Saturday, December 10, 2005

The gods of the cognitive scientists   posted by Razib @ 12/10/2005 06:01:00 PM

Imagine yourself walking in some woods and you see something out of the corner of your eye. What is your thought as you turn toward the object, a bear, or an old refrigerator? I don't know about you, but most "woods" in suburban and small town areas have a greater number of discarded junk, from sofas to fridges, than they do of wild animals (the main exception around here are deer, though to be honest I've seen many more deer ambling down main streets raiding gardens than in wooded areas where the pickings are slim). Even though banal physical objects or processes are more likely to catch your attention out of the corner of your eye or induce leaves to rustle, we still have a bias toward perceiving animate agents when we are unsure (or worry about them!). Now, imagine another scenario, you see someone who holds a strange object, vaguely metallic, and they throw the object to the ground. The object lands with a thud. OK, big deal. Now, how would you react if the object just slipped into the ground and disappeared. Or, what if you heard a scream eminate from it, and it "ran" away from the location where it landed? I think you'd be pretty weirded out. The point here is that humans have innate and reflexive biases in how the world around us is constructed. Metallic objects don't scream or run. The ground is solid. We don't process all the information anew each time we confront a situation. We have an idea that rocks aren't going to start screaming if we sit on them, we don't go into each situation consciously evaluating that possibility.

As regards the reflexive aspects of our mind, I don't mean to imply that a coupling of inputs and outputs are innate, hard-wired or perfect. Many reflexive tendencies do not resemble anything like a hammer-to-the-knee kick. For example, I go to the library a lot, and I have a routine about how I check the books out. Normally I'm thinking about other stuff. A few weeks ago there was a new librarian, and she was a bit slow and did things out of the normal order. She placed the books back where the librarians normally place them after the allotted period of time necessary to check them into the system, so I just picked them up and started walking off. She had to get my attention to stop me from leaving the building because I wasn't really thinking much about what I was doing. Obviously I don't have a hard-wired "library checkout" module in my brain, but the process is so set and pat for me that I'm quite close to being on autopilot. Now I've been on guard when this librarian is working, and she did it again a few days ago and I almost picked the books up and walked off again (I had to restrain myself).

In Why would anyone believe in God? Justin L. Barrett answers that exact question by utilizing banal cognitive truths. Though Barrett's treatment is more focused, the topics addressed and the hypotheses put forward draw deeply upon the research of cognitive anthropologists like Pascal Boyer (Religion Explained) and Scott Atran (In Gods we trust). Nevertheless, Barrett is more sympathetic to theism and less clinical than either of the two thinkers above, which makes sense, as Barrett himself is a theist who is currently employed with the Christian organization Young Life. This is relevant because it reiterates that Barrett is not decomposing a phenomenon which is to him alien and a curiosity, he is in effect explorating cognitive processes which are essential to the core of his own being.1

The two meta-concepts which undergird the ideas within the text (which is only a bit over 100 pages) are the human mental processes which result in hyperactive agency detection device (HADD) and theory of the mind (ToM). These cognitive traits are slotted under the umbrella of hard & fast response reflexive cognition. For example, when someone is gossiping with you you do not spend a great deal of time consciously threading all the personal interrelationships which might come into play as relevant to the various interactions being alluded to. Though there are reflective "Oh, yes!" rememberences (x is y's cousin), many of the cognitive operations are occurring under the hood and are encapsulated away from your reflective mind. This is why many extremely intelligent people are socially inept, they simply do not possess the requisite reflexive cognitive aptitude and so must rely on slow and klunky reflective cognition which simply can't keep up and is highly error prone. Unlike ToM, HADD is a little more obscure. In short, when people say they see design all around them, this is a reflective aspect of HADD at work. But why hyperactive? It is simply the classic false positve vs. true positive contrast. Going back to the example with the bear vs. the fridge, if it does turn out to be a bear you better be ready to run! If on the other hand it doesn't turn out to be a bear, you probably stressed yourself out more than is healthy, but you'll survive. Though there might not be an archetypical "environment of evolutionary adaptiveness" (EEA), I think it is safe to say that sophisticated mental models derived from assumptions about intraspecific and interspecific competition have always been extremely fitness relevant to homonids.2 If you think you saw a scout from an enemy tribe, or that your cousin is trying to screw you over by doing the nasty with your wife while you are out in the fields, it is often better to be safe than sorry. In other words, there might be strong adaptive value in seeing agency all over the place, even if that implies that there are lots of false hits ("only the paranoid survive"). Similarly, ToM is crucial in smoothing over cooperation and modeling the social networks which are relevant to your own fitness within the group, so going a little overboard might give you a reputation as a busy-body, but it might also save your ass if you know that an 'enemy' is trying to get you kicked out of your clique and turn you into a social pariah.3

