He deftly and thoroughly summarizes two distinct literatures that deal with the image restoration problem. First is the rhetorical literature on apologia. Second is the functionalist treatment of accounts. After considering comprehensively the work in both areas, he integrates the findings into a general theory of image restoration.

The argument of the book is particularly clear and easy to follow. Indeed, it might well serve as a textbook model for students for planning a thesis and conducting a literature review, laying out an argument, and keeping the reader always positioned as to the point under discussion. This is especially helpful in the face of considerable abstract detail that is inevitable in summarizing such a quantity of work of this type. In his first chapter, Prof. Benoit establishes the problem that "apology/excuse/account behavior is a common feature of human behavior." Two following chapters review the apologia and account literatures respectively. The fourth chapter proposes the general theory and is followed by four brief case study illustrations (Chapters 5-8) and a conclusion. To accomplish all of this in 167 pages Prof. Benoit has obviously produced a focused and efficient work. The reader quickly gains considerable insight into the ubiquity of the attack/defense condition and the various ways of theorizing our responses. I believe that by the end of the book, most readers will find themselves either satisfied that they understand this phenomenon workably well, and move on; or, they will find themselves challenged with both a new appreciation for the vagaries of human ingenuity and motivated to engage the theory and further interrogate its explanations of this part of our communicational complexity. Though I am in the latter group, I believe we should thank Prof. Benoit for producing a work that can do both.

The entire study turns on a central problem which Prof. Benoit states clearly at the beginning: "Human beings frequently must attempt to restore their reputations after alleged or suspected wrong-doing." The problem of reputation, and more particularly its role in discourse, is an ancient one. Readers will not be surprised to find Aristotle and ethos treated here. The problems of image and face, however - terms Prof. Benoit uses interchangeably - seem to me to be both newer and different. I believe they arise from different philosophical and research traditions. Are they psychological; are they sociological? If they have elements of both does that lead to problems of understanding or explaining how they work? Yes to each.

I believe that Prof. Benoit conflates these terms inappropriately. Certainly, given their importance to the argument, I find the absence of attention to distinguishing them troubling. Here's an example of how they seem to slide around: "These attacks on our reputation are serious matters, for our image or reputation is extremely vital to us. Face, image, or reputation not only contributes to a healthy self-image, but it also can create important favorable impressions on others" (p. 2). Here the syntax tells us that he uses these terms interchangeably and synonymously to refer to one construct. Perhaps we could be persuaded that they are functionally the same for purposes of understanding our discourse, but at least that's an argument that should be made.

I admit that I'm trying to listen carefully here to the inflections in these words, but let me just suggest the shape of my concern. Face, a term from Goffman's interactional sociology, seems to be what communication scholars would consider a strongly interpersonal phenomenon, grounded in interactional moments of presence with others. When acquaintances gather later to offer estimations of a person's character (the kind of thing we want to protect), they talk about reputation, not face. It's true that we might want to restore face in the moment, and restore reputation over time and space (President Kennedy has a reputation for us, but not "face"), but are these really equivalents? Would the same strategies and choices apply? Even if we talk about how X "lost face," would that color his/her reputation? One can lose face without being subject to accusations of wrongdoing.

Image, too, seems different. We're more invested in the "face" of our interlocutor. Though others indeed hold an image of you, it is still more an investment of the individual ego, in a way that face is not. When I interact with you, I share responsibility for your "face" in a way that I don't share responsibility for your image (or your reputation). Nor does image seem equal or equivalent to reputation. Image suggests a malleable production, a self-serving projection. Reputation seems more solid, more earned, less subject (thought not totally proof from) manipulation. Your reputation might well be at stake, and operate discursively, long after you, your face, and your image are gone.

Perhaps I'm on shaky ground with these distinctions. Perhaps they'll invite rejoinder and argument. I don't offer these responses to fault Prof. Benoit; I believe he has opened up a fertile area of inquiry. Given more investigation, will these turn out to be identical or different phenomena? Will his theory of image restoration account for our responses in protecting them? Do the same strategic response choices hold across the three? I hope at some point I'll be tuning into volume two to find out more.

