Michael Shermer, The Believing Brain (2011)

This is a better, more thorough book on the subject of how we come to believe things that are not real than Supersense (see June 5, 2011 post). I say that for multiple reasons. Not just because Michael Shermer avoids a gimmicky term like "supersense," but primarily because the structure of the author's presentation and the evidence marshaled in support of his thesis is more persuasive than Hood's more anecdotal presentation. While there is much in common with the subject-matter of Supersense, in The Believing Brain Michael Gazzaniga's "left hemisphere" interpreter is presented as a candidate for the brain's storytelling capability, which reconstructs events and weaves those events into a meaningful story --- including stories that are simply works of fiction. As I said in the previous post, this storytelling capability is critical to our narratives of belief. Finally, Shermer treats his subject with seemingly more objectivity than Hood. For Shermer, the study of how we believe something is not limited to the strange or unreal --- ghosts, angels, gods, aliens and phantom conspiracies --- but also politics, economics, and even hard science.

According to Shermer (and Hood, see previous post), as sensory data flows into the brain, there is a "tendency" for the brain to begin looking for meaningful patterns in both meaningful and meaningless data. He calls this process patternicity. A second tendency of the brain is to "infuse patterns with meaning, intention, and agency." Shermer labels this tendency agenticity. Because of these tendencies, we form beliefs first and only later do we try to inform our beliefs with facts. The overall process Shermer calls belief-dependent realism (well, maybe there is one gimmicky term). This model not only purports to explain not only how we form entirely mistaken, fantasy beliefs, but also beliefs that are later embraced under the gospel of science (e.g. hypothesis later substantiated by empirical and repeatable testing).

Harkening back to David Hume (see February 27, 2011 post), Shermer asserts that patternicity is premised on "association learning," which is "fundamental to all animal behavior from C. elegans (roundworm) to homo sapiens." Because our survival may depend on split-second decisions in which there is no time to research and discover underlying facts about every threat or opportunity that faces us, evolution set the brain's default mode in the position of assuming that all patterns are real, says Shermer. A cost associated with this behavior is that the brain may lump causal associations (e.g. wind causes plants to rustle) with non-causal associations (e.g. there is an unseen agent in the plants). In this circumstance, superstition --- incorrect causal associations --- is born. "In this sense, patternicities such as superstition and magical thinking are not so much errors in cognition as they are the natural processes of a learning brain." Religion, conspiracy theories and political beliefs fit this model as well.

The difficulty faced by humans applying later learned facts to earlier-formed beliefs is that beliefs are often difficult to shake despite the fact that our factual knowledge is entirely inconsistent with the belief. The difficulty Galileo faced in convincing others that the earth revolved around the sun in violation of the long-held belief in Aristotle's geocentric view and the words of the Bible that the sun revolved around the earth is an example. What Shermer does, however, to document this difficulty is to catalog a series of biases that reinforce beliefs, which he labels cognitive heuristics that confirm the beliefs are correct. A heuristic is a brain's capacity to solve problems through intuition, trial and error, rules of thumb, or other informal means, when there is no formal means for solving the problem. These biases are part of our documented behavior, including the tendency to seek and find confirmatory evidence in support of an existing belief and to ignore or reinterpret disconfirming evidence (confirmatory bias), the tendency to reconstruct the past to fit the present (hindsight bias), the tendency to rationalize decisions after the fact to confirm that our action were the best we could have done under the circumstances (self-justification bias), the tendency to attribute different causes for our own beliefs and action than that of others (attribution bias), the tendency to believe in something because we have already invested so much in that belief (sunk-cost bias), the tendency to opt for whatever we are used to or familiar with (status quo bias), the tendency to value what we own more than what we do not own (endowment effect), the tendency to draw different conclusions based on how data is presented (framing effect), the tendency to rely too heavily on a past reference or on one piece of information (anchoring bias), tendency to miss something obvious while attending to something special and specific (inattentional blindness bias), and the tendency to recognize the power of cognitive biases in others, but to blind to their influence on ourselves (bias blind spot) and others.

