## RS74 - Live! John Shook on Philosophy of Religion

Release date: November 18 2012

Massimo and Julia visit Indianapolis for a heated debate, in this live episode of Rationally Speaking. At a symposium organized by the Center for Inquiry (CFI), they join up with *John Shook*, Director of Education and Senior Research fellow at the CFI, and the author of more than a dozen books on philosophy and religion. Sparks fly as the three debate questions like: Should science-promoting organizations, like the National Center for Science Education, claim publicly that science is compatible with religion? And is philosophy incapable of telling us anything about the world?

*John's pick: **"Meaning and Value in a Secular Age: Why Eupraxsophy Matters—The Writings of Paul Kurtz"*

## Reader Comments (9)

The guest babbled about how religion has been suppressing science for 400 years, but could give no examples of what he was talking about. Galileo and stem cells were mentioned, but these stories are widely misunderstood.

Massimo called Occam's razor a purely esthetic preference. I do believe that it has a statistical justification, related to the size effect discussed by Perfors et al: (2011) in their paper "A tutorial introduction to Bayesian models of cognitive development".

Perfors et al describe a hypothesis as covering part of a parameter space. The more specific a hypothesis, the more potential observations should not occur, and therefore the greater a region of the parameter space the hypothesis excludes. The more general a hypothesis is, the more conceivable observations it permits. The actually observed data therefore cover a greater proportion of the volume of the parameter space consistent with the more specific hypothesis. In the simplest case, the likelihood (probability of the actually observed data given that a hypothesis is true) is the proportion of the volume of the parameter space covered by the observations out of the volume consistent with the hypothesis. The more specific hypothesis therefore will have a greater likelihood for the same data than a more general hypothesis that is compatible with the same data. The implications of that are best examined by an explicit Bayesian calculation.

Assume we have two mutually exclusive scientific hypotheses, and for the sake of simplicity we divide the parameter space into four discrete hypotheses. Hypothesis A can only be true if one particular observation occurs. Excluding measurement errors, then p(observation1|hypothesisA) = 1. If we allow for some measurement error, we must allow p < 1, but we can neglect that for the purpose of discussing the size effect. The more general hypothesis B is compatible with all four mutually exclusive possible observations. Assume hypothesis B predicts an equal probability of 0.25 for all potential observations. Then we observe data set 1 that is compatible with both hypotheses. Bayes' theorem in odds form states that: p(hA|d1)/p(hB|d1) = p(hA)/p(hB) * p(d1|hA)/p(d1|hB) = 1/0.25 = 4. The likelihood of the actually observed data for the more specific hypothesis is four times greater than the likelihood for the more general hypothesis. Therefore the posterior odds shift by that same factor in favour of the more specific hypothesis when compared to the prior odds.

The connection I see to parsimony is that the more parameters a hypothesis has, the more observations it usually can account for. So as far as I can see, the more complex hypotheses are more general. If that is correct, then a preference for the simpler hypothesis will often be a preference for the more specific hypothesis.

Wow, Robert. I really appreciate your post. It's very clear and thoroughly walks through a very important bit of information that I think has yet to reach the minds of most scientists. I'm thankful some of us (like you!) are paying close enough attention to the progress of Bayes and its applications in science, philosophy, and history.

I'm only up to minute 33, just after Shook says science can only help cultural practices and religion has only hindered them. Totally agree, btw. But I'm writing now to say that I think Massimo is unhelpful with always bringing up logic and mathematics in philosophical debates. Looking forward to listening to the rest and commenting about all of it.

BOOYA! Loved Shook. 400 years of failure sure is empirical evidence. His points about appealing to values and realizing people are emotional decision-makers are spot on and deserve a full episode of a RS podcast.

I also think it is disengenuous to require a particular type of empirical data to allow you to say that there is no empirical data available. I'm not exactly sure what Massimo is thinking about, but it sounded like a forward looking study with some type of control or at least a comparison case. How feasible is that? Are we ever going to have something like that for accommodationism? And even if we did, would it be worth anything?

Related point, Massimo, do you consider the I-was-jolted-out-of-religious-belief-because-of-confrontation (along with or because of accepting evolution) letters to be empirical data touching on this topic, accommodationism?

The distinction between mathematics and chess -- that one thing is discovering and the other isn't -- is specious. Shook is insisting that games are different, but why? Chess has a perfectly understandable mathematical description, and indeed all possible chess games can be written out as a simple tree (though a large tree, with some special treatment for non-terminating situations). Players are merely jointly taking a path down that tree. There's nothing more to it. All games are "boring" in this sense; they completely understood from a mathematical perspective. We can prove theorems about particular paths in the tree and so forth.

Nothing said in the podcast had anything to do with the philosophy of religion.

If you are looking for a Way to unite science philosophy and religion it can be done simply and most elegantly with truth.

Descartes' method worked for me. Einstein was on the path as well. Truth is much more simple than thought.

Remove any uncertainty or doubt from the equation and equal is all that remains.

Truth is,

=