Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Sociol Health Illn. Author manuscript; available in PMC 2013 March 27.
Published in final edited form as:
PMCID: PMC3609551

An interactional approach to conceptualising small talk in medical interactions


In medical interactions, it may seem straightforward to identify ‘small talk’ as casual or social talk superfluous to the institutional work of dealing with patients’ medical concerns. Such a broad characterisation is, however, extremely difficult to apply to actual talk, and more specificity is necessary to pursue analyses of how small talk is produced and what it achieves for participants in medical interactions. We offer an approach to delineating a subgenre of small talk called topicalised small talk (TST), derived on the basis of conversation analytically-informed analyses of routine consultations involving orthopaedic surgeons and older patients. TST is a line of talk that is referentially independent from their institutional identities as patients or surgeons, oriented instead to an aspect of the personal biography of one (or both), or to some neutral topic available to interactants in any setting (e.g. weather). Importantly, TST is an achievement of both patient and surgeon in that generation and pursuit of topic is mutually accomplished. In an exploratory but systematic analysis, when this approach was applied to a purposive sample of surgeon-patient interactions, TST was much more prevalent in visits with White than African American patients. Accounts for possible ethnic differences in TST are suggested.

Keywords: small talk, medical interaction, ethnic differences in talk, conversation analysis


Our project is an iterative one concerning what is commonly known as ‘small talk’. In prior work, we described how small talk can be used by patients and physicians to exhibit a form of disattentiveness in medical interactions (Maynard and Hudak 2008: 663). Yet, what was clear is that not all episodes of small talk are used to disattend, and that there is significant variability in its content and placement as well. We also noticed in our patient-surgeon data an apparent asymmetry in the prevalence of what we referred to broadly as small talk, with a greater propensity for such episodes to occur in interactions with White patients than with African American patients. These findings led us naturally to question what other social actions small talk achieves in talk, and how widespread its use is, both as a practice for disattending and more generally. To systematically address these questions, we appraised the existing scholarly work on small talk, particularly to see how distinctions with the wide domain it covers have been conceptualised. Because we wanted to measure systematically the distributions of various kinds of small talk, the approach to conceptualising it presented here arose as a response to what we felt were gaps of definition and operationalisation.

Small talk is a category for varieties of conversational interaction occurring in many settings, including medical ones. Research in clinical settings has relied heavily on commonsense notions of small talk as social conversation or chitchat, superfluous to the institutional, ‘on-task’ work of dealing with patients’ medical concerns. We suggest that this simple on-off task distinction is not always clear because some talk appears to be both on and off-task at the same time. Furthermore, what is typically characterised as off-task is a variety of qualitatively distinct kinds of talk, ranging from greetings and parting exchanges to more personally oriented discussion. This lack of specificity is problematic for attempts to systematically explore the distribution of small talk across clinics and types of patients and physicians, and for sequential analyses of what it achieves for participants in medical settings. Such efforts necessitate as a starting point the identification of sequences of related talk.

This paper contributes to the theorisation and empirical investigation of small talk by critically appraising current conceptualisations, offering a new conversation analytically-informed approach to delineating a subgenre of small talk called topicalised small talk (TST). A well-delineated approach to identifying episodes of TST will enable subsequent analyses of its distribution, and of the different social actions which it can accomplish. We then report on the use of this approach to explore systematically the distribution of TST in audiotaped consultations between orthopaedic surgeons and patients, supplying an example of work called for by Heritage and Maynard (2006: 8) that ‘find(s) a meeting point between the two methodologies of coding and microanalysis’. Because this exploratory analysis suggests that the distribution of TST is strongly asymmetrical on the basis of patient ethnicity, we conclude by considering possible bases for this asymmetry.

What is small talk?

The most recent and comprehensive treatments of small talk are the collections in book and journal format edited by Coupland (2000, 2003). These publications highlight work on small talk from an array of disciplines and in a variety of contexts, both institutional and more casual settings. Coupland (2000: 1) begins with an overview of the origins of this research, summarising the traditional perspective this way:

Small talk has widely been taken, from both academic and popular perspectives, to be a conventionalized and peripheral mode of talk. It seems to subsume ‘gossip’, ‘chat’ and ‘time-out talk’ for example, although it is not helpful to try to impose firm and final definitions on these generic labels. What the labels point to is a range of supposedly minor, informal, unimportant and non-serious modes of talk, linked to the general communicative function sometimes characterized as ‘talking to avoid other problems’ (see Robinson 1972).

At least three interrelated assumptions are embedded in this perspective, some of which Coupland and her contributors (2000) begin to problematise and challenge. The first assumption is that small talk is peripheral, a descriptor which draws on a spatial metaphor to evoke the idea of small talk as being on the edge or border of other talk rather than in a central place. This positioning contributes to characterisations of small talk as minor, informal, and unimportant – that is, as trivial and superficial rather than central and essential. The idea of talk being peripheral versus central suggests a clear divide, even though there is recognition of an intimate connection between small talk and work talk in the sense that the former can help constitute social relationships that facilitate instrumental tasks (i.e. small talk ‘oils the social wheels’ of work-related discourse, Holmes 2000: 57). In this capacity, small talk is neither unimportant nor peripheral.

