The Language Teacher

What would you think if you were told that students in Japan, even junior high students, are already familiar with more than a third of the most useful words of English. Unbelievable? Actually there is much truth to this outrageous statement.

Taro's teacher writes advice on the blackboard, a high frequency word in English-one of the 2,000 most common and useful headwords in the lexicon (Nation, 1990, p. 14). As a headword, advice is associated with eleven other words and word forms, such as advisability and advises. These and other high frequency words are the most likely to occur in the English we read and listen to.

Fortunately, Taro's lexicon already contains many English words that have been added to Japanese and have become enduring parts of its loanword lexicon. For Taro, adobaisu is pretty close in sound and meaning to the advice that the teacher expects him to learn.

But, as any English teacher who has heard Retsu ingurishu! knows, loanwords have been transformed to various degrees and are different from the English basewords from which they are derived. Nevertheless, this paper shows that Japanese loanwords are a preexisting lexical resource that students can employ in more effective ways to improve their acquisition of English vocabulary.

First, we will briefly review some of the changes basewords have undergone to become Japanese loanwords. Next, after surveying the research on the influence of Japanese loanwords on English vocabulary acquisition, my own classroom research data will help confirm that the recall and recognition of lexical items with loanword cognates, i.e., basewords, is considerably better than for those without, i.e., nonbasewords. Finally, an estimate of the number of loanwords that correspond to a corpus of 1,942 high-frequency English vocabulary will open possibilities for new teaching approaches.

The Transformations of English to Japanese

Anyone knows that the average Japanese rice farmer would have extreme difficulty communicating if suddenly confronted with a native English speaker. This is despite the fact that the rice farmer's native Japanese is chock full of English, to the tune of 20,000 or so words added over the years (Miura, 1979, p. 7) or almost 10 percent of the Japanese language (Shibatani, 1990, p. 153).

This is due to the process that transforms the English language into Japanese. Below are but four of the many changes English words may undergo (Daulton, 1995, pp. 133-138),

Rephonalization

When listening to English spoken with a strong Japanese accent, as in Ai rabu sukyuuba daibingu, it undergoes a bewildering transformation.

Rephonalization results from the radically different Japanese phonological system, which governs the way in which foreign words re transliterated and pronounced (Yamagiwa, 1942). Notably, Japanese is a consonant/vowel (C/A) language (e.g. neko and banana) whereas English consonants and vowels are combined in a variety of patterns, as in "perplexing" (Romaine, 1989). Therefore, McDonald's, rephonalized becomes makudonarudo.

Truncation (Shortening)

English words are often shortened in ways perplexing to native speakers (Shibatani, 1990). For example, department store becomes depaato and television becomes terebi. An even more extreme example is jomo, from joy of motoring. The most semantically important morphemes, generally the first, are usually all that remain of the original baseword.

Speech Part Modification

A baseword's part of speech is not necessarily considered when the loanword is used in Japanese (Henderson, 1948). For example, the English noun, harmony, is used and conjugated like a regular Japanese vowel in haamoru (to harmonize). Many other foreign words, especially nouns, can be verbalized by adding suru (Park, 1987, p. 36). For example meiku suru, literally means to do (put on or wear ) make up.

Semantic Modification

To the vexation of many English nitpickers, the meanings of English words are modified freely to meet Japanese lexical needs. For example, the word demagogue has been shortened to dema, and then assigned the quite different meaning of "a false rumor" (Park, 1987, p. 99).

In a phenomenon called "semantic narrowing," only one of the possible English meanings is adopted; for example,furii (free) generally means unrestrained in Japanese but not gratis (Shibatani, 1990, p. 151). In this case,furii has a "shifted" meaning, one that is strictly limited. There are many other types of semantic change.

As we see, English loanwords in Japanese are different in form, function, and/or meaning from their English basewords. Despite this discouraging picture, we shall find that the influence of loanwords on English vocabulary acquisition can be highly beneficial.

Japanese Loanword Influence on English Vocabulary Acquisition

Although even a TESOL specialist would be hard pressed to explain the precise mechanisms of language transfer, few would disagree that L2 vocabulary learning is influenced by L1 vocabulary. Nation further asserts that there is strong evidence that L1 and L2 vocabulary are stored together in a state which encourages borrowing and interference (1990, pp. 32-33). Furthermore, Nation states that when an L2 word resembles a word in the learners' L1, that it will have a lighter "learning burden" (p. 35). For example, just as French speakers learning English find the learning burden of words like table, elementary, and dentist very light because of the existence of table, elementaire and dentiste in their own language, Indonesian speakers find the word communication easy to learn because of the Dutch loanword komunikasi in the Indonesian language (Nation, 1990, pp. 35, 40).

