Part VI The Human Sciences

The Third Wound

Has Psychology Banished the Ghost from the Machine?

DYLAN EVANS

Introduction

The advance of science has challenged some of humanity’s most cherished beliefs. Sigmund Freud noted three such scientific disturbances, describing them as “severe wounds” (schwere Kränkungen¹). The first was the cosmological wound inflicted by Copernicus. Then there was the biological injury caused by Darwinism. The third wound was the psychological one which Freud modestly attributed to his own theory of psychoanalysis (Freud 1917). For Freud, the victim of these three assaults was “the universal narcissism of men,” but it would be more historically accurate to identify the Christian faith as the real victim of these particular scientific revolutions. It would also be more accurate to credit psychology in general, rather than psychoanalytic theory, with inflicting the third wound, but Freud was right in thinking that this was the deepest cut.

Despite what churchmen thought in the fifteenth and sixteenth centuries, no deep theological question hinges on whether the sun orbits the earth or vice versa. Nor is there any fundamental incompatibility between belief in a Christian God and Darwin’s theory of evolution by natural selection. As I will argue in this chapter, however, the various developments in scientific psychology over the past century are far more corrosive of Christian faith than the earlier revolutions in astronomy and biology. I will focus on three strands of psychological research in particular: the study of human error, research into the placebo effect, and the demonstration that psychological properties can be exhibited by purely physical systems.

Human Error

One way in which psychology tends to undermine Christian faith is by exposing the extent of human error. In particular, the “feeling of knowing” often fails to correspond with real knowledge (Koriat et al. 1980). The fact that people can feel utterly convinced of something and yet be completely mistaken was of course well known long before the advent of scientific psychology in the mid-nineteenth century. Descartes famously illustrated the problem by imagining a malignant demon, “who has employed all his energies in deceiving me,” and proceeded to ask, rhetorically: “could not an all-powerful demon make me believe those propositions are true when, as a matter of fact, they are not?” (Meditations, AT 22). In the past few decades, however, psychological research has gone well beyond such blanket forms of skepticism, identifying certain systematic patterns of error which cast more empirically founded doubts on Christian beliefs.

One way of categorizing types of error is to divide them into two groups: false positives (in which x is mistakenly identified as y) and false negatives (in which x is not recognized as y, even though it is). Psychologists have identified many cognitive systems in which these errors are not randomly distributed but systematically skewed in one direction (Haselton and Nettle 2006). That is, instead of making an equal number of each type of mistake, many cognitive systems make more of one kind than the other. For example, the psychological mechanisms responsible for spotting patterns are biased towards false positives; they are more likely to see a pattern in what is, in fact, a random series than to mistake order for chaos. A classic example is the so-called “clustering effect,” in which people are more likely to mistake a random sequence of coin tosses for a nonrandom one than vice versa.

Experimental evidence suggests that there are similar biases in the cognitive mechanisms for spotting causal relations and intentional agents. That is, we are more likely to imagine that things are causally related when they are not, rather than vice versa, and more likely to see an intentional agent where there is none than fail to spot one that is there. Several psychologists and anthropologists have argued that these cognitive biases provide a naturalistic explanation for the origin of many religious beliefs. Skinner, for example, argued that the tendency to see causal relations where there is none, which is also evident in nonhuman animals, was at the root of many superstitions. In a famous experiment, Skinner placed hungry pigeons in cages fitted with automatic mechanisms which delivered food to the pigeons “at regular intervals with no reference whatsoever to the bird’s behavior” (Skinner 1948, 168). He found that the pigeons associated the delivery of the food with whatever chance actions they had been performing as it was delivered, and that they subsequently continued to perform these same actions:

One bird was conditioned to turn counter-clockwise about the cage, making two or three turns between reinforcements. Another repeatedly thrust its head into one of the upper corners of the cage. A third developed a “tossing” response, as if placing its head beneath an invisible bar and lifting it repeatedly. Two birds developed a pendulum motion of the head and body, in which the head was extended forward and swung from right to left with a sharp movement followed by a somewhat slower return.

