The Existential Theory of Mind

Jesse M. BeringFlorida Atlantic University

The primary causal explanatory model for interpreting behavior, theory of mind, mayhave expanded into corridors of human cognition that have little to do with the contextin which it evolved, questioning the suitability of domain-specific accounts of mindreading. Namely, philosophical–religious reasoning is a uniquely derived explanatorysystem anchored in intentionality that does not clearly involve behavior. The presenceof an existential theory of mind (EToM) suggests that individuals perceive somenondescript or culturally elaborated (e.g., God) psychological agency as having en-coded communicative intentions in the form of life events, similar to a person encodingcommunicative intentions in deictic gestures. The emergence of EToM is discussedfrom ontogenetic and phylogenetic perspectives; autism is examined to determinewhether alternate core explanatory models (e.g., folk physics) are used by those withdeficits in theory of mind to derive existential meaning.

The search for a precise understanding ofhow the human brain translates observable ac-tions into unobservable intentions has come todominate several major disciplines of the cog-nitive sciences. Cognitive developmentalists,comparative psychologists, and evolutionarypsychologists are all enamored with the studyof mind (Cheney & Seyfarth, 1990; Cosmides& Tooby, 1992; Flavell, 1999; Sperber, 1994;Tomasello & Call, 1997; Wellman, 1990), andrightly so. Successfully navigating through eventhe everyday, humdrum social world of humanrelations requires nothing less than a mechan-ism of such finesse and precision that it func-tions outside the boundaries of conscious aware-ness. A threatening glance from a stranger, agentle tug of a child’s hand on one’s sleeve, andthe rolled eyes of a disgruntled student are notisolated, meaningless behaviors emitted by me-chanical agents but the symbolic manifestationsof mental states intentionally transmitted asmessages from complex psychological agents tobe received by other complex psychologicalagents. Yet, even these peripherally social be-

haviors are, for most people, effortlessly en-riched with meaningful mentalistic interpretations.

Indeed, it is partially because of this extremeefficiency in the area of “mind reading” thatsome researchers have suggested that, perhaps,such processes are part of a larger adaptivesystem that has become modularized over evo-lutionary time (Baron-Cohen, 1995; Leslie,1987, 1991; Whiten, 1998). Because of its pri-mary importance for human sociality, this “the-ory of mind” system, as it has come to be called,has become an obvious candidate for the originsof sophisticated human social cognition and thecultural institutions created thereof. Some the-orists have gone so far as to claim that it, in fact,was the single evolutionary novelty that gavebirth to culture, adding such critical elementsinto the primate social melange as deception(Byrne & Whiten, 1988; R. Mitchell & Thomp-son, 1986), imitation (Tomasello, Kruger, &Ratner, 1993), teaching (Boesch, 1991), andmost notably language (Pinker, 1994; Toma-sello, 1999).

The intuitiveness of religious thought and itsrelationship to theory of mind, however, has inlarge measure been a stone unturned in thesedebates, or at least it has been kicked about onlyhalf seriously. Yet, philosophical–religious the-orizing seems an important link in the processof becoming human; for better or worse, it hasserved to psychologically extricate the humanorganism from the raw biological heritage of itsmind-blind past. Throughout history, it has

I am grateful to David Bjorklund, Alice Bering, JustinBarrett, Alan Nash, Jennifer Yunger, Edgar Schneider, andSymone Carroll for their discussions and very helpful com-ments on earlier versions of this article.

Correspondence concerning this article should be ad-dressed to Jesse M. Bering, Department of Psychology,Florida Atlantic University, Boca Raton, Florida 33431–0991. E-mail: jber4317@fau.edu

Review of General Psychology Copyright 2002 by the Educational Publishing Foundation2002, Vol. 6, No. 1, 3–24 1089-2680/02/$5.00 DOI: 10.1037//1089-2680.6.1.3

3

prospered into the higher arts of literature, mu-sic, and, even science. But even more important,it has reorganized the way individuals can rep-resent the unique experiences that composetheir own narratives.

In this article, I attempt a rapprochement be-tween the theory of mind system that has beenpopularized in contemporary cognitive aca-demia and a very basic, species-wide existentialtheory of mind (EToM) proposed as an indepen-dent system that, although built on the founda-tions of theory of mind, serves not to explain orpredict behavior but, rather, to allow individualsto attribute meaning to certain classes of auto-biographical experiences. This is by no meansthe first attempt of its kind, and others have infact seen a relationship between theory of mindand theism (e.g., J. L. Barrett & Keil, 1996;Boyer, 1994; Maser & Gallup, 1990); however,it is the first attempt, so far as I can gather, toclearly articulate the issues involved.

Existential Theory of Mind:Distinctiveness and Contextual Issues

I define EToM, in a purposefully generalsense, as a biologically based, generic explana-tory system that allows individuals to perceivemeaning in certain life events. The term mean-ing is used throughout this article in the sensethe philosopher Grice (1957) adopted when hespoke of nonnatural meaning, which he de-scribed as occurring wherever an agent “in-tended the utterance of x [message] to producesome effect in an audience by means of therecognition of this intention” (p. 385). In thecontext of EToM, meaning is viewed as anembedded conceptual property that a life eventeither does or does not possess, in the samemanner as the Gricean law just described.Whether it is present in a given situation de-pends solely on the attributions of the personexperiencing the event. A meaningful life eventis one that implies purpose or intention as thecausal force (e.g., “I was in a bad car accidentwhen I was a teenager because I needed to learnthat my life is fragile”). A meaningless lifeevent is one that does not imply purpose orintention as the causal force (e.g., “I saw theboat leave the dock”).

Underlying EToM is the interpretive percep-tion of natural events as symbolic of the com-municative attempts of some nondescript or cul-

turally elaborated (e.g., God) psychologicalagency. That is, an agent that does not fit intothe ontological category of known behavior-driven agents (i.e., humans, animals, or “inten-tional” artifacts such as robots) is perceived tohave orchestrated the life event of its own vo-lition and has framed the event as a symbolicdevice to transmit information to the experienc-ing human. For the meaningful case, an implicitquestion is the following: “What, or who, setout to intentionally teach the teenager about hisor her mortality?” To the religious individual,the answer might be God, but note that thestatement could just as easily have been madeby a nonreligious individual without explicitlyrepresenting a specific supernatural agent. Forthe meaningless case, the only agency involvedin moving the boat out of the dock is the humanorganism at the helm; the occurrence of theevent itself is not likely to be perceived asguided by any intentional source or as servingany communicative purpose. (An exception tothis, of course, would occur if for some reasonseeing the boat leave the dock was an especiallyimportant episode in the observer’s life, forexample, the observer’s daughter was aboardthe boat and attempting to run away from homeand the observer happened to see the boat slink-ing away from the shoreline.)

Closely related to the perception of meaningin random life events is the perception of mean-ing in the behaviors of other organisms. Ameaningful behavior might be “John is pointingto the corner ceiling of the room because hewants to show me where the spider made itsweb,” whereas a meaningless behavior might be“John extended his arm and index finger andheld it upward at a 45 degree angle 58 cm fromhis face.” (This might be something similar towhat people with autism see.) In the first case,attached to John’s pointing gesture are John’sintentions to show the observer the web; thisrepresentation of intentionality behind the be-havior lends the gesture itself meaning. In thesecond case, however, John’s mechanistic posi-tioning of his index finger does not encode anycommunicative purpose, and hence the behavioris meaningless.

According to this description, then, meaningstems from the mind of some intentional source.The reason the preceding analogy between theperception of ultimate meaning in life eventsand the perception of meaning in the behavior

4 BERING

of others is useful is that both tap reasoningprocesses involving mental causality. Note thatin the case of the meaningful life event, theindividual would not be satisfying the existen-tial question by answering that the car accidentoccurred because the other driver did not flashhis turn signal before changing lanes or becausehis tire blew out; rather, we want to know whathe perceives to be the ultimate meaning of theaccident (how the individual makes sense ofthe occurrence of the event itself). Likewise, inthe meaningful behavior case, when we ask theindividual why John is pointing, it is not appro-priate for him to answer “Because the axons ofthe corticospinal tract formed synapses withmotor neurons in the spinal cord and triggeredJohn’s arm and finger extension.” Mechanicalaccounts of behavior are of little help in maneu-vering through the social world because they donot satisfy a search for intentional causality.Ultimately, it is up to the individual to interpretthe behaviors of others in terms of underlyingmental states if he or she is to function normallyin society. The unfortunate impoverishments injust this area of reasoning render people withautism greatly impaired in their communicativeinteractions with others (see Baron-Cohen,1995; Frith, 1991; Leslie, 1991, 1994).

To perceive the symbolic referent in eitheractions or life events, one must be able to at-tribute mental states to the psychological agentsusing these devices as methods of symboliccommunication. It involves, in a sense, unwrap-ping the package that the communicatory de-vice arrives in—gesture, language, eye gaze, orlife event, among others—and arriving at thesemantic core of the primary perceptual experi-ence. However, although EToM shares many ofits core processes with those governing every-day social cognition and is a social system initself, it is important to stress that it workslargely in a different domain: that of experience,not behavior. It takes an entire complex of agiven life event, including the emotional prod-ucts associated with such an event, and trans-lates that complex into a symbolic message forthe self. In effect, it answers the question “Whydid this happen to me, of all people?” In con-trast, theory of mind is normally conceptualizedas being triggered only in response to action-related changes in the environment, such asbehaving agents or the self-propelledness ofinanimate objects.

EToM, then, describes a prepared drive in thehuman mind to represent some nondescriptagency—essentially, disembodied mind—andto view this agency as causing experiences.Although the most obvious example of a world-view directly dependent on EToM is a religiousinstitution that posits that gods or spirits ordeceased relatives are the main causative agentsin the daily affairs of an individual, EToM neednot be confined to such typically religiousmind-sets. EToM is not a set of ideas, nor is itnecessarily transmitted or mediated by socio-cultural processes such as are the dictates ofinstitutionalized religions or the highly formal-ized laws of established philosophical schools.Although these dictates and laws may involveEToM by postulating the form of agency thatdistinctively characterizes it, they do not cap-ture the naturalness of an evolved EToM, whichdoes not need such explicit causal reasoningstrategies to be instantiated in the individualmind.

Rather, ostensibly all that is required to gen-erate EToM is that individuals have alreadybuilt up their understanding of intentional rela-tions. Any person who has ever harbored ac-counts of his or her life that do not center on thephilosophical school of existentialism has em-ployed his or her EToM. Among other knee-jerk responses, anyone who has ever entertainednotions of individual purpose, questioned theequity of unfortunate biological or environmen-tal circumstances (e.g., disease, death, or naturaldisasters), attempted stoicism, succumbed to theappeal of the concept of destiny, found patternsor purpose in random life episodes, seenplanned order among nature, or felt compelledto decipher the deontic codes of human livinghas tacitly reflected upon the mind of someimmaterial agency and has placed—or at-tempted to place—meaning on the otherwisemeaningless.