Likewise, the sparse research available that focuses specifically on the affect of L1 Japanese knowledge effecting L2 English vocabulary acquisition has shown a generally positive effect of Japanese loanwords on English vocabulary acquisition (Brown & Williams, 1985; Kimura, 1989; Yoshida, 1978).

The Effect of Japanese Loanwords on a Child's ESL Vocabulary Acquisition

Yoshida (1978) found that English loanwords helped a Japanese-speaking child living in the United States acquire the related English basewords quickly (p. 100). The subject, Mikihide, was three years and five months when the observation started (p. 92). He had had no previous English study in Japan.

Yoshida found that loanwords in Japanese helped Mikihide learn English words more quickly at his nursery school because of their similarity as cognates (1978, p. 99). The cognates were particularly helpful in enlarging Mikihide's receptive vocabulary. In particular, loanwords were helpful for comprehending new English vocabulary items. Presented with 22 English basewords in a Peabody Picture Vocabulary Test (PPVT), Mikihide comprehended 19 words such as table (teeburu) and orange (orenji). No figure was given for nonbasewords.

Regrettably, in production of these basewords, Mikihide's pronunciation was not always recognized by English speakers (Yoshida, 1978, p. 100). Needless to say, this is because some of the basewords were pronounced using the Japanese sound system. For example, table was changed to English /teybl/ from the Japanese teeburu, but orange remained as orenji.

The Effect of Loanwords on College-Level ESL and EFL

Vocabulary Acquisition While Yoshida's study dealt with a young child, Brown and Williams (1985), and Kimura (1989) studied college-level, Japanese learners of English.

Brown and Williams (1985) tested whether EFL students, on hearing an English word, understand the word better if it is a loanword cognate (p. 133). The subjects, second-year English majors in Japan, selected from four choices the correct Japanese definition of the English word they heard on a tape.

The researchers found that the Japanese students of English, when they heard English words, understood the meanings of basewords better than nonbasewords (Brown & Williams, 1985, p. 144). Although scores were better for all basewords than for nonbasewords, students did better when not told that the correct responses would be basewords; students scored 5% and 3% better on basewords respectively (p. 140). Brown and Williams state, "Students may do better when they make the English association on their own" (p. 141).

Kimura (1989) studied college-level Japanese EFL and ESL groups. Kimura employed a multiple-choice test that matched English words to three definitions offered for basewords and nonbasewords. For basewords, the definitions were: 1) the typical Japanese loanword meaning (which may or may not have been adopted by Japanese); 2) a correct English meaning that was different from the Japanese loanword meaning, and 3) a nonsense meaning (p. 30). For nonbasewords, two correct English definitions and a nonsense one were used (1989, p. 20). Students chose the appropriate definitions.

Kimura (1989) found that both ESL and EFL groups had significantly better scores for English basewords than for nonbasewords (p. 77). Both groups scored about 5% better for basewords over nonbasewords, as in the study by Brown and Williams (1985). Kimura asserts that because basewords are the most accessible, they can be an effective tool for learners to learn related vocabulary (p. 17).

Kimura (1989) further determined that English loanwords might even assist acquisition of the correct range of meanings of English words (p. 49). This contradicts Lado's (1972) assertion that loanword knowledge limits the range of English meanings known to learners (p. 285). Kimura proposed that loanword knowledge can encourage learners to add more meanings to the loanword meanings they already know, although they may lack confidence in using the loanword lexicon as a resource (p. 80).

Based only on the studies of Brown and Williams (1985), Kimura (1989), and Yoshida (1978), a relationship where English loanwords in Japanese aid the acquisition of their baseword counterparts emerges. Because of the lack of other experimental research (also mentioned by Brown and Williams, 1985, p. 130), I conducted the following study which substantiated that Japanese loanword cognates are a resource in English vocabulary learning.

Acquisition Patterns of Baseword Vocabulary vs. Non-Baseword Vocabulary

Research Goal

According to the above three studies by Brown and Williams (1985), Kimura (1989), and Yoshida (1978), basewords appear to have a positive effect on English vocabulary acquisition. I attempted to confirm with my research that recall and recognition of lexical items with loanword correlations is better than for those without.