(Skinner 1948, 168)

Skinner suggested that the pigeons behaved as if they were influencing the automatic mechanism with their rituals and that this experiment shed light on human behavior:

The experiment might be said to demonstrate a sort of superstition. The bird behaves as if there were a causal relation between its behavior and the presentation of food, although such a relation is lacking. There are many analogies in human behavior. Rituals for changing one’s fortune at cards are good examples. A few accidental connections between a ritual and favorable consequences suffice to set up and maintain the behavior in spite of many unreinforced instances. The bowler who has released a ball down the alley but continues to behave as if she were controlling it by twisting and turning her arm and shoulder is another case in point. These behaviors have, of course, no real effect upon one’s luck or upon a ball half way down an alley, just as in the present case the food would appear as often if the pigeon did nothing – or, more strictly speaking, did something else.

(Skinner 1948, 171)

Superstition might then be a natural result of a “hyperactive causation detector.” In a similar way, Scott Atran and Pascal Boyer have argued that beliefs in spirits and gods may be a natural result of a parallel tendency to mistake inanimate things for intentional agents (Boyer 2001; Atran 2002). Justin Barrett attributes this tendency to what he calls a “hyperactive agent detection device,” or HADD, which errs on the safe side by mistaking sticks for snakes rather than vice versa, but which also leads us to attribute agency to thunder and lightning (Barrett 2000).

It is interesting to note that in these examples of systematic error, all the cognitive biases favor religious beliefs rather than nonreligious ones. This poses a serious challenge to the Christian notion that God has left man free to disbelieve. A Christian who accepts the evidence from psychology would have to concede that the creator had endowed the human mind with a strong innate tendency to believe in him. Far from allowing his creatures the freedom to make their own minds up on matters of faith, then, it would appear that God had heavily loaded the dice in his favor.

Similar biases can also shed light on cases of pareidolia, in which people perceive religious imagery in natural phenomena. In 1978, for example, a New Mexican woman found that the burn marks on a tortilla she had made appeared similar to the traditional Western depiction of Jesus’ face. Thousands of people subsequently came to see the framed tortilla. In 2009, an equally abundant number of Irish Catholics flocked to a church in Rathkeale, County Limerick, to pray at the stump of a recently cut willow tree in which many observers claimed to see the silhouette of the Virgin Mary. It would appear that the cognitive mechanisms responsible for face recognition are just as prone to false positives as the mechanisms for spotting causal relations and intentional agents.

Although these biases are found in most people, there are of course individual differences in the degree to which they influence cognition. In a particularly revealing experiment carried out by Peter Brugger, a neurologist at the University Hospital in Zurich, volunteers were asked to distinguish real faces from scrambled faces as the images were flashed up briefly on a screen, and then to distinguish real words from made-up ones (Krummenacher et al. 2010). Those who believed in the paranormal were much more likely than skeptics to see a word or face when there was not one, whereas skeptics were more likely to miss real faces and words when they appeared on the screen. The researchers then gave the volunteers a drug called L-dopa, which is usually used to relieve the symptoms of Parkinson’s disease by increasing levels of dopamine in the brain. Under the influence of the drug, both groups made more mistakes, but the skeptics became more likely to interpret scrambled words or faces as the real thing. This suggests that paranormal thoughts are associated with high levels of dopamine in the brain and that L-dopa makes skeptics less skeptical. More generally, people with higher levels of dopamine are more likely to find significance in coincidences, and pick out meaning and patterns where there are none. This fits in well with the dopamine hypothesis of schizophrenia, according to which some of the symptoms of disease are caused by hyperactive dopaminergic signal transduction; paranoid delusions often involve the attribution of sinister meanings to random events such as lottery numbers or freak accidents.

