Wednesday, March 30, 2011

History is over, if you want it

A recent paper by is making the rounds, concluding that religion is heading for "extinction" in certain parts of the world. Having built up half a head of steam to write a few words on it I see that Brandon Watson has beat me to much of what I was going to say. (See also John Wilkins and John Pieret, though whom I originally became aware of it).

In brief, the paper, by Daniel M. Abrams, Haley A. Yaple, and Richard J. Wiener, argues that it can be mathematically demonstrated that current declines in religious affiliation in several secular democracies will trend to zero in coming years, owing to the same sort of diminishing returns one sees when fewer and fewer people speak a traditional language (e.g. Quechua).

Brandon begins with the most immediate and obvious objection that religious affiliation is not coextensive with religious belief or observation. To underscore this, I would note the commenter on the BBC website who writes "I don't see the decline of religion to be a particularly bad thing. And I say this as a Born-again follower of Jesus."

We should keep in mind that the study relies on census data to track membership in a religion. This data is probably accurate enough as far as it goes (despite complicating factors like the Jedi Census phenomenon), but it primarily measures social labels, not belief or practice. There is. obviously, a widespread discouragement with "organized religion" in the West, though in it sell this no more signals a disinterest in religion or spirituality than the rise of the political "Independent" in the US signals a disinterest in politics.

I have a little bit more trouble with Brandon's second point, that the authors of the study don't really mean what they say when they suggest religion will trend to zero, because their conclusions are based on an overly idealized model. He writes that their conclusions
should really should be understood to be qualified by "if no significant countervailing factors arise." (The authors are quite above-board about the fact that they are abstracting from things that could have real effect -- this is an 'assume a cow is a perfect sphere' sort of exercise, to get an idealized model that is at least reasonably close to serve as a starting point for further work.)
True, they are "above-board" about the fact that they are eliding any complicating factors, but only deep within the bowels of the paper. In the summary, introduction, and conclusion, they clearly imply that their model validates to the real world, writing, for example, that the "data suggest a particular case of our general growth law, leading to clear predictions about possible future trends in society." (Namely, the "continued growth of non-affiliation, tending toward the disappearance of religion.") This is an extremely strong claim, which the authors take few pains to qualify with disclaimers that it is a preliminary finding.

When we talk about idealized theories that are later subject to fine tuning on the particulars, we usually don't have in mind the kinds of fine tuning that completely shatter the premises of the theory. Copernicus and Galileo, for example, theorized that the earth travelled around the sun in a circular orbit. The math didn't work out, because, as Kepler showed, this orbit was actually an ellipse, not a perfect circle. But the main point, that the earth was not the stationary center of the universe, proved sound.

The main point of Abrams et al is that religion is subject to being driven to extinction by secularism in a Darwinian contest. This presupposes at least two major things that are not controlled for in the study:

(1) That people choose (and cleave to) their religion for its "perceived utility."
(2) That adherence to religion in the general sense can be extrapolated from individual instances, like Catholicism.

Brandon writes that he finds the first assumption plausible (though insufficient.) This seems to me quite an understatement. Granted, there are numerous historical cases of people converting for self-interested reasons. In the 16th century, the upper classes in the Balkans, for example, living under the Ottomans, were able to get a far more favorable tax rate if they converted to Islam, which large numbers of them did. And there is at least one case of even a putative Messiah, Sabbatai Zevi, converting from Judaism to Islam to save his neck, in 1666.

But for the model of Abrams et al to work, this motive has to function at the exclusion of all others. This requires stretching the concept of "perceived utility" so that it is all-encompassing. How to explain the martyrdom of so many of the early Christians, when it would have been so much easier to go along as a pagan? We can invoke the rewards of the afterlife in this case, perhaps, but only at the price of the presumption of Darwinian zero-sum competition between social groups. Similar problems are presented by the Falun Gong in China, the Crypto-Jews in Spain and the New World, and myriad similar cases. If "perceived utility" is allowed to become a placeholder for "the rationale for whatever people ultimately decide to do," it loses much, if not all, of its predictive power.

Metaphysical beliefs, whether "religious" or not, indicate one's orientation to the profoundest truths, whether that means that the universe is meaningless, that earthly life is all there is, that there is a loving god, a natural order, that "all is one," that there is an eternal return, or that a race of immortals mocks us from on high. Brandon is absolutely correct to suggest that, under pressure, these beliefs are subject to modulation. Some of these pressures are private--we can "lose our faith" or "get religion." Some are social, as in Brandon's example of anti-Catholic legislation, though I think he overstates this influence on actual beliefs and observances, as opposed to publicly visible religious identification. But "perceived utility" cannot be the primary driver of these beliefs, as I think would become clearer to the authors if they would inquire into their own reasons for believing as they do.

The second assumption is harder to parse, mostly because the very concept of "religion" itself is such a recent one. There is no secular/religious divide in most pre-modern societies. The Greeks had no cognate for it, nor the Chinese. Pascal Boyer goes so far as to suggest there is really no such one thing as "religion," and I while I think the term still has a lot of value,  I think Boyer's argument is helpful in pointing out just how many disparate social forms are subsumed in a single word. In a passage that demonstrates just how greatly their work could have benefited with a greater intimacy with the history and sociology of religion, Abrams et al write
We speculate that for most of human history, the perceived utility of religion was high and of non-affiliation low. Religiously non-affiliated people persisted but in small numbers. With the birth of modern secular societies, the perceived utility of adherence to religion versus non-affiliation has changed significantly in numerous countries[11], such as those with census data shown in Fig. 1, and the United States, where non-affiliation is growing rapidly[18].
There's no need to speculate. This account bears little resemblance to the actual history of religious belief and observance.  It's important to note here two things: the alleged existence, in "small numbers" of a quasi-atheist vanguard waiting since the beginning of history for the "birth of modern secular societies" to give rise to the conditions of their ascendance, and (2) the implicit fealty among the religiously affiliated, at the time of this birth, not just to their own beliefs but to "religion" itself. Dennett calls this phenomena "belief in belief," a phrase that only makes sense if you subscribe to the notion that to be secular is to have no faith at all--to believe in nothing for which you have no evidence. This is the "post-metaphysical" myth ginned up by the logical positivists a century ago, and still embraced by those in the thrall of the "Conflict Thesis" (among whose number would appear to be the authors of this study), wherein science and reason are the crucibles of truth, and religion is the sum of all the delusional impurities waiting to be boiled away.

I've said more on this elsewhere. I mention it now only to show the unfortunate influence of the Conflict Thesis on Abrams et al's reasoning, leading them to suggest there is something exceptional and irreversible about the transition in Western Europe from Christianity to secular modernism, something that makes it different from  (for example) the transition from paganism to Christianity. The authors predict not just the inevitable extinction of the local religion, but the extinction of Religion, period. This would imply that the conditions favoring the perceived utility of secularism, are not, like every ideological movement that preceded it, subject to ebb under the influence of some future scheme--that secularism is not merely non-religious, but post-religious. This assumption may in fact be true, but there is no explanation for it in the authors' model.

Finally, a short word on the strange passivity of Abrams et al's approach to the phenomenon of social change. The existence of religion and secularism are no doubt valuable, in differing ratios, to the readers of their paper, all of whom comprise part of the social ecology that would contribute to the "perceived utility" of adhering to these allegedly competing groups. If the authors had written on the decline of trade unionism in the West over the last half century, and predicted, either blithely or wistfully, its ultimate demise, would we nod sagely at the accuracy of the modeling? Or would we remember that as stakeholders in our own society we can participate in its direction, through rhetoric, political action, economic action, and a number of other social modes (including the writing of scientific papers)? Culture and society are a conversation we have, not a movie or slide show we passively watch. Predicting its trajectory is fine for Vegas oddsmakers, but the rest of us can never have too many reminders that the world is not just the aggregate of what other people do. There is no person named “Society” whose behavior we can observe from behind a two-way mirror. I have no serious problem with the secularization of Europe, nor, depending on how you conceive it, with the end of religion as we know it. But any prediction of what is going to happen by our own collective hand, for good or ill, that doesn’t take into account the facts of agency or intention, is of very dubious worth.

6 comments:

Yemon Choi said...

I'd noticed the arXiv paper that you mention and after a quick scan was minded to dismiss it as one of a genre, whereby people who know some techniques for mathematical modelling try to apply it to "real-world" (=research fundable) settings. So I think you may be in danger of reading too much into the thought processes of its authors.

(I speak as someone with no real training in history or philosophy, of religion or of anything else, but as a mathematician who has read and written his share of "need to find a hook to sell this thing at the bottom of my desk drawer" papers.)

underverse said...

I'm not sure that's much of a defense. People in the real world take what scientists say seriously. Indeed this paper is being widely reported on in the major media. So if they just intended it as a funding hook (which I hope is not the case), there seem to be some pretty negative externalities associated with that strategy.

Yemon Choi said...

I would not like to give the impression I approve of what the authors seem to have done. However, academic courtesy, and the knowledge I don't know enough about their prior work or motivations, mean I don't want to be as blunt here as I might be.

Let's just say that I am not asserting-with-100%-confidence that their paper is a fancier version of "formulas for optimal dunking of biscuits in tea". Nor am I claiming that such papers must necessarily be *entirely* meretricious...

Yemon Choi said...

See also http://xkcd.com/793/ which was my first reaction to that paper...

Jeff Engeman said...

Hurrah! Underverse is back. I keep faithfully checking day in and day out...I really enjoy your blog.

Thanks for this post and for the link to the Boyer post. It's helping me articulate to myself some rather vague thoughts that pop into my head when my friends say "Religion is Bad, mmm...kay." (Those vague thoughts being: "but, but, religion is such a BIG BASKET of a word! How can it be 'bad'?"

I'm not all that convinced of Boyer's contention that "Religion" doesn't exist. I'll have to read his post more carefully later, but my first impression is that if "Religion" doesn't exist, then similarly "Government" or "Law" or other phenomena represented in language with BIG BASKET words don't exist either.

underverse said...

I think Boyer's point is a rhetorical one, that we should be careful not to oversimplify everything in the "religion" basket as being of a singular nature. Maybe he's just doing penance for letting his publisher talk him into naming his book "Religion Explained."