Some Preliminary Notes on Research in Consumer Esthetics

Morris B. Holbrook, Columbia University
ABSTRACT - These introductory comments delineate the area of consumer esthetics and suggest some theoretical developments, methodological issues, and marketing applications that have characterized work in this field to date. The practical importance of this research area is emphasized, with a partial explanation offered for its otherwise surprising past neglect. Arguments are advanced for the needed development of theory in consumer esthetics. Several methodological hurdles are discussed. And the nature of marketing applications--as illustrated by contributions to this session--is described.
[ to cite ]:
Morris B. Holbrook (1980) ,"Some Preliminary Notes on Research in Consumer Esthetics", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 104-108.

Advances in Consumer Research Volume 7, 1980     Pages 104-108

SOME PRELIMINARY NOTES ON RESEARCH IN CONSUMER ESTHETICS

Morris B. Holbrook, Columbia University

[The author gratefully acknowledges the support of Columbia University's Faculty Research Fund in organizing this session and preparing these introductory remarks. This paper is itself a highly condensed synopsis of a much more extensive and detailed review currently being undertaken by the author. Because of the resulting degree of elision and abbreviation, explicit references are made only to the papers contained in this session. Comprehensive references and a fuller consideration of the issues raised herein will appear in the final complete review.]

ABSTRACT -

These introductory comments delineate the area of consumer esthetics and suggest some theoretical developments, methodological issues, and marketing applications that have characterized work in this field to date. The practical importance of this research area is emphasized, with a partial explanation offered for its otherwise surprising past neglect. Arguments are advanced for the needed development of theory in consumer esthetics. Several methodological hurdles are discussed. And the nature of marketing applications--as illustrated by contributions to this session--is described.

INTRODUCTION

Definition of Consumer Esthetics

For purposes of the present discussion, consumer esthetics may be defined as the study of the buyer's cognitive, affective, and behavioral responses to media, entertainment, and the arts. Though this is an intentionally broad definition, it does confine attention to those products likely to prompt a uniquely "esthetic" experience. The phenomenological nature of such an experience is, of course, a matter for debate. There appears to be some consensus among philosophers devoted to the subject, however, that esthetic experience involves attending to, perceiving, and appreciating an object-for-itself, without regard to whatever utilitarian function it might perform. Viewed in this light, virtually any product--for example, a chair or a bottle of wine--could be regarded as an artistic object and appreciated accordingly. In this sense, all merchandise involves certain esthetic aspects in its design, packaging, and promotion. Nevertheless, some products--such as Beverly Sills performances, Picasso paintings, Shakespearean sonnets, and Paul Desmond recordings--seem to exist almost solely for the sake of being enjoyed in their own right, as objects-for-themselves. Only in perverse cases could we imagine them being used for other purposes like drowning out the neighbor's dog, serving as an impromptu umbrella, wallpapering the den, or keeping the potted plant from leaving stains on the coffee table. Accordingly, while recognizing the possible significance of such nonesthetic uses of artistic objects and the clear importance of esthetic aspects of otherwise utilitarian products, the present discussion focuses on those media, entertainment, and artistic offerings that can be regarded as primarily esthetic in nature.

Notice that--though the difference has sometimes interested sociologists--no attempt is made here to distinguish between "mass" culture and "high" art. Besides their obvious elitism, such judgments of intrinsic worth are seldom crucial to whether some object prompts an esthetic response as defined above. Thus, the economic destiny of an artistic offering depends less on its ability to win acceptance by art critics than on its degree of appeal to the tastes of consumers (however "popular" in kind) that make themselves felt in the market place.

The Paradoxical Neglect of Research on Consumer Esthetics

The recent appearance of a rock band on the cover of Fortune reminds us that, however defined, the esthetic activities of consumers exert a major impact on the American economy. By any reasonable estimate, the (roughly) $25 billion "esthetics" industry is a vast economic enterprise, with annual consumer expenditures rivaling or exceeding those for such products as household appliances ($11 billion), furniture ($20 billion), shoes ($15 billion), tobacco ($20 billion), liquor ($13 billion), or other alcoholic beverages ($19 billion). Certainly, marketing researchers have devoted a large amount of attention to studies on the consumption of beer, cigarettes, appliances, furniture, clothing, and other manufactured goods. Yet, despite its obvious importance, there has been comparatively little published research on consumer esthetics. Even if the present session offers some hope that the tide may be turning, one must still raise the question of what has caused this historical neglect of consumer esthetics research.

