‘Big Gods’ book club #2: Analytic atheism and the puzzle of apologetic

(This week, cognitionandculture.net is hosting a “book club” webinar discussing Ara Norenzayan’s latest book [1] with its author. This précis introduced the discussion.)

Norenzayan’s Big Gods presents an elegant account of how belief in super-knowing, powerful and morally concerned gods emerged as a result of normal individual cognitive processes and cultural selection. Any comprehensive account for the success of religion should be able to explain the wide distribution of atheism, which is currently the fourth most popular ‘religious’ outlook. Norenzayan proposes four roads to atheism: atheism caused by deficits in desire-belief reasoning, indifference towards religion, lack of exposure to public displays of religious acts, and analytic atheism.

Here, I will focus on analytic atheism and present Norenzayan with a challenge for his suggestion that analytic thinking counters religious belief. If analytic thinking decreases religious belief, how can we explain the persistence and cultural success of philosophical theology, which consists of analytic reasoning about God/the gods?

Briefly stated, analytic atheism is the state of mind when one actively reflects and deliberates, and comes to religious disbelief “Religious disbelief arises when there are opportunities to revise or override the cognitive biases that support religious intuitions” (from the précis). Norenzayan endorses the view—standard in cognitive science of religion—that intuitive thinking mechanisms, such as mental state attribution, push our brains towards religious beliefs. Analytic reasoning, by contrast, encourages believers to revise or override the cognitive biases that support religious intuitions. Norenzayan cites several experiments (conducted by himself and others) that demonstrate a causal link between analytic reasoning and decreased belief in God. Even just being primed with this style of thinking, as e.g., in a presentation of Rodin’s Thinker – has this result. On the basis of this, Norenzayan makes the following interesting prediction: “apologetics is doomed to failure as a philosophical enterprise because it fails to capture how our minds accept the plausibility of religious belief.” (p. 181).

However, I am not convinced that the evidence is compatible with this prediction. Many societies that worship Big Gods have developed sophisticated theological systems that include religious arguments, such as the European middle ages and early modern period, the Arabic Islamic world and classical Hindu theology see e.g., here [2]. Today, the interest in reflecting on religion is widespread, with a large output of philosophical theology in academic and more popular writings. How can we explain this anomaly? I propose four possible reasons:

1. The association between analytic thinking and atheism is the result of culturally contingent factors, rather than of a cognitive incompatibility between theism and reflective reasoning. While current westerners may see religion as opposed to reason, this is not how medieval theologians saw it (they spoke about “faith seeking understanding”). An author like Aquinas attempted to make his theology compatible with the best natural philosophy of his time. Today, as a result of fundamentalism in Christianity, Islam and other Big God religions that have sacred scripture, science and religion are often pitted against each other. To examine whether the opposition between religious and reflective reasoning is peculiar to current western culture, more research needs to be done with people who are not WEIRD (i.e., not from western, educated, industrialized, rich and democratic nations). Do people from non-western cultures also show decreased belief in God when they reason analytically, or when they are primed with analytic reasoning? It would be important to know this to establish whether analytic atheism is in fact a cause of atheism, rather than a culturally contingent byproduct.

2. Supposing there is a real cognitive tension between analytic reasoning and religious belief, why then do religious arguments occur so often? To explain this, we might have to look at the social context of religious argumentation. Very rarely do religious beliefs have an absolute monopoly. According to Mercier and Sperber’s social account of reasoning, reasoning has a primarily social, argumentative function. If a religious belief system has to compete with other belief systems, it might be more viable if its adherents can present compelling reasons for why the beliefs are rational.

3. Apologetics, while using the tools of reason, paradoxically often encourages people to hold religious beliefs without further reasoning. For example, a flourishing branch of recent philosophy of religion is so-called Reformed epistemology. Reformed epistemologists argue that people are rational to believe that God exists, even if they have no evidence for their beliefs, just as they are rational to hold other “basic” beliefs, such as that the world exists. This may explain the recent popularity of Reformed epistemology in Anglosaxon theology. In effect, Reformed epistemology says it’s OK to go with the intuitive, unreflective cognitive processes as belief in God is concerned (see here).

