Theism and Expert Knowledge


+10 rating, 16 votes
Loading...

Theism and Expert Knowledge
Rationality
70% theists among philosophers of religion: me worry?
Share

Introduction

Post last updated: 12.02.2015. Thanks to Kevin Corbett, Michael Tilley, M, and other commentators who helped us improve this post!

Suppose you are a layman when it comes to computers and artificial intelligence and you are curious about the time of arrival of human-level artificial intelligence, computers which are at least as smart as humans are. How can you arrive at realistic estimates? A sensible option is to turn to experts in the field, since their opinion is informed by their expert knowledge.

Now suppose you want to figure out whether God(s) exist, but you’re not an expert on the question. How can you make up your mind without reading thousands of pages on arguments for and against theism? You might want to turn to experts, in this case philosophers, or more specifically, philosophers of religion. What would these experts tell us?

The Empirical Data

A survey conducted among philosophers in 2009 shed some light on this question. Of all the faculty members among the 3226 philosophers who took the survey, 72.8% were atheists and only 14.6% theists. (The remaining 12.6% chose another option.) This is especially interesting when contrasted with the percentage of atheists among the global population: a paltry 2.01% in 2010.

But if we focus more narrowly on philosophers of religion, the experts among philosophers on the question of theism/atheism, we’re in for a surprise. Among philosophers specialized on philosophy of religion only 19.1% accept or lean toward atheism, while 72.3% accept or lean toward theism. This is the highest effect any kind of specialization has on any of the questions in the survey. In a similar study by Helen de Cruz only 15.7% of the philosophers of religion accepted atheism and 5% were agnostics while 57.7% self-identified as Christian theists. 17.6% did not classify themselves as falling in any of those categories.

What should we make of this? Does this mean that theists can justify their beliefs by appealing to the authority of philosophy of religion experts? Should atheists be worried since they contradict the majority view among experts on the topic?

On the face of it, there are two hypotheses which could explain the data, one of them worrying for atheists, the other less so:

Expert Knowledge: Philosophers of religion possess expert knowledge on the arguments for and against God’s existence. The arguments for God’s existence are just overall more convincing and render God’s existence more probable than not.

Selection Bias: People often become philosophers of religion because they are religious, or at least have a high credence in God’s existence. Theists often become philosophers of religion, not the other way around.

It should be noted that these two cases are not mutually exclusive. Adherents of the expert knowledge hypothesis could in principle admit that there is a selection bias and claim that expert knowledge still favors theism over atheism. This would then have to be reflected not in the overall theists to atheists ratio among philosophers of religion, but in how exposure to philosophy of religion changes the theists to atheists ratio. So we have to keep in mind in the next section that diagnosing a selection bias is not sufficient to undermine the expert knowledge hypothesis.

Expert Knowledge or Selection Bias?

To draw further conclusions concerning the two hypotheses we need more empirical data, as the 2009 philsurvey does not contain sufficient information to determine their truth or falsity. Helen De Cruz’s study, however, contains qualitative data on why philosophers started doing philosophy of religion as well as quantitative data on how their beliefs concerning theism and atheism developed over time. This is exactly what we need.

Motivation for doing philosophy of religion: The study brings to light three main reasons for doing philosophy of religion. The most prevalent is described by the author of the study as “faith seeking understanding”, religious people who want to better understand their own belief. The second frequently cited reasons was Proselytism and witnessing: many philosophers of religion felt that doing philosophy of religion was part of their calling as religious people. One philosopher for example wrote: “My religious commitment helps to motivate some of the work I do (part of which involves defending and explicating Christian doctrine)”. The third most cited reason was a fascination with religion as a cultural phenomenon. These results give some support to the selection bias hypothesis, but since the study does not contain the numbers for each of these responses it is too early to tell.

Belief-revision: The study contains numbers on how philosophers engaged in belief-revision due to their engagement with philosophy of religion. Of all the respondents (136 of 151 answered the questions on belief-revision) more than 75% underwent some degree belief-revision on topics in philosophy of religion, and around 67% of all participants observed a change in their beliefs which can be attributed to philosophy:

no change: 24.3%

belief revision to atheism or agnosticism: 11.8%

belief revision to theism: 8.1%

philosophy polarized: 9.6%

philosophy tempered: 25%

other change: 12.9

change, but not attributed to philosophy: 8.1%

These numbers show that there was an overall shift toward atheism/agnosticism of 3.7% if we compare both directions of belief-revision: the direction of belief-revision was most frequently in the direction of atheism/agnosticism.

