Sociality,spirituality, and meaning making

Sociality, Spirituality, and Meaning Making: Chicago Health, Aging,and Social Relations Study

John T. Cacioppo, Louise C. Hawkley, Edith M. Rickett, and Christopher M. MasiUniversity of Chicago

Scientific theories in the natural sciences posit invisible forces operating with measur-able effects on physical bodies, but the scientific study of invisible forces acting onhuman bodies has made limited progress. The topics of sociality, spirituality, andmeaning making are cases in point. The authors discuss some of the possible reasonsfor this as well as contemporary developments in the social sciences and neurosciencesthat may make such study possible and productive.

In approximately 600 BCE, the Greek philos-opher Heraclitus referred to the mind as anoverwhelming space whose boundaries couldnever be fully comprehended. For thenext 2,300 years, little changed in this regard.Indeed, before the enlightenment of the 18thcentury, scholars generally believed thatthought was instantaneous and that action wasgoverned by an indivisible mind separate fromthe body. As a result of the belief that the mindwas infinitely fast and essentially unanalyzable,there was no point in trying to understand itusing scientific means. The human spirit—en-compassing the qualities of kindness, mercy,empathy, trust, compassion, justice, love,friendship, devotion, and hope—was champi-oned in art, literature, and religion but was atbest ignored by the scientific community.

The past three centuries have been a period ofunparalleled advance in science. Scientific the-ories of magnetism, gravity, and dark matterhave emerged to posit invisible forces operatingwith measurable effects on physical bodies.During this same period, the scientific study ofinvisible forces acting on human bodies has

made limited progress. Research on the dimen-sions of sociality, spirituality, and meaningmaking for instance, has been blunted by biasesagainst what were regarded as soft or religioustopics, misguided by metaphors such as thehuman brain as a solitary computer, and over-looked in a funding climate that demands timeand attention be given to societal, psychologi-cal, and physiological deficits rather than capac-ities. Although the history and causes of suchbiases are complex, little is achieved to mitigatethese biases when the constructs of sociality,spirituality, and meaning making are definedimprecisely, their predicted effects are not fal-sifiable or are difficult to replicate, or the un-derlying mechanisms for such effects are notdelineated within a framework that is recog-nized by the scientific community.

The dawn of the 21st century may herald aparadigm shift in constructs deemed amenableto scientific inquiry. With new developmentsand instruments in genetics, neuroscience, brainscience, and behavioral science, long-standingscientific biases are being challenged by rigor-ous quantitative analysis. Fueling these chal-lenges is the recognition that the traditionalfocus by founding and federal funding agencieson maladies and disease misses the mark onsome of the most complex yet pressing andenduring questions of humankind.

The guiding metaphors are also undergoingdramatic transformations. Now that the humangenome has been sequenced and has been foundto involve fewer genes than anticipated, it hasbecome apparent that humans are not inextrica-bly determined by their genotype irrespective oftheir social environment. In addition, the ex-

John T. Cacioppo and Louise C. Hawkley, Department ofPsychology and Center for Cognitive and Social Neuro-science, University of Chicago; Edith M. Rickett, Depart-ment of Psychology, University of Chicago; Christopher M.Masi, Department of Medicine and Center for Cognitive andSocial Sciences, University of Chicago.

This research was supported by National Institute onAging Grant PO1 AG18911.

Correspondence concerning this article should be ad-dressed to John T. Cacioppo, Department of Psychology,University of Chicago, 5848 South University Avenue, Chi-cago, IL 60637. E-mail: [email protected]

Review of General Psychology Copyright 2005 by the Educational Publishing Foundation2005, Vol. 9, No. 2, 143–155 1089-2680/05/$12.00 DOI: 10.1037/1089-2680.9.2.143

143

tended period of utter dependency of the humaninfant has led to the recognition that humangenetic transmission is based not on an individ-ual’s selfish ability to reproduce but on thesuccess of offspring to reproduce (Dawkins,1976). Accordingly, we have a better apprecia-tion of how humans have evolved to be aninherently social, meaning-making species.

Moreover, the notion of the solitary com-puter—the dominant metaphor for the humanmind for more than a quarter century—sud-denly seems dated. Computers today are mas-sively interconnected devices with capacitiesthat extend far beyond the resident hardwareand software of a solitary computer. How ironicthat although the telereceptors of the humanbrain have provided wireless broadband inter-connectivity to humans for millennia, it took theadvent of technological innovations for it tobecome patently obvious that the isolated com-puter is a poor metaphor for the human mind.

The qualities of kindness, mercy, empathy,trust, compassion, justice, love, friendship, de-votion, and hope—qualities attributed in art,literature, and religion to the human spirit—didnot comport well with the dominant metaphorsof the mind in the latter part of the 20th century.Today, these same qualities are simple to incor-porate into a metaphor of the mind as a mas-sively networked portable computer, given therecognition that human genetic transmission re-lies in part on cooperation and nurturance.

Even useful metaphors have limited utility,however, if the constructs under study arepoorly defined. In our study of sociality, spiri-tuality, and meaning making, we define social-ity as the need for the company of others, mea-surable as the tendency to form social connec-tions, to react to perceived social isolation, andto exhibit the long-term effects of social con-nectedness on health, successful aging, andwell-being.

By spirituality we mean a theoretical con-struct that represents what is common in humanqualities such as kindness, mercy, virtue, empa-thy, trust, compassion, justice, love, friendship,devotion, and hope. Accordingly, sociality mayunderlie spirituality in at least two senses. Aswe explain subsequently, we posit that the qual-ities ascribed in art, literature, and religion tothe human spirit evolved because humans arefundamentally a social species. Human virtuesare social virtues, potentially lending spiritual-ity and sociality a common substrate. In addi-

tion, sociality may serve as a model for spiritu-ality, as when people form a personal relation-ship with a deity. As in the case of sociality, thenature and existence of the deity with whom aperson forms a relationship is less importantthan the person’s mental representations of thedeity and relationship to the deity. The measur-able effects of spirituality also mirror those forsociality, including the tendency to form a men-tal representation that relates the self to a deity;to exhibit the long-term effects of such a rela-tionship on health, effective aging, and well-being; and to react to perceived separation fromor loss of worth in the eyes of the deity withbehavioral, physiological, and emotional re-sponses. Spirituality and sociality overlap inmeasurable ways such as religious affiliation,church attendance, and church participation; ac-cordingly, these variables are also investigated.

Central to sociality and spirituality is mean-ing making, defined as the construction of anaccount or recital of an event or a series ofevents, either true or fictitious, that serves toorganize or structure life. Our definition ofmeaning making is compatible with current the-ory in perception and cognition, in which evenvisual percepts are generated according to awholly empirical strategy that signifies to theindividual the empirical significance of thestimulus rather than its properties as such(Purves, Lotto, & Nundy, 2002).