For what or, better, who are such intuitivethoughts appealing to when they flitter, perhapsunconsciously, through our heads? What inten-tional source do we perceive to have causedsuch purposeful design in our experiences?Meaning arises through mind (Grice, 1957), andif meaning is ever perceived to have an exis-tence apart from human minds by becomingembedded in life events or existential signifiers,then EToM is implicated. As argued later, thenotion of God (and other culturally suited su-

5EXISTENTIAL THEORY OF MIND

pernatural agents, deities, or cosmic forces) maybe less an idea stemming from external sources(cf. Boyer, 1994; Sperber, 1996) than a descrip-tor of natural inferential processes. If belief inGod is any measure of the accuracy of theseclaims (and I believe it is), the overwhelmingmajority of people view the universe as beingtouched by the hands of psychological agency,precisely because humans happen to specializein the area of mentalistic causal reasoning. Theapplication of such reasoning to personal expe-riential factors in the context of self-narrative iswhat gives EToM its distinctiveness as a cog-nitive system.

But from whence did such an interpretivemeaning-based process arise? Certainly com-mensurate natural, random events took placelong before modern humans came onto thescene around 180,000 years ago impingingupon the livelihood of our ancestors. Primordialorganisms were displaced and killed in ancientearthquakes and wildfires, and creatures of theMiocene had their homes destroyed and theiroffspring injured by wild, prehistoric tornadoes.Australopithecines suffered the unpredictableelements, Homo habilis successfully caught andslaughtered prey animals or became prey itself,and Homo ergaster searched for mates and wascrippled prematurely by disease. But did thebrains of human ancestors see design in life-altering events or pattern in the vicissitudes ofbeing? What about other extant species, such aschimpanzees? Among modern humans, at whatontogenetic stage are people likely to developEToM, and how does it typically express itselfduring the course of normal development? CanEToM be stunted—or prevented—by a devel-opmental disorder that severely impoverishesthe ability to reason about the unobservablecauses of behavior, or do people with autismexpress an entirely different type of philosophi-cal–religious reasoning altogether? We are pre-vented from arriving at suitable answers formany of these questions because the issues havenot yet been empirically explored or have onlyjust begun to be explored.

We can, fortunately, find answers to morespecific related questions, and by so doing wemay begin to build a bridge to the larger onesposed earlier. In what follows, I review portionsof the literature on the evolutionary and devel-opmental emergence of theory of mind thatseem to provide some important insight into

EToM and usher in related theoretical viewsconcerning philosophical–religious reasoningwherever they might apply. Also, I offer sometentative hypotheses regarding what may hap-pen to the EToM “pull” when social–cognitivedevelopment goes astray, namely in individualsafflicted with autism.

EToM in Phylogeny

Previous authors have attempted to sketch aplausible scenario for the evolution of social–cognitive systems based on the fossil record anda smattered assemblage of prehistorical materialartifacts (Donald, 1991; Mithen, 1996). Al-though such an approach is highly worthwhileand can be exquisitely argued, the degree ofspeculation and creativity involved is a cause ofserious concern for many scientific investiga-tors, who largely prefer to deal with extantbrains to extinct ones. Fortunately, we can in-vestigate the evolution of human social cogni-tion through experimental means by studyingour closest living relatives, the great apes.

Modern estimates put the phyletic divergenceof humans from the African ape lineage at ap-proximately 5–7 million years ago. Becausechimpanzees (Pan troglodytes) are envisionedto be a conservative species in that they seem tohave mostly retained the morphological charac-teristics of the common ancestor and have alsoprobably retained a similar lifestyle, they areespecially important to comparative psycholo-gists interested in tracing the evolution ofthe human mind. The underlying assumptionamong such investigators is that if a cognitiveskill is markedly absent in the chimpanzee mindbut fundamental to the human mind, then theconceptual architecture supporting such a skillprobably arose at some point after the hominidsseparated from the rest of the great apes. Thus,by conducting carefully controlled experimentalstudies with chimpanzees, we may be able tocatch a glimpse of the way the common ances-tor might have viewed the world. Whether suchan assumption is warranted is still a questionopen to debate, but the consensus among mostscientists is that this methodological approach isboth highly valuable and theoretically justified(for a review, see Zihlman, 1996).

6 BERING

Representation of Intentionality inChimpanzees

There is something of a firestorm raging inthe comparative sciences, the center of the blazelying in the matter of whether chimpanzees canrepresent the intentional states of themselvesand others. Some theorists ardently claim thatindeed they can do so (Suddendorf & Whiten,2001), much in the manner of 2-year-old hu-mans; others (Heyes, 1998; Povinelli, Bering, &Giambrone, 2000; Tomasello & Call, 1997),often referred to as advocates of the killjoyhypothesis, forcibly argue the other side byholding that chimpanzees’ representational ca-pacity in the area of intentionality bears littleresemblance to that seen in toddlers. For presentpurposes, I need not get too deep in the concep-tual issues, but I must nonetheless get deepenough to convey what is involved in the vari-ous arguments.

To start with, a distinction must be madebetween those skills involved in secondary rep-resentation and metarepresentation. Accordingto Perner (1991), secondary representation oc-curs insofar as one is capable of representing asingular object of attention, say, a photograph,in multiple ways. Thus, a picture of a dog is, tothe possessing individual, both a representationof the dog and a four-cornered piece of glossypaper. This requires, according to Leslie (1987),a process of “decoupling” in which the primaryrepresentation must be temporarily held apartfrom the thing that it is actually about.

There is sufficient evidence to argue thatchimpanzees, in fact, are capable of this form ofsecondary representational thought (for a re-view, see Suddendorf & Whiten, 2001). Exper-imental apes’ imperative usage of ideographiclanguage systems, based mostly on arbitrarylexigrams, suggests a general symbolic capacity(Savage-Rumbaugh, 1986), as does their abilityto convert abstract numeric quantities to Arabicnumerals (Boysen, 1993). In addition, Kuhl-meier, Boysen, and Mukobi (1999) showed thatchimpanzees can solve scale-model, hidden re-ward tasks; after witnessing an experimenterhide a food reward in a miniaturized version ofthe laboratory, subjects successfully retrievedthe object in the real location that the model hadserved to represent.

The evidence that chimpanzees are able torepresent the desires and intentions of others

is much more tenuous than the preceding.Whereas findings from several studies suggest,for instance, that apes are able to distinguishbetween accidental outcomes and intentionalones, data from other experiments show quitethe opposite. Povinelli, Perilloux, Reaux, andBierschwale (1998), for example, discoveredthat their group of 6 juvenile chimpanzees wasjust as likely to select an experimenter who hadearlier maliciously poured out a juice reward asone who had stumbled and accidentally spilledthe juice.

More solid evidence for secondary represen-tation in the area of mental states comes fromgaze-following behavior. Several studies (Pov-inelli & Eddy, 1996; Tomasello, Hare, & Ag-netta, 1999) have independently shown thatchimpanzees visually track experimenters’ gazearound occlusive barriers, suggesting that theyare representing the psychological referent ofthe gaze rather than simply following the visualtrajectory in an automatic fashion (Whiten,1998). Yet this, in turn, does not agree with thelarger suite of findings reported in Povinelli andEddy’s (1996) monograph in which their groupof chimpanzee subjects did not seem able torepresent the perceptual state of seeing; the an-imals were just as likely to solicit an out-of-reach food item from an experimenter with ablinding bucket over her head as they were fromone paying rapt attention to them. But the mat-ter still does not rest here, it seems. There havebeen a number of writers (for a review, seeHare, 2001) who have called Povinelli and Ed-dy’s (1996) findings into question. Chief amongtheir concerns were the fact that Povinelli’schimpanzees had been reared in atypical envi-rons and the fact that the measures used werehighly artificial (chimpanzees in the wild, thesecritics claim, seldom encounter bucket-wearinghumans ready to relinquish food).

Recently, Hare, Call, Agnetta, and Tomasello(2000) provided empirical support for these re-joinders (see also Karin-D’Arcy & Povinelli,2001). When the testing conditions are embed-ded in a context of conspecific competition, aswhen a subordinate animal has visual access toa food item but the dominant does not, chim-panzees do in fact seem to appreciate the per-ceptual state of seeing (e.g., they will inhibittheir behavior toward the food while in thedominant’s presence).

7EXISTENTIAL THEORY OF MIND

However, what this implies for the issue ofmetarepresentation is unclear. Metarepresenta-tion, according to Perner (1991), involves theexplicit representation of mental states, partic-ularly those issuing from epistemic relations tothe perceptual world, such as knowledge andignorance. Whether Hare et al.’s (2000) apeswere operating at this level is still a questionopen for debate and one that awaits furtherreplication (Bering & Povinelli, in press; Karin-D’Arcy & Povinelli, 2001), but their findingsare greatly overshadowed by the larger numberof studies showing negative performance onmetarepresentational tasks. For instance, theonly carefully controlled study (of which I amaware) on false-belief understanding (false be-liefs are discussed later in the Emergence ofTheory of Mind in Ontogeny section) in chim-panzees produced no evidence of metarepresen-tation (Call & Tomasello, 1999). Chimpanzeesalso do not appear able to understand the com-municative intentions of others through gesturessuch as pointing (for reviews, see Povinelli,Bering, & Giambrone, in press; Povinelli,Reaux, Bierschwale, Allain, & Simon, 1997;Tomasello, Call, & Gluckman, 1997), nor dotheir own gestures appear to be regulated or toarise through an understanding of intentionalrelations between parties (Tomasello, Gust, &Frost, 1989). If intentional communication is tooccur, one agent must understand that the otherhas a message (secondary representation) toshare in the form of a gesture or vocalization(primary representation), and the other agentmust understand that the communicative partnerhas the mental wherewithal to receive thatmessage.

Although the jury is still out, I suspect that inthe area of sociality, chimpanzees do not takeothers’ mental states into consideration at anexplicit level; rather, they depend on evolvedbehavioral mechanisms in collaboration withunique ontogenetic experiences in response tothe observable actions exhibited by others. Forexample, mother chimpanzees evolved to besensitive to milk requests by their offspring, andjuveniles evolved to solicit milk when hungry.The particular way to go about this, however,might vary between mother–offspring dyadssuch that one juvenile might have initially got-ten milk by tugging on the mother’s arm,whereas another had to crawl upon the mother’sstomach for access to her mammary glands. The

first may eventually need to merely touch themother’s arm for her to position herself fornursing, because the entire sequence, as a resultof its repetition and the anticipation of themother, has been truncated. The second juve-nile, in contrast, may only have to approach themother ventrally. In both cases, we need notargue that the mother understands that her baby“wants” milk when it makes such approaches orthat the gesture is “about” wanting milk. Rather,the entire sequence becomes, borrowing fromTomasello and Call (1997), ontogenetically rit-ualized, such that both parties come to be suc-cessful participants in an established communi-cative set without really ever understanding theintentions of the other’s gestures.

The absence of metarepresentation in chim-panzees can explain why there are no definitivereports of true imitation, teaching, declarativepointing, or progressive cultural change amongwild populations, because all of these thingsdepend ultimately on the ability to explicitlyentertain what another individual does or doesnot know (Tomasello, 1999; Tomasello et al.,1993). Note, however, that I am not implyingthat chimpanzees lack symbolic competenciesaltogether. To the contrary, as argued earlier, itseems as if they have intact secondary represen-tational skills when applied to matters outsidethe mind. And, as Hare et al.’s (2000) resultshave shown, they may perhaps even have someimplicit awareness of goals and perceptualstates such as seeing.1

Most important for present purposes, how-ever, there is no reason to suspect that chimpan-zees can represent epistemic states, such asknowledge and ignorance, or other higher ordercognitions, or that they are able to representintentions at higher orders of recursiveness(e.g., “She knows that I know that she knows,”and so on). It is at this level, at which represen-tations of mental states have been “redescribed”(Karmiloff-Smith, 1992) to a degree allowingconscious reflection, that EToM is employed, in

1 There are as yet no grounds to claim that apes under-stand emotions. In contrast to Suddendorf and Whiten(2001), I find the evidence for empathic consolation,wherein chimpanzees have been observed to pacify themental duress of others, to be thoroughly unconvincing; onecan easily imagine how reconciliation and pacification be-haviors could have evolved in a highly political, socialspecies without invoking theory of mind to explain them.

