An analogy is an argument where two things are compared and a conclusion is drawn that what is true of one is true of the other. If an analogy were to be spelled out completely (which they rarely are), it would take a form such as this:
* Premise 1: A and B are similar in certain relevant respects.
* Premise 2: A has some further feature x.
* Conclusion: B (probably) has x as well.
Here’s an example:
“When Otis found a dead mouse in his Big Mac, McDonald’s paid him $5,000 not to sue. I’ll bet I can get $5,000 from Burger King for these dead roaches I just found in my Whopper!”
The fallacy of questionable analogy refers, not surprisingly, to an analogy that fails to justify its conclusion.
Assessing the strength of an analogy is a matter of determining how similar the cases are in the relevant respects. Key word here being “relevant.” The cases don’t have to be similar in all respects, nor even in all “important” respects. But they do have to be similar in the respects that are relevant to the specific claim being made.
Going back to our example, how much support does the point about the dead mouse provide for the conclusion that the speaker will be able to get $5,000 from Burger King?
Well, the kind of questions you’d need to ask would be how relevantly similar Otis and the speaker are, McDonald’s and Burger King are, a dead mouse and dead roaches are, etc., plus how relevantly similar other unmentioned details of the cases are.
If, for instance, the speaker has been convicted of fraud in the past for faking similar incidents and suing, whereas Otis had no such history, that would seriously weaken the analogy because one would expect a company to be a lot more reluctant to pay up in the former case.
But, again, the differences have to be relevant. It may be that Otis is a midget and the speaker is not, or Otis had an unhappy childhood and the speaker did not, or Otis plays the violin in a symphony orchestra and the speaker does not, or Otis gives thousands of dollars a year to charity and the speaker does not, but even though in their way these are all quite important differences between these two individuals, they (probably) have no relevance to the specific issue at hand.
McDonald’s and Burger King certainly are similar in a lot of ways, but if it turns out that one of them has a corporate policy of quickly paying off claims like this, and the other has a corporate policy of aggressively defending against such claims (and a history of winning when they do), then that could be a relevant difference between the cases.
Or, as I say, there may be other unmentioned aspects of the incidents themselves that are relevantly dissimilar. Let’s say Otis, for instance, discovered the dead mouse when he opened up his Big Mac at the counter in full view of a dozen witnesses. Whereas the speaker got his Whopper to go, took it home, and then returned to the restaurant an hour later waving it around angrily and showing people the dead roaches in it. That would mean Otis got $5,000 for a much stronger, more easily provable, case, and would render it less likely the speaker could get the same amount for his rather suspect case.
Analogies are ubiquitous:
“Look at Jones and his stats compared to Smith. Smith is in the Hall of Fame, so Jones deserves to be in the Hall too.”
“Israel has the right to cross into Lebanon and deal militarily with the people shooting rockets into its territory. Just think what the U.S. would do if there were some group just over the border in Canada firing rockets into this country.”
“You can see it coming already. Once Dakota Fanning gets a little older, she’s going to fall apart spectacularly and publicly like Brittany Spears and Lindsey Lohan.”
“But you let Raven stay out until midnight when she was 17! Why won’t you let me!”
“Teaching Creationism as an alternative to evolution is no more justified than teaching the Stork theory as an alternative to biological theories of human reproduction.”
In each instance, you need to examine if the cases are similar enough in relevant respects to conclude that what’s true of one likely is also true of the other.
As is generally the case with the different fallacy categories, it’s not so much that an argument either is or isn’t a fallacy-it’s not black and white like that. It’s a matter of the degree of strength of the argument, of the analogy. The weaker it is, the more we can say it commits the questionable analogy fallacy.