Apparel As Communication

APPAREL AS COMMUNICATION

Rebecca H. Holman, Young and Rubicam, Inc.

[At the time that this paper was prepared the author was Assistant Professor of Marketing at The Pennsylvania State University. The research presented in this paper was made possible through a Faculty Projects Grant from the Center for Research, College of Business Administration, The Pennsylvania State University. The author acknowledges the helpful comments on an earlier draft on the paper made by Jerome Greenberg, Elizabeth C. Hirschman, Morris B. Holbrook, and Gerald Zaltman.]

ABSTRACT -

This paper discusses several attributes of how apparel serves as communicational signs. It presents a taxonomy of apparel, perspectives on its study, and suggestions for building and testing theories of apparel as communication. An empirical analysis of women's clothing use illustrates an approach to the topic.

INTRODUCTION

The human body is the primary medium for human interpersonal communication. And yet, communication researchers have for many years focused upon only one part of the body: the vocal apparatus and the messages produced there (i.e., upon human speech). With the recent popularity of nonverbal communication research, the means by which people communicate using other parts of their bodies has emerged as a viable topic for study. For example, a large group of researchers has elected to study human movements (called "kinesics") and another has looked at body position relative to other people and to objects (called "proxemics"). (See Harper, Wiens, and Matarazzo 1978 for a recent review of research in these two areas.) Although suggested by most who have speculated upon the domain of both verbal and nonverbal human communication with the body, (e.g., Duncan 1969; Barker and Collins 1970; Scheflen 1974; Knapp 1978) relatively few authors have actually studied how body adornment may serve as a form of communication. Such communicative use of products which dec,-,ra~e or enhance the bod%, is the topic of the current paper.

"Apparel" may be defined as the set of products (material artifacts) that adorn the body. (This definition is elaborated in a taxonomy of apparel developed later.) Not all apparel serves as communication and one purpose of the paper is to examine the necessary and sufficient conditions for such a use of apparel. Research investigating the communication function of apparel is reviewed briefly. The paper also illustrates one approach to research on apparel as communication by presenting an empirical study of women's clothing usage.

DEFINITION OF APPAREL

Apparel was defined as the set of products that enhance or adorn the body. The body appearance (including smell and tactile attributes) is temporarily changed through their use and this "parasomatic" [The word "parasomatic" was coined to refer to anything that modifies or is auxiliary to the body.] function may be one of the most important functions of apparel. Other functions include utilitarian (e.g., clothing as protection of the skin, a watch for time keeping) and aesthetic (e.g., wearing clothing or jewelry because it is beautiful. In addition, apparel can serve as a reminder of some activity previously experienced by the user or of some previous (and at times on-going) emotional state; both of these are referred to under the title "mnemonic function." Apparel may also be used as "emblems" or as "illustrators" (Ekman and Friesen 1969 quoted in Knapp 1978, pp. 13-h,), functions common to other nonverbal [Apparel is conceptualized here as primarily nonverbal although apparel can also convey verbal messages: brand names on clothing, words written on jewelry, jewelry which spells out names or initials, and words on Ishirts are all examples of verbal aspects of apparel use.] communication systems. [Ekman and Friesen (1969) also listed two other categories of nonverbal communication. These were regulators to mediate turn-taking in conservation, and adaptors a way of managing one's emotions. It is unlikely that apparel could serve either of these functions although other types of products are often used with adaptive behavior (Knapp 1978, pp. 13-18).] Each of these functions is discussed separately and illustrated in Table 1.

The Parasomatic Function of Apparel

Apparel which temporarily modifies or alters some attribute of the body performs a parasomatic function. The parasomatic function takes two different forms: camouflage and display. The body is camouflaged when its true nature is hidden (e.g., loose clothing covering the torso, deodorants supplanting actual body odors) or when its form is altered (e.g., cosmetics adding color to the skin, prostheses substituting for missing or inadequate body parts, girdles and corsets squeezing the flesh to make it appear more compact or rounded than it really is). [Body form can also be altered temporarily by products which are not items of apparel (e.g., products which bleach, curl, straighten, or remove the hair). Hie body can also be altered permanently through surgery (e.g., a face lift). These parasomatic activities are outside the domain of the present paper.]

The body is displayed when the viewer's attention is directed toward it or some part of it. This can occur when apparel reveals the body's dimensions through exposure (e.g., transparent or "revealing" clothing) or through boundary definition (e.g., tight-fitting clothing). Display also is the result when apparel draws attention because it is colorful, flashy, or contrasts with the background (e.g., rings and bracelets which direct attention to the user's hands and wrists). Direction of attention away from an unattractive body part is another use of apparel for display.