4. Apologetics uses reflective reasoning, but it does depend on intuitive reasoning to supply content for its premises. Think about the design argument, which infers the existence of an intelligent designer from the design of the natural order. Or consider the cosmological argument, which assumes that the universe must have had an eternal and personal cause. The premises that underlie such arguments (e.g., purposive design requires a designer, a temporal thing that exists requires an external cause) resonate with intuitions that very young children already hold. As I will be arguing in a book forthcoming with MIT Press, A natural history of natural theology, religious arguments often appeal to our intuitive reasoning processes, as well as to the reflective ones.

These reflections on apologetics encourage us to rethink the relationship between atheism and reflective reasoning. How universal and cross-culturally robust is this relationship? What makes some intelligent, reflective people engage in philosophical theology, and what are the cognitive dynamics underlying this enterprise? Ultimately, I think that a better understanding of this will foster a better view on the cognitive factors that underlie atheism and theism.


[1] Norenzayan, A. (2013). Big gods: How religion transformed cooperation and conflict. Princeton University Press.

[2] Brown, C. M. (2012). Hindu perspectives on evolution: Darwin, dharma, and design. Routledge.

7 Comments

  • comment-avatar
    Hugo Mercier 10 December 2013 (07:04)

    Thank you for this interesting post Helen!

    The historical record does seem to support you against Ara regarding the prediction that: “apologetics is doomed to failure as a philosophical enterprise because it fails to capture how our minds accept the plausibility of religious belief.” As you say, apologetics has been around for a while, and it might exist as long as organized religion exists.

    I’m just curious about whether you agree with the weaker claims Ara makes regarding the association between atheism and reflective thinking. Historically, it seems intuitively (?) to be true that in monotheistic cultures at least, atheism developed first among the most reflective members of the society. However, would it be possible to detect signs of atheism among any but these people? Are we assuming that no peasant, merchant or craftsman ever questionned religion throughout the middle ages and early modern period? Do you know anything about that?

    And as atheism spreads, one could argue that it will stop to require any more reflective thinking than religious belief: it will simply be what one is taught at home, in school, by one’s peers, etc. If Ara is right, one could imagine cyclical phenomena: it takes a lot of reflective thinking, by a lot of people, to make most people atheists, but then the next generations don’t have to bother with this, they forget about the reasons why they should be atheists, and they revert to some form of intuitive religion (wether it’s theism or a something closer to what is found in traditional cultures). Do you know if we see signs of this? For instance, that people brought up as atheists are more likely to become religious than people who ‘converted’, so to speak, to atheism are to return to religion?

    (Obviously these questions, while addressed at Helen, would welcome anybody’s answers!)

  • comment-avatar
    Helen De Cruz 10 December 2013 (14:14)

    Hi Hugo: thanks for your reflections!
    I think Ara is correct in arguing that atheism is more prevalent among the more reflective members of society, but I’m wondering about the causal relationship. Perhaps reflection makes it more likely that one adopts beliefs different from those in one’s surrounding culture. So for instance I am currently looking at results of a survey I conducted with philosophers of religion, and I was surprised to see the diversity of heterodox ideas they have (e.g., pantheism, polytheism). It is not a surprise, I think, that critical thinkers of the past like Newton, Bruno, etc were heterodox religious believers rather than atheists (e.g., Newton did not hold a traditional trinitarian view). Atheism was perhaps too big a jump for some of these authors, given the cultural context.
    It is also hard to precisely find out how pious people in the Middle Ages etc were, especially people who did not contribute to written history. Muchambled in his History of Violence seems to suggest that swearing was very common in the middle ages, e.g., people who invite God to strike him down, and “I renounce God’ was a common swear term, uttered on various occasions.
    It is quite conceivable, as you point out, that atheism can be just as unreflective as theism. In a society where almost everyone is an atheist, there are not cultural displays (where Ara also talks about) and no cultural context to interpret experiences one may have such as the detection of a purported agent (as Guthrie describes). It would be interesting to see then if this is a stable position, or if people revert back to theism. Japan, for instance, has a stable population composed mostly of atheists, and this really seems to be an unreflective form of atheism. I did not make a detailed study of the religious situation in Japan, mind, I have this from a book I recently read, written by a Shinto priest and expert on religion. The priest laments the fact that people in Japan just take naturalism as the default stance, and that their visiting Shinto temples is more a form of respect for traditions and “Japaneseness” than the expression of religious sentiment. Indeed, so ingrained is the idea that natural is the only game in town, that when Stewart Guthrie (personal communication) spoke to them about “supernatural’, they just thought it meant ‘very natural.
    I recently saw a study on atheism and non-affiliation in Britain where it was shown that churchgoers have a bit less than 1/2 chance to transmit their beliefs to their kids. However, atheists just bring up atheists. If this trend continues, Britain will soon consist solely of non-religious affiliates.
    I think ,therefore that atheism can be a stable situation, and that without religious displays or incentives (as Ara describes) a population can remain atheist.