This supports the view that the theists to atheists/agnostics ratio is even higher before exposure to philosophy of religion and confirms the impression we got from considering philosophers’ motivations for doing philosophy of religion: most philosophers of religion were already theists when they started, so there is a strong selection bias at work.

Moreover, there are more philosophers of religion updating their beliefs toward atheism and agnosticism than toward theism. This seems to weaken the hypothesis that although there is a strong selection bias, expert knowledge favouring theism is still reflected in the fact that philosophers of religion convert more often to theism than to atheism/agnosticism while acquiring expertise in the field. The numbers show that the ratio of theists to atheists/agnostics declines with exposure to philosophy of religion.

This verdict is confirmed if we look at the percentage of theists who report that exposure to philosophy of religion tempered their beliefs and the percentage of atheist who reported a tempering of their beliefs. Of the theists, 33.7% reported a tempering influence, whereas only 10.3% of the atheists reported a similar influence. In other words: a higher proportion of theists become less sure about beliefs such as taking the Bible to be literally true, accepting the Fall, regarding Catholics as heretics, etc. than atheists who become less sure about aspects of their atheism or more appreciative of theist views.

Despite these statistical conclusions it is still possible that most theists remain theists due to strong arguments for theism, and those atheists/agnostics who convert to theism do so for the same reason, while conversion to atheism/agnosticism happens due to weak arguments. This may be very implausible, but it is an epistemic possibility.

However, it seems that the fact that this epistemic possibility is highly unlikely is already sufficient to undermine appeal to authority arguments in this domain. In most cases it is truth-tracking to reject appeal to authority arguments if strong selection biases are at work in the field and the experts in question are more likely to reject the view in question after acquiring expert knowledge on the topic. Even if it just so happens that this is not the case when it comes to theism, it is still reasonable to reject appeal to authority arguments for theism, because rejecting such arguments is overall more likely to promote true beliefs. Long story short: atheists should not be worried about the theists to atheists ratio in philosophy of religion.

General Remarks on Philosophy of Religion

The findings of this post fit with other findings on reasoning and religion: It has been argued that many philosophers of religion suffer from cognitive biases and group influence, and that the field as a whole is too partisan, too polemical, too narrow in its focus, and too often evaluated using criteria that are theological or religious instead of philosophical. Recent work in cognitive science of religion suggests that analytic thinking is a pathway to atheism (Norenzayan and Gervais 2013), and it has been observed that analytic thinkers show weaker religious belief and tend to lose their religious fervour, even if they were originally raised in a religious environment (Shenhav et al. 2012). Experimental work supports these correlations and provides additional evidence for causal connections between analytic thinking and erosion of religious beliefs (Gervais and Norenzayan 2012).

An especially critical stance on philosophy of religion has been taken by one of the participants in De Cruz’s study: “I would not be the first to say that philosophy of religion, especially “analytic theology”, is simply not philosophy. It’s Christian apologetics, and it often is poorer philosophically because of that. A Christian bias pervades everything, and, once one becomes a non-Christian, the irrational faith-based assumptions and intuitions start to stand out.” More work would need to be done to determine whether things really are that bad in philosophy of religion, but the empirical data are already sufficient to render the expert knowledge hypothesis improbable and hence undermine the appeal to authority arguments in the theism/atheism debate.

References

Gervais, W. M., & Norenzayan, A. (2012). Analytic thinking promotes religious disbelief. Science, 336(6080), 493-496.

Norenzayan, A., & Gervais, W. M. (2013). The origins of religious disbelief. Trends in cognitive sciences, 17(1), 20-25. http://www2.psych.ubc.ca/~ara/Manuscripts/Norenzayan&Gervais_2013.pdf

Shenhav, A., Rand, D. G., & Greene, J. D. (2012). Divine intuition: cognitive style influences belief in God. Journal of Experimental Psychology: General, 141(3), 423.

This article has 22 comments

  1. […] religion, though, 72.3% accept or lean towards theism. Adriano Mannino considers the question in a post at the group blog Crucial Considerations. Of these figures, he […]

  2. Are you taking into account the base rate of those entering as theists/atheists? If, on the one hand, approximately 53% (the percentage of theists 61% minus the percentage of theists who had a belief revision to theism 8%=53%) of the total participants in the study begins as a theist, and approximately 12% have a belief revision to agnosticism or atheism (some of these could be from the 18% who were neither theists nor agnostics/atheists, but I’ll assume they were all theists), that would be only about 23% of the total number of those surveyed who came in as theists and had a belief revision to atheism/agnosticism.

