You are on page 1of 18

How Is a Possession ''Me" or ''Not Me"?

Characterizing Types and an Antecedent


of Material Possession Attachment
SUSAN SCHULTZ KLEINE
ROBERT E. KLEINE III
CHRIS T. ALLEN*

Material possession attachment, a property of the relationship between a specific


person and a specific object of possession, reflects the extent of "me-ness" as-
sociated with that possession. The two Q-methodological studies reported here
investigated the nature of this me-ness (and "not me-ness"). Study 1 explores
different types of attachment and how these types portray various facets of a per-
son's life story (i.e., identity). It shows how strong versus weak attachment, affiliation
and/or autonomy seeking, and past, present, or future temporal orientation combine
to form qualitatively distinct types of psychological significance. Study 2 begins
development of a nomological network encompassing attachment by showing how
mode of gift receipt (self-gift vs. interpersonal gift), as an antecedent, influences
attachment type. Study 2 also examines aspects of successful and unsuccessful
gifts. Both studies demonstrate that unidimensional affect fails to adequately descritje
or explain attachment. Together, the two studies suggest a more parsimonious way
to represent person-possession relationships than has been offered in previous
studies. Moreover, the findings heip delineate the boundaries of attachment (e.g..
What does it mean to say a possession is "not me"?).

M aterial possession attachment is a property ofthe


relationship between a specific individual and a
specific material possession. Possessions of strong at-
one time in my life," a daughter's ring portrays her
mother's love, and another person's piece of furniture
reflects his family heritage. Another kind of attachment
tachment are more closely held to the proximal self, portrays key' aspects of a person's individuality. Ex-
are more affectively charged or cathected (Ball and Ta- amples include a suit that "makes me feel good about
saki 1992; Belk 1987, 1988; Wallendorf and Arnould myself"; a ring that signifies "my first real job"; or pos-
1988), and are more positively valenced (Schultz, sessions that fit "me." Least favorite possessions (weak
Kleine, and Kernan 1989) than objects of lesser attach- attachments) often represent a period in life from which
ment. a person wishes to disconnect, or objects that are "not
Possessions to which there is attachment help narrate me." In this way, attachments help narrate the devel-
a person's life story; they reflect "my life." One kind of opment of a person's life story.
strong attachment reflects a person's desirable connec- Conceptualized as artifacts of life stories, strong at-
tions with others. For example, one person's photo- tachments are characterized by afliliation and auton-
graphs signify "people who were important to me at omy seeking and a specific temporal orientation (e.g.,
past, present, or future; Schultz et al. 1989). Different
combinations of these characteristics yield qualitatively
*Susan Kleine and Robert Kleine are assistant professors of mar- distinct types of attachment; each type is used for cul-
keting, College of Business, Arizona State University, Tempe, AZ tivating a different facet of one's life story, as the two
85287-4106. Chris T. Allen is professor of marketing. University of
Cincinnati, Cincinnati, OH 45221. Study 1 is based on the first au- studies we describe below portray. Study 1 describes
thor's dissertation (University of Cincinnati, 1989), which Chris Allen qualitatively distinct types of attachment that individ-
supervised; Karen A. Machleit and David C. Lundgren were com- uals use to portray their life stories. Study 2 begins de-
mittee members and their contributions gratefully are acknowledged. veloping a nomological network encompassing attach-
The authors thank Dwayne Ball, Marsha Richins, and Peter Reingen
for their many helpful comments on earlier drafts, and Deborah Lav- ment by investigating mode of gift receipt as an
erie and Julie Huntley for their help with study 2 data collection. antecedent that influences type of attachment. Study 2
Last, but not least, the authors sincerely appreciated the many helpful also examines aspects of successful and unsuccessful
comments of the editor, associate editor, and two reviewers. Their gifts. Together, the two studies offer a more parsimo-
suggestions led to substantial improvements in the article.
nious way to represent material possession attachment
327
© 1995 by JOURNAL O F CONSUMER RESEARCH. Inc. • Vol. 22 • [Jccember 1995
All rights reserved. 0O93-5301/96/2203-OOO7$2.0O
328 JOURNAL OF CONSUMER RESEARCH

than has been offered in previous studies. We also dem- Affiliation and autonomy seeking underlie life-stage
onstrate that attachment is not unidimensional evalu- development theories (see, e.g., Kegan 1982; McAdams
ative affect (cf. Wallendorf and Arnould 1988). Finally, 1985). At certain times in our lives we emphasize one
as we examine both strong and weak attachments, our task or the other. For example, the identity stages of
findings illuminate what it means to say a possession is toddlers or adolescents emphasize the autonomy-seek-
nor me (Cohen 1989). ing theme; at other times affiliation seeking dominates
a particular life stage. Maturity is defined by an ability
LIFE STORIES AND TYPES to engage both aspects of one's identity when needed
OF ATTACHMENT (Altman 1976; Kegan 1982). Although cultural differ-
ences affect the degree to which each mode is valued,
Investigators generally agree that individuals use at- both are experienced by men and women of all ages
tachments to define and maintain their identities (e.g.. (Kegan 1982). Indeed, our constant struggle to negotiate
Ball and Tasaki 1992; Belk 1988; Csikszentmihalyi and affiliation and autonomy seeking "may be our experi-
Rochberg-Halton 1981;Dittmar 1992; Kamptner 1991; ence of the unitary, restless, creative motion of life it-
Myers 1985; Schultz et al. 1989; Wallendorf and Ar- self" (Kegan 1982, p. 107).
nould 1988). Identity is reflected in one's life narrative, Thus, if possessions to which there is attachment re-
or life story, capturing various roles including past, flect identity development, they should exhibit affilia-
present, and anticipated future selves. My life narrative tion and/or autonomy seeking (Csikszentmihalyi and
describes the path of my identity development; it defines Rochberg-Halton 1981; Schultz etal. 1989; Wallendorf
who I am, who I have been, who I am becoming, and/ and Arnould 1988). Conversely, affiliation or autonomy
or who I am no longer. will tend not to be associated with possessions to which
Identity development is often examined through life individuals are relatively unattached (i.e., not me;
stages (e.g., Erikson 1968; Kamptner 1991). Yet there Schultz et al. 1989). An object of low attachment thus
are basic, archetypal modes of existence that underlie indicates the absence of identity relevance or, in some
life stages that motivate identity development (Kegan cases, its "has been/it's not me any longer" quality.
1982) and are reflected as basic themes in life narratives
(McAdams 1985). These provide the fundamental di- Archetypal Theme 2: Managing Identity Change ver-
mensions of attachment. sus Stability. A second protean theme in the identity
literatures is the evolutionary character of life stories.
Themes Reflected in Life Stories Life narratives include an individual's current under-
standing of the present self, of past selves that are being
A broad survey of the literatures about the self and carried on, of past selves that are being let go, and of
human nature reveals two archetypal modes of human anticipated selves (Erikson 1968; McAdams 1985). The
behavior that motivate development of the self: (1) af- dialectic tension between stability (i.e., maintaining a
filiation versus autonomy seeking and (2) temporal facet of identity) and change (i.e., acquiring or discard-
change versus stability management. Although there are ing a facet of identity) is another motivator of devel-
ostensibly other themes in the self literatures, one is opment of the self.
struck by the universal use of these two (e.g., Kegan
1982). Moreover, life narratives reflect these two themes Temporal change versus stability thus should be re-
flected in possessions to which there is attachment (Belk
(McAdams 1985).
1988, 1990; Schultz et al. 1989). By their very nature,
Archetypal Theme 1: Affiliation versus Autonomy concrete material things help us to maintain a "personal
Seeking. People are motivated universally to establish archive or museum" (Belk 1988, p. 159), providing a
and maintain a personal and unique identity, distinct sense of permanence in the world. Material objects are
from that of others (i.e., autonomy seeking), while at also used in the processes of becoming by signaling that
the same time they are motivated to maintain inter- a desired identity is developing (Markus and Wurf 1987;
personal connections that also define the self (i.e., af- Solomon 1983; Wicklund and Gollwitzer 1982). Fi-
filiation seeking). Possessions reflect autonomy seeking nally, we sometimes use material possessions to dis-
when they evidence individual accomplishments, dis- connect from, or to shed, old selves (Belk, Sherry, and
tinctiveness, uniqueness, independence, self-control, or Wallendorf 1988; Myers 1985). By dispossessing our-
other aspects of individual integrity (Schultz et al. 1989). selves of a symbol of a former self, we aid the process
Likewise, affiliation seeking is apparent when posses- of leaving the past behind. Thus, possessions create a
sions reflect connections with others, with one's heritage tangible residue of past, present, and possibly antici-
or tradition, or with occasions spent with important pated future identity development.
others or reflect being in touch with or cared for by A special possession, therefore, could facilitate self-
others (Schultz et al. 1989). Although Western psy- continuity by connecting a person with a desirable past
chology traditionally equates the autonomous self with self (e.g., memories), a present self (me now), or a future
a person's "me," this conceptualization suggests that self (who I am becoming). A least favorite possession
affiliation-oriented possessions are just as much me as could aid self-change by disconnecting an individual
those that reflect a person's individuality. from a past, "has been," no-longer-desired self. Thus,
MATERIAL POSSESSION ATTACHMENT 329

