On Arguments by Analogy

by Scott Aikin and Robert Talisse

ChainsawArgument by analogy is like a powerful chainsaw. If you know how to use it, you can do some nice work. But if you aren't careful, you can make a big mess, and maybe hurt yourself as well.

The core form of argument by analogy is to infer from two things' acknowledged similarities that those things have further, as yet unacknowledged, similarities. One begins with a generally familiar phenomenon (the proximal object of the analogy), and then attributes salient features of the proximal object to another, less familiar matter (the distal object of the analogy). One the basis of the analogy, it is established that what is true of the proximal object is also true of the distal object.

Argument by analogy is a particularly widely used tool throughout Philosophy. Plato's analysis in The Republic of the good man runs on an analogy between justice in the soul and justice in the city. Plato's argument is that, as justice in the city consists of a certain kind of hierarchical order, the just man's soul manifests the same structure. Judith Jarvis Thomson's famous violinist case analogizes unplanned pregnancies to being kidnapped, taken to a hospital, and hooked up to a famous violinist who needs use of your organs in order to stay alive. And Paley's argument from design starts with the hypothesis that the world's functions (and those of many bodies within the world) are like those of a watch, whose very existence suffices to demonstrate the existence of a watchmaker.

It is not difficult to see how arguments by analogy may be challenged. These arguments depend on there being good reason for accepting the proposed analogy in the first place. Accordingly, many of those who reject the conclusions of Plato's, Thomson's, and Paley's arguments contend that the analogies themselves are at least as controversial as the conclusions they are supposed to support. Consider Paley's watchmaker argument. The premise that the world is analogous to an artefact goes a long way towards reaching the conclusion. If you thought the conclusion unacceptable, then the analogy just won't look right.

Hence arguments by analogy face a particular dialectical obstacle. Only those already roughly in agreement with the desired conclusion will readily grant the initial analogy. And so the analogies are either unnecessary or they are proxies for arguments that establish the appropriateness of the analogies. Arguments by analogy, then, are either otiose or insufficient.

Arguments by analogy commonly deployed in debates concerning same-sex marriage are exemplary of the problem. Analogizing laws against same-sex marriage to anti-miscegenation laws telegraphs the conclusion, as does the analogy between same-sex marriage and bestiality. In both cases, when one announces the analogy one is less arguing for a conclusion than situating one's conclusion within a broader worldview. The consequence, as we see it, is that arguments by analogy are better understood as heuristic devices that take argumentative form. That is, arguments by analogy do not provide reasons for adopting a conclusion. Rather, they reveal or make manifest the way the person wielding the analogy represents an issue. By attending to an interlocutor's argument by analogy, we do not expose ourselves to new reasons concerning the matter in dispute, we rather come to see how the interlocutor understands what's at issue.

Arguments by analogy confront a second dialectical complication, one that arises from the background presumption that beliefs about the proximal phenomenon are more settled or obvious than those about the distal. Consider Lucretius' famous parity of reasons argument as to why one should not fear death. He reasons that if two things are alike, we should have the same attitude about them. Since being dead is just like the time before you were born, and you don't feel bad about the time prior to your birth, you shouldn't have bad feelings about the time after your death. It's a nice argument. But what if someone responded by saying: Yeah, but my more stable reaction is feeling really bad about just not being, so instead of changing my attitudes about the time after my death, I'll change my attitudes about the time before my birth. No I'll feel bad about that, too!

Call this the backfire effect, and it is a challenge for anyone who runs a parity of reasons argument. Arguments by analogy are especially vulnerable to backfire. They depend on our attitudes about the proximal phenomenon being stable, and this stability is what allows one to extend them to the distal phenomenon. But if those proximal attitudes are not stable, the argument backfires. And note, this backfire effect has begun to bear fruit in ways in the political realm. Analogizing President Trump's behavior to that of Nixon depends on stable anti-Nixon sentiments. But if for a particular interlocutor those sentiments aren't that stable (and the more robust attitudes are pro-Trump) the analogy backfires; the interlocutor is encouraged to adopt more sympathetic thoughts about Nixon.

So, despite their pervasiveness in dialectical exchange, arguments by analogy rarely cut any argumentative ice. The reasons why one regards the initial analogy as salient are often cut from the same cloth that supplies the reasons why one embraces the conclusion. From the perspective of one's interlocutor, arguments by analogy simply beg the question. Although they may be useful ways to signal or express one's general viewpoint, arguments by analogy most frequently offer one's interlocutor no reason to change her position. In fact, in many of the contexts where arguments by analogy are most pervasive, they are also most prone to occasion backfire. In these cases, like with badly used chainsaws those wielding the argument saw off the limb they're standing on.