1.
We Bayesian statisticians like to say there are three kinds of statisticians:
a. Bayesians;
b. People who are Bayesians but don’t realize it (that is, they act in coherence with some unstated probability);
c. Failed Bayesians (that is, people whose inference could be improved by some attention to coherence).
So, if a statistician does great work, we are inclined to claim this person for the Bayesian cause, even if he or she vehemently denies any Bayesian leanings.
2.
In his autobiography, Bertrand Russell tells the story of when he went to prison for opposing World War 1:
I [Russell] was much cheered on my arrival by the warden at the gate, who had to take particulars about me. He asked my religion, and I replied ‘agnostic.’ He asked how to spell it, and remarked with a sigh: “Well, there are many religions, but I suppose they all worship the same God.” This remark kept me cheerful for about a week.
3.
In an op-ed today, Ross Douthat argues that celebrated atheist Christopher Hitchens actually “intuited” that there was a God:
In his very brave and very public dying, though, one could see again why so many religious people felt a kinship with him. When stripped of Marxist fairy tales and techno-utopian happy talk, rigorous atheism casts a wasting shadow over every human hope and endeavor, and leads ineluctably to the terrible conclusion of Philip Larkin’s poem “Aubade” — that “death is no different whined at than withstood.”
Officially, Hitchens’s creed was one with Larkin’s. But everything else about his life suggests that he intuited that his fellow Englishman was completely wrong to give in to despair.
I have a horrible feeling that if Bertrand Russell had supported the Vietnam war, Douthat would be enlisting him on the side of Christianity also.
P.S. Yeah, I read that Ray Monk biography too. Bertrand Russell had a lot of problems. Still, that was a great quote.
What a horrible and remarkable misreading of a very nice article on Hitchens.
Thomas:
I think you should be arguing not with me but with Douthat. He’s the one who speculated on what Hitchens was intuiting!
By saying, repeatedly, that Hitchens was an atheist? Perhaps Douthat has a broader understanding of what it means to be a Christian than most.
I didn’t read it as Douthat saying Hitchens was really a theist. Ross said Hitchens’ atheism was motivated by rebellion, rather than a presumably depressing materialism. So even if they disagreed on god, theists and Hitchens had a kinship in that they didn’t resemble the straw-atheist Douthat has imagined.
“everything else about his life suggests that he intuited that his fellow Englishman was completely wrong ”
“I work all day and get half drunk at night” sounds fairly Hitchenish. Douthat’s picture of atheists (“bloodless prophets of a world lit only by Science” who have no time for literature) is juvenile.
http://theleastobviousanswer.blogspot.com/2011/12/in-heaven-christopher-hitchens-debates.html
Enjoyed the title, reminding me of the alternative corners of the analogy, e.g. “Ross Douthat is a Bayesian” or “Coherent Frequentists are Christians”.
On the other hand following the Douthat line, all Bayesians are secretly making Minimum Description Length inference – they don’t appreciate it yet. Poor deluded folk. Still, they’ll come around. What’s more, any amount of protest that that is not what they are doing proves that they really are doing it. Handily, these arguments also work to prove the converse.
But maybe they won’t ‘come around’ and that isn’t ‘what they’re doing’ because the first argument is an instance of the form ‘people whose behaviour we approve of must share our principles’ which only works when one also assumes that those principles and no others uniquely imply the behaviour, and the second argument is plainly unfalsifiable.
I think it’s broader than Bayesians or other types of believers. We humans tend to think that people we like and admire are like us in ways that they are not. The Bayesian thinks if a statistician does great work he/she must be a Bayesian (even if unknown); the Christian thinks thus about atheists who are prolific, profound or charming. But so too do liberals and conservatives, for example.
Maybe this is part of the halo effect?
More generally it’s a data imputation issue: I impute your principles by assuming that they are MAR with respect to an imputation model that describes how principles translate into behaviour _that is true of me_. This is a dumb strategy when it disobeys the rule of imputation models, which is that they should not be _less_ representationally flexible than the model you use for inference. Here the imputations bias the inference by simply reading themselves into the data.
The more philosophically inclined will recognise this as an application of a Quine / Davidson ‘principle of charity’ where, in a ‘from scratch’ interpretation problem, I assume you think what I think in order to start understanding what you saying. The idea, roughly speaking, is that I have to hold constant by assumption one of a) what you believe and b) how you talk about it in order to identify divergence from you with respect to the other one. But this is meant to be just about getting going with interpretation. Just assuming that plentiful linguistic evidence, about atheism in this case, is so unexpected it should be disregarded is just silly.
We Bayesian statisticians like to say there are three kinds of statisticians:a. Bayesians;
b. People who are Bayesians but don’t realize it (that is, they act in coherence with some unstated probability);
c. Failed Bayesians (that is, people whose inference could be improved by some attention to coherence).
So, if a statistician does great work, we are inclined to claim this person for the Bayesian cause, even if he or she vehemently denies any Bayesian leanings.
Oy! And you want to know why Bayesians so often are regarded as dogmatic?
What I don’t understand is this need to try and find that “even so and so is a Bayesian”—as if that gives reasons. Why so much insecurity?
Mayo:
I was joking! I was analogizing Douthat’s very real claim that an avowed atheist is really an “intuitive” believer, to the joke that Bayesians are religious proselytizers. To just bang on Douthat seemed boring so I made the Bayesian connection.
I don’t really believe everyone’s a Bayesian. What I believe is that thinking about Bayesian models can be a good way to understand statistical methods, even those methods that are avowedly non-Bayesian. Similarly, Douthat may very well feel that Christian belief is a useful filter for understanding the behavior of nonbelievers. My problem is he didn’t stop there, he couldn’t resist making the evidence-free claim that Hitchens “intuited” something contrary to everything Hitchens had actually said. This seemed to me to be rude and clueless.
The best anyone could do during Hitchens’ lifetime was to get him to say that he might be a hermaphrodite. Now that he’s dead they can finally get him to admit that he’s religious.
To be fair to Douthat, the implication was more the intuition that “life is meaningful” or that “despair is, in some objective/’official’ sense, ‘wrong'”, rather than “there is a God”. He did seem to be trying to steer clear of the “redeeming Hitchens” approach for most of the article at least.
(Then again, about 5 or 6 people have said this already here.)
After seeing your comment I read Douthat’s piece a few more times and I can’t agree. Douthat argues Hitchens was “not so much a disbeliever as a rebel”, and follows it up with the suggestion that Hitchens thought “that his atheism was mostly a political romantic’s attempt to pick a fight with the biggest Tyrant he could find”. Douthat is clearly suggesting that Hitchens thought the tyrant exists. If Douthat thought Hitchens’ intuition was more about “life is meaningful”, why does he write “When stripped of Marxist fairy tales and techno-utopian happy talk, rigorous atheism casts a wasting shadow over every human hope and endeavor…”?
I just can’t see Douthat’s piece as an expression of goodwill: it’s too underhanded.