The Theory of Image Repair Discourse and Other Research
This section discusses the relationship of this theory to other work in the literature. It should be obvious that the theory of image repair discourse is heavily indebted to previous work. I have adopted categories directly from the works of the other scholars (e.g., Burke, 1970; Scott & Lyman, 1968; Ware & Linkugel, 1973) who made extremely important contributions. Toner (2009) reviews the literature on apologia, image repair, and reconciliation. This means the theory of image repair discourse is compatible in many regards with much research.
Rosenfield's Theory of Mass Media Apology
Rosenfield's (1968) theory offers four characteristics of mass media apologies, one of which describes the situation (a “short, intense, decisive clash of views,” p. 449), and two of which concern content at a very general level (facts cluster in the middle; reusing arguments). Invective, or attacking one's opponent, is an option for those attempting to repair their reputations. His theory offers an important starting point.
Burke's Theory of Guilt
Burke (1970) discusses only two strategies for reducing guilt: victimage and mortification (Brummett , drawing on Burke, adds transcendence). Although it is clear that these are important strategic options, they simply do not exhaust the possibilities available for image repair. Burke discusses the purgative-guilt cycle, where humans inevitably violate the social order, requiring redemption. Guilt is an important motivation for image repair.
Burke sees an important similarity in the way in which mortification and victimage deal with guilt, symbolically “killing” it. Burke suggests that they are both a form of death: mortification a kind of suicide and victimage a kind of homicide (1970, p. 248). I separate these image repair strategies, however, because of the effects they engender: Mortification accepts the blame (placing it on one's “bad” self) and begs forgiveness, while victimage shifts the blame elsewhere to a scapegoat. Hence I consider victimage, or shifting the blame, as closer in effect to denial than mortification.
Ware and Linkugel's Theory of Apologia
Ware and Linkugel's (1973) approach offers more strategies for image repair (which they term apologia): denial, bolstering, differentiation, and transcendence (these concepts were derived from Abelson, 1959). Still, the strategies of attacking one's opponent, shifting the blame, mortification, minimization, and compensation are not discussed. One difference between Ware and Linkugel's (1973) theory and the theory of image repair lies in the treatment of good intentions. They consider good intentions to be simply a part of denial (p. 276). However, I argue that an apologist who says, “I did not do the bad thing you accuse me of,” is employing a distinctly different approach than one who says, “Yes, I did the bad thing you accuse me of, but I didn't intend for any harm to come from it.” Another difference between the theory of image repair and Ware and Linkugel's theory of apologia concerns their use of postures. Although they admit that more than one strategy may be present, they assert that two will predominate. Perhaps this was true in the speeches they examined to illustrate the initial presentation of their theory, but no conceptual justification was presented for this assertion.
Ware and Linkugel advance four postures—absolutive, vindicative, explanative, and justificative—created by combining either denial or bolstering with either differentiation or transcendence. These postures are puzzling. For example, it is not clear how the source who denies that he or she is associated with an action (or object) that repels the audience would benefit from associating that action with a broader, more positive context (transcendence) or from distinguishing it from other similar but less desirable objects (differentiation). If she or he did not do it, what is the point of differentiating or transcending the action? Yet the vindicative posture, one of the four fundamental apologetic postures, employs denial and transcendence, and the absolutive stance combines denial and differentiation. On the other hand, it seems reasonable to me that an apologist might wish to bolster his or her reputation after denying wrongdoing. However, they identify no stance that relies primarily on denial and bolstering. Therefore the theory of image repair, while adopting Ware and Linkugel's four basic strategies (denial, bolstering, differentiation, and transcendence) as important means for image repair, eschews their four postures.
Ware and Linkugel begin the title of their landmark essay with the words “They spoke in defense of themselves.” Most image repair efforts are performed by the accused. However, as explored in chapter 7, image repair can also originate with other people and organizations besides, or in addition to, the accused. Image repair theory encompasses self-defense but also covers image repair discourse from other sources.
