Introduction

One of the hallmarks of intentional agents is the fact that natural selection has rubber-stamped their causal systems with an approximate expiration date. Miracles and medical caveats aside, brains become irrevocably defunct at biological death. Dying therefore makes intentional causality special because mental states, unlike other causal states, are absolutely absent in dead organisms. Consider the death of a human being. A carcass might tumble out of its coffin (thus the body retains physical causality); or enzymes from a corpse might seep into the soil of its grave (thus the body retains biological causality). But alas, the body has entirely lost its capacity for intentional causality. This is because the brain —  that “thing” which in better days caused this body to stand up, to dance, to cry, to laugh — abruptly halts its production of mental states whenever it expires or exceeds its tolerance for assault.

All of this is bound to strike a scientific audience as painfully obvious. Indeed, one need not have extensive training in cognitive neuroscience to recognize death’s impact on the mind; anybody who has thought about the topic with any empirical gusto is likely to find themselves adopting a materialist stance. But this “Duh” effect is precisely why the study of death should be of interest to cognitive scientists. If it is so obvious that mental states are obliterated by death, then why do people the world over believe and behave otherwise?

Let me not prematurely raise hopes: We don’t really know yet. But as I discuss in this article, there’s reason to be optimistic that cognitive science holds the key to answering this question. Of course, cognitive scientists are not the first to take an interest in this topic; we are in fact nipping at the heels of over a half century’s worth of psychology that has done the emotional bricklaying. From Freud’s more vacuous concept of “wish fulfillment” to Ernest Becker’s inspired “Terror Management Theory” and its huge catalogue of elegant experiments on death anxiety (e.g., Dechesne et al., 2003), psychologists in this area historically have focused almost exclusively on fear of death. Central to these models is the notion that human beings have ‘invented’ the idea of an immortal soul to allay their concerns about nonexistence. (Below I review recent evidence that calls into question this popular model of afterlife beliefs.) The reason that cognitive science is now nipping at these heels is to help orient psychologists to the fact that any human emotions that play a role in reasoning about the mind’s fate after death are dependent on the more basic capacity to think about minds.

A Lacuna of Comparative Investigations

One might therefore begin by asking whether other species reason about death in a fashion similar to human beings. Miguel de Unamuno, the Spanish philosopher who penned the notoriously brooding The Tragic Sense of Life, suggested that “The gorilla, the chimpanzee, the orangutan, and their kind, must look upon man as a feeble and infirm animal, whose strange custom it is to store up his dead. Wherefore?” (1926, p. 20; italics added). From an evolutionary perspective, this is an important question because it bears on the phylogeny of mental representational systems. If the soul is, in essence, the disembodied mind, then the capacity to reason about this mythical entity is quintessentially rooted in evolved social cognition. Comparative studies that investigate other species’ reasoning about death would therefore be fruitful if they helped to map out the evolutionary origins of cognitive traits such as theory of mind.

Unfortunately, to my knowledge no comparative psychologist has taken on this specific task, so we have only anecdotal reports to frame our current understanding. When reviewing these cases, we are wise to be wary of the human penchant for seeing our own minds mirrored in the behaviors of other species. E. O. Wilson (1971) offers a good cautionary tale about the hazards of extravagant theorizing in this area (read the many popular science accounts of elephants’ fondling their grandmothers’ bones and great apes mourning themselves sick). He once described how ants (Pogonomyrmex barbatus) routinely inspected with their antennae day old corpses of nest mates that had been decomposing in the open air. The first ant to find the corpse picked it up and carried it off to the refuse pile, where it was safely removed from healthy sister workers. Do ants therefore understand death as death? Probably not. As Wilsonobserved, it is just natural selection’s way of keeping inclusive fitness levels at a healthy high. A closer look revealed that these predictable behavioral responses were motivated by biochemical cues; bits of paper daubed with acetone extracts of ant corpses were treated by the workers in an identical fashion — they too were carried off to the refuse pile!