Subjects

The research subjects were 27 Japanese first-year junior college English majors. Their average TOEFL score, taken within the last six months, was 415. The low score was 367, the high score 477, for a mean TOEFL of 422. These subjects were comparable to those of Brown and Williams (1985) and Kimura (1989).

Testing Instnlment

Brown and Williams' test consisted of 200 test items, half basewords and half nonbasewords (1985, p. 137), while Kimura's test consisted of 34 English words (1989, p. 29). My testing instrument consisted of only 60 words for two reasons. First, Nation recommends between 60 and 100 test items as an auspicious size for tests of this nature (1990, p. 78), and second, selecting appropriate test items, especially locating high-frequency English words that have not been borrowed into Japanese, is quite difficult (James & Brown, 1985, p. 137).

The prompt for each fill-in was the first Japanese definition listed for it in Obunsha's Comprehensive English Japanese Dictionary (1995). (The prompt was given in the original Japanese kanji and hiragana.) I used the newer edition of the same dictionary used by Kimura (1989) in his research. An example test item (to elicit beautiful) is as follows:

The test consisted of four pages. The first page contained fifteen adjective baseword items and the second contained fifteen nonbasewords. The third page contained 15 non-baseword adjectives and the fourth page 15 non-baseword nouns.

Intrinsic vocabulary difficulty was controlled for by using the Obunsha's Comprehensive English Japanese Dictionary (1995), which ranks the difficulty of English vocabulary for Japanese learners using asterisks. Five words on each page were rated at "junior high level" (***)-as determined by the makers of the dictionary, the expectation being a 1,500-word vocabulary upon graduation (p. 2). Five more on each page were rated at "high school level" (**) for both basewords and nonbasewords-considered appropriate for students with a vocabulary of about 4,500 words (including the first 1500 words). The final five words on each page were rated at "university level" (*), which represents a vocabulary of about 6,000 words.

For the items at the university level, I tried to provide two definitions when available. This addressed the possibility, because the words had become less common, that students might not know the meaning in Japanese.

I used blank-filling questions, where the first and last letter of each item is provided, to prevent the students from guessing another acceptable answer, at which point they would quit looking for the target response. Because the first and last letter of the ideal answer were provided, the test employed recognition as well as recall.

Finally, the vocabulary was chosen at random, in a fashion similar to the system employed by Kimura (1989, p. 30) . I simply turned to a page at random in Obunsha's Comprehensive English Japanese Dictionary, and picked out the first word hat met the minimum requirements (i.e., difficulty level, adjective or noun, baseword or non-baseword).

Test Procedure

I divided the testing into two parts, in order to avoid test subject fatigue. Students were given 10 minutes to complete each half of the test, for a total 20 minutes over two days. Originally I had allotted more time, but at the 10-minute mark, all students appeared to have finished and, when asked, no one desired more time. The baseword part preceded the non-baseword one, and students were not told of the loanword connection to basewords.

Afterwards, answers that were correct but misspelled were counted. The spellings had to be close enough as to show the students understood the correct pronunciation of the word, and could theoretically utter it comprehensibly. Lastly, answers that were also spelled correctly were counted (these scores would naturally be either the same or lower than the above).

I also made notes of words whose spellings were partial or malformed-incorrect spellings that, although indicating the students could not pronounce the word, showed that at least some lexical memory had drawn them "into the ballpark."

Test Results

By both the measures of remembering a word's pronunciation (i.e. correct but misspelled), "Type 1," and remembering the correct spelling of a word, "Type 2," the student's performance was better with basewords over nonbasewords at every difficulty level. Table 1 gives a summary of the results.

Table 1: Summary of Test Results

		Type 1			Type 2
Subject of Comparison	BW	NBW	Difference	BW	NBW	Difference
Junior High School	92%	90%	2%	92%	83%	10%
High School	59%	14%	421%	44%	10%	440%
University	30%	3%	1000%	6%	2%	1300%
Overall Results	61%	36%	69%	51%	33%	55%

Note: BW = basewords; NBW = nonbasewords

Both baseword and nonbasewords performed well at the junior high school level. The difference by the first measure was slight (2%), and even when spelling was considered, basewords outperformed nonbasewords by only 10%.