8 BERING

that most of the time it involves a deliberate adhoc attempt to understand life experiences.

Proto-Religion in Chimpanzees

Boyer (1994) has convincingly argued that arecurrent component of all religions is the pos-tulation of nonphysical psychological agents.These transcendent agents are endowed with thefolk psychologies of humans but are somewhatless likely to be represented as possessing folkbiological traits such as the need to eat or drinkor being subject to disease and death. Thus,explicitly defined and culturally elaborated de-ities, spirits, demons, or ancestral relatives arerepresented as crystallized mind; they are con-ceptual encapsulations of distinct personalitiesbuilt of mental states and devoid of physicalbody. Deference, supplication, or even threattoward such agents demands the capacity toattribute mental states such as beliefs, knowl-edge, and intentions.

Irrespective of claims made earlier that chim-panzees probably are unable to intentionallytransmit cultural information by means ofshared representations (and hence the improba-bility of their mutual construction of a transcen-dent hierarchy), without an ability to metarep-resent chimpanzees cannot possess even the ru-diments of religious cognition (Bering, 2001;for an alternative account, see Maser & Gallup,1990). I propose that the causal reasoning pro-cesses involved in perceiving meaning in lifeevents were therefore probably not developedbefore the evolution of hominids. The intuitiveperception of meaning in particular event epi-sodes that characterizes EToM is an extensionof the theory of mind system and is tied to aneurological–cognitive apparatus that may nothave arrived until very recently in primate evo-lution. Chimpanzees cannot represent the ac-tions of others (e.g., gestures) as intentionallycommunicative (Povinelli et al., 1997; Toma-sello et al., 1989; for a review, see Povinelli etal., in press), nor can they understand the com-municative function of external symbols, eveniconic ones (Tomasello et al., 1997). Withoutthese inferential proclivities in the area of com-munication, there is no means for organisms toperceive “aboutness” or meaning in random lifeepisodes; such intuitive reasoning implies theperception of an external mind that intentionally

generates sequences of experiences for theindividual.

Does this mean that chimpanzees are “mere”existentialists, of the same ilk as Sartre andCamus? Not quite. Philosophical existentialismdemands an enormous assemblage of metarep-resentations to impose self-derived meaning onlife events. In addition, it is not a natural statebut an adopted one and is thus often initiallyaccompanied by a severe psychological crisis inthe human mind as EToM interpretations areforcefully shed (see subsequent discussion).

We can be fairly confident, therefore, in es-timates that the earliest forms of EToM did notcome into our ancestral portrait until approxi-mately 5–7 million years ago. Such a wide timeframe leaves us largely unsettled; however, it isbut a shard of geologic time in terms of theentire history of life on earth. And there is goodreason to suspect that EToM is more recentstill. The cultural explosion occurring between30,000 and 60,000 years ago in modern Homosapiens reveals the earliest material products ofart, technology, and animistic religions andmight well mark such a dramatic transition tometarepresentation (Mithen, 1996).

Yet some theorists, in contrast to this posi-tion, have attempted to argue that the cognitivebases of religion arose before the severing of thehuman lineage from the rest of the primate lineand that, accordingly, chimpanzees express re-ligious behaviors (Goodall, 1975; Guthrie,1993). Goodall (1975) described how chimpan-zees engage in elaborate threat displays in thepresence of rushing streams and in the midst ofloud thunderstorms, and she took such behavioras evidence of proto-religion in the species (seealso Whiten et al., 1999). Such displays, accord-ing to this rationale, are rituals directed towardthe streams or storms themselves, or else di-rected to a causal transcendent force in an at-tempt to ward off its admonition. The implica-tion of such a grand claim, of course (furtherelaborated by Guthrie, 1993), is that the originsof the cognitive underpinnings of religious be-havior are quite primitive, present in the mentalarchitecture of the common ancestor of chim-panzees and humans. However, this argument isnot borne out if we adhere to the earlier-de-scribed empirical findings of work related totheory of mind in chimpanzees. Because a foun-dational characteristic of all religious behaviorsinvolves representation of psychological agency,

9EXISTENTIAL THEORY OF MIND

and because chimpanzees seem to lack the ca-pacity to represent even embodied minds, thereis little justification for claims of proto-religionin Pan.2

According to Lawson and McCauley (1990,p. 54), “Religious rituals . . . involve communi-cation in the sense that they contain informationand transmit it.” Following this definition, ifchimpanzee “rain-dancing” is to qualify as re-ligious ritual, those enacting it must be inten-tionally sending a message they have encodedin symbolic action to be received and decodedby the mind of either the animated event or thetranscendent force generating it; they must alsorepresent the recipient as an intentional agentcapable of knowledge acquisition. I find it ex-ceedingly difficult to believe that this is whatthey are doing. Rather, what is arguably occur-ring in the rain-dancing case is that the chim-panzees are experiencing the rather primitivecognitive phenomenon of reflexive animism(Bering, 2001).

Reflexive animism is conceived as an adap-tation of the sympathetic nervous system toserve the individual organism in avoiding po-tential predation or injury from other organisms.It is exemplified in the cases of the vervet mon-key that emits an eagle alarm call in response toa falling leaf and the farmer who (much to hislater embarrassment) uses his pitchfork to punc-ture the garden hose that has taken on the move-ment of a snake while wrapped around his foot.In reflexive animism, inanimate objects are im-bued with the animate correlates of biologicalagency, such as unpredictable movements andloud noises. However, attributions of psycho-logical agency, such as higher order intentions,knowledge, and beliefs, do not occur. Such at-tributions, I have argued, are reserved to themore phylogenetically recent cognitive phe-nomenon of reflective animism, in which hu-mans quite seriously and deliberately endowinanimate objects, such as religious statues andidols, vehicles, computers, or weather episodes,with mental states (for a more comprehensivedistinction between reflexive and reflective an-imism, as well as a detailed theoretical treat-ment of religious percepts in chimpanzees, seeBering, 2001).

In the matter of ritualistic rain-dancing inwild chimpanzees, then, the persistent and ex-treme stimulus inducement inherent in treach-erous weather (shaking branches, loud noises,

and unpredictable movements) and loud, rush-ing streams probably triggers physiologicalarousal in chimpanzees that leads to displaybehavior, a hardwired response to rival chim-panzees or other perceived animate threats.Support for this interpretation of rain-dancing inchimpanzees comes from the fact that only adultmales have been seen to engage in the behavior(Andrew Whiten, personal communication,July 5, 2001). Interestingly, although physiolog-ical arousal might have much to do with EToMin organisms that specialize in mentalistic cau-sality, it is neither necessary nor sufficient forintuitive philosophical–religious reasoning andcan do nothing but engender scripted behavioralroutines in creatures without metarepresenta-tional skills. A theory of mind must serve as thescaffolding for such experiences if they are tobe perceived as meaningful.

EToM in Development

In significant contrast to chimpanzees, hu-mans are undeniably adept at psychologicalcausal reasoning. Indeed, so encompassing isour capacity to explain the behaviors of othersand ourselves in mentalistic terms, so much hashuman evolution invested in the theory of mindsystem, that the drive to attribute mental stateshas expanded into corridors of human cognitionthat probably had nothing whatever to do withthe system’s initial selection. In so doing, Iargue, it may have diversified into at least onewholly different animal—EToM—as it hascome to operate simultaneously in an entirelydifferent domain and with an entirely differentclass of input. Before addressing these issues,however, it is necessary to present a brief re-view of the developmental emergence of thetheory of mind system.

2 Interestingly, Maser and Gallup (1990) have argued thatchimpanzees do indeed possess a theory of mind and, there-fore, the necessary cognitive substrate for entertaining the-istic notions. However, because they do not have an aware-ness of their own deaths (which does little to support theauthors’ simulation-like model of theory of mind), arguedthe authors, chimpanzees have no reason to invoke theism topalliate existential anxiety. Although I agree with them thatchimpanzees are not explicitly aware of their impendingdeaths (see Bering, 2001), I disagree with them that deathawareness is necessary to instantiate representations of thesort involved in theism.

10 BERING

Emergence of Theory of Mind in HumanOntogeny

From early on, young infants are sensitive tothe differences between inanimate objects andanimate beings. A chair, a book, and a pumpkinare quite different things altogether from a rab-bit and a big sister, not, per se, simply becauseof some intrinsic taxonomy based on superficialcharacteristics (although this probably has a lotto do with it; see Bering, 2001) but because thelatter two are perceived as agents. That is, rab-bits and big sisters are capable of self-propelledmotion, and they are directed, goal-driven ani-mate beings, whereas the objects listed are sim-ply that. If a pumpkin were to scurry awaywhenever the youngster cried, it would be quitea strange sight indeed.

In fact, there have been a number of studiesthat have investigated whether young infantsare “surprised” by the ostensibly animate be-havior of inanimate objects (Golinkoff, Hard-ing, Carlson-Luden, & Sexton, 1984; Leslie,1982); the data suggest that they are, insofar aspatterns of dishabituation are accurate measuresof violated expectations. Hauser (1998) hasshown that this is something shared with otherprimates as well. In addition, when visual dis-plays on a computer screen are depicted asbehaving agents (e.g., a circle skirting around apartition), young infants readily dishabituate toseemingly anomalous “behavior” (e.g., the par-tition is removed but the circle still moves as ifits path were blocked; Gergely, Nadasdy, Csi-bra, & Biro, 1995). Such evidence argues for anunderstanding of agents as goal-directed, ani-mate beings in early infancy and perhaps inother species.

Such an understanding of other agents is stillprimitive, however, because at this level infantsprobably are only reasoning about the primaryrepresentational features of the immediate envi-ronment, such as movement or behavior, andhow these features conflict with an intuitive setof expectations about animacy. It is not until18–24 months of age that full-fledged second-ary representational abilities in the area of in-tentionality are clearly expressed, marked bysuch things as pretense (Leslie, 1987). The2-year-old observing her father dancing in thekitchen with a mop may find the whole scenequite funny because she is now able to tempo-rarily set aside the fact that the mop is just an

insensate object from the knowledge that herdad is simply pretending that the mop is adancing partner with a full head of hair. Toma-sello (1999) quite effectively argued that suchabilities might be driven by an understanding ofthe self as an intentional agent; because thechild knows what it is like to see one thing andact as if it is quite another, it is not such a greatleap for her to imagine that other people mightdo the same as well.

