Friday, July 27, 2007

'The Cancer of Superstition' II

First up, a big thank you to GB Steve (one of the moderators over ay who informed me in a comment to my last entry (which I have since revised due to the number of grammatical errors it contained)that he has syndicated the blog's RSS feed to a group of his friends. Steve also raises a question about my previous claims regarding the 'naturalness' of religion, providing me with the opportunity to do something I've been meaning to do for a while here: give a somewhat detailed overview/summary of current cognitive theories of religion - a set of theories which I generally subscribe to. Again, this is a far cry from the usual Lovecraftian goodness we like to provide you with at Ghooric Zone central. That said, I'll reiterate the point that the following fits generally with Lovecraft's own materialism and anti-supernaturalism, and in fact is a contemporary take on Lovecraft's own views on religion as expressed in some of his letters and in parts of 'The Whisperer in Darkness'. Also, some of the ideas contained herein will probably inform later ramblings dealing with the Lovecraft-occultural interface. That said, I'm pretty busy with a number of other projects at the moment (painting miniatures as well as writing my book on Lovecraftian occultures!) so although this entry will be far from brief, it is not going to be thoroughly or properly referenced or contain footnotes, etc.

The 'Naturalness' of Religion
First off, I think Steve is quite right to question the notion that religion is 'natural' because to assume such is the case means that us atheists are somehow 'unnatural' or 'abnormal'! In any case, such a claim is something that definitely requires clarification. Here I’m following Pascal Boyer (see his book Religion Explained) who refers to the ‘naturalness of religious ideas’ in the following sense: it is not so much the case that it is ‘natural’ to believe in the supernatural because of how our brain is wired; rather it is the case that religious concepts can come to seem ‘natural’ and acceptable because the ability to conceptualise the supernatural is an offshoot or by-product of the very ordinary, mundane, and ‘natural’ but often unreflective aspects of our day-to-day cognitive functioning (what Boyer calls our ‘ontological intuitions’). This is an important distinction - that religion is a by-product but not a function of how our minds work. Hopefully this will be further clarified at various points throughout this post. In fact, Boyer and some of the other theorists I'm going to be talking about below are perhaps less concerned with why people believe than with the cognitive capacities that enables the human mind to conceptualise and entertain ideas about the supernatural in the first place.

Also I'm not convinced that it is always fruitful to encourage people to critically review their beliefs on the basis that said beliefs are irrational (even though I perosnally believe this to be the case). In part, this is probably down to my background in anthropology which emphasises sensitivity to other 'ways of seeing'. In this respect I think Dawkins 'evangelical' evolutionism tends to do more damage to his own argument than is necessary. This is because all of us in all sorts of ways apply and demonstrate counter-intuitive and counterfactual ‘magical’ or 'religious' thinking on a daily basis: for example we probably all unreflectively and implicitly attribute human-like characteristics to non-human things or inanimate objects (getting annoyed with your computer when it crashes, or with your car when it won’t start). As Marx also pointed out, under capitalism people tend to attribute organic properties of self-replictation to money (bank advertisements talking about ‘letting your money grow’ and so on), obscuring the actual exploitation upon which the accumulation of capital depends.

Returning to the important distinction above, this doesn’t of course mean that we actually come to believe that money is a living, breathing entity, simply that we are capable of thinking in that way, and that the ability to think like this may be the consequence or by-product of a helpful survival strategy by which we deal more effectively with our environment by projecting human-like qualities on to it (partly a consequence of the complex forms of 'social intelligence' which humans possess). In the theory I’m going to summarise below, this propensity for anthropomorphism is (according to one anthropologist at any rate) the foundation upon which all complex religious doctrines and supernatural beliefs are built. Importantly, its root cause may be the evolved but ‘hyperactive’ ability humans possess for detecting predators.

I'd also like to emphasise the fact that religion is not being treated here as a functional aspect of how our minds work - humans have not 'evolved' the capacity for religion because it is 'useful'; rather, religion is a spandral or by-product of our evolved cognitive architecture. Which is to say that humans beings, once they 'have' religion, can put it to all sorts of uses (engendering social cohesion, for example). However, religion, as a by-product of 'normal' brain functioning, is not in itself inherently useful.