We begin with a brief evolutionary argumentfor the importance of sociality in the survival ofthe human species. We then describe a neuro-scientific approach to the study of molar con-structs such as sociality, spirituality, and mean-ing making, and we review recent findings fromour program of research on social isolation andloneliness.

Sociality, Spirituality, and MeaningMaking as Components of Human Nature

The genetic constitution of Homo sapiensderives not simply from an individual’s repro-ductive success but more critically from thesuccess of one’s children to reproduce. Thehuman infant is born to an extended period ofcomplete dependency. If infants do not elicitnurturance and protection from caregivers, or ifcaregivers are not motivated to provide suchcare over an extended period of time, the infantsperish along with the genetic legacy of the par-ents. Hunter/gatherers who, in times of danger

144 CACIOPPO, HAWKLEY, RICKETT, AND MASI

or famine, chose not to return to share their foodwith mother and child may have survived tohunt another day, but the genetic constitutionthat enabled them to feel so little humanity alsomade it less likely their genes were propagated.In contrast, those who yearned to return despitepersonal hardship, and individuals who pro-tected and nurtured those close to them, weremore likely to have offspring who survived topropagate.

Even as adults, humans are not particularlystealthy, strong, or fast relative to other species.It is their collective action—their ability tothink, communicate, and work together—thatmakes Homo sapiens such a formidable species.Because human collective action provides anevolutionary advantage over other species andbecause genetic transmission is based not onone’s ability to reproduce but on the success ofone’s children to reproduce, Homo sapiens arethought to have evolved to be an inherentlysocial, meaning-making species with qualitiesascribed in art, literature, and religion to a hu-man spirit. In short, we posit that tens of thou-sands of years of evolution have deeply plantedsociality, spirituality, and meaning making inour genome and in our societies. Accordingly,humans are posited to have evolved a brain andbiology whose functioning benefits from theformation and maintenance of social bonds, hu-man/social virtues, and organizational life nar-ratives. The deprivation of any of these ingre-dients—such as ruptures of social connected-ness that result from relocation distant fromfriends and family—produces feelings of isola-tion and dysphoria, physiological alterations,and a motivation to reinstate connections. Giventhe evolutionary origins of these effects, thereshould also be heritable individual differencesin the extent to which social disconnectednessproduces such effects.

The extant evidence clearly shows that hu-mans quickly learn to attend to faces, perceivecommunicative displays, comprehend social hi-erarchies, and form causal attributions (see Ca-cioppo & Berntson, 2004). People develop atheory of mind by which the traits, intentions,and emotions of others are inferred; they com-municate with others and they sometimes hideor miscommunicate their own mental contentsfrom others; they form relationships, unions,and alliances; and they search for meaning inevents and patterns. Meaning making and soci-ality are such fundamental components of hu-

man nature that people perceive these charac-teristics in the movements of simple inanimateobjects. Heider and Simmel (1944), for in-stance, produced a short film of the movementof a small triangle, a small circle, and a largetriangle around and into a large rectangle. Theanimated film consisted only of these geometricshapes, yet everyone who viewed the film“saw” a social drama complete with intentions,plans, and emotions. Only SM, an individualwithout functioning amygdala—an almond-shaped pair of nuclei deep in the medial tem-poral lobes of the brain—failed to perceivethese movements as occurring within a socialarena (Adolphs, 1999).

Given our evolutionary history, we wouldcontend that people’s conceptual representationof sociality has a specifiable, generalizable, andpossibly universal structure (Hawkley, Browne,& Cacioppo, 2004). In the first study underlyingthis claim, 2,531 undergraduates completed ascale consisting of 20 questions that differenti-ated people who perceived themselves to besocially isolated from those who perceivedthemselves to be socially integrated (Hawkleyet al., 2004). An exploratory factor analysis onhalf of the sample revealed three correlateddimensions generalizable across gender, and aconfirmatory factor analysis on the other half ofthe sample corroborated this finding: The col-lege students’ beliefs, feelings, and ideas abouttheir social connectedness were structured asthree separable but related (oblique) dimen-sions: isolation, relational connectedness, andcollective connectedness.

In a population-based follow-up study, wetested a very different sample of individuals, astatistically representative sample of 230 menand women (one third African American, onethird Hispanic, one third Caucasian) from CookCounty, Illinois, born between 1935 and 1952(M age � 57.5 years; Hawkley et al., 2004).Despite the fact that this sample was an olderurban sample of varying ethnic backgrounds,socioeconomic status (SES), and occupations,the confirmatory factor analysis supported thesame three-factor structure found for the collegestudents. Their feelings of social disconnected-ness consisted of three related factors: feelingsof isolation/intimate connectedness, feelings ofrelational connectedness, and feelings of collec-tive connectedness.

Importantly, we also found statistically sepa-rable predictors of each of these facets of soci-

145SPECIAL ISSUE: SOCIALITY, SPIRITUALITY, AND MEANING MAKING

ality. In our urban sample of older adults, wefound marital status to predict feelings of isola-tion, contact with friends and family to predictfeelings of relational connectedness, and mem-bership in voluntary groups to predict collectiveconnectedness (Hawkley et al., 2004). Tests ofthe universality of this structure (isolation, re-lational connections, collective connections) areneeded, but it is worth noting that psychologicaltheories of the self, which traditionally havefocused on a person’s sense of unique identitydifferentiated from others, now distinguishamong the personal self (individual level ofanalysis), relational self (interpersonal level ofanalysis), and collective self (group level ofanalysis; Brewer & Gardner, 1996).

Individuals who score low on personal, rela-tional, and collective dimensions of socialityalso perceive themselves to be social isolatesand report intense feelings of loneliness anddysphoria (Cacioppo, Ernst, et al., 2000; Ca-cioppo et al., in press; Russell, Peplau, & Cut-rona, 1980). Individual differences in the netcontent of this structure are about 50% heritableand 50% environmental (Boomsma, Willemsen,Hawkley, & Cacioppo, 2004; S. McGuire &Clifford, 2000). In an early study of the originsof these feelings, S. McGuire and Clifford(2000) examined the heritability of loneliness inchildren. In their first study, 69 biologicallyrelated sibling pairs and 64 unrelated pairs inadoptive families in the Colorado AdoptionProject completed an 8-item loneliness scalewhen they were 9, 10, 11, and 12 years of age.In a second study, 22 monozygotic twins, 40dizygotic twins, and 80 full siblings 8–14 yearsof age completed a 16-item scale to assess lone-liness in relation to their schoolmates. Resultsrevealed significant genetic (h2 values of 55%and 48%, respectively, in Studies 1 and 2) andunshared environmental contributions to indi-vidual differences in loneliness.