Also, 18-month-olds, but not younger chil-dren, understand that others might have desiresdifferent from theirs (Repacholi & Gopnik,1997), and older infants can infer what othersare intending to do, even if these others makemistakes or are unsuccessful in their attempts atachieving a goal (Meltzoff, 1995). The 2-year-old understands others as agents who can havetheir attention directed to entities and events inthe outside world (Baldwin & Moses, 1994;Butler, 1999; Flavell, 1992), and thus it is at thisage that protodeclarative gesturing, wherein in-dividuals seek to redirect an observer’s attentionto an outside event by pointing or showing forpurposes of sharing experience, emerges (Bates,1976; Kita, in press; Moore & Dunham, 1995).

In short, the child evolves over the first 2years of life from understanding that the selfand other agents do things (many of which areintuitively expected) to understanding that theself and other agents do things because. It is notfor several years more, however, that childrenbecome able to appreciate knowledge states atan explicit level. That is, although younger chil-dren have a concept of attention, intention, be-lief, and desire, they do not yet seem to connectthese states to what agents actually know aboutthe world, at least not at an explicit level (e.g.,see Clements & Perner, 1994). To test this,experimenters have used various versions of thefalse-belief task, which usually involves somenaive sap who has been played a trick on bychanging the location of, say, his candy bar,when he leaves the room, say, from the fridge tothe pantry. Where will he look for it when hereturns? Most 3-year-olds will incorrectly an-swer that he will look in the pantry, as if theactor had actually seen the hiding event,whereas 4-year-olds get it right and insist thatthe actor will think that his candy bar is still inthe place he had left it.

Thus, the older children seem to appreciatethe fact that others can harbor false beliefs and

11EXISTENTIAL THEORY OF MIND

will therefore be more likely to reason andexplain the behaviors of others correctly (forexcellent reviews, see Flavell, 1999; P. Mitch-ell, 1996). By no means have preschoolers ar-rived at an end point in conceptual development(e.g., see Bartsch & London, 2000)—their un-derstanding of minds will extend well intoadulthood—but they have indubitably reacheda developmental period that marks a point of noreturn. Older children cannot turn off theirmind-reading skills even if they want to. Allhuman actions are forevermore perceived to bethe products of unobservable mental states, andevery behavior, therefore, is subject to intensesociocognitive scrutiny.

Intentionality Outside Behavior

It seems not to be just behavior, however, thatbecomes the object of causal analyses. Ofcourse, I am not the first to propose that theoryof mind becomes generalized to other domains(see Carey, 1985, 1995; Clark, 1994; Kelemen,1999a, 1999b). Carey (1985, 1995), for in-stance, has suggested that children’s folk psy-chology leads them to become animists of thesort described by Piaget (1929). That is, chil-dren’s knowledge of the biological world (suchas what distinguishes the sun from a scorpion) isunderpinned by their theorizing about psycho-logical causality in both animate beings andinanimate objects. Thus, the animistic childmight say “The rock tumbled down the side ofthe mountain because it ‘wanted’ to get to thebottom.” Surely such an interpretation showsthat the child is inferring intention in the rock’smovement.

However, children’s knowledge about onto-logical categories seems to be more sophisti-cated than Carey (and Piaget) initially supposedand is precocious enough to suggest to Atran(1995) and others (e.g., Keil, 1989; Wellman &Gelman, 1992) that humans possess a folk bi-ology that is modular in its own right. Youngchildren have a rudimentary understanding ofgenetic transmission (Keil, 1989), understandthe biological correlates of living (e.g., need forsustenance) and the inevitableness of death forall living organisms (H. C. Barrett, 1999;Slaughter, Jaakola, & Carey, 1999), and, whenthe study design is sufficiently sensitive to im-plicit beliefs, evidence a firm understanding ofwhat separates living things from inanimate ob-

jects. Namely, they do not easily confuse theanimate world with the inanimate world unlessthere are misleading cues typical of living kinds(e.g., self-propelled movements or animatephysical features; see Premack & Premack,1997), and, for the most part, they reserve seri-ous mental state attribution to explaining thebehaviors of intentional agents.

But Carey might be on the right track afterall, just searching for overextension of mental-istic explanation in the wrong place. The novelcontribution of EToM rests in the fact that it canaccount for why life episodes, not the behaviorsof agents in the environment (either animate orinanimate with animate characteristics), are sooften explained by invoking intentional agency.The perception of meaning is, in fact, persevera-tive in the sense that it finds its way into expe-riential categories that, intuitively, have little todo with the behavior-reading context in which itpresumably evolved. By taking the intentionalstance, the human perceptual system recognizesnot only human action, with its characteristicunpredictability, but ambient life experiences,also with their characteristic unpredictability, asintentional behaviors. That is, a wide net is castover what is envisioned to be purposeful action:Human cognition situates the random churningsof the cosmos into the same framework inwhich it has placed human behavior. Mind isperceived to be the causal force behind bothcategories. The human brain effectively piratedthe theoretical system designed for interpretingaction (theory of mind) in its attempt to harnessand make sense of the unpredictable nature ofnature. And, indeed, it has succeeded in doingjust this. As the theory of mind system matures,it engrains itself into the individual’s perceptionof event-related phenomena, lending meaningto the experiences that come to define personalnarratives.

To illustrate the difference between EToMand animism, the girl witnessing the rock tum-bling down the side of the mountain may or maynot be attributing mental states to the rock itself,but she very well may perceive intention behindthe occurrence of the event. If we ask her “Whydid this occur, just at the time you happened tobe walking along the road?” we might receivean EToM response in which she reasons that ithappened “to tell her to go back home.” Notethat it is not the rock itself that intended to giveher this message; no animistic force propelled it

12 BERING

forward. Rather, some external agency causedthe event to occur.

Experimental support of the developmentalorigins of representing this form of agency canbe discovered in Kelemen’s (1999a, 1999b)findings of a “promiscuous teleology” in pre-schoolers. This term refers to the tendency ofyoung children to apply teleological–functionalpurposes to objects in both the artifactual world(e.g., chairs, cars, and forks) and the naturalworld (e.g., oceans, rocks, and dirt). Kelemen(1999b) showed that when asked questionsabout the purpose of inanimate objects, 4- and5-year-old children, for instance, stated thatclouds were “for raining” and that mountainswere “for climbing.” She concluded that “youngchildren promiscuously assert that entities of alltypes, including non-living natural objects, are‘made for something’ ” (Kelemen, 1999a, p.467).

The question of interest for present purposesis the following: “If such objects were designedfor a purpose, then whose intentions do theyreflect?” Evans (2000) and Gelman and Kremer(1991) found that a significant portion of youngchildren reasoned that natural objects weremade by God. The importance of these re-sponses resides not in the children’s theisticanswer but in that God is a culturally specificrepresentative of the form of psychologicalagency later involved in EToM. As a youngchild, the linguistic isolate Helen Keller, keptfrom the cultural realm of explicit ideas as aresult of her congenital deafness and blindness,reportedly spontaneously pondered the samequestions posed to the children in the studiesjust mentioned: “Who made the sky, the sea,everything?” (Bovet, 1928). What is intriguingabout her case is that, because others did nothave a means to communicate and share specificreligious beliefs with her, she could not applythe label “God” to this causal force but couldsimply represent some unspecified psychologi-cal agency (for a similar case, see James, 1892).

The functional explanations children give fornatural objects are attempts to make sense of theintentions of the perceived agent that createdthem. Whenever theistic answers are given forfundamental “why” questions, respondents areessentially exercising their theory of mind. Thisearly teleology coincides with the emergentmetarepresentational capacity children use toaccount for behavior in the social domain and is

sufficiently developed by 4–5 years to allowthem to successfully receive religious instruc-tion. Interestingly, this time period correspondsto the age at which children understand therelationship between God and prayer (seeWoolley, 2000).

When children evolve into mentalistic spe-cialists, they simultaneously step into nothingless than a theater of the mind. The world hasbecome a stage, occupied by the performance ofrandom events. The ironies; the hardships; theprosperity; the loneliness; the wealth of friend-ship, suffering, and dying; the accidents thejoys; the losses—the blueprints to a human bi-ography!—are perceived as intentionally driven.No longer do they “just happen” inside a vac-uum, but to the mind-reading child and adult,these experiences “happen because” and areimbued with purpose and, hence, meaning. Asdiscussed earlier, meaning in the sense usedhere must come from mind, and it rests upon therepresentations shared between members of aninteractive dyad. To extrapolate meaning fromsuch experiences, the recipient of the “message”must decode the symbolic device that the com-municator encoded it in. Only by so doing canthe recipient infer the intent of the sender. Inmany cases, the identity of the communicatormay be unknown, and literally unprocessed, bythe recipient of the message, as in the case ofproclaimed atheists or linguistic isolates. Butthe possible presence of an intuitive perceptionof intentional agency in the external world asdriving the experiences of the individual seri-ously challenges assertions that gods infiltratedhuman culture by way of idea transmission (cf.Boyer, 1994; Sperber, 1996). Rather, I arguethat the gods were already in our brains and thatculture merely gave them names.

EToM in Autism

Autism is a developmental disorder thatseems to impair the specific nexus of social–cognitive skills that are guided by theory ofmind. Whereas the vast majority of those withautism are also profoundly retarded, some au-tists are left with intact general intellectualskills; indeed, in such cases, their IQs frequentlytest well above average. Yet, even these indi-viduals cannot typically complete false-belieftasks in a successful manner, and they haveproblems in other areas that necessitate having

13EXISTENTIAL THEORY OF MIND

mind-reading abilities (Baron-Cohen, 1995;Frith, 1991; Happe & Frith, 1995; Leekam &Perner, 1990; Leslie, 1991, 1994; Swettenham,Baron-Cohen, Gomez, & Walsh, 1996; but seeFrith & Happe, 1995). Baron-Cohen (1990) ac-cordingly coined the term “mindblindness” as asynonym for autism but is sure to point out thatthe disorder should be conceptualized as involv-ing mindblindness to varying degrees. For in-stance, people with Asperger syndrome, a formof very high-functioning autism, might be ableto complete second-order false-belief tasks butwould have trouble making out the underlying,foggier shades of intention in dialogue, miss-ing faux pas in others’ speech and violatingGricean conversational maxims (Baron-Cohen,O’Riordan, Stone, Jones, & Plaisted, 1999;Fine, Bartolucci, Szatmari, & Ginsberg, 1994;Happe, 1993; Landa, 2000). It might be better toconceptualize higher functioning patients, then,as having never developed a fully erect inten-tional stance. Their sensitivity to intentionalityis seemingly attenuated rather than missingaltogether.

Baron-Cohen (2000) and his colleagues (Bar-on-Cohen, Wheelwright, Stone, & Rutherford,1999) have recently advanced the compellingargument that, although people with autism arenot mentalistic specialists and generally func-tion poorly in the social realm, their understand-ing of physical causality might actually be su-perior. That is, they may not be much in the wayof folk psychologists, but they excel in thedomain of folk physics. Folk physics, accordingto Baron-Cohen (2000, p. 1253), “is our every-day ability to understand and predict the behav-ior of inanimate objects in terms of principlesrelating to physical causality.” To support thisidea of superior folk physics in autism, he takesas evidence, among other things, the fact thatthe targets of obsessive–compulsive interests inautistic children overwhelmingly cluster aroundmachines and physical systems. In short, autistsare preoccupied with the way things work, suchas cameras, in terms of how they do so, not interms of why they do so.