Hearit and Crisis Management by Apology
Hearit (2006; see also 1994; 1995a; 1995b; 1996; 1997; 1999) discussed crisis management, indicating that there are three responses to guilt: denial, shifting blame, and mortification. He talked about legal liability and ethics and crisis communication. Then he investigated three general contexts: individual (politicians, athletes, and celebrities), organizational (retail and manufacturing businesses and nonprofit organizations), and institutional (governments and their agencies, including universities and religious organizations). This work is a useful addition to the literature, but the list of strategic options is incomplete.
Situational Crisis Communication Theory
Coombs's Situational Crisis Communication Theory (SCCT) is an important and popular approach to image repair by organizations. I will describe elements of this theory and then evaluate it.
The Nature of SCCT
Coombs and Holladay (1996; see also Coombs, 1995; 2013) identify four basic crisis types based on attribution theory: accident (unintentional and internal), transgressions (intentional and internal), faux pas (unintentional and external), and terrorism (intentional and external). Coombs explains that “crisis type is the frame that is being used to define the crisis” (2013, p. 264). Five crisis response strategies are identified: denial (labeled nonexistence in Coombs, 1995), distance, ingratiation, mortification, and suffering of the accused (p. 284). Then a matrix is developed relating the first two crisis types (accidents, transgressions) and the dichotomy of a one-time offensive act versus a repeated offensive act with the appropriate crisis response strategy (p. 286; see Table 2.4). Coombs (2012) also observes that this study establishes the recommendation that “as reputational threat increases, crisis teams should use more accommodations” (p. 158). He also offers a much more complex set of response contingencies (1995; 2012, Table 8.3, p. 159).
Another study (Coombs & Holladay, 2002; see also Coombs, 2012) employed factor analysis to identify three crisis types based on perceived responsibility. These crisis types are victim (very little attribution of responsibility to organization), accident (low attribution of responsibility), and preventable (high attribution of responsibility).
Table 2.4. SCCT Crisis Type and Appropriate Crisis Response Strategy
|Crisis type||Crisis response strategy|
|One time, accident||Excuse: no intent to do harm|
|One time, transgression||Remedial|
Coombs (2012, p. 155; see also 2013) also offers a list of crisis response strategies organized into four postures. These are displayed in Table 2.5. This list is consistent with the five response strategies identified in Coombs and Holladay (1996); noticeably absent from both is corrective action, a strategy that was included in Coombs (1998) and Coombs and Holladay (2004).
SCCT (Coombs, 2013, p. 267) also identifies three other situational factors: veracity of evidence a crisis exists (true, ambiguous, false), severity of the damage from the crisis (major or minor), and performance history (prior reputation). He also identifies two potential audiences: “victims of the crisis or nonvictims” (p. 267).
As Coombs and Holladay note, “One consistent theme in communication research is that situations influence the selection of communication strategies (Bitzer, 1968; Black, 1965; Metts & Cupach, 1989; Ware & Linkugel, 1973; Wilson, Cruz, Marshall, & Rao, 1993)” (p. 281). An important advance was the incorporation of attribution theory: Clearly the perceived blame of the accused for the offensive act matters. Further, there is no doubt that the situation in which discourse arises is a very important factor in understanding messages; however, the situation is only one of several factors influencing the production of discourse (see Benoit, 2000b). I am also very sympathetic to the motivation for Situational Crisis Communication Theory, which attempts to determine which communication strategies are most effective in repairing an image. However, I have several fundamental reservations about this approach.