Neurotheology, the God Module and the 'Function' of Religion
A counter-argument to this is sometimes referred to as the 'committment' theory of religion, or more popularly as the 'God Module' theory: a ‘hardwired’ or ‘neurotheological’ view of religion which I am suspicious of. According to two of its main exponents, Eugene D’Aquili and Andrew Newberg, humans have evolved as part of their brain structure a ‘God Module’. In their 1993 article ‘Religious and mystical states: A neuropsychological model’ in Zygon: Journal of Religion and Science) D’Aquili and Newberg have also argued that various religious practices (meditation, trancing, etc.) generate neurological changes, including an increase in activity in the frontal lobes. It also leads, they claim, to decreased activity in the superior parietal lobe (the part of the brain which generates an awareness of the physical limitations of the self)which in turn leads to a sense of transcending material existence, or a feeling of unity with god, oneness with the universe, etc.
D’Aquili and Newberg go on to argue that experiences such as these are more likely to be generated in ritual context, where they create a feeling of unity and connectedness with others. As such, the ‘god module’ may be an evolutionary adaptation to enhance social co-operation. In effect, neurotheology has tried to place Emile Durkheim’s claim that religion is concerned with social cohesion on a neurological footing. This is perhaps a useful way of understanding how ritual functions to facilitate group bonding; this 'neurotheological' approach also suggests that religion is an ‘illusion’ created by neurological processes. However, it also leaves open the doorway for ‘intelligent design’ for the more savvy of creationists: namely that god has used evolutionary processes to open a neurological doorway to communion with he/she/it (the fact that god tends to be gendered is an important point to which I’ll return later). It also posits a hardwired view of religion that does, again, implicitly suggest that us atheists are not functioning properly.

The prevailing view (at least as I see it) is that whilst religion is an outcome of evolutionary processes, it is not an evolved adaptation per se. Steven Pinker has, for example, argued that:

‘Many of our faculties are adaptations to enduring properties of the real world. We have depth perception, because the world really is three-dimensional. We apparently have an innate fear of snakes, because the world has snakes and they are venomous. Perhaps there really is a personal, attentive, invisible, miracle-producing, reward-giving, retributive deity, and we have a God module in order to commune with him. As a scientist, I like to interpret claims as testable hypotheses, and this certainly is one. It predicts, for example, that miracles should be observable, that success in life should be proportional to virtue, and that suffering should be proportional to sin. I don't know anyone who has done the necessary studies, but I would say there is good reason to believe that these hypotheses have not been confirmed’. ( media/2004_10_29_religion.htm)

Scott Atran also suggests that god module theory is profoundly flawed: the claim that religion offsets apparent disadvantages is countered by the fact that it creates other disadvantages. Thus most theories that have been posited to explain religion in functional terms can be countered by showing how religion also does the exact opposite:

‘It tries to answer the question Why? It prevents answers to the question Why? It creates meaning for an arbitrary world. It postulates and imaginary world that hides reality’s reason. It discovers the origin of nature’s regular occurrences…It disguises the origin of nature’s regular occurrences…It relieves anxiety. It terrorizes…It aims to overcome evil, suffering, misfortune and injustice among believers. It aims to cause evil, suffering, misfortune and injustice among nonbelievers…It benefits elites. It benefits the downtrodden…It’s the workhorse of war. It’s a player for peace’ (Atran, 2002. In Gods We Trust, pp. 6-7).

In their joint article ( Atran and Norenzayan go on to claim that whilst ‘commitment’ theories of religion like the 'god module' are useful for understanding how religion can faciltate the kinds of non-kin-based and non-reciprocal altruism necessary for the survivial of human grousp, they are not sufficient explanations of religion in and of themselves. The problem being that such theories don’t account for the apparently universal belief in supernatural beings: the existence of Marxism and humanism as non-religious moral systems suggest non-religious moral systems that do not invoke supernatural agents can be sufficient to engender human co-operation. As such, commitment theories are incomplete explanations of religion because they fail to account for the specificity of its supernaturalist elements.

