Possessions and the Extended Self RUSSELL W. BELK* Our possessions are a major contributor to and reflection of our identities. A variety of evidence is presented supporting this simple and compelling premise. Related streans of research are identified and drawn upon in devetopJng this concept and implications are derived for consumer behavior. Because the construct of exterxJed self involves consumer behavior rather than buyer behavior. It apjpears to be a much richer construct than previous formulations positing a relationship between self-concept and consumer brand choice.

Hollow hands clasp ludicrous possessions because they are links in the chain of life If it breaks, they are truly losL—Dichlsr 964 W e cannot hope to understand consumer behavior without first gaining some understanding of ihe meanings that consumers attach to possessions. .• key to understanding what possessions mean is recognizing thai, knowingly or unknowingly, intentionally or unintentionally, we regard our possessions as parts of ourselves. As Tuan argues, “Our fragile sense of self needs support, and this we get by having and possessing things because, to a large degree, we are what we have and possess” (1980. . 472). That we are what we have (e. g.. Van Estcrick 1986; Feirsiein 1986; Rosenbaum 1972) is perhaps the most basic and powerful faci of consumer behavior. The premise that we regard our possessions as parts of ourselves is not new. William James (1890, pp. 291-292), who laid the foundations for modern conceptions ofself, he! d that: a man’s Self is the sum total of all that he CAN call his, not only his body and his psychic powers, but his clothes and his house, his wife and children, his ancestors and friends, his reputation and works, his Jands, and yacht and bank-account.

All these things give him the same emotions. If they wax and prosper, he feels triumphant; if they dwindle and die away, he feels cast down,—not necessarily in the same degree for each thing, but in much the same way for all. ‘ If we define possessions as things we call ours, James was saying that we are the sum of our possessions. The purpose of this article is to examine the relationship between possessions and sense of self. It is based not only on the premise that this relationship is •RusscK W. Belk is the N. EJdoa Tanner Professor of Business Administration.

Graduate School of Business, University of Utah, Salt Lake City, UT 84060. The author wishes to thank Melanie WaUendorf, Floyd Rudmin. and Grant McCracken for their commenis on an earlier version of this anicle. of imponance to understanding consumer behavior, but also on the premise that understanding the extended self wi]] help us learn how consumer behavior contributes to our broader existence as human beings (Belk ! 987a). The first section considers various evidences that possessions are an important component of sense of self.

The most direct form of evidence is found in the nature of self-perceptions. Additional, especially striking evidence is found in the diminished sense of self when possessions are unintentionally lost or stolen. More evidence ofthe role of possessions in sense of self comes from anthropological studies ofthe way possessions are treated ritually and after death. Because extended self is such a broad topic, several unreviewed areas of evidence on the extent and nature of the relationship between possessions and sense of self also are identified. In so doing, the scope ofthe present treatment is also defined.

The question of what functions the extended self serves is addressed in the second section, which begins with abrief review of the basic states of our existence: having, doing, and being. These states are relevant to the question of how we define who we are. Next, the functions of possessions in human development are considered. Four stages are identified: (1) the infant distinguishes self from environment, (2) the infant distinguishes self from others, (3) possessions help adolescents and adults manage their identities, and (4) possessions help the old achieve a sense of continuity and preparation for death.

Finally, the role of possessions in creating or maintaining a sense of past is considered. The third section examines several processes involved in self-extension. One process is the initial incorporation of objects into our extended selves. A number of incorporation processes are discussed, not all of which involve possession in the sense of individual ownership. A particular process of self-extension ‘James calied his text an encyclopedia of psychology and quolcs Hcrr Horwicz’s Psychologische Analysen (no date or publisher given)asasourceof many of his ideas on self. 39 ? JOURNAL OF CONSUMER RESEARCH • Vol. 15 • September 1988 140 THE JOURNAL OF CONSUMER RESEARCH that is considered in some detail is contamination. In contamination, both good and bad aspects of objects are seen to attach to us through physical contact or proximity. A final process theorized is the maintenance of multiple levels of the self, such as viewing our family, city, and nation to be a part of who we are. The fourth section ofthis article focuses on a number of special categories of possessions that are commonly incorporated into the sense of self.

These categories are collections, money, pets, other people, and body parts. In each case, research is reviewed supporting the contention that this category’ of objects is a part of the extended self and is therefore treated differently from objects not considered to be a part of self. The final section discusses implications ofthe extended self formulation for consumer research. The areas of implications outlined include gift-giving, vicarious consumption (generally through other family members), care of possessions, organ donation, product disposition, and the contribution of extended self to defining meaning in life.

The latter topic elevates the focus of consumer behavior research xo a level of greater significance than satisfaction with product performance. Following the final section, the formulation ofthe extended self is reviewed briefly and conclusions are offered. EVIDENCES Possessions in Self-Perception Research The term extended self has not been applied previously to the conception of self-plus-possessions, but Rochberg-Halton (1984, p. 335) comes close: Valued material possessions. . . act as signs ofthe self that are essential n their own right for ils continued cultivation, and hence the world of meaning that we create for oursleves, and that creates our selves, extends literally into the objective surroundings. One difference in the present view is that the extended self is seen not to be limited to external objects and personal possessions, but also includes persons, places, and group possessions as well as such possessions as body parts and vital organs. The notion of extended self is a superficially masculine and Western metaphor comprising not only that which is seen as “me” (the self), but also that which is seen as “mine. As James (1890, p. 291) notes, the two concepts are interwoven in the way we think of our selves: The Empirical Self of each of us is all that he is tempted to call by the name of me. But it is clear that between what a man calls me and what he simply calls mine the line is difficult to draw. We feel and act ahout certain things that are ours very much as we feel and act about ourselves. Our fame, our children, the work of our hands, may be as dear to us as our bodies are, and arouse the same feelings and the same acts of reprisal if attacked. And our bodies themselves, are they simply ours, or are they us?

Certainly men have been ready to disown their very bodies and to regard them as mere vestures, or even as prisons of clay from which they should some day be glad to escape. Although prior theories and research on consumer self-concept (see Sirgy 1982 for a review) are moderately supportive of the contention that possessions are incorporated into self-concept, this research probably considerably underestimates the extent to which this is true. One reason is that prior research methods generally attempt to find a correspondence between perceived characteristics of these objects and perceived characteristics ofthe self.

