Exemplification in Communication

the influence of Case Reports on the Perception of Issues

Dolf Zillmann, Hans-Bernd Brosius

• Publisher: Routledge
• Year: 2012
• Print ISBN: 9780805828115
• eBook ISBN: 9781135681609
• Topic: Languages & Linguistics
• Subtopic: Communication Studies
• Index: Languages & Linguistics

This volume offers a new conceptual framework for exemplification, a coherent theoretical approach based on contemporary psychological models of information processing, and an exhaustive integration of the pertinent research demonstrations. Focus is on the news media, but the influence of fiction and quasi-fiction is also considered. The informational competition between concrete, verbal, or pictorial exemplification and abstract, mostly quantitative exposition is analyzed. Implications for issue perception, including delayed consequences are also examined.

Exemplification is subjected to conceptual scrutiny and a new theoretical framework is offered. Contemporary psychological paradigms are applied to predict effects of various forms of exemplification. Perhaps most important, novel experimental research is presented to document the specific consequences of exemplifications featured in the news, even of those featured in fiction. Finally, recommendations for information providers and recipients are derived from the research demonstration in order to advance media literacy specific to exemplification.

This unique volume:
* provides a comprehensive account of the power of case-report selection in the manipulation of perceptions of social issues,
* addresses exemplification in communication, i.e., the influence of case reports in the news media, primarily, on the perception of pertinent social issues,
* offers an empirical assessment of the practice of issue exemplifying by the media,
* gives an exhaustive account of representative research on exemplification effects on issue perception--primarily by the news media, but also by the entertainment media, and
* includes a compilation of guidelines for information providers and recipients in efforts at creating media literacy with regard to exemplification.

One
Exemplification in Communication

• Conceptual Considerations
• Definition of the Exemplification Process
  • Exemplification of Known Event Populations
  • Exemplification of Unknown Event Populations
• Exemplification in Different Domains of Communication
• Exemplification in Personal Experience
• The Interface Between Direct and Mediated Experience

Everybody is familiar with examples. Everybody has been given examples, and everybody has related examples to others in efforts of elucidating a broader concept or issue. Everybody, then, has some tacit understanding of a relation between an example and a larger entity to be exemplified by it. What is implied is that more than one example exists and that several examples tend to do a better job than just one in explaining aspects and features of the exemplified entity. What is also implied is that an utterly unique, singular incident, such as the first moon landing by humans, could only exemplify itself; hence, it could not serve as an example of other first moon landings by humans. It may, however, serve as an example of other landings of spacecraft, if the expositional focus were on aspects of the event that exhibit a degree of similarity with other events under consideration.

Tacit understanding of exemplification thus entails recognition of shared features between an example (also called exemplum or exemplar) and the exemplified, as well as between all possible examples (also called exempla or exemplars) of the exemplified. In simple terms, such sharing amounts to similarity between exemplars and the exemplified. Lexical definitions focus on this similarity by stipulating that the exemplar be typical or characteristic of exemplified entities. The highest degree of similarity is demanded by definitions that specify the exemplar as a case in point or an instance of the exemplified entities. Both specifications suggest an array of identical entities from which any single one can be taken to exemplify all others. Each and every instance may be singled out to inform about all other instances.

The stipulation that all exemplars be identical may seem overly stringent but actually is not when one considers that the stipulation need not be applied to all features of an entity. We can speak of New Yorkers as identical entities in that their domicile is New York. They may differ in any other regard. It is important, then, to distinguish a set of features for which interexemplar similarity is required from a set of features that are free to vary. The latter features are immaterial in considering a particular instance an exemplar of other instances subsumed in the exemplified group of instances. As a consequence, exemplars are to be considered instances of whatever kind that are capable of representing other instances only to the extent that they share with them all defining features.

The specification that exemplifying and exemplified instances may be of whatever kind needs elaboration. The features of perceptible entities and events obviously can be represented to percipients. Features of objects such as trees, houses, and bridges can be exemplified, as can dogs, people, and computers, along with their perceivable actions. The overt “behavior” of entities is, of course, ready subject to exemplification. Less obvious might be the representation of abstractions and concepts that elude direct perceptual control. Extracted relations between entities, either in the form of covariations or causality, may be exemplified nonetheless. Similarly, the match between intent and performance, or that between precept, action choice, the expectation of consequences, and actual consequences may also be exemplified. For instance, children who consume a fair amount of fairy tales readily appreciate that elderly women with a deformity of the back and a screechy voice harbor hostile intentions and, given the opportunity, act on them. These children also appreciate that actions, such as a witch’s efforts to make a meal of Hansel and Gretei, are in violation of moral precepts and call for punishment. Moreover, they are able to spot a match between punitive precepts and punitive actions against a wrongdoer and hence can rejoice when the witch gets her just deserts. All of these assessments and judgments are exemplars of conduct—in particular, of socially reproached and punished conduct as well as of socially approved and rewarded conduct.