The Utilitarian Function of Apparel

Apparel often serves some rational or utilitarian purpose. "Supportive" underwear purports to prevent (or retard) sagging flesh; coats, gloves, scarves, and hats protect the skin from cold, wet air; a watch provides a convenient and portable clock. in the case of clothing, Laver (1969) has argued that the utilitarian properties are much less important than the social (communicative) ones; it seems likely that this is the case for many other categories of apparel, as well, although utilitarian properties should not be overlooked entirety.

TABLE 1

A TAXONOMY OF FUNCTIONS PERFORMED BY APPAREL

The Aesthetic Function of Apparel

Given a wide range of choices for apparel, it seems likely that the decisions about such things as color, fabric style, and ensemble configuration are made on the basis of "personal taste," the aesthetic dimension. Apparel is chosen often because it is itself beautiful, or pleasing in some non-visual way (i.e., olfactory, tactile and the user wishes to display that beauty, or because it improves the user's overall beauty or attractiveness through the parasomatic camouflage function mentioned above.

The Mnemonic Function of Apparel

Some pieces of apparel are worn for the purpose of reminding either the user or those observing the use about some event that occurred in the past or one enduring into the present. These may take on a "souvenir" quality (e.g., items purchased while visiting far-distant places, gifts commemorating various rites of passage) or may serve as affect display either for past emotional states (e.g., items celebrating a romantic love affair, paraphernalia symbolic of support for an athletic team) or for some event occurring in the past for which the emotional reaction, when remembered, is still strong (e.g., mourning costume, wedding rings). The characteristic linking these uses of apparel is that they trigger memories bringing certain thoughts into top-of-mind awareness.

The Emblematic Function of Apparel

A nonverbal sign acts as an emblem when the sign has a direct verbal translation. In common terms. this is what use of the sign "says" about the user. Apparel that belongs to a uniform (e.g., a fireman's clothing), apparel that is a material symbol of some group (e.g., a sorority or fraternity pin), or apparel that strongly suggests some personal characteristic of the user (e.g., apparel normally worn only by women) are examples of apparel's emblematic function.

The verbal translation of apparel used as emblems may either be denotative or connotative. A denotative translation occurs when the apparel has been formulated into a symbol, such as when a uniform is worn only by members of some organization or profession and when the details of that uniform are subject to strong normative control. (Military uniforms are one example; a policeman's badge is another.) Such translations span time, location, ana social ,roup within the larger culture as a function of their degree for formalization.

A connotative translation occurs when the apparel serves only as a sign not as a symbol. The nature of the translation is dependent upon the interpreter and if individuals come from different social systems, the translations made by them may be quite dissimilar. Clothing, for example, connotes social status (Veblen 1953), but as pointed out by Blumberg (1974) the particular pieces of clothing used to indicate a particular status change over time and also differ across social groups.

The Illustrative Function of Apparel

A nonverbal sign is an illustrator if it punctuates or accentuates verbal communication. An item of apparel can serve as an illustrator if its use can take on a dynamic quality. For many items of apparel (e.g., virtually all cosmetics, most clothing, most scents, and some hair decorations) the use is static: one either wears them or not, and they cannot normally be applied or removed during an interaction. When they can be changed during communication, they probably serve as very strong illustrators (e.g., a woman putting on lipstick, a person removing pieces of clothing or using a breath freshener), possibly because apparel is not typically used in this way.

An interesting subset of apparel serving an illustrative function is composed of those items that achieve their greatest significance when they are perceived in motion. Examples include fringe on clothing, "dangling" earrings, fabrics which accentuate walking movements of the body, "jangling" bracelets, and "squeaky" shoes. These items may or may not illustrate speech, but they definitely illustrate the kinesic (body movement) codes. They incorporate both visual and auditory stimuli and could conceivably incorporate olfactory cues as well (e.g., a hypothetical perfume in micro-fragrance capsules which dissolve when scratched or when the body temperature changes).

Requirements for Apparel To Serve As Communication

Two conditions are necessary for any behavior to serve as communication: the behavior must be visible or perceivable through some sensory apparatus, and the behavior must be variable (often within a narrowly-specified range of "acceptable" behavior). A further condition is necessary for apparel (and all products) to serve as communication: the use must be personalizable or linkable to the individual user. Each of these conditions is discussed separately below: together they form the sufficient conditions for apparel becoming communicative.

