Why Are So Many Philosophers of Religion Theists?
72.8% of the 3226 philosophers who took the PhilPapers survey in 2009 said that they accept or lean towards atheism. Among philosophers of religion, though, 72.3% accept or lean towards theism. What explains this difference? Adriano Mannino considers the question in a post at the group blog Crucial Considerations. Of these figures, he writes:
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. Theist often become philosophers of religion, not the other way around.
He then makes use of the results of the data from the study by Helen De Cruz (VU University Amsterdam) of why philosophers of religion went into that field and how their beliefs concerning theism and atheism changed over time. He ends up concluding that the evidence is best explained by the “selection bias” hypothesis. He says:
The theists to atheists/agnostics ratio is even higher before exposure to philosophy of religion. This 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, 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. The numbers show that the ratio of theists to atheists/agnostics declines with exposure to philosophy of religion.
The whole post is here. Some previous discussions of philosophy of religion at Daily Nous are here and here.
(via Helen De Cruz)
I would think that if you are not at least an agnostic, then a lot of the philosophical discussions about the nature of the supernatural would seem kind of pointless; how many angels can dance on the head of a pin seems even sillier if you don’t believe angels exist. I would think if you are interested in the social/cultural, historical, or literary aspects of religion then rather than gravitate towards philosophy you would go into sociology/anthropology/religious studies, history, or literature respectively.Report
Are philosophers of religion, on average, older than other philosophers? http://www.smbc-comics.com/?id=3626Report
There are, in fact, other religions aside from Christianity. How many philosophers of religion study Hinduism or Buddhism? For that matter, how many philosophers of religion are Jewish or Muslim? The selection effects are fairly obvious, and I’m glad they’ve been confirmed empirically. But more can be done to make the field better. Religion is a very important philosophical topic, so we should be concerned about whether it is being studied in a rigorous and unbiased way. The phrase “philosophy of religion” should not be synonymous with “Christian philosophical theology”.
(And for the record, I have no problem whatsoever with philosophical theology. It should simply be labeled as such.)Report
In other news, grass is green, water is wet, and Bill Belichick cheats.Report
What about the following alternative proposal:
On the face of it, there are two hypotheses which could explain the data, one of them worrying for theists, the other less so:
Expert Knowledge*: Philosophers possess expert knowledge on the arguments for and against God’s existence. The arguments against God’s existence are just overall more convincing and render God’s nonexistence more probable than not.
Selection Bias*: People often become philosophers because they are not religious, or at least have a low credence in God’s existence. Atheists often become philosophers, not the other way around.
Of course, to test which hypothesis is better, we would need data on why individuals have chosen to go into philosophy, and how their beliefs concerning theism and atheism have changed over time.Report
I agree with Shea. “Philosophy of religion” is a terrible label for what goes on under that heading. I, and most other folks who work in ‘philosophy of religion,’ could not care less about religion as a category. What we care about is theism, its warrant and implications, and for most of us the sort we are concerned with is Christian theism. It is not as if we are going to change what we are interested in because the label stinks. Those of you who think that religion as such is worthy of philosophical investigation can go for it, regardless of what the rest of us are up to.Report
Suppose that one received some empirical data that engaging in philosophical investigation tends to move one toward substantive moral views that are less characteristic of common sense. Suppose you are more of a common-sense moralist. How worried should you be by this data? My own thought is that this movement could be explained in many ways other than the force of reason — e.g. by distorting influences of philosophical investigation (that, for example, those that engage in such investigation for a living tend to overestimate the amount of insight that one gets through argument and to underestimate the value of lived experience in moral communities). In short, there are a lot of ways to explain the data; how about, instead, just showing me the arguments themselves that, if this movement away from common sense morality is rational, is the explanation of that movement? The same goes for theism and philosophy, I think.Report
Maybe I’m just insufficiently interested in backing up my beliefs with statistical evidence, but the explanation here strikes me as so obvious as to not warrant discussion. A disproportionate number of philosophers of religion are religious for the same reason that a disproportionate number of feminist philosophers are women, and philosophers of race are racial minorities, and (I would bet) philosophers of art are artists, and philosophers of literature are writers, and philosophers of sport are sports fans. If X is an important part of your life, and you are a philosopher, then you are more likely to be interested in the philosophy of X. Is there anything mysterious or surprising about this?Report
I think there is a selection bias in philosophy of religion, but I don’t think the belief-revision numbers show it. It looks like the base rate is doing all the work here. I’m going to set out my thinking here. If anyone sees any errors in my reasoning, please point them out!
Using percentages that Helen De Cruz reports in the comments on her article at Prosblogion, we have roughly the following numbers out of 151 surveyed philosophers of religion:
37 no change
18 change to atheism/agnosticism
12 change to theism
19 other change
12 non-philo change
These sum to 150 owing to rounding, since I am working backwards from percentages to get counts, instead. Assume that changes to atheism/agnosticism are always from theism. The raw number of theists who change their beliefs toward atheism (18) is larger than the raw number of atheists/agnostics who change their beliefs toward theism (12). But we should expect the raw number of theists converting to be larger than the raw number of atheists/agnostics converting simply because there are so many more theists in the initial pool. When you look at how many atheists/agnostics there are and how many change their beliefs versus how many theists there are and how many change their beliefs, you see that the conversion rate is much higher from atheism/agnosticism to theism than from theism to atheism/agnosticism. (I’m not terribly *surprised* by this because I agree that there is a strong selection effect … I just don’t think the numbers confirm that there is a selection effect or, really, show anything interesting about the power of arguments in philosophy of religion.)
