Friday, October 25, 2013

Shouldn’t I Agree with David Chalmers?

by Matt DeStefano

David Chalmers and I disagree about issues in philosophy of mind. Given that I believe David Chalmers is an expert, and I am not, shouldn’t I revise my belief to agree with Chalmers? After all, he probably knows better than I do.

A lot of attention in philosophy has been paid to peer disagreement, and how one ought to revise her view in light of knowing that an epistemic peer disagrees with her (an epistemic peer being someone with the same evidence and same reasoning abilities as you). There is an additional question about how we should respond when experts disagree with us.

There are some conditions under which disagreement might serve as a defeater for a position you hold. For instance, imagine that Katie and Andy are watching a horse race. Katie thinks that the horse named Ain’t Misbehavin won, while Andy thinks that Tango Goin Cash has won. (I found these names by Googling “horse names”, it was very entertaining.) Katie knows that Andy's vision is about as good as hers, was watching the race just as closely, and is her “peer” in the relevant sense. They are both equally likely to have made a mistake in declaring a winner. In this case, we might think that Andy’s belief that Tango Goin Cash won serves as a defeater for Katie’s belief that Ain’t Misbehavin won - and vice versa. At the very least, we might encourage Katie and Andy to suspend belief about what horse won (perhaps until the results are revealed by an official).

What about expert opinion? Suppose Katie is watching the race with Susan, an expert analyst, who has a proven track record of correctly guessing winners as they cross the finishing line. Susan has excellent vision, and has been watching horse races for her entire life. If Susan were to say that Tango Goin Cash has won, we might think that Katie ought to revise her view to agree with Susan, since she is an expert. At least, it seems she should be more inclined to revise her belief in this case than in her disagreement with Andy.

There are other cases where it seems we are justified in believing something based on expert opinion alone. For example, I am feeling sick and decide to go to four doctors to get different opinions about my ailment. Three of them tell me I have the flu, while one says I have allergies. Most people would say that I am justified in believing that I have the flu, based on the majority opinion of the medical professionals.

Consider philosophers who are not epistemic peers, such as David Chalmers and myself. He has been studying issues in philosophy of mind for far longer than I have, he is more knowledgeable about the relevant arguments, and more intelligent than I am. When I find out that David Chalmers disagrees with me about some important issue in the philosophy of mind, should I revise my belief according to his expertise? Intuitively, this seems like a rational course of action. However, there are plenty of experts who disagree with Chalmers. In fact, Chalmers’ positions are in some ways minority positions among relevant experts. Considering this, we might think that the proper way to decide what we are justified in believing in is to count the number of philosophers who hold each position, and whichever position the majority of philosophers believe is the one we are justified in holding. This is a truly ridiculous way to determine our beliefs, and we need a better way to understand the influence of expert opinion on our beliefs.

Philosophers have taken a number of different positions about how we ought to treat peer disagreement. Some have argued that one ought to “split the difference” between the two views, and either suspend judgment or come to a middle position. Others have argued that one should “stick to their guns”, and believe what they believe despite disagreement. These views do not easily capture how one ought to respond to expert disagreement.

One view, the Total Evidence View, presents a framework for handling disagreement as a form of evidence. Thomas Kelly has argued that peer disagreement should count as higher-order evidence, and we should revise our beliefs based on our total evidence. Roughly, this position argues that “Rather, what it is reasonable to believe depends on both the original, first-order evidence as well as on the higher-order evidence that is afforded by the fact that one’s peers believe as they do.” (p. 32 of linked paper)

We can extend this “Total Evidence View” to our own disagreement with experts. This is useful in understanding why I might be permitted to believe (and even be justified in believing) my diagnosis on the basis of a majority medical opinion, but not permitted to believe (and certainly not justified in believing) what the majority of philosophers believe. My first-order evidence in the case of my illness are my symptoms, and a very limited idea of what these symptoms might entail. My higher order evidence is that these doctors are much, much better at diagnosing illnesses (this might also be first-order evidence), and that a majority of them seem to agree that I have the flu.

On the other hand, I have much more evidence for my positions in philosophy of mind. While I do not have the same breadth of evidence, or ability to discern what position it supports as Chalmers might have, I have enough to form a reasonably justified opinion about these matters. In this case, expert opinion should not weigh nearly as heavily in determining my beliefs as it did in the diagnostic case. If we consider disagreement as higher-order evidence, we can much better decide how to handle disagreement among both peers and experts. It also eases my personal discomfort about disagreeing with Chalmers.

Matt DeStefano
Sacramento State Philosophy Alumn
Philosophy Graduate Student
University of Missouri, St. Louis


  1. Matt, you are a chicken sexer with years of experience, and a darned good one. Let's say you are typically 95% reliable in your ability to identify the sex of a baby chick (which, as I'm sure you know, untrained people can do no better than chance.)

    You are at the Annual World Chicken Sexers Convention and at one of the convention smokers (before the drinking starts) you can not resist entering a chicken sex-off. About 20 of you pass around a baby chick. You look at it, confidently assess its sex, and all the others, who are just as skilled as you, do the same.