Apparel Use Must Be Visible

As pointed out by Robertson (1971) products (and items of apparel) differ in their degree of visibility in use. Apparel which is kept out of sight during most interactions (e.g., underwear) or which is not meant to be noticed (e.g., hair pieces) is low in visibility; apparel which is meant t be noticed (e.g., jewelry) is high in visibility. items of apparel that are low in visibility will not serve as communication for the simple reason that an observer cannot know whether those items are being used and attributions about the user cannot be formed on the basis of the use unless it is visible.

Items of apparel performing most parasomatic camouflage functions are low in visibility and are thus unlikely to be used a s communication. However, if the social system demands that these items be 'Kept hidden and they are inadvertently made visible (e.g., when a woman's slip is showing), then communication occurs, and it is likely that negative attributions are made. Those items of apparel that perform parasomatic display functions and those that are emblematic or illustrative are definitionally highly visible and probably serve as communication. (Apparel which is utilitarian, mnemonic, or aesthetic could vary in visibility and thus no generalizations are offered.)

Apparel Use Must Be Variable

If any form of human behavior is common, without variation, to all members of a group, then within that group the behavior is without significance ("everybody does it"). From outside the group, the behavior may identify group members, but it can do so only if those outside the group do not also engage in the same behavior without variation. In sum, behavior nas significance if it differs in some way among or within individuals. The variation it. behavior is communicative if there is a systematic relationship between the variation and characteristics of individuals.

Apparel use may be variable in terms of product form (e.g., color texture, or aroma) or in the way apparel ensembles are constructed (i.e., the choice of items to wear). Thus within certain groups there will be invariability in certain items of apparel (e.g., clothing uniforms, jewelry), but it is difficult to imagine no variation in apparel ensembles, even within very small groups. (Identical twins or any group of people who dress alike including use of fragrances and jewelry is a possible exception.)

For apparel to perform an emblematic function its use must be variable and systematically related to characteristics of users. Likewise those with common perceptions of beauty will presumably behave similarly toward apparel performing an aesthetic function. Similar generalizations about apparel that fulfills other functions may be possible after empirical research, but cannot be made at this point.

Apparel Use Must Be Personalizable

For some products, it is difficult to trace ownership to one individual (e.g., house furnishings for rooms shared by all family members). For others, the observer infers that the user did not have a choice about the items (e.g., actors in costume, others wearing uniforms), but must use them in order to keep the job. The amount of information conveyed in these cases about individual users is minimal. Only when the use is personalizable can anything other than the most general type of information be conveyed.

Most items of apparel are personalizable as they are worn by individuals and therefore "traceable" to one person. Even in the case of uniforms, the other items of apparel not covered by the uniform specification may effectively personalize a uniform. Likewise, the "condition" of the apparel (e.g., new, worn, soiled) can add information which personalizes. It is therefore unlikely that most types of apparel, excepting actors' stage costumes, can fail to meet this criterion for communicability, and even the stage costume conveys information about the character being portrayed.

RESEARCH ON APPAREL AS COMMUNICATION

Review of Previous Research

The empirical studies of apparel as communication conducted thus far have looked almost exclusively at the emblematic - connotative function. (Nineteen of these studies are summarized in Table 2.) As can be seen, Reynolds and Darden (1972), Coursey (1973), and Rosenfeld and Plax (1977) are exceptions. Reynolds and Darden attempted to explain why a style innovation in clothing was not adopted at that point in time. Coursey investigated the effects of wearing a uniform versus ordinary clothing upon perceptions of the user. Rosenfeld and Plax tied stated clothing behavior to personality traits. These and other functions of apparel were discussed by Levi-Strauss (1942) and Wax (1957) and it is unfortunate that they have not been studied more thoroughly.

Clothing has been the type of apparel most commonly studied (fifteen of the articles) although several researchers (e.g., Hamid 1968; Lefkowitz, Blake and Mouton 1955) have attempted manipulations that have more external validity by also including accessories and other apparel in their studies. The effects of cosmetics and glasses have been researched twice, but no empirical studies of the communicative effect of odor-manipulating products (e.g., perfume) have been reported.

TABLE 2

OVERVIEW OF EMPIRICAL STUDIES OF APPAREL AS COMMUNICATION

Three major types of attributions due to apparel have been studied in the past. (These attributions are conceived as the meaning connotated by the apparel use.) Social status of the user is a typical choice for researchers, perhaps because of the strong arguments of Veblen in the 1890's. Other researchers have elicited attributions about personal characteristics of the apparel user (referred to in Table 2 as personality). The third group has focused on how apparel conveys some form of social interaction attributes about the user (e.g., 'social ratings," Hoult 1954; political attitudes, Suedfeld, Bochner, and Matas 1971, Darley and Cooper 1972; and lifestyle choices, Belk 1980).