Although small talk in this perspective includes subgenres like gossip and chat, a second assumption is that firm definitions are counterproductive. The argument is that ‘defining small talk too rigidly as a bounded mode of talk will constrain the analysis of its social functions’ (Coupland 2000: 13). Although firm and final definitions may be premature, broad characterisations can also be problematic, particularly for analyses of social functions – first, because inclusive, somewhat vague characterisations of small talk can be extremely hard to apply to actual talk, and second, because distinct social actions and conversational sequences are combined under its rubric. We recently identified, in a broadly defined sequence of small talk prior to the initiation of a visit between a physician and patient, five distinct sequences and ‘small talk’ social actions: (1) apology and acceptance; (2) appreciation and acknowledgement; (3) joking about shared ethnicity and a self-deprecating receipt; (4) a joking compliment and a ‘modesty-display’ receipt; and (5) a ‘how are you?’ and its reply that starts the medical interview per se (Maynard and Hudak 2008). More generally, the availability and use of labels like gossip, teasing and so on in common parlance both imply some existing appreciation for distinctions within small talk and suggest that a practical grasp of these variations may be possible.

These first two assumptions – that small talk is peripheral and that it needs to be broadly defined – relate to a larger issue of what constitutes or counts as small talk. Some authors have begun to redress the problem of vague definitions by attempting more specific categorisations. One of the more detailed in this regard is Holmes (2000: 37) who conceptualises talk in the workplace as existing along a continuum with core business talk and phatic communion at the extremes, interspersed by work-related and social talk. Core business talk is crucially informative, highly focused, on-topic in terms of the agreed agenda for a particular meeting and workplace, and directly serves the organisation’s goals. Phatic communion is talk which is independent of any specific workplace context, atopical, irrelevant in terms of workplace business, and has relatively little referential content or information load. Small talk, for Holmes (2000: 38), includes both social talk and phatic communion.

The idea of a continuum works precisely because many interactions do not fit neatly into the extreme categories of business talk and phatic communion; that is, ‘distinctions between the categories are sometimes difficult to draw’ (Holmes 2000: 56) and thus largely a matter of degree. Although Holmes’ approach is an advance beyond overarching definitions, small talk continues to cover a range of different types of social talk, from ‘narrowly defined formulaic greeting and parting exchanges to more expansive personally oriented talk’. A similar range is covered by the Roter Interaction Analysis System (RIAS), a widely used quantitative coding scheme in which each complete thought expressed by the patient or physician is categorised into mutually exclusive and exhaustive categories (Roter 2001). Examples of these categories are statements of empathy, asking closed or open-ended questions, showing concern, and so on. Small talk is catalogued into the RIAS category of ‘personal remarks/social conversation’ which includes greetings, goodbyes and ‘conversation on weather, sports or any non-medical or social topic … not related directly to the discussion of general health’.

The RIAS approach and Holmes’ (2000) continuum leave incomplete an understanding of small talk in terms of specificity and sequencing. With regards to specificity, distinctive kinds of talk – gossip, verbal jousting, greetings and so on – are obscured when small talk is conceptualised in umbrella fashion. With regard to sequencing, the sequences by which small talk is collaboratively constituted and the extent to which it is co-produced, whether there are few or many turns and sequences, are missed when it is defined solely based on content.

A third assumption, critiqued by Coupland, is that small talk functions generally ‘to avoid other problems’ (Coupland 2000: 1; Robinson 1972). Generalising in this way can potentially limit seeing or searching for ways that small talk may actually be used as a resource for doing something else. As Coupland (2000: 5) remarks, ‘There is a great deal we need to discover about the local dynamics of small talk in its specific domains – how small talk is achieved interactionally, turn by turn, and what it therefore achieves for participants in situ’ (italics added).

In prior work (Maynard and Hudak 2008: 663), we described how small talk can be used to disattend the instrumental task in which one or other participant is engaged. Surgeons may use small talk to focus away from psychosocial or other patient concerns that, from the physician’s perspective, may be tangential to the central complaint or recommendation related to that complaint. Patients, too, may use small talk to disattend physician recommendations regarding disfavoured therapies (like exercise). Overall, small talk may be used to ignore, mask, or efface certain kinds of agonistic relations in which physician and patient are otherwise engaged. For this work, we suggested a structural definition of small talk, consisting of concrete sequences forming pro-social actions of one kind or another that move away from instrumental tasks as such. In this paper, we extend this work by paying closer attention to both the sequencing and substantive content of small talk as it occurs in orthopaedic surgery consultations. Below are the results of this endeavour.