First, SCCT assumes the crisis type can be determined a priori. At times it is easy to identify the crisis type of a given situation. However, our perception of reality, such as whether blame should be internal
Table 2.5. SCCT Postures and Crisis Response Strategies
|Posture||Crisis response strategy|
Attack accuser: confront source who identifi ed crisis; can include threats, such as lawsuits
Denial: claim that no crisis exists
Shift blame: blame person or group outside organization for crisis
Excuse: minimize responsibility
Justification: minimize damage from offense
Compensate: offer gifts to victims
Apologize: take responsibility and ask for forgiveness
Bolstering: remind stakeholders of organization’s past good deeds
Ingratiation: praise stakeholders
Victimage: claim that organization is victim, not offender
to an organization or assigned to an external target, is socially constructed through messages. Persuaders can use persuasive messages to attempt to change the audience's perceptions—exactly what image repair is about. For example, an organization in a situation for which it appears to be responsible (internal attribution) can try, and sometimes succeed, at persuading an audience that it is not responsible (e.g., shifting blame for the offensive act and persuading the audience that the attribution for the offensive act should be external). In 1982, seven people died after taking Tylenol pain reliever because of cyanide in the pills. At first this appeared to be Tylenol's fault (internal attribution). Tylenol's image repair effort successfully denied that it was to blame, shifting blame to an unknown “madman” (external attribution; Benoit & Lindsey, 1987). Does this mean that an effective defense changed the underlying situation here? Furthermore, audiences are not monolithic; for example, some may think that a crisis is internally caused, whereas others may think it is externally caused. The possibility of multiple audiences with different perceptions of the crisis means than in some cases there is no single crisis type, and no single appropriate defensive strategy can be identified in a particular situation. Furthermore, at times multiple accusations or suspicions concerning a scandal threaten an image. If some of the concerns appear internal and some external, does that mean the individual or organization is simultaneously in multiple situations?
Second, SCCT's crisis response suggestions ignore the audience's beliefs and attitudes. Decisions about which strategy to use in an attempt to persuade an audience ought to be based, in large part, on the audience's attitudes (their beliefs and values about the accused and the offensive act). However, SCCT asserts that the situation type dictates the defensive response without considering the audience's beliefs and values—and without considering the fact that different people in the audience, or different audiences, can have different sets of beliefs and values. SSCT acknowledges that crisis communication can address victims or nonvictims, but (1) it is possible that some victims would have different beliefs and values from other victims and (2) the category “nonvictims” covers a wide array of audiences with potentially different attitudes and knowledge about the crisis. Image repair strategies should be selected that are appropriate to the situation but also that have a reasonable chance of persuading the specific audience addressed. Additionally, at times a person or organization faces different audiences. For example, in the Gulf of Mexico oil spill, several audiences were in play, including BP's stockholders, Gulf residents, those who fished the Gulf, businesses that depended on tourists, government regulators, and potential tourists. The likelihood that a single strategy would be most persuasive for all audiences is very low. The audience's attitudes are a vital element in persuasive communication generally and image repair specifically.
Third, consider two organizations (or two people) facing similar accusations: one innocent and one guilty (and, as suggested earlier, the audience may not be able to determine guilt or innocence). SCCT apparently recommends the same strategy for both. Unless the audience is clearly unwilling to listen to the accused, a person (or organization) who is innocent should use denial (using evidence and arguments to support that denial as available). On the other hand, I think an accused who is guilty should not use denial. Ordinarily I would recommend mortification and corrective action for the guilty, although other considerations, such as the risk of lawsuits following an admission of guilt, might suggest a different defensive strategy or strategies (such as bolstering, minimization, transcendence, or defeasibility). It is a mistake to recommend the same image repair strategies for innocent and guilty alike. I want to acknowledge that guilt or innocence is not always clear; for example, at times even those accused and convicted of a crime may believe they are innocent. Nevertheless, sometimes guilt or innocence is obvious (even to the accused), and it is a mistake to advocate the same defense for both.
Fourth, corrective action is an important potential image repair strategy, identified decades ago in political (Benoit, Gullifor, & Panici, 1991) and corporate (Benoit & Brinson, 1994) image repair. However, SSCT does not discuss this option in its current formulation.
Fifth, research on SCCT has found no difference between crisis response strategies. Coombs and Holladay (2008), arguing that apology is not the “best” crisis response strategy (a claim I fully support), conducted a study of a responses to a chemical explosion. An information only control message was less effective than messages with defensive strategies, but no differences were found between compensation, apology, and sympathy messages (see also Coombs and Holladay  for another study that found no main effect for sympathy versus compensation responses). Of course, I would not argue that there are never differences in persuasiveness between image repair strategies, but this evidence makes it clear that it is a mistake to assume that one strategy is best for a given crisis situation type.