Some anecdotes are relevant to the current discussion, however, because they detail natural behavior patterns that show a reliance on causal agency cues in distinguishing between a dead organism and a live one. Both de Waal (1996) and Goodall (1990), for instance, have described the behaviors of chimpanzees (Pan troglodytes) in response to the deaths of group members as involving (i) close inspection of the corpse; (ii) attenuated social arousal and (iii) momentary group cohesion (see also Teleki, 1973). Miller and Brigham (1998, p. 79) even describe an incidence of “ceremonial gathering” by black-billed magpies (Pica pica) that occurred in response to the mid-flight death of a solitary bird: “Within 5 minutes a flock of 13-14 magpies had gathered in a circle around the body. . .There was no overt sign of aggression nor did the birds make any attempt to scavenge from the carcass. . .The remaining birds stayed around the carcass for at least 5 minutes before spontaneously flying off together.” Barrett and Behne (2005) have recently identified agency detection mechanisms as central to human reasoning about death as well. (To illustrate, simply imagine a runny-nosed second grader using a stick to poke at a dead raccoon on the road.) Corpses are one of the great ambiguities of nature. They place unusual strain on inference systems devoted to reasoning about hidden causes because a body’s absence of action does not necessarily entail an absence of intentional states.

In one of the few serious theoretical papers in this area, Allen and Hauser (1991, p. 231) suggest that our species may be qualitatively unique because we can mentally represent the concept of death. They argue that: “Humans are capable of recognizing something as dead because they have an internal representation of death that is distinct from the perceptual information that is used for evidence of death. It is this separate representation that is capable of explaining the human ability to reason about death rather than merely respond to death in the environment.” Following this logic, the authors proceed to describe several possible experiments that would test the hypothesis that our species is alone in harboring this internal representation. For example, they suggest that researchers might observe the behaviors of female vervet monkeys who are listening to vocalization playbacks of distress calls from their recently dead infants. Presumably, if the mother ignored her dead infant’s distress call, this would be evidence that she has “turned off” her localization and search response because her internal representation allows her to appreciate the finality of death. There may be better — and less cruel! — ways to test such hypotheses, but it is indeed critical to put the many comparative anecdotes on death through the sieve of the scientific method, just as these authors advise.

Autistic Souls?

Because their impoverishments are localized to social cognitive dysfunctions, investigating autistic people’s death concepts may also prove informative. Do — or, more properly, can — autists believe in souls? Unfortunately, as is the case for comparative research on intuitive reasoning about death, no controlled studies have yet been conducted with autistic people. There are a handful of anecdotal reports to pique our curiosity, however. In her autobiography Thinking in Pictures: And Other Reports from My Life with Autism, TempleGrandinsoliloquizes on immortality after describing her invention of a humane slaughtering apparatus for cattle. Indeed, she devotes her final chapter to the subjects of Heaven and God. Of particular interest is a diary entry which blandly reads: “I believe that a person goes on to somewhere else after they die. I do not know where” (1995, p. 197).

Evolutionary Hypotheses

In light of the question I asked at the outset of this paper (“If it is so obvious that mental states are obliterated by death, then why do people the world over believe and behave otherwise?”), it seems reasonable to postulate possible genetic advantages associated with this ubiquitous ‘irrationality.’ In both hunter-gatherer and modern societies, the fear of ghosts abounds (e.g., see Reynolds & Tanner, 1995). In children, this fear rivals such evolutionarily plausible fears as those of snakes and spiders, and it is apparently even more resistant to treatment than fear of strangers (Gullone et al., 2000). Recently, Katrina McCleod, Todd Shackelford and I hypothesized that the fear of ghosts may facilitate the inhibition of selfish behaviors, thus preserving reputation in situations where individuals underestimated the risk of detection by living group members (Bering, McLeod, & Shackelford, in press; see also Boyer, 2001). Sometimes, of course, it pays to cheat, but in general the costs of underestimating the risk of social detection would have been disproportionately greater than the costs of prosocial decisions that were contextually maladaptive.

