Sunday, November 30, 2008

Helping students with analogical reasoning

Just curious to know if anyone out there has tips for helping students develop the ability to appraise arguments by analogy. Analogies are of course all over in philosophy, especially in ethics, but many of my students struggle to understand how analogies are supposed to support their conclusions and how one might go about criticizing a parcel of analogical reasoning.

Here's one problem in particular that students seem to have. Analogical reasoning procceds by arguing that, because A has feature(s) F and G, and B has feature(s)F, then B has feature G. That's a great simplification of course, but it captures the gist of such reasoning. Many students try to criticize analogical reasoning by pointing out some difference or other between A and B, but the difference they point out doesn't seem to undermine the analogy. (Thomson's abortion article comes to mind here; students are good at pointing out differences between the fetus-mother situation and the violinist-plug in situation, but many of the differences they cite don't do much to undermine what, according to Thomson, is analogous about these two situations). It's as if students fail to appreciate what an analogy is, since they think that one can show an analogy to be weak simply by pointing to any difference at all between the two items. But of course, if there weren't some differences between the analogized items, they would stand in a relation of identity and we wouldn't need to invoke an analogy!

But that's just symptomatic of a larger struggle my students seem to have with figuring out what's going on in analogical reasoning and how to critically engage such reasoning. Any thoughts or tips?


  1. On occasion, I ask them to make their own arguments from analogy -- which gives them some charity in their interpretations of others...

    Otherwise, I do what you seem to do -- in that I ask them why they think their objection to the analogy takes out the argument?

  2. I don't think your students are actually doing very badly; it's common enough, I think, to find professional philosophers doing the same thing. (At least, I run into it quite often in articles on Hume's Dialogues Concerning Natural Religion.)

    I'm inclined to think that part of the problem is that there is no agreed-upon account of analogical reasoning in the first place. If we take Hume's account of analogy (in the Treatise), for instance, it is impossible to undermine analogies (Hume actually allows one exception, but it rarely seems to come up): for (almost) any two things you could reason about, it is trivially true that they are analogous in some way. And, of course, where we have no independent information about the conclusion it is impossible to prove the conclusion false. So all you can do is to show the weakness of the inference -- and this is to be done not by showing differences between the analogues but only by running counter-inferences, i.e., inferences based on analogy that yield a contrary conclusion. (It could still be useful to find differences, of course, because these might be the beginnings of counter-inferences.) The ultimate conclusion drawn is based on taking into account all the inferences and all their counter-inferences. I find this a very plausible account, but when I try to convey it to colleagues, I find that they often split between people who think it almost trivially true, on the one side, and, on the other, people who think it obviously false and possibly incoherent. So perhaps the real trouble here is that we struggle with evaluating analogical inferences ourselves? Is there really a hard-and-fast way of refuting an analogical inference, short of simply demonstrating independently that the conclusion is false?

  3. This page from Richard Feldman on arguments from analogies is quite good, IMHO:

    He has an, I guess, different take on what the point of an analogy is and so their role in arguments. Since his notes are easy to read, I won't summarize them.

  4. I've had the same experience when teaching the Thomson article; the same thing also shows up when I discuss Singer's shallow pond case. What I generally try to do in those cases is to change the case a bit to determine whether the difference the students cite makes a moral difference. This strategy is employed very well in Unger's book "Living High and Letting Die".

  5. As ItPF mentions in the first comment, charity is vital, and a more general problem.

    Charity can, I think, only be found where the goal is the search for truth. Absent the search for truth, debate devolves into nitpicking, a sales pitch, or even spamming the opponent with talking points.

    Foster the habit among students of reformulating an opposing argument in the strongest way, if only to make sure it's clearly defeated.

    A stop along the way might be argumentum ad logicam: just because an analogy fails to be persuasive doesn't mean that another analogy will not be.

  6. If you wish to assist you students in the process of analogical reasoning, then you are going to have to know their vocabulary. To do that you must have a grasp of their respective value systems. Once you know what is important to them, then you will be able to provide them with examples of analogical reasoning and they will be able to successfully employ their new found skill.

  7. There is a useful explanation of this type of argument in an anthology by David Boonin and Graham Oddie, "What's Wrong?: Applied Ethicists and their Critics". My own brief summary of this:

    Applied Ethics and the Use of Analogies

    How are analogies used in moral philosophy?
    Philosophers attempt to justify accepting a moral assessment about a practice by comparing it with another practice about which most people already agree.

    Why do philosophers use these types of arguments?
    Philosophers use these in order to resolve a controversial issue by framing it in terms of a less controversial one.

    Critically Evaluating Such Arguments

    1. Challenge the author’s assessment of the “agreed upon” example.
    2. Identify a disanalogy between the example and the issue it is applied to.
    -identify a difference and show that it is morally relevant.

    A Final Point
    The use of such analogies may seem irrelevant, because they are often unrealistic, but the point of doing this is to keep the morally relevant issues present, while avoiding the influence of bias, emotion, and the like.


If you wish to use your name and don't have a blogger profile, please mark Name/URL in the list below. You can of course opt for Anonymous, but please keep in mind that multiple anonymous comments on a post are difficult to follow. Thanks!