affiliation, autonomy, and temporal orientation interact sion is a cherished one"). When participants' Q-sort
to symbolize different facets of a person's life story. data are factor analyzed, each resulting factor represents
a group of people who ranked Q-sample stimuli simi-
Types of Attachments larly. For this study, each factor represents a type of
attachment. Interpretation of each factor is based on
The proposed conceptualization portrays attachment the Q-sample statements that are most and least char-
as though it were located in a multidimensional space. acteristic of that factor (i.e., attachment type). These
A point in this attachment space represents a specific statements come from a Q-sort model, or prototypical
combination of more or less affiliation, more or less Q-sort, constructed for each factor. We also obtained
autonomy, and past, present, or future temporal ori- responses to open-ended questions to more fully inter-
entation. This spatial metaphor suggests there are qual- pret the factors (Schlinger 1969).
itatively distinct types of attachments depending on the
portion of the life story each attachment reflects. For We chose Q-methodology for several reasons. The
example, an individual's family heirloom reflects a past- sorting task serves as a projective device (Kerlinger
affiliation mode. This same person's attachment to an 1988). Through sorting, subjects express the relative
engagement ring reflects a future, affiliative self. A third importance of various factors explaining their attach-
favorite possession, such as a sports trophy, reveals a ment (something impossible with a measurement scale).
present, autonomy-seeking aspect of the person's life This allowed participants to reveal which combinations
story. Finally, an old T-shirt (a least favorite possession) of the proposed attachment dimensions were most use-
represents a past, now soured relationship. ful for describing their subjective viewpoints. Moreover,
the method allowed us to examine whether possession
The point is that individuals possess portfolios of at- liking or disliking is important for explaining attach-
tachments—each attachment reflecting different com- ments. Last, Q-methodology uniquely walks the delicate
binations of affiliation, autonomy seeking, and past, line between discovery and confirmation. The investi-
present, or future temporal orientation. These quali- gator places general premises in the open while bringing
tatively distinct attachment types reflect different facets to the fore subjects' perspectives, which become the ba-
of the life narrative. For example, accomplishing affil- sis for conclusions. Thus, the method is ideal for in-
iative we-ness is a very different, and often opposing, vestigating phenomena, like attachment, that are con-
task from autonomous me-ness. Moreover, temporal ceptually underdeveloped (i.e., have no empirically
orientation helps explain the self-related tasks repre- supported nomological network).
sented by a particular possession. Elements of our past
arefixed,done, and cannot be changed. We may choose Participant Selection. Thirty individuals performed
to keep certain possessions as artifacts of our life stories Q-sorts, a sample size common in Q-method research
or dispose of them. Yet present-oriented possessions (e.g.. Brown 1980; Kerlinger 1988; McKeown and
reflect tasks in the here and now; they reflect issues of Thomas 1988). The participants were college students
who I am now that currently are being cultivated. between the ages of 20 and 28. All were junior and se-
This line of thought suggests that affiliation, auton- nior co-op students working their way through college.
omy, and temporal orientation work together for iden- Several were married or engaged to be married; all came
tifying fully the reasons for a possession attachment and from midwestern or eastern U.S. blue-collar or middle-
how it reflects the life narrative. For the two studies class backgrounds.
reported below, we were interested in exploring how The Q-Sample. The Q-sample is the set of stimuli
temporal orientation combines with affiliation and au- participants Q-sort; we designed 54 written statements
tonomy seeking to describe and explain qualitatively for our stimuli (see the Appendix). To thoroughly op-
distinct attachment types. For example, is affiliation or erationalize our conceptualization of attachment, we
autonomy more often focused on the workable present used a factorial design model to construct Q-sample
or the unchangeable past aspects of one's identity? What statements 1-42 (Brown 1980; Kerlinger 1988). The
self-developmental issues are revealed through least fa- "factors" correspond to the dimensions of attachment.
vorite possessions that people still own? Our design is a 2 (strong attachment/me or weak at-
tachment/not me) X 2 (affiliation or autonomy) X 3
STUDY 1 (past, present, or future temporal orientation) design.
Protocol data from Schultz et al.'s (1989) study pro-
Methodology vided a pool of natural language from which we pur-
Study 1 explored different types of attachment by posively sampled phrases about most and least favorite
eliciting individuals' personal sentiments of favoriteness possessions. In that study, judges coded subjects' open-
and unfavoriteness about a wide variety of possessions. ended responses about most and least favorite posses-
We accomplished this through Q-methodology (Ste- sions into the categories of affiliation, autonomy, and
phenson 1953). Q-Methodology includes the Q-sort, past, present, or future orientation. Use of respondents'
which is a comparative rating task in which individuals terminology facilitates sorting and reduces our chance
rank order a Q-sample (i.e., a set of stimuli) with respect of misinterpreting respondents' meanings (McKeown
to certain instructions (e.g., "Describe why this posses- and Thomas 1988). Finally, Q-sample statements 43-
330 JOURNAL OF CONSUMER RESEARCH

54 (see the Appendix) reflect positive, negative, or neu- their second most favorite thing from their favorite
tral affect that appeared frequently in Sehultz et al.'s possession list to be used for sorting task 1.
data. The statements for liking (no. 43) and disliking
(no. 54) represent Wallendorf and Arnould's (1988) Q-Sort Administration. After filling out the posses-
operationalization of attachment as approach-avoid- sion-elicitation questionnaire, each participant per-
ance affect. formed four sorting tasks: (1) most favorite possession,
The a priori categories for the Q-sample items serve (2) most favorite piece of clothing, (3) car, and (4) least
to ensure domain coverage yet are designated through favorite piece of clothing.
our external observer's viewpoint. The a priori desig- Attached to a table were 54 blank, yellow, 3" X 5"
nations represent a template that is kept in view during index cards, arranged in an 11-column quasi-normal
interpretation, but the investigators' primary respon- pattern into which subjects were asked to fit the 54 Q-
sibility is to let the participants speak for themselves sample statements.^ Each of the 11 columns was marked
through the Q-sorting procedure. Thus, the template with a score from +5 (most characteristic) to - 5 (least
remains fluid and open to change throughout the in- characteristic); the 0 column was labeled "neutral/don't
terpretation of results, which implies that a statement's know/irrelevant."
prior designation may be changed during interpretation. Each participant was given a brief overview of the
procedure. Then the participant was given a well-shuf-
Possession Elicitation. We reasoned that a variety of fled pile of 3" X 5" white index cards. On each card was
possessions would lend itself to exploration of different written one of the 54 Q-sample statements. The partic-
attachment types. Yet, because attachment is not prod- ipant first familiarized himself/herself with the state-
uct category specific, we also wished to see how attach- ments. Next, the participant divided the cards into three
ment might vary within category. Thus, from each par- piles, representing those statements that were (a) most
ticipant, we elicited four possessions: most cherished characteristic of the possession, {b) least characteristic
possession, most favorite piece of clothing, least favorite (most uncharacteristic) of the possession, and (c) the
piece of clothing, and car. Upon arrival at the research remainder. After this initial sorting, the participant then
location, each participant filled out a questionnaire sorted the cards into the 11 columns of" the Q-sort dis-
which elicited the four possessions for which Q-sorts tribution.
would be performed. Investigators often regard posses- The participant was instructed, once s/he had sorted
sions that are "most cherished" or "most favorite" to all 54 cards, to reexamine the sort to be sure it repre-
be strong attachments (e.g., Dittmar 1992; Kamptner sented his/her thoughts and to rearrange statements as
1991; Myers 1985; Sehultz et al. 1989; Wallendorf and necessary. When the participant signaled completion,
Arnould 1988). So, the questionnaire asked the partic- the investigator collected the cards and recorded the
ipant to "please think about the various things you responses. The participant then responded, in writing,
have—your material possessions. Think about your to open-ended questions that asked (for the most cher-
most favorite, most cherished, most difficult to part with ished possession sorting task): (1) "How did you acquire
material possessions. These things do not have to be it?" (2)"Where do you keep it and why?" and (3) "Why
expensive or worth a lot of money—^just anything you is it a favorite?" For the most and least favorite clothing
have which is very important to you. List below your sorting tasks, participants were asked, (1) "How did you
most favorite, most cherished, most difficult to part with acquire it?" and (2) "Why is it (not) a favorite?" For
material possessions." The participant listed two to five the car, the questionnaire asked, "Why did you sort the
most cherished possessions. The next instruction stated: cards the way you did?" This sequence was repeated
"Now, please go back and select the one possession for each sorting task. Each participant was then de-
which is your most favorite of those you listed. Then, briefed.
circle that object." Participants then answered similar
questions that asked for "your most favorite piece of
clothing" and "your least favorite piece of clothing."' Data Analysis
Clothing was selected because it has abundant meaning
for the self (Solomon 1985), and pretesting revealed Factor Analysis of Q-Sort Data. The data for all
that subjects could respond about their attachment to four sorting tasks combined were analyzed according
most or least favorite clothing. If the participants listed to standard Q-methodology procedures (see, e.g.. Brown
a car as their most favorite, they were asked to circle 1980). First, we performed a principal components
analysis on the 54 (number of items in the Q-sample)
X 111 (number of participants' possessions) data ma-

'In another study, student participants sorted a set of 48 ordinary


possessions from "like me" to "not like me" (Belk 1987, 1988). This ^The number of statements (bottom row below) placed beneath
possession set included commonly owned objects—e.g., photographs, each of the 11 columns (-5 to +5 in top row below) appeared as
car, wallet, and jeans—and included a participant's most and least follows:
favorite possessions. Results showed that favorite possessions are more
central (like me) and least favorites less central (not like me) to the -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5
self. 3 3 4 6 8 9 8 6 4 3 3
MATERIAL POSSESSION ATTACHMENT 331

TABLE 1
STUDY 1 OBJECTS LOADING ON EACH FACTOR

Keeping memories of others It's me and 1 like it It's not me anymore Utiiity Breaking away