But, one can hold an object like the Statue of Liberty to be a part of one’s identity without having to hold a self-concept composed of characteristics attributed to this statue. Second, as argued by Belk (1984b), the focus of these studies on brand images prior to acquisition is too limited. Both nonbrand images (e. g. , cigarette smoker, wine connoisseur) and post-acquisition object bonding (e. g. , with one’s pet) may contribute strongly to the sense of self. Third, as argued by Belk (1984b) and Solomon and Assael (1988), rather than a single product or brand representing all of one’s selfconcept, only a omplete ensemble of consumption objects may be able to represent the diverse and possibly incongruous aspects of the total self. For all of these reasons, the present focus on extended self is substantially different than prior consumer self-concept research. For research applications within the perspective advocated here, see Belk (1987b, 1988) and Belk and Austin (1986). This more expansive view ofthe extended self can be examined in light of several prior conceptualizations and studies focusing on distal elements ofthe self.

McClelland (1951) suggested that external objects become viewed as part of self when we are able to exercise power or control over them, just as we might control an arm or a leg. In the case of tools, instruments, and weapons, envisioning the basis for the extended self metaphor is easy. The greater the control we exercise, the more closely allied with self the object should become. This principle led McClelland to hypothesize the following hierarchy of most to least closely self-allied object categories: (1) me, my “free will,” (2) my body, my conscience, (3) my belongings, (4) my friends, and (5) strangers, physical universe.

The predicted closer alignment of self to belongings than to friends recognizes the “free will” of people (friends) that is lacking in most belongings. Prelinger (1959) tested James’s premise that possessions are viewed as parts of self and McClelland’s hypothesis that control dictates the strength ofthis linkage. He had subjects sort 160 items onto a four-position (zero to three) continuum of not-self to self The items were selected so that each of eight conceptual categories was represented by 20 items. These categories and the mean “self scores for the items within them were in descending order;

POSSESSIONS AND THE EXTENDED SELF 141 that the object is “me. ” McCarthy (1984) concludes that such objects act as reminders and confirmers of our identities, and that our identities may reside in objects more than they do in individuals. Allport (1937) hypothesized that the process of gaining an identity, and in so doing gaining self-esteem, progresses from infancy by extending self via a continuously expanding set of things regarded as one’s own. This hypothesis was tested by Dixon and Street (1975) who conducted an approximate replication of” Prelingers’ study among 6- to 16-year-olds.

They found essentially the same rank ordering of item categories regarded as “self,” but found only two categories for which this tendency changed significantly with age: other people and possessions. In both cases, older children were more likely than younger children to categorize such objects as being part of self (“you”). In a three-gene rational study of favorite possessions, Rochberg-Halton (1984, 1986; Csikszentmihalyi and Rochberg-Halton 1981) found that as we age the possessions that people cite as “special” tend increasingly to be those that symbolize other people (e. . , gifts from people, photographs of people). Their interpretation of these findings suggests an age-related widening ofthe boundaries of self (RochbergHalton 1984, p. 352). These findings also may suggest that possessions are regarded not only as a part of self, but also as instrumental to the development of self. Other work on the role that special possessions play in easing life transitions also suggests that possessions can be instrumental to maintenance of self-concept (e. g.. McCracken 1987a). A study by Belk (1987b. 988; Belk and Austin 1986) examines the self-defining role of places, public monuments, experiences, time periods, television programs, motion pictures, and public figures, in addition to the sort of objects, persons, and traits studied in prior research. Ofthe additional extended self categories considered, places and experiences tend to be seen as most clearly a part of extended self. Added to the previously noted findings then, we may summarize the major categories of extended self as body, internal processes, ideas, and experiences, and those persons, places, and things to which one feels attached.

Of these categories, the last three appear to be the most clearly extended. However, given the difficulties in separating mind and body in philosophies and psychologies of tbe self (e. g. , Campbell 1984; Englehardt 1973; Tuner 1984), objects in all of these categories will be treated as potential parts ofthe extended self. In conversations in English (although less true in some other languages such as Japanese), ii is clear that some objects in the former categories are treated as both a part of extended self and a part of essential unextended self.

For instance, saying I have a dark tan or my body is tan (possessive and extended uses) is more usual than saying I am a tan body (a nonpossessive and an unextended usage). However, saying I am 1. Body parts (e. g. , the skin, the genital organs), 2. 98; 2. Psychological or intraorganismic processes (e. g. , the conscience, an itching on the sole of the foot), 2. 46; 3. Personal identifying characteristics and attributes (e. g. , age, occupation), 2. 22; 4. Possessions and productions (e. g. , watch, perspiration, toilet articles), 1. 7; 5. Abstract ideas (e. g. , the morals of society, the law), 1. 36; 6,. Other people (e. g. , the people in your hometown, father), 1. 10; 7. Objects within the close physical environment (e. g. , dirt on the hands, furniture in this room), 0. 64; 8. Distant physical environment (e. g. , the adjoining room, the moon), 0. 19. Although it is unfortunate that Prelinger grouped some autonomic bodily productions with possessions, these findings still support James’s contention that possessions are seen as part of self.

They also suggest an ordering ofthe “selfness” of these object categories that is parallel to the hierarchy suggested by McClelland. To test McClelland’s control hypothesis more directly. Prelinger had five judges separate the 160 items into three groups: those that are predominantly under the control of people, those that primarily control or affcci people, and those that are predominantly neutral in both regards. The first two categories both received high mean “self scores (over 1. 8) from subjects, while the neutral items clearly received “non-self scores (mean less than 0. ). These findings suggest that besides control over objects, control by objects may also contribute to an item being viewed as part of self. That is, we may impose our identities on possessions and possessions may impose their identities on us. Interestingly, control also has been suggested to be the critical determinant of feelings of possession (Furby l978;Tuan 1984). If both hypotheses are correct, the more we believe we possess or are possessed by an object, the more a part of self it becomes.