Though any attempt to address this question must necessarily be speculative, the author believes that a partial answer may be found in the attitude of many arts administrators toward their profession. As adumbrated by various marketing commentators, the logic of arts management appears to run something like this: (1) creative integrity requires the artist or artistic director to be essentially product-oriented; (2) the resulting artistic offering will tend to fail in the market place; (3) such activity thus requires and receives government support; (4) research on consumer esthetics is therefore fairly irrelevant to the success of artistic enterprises.

Though glib, such an argument is obviously riddled with fallacies. First, some artists are strongly afflicted with the profit motive--as indicated in a recent remark by a well-known jazz-rock musician: "Don't let your taste get in the way of reaching a broader audience" (Newsweek 4/2/79). But even the intensely self-directed artist can create more intelligently the better he understands the nature of the receiver's response. If one's purpose is to astound and repel one's audience, then research on esthetic processes can shed light on how to accomplish exactly that. Moreover, once an artistic product (however inaccessible) has been created, consumer research can help direct it toward the market segment most likely to appreciate it. Such an effective marketing strategy is essential since the government, in fact, spends only about $200 million on the arts each year. This paltry support for artistic endeavors is a mere drop in the economic bucket (less than 1% of overall esthetic expenditures), and even this limited aid is often made contingent upon demonstrable market acceptance in the form of attendance figures or other indicants of audience involvement. In such circumstances, it follows that research on consumer esthetics attains an importance bordering on urgency and cannot reasonably be eschewed on the strength of any elitist rationale such as that paraphrased above.

This realization was the original impetus for organizing the present session on consumer esthetics. We have been fortunate in engaging the participation of a number of informed and enthusiastic contributors. Their papers, together with the author's own efforts in preparing a review of the area, have raised a number of salient issues vital to the advancement of this field of inquiry. These will be discussed briefly, with references to the session papers where appropriate, in the remainder of these introductory comments.

THEORY

Thus far, consumer researchers have developed little if any theory to deal explicitly with the case of esthetic behavior. Not surprisingly, this theoretical vacuum is reflected by the papers in the present session. With one partial exception (Huber and Holbrook), none deals with the conceptual underpinnings of consumer esthetics. Here as elsewhere, then, consumer researchers--far from winning the race toward esthetic theory building--have not even left the conceptual starting gate.

Apparently, what the area needs most, as a foundation for theory development, is a place to begin. Possible directions for this beginning are suggested by the matrix in Table 1, which arrays the semiotic aspects of stimulus signs against the types of consumer responses mentioned earlier.

TABLE 1

POSSIBLE DIRECTIONS FOR THEORY DEVELOPMENT IN CONSUMER ESTHETICS

While some work in other disciplines has been and will continue to be done in each of the nine areas of intersection, it appears to the author that the most extensive and precise theory development has occurred in psychological studies of affective responses to the syntactic properties of artistic stimuli (i.e., the way that signs relate to each other in forming patterns and other complex interactions). Thus, experimental esthetics--as developed and advanced by the late Daniel Berlyne and his colleagues--has focused primarily on the checked cell of the matrix at the intersection of syntactic aspects and affective responses. The question raised by this cell translates directly into an important marketing issue--namely, how to structure the artistic offering so as to prompt the optimal esthetic response. Because this focus has been best developed by psychologists, it appears to be the most reasonable starting point for the investigation of consumer esthetics.

Types of Esthetic Response

Philosophical introspection and limited discussion by psychologists suggest that there are actually two qualitatively different levels of esthetic response of interest. The first, which might be called "hedonic value," refers to the diffuse sense of pleasure or generalized enjoyment that one feels when looking at an attractive painting, watching a well-photographed film, or listening to some "catchy" music. By contrast, the second level, which might be called "profound experience," deals with the feeling of being deeply moved or exultant with a spine-tingling sensation of almost sexual intensity and has, except in the work of Abraham Maslow, been studiously avoided by psychologists--doubtless because of the embarrassing measurement problems it raises.

