Special Session Summary the Analogy and Metaphor Literature: What’S in It For Consumer Researchers?

Jennifer Gregan-Paxton, Washington State University
[ to cite ]:
Jennifer Gregan-Paxton (1998) ,"Special Session Summary the Analogy and Metaphor Literature: What’S in It For Consumer Researchers?", in NA - Advances in Consumer Research Volume 25, eds. Joseph W. Alba & J. Wesley Hutchinson, Provo, UT : Association for Consumer Research, Pages: 210-213.

Advances in Consumer Research Volume 25, 1998      Pages 210-213



Jennifer Gregan-Paxton, Washington State University


A significant proportion of human learning and problem solving is facilitated by cognitive comparison processes, ones that relate the novel to the familiar. For example, the first time home buyer deals with the novelty of the purchase experience by relating it to something more familiar, such as the financing of a car. The objective of these comparisons, which have been extensively investigated in the literature on analogy and metaphor (e.g., Ortony, Reynolds and Arter 1978), is the transfer of meaning and structured knowledge from a familiar to a novel domain. Analogy and metaphor, and the comparisons underlying them, facilitate comprehension by taking advantage of goal-relevant commonalities that exist between otherwise disparate domains.

Many argue that analogy and metaphor represent key mechanisms of knowledge development and category acquisition in children as well as adults (Vosniadou 1989; Gentner and Holyoak 1997). A vast body of theoretical and empirical research suggests that comparison-initiated transfer of structured knowledge plays a critical role in a broad array of cognitive tasks, such as solving math word problems (Reed 1987), forming legal arguments (Marchant et al 1991), and predicting an individual’s behavior (Read 1987). Despite their apparent ubiquity, however, we know very little about these comparisons in a consumer context. We lack even a rudimentary understanding of how, when, and to what extent metaphor and analogy influence consumer behavior (see Gregan-Paxton and John 1997).

This session brought together researchers who are actively addressing this void in our understanding of consumer behavior. Their presentations provided a glimpse at the different research questions that can be addressed by the theoretical perspectve represented in the analogy and metaphor literature. In the first presentation, George Babbes (co-authored with David Aaker) focused on the incidence of conceptual metaphors in advertising, providing data that sheds light on the ways in which conceptual metaphors are used in advertising and whether their use benefits some products more than others. In the second presentation, Kathryn Fitzgerald focused on the processing of analogies appearing in advertisements, presenting evidence implicating such factors as the novelty of an analogy and the degree of verbal description detailing the intended mapping between the source domain and the product as determinants of an ad’s effectiveness. Moving the session out of the advertising realm, in the third presentation, Shi Zhang (co-authored with Arthur Markman) focused on the comparison process underlying analogical learning and similarity processing, suggesting that similarity is like analogy and presenting evidence that the representational relationship between a new brand and existing brands has critical implications for a consumer’s memory for product attributes and for his/her preference for new and existing brands. Addressing the comparison issue further, in the final presentation, Jennifer Gregan-Paxton focused on the processing implications of different types of comparisons, presenting data that identified two distinct ways in which consumers transfer their existing knowledge in the process of evaluating a novel brand extension product. Bringing these researchers together in one session promoted an appreciation for the potential of this literature to serve as a productive guide to consumer research, one that raises new issues and answers a broad range of questions.




George S. Babbes, Cornell University

David A. Aaker, University of California, Berkeley

Although the use of conceptual metaphors in everyday communication has been one of the most researched areas across the cognitive sciences, virtually no work has examined the use of conceptual metaphors in advertising. While several papers have examined metaphor use in advertising based on an Aristotelian view (e.g., Leigh, 1994; McQuarrie and Mick, 1995), this approach does not (1) identify the full extent of use, and (2) consider the full range of cognitive aspects. In this paper- motivated by recent work on conceptual metaphors in cognitive psychology, linguistics and anthropology- we develop and empirically test several hypotheses regarding the use of conceptual metaphors in advertising.

More specifically, recent work on the analysis of metaphor in advertisements still relies on the traditional characterization and classification of this "figure of speech" as a purely linguistic option, independent of cognition, a view essentially unchanged since the ancient Greeks (often called the Aristotelian view). However, this view assumes that the meanings of words, including metaphors, exist in the world independent of human perception and cognition. This is simply not the case. For example, considerable work in cognitive linguistics (e.g., Lakoff, 1993) and cognitive psychology (e.g., Gibbs 1995) has demonstrated quite the opposite. This growing body of work links much of our everyday understanding and reasoning to metaphor-based cognitive structures.

