Superficial intersubjectivity in ELF university dormitory interactions

3.1 Data

The corpus used for this paper comes from ELF interactions collected at an international university dormitory in Japan. At this university, approximately half of the students are non-Japanese, either L1 speakers of English or students with adequate English skills to study at university in English whose Japanese ability is usually very limited at the beginning of the academic year. The other half are domestic Japanese students who all have at least six years of formal English education at the time of entering. Therefore, English was often used as the lingua franca between the Japanese and non-Japanese students, especially during the first few semesters after matriculation.

In the on-campus university dormitory, there were approximately 3,000 residents, mostly first-year students, at the time of data collection. Approximately half of the residents were Japanese and the other half were non-Japanese students from approximately 90 countries around the world. The majority of the students in the dormitory, however, were from the Asia Pacific region, such as China, Korea, India, Thailand, and Vietnam.

Approximately 37 hours’ worth of video conversations were collected between four Japanese students (i.e. core participants) and their conversation partners. The partners of the core participants were 32 different students from 10 countries. The non-Japanese partners represent countries including Botswana, Canada, China, Korea, Indonesia, Romania, Thailand, United States, Uzbekistan, and Vietnam, based on their registration status at the university. However, some participants had experience living in more than one country for an extended period of time or had parents of different nationalities, and thus represent more than one country or culture. All participants’ names used in this paper are pseudonyms.

Video recordings were made by the participants at the dormitory, and the researcher was never present during recordings. The instructions given by the researcher were limited to the positioning of the camera to have all participants visible on the camera screen. The camera was usually set on a desk, and the participants would sit on the bed or the floor while they talked. The conversations were intentionally conducted and recorded, but conversation topics were chosen by the participants and the content was spontaneous. The recordings were transcribed using conversation analysis (CA) conventions (Jefferson 2004). Japanese words have been transcribed in italics with a translation underneath. Transcription conventions are displayed in the appendix.

The analysis was done using a two-stage process, starting with a standard CA (Sacks et al. 1974) followed by a post-analytic researcher observation. CA was selected as the main method of analysis due to its ability to “identify ways in which participants themselves orient to, display, and make sense of one another’s cognitive states (among other things)” (Drew 1995: 79). CA is not able to elucidate the actual cognitive state of individuals, nor understand the intentions of why one used a certain language at a certain point. What CA can do is portray the ways in which the interlocutors are making sense of each other’s utterances and the process of achieving intersubjectivity or socially distributed cognition (Seedhouse and Walsh 2010). Through utterances, participants display their own cognitive states and at the same time display their understanding of the interlocutor’s utterances. In other words, CA takes an emic perspective where the analytical claim is grounded in the participants’ visible orientations (Kasper and Wagner 2014).

The outcomes of the emic analysis of the data were then compared to an etic analysis (i.e. the researcher perspective) of the sequence. This second stage of analysis enabled the researcher to demonstrate possible gaps in participants’ intersubjectivity. The traditional line of CA research adheres to applying an emic perspective to data. However, recent research, such as with feminist studies and cross-cultural cross-linguistic investigations, have explored the possibility of utilizing pre- and post-analyses in seeking the connections of CA findings to theories and political agendas (e.g. Kitzinger 2008), or comparison of sequential structures across different languages (e.g. Fox Hayashi and Jasperson 1996). The analytical method adopted for the current study is similar to these studies in such a way that it utilizes both emic and etic analyses in order to understand the social phenomenon in arguably a holistic way.

3.2 Analysis

Initially, all instances of repair sequences regarding a word were identified from the data and examined. From the data, three cases were identified which were different from other repair sequences regarding a word, and the accuracy of the mutual understanding was questionable during the interaction. This section will analyze these three cases along with one case where a potential superficial intersubjectivity was avoided.

The first example is from an interaction between Hang, from Vietnam, and Ami, from Japan, talking about their cultural foods. After Ami initiates a repair, Hang provides an inaccurate translation in order to self-repair. Ami then accepts this inaccurate translation with no hesitation, co-constructing superficial intersubjectivity.

In line 3, Ami displays trouble understanding the word “vinegar” by repeating it with a rising intonation and adding “tte”, which can be understood as the reduced form of a question tte nani (‘what is it?’) in Japanese. Immediately, Hang follows up on this by repeating the word (line 4), which could be understood as Hang seeing the repair initiation by Ami as a possible hearing problem. However, when there is no immediate uptake from Ami, Hang then provides a translation in Japanese, “mirin” (line 4), which is a sweet sake or a sweet wine used for cooking. After a slight pause, Ami displays acceptance of this candidate understanding and understanding by saying “ah” and repeating “mirin”, which is then followed by laughter (line 5 and 6). This laughter could be interpreted as a display of embarrassment due to the simple answer (Glenn 2003). Hang then joins the laughter and the conversation returns to the main topic.

