Content uploaded by Jeana H Frost
Author content
All content in this area was uploaded by Jeana H Frost
Content may be subject to copyright.
COMMENT
Does Familiarity Breed Contempt or Liking?
Comment on Reis, Maniaci, Caprariello, Eastwick, and Finkel (2011)
Michael I. Norton
Harvard Business School
Jeana H. Frost
VU University Amsterdam
Dan Ariely
Duke University
Reis, Maniaci, Caprariello, Eastwick, and Finkel (2011) conducted 2 studies that demonstrate that in
certain cases, familiarity can lead to liking—in seeming contrast to the results of our earlier article
(Norton, Frost, & Ariely, 2007). We believe that Reis et al. (a) utilized paradigms far removed from
spontaneous, everyday social interactions that were particularly likely to demonstrate a positive link
between familiarity and liking and (b) failed to include and incorporate other sources of data— both
academic and real-world—showing that familiarity breeds contempt. We call for further research
exploring when and why familiarity is likely to lead to contempt or liking, and we suggest several factors
that are likely to inform this debate.
Keywords: familiarity, person perception, impression formation, mere exposure, close relationships
We were both pleased and displeased to see the recent article by
Reis, Maniaci, Caprariello, Eastwick, and Finkel (2011) challeng-
ing the findings and account of our article (Norton, Frost, &
Ariely, 2007)—pleased because it is always a compliment to have
conducted research that irritated very smart people who we respect
a great deal enough to want to follow-up on, but displeased
because we believe that Reis et al. have overstated the generaliz-
ability of their results and underreported other streams of literature
that are in conflict with their account (reviewed below) and, even
more, that Reis et al. missed an opportunity to begin to construct
an account that would integrate their findings, our findings, and the
existing literature into a broader account of the link between
knowledge and liking.
Are the Paradigms Representative of Everyday
Social Interactions?
We absolutely agree with Reis et al. (2011) that our two very
different methods of assessing whether more information leads to
less liking—a trait-based paradigm in which we carefully control
the amount of information presented, and a naturalistic experiment
surveying online daters both before and after dates—are not fully
representative of all of the ways in which people come to know
each other. As a result, we are very much in favor of research that
explores when and how information might lead to more liking
rather than less liking—and in particular, research that explains
why information might sometimes lead to more rather than less
liking. In our mind, however, the results reported by Reis et al. fall
short on both accounts, adding data from two experiments that take
place in very specific settings with a specific participant popula-
tion— but then using the data from these experiments to make the
broad claim, as their title states, that “familiarity does indeed
promote attraction.”
Their first claim is that our two methods of assessing the link
between information and liking are not fully representative of what
usually occurs in social interaction: “natural social interaction
involves contexts and processes not present in Norton et al.’s
research” (Reis et al., 2011, p. 557). We could not agree more—
yet, we also disagree strongly with the implication of their claim
that their paradigms do in fact capture “natural social interaction.”
We would have guessed that if Reis et al. (2011) were trying to
examine natural social interactions, they would not have used a
laboratory paradigm in which undergraduates alternated answering
questions designed to promote relationship closeness (Study 1), or
a paradigm in which undergraduates do not even meet face-to-face
but chat via the Internet. Indeed, the task used in Study 1—the
Relationship Closeness Induction Task (RCIT; Sedikides, Camp-
bell, Reeder, & Elliot, 1998)—was specifically designed to make
people like each other more, building on the well-documented
finding that people tend to like others after disclosing to them (see
Collins & Miller, 1994, for a review). In a follow-up article, the
creators of this disclosure task specifically noted the difference
between their task and real-world relationships:
Michael I. Norton, Marketing Unit, Harvard Business School; Jeana H.
Frost, Department of Communication Science, VU University Amsterdam,
Amsterdam, the Netherlands; Dan Ariely, Fuqua School of Business, Duke
University.