In the S. McGuire and Clifford (2000) stud-ies, the sample sizes were relatively small, andbecause the studies involved adopted children,the representativeness of the sample for a pop-ulation estimate is uncertain. Moreover, herita-bility estimates of complex traits such as lone-liness may change across the life span as thefrequency, duration, and range of exposure toenvironmental influences accrue. We thereforeextended this work using data from the Nether-lands Twin Register Study (Boomsma, Ca-cioppo et al., 2004). Data on loneliness

from 7,665 young adult and adult Dutch twins(average age: 24 years) were analyzed withgenetic structural equation models. The esti-mate of genetic contributions to variation inloneliness in adults was 47%, with the remain-ing variance explained by unique environmentalfactors. Thus, the heritability estimates in adultswere similar to those found previously in chil-dren, and no evidence was found for sex or agedifferences in genetic architecture or configuraleffects of the genes. In an ongoing effort tomore specifically identify the genetic locus,Boomsma, Willemsen, Dolan, Hawkley, andCacioppo (in press) conducted a complete ge-nome scan and found evidence for two quanti-tative trait loci, suggesting that at least two setsof genes with additive effects are involved.

The striking development of the human cere-bral cortex, especially the frontal and temporalregions, is believed to be largely responsible forevolutionary advances in social and cognitivecapacities. The cerebral cortex is a mantle ofbetween 2.6 and 16 billion neurons, with eachneuron receiving 10,000 to 100,000 connectionsfrom other neurons (e.g., Pakkenberg, 1966).The human frontal (front part of the brain) andtemporal (right and left sides) lobes constitute32% and 23% of the cerebral cortex, respec-tively, arguably rendering the sensorimotor cor-tices that dominate most mammalian brains tominority status in the human brain. The expan-sion of the frontal regions in the human brain islargely responsible for the human capacity forreasoning, planning, and performing mentalsimulations, and an intact frontal region contrib-utes to the human ability to reason, remember,and work together, thereby contributing to theevolutionary success of humans (Bar-On,Tranel, Denburg, & Bechara, 2003; Krin-gelbach & Rolls, 2004). The temporal and pa-rietal regions, in turn, play essential roles insocial perception, social reasoning, and commu-nication (cf. Adolphs, 1999, 2001; Berntson,Boysen, & Cacioppo, 1993; Saxe & Kanwisher,2003).

Despite these evolutionary advances, humancognition and meaning making are often biasedand counterfactual. The sensory load from thephysical environment is minor in comparisonwith the quantity and complexity of the infor-mation that comes from other individuals,groups, alliances, and cultures, including thepotential for benevolence or treachery posed byeach. Human cognition is not an objective in-


formation process but instead is rife with theoperation of self-interest, self-enhancement,and self-protective processes. Because humansencounter more information than can possiblybe processed, they tend to economize onthought when forming beliefs that are not per-sonally relevant (Petty & Cacioppo, 1986) andtend to search for and attend to evidence thatconfirms what they already believe to be true(Snyder & Swann, 1978).

Information processing is also biased in waysthat protect the self from symbolic as well asactual threats and that promote reproductivesuccess (e.g., Jones & Berglas, 1978). In fact,people are not particularly good at knowing thecauses of their feelings or behavior (Nisbett &Wilson, 1977). People overestimate theirstrengths and underestimate their faults (M.Ross & Sicoly, 1979). They overestimate theimportance of their input, the pervasiveness oftheir beliefs, and the likelihood that a desiredevent will occur (W. J. McGuire, 1981), allwhile underestimating the contributions of oth-ers and the likelihood that risks in the worldapply to them (L. Ross, Greene, & House,1977). Events that unfold unexpectedly are notreasoned as much as they are rationalized(Aronson, 1968), and the act of remembering isfar more of a biased reconstruction than anaccurate record of actual events (McDonald &Hirt, 1997; Roediger, Buckner, & McDermott,1999). Many of these biases in social cognitionare spontaneous, do not require cognitive effort,and represent normative processing.

A crucial consequence of the nuances of thebiased fashion in which humans make meaningis that humans have much more influence in thecreation of their lives and social relationshipsthan most realize (Downey, Freitas, Michaelis,& Khouri, 1998; Murray, Holmes, & Griffin,1996). If an individual is led to believe a newacquaintance is fun and nice, for instance, theindividual behaves in a fashion that draws outpleasant and enjoyable behaviors from the per-son. If an individual is led to think a child isintelligent, the individual does and says thingsthat make a smarter child than would result ifthe individual was led to believe the child wasof average intelligence. Another instance of bi-ased meaning making is the tendency of peopleto self-handicap when they think they will fail atan important task. By subtly producing insur-mountable obstacles to success, they can at-

tribute their subsequent failure to these obsta-cles rather than to themselves.

Importantly in the present context, bias insocial cognition is also at work when peoplewho feel socially connected construe the worldas presenting challenges to be overcome withthe aid of others and react to interpersonal con-flicts in peaceful and constructive rather thanoffensive and aggressive ways, thereby produc-ing an environment that others want to inhabit(Cacioppo & Hawkley, in press). In contrast,lonely individuals are more likely to construetheir world (including the behavior of others) aspotentially threatening or punitive. Conse-quently, lonely individuals are more likely to besocially anxious, hold more negative expecta-tions for their treatment by others, and adopt aprevention focus rather than a promotion focusin their social interactions. These individualsare also more likely to appraise stressors asthreats rather than challenges and to cope withstressors in a passive, isolative fashion ratherthan an active fashion that includes activelyseeking the help and support of others. To-gether, these differences in social cognition pre-dictably result in an increased likelihood oflonely individuals acting in self-protective and,paradoxically, self-defeating ways (Cacioppo &Hawkley, in press). In each instance, the indi-viduals may be oblivious to the fact that the wayin which they perceived and thought about theirsocial world contributed to their social realities.

We noted at the outset that biases against thescientific study of constructs such as sociality,spirituality, and meaning making are attribut-able to imprecise definitions, nonfalsifiable ornonreplicable effects, a focus on associationsrather than underlying mechanisms, and the ab-sence of a broad scientific framework withinwhich to study abstract constructs of this sort.We next address the latter issue. Specifically,we argue that a neuroscience perspective mayprovide a framework that allows theoreticalconstructs and associations to be rigorously de-fined, tested, and parsed so as to investigatetheir underlying mechanisms.

Social Neuroscience Perspective

During the latter half of the 20th century, thenature of the human mind was plumbed throughclever experimental designs that used measuresof verbal reports, judgments, and reaction time.These methodologies proved limited, however.


Social cognition and interactions range fromaffect laden to habitual, and nuances derivingfrom these features may prove difficult to cap-ture fully using subjective measures and re-sponse latencies to semantic (e.g., lexical deci-sion) tasks alone (LeDoux, 2000; Zajonc,1980). A new approach, termed social neuro-science, first introduced just over a decade ago(Cacioppo & Berntson, 1992), represents a newdevelopment in the study of the human mind,including how sociality, spirituality, and themeaning of life might modulate brain andbiology.

