Volume 10, No. 1, Art. 49 – January 2009
Intercultural Communication Online: Conversation Analysis and the Investigation of Asynchronous Written Discourse
Abstract: This paper works through the methodological issues involved in treating "culture" and "interculturality" as interactionally demonstrable and observable phenomena in written online asynchronous discourse. In particular, the paper explores the ways that conversation analysis (CA) and its focus on sequentiality and membership categorization analysis might aid the analysis of culture as a textural interactional achievement. The paper argues that, while there are some clear differences between sequential talk and written asynchronous discourse, there are still interesting ways in which CA's analytic foci may be worked through in relation to online discourse. Both the concern with sequentiality and with membership categories may well help us to see how the construction of visible and recognizable intercultural discourse practices are accomplished through written modes in online forums.
Key words: conversation analysis; sequentiality; membership categorization analysis; interculturality
Table of Contents
2. Conversation Analysis and the Analysis of Culture
3. Empirical Studies of Interculturality and Conversation
4. Conversation Analysis and Online Discourse
4.1 Sequentiality and turn taking
4.2 Constitutive ordering of posts
4.3 Membership categorization analysis
4.4 Membership categorization analysis and asynchronous intercultural communication
5. Culture as Context or Culture as Practice
The proliferation of new communication technologies such as Web 2.0, handheld mobile digital communication devices, and wireless connectivity infrastructure, have dramatically expanded the opportunities for transnational and, consequently, cross-cultural and intercultural social praxis. It is hard to think of any aspect of "Westernized" social life that has not been touched by this change. Taking Higher Education as an example, the rapid expansion of Virtual Learning Environments in recent years (BROWNE, 2003) is now accompanied by strong interest in other distributed participation spaces, including Blogs, Wikis, file/video/bookmark/photo-sharing, and so on (MALONEY, 2007). The result of the increasingly common use of such technologies in education is that a significant proportion of formal higher education now only gets done in virtual spaces1). 
These dramatic changes have of course sparked much discussion of the shifting (and sometimes even "collapsing") of cultural discourse practices and identities (e.g. AL-SAGGAF & WILLIAMSON, 2004), very often framed in terms of debates about globalization (see BACHMANN's, 2007 review of BLOMMAERT's, 2005, and FAIRCLOUGH's, 2006, approaches to the study of discourse and globalization). 
The term "interculturality" refers, in broad terms to "cultural negotiations between social subjects" CATALAN-ERASO (2006, p.2), the coming together and working out of practices by people from different cultural backgrounds. As GUMPERZ (1994) pointed out some time ago (and before many of the technological moves that have expanded the frequency and prevalence of intercultural exchanges so dramatically), intercultural exchanges of this type are a normal aspect of modern daily life. GUMPERZ (1972) uses the term "speech communities" to describe socio-linguistic groupings such as countries, tribes, religious groups (1972, p.16), and suggests that intersections across such groupings may result in changes to linguistic form/practice. As DURANTI (1997) has argued, even comparatively simple exchanges such as greeting are organized according to complex socio-historic cultural knowledge and are dependent for their interactional accomplishment on participants "sharing" that knowledge. Where this knowledge is not shared, one might expect breaches to these taken for granted linguistic forms, with all kinds of interactional consequences. 
The interest in the shifting parameters within discursive practice, and the politicized nature of such changes (TERBORG, 2006) have also involved a strong concern with the role of communication technologies in the formation and transformation of discourse. Particular areas of interest here include the ways that specific languages operate as an inter-cultural (or perhaps "trans-cultural") discursive medium (WARSCHAUER, SAID & ZOHRY, 2002) and the emergence of new discourse modes through such intercultural discourse encounters (BLOMMAERT, 2005). 
This paper aims to explore the ways in which conversation analysis (CA) may help us to understand these types of changing intercultural discourses in relation to online modes of textual exchange. To inform this discussion, I will draw on my teaching and research experiences of using and analyzing asynchronous discourse in postgraduate education environments. The two key issues that I will be discussing are, firstly, the relevance of CA for discussing written modes of discourse, and, secondly, the ways that CA may contribute to our understanding of intercultural encounters. As a precursor to these aims I will very briefly outline some of the key features of CA and look at some of the existing literature that has used this perspective to examine intercultural communication. 
One of the central concerns in CA is with the operations of culture in action; that is, the accomplishment of culture through "talk in social interaction" (HESTER & EGLIN, 1997; SCHEGLOFF, 1997). CA shares with its "sister discipline" ethnomethodology (EM), an interest in the ways in which societal members display and make "accountable" their actions qua culture-in-action. Accountable, means something like "intelligible" or "understandable" (TEN HAVE, 2002), so EM and CA are directed towards the investigation of how members create "understandability" in social action. CA is particularly concerned with the ways that conversational machinery is used to this end. 
EM and CA have a distinctive orientation to the concept of "culture." Firstly, in EM and CA, empirical examination brackets out the assumption of "shared culture" in the shape of norms and knowledge that are held in common and that operates "behind the scenes" of social action. That is, it suspends the idea that this shared knowledge exists between the participants in a given action/community, and instead, tries to find that knowledge as visible and working practices in real social action. So, instead of explaining action according to abstract cultural rules that operate as a normative framework for action, social action is investigated such that those rules can be made visible as features of social praxis. The aim is to uncover the operations of the "stock of knowledge" as a feature of the "natural attitude of everyday reasoning" (LEITER, 1980; TEN HAVE, 2002). 