Hang’s action in line 4 of providing a translation in Ami’s L1 rather than providing an explanation of the object in English suggests Hang’s understanding of Ami’s turn in line 3 as a display of non-knowing of the English word “vinegar”. This translation-providing action by Hang is similar to what Greer (2008) found in his study where a speaker would code-switch to the hearer’s L1 in order to secure their understanding. As a result, the speaker presents their perception of the hearer as a novice user of the language. This suggests Hang’s perception of Ami as a novice and Ami’s English language ability as being possibly limited. At the same time, Ami accepting the translation, along with the embarrassed laughter, could be seen as signs of alignment to the novice-positioning.

The sequential pattern of this repair sequence is a common pattern found in the corpus, in which one participant would display trouble of a single word through the use of a repetition or explicit question, and initiate a repair. In general, the speaker of the trouble source will repair through the use of explanations, translations, or gestures to achieve intersubjectivity (e.g. Bolden 2012; Funayama 2002; Seo and Koshik 2010). From an emic participant-oriented perspective, one can interpret that the two participants have reached a mutual understanding.

However, conducting a post-CA researcher observation, the translation that Hang has provided (mirin) is not an accurate translation of the word vinegar (su in Japanese). Therefore, despite the repair sequence and both participants displaying agreement, the two cannot be said to have reached an “accurate” mutual understanding. This gap identified in the participants’ understanding through a comparison between an emic and etic analysis is an example of what I term, “superficial intersubjectivity.”

The following two examples also demonstrate interactions where the participants do not seem to prioritize accurate mutual understandings. Excerpt (2) is an example from a recording between Maya and Kei. Maya is from Japan and Kei is from Korea. The excerpt begins where Maya is attempting to explain about the Japanese ‘cute culture’ (kawaii culture), a concept that refers to a cute and sweet fashion style and lifestyle that can be associated with females wearing lace and pink-colored clothes (Borggreen 2013). Maya uses an onomatopoeia “kira kira kira”, to describe the aura that surrounds the people of this cute culture, which is then repaired by Kei.

In order to explain the concept of “cute culture”, Maya introduces associated images such as “pink” (line 2). After relatively long pauses of 1 and 1.2 seconds (lines 3, 5), she introduces the idea of “lace”, however with a Japanese pronunciation, “lehsu” (line 6) with extra vowel sounds. Kei then repairs Maya through introducing a North American pronunciation, “lace” (line 7). Maya repeats this with a similar pronunciation (line 8). This sequence is similar to what Brouwer (2004) calls “doing pronunciation,” often found in L1–L2 interactions where the L2 speakers initiates repair with a rising intonation, then the L1 speaker presents their version of the pronunciation, which is then followed by the L2 speaker imitating the pronunciation of the L1 speaker. The similarity of the sequences suggests Kei’s positioning of Maya as the novice user of English, and Maya aligning to this positioning.

Kei does not respond to provide confirmation (line 9) and Maya attempts to continue with her explanation (line 10), but pauses again and moves her hands along the side of her body while twisting her wrist, suggesting an on-going solitary word search (Goodwin and Goodwin 1986) (line 11, Figure 1). She then presents her candidate word “ahula” with a rising intonation while looking at Kei and using a shorter version of the previous gesture (line 14, Figure 2), which is most likely the Japanese pronunciation of the word “aura.” Therefore, the gesture used in lines 11 and 14 can be understood as Maya attempting to describe the atmosphere surrounding one’s body (i.e. aura).

Figure 1:

Line 11.

Figure 2:

Line 14.

When there is no response from Kei (line 15), she changes the first vowel sound of the word and presents another candidate word “ohla” and waits for a response (line 16). However, Kei again does not respond (line 17). Maya then uses a Japanese onomatopoeia “kira kira kira” (‘sparkle sparkle sparkle’) possibly to describe the atmosphere that surrounds the cute girls, again with a similar gesture used in line 11 (line 18, Figure 3). Code-switching is a common practice used by L2 speakers to initiate a repair sequence when they cannot find the word for which they are searching (e.g. Mori and Hasegawa 2009). However, Maya’s utterance is followed by her laughter (lines 18, 20), thus marking the word and displaying Maya’s understanding of “kira kira kira” not being in the desired language for the interaction and will require some interpretive work by Kei (Firth 2009). After repeating what Maya had said, “ahura” (line 19), Kei then repairs Maya by providing a translation in English, “twinkle twinkle” (line 21). Kei’s repair of Maya’s turn in line 18 shows that Kei understands some Japanese but orients towards having the interaction conducted in English, as also observed through Maya’s laughter in lines 18 and 20. At the same time, Kei shifting the focus to repair Maya’s Japanese rather than attempting to resolve the non-understanding of “aura” suggests his interest in possibly progressing the conversation.