Data Source

We used a subset of data from a study of informed decision making which included audiotapes of office visits between 886 older patients (≥ 60 years) and 89 orthopaedic surgeons in a large Midwestern American city and part of the neighbouring state. Institutional Review Board approval for the study was obtained from all participating institutions and informed consent from patients and surgeons prior to their participation.

Our study sample was identified in the following manner: a ‘complex decision’, defined as a decision about a diagnostic or treatment intervention that is high risk, controversial, or otherwise warrants substantial discussion (e.g. surgery), was present in 206/886 (23%) of patient-surgeon visits. We asked all eligible African American (AA) patients considering a complex decision for permission to audiotape their visit with the surgeon, and to participate in follow up telephone interviews. For each AA who agreed, we sought a White patient of the same educational attainment (at least high school vs. less than high school education). This sampling strategy is based on: (1) evidence of racial disparities in orthopaedic surgery procedures (see Agency for Health Care Policy and Research 1994, Baron et al. 1996, Special Issue – Health Disparities 2004) and the possibility that differences in communication may contribute to disparities; and (2) the assumption that patient educational attainment may influence communication. The result was 59 patient-surgeon visits: 29 White patients (12 high/17 low education) and 30 AA patients (15 high/15 low education). These 59 visits included some surgeons who saw multiple patients: 20 surgeons who saw only one patient, four surgeons who saw two patients each, seven surgeons who saw three patients each, one surgeon who saw four patients and one surgeon who saw six patients. To address the potential for clustering of patients within providers, we replaced the patients seen by surgeons who saw four and six patients with 10 patients from the original 206 visits who were seen by different surgeons yet similar in terms of ethnicity and educational attainment.

Methodological approach

Rather than articulating subgenres of small talk with analyst-defined distinctions as a starting point, we draw upon the methodology of conversation analysis (CA) to develop a CA-informed way of categorising and coding distinctive kinds of small talk in order to: (1) refine what that phrase glosses; (2) identify sequences of related talk; and thereby (3) enable coding and distributional analyses. This work of identifying the ‘domain of occurrence’ (see Schegloff 1996)1 is foundational for more traditional conversation analytic work emphasising the ‘importance of utterance sequencing and placement as a resource for performing social actions and for providing for particular interpretations on the part of coparticipants’ (Gill 2005: 455).

We began by assembling a collection of sequences generously,2 drawing on our commonsense judgment about small talk as concerning a variety of non-task oriented, pro-social topics (Maynard and Hudak 2008). We identified sites in the patient-surgeon visits where the talk was more informal and less institutional, focusing on life outside the clinic or other matters rather than medical concerns. Some episodes were clearly demarcated – the shift to small talk was clear and the content of the discussion as social and casual was also clear; other episodes, however, because of their brevity and minor departures from the medical work or instrumental tasks at hand, were less obvious and tended to include briefer sequences involving what we recognised as humour, banter, repartee, ironic commentary and so on. We used this broad stance as the starting point, but were concurrently oriented to the previously described limitations of overarching definitions. Thus, as this work progressed, we discriminated ‘topicalised small talk’ as a distinct subgenre of small talk. The examples below were selected because they illustrate broader patterns and are brief and straightforward. Each is labelled with the unique identifying number and time stamp of the audiotape, with names and identifying references changed to protect participants’ identities.

Topicalised small talk

We defined sequences of TST as those involving a topic shift or change, initiated by either the surgeon or patient at any point in the visit, to talk that both participants treat as a new line of talk that is referentially independent from their institutional identities as patients or surgeons. These sequences were selected to include the devices or practices used to make a bid or proffer to introduce a topic. Proffers (Schegloff 2007), or bids (Maynard and Zimmerman 1984), which include but are not limited to pre-topical questions, topic initial utterances, topic nominations or mentionables (Button and Casey 1985), are invitations to generate talk with independent topic potential. TST may include setting talk (e.g. of weather, current events), talk related to the personal biographies or interests of patient or surgeon (or both), or talk that displays participants’ shared characteristics or history (common acquaintances, similar interests, prior history). Finally, both participants must collaborate to develop the proposed line of talk.3

In Example 1, the transition to TST is initiated by a proffer from the surgeon at line 28. Prior to this, the patient has been describing how she injured her ankle in an incident involving her neighbour’s dog (lines 11–21). Following her description of having ‘landed on the sidewalk’ at line 21, the surgeon asks about the timing of the injury (line 22).

