Dina R. Yoshimi and Haidan Wang, Editors, 2007
Selected Papers from Pragmatics in the CJK Classroom: The State of the Art
Table of Contents

Read paginated .pdf version


The development of pragmatic competence in children learning Japanese as a second language1

Kimberly Jones
University of Arizona



The Japanese education system today is faced with an increasing number of students who speak little or no Japanese. By 2005, there were over 20,000 such students in the public schools alone. Depending on the school, various arrangements exist for Japanese as a Second Language (JSL) instruction for these children. The primary goal of JSL instruction in the public schools is generally to help students communicate with teachers and classmates and become able to understand academic content. Instruction tends to focus primarily on vocabulary, grammar, and orthography. Although Japanese is a language well-known for linguistically marking a variety of social, situational, and interpersonal distinctions, like their counterparts in Japanese as a Foreign Language (JFL) classrooms, young JSL learners rarely receive much explicit instruction in these more pragmatic aspects of the language. However, being immersed in a Japanese-speaking community, they do have an advantage over JFL learners in acquiring socioculturally appropriate ways of interacting.

This case study discusses three English speaking children's development of Japanese language skills, first touching on their overall acquisition of grammar, then focusing on certain pragmatic aspects of Japanese. Some of the questions considered are: How do young JSL learners acquire socioculturally appropriate ways of speaking? How quickly and how well are various pragmatic distinctions acquired? Do different learners show similar patterns? What implications might the answers to these questions have for language teaching?

The pragmatic aspects analyzed for this paper include the use of interactional particles, different politeness levels, regional versus standard dialect, gendered language, donatory verbs, and person reference and address forms. The three children whose linguistic development was studied varied considerably in how quickly they acquired the ability to use these aspects of Japanese, and some pragmatic features were not learned as quickly or completely as others. Not surprisingly, however, their acquisition of these features was more rapid and complete than what has been described for children (or adults) learning Japanese in classrooms.


Study participants and data

The children whose second language development is described in this paper are three American siblings whose first language is English. Their family moved to western Japan (Shimane Prefecture) for one year when they were 7, 5, and 2. The oldest child, Miranda, is a girl; the middle child, Kent, and the youngest, Eric, are boys.2 The children knew a little Japanese when they first went to Japan, having had private lessons in Japanese for one hour each week for a little over a year before the move.3 In addition to exposure to Japanese during their lessons, the children also heard some Japanese from their mother, a nonnative speaker of Japanese. Their father, although he understood a fair amount of Japanese, did not speak it at home. By the time of their move to Japan, the children had learned a few hundred words in Japanese and were accustomed to being addressed in simple Japanese by their Japanese teachers and their mother. However, the children themselves spoke very little Japanese except during their Japanese classes, and the Japanese utterances they did produce on their own were almost all one-word utterances.

Once in Japan, Miranda attended a public elementary school and Kent and Eric attended a preschool (hoikuen). Their school and preschool experiences, plus after-school time spent playing with children in their neighborhood, resulted in the children being immersed in Japanese for much of the day. In addition, the local school district had a program through which non-Japanese-speaking children were given JSL instruction by volunteer tutors. Through this program, Miranda had individual JSL tutoring at her elementary school. Three different teachers each tutored her once a week for approximately an hour and a half, for a total of about four and a half hours of instruction each week. Kent and Eric had no formal Japanese instruction during the year in Japan.

After one year in Japan, the family returned to the United States. The children continued to use Japanese in weekly Japanese classes for heritage learners, once a week with a Japanese babysitter, and during the summers, which their family continued to spend in Japan.

This preliminary analysis of the children's pragmatic development in Japanese is based on the analysis of two sets of data: 1) videotapes of family dinner conversations that were recorded in the children's home (generally with only the family present, although on two occasions there were Japanese visitors present as well) and 2) a journal kept by their mother in which she recorded information about the children's language use. The dinner conversations were recorded at 12 different points during the initial year of residence in Japan, beginning with week 11 and ending with week 52.4


Overall second language development

The children's linguistic development in Japanese followed a typical pattern. Their first attested utterances were phrases and routines that the children had presumably heard and memorized and were repeating in an appropriate context (or at times were simply practicing in an unrelated context).5 All three children were observed using these phrases and routines from about their fourth week in Japan. For example, by that time the children were easily able to produce the minimal verbal accompaniments needed for playing hide-and-seek in Japan (counting to twenty, moo ii kai 'are you ready?', and mada da yo 'not yet', etc.).6 Other examples of this type of language use included those in examples 1 through 7 below.7

1) Miranda: bikkuri shita. [week 4]
(I) was surprised.
2) Miranda: abunai, abunai! [week 4]
Watch out, watch out! (literally: 'dangerous, dangerous!')
3) Miranda: onaka ippai tabeta. [week 4]
(I) ate (till I was) full.
4) Kent: dame! [week 4]
(That's) no good.
5) Eric: chotto matte! [week 4]
Wait a second!
6) Kent: suki ja nai. [week 5]
(I) don't like (it).
7) Kent: doo itashimashite. [week 5]
You're welcome.

Within a few more weeks, the children's Japanese production evidenced a growing ability to produce novel sentences. By week 8 they began to alter set verbal routines in order to express new meanings. For example, before dinner on several evenings during week 8, the children demonstrated and rehearsed the routines that they used with their teachers and classmates at school before eating lunch. One night Kent initiated a change in the final line of his routine. Instead of oishii kyuushoku itadakimasu! (We eat our delicious school lunch!'), he instead proclaimed oishii gohan itadakimasu! (We eat our delicious meal!). The next evening, with pizza on the menu, this phrase underwent yet another transformation, becoming oishii pizza itadakimasu! (We eat our delicious pizza!).

The following example, taken from a family dinner tape, shows some of this sort of language practice. In it, Kent not only practices the routine, changing kyuushoku to gohan, but also prompts his parents to take the teacher's role in the routine.

8) [week 8] shem
shem 1 Kent: te o awase-, te /o awasemasho./
Put our hands-, let's put our hands together.
shem 2 Miranda: /o awasemasho./
put them together.
shem 3 Kent: patchin.
[onomatopoeia to represent putting hands together]

oishii gohan, itadakimasu!
We eat our delicious meal!

shem 4 Miranda: Let me do it my way now. te o awase/mashoo./
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemLet's put our hands together.
shem 5 Kent: /No no/, somebody is supposed to say "doozo".
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemPlease eat.
shem 6 Mother: hai /doozo./
Yes, please eat.
shem 7 Father: /doozo./
Please eat.

Because the elementary school and the preschool had slightly different routines, Miranda was eager to practice her routine as soon as Kent finished his. Note, however, Kent's insistence on having the entire routine completed appropriately before he was satisfied. A few minutes after the previous exchange, and well after the family had begun eating, Eric suddenly launched into the same routine. He found it rather difficult, however, pausing several times and relying on some coaching from Kent before he was able to manage more or less to alter the routine so that it was appropriate for the dinner context.