    If, one the other hand, atheists and agnostics only accounted for approximately 9% of those entering the field already as an atheist or agnostic (the 21% percent of atheists/agnostics minus the 12% percent of atheists/agnostics who had a belief revision to atheism/agnosticism = 9%). Now, at least some of those who had a belief revision to theism had to come from the underdecided category, but it could also very well be the case that a greater percentage of atheist/agnostics entering the field have a belief revision to theism than vice versa. Furthermore, even if it were the case that every single belief revision to theism was from the undecided category and every single belief revision to agnosticism/atheism was from someone who began as a theist, I’m concerned that the there could still be substantial support for the expert hypothesis.

    In short, I’m concerned about the use of raw percentages to make the argument, since so many more came into the discipline as theists already.

    • Thanks Michael and Joel for your comments! It seems to me that the ratio of theists to atheists in the phil paper survey + the fact that the ration declines with exposure to philosophy of religion is sufficient to support the selection effect hypothesis: the ratio in the philpapers survey is due mostly to a very strong selection effect.

      However, as is pointed out in the post, the selection effect and expert knowledge hypotheses are not mutually exclusive, and evaluating the expert knowledge hypothesis is trickier. I agree that it would be good to take the base rates into account. The absolute numbers seem to be sufficient to show that the theists to atheists ratio declines with exposure to philosophy of religion, but they are not sufficient to show that it would also decline if the initial ratio were more even.

      The numbers we get from the original blog post at problogion are not sufficient to keep track of where the converts come from and hence what percentage of the original groups converted, but Helen De Cruz study has not yet been published. Maybe the published version will contain more details. I’ll send her an email to point her to this question. And I’ll make sure we’ll add any additional numbers the published paper may contain to this post.

  3. Maybe of interest: there is a forthcoming volume of essays, edited by Kevin Timpe and Daniel Speak on the connection between theism and libertarianism about free will, and several of the essays engage with some related topics here about religious belief, philosophical views, and the relationship between them. The table of contents can be found here: .

  4. Hi Adriano, thanks for this interesting post. A few points:
    1) To me, Helen’s comment (in the article you point to) that most probably, atheists are unlikely to specialise in philosophy of religion because they have no inner trigger to engage in it (no more than people who are not musicians and do not enjoy music might be interested in philosophy of music and so on) seems completely convincing.
    2) Part of the critiques against philosophy of religion being a disguised apologetics should apply to many other fields of philosophy when they specialise on a single set of data (e.g., philosophy of aesthetics if it would only discussing classical paintings from the Renaissance). Perhaps philosophers of religion should engage more often with theologies and religions different than the mainstream Christian one (I have argued, on a similar vein, that logicians should also engage with texts remote in time or space: http://philosopherscocoon.typepad.com/blog/2015/01/why-testing-logic-or-philosophy-in-general-on-non-western-ideas.html)

  5. Moreover, there are more philosophers of religion updating their beliefs toward atheism and agnosticism than toward theism, so we can reject the hypothesis that although there is a strong selection bias, expert knowledge favouring theism is still reflected in the fact that philosophers of religion convert more often to theism than to atheism/agnosticism while acquiring expertise in the field.

    This inference seems to me unwarranted. There are more theists to begin with, so there’s a larger pool to convert to atheism/agnosticism. If 21% of philosophers of religion are atheists/agnostics and 58% are Christian theists, then 11.8% moving towards atheism/agnosticism is roughly 1/5 of Christians becoming atheists/agnostics, whereas 8.1% revising to theism is over 1/3 of atheists/agnostics becoming theists.

  6. “These numbers show that there was an overall shift toward atheism/agnosticism of 3.7% if we compare both directions of belief-revision: the direction of belief-revision was most frequently in the direction from theism to atheism/agnosticism.”

    A revision to atheism or agnosticism does not imply that all such moves are from *theism* to one of those. After all, one could’ve begun as an atheist but moved to agnosticism (which I take to be suspending judgment on the question of theism); this would be a move *away* from atheism. Unless the questions have data on the position from which one is moving, we cannot conclude that all of the movement is from theism to one of [atheism or agnosticism].

    • I would have thought that the category “belief-revision toward atheism/agnosticism” excludes cases where someone moves from atheism to agnosticism and only includes cases where someone moves to one of these views from another category. But now that you point it out it looks like you can understand the category in two ways.