Example 1

An external file that holds a picture, illustration, etc.
Object name is nihms443342f2a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f2b.jpg

11500008, 0:13–0:49

After the patient provides a dating of the incident (line 24), the surgeon produces a transitioning ‘okay’ (Beach 1993) (line 25) and offers an assessment of her x-ray (line 27). This assessment is more positive than the one at line 4 and may represent an optimistic trajectory that is closure relevant; indeed, it is immediately followed by a topic proffer in the form of a question about the patient’s place of work at line 28 (‘you’re over at nickels’), by which the surgeon invites the patient to talk about her workplace in a way that is not relevant to her status as a patient or to the institutional work of the visit. In lacking a tie to the immediately prior talk, this shifting of topic is disjunctive (Jefferson 1984), but orderly, fitting a pattern in which initiation of small talk can be disattentive to a patient’s complaining. The topic is then collaboratively developed: after the patient’s token confirmation at line 29 (‘↑mm: ↑hm’), the surgeon pursues the topic of her workplace with a topicaliser – ‘that’s a busy time of the year for ya’ (line 30) – which elicits further topical talk. The patient acknowledges the surgeon’s claim(line 31) and talk concerning her work and husband’s recent death follows, extending over multiple turns. The sequence of TST ends at line 50 when the surgeon transitions back to institutional talk with ‘WELL [ANYWA:Y’ and a second assessment of the patient’s x-ray.

In contrast to the disjunctive move in Example 1, the topic change from medical work to TST in Example 2 involves a stepwise transition that gradually disengages from the prior talk. After discussion (not shown) of the patient’s x-rays and shoulder problems, the surgeon advances the ‘choices’ for treatment. The surgeon proposes to rule out doing nothing (lines 1–3). Following a silence (line 4), he starts to provide an account for this proposal. As he starts to project what will happen if she tries to ‘git normal’ (lines 5–6), the patient interrupts to state that that’s what she wants to ‘go ba:ck to’ (lines 7–8). This occasions an inference regarding an activity that may be difficult for her (doing her hair) and attempted empathy from the surgeon (lines 12–14, 16).4 The patient produces a disaffiliative ‘nwe:::ll’ in overlap at line 15, and in lines 17–21, a contrast version of her pain – it’s not getting up into that zone when she’s doing her hair, but rather coming down while at work that is bothersome.

Example 2

An external file that holds a picture, illustration, etc.
Object name is nihms443342f3a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f3b.jpg

10110008, 4:43–5:22

The complaint about symptom causation (lines 17–21) is an activity formulation that is inferentially category-bound. That is, hearing the activity formulation, the surgeon proposes relevant membership categories (‘teacher, librarian’, line 22) for her in a way that, although roughly on topic, addresses a matter that is ancillary (Jefferson 1984) to her pain, insofar as her turn was a complaint about pain occurring in the context of her work and not about her occupation as such. The patient self-identifies as a textbook librarian (line 23). This question-answer sequence shifts the interaction away from discussion of her symptoms to her work. Following a pause (line 25), the surgeon asks a question about high school kids and their use of books, and an extended sequence of TST about students and heavy textbooks follows.

The transition to TST in Example 3 is initiated by a proffer from the patient at line 36. Prior to this, the patient and surgeon were discussing arrangements for the patient’s spine surgery. The surgeon requests that, prior to surgery, the patient’s ‘second’ (the individual who will assist him after surgery), be present during a clinic visit to reassure the surgeon of their commitment (‘I wanna make sure your second’s gonna be there’, line 9). The patient is confident she will not ‘throw in the towel’ and mentions her brother as a back up (line 15). This discussion gives rise to a long stretch of mutual laughter, ending with the surgeon’s offer to meet with both potential assistants as they support the patient in making the ‘biggest decision’ of his life (line 29). The patient quietly agrees (line 30), pauses, and then asks about the surgeon’s daughter (line 36), which in this environment is a disattentive topic change (Maynard and Hudak 2008).

Example 3

An external file that holds a picture, illustration, etc.
Object name is nihms443342f4a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f4b.jpg

10280004, 26:15

The surgeon displays appreciation for the patient’s inquiry (lines 38, 40), and what ensues is an extended sequence of TST. The surgeon takes the opportunity to ‘brag’ by showing pictures of his grandchildren, including two who apparently are twins and difficult to tell apart. Again, this sequence is punctuated by mutual laughter as the two discuss strategies to avoid making identification mistakes regarding their grandchildren (lines 67–82). The visit ends shortly thereafter with the surgeon’s request that the patient inform him about a pending urology appointment.

In summary, the features of TST are present in each of Examples 1 to to3:3: TST is a line of topical talk arising from a topic proffer that: (1) is outside the domain of the medical interview; (2) is collaboratively developed by the interactants; and (3) extends for multiple turns.

Distinguishing TST from other types of small talk

Topic proffers that are not collaboratively developed or that are not clearly outside the domain of the medical interview are not TST. Accordingly, sequences of brief or minimal small talk, co-topical talk (combining instrumental and small talk), and sequences of talk addressed solely to third parties are all within the overall domain of small talk but are not TST.