Again, assume that every conversion to agnosticism/atheism was conversion from theism. In the sample of 151, there were (as far as I can tell from the write-up at Prosblogion — I actually found this especially unclear, given that 18% of respondents were supposed to have given answers that didn’t really fall on the theism-atheism continuum) 92 theists, 32 atheists/agnostics, and 27 uncategorized. Working backwards based on the belief change numbers, that pool started out looking like 98 theists and 26 atheists/agnostics. So it’s true that the ratio changed toward atheism/agnosticism. However, the rate of change within each pool strongly favored theism: 18/98 (18.4%) converted from theism to atheism/agnosticism while 12/26 (46.2%) converted in the other direction.
I doubt that the conversion number is a reliable estimate of the power of the arguments to convert people one way or the other. But if it *were* such an estimate, then if the initial pool entering philosophy of religion were balanced on the question of theism versus atheism/agnosticism — 75 theists and 75 atheists/agnostics, say — then after exposure to the arguments, the numbers would look like 96 theists and 54 atheists/agnostics. So, if the estimate of the conversion rate were a good one, the fact that the overall ratio moves towards atheism/agnosticism in the actual case would only reflect the fact that in the actual case the initial pool so strongly tilts towards theism.Report
Jonathan Livengood: Thanks for unpacking your reasons for your claim. You’ve convinced me.
Anonymous # 4:59: This is an important point. Thank you for highlighting it.
I feel sheepish doing this, but I will add that in one analysis of correlations between philosophical training/selection, cognitive style, and PhilPapers responses (n=500+), I found the following: “having or being a candidate for a PhD in philosophy [vs. not] was not signiﬁcantly related to one’s inclinations about the existence of a god” (here, page 28). Curiously, one’s cognitive style (intuitive vs. reflective) was significantly related to one’s view about God: “each intuitive response on the [cognitive reflection test (Frederick, 2005)] was signiﬁcantly related to being more likely to report either leaning toward or accepting theism; F (1, 559) = 7.3, p < 0.01, d = 0.16, b = 0.12" (page 22). This does not conclusively rule out either a selection effect or an expertise effect, but it might be worth throwing into the data/analysis being considered. It would be interesting to see if cognitive style also predicted one's interest in or AOS/AOC in philosophy of religion since this might support some sort of selection effect — although not necessarily the same selection effect being proposed here. I hope to ask these questions (and many more, including questions about personality effects, etc.) of the dataset in the future.
One more query: I am looking for a sense of statistical significance and effect size in Helen De Cruz's analysis or Adriano Mannino's analysis. But it's entirely possible that I just failed to see or find them. If anyone could point me in the direction of these numbers, I'd be much obliged.
Thanks Helen, Adriano, Jonathan for your interesting work in this area. Thanks also to Shea, DC, M, and the various Anons for the helpful discussion. Peace!Report
Jonathan Livengood on Jan 31, 2015 • 1:37 pm at 1:37 pm:
“… However, the rate of change within each pool strongly favored theism: 18/98 (18.4%) converted from theism to atheism/agnosticism while 12/26 (46.2%) converted in the other direction.”
– it seems you assume that all those converted to theism started out as atheists/agnostics, and vice versa?
By contrast, suppose, for example, all those converted to theism had fallen in the “uncategorized” category (e.g. deists, buddhists, shamanists, etc.) whereas all those converted to atheism/agnosticism had been theists originally. Then we would have 18/98 (18.4%) converted from theism to atheism/agnosticism while 0/26 (0%) converted in the other direction (with some atheists/agnostics converted to various beliefs other than theism).Report
This makes me wonder what the numbers are for feminist philosophy. I’d bet among people working on feminism, 99.99% are or lean towards being feminists. What might explain that?Report
So I take the data as presented to serve as a sort of undercutting defeater for a specific kind of evidence for theism, i.e. evidence from professional consensus. Nothing more, nothing less. The presence of selection effects does not thereby also act as a defeater for theism simpliciter. To be frank, I’m not sure who was appealing to consensus in the first place. I’m fairly sure that people like Plantinga think the primary evidence for theism consists of their arguments, not in the fact that they believe their arguments. So it may be that the target of the experiment is a bit of a philosophical bogeyman, unless there are a bunch of theists running around and appealing to consensus within philosophy of religion that I’m not aware of (which could very well be the case).Report
“It seems you assume that all those converted to theism started out as atheists/agnostics, and vice versa?”
Yes, that’s right. (I thought I was explicit about that assumption?) Given just the numbers reported, your reconstruction is possible. And for all the numbers say, it could be that all of the conversions to atheism/agnosticism came from the uncategorized group and all of the conversions to theism came from the atheist/agnostic group. Both those seem implausible to me and the flat 0% claims are both ruled out by things that Helen De Cruz says in her post.
