Symbolism and Technology As Sources For the Generation of Innovations

Elizabeth C. Hirschman, New York University
ABSTRACT - This paper has proposed that product innovations may arise from two independent sources - symbolism (intangible attributes) and technology (tangible attributes). Symbolic innovations are those which result from the reassignment of social meaning to an existing product, generating a secondary diffusion for it among those identifying with the relevant reference group. Technological innovations are those that spring from the addition or alteration of tangible features in a product that serve to distinguish it from prior motels. Innovations may arise in given product classes from either or both of these two sources. It is argued that symbolic innovations will diffuse primarily due to their association with a given reference group; whereas technological innovations will diffuse primarily as a result of consumer perceptions of and need for their superior performance.
[ to cite ]:
Elizabeth C. Hirschman (1982) ,"Symbolism and Technology As Sources For the Generation of Innovations", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 537-541.

Advances in Consumer Research Volume 9, 1982      Pages 537-541


Elizabeth C. Hirschman, New York University


This paper has proposed that product innovations may arise from two independent sources - symbolism (intangible attributes) and technology (tangible attributes). Symbolic innovations are those which result from the reassignment of social meaning to an existing product, generating a secondary diffusion for it among those identifying with the relevant reference group. Technological innovations are those that spring from the addition or alteration of tangible features in a product that serve to distinguish it from prior motels. Innovations may arise in given product classes from either or both of these two sources. It is argued that symbolic innovations will diffuse primarily due to their association with a given reference group; whereas technological innovations will diffuse primarily as a result of consumer perceptions of and need for their superior performance.


The diffusion of product innovations has generally been described using a paradigm originating in rural sociology (Ryan and Gross 1943) and applied in such diverse fields as medicine, communications, anthropology and education. This model of the diffusion process at present dominates the design and conceptualization of the majority of research conducted on new product diffusion and adoption (Rogers and Shoemaker 1971, Robertson 1971).

Research conducted using the traditional innovation diffusion model has centered on characteristics of two primary entities in the diffusion process - adopters of innovations and the innovations, themselves. For example, innovation adopters are typically classified according to the time they adopted an innovation as measured: (1) from the innovation's introduction into their social system or (2) from their first learning of the innovation (see Midgley and Dowling 1978, Wallendorf 1979, Zaltman and Wallendorf 1979 for recent summaries of this research). Typically, an 'innovator', 'early adopter', 'early majority','late majority', 'laggard' categorization scheme is applied to adopters based on the rapidity of their adoption (Rogers and Shoemaker 1971).

The characteristics of innovations that affect their diffusion have received a similar level of attention (c. f. Robertson 1971, Rogers and Shoemaker 1971, Zaltman and Wallendorf 1979). The attributes of innovations identified in research as possessing importance for their potential diffusion include: (a) relative advantage, (b) compatibility, (c) complexity, (d) trialability and (e) observability (Rogers and Shoemaker 1971).

Despite the utility of these two research streams for increasing our understanding of the diffusion process, there is a third aspect of innovation adoption and diffusion that has received considerably less attention in the literature but which may possess large importance. This aspect is the dimension(s) along which innovations are generated within given product classes. In other words, a valuable path for investigation may be how innovations are created for a given type of product; what type or source of change from pre-existing forms is required in a later form for it to be considered new or innovative.

Robertson (1971) has addressed this issue by conceptualizing a continuum of innovations from "radically discontinuous" to "continuous". The former term designates an innovation differing from earlier forms in several relevant features (attributes); while the latter term describes innovations possessing a majority of features in common with earlier products in addition to some novel features. The central thesis of Robertson's approach is that innovations may be conceived as more or less innovative depending on the proportion of features they share in common with earlier motels. This conceptualization is somewhat analogous to Tversky's (1977) propositions concerning similarity and dissimilarity of objects as measured by their congruent and noncongruent attributes. In this perspective, innovations are considered more, or "incrementally", innovative to the extent that they possess attributes not isomorphic with prior models.