So how does ToM and HADD intersect to help generate G-O-D? First, it is gods, not one God. "Higher" theism is a particular subset of the general tendency toward acceptance of supernatural agents which seems to be a human norm. I have a friend who works at a place which is dominated by non-Christian liberals, and yet they are convinced about the reality of astrology. When something good or bad happens they appeal to astrological explanations (i.e., the moon is full, Jupiter is ascendent, etc. etc.). They might not be Christian, but don't tell me they don't have religion. When things happen people have a bias toward attributing agency, someone was responsible, and better get credit or take the fall. Barrett gives the example of a friend who was stuck in an exploding grain silo. Somehow he was lifted 12 feet off the ground and went through a window and survived. He recalled that when he was stuck inside he muttered, "Take me home God," but he heard someone whisper "Not yet." His hypothesis was that angels had lifted him up and saved his life. What to think of this? I clearly am skeptical, nevertheless, it does sound like his survival was improbable. The first propane explosion should have killed him from asphyxiation according to the doctors who examined this individual. Additionally, it is not known that humans can leap 12 feet into the air. Flying through a window is also rather dangerous, but he didn't exhibit many scratches. What happened? The joint probability of the physical events that would have had to occur, i.e., the first explosion throwing him up and through the window, his physiology managing to tolerate proprane poisoning, etc., is low, but I conclude that that is what happened. Why? Well, I don't believe in angels, for a variety of reasons. The individual in question though did believe in angels, so it was entirely rational for him to think that angels might have saved his life (if you do believe in angels, what is more improbable, a succession of unlikely physical events or angels coming to your rescue?). Even if the individual had not believed in angels before the event, I am not convinced that he would not find angelic rescue a plausible hypothesis post facto. There is the common report that life-after-death experiences tend to strong biased toward the cultural background of the individual experiencing it. Americans see Jesus, Japanese see a Bhoddisattva, and so forth. No matter if they had strong beliefs a priori, their mind was probably interpreting the experience filtered through the ideas floating around in their culture. What is going on here is that the individual is engaging in abduction, reasoning to the best possible explanation that comes to mind for a given set of facts. The reflexive ToM and HADD all tell you that being lifted up and having your life saved is what agents do, not physical processes (rocks don't lift you up, people do, or super-people, supernatural agents). If you are not a believer in supernatural agents, you might consider that an extremely unlikely chain of physical events had occurred to save your life, but here you are weighing reflexive biases which tell you an agent helped you against reflective reasoning based on a conception of how the world works (i.e., angels don't come to save the day except in The Left Behind series). Consider something else, how many times do people attribute winning the lottery to their faith in God? We know pretty much how people win the lottery, it is a randomized process. Yes, your chance of winning is 1 in 100 million, but the chances are good that someone has the winning ticket, it isn't an unexplainable miracle. Or is it? The person who won must wonder "why me?" A naturalistic explanation doesn't offer much in terms of ontological "whys," reflectively you might wax on about the randomness of the quantum world as being the root of the non-deterministic nature of the universe, but that has little appeal for many with ToM and HADD is firing "something special happened! Someone cares! Someone intervened!" Shit doesn't just happen.

One could argue that the miracle of existence is amazing enough to start triggering ToM and HADD. But why supernatural agents? Because they have innate inferential richness, and you know people didn't make you, so by elmination it has be something supernatural. Saying that the universe emerged inexplicably and crossed the first Plank unit threshold doesn't tell you much, at least if you aren't a theoretical physicist. On the other hand, our ToM and HADD can automatically generate many inferences from the idea that a supernatural agent created the universe, that a supernatural agent cares about you. It is orders of magnitude more informative than "shit happens," and far less intellectual taxing and inscrutable than probability theory, and in the end, it offers the possibility of final ultimate answers which are intuitvely satisfying and generate a cogent and flexible model of the world.