9) [week 8] shem
shem 1 Eric: [hands together in front of him] oishii kyuu-, o-, oishii,
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemour delicious school l-, del-, delicious,
shem 2 Kent: gohan.
shem 3 Eric: o-, gohan, itadakimasu.
del-, we eat, our meal.
shem 4 Mother: hai doozo.
Yes, please eat.
shem 5 Miranda: I guess in Eric's they don't do…

Miranda started to make a comment on how Eric's lunch routine differed, but then trailed off. After a brief pause, Kent then made a comment on his sister's shirt that triggered an interaction in which Miranda tried to teach her brother how to produce some simple sentences in Japanese.

10) [week 8] shem
shem 1 Kent: I like that snowman, on your shirt.
shem 2 Miranda: yukidaruma. (5) yukidaruma suki? (3)
Snowman. (You) like (the) snowman?
shem 3 Eric: I like that /snow/man on your shirt.
shem 4 Miranda: /hai./
shem 5 Kent: yukidaruma o, um daisuki.
(I) really like (the) snowman.
shem 6 Miranda: yukuuidaruma, suki desu. If I say, Kent, if I say,
(I) like (the) snowman.

yukidaruma suki desu ka, you say,
(Do you) like (the) snowman?

yuki-, hai, yuuki daruma sukii desu.
Snow-, yes, (I) like (the) snowman.

shem shem
[2-second pause]
shem 7 Kent: O.K.
shem 8 Miranda: and if you don't like it, you say,

iie, yuki daruma suki [laughing] arimasen. (2) I think.
No, (I do) not like (the) snowman.

A number of incorrect utterances can be observed in this interaction. In line 5, Kent uses the case particle o, typically used to mark direct objects, instead of the "subject particle" ga that would be more appropriate, then hesitates before producing the predicate daisuki. In line 6, Miranda elongates vowels oddly in two words (yuuki instead of yuki 'snow' and sukii instead of suki 'like') and mispronounces another word (yukuuidaruma instead of yukidaruma 'snowman'), and in line 8, she produces suki arimasen instead of the appropriate negative form, suki ja arimasen. Interestingly, using o instead of ga to indicate the item that is liked, and problems with negation, such as omitting the (less perceptually salient) ja in ja arimasen, have both been noted in L1 learners of Japanese (Clancy, 1985). Another of Kent's early attempts at a sentence can be seen at the beginning of a mixed Japanese and English utterance in example 11.

11) Kent: moo okke nai. Right? It's not finished. It's not OK. [week 10]
already not OK

Judging from the context in which this was produced, and from the English in the second half of the utterance, Kent's intended meaning seems to have been something like mada ookee ja nai (it's not OK yet). However, instead of mada ([not] yet), he used moo (already). He also mispronounced the word 'OK' and omitted ja from the negative form.

Eric lagged very slightly behind his older siblings in attempting to produce novel sentences, at least as far can be determined by his production of Japanese at home or on other occasions when his mother was present to observe his language use, but he too can be observed using simple sentences in the data beginning in week 10:

12) Eric: kani-san no nai. [week 10]
(It) is not (a) crab.

Eric's intended meaning seemed to be kani-san ja nai (it's not a crab). Negating nouns by adding no nai was a common pattern for him right around that time.

As they began to show a growing knowledge of Japanese grammar, the children also began to use various compensatory strategies to communicate content that was too complex for what they were capable of doing grammatically. These strategies included stringing words or phrases together, using gestures and pantomime to act out what they were trying to express, and utilizing different voices. This was especially true of the oldest child, perhaps because she was at a cognitively higher level than her brothers and had more complex meanings that she wanted to express. For example, Miranda had been in Japan for 10 weeks when she encountered some classmates near her brothers' preschool as she and her father were on their way to pick up the boys. Attempting to explain why she was there, she said:

13) Miranda: otooto, futari, koko de. [week 10]
little brother, two, here ["My two little brothers are here."]

Similarly, the following week, she saw a neighborhood friend, Ruka, coming home from school just as she and her mother were leaving to pick up her brothers. Trying to say that after she went to Himawari preschool with her mother, she would go to Ruka's house to see whether she could play, Miranda produced the following utterance:

14) [week 11] shem
shem Miranda: Miranda-san, okaasan, Himawari. Miranda-san, Ruka-san no uchi,
Miranda, Mother, Himawari. Miranda, Ruka's house,
shem shem [miming ringing a doorbell] din-don!
shem shem [changing to enthusiastic voice] Ruka-san, asoboo!
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemRuka, let's play!

The children's ability to produce more accurate and more complex Japanese sentences gradually improved. There were clear differences in the pace of that improvement, however. Miranda became fluent much faster than her brothers. By the time the family had been in Japan for about nine months, she was rarely observed to make grammatical mistakes in Japanese and was able to communicate quite comfortably. At that point, Kent had just recently begun using more complex constructions and case particles, but his Japanese output was still relatively full of errors, primarily mistakes with verb inflections. Eric's Japanese developed the most slowly of the three children.

The children's use of Japanese at home increased dramatically over the course of the year. In the beginning videotapes, they spoke almost entirely in English. During the last dinner that was recorded, which took place during week 52, over 90% of Miranda's and Kent's speech and over 80% of Eric's speech was in Japanese. This tendency to use Japanese even at home was probably reinforced by the fact that their father had returned to the United States the previous week, and their mother was speaking to them almost entirely in Japanese at that point.


Development of second language pragmatics

The children's use of various pragmatic aspects of Japanese also developed gradually over the course of the year. Below, I will look at various aspects in turn, focusing on how, when, and to what degree each child displayed an understanding of various pragmatic factors in his or her production of Japanese.

Interactional particles

Interactional particles in Japanese are notoriously difficult for adult learners, who tend to under-use them and use them inappropriately for a long time. (See, for example, Ohta, 1993, 1994; Sawyer, 1992; and Yoshimi, 1999.) All three children began to produce the common interactional particles ne(e) and yo relatively early and used them quite naturally from the beginning. As was true of the earliest multi-word utterances that the children produced, the earliest attested uses of interactional particles were not in novel utterances, but in phrases that the children had almost certainly heard before. This is despite the fact that they were experimenting with producing novel sentences by the time they began to use interactional particles around week 12. The following are some examples of this very early particle use, all involving either ne(e) or yo.

15) Eric: ii yo. [week 12]
That's OK!
16) Eric: soo yo. [week 12]
That's right!
17) Miranda: chotto matte yo. [week 13]
Wait a second!
18) Miranda: samui desu ne. [week 15]
It's cold, isn't it.
19) gomen ne. [used by both Miranda and Eric during week 15]

From the data examined for this paper, Kent appeared to lag behind Miranda and Eric in producing ne(e) and yo, with his first attested use of both occurring relatively late, during week 28 (see Figure 1). However, he used very little Japanese in the home conversations recorded during the first seven months of the year, and it is possible that he was using these particles in all-Japanese contexts well before he was first observed using them at home.

This image shows Figure 1, as described in the text
Figure 1. First attested production of common interactional particles

After they began reproducing commonly heard sentences with interactional particles, the children began using the particles on the end of what seemed to be novel utterances. They also gradually began to produce particles other than ne(e) and yo, such as the reflective kana and na and the feminine rising intonation wa (used only by Miranda). The order in which the children began using the particles is similar to what is seen for children acquiring Japanese as a first language (Clancy, 1985), who typically use yo and ne(e) at a very early age, generally before age 2, and then begin using kana at a slightly later stage, typically right around age 2. Na(a) also comes in at that stage as well. (Miranda and Kent actually began to use na[a] slightly later than kana, and contain no instances of Eric using na[a].) Examples 20 through 27 show the use of interactional particles with more complex (though not always grammatically accurate) sentences and the use of particles other than yo and ne(e). The particle gaa, which is common in the regional dialect of the area in which the children lived, will be further described below.