Human nature is recognized as being com-plex. To simplify the study of human nature,neuroscientists in the past century tended toignore or hold constant social influences,whereas cognitive and social scientists tendedto ignore the biological constraints on andmechanisms through which cognition, affect,and conation are expressed. In the neuro-sciences, the architects of development and be-havior were conceived as anatomical structuresand genetic strings sculpted by the forces ofevolution operating over millennia, the builderswere cast as encapsulated within living cells farfrom the reach of social influences, and thebrain was treated as an analytical information-processing machine. The additional informationthat might be attributable to the social worldwas thought to be best considered last, if theneed arose. Social factors, the reasoning oftenwent, had minimal implications for basic devel-opment, structure, or processes of the brain ormind, in which case the consideration of socialfactors is entirely irrelevant. And even if rele-vant, consideration of social factors may renderthe study of the human mind and behavior toocomplicated to sustain scientific progress.

The attitude toward the neurosciences amongcognitive and social scientists throughout mostof the 20th century was no less antagonistic(Berntson & Cacioppo, 2000; Cacioppo, 2002).World wars, the Great Depression, and civilinjustices made it amply clear that social andcultural forces were too important to address toawait the full explication of cellular and molec-ular mechanisms. Biological constraints, mech-anisms, and insights tended to be ignored, oftenunder the misguided auspice of protecting thebehavioral sciences from the onslaught of re-ductionism. The specialized knowledge andfundamental research that were required to cul-tivate descriptive taxonomies, theoretical for-

mulations, and methodologies—coupled withan early emphasis on isolated scientific work—all but ensured that social and biological per-spectives would evolve insulated from develop-ments in the other.

For decades, the central precept of molecularbiology was that all of the information we needto construct a mammalian body, whether man ormouse, is contained in the approximately hun-dred thousand genes of mammalian DNA andthat a set of master genes activates the DNAnecessary to produce the appropriate proteinsfor development and behavior (Crick, 1970;Panksepp, 1998). In this scheme, DNA encodesthe sequence of amino acids in proteins andpeptides using the sequence of nucleotides inthe gene as a template. The DNA sequentialcode is transferred to messenger RNA (mRNA)via transcription, a process involving enzymes(RNA polymerases), followed by translationfrom mRNA to polypeptide chains (proteinpieces) and proteins. The process of DNA toRNA transcription has been assumed to be re-stricted to the confines of living cells outside theinfluence of personal and social ties.

As neuroscientific approaches have been ap-plied to more complex questions, however,these presumably basic principles have begun tobe questioned. Recent research suggests thateven DNA to RNA transcription can be subjectto modulation by the social environment (Ca-cioppo, Berntson, Sheridan, & McClintock,2000). In an illustrative study, we investigatedthe DNA to RNA transcription for a growthhormone that occurs within a type of immunecell called a lymphocyte. The production of thisgrowth hormone is of interest because it isthought to be involved in the effectiveness oflymphocytes to combat pathogens. We recentlyfound that caregivers of spouses with Alzhei-mer’s disease (AD) had markedly suppressedlymphocyte growth hormone mRNA levels rel-ative to age- and gender-matched controls (Wuet al., 1999). It is reasonable to assume that thespouses of AD patients were essentially ran-domly assigned to caregiver or noncaregiverroles by their spouse’s unexpected developmentof AD, in which case these results indicate thatsocial roles can modulate DNA to RNA tran-scription processes.

The behavior of strains of mice with specificgenes inactivated (i.e., knockout mice) has beenknown to depend on genetic background,whereas the effects of the social context have


been thought to be unimportant. Contrary to thislatter belief, however, Crabbe, Wahlsten, andDudek (1999) found that the specific behavioraleffects associated with a given knockout couldvary dramatically across experimenters, testingenvironments, and laboratories. The implicationof these and related studies is that aspects ofgenetic expression, which were thought to beencapsulated within each living cell far from thereach of personal ties or social influences, are infact subject to modulation by the socialenvironment.

In the twilight of the 20th century, neurosci-entists and cognitive scientists began to collab-orate more systematically, united by the com-mon view that information processing couldbest be understood by appeal to the brain as wellas its emergent manifestation in mind and by thegoal of understanding how the mind works.These collaborations have helped unravel puz-zles of the mind including aspects of perception,imagery, attention, and memory. Many aspectsof human nature require a more comprehensiveapproach, however. These include aspects of thehuman spirit such as sociality, altruism, affilia-tion, attachment, kin recognition, social identi-fication, communication, cooperation, com-merce, empathy, morality, contagion, love, nur-turance, kindness, mercy, compassion, justice,friendship, and hope. All can be conceptualizedas invisible forces emanating from the operationof the human brain within a social arena withmeasurable effects not only on subjective well-being but on brain, biology, and health.

Brain imaging studies clearly point to theattention given to social stimuli. If an individualviews a novel picture of an evocative image(e.g., snowcapped mountains) or an equallypleasant, arousing, and statistically infrequentimage of a person (e.g., a smiling baby), thesocial stimulus elicits a larger amplitude latepositive electrocortical (event-related brain) po-tential that peaks about 550 ms after stimulusonset (Ito & Cacioppo, 2000). In addition, thedifferential event-related response emergeseven when the task is simply to classify whetherthe picture is pleasant or unpleasant. That is, thebrain is spontaneously extracting informationabout the presence of conspecifics even whenthe explicit task has nothing to do with thisdistinction (Ito & Cacioppo, 2000). Functionalmagnetic resonance imaging studies have simi-larly suggested that social perception and socialreasoning are robust elicitors of localized re-

gions of brain activation and reflect distinctaspects of social and emotional processing (e.g.,see Blakemore, Winston, & Frith, 2004; Ca-cioppo & Berntson, 2004).

The field of social neuroscience, of course,stretches far beyond social cognition, humanstudies, or brain imaging methodologies (seeCacioppo & Berntson, in press-b; Cacioppo,Berntson, et al., 2002). As the 21st centurydawns, there is a recognition that much of thegroundwork for multidisciplinary scientific col-laborations has been laid by the giants of thepreceding three centuries. Neuroscientists, cog-nitive scientists, and social scientists are mov-ing beyond simplifying assumptions and areplacing less emphasis on the arbitrary divisionbetween the social and the biological sciences towork collaboratively toward developing morecomprehensive theories of mind, brain, biology,and behavior. Through the efforts of such indi-viduals, the broad multidisciplinary perspectiveof social neuroscience is gaining momentum,making more feasible the rigorous scientificstudy of questions such as how sociality, spiri-tuality, and the meaning of life operate.Throughout this article, we have described suchexamples from our own research, ranging frombehavioral genetics to brain imaging. In theremainder of the article, we survey results fromour multidisciplinary research to examine themechanisms underlying the association betweensocial connectedness and longevity andwell-being.