  7. Also possibly interesting: a study from my MA thesis was designed similarly to Shenhav et al’s first study. In my own study I look at correlations between philosophical training/selection, cognitive style, and PhilPapers responses. In the discussion section of the study, I (among other things), I address the prevalence of atheism/agnosticism in philosophy. While I am not able to conclusively isolate selection effects from learned/training effects, I end up leaning toward the selection hypothesis.

    The paper: https://www.academia.edu/10329040/Intuitive_And_Reflective_Responses_In_Philosophy

    Thanks for revisiting this interesting question. And thanks to Helen De Cruz for her important work!

  8. […] tendency for believers to get  involved in the philosophy of religion. A new study has been done confirming the ‘selection bias’ explanation. It’s nice to have the reassurance that random selection of experts reflects my opinion. That […]

  9. […] I came back to the same thought while reading Adriano Mannino’s post about the diffusion of theism among philosophers of religion. Adriano discusses the worries of some […]

  10. I wanted to point out an inaccuracy: the figure of 72.8% in favor of atheism for the PhilPapers survey is not out of the 3226 who took the survey, which includes undergraduate and graduate students, but from the 933 target faculty the creators of the survey selected. The percentage of atheists for the whole 3226 was 66.2%

  11. […] god or gods? Philosophers of religion? A recent discussion of “Theism and Expert Knowledge” is available here. The stats are interesting: “Among philosophers specialized on philosophy of religion only 19.1% […]

  12. […] of our confort-zones and test our ideas outside them. A few days before, Adriano Mannino had posted here his comments on a study by Helen De Cruz and asked whether philosophy of religion is more than […]

  13. […] I came back to the same thought while reading Adriano Mannino’s post about the diffusion of theism among philosophers of religion. Adriano discusses the worries of some […]

  14. […] a few articles (see here, here and here) have raised the issue of whether philosophy of religion is not really little more than […]

  15. No offense to the author, but it is not surprise that he would write an article like this that skips details of the surveys by Dr. Helen Cruz without reading her conclusions first.

    If the author had taken his time to read her responses to others on her blog and what she writes on these surveys, then you will see she does not arrive at the same conclusion the author does: mainly that pre-determined beliefs exclude expertise.

    Now here is what Helen says on this here:(http://prosblogion.ektopos.com/2012/02/25/one_of_the_stri/)

    “The interrelated questions seem to be whether
    – self-selection precludes expertise. I don’t see why this would be the case. In PoR there is a domain of relevant expertise, namely, amongst others, expertise in arguments for/against the existence of God. Regardless of whether God exists, one needs to acquire a competence in specific defenses, attacks, counterarguments, etc and considerations of internal consistency, explanatory potential etc. in questions on His existence.
    – whether, after taking into account self-selection there is still some residual confirmation left over.”

    Notice that she is responding to a blog commenter who misreads or misjudges Helen’s survey or content like Adriano did, and yet, she completely disagrees with him. She notes that PoR are the experts, they are mostly theists, and that they are the experts because of their expertise and competence.

    Then she says that:

    “I am inclined to think that the relevant expertise in philosophy of religion is indeed something along the lines of what Scott and Jordan suggest: expertise in PoR is expertise on arguments for/against the existence of God, next to arguments for/against the reasonableness of religious belief, arguments for/against the internal consistency of certain forms of religious beliefs, e.g., Christianity with its specific doctrines on Incarnation, the Trinity etc.
    If this is the case, and self-selection does not wholly explain the prevalence of theism in PoR, then perhaps the prevalence of theism does have some evidential weight in favor of theism. Before I got into PoR I was decidedly a non-theist (although religion has always fascinated me) but I have since revised my beliefs in the light of particular arguments. Perhaps this is atypical, but I am not alone in this (see some of the comments in this thread and threads at NewApps and elsewhere).”

    “no change: 24.3%
    belief revision to atheism or agnosticism: 11.8%

    belief revision to theism: 8.1%
    philosophy polarized: 9.6%
    philosophy tempered: 25%
    other change: 12.9
    change, but not attributed to philosophy: 8.1%
    The percentages of belief revisers within subgroups (atheists/theists) was as follows:

    of the theists, 33.7% reported a tempering influence of philosophy such as no longer taking the Bible literally, not believing in the Fall, not regarding Catholics as heretics etc, becoming universalist
    of the atheists, 10.3% reported a tempering influence of philosophy, e.g., become more tolerant of religious believers, become appreciative (although not convinced) of the arguments for theism.

