[X posted on Prosblogion] Earlier on this blog, I have reported results of a survey on natural theological arguments , see here and here. To briefly recall, the survey asked philosophers to rate the strength of natural theological arguments, grouped into 8 arguments that seek to support belief in the existence of God, and 8 arguments that seek to support belief in metaphysical naturalism. My initial analysis indicated that religious belief (theism, atheism or agnosticism) is a strong predictor of the extent to which participants evaluate these arguments. However, in my analysis I examined only the effects of religious belief on the total overall assessments, not the arguments individually. In this post, I will report some fine-grained analyses on how philosophers evaluate individual arguments, as a function of their religious belief, gender and whether or not they specialize in philosophy of religion. Since the statistics are quite detailed, I will make this a two-part post, starting out by the positive arguments. The analyses have been conducted by Robert O'Brien, a statistician at the University of Miami.
Robert pointed out that I could test for the effects of gender and theism as covariates. I then suggested to him we could also see whether philosophy of religion made a difference. So he used an ordered logit/cumulative logit model to conduct the tests. I will here summarize his analyses for the positive arguments. To keep things brief, and within the bounds of what's reasonable for a blog post, I'll only report the significant differences.
Brief summary:
Religious belief strongly affects the extent to which participants rate each of the arguments. Controlling for this, philosophers of religion evaluate the following arguments more favorably: design argument, cosmological argument, argument from religious experience, and the argument from miracles. We also found some gender effects. Women rate the ontological argument and the pragmatic argument more favorably than men, men rate the cosmological argument more favorably than women, and there is an intriguing interaction between gender and religious belief for the argument from beauty and the argument from miracles.
Discussion:
As you can see in the detailed report below, theism/atheism/agnosticism is in most cases the strongest predictor of how favorably participants rate arguments. In some cases, a belief by gender interaction proved to have a particularly strong effect (see e.g., the odds ratios for female theists compared to male atheists for the argument from beauty: female theists are 33.34 times as likely as male atheists to assess this argument favorably).
What I found particularly interesting was that philosophy of religion had a significant effect in a positive direction for several arguments. So regardless of whether or not philosophers of religion are theists, they tend to rate several arguments more favorably, with the strongest effects for the design argument (odds ratio 1.76) and the cosmological argument (odds ratio 1.53). I am wondering for possible explanations for this. It is probably not the technical nature of these arguments, as the ontological argument in its various forms is also quite technical. I am inclined to think that this provides some support for these arguments, since even atheist philosophers of religion are more inclined to rate these arguments strongly, but I am aware that many NewApps readers would disagree! It does tie in with my anecdotal observation that non-PoRs frequently dismiss arguments (e.g., only recently, a philosopher of science said to me "The fine-tuning argument, it's just ridiculously weak"). Perhaps one needs to be acquainted with arguments in order to properly assess their full strength, in which case expertise could explain the difference (it will be interesting to consider if the same holds for the negative arguments - which we will see next week.)
I cannot even begin to explain the gender effects. Recent work by Buckwalter and Stich found gender differences in philosophical intuitions, so I suppose this work ties in with this. It could also be due to sampling bias (remember my difficulties in finding female participants). In any case, the stronger evaluation of philosophers of religion of positive arguments, regardless of their religious beliefs, is interesting for the debates on disagreement in philosophy of religion. Rowe argued in 1979 in defense of friendly atheism: according to him, atheists could maintain, coherently, that they have strong evidential grounds for atheism *and* that theists are reasonable in disagreeing with them. In the next installment, I will discuss whether philosophers of religion also rate some individual negative arguments more strongly.
Detailed results
- Design argument: The strongest predictor of positive assessment of the design argument is religious belief. In the sample, theists are 12.15 times as likely as atheists and 5.84 times as likely as agnostics to rate the design argument more favorably (p < 0.0001 for both statements). Agnostics are 2.08 times as likely as atheists to rate the design argument more favorably (p < 0.0001).
Controlling for theism, philosophy of religion also influences assessment of the argument. Participants who are philosophers of religion are 1.76 times as likely as those who are not to rate the design argument more favorably (p-value = 0.0005). - Cosmological argument: Again, religious belief is the strongest predictor. In this sample, theists are 23.12 times as likely as atheists and 6.93 times as likely as agnostics to rate the cosmological argument more favorably (p < 0.0001 for both statements). In this sample, agnostics are 3.34 times as likely as atheists to rate the cosmological argument more favorably (p < 0.0001). Philosophy of religion is also a predictor: participants who are philosophers of religion are 1.53 times as likely as those who are not to rate the cosmological argument more favorably (p= 0.01). Interestingly, gender is also a significant predictor: males are 1.52 times as likely as females to rate the cosmological argument more favorably (p= 0.01).