An alternative way of approaching this same issue is to focus not on the proportion of common or noncommon attributes that innovations possess with earlier products, but on the dimensions along which those attributes are added to the product. It is posited that product innovations are generated primarily along two major dimensions: (1) symbolism and (2) technology; and that innovations arising from one or the other of these dimensions possess fundamentally different properties and diffuse according to fundamentally different principles. These two dimensions for generating innovations are described below as dichotomous "ideal types". Of course, in actuality both dimensions are continuous and may be interrelated for some product classes.


A symbol is an entity that stands in place of or represents another entity (Smelser 1972). An innovation that is generated primarily through symbolic changes is one which communicates a different social meaning than it did previously. Its physical form remains predominately unchanged, but the meaning assigned to that form is novel. Stated in cognitive terminology, a symbolic innovation is one that possesses different intangible attributes than it did in a previous stage. An intangible attribute is one which is associated with the object by consumers, but which does not arise from the physical nature of the object itself (Hirschman 1980, 1981a, 1981b). For example: sexiness, conservatism, and prestige are intangible attributes that may be associated with products.

That products may serve as symbols has been an accepted proposition of behavioral research for many years (Barber and Lobel 1954) and was perhaps first extensively examined within marketing under the auspices of "motivation research" in the 1950's. Despite the implicit understanding of the socially symbolic role played by many products, the influence this may have on their adoption and diffusion has been rarely incorporated in innovation research. The value of using such a perspective becomes apparent if one views symbolic innovations as communicative devices representative of different lifestyles. These lifestyles may serve, in effect, as reference groups for the adopting consumer. Therefore, the consumption of symbolic innovations may be viewed within a sociological context as representing the individual's attempt to assimilate roles and to communicate reference group identification(s) to others (Chapman and Volkman 1939, Hyman 1942, Merton and Kitt 1950. Sherif 1953. Shibutani 1973).

As Grubb and Grathwohl (1971, p.24) have noted, "A more meaningful way of understanding goods as social tools is to regard them as symbols serving as a means to communicate between the individual and his significant references... If a symbol is to convey meaning it must be identified by a group with which the individual is associated... and the symbol must communicate similar meaning to all within the group." This idea can be extended to include the notion that innovations are probably even more useful as social symbols if their intangible meanings are known to those outside the reference group, as well.

Products may vary in the amount of socially symbolic (i.e., intangible) meaning they convey and in the degree to which socially symbolic meaning contributes to their perceived innovativeness. For example, an innovation such as interferon, a new drug for use in the treatment of cancer, possesses little meaning as a social referent and it is unlikely that considerations of life style would figure prominently in its adoption by an individual or its diffusion throughout society. Conversely, innovative products in such areas as hairstyles, jewelry, and apparel may be adopted largely as a result of reference group influence (Schenk and Holman 1980). Such socially symbolic innovations may be adopted as a result of socialization processes and the desire to express self-identity; that is, consumers may find the intangible attributes (e.g., youthfulness) associated with the innovation to be congruent with the self-image they wish to convey and adopt the innovation accordingly (Brim 1960, Gergen 1971, Mortimer and Simmons 1978, Turner 1975).

A central idea in viewing symbolic innovations is that they may have been physically present in society for an extensive period of time, yet be considered innovations at the present time. This is because their innovativeness arises not from the novelty of their tangible features (which are already "known"), but rather from a change in the social meaning (i.e., intangible attributes) assigned to the product, which makes it appear novel as a symbol. For example, corn-rowing of hair has been known and present in American society ever since Africans, to whom it is indigenous, arrived here in the 1600's; yet its widespread diffusion as an innovation among white women was not experienced until the late 1970's.

The symbolic meaning assigned to the corn-row hairstyle changed from its being associated with the Black ethnic subculture to being viewed as an innovation appropriate for white women who desired to emulate the life style exhibited by Bo Derek in the movie '10'. The corn-row hairstyle "innovation" was already present in society; however, its innovativeness arose from the fact that its social meaning had been altered ant, hence, was novel. In essence the style was relabeled as a product appropriate for a different reference group than that in which it originated. That is, the intangible attributes it had identifying it with a particular reference group--were replaced with a different set of intangible attributes which made it appropriate for a second reference group. Subsequently, it diffused secondarily among those identifying with this latter reference group.

