Innuendo and Damage to Reputations

Daniel H. Wegner, Trinity University
ABSTRACT - An innuendo is a statement about something combined with a qualifier about the statement. "Brand X is not dangerous," therefore, is an innuendo; it can be decomposed into a statement (Brand X is dangerous) and qualifier (This statement is false). In the series of studies reviewed here, it is found that people are remarkably insensitive to innuendo qualifiers, basing their impressions instead on innuendo statements. The conditions under which this phenomenon can promote damage to the reputations of people, organizations, or products, and the steps that may be effective in avoiding such damage, are the principal concerns of the research.
[ to cite ]:
Daniel H. Wegner (1984) ,"Innuendo and Damage to Reputations", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 694-696.

Advances in Consumer Research Volume 11, 1984      Pages 694-696


Daniel H. Wegner, Trinity University


An innuendo is a statement about something combined with a qualifier about the statement. "Brand X is not dangerous," therefore, is an innuendo; it can be decomposed into a statement (Brand X is dangerous) and qualifier (This statement is false). In the series of studies reviewed here, it is found that people are remarkably insensitive to innuendo qualifiers, basing their impressions instead on innuendo statements. The conditions under which this phenomenon can promote damage to the reputations of people, organizations, or products, and the steps that may be effective in avoiding such damage, are the principal concerns of the research.


Over the past several years, a number of studies have been conducted to examine the psychology of media innuendo. Such research has focused on the impressions people form when they are exposed to a particular kind of incriminating or damaging information about a target: information that is qualified, denied, or otherwise called into question. Some of these studies have been aimed at learning the degree to which innuendo can harm reputations; they ask "Does it work?" Other studies have focused on the extent to which innuendo has impact across a wide range of audiences, sources, and targets; they ask "When does it work?" And, yet other studies have examined potential explanations for innuendo effects; they ask "Why does it work?" This report is a summary of the research on these topics to date, with special emphasis on the application of the findings to the potentially damaging role of innuendo in consumers' impressions of products and their producers.


The initial research on innuendo was conducted to estimate the impact it might have when used by the media. Wegner, Wenzlaff, Kerker, and Beattie (1981) showed research participants several newspaper headlines that conveyed damaging information about people in different ways. Some headlines were direct incriminating statements (e.g., "Bob Talbert Linked with Mafia"); others were incriminating questions (e.g., "Is Bob Talbert Linked with Mafia?"); yet others were incriminating denials (e.g., "Bob Talbert Not Linked with Mafia"). Finally, participants also saw several neutral headlines included for comparison (e.g., "Bob Talbert Arrives in City"). When their impressions of the headline targets were assessed, the participants revealed that innuendo was indeed very powerful. As compared to their impressions of the people mentioned in neutral headlines, their impressions of people who were incriminated by innuendos--such as questions and denials--ere regularly negative. Moreover, this negativity was comparable to the level produced by direct incriminating assertions.


The natural next step in the study of innuendo was the examination of the conditions under which it is influential. Research by Wegner, Kerker, and Beattie (1978) revealed that it could be remarkably effective in a setting quite outside the arena of media commentary OD public figures. This study called for undergraduate students to examine the application materials of several potential graduate students in psychology, and to judge the suitability of each applicant for admission to a graduate program. Amidst a wide array of other information sources on each applicant, a brief innuendo statement was inserted in a letter of recommendation (e.g., "I don t believe that Paul was responsible for the loss of the laboratory tape recorder"). Such a statement was enough to diminish significantly the judged acceptability of a graduate school candidate. This result indicates, then, that innuendo is not the province only of the press.

In the press, however, innuendo seems to be particularly useful. It seems to remain quite effective, for example, despite dramatic variations in the credibility of the particular media source from which it emanates. The research by Wegner et al. (1981) varied source credibility in two studies to find that such variation had little moderating influence on innuendo. Whether a particular innuendo was attributed to the New York Times or the National Enquirer, it appeared to be equally effective. This finding introduces an interesting irony. Because direct accusations were found to be less impactful when used by low credibility sources--while innuendo was not--it appears that the use of innuendo by media sources of poor repute may actually help them to get their incriminating messages across more persuasively than would the use of direct accusations. In a way, innuendo seems to serve as a buffer for the poor audience acceptance that might otherwise be expected when sensationalist sources make extravagant claims.