Regrettably, Shermer does not provide a biological explanation for these biases (or "tendencies," as he seems inclined to refer to them). And "bias" may be too charged a term. Bias is a term that arguably reflects a cultural or community predisposition of some sort (whether the community consists of a family, a village, a region, nation or species), which is really within the province of nurture rather than nature. These "tendencies," however, can also be predispositions grounded in our biology, particularly our neurobiological system. I strongly suspect that if we dissected human brains and a network of connected neurons from a representative sample of humans, we would find a very high level of near identity among brains. There will be some differences due to DNA, and there will be some pathological differences as well, perhaps caused during embryonic development. But I believe that by and large we will find that human brains, neuron by neuron, are organized and folded and layered in substantially identical ways. This is the product of evolution. Some of these neuronal networks are unconsciously and automatically operational as part of what Damasio calls the protoself -- the body monitoring itself, (see April 8, 2011 post) and are functioning and repeating the transmission of electro-chemical signals to and from the brain even before we are born. Other neural networks are only activated in response to sensory experience with the external physical world as part of what Damasio calls the "core self," which process begins immediately with our birth. Finally, when the neural networks within the cerebral cortex are activated to trigger autobiographical memory and our storytelling capability, we find our autobiographical self. While much of the neuronal activity in the cerebral cortex is connected to the automated, unconscious neuronal networks tied to monitoring our own body and our external sensory experiences that are part of our core self experience, there is much that is influenced more by nurture than nature --- culture and learning at the social level. At the level of the protoself and the core self, except to the extent that a particular individual may have a variant pathological condition, there is a high level of commonality in the activation of neuronal networks repeatedly sending and receiving electro-chemical impulses. This is reflected in the high level of commonality of similar emotional and feeling experiences humans share in response to the same stimuli --- whether the stimuli is within the body or outside of it. There will be more differences among us at the level of the core self, because we will not always share sensory experiences with the same level of frequency as all other humans. I am now in speaking in terms of the brain's capacity for memory and learning, which, as Eric Kandel demonstrated, is influenced by synaptic connections strengthened through repeated experience and learning. In contrast, at the level of the neural networks in the cerebral cortex, there is likely to be greater disparity among us in the precise neural networks activated, because we are going to have significantly different cultural and social learning experiences. It is here where biases may be formed. But at the level of the protoself and core self, important biological predispositions or tendencies have significant influence on instincts and intuition that are part of the belief-forming process. And the tendencies that Shermer identifies may be the outcome of both of biologically hardwired predispositions and culturally learned biases.

The Believing Brain has a discussion entitled "Synaptic States and Believing Neurons" in which Shermer recognizes that neurons communicate information in three ways: (1) firing frequency, (2)firing location, and (3) firing number. This discussion also includes a recognition that certain types of chemicals in the neuronal network act to reinforce learning and belief. This discussion, in my view, supports what I just said. Shermer, on the other hand, does not incorporate this discussion into his discourse on biases, and this is a shortcoming of his discussion on biases.

Our senses evolved, Shermer observes, for "perceiving objects of middling size." I wrote in an earlier post (see August 31, 2009 post) that in my worldview, there are three very big subjects for human inquiry: 1) the realm of the very large --- the universe (and whether there is more than one, making the word universe a possible oxymoron), its origin and history; 2) the realm of the very small -- the smallest molecular (sub)units, and their behavior; and 3) the human mind --- how it works, consciousness, intentionality. "We are not equipped to perceive atoms and germs, on the one end of the scale, or galaxies and expanding universes, on the other end," writes Shermer. I disagree, but I understand Shermer's point. There are several cognitive biases, listed above (e.g., the status quo bias, the inattentional blindness bias), that discourage the general population from looking beyond objects of middling size to the realms of the very large and the very small. But frankly, most of us simply lack the education, training, and time to focus on the realms of the very large and the very small. We do not understand them. Shermer seems to have concluded that only "scientists," those trained in the scientific approach to knowing and understanding the world, are capable of neutralizing these various biases. But the substantial variation among the human population in education, training, and the amount of time we have to focus on the large and small confirms that our biases are significantly tied to culture and nurture. Are we all born with the capacity to become an Einstein or a Steven Weinberg? I am not saying that. But I don't believe Einstein became Einstein entirely on his own.