The Evolution of Esthetic Theory

The development of theory to explain these phenomena appears to have progressed from two extremes toward the middle. On one side, beginning with the Greek ideals of structural order and continuing into the modern corollaries posed by Gestalt psychology, the principle of unity has argued that the value of art lies in its imposition of coherence on otherwise chaotic material. Conversely, students of cognitive motivation have suggested that the variety-seeking tendency of humans causes them to value complexity or so-called "collative" properties so that processes described by information theory or, later, Chomskian generative grammar can best account for the determinants of esthetic pleasure.

Increasingly, both sides of the debate have yielded to a more balanced view recognizing that structure and complexity are merely two sides of the same artistic coin (as formalized, for example, in the inverse relationship between information-theoretic redundancy and uncertainty). The resulting integrative models can be addressed to each of the two levels of esthetic response mentioned earlier.

Case 1

Case 1 concerns the determinants of hedonic value and argues that neither extreme of rigid organization or total chaos is pleasing, but that maximal enjoyment occurs somewhere in between at some optimal level of complexity. This hypothesized nonmonotonic relationship is represented by Berlyne's "Wundt curve," sketched in Figure 1 and discussed further in the paper by Huber and Holbrook.

FIGURE 1

CASE 1: THE BERLYNIAN "WUNDT CURVE" RELATING COMPLEXITY TO HEDONIC VALUE

FIGURE 2

CASE 2: HYPOTHESIZED PROCESS DETERMINING PROFOUND ESTHETIC EXPERIENCE

Case 2

Case 2 deals with the more profound type of esthetic experience. Here the theorist is on thinner ice--some would say quicksand--because of the aforementioned historical neglect of this topic by psychologists. However, based on Berlyne's concept of the "arousal jag" and on penetrating discussions offered by estheticians such as Leonard Meyer, one might speculate that Case 2 involves a dialectic process, unfolding over time, as diagrammed in Figure 2. The hypothesized process begins with an awareness of structure involving the formation of probabilistic expectations (thereby increasing information-theoretic redundancy). Such expectations are violated, however, by the occurrence of some deviation that produces surprise (ex post uncertainty in information-theoretic terms). The resulting arousal motivates the reinterpretation of the surprising event and its cognitive integration into a newly perceived, larger pattern that, in turn, becomes the structural framework for the next cycle of the esthetic process (via the feedback loop). The corresponding reduction in arousal is so satisfying as to create a deeply felt emotional response (via the "arousal jag").

METHODOLOGY

As in other areas of marketing research, studies on consumer esthetics have employed both field survey techniques and experimental laboratory designs. Because methodological problems differ noticeably between the two approaches, they will be discussed separately.

Methodological Problems in Esthetic Surveys

Field research on consumer esthetics may be further divided into two sub-types: (1) audience studies wherein those attending an event are questioned concerning related characteristics, attitudes, and behaviors (Semenik and Young; Sexton and Britney); and (2) cross-sectional surveys in which the sample is drawn from a wider population of potential respondents, such as the residents of a particular market area (Belk and Andreasen).

Both approaches have tended to measure consumer responses in terms of simple behavioral criteria like claimed attendance rates or subscription holding (Belk and Andreas-en; Semenik and Young; Sexton and Britney). Clearly, however, such measures do not form direct links to the esthetic models intended to describe Cases 1 and 2.

As indicated by Semenik and Young, much audience research has been excessively "descriptive" in nature--simply delineating the demographic, socioeconomic, and/or psycho-graphic profile of some particular set of arts patrons and almost invariably finding them more "upscale" than the general population. Such studies may be of interest for purposes of market segmentation, but tell us little about the underlying dynamics of esthetic behavior. Moreover, audience studies suffer from the problem of respondent self-selection since they sample only those who already attend arts events. This difficulty suggests that---where one's purpose is to investigate the underlying determinants of esthetic activity--the cross-sectional approach, while much more expensive, may be better suited to the researcher' s objectives.