Objects loading on factor Pictures, snapshots, photo Shorts, stereo, car, summer Blouse, s/i/rt," sundress,' Car, car, car. Car, T-shirt, car.
aibums, wedding ring, family dress, jeans, jeans. T-stjirt,' sport coat,' car, car. -shorts,'
pictures, computer, car, car. jeans, shirt, suit. sttirt, skirt outtit,' car, car. -wool jacket
photo aibums, stereo. overalls," car, red suit. sweatstiirt,' vest,' -car
sweater,* car, diamond ring," T-sttirt, phone, tie, dress. skirt,' Hawaiian stiirt.
ring," music, fishing rod, car. sweater, biouse, car. sweater, car, jacket,'
coin collection, summer stereo, jeans, car, shirt. jeans, gym shorts.
dress, car, dress,* T-shirt, jacket, motorcycle, car. car, sweater, jeans.
bracelet," sweater, letterman golf clubs, car, cassette pink shirt,' sweater,'
jacket, engagement ring, fiy recorder, ring skirt, old shorts, -car
rod, dog, higt> school
uniform, car, car, trophies.
boxer shorts, dad's wailet.
shorts, car, flannel shirt," suit
pants, —television, —suit
vest, -T-stiirts,' -famiiy
pillow, —car, —car
Percentage gifts 21 4 52 n.a. 33
(7/34) (1/24) (11/21) (1/3)
Percentage most favorite 88 100 9 n.a. 67
(30/34) (24/24) (2/21) (2/3)
Percentage female 43 43 37 37 40
(19/44) (13/30) (9/24) (3/8) (2/5)

NOTE. Least favorite objects are in //a//cs. Minus sign before object indicates it loaded negatively on the factor. Percentage gifts and percentage most favorite
objects loading on a factor are caicuiated after removing the number of cars that loaded on that factor. The car, as is described in the text, was not specified a priori
as most or least favorite, and we did not ask participants to specify whether it was received as a gift, n.a.. Not applicable.
"Object received as a gift.

trix.^ Note that, because the possessions are the "vari- pretation of each factor's viewpoint is based on that
ables," sample size for analysis is determined by the 54 factor's Q-sort model.
Q-sample stimuli. This analysis yields factors of pos- Constructing the Q-Sort Models. A Q-sort model is
sessions. Participants whose possessions group on a constructed for each factor. Conceptually, the Q-sort
particular factor sorted the statements essentially the model is a distribution of Q-sample statements proto-
same way, from the many thousands of possible ways typical of the viewpoint of those individuals who load
to sort the 54 Q-sample statements."* heavily on the factor. The Q-sort model is constructed
In this study, each factor represents a type of posses- by first computing factor scores for each Q-sample
sion attachment. If there were only one type of posses- statement. The larger the magnitude of a statement's
sion attachment, all the participants would sort the Q- factor score (i.e., absolute value), the more characteristic
sample similarly. This would yield one factor with most or prototypical that statement is of the factor's view-
favorites loading positively and least favorites nega- point. Next, on a factor-by-factor basis, the Q-sample
tively. Alternatively, if participants' did not find the Q-
sample statements useful for expressing their attach- statements are arranged according to factor score—from
ments, we would not be able to interpret the factors. highest positive to lowest negative score. Finally, to fa-
The conceptualization is upheld if the factors "corre- cilitate interpretation and presentation, each ofthe 54
spond to major variables of the theory" or conceptu- Q-sample statements is assigned to a column in the Q-
alization (Kerlinger 1988, p. 517). The study 1 data sort distribution used in data collection (i.e., assigned
yielded five interpretable factors (see Table 1). Inter- a score from +5 to —5). This is done by assigning the
three Q-sample statements with the largest positive fac-
tor scores to the +5 (most characteristic) column ofthe
••Participants who failed to sort into the Q-sort distribution were
Q-sort distribution. The next three largest positive factor
omitted from analysis. Resulting sample sizes were sorting task 1 scores are assigned to the next (+4) column in the Q-
= 29, sorting task 2 = 28, sorting task 3 = 29, and sorting task 4 sort distribution. This process continues until all that
= 25, remains are the three Q-sample statements with the
'Varimax rotation consistently gave superior results with these data. most negative factor scores; these statements are as-
Factors are retained more for their theoretical, as opposed to statistical,
significance (Brown 1980, p. 46). Heuristics used to determine the signed to the least characteristic (—5) column. The re-
number of factors to retain include (a) Retain factors on which at sulting Q-sort model reveals a type of attachment.
least two persons load highly (Brown 1980), {b) Use Schlinger's (1969)
index numbers, and (c) Retain conceptually sensible factors (Brown
Interpreting the Q-Sort Model. Factor interpreta-
1980; Kerlinger 1988; McKeown and Thomas 1988). Scree plots tion is based on the Q-sample statements most and least
(Kerlinger 1988) also can be used as rules of thumb. indicative of a factor. Each statement's a priori category
332 JOURNAL OF CONSUMER RESEARCH

serves as a template for interpretation. But until the Responses to the open-ended questions reinforce this
researcher observes how participants use each statement viewpoint. Factor 1 possessions signify "people who
vis-a-vis the entire Q-sample, the template remains fluid were important to me at one time" (participant 5's
and open to change. For example, statement 4, "It ties [P5's] photographs) or "belonged to someone very dear
me with memories and experiences," was frequently to me" (P25's grandmother's charm bracelet). Partici-
used in pretest protocols coded as autonomy seeking. pant 29's computer came from his brother, who "taught
Yet subjects might use statement 4 in conjunction with me how to use it before he passed away." Participant
affiliation statements, whereupon we would consider it 19 acquired interest in his favorite music from his older
to be portraying an aspect of affiliation. Participants' brother and friends; PI 8 acquired his varsity letterman's
written responses to the open-ended questions supple- jacket through team participation. The eight cars load-
ment interpretation. ing positively on Fl have sentimental associations such
as P82's: "It's an old car and I believe one should not
Results forget the past. My father likes it and has worked on it
. . . it reminds me of him."
Through factor analysis we identified five groups of The past orientation of Fl possessions provides a
participants, each representing a different type of at- sense of personal history:
tachment (see Table 1). Table 1 shows each possession
and whether the possession was received as a gift. 4. It ties me with memories
As we describe each factor's viewpoint, we report the and experiences. +5 0+4 0 +5
most and least indicative Q-sort model statements. We 44. It makes me feel senti-
list to the right of each statement five factor scores rep- mental or nostalgic. -¥4 - 1 +1 +1 +3
resenting the ranking of the statement for each of the 1. It reminds me of past ac-
five factors. An italicized factor score corresponds to complishments or other
the factor being discussed. A large positive score (+5, important events in my
+4, +3) indicates a highly indicative statement; a large life. +5 0+1 - 1 +1
negative score (-5, - 4 , -3) shows a most opposite —25. It doesn't have anything
statement for a factor. Reporting all five scores aids in to do with past events in
cross-factor comparisons. For example, we report my life. -3 0-3 0 -5
statement 8 as -34. It does not remind me of
8. It fits my image. 0 +5 - 5 - 2 +4 anyone from my past. - 5 +2 - 1 - 3 -2
So statement 8 is not indicative of factor 1 (0) or factor The past affiliation may be vvith important others "who
4 (-2), most indicative of factor 2 (+5) and factor 5 I never got to know" (P26's photographs), or with
(+4), and the opposite of factor 3 (-5) for which it someone who remains part of the participant's life
indicates a possession that "does not fit my image." (P27's wedding ring). Moreover, the intent to carry per-
sonal history into the future reveals the self-continuity
Factor 1 (Fl): Keeping Memories of Others. Forty- theme of Fl:
four possessions loaded on this factor, six of them neg-
atively. Participants ascribed affiliation-seeking motives - 3 3 . I probably will not care
to these possessions: about it in the future. -3 +2 +4 +2 +2
31. I do not plan to have it in
16. It reminds me of a person the future. -3 +1 +3 +4 0
who is important to me
now; +5 - 1 +2 - 2 - 4 Affect statements indicate the possessions' emotional
17. It reminds me of my par- significance:
ents or other family mem- 46. It makes me happy. . + i +4 - 3 - 2 0
bers. +5 - 1 +1 +1 - 4 45. It reminds me of some-
15. It reminds me of good thing I feel proud of. +3 0 -2 +1 -1
times with family or —49. I would not be upset if it
friends. +•# +1 +2 - 1 +1 were lost or stolen.
18. It makes me feel loved or -4 +1 +5 -5 -1
-48. It has no special meaning
cared for. +4 - 2 - 1 - 2 - 4 for me—it's just some-
12. It reminds me of an im- thing I have. -5+1+1+3-3
portant friend from my —47. I have no feelings for it. —4—\ +2 - 4 +1
past. +5-3 0-1+3
- 5 1 . I really do not like it at all. -3 - 4 +4 - 4 - 2
—37. It does nothing to help me
feel loved or cared for. -5+1 0+1+3 Yet liking of an Fl possession (Q-sample statement 43)
-40. It is unlikely that I will does not explain its significance. Affiliation and self-
keep it as a reminder of maintenance motives are better explanators ofthe Fl
my friends or family. -4 +3 + 2 + 2 0 type of attachment.
MATERIAL POSSESSION ATTACHMENT 333