When contrasting the performance of basewords and nonbasewords, the low differences for junior high-level vocabulary of 2% and 10 % resemble the 5% and 3% found by Brown and Williams (1985, p. 140) and the 5% found by Kimura (1989, p. 47). This may be due to the similarity of test item selection. Brown and Williams chose words at or below the 2,000 (most-common-English-word) level as listed by a common English-Japanese dictionary for learners (p. 137). Likewise Kimura (1989) chose three-fourths of his basewords according to a loanword dictionary that lists the most common 908 loanwords, the remaining one-fourth being based on less common loanwords; the nonbasewords were then chosen at a level comparable to these baseword selections. Thus, the word selection of both these research groups would tend to lower the difficulty level to that of my junior high school level, which was at the 1,500 word level.

Much more conspicuous results could be seen at the high school and university levels, which represented vocabulary levels of 4,500 and 6,000 words, based on Obunsha's Comprehensive English Japanese Dictionary (1995). Taken together, the basewords out-powered nonbasewords by about five times for both measures.

Looked at separately, high school, nonbasewords, which students had almost definitely encountered as part of their previous curriculums, performed poorly and university level nonbasewords, which students had had limited or zero contact with, performed miserably when compared with basewords.

Below the threshold of the "Type 1" and "Type 2" measures (i.e., whether a word's pronunciation was remembered and whether its spelling was also correct), many answers were partial or malformed. Examples of partial/malformed answers are octurve (octave) for basewords and immuency (immunity) for nonbasewords. It seems that the loanword lexicon was causing better performance even at this level, as partial and malformed (near-miss) answers were almost twice as likely to occur in baseword answers than in non-baseword answers (the alternative being to leave the question unanswered). There were 22 such answers for basewords, yet only 12 for nonbasewords (83% difference).

Many students groaned loudly about how much harder the non-baseword test was than the baseword test had been. This is despite the fact that the level of difficulty of both tests had been controlled for and was ostensibly the same.

Research Summary

As demonstrated, these junior college students answered blank-filling vocabulary questions for basewords with much greater accuracy than for nonbasewords by both the measures of (incorrect but) intelligible ,spelling and correct spelling. While the difference was relatively small for junior high school-level vocabulary, it was considerable when looking at the high school and university levels.

Given the superior performance and apparent learnability of basewords in this and other research, an astute teacher may wonder how many of the 2,000 high frequency words of English have conveniently made the voyage to the Japanese archipelago.

Loanword Cognates for High-Frequency English Words

Many English loanwords in Japanese came from high-frequency English basewords. Thus, given that Japanese loanword knowledge can be an effective instrument for English learners, a powerful arsenal of loanword cognates to high-frequency English vocabulary may be at hand.

Nation states that with a vocabulary of just around 2,000 high-frequency headwords, a learner can read and understand about 87 percent of the words of any given text (1990, p. 14). (Remember that headwords such as absorb are associated with a group of words like absorption and absorbent.) Regarding high-frequency words, he writes, "Any time spent learning them will be well repaid because they cover a lot of text and will be met often" (p. 14). Among the high-frequency word lists available, Nation feels that West's General Service List (GSL) of English Words (West, 1953), which contains 1942 high frequency headwords, has yet to be replaced as the most useful collection of vocabulary because of its realistic choice of words, the grouping of headwords with their various forms, and the information on frequency it readily provides. The results indicated that 734 of the headword groups in the GSL correlated to at least one loanword, at a rate of 38%. (See Appendix 1 for a detailed breakdown of the results).

Of course, because of those transformations that basewords undergo during their incorporation into Japanese (e.g.,rephonalization, shortening, semantic modification, and speech part modification), the level of resemblance of those high-frequency basewords to their loanwords' counterparts varied.

To estimate the similarity between loanword cognates and basewords, I examined the loanwords cognates for English basewords whose spelling begins with "a". Twenty-four loanwords correlated to 20 headword groups. These were contained in the "a", "e" and "o" sections of A Dictionary of Loanwords Usage (Motwani, 1991, pp. 1-10; 43-48;132-139) which correlates to the "a" section of the dictionary's index (pp. 230-231). (See Appendix 2 for detailed results of this comparison.)

There is research that claims that the primary meaning of a word is more transferable to another language (Kimura, 1989, p. 48). Indeed, in the case of those 24 "a" loanwords cognates, only akushon has a "shifted" meaning; shifted meanings being either severely limited or totally different from native English. In this case, akushon is always a modifier that refers to movies (Motwani, 1991).

