Sep 30, 2012

Haidting on reason

Last Sunday, I attended an oddly irritating talk. The speaker was smart and articulate and I am sure I would agree with her on 90% of randomly chosen social, political, or philosophical issues. The presentation was effective and well done. The topic was certainly important—how to talk with people who disagree with us. And yet, there was so much wrong with it, I felt the need to emerge from blogging hibernation just to respond to it. (I have invited the speaker to visit this blog and comment, so I may not go back to hibernation.)

The talk was largely based on Jonathan Haidt's new book, The Righteous Mind: Why Good People are Divided by Politics and Religion. So I should begin with a disclosure: I dislike Haidt. Although I haven't read his new book, I have followed his writing for years, and have consistently found him to be a member of the most annoying of tribes, the Self-Hating Liberals.

Don't get me wrong: Haidt is worth reading. His analysis of happiness and morality draws on recent research in various disciplines, including neuroscience, and almost everybody will learn a lot from it. He writes well and his arguments often challenge the conventional wisdom. Engaging in a debate with him requires clear and disciplined thinking. But the converse holds as well: clear and disciplined thinking requires engaging in a debate with Haidt, because his most prominent arguments are ultimately destructive.

Haidt got a lot of press a few years ago, when he—a declared atheist—defended religion's contribution to people's happiness. I think it was somewhat unfortunate for his own opus, as most of his ideas are not as inchoate, but he never seemed to object to earning much of his relative celebrity status due to this particular idea, so I must assume that he thinks of it as representative enough of his work. Now he has written a book in which he—a declared liberal—defends the conservative values as an integral part of the "moral matrix" our species has evolved with.

There is a pattern here: Haidt distinguishes himself by asking questions of the type "What if we are wrong?", where "we" are those with whom he identifies through fundamental viewpoints: atheists, secularists, liberals. This is, of course, fine in principle, even desirable, but if one builds a career on those questions, one's success becomes dependent on the challenge actually showing the prevailing views wrong. Merely casting doubt does not make you a famous iconoclast; for that, you must break something.

But Haidt doesn't want atheists and liberals to lose arguments. He still reaffirms his atheism and his liberal values. So he shifts focus from the question of who is right to the one of why people hold the views they hold. He wants to study morality scientifically, dispassionately. That's great, but is has a side-effect: it leads to moral relativism. Haidt is not comfortable with that:
I hope you'll accept that as a purely descriptive statement. You can still reject the three binding foundations normatively—that is, you can still insist that ingroup, authority, and purity refer to ancient and dangerous psychological systems that underlie fascism, racism, and homophobia, and you can still claim that liberals are right to reject those foundations and build their moral systems using primarily the harm/care and fairness/reciprocity foundations.
But this discomfort muddles the argument. It becomes less and less clear what the goal is. We should consider that we may be wrong... But we can still believe we are right... But we should understand those who are wrong different... Maybe we are wrong in the way we interact with them... Can one go down this path and remain coherent?

For those unfamiliar with Haidt's work, I suggest this essay and the responses to it as a good introduction. (The quotation above is from it.) The flaws in Haidt's essay are pointed out very well by Sam Harris, PZ Myers, and Marc Hauser, but they all miss an important point—that Haidt uncritically relies on biased sources—which I pointed out in this blog. (Note that Brooks has since solidified his hack credentials, as he has become the president of the AEI.)

As I said, I haven't read Haidt's new book, so I cannot criticize it directly. What I will criticize is the speaker's representation of it. I believe, however, that it is a fair representation, based on earlier articles that I have read, and which were clearly stages of the road to The Righteous Mind. An example is here. Again, read the responses, too; I particularly like those by Sam Harris (again!) and Roger Schank.

But on to last Sunday's talk. Its frame was a question: How do you talk with your "nut job" relatives, friends, or associates? Yes, we all have people we care about (or at least need to keep constructive working relationships with) who hold (or, worse, spill) ideas that are crazy, ludicrous, idiotic, or of some other similar kind. How can we have a conversation with them?

The speaker identified two major errors that derail such conversations. One, in our "toxic culture", we surround ourselves with those who think alike, and in our minds dehumanize The Other, those with whom we disagree. Two, we overestimate the role of rationality and falsely believe that our values have a rational basis, when in fact we use reason to justify our moral views we have already formed based primarily on emotions. The latter is illustrated by Haidt's metaphor of the rider and the elephant: the rider (the rational functions of our mind) thinks he is in charge, but the elephant (the irrational) is far more powerful.

To avoid these errors, we need to understand human morality: it is rooted in emotions, which give rise to six basic moral values: caring, liberty, fairness, loyalty, authority, and sanctity. Those six values form what Haidt calls the moral matrix. (As I understand, that concept is the main new idea in The Righteous Mind.) According to Haidt's research, liberals mainly care about the first three of those values, while conservatives care about all of them.