43. I really like it. +2 +5 -5 +2 +2 Temporally, the F2 type of attachment reflects the
here and now. And, although there is hint of the future,
Two least favorite clothing items loaded on this fac- the possession is not central to an aspirational self:
tor. Participant 9O's high school uniform (that she keeps
yet no longer uses) reflects sentimental associations, 11. It shows who I am becom-
while P99's uncomfortable suit pants have affiliative ing. 0 + i - 2 - 4 +1
connotations because he received them as a gift. 21. It makes me similar to
Six possessions loaded negatively on this factor, re- people I aspire to be like. 0 +J —3 —3 — 1
flecting a viewpoint that is the opposite of the Fl view- 27. I could easily be myself
point just described. Participant 97's least favorite without it. +2 +4 +5 0 +2
clothing item (T-shirts) "represent an image and a pe- 32. I could become who I
riod in life which I prefer not to be associated with," want to be without it. +1+5+5 0+1
demonstrating an intent to disconnect from the past, Consistent with F2's self-esteem connotations is a
or a self-change motive. Thus, we conclude that Fl rep- tone of liking and contentment:
resents a type of attachment that is characterized by
affiliation and self-continuity seeking. 43. I really like it. +2 +5 - 5 +4 +2
- 5 1 . 1 really do not like it any-
Factor 2 (F2): It's Me, and I Like It! For its 30 more. - 3 -4 +4 - 4 - 2
possessions, F2 reflects autonomy seeking with a present 46. It makes me happy. +3 +4 - 3 +2 0
temporal orientation. In contrast to Fl, F2 has a dis- -52. It makes me sad. 0 - 5 - 1 - 1 +1
tinctive me orientation: - 5 3 . It reminds me of some-
thing I feel bad about. - 2 - 5 - 1 - 2 +2
8. It fits my image. 0 +5 - 5 - 2 +4
5. It is a statement of who I A number of the F2 possessions reflect a realized de-
am now. 0 +4 - 4 - 3 0 sire such as P28's Guess overalls, which "I wanted really
-26. It doesn't fit my image any- bad." Again the self-assertion of one's tastes, style, and
more. 0-4+3+2-1 personality—one's me—is evident. Thus, we conclude
that F2 reflects present-oriented autonomy seeking.
-24. It used to be me, but not
anymore. -2-3 0-1+2 Factor 3 (F3): It's Not Me Anymore. Twenty-three
-38. It makes me feel out of possessions loaded positively on this factor; one loaded
style. -1-^+2+1 0 negatively. Participants would not mind if these pos-
sessions were removed from their lives and seem to be
Factor 2's self-esteem flavor reflects the me of the pres- contemplating disposition:
ent, an autonomous self, able to reflect on, appreciate,
and enjoy its own image. A blouse makes P2 "feel ex- 49. I would not be upset if it
pensive and unique"; P29's red summer suit "makes were lost or stolen. -4+1+5-5-1
me feel good about myself (more confident) when I 33. I probably will not care
wear it." about it in the future. - 3 +2 +4 +2 +2
Person-possession "fit" is also evident in P37's Notre 31. I do not plan to have it in
Dame sweater, which "represents a college favorite," the future. - 3 +1 +3 +4 0
and P24's old concert T-shirts, which represent "my These possessions do not fit. They are "not me now"
favorite group." Golf clubs (P33), stereos (P9 and P24), and "not who I am becoming":
a motorcycle (P7), or cassette recorder (PI 1) represent
playful activities that define me (I am a golfer or I like 26. It doesn't fit my image any-
The Who). Leisure pursuits or athletic play may also more. 0-4+5+2-1
be motivated by unique opportunities for accomplish- - 8 . It fits my image. 0 +5 - 5 - 2 +4
ment (i.e., an element of autonomy seeking). Moreover, - 5 . It is a statement of who I
F2 is distinctively not affiliation oriented: am now. 0 +4 -4 - 3 0
27. I could easily be myself
40. It is unlikely that I will without it. +2 +4 +5 0 +2
keep it as a reminder of 32. I could become who I
friends or family. - 4 +3 +2 +2 0 want to be without it. +1+5+5 0+1
- 1 3 . It reminds me of an im- 41. It has nothing to do with
portant family member my future. +1 0 + i +1 - 2
who is gone now. +1 - 5 - 4 +2 - 1 -10. It is a statement of who I
-20. It will help me carry on a would like to be. -1+2-5-5 0
tradition. +2 -3 - 3 - 1 0 - 9 . It has to do with my goals
—35. It reminds me of some- for the future. 0 + 1 -4 - 3 - 2
one from my past I —21. It makes me similar to
would rather forget. - 2 -3 - 1 - 2 +5 people I aspire to be like. 0 +3 —J —3 — 1
334 JOURNAL OF CONSUMER RESEARCH

Participant 22's collarless blue shirt "doesn't do me 50. It is basically a functional,


any justice" and P34's pink dress shirt is "not very utilitarian object. -1 -1 0+5-1
masculine." Generally, these possessions are out of style 29. It is strictly utilitarian. -1-2 0+5-3
and do notfitthe participant's self-image, taste, or pref- 3. It has always related well
erences. Any connection with the object is a carryover to my needs. +1 +2 - 2 +J - 3
from the past: 6. It makes me feel indepen-
dent and self-sufficient. + 1 + 2 - 2 +5 +3
4. It ties me with memories
and experiences. +5 0+4 0 +5 Questionnaire responses reinforce this utilitarian ori-
-25. It doesn't have anything to entation. The car is "practical, basic transportation to
me" (P24), and "it gets me where I want to go and
do with past events in my
doesn't leave me stranded" (PI2). These participants
life. -3 0-3 0 -5 like their cars, but there is no evidence of attachment:
For example, P38 says "I used to like v-neck sweaters 43. I really like it. +2 +5 - 5 +4 +2
but now I very much dislike them." Participant 28's —51. I really do not like it any-
jean skirt is "no longer me." And PlO's secondhand more. - 3 - 4 +4 -4 - 2
sports coat from his grandfather captures "bad mem- 48. It has no special meaning
ories." Not surprisingly, F3 possessions are associated to me—it's just something
with negative aflfect: I have. - 5 +1 +1 +3 -3.
51.1 really do not like it any- Furthermore, F4 cars are neither affiliation nor au-
more. tonomy oriented:
- 3 - 4 +4 - 4 - 2
—43. I really like it. +2 +5 - 5 +4 +2 36. It has nothing to do with
-46. It makes me happy. +3 +4 -3 +2 0 my heritage. -2+1 0 +'^ - 2
—22. It has to do with an im-
Why, then, do participants keep these possessions? portant relationship. +1 - 2 - 2 -5 -3
We had asked participants how they acquired each of - 5 . It is a statement of who I
the possessions (except for the car sorting task). We un- am now. 0 +4 - 4 -3 0
covered an interesting pattern with respect to gift receipt -34. It does not remind me of
(see Table 1): 52 percent of the F3 (not me) noncar anyone from my past. - 5 +2 - 1 -3 -2
possessions were gifts, whereas, of the noncar posses-
sions associated with affiliation seeking (Fl), 21 percent These possessions may be obstacles to emerging
were received as gifts, and only 4 percent of the noncar selves:
autonomy possessions (F2) were gifts. Thus, having re- 39. It makes me diflferent
ceived a possession as a gift is more strongly associated from people I aspire to be
with weaker, not-me attachment (F3) than with stronger like. -1 0 0 + i +2
attachment (x2(F3 vs. Fl) = 6.41, clf= \,p< .05, Cra-
—21. It makes me similar to
mer coefficient C = .34; X^(F3 vs. F2) = 13.32, df= 1,
p < .05, C = .54). Apparently, gift receipt adds a layer people I aspire to be like. 0 +3 - 3 - J +1
of affiliation that makes even not-me objects more dif- 31. I do not plan to have it in
ficult to dispossess. Moreover, gift receipt may distin- the future. -3+1+3+4 0
guish types of attachment (X^(F1 vs. F2) = 3.99, df= 1 -10. It is a statement of who I
p < .05, C = .26). would like to be. -1+2-5-5 0
— 11. It shows who I am becom-
Two more of the F3 possessions connect their par- ing. 0 +3 - 2 -4 +1
ticipants to other people: P27's Hawaiian shirt, which —9. It has to do with my goals
he bought for a dance he attended with a girlfriend, and for the future. 0 + 1 -4 - i -2
P31 's pea-green blouse, which she bought for a Hallow-
een party. These affiliative connotations feed F3's am- Qverall, F4 represents possessions that serve present
bivalence—"It's no longer for me, but I can't get rid of utilitarian needs, and there is no evidence of emotional
it"—which is unique to this type of attachment. In fact, connections to them. Factor 4 portrays functional pos-
we have discovered that "least favorites" are not nec- session importance that is not explained by affiliation,
essarily weak attachments, but instead a different type autonomy, or temporal orientation.
of attachment that reflects emerging or anticipated self- Factor 5 (F5): Breaking Away. Three participants
change. loaded positively and two negatively on F5. Foremost,
F5 has a heavy past element:
Factor 4 (F4): Utilitarian Possessions. Seven cars
loaded positively on this factor; one loaded negatively. 4. It ties me with memories
These cars meet participants' basic needs for transpor- and experiences. +5 0 + 4 0 + 5
tation: 44. It makes me feel senti-
mental or nostalgic. +4 - 1 +1 +1 +3
MATERIAL POSSESSION ATTACHMENT 335