Our study was partially inspired by another recent study by Burnham and Hare (in press), who reported that, in anonymous and final interactions, participants contributed significantly more to a public good when ‘watched’ by a robot with large, human-like eyes (see also Haley & Fessler, 2005). Although their hypothesis was that human eyes would trigger non-conscious mechanisms that gauge privacy, and thus serve to elicit prosocial behaviors, we suspected that the presumed presence of a ghost in the room may similarly prime cooperative effort.

Our experimental design presented undergraduate students with several opportunities to cheat at a challenging spatial intelligence test, which promised the highest scorer a generous monetary prize. On a random sample of mental rotation items, the correct answer was “accidentally” revealed on the screen prior to the question. In both the written and verbal instructions, participants had been informed earlier that, due to a glitch in the computer program, this would periodically happen. If it did, they were told, they should immediately press the space bar so that they could “answer the question honestly.” The catch was that a third of these people had also been told earlier that the ghost of a dead graduate student had recently been spotted in the laboratory. Another group simply read an In Memoriam announcement in honor of this dead student, while the remainder heard nothing about the fictitious decedent. As predicted, there was a significant effect of condition on latency of response (i.e., how quickly the participants pressed the space bar to clear the screen of the correct answer) on glitch items. Participants who were randomly assigned to the ghost condition had faster response latencies than those in the control group and those in the In Memoriam group.

‘Signs’

This inhibitory effect hypothesis of ghost primes has additional adaptationist veracity when considering that many people are under the impression that the dead are arbiters of fortune; they are often seen as meting out punishment for social transgressions. In a recent review, Dominic Johnson and I poured through the Human Relations Area Files (a vast archive of cultural ethnographies that includes many societies which have since gone extinct) for evidence of the cross-cultural tendency to ascribe to the dead responsibility for natural events (Bering & Johnson, 2005). To say that this is a culturally recurrent phenomenon would be an understatement. Thus, not only do many people believe in ambient spirits, they also believe that these spirits are causal agents in the environment, intentional agents that use natural events as signs to communicate messages (often symbolizing their discontent).

Given the right emotional construal, I’d wager that everybody — materialist and dualist alike — is susceptible to knee-jerk attributions of intentions to disembodied minds, whether they profess to ‘believe’ in such things or not. For instance, I am an unswerving atheist and I fully ‘believe’ that my mind will burn out in harmony with the death rattle of my brain cells. Furthermore, I confess that on more than one occasion I have shivered with shock and repugnance at the dogged naiveté of the faithful. And yet — and yet — on the eve of my mother’s death, when the wind chimes outside her window began to sound with no noticeable breeze in the air, I intuitively joined my more religious siblings in their chorus of supernatural affirmation. Under similar circumstances, who among us wouldn’t fleetingly feel that this was their dead mother’s signal from the great beyond? (Rest assured that my consciousness immediately smothered this intuition with sound scientific reasoning.)

There were at least three possible ways that I might have reasoned about the cause of the wind chimes’ movement that evening. I might have: (i) intuitively thought that the movement was brought about by some imperceptible brush of wind, or perhaps a falling branch tickling its strings on its way to the ground (natural cause); (ii) intuitively thought that my mother’s spirit intentionally caused the wind chimes to move, for no particular reason other than that she was traipsing about outside and wanted to make them move (intentional cause), or (iii) intuitively thought that she caused the wind chimes to move, but that she did so in order to communicate with us: to ‘tell’ us that her soul had transitioned safely (“It’s okay, kids; all is swell on the other side of life!”) (declarative cause). Recall that, in the emotional construal of the moment, my intuitive causal ascription fell along the lines of the third variety.

