Attention conservation notice: A couple thousand words that can be summarized as: “Someone is wrong on the internet.”
Jerry Coyne has a longish reply to my post yesterday. He seems quite upset about it. He seems to think I’m very, very wrong. And yet he cannot manage to characterize my argument correctly, and offers nothing at all that would refute my argument (or even refute his mischaracterization of that argument). It’s odd. He’s basically conceding that I’m right that no evidence was offered, and he knows of no evidence. He seems upset that I bothered to point out that lack of evidence.
Ophelia’s reply at her blog and in comments here takes a smarter tack (as she usually does). She hasn’t convinced me that I’m wrong, but she and I are narrowing down our disagreements to a quibble over how strongly to interpret a couple of her sentences, and how strongly she meant to assert some causal links. Further discussion might even get us to some useful common ground.
Coyne, not so much. His title indicates his opinion that I have what he considers a “strange attitude toward evidence” (to wit: I would like to see some):
Over at Thoughts from Kansas, Josh Rosenau vigorously disputes a post by Ophelia Benson
And already we have a problem. I didn’t “vigorously dispute” her post. I said “I don’t know if Ophelia is right,” and that I meant no criticism by saying so. My “strange attitude toward evidence” is that one should offer evidence for empirical claims, and one shouldn’t say things directly contradicted by the evidence. Jerry’s already off to a bad start on both fronts. Anywho…
What’s Rosenau’s beef? That he sees no evidence for this Gnu Effect:
Not quite, but correcting every error will get tedious. Coyne’s reply:
Well, yes, there are no formal surveys about the effect of Gnus on popular perception of religion.
Thank you. We agree that no evidence is on offer, and none seems to exist. What’s his beef, then?
it’s curious for Rosenau to criticize this claim on the basis of a lack of evidence…
“Lack of evidence” having nothing to do with gnu atheists’ common main point about theism, and therefore not worth noting in gnu atheist arguments…
…when for several years he’s been claiming that accommodationism weans people from creationism much more easily than does vociferous atheism—on the basis of even less evidence!

Less evidence than nothing?
I don’t think I’ve made claims about what converts people from creationism (Coyne offers no evidence, natch), so much as how to sway the undecided middle. Which is different, and it’s important to be accurate. At least, that’s my “strange attitude.”
Coyne blunders on:
the only thing Rosenau has ever offered in support of accommodationism is a study showing that people tend to trust experts more when those experts share more of their cultural values.
Had Coyne taken the time to read to the end of the post he links, he’d see that I linked to two other posts in which lots of other research is synthesized to show the same basic result. It’s what scientists call “consensus.” I wrote:
There’s a body of relevant research out there that addresses questions like the one Coyne posed above, and if they [gnus] want to argue about communications strategy, they would do well to address that literature.
One study plus dozens of unrefuted studies does not, according to my “strange attitude toward evidence,” equal one study.
He then quotes from Jason Rosenhouse’s reply to that post, without mentioning that I have addressed Rosenhouse’s points. Jason and I still disagree about the issues raised there as do the commenters here, but it was hardly the rout Coyne would have you believe. Again, Coyne seems to be representing the evidence in a less-than-accurate manner, which I naively still find strange.
He proceeds:
Are there any data bearing on this? Well, mostly anecdotes…
Which are famously not data…
… But let’s look at the anecdotes …
Or, you know, not. I want data.
He cites a debunked story on a blog, and various stories on another website which no one has bothered trying to debunk (or verify) yet. Meh. Altar calls are not the most reliable form of evidence for anything.
Rosenau demands the highest standard of evidence from Gnus to support their tactics, but feels that bald and unsupported assertion suffices to support his own.
No. I cite peer reviewed research that supports my hypotheses, and acknowledge that I haven’t got data when I haven’t got data. In exchange, I ask not for the highest standard of evidence, but some kind of evidence. The end of my allegedly vigorous disagreement with Ophelia said:
Maybe the evidence is there. If it is, I don’t know what it shows. Smallish psychology experiments could probably give some useful data to test the hypothesis … If they’ve been done, I don’t know the results. Maybe someone did that experiment, and some other people confirmed it independently. If so, I don’t know about it. The result isn’t utterly implausible, and I’m not saying it’s false. I’m saying I don’t know, and I tend not to trust people who confidently assert empirically measurable facts without actually offering data to support the claim.
This is hardly an endorsement of bald and unsupported assertions, nor a demand for excessive evidence. Indeed, someone interested in evaluating evidence fairly would have to conclude that it’s the exact opposite of what Jerry says. And he thinks I have a “strange attitude toward evidence”?
We’ll skip Coyne’s quotation from Herbert Spencer. It was written in 1852, long before anything like modern creationism, or the modern theory of evolution, existed. Spencer’s specific point is a lot more reasonable today than it was in 1852. Spencer’s broad argument works: the evidence for transmutationism is still being assembled, creationists are offering no evidence, their model is implausible, and what evidence exists supports transmutation. That’s a lot more like my argument than Coyne’s, except there’s more experimental evidence to support accommodationism today than had been published in support of evolution in 1852.