12. It reminds me of an im- capture a type of attachment that reflects disconnection


portant friend from my from a part of one's self (self-change). Ambivalence,
past. +3-3 0-1+5 evident in F3 and F5, portrays the tension between self-
2. It reminds me of who I change and stability seeking through disconnection
used to be. +1 - 2 +1 0 +4 from old aspects of me (F3) or old relationships (F5).
-25. It doesn't have anything to The combinations of affiliation, autonomy, and past,
do with past events in my present, or future temporal orientation characterize
life. -3 0-3 0 -5 these qualitatively distinct types of attachment. Also,
- 2 3 . It has nothing to do with because cars reflected various types of attachments, the
who I was in the past. —1 0 0 0—5 results illustrate that attachment is not inherent in a
-17. It reminds me of parents product category.
or other family members. +5 —1 +1 +1 —4 Additionally, our results suggest that affect is inad-
equate for describing or explaining attachment, in con-
These possessions are linked to the participants'
trast to Wallendorf and Arnould's (1988) operational-
selves: ization of attachment as approach-avoidance affect.
8. It fits my image. 0 +5 - 5 - 2 +4 Liking occurred with the me attachments (F2), but not
-29. It is strictly utilitarian. -1-2 0+5-5 with the affiliative ones (Fl). Thus, liking does not
identify attachment. Moreover, liking was associated
Factor 5 possessions make the participants feel dis- with utilitarian possessions (F4), so liked objects are
tinct, typically an autonomy-seeking task:
not necessarily attachments.
7. It makes me feel unique, Finally, having received a possession as a gift is as-
not like everyone else. +2 +2—1 —2 -\-4 sociated with attachment in two ways: (1) gift receipt
-30. It is easy for anyone to is more strongly associated with weaker not-me attach-
have; it's not very unique. 0 +3 +3 +3 - 5 ment (F3) than with stronger attachment (Fl or F2),
and (2) gift receipt may distinguish between types of
Yet F5 is associated with negative feelings and bad
strong attachment (Fl vs. F2). We found this discovery
memories: especially intriguing and conducted study 2 to further
54. It reminds me of frustra- examine how mode of possession acquisition influences
tion or stress. - 2 - 2 - 2 +2 +5 attachment type.
35. It reminds me of someone
from my past I would
rather forget. - 2 - 3 - 1 - 2 +5 STUDY 2
37. It does nothing to help me
feel loved or cared for. -5+1 0 +1 +5 Our goal in Study 2 was to begin developing a no-
-18. It makes me feel loved or mological network encompassing attachment by ex-
cared for. + 4 - 2 - 1 -2 -4 ploring how mode of gift receipt, as an antecedent, in-
fluences attachment. The literature identifies two types
Participant 27's T-shirt "reminds me of an old very of gifts based on the giver: interpersonal gifts (see, e.g..
serious girlfriend who moved away." Participant 29 has Sherry 1983) and self-gifts (Mick and DeMoss 1990).
"various feelings about my car. I really like it yet it Things we receive from others will tend to symbolize
makes me feel so frustrated at times." As we would our relationship with the giver, especially if they are
expect from this ambivalence, neither liking nor dislik- more cherished possessions. Hence, we expected that
ing reflects this factor. Factor 5 is a type of attachment interpersonal gifts ("other" gifts) that become most fa-
that reflects the negotiation of self-change and the psy- vorite possessions would be more likely to reflect affil-
chological challenges of letting go. iation themes and to be associated with self-continuity
seeking, reflected in a past to present temporal orien-
Summary and Discovery tation.
Analyzing data from the four sorting tasks together In contrast to other-gifts, self-gifts are more me. They
revealed that the largest difference in the data was strong are things we buy for ourselves as rewards for accom-
versus weak attachment. We uncovered four types of plishments, to reinforce self-images of worth and com-
attachment: strong, affiliative, past-oriented attachment petency, or for self-comfort (Mick and DeMoss 1990).
(Fl); strong, present-autonomy-oriented attachment Thus, in accordance with the me-now orientation of
(F2); weak, not-me attachment (F3); and a "mixed" self-gifts, we anticipated that a favorite self-gift is more
(stronger than I like it) attachment (F5). A fifth factor likely to reflect autonomy seeking and be more present
reflected utilitarian (nonattachment) possessions (F4). oriented. Finally, we expected that least favorite self-or
Most favorite objects captured desirable connections other-gifts would reflect an absence of affiliation or au-
(self-stability) with others (affiliation) or reflected as- tonomy themes and a desire to disconnect from the
pects of one's individuality (autonomy). Least favorites past.
336 JOURNAL OF CONSUMER RESEARCH

Methodology ignated - F l ) . Although the +F1 and - F l groups rep-


resent two ends of a common continuum, we will
Again, Q-methodology was used to elicit participants' interpret them as distinct groups to aid expositional
sentiments about most or least favorite self- and other- clarity.
gifts. A purposively selected convenience sample of 24
adults each performed two Q-sorts (for a most and a Positive Factor 1 (+FI): Keeping Memories of Oth-
least favorite possession). Two participants were ers. Seven most favorite other-gifts, plus one least fa-
dropped from the final analysis for deviating widely vorite other-gift, load positively on Fl. Each of these
from the Q-sort distribution. The remaining sample of favorite other-gifts also cross loads onto F2 (see Table
22 adult participants varied widely in age (from 21 to 2; we explain in the discussion the importance of these
71 years, X = 37 years) and life experience. It included cross-loadings). As anticipated, participants ascribed af-
10 males and 12 females. filiation-seeking motives to these mostly other-gifts:
Each of the participants was randomly assigned to
one of two versions of a questionnaire that elicited the 16. It reminds me of a person
possessions used as stimuli for the Q-sorts. Question- who is important to me
naire version A elicited first a most favorite self-gift and now. +5 -1 +3 +4
then a least favorite other-gift. Version B elicited first 22. It has to do with an im-
a most favorite other-gift and then a least favorite self- portant relationship. +5 -4 +2 -2
gift.^ Instructions to elicit most favorite self-gifts were 18. It helps me feel loved and
modeled after Mick and DeMoss (1990): "Please think cared for. +5 -1 +1 +2
of possessions you have that are: (a) things you bought 17. It reminds me of my par-
for yourself for having accomplished a personal goal, ent(s) or other family
to cheer yourself up because you were feeling down, for members. +4 0 0+2
your birthday, or when you had extra money to spend, 13. It reminds me of an im-
(b) most cherished, most favorite, (c) tangible, (d) things portant family member
you still have." We asked participants to identify two who is gone now. +4 -3 +2 +2
to five of these possessions. The participant then chose . 19. It helps me feel in touch
one of those possessions, described it in writing, and with something.
explained why he or she bought it. The other possessions +3 +1 +2 0
15. It reminds me of good
were elicited similarly, with instructions adjusted ap-
propriately for whether the desired possession was a times with family and
most or least favorite, and a self- or other-gift. friends. +3 +3 +4 0
—37. It does nothing to help me
After describing the most favorite self-gift (or other- feel loved and cared for. -4 0 0+1
gift), the participant then performed a Q-sort for that
possession following procedures used in study 1. The Positive factor 1 possessions reflect family affiliations,
study 1 Q-sample was augmented for study 2 by adding as illustrated by P19's gold Mickey Mouse watch, re-
three items—taken from or inspired by Bloch's (1983) ceived as a Father's Day gift on a family trip to Dis-
enduring involvement scale^designed to elucidate the neyland, and P30's love letters from her husband. Par-
hedonic or pleasure-seeking motive that might underlie ticipant 17's garnet ring from her mother represents a
attachments (see the Appendix). Data analysis and in- "gesture of her love." And PI 5's out-of-print book was
terpretation followed the same procedures used for specially obtained for him by his wife. Even P16's sun-
study 1. glasses (least favorites) were from his wife. Moreover,
+F1 possessions are linked with heritage and tradition
Results that the participants wish to maintain:
Principal components analysis with varimax rotation 4. It ties me with memories
uncovered four factors, or groups of possessions. Table and experiences. +3 +5 +5 +5
2 shows the factor pattern matrix, each possession, the 14. It involves my heritage or
participant's age and gender, whether the possession a tradition. -\-4 0 - 1 +5
was a self-gift or other-gift, and whether it was most" or 20. It will help me carry on a
least favorite. tradition. +J 0-5 0
-36. It has nothing to do with
Keeping (Disconnecting from) Memories of Others my heritage. -J 0 0 0
(+F1 and -Fl). Thefirstfactor reveals two groups of - 3 3 . I will probably not care
possessions: one that loads positively on the factor (des- about it in the future. - 5 +1 +1 +1
ignated +F1) and another that loads negatively (des- -40. It is unlikely that I will
keep it as a reminder of
'A most favorite was always elicited first because pretesting revealed friends or family. —5—1 0 0
respondents found the task too difficult if asked first about a least -32. I could become who I
favorite self- or other-gift. want to be without it. —5 +4 +1 +2
MATERIAL POSSESSION ATTACHMENT 337

TABLE 2
STUDY 2 VARIMAX ROTATED FACTOR PATTERN MATRIX

Person identification
Condition number Possession Affiliative memories It's me! Not me! Bad past Gender

M-O 19 Mickey Mouse watch .80 .35 01 -.20 M


M-0 30 Love letters .69 .48 .06 .03 F
M-0 01 Bookcase .66 .50 -.19 .26 M
M-0 17 Gold garnet ring .66 .54 -.07 .00 F
M-0 46 Diamond earrings .65 .35 .10 .09 F
M-0 48 Corner cupboard .63 .57 -.07 -.09 F
M-0 15 Out-of-print book .61 .39 .07 .11 M
L-0 16 Oakely sunglasses .43 .18 .34 .26 M
L-0 12 Chinese egg -.45 -.13 .22 .30 F
L-S 15 Telescope -.55 .10 .09 .16 M
L-S 46 Orange sweater -.56 .01 .16 .37 F
L-O 02 Dining table -.57 -.25 39 -.02 M
L-S 29 Pair of sculptures -.59 .04 .462 .17 M
L-S 19 Radio-controlled car -.61 -.04 .24 -.42 M
L-S 01 Sofa -.65 -.28 .22 .01 M
L-S 03 Computer software -.70 -.05 -.08 .13 M
L-S 17 Shorts -.70 -.15 .21 .09 F
M-S 04 Tom-tom thing .19 .77 -.08 .20 F
M-0 21 Watercolor painting .28 .75 -.03 .13 M
M-0 29 Stereo .16 .74 -.01 -.35 M
M-S 12 Limited edition print .19 .72 -.03 -.22 F
M-0 03 Mount Hood picture .45 .70 -.06 .09 M
L-S 21 College ring .21 .70 -.06 .26 M
M-0 11 Garnet necklace .28 .66 -.08 -.20 F
M-S 16 Car compact disc player .06 .62 -.29. -.37 M
M-S 14 Family portrait .47 .58 .07 .12 F
M-S 06 Curio cabinet .35 .57 -.09 -.19 F
M-S 02 Denim jacket -.19 .56 .04 -.20 M
M-S 52 Sapphire ring .00 .53 .08 -.18 F
M-S 50 Mahogany bureau .45 .51 -.37 .00 M
L-O 04 Compact disc -.10 .48 .23 -.03 F
M-O 08 Tire .36 .47 .03 -.36 • F