The approach outlined below is, I think, incredibly useful in helping to demonstrate how and why people potentially acquire religious concepts by understanding religious thought not as 'innate', 'natural' and necessary but in naturalistic terms which are explicable via material processes and by a secular scientific understanding of the universe. In brief, this approach states that religion can be both extremely useful to humans and highly damaging, but in either case it is not necessary for humans to function and survive. This is also sometimes refered to as the 'spandrel' theory of religion: a spandrel being an architectural term for describing an aspect of a builing which was not designed with a function in mind, but is a by-product of the functionality of some other aspect of the building (the space under the stairs is often used as storage, but it was not designed as a storage space; rather, it is necessary to have a space in the building at that point as a consequence of the stairs).

The 'Spandrel' Theory of Religion
An initial step in understanding how people are capable of conceptualising the supernatural begins with Theory of Mind. Theory of Mind(ToM) is the remarkable ‘mind reading’ ability that most humans possess which allows us to impute motivations to others. By ‘mind reading’ I am referring to our apparently natural aptitude for inferring what other people are thinking and feeling and predicting their behaviour accordingly, based on visual cues, body language and the like rather than any kind of genuine ‘psychic’ ability! Another key definition of ToM (which I will return to in the epilogue to this piece) is the ability to detect false beliefs in others (I think that she thinks so-and so, even though I know so-and-so to be false). In fact, this is how the 'existence' of ToM is generally demonstrated: if you show a child of around 5 years of age a box of sweets and ask that child what they think is in the box, the will usually say 'sweets!'. Then you show the child that the box actually contains pencils. A second child is brought into the room. The first child is then asked what he or she thinks that the second child thinks is in the box. A child who demonstrates ToM will reply sweets, even though they know it contains pencils. What is demonstrated here is the ability to think about what other people are thinking, even if you know that the other people's beliefs about the world are mistaken.

All in all, ToM is a pretty remarkable ability, but one that we take for granted. This kind of slippage towards taken-for-grantedness is, I think, important in understanding why people are often likely to accept religious ideas and supernatural beliefs (i.e. the dominant beliefs in most cultures have a taken-for-grnated character). This is probably because ideas and beleifs about the supernatural emerge from very banal taken-for-granted cognitive capacities that most of us possess.

With regard to religion, the important thing about ToM is that it allows us to impute a seemingly invisible motivational force to other human beings (‘mind’). We are used to dealing with other humans who have corporeal bodies but are also agents with recognisable motivations, feelings etc. When we encounter a dead body, we are suddenly faced with the corporeal body that is lacking agency and a motivating force. As Pacal Boyer has it, it is then only a small step to inferring the existence of this ‘invisible’ and seemingly non-corporeal mind as something which has ‘left’ the body, and which is, indeed, independent of the body. ToM is one factor that allows us to hypothesis the existence of human-like but non-corporeal entities (i.e. gods, spirits and demons). However, ToM does not demand beleif in said hypothetical otherworldly agents.

The complex kinds of ‘social intelligence’ demonstrated by humans (which allow us to establish massive coalitions such as nation states or global religious communities)are fundamental to our survival (which, for humans is massively dependent on social co-operation - especially in terms of parental investment for human infants - even though humans still remain the main predators of other humans). As such, the social intelligence we possess presumably spread through populations in our evolutionary past as a consequence of its usefulness. However, one of the things about social intelligence is that we simply can’t ‘switch it off’. A consequence of this is that we tend to maintain social relationships with the dead (visiting graves, etc.): especially with members of our immediate family or close friends (with who we have shared deep emotional ties and who, presumably, we depended on in various ways for our daily well-being) even after they have died.