Figure 3:

Line 18.

When Maya displays non-understanding or uncertainty through a repetition with a rising intonation (line 22), Kei introduces a different expression of “bling bling” (line 23). Maya displays recognition and acceptance of the candidate understanding through “ah” and repetition of the word “bling”. Kei does not explain what “twinkle” or “bling” mean, suggesting that he understood Maya’s “ah” and repetition as a confirmation and a sign of mutual understanding. Similarly, Maya does not further pursue the meaning of “twinkle twinkle” or “bling bling”, and displays that she has completed her explanation (line 28). Therefore, from a participant-oriented perspective, it can be assumed that the two have reached a mutual understanding of “kira kira kira” as equivalent to “bling bling”.

In this excerpt, Kei repairs Maya and provides possible English translations in an attempt to understand the meaning of Maya’s utterance (lines 21, 23). Maya then displays agreement to the suggested translation (line 24). The participants in the interaction therefore have agreed on the content, and in theory, have a shared understanding. However, switching to an etic perspective, one would notice that the participants might have slightly different understandings of what is being agreed upon. “Kira kira kira” and “bling bling” hold different connotations and are used to describe different attributes of a person. For instance, a girl who is “kira kira kira” could mean someone wearing cute pink clothes and having curled long hair, while “bling bling” could refer to a person wearing sexy clothes and gaudy jewelry. That said, the hand gesture used by Maya in lines 11, 14 and 18, may have been interpreted by Kei as jiggling bangles. In addition, the protracted repair sequence may have led Maya to accept any sign of Kei’s understanding in order to progress the interaction. Nevertheless, this example again shows that even though the interactants indicate that they have reached a mutual understanding as a sequence, it may not be an accurate understanding (i.e. superficial intersubjectivity).

Observing Excerpts (1) and (2), it appears that superficial intersubjectivity arises during repair sequences where there is a difference in the interactional positioning of the speakers in regards to linguistic ability. In Excerpt (1), Hang displays her perception of Ami’s novice position through providing a translation of the trouble source in Ami’s L1. The translation is then accepted by Ami without question, displaying Ami’s alignment to the novice positioning. In Excerpt (2), Kei demonstrates his expert positioning in English and positions Maya as the novice through repairing Maya’s pronunciation and providing English translations for a Japanese word. Similar to Ami, Maya aligns to the positioning through repairing her pronunciation to sound similar to Kei’s, and accepting the translation without question. Furthermore, in both Excerpts (1) and (2), the Japanese participants display their relative novice positions in terms of the English knowledge through requesting for repair to their interlocutors and marking their selected words, or by displaying embarrassment through laughter after the use of Japanese.

Kamio (1994, Kamio 1997) discusses interactants having domains of information (i.e. “territories of information”) that are relatively close or distant to them compared to their interlocutor. He describes the knowing and unknowing positions being on a continuum, and argues that this concept can broaden to understand not only who knows, but also who has the right to know the information and the right to express it (Kamio 1997). In the case of ELF interactions, there is an interactional hierarchy or expert–novice positioning in regards to the knowledge of the English language, especially between L1 and L2 speakers (Hosoda 2006; Kurhila 2006), but also between L2 speakers (Kaur 2011a; Siegel 2015). Based on Excerpts (1) and (2), this also seems to be true in international dormitory interactions during repair sequences regarding a word, and the speaker who is positioned as the relative expert seems to hold influence over the repair process.

However, superficial intersubjectivity seems to occur in slightly different interactional contexts as well. Excerpt (3) is between Tomoko, from Japan, and Maria, from the United States, talking about foods they dislike. Tomoko uses the Japanese word “pikurus”, a loan word from English for the word ‘pickles’, which leads to a possible superficial intersubjectivity when Maria identifies it as a trouble source and requests for repair.