It is telling that the categories of extended self just noted correspond quite closely to the areas in which Ellis (1985, pp. 115-117) found evidence of human possessiveness (no hierarchical ordering was reported): (1) one’s body, (2) personal space, (3) ingestibles, (4) territory, (5) domicile, (6) copulatory partners, (7) offspring, (8) friends, (9) tools, and (10) objects of aesthetic appeal, play and amusement, pets and mementos. Nuttin (1987) finds that even the tetters in our names are viewed possessively. Apparently, in claiming that something is “mine,” we also come to believe 42 THE JOURNAL OF CONSUMER RESEARCH tired (unextended) is more common than saying my body is tired (extended). Even greater complications in making distinctions between extended and unextended selves are found with asomatognostics who cannot apprehend the existence of parts of their bodies (Litwinski 1956; Sacks 1985), amputees who develop phantom limbs (Plugge 1970), and recent treatments of beliefs as possessions (Abelson 1986; Abelson and Prentice f’orthcoming). From the present perspective, the issue is an empirically resolvable one that depends upon perceptions.

For instance, Belk and Austin (1986) found the following mean scores for various body parts on a four-point scale of “selfness,” where four is the highest possible score: eyes 3. 5, hair 3. 2, heart 3. 1, legs 3. 1, hands 3. 1, fingers 3. 0, genitals 3. 0, skin 3. 0, nose 2. 7, knees 2. 7, chin 2. 6, kidneys 2. 6, liver 2. 6, and throat 2. 5. For this sample, it seems best to conclude that none of these body parts is necessarily an inherent part of unextended self, but that eyes, hair, and heart are more likely to be treated in this way than are kidneys, liver, and throat.

The study also found some evidence of sex and age differences in the incorporation of body parts into sense of self. Furthermore, it is likely that those who have undergone such elective body alterations as plastic surgery and tattooing are likely to view the affected body parts as being more a part of self (e. g. , Sanders 1988). seen to be bestowed by the organization. The present focus would suggest that those who have less of their extended selves stripped from them may adjust more readily to such situations.

Another instance in which nonvoluntary loss of possessions may bring about a diminished sense of self is when possessions are lost to theft or casualty. In the case of burglary victims, Rosenblatt, Walsh, and Jackson (1976) suggest that a process of grief and mourning may follow the discovery of theft, just as one might grieve and mourn the death of a loved one who had been a part of one’s life. What is lost in both cases may be a part of self. As the college student victim of a bicycle theft accuses the unknown thief, she reveals the identity invested in the bike (Donner 1985,p. 1): It hurts to think that someone else is selling something that for me is more precious than money . . . Everyone who owns a bike has their own story that makes their bike more than just machinery to them. And you ripped it off. You stole a piece of my life. You didn’t just steal a chunk of metal to sell. . . You walked off with my memories. The present author conducted a small-scale test of this hypothesis using data from interviews with a nonrepresentative sample of 20 burglary victims who were asked in open-ended questions to recall their initial thoughts and feelings upon discovering the loss.

Following anger and rage, the most commonly reported reactions were feelings of invasion and violation. In fact, eight of the 11 females in the sample spontaneously suggested that it was as though they had been violated, polluted, or raped. There are similar reports in Maguire’s (1980) study of British burglary victims, although only 12 percent ofthe females in his study suggested such a feeling. Additional confirmation ofthis feeling of personal violation is found in studies by Korosec-Serfaty (1985) and Paap (1981).

There are also reports of feelings of loss of a part of self among victims of natural disasters. McLeod (1984) found that those who lost possessions to a mudslide went through a process of grief similar to that in losing a loved one—moving from denial to anger, to depression, and finally to acceptance (often after many months). The author joined several other researchers in conducting depth interviews with flood victims during the summer of 1986, and found that after six weeks most victims were still in the early stages of grief and often could not talk about the disaster or cried while attempting to do so.

Fieldnotes from one such interview include this account: The losses that concerned (the flood victim) most were those of his record collection,. . a first edition book collection, . . . the tools that his father—the cabinet maker—had used, . , . the ceiling and paneling of the basements that he had installed with the help and advice of his father, and (upstairs), the hutch, lowboy, and stereo cabinet that his father had made. Loss of Possessions If possessions are viewed as part of self, it follows that an unintentional loss of possessions should be regarded as a loss or lessening of self.

Goffman (1961) provides a thorough review ofthe evidence of deliberate lessening of self brought about in such institutions as mental hospitals, homes for the aged, prisons, concentration camps, military training camps, boarding schools, and monasteries. One ofthe first steps in receiving new members into these institutions is to systematically deprive them of all personal possessions including clothing, money, and even names. Their bodies may be standardized to some degree, as with military haircuts, and their behaviors and conversations may be severely restricted.

They are reissued standard wardrobes and minimal possessions to aid in rebuilding a new standardized identity. The result ofthis systematic substitution of standardized “identity kits” for former possessions is an elimination of uniqueness (Snyder and Fromkin 1981) and a corresponding and often traumatic lessening ofthe individual’s sense of self. Although the new, more standardized possessions that are substituted may eventually restore some sense of self, the new self should necessarily be less unique and involve more of a shared group identity.

Furthermore, the individual typically becomes a user of these new objects rather than an owner of them. Because control is restricted and the organization remains the owner, identity is POSSESSIONS AND THE EXTENDED SELF Clearly what is mourned here is a loss of seif. Similar findings were obtained in the Buffalo Creek flood (Erikson 1976). AsGeorgSimmel observes, “material property is, so to speak, an extension ofthe ego, and any interference with our property is, for this reason, felt to be a violation ofthe person” (1950, p. 322).

The flood victim also illustrates how the labor ofthe individual (in this case the victim’s recently deceased father) adheres in the objects produced. In this sense, the loss of possessions was also a further loss of his father’s extended self that reniained in his father’s creations. Besides the more direct loss of self when personal possessions are lost to theft or casualty, the vulnerability revealed in such losses may damage the sense of self derived from the attachments to home and neighborhood. Home (e. g.. Cooper 1974; Duncan 1976; Duncan and Duncan 1976) and neighborhood (e. . , Bakker and Bakker-Rabdau 1973; Gerson, Stueve, and Fischer 1977) have been suggested to be strong sources of personal identity. As with more personal possessions, home and neighborhood have been hypothesized to contribute to sense of self to the degree thai a person feels control over them (Bakker and Bakker-Rabdau 1973; Edney 1975). This may explain why Brown (1982) found that burglary victims report less sense of community, less feeling of privacy, and less pride in their house’s appearance than do their nonburglarized neighbors.