Another instructive example of secondary diffusion for a symbolic innovation as a result of reference group reassignment is that of wire-rimmmed eyeglass frames. This type of eyeglass frame was originally introduced in the late nineteenth century and became widely diffused throughout a variety of social strata (Carson 1974). When plastic eyeglass frames entered as innovations during the mid-twentieth century, the proportion of persons wearing wire-rimmed frames declined and was largely confined to those who had acquired them originally and not adopted plastic frames (Favel 1978). However, toward the end of the 1960's, wire-rimmed eyeglass frames became identified with the student/hippie reference group and diffused widely throughout this subculture, where it was perceived as an innovation

An interesting operational anomaly for the traditional diffusion paradigm is raised by these two examples. Because both the corn-row hairstyle and wire-rimmed eyeglass frames were already present in society prior to their being relabeled in social meaning and experiencing a secondary diffusion process, a definite time of social system introduction can not be designated for them. Unlike fluoridated toothpaste, which did not exist prior to its introduction and diffusion, these two socially symbolic innovations preexisted their latter diffusions by several decades.

This operational complication is further illustrated by the inability of the traditional diffusion paradigm classification scheme to appropriately label adopters of symbolic innovations. Using time of adoption, a consumer who adopted wire-rimmed eyeglasses in the 1920's and laggardly refused to adopt the later plastic frames would be classified as considerably more "innovative" than a college student who adopted wire-rimmed frames shortly after they were symbolically relabeled in the 1960's.

This same problem is encountered in studying the diffusion of many apparel item innovations such as skirts, pants, ties, socks, and shoes in which lengths, widths, and designs have gone "in" and "out" of style in successive cycles or which have been adopted and discarded as symbols of different reference groups on a longitudinal basis. If a researcher uses time of adoption as the criterion for classifying adopters, without regard for the social (in- tangible) meaning the product possesses for the adopter, classification is hopelessly confounded by the multiple identities of the innovation.

There are two major propositions to be derived from these examples. The first is that products whose innovativeness springs from their socially symbolic meaning need not be newly entered into a social system to constitute innovations. Their innovativeness arises from a redefinition of their social meaning--the assignment of a novel set of intangible attributes. They are novel as social symbols; yet are not novel as physical entities. The second proposition is that the diffusion pattern for such innovations will be guided by their use as social referents, rather than technological performance. Consumers who want to identify with the life style they are perceived to represent will choose to adopt them; those who do not wish to emulate such a life style will not adopt them. Symbolic innovations primarily are adopted for their utility as social communicative devices and not for their technological superiority.


The technological innovation is the type of innovation for which the traditional innovation diffusion model principally was invented as a descriptive and explanatory device. The central notion of a technological innovation is that it possesses some tangible features never previously found in that product class. Unlike symbolic innovations whose innovativeness derives from the assignment of novel intangible attributes to a present product, technological innovations have never existed in their present form prior to creation. Hence, their time of introduction into a social system can be calculated with accuracy (theoretically at Least) and adopters can be meaningfully arrayed according to the time of both innovation introduction and adopter awareness. Examples of technological innovations include hybrid seed corn (Ryan and Gross 1943), antibiotics (Menzel and Katz 1960), fluoridated toothpaste, birth control pills, the automobile, the laser, and the atomic bomb. in each of these instances, the innovation had a determinant date of creation and represented a substantially different, tangible entity than its predecessors. Further, in each instance, the innovation exhibited what was perceived to be a superiority in performance over its predecessors. For example, fluoridated toothpaste is believed to result in fewer teeth cavities than is unflouridated toothpaste.

Technological innovations, it may be argued, will diffuse according to a somewhat different set of principles than symbolic innovations for several reasons. First, they may be adopted primarily because of their performance characteristics rather than because of their social symbolism. Thus they may diffuse more according to consumer need for -their technical utility, rather than consumer desires for self-identity.

Second, because they possess some fundamentally new features, rather than simply novel reference group assignment, technological innovations may be more inherently "different" and "discontinuous" than symbolic innovations. Because of this, it is perhaps to be expected that they would diffuse first among the more informed stratum of consumers (as indicated by higher educational attainment and occupational status) and among those more predisposed toward risk-taking, as has been found in the majority of research investigations on this type of innovation (Rogers and Shoemaker 1971).

