Analogical Argument

An analogy is a likeness between things, and an analogical argument is one that exploits an analogy to infer something about one of the like items from premises about the other. For example, if you know that Socrates was sentenced to death, and that Bertrand Russell is like Socrates, you might infer either that Bertrand Russell was sentenced to death or (more plausibly) that something like being sentenced to death happened to Russell.

As stated, neither of the two arguments just given concerning Russell and Socrates is very good. Part of the difficulty is that it is not clear in what respect the premise is claiming that Russell is like Socrates, and this makes it difficult to know what features of Socrates it is reasonable to attribute to Russell. After all, the two were alike in some respects: they were both human beings, males, philosophers, and people with a reputation for holding controversial positions that people worried would have a negative influence on the morals of the young. On the other hand, they are different in many respects: one was Greek and the other British, one lived in the 5th Century B.C. and the other (mostly) in the 20th Century A.D., and their philosophical views were different on many points. Since they are like in some ways and not in others, we would expect some, but not all, of the things that are true of Socrates either to be true of Russell or to have close parallels in Russell’s case.

Now in the case of Socrates’ being sentenced to death, this was a result of his being thought to have controversial views that might corrupt the youth. Since this was also true of Russell, we might expect something somewhat similar to have happened to him. In fact, he was never sentenced to death, but he did suffer professionally in various ways because of parents who were concerned that he might corrupt their children. He was fired from a position at the City College of New York after the parents of some students at the college expressed outrage over his liberal views on sex and marriage. Is this like what happened to Socrates? It is in some ways, but not in others. What happened to the two philosophers is similar in that both suffered negative consequences which were due to concern over worries that they were immoral and would corrupt the morals of young people. But they are different in the nature and severity of the consequences they faced, an in who took action against them. Socrates was put to death and Russell merely dismissed from a job, and the action against Socrates was taken by the government of his city, whereas the action against Russell was taken by his employer.

When we know of a similarity between two items, it is reasonable to expect to find other, related similarities, but it can be difficult to specify the respects in which the items are similar, and to know what sorts of further similarities to expect to find, or how strongly to expect them. All other things being equal, the more one specifies the respect in which the like items are similar, what other similarities one should expect to find between the items, and why one should expect to find them, the stronger the argument from analogy.

Notice, however, that if one fully specifies these things, what one is doing is first inducing from the case of the one item (or set of items) to a universal conclusion, and then using this conclusion as a premise in a deduction about the remaining item. To return to our previous example, we might identify the likeness between Socrates and Russell as that they were both well-known figures with controversial views on moral subjects who were in a position to influence the young. And we might conclude that Socrates’ death sentence was a case of his being treated unjustly at the hands of concerned parents with more traditional moral views. Further, we might conclude from Socrates’ case (perhaps with the addition of a few others) that people with radical moral views in a position to influence young people are treated unjustly by concerned parents with traditional values. This would be an induction. And then we might deduce from this general conclusion and the premise that Russell was a figure with radical moral views in a position to influence the young that he too was treated unjustly by concerned parents with traditional values. In this case, rather than an argument from analogy, we have an induction followed by a deduction. But we arrived at this sequence of arguments by specifying as fully as we could something that began as an analogical argument. So, we might think of analogical arguments as crude forms of induction and deduction, in which one never states explicitly the universal proposition that is the conclusion of the induction and a premise in the deduction. Or, by the same token, we might think of induction and deduction as a refined form of analogical argument.

Analogical reasoning is especially useful in contexts in which one does not know of anything that could have caused an effect that one is trying to explain. Thus, it is often used as part of inference to the best explanation. In the article on inference to the best explanation, I mentioned that someone might infer the existence of a previously unknown type of animal on the basis of knowledge of its tracks. Consider how such an argument might work. We would know that the tracks could not have been made by any of the animals that we know of, but we would try to specify the ways in which the tracks were like those made by various known animals, and we would infer that the animal that made these tracks was like those animals in ways that would lead to the similarities in the tracks.