If metaphor-based cognitive structures exist and form the basis for much of our understanding and reasoning, we would expect to see (1) considerable use of conceptual metaphor in advertising (to effectively establish a cognitive structure), and (2) a substantive connection between the metaphors in advertising and those that underlie our everyday understanding and reasoning (to efficiently establish a cognitive structure). In contrast to the Aristotelian view, this view significantly elevates the role of metaphor in advertising ommunication. That is, this view argues that metaphor is primarily a matter of thought and only derivatively a matter of words.

To examine these hypotheses (among several others), we conducted a content analysis of approximately 1,200 print ads across four product categories (cars, computers, travel services, and financial services). We used a sample frame of six consumer magazines over a 10 month period. Each ad was independently coded by two graduate linguistics students.

Some of the key findings include: (1) conceptual metaphors are used in almost all ads; (2) conceptual metaphors are the dominant source of substantive content for approximately half of the ads; (3) conceptual metaphors are used more often for substantive content in abstract categories than in concrete products; (4) the set of source domains used for a given product category is very constrained (less than 10); and (5) the source domains are all based on metaphors we use in everyday discourse (e.g., Lakoff and Johnson, 1980); that is, they are not deviations from language as the Aristotelian view suggests.

These results offer empirical support in the domain of advertising for one of the most actively researched areas across the cognitive sciences. For the first time, we offer a look at how much, how and when conceptual metaphors are used in advertising. Our results are consistent with and extend metaphor research in linguistics, psychology and anthropology.



Kathryn Fitzgerald, University of Notre Dame

The ability of analogy to suggest a framing for issues and facilitate learning in creative, insightful ways is well recognized in the sciences, religion, education, and in the use of legal precedent (Ortony, 1993; Holyoak and Thagard, 1995). The use of analogies also appears to be a fairly common advertising technique, particularly for products in complex and/or very new categories (Fitzgerald, 1996). This presentation discusses research assessing the use of analogies to assist with the problematic issue of consumer education about breakthrough "really new" products.

Understanding the nature of the thought processes involved in reasoning by analogy is currently one of the most vital streams of research emanating from the field of cognitive psychology. However, much of the knowledge accumulated to date stems from educational problem solving contexts in which persuasion is not a primary concern. Consequently, there has been no theoretical or empirical work devoted to affective responses to analogies suggested by an external source. Concurrently, active research on the related topic of metaphor has explored outcomes that more closely approach advertising goals, such as perceived aptness and memorability, but has not considered situations in which new information is being acquired about an unfamiliar domain. This research drew upon theory and findings from both of these streams of basic research to formulate a theoretical understanding of the process by which analogies impact both attitude formation and the learning of new product information.

Discussion began with the presentation of a general classificatory framework for thinking about the many characteristics of ads and their audience members that can be expected to influence the effectiveness of analogies in producing favorable attitudes and enhancing learning and retention of product information. The framework emphasizes the role of an analogy-based message’s processing complexity as a mediator of the combined influence of these ad and audience variables. Analogy-based ads inducing a moderate level of processing complexity were hypothesized to produce more favorable attitudes and more complete and long-lasting learning than ads that are either too simple or too complex. This research examined the interaction of two specific ad characteristics expected to impact how complex it is to process an analogy-based ad: the novelty of the analogy and the amount of verbal explanatio describing the intended mapping between the product and the source domain. Specifically, ads using more novel analogies were hypothesized to be more complex to process, and ambiguous ads that do not provide a description of the intended mapping were hypothesized to be more complex to process. A moderating interaction was also predicted: novelty was expected to increase processing complexity to a greater extent when ads are verbally ambiguous than when verbally explicit.