Another symptom of schizophrenia which may be caused by high levels of dopamine in the brain is hallucination. Psychologists now think that hallucinations are more common than was previously suspected, and are by no means limited to schizophrenics. Indeed, there seems to be a continuum from the bizarre visions of the latter to the almost imperceptible illusions which are common features of everyday life. Reports of religious statues nodding should not be taken at face value, therefore, even if the observers are not mad. Indeed, the disciples may have hallucinated some of the more bizarre episodes recounted in the gospels, such as the ascent of Jesus into heaven, and the Book of Revelation was almost certainly written by a schizophrenic.

It is not necessary, however, to appeal to such extreme examples of human error as visual hallucinations to explain the many weird stories in the gospels. Since the pioneering work of Sir Frederic Bartlett in the 1930s, psychologists have known that remembering does not involve retrieving an exact copy of the original experience but rather a process of reconstruction (Bartlett 1932). This process is often biased by a tendency to force remembered details into the procrustean bed of pre-existing schemata. Sometimes whole chunks can be fabricated unconsciously in a process that psychologists call confabulation. Given that the earliest of the four gospels – that of Mark – was not written down until at least 40 years after the death of Jesus, there would have been more than enough scope for the Chinese whispers of human remembering and the distortions of confabulation to transform perfectly normal events into magical wonders. I do not mean to imply, of course, that the gospel was the product of a single author who attempted to recall events that he had personally witnessed four decades before; clearly it was a communal effort, and may well have drawn on earlier written sources that are no longer extant. These qualifications, do not, however, alter my point about Chinese whispers; if anything, they reinforce it.

Besides the particular conclusions that can be drawn by applying theories of various types of human error to the gospels and to the faith of ordinary believers, there is a more general lesson that emerges from subjecting religious experience to the same scientific scrutiny as any other natural phenomenon. The pioneer in this endeavor, William James, was careful to state in his lectures on The Varieties of Religious Experience that such research could be conducted without taking any view on whether or not religion was also a supernatural phenomenon (James 1997 [1902]). Nevertheless, both believers and nonbelievers have tended to feel that, by shining a harsh light on the often prosaic nature of religious experience as a natural phenomenon, psychological research could not help but cast doubt on its supposed supernatural aspects. Indeed, this may be why Christians have often resisted scientific scrutiny of their beliefs; as Dan Dennett observes, “the religious … often bristle at the impertinence, the lack of respect, the sacrilege, implied by anybody who wants to investigate their views” (Dennett 2006, 17, emphasis in original). Dennett suggests that this “taboo against a forthright, no-holds-barred scientific investigation of one natural phenomenon among many” is bound together with religion itself in a “curious embrace” (17–18). Part of the strength of religion itself may be due to the protection it receives from this taboo.

The Placebo Effect and “Faith Healing”

Another way in which psychological research tends to undermine Christian faith is by providing naturalistic explanations for cases of healing which believers had thought to be “miraculous.” Many Christians have placed great store on the gospel accounts of how Jesus restored sight to blind people, cured people of leprosy, and restored movement to a paralytic man, merely by saying a few words. The Catholic Church stipulates that at least one miracle is necessary before canonizing someone as a saint, and these miracles are often putative cures. Since the 1860s, thousands of pilgrims have left their crutches and canes at the Sanctuary of Our Lady of Lourdes in France, and hundreds of other visitors claim to have been cured of advanced cancers after visiting the shrine. During the past 50 years, television evangelists such as Oral Roberts have attracted millions of viewers worldwide with their dramatic displays of “divine healing.”

Catholics and evangelicals alike agree that such cases count as miracles only if they cannot be explained in scientific terms. Indeed, for many years they could not. Since the emergence of psychoneuroimmunology in the 1970s, however, a number of biological mechanisms have been identified that could explain a variety of cures previously attributed to divine intervention. Some of these go under the general umbrella of the “placebo effect”; this is a process that begins in the brain with the formation of a particular sort of belief – the belief that one has taken, or been given, a powerful therapy of some sort or another (Evans 2003). The precise nature of the therapy is not important, provided that the person receiving it believes it can help to relieve the condition that he or she is suffering from. Pills, injections, surgery, psychotherapy, and acupuncture are just some of the many medical procedures that, depending on one’s belief in their effectiveness, are capable of inducing placebo responses. When someone comes to believe that he has received some such therapy, this belief tends to activate a cascade of chemical messengers which ends in some (perhaps all) cases with the local release of endorphins, the body’s natural pain killers.