20) Eric: [teasing; he was in the Himeyuri class] himeyuri-san ja nai yo. [week 14]
I'm not in the Himeyuri class!
21) Miranda: ninjin da yo. shinjin /tte itta./ [week 28]
It's ninjin ['carrot']. (You) said shinjin.
22) [week 33] shem
shem 1 Kent: shiroi no choochoo mita yo.
I saw a white butterfly!
shem 2 Miranda: watashi mo (kyoo no asa?) mita wa.
I saw one too, (this morning?).
23) Miranda: hoka ni wasureteru mono ga nai kana. [week 42]
Is there anything else I'm forgetting, I wonder.
24) Miranda: oishii kamo shirenai. un, wakaranai na. [week 42]
It's probably yummy. Hmmm, I don't know…
25) Kent: chanto renrakuchoo yonde ne. [week 43]
Be sure to read my message book, OK?
26) Eric: kore moo ikkai yomoo kana. [week 48]
I wonder if I should read this one again.
27) Kent: mama, moo ikkai pachinko yaritai na. [week 48]
Mom, I want to play pachinko again.

After they had begun using interactional particles in Japanese, the children then extended their use of particles to English sentences. All three children were first observed doing this during the same week, week 29. Following the order in which they began using the particles in Japanese, yo and ne were the first particles that were used with English.

28) Eric: [explaining the word ochinchin] That means 'penis' yo. [week 29]
29) Miranda: Kent, wait for me, ne? [week 29]
30) Kent: Let's see if it works, ne? It will, I hope. [week 29]

Miranda in particular made strategic use of the emphatic yo on the ends of English sentences:

31) [week 33] shem
shem 1 Mother: [talking about a computer game] I'll leave the volume screen up so you can turn it up (if we play any more today.)
shem 2 Miranda: Which we will.
shem 3 Mother: We'll see.
shem 4 Miranda: We will yo.
32) Miranda: [to Kent, whom she thought was eating his ice cream too slowly]
If you don't do bigger scoops, it will turn into soup, yo. [week 35]

Unlike what has been observed for adult foreign language learners, then, Miranda, Kent, and Eric acquired a number of common interactional particles relatively quickly and used them appropriately. As with their overall acquisition of Japanese, routines played a large role in their early production of particles. All three children began to produce the particles in a similar order, and in an order similar to what is observed for Japanese L1 acquisition (Clancy, 1985), raising the question of whether there is a natural order in which these particles tend to be acquired.

Direct versus distal predicates

Like the interactional particles discussed in the previous section, the distinction between direct-style predicates and predicates marked with the distal forms desu and —masu was something the children acquired relatively quickly. It was clear from an early point in their stay that they were aware of both styles, and by the end of the year, their use of these forms seemed natural and appropriate. Example 10 above, recorded during the eighth week in Japan, shows that Miranda, at least, already knew that there were two forms that were in some sense synonymous in meaning.

As is typical of Japanese children, Miranda, Kent, and Eric used primarily direct-style speech, with occasional shifts into distal style. Initially, Miranda's shifting between the two styles seemed quite random. By week 16, however, some of her shifts seemed to be pragmatically motivated, and seemingly random or unmotivated shifts were not observed in her spoken and written Japanese after week 29. Unmotivated shifts to distal forms were not noted in the data from either Kent or Eric, whose linguistic development in Japanese was slower, and who relied on routines and set phrases for longer than did Miranda. This reliance on set phrases may have helped them avoid the inappropriate use of distal-style utterances, as it is possible that they had already acquired an understanding of at least some of the reasons for style shifting by the time they began to use more original sentences.

Various motivations for the use of distal forms can be observed in the data. Of course, some routine daily greetings and school routines were always in distal style, such as saying itadakimasu before a meal. Another function of shifting to the distal style was to display a public persona, as has been noted by Cook (1996a, 1996b, 1997, 1998) such as when the children took a pretend announcer's role in a show they were putting on for their parents, told stories, or played various roles in pretend games.

33) Miranda: kyoo no shoo wa, ookina kabu desu. [week 22]
Today's show is, "The Big Turnip."
34) [week 43] shem
shem 1 Kent: momotaroo no umareta kibidango.
The millet dumpling from which Peach Boy was born. [?]
shem 2 Miranda: momo kara umareta kibidango.
The millet dumpling born from a peach.
shem 3 Eric: momo kara kibidango ga umaremashita.
The millet dumpling was born from a peach.

In example 34, the children retell part of a well-known folk tale in a joking way. In the real story, the infant Momotaroo ("Peach Boy"), who is born when he emerges from a large peach, turns out to be fond of millet dumplings. Kent's meaning in line 1 is not entirely clear, but he seems to be suggesting that Momotaroo was born from a millet dumpling instead of from a peach. Miranda then switches things around in line 2 by suggesting that the dumpling was born from the peach. Finally, in line 3 Eric picks up on Miranda's idea and produces a complete sentence in a distal storytelling style.

Distal forms were also used occasionally for making requests and perhaps more generally to show politeness. For example, Kent's teacher reported approvingly during week 41 that she had noticed that he would shift to distal style when speaking to her. In the following example, Miranda first makes a request in direct style, then amends her request to make it distal. Interestingly, though, instead of repeating the verb to produce the correct form kuremasen ka, she simply adds —masen ka after the initial ending of the sentence.

35) Miranda: [to mother] mizu to koori, totte kuren? —masen ka? [week 49]
shemshemshemshemshemshemshemshemshemshemWon't you bring me some water and ice?

Finally, by week 48 it was not uncommon for the children to shift to distal style in order to express indignation and/or a sense of distance when they were annoyed at each other. This behavior was noted especially between Kent and Eric. The child who was the recipient of such a shift seemed to find it quite upsetting and would often explicitly object to the sibling's use of distal style.

36) Eric: [to Kent, indignantly, on being informed of something]
wakatta yo. wakatte imasu. [week 48]
I knew that. I KNOW that.
37) [week 50] shem
shem 1 Eric: [gloating intonation] okaasan minakatta yo.
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemMom didn't see it.
shem 2 Kent: [deliberately, stressing each word] soo, desu, ka.
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemIs, that, so?
shem 3 Eric: [sounding distressed] Don't say "soo desu ka."
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshem Is that so?

Shifting between direct and distal forms for various social and pragmatic purposes tends to be difficult for JFL learners, at least for adults, who tend to learn distal style first then have trouble using direct style with friends their age. Once they do adapt to using the direct style, however, it is not uncommon for them to find it difficult to shift back to distal when speaking to their teachers. A study by Rounds, Falsgraf, and Seya (1997) based on the analysis of language use in role-play situations shows that children learning Japanese in an immersion elementary school setting are sensitive to the direct/distal distinction, and the authors are relatively optimistic about immersion school learners' prospects for learning to use these styles appropriately based on the input they receive from teachers' alternation between the two styles. However, not surprisingly, immersion school learners' acquisition of this distinction falls far short of what Miranda, Kent, and Eric acquired during a year as second language learners in Japan.