Coyne’s curiously disjointed relationship with the meaning of written words continues:
Rosenau … not only sees no evidence that Gnus have helped erode the respectability of religion, but sees no decline at all in that respectability, Gnu-induced or not:
“Absent some sort of evidence that religion is less intellectually respectable now than it was 10 years ago, this first step in Ophelia’s logical chain fails, and the conclusions go with it. And the paragraph above suggests that intellectual respectability has not been necessary or sufficient for its social desirability in America’s past, so the second link strikes me as dubious and unproven as well.”
Well, I don’t have the statistics at hand, but I suspect there’s plenty of evidence for this.
First, note that I’m not saying there is no evidence, only that I see no such evidence being offered. Maybe respectability is down, maybe gnus had that effect. All I want is for someone to produce data to support the claim.
Second, what Coyne “suspects” will matter a lot more when suspicions become evidence. He justifies those suspicions by noting:
…decline of church attendance … the increase in the number of Americans who characterize themselves as nonbelievers. … the bus campaigns … the Gnu books have been best sellers … the growth of secular, humanist, and skeptical societies…
Meh.
Church attendance is not declining over all, though it is among Catholics. More people have been identifying as nonbelievers for a while, too, and both trends started long before the rise of the gnus. Indeed, both trends seem to be continuing at the same rate, or possibly at declining rates, since the rise of the gnus in 2004. That tends to suggest that the gnus are a result of those secular, secularizing trends, not their cause. Ditto for the bus ads and the best-selling gnu books. Book sales aren’t a measure of epistemic merit, nor necessarily of societal influence. If they were, Richard Dawkins would be far behind Dan Brown and Rick Warren, not to mention the Bible. The fact that there were big markets for those books suggests that the market was primed for atheist books, not that the Horsemen created a new market. Two million English copies worldwide is pretty impressive, but represents a tiny fraction of the English-speaking population. No evidence is on offer that the gnus have done anything other than rile up the already-convinced.
I suspect that if you surveyed the number of colleges who had such societies a decade ago, and compared that to what we have today, you’d see a striking increase. Perhaps somebody can supply this information.
Maybe, maybe not. I don’t know, and neither does Coyne. If those data exist, what does it tell us about the effect of the gnus? I don’t think “I suspect” transforms this sentence into data, and I’m confused that he thinks this would be “a strange attitude toward evidence.” It seems like he’s agreeing with me that someone ought to gather the data, that some relevant statistics may exist, and it would be nice to have them on hand. This was my major point. So why does it seem like he’s trying to disagree with me?
It might be interesting to look at Gallup’s questions about confidence in religion’s ability to answer the problems of the day, which has been steadily declining for a while. There’s more evidence of a post-gnu trend there, than from a question in the same survey about whether people think religion is gaining or losing influence on society (more people thought it was losing influence in the ’90s than today, and the trend has been fairly flat other than a predictable spike for religion’s relevance in 2001). All of this suffers from a post hoc ergo propter hoc error unless someone can separate the many other forces acting on perceptions of religion from the effects of the gnus. And again, Jerry agrees disagreeably.
Now whether the Gnus have contributed to this trend is a different matter, …
No, whether the gnus contributed to the trend is exactly the matter at hand. I was addressing a post by Ophelia Benson which suggests that “gnu atheism can Help” by “relentlessly pointing out that religious belief is not altogether intellectually respectable,” and therefore causing a society in which “religion no longer offers such a desirable kind of identity.” It’s not a claim that there’s a trend independent of gnu atheism, it’s a claim that gnu atheism is a cause. And that’s what I want to see evidence about. In the world of scientists and skeptics where I live, asking for evidence to justify claims is hardly a “strange attitude.”
…but surely there’s evidence for an increased respectability attached to being agnostic and atheist. Can you imagine bus-slogan campaigns 25 years ago? Or a President who asserts the rights of non-believers in his inaugural address?
“Surely” is not evidence either. Maybe atheists and agnostics (Jerry usually calls the latter “faitheists,” so this is a pleasant change) are no less respected, but would like to be. Maybe Barack Obama is a better man than the last few decades’ worth of presidents. The actual data available are slim, but show that willingness to elect an atheist president has declined since 1999 (4 points, right at the margin of error) after plateauing in the low-to-mid 40s. There was no change in attitude towards atheists from 2006 to 2008, the time when gnu atheists were most active. But why must I do Coyne’s homework for him?
Coyne gets upset that I wrote: “Maybe the evidence is there. If it is, I don’t know what it shows.”
I don’t know what it shows? We’re talking about evidence in favor of a thesis! It must show something!
No, we’re talking about evidence. Until we have it in hand, we don’t know whether it’s in favor of a thesis. First you get the evidence, then you decide whether it’s in favor of the thesis, then you decide what, exactly, it shows. Coyne seems to want to do things the other way around, which any thinking person knows is a very “strange attitude toward evidence.”
We can all agree something is surely strange here, but I don’t think it’s my attitude towards evidence.
Joshua Rosenau spends his days defending the teaching of evolution at the