In some high-functioning cases, this form ofexpertise is seemingly translated to problemsolving in the area of social matters, and itaffords individuals the capacity to exploit ob-servable cues emitted by behaving bodies and tosubsequently get by in the real world. Tantum(2000), for example, wrote of one Asperger

syndrome woman’s attempt to decipher implicitsocial codes, such as the average distance peo-ple leave between their bodies when waiting inline to use an automated teller machine.

She had observed that when people lined up, they lefta gap between themselves and the person in front, andthat this gap was substantially larger in the case of menstanding behind women. She used this information tojump lines, looking for this combination and pushingin behind the woman nearest the front who was fol-lowed by a man. (p. 382)

This woman’s understanding of the way peoplework was motivated by a desire to learn howthey typically behaved, not why. Only by as-sessing and becoming extraordinarily sensitiveto the way routines and conventional socialrules intersect with people’s manifest behaviorcould she enter the social environment, albeitinappropriately (she still could not take intoconsideration how others would perceive herbutting into the line).

There have been, to my knowledge, no stud-ies on religion in autism, nor are there anyrelated experimental data (e.g., teleological–functional explanations in autistic children) toshed light on these important questions; thus, itis unclear how those with the disorder entertainphilosophical–religious issues. Yet several re-cent autobiographical accounts provide intrigu-ing hints. For instance, Temple Grandin (1995),a well-known high-functioning autist, wrote inseveral passages of her lifelong struggle withher religious beliefs:

It is beyond my comprehension to accept anything onfaith alone, because of the fact that my thinking isgoverned by logic instead of emotion. (p. 189)

In high school I came to the conclusion that God wasan ordering force that was in everything. I found theidea of the universe becoming more and more disor-dered profoundly disturbing. (p. 191)

In nature, particles are entangled with millions ofother particles, all interacting with each other. Onecould speculate that entanglement of these particlescould cause a kind of consciousness for the universe.This is my current concept of God. (p. 200)

Another telling case comes from autisticmathematician and computer programmerEdgar Schneider’s (1999) book, DiscoveringMy Autism. In chapters devoted exclusively tohis religious beliefs, Schneider wrote:

My belief in the existence of a supreme intelligence(or, if you will, a God) is based on scientific factors.(p. 54)

14 BERING

It must be pointed out explicitly that none of this[religious beliefs] has any emotional underpinnings,but is totally intellectual in its nature. (p. 72)

To me, as far as adherence to a religion (or any othertype of ideology) is concerned, intellectual convictionis a condition that mathematicians call “both necessaryand sufficient.” My religious faith, I guess I could say,is not a gift from God, as so many people say; it is agift I gave to myself. In line with this, I have never feltthe emotional exhilaration that people must feel whenthey have a “religious experience.” This is true evenwhen I receive the sacraments. The only thing that hasdeeply moved me is the reasonableness of it all. (p. 73)

One gets the distinct impression by readingsuch descriptions that religion in autism con-sists of entirely different processes than thenormative experience. Namely, in these pas-sages, God, the sine qua non of Western reli-gious experience, is perceived more as a prin-ciple than as a rich psychological agent. God isperceived as a force in the universe that isdirectly responsible for the organization of cos-mic structure—arranging matter in an orderlyfashion or treating entropy—or has been re-duced to the conceptualization of scientificlogic altogether. To Schneider, instead of expe-riencing the emotional correlates of church rit-ual that often engender the physiology of “reli-gious” feelings, Catholicism is perceived as ananxiety-reducing medium with its formal, pre-dictable procedures and the clarity of its canons.

Although these views present an extraordi-narily sophisticated theological stance, it ringsquite differently from the overarching religioussentiments expressed in William James’s (1902/1994) classic text, Varieties of Religious Expe-rience, or current psychology of religion work(e.g., Hood, Spilka, Hunsberger, & Gorusch,1996). What is noticeably absent in the autisticaccounts is a sense of deep interpersonal rela-tions between the worshipper and the deity, asense of emotional dependency on an inten-tional agent who has control over the experi-ences and existence of the individual. God is notpresented as an agent who communicates inten-tions through indirect symbolic means or whohas a distinct constellation of mental states thatform a global personality structure. Rather, it isas if the autistic algorithmic strategies used todeal with other people have been translated tothe authors’ religious beliefs, wherein deductivelogic is used to lay the groundwork for under-standing existence and to impose order on anotherwise chaotic world. The sense of the nu-

minous, or otherness, inherent in normal reli-gious perception is not there.

This is precisely what we should expect,however, from a cognitive phenotype that ispreoccupied with uncovering fundamental prin-ciples of physical causality. Because peoplewith autistic spectrum disorders have difficultyinterpreting the meaning attached to social be-havior and therefore probably cannot rely on atheory of mind to explain their experiences,their religious beliefs cannot affix to core rep-resentations of psychological agency. The reli-gious beliefs of people with autism could there-fore be envisioned as sliding into conceptualslots provided by the folk physics system, eventhose, such as supernatural agent concepts, thatare traditionally relegated to slots in the folkpsychology system. Thus, supernatural agents,such as God, are perceived as behavioral ratherthan intentional agents.

Unlike people without the disorder, peoplewith autism or Asperger syndrome should beless likely to employ an EToM when confrontedwith life-altering events. This is because theevents themselves will not be understood asthings to be decoded (i.e., to derive meaning).The events will not be perceived as driven bythe mind of some intentional agent. Althoughtheir basic life events will largely mirror thoseof their nonautistic peers, their specific deficit intheory of mind abilities will prevent them fromattempting to interpret the meaning of thoseexperiences. Because autists cannot adequatelyrepresent psychological agency, they will notexperience the EToM pull triggered by particu-lar perceptual inputs to a cognitively normalmind. One man with Asperger syndrome in-quired recently on an Internet bulletin boardwhether others with the disorder were like him,“conscious of no feedback from the divine.”This characterizes the basic problems faced bypeople with Asperger syndrome: They seekmeaningful relationships with others but havegreat difficulty understanding the intentions be-hind people’s behaviors. Thus, translated to thearea of religion, a person with Asperger syn-drome might attempt to engage in ritualisticactivities (e.g., praying for a wife) but will notbe able to decode the symbolic device withwhich a supernatural agent “responds”(e.g., afemale friend’s husband leaves her for anotherwoman).

15EXISTENTIAL THEORY OF MIND

In general, if the thoughts presented hereapproach some measure of truth, then autisticpeople should show relatively little interest inspiritual matters or should explain existentialissues by means of physical causation, as in thepreceding accounts. Those who do adopt insti-tutionalized religious views should do so pri-marily to learn how to engage in acceptablebehavior within the culture. Autistic individualsshould therefore be disproportionately repre-sented in more rule-based, structured religions,especially those that have laid out their socialethical laws in written form and in which ad-herence to such laws is strictly enforced withinthe religious community. There is reason tosuspect that this is indeed the case (A. Attwood,personal communication, October 15, 2000).Also, as a result of the obsessive–compulsiveinterests of autists to rigidly adhere to socialrules once they have been learned, such indi-viduals might run the risk of proselytizing andbecoming exceedingly moralistic when con-fronted with religious rule breakers. This, too,seems to be supported by anecdotal reports.

Domain Specificity, Theory of Mind,and EToM

To account for the striking universal presenceof religion, numerous authors have recentlyspeculated that spiritual beliefs and behaviorshinge upon cognitive mechanisms that evolvedfor other adaptive purposes. Many have evensuggested that humans’ theory of mind can ac-count for the ease with which supernatural agentconcepts are transmitted and represented (J. L.Barrett & Keil, 1996; Boyer, 1994; Boyer &Walker, 2000; Hinde, 1999; Kirkpatrick, 1999).Beliefs in spirits or gods are envisioned tofeed off humans’ understanding of intentionalagents; these beings are endowed with the samegeneral catalogue of mental states (knowledge,intentions, desires, beliefs, and so forth) that abeing with a material brain possesses. Boyer(1994; Boyer & Walker, 2000) has suggestedthat religious ideas are developed within thecultural community and are spread throughmechanisms of social transmission, arguing thatthe ideas are intercepted by individual mindsand squeezed into the appropriate conceptualslots. Those ideas that are most “attention grab-bing,” such as the idea of people being dead andinvisible but also cognitively viable (e.g.,

ghosts), are the ones that will be maintained andincorporated into the religious explanatorysystem.

In a recent commentary on this area of in-quiry in the cognitive sciences, Sperber andHirschfeld (1999) stated that “religious beliefscan be seen as parasitical on domain-specificcompetencies that they both exploit and chal-lenge” (p. cxx). Such a sentiment captures theflavor of the argument being made: Religiousideas, including supernatural agent concepts,are products of cultural immersion that, afterbeing effectively transmitted between minds,settle into host modules such as theory of mind.I am somewhat partial to this sentiment as well.The trouble, however, is that with regard topsychological attributions, these authors havenot taken into account the fact that supernaturalagent concepts deal in a domain different fromthe one in which they envision their “domain-specific” module of theory of mind to operate.The notion of domain specificity crumbles, andthe very idea that theory of mind is modularsuffers a serious blow, when one considers thatintentional explanations can be evoked by en-tirely different classes of input: behavior andexperience.

Also, the arguments of these cognitive an-thropologists overlook the intuitiveness of phil-osophical–religious reasoning. Although ex-plicit religious beliefs indeed stem from thecultural milieu and find their way into the mindsof individuals through seeds planted by sociallearning processes, individuals must alreadycome bearing with them the representation ofsome nondescript psychological agency ascausing events. Rather than imbibing from cul-tural sources qualitatively novel ideas about thepresence of transcendent agents, culture pro-vides labels for perceptual experiences of suchagents that children operating with a theory ofmind are already very familiar with. Just as theperception of the color red is experienced beforethe word red becomes attached to it, perceptionof this agency is experienced in the human mindbefore, for example, the Western label “God”(or the African label “Gamab” or the Eskimolabel “Sila-Pinga”) becomes attached to it. Onlybecause the metarepresentational child can re-flectively attend to the class of perceptual expe-riences do such causal agent terms denote thatreligious pedagogy is successful at indoctrinat-ing new members into belief systems.

16 BERING

What I am suggesting is not reification inas-much as the perception of such agency existsuniversally in humans. Particular agent con-cepts may be reified with cultural elaboration ofthis agency into specific characters, but thisneed not concern cognitive scientists, only theo-logians. An individual’s enculturation withinthe religious community serves to carve up dif-ferent categories of experiences as belonging toparticular transcendent agents, with some reli-gious systems (e.g., Islamic monotheism) attrib-uting nearly all classes of life events to thehandiwork of a single agent and others (e.g.,American Indian polytheism) ascribing certaintypes of experiences to a wide variety of “spe-cialty” gods.3 Only after having been effec-tively communicated do explicit ideas regardingagent concepts work against a default back-ground of folk physics, folk biology, and, ofcourse, folk psychology (e.g., J. L. Barrett &Keil, 1996; Boyer, 1994; Kirkpatrick, 1999; butsee J. L. Barrett, Richert, & Driesenga, 2001),and are probably subject to changes during con-ceptual development (Elkind, 1979; Woolley,1997).

Is EToM an Adaptation?