Discussion then proceeded to the results of the first of two laboratory experiments. Six versions of a mock print advertisement for a fictitious high technology product were created for use in a fully crossed 3 x 2 (novelty of the analogy x verbal ambiguity) between-subjects design. Expertise, interest in related product areas, and need for cognition were also assessed as potential individual difference covariates. The findings generally supported the view of processing complexity as a mediator of the effects of novelty and ambiguity. There was a very strong inverted-U relationship between the perceived complexity of an analogy-based ad and the attitudinal evaluations of both the product and the ad, as hypothesized. Ads that were rated as moderately complex produced more favorable outcomes than ads rated as either extremely simple or extremely complex. Learning measures also followed a strong inverted-U relation to perceived ad complexity. In the case of the learning measures, the advantage of moderate complexity was even stronger than expected. A direct "simpler is better" relation was expected unless measures were taken after a delay and involved more effortful retrieval. However, moderately complex ads produced stronger learning even when the measures were collected immediately following exposure to the ad. As predicted, the novelty of the analogy significantly increased processing complexity. However, the interaction between novelty and ambiguity was more subtle than expected. Ambiguous ads were actually perceived as less complex than ads with added verbal explanation of the intended mapping; the opposite was true for highly novel analogies. The overall pattern of results suggests that moderately novel ads that were ambiguous were processed at a superficial level in which the analogy was simply recognized as an instance of a link between commonly linked domains. Conversely, when the intended particular connections between the two were verbally described, a deeper transfer-of-meaning process was engaged.



Shi Zhang, UCLA

Arthur B. Markman, Columbia University

Analogy is a device for conveying that two domains share relational structure despite arbitrary degrees of differences in the objects that make up the domain. Similarity, as a concept as well as a process, emphasizes the sharing of both the relational structure and the attribute properties of the objects in the domains. This recent view of "similarity is like analogy" (Medin, Goldstone and Gentner 1993, Markman and Gentner 1993) characterizes analogy and similarity as a continuum and proposes that the process of carrying out a comparison is the same in both cases. This process is known as structural alignment, an explicitly articulated mental representations about the relational structure of causal dependencies and descriptive attributes of the objects in the relationship.

In this discussion, we first review the structural alignment theory of analogy and similarity. Crucially, we demonstrate that the analogy-based similarity view emphasizes common relational structures in a comparison. This emphasis on certain commonalities in a relational system gives rise to two kinds of differences: (1) the alignable differences, i.e. differences that are related to commonalities (e.g. different weights on an attribute), and (2) the nonalignable differences, i.e. differences that are not related to the commonalities (e.g. unique attributes).

We then argue that learning of new brands in an existing category isinfluenced by the way the attributes of the later entrants compare to the attributes of the first entrant. In two experiments, we show that attributes that differentiate later entrants from the first entrant are better remembered if the attributes are alignable differences than if they nonalignable differences. The greater memorability of alignable differences than of nonalignable differences interacts with the order of entry of brands: later entrants whose attributes are superior to those of the early entrant can come to be preferred over the early entrant when the attributes are alignable differences, but not when they are nonalignable differences.

We conclude the discussion by suggesting that (1) researchers need to pay attention to the processes and representations of analogy, similarity, metaphor and other higher order cognitive processing; (2) the structural alignment model has profound implications on consumer learning about new brands and products in that mental representations of attribute structure (a hierarchical structure of relational and attributional properties in a comparison) impact both memory and preference judgment.



Jennifer Gregan-Paxton, Washington State University

Research guided by the well-established dual processing approach suggests that product evaluations are generated by either a category-based process, which involves the internal transfer of information from a related product category schema to the new product, or a piecemeal process, which involves the external acquisition and integration of information from the product itself, or some other source external to the consumer (e.g., Fiske and Neuberg 1990; Sujan 1985; Boush and Loken 1991). Although a considerable amount of consumer research has been directed toward external acquisition processes, surprisingly little is known about the process or processes by which the internal transfer of consumer knowledge occurs. This is a troubling oversight when one considers the critical role that knowledge transfer has been assumed to play in areas such as comparative advertising, brand extension processing and country-of-origin effects.

The purpose of this paper is to further our understanding of the process of internal knowledge transfer by examining the transfer that occurs as consumers evaluate novel brand extension products. Building on past research, which suggests that internal transfer occurs by a single process, we propose that internal knowledge transfer actually occurs via two distinct transfer processes. Specifically, drawing on theory and research from the analogical learning literature, we propose that transfer occurs via a schema-based process or similarity-to-exemplar process, with the nature of the transfer process being a function of two interacting factors. The first factor concerns the relationship between the extension product and the brand schema and, in particular, whether the two share primarily attributes or attributes and relations. In our framework, an attribute is any independent property of a product (e.g., Colgate toothpaste is white), whereas a relation refers to how a product or one of its attributes relates to another product or attribute (e.g., Colgate toothpaste fights cavities). The second factor concerns the nature of the consumer’s brand knowledge and, in particular, whether it contains primarily attributes (i.e., a novice) or attributes and relations (i.e., an expert).