A key difference between this scientific account of the placebo response and faith healing from a Christian perspective is that the former views the power of the mind to heal the body as entirely dependent on the various physical mechanisms just described. In the scientific accounts, if there is no chemical messenger to act as a go-between, the brain is powerless to alter the action of the immune system, and even when such molecules do exist, they cannot endow the immune system with supernatural powers. All they can do is tell the immune system to behave in one way rather than another. If something is beyond the power of the immune system altogether, then no amount of chemical messengers secreted by the brain will change this.

The scientific accounts therefore predict that the range of “mysterious” cures will be highly circumscribed. Far from being a universal panacea, “divine” healing will exhibit systematic patterns which coincide with underlying biological mechanisms. Indeed, this is exactly what we find. The shrine at Lourdes proudly displays hundreds of crutches cast off by pilgrims who feel able to walk without them after taking the holy water, but no wooden legs. Nor does the shrine display any glass eyes or prostheses left behind by pilgrims whose eyes and hands have grown back. Nicholas Humphrey calls this the “argument from unwarranted design” (Humphrey 1996).

Of course, before attempting to explain any cure, we should first establish that it did, in fact, occur. This means establishing that the disease was present before the alleged miracle. In his wonderful exposé of faith healing in the US, James Randi reports many instances of evangelists who have cured people of diseases they never had (Randi 1989). Even when a disease really is present before the putative cure, one must take care to verify that it is absent afterwards. More observant visitors have noticed that the crutches on display at Lourdes change with an alarming frequency. That is because many are reclaimed a few days after being cast aside by pilgrims who find that they cannot, in fact, do without them.

Even more fascinating than the scientific explanations for cures that Christians have falsely attributed to divine intervention are the psychological mechanisms that allow the fraudulent healers to act with wholehearted sincerity. Randi has also documented the process by which many faith healers start off with deliberate deception, employing methods which they know full well to be tricks. As their reputation grows and they gather followers, however, these healers gradually come to believe their own lies, and eventually become capable of breathtaking doublethink (Randi 1989).

Nicholas Humphrey has made a persuasive case that this is exactly what happened with Jesus himself (Humphrey 1996). He notes that Jesus’ miracles were entirely typical of the conjuring tricks that were popular around the Mediterranean at the time. Hippolytus, for example, describes a certain Marcus who had mastered the art of turning the water in a cup red by mixing liquid from another cup while the onlookers’ attention was distracted. Humphrey then asks, rhetorically:

Is it possible that, even though Jesus was regularly using deception and trickery in his public performances like any common conjuror, he actually believed that he was more than a conjuror: believed that sometimes he could genuinely exert the powers he claimed?

(Humphrey 1996, 99)

If this is indeed what happened, it would reveal a different kind of pathos in those famous words spoken from the cross, echoing the twenty-second psalm: “My God, my God, why hast thou forsaken me?” If Jesus really did say something like this as he was dying, then might these words express genuine bewilderment on Jesus’ part about why he could not summon up any supernatural help when he most needed it.

Humphrey’s speculations on the psychology of Jesus are just that – speculations. Nevertheless, they do point to another way in which psychology can undermine Christian faith – namely, by providing plausible accounts of how a normal, fallible human being could behave like Jesus, and thus disposing of the false dilemma posed by C. S. Lewis, according to which Jesus must be either mad or divine (Lewis 1952). Instead, he was just mistaken in believing himself to be the messiah and Son of God.

The Mechanical Mind

The third way in which psychology tends to undermine Christian faith is by demonstrating that psychological properties can be exhibited by purely physical systems. Of the three strands of psychological research examined in this essay, this is the one that poses the most serious challenge for Christianity, for it strikes at the very existence of the soul.

