The Existential Theory of Mind

Jesse M. Bering Florida Atlantic University

The primary causal explanatory model for interpreting behavior, theory of mind, may have expanded into corridors of human cognition that have little to do with the context in which it evolved, questioning the suitability of domain-speciﬁc accounts of mind reading. Namely, philosophicalÐreligious reasoning is a uniquely derived explanatory system anchored in intentionality that does not clearly involve behavior. The presence of an existential theory of mind (EToM) suggests that individuals perceive some nondescript or culturally elaborated (e.g., God) psychological agency as having encoded communicative intentions in the form of life events, similar to a person encoding communicative intentions in deictic gestures. The emergence of EToM is discussed from ontogenetic and phylogenetic perspectives; autism is examined to determine whether alternate core explanatory models (e.g., folk physics) are used by those with deﬁcits in theory of mind to derive existential meaning.

The search for a precise understanding of haviors are, for most people, effortlessly en- how the human brain translates observable ac- riched with meaningful mentalistic interpretations. tions into unobservable intentions has come to Indeed, it is partially because of this extreme dominate several major disciplines of the cog- efﬁciency in the area of “mind reading” that nitive sciences. Cognitive developmentalists, some researchers have suggested that, perhaps, comparative psychologists, and evolutionary such processes are part of a larger adaptive psychologists are all enamored with the study system that has become modularized over evo- of mind (Cheney & Seyfarth, 1990; Cosmides lutionary time (Baron-Cohen, 1995; Leslie, & Tooby, 1992; Flavell, 1999; Sperber, 1994; 1987, 1991; Whiten, 1998). Because of its pri- Tomasello & Call, 1997; Wellman, 1990), and mary importance for human sociality, this “the- rightly so. Successfully navigating through even ory of mind” system, as it has come to be called, the everyday, humdrum social world of human has become an obvious candidate for the origins relations requires nothing less than a mechan- of sophisticated human social cognition and the ism of such ﬁnesse and precision that it func- cultural institutions created thereof. Some the- tions outside the boundaries of conscious aware- orists have gone so far as to claim that it, in fact, ness. A threatening glance from a stranger, a was the single evolutionary novelty that gave gentle tug of a child’s hand on one’s sleeve, and birth to culture, adding such critical elements the rolled eyes of a disgruntled student are not into the primate social me«lange as deception isolated, meaningless behaviors emitted by me- (Byrne & Whiten, 1988; R. Mitchell & Thomp- chanical agents but the symbolic manifestations son, 1986), imitation (Tomasello, Kruger, & of mental states intentionally transmitted as Ratner, 1993), teaching (Boesch, 1991), and messages from complex psychological agents to most notably language (Pinker, 1994; Toma- be received by other complex psychological sello, 1999). agents. Yet, even these peripherally social be- The intuitiveness of religious thought and its relationship to theory of mind, however, has in large measure been a stone unturned in these I am grateful to David Bjorklund, Alice Bering, Justin debates, or at least it has been kicked about only Barrett, Alan Nash, Jennifer Yunger, Edgar Schneider, and half seriously. Yet, philosophicalÐreligious the- Symone Carroll for their discussions and very helpful com- orizing seems an important link in the process ments on earlier versions of this article. of becoming human; for better or worse, it has Correspondence concerning this article should be addressed to Jesse M. Bering, Department of Psychology, served to psychologically extricate the human Florida Atlantic University, Boca Raton, Florida 33431Ð organism from the raw biological heritage of its 0991. E-mail: [email protected] mind-blind past. Throughout history, it has

3 4 BERING prospered into the higher arts of literature, mu- turally elaborated (e.g., God) psychological sic, and, even science. But even more important, agency. That is, an agent that does not fit into it has reorganized the way individuals can rep- the ontological category of known behavior- resent the unique experiences that compose driven agents (i.e., humans, animals, or “inten- their own narratives. tional” artifacts such as robots) is perceived to In this article, I attempt a rapprochement be- have orchestrated the life event of its own vo- tween the theory of mind system that has been lition and has framed the event as a symbolic popularized in contemporary cognitive aca- device to transmit information to the experienc- demia and a very basic, species-wide existential ing human. For the meaningful case, an implicit theory of mind (EToM) proposed as an indepen- question is the following: “What, or who, set dent system that, although built on the founda- out to intentionally teach the teenager about his tions of theory of mind, serves not to explain or or her mortality?” To the religious individual, predict behavior but, rather, to allow individuals the answer might be God, but note that the to attribute meaning to certain classes of auto- statement could just as easily have been made biographical experiences. This is by no means by a nonreligious individual without explicitly the first attempt of its kind, and others have in representing a specific supernatural agent. For fact seen a relationship between theory of mind the meaningless case, the only agency involved and theism (e.g., J. L. Barrett & Keil, 1996; in moving the boat out of the dock is the human Boyer, 1994; Maser & Gallup, 1990); however, organism at the helm; the occurrence of the it is the first attempt, so far as I can gather, to event itself is not likely to be perceived as clearly articulate the issues involved. guided by any intentional source or as serving any communicative purpose. (An exception to Existential Theory of Mind: this, of course, would occur if for some reason Distinctiveness and Contextual Issues seeing the boat leave the dock was an especially important episode in the observer’s life, for Idefine EToM, in a purposefully general example, the observer’s daughter was aboard sense, as a biologically based, generic explana- the boat and attempting to run away from home tory system that allows individuals to perceive and the observer happened to see the boat slink- meaning in certain life events. The term meaning away from the shoreline.) ing is used throughout this article in the sense Closely related to the perception of meaning the philosopher Grice (1957) adopted when he in random life events is the perception of mean- spoke of nonnatural meaning, which he de- ing in the behaviors of other organisms. A scribed as occurring wherever an agent “in- meaningful behavior might be “John is pointing tended the utterance of x [message] to produce to the corner ceiling of the room because he some effect in an audience by means of the wants to show me where the spider made its recognition of this intention” (p. 385). In the web,” whereas a meaningless behavior might be context of EToM, meaning is viewed as an “John extended his arm and index finger and embedded conceptual property that a life event held it upward at a 45 degree angle 58 cm from either does or does not possess, in the same his face.” (This might be something similar to manner as the Gricean law just described. what people with autism see.) In the first case, Whether it is present in a given situation de- attached to John’s pointing gesture are John’s pends solely on the attributions of the person intentions to show the observer the web; this experiencing the event. A meaningful life event representation of intentionality behind the be- is one that implies purpose or intention as the havior lends the gesture itself meaning. In the causal force (e.g., “I was in a bad car accident second case, however, John’s mechanistic posi- when I was a teenager because I needed to learn tioning of his index finger does not encode any that my life is fragile”). A meaningless life communicative purpose, and hence the behavior event is one that does not imply purpose or is meaningless. intention as the causal force (e.g., “I saw the According to this description, then, meaning boat leave the dock”). stems from the mind of some intentional source. Underlying EToM is the interpretive percep- The reason the preceding analogy between the tion of natural events as symbolic of the com- perception of ultimate meaning in life events municative attempts of some nondescript or cul- and the perception of meaning in the behavior EXISTENTIAL THEORY OF MIND 5 of others is useful is that both tap reasoning EToM, then, describes a prepared drive in the processes involving mental causality. Note that human mind to represent some nondescript in the case of the meaningful life event, the agency—essentially, disembodied mind—and individual would not be satisfying the existen- to view this agency as causing experiences. tial question by answering that the car accident Although the most obvious example of a world- occurred because the other driver did not flash view directly dependent on EToM is a religious his turn signal before changing lanes or because institution that posits that gods or spirits or his tire blew out; rather, we want to know what deceased relatives are the main causative agents he perceives to be the ultimate meaning of the in the daily affairs of an individual, EToM need accident (how the individual makes sense of not be confined to such typically religious the occurrence of the event itself). Likewise, in mind-sets. EToM is not a set of ideas, nor is it the meaningful behavior case, when we ask the necessarily transmitted or mediated by socio- individual why John is pointing, it is not appro- cultural processes such as are the dictates of priate for him to answer “Because the axons of institutionalized religions or the highly formal- the corticospinal tract formed synapses with ized laws of established philosophical schools. motor neurons in the spinal cord and triggered Although these dictates and laws may involve John’s arm and finger extension.” Mechanical EToM by postulating the form of agency that accounts of behavior are of little help in maneu- distinctively characterizes it, they do not cap- vering through the social world because they do ture the naturalness of an evolved EToM, which not satisfy a search for intentional causality. does not need such explicit causal reasoning Ultimately, it is up to the individual to interpret strategies to be instantiated in the individual the behaviors of others in terms of underlying mind. mental states if he or she is to function normally Rather, ostensibly all that is required to gen- in society. The unfortunate impoverishments in erate EToM is that individuals have already just this area of reasoning render people with built up their understanding of intentional rela- autism greatly impaired in their communicative tions. Any person who has ever harbored ac- interactions with others (see Baron-Cohen, counts of his or her life that do not center on the 1995; Frith, 1991; Leslie, 1991, 1994). philosophical school of existentialism has em- To perceive the symbolic referent in either ployed his or her EToM. Among other knee- actions or life events, one must be able to at- jerk responses, anyone who has ever entertained tribute mental states to the psychological agents notions of individual purpose, questioned the using these devices as methods of symbolic equity of unfortunate biological or environmen- communication. It involves, in a sense, unwrap- tal circumstances (e.g., disease, death, or natural ping the package that the communicatory de- disasters), attempted stoicism, succumbed to the vice arrives in—gesture, language, eye gaze, or appeal of the concept of destiny, found patterns life event, among others—and arriving at the or purpose in random life episodes, seen semantic core of the primary perceptual experi- planned order among nature, or felt compelled ence. However, although EToM shares many of to decipher the deontic codes of human living its core processes with those governing every- has tacitly reflected upon the mind of some day social cognition and is a social system in immaterial agency and has placed—or at- itself, it is important to stress that it works tempted to place—meaning on the otherwise largely in a different domain: that of experience, meaningless. not behavior. It takes an entire complex of a For what or, better, who are such intuitive given life event, including the emotional prod- thoughts appealing to when they flitter, perhaps ucts associated with such an event, and trans- unconsciously, through our heads? What inten- lates that complex into a symbolic message for tional source do we perceive to have caused the self. In effect, it answers the question “Why such purposeful design in our experiences? did this happen to me, of all people?” In con- Meaning arises through mind (Grice, 1957), and trast, theory of mind is normally conceptualized if meaning is ever perceived to have an exis- as being triggered only in response to action- tence apart from human minds by becoming related changes in the environment, such as embedded in life events or existential signifiers, behaving agents or the self-propelledness of then EToM is implicated. As argued later, the inanimate objects. notion of God (and other culturally suited su- 6 BERING pernatural agents, deities, or cosmic forces) may EToM and usher in related theoretical views be less an idea stemming from external sources concerning philosophicalÐreligious reasoning (cf. Boyer, 1994; Sperber, 1996) than a descrip- wherever they might apply. Also, I offer some tor of natural inferential processes. If belief in tentative hypotheses regarding what may hap- God is any measure of the accuracy of these pen to the EToM “pull” when socialÐcognitive claims (and I believe it is), the overwhelming development goes astray, namely in individuals majority of people view the universe as being afflicted with autism. touched by the hands of psychological agency, precisely because humans happen to specialize in the area of mentalistic causal reasoning. The EToM in Phylogeny application of such reasoning to personal experiential factors in the context of self-narrative is Previous authors have attempted to sketch a what gives EToM its distinctiveness as a cog- plausible scenario for the evolution of socialÐ nitive system. cognitive systems based on the fossil record and But from whence did such an interpretive a smattered assemblage of prehistorical material meaning-based process arise? Certainly com- artifacts (Donald, 1991; Mithen, 1996). Al- mensurate natural, random events took place though such an approach is highly worthwhile long before modern humans came onto the and can be exquisitely argued, the degree of scene around 180,000 years ago impinging speculation and creativity involved is a cause of upon the livelihood of our ancestors. Primordial organisms were displaced and killed in ancient serious concern for many scientific investiga- earthquakes and wildfires, and creatures of the tors, who largely prefer to deal with extant Miocene had their homes destroyed and their brains to extinct ones. Fortunately, we can in- offspring injured by wild, prehistoric tornadoes. vestigate the evolution of human social cogni- Australopithecines suffered the unpredictable tion through experimental means by studying elements, Homo habilis successfully caught and our closest living relatives, the great apes. slaughtered prey animals or became prey itself, Modern estimates put the phyletic divergence and Homo ergaster searched for mates and was of humans from the African ape lineage at ap- crippled prematurely by disease. But did the proximately 5Ð7 million years ago. Because brains of human ancestors see design in life- chimpanzees (Pan troglodytes) are envisioned altering events or pattern in the vicissitudes of to be a conservative species in that they seem to being? What about other extant species, such as have mostly retained the morphological charac- chimpanzees? Among modern humans, at what teristics of the common ancestor and have also ontogenetic stage are people likely to develop probably retained a similar lifestyle, they are EToM, and how does it typically express itself especially important to comparative psycholo- during the course of normal development? Can gists interested in tracing the evolution of EToM be stunted—or prevented—by a devel- the human mind. The underlying assumption opmental disorder that severely impoverishes among such investigators is that if a cognitive the ability to reason about the unobservable skill is markedly absent in the chimpanzee mind causes of behavior, or do people with autism but fundamental to the human mind, then the express an entirely different type of philosophi- conceptual architecture supporting such a skill calÐreligious reasoning altogether? We are pre- probably arose at some point after the hominids vented from arriving at suitable answers for many of these questions because the issues have separated from the rest of the great apes. Thus, not yet been empirically explored or have only by conducting carefully controlled experimental just begun to be explored. studies with chimpanzees, we may be able to We can, fortunately, find answers to more catch a glimpse of the way the common ances- specific related questions, and by so doing we tor might have viewed the world. Whether such may begin to build a bridge to the larger ones an assumption is warranted is still a question posed earlier. In what follows, I review portions open to debate, but the consensus among most of the literature on the evolutionary and devel- scientists is that this methodological approach is opmental emergence of theory of mind that both highly valuable and theoretically justified seem to provide some important insight into (for a review, see Zihlman, 1996). EXISTENTIAL THEORY OF MIND 7