A question inherent in this article is whetherEToM has its own evolutionary history,wrought by the forces of natural selection, orwhether its generalization from theory of mindimplies that it is merely a useless by-product ofthe primary adaptation of seeing intention be-hind behavior. Such questions of evolved traitsare notoriously tricky (see Buss, Haselton,Shackelford, Bleske, & Wakefield, 1998). Busset al. (1998, p. 537) described by-products ofadaptations as “characteristics that do not solveadaptive problems and do not have functionaldesign. They are carried along with character-istics that do have functional design becausethey happen to be coupled with those adapta-tions.” The whiteness of bones, they claimed, isa case in point, in that the color is a by-productof the large amount of calcium that gives bonestheir strength, which was what was selected for.

However, under certain conditions by-prod-ucts can come to have their own functions, asGould (1991) proposed with his concept of ex-aptations, which are “features that now enhancefitness, but were not built by natural selection

for their current role” (p. 47). The feathers ofbirds, originally selected for thermoregulationbut then co-opted for flight, are often used as anexample of an exaptation. Ironically, in contrastto Gould (1991), who cited religion as one ofmany functionless, useless by-products of alarge brain and certainly not an exaptation, itseems quite plausible to me that EToM has beenco-opted from a broad intentional stance takenby our ancestors, the primary adaptation ofwhich was to explain and predict behavior. Inother words, although it might not have been thedriving force behind mind-reading competen-cies in the species, it nevertheless prospered inits own right and in its own domain as it cameto have increasing adaptive utility.

The argument is thus: Nearly every culture, ifnot all, has established religious institutions or(at least) endorses specific religious beliefs thatare often seamlessly woven into the social struc-ture of the community (Boyer, 1994; Crook,1995; Hinde, 1999). Religious attitudes governeverything from the establishment of marriagerituals to the methods of slaughtering animals.The extent to which societal mores and religiousbeliefs are fluid and disparate across cultureshas nearly everything to do with the social ecol-ogy of the particular culture. Religions informtheir adherents of what they ought and ought notdo; these determinations are not based arbi-trarily on some limitless pool of human imagi-nation but exist because proscribed behaviorsare maladaptive for the individual in the contextof the culture.

Among the Sukumu of Tanzania, for in-stance, religious specialists encourage the kill-ing of deformed or retarded offspring, becausesuch children are taken to be evil tokens repre-senting an ancestor’s displeasure (Cory, 1951).In contrast, many Catholics in the United Statesview such children as a special gift from Godmeant to test their religious faith (Zuk, Miller,Bartram, & Kling, 1961). To them, killing in-fants is an atrocious act. Why the disparity?Reynolds and Tanner (1995) made an excellentpoint when they stated that “ethical abhorrence

3 Anthropologists would be hard put, however, to findcultures in which there does not exist some grouping of“second-tier” nonbiologic psychological agents, such asghosts, angels, demons, ancestral relatives, and so forth.

17EXISTENTIAL THEORY OF MIND

of infanticide occurs against a background ofgeneral affluence” (p. 91). That is, the particularopinions of the gods on the matter of infanticidedepend in large measure on whether parents cansupport such children without incurring botheconomic and social burdens that interfere withtheir reproductive success. Americans are morelikely able to shoulder the burden and increasetheir social status by doing so, whereas, from astrictly evolutionary psychological perspective,it would simply be maladaptive in the case ofpeople of less affluent societies to allow thesechildren to live. Therefore, in some hunter-gath-erer societies infanticide can be committedwithout the participants feeling any guilt orshame. A Western mother who kills her infant,in contrast, can expect to feel these emotionsunimaginably strongly and will have a tough goin finding another mate. In both cases, religioninserts itself, but just as the believers speakdifferent languages, so too do their gods. We aretold that good things happen to good people, butgoodness, to the evolutionist, is not absolute.

EToM, then, may function as an adaptivemechanism in the sense that it serves to regulatefluid morality in the social ecology. The factthat immanent justice and just world biases areentrenched in the tenets of world religions hintsat this regulation (see Gilbert, 1998). Actionsthat are maladaptive for the individual are oftenthose that cause disharmony in the group, suchas theft or adultery (Cosmides, 1989). Whensuch actions are labeled “profane” (Wilson,1978) and against the will of the gods, individ-uals become less likely to engage in them, be-cause doing so might be met with inauspiciousevents in their lives, such as sickness or finan-cial ruin. Thus, individuals’ task in society islargely to determine which behaviors have beendeemed wicked, not so much by their peers(because they may believe that they can getaway with these things unbeknownst to theirpeers) but, rather, by their omniscient gods, whohave the power to mete out experiential reper-cussions in response to social transgressions.The system can therefore by viewed as addedenforcement against individuals committingthose actions that might lead to declines instatus, castigation, imprisonment, execution,and other sundry social punishments that cutinto their reproductive success. By working inthe deontic, EToM may have therefore enabledthe explication of the implicit forces governing

behavioral natural selection. As such, EToM is,I argue, an adapted system, albeit a co-optedone, rather than a useless by-product of theoryof mind.

In line with this, the classes of experiencethat employ EToM, then, should have somefunctional bearing on the preceding. I suspectthat EToM should be activated to the extent thata life event is unexpected and (a) either deviatesfrom or conforms to culturally scripted justworld expectations and (b) triggers significantaffect-related change. The “unexpected” aspectis logistically necessary in that, because thecommunication between believer and agencyexists only in principle, it could not haveevolved by means of every social action beingmet immediately with like consequences in thedomain of experience. (This may have, inciden-tally, enforced its effects by promoting intermit-tent belief confirmation.) To the extent that itoperates in a different domain and can thereforebe conceived as an independent system boot-strapped to theory of mind, there also is thepossibility of a sort of existential autism. Thiswould occur insofar as there are individualswho have clearly articulated theories of mind inthe domain of behavior that did not generalizeto the domain of life experience.

EToM and Atheism

Through EToM, individuals are able to adoptthe position that their experiences are intention-ally guided and are occurring for a reason; itwould be difficult for them to view themselvesas “dust in the wind” even if they wanted to.By reflecting back on the salient events of ourown autobiographies—the outcomes of choicesmade, the unanticipated hazards that derailed usfrom a planned track, ironic encounters withfuture mates, the deaths of loved ones—it isexceedingly difficult not to see some intentionalpropulsion behind their occurrences (for aninteresting discussion, see McAdams, 2001).These are the events that lead to the construc-tion of our current selves and subsequently giveour lives meaning. Underlying all of this, ofcourse, is a very basic representation of mindbehind messages perceived to be intentionallycommunicated.

Such sensitivity to external meaning may, tosome limited extent, be forcefully unattended toin adulthood, but the forceful shedding of

18 BERING

EToM will probably lead to a period of intensepsychological suffering and the pangs of exis-tential crises (Frankl, 1963; May, 1960; Yalom,1980). People must adapt a meaning-based ex-planatory system that is wholly reliant upontheir own attribution of purpose to their expe-riences. William James (1902/1994) eloquentlydescribed individuals’ initial perception of ameaningless existence: “The world now looksremote, strange, sinister, uncanny. Its color isgone, its breath is cold, there is no speculationin the eyes it glares with” (p. 170). No longer isthere a transcendent explanation behind naturalevents. When this agency is killed off, peopleperceive the same phenomena as they did asEToM users, but they can no longer decode theintentional messages attached to their experi-ences. What contributes mostly to their dyspho-ria is the loss of ultimate meaning that their lifeonce had. They encounter the same type ofdilemma faced by solipsists in their questioningof the reality of other human minds. If no otherhuman minds exist, how can the behaviors ofothers have any meaning attached? Likewise, ifthere is no mind behind our existence, so tospeak, then how can we find purpose or mean-ing in the things that happen to us and, moregenerally, in our lives?

Certainly, there are individuals (not leastamong them scientists!) who claim not to sub-scribe to such seemingly romantic workings ofthe mind, but it is my impression that we wouldbe hard pressed to discover an individual ofnormal cognitive functioning who has neverexercised his or her EToM. Imagine the follow-ing: You are on a crowded bus, lost in thenewspaper before you, when suddenly you arecaught in a dizzying fury of screams, blackness,and crushing metal. Your bus has crashed andflipped over a steep embankment. You crawlout a window, dust yourself off, and realize thatyou are the sole survivor out of dozens of otherpassengers. If a week from now, or a year ordecade later, you find yourself asking “Whyme?” then quite simply you have an intactEToM. Even if you brush such questions asidebecause you consider them rather foolish, youstill betray your EToM insofar as you can en-tertain this type of question in the currentcontext.

To reiterate: We are dealing in the domain ofexperience, not behavior. Individuals may verywell be ideological atheists, and I do not doubt

their sincerity on the matter; however, much assolipsists cannot switch off theory of mindmechanisms, neither can EToM be totally dis-engaged in response to our acquired views onreligion. Indeed, experimental data (Weeks &Lupfer, 2000) reveal that individuals professingto be nonreligious are just as likely to makeintentional attributions in the domain of expe-rience as religious people are. The only differ-ence is that whereas believers turn to God inexplaining ironic or life-altering events, nonbe-lievers turn to a belief repertoire including suchelements as fate and immanent justice. In addi-tion, these nonreligious immaterial belief struc-tures are found cross culturally (Pepitone,1997).

An analogy may be helpful. Much as peoplethe world over have different religious beliefs,so too do they have different stylistic prefer-ences, and thus they decorate their homes withdifferent styles of furniture. Some prefer mod-ern design, others like Victorian furniture only,and still others are eclectic and prefer a widemixture of different furniture styles. When itcomes time to sit or sleep, or eat or study, theoccupants of each home use the furniture thatthey have chosen to perform these functions. AVictorian chair and a modern chair might lookvery different next to each other, but, in the end,they are both expert at doing what they weredesigned to do.

What about prisoners living in a barren prisoncell, however? When it comes time for them tosit down, as it surely will, they will do so. Butthey will not sit down on anything in particular.Nevertheless, the constraints on their body to sitwill override the fact that they have no special-ized furniture to accommodate their needs. Sim-ilarly, individuals who have not been exposed to(or have not accepted) religious beliefs or storedthese beliefs for later retrieval face the samecognitive constraints as those who have and willprobably still perceive psychological agency asdriving their experiences in some generic fash-ion. Simply because these individuals do nothave specific religious beliefs to accommodate acore causal attribution style does little to wearaway the core. The human brain is designed tosee meaning behind random events, even underthe wary, conscious eye of an intellect thatthinks it knows better.

19EXISTENTIAL THEORY OF MIND

Concluding Remarks

Only by examining alternative ways of per-ceiving behavior, such as in other species or inautism, are we able to recognize how the humancognitive system was specially adapted to rep-resent mental agency as driving action. At somepoint in evolutionary history after the humanlineage split from the African apes, I have ar-gued, the intentionality framework expanded toinclude those ambient life experiences that hu-mans had little or no control over. Just as it isalmost impossible to predict what another per-son is going to do or to be certain of whysomeone has just behaved in a particular way, itis similarly impossible to predict what naturehas in store for us or to understand why we havejust experienced something that may alter thecourse of our lives. EToM functions as a philo-sophical–religious explanatory system that al-lows us to see meaning in some of the thingsthat happen to us, affords us some sense ofperceived psychological control over what islikely to happen, enforces cultural mores thatadapt the individual to the group, and guardsagainst those behaviors that are maladaptive.

