Andrew Gelman has a post up titled Difficulties in trying to understand the views of others, responding to a Robin Hanson taxonomy outline the motivations of liberals, conservatives and libertarians. Gelman is skeptical of Hanson's glosses of each group. The human ability to engage in Meta-Representation is one of the hallmarks of our species. We can analyze abstract ideas, take the positions of others, examine counter-factuals and what-if's. In terms of core competencies our Theory of Mind is a sharp knife, we are unparalleled at modeling social relations contingent upon the mental states of other human beings and how they might react to a huge range of inputs. Many scholars have made the case that core first order cognitive competencies such as Theory of Mind, Folk Physics, Meta-Representation, etc., are the units from our more complex mental activities or cultural productions are derived. Scientific models for example are abstractions of reality which allow us to examine alternative outcomes as we shift the parameters of the system. Many cognitive scientists would argue that our belief in supernatural agents, gods and ghosts, is only possible because of our well developed Theory of Mind. What does this have to do with Robin Hanson's post and Andrew Gelman's response? I believe that the ability to understand immediately how your friend would react if you told them that you observed their significant other being very physically affectionate with someone who was not known to you in public often misleads people into thinking they can also predict and model the reasons why someone is a liberal, or an atheist, or a theist, or decides to join the military. The powers of perception we have about social relations when it is analogous to what one sees among hunter-gatherers does not mean that we have a good understanding of identities and opinions people who grow up and develop in mass societies with many competing institutions and beliefs. An average human in a medium sized city can speak to more humans in one day than a hunter-gatherer of the past might have in their whole lifetime. I know this problem of modeling the reasons behind the views of others from firsthand experience. Over the past 7 years of my involvement in the blogosphere I have run weblogs with a relatively mixed political orientation, and this has caused problems. Granted, there are more libertarians among my readership than is the norm among the general population, but 2/3 of my readership is probably Right (libertarian + conservative) and 1/3 Left (Left-liberal, socialist, etc.). This is a balanced enough ratio that if a politically relevant topic does come up the arguments and misunderstandings can become rather intense. On many political weblogs there are trolls of the "opposition," but in general the regulars agree on the core principles. This just isn't so on my weblogs because most of the time when talking about scientific or other intellectual issues where politics is not relevant. Only when posts, or discussion threads, go into political territory do the differences jump out. One of the interesting things I've noted is that instead of offering up a positive assertion of one's own position, many of the arguments involve characterizations of one's interloculater. At which point said interloculater has to correct what they perceive are mischaracterizations of their own positions and motivations, as well as get into the game of deconstructing their opponents motives. I believe that the power of introspection in social relations fools people into thinking that they understand on some deep level why someone else holds to a particular position on more abstruse topics of less immediate personal impact. There are several problems with this: 1) Even at its simplest our Theory of Mind does not give perfect answers. Humans have varied dispositions, so people often misjudge each other. That's fine as it goes when it comes to a given social situation because there aren't iterations of assertion, response, counter-assertion, etc. When it comes to more complex verbal arguments the back and forth errors propagate as people misjudge each other, to the point where the conversation is useless by the end as people don't know what the other person is even trying to say. 2) Variation in disposition when it comes to anger, falling in love, and so forth, seems to exhibit a narrower range than dispositions which affect political viewpoints. It may be that evolutionarily there are different personality morphs when it comes to traits like openness to novelty that affect political viewpoints, and these morphs are different because each has an equilibrium fitness at a given frequency. Or it may be that political attitudes very a lot because they're a compound of many traits, and so the resultant combinations produce different outcomes, and the more varied traits you have the more combinations you can produce. In any case, it seems that there is less difference between any two individuals when it comes to what their response to a personal betrayal would be than how they view marginal tax rates. 3) People falsely perceive themselves to be totally rational and reflective beings who have absolute conscious control of own own cognitive processes. The reality is that there is a strong tendency to accept the ideas and values of your ingroup, and to be shaped by hidden background variables which you are not aware of it (whether it be innate dispositions or particular experiences or socialization). This isn't perfect, people who were raised one way do and can shift in other directions. But many times even this evolution isn't described correctly by the person doing the evolving. One example. In Daniel Schacter's work he reports on a study program which tracked Southern whites in their attitudes toward race relations from the late 1960s to the early 1980s. Participants correctly reported the social consensus of whites in terms of attitudes toward issues such as segregation and black civil rights at each time period. But, many of the whites who expressed pro-segregation viewpoints in the late 1960s reported that actually they were in the minority who were pro-civil rights at that time when asked in the early 1980s. In other words, they simply reedited their position in their memories. If you asked any given person why they were a Democrat or a Republican, they would probably give you a list of positions. But the reality is that the #1 reason people are of a given party is their family is of a given party. This is a case where most people will ascribe their own stance to situation, values, reflection, but generally accept as an empirical fact that for much of the population family tradition holds greater weight than anything else. 4) We tend to socialize with people like ourselves, so we get a false sense of the texture of human variation in attitudes. For example, the detail and sharpness with which one perceives political difference drops off in proportion to distance from one's own position. That is, liberals may see the incredible differences on the Left between its factions, but clump all conservatives as an amorphous whole because the latter are simply abstractions with whom they have minimal experience. And the converse. This is probably largely a function of assortative socialization within political groups, so that the differences liberals see are between genuine Leftists and centrist liberals, or labor liberals vs. environmentalists. When Tom Ridge was nominated to head the Office Homeland Security I recall Alan Derschowitz complain that he was a "harsh right-winger." If Derschowitz was a Republican he would have known that Ridge is a pro-choice moderate who caught flak in the 1980s for opposing some of Ronald Reagan's defense programs. But as it is Derschowitz simply assumed that since Ridge was a Republican with the George W. Bush seal of approval he was a "harsh right-winger." That's fine as it goes if you limit yourself to your own ideological kind, but these rules-of-thumb are probably going to be a major issue if you want to have a discussion across ideologies, as no one likes to be caricatured. There's obviously a lot more here. And the points I'm making don't just apply to politics. They probably apply more to religion. Many atheists and theists are convinced they know why the other side believes what it does. I was like that at one point as well. But examining the empirical distribution, I think the reality is that most people misunderstand the motivations of others, and, importantly most people probably misunderstand their own motivations! When it comes to intellectual topics that deal with humanity we pretend as if it's like physics, with billiard balls following predictable paths. The reality is very different, human affairs, human motivation, and so forth, are much more complex. Not only is it important to move beyond our own projections derived from our own mental models, but we have to acknowledge that individuals themselves have a difficult time in teasing apart the variables which lead them to conclude what they do. We have such a tenuous grasp on our own motivations when it comes to religion and politics that it is folly to presume that we'd be any good at deconstructing others.