M-S 10 Antique dresser .36 .45 .01 -.05 F


L-O 50 Sweater -.17 -.50 .34 .46 M
L-0 07 Computer game .09 .04 .68 .06 M
L-0 14 Jewelry organizer -.26 -.25 .60 .22 F
L-S 48 Blonde wig -.19 .12 .57 -.43 F
L-0 52 Picture frame -.24 .25 .50 .13 F
L-S 08 Piggy bank -.38 -.08 .47 .07 F
L-O 10 Living Bible -.00 -.03 .41 -.04 F
L-S 30 Chocolate turtles -.32 -.33 -.05 .72 F
L-S 11 Black pump shoes -.24 .02 -.22 .47 F
L-O 06 Broken antique doll .16 -.10 .29 .39 F
M-S 07 Computer -.00 .03 -.27 -.60 M

NOTE. Condition assignments are designated as most (M) or ieast (L) favorite foiiowed whether the possession is a self (S) or other (O) gift. Gender is designated
as maie (M) or female (F). Italicized numbers designate the factor to which a possession was assigned.

- 4 1 . It has nothing to do with -47. I have no feelings for it. -4 - 3 +3 - 1


my future. -4 +1 -1-3 +1 -50. It is basically a functional,
utilitarian object. -4 -2 +2 +3
To illustrate, Pi's bookcase was his great-grand- —49. I would not be upset if it
mother's and "carries a lot of heritage." Participant were lost or stolen. -4-1-2 0
46 describes the diamond earrings, received from her Positive factor 1 objects recall gestures of love and
grandmother for her sixteenth birthday, as especially portray an aspect of each participant's ongoing self
significant because her grandmother received them through familial connections. In accordance with our
from her mother on her sixteenth birthday. These Q- expectation that other-gifts would tend to reflect affil-
sample statements reveal the cherished character of iation seeking and self-continuity maintenance, eight
these possessions: ofthe nine +F1 objects are most favorite other-gifts.
338 JOURNAL OF CONSUMER RESEARCH

Negative Factor 1 (-F1): Nonaffiliative, Disconnec- least favorites loading positively on F3 also commem-
tions. In contrast to the most favorite other-gifts of orate past events. For example, P21 bought his college
+F1, —Fl includes nine least favorite possessions, in- ring "to mark the completion of my Bachelor's degree."
cluding seven self-gifts. These possessions fail to serve In accordance with a me orientation, F2 is not affiliation
affiliation needs and reflect self-discontinuity. For ex- oriented:
ample, P46's orange sweater (a self-gift) reminds her of
"my awful summer job in Philly—an awful summer 42. It does nothing to help me
and my old boyfriend." Of the radio-controlled car P19 feel connected with others. +1 +5—1 - 5
bought with birthday money, he says, "I thought it -22. It has to do with an im-
would be fun, but I almost never use it." Similarly, P3's portant relationship. +5 -4 +2 - 2
computer simulation game was bought as a self-reward — 13. It reminds me of an im-
but it "became total frustration . . . the longer I portant family member
messed with it, the more I hated it." Participant 17 never who is gone now. +4 -3 +2 +2
wears the shorts she bought as "a little pick-me-up" Moreover, F2 possessions are liked and enjoyed, yet
because they "make me look fat." Participant 12's with some reservation:
Chinese egg knickknack (although an other-gift) rep-
resents her former boyfriend overlooking her tastes. As 43. I really like it. +2 +4 0 +1
expected, these least favorites evidence themes of a lack 56. It is for enjoyment. 0 + i +4 - 4
of affiliation and autonomy and portray self-disconti- -52. It makes me sad. - 1 -4 - 3 +3
nuity. Moreover, these items failed to be the escape, -53. It reminds me of some-
treat, or reward for which self-gifts are intended (Mick thing I feel bad about. 0 -3 - 5 +3
and DeMoss 1990). As gifts, these possessions missed -46. It makes me happy. - 4 -3 +2 - 1
the mark.
And F2 possessions reflect strong attachments:
Factor 2 (F2): It's Me. Sixteen possessions loaded -29. It is strictly utilitarian. - 2 -4 -2 0
positively on F2; one loaded negatively. Nine of the 16 -47. I have no feelings for it. - 4 -3 +3 - 1
are most favorite self-gifts, consistent with F2's dis-
tinctive me-now orientation: The five other-gifts grouping on F2 also reflect the
receiver's distinct identity, suggesting that gifts from
8. It fits my image. - 2 +5 - 1 - 1 others sometimes reflect me more than the giver-re-
5. It is a statement of who I ceiver relationship. Examples include P3'^ picture of
am now. - 2 +4 -5 0 Mount Hood, which shows "I love mountain climbing,"
7. It makes me feel unique, and PI l's garnet necklace, which says "I love garnets."
not just like everyone else. +1 +3—1 — 1 To summarize, nine of the 10 most favorite self-gifts
Factor 2 possessions mark past events and personal ac- load on this present, autonomy-oriented factor, strongly
complishments that continue to be me: supporting our expectations. Ourfindingsthus converge
with Mick and DeMoss's (1990) theorizing about the
4. It ties me with memories self-esteem and identity maintenance motives behind
and experiences. +3 +5 +5 +5 self-gifts. The results also demonstrate that other-gifts
1. It reminds me of past ac- can serve these same me-oriented motivations.
complishments or other
important events in my Factor 3 (F3): It's Just Not Me. Six least favorite
life. 0 +5 - 1 +4 possessions cluster on F3. Four are other-gifts and two
15. It reminds me of good are self-gifts. Factor 3's viewpoint portrays possessions
times with family or for which there was never a good person-possession fit:
friends. +3 +3 +4 0 - 3 . It has always related well
-25. It doesn't have anything to to my needs. 0 + 1 -3 -3
do with past events in my 26. It doesn't fit my image any-
life. 0-5 0-5 more. 0 0+5+2
-23. It has nothing to do with 27. I could easily be myself
who I was in the past. +1 - 5 +1 - 2 without it. - 2 +2 -\-4 +1
Participant 4's "tom-tom thing," bought on a trip to 28. It has nothing to do with
France, "symbolizes that I had actually made it to (what who I am now. +1 —2 + i —3
seemed at the time) a very far, distant place which I —5. It is a statement of who I
had only dreamed of before." Participant 52's sapphire am now. - 2 +4 - 5 0
ring celebrated "the fact that I had just obtained my —2. It reminds me of who I
first real job." A compact disc player was PI 6's "gift to used to be. - 2 +2 -3 +5
myself after getting a patent on a new product I had Factor 3 possessions are least favorites that are "just
been working on as a pat on the back" (PI6). The two not me." Participant 52's picture frame (a wedding gift)
MATERIAL POSSESSION ATTACHMENT 339

does not match her decor or fit pictures she has. Par- —55. It offers me relaxation and
ticipant lO's beliefs do not align with the translation of fun when life's pressures
the Bible she received from a friend. Participant 48's build up. +2 +1 0 -4
blonde wig "no longer looks stylish, especially on me." -56. It is for enjoyment. 0 +3 +4 -4
These possessions are also characterized by weak at- —57. It sets me free. - 1 +1 - 4 -3
tachment; participants see no future for the object, yet Two least favorite self-gifts reflect these bad feelings.
evidence the paradox that they still own it: Participant 3O's chocolate turtles were bought as part
31. I do not plan to have it in of a fund-raiser for her dance studio: "Parents were not
the future. - 3 - 1 +5 -HI cooperative . . . we had many unpaid bills and it was
41. It has nothing to do with just a mess. It wasn't a wise choice." Participant 11
my future. - 4 +1 + i +1 purchased her black pump shoes for a special occasion,
—20. It will help me carry on a but they turned out to be uncomfortable and "were a
tradition. +3 0 - 5 0 waste of money." This sense of regret is associated with
- 9 . It has to do with my goals past events or relationships:
for the future. +2 0-4-1 2. It reminds me of who I
-10. It is a statement of who I used to be. - 2 +2 -3 +5
would like to be. +1 0 - i - 2 4. It ties me with memories
Yet F3 includes neither dislike nor negative feelings: and experiences. +3 +5 +5 -1-5
14. It involves my heritage
56. It is for enjoyment. , 0+3+4 —4 or a tradition. +4 0-1-1-5
—53. It reminds me of some- 1. It reminds me of past
thing I feel bad about. 0 - 3 - 5 +3 accomplishments or
—52. It makes me sad. - 1 - 4 -3 +3 other important events
Interpersonal connections may bring participants to in my life. 0+5-1+4
keep these possessions, as reflected in the fact that four 25. It doesn't have anything
of the six possessions were other-gifts. to do with past events in
my life. 0-5 0-5
15. It reminds me of good
times with family or At least part of the reason for keeping F4 possessions
friends. -t-3 +3 +4 0 may be their utilitarian value:
16. It reminds me of a person 50.. It is basically a functional,
who is important to me utilitarian object. - 4 - 2 +2 +3
now. +5 - 1 +3 +4 48. It has no special mean-
-35. It reminds me of someone ing—it's just something I
from my past I would have. - 3 - 2 +2 +3
rather forget. +1 - 5 -4 +2
Yet, again, affiliations are evident in these least favor-
To summarize, all F3 possessions are least favorites, ites:
and most of them are other-gifts. The F3 viewpoint re-
veals possessions that are not useful for autonomy- 16. It reminds me of a person
seeking purposes, supporting our expectations about who is important to me
least favorites. However, F3 possessions do evidence a now. +5 - 1 +3 +4
slight affiliation motive, consistent with their being -42. It does nothing to help me
mostly other-gifts, which may explain why the partic- feel connected with others. +1 +3 - 1 - 5
ipants still own them. -34. It does not remind me of
anyone from my past. 0—2 0 —-^
Factor 4 (F4) Bad Feelings from the Past. Four
possessions loaded on this factor: two least favorite self- To illustrate, P6 has mixed feelings about an antique
gifts, one least favorite other-gift, and a negatively load- doll with a broken head, given to her in memory of a
ing most favorite self-gift. Factor 4 possessions are dis- deceased friend.
liked, convey bad feelings, and do not serve any escape To summarize, in F4 we observe that least favorites
or reward purposes: tend not to reflect autonomy seeking, which is consis-
tent with our expectations. The feature that distin-
51. I really do not like it any- guishes F4 from other least favorites is its especially
more. 0 - 2 - 1 +4 strong negative overtone—its sadness and regret.
53. It reminds me of some-
thing I feel bad about. 0 -3 - 5 +3
52. It makes me sad. - 1 - 4 - 3 +3
Summary and Discovery
—45. It reminds me of some- To summarize, the key findings of study 2 are (a) a
thing I am proud of. +2 +2 +2 -5 majority (7 of 12) of the most favorite other-gifts re-
340 JOURNAL OF CONSUMER RESEARCH