A sixth concern is the limited empirical evidence for SCCT's predictions. The study by Coombs and Holladay (1996), which is cited for two claims—that responses are more effective when matched to crisis type and that higher threat situations need more accommodation— had very low reliability (Cronbach's α of .57 for external control and .44 for stability; apparently a “final” measure was created, but I cannot find a report of reliability for that measure) and accordingly does not provide a firm basis for this claim. Furthermore, a study by Clayes, Cauberghe, and Vyncke (2010) found that “matching crisis types and crisis responses does not lead to more positive perception of firm reputation than nonmatches” (p. 261; see also Brown & White, 2011). So questions exist about the empirical support for SCCT's predictions.
Finally, SCCT does not consider the “truth,” by which I mean here what the accused believes is true in a given case. For example, SCCT states that organizations should “use denial strategies in rumor crises” (p. 159) without regard for whether the rumor is true or false. We must realize that the “truth” will not always work: Sometimes an audience is utterly close minded. However, I believe it is immoral to lie to an audience (to make arguments that one believes are false). This makes it wrong to issue a blanket declaration about using denial in image repair. Further, in many cases the truth eventually comes out, which means the accused will still face the original accusation and now face an additional one: lying about it (see, for example, the analysis of Lance Armstrong's image repair in chapter 5). This means in addition to being unethical, using denial when the accused is guilty may well backfire when the truth emerges. The truth, insofar as human beings perceive it, must be considered in all persuasion, including image repair. Coombs notes that one situational factor is veracity of evidence about the crisis, but a discussion of evidence does not adequately consider the consequences for image repair of the fact that some organizations are innocent and others are guilty of an offensive act.
At base, SCCT makes the same basic assumptions about persuasive discourse as Bitzer's “Rhetorical Situation.” Bitzer (1968) declares that the situation “dictates the significant physical and verbal responses” (p. 5) and “prescribes its fitting response” (1968, p. 11). He explicitly rules out other potential influences on rhetorical responses: “The situation controls the rhetorical response . . . Not the rhetor and not persuasive intent, but the situation is the source and ground of rhetorical activity—and, I should add, of rhetorical criticism” (p. 6). SCCT does not explicitly rule out other influences as Bitzer does, but the only factor that it suggests should influence selection of crisis response strategy is crisis situation type. In addition to situation, message development should be guided by other factors (e.g., Benoit, 2000b), including the persuader's purpose (e.g., to clear its name or avoid lawsuits), the nature of the persuader (including credibility), the means at hand (including evidence for a crisis response message), and perceptions of the audience's attitudes (beliefs and values).
The Rhetoric of Atonement
Koesten and Rowland (2004; see also Jerome, 2008; Shepard, 2009) advance a theory of “The Rhetoric of Atonement,” arguing that “recently” the focus of public apologies from political and religious leaders “has shifted away from an emphasis on self-defense toward the theme of atoning for past sins” (p. 68). Atonement differs from image repair in that the former uses an indirect approach to obtain forgiveness, orienting toward long-term rather than short-term image repair. They note that “the rhetoric of atonement functions as a purgativeredemptive device” (p. 69). The authors observe, “Many rhetorical critics . . . (see for instance Ware & Linkugel, 1973; Kruse, 1977; Benoit & Brinson, 1994; Benoit, 1995) have focused on apologetic strategies that in some way deny that wrongdoing has been done or that redefine or transcend that wrongdoing” (p. 70). They list the five main strategies from image repair theory (“denial, evading responsibility, reducing offensiveness, corrective action, and mortification as the primary strategies involved in image restoration,” p. 70) but then conclude, “In each case the focus has been on denial, deflection, or justification to restore a damaged image, as opposed to accepting responsibility as a sinner” (p. 70). Many of those accused of wrongdoing deny, whether innocent or not, but it is simply false that “in each case the focus has been on denial, deflection, or justification . . . as opposed to accepting responsibility as a sinner” (p. 70). This claim completely overlooks mortification, despite having just listed it as one of image repair's five major strategic options. Image repair theory's discussion of mortification references Burke's purgative-guilt cycle. Furthermore, mortification has been used to understand defenses in several cases (e.g., Ronald Reagan in Benoit, Gullifor, & Panici, 1991; Hugh Grant in Benoit, 1997a; Bill Clinton in Blaney & Benoit, 2001; AT&T in Benoit & Brinson, 1994; and Dow Corning in Brinson & Benoit, 1996).