Multiplicity of Functions

While each of the functions performed by apparel has been treated separately here (both in Tables I and 2 and in the above review of research) it is clear that such a separation is a simplification for the purpose of illustration. The functions themselves may be independent, but one item of apparel may in fact perform multiple functions as an examination of the examples used in 'Fable I reveals. What is interesting from a research perspective are questions relating to the interactions of functions. Is there such a thing as a "purely" utilitarian item of apparel? If aesthetic considerations are present in virtually every item of apparel, then what is the relative salience of the aesthetic component in communication? Do aesthetically-pleasing apparel items make better illustrators' What is the communication effect when an item is aesthetically pleasing to the user but not to the individual observing the use? What are the connotations for "ugly" emblems? These and other research questions are suggested by a consideration of the interactions of functions and could prove a fruitful area for future research.

Other Opportunities for Future Research

From a theory-building perspective the literature briefly reviewed above suffers from some serious shortcomings. These shortcomings provide an opportunity for future research on apparel as communication. Three of these are discussed briefly here.

Programmatic research is rare. The ideals for theory building and testing set forth by Reynolds (1971), Box (19;6), and others specify a research program in which inductive research generates hypotheses which are tested deductively. This is then followed by more inductive Study to modify the theory and generate more deductively-testable hypotheses, and so forth. Such a program advances the goals of science by systematically explicating, testing, and refining theories until one eventually arrives at clear theoretical statements that have been validated empirically.

With the exception of the three studies by Hamid (1968; 1969; 1972) and the two by Belk (1978; 1980) programmatic research on apparel has not been conducted. In fact it often seems as if researchers duplicate one another's work, apparently without knowledge or intention, due perhaps to a failure to communicate across disciplines. Such research is inefficient as researchers continue to "reinvent the wheel" rediscovering old truths. It does little to extend theory on apparel as communication.

Researchers have often approached the identification of theoretical units casually. As Dubin (1978) points out, the specification of units and their relationships lire at the core of theory building. In studying any form of communication, one must be very careful to select units that reflect the system as it actually exists. This is problematical for two reasons. One is because the researcher is often not a member of the system being studied and may bias the results if meaningless units are selected for study. The other difficulty occurs when the researcher is a member of the system and has difficulty distinguishing elements of superficial structure tram elements of deeper structure. (See Boas 1911, on the latter point.) What this means is that the inductive phases of the research must be conducted very carefully or one risks results that have no true significance.

Most of the previous research on apparel has reported very vague or highly subjective procedures for selecting operational units of analysis. Representative of this is the study by Douty (1963) in which the researcher and stimulus person consulted on clothing to wear for the experiment; unfortunately the reader is not told what criteria were used for selecting the clothing. Worse still is the Hoult (1954) procedure: Hoult allowed stimulus-persons to choose their own clothing, merely instructing them to dress "better" for the second exposure to subjects than they did for the first.

The boundaries of theories have been vaguely or not at all specified. As indicated earlier, the degree to which individuals within a culture agree on the meaning of apparel serving an emblematic function is dependent upon the degree of formalization of the symbol (or sign-unit). Since most of the previous studies have not dealt with symbols but with signs, it is reasonable to believe that the relationships uncovered are bounded by time, space, social system, etc. With few exceptions, however, most researchers have failed to limit their theories or even to acknowledge that there might be limits. Boundary testing is one of the activities involved in sound programmatic research, and it is missing from the literature on apparel as communication.

Communicator Roles

Any researcher studying communication has the option of focusing upon one of three points of view: messages emitted by the decoder, messages received by the decoder, and messages generated by the encoder who keeps in mind the potential for decoding by a particular recipient. In other words, (1) one may focus upon the "actor's" expressions (to borrow the terms used by Goffman 1959), whether those messages are intended or not; (2) one may focus upon the "audience's" responses to the actor; or (3) one may limit oneself to investigation a[ only those messages that the actor intended to send with the objective being some form of influence upon members of the audience. (The latter is often given the name "purposive communication.,.) These three viewpoints will often yield different interpretations of the same phenomenon so one must carefully specify which is the research objective.

Perspectives on Research

[Each perspective is discussed in more detail in Holman (1980c) which also views communicative product use from each perspective.]

As pointed out by Fisher (1978), there are at least four different perspectives (the mechanistic, psychological, interactional, and pragmatic) for envisioning communication. Each perspective requires one to look at the phenomena of communication differently. The perspective identifies the locus of communication (where does communication occur) and ultimately shapes how one chooses units of theory and how one specifies the relationships among units.