Chicago Health, Aging, and SocialRelations Study

In 2002, we began a population-based longi-tudinal study of 230 English-speaking Blacks/African Americans, non-Black Hispanics, andnon-Hispanic Caucasians between the agesof 50 and 67 years from Cook County, Illinois,to investigate the social, behavioral, cognitive,emotional, brain, autonomic, neuroendocrine,cellular, and molecular transduction pathwaysby which the social world affects well-beingand health (e.g., Cacioppo & Berntson, inpress-a, in press-b; Cacioppo & Hawkley, 2003;Cacioppo, Hawkley, & Berntson, 2003). TheChicago Health, Aging, and Social RelationsStudy (CHASRS) was built on our previousresearch on thousands of young adults (Ca-cioppo, Ernst, et al., 2000; Cacioppo, Hawkley,Berntson, et al., 2002; Cacioppo, Hawkley,


Crawford, et al., 2002; Hawkley et al., 2004)and older adults (Cacioppo et al., 1998; Uchino,Kiecolt-Glaser, & Cacioppo, 1994) and onmeta-analyses of the literature (e.g., Uchino,Cacioppo, & Kiecolt-Glaser, 1996).

Among our findings from this earlier workwas that socially connected individuals aremore likely to meet everyday stressors by activecoping and recruiting others’ help. Individualswho feel socially isolated are more likely toconstrue their world (including the behavior ofothers) as threatening or punitive. They aremore likely to appraise stressors as threatsrather than challenges and to cope with stressorsin a passive fashion by isolating themselvesfrom others and withdrawing from the problemsituation. Together, these differences in mean-ing making result in an increased likelihood ofsocially connected individuals acting in a hu-mane, selfless fashion, reinforcing their connec-tions with others and enhancing their self-es-teem, and of lonely individuals acting in self-protective and, paradoxically, socially acidicand self-defeating ways (Cacioppo, Ernst, et al.,2000; Cacioppo & Hawkley, in press; Hawkley,Burleson, Berntson, & Cacioppo, 2003).

Why might lonely individuals be more emo-tionally withdrawn in social settings? Personal-ity differences such as shyness, sociability, neg-ativity, and fear of negative evaluation mayprovide a partial explanation. In addition, be-cause social settings can be overwhelming, so-cial effectiveness depends on an individual’sability to exert voluntary control over his or herattentional focus. Moreover, regulating one’sattention can help one garner social approval.Do lonely individuals differ in their ability tovoluntarily control their attentional focus? As ameans of exploring self-regulation, participantsin our study of loneliness in young adults per-formed a dichotic listening task while at theclinical research center. The dichotic listeningtask required that participants identify the con-sonant–vowel pair presented to their right or leftear. Because the auditory system is predomi-nantly crossed and because language is left-lateralized in most right-handed individuals,right-handed individuals tend to perform betterwhen verbal stimuli are presented to the rightthan left ear. (All of the participants in thisstudy were right-handed.) Superimposed on thisgeneral right-ear advantage, however, are theeffects of attention, as individuals generally per-form better when verbal stimuli are presented to

the ear to which they are focusing their atten-tion. The former effect is said to be data drivenor bottom-up, whereas the latter is said to beconceptually driven or top-down.

As predicted, we found an overall right-earadvantage across loneliness groups (lonely, nor-mal, and nonlonely) and instructional condi-tions (Cacioppo, Ernst, et al., 2000). In addition,however, a significant main effect of attentionalinstruction showed that individuals performedbetter with left-ear stimuli when they were in-structed to focus on the stimuli presented totheir left ear than in the other conditions. More-over, lonely individuals tended to show thestrongest right-ear advantage in the no-instruc-tion condition, presumably reflecting the po-tency of bottom-up (stimulus-driven) atten-tional processing, but failed to shift to the left-ear advantage when instructed to focus on theirleft ear. Specific planned contrasts confirmedthat all three groups showed a significant right-ear advantage during the focus on right-ear con-dition, but only the normal and nonlonely indi-viduals were able to shift to a significant left-earadvantage in the focus on left-ear condition.

Together, these results are consistent with thenotion that attentional control appears compa-rable in lonely and nonlonely individuals untilvoluntary attentional control conflicts with au-tomatic attentional processes, at which pointlonely individuals show an attentional deficit.This result raises the possibility that lonely in-dividuals may feel threatened or overwhelmedand therefore withdraw from the social environ-ment, especially new or complex social envi-ronments, because they have less control overthe focus of their attention than nonlonely indi-viduals. In fact, Baumeister and colleagues re-cently found that experimental manipulations ofsocial exclusion could produce a similar effecton dichotic listening and self-regulation(Baumeister & DeWall, in press).

A measure of the importance of social con-nectedness is that it predicts morbidity and mor-tality from broad-based causes (e.g., Seeman,2000). The reasons for this effect remain un-clear. As noted earlier, few differences in tradi-tional health behaviors (e.g., smoking, exercise,or nutrition) have been found to differentiatelonely and nonlonely individuals, for instance(Cacioppo, Hawkley, Crawford, et al., 2002;Seeman, 2000). We therefore have exploredother possible mechanisms by which lonelinessmay have deleterious effects on health: health


behaviors, cardiovascular activation, cortisollevels, and sleep (Cacioppo et al., 2003).

In one study, we assessed autonomic activity,salivary cortisol levels, sleep quality, and healthbehaviors in undergraduate students selected,on the basis of pretests, to be among the top orbottom quintile in feelings of loneliness (Ca-cioppo, Hawkley, Crawford, et al., 2002). Wefound that the total peripheral resistance toblood flow through the circulatory system in thebody was higher in lonely than nonlonely par-ticipants, whereas cardiac contractility (theforce of a heartbeat), heart rate, and cardiacoutput were higher in nonlonely than lonelyparticipants. Such differences, although not det-rimental in the robust cardiovascular system ofyoung adults, may constitute a source of wearand tear on the vascular system and on regula-tory controls of blood pressure, with cardiovas-cular dysfunctions not appearing until later inlife. We also found that lonely individuals re-ported poorer sleep than nonlonely individuals.Differences in sleep efficiency in young adults,as with differences in cardiovascular function,may have minimal health consequences in theshort term but may influence significant healthoutcomes over time, especially in later lifewhen physiological functions become morefragile.

In a follow-up study, we assessed blood pres-sure, heart rate, salivary cortisol levels, sleepquality, and health behaviors in older adultswhose loneliness was assessed at the time oftesting at their residence (Cacioppo, Hawkley,Crawford, et al., 2002). Results indicatedgreater age-related increases in blood pressureand poorer sleep quality in lonely than non-lonely older adults. These results suggest thatthe stress and dysphoria associated with feelingisolated contribute to wear and tear on the bodyand, over time, to the degradation of central andperipheral regulatory systems.