    You sexed the chick as a hen. But before the betting begins, you learn that only 30% have sexed it has a hen, whereas 70% have sexed it as a rooster. Should you bet on hen or rooster?

    I think it is obvious that you should bet on rooster. Put differently, if I, who am not a chicken sexer was allowed to bet, I'd be sure to go with the majority every darn time. I would be a fool to do otherwise, and you would be a prideful fool not to as well.

    I think you'll agree with what I'm saying, and I'm sure you'll agree it applies equally well to any situation in which acknowledged experts with reasonably high rates of reliability disagree.

    My point here, of course, is that while you are an expert, and I am not, this should not in any way cause us to bet differently when there is disagreement between experts.

    So it seems to me that there is something that you have left out concerning why it is such a bad idea to deal with philosophical disagreements in the same way. I have a view about what you've left out, but I'd rather hear what you have to say.

  2. Thanks for the comment, and you're right to suggest that I've left something out of the picture. You've already hinted at it, but I think one of the biggest differences is the reliability of the experts, and the regularity of the domain which they are making claims or predictions.

    In the case of chicken sexing, many of these experts have reliable histories of making accurate predictions. The same holds for going to the doctor, they likely have histories of being fairly reliable in making accurate diagnoses.

    Philosophers do not have the same type of track records. Arguably, this is because the domain in which they are making claims and/or predictions is not as reliable as the others. Kahneman does a good job of explaining this type of "highly regular environment"* when he's talking about expert intuitions. If we are going to trust the experts, the domain should be such that it is highly regular, and they have had the ability to practice for an extended period of time and check these results against the actual results.

    Philosophy is not a highly regular environment, and doesn't provide us with much data about the reliability of its experts. I'm still inclined to think that there can be philosophical experts, but I don't think they are at all comparable to chicken-sexing experts and others within highly regular domains.

    *I don't have my copy handy, but I think this is the term he uses.

    1. Matt, I am very happy with the depth of your comments on this controversial issue. I wonder what you'd say about a related issue. though. A friend of mine once said to me that she was trying to decide whether to believe in life after death. Since the experts are religious leaders and they almost unanimously believe in life after death, she said, the right thing for her to do is to believe in life after death. I replied, "Aren't some of the philosophers also experts on this issue?" She replied, "Yes, but there are so few of them compared to the religious leaders, and the philosophers don't agree with each other." What do you think about how she should decide whether to believe in life after death (continued personal consciousness after bodily death)?

  3. Matt, that strikes me as a solid answer. But I think it requires you to clarify your claim that Chalmers is an expert and it also requires you to retract (or clarify) your suggestion that you are justified in believing your own contra Chalmers view (since you have just stipulated that it is unlikely to be correct.)

    I am also interested in how you would answer Brad's question.

  4. Thank you for the comments, Prof. Mayes and Prof. Dowden.

    @Prof. Dowden: This is an interesting case, and one that I've certainly encountered some variant of in the past - I'd be interested to hear other opinions.

    As I see it, there are two main questions to ask when determining expertise (1) Is this a domain of knowledge that permits expertise? and (2) What qualifies an expert in this field? I'm inclined to think that the question of whether there is life after death fails at (1). We simply can't test expert predictions against any available evidence. (This might itself be taken as evidence for a position, but that is first-order evidence rather than higher-order evidence.) If a field doesn't permit expertise, then we ought to rely on the available evidence/arguments for and against a position.

    @Prof. Mayes: That's right. I should clarify that I think expertise comes in degrees. A chicken sexer who is accurate 98% of the time is a *very* reliable expert. An expert sports gambler who performs above chance might still be an expert, but is not at all comparable to the chicken sexer.

    The expertise in philosophy is somewhere in the range of the sports gambler. Chalmers might be an expert, but his reliability isn't close to that of a chicken sexer. When it comes to philosophy, higher-order evidence about expert opinion will weigh considerably less than it will when it comes to chicken-sexing. In addition to weighing much less, there is much less agreement in philosophy - it would be nice to have 75% of philosophers agree on anything. So, we've got experts that aren't terribly reliable, and they don't agree on much. In a situation like this, my first-order evidence is going to be what I primarily rely on.

    1. Matt,
      Regarding your response to my question about the woman who said religious leaders are experts on the question of life after death, I would have given the same response as you.

  5. Matt, I agree with with what you said to me, thanks. I'm not sure I agree with what you've said to Brad. It seems to me that if an area doesn't permit expertise, then we are largely justified in ignoring the advice of experts. But it doesn't follow from that that we ought to accept any conclusions we draw on our own. We should just admit that knowledge in this area is not yet attainable, so we are guessing like everyone else.

    I don't know if I think the untestability of a life after death hypothesis is its main problem. It's main problem is that it is incoherent. We have strong reasons for thinking that it is physically impossible and we have no understanding of what this means otherwise, since our only acquaintance with life is physical.