The mechanistic perspective sees communication as being dependent upon the channel used for conveying messages. In a study of apparel as communication, one adopting the mechanistic perspective would assume that different messages were transmitted by different items of apparel. Research questions would focus on determining how a message was changed by using different items of apparel for conveying it independent of characteristics of user or interpreter.

Research from the psychological perspective focuses upon individual communicators, assuming that communication occurs in the minds of individuals. Apparel, from this perspective, would serve as one (of many) stimuli to which individuals could respond. Researchers who adopt this perspective would focus on individual differences in either the use or the interpretation of apparel as communication.

Interactionists assume that communication occurs only as individuals participate in the act of communicating. Apparel serves to define the relationships of interactants to one another and also may mediate in the modification of those relationships. Research on apparel from an interactional perspective would seek to identify how apparel specifies social positions or facilitates role performance.

The pragmatic perspective is a systems view of communication that places primary emphasis upon the overt behavior of communicators. Pragmatists would see apparel as a set of communication systems embedded in a larger s-stem of human communication. The pragmatist would be ma.inly interested in specifying the systematic properties of apparel use, 'now apparel interacts with other human communication systems, and how behavior is affected by apparel use.

Each of these four perspectives may be applied to study of apparel as communication. However, most researchers have adopted either the psychological (e.g., apparel reflects the user's psychological traits, Rosenfeld and Plax 1977) or interactional (e.g., apparel connotes social status, Wise 1974). Rarely has a purely mechanistic approach been taken, although Thornton (194!,), McKeachie (1952), and Reynolds and Darden (1972) are closest. Of the studies cited in Table 2, only Holman (1980a) has adopted a pragmatic perspective, combining it with both psychological and interactional ones. Clearly there are research opportunities for investigation of apparel from either the mechanistic or pragmatic perspective, as well as for creation of new perspectives through combination of these four. (See Fisher 1978, pp. 253-6 on this point).

A STUDY OF APPAREL AS COMMUNICATION

The research presented here illustrates an approach to study of apparel as communication which overcomes some of the difficulties with past research outlined above. It is a study of clothing and thus does little to broaden the study of apparel to other products less-frequently researched, but it is also a follow-up to a previous study (Holman 1976) thus initiating a program of research. Both of these studies are characterized by the following features:

a focus on decoding of messages by members of the audience;

a careful inductive stage in the research to identify clothing-units to serve as stimulus-objects for later deductive testing;

explicit statements of the social and time boundaries of the theory;

a pragmatic perspective containing interactional elements as well;

consideration of the emblematic - connotative function of apparel.

The study described here is an extension and improvement over the earlier one as the manipulation of a stimulus-object allows for unambiguous interpretation where the former study did not.

Methodology

The study consisted of a two-phase, inductive then deductive, design. The first phase was intended to identify the clothing ensemble units (types) worn by women students as they walked between buildings of the University Park campus of The Pennsylvania State University [Referred to as "the campus" throughout the rest of the paper.] during normal class hours for one week in October, 1977. The second phase was to examine a simplifying assumption used in the first phase, thus addressing the issue of the validity of the clothing units identified.

Phase I procedures. Full head-to-toes color photographs of women students who passed a number of high-traffic locations on the campus during one week in October, 1977, were obtained by the researcher and assistants. Greater than ninety-five percent of all women intercepted gave consent for the photography; only two percent refused for other than time-related reasons.

Clothing ensembles of these women were described using a 143-variable coding system that was a revision of the one created by Holman (1976). The system codes seven clothing attributes for twenty areas of the body and has three special codes (whether the clothing is bilaterally symmetrical, contains pants or skirt, and sleeves or a cape).

Two potentially-important features of clothing were not captured by this system: the color and design of the fabric from which the clothing was constructed. Although Compton (1962; 1964) had found relationships between these two attributes and personality traits, it was hypothesized here that these aesthetic features of clothing would be less important in interpersonal communication than would features of form. [The seven features of form recorded by the system were presence of cloth at each area, presence and shape of "attached" layers of cloth (e.g., pockets), fit of as many as three separable layers of cloth, degree to which the cloth made the body appear larger than is actually was, and presence and type of openings/ closing (e.g., zippers).] Addition of codes for color and fabric design would complicate the data transcription, hence this information was omitted as a simplifying procedure.

