabulary base for the Korean learner of English to draw upon to expand their language skills. English vocabulary items that are cognates of EBLs are easy to teach in that the learner already knows a portion of the meaning, pronunciation, and formal and grammatical categories of English cognates through familiarity with their EBL, whereas for totally new vocabulary items, the learner has 100 percent of the aspects of the item to learn. However, because true English-Korean cognates account for only 75 percent of EBLs, the learner will need guidance in their English cognate learning. The learner will benefit from being introduced to learning strategies such as expecting an English cognate to possibly be broader in meaning than its corresponding loanword, as so many cognates that are not true cognates are convergent. The learner would also benefit from knowing that it is the most frequently used English cognates that differ most in meaning and form from the Korean loanwords and that as frequency level decreases, similarities between English-Korean cognates increase. These phenomena concerning English-Korean cognates need to be highlighted through the development of relevant teaching materials for Korean learners of English at all levels. This study has shown that English learners’ use of English cognates of EBLs comprises 5.5% of their total non-function word use in English and that the learner is three times more likely to use a English cognate of a Korean EBL that they are a non-cognate. With this degree of use of English cognates of EBLs and this distinct preference for such cognates, it follows that classroom instruction should include instruction in the correct use of English cognates to rapidly and relatively effortlessly increase learners’

correct use of lexical items. Such instruction has been shown to be necessary as learners could identify only one-fourth to one-half of all English cognates as having modified EBLs and then were able to correctly identify the English cognate’s difference from the EBL (i.e., the EBL’s type of modification) in only about one-third of the cases. In this study, only a few formal and semantic characteristics of English-Korean cognates were studied. The number of features studied needs to be increased to include more semantic and formal features, including phonetic and phonological features, word meaningfulness, word concreteness, word length, collocations, learner proficiency level, and previous contact with English cognates. However, independent studies strongly suggest that early introduction of English-based loanwords can serve to quickly build up the learner’s English vocabulary (Brown, 1995; Brown & Williams, 1985; Daulton, 1998, 1999, 2003, 2005, 2008, 2011; Inagawa, 2010; Karunakaran, 2012; Kay, 1995; Kent, 1999, 2000; Kimura, 1989; Ogasawara, 2008; Uchida, 2002, 2007; Van Benthuysen, 2004). The present study supports this claim. It also agrees with Inagawa (2010) on two major points. The first is that it is desirable for the English cognates of Korean EBLs to be introduced to learners with different degrees of caution based on type of modification. That is, the categories of loanwords with higher degrees of difficulty, as determined by this study, require a more detailed introduction (i.e., more cautionary and detailed instruction is desirable in introducing the phonetic/grammatical modification type than in introducing the morphological modification type). Secondly, in teaching English cognates of loanwords in the learner’s L1, it is desirable and likely more efficient to spend more time highlighting the contrastive differences between loanwords types with a higher degree of difficulty and their English cognates than to spend an equal amount of time on true cognates and English equivalents of morphologically modified loanwords, which have a lower degree of difficulty. In this way, English cognates of EBLs in Korean can be quickly and efficiently introduced to build up the learner’s English lexicon, and with an expanded English lexicon, the learner will become more fluent and more proficient in their English expression.

References Brown, J. B. (1995). Is gairaigo English? The Internet TESL Journal, 1(12). Retrieved from http://iteslj.org/Articles/Brown-Gairaigo/index.htm Brown, J. B., & Williams, C. J. (1985). Gairaigo: A latent English vocabulary base? Tohoku Gakuin University Review: Essays and Studies in English Eibungaku, 76, 129-146. Carroll, S. E. (1992). On cognates. Second Language Research, 8(2), 93-119. Daulton, F. E. (1998). Loanword cognates and the acquisition of English vocabulary. The Language Teacher, 20(1), 17-25. Daulton, F. E. (1999). English loanwords in Japanese: The built-in lexicon. The Internet TESL Journal, 5. Retrieved from http://iteslj.org/Articles/Daulton-Loanwords.html Daulton, F. E. (2003). The effect of Japanese loanwords on written English production. JALT Hokkaido Journal, 7, 3-12. Daulton, F. E. (2005). Common gairaigo corresponding to high-frequency and academic English: Are Japanese students ready for foreign study? JALT Hokkaido Journal, 9, 1-16. Daulton, F. E. (2008). Japan’s built-in lexicon of English-based loanwords. Clevedon, UK: Multilingual Matters. Daulton, F. E. (2011). On the origins of gairaigo bias: English learners’ attitudes towards English-based