- Argument from religious experience. Theists are 12.2 times as likely as atheists and 3.71 times as likely as agnostics to rate the argument from religious experience more favorably (p < 0.0001 for both statements). Agnostics are 3.29 times as likely as atheists to rate the argument from religious experience more favorably (p < 0.0001). Philosophy of religion as an AOS also has an effect. Controlling for religious belief and gender, participants who are philosophers of religion are 1.39 times as likely as those who are not to rate the argument from religious experience more favorably (p-value = 0.042).
- Pragmatic arguments for theism. Both gender and religious belief affect how participants rate pragmatic arguments. In this sample, theists are 4.74 times as likely as atheists and 2 times as likely as agnostics to rate pragmatic arguments in favor of theism more favorably (p-value < 0.0001 and p-value = 0.0004, respectively).Agnostics are 2.37 times as likely as atheists to rate pragmatic arguments in favor of theism more favorably (p-value < 0.0001). Females are 1.59 times as likely as males to rate pragmatic arguments in favor of theism more favorably (p-value = 0.005).
- Ontological argument. Here too, gender and religious belief affect how participants rate this type of argument. In this sample, theists are 7.76 times as likely as atheists and 3.83 times as likely as agnostics to rate the ontological argument more favorably (p-value < 0.0001 for both statements). Agnostics are 2.02 times as likely as atheists to rate the ontological argument more favorably (p-value = 0.0002). Females are 1.43 times as likely as males to rate the ontological argument more favorably (p-value = 0.029).
- Argument from beauty. Here's where things get messy: there is an interaction between religious belief and gender. In this sample, male theists are 26.83 times as likely as male atheists and 7.72 times as likely as female atheists to rate the argument from beauty more favorably (both p-values < 0.0001). Male theists are 5.71 times as likely as male agnostics and 2.79 as likely as female agnostics to rate the argument from beauty more favorably (p-values < 0.0001 and p-value = 0.0007, respectively). Female theists are 33.34 times as likely as male atheists and 9.6 times as likely as female atheists to rate the argument from beauty more favorably (both p-values < 0.0001). Female theists are 7.1 times as likely as male agnostics and 3.46 times as likely as female agnostics to rate the argument from beauty more favorably (p-value < 0.0001 and p-value = 0.002, respectively). Male agnostics are 4.7 times as likely as male atheists to rate the argument from beauty more favorably (p-value < 0.0001). Female agnostics are 9.62 times as likely as male atheists and 2.77 times as likely as female atheists to rate the argument from beauty more favorably (p-value < 0.0001 and p-value = 0.003, respectively). Female agnostics are 2.05 times as likely as male agnostics to rate the argument from beauty more favorably (p-value = 0.035). In this sample, female atheists are 3.48 times as likely as male atheists to rate the argument from beauty more favorably (p-value < 0.0001).
- The argument from miracles Here it gets really complex, as there is a belief by gender by philosophy of religion interaction. In this sample, male theists are 27.27 times as likely as male atheists and 12.7 times as likely as female atheists to rate the argument from miracles more favorably (both p-values < 0.0001). Male theists are 10.26 times as likely as male agnostics and 3.56 as likely as female agnostics to rate the argument from miracles more favorably (both p-values < 0.0001). Female theists are 20.93 times as likely as male atheists and 9.75 times as likely as female atheists to rate the argument from miracles more favorably (both p-values < 0.0001). Female theists are 7.88 times as likely as male agnostics and 2.73 times as likely as female agnostics to rate the argument from miracles more favorably (p-value < 0.0001 and p-value = 0.0074, respectively). Male agnostics are 2.66 times as likely as male atheists to rate the argument from miracles more favorably (p-value = 0.0008). Female agnostics are 7.66 times as likely as male atheists and 3.57 times as likely as female atheists to rate the argument from miracles more favorably (p-value < 0.0001 and p-value = 0.0001, respectively). Female agnostics are 2.88 times as likely as male agnostics to rate the argument from miracles more favorably (p-value = 0.0011). Female atheists are 2.15 times as likely as male atheists to rate the argument from miracles more favorably (p-value = 0.012). Participants who are philosophers of religion, regardless of sex or belief, are 1.42 times as likely as those who are not to rate the argument from miracles more favorably (p-value = 0.043).
Recent Comments