  • comment-avatar
    Bryce Huebner 10 December 2013 (18:09)

    Hey Hellen,

    Really nice and really interesting post. I hadn’t thought too much about this in detail before reading your post, but I have been a bit skeptical of the attempt to link reflective thought to atheism myself. Part of the reason for this is that it’s not at all clear to me that you aren’t going to find a lot of motivated reasoning and attempts at dissonance reduction in the vicinity of antecedent religious (or secular) beliefs. Reasoning is nice, but if a person is an intuitive theist, they are probably going to hunt for rationalizations of their theistic intuitions when they start to reason (similarly, if a person is intuitively secular, we should expect those intuitions to anchor reasoning, with rationalizations being deployed to bolster those intuitions). This is a long way of saying that you might well be right that analytic atheism is a WEIRD culturally local phenomena. But if that’s right, I guess I want to know what it is that’s pushing things in that direction. Early exposure to physicalist hypotheses? The increasing prevalence if Darwinian alternatives? Both coupled to a strong centralized state? Or something else all together?

  • comment-avatar
    Olivier Morin 10 December 2013 (19:20)

    Just a quick word to recall the important work of Carlo Ginzburg (commented and expanded by numerous authors) on spontaneous atheism in popular settings in Renaissance Italy. It rather supports the reflective hypothesis, but it neatly refutes the view that past societies were homogenic as far
    as faith was concerned — or that religious skepticism is strictly an ‘EIRD’ phenomenon.

  • comment-avatar
    Ara Norenzayan 10 December 2013 (20:50)

    Helen is raising an interesting question: if analytic thinking discourages religious belief, how do we explain the persistence of the philosophical tradition of apologetics, which involves analytic reasoning about God? Let me clarify and elaborate a bit here on the claim I made in the book and perhaps offer some additional food for thought.

    1) I think apologetics as a philosophical tradition can surely exist and persist among the elite, but my point and prediction is that it will have little psychological resonance among believers at large. I’m not arguing that believing folk do not engage in doubt or analysis. They surely do. I’m arguing that most believing folk rarely come to arrive and maintain their religious beliefs through that sort of analysis and reasoning – they do so because they have powerful intuitions that make religious belief “feel right,” because they are exposed to powerful cultural influences, and because they live in conditions of uncertainty and threat that make them seek God or gods for succor in times of trouble. So in the book I should have said more accurately, that apologetics is doomed to failure as a psychological enterprise – not, as I wrote, as a philosophical enterprise — because it fails to capture how our minds accept the plausibility of religious belief. I don’t disagree then with Helen that apologetics can be a viable philosophical enterprise. After all, there are numerous philosophical and scientific enterprises that deeply violate natural intuitions we have about the world but continue to persist among the elite.

    2) Is it the case that philosophers and theologians who engage in apologetics are deeply religious? The claim, coming out of the cognitive psychology literature, is that analytic thinking (and to answer Bryce’s question — regardless of the content or subject matter to which it applies) will encourage the tendency to doubt or revise core intuitions such as mind-body dualism, teleology, and mentalizing. To the extent that religious beliefs are rooted in these core intuitions, analytic tendencies will therefore undermine these intuitions, leading to the prediction that, all else being equal, engaging in this tradition, or any tradition that involves habitual analytic thinking will weaken faith – not strengthen it. Maybe Helen knows the answer to this question.

    3) Helen is right that much of the experimental work on analytic atheism is based on WEIRD samples. This work is in its infancy, there is much more to be done! In our research we took great pains to look for socio-demographic moderators, but the point is well taken – we want to cast a wider cultural net to properly test the hypothesis and look for moderators and boundary conditions. As someone who has written about the WEIRD problem in psychology and cognitive science, I would applaud such work. The prediction is that this analytic effect should be specific to beliefs in supernatural agents (that are supported by core intuitions). So I wonder if in religions like Hinduism, where ritual is as (or perhaps more) influential than belief, whether analytic thinking will not be as pivotal of a factor in eroding religiosity.