Representation of Intentionality in is much more tenuous than the preceding. Chimpanzees Whereas findings from several studies suggest, for instance, that apes are able to distinguish There is something of a firestorm raging in between accidental outcomes and intentional the comparative sciences, the center of the blaze ones, data from other experiments show quite lying in the matter of whether chimpanzees can the opposite. Povinelli, Perilloux, Reaux, and represent the intentional states of themselves Bierschwale (1998), for example, discovered and others. Some theorists ardently claim that that their group of 6 juvenile chimpanzees was indeed they can do so (Suddendorf & Whiten, just as likely to select an experimenter who had 2001), much in the manner of 2-year-old hu- earlier maliciously poured out a juice reward as mans; others (Heyes, 1998; Povinelli, Bering, & one who had stumbled and accidentally spilled Giambrone, 2000; Tomasello & Call, 1997), the juice. often referred to as advocates of the killjoy More solid evidence for secondary represen- hypothesis, forcibly argue the other side by tation in the area of mental states comes from holding that chimpanzees’ representational ca- gaze-following behavior. Several studies (Pov- pacity in the area of intentionality bears little inelli & Eddy, 1996; Tomasello, Hare, & Ag- resemblance to that seen in toddlers. For present netta, 1999) have independently shown that purposes, I need not get too deep in the concep- chimpanzees visually track experimenters’ gaze tual issues, but I must nonetheless get deep around occlusive barriers, suggesting that they enough to convey what is involved in the vari- are representing the psychological referent of ous arguments. the gaze rather than simply following the visual To start with, a distinction must be made trajectory in an automatic fashion (Whiten, between those skills involved in secondary rep- 1998). Yet this, in turn, does not agree with the resentation and metarepresentation. According larger suite of findings reported in Povinelli and to Perner (1991), secondary representation oc- Eddy’s (1996) monograph in which their group curs insofar as one is capable of representing a of chimpanzee subjects did not seem able to singular object of attention, say, a photograph, represent the perceptual state of seeing; the an- in multiple ways. Thus, a picture of a dog is, to the possessing individual, both a representation imals were just as likely to solicit an out-of- of the dog and a four-cornered piece of glossy reach food item from an experimenter with a paper. This requires, according to Leslie (1987), blinding bucket over her head as they were from a process of “decoupling” in which the primary one paying rapt attention to them. But the mat- representation must be temporarily held apart ter still does not rest here, it seems. There have from the thing that it is actually about. been a number of writers (for a review, see There is sufficient evidence to argue that Hare, 2001) who have called Povinelli and Ed- chimpanzees, in fact, are capable of this form of dy’s (1996) findings into question. Chief among secondary representational thought (for a re- their concerns were the fact that Povinelli’s view, see Suddendorf & Whiten, 2001). Exper- chimpanzees had been reared in atypical envi- imental apes’ imperative usage of ideographic rons and the fact that the measures used were language systems, based mostly on arbitrary highly artificial (chimpanzees in the wild, these lexigrams, suggests a general symbolic capacity critics claim, seldom encounter bucket-wearing (Savage-Rumbaugh, 1986), as does their ability humans ready to relinquish food). to convert abstract numeric quantities to Arabic Recently, Hare, Call, Agnetta, and Tomasello numerals (Boysen, 1993). In addition, Kuhl- (2000) provided empirical support for these re- meier, Boysen, and Mukobi (1999) showed that joinders (see also Karin-D’Arcy & Povinelli, chimpanzees can solve scale-model, hidden re- 2001). When the testing conditions are embed- ward tasks; after witnessing an experimenter ded in a context of conspecific competition, as hide a food reward in a miniaturized version of when a subordinate animal has visual access to the laboratory, subjects successfully retrieved a food item but the dominant does not, chim- the object in the real location that the model had panzees do in fact seem to appreciate the per- served to represent. ceptual state of seeing (e.g., they will inhibit The evidence that chimpanzees are able to their behavior toward the food while in the represent the desires and intentions of others dominant’s presence). 8 BERING