Brief and minimal small talk

Although providing an opportunity for a new line of talk, there is no guarantee that the attempt to generate talk that a topic proffer provides will be taken up. Proposed topics may not be fully topicalised. A participant may respond to an announcement with a news receipt (like ‘oh’, cf. Heritage 1984), that discourages elaboration (Maynard 2003: 100), or may simply answer a question in a ‘type conforming’ (Raymond 2003), or other manner that is also discouraging of development. Or there may be ‘activity-contamination’ (Jefferson and Lee 1981) – for example, in Example 4, a topic proffer about a common acquaintance becomes part of a complaint sequence concerning something relevant to the participant as a patient rather than a discussion about that acquaintance. Thus, topic proffers may contain ‘mentionables’ (Schegloff and Sacks 1973), with independent topic potential but that potential may be forestalled by the pursuit of medical concerns also related to the mentionable.

We distinguished TST from other small talk in the following way. Brief small talk happens when a topic proffer is met with no uptake (i.e. the proffered topic is rejected at first possible chance with a discouraging response). Minimal small talk occurs when a proffer is done across more than one turn but meets with minimal response such that, beyond two to six turns, there is no subsequent collaborative engagement in the topic.5 The key point is that at least one of the participants works to, and succeeds in, discouraging or declining rather than encouraging or accepting the proffer for small talk.

Example 4 is an example of brief small talk; the surgeon’s topic proffer about a potential common acquaintance at line 13 (‘I got a– friend who works out the:re’) does not generate TST.

Example 4

An external file that holds a picture, illustration, etc.
Object name is nihms443342f5a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f5b.jpg

0890004, 0:15–0:51

The patient responds to the surgeon’s topic proffer about a potential common acquaintance with ‘ye:h, hope he isn’t- doctor Br:ant’ (line 14) – a complaint that is relevant to his identify as a patient which rejects the surgeon’s proffer of small talk regarding the friend. This occasions the surgeon’s disconfirmation and request for an account (line 16, embedded with laugh tokens). The surgeon also does not align sympathetically to the complaint but rather only acknowledges it (lines 20, 24), and then re-introduces (i.e. re-proffers) the topic of his friend (lines 26–27). This potential topic is also quickly abandoned once the patient establishes that he does not know the surgeon’s friend (line 32), and the surgeon then initiates the medical interview itself.

This next example is of minimal small talk. This excerpt comes late in the visit after it has been established that the patient plans to go ahead with surgery. The transcript takes up at a point when surgeon and patient are discussing a rehabilitation centre (lines 1–3), and the patient’s performance (lines 5–6):

Example 5

An external file that holds a picture, illustration, etc.
Object name is nihms443342f6a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f6b.jpg

10210004, 24:22–25:14

At lines 9–10, the patient’s wife announces their plans for an upcoming trip. The surgeon’s questions (lines 11, 20), about where they are holidaying and the nature of their accommodation can operate as invitations to the patient or wife to continue the topic. These invitations are answered with short-form replies (lines 13–14), a longer reply (lines 21–22), and then a closure-relevant item (line 25). A question from the wife (lines 28–29), moves the talk back to the prior medically-relevant topic of the surgery date. Thus, rather than occasioning further topical talk, the ‘possible’ topic of Florida is shut down as the patient’s wife works to establish the timing for her husband’s surgery.

Co-topical small talk

In contrast to TST, which is talk that is referentially independent from the institutional identities of the participants (the patient as a patient or surgeon as a surgeon), co-topical talk is instrumentally related to the ongoing medical talk while performing other actions. Nonserious or humorous comments, for example, occasioned by something in the medical work of the visit and yet not strictly instrumental, were common, as were what we labelled bantering and complimenting which invokes physicians’ or patients’ institutional identities.

Co-topical or on-topic utterances may function to proffer a potential new topic (and in this way overlap to some extent with brief and minimal small talk). However, these sequences are distinct from TST because they remain tied to something from the ongoing medical talk – and medically relevant – and do not lead to further mentionables or proceed to independent TST.

In Example 6, the surgeon, after noticing the age of the patient’s crutches (line 20), produces an aphorism (‘they don’t make them like this anymore’, line 24).

Example 6

An external file that holds a picture, illustration, etc.
Object name is nihms443342f7a.jpg
An external file that holds a picture, illustration, etc.
Object name is nihms443342f7b.jpg

10180010 (4:09–5.08)

The patient, at line 25, laughs. The third party proposes to date the crutches (line 26), the patient acknowledges this (line 27), and after a further utterance and laughter (lines 28–34), the surgeon produces a sequence closing item (‘all right’, line 35), and there is no further small talk as the surgeon returns to the medical interview (line 37). Although infrequent, there are examples in our data where co-topical talk evolves into TST, suggesting that co-topical small talk – like joking, teasing or bantering – may be a bid or resource for generating TST, but it is inconsistently taken up.