The Rhetoric of Renewal
The “Rhetoric of Renewal” (e.g., Seeger & Griffin-Padgett, 2010; Seeger, Ulmer, Novak, & Sellnow, 2005; Sellnow & Seeger, 2013; Ulmer, Seeger, & Sellnow, 2007) is proposed as an alternative to image repair discourse. Four characteristics make up this kind of response to crisis situations, as Ulmer, Seeger, and Sellnow (2007; see also Seeger, Ulmer, Novak, & Sellnow, 2005) explain:
Post-crisis communication that focuses on renewal is provisional as opposed to strategic . . . [Renewal] discourse is a more natural and immediate response to an event . . . The second characteristic of this framework is that renewal exhibits prospective rather than retrospective communication . . . [R]enewal is concerned with what will happen and how the organization will move forward. The third characteristic of renewal focuses on the ability of the organization to reconstitute itself by capitalizing on the opportunities embedded within the crisis . . . Finally, renewal is a leader-based communication form. (pp. 131–132) The authors contrast the rhetoric of renewal with image repair theory. This approach is somewhat like the rhetoric of atonement, but it focuses on corrective action rather than mortification. Seeger and Padgett (2010) explain that the rhetoric of renewal “seeks to go beyond the parameters of image restoration to address the communication exigencies associated with rebuilding, recovery, and revitalization” (pp. 132–133). Ulmer, Seeger, and Sellnow (2007) contrast renewal with image repair: The latter “focuses on explaining and interpreting what has happened and who is at fault . . . renewal is concerned with what will happen and how the organization will move forward” (p. 132). Similarly, Seeger, Ulmer, Novak, and Sellnow (2005) declare that “traditional understandings of post-crisis discourse . . . focus almost exclusively on strategic portrayals of responsibility, blame, scapegoating, denial of responsibility, justification, and related strategies” (p. 82). Those accused of wrongdoing, in fact, frequently deny responsibility, shift blame, and offer justifications or excuses. However, corrective action is clearly oriented to the future. It is one of the five main strategic options in image repair theory. Furthermore, case studies have investigated situations in which those accused of wrongdoing employ corrective action (e.g., Tylenol in Benoit & Lindsey, 1987; Ronald Reagan in Benoit, Gullifor, & Panici, 1991; AT&T in Benoit & Brinson, 1994; Hugh Grant in Benoit, 1997a; Firestone in Blaney, Benoit, & Brazeal, 2002; Sears in Benoit, 1995b; Dow Corning in Brinson & Benoit, 1996).
The authors note that “natural disasters and crises that are massively destructive often create a context and a space more conducive to renewal” (Ulmer, Seeger, & Sellnow, 2007, p. 133). I think the best way to conceptualize these two theories is to say that they include a common element, corrective action (a defining characteristic of renewal, an option in image repair), but operate in different realms. The rhetoric of renewal is most suitable for responding to natural disasters, such as fires or floods, whereas image repair is most suitable for responding to threats to image from alleged offensive actions. Recall Figure 1.1; crisis communication includes natural disasters and the effects of terrorism, which is not per se image repair, One can be blamed for an inappropriate response to crisis, and President Bush needed to try to repair his image after the federal government's botched response to Hurricane Katrina (Benoit & Henson, 2009), but he cannot be blamed for the hurricane itself. This approach is similar to the concept of “restorative rhetoric” proposed by Griffin-Padgett and Allison (2010), which is related to discourse responding to natural disasters or terrorism.
Accounts and the Theory of Image Repair
Some theories of accounts choose not to include apologies or mortification (e.g., Sykes & Matza, 1957; Scott & Lyman, 1968; Schlenker, 1980; Tedeschi & Reiss, 1981; Semin & Manstead, 1983), preferring to focus on excuses and justifications. Thus the theory of image repair offers a more complete list of strategies available to the actor. However, Goffman (1971) does discuss apologies, and Schonbach (1980) includes concessions. Only Goffman includes compensation (and then as a component of apology), despite the fact that plans for correcting the problem can be a very important rhetorical strategy for image repair that has been unjustly overlooked in the literature (e.g., Benoit & Brinson, 1994; Benoit, Gullifor, & Panici, 1991; Benoit & Lindsey, 1987). Thus although some discussions of accounts mentioned here are more detailed than the theory of image repair, it includes more general options. Although a matter of preference, I choose to focus at higher levels of abstraction rather than on details of subcategories.
