Conversation analysis

It may be surprising to find conversation analysis, which originated within sociology, included in a discussion of “linguistic approaches” to workplace communication. The reason for doing so is that the micro-analytical approach and fine-grained turn- by-turn method advocated by conversation analysis results in a detailed and linguistically rich description of spoken interactions. As already mentioned, studies of spoken workplace interactions using conversation analysis have a long tradition, and this approach continues to be applied very widely to the analysis of spoken workplace discourse.

While conversation analysis originally set out to explain the “machinery” of everyday conversation, conversation analysts very soon turned to an investigation of “institutional” encounters. Many of these early studies were published in Drew and Heritage’s 1992 volume Talk at Work including analysis of talk in health-care delivery, legal proceedings and news, and job interviews. Using ordinary, naturally- occurring conversation as a benchmark, conversation-analytic studies seek to establish which interactional characteristics distinguish institutional encounters. Some general distinguishing features of institutional talk were identified (Drew and Heritage 1992; Heritage 1997), as discussed in Section 2.1, the key one being goal orientation. Numerous studies of specific institutional and workplace encounters have also sought to identify the ways in which interlocutors orient to the specific goals and activities within these different institutions and what interactional features distinguish these encounters both from ordinary conversation and from one another (see also Hutchby and Wooffit 2008; Pallotti 2009). A detailed discussion of these numerous studies would be beyond the scope of this chapter, but a number of examples from a range of workplace setting will be presented here to illustrate the method and present some of the findings. First, though, the basic methodology and principles of conversation analysis will be briefly explained (see Hutchby and Wooffitt 2008 for a more detailed description of these phenomena).

Conversation-analytic methodology starts from the basic principle of trying to make sense of spoken interactions from the participants’ viewpoint, that is, it takes an “emic” perspective and avoids a priori analytical frameworks and categories. The analyst begins by carefully studying a single recording/transcript or “case”, and tries to identify any recurring patterns in the interaction. Similar interactions are then investigated for these patterns, and a “collection” is built up. Despite this principle of avoiding a priori categories, the foundational studies in conversation analysis identified a number of recurring interactional patterns in conversation that make up the “machinery” of talk and serve as a baseline against which to analyse other encounters, including institutional ones.

These basic patterns include a set of “rules” for “turn-taking” (Sacks, Schegloff, and Jefferson 1974), “adjacency pairs”, “sequences” (Sacks 1992), “preference” (Pomerantz 1984) and “repair” (Schegloff, Jefferson, and Sacks 1977). Adjacency pairs, consisting of a first and a second pair part, are the most basic unit of interaction; typical pair types are question - answer, greeting - greeting, request - compliance. Of course, conversation does not always proceed with such simple, two- turn units, and longer sequences may occur. But once a first pair part is uttered, it sets up the “conditional relevance” for a corresponding second pair part to be produced. This may not happen until several turns later, perhaps after a “side sequence” which might answer preliminary questions relevant for the expected second-part response. On the other hand, if no response at all is forthcoming, this will be noticed. The notion of preference addresses the phenomenon that certain first pair parts (e.g., request, offer) may be followed by either a positive or negative response. “Preferred” responses (e.g., compliance, acceptance) are more frequent, simpler and “unmarked”, whereas “dispreferred” ones (e.g., refusal, rejection) are rarer, more complex and therefore “marked”. Finally, repair involves mechanisms for “fixing” problems in the discourse, such as disfluencies or mishearings.

In workplace conversations, these mechanisms are also in operation, but there may be systematic deviations or specific patterns which are adapted to the specific interactional goals and activities within the organisation. In an example from media discourse, Hutchby (1996) found that telephone calls to a radio phone-in broadcast, which simulate everyday social phone calls, actually display the speakers’ orientation to the institutional context in a number of ways. First of all, the opening sequence of the phone call deviates from the standard format followed in private phone calls (Schegloff 1968) in that the identities of the callers (“host” and “caller”) and the topic of conversation are fixed in advance. Hutchby also observed that the host was able to exploit a sequential feature of the phone calls, namely the fact that callers always take the first turn in the conversation, in which they are expected to state their opinion. This allows the host to comment on or challenge the caller’s contribution in the second turn, thus exerting his/her power through the sequential resource of “second position”. This power difference reflects the asymmetry of roles typical of institutional contexts, in contrast to much ordinary conversation; however, it involves a manifestation of power that is locally and discursively constructed, rather than being “given”. A criticism sometimes levelled at conversation analysis is that it involves an overly mechanistic and uncritical description of the discourse, but this example demonstrates that conversation-analytic methods can indeed be used to explore issues of power in workplace interactions.

In another type of telephone interaction involving service encounters in small to medium businesses, Varcasia (2009) used conversation-analytic methods to examine the response formats used by the service providers to respond to calls from customers. An interesting facet of this study is that the data were collected from three different countries (Britain, Germany and Italy) with interactions conducted in the respective national languages, thus allowing for cross-linguistic comparison. Data were gathered from a variety of small- to medium-sized shops and service providers (e.g., bookshops, florists, clothes shops and hairdressers) with approximately 50 phone calls from each country. The analysis focused on the structure of response formats when the customer’s request could not be satisfied. Three types of response format were identified across the data set:

• - Response + extension;
• - Insertion sequence + response;
• - Simple negative response.

The first type, response + extension, was the most frequent. It involved providing a negative response (that the product or service requested could not be provided) followed by a further action or set of actions. These were themselves classified into four types: 1) providing an apology; 2) offering an alternative solution; 3) providing a reason for non-provision of item/service; and 4) a combination of “extension” types. A simple negative response was by far the least frequent type, accounting for only about 10% of all responses. Here we can observe the interactional phenomenon of preference in operation in a workplace context. The negative responses to service requests display the typical features of a dispreferred response in that simple, direct responses are rare, whereas the majority of responses are more elaborate, following either the response + extension or insertion sequence + response format. These formats include typical features of dispreferred responses, such as delays, apologies and explanations (Levinson 1983: 332-345). Extract 3 shows a response + extension involving type 4 (combination) from an English bookstore including an explanation, apology and an extension (in this case an alternative solution):

• (3) Extract 3
• 003 C: yeah. I’m just making an enquiry do you have hm (.) i don’t know if it’s
• 004 a biography or autobiography for victoria beckham? .hh do you have it in
• 005 st[ock?
• 006 R: [no no we don’t stock ehm general biographies unfortunately.
• 007 C: right-
• 008 R: we could order to you one. e:m do you want it today?
• 009 C: .hh no it’s i- i- i- it’s not necessary, no. ...
• (from Varcasia 2009: 229)

After the caller’s (C’s) enquiry in 003-005, the call receiver (R) produces a negative response, explanation and apology in 006, and then offers an alternative solution in 008. A cross-cultural comparison of the response formats found remarkable similarities across the three data sets, showing dispreference for a simple negative response in all three languages and cultures. Some differences could nevertheless be observed; for example, simple responses were most frequent in the Italian phone conversations, whereas they were rarer in the German ones and did not occur at all in the English phone calls. Varcasia (2009: 237) warns against simply attributing these differences in response formats to cultural differences, but speculates that differences in the professionalization of the services may play a role too. Thus call receivers in UK businesses often had scripts for dealing with customers, whereas the German and Italian ones did not.

The two case studies discussed in this section illustrate how conversation- analytic methodology is well-suited to revealing the specific discourse patterns which characterise different types of workplace interactions. While conversation analysis has a micro-discursive focus, the method does enable insights into “institution- order” issues, such as power, culture and professional practice.