Summary of phenomenon: topicalised small talk

The features defining the domain of TST draw heavily on notions of topic and topicality. First, in medical interactions, TST is a line of talk that is referentially independent from the patient’s medical condition or the surgeon’s medical work: this is talk consisting of mentionables oriented to an aspect of the personal biography of one (or both) of the interactants or to some neutral topic in the repository of topics available to interactants in any setting. Second, TST is an achievement of both the patient and the surgeon in that the generation and pursuit of topic is a mutual accomplishment (Button and Casey 1985, Schegloff 2007). TST requires that a mentionable within medical work/talk becomes transformed through acceptance of a topic proffer into a line of talk independent of the focal medical business. Simply introducing a mentionable does not mean that it will be pursued or topicalised – it may result in brief and minimal small talk only, or it may be co-topical in the sense of being medically relevant as well as prosocial.

Exploratory analysis: distribution of TST by patient ethnicity and educational attainment

The process of delineating TST was iterative, involving multiple analyses of sequences of broadly-defined small talk. Once specified, we reviewed the entire dataset again to systematically apply the defining features to all potential TST sequences. The authors decided together whether an episode of small talk met the criteria of being topicalised.

By design, 30 of the 59 visits in our sample were with African American patients (15 with high and 15 with low educational attainment) and 29 were with White patients (12 with high and 17 with low educational attainment). TST was present in 15 of 59, or 25.4 per cent, of visits. These 15 visits contained 20 episodes of TST: nine visits included a single episode of TST, four visits included two episodes and one visit included three episodes.

In an exploratory analysis, we considered the distribution of visits with TST by patient ethnicity and educational attainment (Figure 1). A strong asymmetry was apparent: 14 of 15 episodes of TST were with White patients (93.3%). In 30 visits involving African Americans, however, there was only a single episode of TST (included as Example 3).

Figure 1
Percentage of visits with topicalised small talk (TST) by patient ethnicity and educational attainment (N=59)

We were unable to assess whether the presence of TST was related to patient sex6 or racial concordance between the patient and surgeon.7


From a broad collection of talk within patient-surgeon interactions that was in some way off-task in relation to the medical work of the visit, we narrowed our focus to one subgenre of small talk called topicalised small talk (TST). Although our approach to defining episodes of TST that were similar in an identifiable and reproducible way is not the only way a more precise specification of the larger phenomenon of small talk could be approached, we believe it has credibility given the nature of our data: the emphasis on topicality that is central to this particular subgenre was prominent in the data and strongly shaped the features used to define TST. Because these features were developed through working with naturally occurring talk rather than on the basis of prior theoretical conceptions of small talk, they are empirically grounded.

In our approach, TST was distinguished from brief and minimal small talk. At the root of this distinction is a displayed willingness on the part of participants to engage in TST. Why might this matter? Scholars have commented on how small talk ‘oils the social wheels’ of work-related discourse (Holmes 2000: 57), helping to constitute social relationships that can facilitate instrumental tasks. In clinic visits, instrumental tasks that may be facilitated by a social relationship between the surgeon and patient include willingness to disclose relevant health-related information, including potentially relevant aspects of personal biography. Engaging in TST may be important to the extent that it allows patients and surgeons an avenue for a degree of intimacy in relationship (Maynard and Zimmerman 1984: 305, cf. Goffman 1983). Even when invoked as a means for disattending possibly relevant medical issues, TST allows for this action to be accomplished in a way that is still attentive to the patient as a person (i.e. the topic put forth is recipient-oriented, Schegloff 2007: 170).

Our analysis of asymmetry in the distribution of TST by ethnicity was exploratory but found large and clinically meaningful differences. While these findings need to be confirmed in larger prospective studies, the magnitude of the effect is very large and potentially meaningful, leading us to question what is generative of TST, and what (if anything) might make it easier for individuals to engage in TST. Here, we return to Example 3 for closer scrutiny of what might be gleaned as potentially allowing for TST because it is the only instance involving an African American patient. In this visit, TST emerged when the patient asked the surgeon (who was White), about his daughter. Although a question like ‘how’s your daughter?’ exhibits personalised knowledge about the surgeon’s family, it may also be interactionally risky to pose. However, this topic proffer was preceded by banter, much of which was generative of shared laughter (see Lines 15–21, as well as the surgeon’s other-attentive query regarding the presence of two parties for the patient’s surgery at lines 21–29). Our point is that topic proffers that are possibly interactionally risky, but also potentially generative of TST, may require a ‘safe’ or reliable environment for their success. In Example 3, that environment may have been created by: (1) prior knowledge between patient and surgeon; (2) prior joking, and other nonserious comments in which there is mutual laughter or other indications of relative intimacy; and (3) other-attentive talk by the surgeon regarding the circumstances of the patient’s upcoming surgery. Worthy of fuller consideration is whether shared ethnicity by itself can provide a safe or reliable environment for the generation of TST, whereas the absence of shared ethnicity is less reliable for the success of TST and decreases its probability of occurrence.8