As with concerns about human error, the possibility that the human mind might be entirely mechanical was also anticipated by Descartes. In the Discourse on Method, Descartes was happy to admit that bodies – whether of humans or other animals – were just machines:

You won’t find that at all strange if you know how many kinds of automata or moving machines the skill of man can construct with the use of very few parts, in comparison with the great multitude of bones, muscles, nerves, arteries, veins and all the other parts that are in the body of any animal, and if this knowledge leads you to regard an animal body as a machine. Having been made by the hands of God, it is incomparably better organised – and capable of movements that are much more wonderful – than any that can be devised by man, but still it is just a machine.

(Discourse on Method, AT 55)

Descartes was also prepared to concede that the behavior of all nonhuman animals could be controlled by their biological machinery. Aware of the slippery slope that this train of reasoning threatened to open up, Descartes argued that there were aspects of human behavior that could not be attributed to purely physical or biological mechanisms, such as conversation:

We can easily conceive of a machine so constructed that it utters words, and even utters words that correspond to bodily actions that will cause a change in its organs (touch it in one spot and it asks “What do you mean?”, touch it in another and it cries out “That hurts!”, and so on); but not that such a machine should produce different sequences of words so as to give an appropriately meaningful answer to whatever is said in its presence – which is something that the dullest of men can do.

(Discourse on Method, AT 56)

But this argument is hostage to fortune. If a machine could “give an appropriately meaningful answer” to any utterance, Descartes would have to concede that humans are just machines, too. Three hundred years later, the very same test was proposed by the British computer scientist Alan Turing, but by then the presumption had been reversed. Turing assumed that the test would eventually be passed (Turing 1950).

Like Descartes, Turing noted the theological problems that this would pose. The very idea of a humanlike robot suggests that we too might be purely physical entities, devoid of souls. People seem to perceive this instinctively, even without conscious reflection, and this may be the reason why androids so often evoke spooky feelings, as Freud noted in his essay on “The Uncanny” (Freud 1919). A story from Descartes’s own life illustrates the visceral reactions aroused by such automata. While sailing to Sweden, some sailors crept into Descartes’s cabin and discovered a humanoid robot which the philosopher had constructed himself. When shown the contraption, the ship’s captain was horrified, and – convinced that it was some instrument of dark magic – ordered it to be thrown overboard (Wood 2002).

Other historical episodes provide further support for the idea that this horror of robots has theological roots. Four centuries before Descartes, Albert Magnus is said to have made several metal androids, which could move, answer questions, and solve problems. His android butler, 30 years in the making, could answer the door at the sound of a knock. Thomas Aquinas, his pupil, destroyed the android, thinking the devil was in it. In 1727, the French inventor Jacques de Vaucanson built some clockwork androids which could serve dinner and clear the tables for his fellow monks (he was a novice in the order of the Minimes at the time). When the head of the order saw the automata, he declared that Vaucanson had profane tendencies, and ordered that his workshop be destroyed (Wood 2002). Some years later, the French physician Julien Offray de La Mettrie had to flee Holland when the Church forced his publisher to deliver up all the copies of his new book, L’Homme machine (Man a Machine), for burning.

The cognitive revolution which swept through psychology in the 1960s, displacing the behaviorist paradigm that had held sway since the 1920s, is based on the assumption that the human mind can be modeled entirely in computational terms. If one day had to be singled out as the birthday of cognitive science, it is surely September 11, 1956. It was on that day that three seminal papers were presented at a historic meeting at the Massachusetts Institute of Technology (MIT). One of these was the paper by Allen Newell and Herbert Simon about a “logic theory machine,” which inaugurated the modern discipline of artificial intelligence (Newell and Simon 1956).

The contemporaneous invention of integrated circuits, in which large numbers of components were packed onto a small silicon chip, gave cognitive scientists the means to turn their computational models into real machines. An early example of the new generation of “cognitive robots” was Shakey. Developed at the Artificial Intelligence Center at Stanford Research Institute between 1966 and 1972, Shakey had a complex cognitive architecture in which distinct functions such as perception, planning, and natural language processing were implemented by separate programs, which reflected the emphasis of cognitive psychology on the functional decomposition of mental processes.