The speaker concluded with guidelines for better conversation with "the nut jobs":
  1. Find the humanity in each other. Stop thinking about them as crazy.
  2. Put aside right and wrong. (This includes facts!) Start by making a connection.
  3. Understand their values and speak to them. Listen. Don't stereotype them or assume they think X because they are conservatives (or liberals, etc.)
  4. Don't fall into cynicism (resignation? pessimism?). Do something!
  5. Get out of your box. Speak with people of different opinions. Get out from behind your computer!
After the talk, there was some opportunity for response from the audience. My elephant wanted to talk for an hour, but that would have been rude and against the rules, so I limited myself to expressing disagreement over equating thinking of somebody as mentally ill and dehumanizing that person. Mentally ill people are no less human than the rest. I do think that calling Tea Partiers "crazy" may sometimes be a comparison offensive to the crazy folk, but that's a different issue. I also believe that conservatism will one day be understood as related to conditions currently classified as psychiatric disorders. But anyone who sees the "crazy" people (of any kind) as less than human has a problem more acute than how to speak with ideological opponents.

So what are my other problems with the talk? My guess is that at this point, having read the summary of the talk, but not having heard my objections, most readers think it was perfectly fine. OK, there are some details I haven't mentioned yet, but let's go over the main points first.

My first problem with the talk as a whole parallels a problem I have with Haidt: it is not clear what the goal is. Are we trying to persuade the "nut job relative" we are talking with? Or contribute to the political success of "our" candidate? Or be a better person, not act like a jerk, have better relationships? The speaker mentioned all of the above, so I guess that would be her answer, but those are distinct goals and I am not sure they are always compatible. Certainly the strategy in achieving them ought to be different. Political persuasion usually benefits from some degree of deception, while I would think that honesty is a crucial part of being a good person.

Moreover, I would be close to answering that question with "none of the above". I discuss political issues often, but I don't usually expect to change people's views or influence elections, and I don't think the style of the conversation has much to do with me becoming a better or worse person. I like to test my own views by exposing them and defending them in a debate. It helps me grow, which may be akin to becoming a better person. Hopefully, it helps the other person grow, too, if they are willing to use the opportunity. I also believe that reason has an intrinsic value, and it is reason I want to triumph in a debate. If I happen to be on its side, that's an added benefit, but being right after a debate is the next best thing—and when you add the value of learning, it may be the best outcome.

A direct consequence is that I enjoy serious discussions with people who hold different views and can articulate good arguments in support of those views. If their arguments are poor, I see a serious discussion as a waste of time. Mocking their views becomes much more attractive. It may not be nice, but humor and satire have an intrinsic value, too. Or, if practical considerations preclude that (say, you would offend someone you need to work with), then it is best to avoid talking about contentious issues with them, period.

The second fundamental problem I have with the talk is that it promotes relativism. Not only moral relativism, but even the relativism of facts. I've already mentioned how Haidt ends up stepping into relativism and being uncomfortable with it. The speaker may feel similarly, as she unambiguously identified with one side of the political spectrum. So she seems to carry this relativism half-way down several diverging paths.

One is the idea that facts don't really matter to people if they challenge their views. Yes, this is important if your goal is political persuasion. But nothing in this talk, or in Haidt's work, is particularly useful for political strategy. If you want to get your favorite people elected and your preferred policies enacted, the book you should read is Moral Politics by George Lakoff. You won't learn much about neuroscience research from it, but you will understand how to connect with a conservative (or liberal) audience. Lakoff's explanation of two political camps based on two models of moral values ("strict father" and "nurturing parent") is far more parsimonious than Haidt's six-dimensional moral matrix.

This path also leads to a cliff. In an attempt to convince us how little people care about facts, the speaker tried to demonstrate that "we" reject inconvenient facts just like the "nut jobs" do. Pox on both houses! Both sides do it! Of course, that is my biggest pet peeve. I am happier listening to five ultraconservative nut jobs than to one liberal who keeps saying "both sides do it". So the speaker asked:
When you hear a study that confirms your views, do you ask "What methods did they use? What was that person's background?"
I don't know what irked me more: that she asked that, assuming the answer was obvious (and negative), or that the audience laughed, signaling acceptance of the assumption. But the assumption is preposterous: my answer is clearly "yes". I even yelled from my seat: "Some of us do!" If that was uncivilized, so be it.

Ironically, in another part of the talk, the speaker advised against assuming what other people think:
If there's anything that drives me more nuts than being told I'm wrong, it's being told "Well, you think…" (...) Don't tell me what I think because you're stereotyping me."
Guess what? You told me what I think, and you assumed wrong. I always scrutinize studies, regardless of which side their results support. A meta-study once showed that the majority of papers published in top-tear medical journals are wrong. I am not aware of a similar meta-study in social sciences, but I would bet the results would be even worse. There are a lot of bad studies; there is even more bad journalistic reporting about studies. You should always be skeptical. "Constant vigilance!" as Mad-Eye Moody would say.