Correspondence concerning this article should be addressed to Michael
I. Norton, Harvard Business School, Soldiers Field Road, Boston, MA
02163. E-mail: mnorton@hbs.edu
Journal of Personality and Social Psychology © 2011 American Psychological Association
2011, Vol. 101, No. 3, 571–574 0022-3514/11/$12.00 DOI: 10.1037/a0023202
571
At the same time, naturally occurring friendships differ in important
ways from induced closeness. Most notably, friendships extend across
time and setting, whereas induced closeness exists only in the short
duration of the RCIT (i.e., 9 min) and in the laboratory. (Campbell,
Sedikides, Reeder, & Elliot, 2000)
Although Reis et al. (2011) noted this limitation, they then
moved to a task in Study 2 that introduces a different but equally
problematic source of induced closeness: online chatting. A large
body of research has documented the ways in which the relative
anonymity of online communication can lead to “hyperpersonal
communication,” in which the ambiguity of cues in online com-
munication (a result of the lack of face-to-face contact) leads
receivers to overperceive similarity in their partner and then to
engage in strategic self-presentation to match that similarity, lead-
ing to relationships that become artificially close and intense in
a very short period of time (Turner, Grube, & Meyers, 2001;
Walther, 1996, 1997). Thus, the paradigms in both studies,
likely unintentionally, serve to increase perceived similarity
and liking by including elements quite different from natural
social interactions.
Even if we allow that Reis et al.’s (2011) paradigms somewhat
mirror natural interactions—which we believe is suspect—the
assumption that the two kinds of interactions utilized are the most
common or paradigmatic ways of learning about others is also
problematic. Frequently, for instance, we learn information about
others before we meet them through our social networks— either
through our friends telling us about them or increasingly via social
networking websites, such as Facebook, where we view people’s
likes, dislikes, education, employment, and so on before meeting.
In this sense, our trait paradigm—which again we agree is far
removed from everyday social interaction—is not so unlike how
people often learn about others, obtaining information from
sources other than talking directly to that person. Indeed, some
45% of employers in a survey conducted by Harris Interactive for
CareerBuilder.com reported gathering information about potential
hires from their social media profiles (Wortham, 2009).
This last point also relates to Reis et al.’s (2011) critique of our
online dating data, which they have described as “a special case
that cannot be generalized to other forms of attraction and inter-
action. Online dating emphasizes evaluation, because participants
typically choose among many alternative partners” (p. 559). We
agree that online dating is not the sole paradigmatic case of
learning about others—though again laboratory settings and online
chats are not either— but we take issue with the idea that everyday
interaction does not involve evaluation. Given limitations on the
number of social contacts humans are equipped to handle—
Dunbar’s number, usually estimated at roughly 150 (Dunbar,
1992)— humans are of necessity constantly screening people for
whether they will “make” our exclusive set of 150. This limit has
been shown in a number of contexts, with estimates ranging into
the 200s in some investigations but no higher (e.g., Bernard,
Shelley, & Killworth, 1987). Indeed, even analyses of network
data from Facebook—seemingly the way to manage limitless
connections—show that the average person has less than 200
friends and regularly communicates with just 35– 45 friends in any
2-month period (Burke, Marlow, & Lento, 2010).
More broadly, it is simply unclear to us how humans could be
built such that when we meet people, the more we learn about
them, the more we like them. Would this mean that when we meet
someone at any given party and acquire one piece of information
about this person, we become transfixed by him/her and like
him/her more and more over the evening, never talking to anyone
else? Would we expect even a randomly selected 16-year-old male
adolescent and a 75-year-old woman to like each other more and
over the course of chatting? We suspect both of them might seek
to end the interaction as soon as possible.
Do Other Sources of Evidence Show That Familiarity
Breeds Contempt or Liking?
Thus, for a number of reasons, we believe that the paradigms
utilized by Reis et al. (2011) are not ideal for providing a critical
test of whether and when familiarity breeds contempt or liking. We
admit, however, that our paradigms are not ideal either, because
they suffer from some limitations as well. As a result, we searched
for additional sources of data— both academic and real-world—
that addressed this issue. As we outline below, the sources we
found tend to support the notion that more information leads to less
liking.