This personal experience with my dead mother’s ghost was fodder for the experimentalist in me. And so, naturally, I decided to properly investigate this phenomenon in my laboratory. In a recent study, Becky Parker and I had 3- to 9-year-old children play a simple guessing game in which they were to find the location of a hidden ball inside one of two boxes (Bering & Parker, in press). Sticker prizes served as rewards for all correct guesses. To choose a box, children had only to place their hand on top of that box, and in fact they were given 15-s per trial in case they changed their mind and wished to move their hand. The methodological rub was that half of the children were informed that a friendly invisible princess (“Princess Alice”) in the room would help them find the ball by telling them, somehow, when they chose the wrong box. Then, on a counterbalanced 2 of the 4 trials, experimenters in an adjacent room triggered an unexpected event in the laboratory —a table lamp flickering and a picture crashing to the floor —as soon as the child’s hand first made contact with one of the two boxes. If children regarded these events as declarative messages from Princess Alice, then compared to the nonevent trials they would be more likely to move their hands and their verbal judgments would reflect this type of causal attribution (e.g., “it happened because I picked the wrong box”). In contrast, if the events were seen as only intentionally caused by Princess Alice, but not as communicative signs, then children shouldn’t move their hands in response to them nor explain the events as being about their choice of box. Results showed a significant effect of condition by age group. Only the oldest children (M = 7 years) who were assigned to the Princess Alice condition moved their hands and made verbal judgments reflecting declarative causal reasoning. Younger children (M = 5 years) failed to move their hands in response to the events; instead they saw Princess Alice as a trickster who caused the events only because she wanted to (e.g., “because she likes the picture better on the ground”). Finally, the youngest children (M = 4 years) either shrugged their shoulders or gave good scientific answers (e.g., “because the light’s broken”).

These age differences are puzzling given that even 2.5-year-olds see deictic gestures such as eye gaze and indexical pointing as referential and declarative. Only follow-up studies will help to disentangle alternative interpretations of these data. At the moment, however, I suspect that second-order theory of mind plays a role in this domain of causal reasoning (e.g., “As she can see from my behavior, Princess Alice knows [I don’t know] where the ball is actually hidden; thus, that event is her informing me that I have a false belief”).

On ‘Being’ Dead

But all this still begs the question: “If it is so obvious that mental states are obliterated by death, then why do people the world over believe and behave otherwise?” In much of my own work in this area, I have argued for something called the “simulation constraint hypothesis” to account for people’s intuitive reasoning about dead agents’ minds (which is of course an oxymoron!). Ever wonder why you can never actually die in a dream sequence? According to the simulation constraint hypothesis, it is because it is impossible to ever know what it is like to be dead; our phenomenological systems are literally forced to construe theoretical models of a subjective existence beyond death. For my doctoral dissertation, I had people reason about the psychological functioning of a protagonist who had just died in a car accident (Bering, 2002). For example, could he taste the breath mint he ate right before he died? Could he experience lust? Did he know that he was dead? Here is how one young ‘materialist’ answered this last question: “Yeah, he’d know, because I don’t believe in the afterlife. It is non-existent; he sees that now.” In an age-modified task in which a puppet mouse was killed by a puppet alligator, children similarly found it hard to disavow themselves of the possibility that the mouse continued to experience psychological states after its death (particularly emotional, desire, and epistemic states) (Bering & Bjorklund, 2004; Bering, Hernández-Blasi, & Bjorklund, in press).

In fact, the younger the child, the more likely he or she was to say that the dead mouse retained various aspects of its consciousness, which is precisely the opposite pattern that one would expect to find if the origins of such beliefs could be traced exclusively to cultural indoctrination. In fact, religious-type answers (e.g. Heaven, God, spirits, etc.) among the youngest children were extraordinarily rare. Recent findings by Paul Harris and Rita Astuti, however, show that the social context can also be highly influential when it comes to children’s reasoning about the afterlife. For instance, children are more likely to say that consciousness survives death when the story involves a religious figure (e.g., a priest or medicine man) rather than a doctor (Astuti, 2005; Harris & Giménez, 2005).