The same phenomenon has been observed in those displaced by slum clearance, even when they were relocated to “better” housing (Fried 1963), In the words of Peter Marris, ‘They identify with the neighborhood: it is part ofthem, and to hear it condemned as a slum is a condemnation ofthemselves too” (1986, p. 55). Besides loss of possessions to theft or casualty, others have maintained that whenever the functions and property of individuals are taken over by institutions, such as government and schools, there is a regrettable loss of self (Dietze 1963; Wiggins 1974).

Although the intent of these institutions is presumably not to lessen others’ sense of selves, Wiggins (1974) suggests that there are instances in which a person’s possessions are damaged with the intent of diminishing the owner. He gives as one example a cbiJd who destroys the property of a larger child or of an inviolable sibling in an effort to more effectively direct aggression at this person. Vandalism may be motivated similarly with the targels being society, those who seem to be more fortunate, or public institutions (e. . , Chester 1976; Fisher and Baron 1982). The trauma that may attend involuntary loss of possessions normally is not present in voluntary disposition of possessions. Indeed, La Branche (1973) observes that when possessions are recognized as inconsistent with our images of self, we gladly neglect or dispose of them. But, when the disposition is forced, as by economic circumstances, the parting likely brings sorrow. As one elderly respondent pawning possessions to make it through the winter reflects (Cottlel98l,p. 8): i stand in those lines with my suitcase full of things to praciicaJiy give away; 1 stand in that hock shop, and I tell myself that my entire life is being sold . . . Don’t make me hock my life away, I beg you. Of course, there is a more utilitarian explanation of the feelings of resentment at the loss of possessions. In this more utilitarian view, we merely regret the loss of valued possessions because ofthe benefits they provide rather than from any feelings of self erived from or mingled with these objects. James (1890, p. 293) challenges the sufficiency ofthis view: although it is true that a part of our depression at ihe toss of possessions is due to our feeling that we must now go without certain goods that we expected the possessions to bring in their train, yet in every case there remains, over and above this, a sense ofthe shrinkage of our personality, a partial conversion of ourselves to nothingness, which is a psychological phenomenon by itself.

Extreme examples ofthis partial annihilation of self are cited by Beaglehole (1932) and Rigby and Rigby (1949) in accounts of art collectors who have gone to such great lengths as suicide to avoid facing the forced breakup of their collections. Less extreme examples are found in the simple nostalgic regret at the disposal of wornout clothing and similar items that have been associated with pleasant memories of one’s past (e. g. , Lurie 1981, p-33; Rooney 1984, pp. 3-4).

If involuntary loss of possessions causes a loss of self, one ofthe primary reactions following such loss should be an attempt at self-restoration. This phenomenon has been observed in psychoanalysis and has led to the hypothesis that, along with body loss, object loss is the fountainhead of creativity (Niederland 1967; Niederland and Sholevar 1981). Body loss refers to some real or imagined physical deformity or bodily imperfection that detracts from sense of self. Object loss normally refers to the death of a close family member, but is also used by Niederland lo refer to the traumatic loss of possessions.

In body and object loss, the creation of art, craft, concept, or writing is seen as an attempt to extend the self in new ways that make up for the loss and restore the self to wholeness. That is, periods of creativity may follow the loss of one’s possessions. Niederland and Sholevar (1981) also suggest that for many young American males, the automobile is a part of their extended selves and their ego ideals. This view is supported by consumer self-concept research (e. g. , Bloch 1982; Grubb and Hupp 1968; Jacobson and Kossoff 1963).

The processes of creating and nurturing extended self through an automobile may be seen in customizing (personalizing) the car and in lavishing great care on its maintenance. When such a car is damaged, the owners react as if their own bodies i44 THE JOURNAL OF CONSUMER RESEARCH have been injured. Consider the sense of personal injur;’ described by Bellow (1975, p. 36) after a treasured car was assaulted: Someone had done to my car as rats, I had heard, did when they raced through warehouses by the thousands and tore open sacks of flour for the hell of it. I felt a similar rip at my heart . . I had allowed the car to become an extension of my own self. . . , so that an attack on it was an attack on myself. It was a moment terribly fertile in reactions. Furthermore, the possessors of such damaged treasures are anxious to either restore the auto to its former perfection or replace it with a more perfect substitute. These reactions reflect the desire to restore the damaged sense of (extended) seif caused by the injury to the automobile. Investing Self in Objects The idea that we make things a part of self by creating or altering them appears to be a universal human belief.

Anthropologists generally agree that the maker of an object, the user of land, and the cultivator of a plant are regarded as being entitled to the product of their labor (e. g. , Herskovits 1952; Lewinski 1913). Locke (1690) made this the foundation for his views on property and government, explaining the “natural basis” for private property in three steps: (1) we own ourselves (see Wikse 1977), (2) therefore we own our labor (what we direct our bodies to do), and (3) therefore we own what we produce from our labor out of the unowned materials of nature.

Csikszentmihalyi and Rochberg-Halton (1981) provide a more psychological explanation in suggesting that we invest *’psychic energy” in an object to which we have directed our efforts, time, and attention. This energy and its products are regarded as a part of self because they have grown or emerged from the self. The same principle has been suggested to apply to objects that are forcefully appropriated from others (Veblen 1898). After the development of money payment for labor, purchasing objects offers another means for investing self (in this case more symbolically) in possessions.

Beaglehole (1932) reviews other anthropological evidence ofthe link between possessions and self. The almost literal incorporation of objects into self and self into objects is shown in various practices of traditional peoples. These practices include licking new possessions, burying the umbilical cord on tribal land, inserting removed foreskin beneath the bark of a personal tree, eating or taking the name of conquered enemies, burying ancestors on sacred tribal land^ and claiming ownership of new land or artifacts by touching them, naming them for a part ofthe person’s body, leaving a lock of hair on them, or shedding blood on them.