Academic Research
In Reis et al.’s (2011) studies, participants are in some sense
stuck in interactions that are difficult to exit—they have to stay
until the end of the laboratory session in Study 1, and they do not
get paid as much if they do not complete their chats in Study 2.
Even if Reis et al. are claiming that their effect holds only in forced
interactions that people cannot leave—which again calls into ques-
tion how natural and spontaneous their paradigms are—there are a
number of studies that show that even for people forced to interact,
liking decreases over time. Berg (1984); Shook and Fazio (2008);
and West, Pearson, Dovidio, Shelton, and Trail (2009) all have
shown main effects such that college roommates on average like
each other less over time. Thus, using the same sample as Reis
et al.— college undergraduates— but with a longitudinal and more
externally valid paradigm, these articles demonstrate that greater
interaction with others in very naturalistic settings leads to less
liking. In fact, in Berg’s study, roommates who chose not to
continue as roommates after one semester showed decreases in
liking over time, and even those who chose to continue as room-
mates showed either no increase in liking or even a slight decrease.
Reis et al. dismissed the articles by writing that “because none of
these studies were true experiments, their interpretations are po-
tentially ambiguous” (p. 558). It may very well be true that these
articles are not true experiments—though many college room-
mates are randomly assigned, and often when they are not, it is on
the basis of trying to increase similarity (putting morning people
together and night owls together)— but how can the authors ac-
count for this wealth of academic data in opposition to their
account? In other words, if their claim is that interaction leads to
liking, what is it about the interactions between roommates in these
investigations that make them lead to less liking?
Real-World Data: Marriage
Consider an additional source of real-world data on familiarity
and liking: divorce rates. Married couples, of course, are often
572 NORTON, FROST, AND ARIELY
quite similar to each other—indeed, similarity is a key predictor of
relationship formation (Byrne, 1971)—and have virtually unlim-
ited opportunities for spontaneous, natural interactions. In addi-
tion, unlike online daters—who we agree may be a self-selected
group of people—most Americans will be married at some point in
their lives. However, as with data from roommates in college,
divorce rates suggest that familiarity often does not lead to liking:
For marriages that occurred in the 1970s, nearly half— 48%—
ended in divorce within 25 years. Although divorce rates appear to
be declining slightly, some 20% of marriages that took place in the
1990s had already dissolved within just 10 years (Stevenson &
Wolfers, 2007). Of course, these percentages reflect only the
couples who dislike each other strongly enough to actually get
divorced; there are likely other couples who like each other less
than when they got married who do not get divorced—though we
admit (and hope) that there are some who like each other more.
Real-World Data: Politics
As a final piece of data, consider liking for individuals we do not
meet face-to-face, yet about whom we acquire a great deal of
information about over time: U.S. presidents. Gallup poll data
available for all presidents from Harry Truman to George W. Bush
indicate that 10 out of the 11 left with higher disapproval ratings
than when they started—many substantially: Truman moved from
5% disapproving to 65%, George H. W. Bush moved from 10%
disapproving to 40%, and George W. Bush moved from 25%
disapproving to 60%. The trend is true even for presidents we
might consider “popular,” with Eisenhower going from 8% disap-
proving to 28%, and Reagan going from 15% to 30%. (The only
exception is Bill Clinton, who merely managed to hold constant at
30% disapproving.)
1
Finally, another recent article showed that
people who know the most about Congress like Congress the least
(Mondak, Carmines, Huckfeldt, Mitchell, & Schraufnagel, 2007).
Conclusion and Future Directions
Our goal is not to claim that any one of the sources of data
reviewed above— college roommates, marriages, politicians, on-
line daters, or laboratory trait paradigms—is the source of data to
“trust,” but rather that on the whole, these sources of data point in
the direction of more knowledge leading to less liking. Of course,
the results of Reis et al. (2011) and some of the articles they cite
point in the direction of greater familiarity. We suggest that at
minimum, the penultimate statement of Reis et al.—“in spontane-
ous, everyday social interactions among newly acquainted peers,
familiarity does indeed tend to breed liking rather than contempt”
(p. 567)—warrants qualification: Their paradigm does not capture
spontaneous (their participants are forced to interact) or everyday
(chatting about prescribed topics, or online chatting) social inter-
actions, and the data from new college roommates suggest that
interactions among newly acquainted peers do not always lead to
more liking.