A number of studies have used CA's concern with the examination of the operations of common sense understanding within conversational practice to look at the intersections of different cultures in a common language. SCHEGLOFF's, JEFFERSON's and SACKS' (1977) early work in CA, and their observations about the organizational structures of American English conversations led to a number of studies looking at the variation of these practices across different language contexts, (e.g. BOWLES & PALLOTI, 2004; SIFIANOU, 1989, HOSODA, 2006) as well as in contexts of intercultural common language communication. HALMARI (1993) analyzed the differences between Finnish and English speakers in the context of business meetings conducted in English in the U.S. and showed that there were marked differences in the organization of the speech encounters. English speakers used a "how are you" sequence as part of the formal opening of the talk, and as a brief sequence used to get to the "business at hand" (see BUTTON & CASEY, 1984, on the organization of topics in conversation). In contrast, Finnish speakers treated such "introductory issues" as more lengthy topical sequence in their own right, requiring quite detailed exploration. It is easy to imagine comedy sketches playing on this subtle difference, and the difficulties it may create for the hapless participants. This is not by any means intended to trivialize the matter: the apparently "minor" problem of "doing introductions" has serious consequences for basic communicative exchanges. Indeed, YOUNG (1994) has looked at the ways in which Chinese communication strategies and their "translation" into English as a second or other language have impacted on the perceptions of the Chinese by U.S. Americans. YOUNG shows that these simple cultural "ways of doing" create very tangible problems for communication that result in very real cross-cultural misunderstanding. 
There is also a growing body of work that looks at the interaction between second language and native language speakers, often in educational contexts. MORI (2003) examined the achievement of intercultural discourse and the ways in which it is worked through in the moment-by-moment shifts of discourse. MORI explored co-presence question-answer sequences between first and second language Japanese speakers in multi-party conversations and analyzed the ways that cultural differences were organizationally dealt with. MORI showed that the strategies used by first language speakers when asking questions served to categorize the intended answerers (the second language speakers) as linguistic novices. More generally, MORI's work illustrates that cultural differences as represented in linguistic ability are visible aspects of talk, that are used by participant's to structure their conversations in orientation to their understandings of the "others" they are orientating to. A similar point is made by PARK (2007), who looks at the ways that the categories Native Speaker (NS) and Non-Native Speaker (NNS) are made "procedurally relevant" to and in conversational interaction. Park shows that NS and NNS are related to identities (categories) of "expert" and "novice" that are treated as emergent, sequentially realized and negotiated feature of the talk. 
CHOO, AUSTIN and RENSHAW (2007) looked at the ways in which Chinese and Vietnamese teachers and parents used cultural categories as a means of accounting for everyday educational decisions about children. The aim of the analysis was to see the types of categories that were relevant to the participants, and how they put them to use. The authors showed the nuanced ways in which different categories were used as tools for reasoning. For example, the authors illustrate how the association of students with particular nationalities was used as a means of giving explanations for perceived differences between students, e.g. how students' behavior as "fitting in," "answering back" and "speaking English" were related to the category of "Australian." In these sorts of ways, the category(s) of "nationality" were put to work as mechanisms for explanation of differences between students. 
Studies such as these, with their focus on the mechanisms of conversational structure as manifestations of a "stock" of knowledge, help to reveal some of the issues that can emerge through intercultural negotiations. In this way the problematic of cultures coming together in intercultural exchanges gains definition as the investigation of contexts for conversational praxis reveal distinctive issues that are a property of distinctive "speech communities" (FISH, 1978) working through discourse together. 
I will now discuss the relevance of the conversation analytic perspective for the examination of online written discourse. 
Conversation analysis is directed towards the examination of talk as a constitutive site of culture: as SCHEGLOFF (1997) has famously put it, the interest is in the mechanisms of "talk in social interaction." In the context of studies of Computer Supported Collaborative Work (CSCW), CA has been used to look at the ways that participants collaborate through synchronous distributed communication fora such as teleconferencing (e.g. RUHLEDER & JORDAN, 2001) and online games (e.g. MOORE, DUCHENEAUT & NICKELL, 2006; MOORE, GATHAM, DUCHENEAUT & NICKELL, 2007). While there are some examples of work within ethnomethodology that have concentrated on the processes of making sense of texts (e.g. McHOUL, 1982; LIVINGSTON, 1987), there have been very few examples of studies that have used a specifically CA perspective to look at written modes of communication. TEN HAVE (1999) suggested that the CA approach could be very useful for looking at online discourse, and the ways that postings are structured in order to produce particular kinds of readings. The remainder of this paper explores the ways in which the concepts of sequentiality (the turn-taking patterns of conversation) and membership categorization analysis (cultural categories and classificatory schemas) may help to make sense of the organization of asynchronous written discourse in intercultural settings. 