Another exatnple, perhaps repugnant to Western observers, is the drinking ofthe urine of Vedic priests to partake of the psychogenic properties ofthe Amanita muscaria mushroom that these priests ritually consume (Wasson 1972). Each of these practices suggests the desire to tap into the life force of nature or other people by symbolically merging with these forces. In addition, the association of people and possessions is shown in the practice of burying the dead with their possessions. This practice began at least 60,000 years ago (Maringer 1960) and perhaps more than 100,000 years ago (Leaky 198i).

Alekshin (1983) compared the grave goods of men and women in Europe and found evidence that suggests women only began to experience inferior status in the third millennium B. C. (i. e. , the number and quality of their grave goods did not differ until then). Rathje and McGuire (1982) have performed similar analyses of grave goods ofthe Maya. That anthropologists assume that possessions tell us about their possessors is itself evidence ofthe tendency to see possessions as symbols of self.

The inference process is not unlike that of police detectives who attempt to construct an identity for unknown corpses by using the corpses’ possessions (Pogrebin, Poole, and Regoli 1986). In more recent traditional societies, using the clothing or possessions ofthe dead is often a taboo. Until outlawed 100 years ago in India, the wife, as “property” of a deceased husband, was expected to join him in death (Bordewich 1986). Such notions of possession surviving even death suggest a strong association between self and possessions.

To the extent that other people can be viewed as possessions (this point will be pursued in a subsequent section), mourning for dead loved ones also may be interpreted as grieving for a loss of self. The prior possessions ofthe deceased can be powerful remains of the dead person’s extended self. These remains are often the focus of normal and pathological mourning (Volkan 1974). The same association is shown in sympathetic magic in which malevolence is directed at a person through their clothing, hair or nail clippings, or other belongings (Clodd 1920).

Evidence ofthe power of possessions to capture the extended self is also shown in the angry destruction of objects left behind by the Shah of Iran and Ferdinand Marcos in the Philippines after they were deposed (Goldstein 1987). Contemporary consumption also shows that the feeling of identity invested in material objects can be extraordinarily high. For instance, Ames (1984, pp, 30-31) records feelings attached to a 19th century purchase of a parlor organ: Buying a prominent object like a parlor organ might initiate a new chapter in a set of Uves, not only by providing a new way to use time but also a new tool to measure time.

In later years the object would serve to remind its owners ofthe day it first entered their home and of the time thai had passed since then, it would not only structure their present but also their perceptions of their own past. POSSESSIONS AND THE EXTENDED SELF They knew from experience that purchasing a major object could be a significant and momentous occasion in itself, a time of heightened positive emotions and feelings of well-being and importance . . . a major purchase would transform them in their own eyes and in the eyes of others. They would become worth more . . andacquiregreaterstatus. By so doing they would receive more respect and deference from others which would, in turn, make them feel better about themselves. Buying a parlor organ would make them something they were not before. !45 One ofthe modern equivalents ofthe parlor organ in terms of impact on extended self is the automobile, especially for males (e. g. , Myers 1985; Weiland 1955). The owner of an expensive Porsche describes his attachment in this way (Stein 1985, p. 30): Sometimes I test myself. We have an ancient, battered Peugeot, and I drive it for a week.

It rarely breaks, and it gets great mileage. But when I pull up next to a beautiful woman, I am still the geek with the glasses. Then I get back into the Porsche. It roars and tugs to get moving, h accelerates even going uphill at 80. It leadeth trashy women . . . to make pouting looks at me at stoplights. It makes me feel like a tomcat on the prowl,. . . Nothing else in my life compares—except driving along Sunset at night in the 928, with the sodium-vapor lamps reflecting off the wine-red finish, with the air inside reeking of tan glove-leather upholstery and the . . .

Blaupunkl playing the Shirelles so loud it makes my hair vibrate. And with the girls I will never see again pulling up next to me. giving the cara once-over, and looking at me as if I were a cool guy, not a worriedoverextended 40-year-old schnook writer. cal emphasis or with the focus on consumer behavior. Future research seeking a broader perspective would benefit from consulting the additional literatures in Marxism and neoMarxism, critical theory, folklore, political philosophy, environmental psychology, macromarketing, semiotics, impression management, and collective memory.

The literature on property, ownership, and possession also provides a wealth of relevant material (see Rudmin, Belk, and Furby J987). The scope ofthis artical also is delimited by its predominant focus on societies that hold an individualistic concept of self As Belk (1984c) suggests, there are times and, places in world history during which the operative notion of self is more collective than individual. For a series of excellent discussions of the emergence of the individual self, see Campbell (1987), Carrithers, Collins, and Lukes (1985), and Macfarlane (1978).

The present discussion addresses collective selves in a section dealing with levels ofthe self, but the primary focus is on the individual. Most ofthe present formulation also applies in instances of collective conceptualizations of the self, but collective self involves additional concepts not addressed here—for instance, group rituals for fusing a new object into collective identity. Thus, an adequate theoretical formulation of collective extended self must await further work.

In the following section on the functions of extended seif, social functions of this construct largely are ignored. FUNCTIONS OF EXTENDED SELF Having, Doing, and Being Objects in our possession literally can extend self, as when a tool or weapon allows us to do things of which we would otherwise be incapable. Possessions can also symbolically extend self, as when a uniform or trophy allows us to convince ourselves (and perhaps others) that we can be a different person than we would be without them. Tanay (1976) suggests that handguns represent a symbolic penis for their owners.

However, Kates and Varzos (1987) challenge this interpretation and instead emphasize the real rather than symbolic power given by guns. This sense of enhancement of personal power is what made the sixgun the “equalizer” in American Western lore. Tanay’s symbolic interpretation focuses on the sense of being presumably provided by such a weapon, whereas this alternative interpretation maintains that it is what one can cio with a gun that contributes to sense of self. Thus, having possessions can contribute to our capabilities for doing and being.

The relationships among having, doing, and being are strong and have been most fully explored by existential psychologist and philosopher Jean-Paul Sartre. In his major work. Being and Nothingness, Sartre (1943) suggests that doing is merely a transitional As these examples suggest, the degree to which self may become extended into possessions can be great. In extreme cases, we again may note McCarthy’s (1984) contention that identity sometimes may lie more in extended self than in unextended self. Relevant Perspectives and Domain

The preceding discussion has presented eclectic evidence supporting the proposition thai we regard our possessions as parts ofour selves. As this article develops a deeper theoretical understanding of this phenomenon, it will continue to draw upon a broad base of literature from psychology, consumer research, psychoanalytic theory, material and popular culture studies, feminist studies, history, medicine, anthropology, and sociology. These areas and particular studies within them deal with constructs that are useful in advancing the arguments and explanations of the following sections.