Is innuendo effective only because it trades on incrimination? Negative information, after all, is often more relevant to one s immediate welfare than positive information, and it could be that the indirect suggestion of negative information is powerful precisely because such bad news is salient and attractive to audiences. This possibility was explored in research by Beattie and Wegner (1980). Subjects were given a variety of statements about people to read, some of which conveyed positive information (e.g., "John gave an elderly alan some money") and some of which conveyed negative information (e.g., "Diane slapped her little brother"). Amidst these statements were others that represented explicit denials of each (e.g., "Roy did not give an elderly man some money," "Jeannie did not slap her little brother"). We reasoned that if these denials acted as they should, leading subjects to believe that the denied information was false, there should be a considerable difference in the ratings of people characterized by a statement and its denial (e.g., the ratings of John and Roy). This was not the case; for the majority of the statements we examined, people reached the same impression from both a statement and its denial. And, this was true for both positive statements--the denial yielded a positive impression--and negative statements--the denial yielded a negative impression. So, the potentially greater salience of negative than positive information did not seem to make a difference. Innuendos about positive features of people were just as influential as those about negative features.

This study also indicated, however, a variable that did seem to mediate the innuendo effect. Innuendos based on concrete statements were regularly more effective than those based on abstract statements. Thus, for instance, people were more likely to think that Harry was undesirable on reading that he "did not hold up a gas station" than on reading that he "was not cruel." Apparently, denying an abstract term like "cruel" leads people to appreciate the opposite, perhaps then thinking that the person is "kind." But there is no handy opposite for concrete information like "held up a gas station," and for this reason, a denial leaves the audience with only the incriminating statement to think about.

Further research in other contexts has demonstrated, though, that innuendos about abstract features of people can still be effective under certain conditions. Swann, Giuliano, and Wegner (1982), for example, arranged for observers to hear an interview in which one person asked another a series of leading questions. These questions either suggested that the target was an introvert (e.g., "What is it that makes you anxious in social situations?") or suggested that the target was an extrovert (e.g., "What would you do to liven things up at a party?"). Observers drew the conclusion that the target was indeed a person with the suggested trait, and did so no matter what evidence from the interview they heard. They came to believe that the person was the suggested introvert or extrovert on hearing only the questions, on hearing only the target s answers, or on hearing both questions and answers. Apparently, the occurrence of innuendo sets in motion a process whereby innuendo targets can come to portray themselves exactly as the innuendo suggests.

This research points also, however, to an interesting mediator of the effect. In a second experiment, Swann et al. gave observers evidence about the origin of the leading questions. The observers were told in one experimental condition that the questions had been "drawn at random from a fishbowl." Under these conditions, the observers lent far less credence to the questions alone, and so did not attribute introversion or extroversion to people who had been questioned along these lines. On hearing targets answers, though, or on hearing the questions and answers together, observers again reported that the target was characterized by the suggested trait. This finding suggests that innuendo alone is powerful when it is presumed to emanate from some knowLedge or expectation on the part of its communicator; when it stems from a "fishbowl," it loses its impact. But even when innuendo appears to come from thin air, it can lead targets to behave so as to verify its message.


The studies of the limiting conditions of the innuendo effect suggest some ways of explaining it. Certainly, one might focus on information concreteness as a potentially important explanatory variable. It is tempting, too, to point to the presumption people commonly make about the validity of an innuendo message. They seem to assume that innuendo does not come from a "fishbowl," but rather that the communicator had some valid reason for bringing it up--if only to deny it. These lines of explanation are perhaps too limiting, however, for there is yet a more encompassing way of understanding the phenomenon. Quite simply, innuendo effects may occur because people do not fully comprehend the meaning of qualifiers attached to information statements

The psychological study of reasoning processes lends support to this idea. Wason and Johnson-Laird (1972), for example, have documented in a variety of experiments the difficulty with which people understand simple negation. It seems that once an idea is "let out of the bag," attempts to qualify it, deny it, or otherwise temper its truth value are likely to be less than fully successful. Information that is conveyed through questions, denials, or even qualifications (e.g., This possibility may not be true) is understood first, and is then cognitively re-processed and imbued with the internal equivalent of the logical "not." The "not" qualifier becomes, then, what may be a superfluous addendum to the stored information. When negations are thus incompletely understood in the context of syllogistic reasoning, they produce a wide spectrum of logical errors. When they are misunderstood in the context of media reporting, they produce innuendo effects.

This explanation for the innuendo phenomenon gave rise to a recent study by Wegner, Coulton, and Wenzlaff (1983). The purpose of the research was to examine the source of what has come to be known as "impression persevere Ice." This effect was initially uncovered by Ross, Lepper, and Hubbard (1975), and occurs when people are given some performance information about themselves or another person. If that information is subsequently shown to be false--as when an experimenter in the study indicates that performance feedback a subject received was entirely fabricated--subjects typically continue to believe it. They make predictions about future performance that are consistent with the feedback, despite the fact that the experimenter has admitted to its complete inauthenticity. Ross et al. explained this effect by suggesting that people develop an interpretive structure for the information as it is encountered, and that this set of attributions serves to maintain belief in the information even when the information is shown to be false.