However, what this implies for the issue of first may eventually need to merely touch the metarepresentation is unclear. Metarepresenta- mother’s arm for her to position herself for tion, according to Perner (1991), involves the nursing, because the entire sequence, as a result explicit representation of mental states, partic- of its repetition and the anticipation of the ularly those issuing from epistemic relations to mother, has been truncated. The second juve- the perceptual world, such as knowledge and nile, in contrast, may only have to approach the ignorance. Whether Hare et al.’s (2000) apes mother ventrally. In both cases, we need not were operating at this level is still a question argue that the mother understands that her baby open for debate and one that awaits further “wants” milk when it makes such approaches or replication (Bering & Povinelli, in press; Karin- that the gesture is “about” wanting milk. Rather, D’Arcy & Povinelli, 2001), but their findings the entire sequence becomes, borrowing from are greatly overshadowed by the larger number Tomasello and Call (1997), ontogenetically rit- of studies showing negative performance on ualized, such that both parties come to be suc- metarepresentational tasks. For instance, the cessful participants in an established communi- only carefully controlled study (of which I am cative set without really ever understanding the aware) on false-belief understanding (false be- intentions of the other’s gestures. liefs are discussed later in the Emergence of The absence of metarepresentation in chim- Theory of Mind in Ontogeny section) in chimpanzees can explain why there are no definitive panzees produced no evidence of metarepresen- reports of true imitation, teaching, declarative tation (Call & Tomasello, 1999). Chimpanzees pointing, or progressive cultural change among also do not appear able to understand the com- wild populations, because all of these things municative intentions of others through gestures depend ultimately on the ability to explicitly such as pointing (for reviews, see Povinelli, entertain what another individual does or does Bering, & Giambrone, in press; Povinelli, not know (Tomasello, 1999; Tomasello et al., Reaux, Bierschwale, Allain, & Simon, 1997; 1993). Note, however, that I am not implying Tomasello, Call, & Gluckman, 1997), nor do that chimpanzees lack symbolic competencies their own gestures appear to be regulated or to altogether. To the contrary, as argued earlier, it arise through an understanding of intentional seems as if they have intact secondary represen- relations between parties (Tomasello, Gust, & tational skills when applied to matters outside Frost, 1989). If intentional communication is to the mind. And, as Hare et al.’s (2000) results occur, one agent must understand that the other have shown, they may perhaps even have some has a message (secondary representation) to implicit awareness of goals and perceptual share in the form of a gesture or vocalization states such as seeing.1 (primary representation), and the other agent Most important for present purposes, how- must understand that the communicative partner ever, there is no reason to suspect that chimpan- has the mental wherewithal to receive that zees can represent epistemic states, such as message. knowledge and ignorance, or other higher order Although the jury is still out, I suspect that in the area of sociality, chimpanzees do not take cognitions, or that they are able to represent others’ mental states into consideration at an intentions at higher orders of recursiveness explicit level; rather, they depend on evolved (e.g., “She knows that I know that she knows,” behavioral mechanisms in collaboration with and so on). It is at this level, at which represen- unique ontogenetic experiences in response to tations of mental states have been “redescribed” the observable actions exhibited by others. For (Karmiloff-Smith, 1992) to a degree allowing example, mother chimpanzees evolved to be conscious reflection, that EToM is employed, in sensitive to milk requests by their offspring, and juveniles evolved to solicit milk when hungry. 1 There are as yet no grounds to claim that apes under- The particular way to go about this, however, stand emotions. In contrast to Suddendorf and Whiten might vary between motherÐoffspring dyads (2001), I find the evidence for empathic consolation, such that one juvenile might have initially got- wherein chimpanzees have been observed to pacify the mental duress of others, to be thoroughly unconvincing; one ten milk by tugging on the mother’s arm, can easily imagine how reconciliation and pacification be- whereas another had to crawl upon the mother’s haviors could have evolved in a highly political, social stomach for access to her mammary glands. The species without invoking theory of mind to explain them. EXISTENTIAL THEORY OF MIND 9 that most of the time it involves a deliberate ad generates sequences of experiences for the hoc attempt to understand life experiences. individual. Does this mean that chimpanzees are “mere” existentialists, of the same ilk as Sartre and Proto-Religion in Chimpanzees Camus? Not quite. Philosophical existentialism demands an enormous assemblage of metarep- Boyer (1994) has convincingly argued that a resentations to impose self-derived meaning on recurrent component of all religions is the pos- life events. In addition, it is not a natural state tulation of nonphysical psychological agents. but an adopted one and is thus often initially These transcendent agents are endowed with the accompanied by a severe psychological crisis in folk psychologies of humans but are somewhat the human mind as EToM interpretations are less likely to be represented as possessing folk forcefully shed (see subsequent discussion). biological traits such as the need to eat or drink We can be fairly confident, therefore, in es- or being subject to disease and death. Thus, timates that the earliest forms of EToM did not explicitly defined and culturally elaborated de- come into our ancestral portrait until approxi- ities, spirits, demons, or ancestral relatives are mately 5Ð7 million years ago. Such a wide time represented as crystallized mind; they are con- frame leaves us largely unsettled; however, it is ceptual encapsulations of distinct personalities but a shard of geologic time in terms of the built of mental states and devoid of physical entire history of life on earth. And there is good body. Deference, supplication, or even threat reason to suspect that EToM is more recent toward such agents demands the capacity to still. The cultural explosion occurring between attribute mental states such as beliefs, knowl- 30,000 and 60,000 years ago in modern Homo edge, and intentions. sapiens reveals the earliest material products of Irrespective of claims made earlier that chim- art, technology, and animistic religions and panzees probably are unable to intentionally might well mark such a dramatic transition to transmit cultural information by means of metarepresentation (Mithen, 1996). shared representations (and hence the improba- Yet some theorists, in contrast to this posi- bility of their mutual construction of a transcen- tion, have attempted to argue that the cognitive dent hierarchy), without an ability to metarep- bases of religion arose before the severing of the resent chimpanzees cannot possess even the ru- human lineage from the rest of the primate line diments of religious cognition (Bering, 2001; and that, accordingly, chimpanzees express re- for an alternative account, see Maser & Gallup, ligious behaviors (Goodall, 1975; Guthrie, 1990). I propose that the causal reasoning pro- 1993). Goodall (1975) described how chimpan- cesses involved in perceiving meaning in life zees engage in elaborate threat displays in the events were therefore probably not developed presence of rushing streams and in the midst of before the evolution of hominids. The intuitive loud thunderstorms, and she took such behavior perception of meaning in particular event epi- as evidence of proto-religion in the species (see sodes that characterizes EToM is an extension also Whiten et al., 1999). Such displays, accord- of the theory of mind system and is tied to a ing to this rationale, are rituals directed toward neurologicalÐcognitive apparatus that may not the streams or storms themselves, or else di- have arrived until very recently in primate evo- rected to a causal transcendent force in an at- lution. Chimpanzees cannot represent the ac- tempt to ward off its admonition. The implica- tions of others (e.g., gestures) as intentionally tion of such a grand claim, of course (further communicative (Povinelli et al., 1997; Toma- elaborated by Guthrie, 1993), is that the origins sello et al., 1989; for a review, see Povinelli et of the cognitive underpinnings of religious be- al., in press), nor can they understand the com- havior are quite primitive, present in the mental municative function of external symbols, even architecture of the common ancestor of chim- iconic ones (Tomasello et al., 1997). Without panzees and humans. However, this argument is these inferential proclivities in the area of com- not borne out if we adhere to the earlier-de- munication, there is no means for organisms to scribed empirical findings of work related to perceive “aboutness” or meaning in random life theory of mind in chimpanzees. Because a foun- episodes; such intuitive reasoning implies the dational characteristic of all religious behaviors perception of an external mind that intentionally involves representation of psychological agency, 10 BERING and because chimpanzees seem to lack the ca- and unpredictable movements) and loud, rush- pacity to represent even embodied minds, there ing streams probably triggers physiological is little justification for claims of proto-religion arousal in chimpanzees that leads to display in Pan.2 behavior, a hardwired response to rival chim- According to Lawson and McCauley (1990, panzees or other perceived animate threats. p. 54), “Religious rituals...involve communi- Support for this interpretation of rain-dancing in cation in the sense that they contain information chimpanzees comes from the fact that only adult and transmit it.” Following this definition, if males have been seen to engage in the behavior chimpanzee “rain-dancing” is to qualify as re- (Andrew Whiten, personal communication, ligious ritual, those enacting it must be inten- July 5, 2001). Interestingly, although physiolog- tionally sending a message they have encoded ical arousal might have much to do with EToM in symbolic action to be received and decoded in organisms that specialize in mentalistic cau- by the mind of either the animated event or the sality, it is neither necessary nor sufficient for transcendent force generating it; they must also intuitive philosophicalÐreligious reasoning and represent the recipient as an intentional agent can do nothing but engender scripted behavioral capable of knowledge acquisition. I find it ex- routines in creatures without metarepresenta- ceedingly difficult to believe that this is what tional skills. A theory of mind must serve as the they are doing. Rather, what is arguably occur- scaffolding for such experiences if they are to ring in the rain-dancing case is that the chim- be perceived as meaningful. panzees are experiencing the rather primitive cognitive phenomenon of reflexive animism (Bering, 2001). EToM in Development Reflexive animism is conceived as an adaptation of the sympathetic nervous system to In significant contrast to chimpanzees, hu- serve the individual organism in avoiding po- mans are undeniably adept at psychological tential predation or injury from other organisms. causal reasoning. Indeed, so encompassing is It is exemplified in the cases of the vervet mon- our capacity to explain the behaviors of others key that emits an eagle alarm call in response to and ourselves in mentalistic terms, so much has a falling leaf and the farmer who (much to his human evolution invested in the theory of mind later embarrassment) uses his pitchfork to punc- system, that the drive to attribute mental states ture the garden hose that has taken on the move- has expanded into corridors of human cognition ment of a snake while wrapped around his foot. that probably had nothing whatever to do with In reflexive animism, inanimate objects are im- the system’s initial selection. In so doing, I bued with the animate correlates of biological argue, it may have diversified into at least one agency, such as unpredictable movements and wholly different animal—EToM—as it has loud noises. However, attributions of psycho- come to operate simultaneously in an entirely logical agency, such as higher order intentions, different domain and with an entirely different knowledge, and beliefs, do not occur. Such at- class of input. Before addressing these issues, tributions, I have argued, are reserved to the however, it is necessary to present a brief re- more phylogenetically recent cognitive phe- view of the developmental emergence of the nomenon of reflective animism, in which hu- theory of mind system. mans quite seriously and deliberately endow inanimate objects, such as religious statues and idols, vehicles, computers, or weather episodes, 2 Interestingly, Maser and Gallup (1990) have argued that with mental states (for a more comprehensive chimpanzees do indeed possess a theory of mind and, there- distinction between reflexive and reflective an- fore, the necessary cognitive substrate for entertaining theistic notions. However, because they do not have an aware- imism, as well as a detailed theoretical treat- ness of their own deaths (which does little to support the ment of religious percepts in chimpanzees, see authors’ simulation-like model of theory of mind), argued Bering, 2001). the authors, chimpanzees have no reason to invoke theism to In the matter of ritualistic rain-dancing in palliate existential anxiety. Although I agree with them that chimpanzees are not explicitly aware of their impending wild chimpanzees, then, the persistent and ex- deaths (see Bering, 2001), I disagree with them that death treme stimulus inducement inherent in treach- awareness is necessary to instantiate representations of the erous weather (shaking branches, loud noises, sort involved in theism. EXISTENTIAL THEORY OF MIND 11