Prof. Benoit's theory itself is a model of brevity. He makes a self-conscious attempt to achieve a parsimony of categories. This is especially welcome after reading through the pages of account typologies (which he dutifully and usefully reproduces in tabular form) that become so cumbersome you begin to wonder how useful they can possibly be. Even he seems exasperated at points with the proliferation of distinctions that don't seem to make a difference. At any given opportunity there are countless possible responses to an attack on our reputation. The list of all possibilities is far less useful than our confidence that some are more likely responses than others; indeed, that in these circumstances, responses x, y, and z are most likely of all - and most (or least) likely to succeed.

The theory flows well from the model of the attack/response event that Prof. Benoit begins with. He suggests that we have a limited number of basic postures: we can deny it happened; we can admit it happened but exculpate ourselves from blame; we can admit and be blamed but make the event less offensive; we can take corrective actions; or we can admit guilt and undergo some sort of a Burkean "mortification" ritual to expunge our guilt. Only three of these categories have subdivisions. For example, he offers six general ways we can reduce the offensiveness of the event. He argues, and I agree, that until such time as these categories exhaust their explanatory usefulness we should eschew proliferating ever smaller categorical distinctions.

The utility of the theory depends, as Prof. Benoit makes clear, upon the significance to us of maintaining (hence restoring) a favorable reputation (saving face, protecting our image). The theory also depends, though Prof. Benoit does not pay much attention to it, upon the causal or mediating role that image plays in the outcomes that follow a rhetor's defense. While most would agree that we all have occasion to engage in image restoration from time to time, is this a sufficiently powerful motivation to account for our discourse across all attack/defense situations? Prof. Benoit's examples raise some doubts for me. He frequently refers to President Nixon's famous (and widely studied) "Checkers" speech. It seems to me that more was at stake for Nixon than merely image restoration. There was the threat of illegality in the existence and use of the milk fund. Nixon's very ability to continue as Eisenhower's running mate depended upon getting himself out of that situation. Obviously a positive image would help, but he had to deal with other material, legal issues, regardless of image. And are we confident that the improvement of his image would mediate changes in audience confidence in him or votes for him?

A similarly ambiguous example is the response of Union Carbide to the Bhopal tragedy. Prof. Benoit admits repeatedly that Union Carbide's responses may well have been constrained by advice from their legal counsel about what they could or couldn't publicly admit. Nixon's responses to Watergate needed to keep him out of jail, not just restore his image. Reagan's responses to Iran-Contra arms dealing were similarly implicated with more than just image or reputation. Even though our common sense tells us that we generally prefer a positive reputation (or a positive image) with our peers, how can we be confident that image or reputation concerns are the causal motivators of our responses. And, if we get ourselves off the hook, so to speak, how can we be confident that restoring our image was the causal, or even the mediating factor? Nixon, for example, ran and won with Eisenhower, but didn't necessarily improve his image with a lot of people. Reagan is still quite popular though many continue to believe him guilty of a serious offense in the illegal arms dealing. So, how much confidence do we have in the efficacy of attempts to restore our image?

I believe that further development of Prof. Benoit's theory should work to help us better come to grips with some of these finer points about the underlying model: what role does image or reputation play in causing, influencing, or mediating the outcomes of such rhetorical events. And here, of course we must consider both the audience as well as the rhetor. It seems quite possible that audience members may not care about, or have the same kind of investment in, the restoration of the rhetor's image as the rhetor has. For example, I don't believe that any kind or amount of image restoration work will rehabilitate Reagan's image or reputation in my eyes. This is obviously an ideological position and commitment on my part. How does image restoration theory deal with this (common enough) condition - or can it? More recently, the discourse surrounding Clinton's image or reputation in the Monica Lewinsky scandal generated the paradoxical result of his image or reputation apparently plummeting with people who also simultaneously attacked his attacker and defended Clinton as both an individual and as President. What role did image and/or image restoration play for that audience? Though Prof. Benoit does attempt, where possible, to consult public opinion data to ascertain audience reactions to image restoration discourse, it is not data gathered for this study, nor does it begin to deal with the specific role or status of image/reputation as affected by the discourse. In short, is an "approval rating" a satisfactory measure of image, reputation, or image restoration? And, if approval ratings change, is such change mediated by the change of image perception? Short of significant qualitative audience research, these may be the best measures we have. But Prof. Benoit does not make these arguments and, in general, it is perhaps the linkage of the theory to audience factors that is weakest in this book. In fairness, Prof. Benoit recognizes this and points in his conclusion to the need for further work in this area.