The experiment by Wegner, Coulton, and Wenzlaff (1983) replicated the research by Ross et al. precisely--with one exception. Some subjects were presented not with a denial of their performance feedback after their performance, but rather b fore their performance. So, instead of being debriefed, these subjects were briefed; they were told in advance that the success/failure information they would be receiving for the task they were to do would be entirely fabricated by the experimenter. These subjects were thus presented with the job of understanding a blatant denial. And they behaved just as our explanation of the innuendo effect would suggest they should. Even on being warned in advance that the performance feedback was false, subjects given success feedback predicted future success, whereas those given failure feedback predicted continued failure.

These findings indicate that the attributional processes postulated by Ross et al. may have little role in the production of impression perseverance. Rather, this effect can be counted as an instance of the innuendo phenomenon--and chalked up to the failure of individuals to appreciate fully the meaning of information qualifiers. Unlike most computers, people do not have a "reset" button that allows their minds to be wiped clean of information. Rather, they understand any attempt to qualify an assertion as an additional piece of information that should be considered in concert with the assertion. Such combination never seems to remove entirely the mark of the original assertion.


The innuendo research suggests that in a broad range of everyday instances, the least hint that something is possible will be enough to convince some proportion of people that it is true. It would not be surprising, therefore, to find that innuendo influences consumers--about products (e.g., "Does Brand X cause cancer?"), about manufacturers (e.g., "The corporation did not engage in an illegal kickback scheme"), or about retailers (e.g., "The store may have a rat problem"). Although innuendo research has yet to be conducted in this arena, it is possible to wonder even at this point how innuendo effects might be prevented or overturned when they threaten to damage product or producer reputations.

Perhaps the most obvious strategy is prevention. Although rumors and media innuendos can arise without any foundation, they may oftentimes begin with some possibility disseminated by producers themselves. The spokesman for the Three Mile Island nuclear facility who indicated at the height of their crisis that "we are not in a China Syndrome situation" probably created more panic than he quelled. It may often be tempting to comment at length on some problem, but this tactic could easily serve as innuendo to fuel public concern far beyond what might otherwise arise.

Once an innuendo is made public, different strategies are needed. Research on the effectiveness of these is still ongoing, but at this point at least some suggestions can be made. Wegner and Wall (1983) investigated the effectiveness of some seven strategies for repairing reputations damaged through innuendo. The study exposed subjects to the suggestion that a man was being unfaithful to his wife, and examined their judgments of him after he had offered a reply. Although his brief denial (e.g., "No, I didn't do that") was more successful than several more elaborate tactics (i.e., introducing humor, becoming angry, offering a lengthy denial, etc.), it still did not bring subjects impressions of his culpability back to pre-innuendo levels. Similarly, while it was not entirely effective, a strategy of introducing an alternative explanation for the events noted in the innuendo reduced the damage to his reputation.

The study of the innuendo effect, in sum, indicates that it is fairly pervasive. Innuendo can be limited by some features of the context in which it occurs, but is not limited by more other contextual features that common sense might suggest would bring it under control. Most probably, this is because the phenomenon results from a basic difficulty in the design of the human information processor--an inability to lose information completely, even when that information is known to be false. Additional inquiry along this line could be profitably directed toward the strategies whereby incriminating innuendo might be met with effective replies.


Beattie, A. F., & Wegner, D. M. (1980), "The denial innuendo effect in impression formation," Presented at the meeting of the Eastern Psychological Association, Hartford .

Ross, L., Lepper, N. R., & Hubbard, N. (1975), "Perseverance in self perception and social perception: Biased attributional processing in the debriefing paradigm," Journal of Personality and Social Psychology, 32. 880-892.

Swann, W. B., Jr., Giuliano, I., & Wegner, D. M. (1982), "Where leading questions can lead: The power of conjecture in social interaction," Journal of Personality and Social Psychology, 42, 1025-1035.

Wason, P. C., & Johnson-Laird, P. N. (1972), Psychology of reasoning, Cambridge, MA: Harvard University Press.

Wegner, D. M., Coulton, G., & Wenzlaff, R. (1983), "Processing denials: The impact of briefing in the debriefing Paradigm." manuscript in Preparation.

Wegner, D. M., Kerker, R. M., & Beattie, A. E. (1978), "Innuendo in impression formation: When a question is an answer," Presented at the meeting of the Southwestern Psychological Association, New Orleans.

Wegner, D. M., & Wall, M. (1983), "I walk the line: Protecting one's reputation as a faithful intimate," Presented at the meeting of the Midwestern Psychological Association, Chicago.

Wegner, D. M., Wenzlaff, R., Kerker, R. M., & Beattie, A. E. (1981), "Incrimination through innuendo: Can media questions become public answers?" Journal of Personality and Social Psychology, 40, 822-832.