Third, technological innovations result from the accretion of scientific knowledge; they cannot exist before the discovery of certain information needed to create them. Thus, prior to their creation they are unknown and unknowable to the society generating them. After introduction, their long-term effects may be largely unpredictable and, hence, more uncontrollable (e.g. nuclear fission, birth control Dills, recombinant DNA).

In contrast, the physical properties of symbolic innovations are generally known to all or many members of a society; hence, their effects may be more predictable, despite the reassignment of their symbolic meaning.


The implications of these two dimensions of innovation generation may be discussed by viewing them as orthogonal axes along which various product classes can be arrayed. Such an arrangement is depicted in Figure l.

Products designated as Cluster "A" in Figure l are those that are high in social symbolism but low in technology as bases for innovation within the product class. That is, innovation is generally effected via the reassignment of intangible attributes to presently existing products. Examples include apparel, beer, cigarettes, and hairstyles. In each of these instances, products are frequently repositioned by marketer-induced changes in their intangible attributes. A product can be made to seem new by use of a promotional campaign that provides it with novel symbolic features. Cluster "B" contains products such as medical equipment, the laser, and computer systems--in which technological changes generally dominate the symbolic dimension in generating innovation. Cluster "C" contains products for which both the symbolic and technological dimensions are used in balanced proportions to effect differentiation and generate innovation. Included in Cluster "C" would be such products as automobiles, stereo systems, watches, and television sets. Cluster "D" represents products that are relatively lower in both the symbolic and technological differentiation dimensions. Falling into this category are such products as sugar, soap, hardware, and fertilizers, which experience relatively low rates of innovation within the product class.

Products in Cluster "A", it is assumed, will diffuse primarily as a result of reference group influence. Consumers will adopt product innovations in this quadrant because they wish to identify with the reference group seen as "sponsoring" the innovation at the present point in time; that is they wish to use the intangible attributes associated with the product to communicate their own identity to others. This, of course, leaves open the possibility that the consumer knew about, but avoided adopting, the product at an earlier point in time because s/he did not identify with the reference group then sponsoring it, and also the possibility that the consumer may discard the product at a future point in time, because the product has been reassigned to a reference group with which s/he does not wish to identify. The diffusion of such innovations is believed determined predominantly by the diffusion of their sponsoring reference groups. If a reference group experiences an increase in popularity, the diffusion of the product symbols associated with it will increase as well.

Products in Cluster "B", it is assumed, will diffuse primarily as a result of their perceived superior technical performance relative to other products. Innovations in Cluster "3" are relatively (though not completely) free of reference group affiliation. Their utility to the consumer centers amount their novel technological superiority and not their value as social referents. The consumer presumably adopts them to fulfill a physical need (e.g., a cure for cancer or diabetes) rather than to communicate self-image to others. Hence, the diffusion of such products should be largely predictable through identification of consumers with relevant needs and the ability to perceive the utility of the innovation.

Products in Cluster "C" experience innovation on both the symbolic and technological dimensions. Hence, new products may possess both a reference group affiliation and some novel technological features. To predict diffusion patterns for these products, information on the consumer's self-image, ability to perceive technological change, and need for the performance improvements inhering in the innovation would be required. It is possible that technologically superior innovations in this cluster may be rejected by some consumers, because they are perceived to be reference-group-inappropriate. Further, some consumers may adopt innovative products from this cluster because they are viewed as representing the appropriate reference group even though their technological superiority is undetected.

Finally, product classes represented in Cluster "D" are not generally associated with a given reference group and hence communicate little social meaning. Further, they are products for which few technological advances are sought ant/ or possible. As a result, few innovations are generated within Cluster "D". Their diffusion is likely to be stable and dependent primarily on the distribution of the consumer need they are fulfilling.


This paper has proposed that product innovations may arise from two independent sources - symbolism (intangible attributes) and technology (tangible attributes). Symbolic innovations are those which result from the reassignment of social meaning to an existing product, generating a secondary diffusion for it among those identifying with the relevant reference group. Technological innovations are those that spring from the addition or alteration of tangible features in a product that serve to distinguish it from prior models. Innovations may arise in given product classes from either or both of these two sources.