Many of those who write about accounts have developed much more extensive lists of image repair strategies than have rhetorical theorists and critics. Although this work may include illustrations, unlike rhetorical criticism their focus tends not to be on the discussion of particular rhetorical artifacts. These writers often incorporate previous lists in their new work, adding further refinements in the form of additional categories and subcategories. For example, Scott and Lyman (1968) discussed defeasibility, which concerns lack of knowledge or will, as a possible accounting strategy. Tedeschi and Reiss (1981, p. 282) break this category down into failure to foresee consequences (with eight subcategories, including both “mistake” and “inadvertancy”) and lack of volition (including four physical and six psychological varieties). A limitation of these lists is that their complexity renders them unwieldy. It would be fruitless to deny, for example, that a source can develop defeasibility as an account in a variety of ways (just as a source can bolster in many different ways), but it seems preferable to me to simply group these variants of defeasibility together, rather than list drugs, alcohol, illness, and so forth as separate subcategories. Unless we have evidence that, say, drugs is a more readily acceptable excuse than alcohol, how important is it to devote separate categories to these variants of defeasibility? Similarly, Schonbach (1980) distinguishes between accounts based on past restitution or compensation and those based on future restitution or compensation. It is not clear what advantage is gained from this distinction. We could also divide such offers in other ways (e.g., compensation worth less, the same, or more than the injury), but these sorts of choices can add needless intricacy. If we had evidence that, for example, future compensation was more (or less) persuasive than past compensation, this could be a useful distinction. However, it is not clear how the lists of image repair strategies benefit from some of the fine nuances of these lists of accounts.
Of course, those who desire extremely detailed lists of these strategies can consult Schonbach (1990), who lists almost 150 categories and subcategories. There is certainly a place for such exhaustive analyses. However, I find it more useful to list image repair strategies at a higher level of abstraction. Taking this approach results in a list of image repair strategies that is exhaustive at a more general level and is arguably easier to conceptualize. It is clearly a matter of preference and convenience rather than a theoretical or a conceptual decision, though.
Benoit et al. (2014) report a meta-analysis of the effects of accounts on perceptions of the offender. Providing an account improves perception of the communicator across situation (organizational, interpersonal, accident victims, trials, and church leaders). Accounts help with violations of high, medium, and low severity. Accounts reduced punishments recommended for the offender, improved communicator credibility, created impressions of remorse, reduced perceived blame, and increased perceived morality of the offender. Accounts helped across type of violation (relational/personal, role/organizational, and public violations). The data show that image repair messages are capable of persuading the audience.
This chapter articulates the theory of image repair discourse. Two key assumptions are outlined (communication is goal-driven; identity maintenance is a key goal of communication), and a list of strategies for repairing a damaged reputation is developed. An analysis of attacks as comprising blame and offensiveness is advanced to explain how strategies function. The relationship of this analysis to Fishbein and Ajzen's theory of reasoned action (2010) explains that blame is a belief and offensiveness is a value. This analysis also explains why the audience, and the beliefs and values held by the audience, are vital to effective image repair. Image repair theory stresses the importance of perceptions: The accused's perceptions of audience attitudes motivate image repair and help develop image repair messages; the audience's perceptions determine how well the image repair effort works. Stein's concept of antapologia extends our understanding of this process; it is too limiting to assume that image repair always consists of a single attack followed by a single defense. This chapter ends by comparing image repair theory with other approaches found in the literature.
It is important to acknowledge that this theory has a limited domain. It does not address related questions, such as the initial development of a positive image or reputation. Similarly, while acknowledging the importance of persuasive attack to understanding image repair (and discussing the basic elements of an attack), it does not develop a typology of attacking strategies. Nor does image repair theory attempt to address other forms of crisis beyond threats to image, such as responses to terrorism or natural disasters. The purpose of this theory is to understand how messages can respond to accusations or suspicions of wrongdoing.