But again, my overall point was that we can’t say much interesting on the basis of the reported numbers. My reconstruction merely illustrates why one shouldn’t infer from the fact that exposure to philosophy of religion decreases the ratio of theists to atheists/agnostics to the claim that atheism/agnosticism is better supported by the arguments. If we knew the actual conversion rates, we might be tempted to re-run the argument. Especially if the numbers ended up looking the way you suggest. But as I intimated earlier, even if we knew the actual conversion rate, I’m not sure it would tell us anything interesting. It seems to me that we only have something interesting if the measured conversion rates reflect something about the power or quality of the arguments. And I’m not sure I would want to make that assumption.Report
Selection bias probably plays a bigger role in a lot of areas. I don’t this this as a special problem for the philosophy of religion. It’s a challenge for science in general. A lot of climate scientists sign up for grad school because they want to save the planet, and bring with them a large and possibly un-critical explanatory framework to the problems they face. This doesn’t mean that their work isn’t credible, but the mere fact of agreement isn’t, by itself, evidence for a given thesis. (See this DN’s previous posts about gay marriage supposedly being a settled ethical issue.)
The bottom line is that while scholarly agreement should, in many cases (not always), be good enough for non-experts, experts always have the responsibility to keep critical distance and revisit assumptions. No brainer for philosophers, right?Report
I look forward to Mannino’s forthcoming paper on the prevalence of Catholicism among Popes.Report
I’d also confess to being pretty unsurprised by these results. But I’d be interested in analogous data about other subfields of philosophy as well. A parallel that comes to mind would be to ask how many ethicists are moral realists, as compared to the proportion of the whole profession that favors that position. I wouldn’t be surprised to see similar results for that question, although it would be interesting if they were totally different as well. If we found similar results, I’d be inclined to say that that the selection bias hypothesis seems correct here too, but I wonder whether other people’s intuitions would be the same.
Clear-cut analogies for other subfields don’t obviously suggest themselves to me, but some possibilities might be,
what proportion of philosophers of science are naturalists?
what proportion of epistemologists deny skepticism?
I think this is a little different than anon’s question above, because it’s about whether prior held beliefs, not identities, are relevant to people’s decision to enter a particular field, and whether those priors prove resistant to change or not, which would have interesting implications for the way we think about specialization within philosophy.Report
They omitted to remind people that the number of atheists in the general public is 4.1% in the US, that the respondents are mostly all based in the US, and that even if Religion Of Philosophy does have just 20% of respondents being atheist this is 6 times the norm. So the selection bias must be the other way around. 6 times more atheists opt for ‘Philosophy of Religion’ than are present in the base population.
I’d dearly love to know what ‘Other’ meant in this statistic when one is either an atheist or a theist…. whats the third option marked ‘other’?Report
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%
Now, here I could say, the author of this point, being engaged in analytic philosophy and thus probably more inclined to atheism, deliberately misrepresent the statistics in order to support a conclusion he already believed before writing the article. But because I think psychoanalyzing people on their views as a means of attacking those views instead of trying to attack them directly a low and unworthy tactic, I’m going to assume it was just a mistake.Report
Hmmm… This does of course mean that by the time a student qualifies 4 years after their freshman year, that almost 12% of them have deconverted from theism? It rises from 66 to 78 percent based on qualifications… hardly a good thing to ignore if one has an agenda right? Maybe it was an analytically minded person with no skill in analytics or common sense or that matter.
If I were an analytics expert, well since other folks are going to read this, I wouldn’t want to hide that fact.
They also pointed out the whole ‘Philsosphy of religion’ thing with a sample size of just 47! With that low a number 1.5% of the same size, you’d be more accurate throwing darts at the wall blindfolded.
Plus even if the entire set showed just 10% f all philosophers in the sample were atheist, this is still three times the base populations cohort at 3.1%. So you don’t need to massage the figures…. the result is orders of magnitude above the base figure for atheists in the general population. Reversing the figures to show only 20% of Philosophy of Religion were atheists isn’t very helpful there… thats 6 times above the base population too!
Its a little like arguing which would kill you more… and atomic bomb or strangulation…. your dead what would it matter?Report
Even mathematically I have to disagree with this.
of 3200 philosophers 78% are atheist.
Most of the Respondents are American philosophers. I have no idea why this is the case, perhaps Philpapers is just a thing in the US and not so much in Europe.
The actual list is published of respondents.
That means that overall the number of philosophers that are atheist is 26 times higher in philosophy than the general population. And absolutely staggering statistic.
Now in terms of ‘Philosophy of religion’ well only 47 of those 3200 mentioned philosophy of religion… So if we just look at those the sample isn’t representative enough to draw conclusion. Because just one person in that represents 47 a few percent either way. but lets continue and ignore the fact the sample is too small.
That means that the number of atheists that are ‘philosophers of religion’ is still ten times the number in the base population.
How is this in any way an indication that ‘Philosophy of Religion’ leads to religion… It demonstrably and mathematically does not!Report