It is argued that symbolic innovations will diffuse primarily due to their association with a given reference group; whereas technological innovations will diffuse primarily as a result of consumer perceptions of and need for their superior performance. While the traditional.diffusion model has served as an adequate paradigm for technological innovations (for which it was primarily designed), it is argued that problems in measuring time of introduction and in classifying adopters make it inappropriate for describing the diffusion process for symbolic innovations. It is suggested that reference group theory may serve as a more appropriate framework for understanding this latter type of innovation

There are several conceptual and methodological implications to arise from this framework. Let us first consider those regarding symbolic innovations. Because symbolic innovations require no novel technological attributes, they are perhaps easier to create and manipulate than technological innovations. This implies that scientific expertise is not required for their construction. Because of this trait, symbolic innovations may arise from two sources not traditionally considered as the birthplace of innovation consumers and advertisers. Consumers, themselves, are capable of creating a symbolic innovation by undertaking a consensual decision to redefine the social meaning of a cultural object. For example, the recent use of razor blades and chains as jewelry by members of the "punk rock" subculture reflects this symbolic innovation phenomenon. Quick-thinking manufactures s were able to capitalize on this trend by generating "razor blade and chain" jewelry offerings (with dull edges for the timid) which were purchased by consumer-followers of the punk rock trend.

Similarly, advertisers can create a symbolic innovation by relabeling an existing product or by providing differentiating intangible attributes to a product that is physically identical/similar to its competitors. For example, a cigarette can be made to seem new if it is shifted from feminine to masculine (i.e., Marlboro). Similarly, two physically identical automobiles can be made to appear as distinct innovations via the image created through advertising (e.g., Ford Escort vs. Mercury Lynx).

This implies that research and development resources need not be expended by marketers in the generation of symbolic innovations; rather, only a competent promotional campaign is required. Further, it also implies that marketers do not control all symbolic innovations; in many instances consumers have been the source of creativity in reassigning social meaning to objects and providing them with innovation status.

Turning now to technological innovations, we should first note that their dependance upon physical alteration/improvement makes them more likely to be controlled by formal production systems. That is, formal research and development departments staffed by experts are generally necessary to the generation of technological innovations. This has the disadvantage of a probable higher origination cost for a given technological innovation . However, it also implies the corresponding benefit (to marketers) of insuring greater direct control over the distribution and availability of the innovation.

For example, the availability of popular technological innovations can be purposely restricted by their manufacturers in order to maintain high profit margins which will not only cover R & D expenses, but contribute to the capitalization of the company. In contrast, the control of some symbolic innovations may lie more with consumers than manufacturers (e.g., the long hair and faded jeans of the hippie era), thus preventing or inhibiting the ability of marketers to generate profits from them.

These ideas are but a few that may be generated using the symbolic/technological innovation framework. Although this framework is still in its formative stages, it is believed to provide a novel perspective of the innovation process. Its central thesis is that innovations may be brought into being by two substantially different means symbolism or technology and that the differences between these modes of innovation production can have important implications for measuring and explaining the adoption/ diffusion process. It is hoped that researchers of innovation-related phenomena will find it to be a useful vantage point.




Barber, Bernard, and Lobel, Lyle S. (December 1952), "Fashion in Women's Clothes and the American Social System," Social Forces, 31, 124-131.

Brim, P. G. Jr. (1977), "Socialization Through the Life Cycle," in Socialization After Childhood (ed.), P. G. Brim Jr., S. Wheeler, new York: Wiley, pp. 1-49.

Brim, O. (1960), "Personality Development as Role Learning," In Personality Development in Children (ed.) I. Iscoe, H. Stevenson (Austin, TX: University of Texas Press).

Carson, Richard (1974), Fashion in Eyeglasses (London: Owen).

Chapman, D. W., and Volkmann, J. (1939), "A Social Determinant of the Level of Aspiration," Journal of Abnormal and Social Psychology, Vol. 34, 225-238.

DiRinzo, G. (1977) a "Socialization, Personality and Social Systems," Annual Review of Sociology, Alex Inkeles, James Coleman, Neil Smelser (eds.), Annual Reviews, Inc., pp. 261-95.