It is perhaps wise, given the subject matter ofthis article, to remember that the naturalisticfallacy is a double-edged sword. Just as atheismmay not exist except in principle, the gods maynot either. My intention has merely been toshow that life events are often perceived to beproducts of mental causality. EToM is a co-opted by-product of natural selection and occursinside our heads. Comfortable in my scientistskin, I will remain mute as to whether I believesuch perceptions are philosophically justified.

Nevertheless, such a large theory as the one Iam offering demands a large bed of data tosupport it, and I have done little but try to workin experimental evidence and case reports that,more often than not, only indirectly corroboratemy claims. Fortunately, the hypotheses embed-ded in this article are fully testable, and deter-mining the merit of these arguments should notbe impossible. For instance, whether the reli-gious experiences of people with autism arequalitatively different from those of peoplewithout the disorder is an empirical questionopen for exploration.

I have only now to summarize some of thepoints addressed in this article with a hypothet-ical example that should, I hope, clarify any

remaining confusion. If we are given four casesof mothers—a language-trained chimpanzeemother, an autistic mother, a religious mother,and a nonreligious mother—and told that eachof them has just sadly lost her infant as a resultof a disease, we might expect the followingresponses (or something similar) after askingthem why the death occurred. The chimpanzeewould grieve for the loss but otherwise look atus blankly; the autistic mother would speculatethat cancerous lesions had gotten a strongholdon her baby’s immunosuppressive system; thereligious mother would tell us that it was thewill of God; and the nonreligious mother wouldtell us that her baby died so that she could helpother bereaved mothers.

Whether this rings true for particular readerswill rest, I believe, on whether or not they havefound the general arguments presented in thisarticle to be convincing. But even if they findthese ideas still wanting, I would hope that thequestions raised here continue to be asked andthat evolutionary cognitive science opens itsdoors wider for such inquiry, because theseissues truly strike at the heart of what it meansto be human. Above all else, we are a specieswhose members are ravenous to understand thedrama of our own lives.

References

Atran, S. (1995). Causal constraints of categories. InD. Sperber, D. Premack, & A. J. Premack (Eds.),Causal cognition: A multi-disciplinary debate (pp.263–265). Oxford, England: Clarendon Press.

Baldwin, D. A., & Moses, L. J. (1994). Early under-standing of referential intent and attentional focus:Evidence from language and emotion. In C. Lewis& P. Mitchell (Eds.), Children’s early understand-ing of mind: Origins and development (pp. 133–156). Hillsdale, NJ: Erlbaum.

Baron-Cohen, S. (1990). Autism: A specific cogni-tive disorder of “mindblindness.” InternationalReview of Psychiatry, 2, 79–88.

Baron-Cohen, S. (1995). Mindblindness: An essay onautism and theory of mind. Cambridge, MA: MITPress.

Baron-Cohen, S. (2000). The cognitive neuroscienceof autism: Evolutionary approaches. In M. Gazza-niga (Ed.), The new cognitive neurosciences (2nded., pp. 1249–1257). Cambridge, MA: MIT Press.

Baron-Cohen, S., O’Riordan, M., Stone, V., Jones,R., & Plaisted, K. (1999). Recognition of faux pasby normally developing children with Aspergersyndrome or high-functioning autism. Journal of

20 BERING

Autism and Developmental Disorders, 29, 407–418.

Baron-Cohen, S., Wheelwright, S., Stone, V., &Rutherford, M. (1999). A mathematician, a phys-icist, and a computer scientist with Asperger syn-drome: Performance on folk psychology and folkphysics tests. Neurocase, 5, 475–483.

Barrett, H. C. (1999). Human cognitive adaptationsto predators and prey. Unpublished doctoral dis-sertation, Department of Anthropology, Universityof California, Santa Barbara.

Barrett, J. L., & Keil, F. C. (1996). Anthropomor-phism and God concepts: Conceptualizing a non-natural entity. Cognitive Psychology, 3, 219–247.

Barrett, J. L., Richert, R. A., & Driesenga, A. (2001).God’s beliefs versus mother’s: The developmentof non-human agent concepts. Child Development,72, 50–65.

Bartsch, K., & London, K. (2000). Children’s use ofmental state information in selecting persuasivearguments. Developmental Psychology, 36, 352–365.

Bates, E. (1976). Language in context: The acquisi-tion of pragmatics. New York: Academic Press.

Bering, J. M. (2001). Theistic percepts in other spe-cies: Can chimpanzees represent the minds of non-natural agents? Journal of Cognition and Cul-ture, 1, 107–137.

Bering, J. M., & Povinelli, D. J. (in press). Compar-ing cognitive development: From bridging gaps toaccepting diversity. In D. Maestripieri (Ed.), Pri-mate psychology: Bridging the gap between themind and behavior of human and nonhuman pri-mates. Cambridge, MA: Harvard University Press.

Boesch, C. (1991). Teaching among wild chimpan-zees. Animal Behaviour, 41, 530–532.

Bovet, P. (1928). The child’s religion: A study of thedevelopment of the religious sentiments. London:Dent.

Boyer, P. (1994). The naturalness of religious ideas:A cognitive theory of religion. Berkeley: Univer-sity of California Press.

Boyer, P., & Walker, S. (2000). Intuitive ontologyand cultural input in the acquisition of religiousconcepts. In K. S. Rosengran, C. N. Johnson, &P. L. Harris (Eds.), Imagining the impossible:Magical, scientific and religious thinking in chil-dren (pp. 130–156). New York: Cambridge Uni-versity Press.

Boysen, S. T. (1993). Counting in chimpanzees:Nonhuman principles and emergent properties ofnumber. In S. T. Boysen & E. J. Capaldi (Eds.),The development of numerical competence: Ani-mal and human models (pp. 39–59). Hillsdale, NJ:Erlbaum.

Buss, D. M., Haselton, M. G., Shackelford, T. K.,Bleske, A. L., & Wakefield, J. C. (1998). Adapta-

tions, exaptations, and spandrels. American Psy-chologist, 53, 533–548.

Butler, S. C. (1999). Early precursors to theory ofmind: Infant understanding of the referential na-ture of visual perception (comprehension). Disser-tation Abstracts International, 60(4-B), 1884.

Byrne, R. W., & Whiten, A. (1988). Machiavellianintelligence: Social expertise and the evolution ofintellect in monkeys, apes, and humans. NewYork: Oxford University Press.

Call, J., & Tomasello, M. (1999). A nonverbal falsebelief task: The performance of children and greatapes. Child Development, 70, 381–395.

Carey, S. (1985). Conceptual change in childhood.Cambridge, MA: MIT Press.

Carey, S. (1995). The growth of understandings ofnatural kinds. In D. Sperber, D. Premack, & A. J.Premack (Eds.), Causal cognition: A multi-disci-plinary debate (pp. 268–302). Oxford, England:Clarendon Press.

Cheney, D. L., & Seyfarth, R. M. (1990). How mon-keys see the world. Chicago: University of Chi-cago Press.

Clark, A. (1994). Beliefs and Desires Incorporated.Journal of Philosophy, 91, 404–425.

Clements, W. A., & Perner, J. (1994). Implicit un-derstanding of belief. Cognitive Development, 9,377–395.

Cory, H. (1951). The Ntemi: Traditional rites of aSukuma chief in Tanganyika. London: Macmillan.

Cosmides, L. (1989). The logic of social exchange:Has natural selection shaped how humans reason?Cognition, 31, 187–276.

Cosmides, L., & Tooby, J. (1992). Cognitive adap-tations for social exchange. In J. Barkow, L. Cos-mides, & J. Tooby (Eds.), The adapted mind:Evolutionary psychology and the generation ofculture (pp. 163–228). New York: Oxford Univer-sity Press.

Crook, J. H. (1995). Psychological processes in cul-tural and genetic evolution. In E. Jones & V.Reynolds (Eds.), Survival and religion: Biologicalevolution and cultural change (pp. 45–110).Chichester, England: Wiley.

Donald, M. (1991). Origins of the modern mind.Cambridge, MA: Harvard University Press.

Elkind, D. (1979). The child and society. New York:Oxford University Press.

Evans, E. M. (2000). The emergence of beliefs aboutthe origins of species in elementary school chil-dren. Merrill-Palmer Quarterly, 46, 2.

Fine, J., Bartolucci, G., Szatmari, P., & Ginsberg, G.(1994). Cohesive discourse in pervasive develop-mental disorders. Journal of Autism and Develop-mental Disorders, 14, 315–329.

Flavell, J. H. (1992). Perspective on perspective tak-ing. In H. Beilin & P. Pufall (Eds.), Piaget’s the-

21EXISTENTIAL THEORY OF MIND

ory: Prospects and possibilities (pp. 107–139).Hillsdale, NJ: Erlbaum.

Flavell, J. H. (1999). Cognitive development: Chil-dren’s knowledge about the mind. Annual Reviewof Psychology, 50, 21–45.

Frankl, V. E. (1963). Man’s search for meaning. NewYork: Simon & Schuster.

Frith, U. (1991). Autism and Asperger’s syndrome.Cambridge, England: Cambridge University Press.

Frith, U., & Happe, F. (1995). Autism: Beyond “the-ory of mind.” In J. Mehler & S. Franck (Eds.),Cognition on cognition (pp. 13–30). Cambridge,MA: MIT Press.

Gelman, S. A., & Kremer, K. E. (1991). Understand-ing natural cause: Children’s explanations of howobjects and their properties originate. Child Devel-opment, 62, 396–414.

Gergely, G., Nadasdy, Z., Csibra, G., & Biro, S.(1995). Taking the intentional stance at 12 monthsof age. Cognition, 56, 165–193.

Gilbert, P. (1998). Suffering, evolution, and psycho-therapy. In C. Feltham (Ed.), Witness and vision ofthe therapists (pp. 94–122). Thousand Oaks, CA:Sage.

Golinkoff, R. M., Harding, C. G., Carlson-Luden, V.,& Sexton, M. E. (1984). The infant’s perception ofcausal events: The distinction between animateand inanimate objects. In L. P. Lipsitt (Ed.), Ad-vances in infancy research (Vol. 3, pp. 145–151).Norwood, NJ: Ablex.

Goodall, J. (1975). The chimpanzee. In V. Goodall(Ed.), The quest for man (pp. 131–170). NewYork: Praeger.

Gould, S. J. (1991). Exaptation: A crucial tool forevolutionary psychology. Journal of Social Issues,47, 43–65.

Grandin, T. (1995). Thinking in pictures: And otherreports from my life with autism. New York:Doubleday.

Grice, H. P. (1957). Meaning. Philosophical Review,64, 377–388.

Guthrie, S. (1993). Faces in the clouds: A new theoryof religion. New York: Oxford University Press.

Happe, F. (1993). Communicative competence andtheory of mind in autism: A test of relevancetheory. Cognition, 48, 101–119.

Happe, F., & Frith, U. (1995). Theory of mind inautism. In E. Schopler & G. B. Mesibov (Eds.),Learning and cognition in autism (pp. 177–197).New York: Plenum.

Hare, B. (2001). Can competitive paradigms increasethe validity of experiments on primate social cog-nition? Animal Cognition, 4, 269–280.

Hare, B., Call, J., Agnetta, B., & Tomasello, M.(2000). Chimpanzees know what conspecifics doand do not see. Animal Behaviour, 59, 771–785.