Of course gods, even conventionally conceived, are mildly weird entities. That is, they often exhibit counterintuitive tendencies. Barrett labels them 'minimally counterintuitive,' to indicate that there are strange enough to remember, but normal enough to relate to. In short, they inhabit an optimum mental locus where their peculiarities induce awe, rememberence, and powers that make them relevant to our lives, but their conventionalities allow us to seemlessly generate inferences and 'explanations.' 'What would Jesus do?' is a much easier on-the-fly mental computation than 'What would Kantian ethics imply?' In the former case you simply mobilize ToM, put yourself in the position of Jesus (in a fashion) and slot in the general characteristics of that individual and his relation to the world around him, and you are good to go. In the latter case, it isn't so simple.

There is more to Barrett's work, but that is the broadest outline. He goes on to to explain why Abrahamic monotheism is appealing, the role of ritual and moral ethics in religion. These details are for another day. I will jump to the second to last chapter where Barrett addresses the question, why would anyone be an atheist? Seeing as how a majority of gnxp readers are unbelievers, I suspect this would interest us. One simple answer is that you have to go back to ToM and HADD. It might simply be that atheists usually have underdeveloped aptitudes in both traits. This has been suggested by others, and Simon Baron-Cohen has even hinted that autistics tend to be religiously unfocused and apathetic. In short, ToM is important in establishing a relation and relevance to supernatural agents, so if it is weak, there is a high chance that one can reject god hypotheses because they just aren't as intuitive as they should be, and they lack inferential power (i.e., just another irrelevant weird idea). Barrett points out that this can explain a curious cross-cultural fact: women are more religious than men on a host of metrics. Males might dominate the professional aspects of organized religion, but women are more reliable consumers. Barrett posits that this is a function of the fact that a greater proportion of males have underdeveloped ToM and HADD. That is, men are less concerned about the implications of agents and their relation to the networks of interaction that these agents engage in and in which they are embedded.

Barrett also lists off other factors which would bias someone toward atheism:

  1. Consider other explanatory candidates. In The Blind Watchmaker Richard Dawkins famously declared that Darwin's theory of evolution allowed atheists to be intellectually satisfied. Whether you agree with this or not, the general idea is that more powerful hypotheses to explain a given phenomenon (in that case, adapted species) can supersede the reflexive constructs generated by ToM+HADD+culture. For example, many people now discount that shooting stars are messages from the gods, because the physical process by which they occur has been elucidated.
  2. Minimize the outputs from the reflexive mind. For example, if HADD tends to go into overdrive because of concerns about safety at the hands of other agents, avoid a life where you are at others peoples' mercy so that you become habituated toward not being on alert for them, or at least concerned about their actions. The modern Western life, where even government welfare is sufficient to induce obesity, is an example of this, as dangers from interpersonal relationships (vendettas, retribution against your clan, etc.) are minimal, and from wild animals are non-existent. Additionally, immersing ourselves in an artificial urban world filled with exotic but explicable phenomena might also dull the tendency for HADD to get triggered excessively. Not only is everything around us obviously human generated (so not supernatural), but many of the objects behave as if they did have agency (a car with dark tinted windshields so you couldn't have drivers), but you know they don't.
  3. Reduce reliance on secondhand accounts. Barrett points out that in homogenous societies where common idioms and experiences are shared people tend to trust each other's veracity more. In a modern world filled with occupational, social and regional diversity, as well as fluid mobility, we tend to give less credence to local superstitions. In a world filled with novel experiences it is easy to discount eye witness accounts (I had a friend once tell me he saw a UFO in the hills above Pendleton, but I asked another friend about this and he noted this was a common assumption, but the US army had some searchlights going there a lot, so that's probably what my other friend saw).
  4. Fixate on reflective thought so that you discount reflexes. This is pretty straight forward. If you have read Critique of Pure Reason a dozen times, 'What would Kantian ethics imply?' might be a much more relevant starting point. If you are embedded in a controlled world where everyone else privileges detached rationality and distrusts 'common sense' and 'instinct' than you have internalize a reflexive anti-reflexiveness.