loanwords in Japanese. The Language Teacher, 35(6), 7-12. Inagawa, M. (2010). Impacts of loanwords upon second language learning: Application of Englishderived words in English vocabulary learning and teaching in Japan. Berlin, Germany: Lambert Academic Publishing. Karunakaran, T. (2012, March 31). The use of English loan words in teaching ESL in Jaffna [Web log post]. Retrieved from http://neltachoutari.wordpress.com/2012/03/31/the-use-of-english-wordsin-teaching-esl-in-jaffna/ Kay, J. (1995). English loanwords in Japanese. World Englishes, 14(1), 67-76. Kent, D. B. (1999). The discourse of Konglish: A socio-historic analysis of English as a native part of the South Korean vernacular (Unpublished doctoral dissertation). Knightsbridge University, Torquay, UK. Kent, D. B. (2000). Speaking in Tongues: Chinglish, Japlish, and Konglish. KOTESOL Proceedings PAC2, 1999, 197-209. Kimura, M. (1989). The effect of Japanese loanwords on the acquisition of the correct range of meanings of English words (Unpublished master’s thesis). Brigham Young University, Provo, Utah, USA. Korean practical dictionary [Gukeo silyong sajeon], rev. ed. (2000). Seoul, Korea: Kyohak-sa. Nation, I. P. S. (1990). Teaching and vocabulary. Boston, MA: Heinle & Heinle. Ogasawara, M. (2008). The importance of raising awareness of English loanwords in Japanese. Journal of Aomori University of Health and Welfare, 9(1), 29-34. Uchida, E. (2001). The use of cognate inferencing strategies by Japanese learners of English (Unpublished doctoral dissertation). University of Essex, UK. Uchida, E. (2002). The role of English loanwords in ELT classrooms. Journal of Liberal Arts, 113, 19-43. Uchida, E. (2007). Oral and written identification of L2 loanword cognates by initial Japanese learners of English. The English Teacher, 31(9), 19-22. Van Benthuysen, R. V. (2004). Japanese EFL students’ awareness of English loanword origins. Bunkyo Gakuin Daigaku Gaikokugo Gakubu Bunkyo Gakuin Daigaku Tankidaigaku Kiyo, 4, 169-174.

Name Affiliation Email

David E. Shaffer English Language Dept., College of Foreign Languages, Chosun University [email protected]

APPENDIX A Individual Cognate Quality Scores of a Random Sampling of English Cognates Borrowed Word

Frequency Level

Type Score

Definition Score

Shortening Score

Group

1K

1

0

3

Handle

1K

2

0

3

Sale

1K

1

0

3

3.5

Camera

2K

1

3

3

5.0

Card

2K

1

0

1

1.5

Shower

2K

1

0

3

3.5

Heel

3K

1

1

3

Counter

3K

1

0

3

3.5

Initial

3K

1

2

3

4.5

APPENDIX B Multiple-Choice Cognate Questionnaire 1. Let’s go to the department store and look at the items on _______. a. sale b. display c. the racks d. the shelves 2. I think he’s trying to make _______. a. an impression b. a tribute

c. soap

d. a joke

3. You’re such ________. a. a dancer b. a gentleman

c. an expert

d. a hypocrite

4. He gave me his _________. a. ranch b. card

c. disease

d. patent

5. My parents gave me some money for _________. a. glasses b. lunch c. a jacket

d. the movies

6. The _______ amused the girl for hours. a. confusion b. singing

c. recreation

d. game

7. His name appears in the _________. a. movie b. newspapers

c. magazine

d. dictionary

8. The _______ earned his position through hard work. a. chairman b. minister c. teacher

d. professor

Restrictions (-1)

Total 3.5

-1

-1

3.0

3.0

9. We know we belong to the _______. a. society b. group

c. church

d. company

10. He imagined where the _______ might be buried. a. knife b. youth c. governor

d. remains

11. He was standing on the ________ when I noticed him. a. box b. desk c. chair

d. steps

12. His death came as a _________ to everyone. a. surprise b. shock c. blow

d. relief

13. The ________ will need to cook for at least an hour. a. meat b. peas c. chicken

d. fish

14. We plan to go to the _________ this weekend. a. performances b. benefits c. festival

d. exhibit

15. The _______ needed a large mansion to live in. a. doctors b. lawyers c. teams

d. giants

16. His _______ was educated in France. a. assistant b. partner

c. companion

d. boss

17. They changed their _______ easily. a. mood b. beliefs

c. opinions

d. attitudes

18. That’s a great _______. a. value b. idea

c. community

d. position

APPENDIX C Survey Questions for Identification of Modifications in EBLs 1. McDonald’s is a famous FAST FOOD restaurant. (맥도날드는 *패스트푸드* 음식점으로 유명하다.) 2. I didn’t recognize him because he was wearing SUNGLASS. (*썬글라스*를 껴서 그를 알아보지 못했다.) 3. It’s hot in here. Can you turn on the AIRCON? (더운데 *에어컨* 좀 틀어줄래?) 4. I joined the tennis CIRCLE. (나는 테니스 *서클*에 가입했다.) 5. I love SOFT CREAM. (나는 *소프트크림*을 좋아한다.)