fleeted affiliation and self-continuity seeking (+F1), (b) We observed that certain possessions to which there
9 of 10 most favorite self-gifts reflected a present au- is attachment capture the we facet of identity; they carry
tonomy theme, and (c) least favorite gifts most often relationships from the past into the present life story.
(19 of 22) reflect neither autonomy nor self-continuity Other attachments focu5 on current identity work (me
seeking (-F1, F3, or F4), although F3 and F4 revealed now); they serve to connect the person to a significant
slight degrees of affiliation. This pattern of results sup- event, accomplishment, or other signifier of self-esteem.
ports our premise that mode of gift receipt tends to Some attachments (like favorite other-gifts) portray
influence attachment type (X^(gift type X attachment both we and me, indicating the unique contribution of
type) = 9.66, df= 4, p < .05, Cramer coefficient C the affiliation and autonomy dimensions to under-
= .47). standing attachment. Yet another type of attachment
Study 2 also produced a provocative finding that portrays "not me," apparently used for removing an
might have been anticipated from the gift-receipt lit- old, less desirable, piece of the life story.
erature (e.g., Mick and DeMoss 1990; Sherry 1983).
We observed that the meaningfulness of other-gifts is Successful versus Unsuccessful Gifts
reflected through "deep emotions" resulting from the
"conjoining of giver, receiver, and gift" (Sherry 1983). Our results empirically demonstrate certain aspects
All most favorite other-gifts loading on Fl (keeping of gift-giving success as perceived by the gift receiver.
memories of others) also load strongly on F2 (it's me), Unsuccessful (least favorite) gifts are gifts described as
identifying that these possessions reflect both the Fl not me. These included interpersonal gifts that did not
and F2 viewpoints. These are successful gifts because reflect the receiver's me or self-gifts that failed to provide
they simultaneously symbolize the relationship between the anticipated escape or reward. Additionally, we
giver and receiver (Fl) and a match between gift and found two kinds of successful (most favorite) gifts: (1)
receiver (F2). Put differently, these gifts reflect simul- a gift that represents a receiver's me (usually a self-gift)
taneously we and me. Gifts that primarily serve affili- and (2) a gift that primarily symbolizes the relationship
ation seeking can also serve autonomy purposes, but between giver and receiver and also fits with me (these
not vice versa. Additionally, this provocative discovery were always interpersonal gifts). We had expected that
illustrates the versatility of Q-methodology for exploring the favoriteness of other-gifts would be driven by affil-
phenomena such as attachment and gift-giving. iation, and that appears to be so. However, the novelty
in our findings is that affiliation is often not enough for
Study 2 also reveals that an unsuccessful other-gift, gift-giving success; the gift must also fit the receiver's
evidenced by being a least favorite, is such because the personal identity.
possession is a poor symbol of the recipient's identity
(not me) and/or of the giver-recipient's relationship (not The conventional wisdom that gift receipt is one way
we). Identity and/or relationship appropriateness ap- individuals incur attachments might suggest that gifts
pear fundamental to gift-giving success. should be overrepresented among possessions to which
Mick and DeMoss (1990) suggest that the extra there is strong attachment. Yet we found the opposite:
meaningfulness (and by inference, the success) of a self- a greater proportion of weak-attachment possessions
gift results from its being a deserved change from the were received as gifts. The affiliative connections ac-
ordinary. Our data reveal that self-gifts become least companying these gifts appears to have more to do with
favorites (possibly despite initial rapture) for several why people keep not-me possessions than why people
reasons: they fail to provide the sought-after reward incur strong attachments.
(e.g., - F l ) , they come to symbolize an unpleasant as-
pect of one's past life story (F4), or they fail to symbolize Generalizability
identity (e.g., F3). Our findings are derived from a wide variety of prod-
uct categories, which contributes to their generaliza-
DISCUSSION AND CONCLUSION bility. Nevertheless, theorizing about attachment would
be enhanced by further investigation in which subjects'
Two studies investigated participants' attachments
characteristics and/or possession categories are system-
to a wide variety of most and least favorite possession
atically varied to assess their influence on attachment
objects, including various special possessions, clothing,
outcomes. Additionally, although American culture
and cars. The results illustrate how individuals use dif-
emphasizes autonomy seeking, our participants did ex-
ferent attachment types to narrate different facets of
press affiliative attachments. Future investigators might
identity. Archetypal identity tasks of affiliation versus
explore this further and examine how attachment types
autonomy seeking, along with identity change versus
(e.g., past- vs. present-oriented autonomy) play out in
stability management, combine to form qualitatively
cultures that place less emphasis on autonomy.
distinct types of psychological significance. Addition-
ally, by showing that mode of gift receipt (self- vs. other-
gift) predicts attachment type, the findings contribute Q-Methodology
to developing a nomological network encompassing at- The two studies illustrate how Q-methodology is par-
tachment. ticularly useful for examining subjective consumption
MATERIAL POSSESSION ATTACHMENT 341

phenomena. Our findings demonstrate the methodol- infer meaning of self when they are especially useful
ogy's discovery potential and its use for investigating for telling stories ofthe self.
underdeveloped concepts (i.e., concepts having no a
priori, or empirically supported, nomological network). Characterizing Attachment and Its
Additionally, the methodology could be used to inves- Relationship to Consumption Behavior
tigate consumers' subjectivity, not only for possessions,
but also for meanings of a variety of objects, people, or We characterize attachment as a multidimensional
events. signifier of self-extension. Diflerent kinds of attach-
ments result from combinations of these dimensions:
How Is a Possession Me or Not Me? affiliation, autonomy, and past, present, or future tem-
poral orientation. The temporal orientation helps
Attachment and Self-Extension identify whether a possession reflects self-continuity or
How do people use their relationship to certain pos- self-change, and suggests where in the attachment-de-
sessions, objects, or events to infer the meaning of self? tachment cycle the person-object relation lies (Ball and
The conventional view of person-possession relations Tasaki 1992).
portrays the self as an inner core (the "true" self) sur- The future orientation of the proposed conceptual-
rounded by layers of decreasing authenticity. In this ization was not supported by either study. Apparently,
gravitational metaphor, an object's phenomenological people do not become attached routinely to possessions
distance from the core self is related inversely to the that represent emerging aspects of themselves. Attach-
degree to which it reflects one's authentic or true self. ments form with things that mark a realized identity
Self-relevant objects are those perceived as close to the development path, so it makes sense that attachments
core self. This way of thinking implies a direct link be- are usually past or present oriented. Yet we are not
tween self and object and suggests a unidimensional ready to dismiss the possibility of future-oriented at-
view of attachment, which our data show is problematic. tachments. For example, several of our protocols re-
This approach also emphasizes self as me, ignoring the flected a distinct future orientation as they represented
importance to self-conception of relations with others. an emerging relationship (e.g., engagement rings). Fu-
Moreover, this model of the self has provided limited ture studies might examine the identity-development-
opportunity to conceptualize, or empirically examine, related meaning of such attachments.
why certain possessions become strong attachments and
others do not. Antecedents ofAttachment. Our findings show that
manner of possession acquisition predicts attachment
Applying an alternative self-metaphor, we have type. Also, the findings show that attachment is not
demonstrated the utility of the life narrative to more predicted by product class, possession type, or person
usefully conceptualize, and empirically examine, the type; a specific person-object pairing is idiosyncratic to
question of me-ness. The self-narrative captures the the specific life story episode it narrates. Moreover, like
output of a person's progression through various life other emotionally complex consumption phenomena
stages. It is an integrated, sociohistorical self-construal, (Allen et al. 1992), attachment is not adequately cap-
weaving together stories that portray significant life ep- tured by unidimensional affect. Yet affect may be a pre-
isodes reflecting self-change and self-continuity. No dictor of attachment type (e.g., autonomy-oriented
single facet ofthe narrative is more authentic than an- possessions are more strongly associated with liking).
other because one's life story is identity. A possession's Other predictors of attachment type may add to this
potency for self-significance arises indirectly via its link nomological net. For example, what kinds of life epi-
to a meaningful life narrative episode. Thus, possessions sodes predict certain types of attachment? It seems that
are not literally the self, but artifacts of the self. discontinuous life changes offer especially potent op-
Self-artifacts help narrate stories ofthe self and reflect portunities to observe attachment.
self-developmental tasks similar to those underlying life
narratives. Self-identifying possessions reflect who I am Consequences of Attachment. We found intriguing
as a unique individual, and/or who I am as I am con- the implication that affiliative associations can make
nected to others. Possessions that mark who I am not, even a not-me possession more difficult to dispossess.
or who 1 was but am no longer, also signify identity. How does attachrhent type affect the length oftime a
The remainder of a person's possession portfolio in- possession is kept, how it is kept or displayed, or how
cludes things that are not self-identifying (e.g., utilitar- it is disposed of? Are affiliative attachments more tem-
ian only). porally robust or emotionally potent than autonomy-
Self-signifying possessions help mark a path, or trail, seeking ones? What emotions do certain types of pos-
as it were, along which we have traveled in arriving at sessions stimulate?
the current me, placing identity in its sociohistorical The findings about gifts and attachment demonstrate
context. Thus, we conclude that people use possessions, how attachment can inform other aspects of consumer
events, or places to infer the meaning of self when those behavior. In addition to gift giving, studies of attach-
entities are connected to a developmental episode in ment may inform such areas as disposition (e.g., why
the life story. That is, people use external objects to people keep nonfunctional items and how attachment
342 JOURNAL OF CONSUMER RESEARCH