Emergence of Theory of Mind in Human insensate object from the knowledge that her Ontogeny dad is simply pretending that the mop is a dancing partner with a full head of hair. Toma- From early on, young infants are sensitive to sello (1999) quite effectively argued that such the differences between inanimate objects and abilities might be driven by an understanding of animate beings. A chair, a book, and a pumpkin the self as an intentional agent; because the are quite different things altogether from a rab- child knows what it is like to see one thing and bit and a big sister, not, per se, simply because act as if it is quite another, it is not such a great of some intrinsic taxonomy based on superficial leap for her to imagine that other people might characteristics (although this probably has a lot do the same as well. to do with it; see Bering, 2001) but because the Also, 18-month-olds, but not younger chil- latter two are perceived as agents. That is, rab- dren, understand that others might have desires bits and big sisters are capable of self-propelled different from theirs (Repacholi & Gopnik, motion, and they are directed, goal-driven ani- 1997), and older infants can infer what others mate beings, whereas the objects listed are sim- are intending to do, even if these others make ply that. If a pumpkin were to scurry away mistakes or are unsuccessful in their attempts at whenever the youngster cried, it would be quite achieving a goal (Meltzoff, 1995). The 2-year- a strange sight indeed. old understands others as agents who can have In fact, there have been a number of studies their attention directed to entities and events in that have investigated whether young infants the outside world (Baldwin & Moses, 1994; are “surprised” by the ostensibly animate be- Butler, 1999; Flavell, 1992), and thus it is at this havior of inanimate objects (Golinkoff, Hard- age that protodeclarative gesturing, wherein in- ing, Carlson-Luden, & Sexton, 1984; Leslie, dividuals seek to redirect an observer’s attention 1982); the data suggest that they are, insofar as to an outside event by pointing or showing for patterns of dishabituation are accurate measures purposes of sharing experience, emerges (Bates, of violated expectations. Hauser (1998) has 1976; Kita, in press; Moore & Dunham, 1995). shown that this is something shared with other In short, the child evolves over the first 2 primates as well. In addition, when visual dis- years of life from understanding that the self plays on a computer screen are depicted as and other agents do things (many of which are behaving agents (e.g., a circle skirting around a intuitively expected) to understanding that the partition), young infants readily dishabituate to self and other agents do things because.Itisnot seemingly anomalous “behavior” (e.g., the par- for several years more, however, that children tition is removed but the circle still moves as if become able to appreciate knowledge states at its path were blocked; Gergely, Na«dasdy, Csi- an explicit level. That is, although younger chil- bra, & Biro«, 1995). Such evidence argues for an dren have a concept of attention, intention, be- understanding of agents as goal-directed, ani- lief, and desire, they do not yet seem to connect mate beings in early infancy and perhaps in these states to what agents actually know about other species. the world, at least not at an explicit level (e.g., Such an understanding of other agents is still see Clements & Perner, 1994). To test this, primitive, however, because at this level infants experimenters have used various versions of the probably are only reasoning about the primary false-belief task, which usually involves some representational features of the immediate envi- naive sap who has been played a trick on by ronment, such as movement or behavior, and changing the location of, say, his candy bar, how these features conflict with an intuitive set when he leaves the room, say, from the fridge to of expectations about animacy. It is not until the pantry. Where will he look for it when he 18Ð24 months of age that full-fledged second- returns? Most 3-year-olds will incorrectly an- ary representational abilities in the area of in- swer that he will look in the pantry, as if the tentionality are clearly expressed, marked by actor had actually seen the hiding event, such things as pretense (Leslie, 1987). The whereas 4-year-olds get it right and insist that 2-year-old observing her father dancing in the the actor will think that his candy bar is still in kitchen with a mop may find the whole scene the place he had left it. quite funny because she is now able to tempo- Thus, the older children seem to appreciate rarily set aside the fact that the mop is just an the fact that others can harbor false beliefs and 12 BERING will therefore be more likely to reason and jects. Namely, they do not easily confuse the explain the behaviors of others correctly (for animate world with the inanimate world unless excellent reviews, see Flavell, 1999; P. Mitch- there are misleading cues typical of living kinds ell, 1996). By no means have preschoolers ar- (e.g., self-propelled movements or animate rived at an end point in conceptual development physical features; see Premack & Premack, (e.g., see Bartsch & London, 2000)—their un- 1997), and, for the most part, they reserve seri- derstanding of minds will extend well into ous mental state attribution to explaining the adulthood—but they have indubitably reached behaviors of intentional agents. a developmental period that marks a point of no But Carey might be on the right track after return. Older children cannot turn off their all, just searching for overextension of mental- mind-reading skills even if they want to. All istic explanation in the wrong place. The novel human actions are forevermore perceived to be contribution of EToM rests in the fact that it can the products of unobservable mental states, and account for why life episodes, not the behaviors every behavior, therefore, is subject to intense of agents in the environment (either animate or sociocognitive scrutiny. inanimate with animate characteristics), are so often explained by invoking intentional agency. Intentionality Outside Behavior The perception of meaning is, in fact, persevera- tive in the sense that it finds its way into expe- It seems not to be just behavior, however, that riential categories that, intuitively, have little to becomes the object of causal analyses. Of do with the behavior-reading context in which it course, I am not the first to propose that theory presumably evolved. By taking the intentional of mind becomes generalized to other domains stance, the human perceptual system recognizes (see Carey, 1985, 1995; Clark, 1994; Kelemen, not only human action, with its characteristic 1999a, 1999b). Carey (1985, 1995), for in- unpredictability, but ambient life experiences, stance, has suggested that children’s folk psy- also with their characteristic unpredictability, as chology leads them to become animists of the intentional behaviors. That is, a wide net is cast sort described by Piaget (1929). That is, chil- over what is envisioned to be purposeful action: dren’s knowledge of the biological world (such Human cognition situates the random churnings as what distinguishes the sun from a scorpion) is of the cosmos into the same framework in underpinned by their theorizing about psycho- which it has placed human behavior. Mind is logical causality in both animate beings and perceived to be the causal force behind both inanimate objects. Thus, the animistic child categories. The human brain effectively pirated might say “The rock tumbled down the side of the theoretical system designed for interpreting the mountain because it ‘wanted’ to get to the action (theory of mind) in its attempt to harness bottom.” Surely such an interpretation shows and make sense of the unpredictable nature of that the child is inferring intention in the rock’s nature. And, indeed, it has succeeded in doing movement. just this. As the theory of mind system matures, However, children’s knowledge about onto- it engrains itself into the individual’s perception logical categories seems to be more sophisti- of event-related phenomena, lending meaning cated than Carey (and Piaget) initially supposed to the experiences that come to define personal and is precocious enough to suggest to Atran narratives. (1995) and others (e.g., Keil, 1989; Wellman & To illustrate the difference between EToM Gelman, 1992) that humans possess a folk bi- and animism, the girl witnessing the rock tum- ology that is modular in its own right. Young bling down the side of the mountain may or may children have a rudimentary understanding of not be attributing mental states to the rock itself, genetic transmission (Keil, 1989), understand but she very well may perceive intention behind the biological correlates of living (e.g., need for the occurrence of the event. If we ask her “Why sustenance) and the inevitableness of death for did this occur, just at the time you happened to all living organisms (H. C. Barrett, 1999; be walking along the road?” we might receive Slaughter, Jaakola, & Carey, 1999), and, when an EToM response in which she reasons that it the study design is sufficiently sensitive to im- happened “to tell her to go back home.” Note plicit beliefs, evidence a firm understanding of that it is not the rock itself that intended to give what separates living things from inanimate ob- her this message; no animistic force propelled it EXISTENTIAL THEORY OF MIND 13 forward. Rather, some external agency caused sufficiently developed by 4Ð5 years to allow the event to occur. them to successfully receive religious instruc- Experimental support of the developmental tion. Interestingly, this time period corresponds origins of representing this form of agency can to the age at which children understand the be discovered in Kelemen’s (1999a, 1999b) relationship between God and prayer (see findings of a “promiscuous teleology” in pre- Woolley, 2000). schoolers. This term refers to the tendency of When children evolve into mentalistic spe- young children to apply teleologicalÐfunctional cialists, they simultaneously step into nothing purposes to objects in both the artifactual world less than a theater of the mind. The world has (e.g., chairs, cars, and forks) and the natural become a stage, occupied by the performance of world (e.g., oceans, rocks, and dirt). Kelemen random events. The ironies; the hardships; the (1999b) showed that when asked questions prosperity; the loneliness; the wealth of friend- about the purpose of inanimate objects, 4- and ship, suffering, and dying; the accidents the 5-year-old children, for instance, stated that joys; the losses—the blueprints to a human bi- clouds were “for raining” and that mountains ography!—are perceived as intentionally driven. were “for climbing.” She concluded that “young No longer do they “just happen” inside a vac- children promiscuously assert that entities of all uum, but to the mind-reading child and adult, types, including non-living natural objects, are these experiences “happen because” and are ‘made for something’” (Kelemen, 1999a, p. imbued with purpose and, hence, meaning. As 467). discussed earlier, meaning in the sense used The question of interest for present purposes here must come from mind, and it rests upon the is the following: “If such objects were designed representations shared between members of an for a purpose, then whose intentions do they interactive dyad. To extrapolate meaning from reflect?” Evans (2000) and Gelman and Kremer such experiences, the recipient of the “message” (1991) found that a significant portion of young must decode the symbolic device that the com- children reasoned that natural objects were municator encoded it in. Only by so doing can made by God. The importance of these re- the recipient infer the intent of the sender. In sponses resides not in the children’s theistic many cases, the identity of the communicator answer but in that God is a culturally specific may be unknown, and literally unprocessed, by representative of the form of psychological the recipient of the message, as in the case of agency later involved in EToM. As a young proclaimed atheists or linguistic isolates. But child, the linguistic isolate Helen Keller, kept the possible presence of an intuitive perception from the cultural realm of explicit ideas as a of intentional agency in the external world as result of her congenital deafness and blindness, driving the experiences of the individual seri- reportedly spontaneously pondered the same ously challenges assertions that gods infiltrated questions posed to the children in the studies human culture by way of idea transmission (cf. just mentioned: “Who made the sky, the sea, Boyer, 1994; Sperber, 1996). Rather, I argue everything?” (Bovet, 1928). What is intriguing that the gods were already in our brains and that about her case is that, because others did not culture merely gave them names. have a means to communicate and share specific religious beliefs with her, she could not apply EToM in Autism the label “God” to this causal force but could simply represent some unspecified psychologi- Autism is a developmental disorder that cal agency (for a similar case, see James, 1892). seems to impair the specific nexus of socialÐ The functional explanations children give for cognitive skills that are guided by theory of natural objects are attempts to make sense of the mind. Whereas the vast majority of those with intentions of the perceived agent that created autism are also profoundly retarded, some au- them. Whenever theistic answers are given for tists are left with intact general intellectual fundamental “why” questions, respondents are skills; indeed, in such cases, their IQs frequently essentially exercising their theory of mind. This test well above average. Yet, even these indi- early teleology coincides with the emergent viduals cannot typically complete false-belief metarepresentational capacity children use to tasks in a successful manner, and they have account for behavior in the social domain and is problems in other areas that necessitate having 14 BERING mind-reading abilities (Baron-Cohen, 1995; syndrome woman’s attempt to decipher implicit Frith, 1991; Happe« & Frith, 1995; Leekam & social codes, such as the average distance peo- Perner, 1990; Leslie, 1991, 1994; Swettenham, ple leave between their bodies when waiting in Baron-Cohen, Gomez, & Walsh, 1996; but see line to use an automated teller machine. Frith & Happe«, 1995). Baron-Cohen (1990) ac- She had observed that when people lined up, they left cordingly coined the term “mindblindness” as a a gap between themselves and the person in front, and synonym for autism but is sure to point out that that this gap was substantially larger in the case of men the disorder should be conceptualized as involv- standing behind women. She used this information to ing mindblindness to varying degrees. For in- jump lines, looking for this combination and pushing stance, people with Asperger syndrome, a form in behind the woman nearest the front who was fol- lowed by a man. (p. 382) of very high-functioning autism, might be able to complete second-order false-belief tasks but This woman’s understanding of the way people would have trouble making out the underlying, work was motivated by a desire to learn how foggier shades of intention in dialogue, miss- they typically behaved, not why. Only by as- ing faux pas in others’ speech and violating sessing and becoming extraordinarily sensitive Gricean conversational maxims (Baron-Cohen, to the way routines and conventional social O’Riordan, Stone, Jones, & Plaisted, 1999; rules intersect with people’s manifest behavior Fine, Bartolucci, Szatmari, & Ginsberg, 1994; could she enter the social environment, albeit Happe«, 1993; Landa, 2000). It might be better to inappropriately (she still could not take into conceptualize higher functioning patients, then, consideration how others would perceive her as having never developed a fully erect inten- butting into the line). tional stance. Their sensitivity to intentionality There have been, to my knowledge, no stud- is seemingly attenuated rather than missing ies on religion in autism, nor are there any altogether. related experimental data (e.g., teleologicalÐ Baron-Cohen (2000) and his colleagues (Bar- functional explanations in autistic children) to on-Cohen, Wheelwright, Stone, & Rutherford, shed light on these important questions; thus, it 1999) have recently advanced the compelling is unclear how those with the disorder entertain argument that, although people with autism are philosophicalÐreligious issues. Yet several re- not mentalistic specialists and generally func- cent autobiographical accounts provide intrigu- tion poorly in the social realm, their understanding hints. For instance, Temple Grandin (1995), ing of physical causality might actually be su- a well-known high-functioning autist, wrote in perior. That is, they may not be much in the way several passages of her lifelong struggle with of folk psychologists, but they excel in the her religious beliefs: domain of folk physics. Folk physics, according It is beyond my comprehension to accept anything on to Baron-Cohen (2000, p. 1253), “is our every- faith alone, because of the fact that my thinking is day ability to understand and predict the behav- governed by logic instead of emotion. (p. 189) ior of inanimate objects in terms of principles In high school I came to the conclusion that God was relating to physical causality.” To support this an ordering force that was in everything. I found the idea of superior folk physics in autism, he takes idea of the universe becoming more and more disor- as evidence, among other things, the fact that dered profoundly disturbing. (p. 191) the targets of obsessiveÐcompulsive interests in In nature, particles are entangled with millions of autistic children overwhelmingly cluster around other particles, all interacting with each other. One machines and physical systems. In short, autists could speculate that entanglement of these particles could cause a kind of consciousness for the universe. are preoccupied with the way things work, such This is my current concept of God. (p. 200) as cameras, in terms of how they do so, not in terms of why they do so. Another telling case comes from autistic In some high-functioning cases, this form of mathematician and computer programmer expertise is seemingly translated to problem Edgar Schneider’s (1999) book, Discovering solving in the area of social matters, and it My Autism. In chapters devoted exclusively to affords individuals the capacity to exploit ob- his religious beliefs, Schneider wrote: servable cues emitted by behaving bodies and to My belief in the existence of a supreme intelligence subsequently get by in the real world. Tantum (or, if you will, a God) is based on scientific factors. (2000), for example, wrote of one Asperger (p. 54) EXISTENTIAL THEORY OF MIND 15