Now, being skeptical toward all studies is one thing; thinking that both sides cheat equally is quite another. Note, for example, that Paul Krugman often shows the models and reasoning behind his positions in his blog. His critics, not so much. There is a reason that the saying "facts have a well-known liberal bias" has become a cliché.

(At one point the speaker showed a slide juxtaposing the photos of Ann Coulter and Michael Moore. That was so wrong I wanted to throw eggs. When a conservative does that to me, it is a conversation stopper: the parallel is crazy; if you don't see it, you don't live on the same planet as me.)

Another path to relativism is inherent in the idea of understanding each other's values and finding commonalities instead of thinking of the other as mentally deficient. This is fine, but Haidt's own work shows that it is completely illogical as an advice to liberals. Namely, if there are six fundamental values, and liberals care about three of them, while conservatives care about all six, then the commonalities lie in the three values liberals care about. Conservatives need to understand that only those three are universal values. Don't tell me to look for commonalities in things we don't have in common! In fact, I abhore the values of authority and sanctity. (Loyalty sounds ambivalent: usually good in the private sphere, but bad in the public sphere. But as I've seen Haidt denoting it "ingroup/loyalty", I suppose I'd abhore it, too, if I saw the full definition.) Those "values" are harmful and destructive. Haidt acknowledges that, as I've pointed out. So why should we try to connect with them rather than fight against them? On the other hand, conservatives already care about liberal values, so they should focus on what we have in common and shut up about the rest.

That was tongue-in-cheek, of course. I cheated, by pretending that I accept Haidt's moral matrix. In fact, I find it highly suspect. Liberty? Makes me wonder if he has ever discussed liberty with a conservative. Because the word means very different things to liberals and to conservatives. Ditto for fairness. I don't know, if I read the book, maybe I would find the substance of the moral matrix concept convincing. But, at the very least, Haidt chose unfortunate terms, which do not contribute to a better understanding between liberals and conservatives.

Related to "commonalities" is the advice to speak to the others' values. There were some good examples (although I'd say in each case the victory facilitated by this strategy was minor), but also an example I detested. It was about the environment, and the thesis was that it is the area in which secular people can understand sanctity. Well, this secular person (and liberal, and environmentalist) does not! Please do not ascribe New Age nonsense to me. Please follow your own advice and don't tell me I think X because I am liberal. There is nothing sacred about nature. I just want my descendants to have decent lives, for which they'll need a livable environment. As for appealing to the sanctity of Creation to get the Evangelicals to size down their SUVs and reduce dumping chemicals on their lawns, that's fine if it works, but be aware that it's deception. I have no problem with that; I'll take Machiavelli over Kant anytime. Just don't deceive yourself that you are connecting when you are in fact manipulating.

Yet another path to relativism was short, but steep. Part of advice #2 ("put aside right and wrong") was not to say "You are wrong!" when someone gives you, in speaker's words, "facts you completely disagree with". Ugh. Facts are true or false. My agreement with them is irrelevant. And there is no hope of reaching any kind of understanding if we don't agree on facts. While this may be a good tactical advice for political persuasion, it is a bad advice for true understanding. One thing I agree with, though. I don't say "You are wrong!" Rather, I say "That's not true." It is good to keep disagreement from becoming personal. I try to follow that principle knowing full well that most people won't notice the difference. They'll hear it as "You are a liar!" Nevertheless, let the error be theirs!

I do not mean to imply that the speaker holds a post-modern, relativist view of factual truth. I did not get that impression at all. The message I got was primarily that facts don't convince people, and that's fair. But it's only true in a limited sense. If someone is convinced that Obama is a foreign-born Muslim, there may be no way to disabuse them of that nonsense, but five years from now it will be irrelevant, so insisting on facts now may be a waste of time. But if someone believes the world is 6,000 years old and it will end in our lifetime, that necessarily shapes their views of ethics and policy, and that will not change when someone else is in the White House.

In discussing my two fundamental problems with the talk (incoherence of goals and promotion of relativism), I touched on a lot of its specific points. I'll finish by addressing three more specific claims, which I recognize as versions of oft-stated, but poorly supported chunks of conventional wisdom. (I am not going to quote exactly, but I believe I am paraphrasing fairly.)