Added to this the human ability for anthropomorphism which cognitive archaeologist Steven Mithen claims is probably a result of ‘cognitive fluidity’. Cognitive fluidity is tied to the idea of ‘modularity of mind’: that the human mind is not some big general all purpose learning tool - if it was, we probably wouldn’t be able to communicate meaningfully via language. The fact that we have language is, according to the 'modularity of mind' theory, a consequence of our having developed a specialised ‘lingusitic intelligence’ mental module. A useful analogy here is to think of a non-modular mind as being like a toolbox containing some nails, a hammer, a saw, a screwdriver and a wrench. There’s lots of stuff you can usefully do with all of this equipment (cut down a tree and build a shelter), but the toolbox has rather limited application when your computer breaks down. For this kind of repair job, you need a specialised toolkit. Understanding weather patterns may be important for successfully undertaking activities like hunting and farming, but the weather doesn't really impact upon the way we learn language. However, the claim is that if we don't possess modularity of mind, we simply would not be able to realise what factors are important when it comes to hunting, and what are important when it comes to learning language. The net result being that it takes much longer to learn how to farm and to learn how to speak a language. Human children appear to be born with a 'technical intelligence' module which means that even very young children find it easy to learn basic tool use. However, the same does not appear to be true of infant chimps who often take a very long time to master using rocks in a simple hammer-and-anvil technique to break open nuts.

What has this to do with religion? What humans also seem able to do think across these mental modules or domains of knowledge - i.e. cognitive fluidity. The dates remain contentious, but this may have been one of the major evolutionary shifts that brought about the 'cultural revolution' or explosion that seems to have taken place somehwere around 70-40,000 years ago. Generally speaking, this period also evidences the emergence of religious thinking amongst anatomically modern human beings (burial sites, rock art, etc.). In any case, cognitive fluidity allows for analogical thinking which is an incredibly useful tool. Not only does the ability to think analogically enable technical development (thus greater human control over the environment and increased chances of survival), it also seems to underlie a great deal of scientific thinking (looking for universal principles of motion by comparing a birds wing to a fish’s fin).

Cognitive fluditiy also means that when social intelligence (the part of the mind we use for dealing with other people, i.e. ToM and the like) interacts with natural history intelligence (the part of the mind we use for dealing with, thinking about and categorising our environment) we can come up with concepts such as an animal who thinks and acts like a human. This can be an incredibly useful way of interacting with one’s environment, and it certainly seems to be the case that human success in hunting comes from attributing human-like qualities to prey. However, it also allows us to attribute a human-like mind to a tree or a statue. In brief, this ability for analogy and anthropomorphism may have tangentially given rise to the ‘earliest’ kind of religious concept: animism. To reiterate, cognitive fludity does not require religion; rather, religious concepts (along with art, poetry, etc.) are by-products of the evolved capacity for cognitive fluidy (which in and of itself is a useful day-to-day survival strategy).

In enabling symbol use, analogy, metaphor, etc. (‘the wine dark sea’) cognitive fluidity also faciltated the development of complex forms of communication that have also allowed humans to establish and maintain equally-complex forms of social networks and relationships previously mentioned. Things like nation states and religious communities are often dependent on ideas about 'fictive' or 'metaphorical' kinship. To the extent that symbol use, analogy and metaphor are also central to the development of various artistic forms, I wonder if the links between cognitive fluidity, anthropomorphism and animism indicates why, for much of the time we've been on the planet, human artistic endeavours (whether literary, visual or plastic) often deal with religious or supernatural themes. This is also one of the reasons why I think that there is a significant overlap between contemporary occultism and the literature of the fantastic. However, one thing I’m not claiming here is that art, poetry, etc. are inherently religious pursuits.

Tying all of this in a roundabout way to the racist invective we find in Lovecraft, Steven Mithen also makes the interesting point that prior to acquiring cognitive fluidity, humans simply may not have been capable of racist thinking: while cognitive fluidity allows us to think of animals as being human-like, it also allows us to conceptualise other humans as animal-like or less-than human (this doesn’t, of course, mean that animosity between humans wouldn’t have existed prior to cognitive fluidity; but according to Mithen such animosity wouldn’t have been given expression through the discourse of race).