It must be pointed out explicitly that none of this minous, or otherness, inherent in normal reli- [religious beliefs] has any emotional underpinnings, gious perception is not there. but is totally intellectual in its nature. (p. 72) This is precisely what we should expect, To me, as far as adherence to a religion (or any other however, from a cognitive phenotype that is type of ideology) is concerned, intellectual conviction is a condition that mathematicians call “both necessary preoccupied with uncovering fundamental prin- and sufficient.” My religious faith, I guess I could say, ciples of physical causality. Because people is not a gift from God, as so many people say; it is a with autistic spectrum disorders have difficulty gift I gave to myself. In line with this, I have never felt interpreting the meaning attached to social be- the emotional exhilaration that people must feel when they have a “religious experience.” This is true even havior and therefore probably cannot rely on a when I receive the sacraments. The only thing that has theory of mind to explain their experiences, deeply moved me is the reasonableness of it all. (p. 73) their religious beliefs cannot affix to core representations of psychological agency. The reli- One gets the distinct impression by reading gious beliefs of people with autism could there- such descriptions that religion in autism con- fore be envisioned as sliding into conceptual sists of entirely different processes than the slots provided by the folk physics system, even normative experience. Namely, in these pas- those, such as supernatural agent concepts, that sages, God, the sine qua non of Western reli- are traditionally relegated to slots in the folk gious experience, is perceived more as a prin- psychology system. Thus, supernatural agents, ciple than as a rich psychological agent. God is such as God, are perceived as behavioral rather perceived as a force in the universe that is than intentional agents. directly responsible for the organization of cos- Unlike people without the disorder, people mic structure—arranging matter in an orderly with autism or Asperger syndrome should be fashion or treating entropy—or has been re- less likely to employ an EToM when confronted duced to the conceptualization of scientific with life-altering events. This is because the logic altogether. To Schneider, instead of expe- events themselves will not be understood as riencing the emotional correlates of church rit- things to be decoded (i.e., to derive meaning). ual that often engender the physiology of “religious” feelings, Catholicism is perceived as an The events will not be perceived as driven by anxiety-reducing medium with its formal, pre- the mind of some intentional agent. Although dictable procedures and the clarity of its canons. their basic life events will largely mirror those Although these views present an extraordi- of their nonautistic peers, their specificdeficit in narily sophisticated theological stance, it rings theory of mind abilities will prevent them from quite differently from the overarching religious attempting to interpret the meaning of those sentiments expressed in William James’s (1902/ experiences. Because autists cannot adequately 1994) classic text, Varieties of Religious Expe- represent psychological agency, they will not rience, or current psychology of religion work experience the EToM pull triggered by particu- (e.g., Hood, Spilka, Hunsberger, & Gorusch, lar perceptual inputs to a cognitively normal 1996). What is noticeably absent in the autistic mind. One man with Asperger syndrome in- accounts is a sense of deep interpersonal rela- quired recently on an Internet bulletin board tions between the worshipper and the deity, a whether others with the disorder were like him, sense of emotional dependency on an inten- “conscious of no feedback from the divine.” tional agent who has control over the experi- This characterizes the basic problems faced by ences and existence of the individual. God is not people with Asperger syndrome: They seek presented as an agent who communicates inten- meaningful relationships with others but have tions through indirect symbolic means or who great difficulty understanding the intentions be- has a distinct constellation of mental states that hind people’s behaviors. Thus, translated to the form a global personality structure. Rather, it is area of religion, a person with Asperger syn- as if the autistic algorithmic strategies used to drome might attempt to engage in ritualistic deal with other people have been translated to activities (e.g., praying for a wife) but will not the authors’ religious beliefs, wherein deductive be able to decode the symbolic device with logic is used to lay the groundwork for under- which a supernatural agent “responds”(e.g., a standing existence and to impose order on an female friend’s husband leaves her for another otherwise chaotic world. The sense of the nu- woman). 16 BERING

In general, if the thoughts presented here ghosts), are the ones that will be maintained and approach some measure of truth, then autistic incorporated into the religious explanatory people should show relatively little interest in system. spiritual matters or should explain existential In a recent commentary on this area of in- issues by means of physical causation, as in the quiry in the cognitive sciences, Sperber and preceding accounts. Those who do adopt insti- Hirschfeld (1999) stated that “religious beliefs tutionalized religious views should do so pri- can be seen as parasitical on domain-specific marily to learn how to engage in acceptable competencies that they both exploit and chal- behavior within the culture. Autistic individuals lenge” (p. cxx). Such a sentiment captures the should therefore be disproportionately repre- flavor of the argument being made: Religious sented in more rule-based, structured religions, ideas, including supernatural agent concepts, especially those that have laid out their social are products of cultural immersion that, after ethical laws in written form and in which ad- being effectively transmitted between minds, herence to such laws is strictly enforced within settle into host modules such as theory of mind. the religious community. There is reason to I am somewhat partial to this sentiment as well. suspect that this is indeed the case (A. Attwood, The trouble, however, is that with regard to personal communication, October 15, 2000). psychological attributions, these authors have Also, as a result of the obsessiveÐcompulsive not taken into account the fact that supernatural interests of autists to rigidly adhere to social agent concepts deal in a domain different from rules once they have been learned, such indi- the one in which they envision their “domain- viduals might run the risk of proselytizing and specific” module of theory of mind to operate. becoming exceedingly moralistic when con- The notion of domain specificity crumbles, and fronted with religious rule breakers. This, too, the very idea that theory of mind is modular seems to be supported by anecdotal reports. suffers a serious blow, when one considers that intentional explanations can be evoked by en- Domain Specificity, Theory of Mind, tirely different classes of input: behavior and and EToM experience. Also, the arguments of these cognitive an- To account for the striking universal presence thropologists overlook the intuitiveness of phil- of religion, numerous authors have recently osophicalÐreligious reasoning. Although ex- speculated that spiritual beliefs and behaviors plicit religious beliefs indeed stem from the hinge upon cognitive mechanisms that evolved cultural milieu and find their way into the minds for other adaptive purposes. Many have even of individuals through seeds planted by social suggested that humans’ theory of mind can ac- learning processes, individuals must already count for the ease with which supernatural agent come bearing with them the representation of concepts are transmitted and represented (J. L. some nondescript psychological agency as Barrett & Keil, 1996; Boyer, 1994; Boyer & causing events. Rather than imbibing from cul- Walker, 2000; Hinde, 1999; Kirkpatrick, 1999). tural sources qualitatively novel ideas about the Beliefs in spirits or gods are envisioned to presence of transcendent agents, culture pro- feed off humans’ understanding of intentional vides labels for perceptual experiences of such agents; these beings are endowed with the same agents that children operating with a theory of general catalogue of mental states (knowledge, mind are already very familiar with. Just as the intentions, desires, beliefs, and so forth) that a perception of the color red is experienced before being with a material brain possesses. Boyer the word red becomes attached to it, perception (1994; Boyer & Walker, 2000) has suggested of this agency is experienced in the human mind that religious ideas are developed within the before, for example, the Western label “God” cultural community and are spread through (or the African label “Gamab” or the Eskimo mechanisms of social transmission, arguing that label “Sila-Pinga”) becomes attached to it. Only the ideas are intercepted by individual minds because the metarepresentational child can re- and squeezed into the appropriate conceptual flectively attend to the class of perceptual expe- slots. Those ideas that are most “attention grab- riences do such causal agent terms denote that bing,” such as the idea of people being dead and religious pedagogy is successful at indoctrinat- invisible but also cognitively viable (e.g., ing new members into belief systems. EXISTENTIAL THEORY OF MIND 17

