Focus group research has a bad reputation among many sociologists. As a method that has been particularly successful in the Anglo-American world of marketing and business research – “serious” academics tend to turn up their noses at this method. The results of focus group research often seem too predictable and too easily manipulated too be in accordance with rigorous science. Why, we could ask, should anyone believe that the things uttered in such an artificial situation as the focus group counted as social data and can be regarded as an indicator of that ephemeral and spectral thing that we have grown up to know under the name of “public opinion”?
Greg Myers book is driven in large part by this methodological scepticism that is characteristic of ethnomethodological research. He turns the readers’ attention to the notion of “opinion” and tries to develop an argument that goes beyond the dichotomy of realism and scepticism: Proponents of focus groups praise the method for simulating a more “natural” speech situation than the individual interview. They are “realists” in this sense. They attempt to show that groups of people really have “opinions” – even if they are only “social facts” – and these opinions have some degree of exteriority and objectivity. A sceptic, on the other hand, might argue that opinions are always situated, socially constructed and, ultimately, tell us more about the focus group’s moderator and sponsors than about something of such dubious existence as an “opinion”.
Myers stance – and this is again “typically ethnomethodological” – is a reflexive one. Opinions, Myers suggests, are neither fully real, nor fully unreal. They are paradoxical things:
"We cherish our own opinions, but we can also dismiss opinions as a poor substitute for facts. Opinions are meant for public discussion, but are also private, individual, protected. Opinions are personal, but shared with a group. Opinions display one point of view, but the same speaker can express two contradictory opinions. Expressions of opinions are assumed to be ephemeral, but are also part of the on-going structure of society."
These paradoxes might explain why “opinions” lend themselves more easily (or intuitively) to ethnomethodological investigation than other objects. (Let’s remember how much uproar it caused when Latour (1986) started reminding people that the word “fact” derives from “facere”, i.e. to make or do.) Of course, readers of both Luhmann (1987) and Alice in Wonderland will be little surprised about the fact that paradoxes can be immensely productive. Myers initial intuition is, therefore, absolutely correct when he opens up the concept of opinion and follows it down the rabbit hole. Unfortunately, not unlike Alice, his engagement with opinions occasionally strolls off a little making the book and insightful and entertaining read but leaving much to be desired in terms of the consistency of its argument.
After having discussed some versions of the paradoxes of opinions in chapter 1, Myers describes an inventory for analysing talk-as-interaction that will be familiar to most interpretative researches (turn-taking, member categories, participant roles, discourse markers, rhetoric, transcription). In chapter 3 and 4, he returns to the notion of opinions and enlarges upon their contextual and institutional framing. Following Goffman, Myers points out that during focus groups participants constantly deal with the question “What is it that’s going on here? (Goffman 1986: 8)”. They need cues to find an – at least provisional – definition of the situation so they can model their contributions accordingly. But focus groups are not a very common situation for most people and different cues evoke different situations that blend into each other (e.g. raised hands invoke a classroom, while signalling agreement to close a topic invokes a meeting of some sort). This ambiguity can be productive if it prevents any specific situational definition to prevail. It can provoke many modes of contextualizing and avoids the discussion from becoming a too closely-bounded event. In chapter 4, Myers reformulates the above mentioned paradox of the opinion and marks off his approach from the “cognitive view of opinion” which conceives of opinions as distinct quantifiable entities. Instead, he refers to Blumers classic essay (1948) to describes the “social view” which describes public opinion as interactional and institutional. At this point, however, it becomes already less clear, to what degree Myers can resolve the paradoxes of opinions or indeed, whether he even tries to. It seems to me, that the tensions between the interactional and the institutional aspects of opinions only take the form of a restatement of the original paradox. And while Myers then manages to flesh out significant empirical detail by following this tension, the way his analysis relates to interactional and institutional aspects of opinions leaves some promises unfulfilled (as I will argue below).
Chapters 5-8 include the bulk of Myers’ empirical work and as such cannot be reproduced in detail here. Chapter 5 deals with the ways topics structure the interaction of the focus group, how they can be opened, acknowledged, interpreted, rejected, changed, closed or reopened. Chapter 6 is making interesting points about the ways dis-/agreement is managed in a focus group. As in natural talk more generally, disagreements tend to be dispreferred in focus groups (Sacks 1987: 57). The moderator (or facilitator) plays a crucial role in encouraging disagreement and extending the “range” of contributions (Merton et. al. 1956/1990). Myers outlines a series of techniques to make disagreement more likely: using a weak agreement first, using the moderator as a buffer, constructing others and then attributing one’s opinion to them. He makes the general point that opinions are often imagined in opposing pairs. As such, their empirical occurrence strongly depends on the moderator, on the flow or sequence of the utterances and interactional differences between groups. Chapter 7 deals with the phenomenon of reported speech. Disagreement was one example in which real or fictional others can be used to set up an argument (“Some people might find this offensive”). Reported speech in this case has the function of rendering it less serious – Myers refers to this as “detachment”. Reported speech can equally have the function to render talk more persuasive by invoking direct experience (“And I said: ‘Hell no!’”). Myers points out that many things said in focus groups cannot unambiguously be attributed to individuals as their opinion – as something they can be said to “have”. These shifts must be seen from the perspective of interactional shifts within the focus group setting. Finally, chapter 8 brings up the topic of expertise which in many cases occurs as reported speech, too (“they say”). Of the four, this chapter is the least convincing and manages to only hint at the many and confusing ways expertise is invoked to manage interactional challenges. Myers suggests to regard invocations of expertise as “claims to entitlement” (177). It represents a shift in “footing” (Goffman 1981: 128) that can reference forms of knowledge and experience that have their origin from outside the focus group.