In line 1, Tomoko hesitates and then completes her turn with a candidate word, “pikurus” with a rising intonation. This hesitation could be displaying thinking of the answer to the question rather than searching for a word. However, Tomoko completes her turn with a rising intonation, which also could be understood as a lack of confidence in her word choice and marking it as a potential trouble source (Firth 2009). Maria then explicitly requests repair of the word by asking, “what is pikurus” (line 3). Tomoko does not explain what the object is, but rather, defends her word choice by explaining that it is an English word (lines 5 and 6), despite the fact that Maria is an L1 speaker of English. Here, we can observe that Tomoko, in contrast to Ami or Maya in the previous extracts, seems to position herself as an expert English user.

When the repetition does not solve the non-understanding issue (line 8), Tomoko uses what Goffman (1981) calls a “response cry,” “eh” (Hayashi 2009) to initiate a new turn and displays that she is surprised with Maria’s response of not knowing the word (line 12). Tomoko then moves on to explain the object (being in a hamburger) and uses a gesture to show the small rounded shape and demonstrates putting the pickle on the hamburger patty (Figure 4). Maria uses “oh” to suggest that she has understood what Tomoko is referring to; however, she then presents her candidate understanding of “cow meat” (line 15) and “pickles” (line 14) with a rising intonation, displaying uncertainty. This is similar to Wong’s (2000) finding where participants claim their understanding (e.g. oh, uh-huh), but soon after initiate repair.

Figure 4:

Line 16.

Tomoko does not confirm any of the candidate words suggested by Maria (line 20), indicating that they are not the words that describe the object she is referring to. Tomoko continues to provide more hints such as “the green one” (line 21). Maria again shows that she has hit upon an idea (“oh”) and presents her candidate understanding of “relish” (line 22). This time, the candidate understanding is presented with no rising intonation and Maria continues her turn to display agreement with Tomoko’s dislikes (“yeah”) (line 22). Tomoko repeats Maria with a rising intonation with a slightly different pronunciation, which is then confirmed and repaired by Maria (lines 23 and 24). Unlike Maya in Excerpt (2), Tomoko does not self-repair her pronunciation, but nods and signals the end of the repair sequence, suggesting that she does not position herself as a novice English user in the interaction.

Again, from an emic CA perspective, one can claim that the two participants are using various strategies and affordances, and through cooperative collaboration the two were able to reach a mutual understanding of “pikurus” being “relish”. However, switching to an etic researcher perspective, the accuracy of the content is questionable: while “pickles” usually refers to sliced pickled cucumbers or other vegetables, “relish” typically refers to finely chopped pickles mixed with other herbs and spices. Thus, the two have reached a superficial intersubjectivity.

In this case, however, the cause of the superficial intersubjectivity does not seem to be due to the unequal positioning or access to linguistic knowledge. Rather, it seems to be related to Tomoko’s claim of English knowledge. In contrast to previous ELF studies and previous excerpts, Tomoko does not seem to demonstrate cooperativeness in achieving an accurate intersubjectivity. Tomoko in line 6 claims her accuracy of her word choice by saying “pikurus” is “English”, does not display any uptake of Maria’s candidate understanding of “pickles” in line 19, and then does not self-repair her pronunciation or initiate a repair to confirm the meaning of the word “relish” (line 25). Thus, the sequence can be seen as Tomoko avoiding utilizing the sequence as a possible learning opportunity (Brouwer 2003). Despite her inaccurate knowledge of English in this case, Tomoko appears to maintain her position as an expert of the English language, even when speaking with an L1 speaker of English, and prioritizes progressivity over an accurate intersubjectivity. As a result, she displays lack of interest and attention in gaining the accurate word they are searching for, leading to an instance of superficial intersubjectivity.

In the corpus, however, there was a single case where a potential superficial intersubjectivity was avoided. Excerpt (4) is from the same recording as Excerpt (2) between Maya and Kei. They are talking about what Maya learned in her environment class regarding carp, a type of fish. When Maya displays uncertainty with her pronunciation, Kei repairs Maya. However, the two then continue at length to reach an accurate understanding.

Figure 7:

Line 51.

In line 1, Maya introduces the topic “carp”, but with cut-offs, pauses and hesitation, and a rising intonation displaying a possible ongoing word search and uncertainty of her word choice. Maya provides a word association, “fish” and a gesture in an attempt to gain mutual understanding (line 3, Figure 5). Kei, as his frequent practice, repairs Maya with “cod” (line 4). When Maya does not repeat it in the same manner (line 5), Kei presents his version of the pronunciation again (line 6). Maya does not respond to this second repair immediately and displays thinking through “ah”, silence, and repeats “cod” (lines 7–9, 11). It is unclear here if Maya is thinking of the word “cod” or if she is returning to the main topic and thinking of what to say next. However, Kei then presents the spelling “C-O-D” with a rising intonation to prompt a response and to confirm his understanding (line 10).

