I’ve been trying to figure out what precise kind of fallacy is the line that runs “Hitler wore pants, you wear pants, therefore you’re like Hitler.” It’s not a reductio, which would be closer to “Hitler wore pants, therefore wearing pants is bad.” It kind of looks a lot like a false analogy – hat tip to Zoltan Majdik for pointing that out – but if it is, it’s in a weird form. GT Goodnight suggested it’s a fallacious syllogism that’s been rendered enthymematically, where the problem is that the middle wasn’t distributed correctly. That seems close too, but the statement does bear an awful strong familial resemblance to false analogy, which would put it on the side of induction.
What it actually looks like is a fallacious induction that’s been rendered “enthymematically,” where the specific (informal) fallacy is that it’s a false analogy. So two more questions to hash out in the comments, in addition to figuring out what statement technically is: (1) is there a name for the inductive equivalent of an enthymeme, where steps are suppressed for rhetorical purposes, and (2) is there a taxonomy of specific inductive fallacies that’s more detailed than the usual hasty generalization, unrepresentative sample, false analogy, slothful induction, and fallacy of exclusion?
In a standard false analogy we (tenably) reason that Y is like X because Y shares certain (presumably salient) properties with X. But then we add the (informal) fallacy of insisting that therefore Y has a specific negative property which is actually unique to X:
X has properties {P1…Pn} of which one is Q
Y has at least one property {P1…Pn}
Therefore Y is like X
Therefore Y has property Q
To get back to the Hitler line, that would roughly be:
X (Hitler) has P (wears pants) and Q (is really bad guy)
Y (you) has P (pants)
Therefore Y (you) is like X (Hitler)
Therefore Y (you) has Q (is really bad guy)
But the statement is very importantly not that, which is rhetorically critical. Once you make Q explicit – once you enunciate the stuff about being a really bad guy – the reasoning becomes obviously fallacious. So instead any mention of Q is totally suppressed, both in relation to X and in relation to Y, and we have:
X has P
Y has P
Therefore Y is like X
That’s the form of “Hitler wore pants, you wear pants, therefore you’re like Hitler.” Again this is doubly weird because we’re dealing with an induction that’s fallacious. So even if there was a taxonomy of inductive enthymemes that symmetrically mirrored the standard enthymemes, we wouldn’t get that here because the fallacy is screwing up the symmetry.
In fact just by suppressing Q we already get three different rhetorical versions:
[A] Q’s relation to both X and Y is suppressed (that’s the one we just did, which matches “Hitler wore pants, you wear pants, therefore you’re like Hitler”):
X has P
Y has P
Therefore Y is like X
[B] Q’s relation to X is suppressed, but Y having Q is explicitly stated:
X has P
Y has P
Therefore Y is like X
Therefore Y has Q
[C] Q’s relation to Y is suppressed, but X having Q is explicitly stated:
X has P and Q
Y has P
Therefore Y is like X
There are obviously more elements to suppress which would create more combinations. You can rhetorically skip the statement “therefore Y is like X” and jump directly to “therefore Y has Q,” for instance.
But the questions are: (1) is “Hitler wore pants, you wear pants, therefore you’re like Hitler” bad induction or bad deduction, (2) what kind of bad induction or deduction is it, (3) does anyone have a detailed taxonomy of fallacious inductions?
References:
* zoltanmajdik [Twitter]
* goodnight [Twitter]
* Fallacy of the undistributed middle [Wikipedia]
* Inductive Fallacies [One Good Move]
Related Icon Index Symbol Categories:
* Classical Rhetoric
* Argumentation
* Rhetoric and Philosophy
Not to get all warm and fuzzy, but are you sure it’s a fallacy? I think you’re right that rhetorically it’s in the family of false analogies, but I’d be willing to suggest that it’s actually true.
So, perhaps a “dangerous” rather than false analogy. Dangerous, because our usual chain of inference would take us, by selective editing, to an erroneous conclusion.
Greed is bad.
Adam is greedy.
Adam wears dockers.
Joe wears dockers.
Therefore, Joe is greedy?
The error is probably closer to one of psychological bias: our tendency to see the world through the lens of our recent experiences. Since the only salient things you know about Adam are that he’s greedy and that he wears dockers, if I bring up dockers, then your mind is tempted to make the connection to greedy, even though lots of other things are connected to dockers too. It’s an artifact of the tendency of the our minds to make spurious associations.
But, as I said, are we sure it’s a fallacy? It is true that Adam and Joe wear dockers: they have that in common. Perhaps the problem is in the lack of logic in the listener, who we cannot count on doing the work of recognizing that pants are orthagonal to the issue of greed.
Right, but that’s always a problem with induction. You’re going from examples to rules rather than rules to examples – or samples to population rather than population to sample – and so you’re always in danger of making a false inference. But what makes this an (informal) fallacy is that the attribution of Q is done… shall we say, with a particular motivation?
And it’s also absolutely right that there’s a psychological bias in play, since we’re hardwired to make quick generalizations. But we’re hardwired to draw all kinds of different inferences. The point is abstracting the intuitions and then systematizing them.
1. re: Adam. Whether it’s a fallacy or not is irrelevant to it being true. Logic should only consider the relationship between propositions… the relationship between the propositions and the world is a separate kind of concern.
2. If I had to label it, I’d go with Hasty G (so I’d view it as an egregious induction). But i see the problem… maybe there’s more going on in the background than an induction with a small n. I think it’s clumsier to say it’s enthymeme (because we don’t know anything abou the audience, so it’s kind of fuzzy and pointless) than to just say that “like” is vague. So Adam, I think correctly, sees it as potentially true… given his uptake of “like”… while the presumed enthymeme would have the likeness be, presumably, genocidal racism (though, as i say, depending on audience it could also be strange facial hair, failed art students, people with one Ball, etc.).
3. There’s a formal version of this kind of thing, but with the middle term being ambiguous. That’s probably a more common problem… using a squishy middle term to allow properties to mistakenly transfer down the chain. I couldn’t find reference to one where the conclusion itself contained the errant ambiguity… Though lists of informal fallacies often include “equivocation” or “ambiguity” in some form: eg http://www.fallacyfiles.org/equivoqu.html
4. I guess the most solid historical reference to this kind of fallacy, though not very tight at all, would be aristotle’s fallacy of accent. That was a peculiar problem with how ancient greek worked with pronunciation and meaning… but the logical problem seems similar. Without the context of utterance some terms could not be accurately unpacked… In this case there aren’t just 2 lexically distinct possibilities, but as many as there are interpreters and properties.
5. I’m sure i’m missing something, because we’ve been drawing up taxonomies of these suckers for 2500 years… but you’re using an example drawn from comic hyperbole, so maybe it’s your fault. If there isn’t a category, though, maybe “Criminal Vagueness”.
6. There’s a slightly interesting sidenote about whether it’s really vagueness at work here, or generality… but I’ll leave that aside… with the comment that it’s both but that the logical error results from vagueness (which is what causes the enthymeme stuff the be relevant).
7. Fix the link to my new blog, fool.
It’s not exactly accent, although I got sucked into that possibility for a while. I think accent leans way too heavily on verbal emphasis. There might be an analogy to accent, but that still begs the question of whether this kind of line has been cataloged as an inductive fallacy.
And let’s not be too harsh on taxonomic work. Of course it dangerously flirts with ossification during those times when boldness in the humanities is lacking, and when scholars are being too modest with their claims. But is that really the problem with the contemporary academy?