The other 23 "a" loanword definitions, as listed in A Dictionary of Loanwords Usage (Motwani, 1991) matched one of the first three definitions listed in the Random House Webster's Dictionary (Braham, 1996). Furthermore, as many as 18 of the 24 "a" loanwords had the same definition as the first meaning listed in this English dictionary. Of these 18 loanwords, 11 words, including aachi (arch), adobaisu (advice), and akuchibu (active) had no morphological restrictions such as speech part modification or shortening-that is they were quite similar to their baseword counterpart in both form and meaning. Of those 11 high-quality cognates, 8 loanwords (including aachi and adobaisu but not akuchibu) matched to high frequency headwords (which may be more readily associated with other words in the headword group), and could be considered prime cognates.

Implications for Vocabulary Instruction in Japan

Kimura (1989) asserts that although none of the recent learning strategies for acquiring English vocabulary appear to speed acquisition significantly, the fact that English basewords that are similar to Japanese loanwords can be acquired more easily opens new possibilities for enhanced vocabulary acquisition (p. 2). Nation (1990) concurs that, because of their light learning burden, basewords can be learned very quickly by especially beginners (p. 40).

As 734 high-frequency English headword groups correlate to loanword cognates, the loanword lexicon can be tapped to allow learners to gain a large number of highly useful lexical items, particularly nouns, in a short period of time, saving harder ones for later. For more advanced English learners, the same approach could be taken to tackle, for example, the additional 800 "university-level" high-frequency words described by Nation (1990, p. 24).

To estimate how many of these university-level headword groups have loanword correlations, I did a rough estimate. The average percentage of correlations for headword groups beginning with vowels after sampling "a," "i" and "u" was 19%; the average for headword groups beginning with consonants, after sampling "g," "m," "n" and "r," was 28%. With weighing for the number of vowels and consonants in the alphabet, about 26% of university-level headword groups should correlate to loanwords. Thus there should be valuable matches for about a quarter of the 800 university-level, high-frequency headword groups.

Students should be made aware of the loanword resource that they possess. They should learn to have more confidence in their intuitions about new English vocabulary. Kimura (1989) proposes that the loanword lexicon may even be used to develop a native-like semantic intuition (pp. 79, 89). To this end, Kimura advises that teachers and learners pay special attention to loanwords in formal instruction. Nation (1990) notes, "The more the teacher or the course designer draws attention to the similarities and patterns (between L1 and L2 vocabulary), the greater the opportunity for transfer" (p. 49). Brown and Williams (1985) warn, however, that while awareness of the loanword resource is helpful, explicitly associating particular English vocabulary to Japanese loanwords may diminish any potential benefits (p. 133).

Therefore, at the junior high school level, for example, where curricula and vocabulary to be taught are predetermined and basewords and nonbasewords appear together in texts, teachers can assume that basewords are understood and focus all their vocabulary instruction attention on nonbasewords. The basewords, which are the most familiar English words (Kimura, 1989, p. 17), will provide contextual clues to the nonbasewords that neighbor them. This technique both avoids the possible confusion found by Brown and Williams (1985), as well as allots time more efficiently to where it is needed.

As we have seen, basewords and loanwords are seldom identical. However, the points at which loanwords and basewords differ significantly enough to cause confusion can become focal points for learning (Nation, 1990, p. 35). This turns a potential risk of negative transfer and confusion into a classroom asset.

In particular, the pronunciation differences between basewords and loanwords can be a stumbling block. Yoshida (1978) warns that word stress "is important if the learner is to be understood by native speakers" (p. 99). Thus pronunciation instruction should accompany vocabulary instruction (see Daulton, 1997).

Teaching English vocabulary using Japanese loanword cognates naturally requires some knowledge of Japanese language. Thus some researchers such as Topping (1962) urge teachers to become familiar with loanwords in their students' native language (p. 287).

This paper has focused mostly on how loanwords in Japanese can aid the acquisition of English vocabulary. Perhaps the converse is also true. Since the Japanese loanword lexicon consists of many of the high-frequency vocabulary of Japanese, native English speakers can employ their own (baseword) knowledge to quickly learn many valuable Japanese vocabulary items such as taimu rimitto (time limit).

I would like to thank Paul Nation, Victoria University of Wellington, for his advice and encouragement.

Braham C. G. (editor in chief) (1996). Random HouseWebster's Dictionary. New York: Ballantine Books.

