This is a study of the interface between power and politeness as manifested in verbal disagreements. It begins with a number of chapters devoted to a discussion of the key concepts that define the author's theoretical conceptualization of social interaction. These concepts are power, communication, relational work, and politeness. The author then applies the model to four (transcribed) interactions. The first, and most extensively analyzed, is a dinner conversation among friends and acquaintances. The second is an organizational meeting in a physics laboratory. The final two come from official transcripts of a radio interview of President Clinton and Supreme Court proceedings relating to the vote count in the U.S presidential elections of 2000. Common to them all, in the author's view, are instances of disagreement.

Theoretically, interactants are understood to engage in two broad functional types of discourse: transactional and interactional. The former is defined as “optimally efficient transmission of information,” and the latter aims at the “establishment and maintenance of social relationships” (pp. 50–51). These discourse types are understood to overlap functionally, and their overall meanings are interpreted through a process of “contextualization.” Committed in general to a theory of the culturally constructed nature of social “reality” (45), the author embeds interactants in a complex web of sociocultural constructs. She thus locates the interactants' contextualization efforts at the intersection of event frames and norms. These are understood to incorporate knowledge of the “what, where and when” of event types as well as cultural knowledge, both conscious and unconscious, about social identities, gender, speaking styles, history, and the unfolding interactional context itself (57). Contextualization thus takes place when interactants judge behavior against this cultural world of frames and norms and classify it as particular kinds of events that are, more broadly speaking, negative or positive, marked or unmarked, and appropriate or inappropriate (48).

In addition to this general list of contextualizing influences, the author also, and somewhat problematically, relies on more familiar academic ideological constructs. In the constant struggle between providing information and negotiating relationships, individuals attempt to protect whatever social “face” they may take on in a particular interaction by attempting to balance two competing functions: involvement and independence. Though interestingly distinct from some prior formulations of this now classic academic take on social interaction, the basic model here has not changed. Individuals must mediate their “efficient” transactional flow of information against the kinds of challenges that such statements can present to one's own face or that of one's interactants. Such “face-threatening acts” are thus to be found in the form of linguistic expressions of disagreement that further index higher-order categories of “politeness” and “power.”

In closing in on the target phenomena, Locher puts much effort into arriving at working definitions of power and politeness. She draws on an impressive array of social scientific approaches to the concept of power. In the end, she concludes with a series of principles about the nature of power in social interaction. Some of these are very general. For example, she argues that power is relational, dynamic, and contestable, and can be expressed through language (39). The three specific principles that have the potential to be applied to actual instances of verbal interaction are that freedom of action is needed to exercise power; that the restriction of an interactant's “action-environment” often leads to the exercise of power; and that the exercise of power involves a latent conflict and clash of interests that can often be obscured because of society's ideologies (40).

Politeness is theorized as “marked relational work” in terms of the concept of interactional face. Interactants mediate their face concerns based on the often competing functional pulls of involvement and independence. Locher calls this “relational work” (57). A polite utterance is thus understood at the theoretical level to be a “speaker's intended, marked and appropriate behavior which displays face concern” and “the motivation for it lies in the possibly, but not necessarily, egocentric desire of the speaker to show positive concern for the addressee and/or to respect the addressees' and the speaker's own need for independence” (91). Politeness is thus a linguistic strategy for mediating between the opposing functional forces of involvement and independence. It is what happens when one's utterances are called back to a “social equilibrium” between these two forces (4, 99–100). It is marked linguistic usage because it still reflects a lack of “just saying” what we mean (and being “direct”).

The abstract form of Locher's theoretical constructs makes operationalizing them for use in analyzing particular transcripts quite challenging. The first event studied, a dinner conversation among friends and acquaintances, explores power by beginning with word counts. Following this section, about which she is herself a bit apologetic, the author turns to a detailed empirical exploration of particular “strategies” found in the transcripts. Though the link to the earlier theoretical work is unclear, she reviews and documents many interesting types of linguistic use: hedges, giving personal or emotional reasons for disagreeing, modal auxiliaries, shifting responsibility, objections in the form of a question, the use of but, repetition of an utterance by a next or same speaker, and unmitigated disagreement. Her empirical work here allows her even to capture some of the specific differences in “verbal style” among the interactants.