Our culture has become toxic because we surround ourselves with people who think alike. This is evident in political polarization by counties or ZIP codes. I don't buy this. The fact is that our political parties have become more aligned with ideology (which may be good or bad, but at least it makes sense) and with urban and near-suburban vs. exurban and rural interests. So it is probably true that voting patterns within ZIP codes have become more homogeneous. (I haven't seen data on that, but I find the claim plausible.) But it says nothing about people intentionally segregating by ideology. You can't jump to that conclusion from voting data, and I doubt there are good ZIP-level repeated surveys of ideological positions. Even if ideological segregation were increasing, would that make the society more "toxic" (in the sense of dehumanizing "others", i.e., tribal)? While I can trace the logic of ideological segregation leading to equating belonging to a community with belonging to an ideology or party, and that leading to reinforcing tribalism, it seems like a stretch, especially in a highly mobile society like ours. Finally, is our culture more toxic than it used to be? I'd say the general trend is quite the opposite. We have become less racist, less homophobic, less distrusting of people with different religious backgrounds, more respectful of people with disabilities, etc. True, we are witnessing an explosion of racism disguised as conservative politics in response to the election of the first black president, but I don't think it can reverse the long-term trends, as evidenced by the fact that virtually all culprits feel a need to disguise their bigotry.

Read and watch the other side's media. You don't have to watch Fox News or listen to Rush Limbaugh; maybe start with George Will. But stop reading just things that reinforce our own beliefs. We need to understand what the argument is on the other side and where they're coming from. This is partly terrible advice and partly just doesn't make any sense. Let's start with things that make no sense. A clear implication is that we are watching "our side's media". But that beast doesn't exist. There is no mainstream liberally-biased media, with the exception of a few prime-time shows on MSNBC, and their bias generally stops at the choice of topics and questions. How many examples can you cite of Rachel Maddow telling lies on her show? (I don't watch any of it, so my question is not rhetorical. But I bet the answer would be very short.) In fact, the only media I listen to is the other side's media. I listen to NPR on my commute—Morning Edition on the way to work, Marketplace on the way home—and every day I hear some right-wing crap. Yes, most of it is crap because there are no good right-wing arguments. And the ones aired on NPR are probably as good as they get.

(OK, I occasionally switch to Democracy Now! in the morning, if I happen to be driving between 8 and 9. They report the facts, unlike other news programs, and then occasionally blame the facts on Global Capitalism, which I guess makes them left-wing. So again, how many times did you hear Amy Goodman lie? Not to mention that few liberals I know ever listen to that program.)

First of all, if you want to understand anything, you have to stop watching all mainstream news programs. You can read the New York Times, preferably supplemented with some foreign news sources. And you need to read a variety of blogs and follow the discussions among them. But if you are a liberal Democrat and hope to find the argument on the other side, you'll be disappointed. There is no quality argument that can reasonably be associated with "the other side". Reasoned conservatives have generally been ostracized by the Republican Party, so they are really in no-man's land. Reading them may be intellectually satisfying, but it won't give you any insight about "your nut job relatives" because I guarantee that your nut job relatives don't get their ideas from, say, Bruce Bartlett.

It's bad advice to read nonsense and try to understand it. At best, you'll waste your time; at worst, your health will suffer. You may as well try to attend fundamentalist sermons and try to understand where they are coming from. If you are not very good at spotting bullshit, you can use it for training, but that's all it can be good for.

I can foresee responses saying that this is too blunt or not constructive because it is just calling those people crazy or stupid. Well, yes, I wasn't the one proposing that we shouldn't. Or at least their arguments. I do believe many right-wing writers are sane and intelligent. They'll just write whatever works to establish them as conservative pundits.

Which brings us to George Will. George Fucking Will!? You gotta be kidding me. Nobody should read that dishonest piece of shit, ever. When did he write anything other than lies and manipulation? And how would reading him be constructive? If you took him as representative of conservative thought, what could you conclude other than that conservatives are lying assholes? (One exception is when Will writes about people with disabilities. Then he can show some empathy, respect, and decency. That is because he has a son with with a disability. I don't know if this makes him less bad or even worse.)

Anyone objecting to these characterizations, I challenge you to provide examples to the contrary.

And the last myth: We need to get out from behind our computers. If we tell a story with a human face from our side, it won't be from behind the computer. Oh, I guess that invalidates this entire post, as I obviously wrote it from behind my computer. It also contradicts the advice to read conservative pundits; guess where they write their columns. Seriously, abstaining from any one mode of communication, let alone the most pervasive one, will not help anyone get their message across. Sure, communication on the Internet includes idiotic flame wars and posting comments on articles that get thousands of comments that no one sane reads. But writing a blog and judiciously posting comments on blogs is a way to reach a much wider audience than most people can realistically hope to reach face-to-face. Same holds for message boards. Even Facebook is useful, if nothing else, to avoid repeating the same thing dozens of times to people who may or may not want to hear it. In fact, as soon as I publish this post, I will mention it on Facebook. It will be the fastest and least intrusive way to invite the people who may be the most interested in reading it.