A number of other areas of inquiry as well as omitted subfields from these areas just noted are potentially relevant to the study of extended self, but have been excluded either because of space considerations or because ofthe areas’ perspectives being less compatible with the present theoreti- 146 State or a manifestation ofthe more fundamental desires 10 have or to be. Further, Sartre maintains that the only reason we want to have something is to etilarge our sense of self and that the only way we can know who we are is by observing what we have.

In other words, having and being are distinct but inseparable. When an object becomes a possession, what were once self and not-self are synthesized and having and being merge. Thus, according to Sartre, possessions are all-important to knowing who we are. People seek, express, confirm, and ascertain a sense of being through what they have. Other people also affect relationships among having, doing, and being, according to Sartre. Besides others sometimes serving in an object capacity as possessions, others are an important mirror through which we see ourselves.

These others first come to associate possessions and possessor and then, depending upon which is known best, either come to infer the traits ofthe person from the nature of the possessions or the nature ofthe possessions from the traits of the person (Belk 1978). Belk, Bahn, and Mayer (1982) and Holman (1981) review abundant buyer behavior literature supporting this veiw. However, as Douglas and Isherwood (1979, p. 72) remind us, to think that a single item can successfully inform others about us is equivalent to thinking that a single word from a poem can convey the meaning it creates in the context ofthe poem.

Sartre’s view that having and being are the central modes of existence contrasts with Karl Marx’s view that doing, and particularly working, is central to existence and self-worth. The problem with having, in Marx’s view, is that it produces a false path to happiness through “commodity fetishism” (Marx 1978). In commodity fetishism, consumers worship goods and believe that goods have magical powers to bring happiness, provoking a pervasive and ongoing expectation that happiness lies in the next purchase or “I wouldbehappy if 1 could just have. . . “Marx suggests instead that real happiness is achieved through doing meaningful and properly rewarded work (Marx 1967). Accordingly, the perspective advocated by Marxists is that we should live to work rather than work to live (Dyke 1981). This is also the major basis for the Marxist objection to capitalism. When the capitalist owns the products of a worker’s labor, the worker has been alienated from that which s/he has created. The worker has been robbed of a part of selfThe capitalist, in Marx’s view, is seen not only as an exploiter of labor, but also as a thief of the worker’s verj self (Marx 1964).

Fromm (1976) instead advocates being as the preeminent form of existence. Like Marx, Fromm attacks “radical hedonism,” or concentration on having, as being unrewarding. He suggests that this view promotes a having mode of existence that views things, experience, time, and life itself as possessions to be acquired and retained. In the alternate being THE JOURNAL OF CONSUMER RESEARCH mode of existence that Fromm proposes, this orientation to have is rejected in favor of an opposing orientation to share, to give, and to sacrifice.

The outcome of practicing this being mode of existence, according to Fromm, is to realize one’s identity without the threat of losing it, a threat that is inherent in the having mode—for which he asks “If I am what I have and if what I have is lost, who then am I? ” (1976, p. 76), The views of Sartre, Marx, and Fromm on having, doing, and being present significant questions that are not necessary or possible to resolve here. All acknowledge, however, that having possessions functions to create and to maintain a sense of self-definition and that having, doing, and being are integrally related.

Mastery of Possessions and Human Development Self Versus Environment. The functions that possessions fulfill in our lives are not constant over our life spans. According to Freudian and other psychoanalytic theories (e. g. , Erikson 1959), the infant begins life being unable to distinguish self from the environment, including mother. As Ausubel, Sullivan, and Ives (1980) point out, this may be seen as a perceptual problem in distinguishing figure from ground. Others suggest that the distinction soon emerges as a result ofthe contingency and kinesthetic feedback produced by the infant’s actions (Lewis and Brooks 1978; Seligman 1975).

That is, as the infant’s motor skills develop, those objects that can be controlled come to be seen as self and those objects that cannot be controlled come to be seen as environment. According to Isaacs (1933. p. 226), the mother’s caregiving also produces the first sentiments of ownership: In Ihe case ofthe infant at the breast, lo have is literally and simply to take into oneself, into one’s mouth. The nipple is only here at all when it is in my mouth, when it is (in feeling) a pan of me. And to bite and swallow athingisforlongtheonly sure way of retaining it.. . .

This is the ultimate form of ownership, from which all others are derived. Even though the infant’s mother provides care, nourishment, and security, her lack of perfect responsiveness to the infant’s desires makes it likely that she is the first object that the infant regards as not self The separation from mother also has led others to suggest that the “security blanket” serves as a transitional object helping the child to feel the security of the mother through an object that symbolizes her (e. g. , Furby and Wilke 1982; Weisberg and Russell 1971; Winicott 1953).

Bowlby (1969) suggests that such material objects often aid in identity formation when children recognize their independence and sep~ arateness from their mothers. If the early changes in person-object relationships may be described as moving from being one with the POSSESSIONS AND THE EXTENDED SELF 14? environment to having objects that aid the transition to a world where self is distinct from the environment, then the next changes may be characterized as moving from having transition objects to doing things with or lo Ihe environment.

This motivation is labeled “competence” or “mastery” motivation (White 1959). Furby (1980) expanded this concept by suggesting that we develop a stronger sense of self by learning to actively control objects in our environment rather than feeling controlled by them. Furby and Wilke (1982) presented evidence showing that until six months of age the child may be most interested in simply controlling an object, whereas by twelve months the child is more interested in practicing emerging skills (e. g. , with blocks).

In both cases, producing some intended effect by doing something with an object is the goal. Self Versus Others. Data from Kline and France (1899, pp. 446-447) and Isaacs (1935) suggest that the relationship between a person and an object is never as simple as a person-thing bond, because other people often seek to control these objects: a great part ofthe value oflhose things which iittie children want to own is far from intrinsic. It arises directly from the fact that others have or want the object.