A key concern in CA has been with the participants in conversation create sequences of talk by taking turns at speaking. Turns are constructed by participants orientating to tacit knowledge about how turns operate: In his lectures, SACKS (1992) proposed a number of maxims that can be seen to operate as general procedures for talk. Three of the most basic of these are: (1) that one person speaks at a time; (2) that conversational turns do not overlap; (3) that people take turns at producing turns (see also SACKS, SCHEGLOFF & JEFFERSON, 1974). There are lots of other maxims that participants use to decide, inter alia, who's turn it is next, when it is their turn, when might be a good time to make a conversational turn, what kinds of topics those turns might reasonably deal with, how turns can be organized to bring about an opportunity to talk about something, and so on. These basic maxims and conversational mechanisms are used to "read" contexts, conversational participants and their interactional "intentions." So, the ways in which participants organize their talk will tell you something about their role in that setting, their expectations of other people's roles in that setting, their intentions for what the setting should accomplish, and so on. 
For example, in face-to-face postgraduate educational seminars, teachers and students have been shown to use their conversational turns in distinctive ways. Teachers have—or "show" or "are treated as having" or "regularly display"—greater interactional rights in terms of the selection of next speaker than students do, and can "close down" other people's talk in ways that student cannot (GIBSON, HALL & CALLERY, 2006). Those particular educational contexts are comprised of and constituted by those kinds of conversational mechanisms. The very fact that participants regularly display preferences for and deference to these turn-taking procedures is, in part, what distinguishes those interactive environments from any other; in other words, these types of conversational practice constitute educational talk. 
There are clearly problems with moving the conversation analytic concern with "working out the organization of talk through sequential turns" to online asynchronous discussions. We might wonder about the relevance of the first conversational maxim ("one person speaks at a time") in a context in which conversational participants don't "speak" and certainly don't do so, or don't typically do so, at the same interactional moment. Conversations online are normally distributed across time, and do not occur at a given temporal point (I say "normally," because there are instances that by design or happenstance, two or more participants may be holding a conversation via postings at the same time). So, the maxim "one person speaks at a time" would seem an odd one for characterizing online asynchronous discourse. For the same reason the second maxim ("conversational turns do not overlap") is somewhat inappropriate. 
Elsewhere (GIBSON, forthcoming) I have illustrated that there is some evidence that the negotiation of turns and the taking of turns has been of concern in online postgraduate discussion forums, which may mean that the third maxim ("people take turns at producing turns") could have some interactional relevance. I showed that there were instances of discourse online where the ordering of turns displayed a preference that conventional discourse interaction rights were maintained, and where posting "turns" followed a similar pattern to similar exchanges in spoken talk. However, there are certainly also very many examples in which individual contributors do post "turn" after "turn" in a way that seems to breach maxim three. 
So, apart from the fact that these online asynchronous conversations are posted one after another is there any warrant to suggest that these turns bear any resemblance to conversational talk, or that CA may be useful for examining this mode of discourse? In order to answer this question I will briefly talk about another key concept in CA as introduced by SACKS (1992): adjacency pairs. An adjacency pair is a sequence of conversational turns that are tied to each other in which the former calls forth the latter. Examples of an adjacency pairs include "question—answer"; "greeting—greeting"; and "request—reply." In all cases, a turn that is readable as the first part of a pair creates a strong interactional preference that it is followed by the second part of a pair. In my analysis of online discussions (GIBSON, forthcoming) I found that one of the methodological challenges was to uncover the logical sequence of the posts. Precisely because postings are not synchronous, participants do not always take turns at relevant "places" but "take turns when they can" (e.g. when they log on), which means that "turn placement" is often rather haphazard in comparison to the order of turns in face-to-face conversation. However, participants nonetheless do display what TEN HAVE (1999) has described as a "reading path," and demonstrate through the organization of their posts how they functionally relate to other contributions. Answers are readable as such wherever they occur in a sequence of posts; postings in "answer positions" (next turns) that are not answers, are readable as such even though they do occur in answer positions; postings that provide answers to more than one question or that index multiple posts in some way, are readable as such. 
To give a small example, Figure 1 shows a post in an online reading group discussion board. The posting is the first post in this particular discussion board, which includes twelve postgraduate students and a tutor. The post outlines two distinct questions. The five posts that immediately follow this (as "replies" to the post—i.e. as representationally linked to it), do not readably relate to the post as they make no reference to it. For example, Figure 2 shows the post that occurred immediately after Figure 1 and which reports on the author's experiences of trying to access the online library, and subsequently asks the other students for help. The four posts following this involve some answers to Thomas' questions in Figure 2 and some descriptions of their shared experiences. Finally, six posts after Sarah's initial question in Figure 1, the tutor posts a response that encourages the other participants to answer Sarah's questions. This post contains a typical "tutor like" move, of asking the other participants to try to answer the question. The discourse immediately "looks familiar" as a kind of classroom dialogue, where participants have observable interactional rights to, for example, re-direct questions. The tutor's post can be seen quite clearly as a move that topicalizes the questions as providing relevant points of discussion for the rest of the group. It can be read as doing this even though the post does not directly follow the question as adjacency pairs do in spoken conversation (SACKS, 1992).
Forum: Activity one
Date: 07-14-2006 10:29
What does Berliner mean by "organisational competencies"? How does that relate to Becker's notion of normative practice?