Erickson and Schultz’s (1982) work on interactions between high school guidance counsellors and students refers to ‘co-membership’, a concept which may also shed light on what is generative of TST. In counselling interviews, small talk was used to communicate co-membership, defined as an aspect of performed social identity that involves attributes of status shared by the counsellor and student (race, ethnicity, sex, common acquaintances, shared interests, and so on). High co-membership was more common in counsellors and students of the same race than those of different races. In a similiar vein, Maynard and Zimmerman (1984) noted how co-membership can consist of shared category experience, knowledge, or affiliations, and include common interests, acquaintances, activities and so on – broadly speaking, matters that participants may discover as shared and which are potential mentionables. In Example 2, the participants exhibit shared views concerning students and textbooks (lines 26–39), as well as the need for replacement of actual books with electronic copies (not shown). This surgeon and patient also had common acquaintances and knowledge of a local high school environment (where the patient is an employee and the surgeon a healthcare provider for athletes). In other visits, surgeons and patients talk about Tommy Hilfiger socks, the wisdom of riding motorcycles, and views on local theatrical productions, to name a few successful instances of TST. The possibility that physicians and patients can co-produce the visibility of the ‘ways in which we are alike’ through TST suggests a question for future work: whether the sharing of membership in the same racial/ethnic category provides an immediate basis for TST between patients and surgeons that is not there, or requires more work to establish or invoke, when such membership is not shared. This has resonance with the work of Erickson and Schultz (1982), who demonstrated that:

…among interactional partners whose positions in the social world outside the encounter are very dissimiliar (in terms of ethnicity, race, social class, and other kinds of potential comembership categories), it may be necessary to spend more time in small talk to establish comembership, and to do such self-disclosing talk about more universally shared commonalities among people (Erickson and Shultz 1982: 199).

In addition to probing the suggestion of differences in TST among ethnic groups, future research could usefully focus on brief or minimal small talk, and co-topical small talk which are defined here but not subjected to comprehensive analyses.

In sum, specifying the features of topicalised small talk (TST) is a first, foundational step for subsequent analyses of what TST achieves for surgeons and patients in their relationship both socially and instrumentally. Our work on how small talk can exhibit a form of disattentiveness is one example, but further systematic study may reveal TST as a resource for additional social actions, including those related to enhancing relationships between physicians and patients.


This study was supported by Grant #R03 HS015579-01A1 from the Agency for Healthcare Research and Quality, and Grant #R01 AG018781 from the National Institute of Aging. Dr. Hudak is a recipient of a Career Scientist Award from the Ontario Ministry of Health and Long-Term Care.


1We followed the approach to outlining the set of discriminations defining a phenomenon’s domain of occurrence presented in Schegloff (1996).

2Coupland et al.’s (1994: 104) paper on frame negotiation in doctor-elderly patient consultations includes a lengthy extract illustrating doctor’s occasional ‘willingness to pursue non-medical topics’. This excellent example of what we term TST highlights points of overlap between TST and these authors’ discussion of socio-relational framing in patient-physician talk.

3Sequential analysis for excerpt 10110008 is presented in Maynard and Hudak (2008).

4The physician’s statement is what Labov and Fanshel (1977) call a ‘B-event’, an experience that is within the domain of someone other than the speaker and to which the speaker has limited access. We regard his utterance as offering empathy insofar as it is an inference about the patient’s experience using an assessment (‘tough’) formulating the B-event as a challenging one.

5In our data, we have two instances of brief and six of minimal small talk. We settled on the number of turns as being between two and six because none of the instances of minimal small talk was beyond the higher end of this range (beyond six turns, participants were displaying a willingness to engage in the topic rather than to shut it down). Within these turns, whether the response is composed of a single turn constructional unit (TCU) or several brief TCUs, the key is that these are ways of embodying what Schegloff (2007: 171) describes as a minimal response to a topic proffer: ‘A third feature, and often the key one, is whether the response turn is constructed to be minimal (or minimised – i.e. analyzably kept short, even if not as short as possible) or expanded. Here turn organization plays a strategic role; response turns composed of a single TCU, and a brief one, or several brief TCUs (especially if they are redundant or repetitive) are ways of embodying minimal responses. Turns built of more than one TCU, and including one whose grammatical composition is more elaborate or inflected and designed to add to what the prior TCU has done, are ordinarily ways of doing expanded responses’.

6The distribution of TST by patient sex is as follows: six of 15 episodes are with men, nine with women. Overall, 78 per cent of visits are with women (45/59). All surgeons in this sample are men.

7It is reasonable to presume that surgeon race, and racial concordance (Cooper-Patrick et al. 1999) between patient and surgeon may contribute to the presence of TST. However, because we did not select our sample to allow for comparisons between race-concordant and discordant patient-surgeon pairs, the ethnicity of surgeons is reflective of orthopaedic surgeons in the US generally: they are predominantly White and male. Our sample of 39 surgeons included 32 of White descent, two African Americans and five of other ethnicities. Only three of the 30 African American patients (10%) were in racially concordant visits as compared to 24 of 29 White patients (82.8%). Thus, we cannot comment on the incidence of TST in relation to racial concordance between patient and surgeon; what we can report relates to White surgeons only. In our 15 visits with TST, 13 were with White surgeons, none with African American surgeons, and two with surgeons of another ethnicity. Of the 13 episodes of TST with White surgeons, one was with an African American patient and the remaining 12 were with White patients. This asymmetry was not a function of opportunity: overall, the proportions of White surgeons attending only African Americans or only White patients were similar (40.6%and 46.9%, respectively). The number of surgeons attending both African Americans and White patients was small (4 of 32, 12.5%).