Consistent with our framework, the results suggest that when consumers draw upon existing knowledge in the course of evaluating a new product, they do so by one of the two proposed processes. Schema-based processing occurs when the new product can be mapped onto an existing knowledge structure in terms of relations as well as attributes. Because this internal transfer process requires relational mappings to be constructed, it is used primarily by consumers with well developed knowledge structure. In contrast, similarity-to-exemplar processing occurs when the new product can only be mapped onto existing knowledge structures in terms of attributes. Although primarily used by consumers with impoverished knowledge structures, similarity-to-exemplar processing is also used by consumers with a higher level of knowledge when the absence of a relational match between a new product and an existing knowledge structure blocks the use of schema-based processing. A key implication of our study is that existing knowledge is likely to play a larger role in the evaluation of new products than research guided by the dual processing paradigm has led us to believe.



Deborah Roedder John, University of Minnesota

The discussion leader reflected on the occurrence of analogies and metaphors in the consumer domain, offering several observations designed to spark additional interest in this nascent stream of research. First, it was noted that some analogical and metaphorical thinking occurs spontaneously (it is self-directed), whereas other analogical and metaphorical thinking occurs because of prompts provided by an external source (it is other-directed), such as a marketer. Second, it was pointed out that some domains appear better suited to serve as the source (or base) of an analogy or metaphor than others. As an example, the domain of baseball was recognized as the source of numerous analogical or metaphorical comparisons, such as "step up to the plate" and "you really struck out with them." In contrast, other sports domains, such as hockey, rarely serve as the source of analogies or metaphors. Finally, it was proposed that individuals may differ in their propensity and/or ability to engage in the kind of comparative thinking underlying analogies and metaphors. Constructing an analogy to research exploring "need for cognition," Debbie proposed a new construct, which she termed "need for analogy," to capture this potentially important individual difference variable.


Boush, David and Barbara Loken (1991), "A Process Tracing Study of Brand Extension Evaluation," Journal of Marketing Research, 28 (February), 16-28.

Fiske, Susan and Steven Neuberg (1990), "A Continuum of Impression Formation, From Category-based to Individuating Processes: Influences of Information and Motivation on Attention and Interpretation," in Advances in Experimental Social Psychology, vol. 23, ed. M.P. Zanna, Orlando, FL: Academic Press, 1-74.

Fitzgerald, Kathryn A. (1996), "Ana-logic Theory and Practice: How and When Do Advertisers Use Analogies?" Working Paper.

Gentner, Dedre and Keith J. Holyoak (1997), "Reasoning and Learning by Analogy," American Psychologist, 52, 1, 32-34.

Gregan-Paxton, Jennifer and Deborah Roedder John (1997), "Consumer Learning by Analogy: A Model of Internal Knowledge Transfer," Journal of Consumer Research, 24 (December).

Holyoak, Keith and Paul Thagard (1995), Metal Leaps: Analogy in Creative Thought, Cambridge, MA: The MIT Press.

Marchant, Garry, John Robinson, Urton Anderson, and Michael Schadewald (1991), "Analogical Transfer and Expertise in Legal Reasoning," Organizational Behavior and Human Decision Processes, 48, 272-290.

Markman, Arthur and Dedre Gentner (1993), "Structural Alignment During Similarity Comparisons," Cognitive Psychology, 25, 431-467.

Medin, Douglas, Robert Goldstone and Dedre Gentner (1993), "Resects for Similarity," Psychological Review, 100, 2, 254-278.

Ortony, Andrew (1993), Metaphor and Thought, New York, NY: Cambridge University Press.

Ortony, Andrew, Ralph Reynolds, and Judith Arter (1978), "Metaphor: Theoretical and Empirical Research," Psychological Bulletin, 85, 5, 919-943.

Read, Stephan (1987), "Similarity and Causality in the Use of Social Analogies," Journal of Experimental Social Psychology, 23, 189-207.

Reed, Stephen (1987), "A Structure-Mapping Model for Word Problems," Journal of Experimental Psychology: Learning, Memory and Cognition, 13, 1, 124-139.

Sujan, Mita (1985), "Consumer Knowledge: Effects on Evaluation Strategies Mediating Consumer Judgments," Journal of Consumer Research, 14 (December), 372-378.

Vosniadou, Stella (1989), "Analogical Reasoning as a Mechanism in Knowledge Acquisition: A Developmental Perspective", in Similarity and Analogical Reasoning, Stella Vosniadou and Andrew Ortony, Cambridge: Cambridge University Press, 413-437.