New APPS readers probably remember Helen De Cruz's excellent post on the polarized debate surrounding evolutionary science (which was picked up by NPR), as well as Roberta Millstein's follow-up post on the perhaps equally polarized debate concerning climate change. Both posts cite the work of Dan Kahan, who has a distinct take on these issues:

"I study risk perception and science communication. I’m going to tell you what I regard as the single most consequential insight you can learn from empirical research in these fields if your goal is to promote constructive public engagement with climate science in American society. It's this: What people “believe” about global warming doesn’t reflect what they know; it expresses who they are."

I just attended a talk by Michael Ranney, who opposes Kahan's position. In Ranney's view, communicating the mechanism of global climate change is enough to change the minds of people on both sides of the political spectrum. (Check out the videos!) Ranney shows, surprisingly, that just about no one understands the mechanism of climate change (Study 1). Further, he shows that revealing that mechanism changes participants' minds about climate change (Study 2). 

Why might this be?


Ranney claims that mechanistic explanations are more convincing than other explanations, citing the work of Frank Keil. This alone is an interesting possibility. But his explanatory statements and videos (see also this page) do more than just offer mechanistic explanations. First, they provide short explanations. As Rottman and Keil put it, "We are frequently confronted with complex scientific explanations of phenomena such as global warming, diabetes, and nuclear reactors. Given the immense potential breadth and depth of such explanations, we need ways to avoid being overwhelmed by the information." A short, salient explanation makes it possible for participants to avoid being overwhelmed by details. Second, they use a rhetorical strategy as old as Socrates–put your interlocutor on their back foot. In every video but the shortest, the videos start by asking the viewer to come up with their own explanation of global climate change or global warming (see, e.g., the 2-minute version), then the voiceover proceeds to explain that almost no one understands global warming, finally providing the explanation required. One member of the audience wondered whether this rhetorical strategy would yield the changes found in this study even without the mechanistic content. So here we have three possible explanations, which may well be working together in this case: our beliefs are more easily changed when confronted with 1) short, 2) mechanistic explanations of the phenomena that 3) begin with the rhetorical strategy of showing us that we don't know something we think we already understand. I suspect that philosophers might have something to say about any one of these three possibilities. But I want to say a little more about the first explanation–that shorter, more salient explanations are more likely to change one's beliefs. 

Why do shorter, more salient explanations move us? Perhaps because we regularly confuse cognitive fluency (e.g. familiarity) for truth. I am reminded of this piece in the Boston Globe: "The persuasive power of repetition, clarity, and simplicity is something that people who set out to win others’ trust – marketers, political candidates, speechwriters, suitors, and teachers – already have an intuitive sense of if they’re good at what they do. What the fluency research is showing is just how profound the effect can be, and just how it works." This research might explain why the style of what is often called "analytic philosophy" is so compelling, for instance. And, yet, the article goes on to explain the problem with cognitive fluency–we accept its content too readily: "In other words, to get people to think carefully and to prevent them from making silly mistakes, make them work to process the question: make the font hard to read, the cadence awkward, and the wording unfamiliar." The fact that something is easier to read might make us more willing to accept it, but a deeper reading sometimes requires more superficially difficult writing. Perhaps this is a partial exoneration of the style of what is often called "continental philosophy"? At the very least, as we learn what helps us to change the minds of others, we also learn when to be cautious of being so changed. (But, again, check out those videos!)

Posted in , , ,

8 responses to “Mechanism, Salience, and Belief Change”

  1. Roberta L. Millstein Avatar

    Very interesting, Carolyn, thanks!
    My own anecdotal experience teaching the evolution/creationism debate supports this to some extent. I have had skeptical students say, “Really? That’s it?” when presented with the basics of Darwin’s theory of natural selection. In this case, it helps that the mechanistic explanation [setting aside philosophical debates about what a mechanism is] is so straightforward. I’m not saying that creationists become evolutionists after one quick explanation. But it can definitely get the wheels turning for some. Here, part of the problem simply seems to be an overly-complex, and probably mistaken, view of what natural selection really is.
    My second thought is that we have hammered home the short and simple “correlation is not causation.” So, if people are learning about the correlation between CO2 and global warming without learning the mechanism (or having mistaken ideas about the mechanism, e.g., thinking it has to do with the ozone hole), then I can see where it is harder to accept the reality of global climate change.
    Anyway, I guess what I am saying is that if it is true that short, mechanistic explanations move us, that doesn’t strike me as a terrible thing, and could turn out to be truth tracking in cases where the truth really is simple (which is not all cases, of course).

    Like

  2. Aaron Lercher Avatar
    Aaron Lercher

    Wonderful videos. Thank you, Michael Ranney.
    I still do not understand how climate change happens, so I can not with any honesty teach about it.
    In an effort to find an analogous process that was easier to understand and to teach I had considered teaching how atmospheric oxygenation occurred (as a example of a research project in an “information literacy” course – I’m a librarian). After some research, I gave up on that.
    Like Roberta, I never had trouble teaching something about evolutionary theory, which came up in different ways in philosophy when I taught that, and in “information literacy” when I taught that. It isn’t hard to explain to creationist students. It is possible to point to familiar consequences of the facts of evolutionary theory. For instance, species change isn’t just a theory Darwin hypothesized, it’s a freaking giant industry, producing the food we eat, etc.
    About these climate change videos, it is important, I think, to emphasize to students that the theory sketch in the videos organizes a huge amount of complex information. The videos tell us where to look for comfirmation of that this mechanism exists: the chemistry of light absorption and emission spectra by many different kinds of molecules both on earth and those in the atmosphere. The videos claim that these absorption and emission processes tend to add up in one way.
    That’s the simplification achieved by the videos. That’s a huge step. A layperson like me still cannot see how this immense equation works out. But I at least I have been told in the simplest terms what kind of thing is added up.
    Also the videos implicitly claim that the all the other mechanisms involving volcanoes, dust, clouds, water temperature, ocean currents, etc. are minor details compared to the mechanism described in the videos. That’s worth mentioning too.