Set aisatsu (greetings)and classroom routines that entail switching to distal style (such as the school lunch routines mentioned above, or the routines that punctuate the school day as a class formally switches from one class or activity to the next) are ubiquitous in Japanese schools, and may help to make the direct/distal distinction particularly salient to JSL learners. In addition, Japanese teachers sometimes insist that children use distal style to them in certain contexts, providing more explicit input on appropriate language use.

Regional dialect

The regional dialect of the area in which the children lived has been undergoing gradual standardization, but it still differs from the dialect that is considered to be standard Japanese. The children's use of various regional dialect forms was first attested relatively late in the year. However, they acquired the local dialect well and with a sense of the contexts in which it was appropriate. The fact that dialect forms are attested only late in the year may be due to the fact that the data for this paper come from the children's speech at home, or when not at home, at least in the presence of their mother. Both parents had studied standard Japanese, and though their mother began using some regional dialect forms over the course of the year, standard Japanese was what the children generally heard at home and was always more frequent in their own home speech. Before any regional dialect forms were noted by their mother or observed in the videotapes, however, neighbors had mentioned that the children were using the regional dialect.

Because of the relative lack of dialect use at home, it is impossible to be sure when and in what order the children began to produce various local dialect forms. However, Figure 2 shows the first use attested in the data by each child for several different forms.

This image shows Figure2, as described in the text
Figure 2. First attested production of regional dialect forms

"Neg." refers to the negative forms common in many dialects in western Japan. For example, the negative forms of the verbs suru (do), taberu (eat), and iku (go) are sen, taben, and ikan (shinai, tabenai, and ikanai in standard Japanese). In example 38 Kent uses two of these forms. He was trying to dish up a portion of the last of the soomen noodles that the family was having for dinner. However, they were sticking together and he could not get them separated.

38) [week 42] shem
shem 1 Kent: aa, moo, dekin! motto soomen iru?
Ah, I can't do it! Do you need more noodles?
shem 2 Mother: /Miranda/ moo iranai?
Do you not need any more, Miranda?
shem 3 Kent: /iran?/
You don't need any?
shem 4 Kent: soomen.
shem 5 Miranda: uun wakannai.
Um, I dunno.

Frustrated by his inability to pick up the noodles, Kent seems to be considering taking all of them, but he first asks his sister if she needs any more. He uses the negative verb forms of the local dialect in both line 1 (dekin 'can't do') and line 3 (iran 'don't need').

As can be seen in Figure 2, the data do not include any examples of Eric using ken or Kent using choku. (They may well have been using these forms in other contexts, of course.) Examples 38 through 42 show the children's use of most of the dialect forms shown in Figure 2. Ken is similar to standard Japanese kara (because); choo is a present progressive form, like standard Japanese —te iru; and choku corresponds to —te oku, an auxiliary verb form that refers to doing something for future benefit. Gaa is an interactional particle that typically comes after the predicate of a sentence. Depending on the context, it may be similar to standard Japanese yo, ne, or yo ne.

39) Miranda: tanoshikatta gaa! [week 42]
It was fun, wasn't it!
40) [week 43] shem
shem 1 Eric: risu ga haitchoo.
There's a squirrel in it.
shem 2 Kent: risu ga haitchoo ken.
Because there's a squirrel in it.
41) [week 48] shem
shem 1 Eric: kore totchatta iken gaa. [error for 'kore totchatte iken gaa' ?]
We can't take this, huh!
shem 2 Kent: un. datte, Miranda no da ken nee.
Yeah. Because it's Miranda's, isn't it.
42) Kent: [complaining to Eric] bikkuri shita gaa! [week 50]
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshem You startled me! (literally: I was surprised!)

The relative lack of dialect forms in the home-based data, and the fact that neighbors commented on their use of dialect in other contexts, suggests that from the beginning the children were distinguishing between standard and regional dialect forms and using them in different contexts. This is further substantiated by their language use when they returned to the United States after one year in Japan. Very little time passed before the children had almost ceased to use the dialect forms mentioned above, except for occasional use of the negative verb forms. (These forms occurred in some of the Japanese videos the children watched, which may have reinforced their use.) At that point, the children were speaking Japanese only with their mother, with a Japanese babysitter who came once a week, and with teachers and other students at weekly Japanese classes for Japanese heritage learners that they attended. None of these people spoke a dialect similar to the one that the children had acquired. Two months after their return to the U.S., however, the family had visitors whom the children had known in Japan and who were speakers of another western Japanese dialect. During their visit the children suddenly began using the regional dialect forms again.

Gendered forms

Japanese has often been said to have fairly clearly distinguishable "women's language" and "men's language". As sociolinguists working in this area have pointed out recently, the pervasive perception of certain linguistic forms being used by speakers of one gender only does not reflect the reality of speakers' language use.8 Nonetheless, ideologies concerning gender and language are pervasive and influential and children begin to use gendered language from an early age (Nakamura, 2001; Sakata, 1991). For children, one of the most salient differences is the fact that young boys are typically expected to use the first-person pronoun boku to refer to themselves, while young girls are expected to use watashi. Adults also sometimes use boku to address or refer to boys, or, less commonly, use watashi to refer to girls. Another gender-based distinction that is quite salient to children is that young girls are typically addressed by first name plus the suffix —chan, while boys are most commonly called first name plus —kun (though there are also boys who are called a nickname plus —chan). Other sources of information about societal beliefs about gendered speech include various media aimed at children, such as books (including textbooks) and television shows, which often portray characters speaking in very stereotypically feminine or masculine ways.

Not surprisingly, Miranda, Kent, and Eric quickly became aware of gendered language and the expectations surrounding it. At one point, Kent extended his newfound understanding of gendered language to English:

43) [week 15] shem
shem 1 Kent: What do you call pink in boy-ish?
shem shem
[parents exchange confused looks]
shem 2 Miranda: He means is there another name for pink that sounds more boy-ish-y.

[turning to Kent] Pink is boy-ish-y and girl-ish-y.

Knowing that pink was associated with girls, Kent evidently assumed that the color term was feminine language. Given that, he wondered what word he should use for the color.

In example 44, Miranda was giving Kent a language lesson, trying to explain to him how to say "I'm not afraid of anything." (Unfortunately, her explanation was incorrect.) As he repeated her sentence, Kent automatically substituted boku for watashi.

44) [week 24]
shem 1 Miranda: watashi wa zenbu kowakunai.
Everything is not scary for me.
shem 2 Kent: Yeah. boku wa zenbu kowakunai.
Everything is not scary for me.

In addition to watashi, which is used by adult men in polite contexts, girls and women also use atashi, which is thought of as quite feminine. Miranda began to use this form by the end of the year. Reminiscing a couple of years later, she said she had originally thought it unfair that, depending on the context, male speakers could use either boku or watashi, while she, as a girl, was expected not to use boku. Because of that, she had been happy to find out that there was a pronoun, atashi, that was thought to be only for female speakers, and she enjoyed using it.