Figure 1: The first post in an online asynchronous reading group discussion
Forum: Activity one
Date: 07-16-2006 13:32
Subject: Online Library
Has anyone had problems accessing Has anyone had problems accessing the library? I can't seem to get the articles I am requesting through the databases. How do you get to the actual text? I just get publication information.
Thanks for your help,
Figure 2: The second post in an online asynchronous reading group discussion
Forum: Activity one
Date: 07-18-2006 11:06
Interesting questions Sarah. Thanks for starting us off. I will let the others try to deal with this before I jump in.
Figure 3: The seventh post in an online asynchronous reading group discussion 
This brief example suggests that the organization of written discourse in these visual forums works quite differently to spoken forums. Sequences of exchange can be arranged in non-chronological ways, and still maintain their sense of sequential order. They do this by using similar mechanisms of ordering found in spoken conversation. For example, in Figure 3, the tutor's reference to Sarah indicates the recipient design of the tutor's comments. The questions being referred to are obviously Sarah's questions, and not anybody else's (like Tomas's). The tutor does not refer to the other posts in the discussion board that precede their response to Sarah, but treat their response as if it were sequentially consecutive: they treat their dialogue as a question-answer sequence. 
This small example is intended merely to illustrate that sequence is potentially not only a relevant but an important resource for both readers of and contributors to online asynchronous forums. Without an orientation to sequentiality, there could be no sense of discursive practice: there could be no question-answer patterns for the participants to orientate to, as these are constituted in the discursive operations of the discussion board members. The exploration of these modes of analysis in detail is beyond the confines of this paper. However, I would like to make a preliminary analytic distinction in order to help think about the relevance of these ideas for exploring intercultural discourse: In asynchronous dialogue, there are at least two types of sequential ordering—the relational ordering between posts (e.g. the sequential order that figures one, two and three have in relation to each other), and the constitutive ordering of individual posts (i.e. the order of the discourse within a given contribution). I propose this distinction merely as a means of talking about the ways that discourse is produced rather than as a means of categorizing different types of discourse. 
As we saw at the beginning of this paper, the sequential ordering of talk is a cultural matter—different cultures and groups orientate to the construction of dialogue in different ways, with quite basic features of talk such as greetings taking quite distinctive cultural forms that are constituted in the sequential ordering of talk (YOUNG, 1994; HALMARI, 1993; MORI, 2003). As such, the fact that sequentiality is an important resource for both readers and contributors to online asynchronous discussions means that it may also be a useful way for reflecting on the working out of intercultural practices online. By inspecting the exchanges between participants, and the ways that conversational dialogues are worked out and ordered as sequential matters, we may come to see important cultural differences in newly emerging online discourses, and begin to empirically explore the implications of such intercultural exchange in terms of the sequentioal ordering of online discourse. It is of course well beyond this paper to explore such issues in detail. Instead, I want to show some of the ways in which CA may help to evidence culture as a feature of social interaction. In the following section I wish to provide brief illustrations of how this basic concern with culture might be taken forward in the exploration of constitutive ordering, and then move to look at the relevance of membership categorization analysis (MCA) as an approach to analyzing culture. 
As we have seen, individual posts can be examined to find out the contextualization cues that are used to interactionally situate them. By "interactionally situate" I mean display the interactional context in which they operate. Two of the basic features of an interactional context are the person to whom a conversational remark is addressed (the "recipient design") and the production of a sense of relevance to context (i.e. that what is being said pertains to the context). To return to Figure 1, it is noticeable that the post does not orientate to either of these two basic interactional practices. For example, the post does not contain a specified recipient; all of the other posts in this discussion either directed their questions to particular people, or explicitly identified the post as open to "anyone" or "all." Further, the post does not provide any discourse markers that contextualize the questions, but merely asks them, i.e. it does not start with "I was wondering" or "I have a question" as many other posts did, and does not offer a subsequent context in which the question can be situated, as the post in Figure 2 does (i.e. "I can't seem to get the articles I am requesting through the databases"). 
In both of these respects Figure 1 is an example of a post that breaks important and basic interactional practices (practices that are observable and observed within the same context). The author of this post was not native English speaker. Their written English often contained features that, as with the above example, displayed what we might call distinctive discourse practices. For example, in a number of other posts the author had a very informal style of discourse, that would involve, for instance, long stories about members of her family, and frequently used religious phrases such as "god be with you" as a means of ending posts. It is my no means my intention to offer value judgments on any of these aspects, or to imply a simple correlation between "discursive differences" and cultural differences. Rather, I suggest that the characteristics that I have outlined here, in addition to the conversational "breaches" outlined above, represent discourse features that could be the basis for asking questions about potential cultural differentiation within discourse and their sequential handling. In this way, culture may be treated as an observable phenomenon that is noticeable in the unfolding interaction of participants. I will come back to this point in the last section of this paper. 
Membership categorization analysis involves exploring the categories used in conversation and their role in achieving and working through interactional events. The basic principle is that people display common-sense knowledge through the use of these categories, which can be examined to reveal something about how categories are constructed and used in the achievement and negotiation of interaction; in short "how social phenomena are unavoidable made available in and as their description" (EGLIN & HESTER, 2003, p.125). In intercultural interaction, inquiry becomes directed towards the differential uses of knowledge categories in social interaction, and the ways that these are dealt with by participants (e.g. CHOO et al., 2007). 