Humans are not static mechanical devicesthat simply wear out, but instead, human phys-iology includes anabolic processes that promoterepair and maintenance (e.g., wound healing)and growth (e.g., muscular development) in re-sponse to stressors. Sleep is the quintessentialrestorative behavior, however, and sleep in-volves no obvious social interaction. We there-fore next asked whether the restorative power ofsleep was greater for people who felt low ratherthan high levels of loneliness (Cacioppo, Hawk-ley, Berntson, et al., 2002). All participants

were tested one night in the clinical researchcenter of an academic hospital and several ad-ditional nights in their residence. Results re-vealed that lonely and nonlonely individualsallocated the same amount of time in bed forsleep, but lonely individuals evinced poorersleep efficiency and more time awake aftersleep onset than nonlonely individuals. Theseresults suggest that lonely individuals may beless resilient than nonlonely individuals in partbecause they sleep more poorly. These resultsalso raise the possibility that feelings of socialconnectedness may influence the extent towhich wear and tear on the brain and body canbe slowed or reversed by inherent anabolic pro-cesses. That is, sociality not only may influencethe selection of health behaviors but may mod-ulate the salubrity of restorative behaviors.

We designed the CHASRS to include de-tailed medical histories, health assessments,measures of health care use, health behaviormeasures (both self-report and objective), sleepquality indexes, personality measures (e.g., BigFive), life event assessments, nutrition and ex-ercise measures, markers of metabolic and in-flammatory processes, endocrine and immuneassays, baseline autonomic (especially cardio-vascular) assessments and tests, tests of mem-ory and cognitive function, exposure and reac-tivity to stressors, mood, volunteerism, opti-mism, hopefulness, church attendance, andreligiosity. Finally, economic and sociologicalcomponents (e.g., neighborhood characteristics,education, income, and social networks) havebeen incorporated into the design of theCHASRS. We have used these data, for in-stance, to examine the social, behavioral, cog-nitive, and emotional pathways through whichneighborhoods affect health and well-being. Us-ing data from the CHASRS, we recently foundthat neighborhood contexts predict self-reportedhealth in older adulthood even after controllingfor age and gender (Wen, Hawkley, & Ca-cioppo, 2004). More important, we found thatthe impact of neighborhood SES (e.g., educa-tion and income) on health was mediated in ourpopulation-based study through the subjectiveperception (meaning making) of neighborhood.The data further showed that these neighbor-hood effects on health were mediated indepen-dently by individual SES, loneliness/connected-ness, and depression but not by size of socialnetworks or perceived support.


As a complement to the Chicago population-based longitudinal study and in collaborationwith Linda Waite and M. E. Hughes, a three-item measure of loneliness was included in the2002 wave of the Health and Retirement Study,a longitudinal study of the later life course in anationally representative sample of individualsborn in 1947 or earlier (e.g., Hughes, Waite,Hawkley, & Cacioppo, 2004). At each wave(interview), detailed information is collectedabout the respondent’s health, family relation-ships, employment, income and wealth, and de-mographic background. A nationally represen-tative sample of 2,182 individuals completedthe module administered as part of theCHASRS. The three-item survey scale of lone-liness was found to have adequate internal con-sistency and to correlate highly with the fullrevised UCLA Loneliness Scale (Hughes et al.,2004).

Analyses of the Health and Retirement Studyand the Chicago population-based sample fur-ther indicated that objective social isolation andloneliness are related and that loneliness in bothsamples was similarly associated with poorerself-reported health, poorer self-reported emo-tional health, and a greater number of chronichealth conditions. As noted earlier, the hypoth-esis that the health behaviors of patients accountfor higher levels of morbidity and mortalityamong lonely or isolated individuals has re-ceived weak support. Care for patients has comeunder scrutiny in the past decade because stud-ies have shown that health care in the UnitedStates falls short of basic quality standards andthat health care provision is influenced subtly byvarious extraneous factors such as gender, race,and age. We reasoned that if the health careprovided to patients who appear to be sociallyisolated is of lower quality than that provided topatients with family and friends, then at leastpart of the explanation for the morbidity/mor-tality relationship may be attributable to thedecisions and behaviors of health care providersrather than to the health behaviors of the pa-tients themselves (Cacioppo, Brown, & Hawk-ley, 2004).

As an initial exploration of this hypothesis,we conducted a national study, randomly sam-pling 600 physicians whose practice centers onthe treatment of older adults. The physicianswere asked a series of questions comparing thecare of hospital inpatients with constructive andinvolved family members or friends and inpa-

tients who were socially isolated, along withanother series of questions comparing the careof inpatients with difficult or hostile familymembers or friends and inpatients who weresocially isolated. In each series of questions, thephysicians indicated the health care they ob-served to be provided by nurses, doctors, ancil-lary staff (e.g., physical therapists and nursesaides), and themselves. To allow the same sur-vey question to be used for physicians fromdifferent medical specialties and to address thecare provided by various health care agents(doctors, nurses, and ancillary staff), respon-dents were asked simply to indicate who re-ceived “better care.” Finally, we assessed at-tributes of the physicians such as medical spe-cialty, region of the country in which theypracticed, years of experience, and type of med-ical facility (e.g., hospital) to determine whetherthese factors moderated physicians’ beliefsabout the effects of their patients’ social milieuon health care provision.

Survey results indicated that physicians re-ported that they and other physicians, nurses,and ancillary staff provide better medical care toinpatients with constructive family and friendsthan inpatients without family or friends,whereas inpatients with hostile family andfriends receive treatment comparable to thatreceived by socially isolated inpatients. Thesefindings held for outpatients as well as inpa-tients and regardless of the physician’s medicalspecialty, type of hospital in which the physi-cian practiced, years of experience, and regionof the country. These results indicate that socialisolation may be associated with broad-basedmorbidity and mortality at least in part becausethe care provided by physicians and health careprofessionals differs.

Results from the CHASRS also confirmedthat sociality has profound effects on physio-logical functioning and sleep. As in our earlierstudies, we have observed that lonely olderadults exhibit evidence of chronically elevatedsympathetic activation (e.g., higher overnighturinary epinephrine and resting blood pressure),greater arterial stiffness, and less efficient oreffective sleep. These results suggest that socialisolation and loneliness have the potential tosimultaneously increase physiological load viasympathetic activation and decrease the capac-ity to recover from that load via reduced qualityof sleep.


Conclusion

Although the ancestral heritage of Homo sa-piens has placed an emphasis on sociality,Western thought and culture value individualexcellence over collective achievement. The ev-idence we have reviewed suggests our ancestralheritage is not easily ignored even if its effectsare not immediately obvious. Specifically, wehave argued that the human brain has evolved tofacilitate social information processing and ac-tion and that human contact and nurturance arenecessary for normal brain development andfunction. Loneliness is in part heritable, withindividual differences in the level of social con-nectedness required to feel right. Feelings ofsocial connectedness, or the lack thereof (i.e.,loneliness), were further found to be repre-sented conceptually as three facets of a singleoverarching construct, the facets being isola-tion/intimate connectedness, relational connect-edness, and collective connectedness. Thesefacets are correlated but are nevertheless relateddistinctively and predictably to life circum-stances. In addition, data from the CHASRSindicate that an individual’s self-esteem andsense of purpose in life reflect the individual’sperceived intimate, relational, and collectiveconnectedness. Dysphoria/depression, too, hasbeen found to be a distinct but related theoret-ical construct commonly triggered by social re-jection or disturbances (Cacioppo et al., inpress).