The children also used other forms associated with a particular gender. By the end of the year, Kent and Eric used forms considered to be masculine, such as sugee (sugoi 'amazing') and negative verbs such as shiranee (shiranai 'I don't know'), relatively frequently. They did not use any forms considered to be feminine, except perhaps when mimicking a female character. As for Miranda, around week 32 she began occasionally using the particle wa with rising intonation on the ends of her sentences, a usage associated with femininity. On the other hand, she also experimented with the masculine interactional particle zo between weeks 37 and 41.9 However, zo is not attested in the data after that point. Miranda also occasionally used other "masculine" forms, such as sugee instead of sugoi (amazing) and kimochee instead of kimochii (slang for kimochi ga ii 'comfortable, pleasant'). However, her use of these forms was rare compared to that of her brothers.

Donatory verbs

The Japanese language has a number of different verbs corresponding to English "give" and a number of others corresponding to "receive". The use of these verbs is complex, necessitating a consideration of various pragmatic factors, such as the identities of the giver and recipient, their relationship to each other and to the speaker, and whether honorific verbs are appropriate in a particular speech context. In addition to being complex, these verbs are also ubiquitous, being used not only as main verbs, but also as auxiliary verbs to indicate that the action expressed by the main verb is being done for someone's benefit.

Japanese children are known to have initial difficulties using donatory verbs appropriately, and teachers of Japanese as a second language know that they are one of the more difficult aspects of the language for second language learners as well. Not surprisingly, despite the relative ease with which they seemed to acquire some pragmatic aspects of Japanese, Miranda, Kent, and Eric showed signs of having initial difficulties with these verbs. Eric, in particular, produced only a few relatively basic forms.

The children acquired these verbs in a similar order and manner. As was true overall for their language development, donatory verbs first occurred in set phrases and were only later extended to more productive usage. As noted previously, long before coming to Japan the children were used to saying itadakimasu, a humble form of a verb meaning "receive", before beginning to eat. The next donatory verb they were observed producing was kudasai. This verb was used first in the set pattern X kudasai (please give me X), a pattern that was well established for all three children by week 15.

The next donatory verbs that the children were observed to be using were morau (receive from an out-group person) and ageru (give to an out-group person). These both appear in the data between weeks 22 and 24 for all three children, except that there are no attested instances of Eric using ageru. Early utterances that included these verbs contained some errors:

45) Miranda: moraita. [week 24]
[intended form was moratta 'I received it.']

Example 46 begins as the children's mother teases Eric, telling him she won't share her food with him. The children then began teasing her in the same way.

46) [week 22]
shem 1 Mother: [teasing voice] kore mama no. Erikku-kun ni wa age-, agenai.
This is mine. I'm not going to gi-, give you any.
shem 2 Eric: /Okay./1
shem 3 Miranda: [teasing voice] /kore/1 Miranda no. mama ni agenai.
This is mine. I won't give it to you, Mama.
shem 4 Eric: kore mo /Erikku-kun…/2
This one too, I…
shem 5 Miranda: /kore mo agete/2, chotto.
Give this one too, a bit.
shem 6 Mother: dore?
Which one?
shem 7 Kent: kore wa Kento-kun no.
This is mine.
shem 8 Mother: dore? suupu?
Which? The soup?
shem 9 Kent: mama ageru. kore wa Kento-kun no. mama ageru.
I'll give this to you, Mama. This is mine. I'll give it to you.
shem 10 Mother: ageru?
shem 11 Kent: [loudly, teasing voice] =agena:i.
I won't give it to you!

In line 3, Miranda uses the verb ageru correctly and includes the particle ni to indicate that her mother is the indirect object of the sentence. In line 5, her intended meaning is unclear, so it is hard to tell whether or not she chose the correct verb. She may be intending to tell someone to give her something, in which case, ageru is inappropriate, or she may be telling someone to give something to another person, in which case ageru is correct. Kent also uses ageru in line 9. He seems to be referring to his giving something to his mother, so the verb choice is appropriate, but he omits the particle ni (mama ni ageru) that would be expected if his mother is the indirect object of the sentence. His mother's question in line 10 may be an error, with kureru? (you'll give it to me?) intended, or, given the missing particle, she may be confused and questioning what Kent said, wondering whether he means that she will give something to someone.

Not only did the children make errors in their earlier utterances containing these verbs, occasional mistakes with verb choice persisted for some time. In the following mixed English and Japanese utterance, Kent uses the wrong verb as he begins to list off various origami items that he taught his preschool director to make:

47) Kent: Mommy, today I oshiete moratta-d the enchoo sensei how to make…[week 39]
shemshemshemshemshemshemshemshemshemshemshemshemshemshemshemshemteach-received preschool director
[oshiete ageta 'teach-gave' would imply that Kent was the one who was teaching the preschool director]

Similarly, Miranda produced the following utterance over four and a half months after she was first heard using the verb ageru. (It was relatively rare for her to make a mistake by that point, however.)

48) Miranda: watashi ni ageta. [week 43]
[intended meaning was watashi ni kureta 's/he gave it to me'.]

As far as can be seen from the available data, Eric did not begin to use any other donatory verbs during the year in Japan other than using choodai (which can be used with the verb suru to mean 'receive') to request items a couple of times near the end of the year. Kent and Miranda, however, were next noted using kureru (give to the speaker or another in-group person). Children acquiring Japanese as a first language are said to produce ageru first, and to begin to use morau and kureru only at a later stage, with the order of acquisition of morau and kureru varying (Clancy, 1985). If this is indeed consistent for most first-language learners of Japanese, it is interesting that the order in which these verbs began to be produced is different for these three children learning Japanese as a second language.

Kent was first heard trying to use kureru at week 33. As with the children's early use of morau and ageru, his early utterances with kureru contained some mistakes.

49) Kent: tasukete kureru. [week 33]
[intended utterance was tasukete ageru 'I'll help you.'
50) Kent: kuretai! [week 33]
You want to give it to me! (?)

While the English translation of example 50 may sound marginally acceptable, Kent's utterance is odd in Japanese. He was asking his mother to give him some rice crackers, and his intended meaning seems to have been something like "I want to receive some" or "give me some".

There are no examples in the data of Miranda using kureru before a dinner conversation from week 42. However, at that point she was using this verb frequently, comfortably and accurately, so it is likely that she had been using it for some time before that.

After initially using donatory verbs in set phrases and as main verbs, the children then began to use them as auxiliary verbs. Eric used only kudasai in this way, however, in the formulaic request form of a verb in the —te form plus kudasai, while Miranda and Kent used other verbs as well, such as ageru, morau, and kureru.10

51) Miranda: doite kudasai. [week 15]
Please get out of my way.
52) Eric: chotto matte kudasai. [week 33]
Please wait a second.
53) Miranda: enchoo sensei ga hambun ni otte kureta. [week 42]
The preschool director folded it in half for me.
54) Miranda: matte, Erikku-kun, matte. Hiroko-san ni yarasete morattara? [week 51]
Wait, Eric, wait. How about if you have Hiroko let you do it?
55) Kent: Sakai-san tachi, oshiete ageyoo ka? [week 51]
Sakai and Hiroko, shall I teach it to you?

As can be seen in examples 52 through 54, despite initial errors when they first began to produce these verbs, and more occasional errors that continued to be noted, by the end of the year, the two older children were using a number of donatory verbs with considerable facility.

Person reference and address forms

The children's usage of most of the aspects of Japanese discussed so far quickly became similar to that of other children their age, and even donatory verbs were eventually acquired relatively well by Miranda and Kent within their initial year in Japan. In one respect, however, their production was markedly different from that of Japanese children their age who were acquiring Japanese as a first language. That area of difference was their choice of person reference and address terms.

At the boys' preschool, all of the children were called by their given names (with a suffix such as —chan or —kun following the name) or, in some contexts, by their full names (both surname and given name, again followed by —chan or —kun). At the elementary school, children were more typically called by surname or full name plus —san, though the use of first names was also heard at times. In a meeting with her teacher before starting school, however, when asked whether she would rather be called by her given name or her surname, Miranda said firmly that she would prefer to be called by her given name. Not only did she not want to be called by her surname, in her early weeks at school she even expressed a dislike for having —san attached to her name and commented at home that she wished people would simply call her Miranda.

In referring to themselves, the children primarily used the first-person pronouns discussed above in the discussion of gendered language, with Miranda using either watashi or atashi, and the boys using boku. Early in the year, Miranda also used her own name with —san on the end (as in example 14 above). She did not continue that usage for long however, and she was not observed using either her first name alone, or her first name with —chan attached to the end, as Japanese children sometimes do. Kent and Eric, however, did at times use their own first names, or first name plus —kun (as in example 46 above), to refer to themselves.

Unlike Japanese children their age, during their initial year in Japan, the children were never observed to use oneechan (older sister) or oniichan (older brother) to address their older siblings when speaking Japanese. Instead, they typically used first name plus —chan for Miranda and first name plus —kun for Kent and Eric (with the first names spoken with Japanese pronunciation). Occasionally they would refer to both older and younger siblings by first name only (using Japanese pronunciation, but no —chan or —kun suffix). When speaking to others, whether to their mother or to Japanese friends and acquaintances, they also tended to refer to each other by first name and ——kun or —chan, reserving the use of kin terms such as oneechan (older sister), oniichan (older brother),and otooto (younger brother) for situations in which the person they were speaking to did not know the sibling's name.

In addressing and referring to their parents, the children's usage shifted over the course of the year. Originally, they used papa and mama in Japanese, which was how their mother referred to herself and their father. However, those terms were not common in the area in which they lived, and they soon began to alternate between papa and otoosan (father) and mama and okaasan (mother). At the end of one year, they still addressed their mother as mama at times, but when she later shifted and began calling herself okaasan, the children completely stopped using mama. In addition to okaasan, on one occasion Kent also showed evidence of shifting address forms for pragmatic purposes, using the honorific variant okaasama for a second request when he was particularly concerned that his mother read his renrakuchoo, the notebook in which parents and preschool teachers exchanged daily messages.

56) Kent: [getting into the car after preschool; week 43]
Okaasan, kyoo wa zettai renrakuchoo yomanai to.
Mom, today you absolutely have to read my message book.

datte, karee udon no reshipi ga aru.
Because it has the recipe for the curry udon noodles in it.

shem Kent: [5 minutes later, on arriving home]
Okaasama, eeto ne, chanto renrakuchoo yonde ne.
Mother, um, be sure to read my message book, OK?

Other areas where the children's usage differed from that of Japanese children their age was in failing to use the suffixes —chan, —kun, and —san when it would have been appropriate. For example, when referring to friends' siblings, they at times used a kin term such as imooto (little sister) or otooto (little brother) without attaching a suffix such as —san. Another example of atypical usage was the children's use of the second-person pronoun anata. For example, they might call a friend's mother anata, although a Japanese child would typically either address the other mother as obasan (aunt) or employ the friend's first name, addressing the mother as, for example, Narumi-chan no okaasan (Narumi's mom). The overuse of anata is not uncommon among second language learners of Japanese who are native speakers of English, in which the second person pronoun "you" is used frequently, so transfer from their first language may have played a part in this usage. Miranda also explicitly expressed distaste for calling someone who was not related to her obasan, saying that she thought that was a strange thing to do. Atypical use of anata was also noted within the family as well; Kent called his mother anata at times, and Eric was observed calling his older brother anata, both very atypical address practices.

Explicit correction sometimes helped the children come closer to Japanese pragmatic norms for using person reference and address terms, but conscious knowledge did not always lead to appropriate usage. The following example, recorded during a summer visit to Japan after the family's initial year of residence, shows that Miranda was aware that one should not use the suffix —kun on a sibling's name when talking to people outside one's own family. Her mother was looking over a note that Miranda had written to a friend. In the note, she had refered to her brothers as Kento-kun and Erikku-kun.

57) 1 Mother: You know, about your own brothers you probably wouldn't write,
shem 2 Miranda: kun?
shem 3 Mother: kun.
shem 4 Miranda: Just like you wouldn't say it.

Similarly, after the family's move back to the U.S., Kent at one point began occasionally calling his mother anata. This continued for several weeks, or even months, despite the fact that his mother invariably objected when he did so. The following exchange occurred after Kent's mother had corrected his use of anata for her a number of times.

58) 1 Kent: nani shiteru no, anata no inku jetto purintaa.
What's it doing, your inkjet printer?
shem 2 Mother: What did you just say?
shem 3 Kent: [hurrying out of the room] Nothing!
shem 4 Mother: [with mock fierceness] Did you just call me anata?!
shem 5 Kent: [from down the hall, sheepishly] Kind of…

While some of the children's atypical uses of person reference and address forms may be due to not knowing the appropriate usage or due to transfer from English, in some cases, they quite consciously rejected Japanese pragmatic norms. Although they knew that the use of kin terms for addressing and referring to certain people (such as older siblings and friend's parents) was expected in Japanese, they preferred not to do so. Similarly, they knew that when speaking to non-family members they should refer to each other using kin terms or names without suffixes, but they generally used name plus —chan or -—kun anyway.



Miranda, Kent, and Eric, three children of different ages learning Japanese as a second language, were similar in how their Japanese production progressed, both for overall grammatical development and for the specific pragmatic features of Japanese discussed in this paper. They relied greatly on routines for early language production, and they began to produce forms such as interactional particles and donatory verbs in similar orders. The pace of their language acquisition varied, however, with the oldest child progressing the fastest, followed by the middle child, then the youngest.

Although they were similar to each other, we can see evidence of some differences between their language development and that of both children and adults learning Japanese in a foreign language context. Some aspects of Japanese are typically difficult for classroom learners to acquire, such as interactional particles and the direct/distal distinction. However, these seemed to pose little difficulty for Miranda, Kent, and Eric, who soon began using them in a way that was congruent with the usage of other children their age. Similarly, from the time they began to acquire the local dialect of their area, they seem to have distinguished between it and the version of Japanese that their previous teachers and their mother used and to have used each in different situations.

One reason for the children's success in acquiring interactional particles, the direct/distal distinction, and the regional/standard dialect distinction may be the high amount of input they received for all three of these aspects of Japanese. Interactional particles are extremely frequent in natural spoken Japanese. Occurring with high frequency, coming as they typically do at the end of a phrase, clause, or sentence, and often being pronounced with emphatic intonation, they should be highly salient to learners. As for the distinctions between distal and direct forms, typical children's play, aisatsu, and classroom routines and activities all provide learners with examples of how to shift between the two types of forms. Cook (1996b) observes the constant shifting between distal and direct forms in classroom discourse, with more on-record or public talk being in distal style and other talk being in direct style. Based on my own observations in schools, the same sort of shifting occurs between the use of standard and regional dialects, with teachers using more standard dialect in "presentation" mode and for teacher-centered activities, then shifting to a higher percentage of regional dialect when engaged in one-on-one talk with children or less formal aspects of the class. The routines that punctuate the school day are also typically in distal style and in standard dialect. Thus, learners receive a high amount of input that should help them to distinguish between different styles and dialects, and to learn how to shift between them for various contexts and purposes.

On the other hand, frequency of occurrence and amount of input are clearly not sufficient to insure rapid and accurate acquisition. The various donatory verbs also occur very frequently in Japanese, and yet the children only gradually became able to use them appropriately. Nevertheless, Miranda and Kent both used these verbs comfortably by the end of one year. Eric's ability to use these forms lagged behind theirs, although it is possible that his usage was in line with that of other children his age, since Japanese children too are known to find these forms difficult.

Not only did the children acquire certain pragmatic aspects of Japanese quite easily, they also showed no sign of resistance to certain aspects of Japanese to which adult learners from an English language background often object, such as gendered language and various types of honorific forms (Siegal, 1994, 1996). They never showed any discomfort with the concept that honorifics should be used in certain contexts and/or to address certain people, and they used those forms in a culturally appropriate way. (In the children's case, the use of honorifics was limited to distal forms, a type of addressee honorific, since Japanese children their ages do not typically use referent honorific forms.) Nor did they seem to question the idea that certain linguistic forms might be related to gender or to object to gendered forms, other than Miranda's initial jealousy of the fact that male speakers could use either boku or watashi, which was assuaged when she learned the pronoun atashi.

Although they did not object to gendered language or honorific forms, the children actively resisted some aspects of Japanese pragmatic norms in the area of person reference and address. It is often assumed that second language learners who do not use a language according to the usual pragmatic norms for that language have failed to adequately acquire those norms. However, we need to keep in mind the possibility that learners may be aware of the norms but nevertheless choose not to follow them. What one calls oneself and others is tightly connected to issues of social and cultural identity. The culturally appropriate use of Japanese person reference and address terms necessitates referring to age-based hierarchical differences between siblings and to distinctions between one's own family or group and outsiders. Miranda, Kent, and Eric, culturally American and accustomed to English pragmatic norms, seem not to have been inclined to go along with the emphasis on in-group/out-group distinctions and an age-based hierarchy of siblings that adhering to Japanese pragmatic norms would have entailed.

This study has some implications for second or foreign language teaching. First, it pushes us to consider the fact that different groups of students may have different needs. It also suggests the importance of the role of repetition and routines in language learning, for children, at least, and perhaps for adults as well. As far as the first point is concerned, although we should not assume that learners in one context will all be similar to each other, the differences between the three children's acquisition and that noted previously for foreign language learners does remind us that various groups of language learners — children and adults, and second language learners and foreign language learners — may well have different needs. In deciding how to spend instructional time, it is helpful to have an empirically based knowledge of what is likely to prove easy or difficult for learners of different ages, from different language backgrounds, and learning in a variety of different contexts (in a country where the language is spoken, in an immersion classroom, in a foreign language classroom, etc.). If there are aspects of a language that one group acquires easily, such as the ease with which these children began using interactional particles, then it may not be necessary to spend much (if any) instructional time on them. That time can be better used for more troublesome aspects of the language. On the other hand, for adult foreign language learners, who are known to have trouble mastering the use of interactional particles, teachers might want to be sure that they provide plenty of authentic input that includes the use of interactional particles, and also that they encourage learners to produce the particles in the contexts in which native speakers would do so. In order to target our teaching in this way to the specific points that learners are likely to find difficult, we need to do more detailed research into how various learners acquire a second language.

As for the role of repetition and routines, frequently hearing various set phrases and routines, participating in routines with classmates, and practicing them at home played an important role in the children's acquisition of Japanese. Not only their growing grammatical ability in Japanese, but also many aspects of their pragmatic competence, can be seen to have their roots in routines. The fact that routines recur over and over again, generally in the same or similar contexts, doubtless helped the children make sense of their meanings. It also gave them plenty of opportunities to practice using the formulaic language that comprises them. At a very early stage of their stay in Japan, the children delighted in daily classroom routines, such as the set phrases that the nitchoku-san (the class leaders for that particular day) would use to announce the beginning of a new class period at school and those used by their classes before beginning to eat their school lunch. They loved to rehearse these routines at home and to perform them for their parents and siblings, and they began to produce Japanese by first relying on those routines, then by altering them, and then by producing more creative utterances.

Routines of this sort are built into everyday life in Japan, so second language learners are virtually assured of encountering them. The ease with which the children acquired interactional particles, the direct/distal distinction, and the regional/standard dialect distinction suggests the possibility that students learning Japanese in a foreign language classroom context also might benefit from utilizing some of the same sorts of routines and repetition in the classroom.

Finally, this study reinforces what previous scholars have suggested about the need to avoid essentializing the target language and culture and also to respect the fact that learners may not always want to emulate native speakers of the language (Kramsch, 1993; Kubota, 2003; Matsumoto & Okamoto, 2003; Siegal & Okamoto, 2003; Tai, 2003). Rather than assuming that deviation from target language pragmatic norms represents mistakes or pragmatic failure, we should consider the possibility that learners may be aware of the norms but nonetheless choose not to follow them. Although teachers may want to make learners aware of the ways their utterances might be interpreted, we should also be aware of the danger of attempting to force learners to meet a stereotyped ideal of how the target language should be used. Just as native speakers sometimes flout norms, so too should learners be able to do so.



1. In acknowledging all the assistance I have benefited from in writing this paper, I would like first and foremost to thank Miranda, Kent, and Eric for being willing to have their language use analyzed. I am also particularly indebted to Dina Yoshimi for her initial invitation to attend the Pragmatics in the Chinese, Japanese, and Korean Foreign Language Classroom symposium, which resulted in my research taking an unexpected turn. I will always be grateful for her enthusiastic reception of that research and her unflagging encouragement and thoughtful comments. I also thank an anonymous reviewer and the symposium participants for their helpful comments on my original paper. Any remaining weaknesses are solely my own responsibility.

2. These names, and all other proper names in the paper, are pseudonyms.

3. These lessons were with two different teachers, both Japanese women in their 20s. The first teacher taught the children for several weeks before moving away from the area. After that, the lessons were resumed by another teacher. She taught the children for approximately one year, until they moved to Japan. The youngest child, Eric, was only 1 when the lessons began and was not present during most of the lessons with the first teacher. However, the second teacher came to the children's house, and over the course of the year, he too began to participate in the lessons.