    4) I emphasized above all else being equal, because as I argue in the last chapter of the book, there are multiple channels (which I explain in more detail here) that get us to disbelief. So if there is some religious sentiment among people who practice apologetics, that wouldn’t surprise me, because these intuitive, motivational, and cultural channels will continue to produce religious belief even if someone is immersed in an analytic enterprise. The claim is that their religious faith wouldn’t be the result of analytic reasoning (even if it appears that way to them). But as more of these channels come into force and converge, the likelihood of disbelief increases dramatically. One idea among many cognitive scientists of religion is that atheism is really just an effortful rejection of intuitive theism. The model I am advocating departs from this idea, arguing instead that atheism can achieve a viable cultural equilibrium when these multiple channels converge. So Hugo is right that a lot of atheism today is quite habitual and non-reflective once the right conditions are in place. And as Olivier points out, there are historical examples of atheist individuals (Diagoras of Melos supposedly was one of the first openly atheist Greek philosophers of the classical era). Wherever there has been religion, there has also been those who lacked or doubted the intuitions that make religious belief compelling. But the convergence of these multiple psychological and cultural channels seems to me to be more recent, where you have entire societies distancing themselves from religion.

  • comment-avatar
    Helen De Cruz 11 December 2013 (09:43)

    Thanks for everyone’s thoughtful responses. About point (3) in Ara’s response, there is some cross-cultural work that might be relevant, including the work of McKenzie Brown on the influence of Darwinism and pre-Darwinian non-theistic views on the ideas of creationism and intelligent design in hinduism. There is also Sedley’s book on creationism and antiquity. Both works indicate that when you have a class of people who can engage in reflection and enquiry, they can come up with non-theistic accounts of why the world is as it is today. So you have some non-theistic buddhist schools of thought in India, and naturalistic models (e.g., atomism) in ancient Greece.
    In that respect I think Ara is right that reflective reasoning encourages atheism (also in non-WEIRD groups), but as I said earlier, I am not sure if it a link between atheism and analytic thinking, or more generally between ideas that stray from the orthodoxy and analytic thinking. Paradoxically, once you have such non-theistic views, theists will use the tools of reasoning to come up with arguments for why those theistic views aren’t sound, this is what Ara refers to as apologetics.
    As I see it, apologetics is dependent on the presence of atheism. If everyone is a theist, and believes in the received orthodoxy, there’s no reason to engage in apologetics.
    BTW, I am currently conducting an empirical study on philosophers of religion, their motivations and their reasons that could provide some answers to the question of how reflective reasoning and faith interrelate. Of course, that is a very specific sample (as WEIRD as you can get), but I think the results are very interesting and relevant for the discussion here. I’ll have to first carefully analyze my results and interpret them (which will take some time, it’s a sample of >100 responses to an open survey), but I’ll put them online in due course, making reference to this discussion.

  • comment-avatar
    Ara Norenzayan 11 December 2013 (21:45)

    I think Helen’s alternative account is worth pursuing. If I understood her correctly, it goes as follows. Analytic thinking reduces any kind of normative belief, not just religious belief. Since religious belief is largely normative in Canada and US (where we ran these experiments), this possibility is viable. Whereas the account Will Gervais and I favor is the idea that thinking analytically blocks, overrides, or revises the core intuitions underlying religion (and for that matter, this should be true for any belief that is rooted in System 1 or intuitive processes). Now, we found these effects for every level of prior religious belief, and also found them in Vancouver, Canada, where one in two people are not religious (a very high percentage in North America). But these patterns are suggestive, not conclusive. Helen’s hypothesis is an interesting one, and I suppose can be tested in two ways 1) by running these experiments in overwhelmingly secular societies where religious belief is not normative (for example, Denmark, the Czech Republic, China, Japan, some subpopulations in France), and by measuring non-normative supernatural beliefs (such as belief in the gods of someone else’s religion). Importantly, I don’t think these have to be mutually exclusive accounts (it could be that both hypotheses are supported even in a single experiment).

    I look forward to seeing Helen’s findings with the philosophers and theologians!