What I am suggesting is not reification inas- for their current role” (p. 47). The feathers of much as the perception of such agency exists birds, originally selected for thermoregulation universally in humans. Particular agent con- but then co-opted for flight, are often used as an cepts may be reified with cultural elaboration of example of an exaptation. Ironically, in contrast this agency into specific characters, but this to Gould (1991), who cited religion as one of need not concern cognitive scientists, only theo- many functionless, useless by-products of a logians. An individual’s enculturation within large brain and certainly not an exaptation, it the religious community serves to carve up dif- seems quite plausible to me that EToM has been ferent categories of experiences as belonging to co-opted from a broad intentional stance taken particular transcendent agents, with some reli- by our ancestors, the primary adaptation of gious systems (e.g., Islamic monotheism) attrib- which was to explain and predict behavior. In uting nearly all classes of life events to the other words, although it might not have been the handiwork of a single agent and others (e.g., driving force behind mind-reading competen- American Indian polytheism) ascribing certain cies in the species, it nevertheless prospered in types of experiences to a wide variety of “spe- its own right and in its own domain as it came cialty” gods.3 Only after having been effec- to have increasing adaptive utility. tively communicated do explicit ideas regarding The argument is thus: Nearly every culture, if agent concepts work against a default background of folk physics, folk biology, and, of not all, has established religious institutions or course, folk psychology (e.g., J. L. Barrett & (at least) endorses specific religious beliefs that Keil, 1996; Boyer, 1994; Kirkpatrick, 1999; but are often seamlessly woven into the social struc- see J. L. Barrett, Richert, & Driesenga, 2001), ture of the community (Boyer, 1994; Crook, and are probably subject to changes during con- 1995; Hinde, 1999). Religious attitudes govern ceptual development (Elkind, 1979; Woolley, everything from the establishment of marriage 1997). rituals to the methods of slaughtering animals. The extent to which societal mores and religious beliefs are fluid and disparate across cultures Is EToM an Adaptation? has nearly everything to do with the social ecology of the particular culture. Religions inform A question inherent in this article is whether their adherents of what they ought and ought not EToM has its own evolutionary history, do; these determinations are not based arbi- wrought by the forces of natural selection, or trarily on some limitless pool of human imagi- whether its generalization from theory of mind nation but exist because proscribed behaviors implies that it is merely a useless by-product of are maladaptive for the individual in the context the primary adaptation of seeing intention behind behavior. Such questions of evolved traits of the culture. are notoriously tricky (see Buss, Haselton, Among the Sukumu of Tanzania, for in- Shackelford, Bleske, & Wakefield, 1998). Buss stance, religious specialists encourage the kill- et al. (1998, p. 537) described by-products of ing of deformed or retarded offspring, because adaptations as “characteristics that do not solve such children are taken to be evil tokens repre- adaptive problems and do not have functional senting an ancestor’s displeasure (Cory, 1951). design. They are carried along with character- In contrast, many Catholics in the United States istics that do have functional design because view such children as a special gift from God they happen to be coupled with those adapta- meant to test their religious faith (Zuk, Miller, tions.” The whiteness of bones, they claimed, is Bartram, & Kling, 1961). To them, killing in- a case in point, in that the color is a by-product fants is an atrocious act. Why the disparity? of the large amount of calcium that gives bones Reynolds and Tanner (1995) made an excellent their strength, which was what was selected for. point when they stated that “ethical abhorrence However, under certain conditions by-products can come to have their own functions, as 3 Anthropologists would be hard put, however, to find Gould (1991) proposed with his concept of ex- cultures in which there does not exist some grouping of aptations, which are “features that now enhance “second-tier” nonbiologic psychological agents, such as fitness, but were not built by natural selection ghosts, angels, demons, ancestral relatives, and so forth. 18 BERING of infanticide occurs against a background of behavioral natural selection. As such, EToM is, general affluence” (p. 91). That is, the particular I argue, an adapted system, albeit a co-opted opinions of the gods on the matter of infanticide one, rather than a useless by-product of theory depend in large measure on whether parents can of mind. support such children without incurring both In line with this, the classes of experience economic and social burdens that interfere with that employ EToM, then, should have some their reproductive success. Americans are more functional bearing on the preceding. I suspect likely able to shoulder the burden and increase that EToM should be activated to the extent that their social status by doing so, whereas, from a a life event is unexpected and (a) either deviates strictly evolutionary psychological perspective, from or conforms to culturally scripted just it would simply be maladaptive in the case of world expectations and (b) triggers significant people of less affluent societies to allow these affect-related change. The “unexpected” aspect children to live. Therefore, in some hunter-gath- is logistically necessary in that, because the erer societies infanticide can be committed communication between believer and agency without the participants feeling any guilt or exists only in principle, it could not have shame. A Western mother who kills her infant, evolved by means of every social action being in contrast, can expect to feel these emotions met immediately with like consequences in the unimaginably strongly and will have a tough go domain of experience. (This may have, inciden- in finding another mate. In both cases, religion tally, enforced its effects by promoting intermit- inserts itself, but just as the believers speak tent belief confirmation.) To the extent that it different languages, so too do their gods. We are operates in a different domain and can therefore told that good things happen to good people, but be conceived as an independent system boot- goodness, to the evolutionist, is not absolute. strapped to theory of mind, there also is the EToM, then, may function as an adaptive possibility of a sort of existential autism. This mechanism in the sense that it serves to regulate would occur insofar as there are individuals fluid morality in the social ecology. The fact who have clearly articulated theories of mind in that immanent justice and just world biases are the domain of behavior that did not generalize entrenched in the tenets of world religions hints to the domain of life experience. at this regulation (see Gilbert, 1998). Actions that are maladaptive for the individual are often EToM and Atheism those that cause disharmony in the group, such as theft or adultery (Cosmides, 1989). When Through EToM, individuals are able to adopt such actions are labeled “profane” (Wilson, the position that their experiences are intention- 1978) and against the will of the gods, individ- ally guided and are occurring for a reason; it uals become less likely to engage in them, be- would be difficult for them to view themselves cause doing so might be met with inauspicious as “dust in the wind” even if they wanted to. events in their lives, such as sickness or finan- By reflecting back on the salient events of our cial ruin. Thus, individuals’ task in society is own autobiographies—the outcomes of choices largely to determine which behaviors have been made, the unanticipated hazards that derailed us deemed wicked, not so much by their peers from a planned track, ironic encounters with (because they may believe that they can get future mates, the deaths of loved ones—it is away with these things unbeknownst to their exceedingly difficult not to see some intentional peers) but, rather, by their omniscient gods, who propulsion behind their occurrences (for an have the power to mete out experiential reper- interesting discussion, see McAdams, 2001). cussions in response to social transgressions. These are the events that lead to the construc- The system can therefore by viewed as added tion of our current selves and subsequently give enforcement against individuals committing our lives meaning. Underlying all of this, of those actions that might lead to declines in course, is a very basic representation of mind status, castigation, imprisonment, execution, behind messages perceived to be intentionally and other sundry social punishments that cut communicated. into their reproductive success. By working in Such sensitivity to external meaning may, to the deontic, EToM may have therefore enabled some limited extent, be forcefully unattended to the explication of the implicit forces governing in adulthood, but the forceful shedding of EXISTENTIAL THEORY OF MIND 19

EToM will probably lead to a period of intense their sincerity on the matter; however, much as psychological suffering and the pangs of exis- solipsists cannot switch off theory of mind tential crises (Frankl, 1963; May, 1960; Yalom, mechanisms, neither can EToM be totally dis- 1980). People must adapt a meaning-based ex- engaged in response to our acquired views on planatory system that is wholly reliant upon religion. Indeed, experimental data (Weeks & their own attribution of purpose to their expe- Lupfer, 2000) reveal that individuals professing riences. William James (1902/1994) eloquently to be nonreligious are just as likely to make described individuals’ initial perception of a intentional attributions in the domain of expe- meaningless existence: “The world now looks rience as religious people are. The only differ- remote, strange, sinister, uncanny. Its color is ence is that whereas believers turn to God in gone, its breath is cold, there is no speculation explaining ironic or life-altering events, nonbe- in the eyes it glares with” (p. 170). No longer is lievers turn to a belief repertoire including such there a transcendent explanation behind natural elements as fate and immanent justice. In addi- events. When this agency is killed off, people tion, these nonreligious immaterial belief struc- perceive the same phenomena as they did as EToM users, but they can no longer decode the tures are found cross culturally (Pepitone, intentional messages attached to their experi- 1997). ences. What contributes mostly to their dyspho- An analogy may be helpful. Much as people ria is the loss of ultimate meaning that their life the world over have different religious beliefs, once had. They encounter the same type of so too do they have different stylistic prefer- dilemma faced by solipsists in their questioning ences, and thus they decorate their homes with of the reality of other human minds. If no other different styles of furniture. Some prefer mod- human minds exist, how can the behaviors of ern design, others like Victorian furniture only, others have any meaning attached? Likewise, if and still others are eclectic and prefer a wide there is no mind behind our existence, so to mixture of different furniture styles. When it speak, then how can we find purpose or mean- comes time to sit or sleep, or eat or study, the ing in the things that happen to us and, more occupants of each home use the furniture that generally, in our lives? they have chosen to perform these functions. A Certainly, there are individuals (not least Victorian chair and a modern chair might look among them scientists!) who claim not to sub- very different next to each other, but, in the end, scribe to such seemingly romantic workings of they are both expert at doing what they were the mind, but it is my impression that we would designed to do. be hard pressed to discover an individual of What about prisoners living in a barren prison normal cognitive functioning who has never cell, however? When it comes time for them to exercised his or her EToM. Imagine the follow- sit down, as it surely will, they will do so. But ing: You are on a crowded bus, lost in the they will not sit down on anything in particular. newspaper before you, when suddenly you are Nevertheless, the constraints on their body to sit caught in a dizzying fury of screams, blackness, will override the fact that they have no special- and crushing metal. Your bus has crashed and ized furniture to accommodate their needs. Sim- flipped over a steep embankment. You crawl ilarly, individuals who have not been exposed to out a window, dust yourself off, and realize that you are the sole survivor out of dozens of other (or have not accepted) religious beliefs or stored passengers. If a week from now, or a year or these beliefs for later retrieval face the same decade later, you find yourself asking “Why cognitive constraints as those who have and will me?” then quite simply you have an intact probably still perceive psychological agency as EToM. Even if you brush such questions aside driving their experiences in some generic fash- because you consider them rather foolish, you ion. Simply because these individuals do not still betray your EToM insofar as you can en- have specific religious beliefs to accommodate a tertain this type of question in the current core causal attribution style does little to wear context. away the core. The human brain is designed to To reiterate: We are dealing in the domain of see meaning behind random events, even under experience, not behavior. Individuals may very the wary, conscious eye of an intellect that well be ideological atheists, and I do not doubt thinks it knows better. 20 BERING