More broadly, however, the claim in Norton et al. (2007) is that
familiarity leads to dislike on average, but not in every case—
indeed, even within in our data, some online daters do show very
high liking for their partner after their dates. Surely there are some
situations that increase the likelihood that knowing more about
someone does lead to more liking—and these may be the kinds of
situations in Reis et al. (2011)’s paradigms—and clearly from the
research reviewed above, there are many situations in which know-
ing more leads to less liking. Our goal in writing this comment was
to sketch more fully the various sources of data in favor of and
opposed to the notion that “familiarity breeds contempt” in an
effort to spur future research that explores the factors that influ-
ence the relationship between information and liking, and the
psychological mechanisms underlying those relationships.
Although beyond the scope of this short comment, our review of
our work, Reis et al. (2011), and the sources of data reviewed
above suggest several promising avenues.
The Medium
Even examining just the paradigms in Norton et al. (2007) and
Reis et al. (2011), the medium in which people acquire information
is vastly different, from reading trait information (somewhat akin
to reading information on social networking sites) to chatting
online to meeting in the laboratory to meeting for a lunch date. It
is very likely, given the differing results of our two articles, as well
as the large body of research on the impact of the medium on
social interaction (Walther, 1996), that how people acquire infor-
mation about others will impact the strength and nature of the link
between familiarity and liking.
Interaction Goals
When people meet others, they can have many different goals,
and these goals likely impact both the processing of information
and liking for the other person. Consider merely the difference
between two people choosing to meet for an online date (as in
Norton et al., 2007) and two randomly assigned partners disclosing
information to each other in a laboratory session (as in Reis et al.,
2011). In the real world, some of our interactions are chosen (we
can approach whoever we want in a crowded bar) and some are not
(we are stuck with our siblings). Examining how the perceiver’s
goal for the interaction (e.g., “find my life partner” vs. “just get
through the holidays with the family”) impacts the link between
familiarity and liking is also likely to shed light into the phenom-
enon.
Familiarity Versus Information
Finally, more research is needed on the underlying mechanisms
that predict when familiarity leads to liking and when it leads to
contempt. Both Norton et al. (2007) and Reis et al. (2011) cited the
seminal work of Zajonc (1968) and related follow-up work (e.g.,
Moreland & Beach, 1992); these articles—about the positive ef-
fects of feelings of familiarity— highlight a critical distinction that
is often overlooked, between the feeling of familiarity (e.g., “I feel
like I know this person”) and actual familiarity (e.g., “I know
information about this person”). Parsing the relative contributions
of these two kinds of familiarity—ideally in the same paradigm—
would likely be useful in understanding the broader relationship
between (kinds of) familiarity and liking.
1
For graphical representations of these ratings, see http://en.wikipedia
.org/wiki/United_States_presidential_approval_rating
573
COMMENT ON REIS ET AL. (2011)
References
Berg, J. H. (1984). Development of friendship between roommates. Jour-
nal of Personality and Social Psychology, 46, 346 –356. doi:10.1037/
0022-3514.46.2.346
Bernard, H. R., Shelley, G. A., & Killworth, P. (1987). How much of a
network does the GSS and RSW dredge up? Social Networks, 9, 49 – 61.
doi:10.1016/0378-8733(87)90017-7
Burke, M., Marlow, C., & Lento, T. (2010). Social network activity and
social well-being. In E. D. Mynatt, D. Schoner, G. Fitzpatrick, S. E.
Hudson, W. K. Edwards, & T. Rodden (Eds.), Proceedings of the 28th
International Conference on Human Factors in Computing Systems (pp.