4. I would like to thank Naoko Witzel, Momoko Ushiki, Margaret Camp, Soumantha Colwell-Chanthaphonh, and Yuko Watanabe for their work transcribing tapes. These conversations were taken from a larger collection of videotapes that are still being transcribed. The larger collection contains more tapes of family dinners and of the children's interactions with visiting Japanese children and adults. The journal observations include notes on the children's language usage both with family members at home, and also outside the home and/or with non-family members. Because the bulk of the data for the current study are from talk within the family, it cannot be assumed to be representative of the children's speech in other contexts, such as at school or playing with other children in the neighborhood.

5. A great deal of research has been done on the role that interactional routines can play in language acquisition. For Japanese in particular, see Kanagy, 1999; Kondo-Brown, 2002, 2004; Ohta, 1999; and Yoshimi, 1999. A major question remaining is what conditions might encourage learners to use the formulaic language found in routines as a steppingstone to creative language use. In JFL settings, particularly with child learners, it seems common for learners not to move beyond the stage of producing formulaic utterances.

6. A few of their routines dated from before the move to Japan, as the children had long been accustomed to saying itadakimasu before eating and gochisoosama deshita after eating. While still in the United States, they also occasionally used tadaima to announce their return home, though that usage was not consistent. Hide-and-seek routines were learned within the first few days of the move, as that was one of the earliest games that other children in their neighborhood taught them.

7. In examples 1 through 7, and also in example 10, English words that are not explicit in the Japanese have been put in parentheses to show the relative simplicity of these early Japanese utterances. For a list of other transcription conventions used, please see the appendix.

8. See Okamoto and Shibamoto Smith (2004) for an excellent collection of recent work in this area.

9. Although it may sound masculine in standard Japanese, the particle na, which both Miranda and Kent were observed to use at times, is not associated with a particular gender in the area in which the children lived.

10. The data for kureru are unclear on this point. Since kureru seems to have been acquired later than morau and ageru, and at a point when the children had already recognized that auxiliary usage of morau and ageru was possible, they may have used kureru as an auxiliary as well as a main verb from the beginning.



Clancy, P. M. (1985). The acquisition of Japanese. In D. I. Slobin (Ed.), The crosslinguistic study of language acquisition. Volume 1: The data (pp. 373-524).Hillsdale, NJ: Lawrence Erlbaum.

Cook, H. M. (1998). Situational meanings of Japanese social deixis: The mixed use of the masu and plain forms. Journal of Linguistic Anthropology, 8(1), 87-110.

Cook, H. M. (1997). The role of the Japanese masu form in caregiver-child conversation. Journal of Pragmatics, 28, 695-718.

Cook, H. M. (1996a). Japanese language socialization: Indexing the mode of self. Discourse Processes, 22, 171-197.

Cook, H. M. (1996b). The use of addressee honorifics in Japanese elementary school classrooms. In N. Akatsuka, S. Iwasaki, & S. Strauss (Eds.), Japanese/Korean Linguistics Vol. 5 (pp. 67-81). Stanford: CSLI Publications.

Kanagy, R. (1999). Interactional routines as mechanisms for L2 acquisition and socialization in an immersion context. Journal of Pragmatics, 31, 1467-1492.

Kondo-Brown, K. (2004). Investigating interviewer-candidate interactions during oral interviews for child L2 learners. Foreign Language Annals, 37(4), 602-615.

Kondo-Brown, K. (2002). A longitudinal evaluation study on child JFL learners' oral performances. Japanese Language and Literature, 36(2), 171-199.

Kramsch, C. J. (1993). Context and culture in language teaching. Oxford: Oxford University Press.

Kubota, R. (2003). Critical teaching of Japanese culture. Japanese Language and Literature, 37(1), 67-87. Special issue: Sociocultural issues in teaching Japanese: Critical Approaches.

Matsumoto, Y., & Okamoto, S. (2003). The construction of the Japanese language and culture in the teaching of Japanese as a foreign language. Japanese Language and Literature, 37(1), 27-48. Special issue: Sociocultural issues in teaching Japanese: Critical Approaches.

Nakamura, K. (2001). Gender and language in Japanese preschool children. Research on Language and Social Interaction, 34(1), 15-43.

Ohta, A. S. (1999). Interactional routines and the socialization of interactional style in adult learners of Japanese. Journal of Pragmatics, 31,1493-1512.

Ohta, A. S. (1994). Socializing the expression of affect: An overview of affective particle use in the Japanese as a foreign language classroom. Issues in Applied Linguistics, 5(2), 303-325.

Ohta, A. S. (1993). Activity, affect and stance: Sentential particles in the discourse of the Japanese as a foreign language classroom. UCLA Ph.D. dissertation.

Okamoto, S., & Shibamoto Smith, J. S. (2004). Japanese language, gender, and ideology: Cultural models and real people. New York: Oxford University Press.

Rounds, P., Falsgraf, C., & Seya, R. (1997). Acquisition of sociolinguistic competence in a Japanese immersion school. Journal of the Association of Teachers of Japanese, 31, 25-51.

Sakata, M. (1991). The acquisition of Japanese "gender" particles. Language and Communication, 11(3),117-125.

Sawyer, M. (1992). The development of pragmatics in Japanese as a second language: The sentence-final particle ne. In G. Kasper (Ed.), Pragmatics of Japanese as a native and target language (pp. 83-125) Honolulu: Second Language Teaching and Curriculum Center, University of Hawai‘i at Manoa

Siegal, M. (1996). The role of learner subjectivity in second language sociolinguistic competency: Western women learning Japanese. Applied Linguistics, 17(3), 356-382.

Siegal, M. (1994). Second-language learning, identity, and resistance: White women studying Japanese in Japan. In M. Bucholtz, A. C. Liang, L. A. Sutton, & C. Hines (Eds.), Cultural performances: Proceedings of the third Berkeley Women and Language Conference (pp. 642-650). Berkeley, CA: Berkeley Women and Language Group, University of California: Berkeley.

Siegal, M., & Okamoto, S. (2003). Toward reconceptualizing the teaching and learning of gendered speech styles in Japanese as a foreign language. Japanese Language and Literature, 37(1), 49-66. Special issue: Sociocultural issues in teaching Japanese: Critical Approaches.

Tai, E. (2003). Rethinking culture, national culture, and Japanese culture. Japanese Language and Literature, 37(1), 1-26. Special issue: Sociocultural issues in teaching Japanese: Critical Approaches.

Yoshimi, D. R. (1999). L1 language socialization as a variable in the use of ne by L2 learners of Japanese. Journal of Pragmatics, 31, 1513-1525.



Transcription conventions used:
[ ]Brackets enclose information about when the utterance occurred, voice quality, laughter, etc.
,Indicates a brief pause, with non-final intonation.
.Indicates sentence-final intonation.
?Indicates rising intonation at the end of an utterance.
-Indicates that the utterance was broken off in the middle.
=Indicates "latching," with no perceptible gap between two utterances.
/ /Slashes enclose overlapped speech, words that were spoken at the same time as the next/previous speaker. When an example contains multiple overlaps, they may be numbered to indicate which utterances occurred at the same time.
( )Parentheses around a number indicate a pause, with the number in parentheses representing the length of the pause in number of seconds. Parentheses around one or more words indicate that the utterance was hard to hear and the transcription is the transcriber's guess at what was said.

Table of Contents