As a consequence of ToM, social intelligence (what Boyer generally refers to as 'intuitive psychology'), religious concepts, universally, seem to be constrained by a propensity toward anthropomorphism. The anthropologist Stewart Guthrie (1993, Faces in the Clouds: A New Theory of Religion) reiterates this in claiming that the most efficient interpretation that human beings can place on events in an uncertain world is to assume that those events are structured by a human-like intentional agency: that shape in the distance might be a boulder, or it might be a bear. It is better to assume that it is a bear if you want to ensure your survival. If it turns out to be a boulder, you've lost nothing. If you think it's a boulber but it turns out to be a predator, you're in trouble. In this respect, anthropomorphism may be relate to an evolved adaptation which (if I recall aright) psychologist Justin Barrett calls the 'Hyperactive Agent Detection Device' (HADD) - a biased psychological-perceptual capacity for ‘over detecting’ potential predators. Like social intelligence, this is 'hyperactive' because we simply cannot turn it off, and in the bear-boulder example above it is a useful survival strategy t to start from the assumption that the boulder is a predator. The HADD kicks alerts us to potential danger when, for example you hear a floorboard creaking downstairs at night, or you hear a rustling behind you during a walk in the woods. The floorboard may be creaking because of temperature changes, and the rustling behind you might just be the wind, but our intuitive response is to react as if a potential predator is in the vicinity. The implications being that the HADD often leads us to detect for agents where none exist. In conjunction with ToM and so forth, this can add weight to the notion that non-corporeal human-like agents exist. The reason why it is human-like spirits that are posited is because the main predator of human beings has been other human beings. Thus the HADD is more likely to lead us to infer human-like predators rather than animal-predators.

This returns me to Tylor's minimal definition of religion as a 'belief in spiritual beings' mentioned in my last post: doctrinal aspects of religion do not necessarily reflect what people think at the 'grass-roots' of religious belief and practice. Christianity holds that god is atemporal, omnipotent and omnipresent; however, when praying people often make different inferences about the nature of god - that 'he' experiences time like other human beings, and that he is specifically focusing on the person praying. In European Christian art, God is often represented in gendered, corporeal, tangible and anthropomorphic form (i.e. a bearded old man enthroned in heaven); in the Bible, God also has human attributes (i.e. being jealous and wrathful). Similarly, Justin Barrett found that many Hindu’s applied anthropomorphic qualities to deities in everyday situations, even though when questioned about the nature of deities in a more formal setting would recapitulate formal, learnt theology that claimed otherwise (J. Barrett, 1998. ‘Cognitive Constraints on Hindu Concepts of the Divine’ in Journal for the Scientific Study of Religion). What this indicates is that human religious concepts are, to an extent, constrained and are ‘minimally counterintuitive’. While cross-culturally gods and spirits may be conceived of as all-knowing, there are innumerable instances in the myths of various cultures where humans are able to trick supposedly omniscient gods and spirits.

The Counterintuitiveness of Religious Ideas
Similarly, Pascal Boyer argues that supernatural concepts seem to follow a cross-cultural template, and the nature of such concepts appears to be constrained by existing intuitions about the world which are either innate or which humans acquire early in childhood. These ‘ontological intuitions’ form the basis of what anthropologists refer to as ‘folk knowledge’ - namely implicit assumptions about the way the world is ordered. In brief, humans recognise distinctions between animate objects, inanimate objects, natural kinds, artificial objects, and persons. For example, if we encounter an animate object we have never encountered before, we are able to quickly infer a number of things about it: it has four legs, fur and moves, so it is probably an animal. As an animal, we infer that it is mortal, eats to live, reproduces, etc. We are so used to this that it seems unremarkable, but without these pre-existing ontological intuitions which allow us to cognitively structure our environment in relatively coherent ways (and which are presumably a product of long evolutionary processes) we would have difficulty in recognising that a rock is of a different kind to a fish.

In any case, Boyer what is characteristic of religious representations is that they violate these ontological assumptions in some way. The reason why religious representations become widespread is because they disrupt out ontological intuitions, and as such become more salient or memorable than other kinds of knowledge or ideas, and thus more likely to be passed on. Again, this doesn';t mean that such ideas will necessarily be believed, but it suggests a reason why mental representations of the supernatural become widespread through populations.