    Like

  3. Helen De Cruz Avatar

    Thank you for letting me know about this, Carolyn. I will check out this work when some pressing deadlines have come by, so I haven’t checked it out for myself yet. One thing I’m wondering about is whether this research really invalidates Kahan’s work – it could still be the case that people strongly identify with positions and that it is hard to change their minds. So I’m a bit skeptical (a prima facie skepticism) that simple explanations of mechanism or a socratic-like method, or a combination of these factors could change the minds of anti-vaxxers or fundamentalist Christian or Muslim Young earth creationists. One study that examined the attitudes of anti-vaxxers found that no intervention (not even carefully explaining) could change their minds. I know climate change is also a polarized issue, but it is less polarized than evolutionary theory. It wouldn’t surprise me if opponents to evolutionary theory from the lay audience wouldn’t know how to describe the mechanism, but I doubt that it would change their minds. But one interesting finding is that the techniques used to not involve debating (which polarizes discussion). Perhaps by emphasizing mechanism, we go away from these views as positions that define people’s cultural world views. Work by Larisa Heiphetz and colleagues indicates that people see beliefs along several dimensions: fact, ideology, and preference. http://www.wjh.harvard.edu/~lds/pdfs/Heiphetz_JESP_2013.pdf – for instance, suppose I enthusiastically endorse vaccines and say they are effective, I’m saying something factual (I believe vaccines are effective and pose no health risk), preference (I’d like my own kids to enjoy this protection) and ideology (I find herd immunity etc important, and want to see preventable serious childhood illnesses disappear). Now, when we emphasize mechanism we might increase the salience of belief-as-fact and decrease the belief-as-preference and belief-as-ideology. I am just wondering if this works for things where ideological fault lines run very deep, such as evolution.

    Like

  4. Carolyn Dicey Jennings Avatar

    I realize now that one could read my post as specifically questioning the truth of the provided mechanism, which I did not intend. I think Ranney’s work is great.

    Like

  5. Carolyn Dicey Jennings Avatar

    Good points! And good luck.

    Like

  6. Carolyn Dicey Jennings Avatar

    In discussion, Ranney allowed that coming up with a short, mechanistic explanation for the benefits of vaccination would be difficult. On evolution, Ranney et al. say at least two things that I think are relevant (from http://mindmodeling.org/cogsci2012/papers/0388/paper0388.pdf): 1. “Acceptance beliefs largely follow understanding, given variation in both. Thus, while evolutionary acceptance might be uniformly high or low in a certain classroom or campus, evolutionary biologists largely accept evolution more than do “Intro to Bio” students (Shtulman & Calabi, 2012)” and 2. “Just as knowing “how reproduction works” supports evolutionary acceptance (cf. Shtulman & Calbi, 2012), our studies show that a mechanistic global warming understanding (e.g., a-c above) supports its acceptance.” Thus, just as it seems to me that Kahan’s statement is false in the case of global warming (“What people “believe” about global warming doesn’t reflect what they know; it expresses who they are.”), it also appears false in the case of evolutionary theory, if these references pan out. This isn’t to doubt the validity of Kahan’s suggestion that we should aim to separate knowledge from source in these cases (which the videos do nicely), nor the validity of your suggestion that we should push for change from within the Christian community. I found those suggestions interesting and helpful. But it looks to me that this shows that the evidence can move people–and even people who identify as conservative–if it is presented in the right way.

    Like

  7. Ed Kazarian Avatar

    Quick observation: some of this reminded me a lot of Hume’s emphasis on short chains of reasoning as more convincing–and indeed, on the very fact of long and ‘abstruse’ chains being inherently unconvincing. It’s also interesting from a Humean point of view that mechanistic (which is to say, causal) accounts seem to be the most powerful.

    Like

  8. Jay Odenbaugh Avatar

    This is an interesting post. I suppose I agree with Helen that Ranney’s claims are not inconsistent with Kahan’s work. As I understand Kahan et. al.’s analysis it goes like this. We can group Americans in terms of their views regarding authority and regulation; namely, we have hierarchical individualists and egalitarians. One hypothesis, the science competence hypothesis suggests that as one is more scientifically and quantitatively literate, one comes to believe anthropogenic climate change is occurring and that it is risky. Another hypothesis, the cultural cognition hypothesis, suggests that the increasing said literacy does not change one’s views regarding anthropogenic climate change and the associated risk. Rather, it concerns your social grouping. The data suggests that the best predictor of your views is not your literacy, but your “tribal” affiliation. In fact, increasing literacy has no effect on egalitarians and actually reduces acceptance of anthropogenic climate change and the attendant risk. Strictly speaking, the results have little to do with belief change. The analysis is roughly a static one other than the result that literacy has no effect. Of course, simple mechanistic explanations might change the results but we have little empirical reason to think so. All of this is really interesting and I myself am working on an essay regarding climate skepticism and moral psychology.

    Like

Leave a comment