How we think about people in everyday lifemay be profoundly affected by feelings of socialconnectedness as well. For instance, lonely in-dividuals are more likely than nonlonely indi-viduals to construe their world, including thebehavior of others, as threatening or punitive.Consequently, lonely individuals are morelikely to be socially anxious, hold more nega-tive expectations for their treatment by others,and adopt a prevention focus rather than a pro-motion focus in their social interactions.Lonely, relative to nonlonely, individuals arealso more likely to appraise stressors as threatsrather than challenges and to cope with stressorsin a passive, isolative fashion rather than anactive fashion that includes actively seeking thehelp and support of others. Together, these dif-ferences in social cognition result predictably inan increased likelihood of lonely individualsacting in self-protective and, paradoxically,self-defeating ways. These dispositions, in turn,

activate social neurobehavioral mechanismsthat may contribute to the association betweenloneliness and mortality.

In summary, we would argue that it is time tomove beyond the solitary computer as a meta-phor for the human mind. Computers today aremassively interconnected devices with capaci-ties that extend far beyond the resident hard-ware and software of a solitary computer. In thisdevelopment, contemporary computers are be-coming a better metaphor for the human mind,as the telereceptors of the brain have providedwireless broadband interconnectivity to humansfor millennia.

References

Adolphs, R. (1999). Social cognition and the humanbrain. Trends in Cognitive Sciences, 3, 469–479.

Adolphs, R. (2001). The neurobiology of social cog-nition. Current Opinions in Neurobiology, 11,231–239.

Aronson, E. (1968). Dissonance theory: Progress andproblems. In R. P. Abelson, E. Aronson, W. J.McGuire, T. M. Newcomb, M. J. Rosenberg, &P. H. Tannenbaum (Eds.), Theories of cognitiveconsistency: A sourcebook (pp. 5–27). Chicago:Rand McNally.

Bar-On, R., Tranel, D., Denburg, N. L., & Bechara, A.(2003). Exploring the neurological substrate ofemotional and social intelligence. Brain, 126,1790–1800.

Baumeister, R. F., & DeWall, C. N. (in press). Theinner dimension of social exclusion: Intelligentthought and self-regulation among rejected per-sons. In K. Williams, J. P. Forgas, & W. vonHippel (Eds.), Sydney Social Psychology Sympo-sium. Mahwah, NJ: Erlbaum.

Berntson, G. G., Boysen, S. T., & Cacioppo, J. T.(1993). Neurobehavioral organization and the car-dinal principle of evaluative bivalence. Annals ofthe New York Academy of Sciences, 702, 75–102.

Berntson, G. G., & Cacioppo, J. T. (2000). Psycho-biology and social psychology: Past, present, andfuture. Personality and Social Psychology Re-view, 4, 3–15.

Blakemore, S., Winston, J., & Frith, U. (2004). Socialcognitive neuroscience: Where are we heading?Trends in Cognitive Sciences, 8, 216–222.

Boomsma, D. I., Willemsen, G., Dolan, C. V., Hawk-ley, L. C., & Cacioppo, J. T. (in press). Geneticand environmental contributions to loneliness inadults: The Netherlands Twin Register Study. Be-havior Genetics. Manuscript submitted for publi-cation.

Boomsma, D. I., Willemsen, G., Hawkley, L. C., &Cacioppo, J. T. (2004). Genetic and environmental


contributions to loneliness in adults: The Nether-lands Twin Register Study. Manuscript submittedfor publication.

Brewer, M. B., & Gardner, W. L. (1996). Who is this“we”? Levels of collective identity and self repre-sentations. Journal of Personality and Social Psy-chology, 71, 83–93.

Cacioppo, J. T. (2002). The risks and rewards ofirrational thinking. University of Chicago Record,37(3), 15–16.

Cacioppo, J. T., & Berntson, G. G. (1992). Socialpsychological contributions to the decade of thebrain: Doctrine of multilevel analysis. AmericanPsychologist, 47, 1019–1028.

Cacioppo, J. T., & Berntson, G. G. (2004). Keyreadings in social neuroscience. New York: Psy-chology Press.

Cacioppo, J. T., & Berntson, G. G. (in press-a). Thebrain, homeostasis, and health: Balancing demandsof the internal and external milieu. In H. S. Fried-man & R. Cohen Silver (Eds.), Oxford handbookof health psychology. New York: Oxford Univer-sity Press.

Cacioppo, J. T. & Berntson, G. G. (in press-b). Es-says in social neuroscience. Cambridge, MA: MITPress.

Cacioppo, J. T., Berntson, G. G., Adolphs, R., Carter,C. S., Davidson, R. J., McClintock, M. K., et al.(2002). Foundations in social neuroscience. Cam-bridge, MA: MIT Press.

Cacioppo, J. T., Berntson, G. G., Malarkey, W. B.,Kiecolt-Glaser, J. K., Sheridan, J., Poehlmann,K. M., et al. (1998). Automatic, neuroendocrine,and immune responses to psychological stress: Thereactivity hypothesis. Annals of the New YorkAcademy of Sciences, 840, 664–673.

Cacioppo, J. T., Berntson, G. G., Sheridan, J. F., &McClintock, M. K. (2000). Multilevel integrativeanalyses of human behavior: Social neuroscienceand the complementing nature of social and bio-logical approaches. Psychological Bulletin, 126,829–843.

Cacioppo, J. T., Brown, K., & Hawkley, L. C. (2004).Physicians’ reports of medical care as a functionof their patients’ social isolation. Manuscript sub-mitted for publication.

Cacioppo, J. T., Ernst, J. M., Burleson, M. H., Mc-Clintock, M. K., Malarkey, W. B., Hawkley, L. C.,et al. (2000). Lonely traits and concomitant phys-iological processes: The MacArthur social neuro-science studies. International Journal of Psycho-physiology, 35, 143–154.

Cacioppo, J. T., & Hawkley, L. C. (2003). Socialisolation and health, with an emphasis on under-lying mechanisms. Perspectives in Biology andMedicine, 46, S39–S52.

Cacioppo, J. T., & Hawkley, L. C. (in press). Peoplethinking about people: The vicious cycle of beinga social outcast in one’s own mind. In K. Williams,

J. P. Forgas, & W. von Hippel (Eds.), The socialoutcast: Ostracism, social exclusion, rejection,and bullying. New York: Psychology Press.

Cacioppo, J. T., Hawkley, L. C., & Berntson, G. G.(2003). The anatomy of loneliness. Current Direc-tions in Psychological Science, 12, 71–74.

Cacioppo, J. T., Hawkley, L. C., Berntson, G. G.,Ernst, J. M., Gibbs, A. C., Stickgold, R., & Hob-son, J. A. (2002). Lonely days invade the nights:Social modulation of sleep efficiency. Psycholog-ical Science, 13, 384–387.