The illustration of witch-defining features entails a leap from detecting an abstraction to applying it to other situations and contexts. This extension is part and parcel of exemplification. If the identification of an exemplar in a given context (i.e., the recognition of a resemblance between a familiar abstraction and a particular manifestation of that abstraction) were the terminal stage of the process, exemplification would be of little relevance to communication. As it stands, however, exemplification is mostly the starting point in that the world of exemplars appears to influence our perception and judgment of essentially all phenomena and issues of the so-called real world.

Analogous to forming impressions and dispositions toward entities and happenings on a sampling of pertinent experiences, our perception and judgment of phenomena and issues with which we have little or no immediate contact are bound to be influenced by samplings of mediated events. Such influence can be seen as the result of nonconscious inductive inference (Bargh, 1996; Kissin, 1986; Lewicki, 1986). The more general case is inferred on the basis of limited mediated experience with relevant happenings. For the vast majority of the citizenry, for instance, violent crime is not immediately experienced. Yet perceptions are formed on the basis of news reports, friends’ hearsay, and possibly fictional portrayals. Moreover, judgments of the moral variety are formed on the basis of these perceptions. We may smile or cringe when a child, apparently as the result of frequent witch exemplification in Grimm-style fairy tales, points to a lady in the street and utters, “A witch, a witch!” However, do we routinely, if ever, examine the etiology of our own beliefs and dispositions, especially regarding the likely influence of communication-mediated exemplars? Surprisingly little attention has been paid to this aspect of interpersonal and media influence.

We attempt to correct this neglect by subjecting the influence of communication-mediated exemplification to systematic analysis. We first examine exemplification in conceptual terms, then ascertain what the media do in generating an exemplar flood, and finally explore the effects of exemplification on the perception and judgment of phenomena and issues.

Conceptual Considerations

In conveying information about the flow of happenings in the so-called real world it always has been deemed appropriate, if it was not plainly recognized as a necessity, to cut this flow into manageable chunks and to isolate and focus on some events at the expense of attention to occurrences in between (Burns, 1992; Rosch & Lloyd, 1978; Tversky & Hemenway, 1984). Narratives, as a rule, leap from event to event, irrespective of the events’ locality and position in time. More important here, narration aggregates events that exhibit sufficient phenomenal similarity to warrant their being classified as manifestations of the same kind. Such grouping implies that each and every grouped event, to the extent that it shares all essential attributes with the remaining grouped events, is capable of representing the group at large—meaning that it is capable of providing reliable information about all other events in this group and thus about the group itself. It is this capacity of individual events that defines them as exemplars of an event group. Given that the events in a particular group share all essential attributes, as was stipulated, each and every group member would indeed exemplify the group attributes. If, for instance, it can be considered established that all humans are mortal, then each and every human would exemplify human mortality.

The outlined paradigm of representation is, of course, an abstraction. Since antiquity it has been argued that no two events are truly alike. Grouping, classifying, or categorizing on grounds of likeness has been practiced through the ages nonetheless (Burns, 1992; B. Hayes-Roth & Hayes-Roth, 1977; Mervis & Rosch, 1981). Economy and efficiency of thought and of information conveyance would seem to necessitate it. It should be clear, however, that conditions in which grouped events are fully interchangeable, and thus capable of representing the group without error, exist rarely, if ever. Especially in the realm of human affairs, exemplification is bound to be less than perfect, and a certain degree of imprecision is unavoidable and also may be immaterial for many practical purposes.

The indicated imprecision, formally expressed, derives from the fact that events are necessarily grouped on the basis of a limited number of attributes, with a potentially large number of additional attributes remaining unidentified or being ignored. If n attributes are identified and employed as grouping criteria, m attributes may vary freely. For instance, if the group event is defined as carjacking committed in the United States during the 1990 to 1996 period by men aged 15 to 25 years, any particular carjacking within this group is likely to exemplify the specified crimes only poorly because some perpetrators will have only threatened violence, whereas others may have used force, even deadly force, against the evicted car owner. Exemplification of the crime by any particular crime thus cannot fully and impartially represent the grouped crimes.

Conceptually, precision in exemplification is readily specified. It requires that n, the number of defined attributes employed as grouping criteria, be associated with m=0, the number of undefined attributes that may be pertinent. In case it is recognized that m > 0, m must be reduced to zero by incorporating the m attributes in n. In the carjacking illustration, this would mean that subgroups should be created and that exemplification should be limited to these subgroups. For instance, the group of deadly carjackings would have to be isolated and could be exemplified only by individual cases of deadly carjackings. Carjackings by other means would have to be treated analogously. Such group partitioning by increased definitional specificity would be recursive without apparent end. In the carjacking illustration, deadly force may, after all, have been applied in different ways; for instance, by clubbing, by knifing, or by shooting. The number of grouping attributes would have to be increased again, producing an ever larger number of ever smaller, yet more specific, event groups. The direct, single-case exemplification of the crime of carjacking irrespective of particular manifestations would no longer be feasible because the subgroups, unless their case count is zero in all but one of the groups, are bound to reveal marked differences in the manifestations of the grouped events. Depending on one’s focal interest, then, it will have to be allowed that m, the number of undefined and uncontrolled attributes, be larger than zero. Variation in these attributes will have to be accepted on grounds of practicality, meaning that some degree of imprecision in exemplification is to be tolerated in order to achieve greater efficiency in the conveyance of information about grouped events.

Although some imprecision in exemplification may be unavoidable, the concept of representation implies that the highest degree of precision attainable under given circumstances should be pursued. Common definitions of exemplification stipulate that a group of events need to be represented by single events that are typical and characteristic of the group. Exemplification by atypical and uncharacteristic events is deemed inappropriate because it fails to provide reliable information about the group. The arbitrary selection of “a case in point” can have utility only for the exemplification of a homogeneous event group; that is, for a group with minimal variance in uncontrolled attributes. Whenever such variance is more than minimal, which it is likely to be for most issues of concern, exemplification by arbitrary selection would seem to be unacceptable—even irresponsible, if the object is to provide veridical information about an event group.

For instance, if carjackings with deadly outcome amounted to a trivial number and noninjurious outcomes accounted for almost all crimes of this kind, it would seem inappropriate, indeed, to exemplify the crime of carjacking by the presentation of a deadly case only. Such inappropriate exemplification is bound to mislead the recipient of the information, resulting in erroneous conceptions about the danger associated with the crime at large.

The obvious limitations of exemplification by a case in point (i.e., one case) can be overcome, of course, by exemplification that draws on multiple exemplars. Using several or numerous exemplars does not guarantee greater precision in representation, however. Arbitrary selection of exemplars may lead to duplications of partiality and may thereby escalate misrepresentation. The selection of two or three deadly cases in our carjacking illustration should make that point. Representational accuracy would be better served if exemplars were selected blindly. In our illustration, the most frequently occurring forms of carjacking would likely be drawn to represent the crime, and greater accuracy would be insured, to a point. The frequent case, it should be noticed, defines what is to be considered typical and characteristic. If carjacking, for instance, is mostly injury-free, this attribute is typical and characteristic. Exemplification by a small number of blindly drawn exemplars, say three or five, would bring out this property of carjacking. It would furnish a reliable projection of the crime of carjacking in these terms. However, exemplification would probably not include deadly carjackings because of their extremely low incidence. Representation by a small number of blindly drawn exemplars thus may still be considered imperfect. It is likely to be incomplete, thereby allowing or fostering erroneous assessments of the grouped events. Only the use of large numbers of exemplars would insure the inclusion of infrequently occurring but nonetheless relevant events. Clearly, representational accuracy for a large event group is higher, the more closely the number of blindly drawn exemplifying events approximates the total of grouped events. However, although the employment of such large numbers of exemplars accomplishes great representational precision, it is often or mostly unworkable because it entails a forbidding amount of redundant information about the typical case. In our illustration, the recipient of an account of carjacking might have to be informed about hundreds of frequent cases before being appraised of the fact that carjacking may have deadly outcomes as well.

An apparent way out of this dilemma is, of course, to attach quantitative information to specific exemplars. The typicality of particular occurrences within a group of events would be defined in frequencies or in proportions. The effectiveness of such presentational strategies in terms of acquisition, processing, retention, and retrieval of the supplied information by recipients remains to be seen, however.

Definition of the Exemplification Process

Our introductory comments may have made apparent that the representation of a group of events by exemplar events resembles that of the representation of a population of events by a sample of events. The relation between exemplification and sampling is indeed strong enough to accept the representation of population events by sampled events as a model for some forms of exemplification. We do, in fact, adopt much of the well-established nomenclature and procedures of statistical representation and inference for particular types of exemplification. However, there exist forms of exemplification that differ considerably from the statistical model, and we modify this model as the circumstances require.

We first formalize exemplification processes that are analogous to sampling from a specified population and thereafter consider exemplifications that define populations.

Exemplification of Known Event Populations

A population of events is defined as a usually finite aggregate of events that share a limited number of specified characteristics but that may differ in numerous unspecified properties.

A sample of events is defined as any possible subset of the population events.

A random sample of events is defined as any subset of the population events for which every sampled event had the same chance of inclusion. This is the condition of equiprobability that insures impartial, unbiased representation of the population by the sample.

A subpopulation of events is defined as a subset or a stratum of the population of events. Events subsumed in subsets or in strata are specified by a limited number of characteristics in addition to those specifying the events of the population. Sampling from a subpopulation is analogous to sampling from a population.

In the special case of complementary subpopulations of known size, samples drawn randomly from the subpopulations may be combined in proportion to their size to represent the population. This process, known as stratified sampling, is economical in that it prevents redundant oversampling in large subpopulations while insuring consideration of small subpopulations.

These concepts apply directly to the exemplification of well-defined, known populations and subpopulations of events. A minor adjustment concerns population size. Regarding exemplification, the ...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Preface
• 1 Exemplification in Communication
• 2 Exemplification in Practice
• 3 Information Processing
• 4 Exemplification Effects of the News
• 5 Exemplification Effects of Fiction and Quasi-Fiction
• 6 Toward Exemplification Literacy
• References
• Author Index
• Subject Index