type affects manner of disposition), socialization (how 22. It has to do with an important relationship.
people use objects to narrate developing aspects of
themselves), or involvement (whether attachment is a Weak Attachment
kind of involvement that includes we-ness along with
the me-ness of ego extension). Autonomy Seeking:
In conclusion, our characterization of attachment Past:
goes beyond the unidimensional view (e.g., Wallendorf 23. It has nothing to do with who I was in the
and Arnould 1988) to yield a rich, multidimensional past.
construct that also is more parsimonious than frame- 24. It used to be me, but not anymore.
works offered by others (e.g., Csikszentmihalyi and 25. It doesn't have anything to do with past
Rochberg-Halton 1981; Dittmar 1992). Moreover, the events in my life.
two studies demonstrate that the conceptualization is 26. It doesn't fit my image anymore.
empirically workable and useful for building theory Present:
about possession attachment. 27. I could easily be myself without it.
28. It has nothing to do with who I am now.
29. It is strictly utilitarian.
APPENDIX 30. It is easy f^or anyone to have; it's not very
Q-Sample Statements unique.
Future:
31. I do not plan to have it in the future.
Strong Attachment 32. I could become who I want to be without
Autonomy Seeking: it.
Past: 33. I probably will not care about it in the fu-
1. It reminds me of past accomplishments or ture.
other important events in my iife. Affiliation Seeking:
2. It reminds me of who I used to be. Past:
3. It has always related well to my needs. 34. It does not remind me of anyone from my
4. It ties me with memories of experiences. past.
Present: 35. It reminds me of someone from my past I
5. It is a statement of who I am now. would rather forget.
6. It makes me feel independent and self-suffi- 36. It has nothing to do my heritage.
cient, i^ Present:
7. It makes me feel unique, not just like every- 37. It does nothing to help me feel loved or cared
one else. for.
8. It fits my image. 38. It makes me feel out of style.
Future: 39. It makes me feel different from people I as-
9. It has to do with my goals for the future. pire to be like.
10. It is a statement of who I would like to be. Future:
11. It shows who I am becoming. 40. It is unlikely that I will keep it as a reminder
Affiliation Seeking: of my friends or family.
Past: 41. It has nothing to do with my future.
12. It reminds me of an important friend in my 42. It does nothing to help me feel connected
past. with others.
13. It reminds me of an important family
member who is gone now. Affect Statements
14. It involves my heritage or a tradition.
15. It reminds me of good times with family or Positive:
friends. 43. I really like it.
Present: 44. It makes me feel sentimental or nostalgic.
16. It reminds me ofa person who is important 45. It reminds me of something I am proud of.
to me now. 46. It makes me happy.
17. It reminds me of my parent(s) or other fam- Neutral:
ily members. 47. I have no feelings for it.
18. It makes me feel loved and cared for. 48. It has no special meaning to me—it's just
19. It helps me feel in touch with something. something I have.
Future: 49. I would not be upset if it were lost or stolen.
20. It will help me carry on a tradition. 50. It is basically a functional, utilitarian object.
21. It makes me similar to people I aspire to be Negative:
like. 51.1 really do not like it anymore.
MATERIAL POSSESSION ATTACHMENT 343

52. It makes me sad. Csikszentmihalyi, Mihalyi and Eugene Rochberg-Halton


53. It reminds me of something I feel bad about. (1981), The Meaning of Things. London: Cambridge
University Press.
54. It reminds me of frustration or stress.
Dittmar, Helga (1992), The Social Psychology of Material
Possessions, New York: St. Martin's.
Erikson, Erik H. (1968), Identity: Youth and Crisis, New York:
Hedonic (Added for Study 2) Norton.
55. It oflers me relaxation and fun when life's pres- Kamptner, N. Laura (1991), "Personal Possessions and Their
sures build up. Meanings: A Life-Span Perspective," Journal of Social
Behavior and Personality, 6 (6), 209-228.
56. It is for enjoyment. Kegan, Robert (1982), The Evolving Self Cambridge, MA:
57. It sets me free. Harvard University Press.
Kerlinger, Fred (1988), Foundations of Behavioral Research,
[Received April 1994. Revised June 1995. Brian New York: Holt, Rinehart & Winston.
Sternthal served as editor and John Lynch as Markus, Hazel and Elizabeth Wurf (1987), "The Dynamic
associate editor for this article.] Self-Concept: A Social Psychological Perspective," An-
nual Review of Psychology 38, 299-337.
McAdams, Dan P. (1985), Power, Intimacy, and the Life
REFERENCES Story: Personological Inquiries into Identity, Homewood,
IL: Dorsey.
Allen, Chris T., Karen A. Machleit, and Susan Schultz Kleine McKeown, Bruce and Dan Thomas (1988), Q-Methodology,
(1992), "A Comparison of Attitudes and Emotions as Newbury Park, CA: Sage.
Predictors of Behaviors at Diverse Levels of Behavioral Mick, David Glenn and Michelle DeMoss (1990), "Self-Gifts:
Experience," Journal of Consumer Research, 18 (March), Phenomenological Insights from Four Contexts," Journal
493-504. of Consumer Research, 17 (December), 322-332.
Altman, Irwin (1976), "Privacy: A Conceptual Analysis," Myers, Elizabeth (1985), "Phenomenological Analysis of the
Environment and Behavior, 8 (March), 7-29. Importance of Special Possessions: An Exploratory
Ball, A. Dwayne and Lori H. Tasaki (1992), "The Role and Study," in Advances in Consumer Research, Vol. 12, ed.
Measurement of Attachment in Consumer Behavior," Elizabeth C. Hirschman and Morris B. Holbrook, Provo,
Journal of Consumer Psychology, 1 (2), 155-172. UT: Association for Consumer Research, 560-565.
Belk, Russell W. (1987), "Identity and the Relevance of Mar- Schlinger, Mary Jane (1969). "Cues on Q-Technique," Jour-
ket, Personal, and Community Objects," in Marketing nal of Advertising Research, 9 (February), 53-60.
and Semiotics: New Directions in the Study of Signs for Schultz, Susan E., Robert E. Kleine, III, and Jerome B. Kernan
Sale, ed. Jean Umiker-Sebeok, Berlin: Mouton de Gruy- (1989), "'These Are a Few of My Favorite Things': To-
ter, 151-164. ward an Explication of Attachment as a Consumer Be-
(1988), "Possessions and The Extended Self," Journal havior Construct," in Advances in Consumer Research,
of Consumer Research, 15 (September), 139-168. Vol. 16, ed. Thomas Srull, Provo, UT: Association for
(1990), "The Role of Possessions in Constructing and Consumer Research, 359-366.
Maintaining a Sense of Past," in Advances in Consumer Sherry, John F., Jr. (1983), "Gift Giving in Anthropological
Research, Vol. 17, ed. Marvin E. Goldberg et al., Provo, Perspective," Journal of Consumer Research, 10 (Sep-
UT: Association for Consumer Research, 669-676. tember), 157-168.
-, John F. Sherry, Jr., and Melanie Wallendorf (1988), Solomon, Michael R. (1983), "The Role of Products as Social
"A Naturalistic Inquiry into Buyer and Seller Behavior Stimuli: A Symbolic Interactionism Perspective," Journal
at a Swap Meet," Journal of Consumer Research, 14 of Consumer Research, 10 (December), 319-329.
(March), 449-470. (1985), The Psychology of Fashion, Lexington, MA:
Bloch, Peter H. (1983), "An Exploration into the Scaling of Lexington.
Consumers' Involvement with a Product Class," in Ad- Stephenson, William (1953), The Study of Behavior, Chicago:
vances in Consumer Research, Vol. 8, ed. Kent Monroe, University of Chicago Press.
Ann Arbor, MI: Association for Consumer Research, Wallendorf, Melanie and Eric J. Arnould (1988), "'My Fa-
61-65. vorite Things': A Cross-Cultural Inquiry into Object At-
Brown, Stephen R. (1980), Political Subjectivity, New Haven, tachment, Possessiveness, and Social Linkage," Journal
CT: Yale University Press. of Consumer Research, 14 (March), 531-547.
Cohen, Joel B. (1989), "An Overextended Self?" Journal of Wicklund, Robert A. and Peter M. Gollwitzer (1982), Sym-
Consumer Research, 16 (June), 125-129. bolic Self-Completion, Hillsdale, NJ: Erlbaum.

You might also like