However, the implication of this claim is that the more 'outrageous' or ‘counter-intuitive’ a religious idea is, the more likely it is to be remembered and transmitted. Boyer suggests otherwise: a balance is usually struck in religious concepts between being counterintuitive enough to be memorable and being so counterintuitive that the concept no longer sensibly fits with out intuitive understanding of the world. For example, the belief in ghosts, spirits and deities fits into our intuitive assumptions insofar as these entities are classified as persons (sentient beings that have intentions, desires and motivations); however, they are also counterintuitive according to our folk concepts of biology, psychology and physics (they can move through walls, see everything, etc.).
On the other hand, a chair that exists only on Thursdays or (to borrow Justin Barrett's example) cat that never dies, has wings, is made out of steel, experiences time backwards, lives underwater and speaks Russian is so counterintuitive to our concept of a cat that it could no longer be classified as a cat. It is certainly so outside our experience as to be wholly unbelievable. In brief, only concepts which are counter-factual and counter-intuitive in a limited sense are likely to become transmissible and culturally salient. Beliefs that are ‘excessively’ unusual or violate our ontological assumptions to the extreme are likely to be rejected as unbelievable. As an aside, this is something I hope to revisit with regard to representations of the Old Ones in Lovecraft. Following China Mieville's comments at the Weird Realism conference about the overdetermined nature of Lovecraft's descriptions (despite his use of adjectives such as 'indescribable'), I actually think that Lovecraft's representations of the Old Ones are much more anthropomorphic than is often assumed.

In summary, ideas about gods and spirits - which seem to constitute the 'core' of religious and supernatural beliefs cross-culturally - are 'universal' because they only require a slight modification of Theory of Mind, because of social intelligence and because the posited HADD encourages a belief in non-corporeal persons or agents. However, all of these things together produce religion as a by-product of cognitive functioning, but the production of religion is not the function of these ordinary (although in some senses remarkable) human abilitites. In brief, religion is not an evolved adaptation necessary for human survival or co-operation. Human beings can do quite well without it.

That said, Scott Atran does see scope for incorporating 'committment' theories into a more general understanding of the evolutionary contours of religious thought. Earlier I mentioned that Theory of Mind enables humans to recognise false beliefs in others. This also allows us to represent and communicate counterfactual or false beliefs to others. The capacity for deception is useful for our survival, but as social beings it is also a potential threat to the social order. Scott Atran points out that 'religion often involves hard-to-fake public expressions of costly material commitments to supernatural agents, that is, offering and sacrifice (offerings of goods, property, time, life). A general problem in the maintenance of cooperation is how to distinguish people who are altruistically committed to a coalition from those who are not. One way to test who's genuinely committed is to see who is willing to undertake a costly sacrifice. Painful ritual practices such as tattooing, scarification and circumcision are not the kind of thing that anyone would do unless they took their affiliation with the group seriously.'

Our ability to recognize potentially false beliefs (including supernatural beleifs) also means that most human societies are always under threat of defection. If some better ideology comes along there is no longer any reason to accept the current ideology. Once human beings have the ability to conceptualise supernatural agents, such agents become incredibly useful in encouraging people not to defect from the group, especially if this supernatural agent is believed to possess the power to invisibly check and review people's thoughts and behaviours, and can punish them accordingly. The 'social usefulness' of the supernatural (despite the fact that this can also be incredibly damaging on a personal and individual level) may also be an important component to understanding why religious moral systems are more widespread than non-religious ones, despite the fact that (as Atran himself seems to recognise) non-religious moral systems can also be used to facilitate social co-operation. On this point, I am not advocating the necessity of religion; rather, I'm pointing to a factor that might be important in understanding the social contexts underpinning people's 'preference' for supernaturalism and resistance to atheism.

Here endeth the lesson.


  1. You don't do short, do you?

    In "The counterintuitiveness of Religious Ideas" you talk about ontological intuitions, about how we infer classification. I'm not sure I'd agree with Boyer about how religious representation violate these assumptions. I think that would be the case if the assumptions were limited to "objective reality" but they are clearly not. Our culturally-bound ontological assumption cross many categories. Dogs, for example, food for some, child substitues for others.

    I just don't think outrageousness is a proper yardstick given that most people's beliefs include a fair amount of non-religious outrageousness.

    And getting back to HPL, in Mountains of Madness, talking of the Old Ones, he says, "Radiates, vegetables, monstrosities, star spawn - whatever they had been, they were men!"