Hauser, M. (1998). Expectations about object motionand destination: Experiments with a nonhumanprimate. Developmental Science, 1, 31–38.

Heyes, C. (1998). Theory of mind in nonhuman pri-mates. Behavioral and Brain Sciences, 21, 101–134.

Hinde, R. A. (1999). Why gods persist: A scientificapproach to religion. London: Routledge.

Hood, R. W., Spilka, B., Hunsberger, B., & Gorusch,R. (1996). The psychology of religion. New York:Guilford Press.

James, W. (1892). Thought before language: A deafmute’s recollections. Philosophical Review, 1,613–624.

James, W. (1994). The varieties of religious experi-ence. New York: Random House. (Original workpublished 1902)

Karin-D’Arcy, M. R., & Provinelli, D. J. (2001). Dochimpanzees know what each other see? A closerlook. Manuscript submitted for publication.

Karmiloff-Smith, A. (1992). Beyond modularity: Adevelopmental perspective on cognitive science.Cambridge, MA: MIT Press.

Keil, F. C. (1989). Concepts, kinds, and cognitivedevelopment. Cambridge, MA: MIT Press.

Kelemen, D. (1999a). Functions, goals, and inten-tion: Children’s teleological reasoning about ob-jects. Trends in Cognitive Sciences, 3, 461–468.

Kelemen, D. (1999b). Why are rocks pointy? Chil-dren’s preferences for teleological explanations ofthe natural world. Developmental Psychology, 35,1440–1452.

Kirkpatrick, L. (1999). Toward an evolutionary psy-chology of religion and personality. Journal ofPersonality, 67, 921–952.

Kita, S. (in press). Pointing: Where language, cul-ture, and cognition meet. Hillsdale, NJ: Erlbaum.

Kuhlmeier, V. A., Boysen, S. T., & Mukobi,K. L. R. P. (1999). Scale-model comprehension bychimpanzees (Pan troglodytes). Journal of Com-parative Psychology, 113, 396–402.

Landa, R. (2000). Social language use in Aspergersyndrome and high-functioning autism. In A. Klin,F. R. Volkmar, & S. S. Sparrow (Eds.), Aspergersyndrome (pp. 129–155). New York: GuilfordPress.

Lawson, E. T., & McCauley, R. N. (1990). Rethink-ing religion: Connecting religion and culture.Cambridge, England: Cambridge University Press.

Leekam, S., & Perner, J. (1990). Does the autisticchild have a metarepresentational deficit? Cogni-tion, 40, 203–218.

Leslie, A. (1991). The theory of mind impairment inautism: Evidence for a modular mechanism ofdevelopment. In A. Whiten (Ed.), Natural theoriesof mind: Evolution, development and simulation ofeveryday mindreading (pp. 63–78). Oxford, En-gland: Blackwell.

22 BERING

Leslie, A. (1994). ToMM, ToBY, and agency: Corearchitecture and domain specificity. In L. A. Hir-schfeld & S. Gelman (Eds.), Mapping the mind:Domain specificity in cognition and culture (pp.119–148). Cambridge, England: Cambridge Uni-versity Press.

Leslie, A. M. (1982). The perception of causality ininfants. Perception, 11, 173–186.

Leslie, A. M. (1987). Pretense and representation:The origins of “theory of mind.” PsychologicalReview, 94, 412–426.

Maser, J. D., & Gallup, G. G. (1990). Theism as aby-product of natural selection. Journal of Reli-gion, 90, 515–532.

May, R. (1960). Existential psychology. New York:Random House.

McAdams, D. P. (2001). The psychology of lifestories. Review of General Psychology, 5, 100–122.

Meltzoff, A. N. (1995). Understanding the intentionsof others: Re-enactment of intended acts by 18-month-old children. Developmental Psychology,31, 838–850.

Mitchell, P. (1996). Introduction to theory of mind:Children, autism, and apes. Oxford, England: Ox-ford University Press.

Mitchell, R., & Thompson, N. (1986). Deception:Perspective on human and nonhuman deceit. Al-bany: State University of New York Press.

Mithen, S. (1996). The prehistory of the mind: Thecognitive origins of art, religion and science. Lon-don: Thames & Hudson.

Moore, C., & Dunham, P. J. (1995). Joint attention:Its origins and role in development. Hillsdale, NJ:Erlbaum.

Pepitone, A. (1997). Nonmaterial beliefs: Theory andresearch in cultural social psychology. In C. Mc-Garty and S. A. Haslam (Eds.), The message ofsocial psychology: Perspectives on mind in society(pp. 252–267). Malden, MA: Blackwell.

Perner, J. (1991). Understanding the representa-tional mind. Cambridge, MA: MIT Press.

Piaget, J. (1929). The child’s conception of the world.London: Routledge & Kegan Paul.

Pinker, S. (1994). The language instinct. New York:HarperCollins.

Povinelli, D. J., Bering, J., & Giambrone, S. (2000).Toward a science of other minds: Escaping theargument by analogy. Cognitive Science, 24, 509–541.

Povinelli, D. J., Bering, J., & Giambrone, S. (inpress). Chimpanzee ‘pointing’: Another error ofthe argument by analogy? In S. Kita (Ed.), Point-ing: Where language, culture, and cognition meet.Hillsdale, NJ: Erlbaum.

Povinelli, D. J., & Eddy, T. J. (1996). What youngchimpanzees know about seeing. Monographs of

the Society for Research in Child Development,61, 3.

Povinelli, D. J., Perriloux, H. K., Reaux, J. E., &Bierschwale, D. T. (1998). Young and juvenilechimpanzees’ (Pan troglodytes) reactions to inten-tional versus accidental and inadvertent actions.Behavioural Processes, 42, 205–218.

Povinelli, D. J., Reaux, J. E., Bierschwale, D. T.,Allain, A. D., & Simon, B. B. (1997). Exploitationof pointing as a referential gesture in young chil-dren, but not adolescent chimpanzees. CognitiveDevelopment, 12, 423–461.

Premack, D., & Premack, A. J. (1997). Infants at-tribute value to the goal-directed actions of self-propelled objects. Journal of Cognitive Neuro-science, 9, 848–856.

Repacholi, B. M., & Gopnik, A. (1997). Early rea-soning about desires: Evidence from 14- and 18-month-olds. Developmental Psychology, 33, 12–21.

Reynolds, V., & Tanner, R. (1995). The social ecol-ogy of religion. New York: Oxford UniversityPress.

Savage-Rumbaugh, E. S. (1986). Ape language:From conditioned response to symbol. New York:Columbia University Press.

Schneider, E. (1999). Discovering my autism: Apo-logia pro vita sua (with apologies to CardinalNewman). London: Jessica Kingsley.

Slaughter, V., Jaakola, K., & Carey, S. (1999). Con-structing a coherent theory: Children’s biologicalunderstanding of life and death. In M. Siegal & C.Peterson (Eds.), Children’s understanding of biol-ogy, health, and ethics (pp. 71–96). Cambridge,England: Cambridge University Press.

Sperber, D. (1994). The modularity of thought andthe epidemiology of representations. In L. Hirsch-feld & S. Gelman (Eds.), Mapping the mind: Do-main specificity in cognition and culture (pp. 39–62). New York: Cambridge University Press.

Sperber, D. (1996). Explaining culture: A naturalis-tic approach. Oxford, England: Blackwell.

Sperber, D., & Hirschfeld, L. (1999). Culture, cog-nition, and evolution. In R. A. Wilson & F. C. Keil(Eds.), The MIT encyclopedia of the cognitive sci-ences (pp. cxi–cxxxii). Cambridge, MA: MITPress.

Suddendorf, T., & Whiten, A. (2001). Mental evolu-tion and development: Evidence for secondary rep-resentation in children, great apes, and other ani-mals. Psychological Bulletin, 127, 629–650.

Swettenham, J., Baron-Cohen, S., Gomez, J. C., &Walsh, S. (1996). What’s inside a person’s head?Conceiving of the mind as a camera helps childrenwith autism develop an alternative theory of mind.Cognitive Neuropsychiatry, 1, 73–88.

Tantum, D. (2000). Adolescence and adulthood ofindividuals with Asperger syndrome. In A. Klin,

23EXISTENTIAL THEORY OF MIND

F. R. Volkmar, & S. S. Sparrow (Eds.), Aspergersyndrome (pp. 367–399). New York: GuilfordPress.

Tomasello, M. (1999). The cultural origins of humancognition. Cambridge, MA: Harvard UniversityPress.

Tomasello, M., & Call, J. (1997). Primate cognition.New York: Oxford University Press.

Tomasello, M., Call, J., & Gluckman, A. (1997). Thecomprehension of novel communicative signs byapes and human children. Child Development, 68,1067–1081.

Tomasello, M., Gust, D., & Frost, G. T. (1989). Thedevelopment of gestural communication in youngchimpanzees: A follow up. Primates, 30, 35–50.

Tomasello, M., Hare, B., & Agnetta, B. (1999).Chimpanzees (Pan troglodytes) follow gaze direc-tion geometrically. Animal Behaviour, 58, 769–777.

Tomasello, M., Kruger, A. C., & Ratner, H. H.(1993). Cultural learning. Behavioral and BrainSciences, 16, 495–552.

Weeks, M., & Lupfer, M. B. (2000). Religious attri-butions and proximity of influence: An investiga-tion of direct interventions and distal explanations.Journal for the Scientific Study of Religion, 39,348–362.

Wellman, H. M. (1990). The child’s theory of mind.Cambridge, MA: MIT Press.

Wellman, H. M., & Gelman, S. A. (1992). Cognitivedevelopment: Foundational theories of core do-mains. Annual Review of Psychology, 43, 337–375.

Whiten, A. (1998). Evolutionary and developmentalorigins of the mindreading system. In J. Langer &

M. Killen (Eds.), Piaget, evolution, and develop-ment (pp. 73–99). Mahwah, NJ: Erlbaum.

Whiten, A., Goodall, J., McGrew, W. C., Nishida, T.,Reynolds, V., Sugiyama, Y., et al. (1999). Culturesin chimpanzees. Nature, 399, 682–685.

Wilson, E. O. (1978). On human nature. Cambridge,MA: Harvard University Press.

Woolley, J. D. (1997). Thinking about fantasy: Arechildren fundamentally different thinkers and be-lievers from adults? Child Development, 68, 991–1011.

Woolley, J. D. (2000). The development of beliefsabout direct mental-physical causality in imagina-tion, magic, and religion. In K. S. Rosengren, C. N.Johnson, & P. Harris (Eds.), Imagining the impos-sible: Magical, scientific, and religious thinking inchildren (pp. 99–129). Cambridge, England: Cam-bridge University Press.

Yalom, J. (1980). Existential psychotherapy. NewYork: Basic Books.

Zihlman, A. (1996). Reconstructions reconsidered:Chimpanzee models and human evolution. InW. C. McGrew, L. F. Marchant, & T. Nishida(Eds.), Great ape societies (pp. 293–304). Cam-bridge, England: Cambridge University Press.

Zuk, G. H., Miller, R. L., Bartram, J. B., & Kling, F.(1961). Maternal acceptance of retarded children:A questionnaire study of attitudes and religiousbackground. Child Development, 32, 525–540.

Received March 7, 2001Revision received July 9, 2001

Accepted July 10, 2001 �

24 BERING