After data transcription, the ensembles were grouped into fourteen relatively-homogeneous clusters. Two clustering algorithms (Ward's method, Anderberg 1973; and Forgy's method, Forgy 1965) produced a solution that maximized within-group homogenity. Each clothing cluster obtained in this manner represents distinct clothing-ensemble types (or units) within the limitations of the notation system and the social-temporal clothing use system.

Phase II procedures. Phase II was designed to test the validity of the hypothesis that features of the cloth are less important in interpersonal communication than are features of clothing form. In order to do this, it would be necessary first to demonstrate that the ensemble-types identified in Phase I produced different communication effects from one another; this was accomplished and reported in Holman (1980b). Then, it would also be necessary to demonstrate that the addition of fabric information (either of color or fabric pattern) to a clothing ensemble produced an effect that was less than the effect due to a change in ensemble form. The following procedures were used as an exploratory test of the hypothesis.

A college-age woman [The model used was actually a student at Penn State at the time of the photography but transferred to another university for Fall term, 1978. She was perhaps more attractive than most students, but not noticeably glamourous.] was dressed in clothing typical of one of the clusters obtained in Phase 1. Her clothing consisted of the following pieces:

emerald green blouse with white pointed collar and front pocket band, set-in sleeves, average fit;

reddish-pink crew neck pull-over sweater, long set-in sleeves, loose fit;

long jeans, moderately flared at bottom, average fit through waist and upper thigh, zipper in front;

blue espadrilles, small heel, and a slightly thickened sole.

This woman was photographed on the campus as she walked past an easily-recognizable spot carrying books; care was taken to ensure that there were no people in the background, only buildings and plants. Black and white photography was used to minimize the effect of the above colors.

The woman was also photographed in an ensemble that was identical with the one described above except that the sweater described above was substituted for one like it but the latter's large (5-6 inches in size) counterpane color design. This design was readily noticeable in the black and white photographs.

Two of the photographs taken in these two ensembles were matched for posture (leg and body positions) and facial expressions. These served as stimulus-objects to be viewed by subjects who were students at The Pennsylvania State University in Fall, 1978.

Booklets containing one photograph and a questionnaire were randomly distributed in classes. The questionnaire elicited information about personal characteristics of the subject (e.g., age, sex) and also asked subjects to describe the woman in the photograph in terms of thirty-three attributional descriptors. Subjects were asked to pretend they had just seen the woman walking on campus and were asked for their first impressions of her. A six-point scale ranging from "definitely yes" to "definitely no" was available for subject responses. This procedure was similar to the one described by Holbrook and Hughes (1978). Informed consent [Prior to completing the questionnaire, subjects read a general description of the study which omitted any reference to clothing. This statement had been approved by a University committee as a reasonable procedure. Subjects were promised, and given, a thorough debriefing after all questionnaires were collected.] and completed questionnaires were obtained from forty-eight subjects, divided equally between treatments (the different photographs).

The complete list of attributional descriptors as presented to subjects (who were asked to complete the phrase "The woman in the photograph..."), is found in Table 3 along with the means and standard deviations for each attribute for each treatment. (All attribute scores were keyed so that 1 = definitely yes, 6 = definitely no.)

A multivariate analysis of variance (MANOVA) was performed to test the hypothesis of no difference in the attributions to the two stimulus objects across a linear combination of all attributes (Barr et al. 1976). This analysis produced an F = 3.14, d.f. = (33,10), p = .03 thus allowing rejection of the null hypothesis.

Next, t-tests were performed to determine which variables contributed to the significant F in the MANOVA results. Table 3 contains t-values and probabilities. As can be seen, only one attribute, "does not smoke cigarettes" produced a t significant at or beyond the .05 level.

The finding, that one out of thirty-three attributes is significant at or beyond the .05 level, has no practical significance since such a finding would be expected on the basis of chance alone. Thus it is possible to conclude that there were no genuine differences in the perceptions of the model as she was dressed in clothing ensembles which differed only in terms of the fabric pattern for a sweater.

Discussion

This study has demonstrated an approach to study of apparel as communication that could be generalized to study of. communicative properties of other products as well (c.f. Holbrook and Hughes 1978). Because the data report a comparison of only two clothing ensembles, essentially a test of a single clothing attribute, one would need to be very cautious in concluding that features of clothing fabric are unimportant in interpersonal communication. Yet within the limitations of the current study, that was the case. If nothing else, the study demonstrates an approach that may be taken in further studies of clothing, each of which could add empirical evidence to the hypothesis that fabric features are less important than form features in clothing as communication.