Brown, J. B. and Williams, C. J. (1985). Gairaigo: a latent English vocabulary base? Tohoku Gakuin University Review: Essays and Studies in English Eibungaku, 76,129-146. Sendai, Japan.

Daulton, F. E. (1995). Gairaigo: Japan's adaptation of English. Journal of Nanzan junior College, 23,129-143. Nagoya, Japan.

Daulton, F. E. (1997). Katakana English and the teaching of pronunciation. Journal of Nanzan Junior College, 24, 43-54. Nagoya, Japan.

Henderson, H. G. (1948). Handbook of Japanese grammar. Boston: Houghton Mifflin Company.

Kimura, M. (1989). The effect of Japanese loanwords on the acquisihon of the correct range of meanings of English words. Masters Thesis, Brigham Young University, Department of Linguistics.

Lado, R. (1972). Patterns of difficulty. In K. Croft (Ed.) Readings on English as a second language (pp. 277-91). Cambridge, Massachusetts: Winthrop Publisher, Inc.

Miura, A. (1979). English loanwords in Japanese: A selection. Rutland, Vermont & Tokyo: Charles E. Tuttle Company, .

Motwani, P. (1991). A dictionary of loanwords usage. Tokyo: Maruzen Co., Ltd.

Nation, I.S.P. (1990). Teaching and learning vocabulary. Boston: Heinle & Heinle Publishers.

Obunsha. (1995). Obunsha's comprehensive English-Japanese dictionary. Japan: Obunsha.

Park, W. (1987). Western loan-words in Japanese. Stockholm University, Department of Oriental Languages.

Romaine, S. (1989). Bilingualism. Oxford and New York: Basil Blackwell Ltd.

Shibatani, M. (1990). The languages of Japan. New York: Cambridge University Press.

Topping, D. M. (1962). Loanblends, a tool for linguists. Language Learning, 12(4), 287.

West, M. (1953). A general service list of English words. London: Longmans, Green and Co.

Yamagiwa, J. K. ( 1942). Modern conversational Japanese. New York: Mc.Graw-Hill Book Company, Inc.

Yoshida, M. ( l978). The acquisition of English vocabulary by a Japanese-speaking child. In E. M. Hatch (Ed.) Second language acquisition, (pp. 91-100). New York: Newbury House.

Appendix 1-loanword/baseword

correlations from A to Z

(as a fraction of correlations over number of headword groups and percentage)

a	20/121	17%	n	17/45	38%
b	46/125	37%	o	17/50	34%
c	72/134	54%	p	65/145	45%
d	28/100	28%	q	3/12	25%
e	16/80	20%	r	47/111	42%
f	50/110	46%	s	104/265	39%
g	28/51	55%	t	51/128	40%
h	26/74	35%	u	5/23	22%
i	13/47	28%	v	4/21	19%
j	6/16	38%	w	29/91	32%
k	6/15	40%	x	0/0	n.a.
l	33/76	43%	y	4/10	40%
m	44/92	48%	z	0/0	n.a.

---

Appendix 2-loanword/baseword proximity level

semantic proximity

1. Same meaning as first listed in Webster's Dictionary (1996).
2. Same meaning as second listed.
3. Same meaning as third or later
4. Shifted: strictly limited or totally different usage (Park, p. 59).

headword or associated word match

a headword match

b associated word match

baceword	loanword	match level	morphological restriction
accident	akushidento	1a	.
action	akushon	4b	(always modified)
active	akuchibu	1b	.
address	adoresu	1a	.
advice	adobaisu	1a	.
adviser	adobaizaa	1b	.
adventure	adobenchaa	1a	(always modified)
after	afutaa-kea, afutaa-saabisu	1a	(coined)
agent	eejento	1a	.
air	ea	1a	(always modified)
all-	ooru	1a	.
apple	appuru	1a	.
apply	apurai (suru)	3a	.
appointment	apointo	1b	.
April	eipuriru fuuru	1a	(always "April Fool")
arch	aachi	1a	.
arrange (to plan)	arenji (suru)	3a	.
arrenge (music)	arenji (suru)	3a	.
art	aato	1a	(always modified)
artist	aachisuto	1b	.
association	asoshieeshon	1a	(used as a suffix)
at	atto houmu/ atto randamu	3a	.
attack (e.g. amountain)	attaku (suru)	1a	.
attraction (e.g. circus)	atorakushon	3b	.