Many audience and cross-sectional esthetic surveys have employed multidimensional scaling, multiple regression, discriminant analysis, factor analysis, cluster analysis, or other multivariate techniques typical of consumer research (e.g., Sexton and Britney). Often, however, survey studies have clung to a somewhat primitive bivariate approach--examining relationships one variable at a time in a rather ad hoc fashion. Semenik and Young make a case for such a procedure in their own research. Elsewhere, however, the use of complex multivariate techniques appears essential if one considers the nature of the marketing problem in media, entertainment, and the arts. Unlike the manufacturer of coffee or beer, who typically offers one or two brands that vie for the consumer's purchasing dollar, the TV network or record company or performing arts center tends to produce a large number of esthetic offerings with the hope that the consumer will buy several or, better yet, the whole lot. This situation is best modeled by techniques that portray preference relationships among a large set of esthetic objects simultaneously. The author, for example, has used multidimensional scaling of interobject preference correlations to create evaluative spaces containing as many as 180 entertainers whose relative positions indicate which are likely to be preferred together and by what kind of consumer. If dealing with large numbers of objects is especially desirable in esthetic research, one can take some comfort that judgments provided by consumer respondents are likely to be more valid here than in the typical case. It might be questioned, for example, whether respondents can provide meaningful attribute ratings or similarity judgments on even as many as four or five brands of toothpaste, cigarettes, or soap. These are low-involvement products and may not prompt enough search and brand familiarity to permit the kinds of informed comparisons required by many marketing research paradigms. By contrast, esthetic activities are central to the lives of most consumers and are subject to high levels of ego-involvement. Respondents familiar with a particular class of media, entertainment, or art offering are likely to be capable of rating, judging, or evaluating large numbers of specific examples of that esthetic category. Thus, successful survey research has been performed on broad arrays of television shows, popular singers, and jazz musicians. Little difficulty would be expected in extending such studies to movies, composers, painters, novels, or other esthetic products.

Methodological Problems in Experimental Esthetics

In laboratory experiments, processes determining esthetic perceptions and preferences can be examined in considerably more depth (though purchase behavior is, of course, concomitantly more difficult to investigate). Here, the most troublesome methodological problems concern measures of the syntactic properties of the stimuli used and the affective nature of the responses obtained.

Many measures of esthetic response have been used--with mixed success. For example, psychophysiological reactions (e.g., GSR, EKG, EEG, or pupil dilation) and behavioral measures (e.g., exposure choice or looking time) have been beclouded by ambiguous interpretations, though the growing consensus is that they indicate attention, arousal, or orienting response rather than positive affect or pleasure. Accordingly, psychological experimenters have increasingly reverted to simple verbal measures such as the evaluative component of semantic differential scales like pleasing/displeasing, beautiful/ugly, and good/bad.

In developing measures of stimulus complexity, considerable controversy has occurred over whether they should be objective or subjective, verbal or behavioral. Strictly speaking, information theory applies to the objective stochastic properties of a message generator so that complexity should be manipulated or measured by the generative rules used to construct the stimulus display. Clearly, however, the psychological effects of the display depend upon its subjectively perceived complexity. It follows that measures of subjective complexity may legitimately be used in experimental studies of esthetic responses--a ploy that has the additional benefit of saving information-theoretic measures from Chomsky's attack on Markovian communication models. Such measures of subjective complexity or uncertainty may be based upon either behavioral or verbal responses. The behavioral approach asks subjects to guess at missing parts of the stimulus display and uses the relative frequencies of their responses as measures of subjective probability in the information-theoretic uncertainty formulae. Such behaviorally-based indices of subjective complexity have been employed successfully in the cases of prose and graphic displays, but appear almost insurmountably difficult to apply in the case of real music. By contrast, the verbal approach simply collects ratings of the stimulus object along various scales such as simple/complex, certain/uncertain, and orderly/disorderly. Since such verbal measures are by far the easiest to use, it is fortunate for researchers in consumer esthetics that the various indices of complexity (objective vs. subjective, behavioral vs. verbal) appear to possess strong convergent validity, showing high correlations between measures.

Given adequate measures of stimulus complexity and affective response, appropriate procedures for investigating Case 1 might seem relatively straightforward. One could simply manipulate complexity, for example, and assess its effect on hedonic value. Unfortunately, however, a large number of such experiments--including that of Huber and Holbrook--has not yet succeeded in establishing the validity of the aforementioned nonmonotonic Wundt curve. In part, this discouraging state of affairs has been confused and distorted by the fact that, by judiciously interpreting the range of the independent variable, the hypothesized inverted U-shaped curve is compatible with virtually any data outcome (except, of course, a strict U-shaped relation). Indeed, Huber and Holbrook engage conspicuously in such a "judicious interpretation'' to account for their otherwise puzzling findings.