In his early writing SACKS showed the relevance of category analysis by looking at how people routinely use categories to order information. SACKS introduced the term "membership category device" (MCD) to describe a collection of categories. For example, the terms father, son, daughter, wife, mother, can be heard as part of a more general category collection, "family." SACKS proposed various maxims that operate as analytic expressions of the common sense knowledge that informs the use of such categories and category devices. To give an illustration, SACKS proposed that there is an observable maxim that where one person is described by one category, a second person introduced as part of the same scene, and who is part of the same population as the first person, will be described according to a category in the same MCD. So, a person may describe having seen a "father" and "son" having an argument, or a "teacher" and a "pupil." The categories in each of these pairs are part of some bigger MCD (which in these instances we might describe as "family" and "school members"). Furthermore, where those descriptors imply a relation, then they will be heard as being related (as in the father of the boy; the teacher of the pupil). However, where categories from different MCDs are used, the parties will not be heard as "related." For example, the same adult and child could have been described as a "car driver" and "juvenile," but this would create an entirely different sense, with different implications for what the story was about (even though, it too, could be an accurate description of the same people). The important point to note here is that categories are used and interrogated by people in social interaction to create a "sense" of some telling or other, and to construct scenes in particular ways. 
Another key aspect of this is the idea that categories are bound to particular activities, and there are preferences in the ways in which categories are used in particular settings. So, for example it is better to say "the doctor cured the patient" than "the mother of two cured the patient," even though both may be "reasonable" descriptions of a given interaction. However, membership categories are put to use in contexts, and the using or breaching of a given maxim may be inspected to display a person's design. So, were "mother of two" used in place of "doctor," analysts, as with members, can inspect the talk to find out the relevance of that category in that circumstance. It is well beyond the confines of this paper to outline in detail this perspective (interested readers may find the following particularly useful: EGLIN & HESTER, 1992, 2003; HESTER & EGLIN, 1997; SILVERMAN, 1998; WATSON, 1978 and of course SACKS, 1992, particularly pages 40-48). Instead, I want to turn my attention to a brief exploration of the issues of this mode of analysis in relation to asynchronous written discourse and intercultural exchanges. 
Membership categorization analysis has a history of use to analyze textual resources (EGLIN & HESTER, 1992; LEE, 1984; FRANCIS & HART, 1997). Unlike sequential analysis, this approach is not directed towards understanding how sequential turns at speech are organized, but at how categories function as mechanisms for the construction of intelligibility. This is not, then, a method for examining just talk, but of looking at discourse, more generally conceived. 
Different discursive events are carried out through distinct categories that are put to work in contextually specific ways in the achievement of particular practices. Continuing with my discussion of online asynchronous discourse in postgraduate education settings, a part of the way in which the "reading path" (TEN HAVE, 1999) of a given set of postings is constructed is through the use of categories to "index" one post to another (GIBSON, forthcoming). For example, a posting becomes readable as an "answer" through the use of terminologies and categories that match the MCD implied in the "question" post to which it is a reply. SACKS (1992) proposed that topical ties are created by re-using the categories that have already been introduced by others. The functional intelligibility of a post is brought about through the use of categories as commonsensically linked to another. It is in this respect that the analysis of categories is to be seen as an important component of CA in online discourse. 
This mode of analysis has a very real potential to reveal the ways in which new categorical mechanisms are formed and deployed as constitutive features of local discourse. The incumbent categories that make up the specific sense-producing mechanisms of any given setting are visible, and their functional relevance to one another is recoverable. Similarly, the ways in which people operating with different "taken for granted" categorical knowledge can run into problems when operating in unfamiliar discourse environments is also recoverable. For example, "different English-speaking populations may draw from the same pool of lexical-grammatical material, yet routinely rely on different contextualization conventions and interpretive practices to achieve their interactive ends" (YOUNG, 1994, p.42). The "contextualization conventions" when conceived of as "membership category devices" are retrievable by the analyst inspecting the setting to try to find out the categorical mechanisms that may have produced the "mis-application." Where this relies on talk across languages, the analysis will be more successful where the analyst possesses knowledge of both idioms. 
Figure 4 provides a striking example of a contribution that seems to breach some of the practices of the other posts and to draw on different "contextualization information" about, for example, how online classrooms should operate. The formal means of addressing the tutor was never used by other participants and therefore stands out as a distinctive practice that constitutes a particular kind of relation between student and tutor. The category of address does not neatly fit with the categories used in this particular higher education context, which usually involve the use of first names. Again, one example is not enough to make strong claims about the generality of this as a cultural difference, but it is enough to raise a suspicion about the ways that different modes of discourse might constitute the environment in very different ways.
Forum: G. H. Mead
Date: 01-15-2006 06.53
Subject: Sunday Evening
I have a view. It will be not good to give definitions without examples. We researchers must see people doing peoples things for us to be clear what autonomy could mean. Does it make sense?