Concluding Remarks remaining confusion. If we are given four cases of mothers—a language-trained chimpanzee Only by examining alternative ways of per- mother, an autistic mother, a religious mother, ceiving behavior, such as in other species or in and a nonreligious mother—and told that each autism, are we able to recognize how the human of them has just sadly lost her infant as a result cognitive system was specially adapted to rep- of a disease, we might expect the following resent mental agency as driving action. At some responses (or something similar) after asking point in evolutionary history after the human them why the death occurred. The chimpanzee lineage split from the African apes, I have ar- would grieve for the loss but otherwise look at gued, the intentionality framework expanded to us blankly; the autistic mother would speculate include those ambient life experiences that hu- that cancerous lesions had gotten a stronghold mans had little or no control over. Just as it is on her baby’s immunosuppressive system; the almost impossible to predict what another per- religious mother would tell us that it was the son is going to do or to be certain of why will of God; and the nonreligious mother would someone has just behaved in a particular way, it tell us that her baby died so that she could help is similarly impossible to predict what nature other bereaved mothers. has in store for us or to understand why we have Whether this rings true for particular readers just experienced something that may alter the will rest, I believe, on whether or not they have course of our lives. EToM functions as a philo- found the general arguments presented in this sophicalÐreligious explanatory system that al- article to be convincing. But even if they find lows us to see meaning in some of the things these ideas still wanting, I would hope that the that happen to us, affords us some sense of questions raised here continue to be asked and perceived psychological control over what is that evolutionary cognitive science opens its likely to happen, enforces cultural mores that doors wider for such inquiry, because these adapt the individual to the group, and guards issues truly strike at the heart of what it means against those behaviors that are maladaptive. to be human. Above all else, we are a species It is perhaps wise, given the subject matter of whose members are ravenous to understand the this article, to remember that the naturalistic drama of our own lives. fallacy is a double-edged sword. Just as atheism may not exist except in principle, the gods may References not either. My intention has merely been to show that life events are often perceived to be Atran, S. (1995). Causal constraints of categories. In products of mental causality. EToM is a co- D. Sperber, D. Premack, & A. J. Premack (Eds.), opted by-product of natural selection and occurs Causal cognition: A multi-disciplinary debate (pp. inside our heads. Comfortable in my scientist 263Ð265). Oxford, England: Clarendon Press. skin, I will remain mute as to whether I believe Baldwin, D. A., & Moses, L. J. (1994). Early under- such perceptions are philosophically justified. standing of referential intent and attentional focus: Nevertheless, such a large theory as the one I Evidence from language and emotion. In C. Lewis & P. Mitchell (Eds.), Children’s early understand- am offering demands a large bed of data to ing of mind: Origins and development (pp. 133Ð support it, and I have done little but try to work 156). Hillsdale, NJ: Erlbaum. in experimental evidence and case reports that, Baron-Cohen, S. (1990). Autism: A specific cogni- more often than not, only indirectly corroborate tive disorder of “mindblindness.” International my claims. Fortunately, the hypotheses embed- Review of Psychiatry, 2,79Ð88. ded in this article are fully testable, and deter- Baron-Cohen, S. (1995). Mindblindness: An essay on mining the merit of these arguments should not autism and theory of mind. Cambridge, MA: MIT be impossible. For instance, whether the reli- Press. gious experiences of people with autism are Baron-Cohen, S. (2000). The cognitive neuroscience qualitatively different from those of people of autism: Evolutionary approaches. In M. Gazza- niga (Ed.), The new cognitive neurosciences (2nd without the disorder is an empirical question ed., pp. 1249Ð1257). Cambridge, MA: MIT Press. open for exploration. Baron-Cohen, S., O’Riordan, M., Stone, V., Jones, I have only now to summarize some of the R., & Plaisted, K. (1999). Recognition of faux pas points addressed in this article with a hypothet- by normally developing children with Asperger ical example that should, I hope, clarify any syndrome or high-functioning autism. Journal of EXISTENTIAL THEORY OF MIND 21

Autism and Developmental Disorders, 29, 407Ð tions, exaptations, and spandrels. American Psy- 418. chologist, 53, 533Ð548. Baron-Cohen, S., Wheelwright, S., Stone, V., & Butler, S. C. (1999). Early precursors to theory of Rutherford, M. (1999). A mathematician, a phys- mind: Infant understanding of the referential na- icist, and a computer scientist with Asperger syn- ture of visual perception (comprehension). Disser- drome: Performance on folk psychology and folk tation Abstracts International, 60(4-B), 1884. physics tests. Neurocase, 5, 475Ð483. Byrne, R. W., & Whiten, A. (1988). Machiavellian Barrett, H. C. (1999). Human cognitive adaptations intelligence: Social expertise and the evolution of to predators and prey. Unpublished doctoral dis- intellect in monkeys, apes, and humans. New sertation, Department of Anthropology, University York: Oxford University Press. of California, Santa Barbara. Call, J., & Tomasello, M. (1999). A nonverbal false Barrett, J. L., & Keil, F. C. (1996). Anthropomor- belief task: The performance of children and great phism and God concepts: Conceptualizing a non- apes. Child Development, 70, 381Ð395. natural entity. Cognitive Psychology, 3, 219Ð247. Carey, S. (1985). Conceptual change in childhood. Barrett, J. L., Richert, R. A., & Driesenga, A. (2001). Cambridge, MA: MIT Press. God’s beliefs versus mother’s: The development Carey, S. (1995). The growth of understandings of of non-human agent concepts. Child Development, natural kinds. In D. Sperber, D. Premack, & A. J. 72,50Ð65. Premack (Eds.), Causal cognition: A multi-disci- Bartsch, K., & London, K. (2000). Children’s use of plinary debate (pp. 268Ð302). Oxford, England: mental state information in selecting persuasive Clarendon Press. arguments. Developmental Psychology, 36, 352Ð Cheney, D. L., & Seyfarth, R. M. (1990). How mon- 365. keys see the world. Chicago: University of Chi- Bates, E. (1976). Language in context: The acquisi- cago Press. tion of pragmatics. New York: Academic Press. Clark, A. (1994). Beliefs and Desires Incorporated. Bering, J. M. (2001). Theistic percepts in other spe- Journal of Philosophy, 91, 404Ð425. cies: Can chimpanzees represent the minds of non- Clements, W. A., & Perner, J. (1994). Implicit un- natural agents? Journal of Cognition and Cul- derstanding of belief. Cognitive Development, 9, ture, 1, 107Ð137. 377Ð395. Bering, J. M., & Povinelli, D. J. (in press). Compar- Cory, H. (1951). The Ntemi: Traditional rites of a ing cognitive development: From bridging gaps to Sukuma chief in Tanganyika. London: Macmillan. accepting diversity. In D. Maestripieri (Ed.), Pri- Cosmides, L. (1989). The logic of social exchange: mate psychology: Bridging the gap between the Has natural selection shaped how humans reason? mind and behavior of human and nonhuman pri- Cognition, 31, 187Ð276. mates. Cambridge, MA: Harvard University Press. Cosmides, L., & Tooby, J. (1992). Cognitive adap- Boesch, C. (1991). Teaching among wild chimpan- tations for social exchange. In J. Barkow, L. Cos- zees. Animal Behaviour, 41, 530Ð532. mides, & J. Tooby (Eds.), The adapted mind: Bovet, P. (1928). The child’s religion: A study of the Evolutionary psychology and the generation of development of the religious sentiments. London: culture (pp. 163Ð228). New York: Oxford Univer- Dent. sity Press. Boyer, P. (1994). The naturalness of religious ideas: Crook, J. H. (1995). Psychological processes in cul- A cognitive theory of religion. Berkeley: Univer- tural and genetic evolution. In E. Jones & V. sity of California Press. Reynolds (Eds.), Survival and religion: Biological Boyer, P., & Walker, S. (2000). Intuitive ontology evolution and cultural change (pp. 45Ð110). and cultural input in the acquisition of religious Chichester, England: Wiley. concepts. In K. S. Rosengran, C. N. Johnson, & Donald, M. (1991). Origins of the modern mind. P. L. Harris (Eds.), Imagining the impossible: Cambridge, MA: Harvard University Press. Magical, scientiﬁc and religious thinking in chil- Elkind, D. (1979). The child and society. New York: dren (pp. 130Ð156). New York: Cambridge Uni- Oxford University Press. versity Press. Evans, E. M. (2000). The emergence of beliefs about Boysen, S. T. (1993). Counting in chimpanzees: the origins of species in elementary school chil- Nonhuman principles and emergent properties of dren. Merrill-Palmer Quarterly, 46,2. number. In S. T. Boysen & E. J. Capaldi (Eds.), Fine, J., Bartolucci, G., Szatmari, P., & Ginsberg, G. The development of numerical competence: Ani- (1994). Cohesive discourse in pervasive develop- mal and human models (pp. 39Ð59). Hillsdale, NJ: mental disorders. Journal of Autism and Develop- Erlbaum. mental Disorders, 14, 315Ð329. Buss, D. M., Haselton, M. G., Shackelford, T. K., Flavell, J. H. (1992). Perspective on perspective tak- Bleske, A. L., & Wakeﬁeld, J. C. (1998). Adapta- ing. In H. Beilin & P. Pufall (Eds.), Piaget’s the- 22 BERING

ory: Prospects and possibilities (pp. 107Ð139). Hauser, M. (1998). Expectations about object motion Hillsdale, NJ: Erlbaum. and destination: Experiments with a nonhuman Flavell, J. H. (1999). Cognitive development: Chil- primate. Developmental Science, 1,31Ð38. dren’s knowledge about the mind. Annual Review Heyes, C. (1998). Theory of mind in nonhuman pri- of Psychology, 50,21Ð45. mates. Behavioral and Brain Sciences, 21, 101Ð Frankl, V. E. (1963). Man’s search for meaning. New 134. York: Simon & Schuster. Hinde, R. A. (1999). Why gods persist: A scientific Frith, U. (1991). Autism and Asperger’s syndrome. approach to religion. London: Routledge. Cambridge, England: Cambridge University Press. Hood, R. W., Spilka, B., Hunsberger, B., & Gorusch, Frith, U., & Happe«, F. (1995). Autism: Beyond “the- R. (1996). The psychology of religion. New York: ory of mind.” In J. Mehler & S. Franck (Eds.), Guilford Press. Cognition on cognition (pp. 13Ð30). Cambridge, James, W. (1892). Thought before language: A deaf MA: MIT Press. mute’s recollections. Philosophical Review, 1, 613Ð624. Gelman, S. A., & Kremer, K. E. (1991). Understand- James, W. (1994). The varieties of religious experi- ing natural cause: Children’s explanations of how ence. New York: Random House. (Original work objects and their properties originate. Child Devel- published 1902) opment, 62, 396Ð414. Karin-D’Arcy, M. R., & Provinelli, D. J. (2001). Do Gergely, G., Na«dasdy, Z., Csibra, G., & Biro«,S. chimpanzees know what each other see? A closer (1995). Taking the intentional stance at 12 months look. Manuscript submitted for publication. of age. Cognition, 56, 165Ð193. Karmiloff-Smith, A. (1992). Beyond modularity: A Gilbert, P. (1998). Suffering, evolution, and psycho- developmental perspective on cognitive science. therapy. In C. Feltham (Ed.), Witness and vision of Cambridge, MA: MIT Press. the therapists (pp. 94Ð122). Thousand Oaks, CA: Keil, F. C. (1989). Concepts, kinds, and cognitive Sage. development. Cambridge, MA: MIT Press. Golinkoff, R. M., Harding, C. G., Carlson-Luden, V., Kelemen, D. (1999a). Functions, goals, and inten- & Sexton, M. E. (1984). The infant’s perception of tion: Children’s teleological reasoning about ob- causal events: The distinction between animate jects. Trends in Cognitive Sciences, 3, 461Ð468. and inanimate objects. In L. P. Lipsitt (Ed.), Ad- Kelemen, D. (1999b). Why are rocks pointy? Chil- vances in infancy research (Vol. 3, pp. 145Ð151). dren’s preferences for teleological explanations of Norwood, NJ: Ablex. the natural world. Developmental Psychology, 35, Goodall, J. (1975). The chimpanzee. In V. Goodall 1440Ð1452. (Ed.), The quest for man (pp. 131Ð170). New Kirkpatrick, L. (1999). Toward an evolutionary psy- York: Praeger. chology of religion and personality. Journal of Gould, S. J. (1991). Exaptation: A crucial tool for Personality, 67, 921Ð952. evolutionary psychology. Journal of Social Issues, Kita, S. (in press). Pointing: Where language, cul- 47,43Ð65. ture, and cognition meet. Hillsdale, NJ: Erlbaum. Grandin, T. (1995). Thinking in pictures: And other Kuhlmeier, V. A., Boysen, S. T., & Mukobi, reports from my life with autism. New York: K. L. R. P. (1999). Scale-model comprehension by Doubleday. chimpanzees (Pan troglodytes). Journal of Com- Grice, H. P. (1957). Meaning. Philosophical Review, parative Psychology, 113, 396Ð402. Landa, R. (2000). Social language use in Asperger 64, 377Ð388. syndrome and high-functioning autism. In A. Klin, Guthrie, S. (1993). Faces in the clouds: A new theory F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger of religion. New York: Oxford University Press. syndrome (pp. 129Ð155). New York: Guilford Happe«, F. (1993). Communicative competence and Press. theory of mind in autism: A test of relevance Lawson, E. T., & McCauley, R. N. (1990). Rethink- theory. Cognition, 48, 101Ð119. ing religion: Connecting religion and culture. Happe«, F., & Frith, U. (1995). Theory of mind in Cambridge, England: Cambridge University Press. autism. In E. Schopler & G. B. Mesibov (Eds.), Leekam, S., & Perner, J. (1990). Does the autistic Learning and cognition in autism (pp. 177Ð197). child have a metarepresentational deficit? Cogni- New York: Plenum. tion, 40, 203Ð218. Hare, B. (2001). Can competitive paradigms increase Leslie, A. (1991). The theory of mind impairment in the validity of experiments on primate social cog- autism: Evidence for a modular mechanism of nition? Animal Cognition, 4, 269Ð280. development. In A. Whiten (Ed.), Natural theories Hare, B., Call, J., Agnetta, B., & Tomasello, M. of mind: Evolution, development and simulation of (2000). Chimpanzees know what conspecifics do everyday mindreading (pp. 63Ð78). Oxford, En- and do not see. Animal Behaviour, 59, 771Ð785. gland: Blackwell. EXISTENTIAL THEORY OF MIND 23