Favel, Douglas Borne (1978), Eyeglass Frames, new York: Bristol.

Gergen, R. J. (1977), "The Decline of Character: Socialization and Self-Consistency," in We, the People: American Character and Social Change, G. J. Di Renzo (ed.), Westport, CT Greenwood.

Grubb, Edward L., and Grathwohl, Harrison L. (1967), "Consumer Self-Concept, Symbolism and Market Behavior: A Theoretical Approach," Journal of Marketing, 31, pp. 22-27.

Hirschman, Elizabeth C. (1980), "Attributes of Attributes and Layers of Meaning," in Advances in Consumer Research, Vol. 7, Association for Consumer Research, Ann Arbor, MI.

Hirschman, Elizabeth C. (1981a), "Commonality and Idiosyncracy in Popular Culture," in Elizabeth C. Hirschman and Morris B, Holbrook (eds.), Symbolic Consumer Behavior, Association for Consumer Research, Ann Arbor, MI.

Hirschman, Elizabeth C. (1981b), "Cognitive Complexity and the Perception of Intangible Attributes," Proceedings National Educator's Conference (American Marketing Association. Chicago. IL).

Hymwn, R. R. (1942), "The Psychology of Status," Archives of Psychology, Vol. 38, 15.

Menzel, Herbert, and Katz, Elihu, "Social Relations and Innovation in the MedicaL Profession: The Epidemiology of the New Drug," Public Opinion Quarterly, 19, 337-352.

Merton, R. R., and Kitt, A. (1950), Contribution to the Theory of Reference Group Behavior: Studies in the Scope and Method of "The American Soldier," (ed,) R. R. Merton and P. F. Lasarfeld (The Free Press).

Midgley, David F., and Dowling, Grahame R. (March 1978), "Innovativeness: The Concept and Its Measurement," Journal of Consumer Research, 4, 229-242.

Mortimer, Jeylan F., and Simmons, Roberta G. (1978), "Adult Socialization," in Annual Review of Sociology, Vol. 4, Ralph H. Turner, James Coleman and Renee C. Fox (eds.) Annual Reviews, Inc. (Palo Alto, CA, pp. 421-454).

Ostlund, Lyman E. (September 1974), "Perceived Innovation Attributes As Predictors of Innovativeness," Journal of Consumer Research, Vol. 1, 23-29.

Robertson, Thomas S. (1971), Innovative Behavior and Communication (New York: Holt, Rinehart and Winston, Inc.).

Rogers, Everett M., and Shoemaker, Floyd (1971), Communication of Innovations (Nev York: Free Press).

Ryan, Bryce, and Gross, Neal C. AccePtance and Diffusion of Hybrid Seed Corn in Two Iowa Communities (Ames, IA: Agricultural Experiment Station, Research Bulletin 372).

Schenk, Carolyn T., and Holman, Rebecca R., (1980), "A Sociological Approach to Brand Choice: The Concept of Situational Self ImAge," in Jerry C. Olson (ed.) Advances in Consumer Research, Vol. 7, Association for Consumer Research, 610-614.

Sherif, M., (1953), "The Concept of Reference Groups in Human Relations," GrouP Relations at the Crossroads, (ed.) M. Sherif an G M. Wilson Harper & Row, 203-231.

Shibutani, Tamatsu (1973), "Reference Groups as Perspectives," in Harold H. Kassarjian and Thomas S. Robertson (eds.) Perspectives In Consumer Behavior, Scott, Foresman, and Company (Glenview, Il, 299-309).

Smelser, Neil J. (1973), Sociology Second Edition (New York: John Wiley & Sons).

Turner, J. R. (1974), The Structure of Sociological Theory (Homewood, IL: Dorsey).

Tversky, Amos (1977), "Features of Similarity," Psychological Review, 84, 327-352.

Wallendorf, Melanie (1979), Role Accumulation and the Diffusion of Innovations, unpublished doctoral dissertation Graduate School of Business, University of Pittsburgh, Pittsburgh, PA.

Zaltman, Gerald, and Wallendorf, Melanie (1979), Consumer Behavior: Basic Findings and Managerial Implications, (New York: John Wiley & Sons).