Figure 4: An example of a formal opening in an online discussion 
The concern with cultural categories is also often a visibly negotiated practice. In figures five and six we see an exchange between two participants where the relevance of a particular form of address is worked out between the participants. The first post uses the Japanese suffix "kun" to name the person being addressed by the post. The naming category is given relevance because of the Japanese nationality of the recipient and to the topic of discussion (Japanese education). In Figure 6 the recipient responds and explicitly comments on the appropriateness of this suffix in the context of female friendships in Japan. The participants make the category of "that particular way of naming" available to each other as a matter for negotiation, and produce a context for its use. The participants make the cultural difference between each other a visible discourse artifact.
Date: 02-02-2007 13:32
Subject: Question about the reading
I liked the description of your interests in educational attainment. Could you tell me more about the nature of the system in your own context?
Figure 5: Part one of a two-part exchange in an asynchronous discussion
Date: 02-12-2007 19:41
Subject: Question about the reading
Thanks for your comments Jody,
I am not really interested in education in Japan, but in educational attainment in general. However, let me tell you a bit about Japanese education
[... SECTION OF POST OMITTED ...]
By the way, in Japan, girls often use "kun" with their friends.
Figure 6: Part two of a two-part exchange in an asynchronous discussion 
To conclude this section, while there are clear differences between talk and online asynchronous discourse, the use of sequential concerns together with membership category analysis could be, I suggest, a very productive approach to revealing the ways in which participants create, sustain and negotiate distinct discourse practices in online environments. 
Both sequential analysis and MCD analysis are directed towards looking at culture in action such that the mechanisms and organization of sequential ordering and common sense knowledge and categories can be revealed. They aim to make visible the facticity of social praxis and the methods by which people treat it as a real and constraining feature. This is the motivated gaze of EM/CA: to look at what is "demonstrably relevant" (SCHEGLOFF, 1999) for the parties in the interaction, and the ways that they use sequences and categories as features of their talk/discourse in social interaction. 
One of the issues that this position raises for the examination of culture as a sociological artifact is that, from one perspective at least, culture is not treated as a mechanism that drives action, but as an observable feature of it. As such, in one view (e.g. SCHEGLOFF, 1999) culture cannot be extracted from discursive contexts as a sociological explanation for conversational praxis but can merely be examined as an interactional event. I refer to this as the "culture in praxis" thesis.2) A good example of this problem is found in the reaction of some conversation analysis to the idea that CA might be usefully combined with feminism. From the culture in praxis perspective, CA can never be combined with feminism because the latter involves a motivated analytic gaze that privileges a particular sociological concept as an explanatory factor within discourse (see STOKOE & SMITHSON, 2001; WOWK, 2007; KITZINGER, 2008). In a similar way, culture cannot be privileged in intercultural exchanges as an abstracted conceptual concern, but must simply be an empirical focus and be shown to be of relevance and of issue within and through the analysis of conversation. 
On the other hand, some analysts have argued that talk or discourse itself will not tell you everything you need to know about the cultural operations that inform its organization. In many examples of MCA, the cultural knowledge that forms the organization of categories is produced through the analyst's explanation and not in the talk itself (as in Figure 4). While these forms of cultural knowledge are of course worked out, used and displayable sequentially, and in that respect compatible with sequential analysis (WATSON, 1978), there are implicit cultural rules that are not displayed but are rather orientated to and which require the researcher to go beyond the visible written context to account for them (see McHOUL, RAPLEY & ANTAKI, 2008, for a discussion of these debates). For example, our analysis of the categories of naming for tutors only comes to look strange when the ordinary rules of naming practice are explained. These are not necessarily obviously displayable in the talk, but require a furnishing of the context by, for example, explaining the procedures of discourse in higher education context in that institution. In this approach then, context becomes a mechanism for explanation and perhaps helps culture to be recovered as an explanatory resource rather than just an empirically demonstrable one. However, even from this perspective, culture is typically not regarded as an explanatory mechanism; culture is not treated as a normative practice that "drives" action, but is a performative enactment that constitutes practice. 
The implications of both of these approaches to CA are that the investigation of interculturality must be an empirical enterprise and not a theoretical one. The focus on the praxeological workings of discourse cultures or communities must be shown as practices, and not as idealizations; culturally distinct knowledge must be shown to be in operation and its forms, functions, and implications are the focus of empirical inquiry. The various methodological resources that I have described in this paper are, I suggest, useful mechanisms for empirically pursuing questions related to intercultural practices. 
This article has sought to explore the uses of conversation analysis for exploring "intercultural" discourse in online asynchronous settings. In the context of sequential analysis in CA, there is some difference between the normal concerns with the negotiation of turns at talk and the exchange of written turns online. However, I have also argued that the general interest in the construction of sequential relevance is transferable to these distinctive discursive environments. In relation to intercultural communication, I suggest that cultural practices and cultural differences are also displayable within such discourse. One important mechanism for re-capturing the cultural mechanisms is the use of membership categorization analysis, which helps to display the ways in which commonsense categories are put to work within discourse settings. 
As yet, there are very few studies that use this type of analysis in asynchronous contexts, so it is very difficult to make strong pronouncements about the empirical successes or methodological problematics that could emerge from their application. What I have sought to do in this short article is to simply highlight that the CA concerns with culture and discourse in action are not incompatible with these increasingly popular mechanisms of intercultural discourse. This type of approach may therefore help us to empirically situate some of the "theoretical" problems outlined in the introduction. However, it must be remembered that CA's concern with conversation as culture in action creates limitations for treating culture as a normative drive to action. 