Further Research

Further research on clothing following that presented here would need to establish the generalizability of the finding that cloth pattern is of no significance in the context of first impression attributions about another. In order to do this it would be necessary to determine that the findings were consistent across the following possible sources of variation:

models (i.e., users of clothing)

items of clothing

fabric designs

social settings

observers of the clothing use (i.e., subjects)

temporal dimensions (both seasonal and in the context of fashion cycles)

facial expression and body posture variables.

A complex factorial design using the same structured projective technique used here could accomplish the goal of determining the boundaries of the theory relating clothing form, cloth pattern, and first impression attributions.

SUMMARY AND CONCLUSIONS

The paper has defined an area of study -- apparel as communication. It presented a taxonomy of apparel organized around the functions that apparel may serve. The paper discussed the necessary and sufficient conditions for apparel to become communicative and suggested types of apparel that would and would not satisfy these conditions.

A discussion of research on apparel as communication reviewed previous studies, characterized some of their weaknesses from a theory-building perspective, and identified four communication perspectives for building theories of apparel as communication. An empirical investigation of clothing as communication illustrated some of the features of research discussed previously. Although the results of that study were modest, they did suggest directions for further research on clothing.

TABLE 3

RESULTS OF T-TESTS COMPARING PATTERNED AND UNPATTERNED CLOTHING ENSEMBLES

Clearly the study of apparel as communication is in its infancy. As outlined in the current paper, there are vast gaps in the body of knowledge concerning the topic. It is surprising that more attention has not been directed at the area because of its impact upon first impression formation and management (e.g., Douty 1963) and upon the manipulations of visual imagery in advertising (see Rossiter and Percy 1979, for a discussion of this). Perhaps it is now time to study these product-based systems of communication with the same enthusiasm previously given the purely body-based systems.

REFERENCES

Michael Anderberg, Cluster Analysis for Applications (New York: Academic Press, 1973).

Larry L. Barker and Nancy B. Collins, "Nonverbal and Kinesic Research," in Methods of Research in Communication, ed., P. Emmert and W. W. Brooks (Boston: Houghton Mifflin, 1970).

Anthony J. Barr, James H. Goodnighl, John P. Sall, and Jane T. Helwig, A User's Guide to SAS 76 (Raleigh, NC: SAS Institute, Inc. 1976).

Russell W. Belk, "Assessing the Effects of Visible Consumption on Impression Formation," in Advances in Consumer Research, Vol. 5, ed., H. Keith Hunt (Ann Arbor: Association for Consumer Research, 1978).

Russell W. Belk, "Effects of Consistency of Visible Consumption Patterns on Impression Formation," in Advances in Consumer Research, Vol. 7, ed., Jerry C. Olson (Ann Arbor: Association for Consumer Research, 1980).

Leonard Bickman, "The Effect of Social Status on the Honesty of Others," The Journal of Social Psychology, 85 (1971), 87-92.

Paul Blumberg, "The Decline and Fall of the Status Symbol: Some Thoughts on Status in a Post-Industrial Society," Social Problems, 21(1974), 4?0-98.

Franz Boas, Introduction to the Handbook of American Indian Languages (Washington: Bureau of American Ethnology Bulletin No. 40 (1908) Part 1, 1911).

George E. P. Box, "Science and Statistics," Journal of the American Statistical Association, 81(1976), 791-9.

Norma H. Compton, "Personal Attributes of Color and Design Preferences in Clothing Fabrics," The Journal of Psychology, 54(1962), 191-5.

Norma H. Compton, " Body-Image Boundaries in Relation to Clothing Fabric and Design Preferences of a Group of Hospitalized Psychotic Women," Journal of Home Economics, 56(1964), 410-5.

Robert D. Coursey, "Clothes Doth Make the Man, in the Eye of the Beholder," Perceptual and Motor Skills, 36 (1973), 1259-64.

John M. Darley and Joel Cooper, "The 'Clean for Gene' Phenomenon: The Effect of Students' Appearance on Political Campaigning," Journal of Applied Social Psychology, 2(1972), 24-33.

Helen I. Douty, "Influence of Clothing on Perception of Persons," Journal of Home Economics, 55(1963), 197-202.

Robert Dubin, Theory Building (revised ed.)(New York: The Free Press, 1978).

Starkey Duncan, Jr., "Nonverbal Communication," Psyshological Bulletin, 72(1969), 118-37.

Paul Ekman and W. V. Friesen, "The Repertoire of Nonverbal Behavior: Categories, Orgins, Usage, and Coding," Semiotica, 1(1969), 49-98.