8See Rawls’s (2000) discussion about different ‘interaction orders’ for African Americans and White Americans for the generation of casual talk.


  • Agency for Health Care Policy and Research. Trends in Hospital Procedures Performed on Black Patients and White Patients: 1980–1987. Rockville, MD: Agency for Health Care Policy and Research; 1994.
  • Baron JA, Barrett J, Katz JN, Liang MH. Total hip arthroplasty: use and select complication in the US Medicare population. American Journal of Public Health. 1996;86(1):70–2. [PubMed]
  • Beach WA. Transitional regularities for ‘casual’ ‘okay’ usages. Journal of Pragmatics. 1993;19(4):325–52.
  • Button G, Casey N. Topic nomination and topic pursuit. Human Studies. 1985;8(1):3–55.
  • Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al. Race, Gender and Partnership in the Patient-Physician Relationship. Journal of the American Medical Association. 1999;282(6):583–9. [PubMed]
  • Coupland J, editor. Small Talk. Harlow: Pearson Education Limited; 2000.
  • Coupland J, editor. Special issue: small talk: social functions. Research on Language and Social Interaction. 2003;36(1):1–106.
  • Coupland J, Robinson J, Coupland N. Frame negotiation in doctor-elderly patient consultations. Discourse & Society. 1994;5(1):89–124.
  • Erickson F, Schultz J. The Counselor as Gatekeeper: Social Interaction in Interviews. New York: Academic Press; 1982.
  • Gill V. Patient ‘demand’ for medical interventions: exerting pressure for an offer in a primary care clinic visit. Research on Language & Social Interaction. 2005;38(4):451–79.
  • Goffman E. The Interaction Order. American Sociological Review. 1983;48:1–17.
  • Heritage J. A change of state token and aspects of its sequential placement. In: Atkinson JM, Heritage J, editors. Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press; 1984.
  • Heritage J, Maynard DW. Introduction: Analyzing Interaction Between Doctors and Patients in Primary Care Encounters. In: Heritage J, Maynard DW, editors. Communication in Medical Care Interaction Between Primary Care Physicians and Patients. Studies in Interactional Sociolinguistics. 20. Cambridge, UK: Cambridge University Press; 2006.
  • Holmes J. Doing collegiality and keeping control at work: small talk in government departments. In: Coupland J, editor. Small Talk. Harlow, England: Pearson Education Limited; 2000.
  • Jefferson G. On stepwise transition from talk about a trouble to inappropriately next-positioned matters. In: Atkinson JM, Heritage J, editors. Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press; 1984.
  • Jefferson G, Lee JRE. The rejection of advice: managing the problematic convergence of a ‘troubles-telling’ and a ‘service encounter’ Journal of Pragmatics. 1981;5(5):399–422.
  • Labov W, Fanshel D. Therapeutic Discourse: Psychotherapy as Conversation. New York: Academic Press; 1977.
  • Maynard DW. Bad News Good News: Conversational Order in Everyday Talk and Clinical Settings. Chicago: University of Chicago Press; 2003.
  • Maynard DW, Hudak PL. Small talk, high stakes: interactional disattentiveness in the context of prosocial doctor-patient interaction. Language in Society. 2008;37(5):661–88.
  • Maynard DW, Zimmerman DH. Topical talk, ritual and the social organization of relationships. Social Psychology Quarterly. 1984;47(4):301–16.
  • Rawls AW. ‘Race’ as an interaction order phenomenon: W.E.B. Du Bois’s ‘double consciousness’ thesis revisited. Sociological Theory. 2000;18(2):241–74.
  • Raymond G. Grammar and Social Organization: Yes/No Interrogatives and the Structure of Responding. American Sociological Review. 2003;68:939–67.
  • Robinson WP. Language & Social Behaviour. Harmondsworth: Penguin; 1972.
  • Roter DL. The Roter Method of Interaction Process Analysis. Baltimore, MD: John Hopkins University; 2001.
  • Schegloff EJ. Confirming allusions: toward an empirical account of action. American Journal of Sociology. 1996;102(1):161–216.
  • Schegloff EJ. A Primer in Conversation Analysis. Los Angeles, CA: University of California Press; 2007. Sequence Organization in Interaction.
  • Schegloff EA, Sacks H. Opening Up Closing, Semiotica, VIII. 1973;4:289–327.
  • Special Issue – Health Disparities. Journal of General Internal Medicine. 2004;19(2):101–203. [PMC free article] [PubMed]