With many thanks to Dirk VOM LEHN, Mariana JIMÉNEZ HUERTA and Heladio JIMÉNEZ ZARAGOZA for comments on the paper.
1) If I may use myself as an illustrative case, I run a comparatively small masters course (currently just over 30 participants) that has representation from over 21 countries around the world, with students operating in at least ten different time zones. Towards the end of this paper, I will be using data from my analysis of the discourse in online postgraduate educational environments such as this to help explore the applicability of conversation analysis for examining intercultural communication. All of the data provided is anonymized and has been provided by permission of the participants. <back>
2) See also the articles of OTTEN and GEPPERT (2009), as well as of BUSCH (2009) in this thematic issue. <back>
Al-Saggaf, Yeslam & Williamson, Kirsty (2004). Online communities in Saudi Arabia: Evaluating the impact of culture through online semi-structured interviews. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 5(3), Art. 24, http://nbn-resolving.de/urn:nbn:de:0114-fqs0403247 [Date of access: December 20, 2008].
Bachmann, Iris (2007). Review: Norman Fairclough (2006). Language and globalization / Jan Blommaert (2005). Discourse. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 9(1), Art. 36, http://nbn-resolving.de/urn:nbn:de:0114-fqs0801361 [Date of access: December 20, 2008].
Blommaert, Jan (2005). Discourse. Cambridge: Cambridge University Press.
Bowles, Hugo & Pallotti, Gabriele (2004). Conversation analysis of opening sequences of telephone calls to book stores in English and Italian. Textus, 17(1), 63-88.
Browne, Elizabeth (2003). Conversations in cyberspace: A study of online learning. Online Learning, 18(3), 245-259
Busch, Dominic (2009). The notion of culture in linguistic research. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(1), Art. 50, http://nbn-resolving.de/urn:nbn:de:0114-fqs0901508.
Button, Graham & Casey, Neil (1984). Generating topic: The use of topic initiation elicitors. In J. Maxwell Atkinson (Ed.), Structures of social action (pp.167-190). Cambridge: Cambridge University Press.
Catalán Eraso, Laura (2006). Reflecting upon interculturality in ethnographic filmmaking. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(3), Art. 6, http://nbn-resolving.de/urn:nbn:de:0114-fqs060369 [Date of access: December 20, 2008].
Choo, Juliet; Austin, Helena & Renshaw, Peter (2007). Reasoning with cultural categories in understanding student diversity. International Journal of Educational Research, 46(3/4), 229-239.
Duranti, Alessandro (1997). Universal and cultural-specific properties of greetings. Journal of Linguistic Anthropology, 7(1), 63-97.
Eglin, Peter & Hester, Stephen (1992). Category, predicate and task: The pragmatics of practical action. Semiotica, 88(3/4), 243-268.
Eglin, Peter & Hester, Stephen (2003). The Montreal massacre: A story of membership categorization analysis. Waterloo, Ontario: Wilfrid Laurier University Press.
Fairclough, Norman (2006). Language and globalization. London: Routledge.
Fish, Stanley (1978). Normal circumstances, literal language, direct speech acts, the ordinary, the everyday, the obvious, what goes without saying, and other special cases. Critical Inquiry, 4(4), 625-644.
Francis, David & Hart, Christopher (1997). Narrative intelligibility and membership categorization in a television commercial. In Stephen Hester and Peter Eglin (Eds.). Culture in action: Studies in membership categorization analysis (pp.123-152). Washington D.C.: International Institute for Ethnomethodology and Conversation Analysis and University Press of America.
Gibson, Will (forthcoming). Negotiating textual talk: Conversation analysis, pedagogy, and the organisation of online asynchronous discourse. British Educational Research Journal.
Gibson, William; Hall, Andrew & Callery, Peter (2006). Topicality and the structure of interactive talk in face-to-face seminar discussions: Implications for research in distributed learning media. British Educational Research Journal, 32(1), 77-94.
Gumperz, John J. (1994). Foreword. In Linda Young (Ed.). Cross talk and culture in Sino-American communication (pp.xiii-xix). Cambridge: Cambridge University Press.
Gumperz, John J. (1972). Introduction. In John J. Gumperz & Dell Hymes (Eds.),. Directions in sociolinguistics: The ethnography of communication (pp.1-25). New York: Holt, Rinehart and Winston.
Halmari, Helena (1993). Intercultural business telephone conversation: A case of Finns vs. Anglo-Americans. Applied Linguistics, 14(4), 408-430.
Hester, Stephen & Eglin, Peter (1997). Culture in action: Studies in membership categorization analysis. Washington D.C.: University Press of America.
Hosoda, Yuri (2006). Repair and relevance of differential language expertise in second language conversations. Applied Linguistics, 27(1), 25-50.
Kitzinger, Celia (2008). Developing feminist conversation analysis: A response to Wowk. Human Studies, 31(2), 179-208.
Lee, John R.E. (1984). Innocent victims and evil doers. Women's Studies International Forum, 7(1), 69-78.
Leiter, Kenneth (1980). A primer on ethnomethodology. Oxford: Oxford University Press.