    The percentage within each group of hardening was about the same: 10.3% for the atheists, 8.4% for the theists. 24.1% of current atheists and agnostics in the sample were former theists, losing their faith at least in part through philosophy (others are coded as ‘change, but not due to philosophy, see above).
    By contrast, only 12% of current theists in the sample came to their belief at least in part due to philosophy. This is a significant difference.”

    Notice that only a small number of theists move to theism. The majority of PoR have not changed their opinions on God to atheism/agnosticsm, they are still in the theist camp. So Adriano is quite wrong in saying that “theists are revising their beliefs to atheism in philosophy of religion than the other way around”. The simple fact is that most philosophers of religion have not changed their opinions on theism and have been influenced to theism by the evidence.

    • Jonathan Erhardt
      Tuesday 12 May 2015, 8:00 am /

      Thanks for your thoughts Matthew!

      As far as I can see nothing you say disagrees with Adriano’s post. Adriano emphasizes twice that self-selection doesn’t preclude expertise: He writes “It should be noted that these two cases [Selection Bias and Expert Knowledge] are not mutually exclusive” and later “it is […] possible that most theists remain theists due to strong arguments for theism, and those atheists/agnostics who convert to theism do so for the same reason, while conversion to atheism/agnosticism happens due to weak arguments”. Both statements support your view on the relation of self-selection and expertise.

      You suggest that Adriano is wrong in saying that the direction of belief-revision was most frequently in the direction of atheism/agnosticism. But as far as I can tell the Helen De Cruz quote you post supports Adriano’s interpretation: In absolute numbers more people converted to atheism/agnosticism than the other way around and more theists reported a tempering influence of philosophy than atheists/agnostics.

      You observe that “[T]he majority of PoR have not changed their opinions on God to atheism/agnosticsm, they are still in the theist camp”. That’s true, but I can’t see how this undermines Adriano’s interpretation. It seems to me that all he needs are observations about the relative frequencies of belief-revision. Given the prevalence of various biases we should perhaps expect most people to stick with their original beliefs in areas of inquiry where hard empirical evidence is scarce.

      Moreover, Adriano doesn’t suggest in the post that his interpretation of the study agrees with that of Helen De Cruz. How then does he misread her study? Did he get any of the number from the study wrong? If so, please let us know!

  16. I’m very open to being corrected, but it seems to me that:

    (a) The author comes off as sufficiently biased that it perhaps shouldn’t be surprising if we find things are skewed things in a rather egregious manner. Regarding why I suspect bais, first, notice how only quotes sources saying Philosophy of Religion is apologetics, and slanders it, are given while not quoting anyone saying or implying the opposite. Second, the piece gratuitously also decides to quote studies regarding indicating that studies show, using his words, “analytic thinking is a pathway to atheism” (because there is a correlation among laymen!) This leaves out the fact that theism is sort of the highly intuitive default belief (http://beliefmap.org/god-exists/theism-intuitive/), and analytic thinking is always the pathway out of those… including into solipsism and fringe theories of time. It says nothing about such a correlation among experts, nor anything about the intellectual merits of the relevant ideas. It is 100% irrelevant.

    (b) It is stressed hard that there is an overall shift towards atheism, but it is only 11.8% vs. 8.1%… *crickets chirping* Aside from the fact that that’s a statistically small difference, was it not just noted that the vast majority of individuals entering into PoR are theist? That means at least 8.1% of individuals entering are atheist, because 8.1% are converting to theism. Is that like, almost all the atheists? If I’m reading this right, the percentage of atheists converting to theism vastly outstrips the percentage of theists converting to atheism.

    (c) The article for no reason takes it as being in favor of atheism that theists are moderating their views more than the atheists entering [e.g. dropping inerrancy]. How strange, when that sooner suggests the theists in question would move to atheism if they felt the evidence led that way. That is nothing to do with anything.

    (d) IMO, it’s egregious in general to boldly suggest that so many academics here are not trying to honestly follow the evidence where it leads, all without knowing any of them personally.

  17. Nycce post !

    I’m still confused about how to deal with selection-bias, though. If exposure to a field skews the initial opinions in one direction, that may just be because the expert opinion tends to shift toward the majority view among those experts. So I don’t think that measuring shift in initial beliefs after exposure to studying AI would be good evidence in favour of higher probabilities of AI risk. Though it does seem to apply to the case of theological philosophers, since there the initial beliefs are skewed away from the group norm, the opposite of what group conformity would predict.