It is this strategy of functional decomposition which provides the solid theoretical support for the intuitive notion that cognitive robots pose deep theological problems. By breaking the mind down into successively smaller components, cognitive science avoids the homunculus fallacy – the idea that cognition can be explained by positing the existence of a homunculus (a little man, or soul) inside the head which performs cognitive tasks such as face recognition or logical reasoning. This is, of course, no explanation at all, since it begs the question of how the homunculus is performing the task. Functional decomposition avoids the infinite regress inherent in the homunculus fallacy by stipulating that the components must be simpler at each step, until they become trivial to build (Dennett 1978). At this point there is no place left in the machine for any ghost to hide.

Conclusion

I have argued that the various developments in scientific psychology over the past century pose deep problems for Christianity. These problems are implicit in many areas of psychological research, but for reasons of space I have focused on three strands in particular. The study of human error suggests that many religious beliefs are natural consequences of systematic cognitive biases. Research into the placebo effect provides a biological explanation for many so-called miracles. And the development of cognitive robots demonstrates that psychological properties can be exhibited by purely physical systems. Taken together, these three strands of psychological research are far more corrosive of Christian faith than the earlier scientific revolutions inspired by Copernicus and Darwin.

Note

1 This phrase is translated as “severe blows” by James Strachey in the Standard Edition, but I have chosen to follow Joan Riviere’s earlier translation in using the word “wound.”

References

Atran, Scott 2002. In Gods We Trust: The Evolutionary Landscape of Religion. Oxford: Oxford University Press.

Barrett, Justin. 2000. Exploring the Natural Foundations of Religion. Trends in Cognitive Science, 4, pp. 29–34.

Bartlett, Frederic. 1932. Remembering. Cambridge: Cambridge University Press.

Boyer, Pascal. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New York: Basic Books.

Dennett, Daniel C. 1978. Brainstorms. Cambridge, MA: MIT Press.

Dennett, Daniel. 2006. Breaking the Spell: Religion as a Natural Phenomenon. London: Allen Lane.

Evans, Dylan. 2003. Placebo: The Belief Effect. London: HarperCollins.

Freud, Sigmund. 1917. A Difficulty in the Path of Psycho-Analysis. In J. Strachey, ed. The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 17. London: Hogarth Press, pp. 135–144.

Freud, Sigmund. 1919. The “Uncanny.” In J. Strachey, ed. The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 17. London: Hogarth Press, pp. 217–256.

Haselton, Martie G. and Nettle, Daniel. 2006. The Paranoid Optimist: An Integrative Evolutionary Model of Cognitive Biases. Personality and Social Psychology Review, 10(1), pp. 47–66.

Humphrey, Nicholas. 1996. Leaps of Faith: Science, Miracles, and the Search for Supernatural Consolation. New York: Basic Books.

Koriat, A., Lichtenstein, S., and Fischhoff, B. 1980. Reasons for Confidence. Journal of Experimental Psychology: Human Learning and Memory, 6, pp. 107–118.

Krummenacher, Peter, Mohr, C., Haker, H., and Brugger, P. 2010. Dopamine, Paranormal Belief, and the Detection of Meaningful Stimuli. Journal of Cognitive Neuroscience, 22(8), pp. 1670–1681.

Lewis, C. S. 1952. Mere Christianity. London: Macmillan.

Newell, Allen and Simon, Herbert. 1956. The Logic Theory Machine: A Complex Information Processing System. Information Theory, IRE Transactions on Information Theory, 2(3), pp. 61–79.

Randi, James. 1989. The Faith Healers. Amherst, NY: Prometheus Books.

Skinner, B. F. 1948. “Superstition” in the Pigeon. Journal of Experimental Psychology, 38, pp. 168–172.

Turing, Alan. 1950. Computing Machinery and Intelligence. Mind, 59(236), pp. 433–460.

Wood, Gaby. 2002. Living Dolls: A Magical History of the Quest for Mechanical Life. London: Faber and Faber.