Drawing on extant research in conversation analysis (CA) and other fields, Glenn's entry in the “Studies in Interactional Sociolinguistics” series is both an excellent introduction to CA for those outside the field and an interesting exploration of the social phenomenon of laughter. The first chapter gives a general survey of the research, which has often focused on the physicality of laughter and its relation to humor. Ultimately, however, the chapter suggests a function of laughter beyond expressing amusement: affiliation with co-participants in social interaction. In chap. 2, Glenn provides a helpful outline of CA. He emphasizes CA's focus on participants' perspectives and provides a cogent explanation for CA's meticulous transcription conventions with the example of silences, marked on CA transcripts in tenths of seconds. To leave them out, he argues, would be to assume from the beginning that silences in talk mean nothing (p. 37) – an idea that most competent talkers, let alone CA researchers, would reject out of hand.
This meticulousness extends to the transcription of laughter in CA. Gail Jefferson's early transcriptions included orthographic renderings of laughter, which allowed analysts to see exactly how participants embed laughter tokens in talk and thus mark utterances as “laughable.” Glenn is careful to point out that “laughable” is a retroactive term; it is only from the laughter in a sequence that the laughable can be identified. At times the referent is not clear to participants, who may demand clarification from laughing co-participants. Glenn also describes his own data in this chapter, most of which are audio recordings of telephone conversations, as is common in CA data. While the observer's paradox defined by William Labov may minimally affect recorded conversations, Glenn suggests that most social laughter is already affected by the observer's paradox insofar as people are sensitive to the contexts in which they laugh (41). Participants in Glenn's data were not given subjects to talk about or told that laughter was going to be the subject of study.
Chap. 3 addresses shared laughter, drawing heavily on Jefferson's study (1979) of laughter invitations and acceptances/declinations. The recipient of such an invitation may laugh along, remain silent, or engage in serious talk; only the last definitely rejects the invitation, since speakers may pursue a laughter response in the face of silence. In cases where laughter invites laughter in response (i.e., not in troubles talk or self-deprecation), participants may actively work to extend shared laughter once it has begun in order to display affiliation more strongly. Glenn identifies three devices for this: extended laughter, repetition of the original laughable, and the production of new laughables thematically linked to the original (82). Extended laughter and repetitions can only go so far, but the production of new laughables by participants may be drawn out at some length.
Although laughables may be ambiguous and identified only retroactively, chap. 4 addresses the more concrete question of who laughs first. Glenn briefly returns to review CA by describing the turn-taking system in conversation and the effects of group size on that system. He suggests that a general participant bias against self-praise, described in early CA research (Pomerantz 1978), also constrains “laughing at one's own jokes.” In multi-party interactions where there are several possible laughers, therefore, speakers generally do not laugh first. In two-party interactions, however, he claims that the laughable-producing speaker must laugh first if shared laughter, and the affiliation it promotes, is to occur (101). The ambiguity of some potential laughables makes the possibility that the recipient will “miss the joke” too great. Thus “speaker laughs first” is common in two-party interactions but not multi-party ones. First laughter in Glenn's data works, among other things, to mark previous utterances as non-serious, to show a willingness to laugh at/tease oneself, and to disambiguate problematic or marginal laughables. A minor point that might have produced further interesting results is the role of physical cues in these interactions. Glenn does earlier describe a piece of video data in which smiling indicates receptiveness to joking/laughter (71), and he notes that the majority of his two-party interactions were telephone conversations, while the multi-party interactions occurred mainly face-to-face, but he does not explore the larger role of the body in face-to-face instances of shared laughter.
In chap. 5 Glenn distinguishes between affiliative and disaffiliative laughter. He identifies four keys to distinguishing “laughing at”: Someone co-present is identified as the laughable; someone other than the butt laughs first; second laugh is produced by another co-participant, not the butt – in two-party interactions, there is no second laugh; and subsequent talk references “laughing at.” Both affiliative and disaffiliative laughter may be converted to their opposites. Jokes, which often function as “understanding tests,” may lead to laughter at a participant who doesn't get them, or at an inadequate joke teller. In cases where a co-present person has been made the butt of laughter, she may transform “laughing at” to “laughing with” by participating in subsequent laughable production.
In chap. 6 Glenn addresses the use of laughter in both “laughing along” and resisting teasing to constitute relationships and identities. He cites earlier research on laughter, particularly Jefferson, Sacks & Schegloff 1987, that identifies laughter as a midpoint between disaffiliation and explicit affiliation in response to “improprieties” in interaction. Improprieties, like teasing, may create and display intimacy between participants engaging in a common form of moderately risky play. While laughter signals some appreciation of such utterances, however, it does not in itself agree with them. Accompanying utterances are necessary to clarify any particular laugh. Glenn presents two interesting cases of teasing interactions in which female laughter resists sexual teases from male co-participants. One case involves a long sequence of wordplay in which both participants humorously produce incorrect verb tenses. The wordplay turns sexual with the use of the word come; the female participant laughs but follows up with a non-sexual continuation of grammatical play despite the sexualized referencing in the previous utterance (131–41).
The second case is taken from the popular National Public Radio program Car talk; a female caller's name is marked as interesting in the opening of the call. She claims her (unspoken) last name is “even wilder”; one of the male hosts then produces a sexual joke: “[the caller] is even wilder than the last girl I went out with” (145). His co-host produces the first laugh and is joined by both the original speaker and the caller. The caller ends her laughter first, however, and then produces well in a tone that suggests mock indignation. Early CA research showed that well often prefaces disagreements or dispreferred utterances; together with her tone and cessation of laughter, the female caller effectively closes the sexual joke. The hosts retreat to non-sexual wordplay on her name. Accordingly, the caller is able to acknowledge the humorous intent of the sexual joke without affiliating with it, and to initiate movement away from potential offense.
Glenn discusses Jefferson's work on how men and women laugh differently; she found a “tentative pattern” with women tending to laugh more affiliatively and men more disaffiliatively (154–55). Glenn suggests that male and female laughter differs in courtship-relevant and non-courtship-relevant interactions. Men tend to respond more to women's laughter invitations in non-courtship-relevant interactions; in courtship-relevant interactions, however, women respond more often to men's laughter invitations. Laughter that does not invite further laughter in response (in troubles-telling, etc.) is more common in courtship-relevant interactions than others, is more often produced by women, and is more often at the woman's “expense” (157). Despite these asymmetries, and although gender is an interesting and often powerful construct, we do not “live our communicative lives in the aggregate” (158) – in any single interaction, gender may be more or less relevant to the participants themselves. This work answers a call for more study of talk-in-interpersonal relationships without assuming that variables like gender are omnirelevant. Although Glenn concedes that talk and relationships are different things, he asserts that talk is essential to the constitution of relationships. Laughter plays a large part in characterizing relationships; its ambiguity allows participants to explore relationship possibilities without fully committing to them. Future research, he suggests, might focus on laughter's role in the “social constructions of power in relationships,” particularly in enforcing dominance (169).
Glenn provides thorough background for his own work on laughter, drawing largely from CA sources, but the book offers much more than a summary. Early CA research did address laughter in “general conversation,” but more recent work has mostly addressed it in specific institutional settings (Haakana 2001, Lavin & Maynard 2001). The more we know about “free” laughter in relatively unconstrained naturalistic interactions, the better we can understand how participants use the resource of laughter in settings where it may be more constrained. In this light, Glenn's book is an important contribution. His use of CA highlights the importance of context and sequential placement; it also foregrounds the orientations of participants themselves and presents a compelling argument for increased sensitivity to what individuals do in particular interactions.