Chapters 9 and 10 are excurses into radio phone-ins and vox pop television interviews. While interesting in themselves they constitute somewhat of a distraction from the main body of the book. It is my opinion that its overall purpose could have been achieved more elegantly without them. Leaving the empirical detail of these chapters aside, their main impetus is to ask how such displays of opinion survive even though they can in no way be said to represent “public opinion” adequately. Vox pops, Myers suggests, might survive “because of the concreteness and particularity of the faces, places, and even the little hesitations and quirks” (222). Similarly, phone-ins incorporate “all the conventions and intimacy we associate with the phone” (201) and thereby create a direct, intimate link between the listener – typically a driver in her car on her own – and the person calling in.
The pleasure of opinions
These excurses help one of Myers’ main hypotheses to which we shall turn now. Phone-ins, vox pops and focus groups work, according to Myers, because people derive a pleasure from sociable argument.: “The expression of opinions, the discovery of how one’s opinions dovetail with others (or doesn’t) in a complex on-going interaction, is part of the excitement of being with other people, and the pleasure of talk” (234). These “social games” (Simmel 1949) constitute a special interactional quality. It is unclear what Myers tries to argue by referring to the “affective” quality of exchanges of opinion and it is very unsatisfying as a result of this book for a series of reasons. Firstly, the purpose of focus groups is not the entertainment of its participants. Keeping up their spirits is certainly a necessary tool for an effective focus group but it is not their institutional “goal orientations” (cf. Drew/Heritage 1992: 22) – whereas we might argue that that is exactly the case with vox pops and radio phone-ins. A clearer differentiation between the ways interactional pleasure is institutionally contextualized would have made a clearer case of what this argument adds to our understanding of opinions (there is a possible book to be written here about the use of interactional playfulness for institutional purposes – but it was not this book). Secondly, this argument is strictly speaking not a “finding”. It does not relate to the methods used in any meaningful way and could have been made with similar force by anyone unacquainted with ethnomethodology that knows focus groups, vox pops or phone-ins. In this sense, the argument distracts from the empirical work rather than referencing it.
The “social game”-argument points to a deeper problem of this book. Myers uses the often-mentioned paradoxes of opinion to set up his book. Unfortunately, he ends up getting lost in their maze. Not entirely unlike other CA scholars who praise the special or moral character of face-to-face interactions (e.g. Anne Warfield Rawls 2010), Myers emphasises the pleasantness of exchanging opinions in order to keep up interactional sociability. Interaction thereby becomes unlocked from its institutional context and becomes a “social thing-in-itself”. This tendency is in keeping with the general sense one gets in reading large parts of his book. The interactional achievement of expressing an opinion leaves the concept itself largely untouched. It becomes thereby tempting to read this book as kind of CA-inspired guidebook to focus groups. Indeed, many classics of the trade give advice that sounds not dissimilar to what Myers tells his reader (especially Merton et. Al. 1956). This was clearly not the author’s attention but it casts doubt on the book and on the CA method more generally. Simultaneously, there is an enduring ambiguity running through chapters 5-8 as it is sometimes less clear whether Myers is discussing opinions in focus groups or opinions per se. At many points, it sounds like he is doing the latter. Paradoxically, the focus group setting then takes on the same lab-like naturalising function for his research that it does in original focus group research.
Finally, Myers manages to resolve the paradox and achieve a kind of narrative closure for his book. But the tension within the concept of opinion, I would argue, is not there to be resolved – at least not so easily. In a much too short concluding chapter, Myers comes back to the point that opinions are simultaneously mediated and situated. (the book confirms my more general suspicion that ethnomethodology and systems theory can only really survive if they finally bury their grudges and scholars manage to weave them together in empirical research). He goes into four defining features of opinions: they are interactional, packaged, mediated and intertextual. But his hints at what this means concretely do not go far and the connection to his empirical material is loose at best. If Myers had followed up on his promise to take up a “social view” that would follow the construction of opinions from interaction through to different institutional settings and discursive practices, this could have been a much stronger book. Unfortunately, he stays too close to the conversational data to really achieve this. It is questionable, whether CA on its own is really the right method to approach questions that are trans-situational, institutional, mediated and intertextual (as Yuqiong Zhou (2006) has also pointed out). My strong inclination is to say probably not. As a result, Myers achieves a lot of things in his books. He lays out a convincing and thoughtful way to talk about the interactional qualities of opinions. He develops a conversation analytic perspective on the phenomenon of focus groups. He compares different formats in which opinions are constructed. But Myers cannot deliver all the things he sets out to do initially. Especially the ways in which the interactional aspects of opinion construction links up to institutional contexts and become points of reference that can travel across situations remain unexplored. This leads to some considerable ambiguity which undermines the overall methodological implications of the argument. This being said, these tensions give the reader ample opportunity to contemplate the extent of micro-sociological research, which – combined with its empirical insight – makes Matters of Opinion a more than worthwhile read for anyone interested in conversation analysis, public opinion research or political communication.