B. Aubrey Fisher, Perspectives on Human Communication 'Ne~ York: Macmillan Publishing Company, Inc., 1978).

E. W. Forgy, Cluster Analysis of Multivariate Data (Riverside, CA: Biometric Society Meetings, 1965).

Frying Goffman, The Presentation of Self in Everyday Life (Garden City, NY: Doubleday, 1959).

Paul N. Hamid, "Style of Dress as a Perceptual Cue in Impression Formation," Perceptual and Motor Skills. 26 (1968), 904-6.

Paul N. Hamid, "Changes in Person Perception As a Function of Dress," Perceptual and Motor Skills, 29(1969), 191-4.

Paul N. Hamid, "Some Effects of Dress Cues on Observational Accuracy, a Perceptual Estimate, and Impression Formation," Journal of Social Psychology, 86(1972), 279-89.

Robert G. Harper, Arthur N. Wiens, and Joseph D. Matarazzo, Nonverbal Communication: The State of the Art (New York: John Wiley & Sons, Inc., 1978).

Morris B. Holbrook and Nevill C. Hughes, "Product Images: How Structured Rating Scales Facilitate Using a Projective Technique in Hypothesis Testing," The Journal of Psychology, 100(1978), 323-8.

Rebecca H. Holman, "Communicational Properties of Women's Clothing: Isolation of Discriminable Clothing Ensembles and Identification of Attributions Made to One Person ,~earing Each Ensemble," unpublished doctoral dissertation (Austin: The University of 'Texas at Austin, 1976).

Rebecca H. Holman, "Clothing as Communication: An Empirical Investigation," in Advances in Consumer Research, Vol. 7, ed., Jerry C. Olson (Ann Arbor: Association for Consumer Research, 1980a).

Rebecca H. Holman, "A Methodology for Studying Product Communication Systems," unpublished manuscript (University Park, PA: The Pennsylvania State University, 1980b).

Rebecca H. Holman, "Product Use as Communication: Perspectives on the Neglected Systems," paper presented at the XXX International Conference on Communication, Human Evolution, and Development, Acapulco, Mexico, 1980c.

Thomas Ford Hoult, "Experimental Measurement of Clothing as a Factor in Some Social Ratings of Selected American Men," American Sociological Review, 19(1954), 324-8.

Mark L. Knapp, Nonverbal Communication in Human Interaction (2nd ed.) (New York: Holt, Rinehart and Winston, 1978).

Thomas E. Lasswell and Peter F. Parshall, "The Perception of Social Class from Photographs," Sociology and Social Research, 45(1961), 407-14.

James Laver, Modesty in Dress (Boston: Houghton Mifflin, 1969).

Monroe Lefkowitz, Robert R. Blake, and Jane Srygley Nouton, "Status Factors in Pedestrian Violation of Traffic Journal of Abnormal and Social Psychology, 51 (1955), 704-6.

Claude Levi-Strauss, "Indian Cosmetics," V.V.V., 1(1942), 33-5.

W. J. McKeachie, "Lipstick as a Determiner of First Impressions of Personality: An Experiment for the General Psychology Course," The Journal of Social Psychology, 36 (1952), 241-4.

Fred D. Reynolds and William R. Darden, "Why the Midi Failed," Journal of Advertising Research, 12(1972), 39-44.

Paul Davidson Reynolds, A Primer in Theory Construction (Indianapolis: The Bobbs-Merrill Company, Inc., 1971).

Thomas S. Robertson, Consumer Behavior (Glenview, IL: Scott, Foresman and Company, 1970).

Lawrence B. Rosenfeld and Timothy G. Plax, "Clothing as Communication," Journal of Communication, 27(1977), 24-31.

John R. Rossiter and Larry Percy, "Visual Communication in Advertising," Research Working Paper No. 231A (New York: Columbia University Graduate School of Business, 1979).

Albert E. Scheflen, How Behavior Means (New York: Anchor Press, 1974).

Peter Suedfeld, Stephen Bochner and Carol Matas, "Petitioner's Attire and Petition Signing by Peace Demonstrators: A Field Experiment," Journal of Applied Social Psychology, 1(1971), 278-83.

G.F. Thornton, "The Effect of Wearing Glasses Upon Judgments of Personality Traits of Persons Seen Briefly," Journal of Applied Psychology, 28(1944), 203-7.

Thorstein Veblen, The Theory of the Leisure Class (New York: The New American Library, 1953).

Murray Wax, "Themes in Cosmetics and Grooming," American Journal of Sociology, 62(1957), 588-93.

Gordon L. Wise, "Differential Pricing and Treatment by New-Car Salesman: The Effect of the Prospect's Race, Sex, and Dress," Journal of Business, 47(1974), 218-30.