6. The VARIETY SHOW that was on TV last night was really funny. (어젯밤에 본 *버라이어티쇼*는 매우 재미있었다.) 7. The room is only two METER wide. (방 넓이가 겨우 2 *미터*다.) 8. Would you turn on the TV with the RIMOCON? (*리모컨*으로 텔레비전을 켜줄래?) 9. Can I use this CONSENT for the fan? (선풍기 켜게 *콘센트* 좀 사용할 수 있을까?) 10. Let’s have them give us a MORNING CALL at 07:00. (아침 7 시 *모닝콜*을 신청하자.) 11. The RECYCLING movement is promoted all over the world. (*리싸이클링*은 전 세계적으로 홍보되는 캠페인이다.) 12. It is bad MANNER to talk with your mouth full. (입안에 음식이 가득할 때 말하는 것은 좋지 않는 *매너*다.) 13. She went to a beauty salon yesterday to get a SHORTCUT. (그녀는 어제 미용실에서 *쇼트컷*으로 머리를 손봤다.) 14. He was caught CUNNING on the exam. (그는 시험 중 *컨닝*하다가 걸렸다.) 15. What is her CHARMING POINT? (그녀의 *차밍포인트*는 뭘까?) 16. Would you like an alcoholic beverage, or a SOFT DRINK? (알코올음료와 *소프트드링크* 중 어떤 것을 드릴까요?) 17. How many boxes of chocolates did you get on VALENTINE DAY? (*발렌타인데이*에 초콜릿 몇 개 받았어?) 18. I went to the SUPER to buy some chocolate. (초콜릿을 사러 *슈퍼*에 갔다.) 19. Choi Bul-Am is one of the most famous TALENTs in Korea. (최불암은 한국에서 가장 유명한 *탤런트* 중 한 명이다.) 20. It seems that she cut her hair to IMAGE CHANGE. (그녀는 이미지체인지를 위해 머리를 자른 것 같다.)

21. Good SKIN CARE will keep you looking younger. (동안의 비결은 좋은 *스킨케어*에 있다.) 22. I need to buy a new PAJAMA for winter. (겨울에 입을 *파자마*를 사야겠다.) 23. Prices are high here, but the AFTER SERVICE is really good. (여기 가격은 비싸지만 *애프터서비스*가 훌륭해.) 24. He has good SENSE in clothes. (옷 입는 *센스*가 있네.) 25. SKINSHIP is very important for kids. (*스킨십*은 어린이에게 매우 중요하다.) 26. Can I use this COUPON? (이 *쿠폰* 사용해도 될까?) 27. I always wear a tie with my Y-SHIRTS. (*와이셔츠*를 입을 땐 항상 넥타이를 맨다.) 28. Mary is an INTELLI woman. (메리는 *인텔리* 여성이다.) 29. Turn the HANDLE to the right when you want to turn right. (우회전을 하려면 *핸들*을 오른쪽으로 돌려야 한다.) 30. My next car is going to be an OPEN CAR. (다음엔 *오픈카*를 꼭 장만해야겠다.) 31. I want to have chocolate cake for DESSERT. (*디저트*로 초콜릿케익이 먹고 싶다.) 32. Every morning I have a SCRAMBLE for breakfast. (아침에 항상 *스크렘블*을 먹는다.) 33. I enjoy LEPORTS on weekends. (나는 주말에 *레포츠*를 즐긴다.) 34. I got Tom Cruise’s SIGN. (톰크루즈 *싸인*을 받았다.) 35. Every day begins with a cup of MORNING COFFEE. (매일 아침 *모닝커피* 한잔으로 하루는 시작된다.)

36. He suddenly changed the TOPIC of conversation. (그는 갑자기 대화의 *토픽*을 바꿔버렸다.) 37. I always use my HAND FREE while driving. (운전시 항상 *핸드프리*를 사용한다.) 38. The cost of living went up seven PRO last year. (작년에 생계유지비가 7 *프로* 올랐다.) 39. How does she manage to keep her STYLE when she eats so much? (저렇게 많이 먹는데 어떻게 저런 *스타일*을 유지할 수 있을까?) 40. It’s common for college couples to DUTCH PAY. (캠퍼스 커플끼리는 *더치페이* 하는 것이 일반적이다.) 41. I have to buy shampoo and RINSE because I’m out of both. (샴푸랑 *린스*가 다 떨어져서 사야 돼.) 42. Our whole family likes SKI. (우리 가족 모두가 *스키*를 좋아한다.) 43. INFLAY has worsened for the past three years. (지난 3 년간 *인플레*는 악화되었다.) 44. There are three MENUs for lunch today: spaghetti, fried rice, and fish fillet. (오 늘 점심*메뉴*는 스파게티, 볶음밥, 생선가스 3 가지다.) 45. Always turn your HAND-PHONE off during class. (수업 중엔 *핸드폰*을 꺼두는 것이 좋다.) 46. John is leading 40-LOVE in his first game of the tennis tournament. (존은 테니스 대회 첫 경기를 40 대 *러브*로 앞서고 있다.) 47. "WELLBING” fever has spread throughout Korea. (전국은 지금 *웰빙*열풍에 휩싸였다.) 48. Tom Cruise is still looking for a new CODI. (톰크루즈는 아직도 *코디*를 구하는 중이다.) 49. Hiking three times a week is a good way to DIET. (주 3 회 등산은 좋은 *다이어트* 방법이다.) 50. This is already the third HOME-IN in the last five minutes. (5 분 사이에 벌써 세 번째 *홈인*이다.)