1909 –1912).
Byrne, D. (1971). The attraction paradigm. New York, NY: Academic
Press.
Campbell, W. K., Sedikides, C., Reeder, G. D., & Elliot, A. J. (2000).
Among friends? An examination of friendship and the self-serving bias.
British Journal of Social Psychology, 39, 229 –239. doi:10.1348/
014466600164444
Collins, N. L., & Miller, L. C. (1994). Self-disclosure and liking: A
meta-analytic review. Psychological Bulletin, 116, 457– 475. doi:
10.1037/0033-2909.116.3.457
Dunbar, R. I. M. (1992). Neocortex size as a constraint on group size in
primates. Journal of Human Evolution, 22, 469 – 493. doi:10.1016/0047-
2484(92)90081-J
Mondak, J. J., Carmines, E. G., Huckfeldt, R., Mitchell, D., & Schraufna-
gel, S. (2007). Does familiarity breed contempt? The impact of infor-
mation on mass attitudes toward Congress. American Journal of Polit-
ical Science, 51, 34 – 48. doi:10.1111/j.1540-5907.2007.00235.x
Moreland, R. L., & Beach, S. R. (1992). Exposure effects in the classroom:
The development of affinity among students. Journal of Experimental
Social Psychology, 28, 255–276. doi:10.1016/0022-1031(92)90055-O
Norton, M. I., Frost, J. H., & Ariely, D. (2007). Less is more: The lure of
ambiguity, or why familiarity breeds contempt. Journal of Personality
and Social Psychology, 92, 97–105. doi:10.1037/0022-3514.92.1.97
Reis, H. T., Maniaci, M. R., Caprariello, P. A., Eastwick, P. W., & Finkel,
E. J. (2011). Familiarity does indeed promote attraction in live interac-
tion. Journal of Personality and Social Psychology, 101, 557–570.
Sedikides, C., Campbell, W. K., Reeder, G. D., & Elliot, A. J. (1998). The
self-serving bias in relational context. Journal of Personality and Social
Psychology, 74, 378 –386. doi:10.1037/0022-3514.74.2.378
Shook, N. J., & Fazio, R. H. (2008). Interracial roommate relationships: An
experimental field test of the contact hypothesis. Psychological Science,
19, 717–723. doi:10.1111/j.1467-9280.2008.02147.x
Stevenson, B., & Wolfers, J. (2007). Marriage and divorce: Changes and
their driving forces. Journal of Economic Perspectives, 21, 27–52.
doi:10.1257/jep.21.2.27
Turner, J. W., Grube, J. A., & Meyers, J. (2001). Developing an optimal
match within online communities: An exploration of CMC support
communities and traditional support. Journal of Communication, 51,
231–251. doi:10.1111/j.1460-2466.2001.tb02879.x
Walther, J. B. (1996). Computer-mediated communication: Impersonal,
interpersonal, and hyperpersonal interaction. Communication Research,
23, 3– 43. doi:10.1177/009365096023001001
Walther, J. B. (1997). Group and interpersonal effects in international
computer-mediated collaboration. Human Communication Research, 23,
342–369. doi:10.1111/j.1468-2958.1997.tb00400.x
West, T. V., Pearson, A. R., Dovidio, J. F., Shelton, J. N., & Trail, T.
(2009). Superordinate identity and intergroup roommate friendship de-
velopment. Journal of Experimental Social Psychology, 45, 1266 –1272.
doi:10.1016/j.jesp.2009.08.002
Wortham, J. (2009, August 20). More employers use social networks to
check out applicants. The New York Times. Retrieved from http://
bits.blogs.nytimes.com/2009/08/20/more-employers-use-social-
networks-to-check-out-applicants/
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of
Personality and Social Psychology, 9, 1–27. doi:10.1037/h0025848
Received January 31, 2011
Revision received January 31, 2011
Accepted February 7, 2011 䡲
574 NORTON, FROST, AND ARIELY
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Personality and Social Psychology
This content is subject to copyright. Terms and conditions apply.