Image restoration discourse treads a significant boundary tension between the good of the individual and the good of the community. Aristotle's recognition of the role of ethos, after all, was focused on the importance of wise deliberation in a just state, not on the satisfaction of the individual rhetor, which the more modern concept of image seems to value. This community focus seems capable of being at odds with the needs or desires of the individual. Presidents like Nixon or Reagan may want to cultivate a positive image to help them win (re)election. Such an outcome might, in their view, be for the good of the community. But how does this theory link image and image restoration with the larger good of the community for those with significant ideological differences? For example, I care less how well Reagan restores his image than how much he does to alleviate poverty. The theory, however, measures success in terms of the individual rhetor, not the good of the community. This became particularly glaring to me in the Union Carbide/Bhopal example. The data for the "public reaction to Union Carbide's image restoration efforts" was all apparently U.S. audience data, not the Bhopal audience. While most measures of response to Union Carbide's restoration discourse were negative, the U.S. audience still held a favorable overall opinion of the company. Prof. Benoit recognizes that the "salience of accusations to the audience is an important factor in image restoration." But, speaking ethically, how much concern should we have for the company's image in the eyes of audiences who did not suffer the tragedy?

Prof. Benoit offers four brief case studies to illustrate application of the theory. Each case study represents a rather large-scale event. First, he looks at the wars between Coke and Pepsi, as conducted in a campaign of dueling advertisements in a trade publication. Second, he examines Exxon's responses to the Valdez oil spill. Union Carbide is the third case, followed by Nixon's proactive speech on the invasion of Cambodia. I believe that, as case studies, these are more suggestive and illustrative than adequate analyses of complex events. Indeed, altogether they take up only about one third of the book. But Prof. Benoit makes no claim here to be exhaustive. He merely wants to show that the theory has some utility for identifying and analyzing the strategies rhetors choose. In this respect they are useful for they show the reader clear instances of the strategies discussed in the book. Of course, as I suggested in the beginning, they will tend either to summarize and conclude the demonstration, or to whet the reader's appetite and suggest all sorts of follow-up questions. These questions will likely be motivated by the reader's knowledge of the historical cases, or investment in the larger issues. In my case, there was too much detail about Coke and Pepsi and Nixon's Cambodia speech and not nearly enough about Exxon Valdez or Union Carbide.

The case studies were interestingly varied. Coke and Pepsi really stretched the theory for me, and Prof. Benoit admits that it is not the most prototypical example. He provides ample detail of the back and forth argument between the two soft drink giants, but it is difficult to frame it as image restoration discourse. Since most of us wouldn't have seen this ad campaign, it is fresh and interesting, but not the strongest example of the theory to lead with. Nixon's Cambodia address is the only single speaker example, the others being agency or corporate discourse. However, Prof. Benoit treats this as a special case of preemptive image restoration. That is, he argues that Nixon was trying to prevent the deterioration of his image in advance of being attacked. Again this stretches the theory more than it illustrates it, especially since the preemptive potentials were not addressed anywhere in the development of the theory. For Exxon and Union Carbide I simply wanted more details and more analysis - they were major, far reaching corporate disasters - but I don't fault these chapters in the context here of illustrating the use of strategies. In fact, their clear focus and brevity would make them very useful in the classroom for illustrating the application of theory to case and for generating useful discussion about the issues raised in the book.

One of Prof. Benoit's conclusions is particularly compelling: we lack, he says, "a thorough understanding of accusation, reproaches, or persuasive attacks." He has recognized throughout the book that defensive image responses presuppose, and are set in motion by, such attacks. Clearly we need to work out the details of who attacks, how, and why and link up those findings with the choice of strategies and the receptions and responses of audiences. Only then will we have a comprehensive theory of the role of discourse that invokes, establishes, maintains, and repairs image and reputation. We owe Prof. Benoit our thanks for putting this challenge on the table in such a coherent and provocative way. I believe the book would be quite useful as a text in upper division courses intended to provoke discussion of practical public communication behavior.