Very few people understand how to evaluate analogies properly. An analogy will have explanatory value if the things or experiences or events or ideas likened to one another are indeed alike in the respect called attention to by the analogy. Far too many people think they can deny the validity of an analogy between X and Y by pointing out ways in which X and Y are different. Yes, and if they were not different you couldn’t analogize them because they would be the same thing. In Thomistic terms, you do not discredit an exercise in analogical predication by gleefully announcing that the predication is not univocal.

Here’s the proper way to evaluate an analogy:

1) Ask this question: Does the person making the analogy between X and Y explain the respect in which he or she claims that X and Y are similar?

2) If not, ask the person to clarify that point.

3) If so, think about whether X and Y are indeed similar in the respect specified. If so, the analogy is legitimate. If not, the analogy fails.

4) Feel free at this point to pursue other questions about the analogy, e.g., whether even if legitimate it identifies an important similarity, or whether the analogy does the intellectual work its maker thinks it does.

Thank you for your time. We will now return to our regularly scheduled programming.


  1. Good one Alan. But I think you've neglected the flip side. Yes, helpful analogies are often discredited unreasonably. But the person issuing the analogy also bears a large burden of responsibility. In casual discussions at least, I think the big problem is a lazy analogy that hasn't been well thought out.

    The worst cases are analogies almost designed to inflame instead of illuminate. I'd put comparisons with slavery or the holocaust in this category. And I've definitely heard them before.

    I think the ultimate judge of an analogy is, as you say, it's "explanatory value." It's outcome dependent. If it works, then it works. If it doesn't, then both the issuer and receiver of the analogy should both try figure out why.


  2. Praj, I tried to point to your concern in my number 4. You're right that many, many analogies are badly chosen — and I'm known among my friends for being a comically fierce critic of them — but my point in this post was to emphasize the need to evaluate them properly.

    Freddie, I deliberately avoided examples here because I didn't want to get into the merits of any particular analogy but rather lay down (what I hope is) a useful set of general principles.

Comments are closed.