Leslie, A. (1994). ToMM, ToBY, and agency: Core the Society for Research in Child Development, architecture and domain specificity. In L. A. Hir- 61,3. schfeld & S. Gelman (Eds.), Mapping the mind: Povinelli, D. J., Perriloux, H. K., Reaux, J. E., & Domain specificity in cognition and culture (pp. Bierschwale, D. T. (1998). Young and juvenile 119Ð148). Cambridge, England: Cambridge Uni- chimpanzees’ (Pan troglodytes) reactions to inten- versity Press. tional versus accidental and inadvertent actions. Leslie, A. M. (1982). The perception of causality in Behavioural Processes, 42, 205Ð218. infants. Perception, 11, 173Ð186. Povinelli, D. J., Reaux, J. E., Bierschwale, D. T., Leslie, A. M. (1987). Pretense and representation: Allain, A. D., & Simon, B. B. (1997). Exploitation The origins of “theory of mind.” Psychological of pointing as a referential gesture in young chil- Review, 94, 412Ð426. dren, but not adolescent chimpanzees. Cognitive Maser, J. D., & Gallup, G. G. (1990). Theism as a Development, 12, 423Ð461. by-product of natural selection. Journal of Reli- Premack, D., & Premack, A. J. (1997). Infants at- gion, 90, 515Ð532. tribute value to the goal-directed actions of self- May, R. (1960). Existential psychology. New York: propelled objects. Journal of Cognitive Neuro- Random House. science, 9, 848Ð856. McAdams, D. P. (2001). The psychology of life Repacholi, B. M., & Gopnik, A. (1997). Early rea- stories. Review of General Psychology, 5, 100Ð soning about desires: Evidence from 14- and 18- 122. month-olds. Developmental Psychology, 33, 12Ð Meltzoff, A. N. (1995). Understanding the intentions 21. of others: Re-enactment of intended acts by 18- Reynolds, V., & Tanner, R. (1995). The social ecol- month-old children. Developmental Psychology, ogy of religion. New York: Oxford University 31, 838Ð850. Press. Mitchell, P. (1996). Introduction to theory of mind: Savage-Rumbaugh, E. S. (1986). Ape language: Children, autism, and apes. Oxford, England: Ox- From conditioned response to symbol. New York: Columbia University Press. ford University Press. Schneider, E. (1999). Discovering my autism: Apo- Mitchell, R., & Thompson, N. (1986). Deception: logia pro vita sua (with apologies to Cardinal Perspective on human and nonhuman deceit. Al- Newman). London: Jessica Kingsley. bany: State University of New York Press. Slaughter, V., Jaakola, K., & Carey, S. (1999). Con- Mithen, S. (1996). The prehistory of the mind: The structing a coherent theory: Children’s biological cognitive origins of art, religion and science. Lon- understanding of life and death. In M. Siegal & C. don: Thames & Hudson. Peterson (Eds.), Children’s understanding of biol- Moore, C., & Dunham, P. J. (1995). Joint attention: ogy, health, and ethics (pp. 71Ð96). Cambridge, Its origins and role in development. Hillsdale, NJ: England: Cambridge University Press. Erlbaum. Sperber, D. (1994). The modularity of thought and Pepitone, A. (1997). Nonmaterial beliefs: Theory and the epidemiology of representations. In L. Hirsch- research in cultural social psychology. In C. Mc- feld & S. Gelman (Eds.), Mapping the mind: Do- Garty and S. A. Haslam (Eds.), The message of main specificity in cognition and culture (pp. 39Ð social psychology: Perspectives on mind in society 62). New York: Cambridge University Press. (pp. 252Ð267). Malden, MA: Blackwell. Sperber, D. (1996). Explaining culture: A naturalis- Perner, J. (1991). Understanding the representa- tic approach. Oxford, England: Blackwell. tional mind. Cambridge, MA: MIT Press. Sperber, D., & Hirschfeld, L. (1999). Culture, cog- Piaget, J. (1929). The child’s conception of the world. nition, and evolution. In R. A. Wilson & F. C. Keil London: Routledge & Kegan Paul. (Eds.), The MIT encyclopedia of the cognitive sci- Pinker, S. (1994). The language instinct. New York: ences (pp. cxiÐcxxxii). Cambridge, MA: MIT HarperCollins. Press. Povinelli, D. J., Bering, J., & Giambrone, S. (2000). Suddendorf, T., & Whiten, A. (2001). Mental evolu- Toward a science of other minds: Escaping the tion and development: Evidence for secondary rep- argument by analogy. Cognitive Science, 24, 509Ð resentation in children, great apes, and other ani- 541. mals. Psychological Bulletin, 127, 629Ð650. Povinelli, D. J., Bering, J., & Giambrone, S. (in Swettenham, J., Baron-Cohen, S., Gomez, J. C., & press). Chimpanzee ‘pointing’: Another error of Walsh, S. (1996). What’s inside a person’s head? the argument by analogy? In S. Kita (Ed.), Point- Conceiving of the mind as a camera helps children ing: Where language, culture, and cognition meet. with autism develop an alternative theory of mind. Hillsdale, NJ: Erlbaum. Cognitive Neuropsychiatry, 1,73Ð88. Povinelli, D. J., & Eddy, T. J. (1996). What young Tantum, D. (2000). Adolescence and adulthood of chimpanzees know about seeing. Monographs of individuals with Asperger syndrome. In A. Klin, 24 BERING

F. R. Volkmar, & S. S. Sparrow (Eds.), Asperger M. Killen (Eds.), Piaget, evolution, and develop- syndrome (pp. 367Ð399). New York: Guilford ment (pp. 73Ð99). Mahwah, NJ: Erlbaum. Press. Whiten, A., Goodall, J., McGrew, W. C., Nishida, T., Tomasello, M. (1999). The cultural origins of human Reynolds, V., Sugiyama, Y., et al. (1999). Cultures cognition. Cambridge, MA: Harvard University in chimpanzees. Nature, 399, 682Ð685. Press. Wilson, E. O. (1978). On human nature. Cambridge, Tomasello, M., & Call, J. (1997). Primate cognition. MA: Harvard University Press. New York: Oxford University Press. Woolley, J. D. (1997). Thinking about fantasy: Are Tomasello, M., Call, J., & Gluckman, A. (1997). The children fundamentally different thinkers and be- comprehension of novel communicative signs by lievers from adults? Child Development, 68, 991Ð apes and human children. Child Development, 68, 1011. 1067Ð1081. Woolley, J. D. (2000). The development of beliefs Tomasello, M., Gust, D., & Frost, G. T. (1989). The about direct mental-physical causality in imagina- development of gestural communication in young tion, magic, and religion. In K. S. Rosengren, C. N. chimpanzees: A follow up. Primates, 30,35Ð50. Johnson, & P. Harris (Eds.), Imagining the impos- Tomasello, M., Hare, B., & Agnetta, B. (1999). sible: Magical, scientific, and religious thinking in Chimpanzees (Pan troglodytes) follow gaze direc- children (pp. 99Ð129). Cambridge, England: Cam- tion geometrically. Animal Behaviour, 58, 769Ð 777. bridge University Press. Tomasello, M., Kruger, A. C., & Ratner, H. H. Yalom, J. (1980). Existential psychotherapy. New (1993). Cultural learning. Behavioral and Brain York: Basic Books. Sciences, 16, 495Ð552. Zihlman, A. (1996). Reconstructions reconsidered: Weeks, M., & Lupfer, M. B. (2000). Religious attri- Chimpanzee models and human evolution. In butions and proximity of influence: An investiga- W. C. McGrew, L. F. Marchant, & T. Nishida tion of direct interventions and distal explanations. (Eds.), Great ape societies (pp. 293Ð304). Cam- Journal for the Scientific Study of Religion, 39, bridge, England: Cambridge University Press. 348Ð362. Zuk, G. H., Miller, R. L., Bartram, J. B., & Kling, F. Wellman, H. M. (1990). The child’s theory of mind. (1961). Maternal acceptance of retarded children: Cambridge, MA: MIT Press. A questionnaire study of attitudes and religious Wellman, H. M., & Gelman, S. A. (1992). Cognitive background. Child Development, 32, 525Ð540. development: Foundational theories of core domains. Annual Review of Psychology, 43, 337Ð 375. Received March 7, 2001 Whiten, A. (1998). Evolutionary and developmental Revision received July 9, 2001 origins of the mindreading system. In J. Langer & Accepted July 10, 2001 Ⅲ