Cacioppo, J. T., Hawkley, L. C., Crawford, L. E.,Ernst, J. M., Burleson, M. H., Kowalewski, R. B.,et al. (2002). Loneliness and health: Potentialmechanisms. Psychosomatic Medicine, 64, 407–417.

Cacioppo, J. T., Hawkley, L. C., Ernst, J. M., Burle-son, M., Bernston, G. G., Nouriani, B. & Spiegel,D. (in press). Loneliness within a nomological net:An evolutionary perspective. Journal of Researchin Personality.

Crabbe, J. C., Wahlsten, D., & Dudek, B. C. (1999,June 4). Genetics of mouse behavior: Interactionswith laboratory environment. Science, 284, 1670–1672.

Crick, F. (1970). Central dogma of molecular biol-ogy. Nature, 227, 561–563.

Dawkins, R. (1976). The selfish gene. New York:Oxford University Press.

Downey, G., Freitas, A. L., Michaelis, B., & Khouri, H.(1998). The self-fulfilling prophecy in close rela-tionships: Rejection sensitivity and rejection byromantic partners. Journal of Personality and So-cial Psychology, 75, 545–560.

Hawkley, L. C., Browne, M. W., & Cacioppo, J. T.(2004). Dimensions of loneliness: Perceived socialisolation, relational connectedness, and collectiveconnectedness. Manuscript submitted for publica-tion.

Hawkley, L. C., Burleson, M. H., Berntson, G. G., &Cacioppo, J. T. (2003). Loneliness in everydaylife: Cardiovascular activity, psychosocial context,and health behaviors. Journal of Personality andSocial Psychology, 85, 105–120.

Heider, F., & Simmel, M. (1944). An experimentalstudy of apparent behavior. American Journal ofPsychology, 57, 243–249.

Hughes, M. E., Waite, L. J., Hawkley, L. C., &Cacioppo, J. T. (2004). A 3-item version of therevised UCLA Loneliness Scale for use in tele-phone surveys: Chicago Health, Aging, and SocialRelations Study (CHASRS). Manuscript submittedfor publication.

Ito, T. A., & Cacioppo, J. T. (2000). Electrophysio-logical evidence of implicit and explicit categori-zation processes. Journal of Experimental SocialPsychology, 36, 660–676.

Jones, E. E., & Berglas, S. (1978). Control of attri-butions about the self through self-handicapping


strategies: The appeal of alcohol and the role ofunderachievement. Personality and Social Psy-chology Bulletin, 4, 200–206.

Kringelbach, M. L., & Rolls, E. T. (2004). The func-tional neuroanatomy of the human orbitofrontalcortex: Evidence from neuroimaging and neuro-psychology. Progress in Neurobiology, 72, 341–372.

LeDoux, J. (2000). Emotion circuits in the brain.Annual Review of Neuroscience, 23, 155–184.

McDonald, H. E., & Hirt, E. R. (1997). When ex-pectancy meets desire: Motivational effects in re-constructive memory. Journal of Personality andSocial Psychology, 72, 5–23.

McGuire, S., & Clifford, J. (2000). Genetic and en-vironmental contributions to loneliness in children.Psychological Science, 11, 487–491.

McGuire, W. J. (1981). The probabilogical model ofcognitive structure and attitude change. In R. E.Petty, T. M. Ostrom, & T. C. Brock (Eds.), Cog-nitive responses in persuasion (pp. 291–308). Hills-dale, NJ: Erlbaum.

Murray, S. L., Holmes, J., & Griffin, D. W. (1996).The self-fulfilling nature of positive illusions inromantic relationships: Love is not blind, but pre-scient. Journal of Personality and Social Psychol-ogy, 71, 1155–1180.

Nisbett, R. E., & Wilson, T. D. (1977). Telling morethan we can know: Verbal reports on mental pro-cesses. Psychological Review, 84, 231–259.

Pakkenberg, H. (1966). The number of nerve cells inthe cerebral cortex of man. Journal of Compara-tive Neurology, 128, 17–20.

Panksepp, J. (1998). Affective neuroscience. NewYork: Oxford University Press.

Petty, R. E., & Cacioppo, J. T. (1986). The elabora-tion likelihood model of persuasion. Advances inExperimental Social Psychology, 19, 123–205.

Purves, D., Lotto, R. B., & Nundy, S. (2002). Whywe see what we do. American Scientist, 90, 236–243.

Roediger, H. L., Buckner, R. L., & McDermott, K. B.(1999). Components of processing. In J. K. Foster& M. Jelicic (Eds.), Memory: Systems, process, orfunction? (pp. 31–65). New York: Oxford Univer-sity Press.

Ross, L., Greene, D., & House, P. (1977). The “falseconsensus effect”: An egocentric bias in social

perception and attribution processes. Journal ofExperimental Social Psychology, 13, 279–301.

Ross, M., & Sicoly, F. (1979). Egocentric biases inavailability and attribution. Journal of Personalityand Social Psychology, 37, 322–336.

Russell, D., Peplau, L. A., & Cutrona, C. E. (1980).The revised UCLA Loneliness Scale: Concurrentand discriminant validity evidence. Journal of Per-sonality and Social Psychology, 39, 472–480.

Saxe, R., & Kanwisher, N. (2003). People thinkingabout thinking people: The role of the temporo-parietal junction in “theory of mind.” NeuroIM-age, 19, 1835–1842.

Seeman, T. E. (2000). Health promoting effects offriends and family on health outcomes in olderadults. American Journal of Health Promotion, 14,362–370.

Snyder, M., & Swann, W. B., Jr. (1978). Hypothesis-testing processes in social interaction. Journal ofPersonality and Social Psychology, 35, 656–666.

Uchino, B. N., Cacioppo, J. T., & Kiecolt-Glaser,J. K. (1996). The relationship between social sup-port and physiological processes: A review withemphasis on underlying mechanisms and implica-tions for health. Psychological Bulletin, 119, 488–531.

Uchino, B. N., Kiecolt-Glaser, J. K., & Cacioppo,J. T. (1994). Construals of preillness relationshipquality predict cardiovascular response in familycaregivers of Alzheimer’s disease victims. Psy-chology and Aging, 9, 113–120.

Wen, M., Hawkley, L. C., & Cacioppo, J. T. (2004).Neighborhood environment, individual resources,and health in older adults: Chicago Health, Aging,and Social Relations Study (CHASRS). Manuscriptsubmitted for publication.

Wu, H., Wang, J., Cacioppo, J. T., Glaser, R.,Kiecolt-Glaser, J. K., & Malarkey, W. B. (1999).Chronic stress associated with spousal caregivingfor patients with Alzheimer’s dementia is associ-ated with downregulation of B-lymphocyte GHmRNA. Journal of Gerontology, 54, M212–M215.

Zajonc, R. B. (1980). Feeling and thinking: Prefer-ences need no inferences. American Psycholo-gist, 35, 157–193.

Received June 5, 2004Accepted September 28, 2004 �
