Saturday, April 16, 2016

Sam Harris should communicate social science better

I like Sam Harris a lot. His ability to identify and communicate novel angles to stale topics in the public sphere is impressive: popularizing a neuroscience approach to religion and free will, repackaging meta-utilitarianism and critiquing the is-ought divide, trying to bridge atheism and spirituality. All excellent.

But unfortunately his forays into the realms of social science and public policy haven't been up to par.

His schtick on these topics is a clever one, from a branding perspective. He's taken stands opposing the typical liberal position on social issues that are especially complex and multi-level, deliberately making arguments so simplistic as to be shocking.

The best example is his support for racial profiling in security systems. You can read his position get completely obliterated by Bruce Schneier here, but essentially Harris argues that statistical racism is warranted due to the overabundance of terrorists with the physical trait: 'Middle Eastern-looking male'.

Now. Trivially some statistical racism is a necessary tool in the antiterrorism toolbox. Harris knows that unabashedly discussing statistical racism is exactly the sort of contrarianism that can spark discussion and visibility. Fair enough.

But in this specific case, statistical racism is simply a bad strategy because terrorist behavior isn't set in stone: it's dynamic, and can adjust to security strategies. Any security benefits of profiling would be quickly adapted away, leaving only the social and symbolic costs of a strategy generally regarded as conflicting with core US ideals.

Harris has since generalized his focus on Islam, arguing that its unique qualities warrant a special focus compared to other religions. A comprehensive discussion of his argument was recently made in this Salon interview.

The interview is quite good, and Harris makes some excellent points. But he also makes some terrible points, which undermine the best version of his overall claim. I'd like to unpack this and reframe what I think is his core message.

How Islam relates to outcomes in the world

The first three questions of the interview basically ask the same thing: with so many factors influencing Islamic terrorism and bad outcomes in the Middle East (history, geopolitics, culture, etc.), what makes the causal effect of Islam as a religion especially clear-cut and worthy of special emphasis?

In question #3 we get a proper response:
"I agree that the history of colonialism isn’t pretty. But the example you raise just proves my point. In fact, this practically became a science experiment that dissected out the crucial variable of religion. There are (or were) Christians living in all these beleaguered countries. How many Christian suicide bombers have there been?"
Harris flips the question, using the multitude of factors to support his factual claim about the causal effect of religion. Later in the interview, Harris admits that Islam's strong causal relationship to terrorism and unfreedom is in fact contingent; i.e. its power is enabled and shaped by other factors like history, geography etc.

Here is the best version of Harris' argument (in my view):
  1. Religious doctrine (obviously) has a causal impact on the world, which is mediated by many other variables.
  2. There is no fundamental reason to think that all religions have an equivalent magnitude of causal impact on the world; some might matter more than others because of inherent, contextual, or temporal factors.
  3. Similarly, the net social benefit is not necessarily of equivalent direction for all religions: some might fit better with the current global institutional setup and modern cosmopolitan conceptions of justice.
  4. Religion's causal effects operate on many levels, from the individual to the entire earth, with a high degree of interaction between levels.
  5. Given the evidence, Islam has a strong negative causal impact of the world as measured by universal metrics.
  6. People should work hard to reshape Islam to minimize its negative effects.
The first four points are fairly ironclad logic, basically summed up as, "other stuff matters but religion still matters too." Harris would greatly help his cause by emphasizing this unobjectionable point more. The factual point of #5 and the normative point of #6 are both controversial.

Religion is bad, but for who?

Claims about religion's net negative effect on the world often fail to acknowledge the tremendous social benefits that individual participants receive. This directly runs up against the normative view that we should discourage it.

On the global scale, dogmatic religion probably holds back the wider adoption of the scientific worldview. It also has negative effects on individuals in many cases. That's bad. But although a 'no religion' world might be better for almost everyone, the process of getting there might be worse for almost everyone. Like many strange cultural phenomenon, there's a collective action dynamic operating that makes change exceptionally difficult and perhaps morally problematic.

Unintended consequences

Arguing against religion within the marketplace of ideas is great, but truly vigorous action runs a huge risk of unintended consequences. Most obviously, we might be wrong about the costs and benefits of religion and end up doing more harm than good. Teasing out the total causal effect of religion is essentially impossible, so lots of militant atheists have a certain intellectual arrogance (no more than religious people or reflexively pro-religion leftists, however).

Most anti-religion approaches are also highly susceptible to abuse and co-option. Some of the rhetoric Sam Harris employs is similar in spirit to that used by actual bigots (part of his shock-appeal), and much of the backlash against him is actually a byproduct of the largely benevolent norm against brazen discussions of the moral qualities of culture. Most leftists are wrong to decry Harris' factual points, but they are doing so for essentially the right reasons.

Why emphasize religion?

Even if one thinks religion is terrible and working to abolish it would make the world a better place, incrementally and ultimately, it still might not be the best topic to spend time on, for a few reasons.

First, it sends a bad 'meta-message'. It's like those conservative Christians obsessing over bathroom policy and sexual mechanics--why, exactly, are they so focused on that? It might be unfair or suboptimal, but currently most everyone finds obsessing over religion uncomfortable and mean-spirited. That creates a huge headwind to progress which decreases the social return on investment.

Second, as we've seen, the causal channels through which religion affects the world are complicated and confusing. In any project of social change, messaging is key, and I'm not sure I can think of a more difficult topic to discuss.

In general, lots of other intervention points in society have much clearer relationships to outcomes, and it's no surprise that much of the atheist community has migrated towards things like effective altruism.

In my mind, the best angle of attack against religion is to extend and enhance the neutral principle of church-and-state separation. Lots of countries don't have it, and the countries that do would be well-served by strengthening it. Property taxes for churches, child sexual abuse enforcement, clearer education standards--these are all nominally popular things within the modern liberal democratic world, and their implementation would represent a decent move towards a better world.

No comments :

Post a Comment