Perhaps reflecting this difficulty in establish/rig even the most basic relationship postulated in Case 1, experimental esthetics has relied more and more on powerful multivariate techniques--such as factor analysis, discriminant analysis, and multidimensional scaling--used to represent stimulus perceptions. Note that, in such applications of }fi)S, the requirement for pairwise similarity judgments may impose enormous demands on the subject's time in the cases of literature, music, or other art forms that unfold sequentially. Probably for this reason, the study by Huber and Holbrook appears to have been the first to collect pairwise similarity judgments on excerpts of real music. Encouragingly, their finding that discriminant analysis and MDS produced closely comparable perceptual spaces should prove reassuring to investigators wishing to economize on time by substituting compositional for decompositional methods.

In further attempts to uncover more systematic relationships, experimental research on Case 1 has turned increasingly to the incorporation of situational variables such as social effects, emotional states, environmental distractions, and learning stages. For example, Huber and Holbrook examined the effect of artistic knowledge and training on esthetic preference patterns, but were unsuccessful in establishing such differences between groups of subjects.

When one turns from Case 1 to Case 2, the situation is even more discouraging. The author is aware of no empirical research--in this session or elsewhere--that has addressed the problem of tracing the dialectic process that culminates in the kind of profound esthetic experience described earlier. Yet this is the most powerful esthetic phenomenon of all, the most tied to questions of beauty and greatness, and the most deserving of serious, dedicated study. Though it is probably the most intractable area of consumer esthetics research, it may also be the most likely to yield big payoffs in the long run.

MARKETING APPLICATIONS

Marketing applications of research on consumer esthetics appear to fall into two basic types. The first is intended to facilitate design of the artistic product or offering and assumes an artist, creative manager, or program director who is willing to orient his work toward satisfying the tastes of some target segment. By contrast, the second assumes esthetic offerings whose parameters are already fixed (by creative integrity, artistic stubbornness, or otherwise) and addresses the problem of finding and reaching the target segment(s) most likely to respond favorably. In both cases, applications may be further subdivided into survey and laboratory approaches, thereby creating the typology shown in Table 2 and the resulting classification of papers presented in this session.

TABLE 2

TYPOLOGY OF MARKETING APPLICATIONS FOR RESEARCH ON CONSUMER ESTHETICS

The order of presentation will follow the structure of the typology, as indicated by the parenthetical numbers. To preview briefly, Belk and Andreasen use a cross-sectional survey questionnaire to address the issue of what modifications in the design of theater and symphony offerings would be most effective in selectively broadening attendance and conclude, for example, that symphony audiences might be thus increased by the introduction of more famous conductors and performers.

Though not specifically couched as a segmentation study, the paper by Semenik and Young examines attitudinal and behavioral correlates of opera subscription and attendance. It therefore appears most useful as a basis for directing advertising and other promotional efforts toward particular market segments (such as the cross-arts subscriber). Sexton and Britney present a more conventional approach to market segmentation based on multivariate techniques. Opera, ballet, and symphony audiences are clustered on the basis of arts-attendance patterns, and the contrasting demographic and information-exposure characteristics of these clusters suggest ways of directing promotional efforts at various possible target segments.

The Huber-Holbrook study is somewhat more difficult to classify. Certainly its purpose was not to make unsolicited and probably unwanted recommendations on stylistic changes that might broaden the commercial appeal of various jazz saxophonists. Yet its unsuccessful attempt to isolate preference differences among groups of subjects suggests that--at least for this restricted sample--meaningful segments do not appear to exist on the basis of jazz knowledge, musical training, or listening experience.

DISCUSSION

Based on the foregoing comments, it would clearly be premature to attempt to draw firm conclusions concerning the future course of research on consumer esthetics. Throughout, critical opinions have been offered on various key issues, and there is no need to recapitulate these here in what is already a lengthy set of introductory remarks.

There may be a need, however, to emphasize the keen sense of excitement with which the author and the session participants have approached the uncharted but inviting waters of research on consumer esthetics. For the issues raised in this area seem of genuine importance. The potential rewards to marketers appear enormous. And one senses that, by investigating this intriguing aspect of consumer behavior, the researcher moves a small step closer to understanding the elusive connection between marketing and basic human values.

----------------------------------------