Figure 5:

Line 3.

Although Maya does not provide an immediate or definite response to this question (lines 12 and 13), Kei attempts to resolve the non-understanding through describing the meat color (lines 14, 17, 20). However, again, he fails to do so due to Maya describing the color of the outside of the fish (line 15), and then explaining that she has never eaten the fish (lines 26 and 27). Maya then introduces her “Japanese sound” of the word “carp” again (lines 31–33) and then describes the color (line 34).

When these negotiation turns do not work to resolve the non-understanding, Maya then requests her dictionary (line 38). While she is looking up the word, she asks Kei if he knows the Japanese baseball team called “Hiroshima Carp” (lines 40 and 41). This could be interpreted as Maya demonstrating her knowledge and accuracy of the word “carp” using concrete examples. Kei however shakes his head in response, suggesting that he does not know the name of the team (line 42). Maya then plays the audio of the dictionary for “carp” (lines 44 and 45), and then shows the dictionary to Kei (line 46, Figure 6). This pronunciation from the dictionary is then repeated by Kei (line 47) and Maya (line 48). Only when Kei looks at the dictionary closely (lines 49–61, Figure 7) does he display acceptance or understanding through a soft but elongated “nh” (line 55). After Maya notices that Kei has accepted her self-repair and that they have reached a mutual understanding, she continues with her storytelling.

Figure 6:

Line 46.

Although Maya questions the candidate word “cod”, repeats “carp” several times with varying pronunciations, and provides an example of the baseball team name, Kei does not display understanding. After an extended question and answer sequence, Kei finally displays a change of state when he sees the dictionary entry. This is contrastive with Maya’s actions in Excerpt (2) lines 6–8, 23 and 24, where she accepts Kei’s repair immediately, thus suggesting Kei’s continuing perception and positioning of Maya as the novice English user in the interaction. In other words, the dictionary was the last method that Maya could resort to in supporting her word choice and having Kei understand the object she was referring to. This non-understanding by Kei could be due to Maya’s non-standard pronunciation or Kei’s lack of familiarity with fish names. Nevertheless, until Maya demonstrates and supports the accuracy of her candidate word “carp” accompanied with a dictionary, Kei does not display alignment. Thus, similar to Okada’s (2015) and Reichert and Liebscher’s (2012) findings, the dictionary held a key role in the L2 repair process.

Comparing the four examples of repair sequences regarding a single word analyzed in this study, there seems to be a connection between superficial intersubjectivity, preference towards progressing the conversation, and the speakers’ expert–novice positioning in terms of their linguistic knowledge. In Excerpts (1) to (3), the Japanese participants display trouble understanding an English word or producing a “standard” English word that can be understood by their interlocutor. In addition, in Excerpts (1) and (2), the Japanese participants are positioned by their partners as relative novice English speakers through being provided translations in their L1 or having their pronunciation repaired. When they are provided a candidate word by their relative expert interlocutor, they see it as sufficient and progress the interaction. In Excerpt (3), the Japanese participant maintains her expert positioning despite her lack of English knowledge, and displays preference towards progressing the conversation. Moreover, the Japanese participants accept the candidate words presented by their partner without question. In contrast, in Excerpt (4), Maya is able to refute Kei’s candidate word and convince him of the accuracy of her initial word choice. This was possible because Kei provided the opportunity for negotiation by starting another repair sequence, had a concrete example to support her word choice, and had assistance of the dictionary. However, if she did not have any of these, she may have accepted Kei’s candidate understanding of “cod” and progressed the interaction after Excerpt (4) line 8 when she utters “ah”, and thus possibly leading to a superficial intersubjectivity.

This does not necessarily mean Kei is not cooperative. Maya and Kei collaboratively used candidate pronunciations, spelling, gestures, and word associations in co-constructing an accurate intersubjectivity. In addition, Kei’s action of seeking confirmation in Excerpt (4) line 10 started the relatively long repair sequence and allowed the two to avoid superficial intersubjectivity. When Maya was provided the opportunity after Excerpt (4) line 10, she was able to demonstrate the accuracy of her word choice. This is similar to classroom interactions where students are able to raise questions to the teacher and demonstrate their understanding and non-understandings of the content only when provided the opportunity (Wong and Waring 2009). Otherwise, the novice speaker (e.g. students) may lack the opportunity or ability to raise questions, stay silent, and as a result, mis- or non-understandings may stay unnoticed.