Conclusion

In conclusion, investigating folk beliefs about dead agents’ minds may provide important information concerning the evolution of human cognition. Since the idea of a soul hinges on the cognitive capacity to represent mental states, studies of peoples’ underlying beliefs about the fate of consciousness after death may open an empirical backdoor to this representational system. Reactions to dead bodies, the ability to reason about dead agents’ minds, or beliefs concerning the intelligent design of souls are just some of the empirical topics waiting to be further explored.

References

Allen, C., & Hauser, M. D. (1991). Concept attribution in nonhuman animals: Theoretical and

methodological problems in ascribing complex mental processes. Philosophy of Science, 58, 221-240.

Astuti, R. (2005). Turning belief into ethnography.  Unpublished manuscript.

Bering, J. M. (2002). Intuitive conceptions of dead agents’ minds: The natural foundations of afterlife beliefs as phenomenological boundary. Journal of Cognition and Culture, 2, 263-308.

Bering, J. M., & Bjorklund, D. F. (2004). The natural emergence of reasoning about the afterlife

as a developmental regularity. Developmental Psychology, 40, 217-233.

Bering, J. M., Hernández-Blasi, C., Bjorklund, D. F. (in press). The development of ‘afterlife’ beliefs in secularly and religiously schooled children. British Journal of Developmental Psychology.

Bering, J. M., & Johnson, D.D.P. (2005). “O Lord . . . you perceive my thoughts from afar”: Recursiveness and the evolution of supernatural agency. Journal of Cognition and Culture, 5, 118-142.

Bering, J. M., McLeod, K. A., & Shackelford, T. K. (in press). Reasoning about dead agents reveals possible adaptive trends. Human Nature.

Bering, J. M., & Parker, B. D. (in press). Children’s attributions of intentions to an invisible agent. Developmental Psychologist.

Boyer, P. (2001). Religion explained: The evolutionary origins of religious thought. New York, NY: Basic Books.

Burnham, T., & Hare, B. (2005). Engineering human cooperation: Does involuntary neural activation increase public goods contributions in adult humans? Human Nature.

Dechesne, M., Pyszczynski, T., Arndt, J., Ransom, S., Sheldon, K. M., van Knippenberg, A., & Janssen, J. (2003). Literal and symbolic immortality: The effect of evidence of literal immortality on self-esteem striving in response to mortality salience. Journal of Personality and Social Psychology, 84, 722-737.

Goodall, J. (1990). Through a window. Boston, MA: Houghton Mifflin.

Gullone, E., King, N. J., Tonge, B., Heyne, D., & Ollendick, T. H. (2000). The fear survey schedule in children – II (FSSC-II): Validity data as a treatment outcome measure. Australian Psychologist, 35, 238-243.

Haley, K., & Fessler, D. (2005). Nobody’s watching? Subtle cues affect generosity in an anonymous economic game. Evolution and Human Behavior 26:245-256.

Harris, P. L., & Giménez, M. (2005). Children’s acceptance of conflicting testimony: The case of

death. Journal of Cognition and Culture 5:143-162.

Miller, W. R. & Brigham, R. M. (1990). “Ceremonial” gathering of black-billed magpies (Pica pica) after the sudden death of a conspecific. Murrelet, 69, 78-79.

Reynolds, V., & Tanner, R. (1995). The social ecology of religion. New York: OxfordUniversity

Press.

Wilson, E. O. (1971). The insect societies. Cambridge, MA: Belknap Press.

de Unamuno, M. (1926). The tragic sense of life. London: MacMillan and Co., Ltd.

de Waal, F. B. M. (1996). Good natured: The origins of right and wrong in humans and other animals. Cambridge, MA: HarvardUniversityPress.