And thus we enter the openfieldofrivalry. Not to have what others have, or to have less than they, is to fee) shut out from the love and regard ofthe person giving. It is to betreatedasnot loveworthy (Isaacs 1935, p. 74). In this sense, relationships with objects are never twoway (person-thing), but always three-way (personthing-person). This brings forth a meum ei mum concern with object ownership (Beaglehole 1932). The rivalry aspects of possessions seem clear among young children.

Piaget (1932) reported that 8to 12-month-old children often display violent rage when a toy is taken from them and given to another child, Mueller (1978) and Mueller and Brenner (1977) found that between 80 and 90 percent of social interactions of children up to two years of age are focused on physical objects; the authors did not report what proportions of these interactions involved conflicts. Furby’s (1982) examination of this issue revealed that for 18- to 21-month-olds, more than 85 percent of their object-oriented interactions with peers involved conflict about retaining possession instead of sharing or giving.

Horney (1964) suggested that such competitiveness, along with other evidence of lack of affection from parents or peers, leads the child to compensate as an adult through neurotic strivings for power, prestige, and possessions. Although this may not be a complete explanation of these adult traits, it seems a more plausible basis for adult orientations toward possessions than are explanations via Freudian oral and anal fixations (Belk 1982a). Although receiving material objects may convey a sense of love and worth to the child (substituting ma- erial resources for love resources is difficult according to research by Foa and Foa 1974 and perceptual findings by Brinberg—Brinberg and Castell 1982; Brinberg and Wood 1983), from the parents’ points of view, control of their children’s material possessions offers a means o^ bringing about desired behaviors. Whiting (I960) provides a succinct model ofthis sort of resource mediated socialization: L Parents can use resources to reinforce behavior in three ways— a. Giving (e. g. , a “treat” for being “good”), b. Withholding (e. g. , no dessert until vegetables are eaten), c. Depriving (e. g. no more tc]evisioR viewing—something already “possessed”—until the child “behaves”); 2. Resources involved must be— a. Scarce (i. e. , noX freely available to ihe child), b. Valued (at the time) by the child, c. Controlled by the parent; 3. Anticipations of resource availabiJily m ihe future can also be modified to mediate behavior through— a. Threats to withhold or deprive resources, b. Promises to give resources. The way parents use such resource mediated behavioral modification not only affects behaviors—those concerning possessions as well as other ones—but also creates new attitudes toward the possessions used as reinforcements.

For example, if sweets are withheld or deprived or if threats to do so are made, these actions may enhance the value of sweets, encourage the deiay of gratifications until unpleasant tasks are completed, or instill an attitude that good performance should be followed by indulgence. The potential effects of such socialization on adult material lifestyles are envisioned easily. Adolescence and Adulthood. Erikson (1959) suggested ihat adolescents predictably undergo an “identity crisis. ” One hypothesis is that adolescents at this stage increasingly seek identity through acquiring and accumulating selected consumption objects.

Montemayor and Eisen’s (1911) study, which asked teenagers to describe who they are, found that this is true in early teenage years when respondents cited possessions, name, and location as part of who they are. However, in later teenage years, they found that respondents were more likely to cite skiils (e. g. , athletic, artistic) and traits (e. g. , expressions of moral character, self-sufficiency). A study of 8- to 30-yearold Chicagoans (Csikszentmihalyi and RochbergHallon 1981) found that this generation is more likely than its parents and grandparents to cite as favorite possessions those that reflect skills (e. .. athletic equipment) or objects which they can manipulate or 148 control (e. g. , musical instruments, stereo, pets). Material possessions such as clothing and automobiles are seen as an important source of prestige during high schooJ {Snyder ] 972), but there isprobably some tendency to ascribe such prestige to one’s family rather than to one’s self as an individual. These findings suggest that only certain types of possessions are valued as extensions of self during adolescence and that se/f-ciefinifion through doing things may de preferred to self-definition through having things.

During preretirement adulthood, Csikszentmihalyi and Rochberg-Halton (1981) found that emphasis shifts from defining oneself by what one does to defining self through what one has. Furby (1978) found that 40- to 50-year-oIds are the most likely of all age groups to cite social power and status as reasons to own personal possessions. Csikszentmihalyi (1982, THE JOURNAL OF CONSUMER RESEARCH pp. 5-6) explains: A person who owns a nice home, a new car, good furniture, the latest appliances, is recognized by others as having passed the test of personhood in our society . . , the objects we possess and consume are . . wanted because. . . they tell us things about ourselves that we need to hear in order to keep our selves from falling apart. This information includes the social recognition that follows upon the display of status symbols, but it includes also the much more private feedback provided by special household objects that objectify a person’s past, present, and future, as well as his or her close relationships. Olson (1981, 1985) found that young couples cite as favorite objects in the home those that reflect their future plans and goals, but older couples cite objects that relate to their experiences together.

Cameron (1977) conducted a series of experiments suggesting that having children is a key life event that causes the parents to become less self-focused and more focused on their children. Feibleman (1975) notes the emergence of a tendency of parents by late middle age to live vicariously through their children. At this point, children represent an extension of self, but not lo the exclusion of material possessions. In fact, Belk (1985) found parents to be more materialistic and possessive tban their children and their own parents.

Because of accumulated possessions, well-developed skills, possession of both a past and a future, and parenthood, the middle years of life also are likely to involve the most extended concept of self. Old Age. If the young are future-oriented, the old are past-oriented. Csifcszentmihafyi and RochbergHalton (1981) found that for their Chicago sample, such possessions as photographs, athletic trophies, and mementos are most treasured by grandparents. The reason most often cited for possessions being treasured by this group is that possessions have the ability to symbolize others, often because they are gifts from these important others.

Sherman and fjewman (1977) found that postretirement-age persons who possess such remembrances are happier than those who do not. McCracken (1987a) suggests that homes for the aged would do well to consider the identity deprivation that occurs when these people are made to discard possessions. Places that are especially relevant to one’s past have also been found to be particularly valued by the old because ofthe memories that places can stir (Howell 1983; Lowenthal 1975).