Finally, I want to return to the concept of agenticity -- the tendency to impart patterns with intention and agency, even patterns we perceive in inanimate objects, that lead us to believe that unseen agents control the world. Now we enter the world of spirits, gods, souls, devils, angels, aliens, and yes, intelligent designers. This is the domain of teleology, which I have mentioned in previous posts. (See March 24, 2010 and May 12, 2010 posts), as well as that part of philosophy that discusses intentionality --- how our brain relates to objects external to ourselves. Shermer's discussion of essentialism flows from the writings of Paul Bloom in Descartes' Baby, which concept is elaborated upon by Hood as well. Essentialism refers to the brain's ability to abstract about a physical object and to ascribe to that object an unobservable essence --- beliefs, desires, intentions and goals --- and treat these invisible essences as real. In the case of animate objects like other humans or animals, notes Michael Gazzaniga in Human, we believe they must have beliefs, desires, intentions, etc. just like our own. Anthropomorphism and teleological thinking are born.

Two questions emerge from the development of this discussion: what evolutionary value would essentialism provide to humans? And what is the biological process in the brain that explains it? Gazzaniga explains that our ability to reason about unobservable entities or unobservable essences allows humans to predict and explain events --- to predict the behavior of others. In other words, we can predict the behavior of another animal by inferring its psychological state (sometimes referred to as a theory of mind (TOM)), which would have enormous significance for survival. It also appears to relate to our storytelling capability about things that do not exist. Gazzaniga explains it this way in Human:

"Sometimes our predilection for explaining the cause of things or behaviors with teleological thinking runs amuck. One of the reasons is that the agency-detection device [in our brain] is rather zealous. Barrett calls it hyperactive. It likes to ...find animate suspects even when there are none. When you hear a sound in the middle of the night, the question that first comes to mind is Who is that? rather than What is that? When you see a wispy something moving in the dark, Who is that comes to mind because the detective device is not modern and up-to-date. The detective device was forged many thousands of years ago before there were inanimate objects that could move or make noise on their own. To first consider a potential danger as animate is adaptive. It worked most of the time. Those who did it survived and passed their genes to us. *** The hyperactive detective device, combined with our need to explain and teleological thinking, is the basis of creationism. To explain why we exist, the hyperactive detection device says there must be a Who involved. Teleological reasoning says there must be an intentional design. The cause must be the desires and intentions and behavior of the Who. The we were designed by a Who. *** All of this is reminiscent of what the left-brain interpreter would do, which it has been demonstrated to do in other settings."

The research into so-called mirror neurons (see September 18, 2009 post) puts us on the road to explaining how the brain helps us form beliefs about other people's intentions, and Shermer incorporates this research into his discussion of agenticity. So does Paul Ekman's research into facial descriptions, as described in Dacher Keltner's book Born to Be Good (see July 16, 2010 post).

I disagree with Shermer and others that we are hardwired to believe in God. Hardwired means that it is in our DNA, and I have rejected the idea that there is a "god gene" in prior posts (see November 30, 2009 post). We are hardwired to be disposed to believe first that if we do not immediately perceive a causal explanation for some event, that some unseen agent with intentions and desires and beliefs very similar to our own is the causal explanaton for this event. That unseen agent may be another person we do not see, another animal we do not see, a ghost we do not see, or the devil we do not see. We are hardwired to search for patterns in what we perceive, and we are hardwired with a storytelling capacity to explain what we perceive, and sometimes that story is based on things we can actually see and touch, and sometimes that story is made up out of whole cloth. But even if that story is made up out of whole cloth, it is a story about an agent that thinks, desires and behaves just like ourselves. God was made by humans in our image, not the other way around. In this explanation, we also begin to understand our ability for creativity --- engineering, songwriting, and writing works of fiction.