Livingston, Eric (1987). Making sense of ethnomethodology, London: Routledge and Kegan Paul.
Maloney, Edward J. (2007). What web 2.0 can teach us about learning. Chronicle of Higher Education, 53(18), http://chronicle.com/ [Date of access: February 17, 2009].
McHoul, Alec (1982). Telling how texts talk: Essays on reading and ethnomethodology. London: Routledge and Kegan Paul.
McHoul, Alec; Rapley, Mark & Antaki, Charles (2008). You gotta light? On the luxury of context for understanding talk in interaction. Journal of Pragmatics, 40(5), 827-839.
Moore, Robert J.; Ducheneaut, Nicolas & Nickell, Eric (2006). Doing virtually nothing: Awareness and accountability in massively multiplayer online worlds. Computer Supported Cooperative Work (Online), 16(3), 265-303.
Moore, Robert J.; Gathnam, E. Cabell Hankinson; Ducheneaut, Nicolas & Nickell, Eric (2007). Coordinating joint activity in avatar-mediated interaction. In Bo Begole, Stephen Payne, Elizabeth Churchill, Rob St. Amant, David Gilmore & Mary Beth Rosson (Eds.), Proceedings of the SIGCHI conference on Human factors in computing systems CHI 2007 (pp.21-30). San Jose, California: ACM.
Mori, Junko (2003). The construction of interculturality: A study of initial encounters between Japanese and American students. Research on Language and Social Interaction, 36(2), 143-184.
Otten, Matthias & Geppert, Judith (2009). Mapping the landscape of qualitative research on intercultural communication. A hitchhiker's guide to the methodological galaxy. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(1), Art. 52, http://nbn-resolving.de/urn:nbn:de:0114-fqs0901520.
Park, Jae-Eun (2007). Co-construction of nonnative speaker identity in cross-cultural interaction. Applied Linguistics, 28(3), 339-360.
Ruhleder, Karen & Jordan, Brigitte (2001). Co-constructing non-mutual realities: Delay-generated trouble in distributed interaction. Computer Supported Cooperative Work, 10(1), 113-138.
Sacks, Harvey (1992). Lectures in conversation: Vol 1 & 2. London: Blackwell.
Sacks, Harvey; Schegloff, Emanuel & Jefferson, Gail (1974). A simplest systematic for the organization of turn-taking in conversation. Language, 50(4), 696-735.
Schegloff, Emanuel (1997). Whose text? Whose context? Discourse and Society, 8(2), 165-187.
Schegloff, Emanuel (1999). Naïveté vs. sophistication or discipline vs. self-indulgence: A rejoiner to Billig. Discourse and Society, 10(4), 577-582.
Schegloff, Emanuel; Jefferson, Gail & Sacks, Harvey (1977). The preference for self-correction in the organization of repair in conversation. Language, 53(2), 361-382.
Sifianou, Maria (1989). "On the telephone again" differences in telephone behaviour: England versus Greece. Language in Society, 18(4), 527-544.
Silverman, David (1998). Harvey Sacks: Social science and conversation analysis. New York: Oxford University Press.
Stokoe, Elizabeth & Smithson, Janet (2001). Making gender relevant: conversation analysis and gender categories in interaction. Discourse and Society, 12(2), 217-244.
Ten Have, Paul (1999). Structuring writing for reading: Hypertext and the reading body. Human Studies, 22(2-4), 273-298.
Ten Have, Paul (2002). The notion of member is the heart of the matter: On the role of membership knowledge in ethnomethodological inquiry. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(3), Art. 21, http://nbn-resolving.de/urn:nbn:de:0114-fqs0203217 [Date of access: December 20, 2008].
Terborg, Roland (2006). La "Ecologia de presiones" en el desplazamiento de las lenguas indigenas por el español. Presentación de un modelo. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(4), Art. 39, http://nbn-resolving.de/urn:nbn:de:0114-fqs0604396 [Date of access: December 20, 2008].
Warschauer, Mark; Said, Ghada R.E. & Zohry, Ayman (2002). Language choice online: Globalization and identity in Egypt. Journal of Computer-Mediated Communication, 7(4), http://jcmc.indiana.edu/vol7/issue4/warschauer.html [Date of access: February 17, 2009].
Watson, Rod (1978). Categorization, authorization and blame-negotiation. Sociology, 12(1), 105-113.
Wowk, Maria (2007). Kitzinger's feminist conversation analysis: Critical observations. Human Studies, 30(2), 131-155.
Young, Linda Wai Ling (1994). Crosstalk and culture in Sino-American communication. Cambridge: Cambridge University Press.
Will GIBSON's work explores the applications of ethnomethodology, conversation analysis and symbolic interactionism to a broad range of empirical fields, including online discourse, medical practice, and musical performance. He is interested in the intersection between these approaches to sociological study and to qualitative research methodology more broadly conceived.
Dr Will Gibson
Faculty of Culture and Pedagogy
Institute of Education, University of London
20, Bedford Way, London WC1H 0AL, UK
Tel.: +44(0)20 76126603
Gibson, Will (2009). Intercultural Communication Online: Conversation Analysis and the Investigation of Asynchronous Written Discourse [39 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(1), Art. 49, http://nbn-resolving.de/urn:nbn:de:0114-fqs0901493.