Monday, January 21, 2013

"To save man from the morass of propaganda, in my opinion, is one of the chief aims of education."

Martin Luther King day is providing an excellent opportunity to reflect on some of his writings. The Washington Post has provided an interesting list of quotations. In particular:

Education must also train one for quick, resolute and effective thinking. To think incisively and to think for one’s self is very difficult. We are prone to let our mental life become invaded by legions of half truths, prejudices, and propaganda.  At this point, I often wonder whether or not education is fulfilling its purpose. A great majority of the so-called educated people do not think logically and scientifically. Even the press, the classroom, the platform, and the pulpit in many instances do not give us objective and unbiased truths. To save man from the morass of propaganda, in my opinion, is one of the chief aims of education. Education must enable one to sift and weigh evidence, to discern the true from the false, the real from the unreal, and the facts from the fiction.
I'm afraid that I strongly agree with Dr. King that the education system continues to fail in its purpose, and that both educated people, as well as those not so educated, often seem to fail at thinking both logically and scientifically. I don't really know enough about education policy to address solutions, though it seems that focusing mathematics classes more on statistics and less on other topics like geometry as well as addressing topics such as cognitive biases at the high school level would help, but I do think that addressing the difference between political and empirical questions are worth addressing.

In many of the major topics of ideological division in America, such as taxes, climate change, guns, and the welfare state, we see political considerations and empirical questions conflated. If we are going to have a meaningful conversation on these topics it is necessary to distinguish between the two.

Since I've been writing about it, take gun control for instance (I thought about using global warming, but haven't been writing about it lately). In most discussions of the topic we see the political topic of should we impose controls on firearms conflated with the empirical questions of whether firearms contribute to crime.

This creates opportunities for the propaganda that King references to have an undue influence on the debate and interferes with our efforts to think about the topic objectively. The political question of what we should do about guns is entirely distinct from the empirical question of what effects guns have (guns are hardly unique in this).

Yet, most discussions on this topic make it sound like these are the same question with individuals becoming polarized on both sides around both normative, subjective political questions and objective, empirical questions. This is sloppy thinking, answering the empirical questions regarding the effects of guns does not tell us the answer to the normative question of what we should do about guns, it only allows us to have a grounded conversation about our normative, subjective beliefs regarding how we should react to the data unearthed by our empirical investigation. Without first grounding our beliefs empirically, however, we cannot reach the point of discussing our normative beliefs, we instead get bogged down in pointless discussions of speculative effects rather than discussing how to reconcile our conflicting normative beliefs about what to do in response to these known effects.*

* Not that effects can often be known with certainty. But the weight of evidence of competing claims is an essential element that needs to be added to our normative political discussions. It is simply not true that there is equal evidence to the video games cause homicide and guns cause homicide hypothesis, yet in most common media sources these hypotheses are presented equally simply because someone can be quoted as presenting them. This is a flaw that is in desperate need of correction and better scientific, and especially social scientific, literacy would help to make these discrepancies in evidence clear.


  1. Ugh, the problem with getting people to become social scientifically literate is that too many of them dismiss the field as being pseudo-science with little merit, especially those in the "hard" sciences, and I have no idea how to go about fixing this. I have to confess that when I was younger I felt this way myself until I grew to realize that I (like most of the others in my position) had formed my opinion based on hearsay that social science was this way rather than strong knowledge about what social scientists actually do.

  2. That's a difficult challenge to solve, especially since there is a degree of truth to it. Social science can't give us the kind of concrete results that say, physics, can in allowing us to go from theory to building a nuclear reactor.

    What would help is learning about different kinds of science. Put on a continuum from experimental to observational sciences, physics would be at one end, followed by biology as a hybrid, then astronomy, medicine, and only then the social sciences where experimentation plays a small role (and I've met some hard science types that tend to be sceptical that some sub-fields of biology are a real science). To understand the place of the social sciences it's important to understand the diversity of the applications of scientific reasoning and how it can be differently applied, something that I think science classes could do a better job at both at the high school and university level (I was a biology undergrad, so I have at least anecdotal experience of this).

    I'd also add that social science has a few problems in that well known biases distort people's views of it, it gets mixed up with politics, many groups have interests impacted by its finding, and it often doesn't really answer the questions people want answers to.

    To address that last point, a big problem is that people think in narratives and they want to be treated as individuals. This makes science impossible. Imagine if a chemist was expected to be able to predict which particular oxygen molecule would interact with which two specific hydrogen molecules to form water. I think most scientists would laugh this off as absurd.

    Yet, with social science, you see nonsense like asking whether or not specific gun control measures would have prevented the Sandy Hook shooting. This is as impossible to answer as it would be for a chemist to determine which particular CFC molecule reacted with a particular ozone molecule; it's just not possible. But, like a chemist saying that CFCs react to deplete the ozone, it is possible to say that guns appear to raise the probability of homicide by acting as a catalyst (social scientists should consider using terms like this more often, rather than getting hung up so much on establishing strict causal relationships). Social science can help us to understand rates of interactions and give us some sense of probabilities, but since people like their stories and like thinking about people as individuals they are often uncomfortable with this; they want information on which individual stock to pick or which individual to be wary of. They aren't comfortable with a social scientist being like a chemist who can predict which catalysts make a reaction happen more quickly and result in a better final product and which contaminants make undesirable reactions more likely and result in more waste products but who can't predict which particular individual molecules will react to form final product and which will react in an undesirable fashion to result in industrial sludge that needs to be cleaned out in the end.

    But as scientists, social scientists work with observations of aggregates and can't work to predict the outcomes of individual interactions in a complex system, except in very narrowly defined circumstances, such as Nate Silver and the presidential election (though there are good arguments that he misstated the actual probabilities to make the race look closer than it was and that other academic models were better). As a way forward, more high profile social science founded models like Silver's might be a way to make people pay a little more attention to social science, but we've got a long way to go. Also, I think that more attention to other branches of social science aside from economics will help; economics has some problems and the focus on it is matching a fair amount of convergence among other social sciences with economics being a bit of an outlier in its resistance to using the findings of other fields to inform its research.

  3. One difficulty in separating the two types of questions is that sometimes they have been legislated together. For example, the CDC is prevented, by law, from even considering any issue involving guns. Apparently the Congress felt that having hard data on the subject would be detrimental to their consideration of the questions that they wanted to deal with.

    I suspect that this isn't the only issue where those on one (ideological) side of an issue are determined that they don't want empiracal evidence intruding into the discussion. Which suggests that they, at least, know that reality is biased against them.