In contrast, the youn$ tend to value places according to the activities these places facilitate (Csikszentmihalyi and Rochberg-Halton 1981; Hart 1979). During old age, the sense of one’s own mortality also becomes more and more undeniable. With decreasing future years, declining skills and abilities, and a shrinking network of old friends, sense of self possibly contracts as well. However, this is not necessarily the case. Many people seek to assure that their selves will extend beyond their deaths.

Lifton (1973) suggests five ways through which this extension may be attempted: (1) through one’s children, (2) through belief in a life after death, (3) through one’s works (e. g. , artistic, literary, scholarly), (4) through identification with nature (which -will continue), and (5) through experiential transcendence (e. g. , absorption in music may allow one to transcend the world of here and now and symbolically be reborn). A sixth way, which is not mentioned”, is to have one’s possessions (especially those in collections one has created) “live on” through heirs or museums (Rigby and Rigby 1949).

Based on interviews with persons ages 62 to 85, interviews with their friends, relatives, and acquaintances, and an analysis of lelters, mementos, and conversations ofthe dying and their survivors, Unruh (1983) found evidence ofthe widespread use ofthis strategy. He detected first a solidification of identity through creating letters, journals, memos, and poems ihat were meant io be left behind. Second, artifacts including photographs, scrapbooks, souvenirs, and jewelry were accumulated.

And third, these artifacts were distributed to persons who were believed to be willing to care for them, and in so doing honor and remember the donor. This distribution was accomplished through predeath gifts and wills and testaments. Western society seldom elevates reverence for ancestors to the level of Far Eastern cultures such as Japan and China, but Western society does revere its heroes’ and villains’ possessions, as illustrated by pilgrimages to Elvis Presley’s Graceland mansion and William Randolf Hearst’s castle (Maines 1978). Possessions and the Sense of Past Integral to a sense of who we are is a sense of our past.

Possessions are a convenient means of storing the memories and feelings that attach our sense of past. A souvenir may make tangible some otherwise intangible travel experience. An heirloom may record and recall family heritage just as a historic monument may help to create a sense of a nation’s past. POSSESSIONS AND THE EXTENDED SELF Overall, Csikszentmihalyi and Rochberg-Halton (J 981) found that the three types of possessions that the 3 i 5 Chicago families most frequently cite as treasured are furniture, visual art (including that created by family and friends), and photographs.

In each case, the most frequently given explanation for valuing these objects is the memories they call forth of other people, occasions, and relationships. These reasons overshadow functional explanations for attachments to furniture and aesthetic reasons for valuing art objects and photographs. As one of tbeir informants explains (Rochberg-Halton 1984, p. 171): This [painting] is my great, great grandfather. I’ve had it since childhood. U’s more than just a portrait—it’s a personf I’d grab it right away in a fire. [Without it] my life would be lessened.

I’d go on living, but it would deplete my secure “lump. ” It would mean that I wouldn’t be able to hand it down to my children. The kids already say, ‘Tm gonna inherit this and that. ” . . . It’s part of Ihe continuity of who I am, where I came from, where I’m going. Older respondents are especially likely to link such objects to past experiences. One explanation is that our attachment lo memor>’-evoking possessions grows as we accumulate experiences from our past and reduce the stock of pleasurable experiences likely to occur in our futures. Also, as Kastenbaum 1977) observes, “the older person scans the past for evidence that he once was competent, once was loved, once commanded respect. ” Gifts received from others are one such evidence of love from significant others (Belk 1982c; Wallendorf and Arnold 1988). Thus, cherished possessions are not likely to be a random assortment of items that recall our pasts. Just as we pose family photographs to capture the “good” (happy) moments ofour lives and then selectively edit the best of these into albums (Chalfen 1987; Milgram 1976; Sontag 1973), we are also likely to treasure most tbose possessions associated with pleasant memories.

These possessions are likely to include objects such as newspaper clippings and trophies representing past accomplishments, mementos of past romances, and souvenirs of enjoyable travel experiences, and to exclude others such as belongings of estranged former spouses, poor report cards, and gifts from suitors who later rejected us. Note that social institutions such as museums follow a similar process in selectively retaining aesthetic, scientific, and historical cultural artifacts. Mukerji (i 978) makes a di. -tinction between goods that are initially produced as art works and are acquired and retained based on presumably aesthetic judgments, and goods that are initially produced for more utilitarian purposes but are later regarded as worthy of preservation. Although the retention criteria are somewhat different for the two classes of objects, in both cases the decisions to retain tbe object rather than reject it determine the picture ofour cultural past that is available to future generations. Obviously, we are more ikely to chronicle our cultures’ successes than their failures. The desire to know one’s individual past can explain the retention of personal memorabilia, just as the desire to remember family heritage can explain retention of family heirlooms and the desire to appreciate national history can explain museum patronage and visits to historic sites. However, what can explain the desire to acquire and collect antiques and antiquities from another lime, place, and family? Clearly, it is not a claimable sense of past that is achieved at any except the broadest level of identity.

Part ofthe answer lies in the desire to identify wiih an era, place, or person to which we believe a desirable set of traits or values adheres. At a national level, neoclassical architecture seems to have this objective. At a more personal level, owning artifacts that once belonged to a famous historical figure seems to share this objective (Rigby and Rigby 1949; Waliendorf and Belk 1987). In each case, thereseems tobeadesire to bask in the giory ofthe past in the hope that some of it will magically rub off—a form of positive contaminaiion (Levi-Strauss 1963).

This nostalgic desire to gain the glory ofthe superstar or of a mythical golden age of the past shares something in common with the tendency McCracken (1988) describes as depositing and retrieving cultural meaning in places where it is unlikely to be disturbed by contradictions present in reality (e. g. , Davis 1979). Another reason tor the accumulalion of antiquities that are found or acquired rather than inherited or claimed on the basis of a more direct linkage lo the extended self is that antiques are rare and therefore potentially serve as symbols of status or “status markers” (Douglas and Isheru-ood 1979).

Other motives might be found in the amusement of collecting curiosities, aesthetic preference for antiques over currently produced artifacts, and a preference for handcrafted works over current mass-produced works. However, each of these additional explanations relies on something of the extended self of the previous owner, artist, or craftsperson adhering to the work. Just as we seek to extend our selves by incorporating or owning certain objects, we may stili seek the sympathetic magic (contagion)