With the focus of this part of her analysis on the expression of disagreement, Locher readily admits that power and politeness—her theoretical foci – are considered only “marginally” (113). It is here, though, that we find a central problem with this study. What are these “strategies” expressing, if not power and politeness? And, if they are, shouldn't we be able to see the link between the theory and the empirical data much more clearly? Indeed, perhaps the most troubling fact is that although the examples are linked to a full transcript of the interaction in the book's appendices, the empirical counts – upon which all general conclusions are based – can't be traced in the same way. The all-important contextualization of these strategies as being related to power or politeness thus can't really be cross-checked. Neither are they correlated with self-reports from the interactants (or any other data from native speakers), nor can they be interpretively checked in each particular context. The theoretical link here is simply to translate these strategies into the second-order theoretical language of “mitigating” vs. “non-mitigating” disagreement strategies (143).

The rest of the book focuses on “close readings” of example selections from all of the focal interactions. It is here that the disagreements are to be linked explicitly to power and politeness. Here, however, is where one finds a gap between the empirical work and the theory. Two things occur in these chapters, and they repeat consistently for each of the interactions that are reviewed here. First, analytical examples of disagreements are provided by paying close attention to their linguistic realization (as FTAs of various degrees). These are then translated (problematically) into the theoretical language of mitigated and unmitigated disagreement strategies. Second, the analysis of each interaction closes by reviewing the ways in which the examples generally relate to the principles that were used to define “power” and “politeness.”

With regard to the first, the problem is that quite often the “contextualization” of the examples given easily lends itself to different “close readings.” One example will have to suffice here (163). Certain lines are identified as FTAs that are said to show mostly unmitigated, direct disagreements. It becomes clear later in this segment, however, that they were intended to be humorous. The point though is not what “really” happened but the fact that these “readings,” as the author explicitly calls them, are problematic ones. They are thus even more problematic when they are read through the frame of the larger theoretical model that the author built in the first part of the book. Second, for each interaction, the conclusion that links how politeness and power are interconnected in disagreements comes only in the form of a review of the basic theoretical principles. As these are principles that constitute either indisputable opening assumptions (e.g., power is relative and negotiable) or definitions that are empirically indeterminate (e.g., power requires freedom of action and restricts the environment of the other), the gap between the data and the theory is again visible.

Thus, this book is a thought-provoking but frustrating contribution to the literature on power and politeness. There is, on one hand, much to commend here. There is an attempt to study the empirical realization of interactional categories such as disagreement, power, and politeness. The types of interactions that are selected for study reflect wise methodological choices. The transcriptions are detailed and present material that would be of interest to any student of social interaction. The theoretical reviews of key concepts, such as power and politeness, are thought-provoking and thorough. The subtle theoretical revisions of many theorists working in these areas are important ones. Accordingly, the principles that Locher arrives at are typically important opening assumptions for this kind of linguistic work. Most important, however, is her excellent theoretical understanding of the concept of contextualization (49). She also recognizes how important the process of contextualization is in the emergence of meaning in contexts of social interaction, and thus to the kind of empirical project that she presents. The frustrating irony is that this theoretical understanding of contextualization does not inform the actual analyses of her data. Although she reports that meaning cannot be contextualized without reference to cultural, subcultural, and even individual differences, she goes on to apply a series of academic ideologies to the framing of all her readings. She fails to acknowledge the ways in which they are themselves ideologies that play a reflexive role in her own particular readings of these interactions. With power defined in ways that are essentially always true to a degree – for example, one's actions limiting the freedom of the others – the foci of the study, politeness and power, are only “found” in the ideological terms of the theory itself. There is no way to link the actual utterances to arguments about the regular ways in which linguistic signs in conjunction with native ideological constructions index functional categories like “power,” “politeness,” or even “disagreement” into relative social existence. Instead, we are presented with a “reasonable” ideological construction of these phenomena, but not a reflexive investigation of them (cf. Glick 1996). We are going to have to move beyond analytical strategies that seek to document linguistic functions as if they emerged from clear “models.” We are going to have to allow a more reflexive approach that includes native ideologies in the relative constitution of these kinds of metapragmatic “ways of speaking” (cf. Silverstein & Urban 1996 for examples). This, I believe, will force us to ask what we are trying to do. Is the goal of our analysis to model what we think politeness is? Is it to predict what others think it is? Why? However we answer these questions, unless we change the way we think about such linguistic phenomena, I fear that we are not going to stop seeing ourselves in our theory, and thus in our data.