  2. Hi Steve,

    I probably haven't done justice to Boyer's work, but certainly his notion of 'ontological intuitions' is that they are innate aspects of mind that are principallu concerned with 'objective reality', and are thus somewhat different (though maybe not that much so) from culturally-derived sets of classifications. The latter, I would argue, are more about communicating important social messages and as such are not usually taken as objective mirrors of reality.

    For Westerners, a dog may be treated as a 'symbolic' human being (i.e. pet), but there is no fundamental confusion between dogs and humans. A dog is still intuitivesly recognised seen as being of a different 'living kind' to human beings. As you mention in your comment, they may be treated as a child substitute; however, dogs are rarely mistaken for actual human children! In fact they are still perceived in the same way that dogs are perceived as different living kinds to humans by those cultural/ethnic groups that treat dogs as a category of food. In brief, for Boyer ontological intuitions of this kind (which I think he refers to as 'biological intuitions' and what other theorists of modularity of mind refer to as 'natural history intelligence') is the kind of thing that simply allows us to differentiate animals/living kinds from inanimate objects (rocks, etc.), and to differentiate, say, mammals from reptiles, birds from fish and so on. It is also the 'innate' structuring knowledge that allows us to identify an animal that we have not seen before as a bird, fish, species of dog or whatever. There are a few instances in other cultures of animals that are treated as classificatory anomalies (i.e. mammals that may be classified as repitles or even fish), but I think that Scott Atran has elsewhere marshalled good evidence to support the claim that, cross-culturally, people do tend to classify animal groups in very similar ways (sorry, I don't have the reference close to hand). And this would be the level at which Boyer's ontological intuitions work.

    The symbolic attribution of personhood to dogs is something that probably only comes with cognitive fluidity, whilst from an eolutionary standpoint our ontological intuitions are probably prior to this. Examples such as the dog (alongside the famed taboos on eating pigs in Judaism and Islam and cows in Huinduism)may well be tied to environmental and material necessities, but may not have arisen until human beings were capable of forming complex social groups that were facilitated by symbolic thought rather than pure force. In this case, these cultural classifications have played and continue to play a significant 'symbolic' role in defining group boundaries and limiting social defection from one group to another. If one group treats dogs as symbolic (but not actual) humans, then groups who eat dogs are probably going to be viewed as 'disgusting' because they are 'cannibals' (symbolically speaking) in the eyes of dog lovers.

    I would also say that for Boyer outrageousness/counterintuitivenessisn't the principal yardstick of religion; rather it is one of a number of factors that helps shape the 'religious' or supernaturalist character of certain ideas. Indeed, Boyer argues that there are aspects of science that are counterintuitive (the difference being that science isn't counterfactual, whereas religion is both counterintuitive and counterfactual): the notion that the earth travels round the sun, for example, goes against 'commensense' experience of seeing the sun travel through the sky relative to ones own position. Perhaps an example more relevant to the here and now would be some of the more 'way out' theories hypothesised within quantum mechanics.

    I think that this, again, is quite an important point because the fact that counterintutiveness is found within different domains of knowledge and sets of ideas (not just those classified as 'religious') generally supports Boyer & Atran's view that 'religion' is not a special category of human thought, but is a by-product of extant cognitive processes more generally. I gues there is a significant political point here in so far as it de-emphasises the 'specialness' of religion.Boyer seems to be suggesting, then, that minimal counterintuitiveness makes certain ideas seem salient. (This is an idea also explored in the work of another cognitive anthropologist Dan Sperber in his book 'Rethingking Symbolism'). As such, counterintuitiveness is important in understanding the cultural spread/transmission of religious ideas via there 'memorableness'.

    HPL's description of the Old Ones was precisely what I was thinking of in my post - I intend to pursue the popularity of the mythos (probably with special reference to its co-option within contemporary magic) with regard to Boyer's ideas about 'minimal counterintuitiveness' in a later post. I think the trick that Lovecraft pulls off magnificently is pursuading us that his monstrousities are beyond human comprehension by describing them anthropomorphically and in great detail!