I think it is clear that to be not religious is a phenotype that is probably the outcome of a host of factors which load the probability for or against. Consider American scientists. The most elite, members of the National Academy of Sciences, seem to exhibit only a 1 out of 10 chance of being theists. This is highly atypical. And yet if you look at the conditions above, scientists probably come close to meeting many of them. They tend to be prosperous, urban, live in artificial conditions and fixate to a far greater degree on counterintuitive reflective cognitive states than is the norm. To some extent many scientists wouldn't be scientists if they trusted 'common sense' to the first order, their models are often technical because they are intuitively opaque, and the instrumentation they use often yields up bizarre findings in regards to the fundamentals of the universe around us. The general Ph.D. population is far more religious than NAS members, but they are likely far less monomaniacal in their scientific career. Scientists themselves fuck up constantly and have problems thinking outside the common sense box, especially out of their field. That is why the social system of science is so important to dampen and restrain individual biases and confusions. But I think the most important reason that scientists are less religious is that they simply discount common sense assertions based on gestalt intuition in many areas where others do not because it would be professionally detrimental if they were the type of people who did not explore questions that seemed a priori open to introspective understanding.

Because the reflexive mind tends to be encapsulated and detached from the reflective conscious, the reasons that people give for opinions and decisions are often fabulations that derive from the first thing that pops into their head, or a socially agreed upon convention. I recall that when I was younger Creationist friends would point to a tree and assert, "Look at the Design, of course I believe!" Now, I doubt that the tree was really that important, but it was a token that had gained currency and could be offered as a reason for their inner computations. There is research that suggests that children from non-fundamentalist homes are just as prone to thinking in "kinds" and in a Creationist fashion as fundamentalist children. The reality is that as we grow many of us have to shed our intuitions, our reflexive minds, and put faith in reflective paradigms which we might not fully understand. Ultimately, my rejection of the angel hypothesis for any given scenario is not predicated on that specific scenario, and I am willing to remain an agnostic in how particular miraculous events occur without angels, because I attach great weight to a broader contingent system of how the universe works. That contingent system is only loosely, and not necessarily, tied to my intuitions.

One thing that Barrett's book does bring home though is that science fictional works which depict an atheist future might be just a little less unrealistic than I had thought, because if civilization continues, and our species remains as it is, it seems likely that our day to day world will be more human designed than natural. The reality of design by humans will obviate the need for supernatural agents, because abduction will have a clear best working culprit for agency, our own species. This does not imply that everyone will be an atheist, but it seems plausible that there will be a withdrawl of God from the world that is entirely man-made, as a time might come when someone within a space colony might point to a tree, and wonder at the beauty of genetic engineering which generated it de novo with tailor-made DNA sequences which preadapted it to low g environments....

Related: Inducing disgust, We are born Manichaeans, Reflections on the God Module, Theological incorrectness.

1 - Barrett's undergraduate degree is from Calvin College, the premier Reformed institution of higher education in the nation, which implies he is likely a philosophically serious Christian (Wheaton is more ecumenical and I think less focused on systematic theology). There are strong clues within the text that he is a theistic evolutionist who finds Dawkinsian atheism and its intolerance unacceptable.

2 - Some species, for example herbivores, tend to be constrained mostly by environmental and resource related factors, i.e., drought decreases feed. Cows are basically grass processing machines and they don't have to worry about 'competing' for grass when it is plentiful because it is abundant but low quality in character. The race isn't against other cows, but against the reality that the metabolic and microbial processes needed to extract nutritive value out of grasses is a close thing. In contrast carnivores and omnivores engage in a great deal of competition within and between species, for obvious reasons. Humans are more like the latter than the former, ergo, they better be aware of the agents all around them in a sophisticated fashion.

3 - Being a social pariah is a matter of self-esteem today, but in the past it might have been life or death in communities where variance in hunting and gathering within families was dampened by redistribution within the clan or village from those who were at surplus. In short, without a social network the next famine could kill you.