Joho the BlogThe fallacy of examples - Joho the Blog

The fallacy of examples

Nicholas Kristof has a terrific column today about how the donation of a goat to a family in Uganda ultimately led to one of the children, Beatrice, earning a degree from Connecticut College, and beginning a path of service for her community. It’s a wonderful story, the point of which is what Jeffrey Sachs calls the “Beatrice Theorem” of development economics: “small inputs can lead to large outcomes.”

Well, yes, of course. In fact, small changes have determined the success or failure of us all. And I have no misgivings whatsoever about this past Channukah having given our children certificates announcing that Oxfam had given goats in their name. Yes, I am a goat-giver, and proud of it.


…I’ve noticed in business writing in particular the frequency of what we can call the Fallacy of Examples (a type of Fallacy of Hasty Generalization). You read some story about a successful CEO as if we should learn from his (yes, usually it’s a him) example. But we are struck by examples frequently because they’re exceptional. As exceptions, examples are the last thing you want to learn from.

Not always, though. Sometimes examples are typical. That’s different. The trick is determining which are which.

An even when you can, you’re still not done. Is Beatrice and her goat an exception? Yes. That’s why her story is so inspiring. As an exception, it may be exactly what we should not be emulating. After all, if she’d won the lottery, we wouldn’t think that giving lottery tickets to the poor is a sensible approach to the problem of world poverty. But, even though Beatrice is an exception, the typical effect of donated goats (and other such small-ish gifts) may be quite good.

That’s why the Fallacy of Examples is a fallacy. Reasoning from examples doesn’t always lead to false conclusions. The reasoning just isn’t enough to tell you what the valid conclusions are.

And in the absence of valid conclusions, here’s Kristof’s list of ways to donate goats or their equivalents. And here’s Oxfam’s program. And, because it’s the Internet, here’s samizdata’s warning that goats cause poverty. [Tags: ]

Ethanz brilliantly contextualizes this post. Thanks, Ethan!

15 Responses to “The fallacy of examples”

  1. You may be interested in this book on “evidence-based management”, which talks about deciding certain issues of how to manage companies (e.g., “pay for performance”) based on gathering evidence instead of reading inspiring stories about what this or that heroic CEO did to turn a Fortune 1000 company around.

    (h/t Brad Hicks)

  2. […] Weinberger has an intriguing post up today about the “Fallacy of Examples“. He’s reacting to a column from Nick Kristof in the New York Times titled “The […]

  3. A better example would be, IMO, DB Cooper.

  4. ‘The fallacy of examples’ is a horribly misleading name for a species of fallacy.

    I had a look through my informal logic books and couldn’t find mention it, nor was there a mention of it on the fallacyfiles taxonomy page. There are a couple of hits on Google books though, not that the absence of the name strikes a hit at the fallacy itself. I’m just curious as to where you picked up the name, as it sounds to me like a renaming of the hasty generalisation fallacy or perhaps the overwhelming exception fallacy.

    I know the species of fallacy you are talking about; nevertheless, to name it ‘fallacy of examples’ seems to deny ‘examples’ their power in specific cases of argument. One very important use of examples that comes to mind is counterexamples to refute universals. Take for example your argument. If person A was to say, ‘Giving third world people a goat will never lead to anything good’. Then a counterexample from person B who stated, ‘well there was a case where …’ which refuted that universal claim.

    You also stated, ‘That’s why the Fallacy of Examples is a fallacy. Reasoning from examples doesn’t always lead to false conclusions. The reasoning just isn’t enough to tell you what the valid conclusions are.’

    I would disagree with your terminology. I think it is confusing two different types of argument. Deductive reasoning can be valid. Inductive reasoning, which hasty generalisations are often associated with, cannot be valid. They can only be strong or weak. So the reason that it is fallacious has nothing to do with its validity (if it did, then a hasty generalisation would be a formal fallacy based on the structure of the argument itself).

  5. @Yo There’s no such thing as a definitive list of informal fallacies.

  6. @ Yo Inductive arguments can be valid or invalid. You’re right in saying the strong/weak distinction is present w/ inductive arguments, but they also are either valid or invalid.

    Example of Valid Inductive Argument:

    1. If you’re a politician, you probably lie well.
    2. You’re a politician.
    Therefore, you probably lie well.

    Example of Invalid Inductive Argument:

    1. If you’re a politician, you probably lie well.
    2. You’re a mortician.
    Therefore, you probably lie well.

    The conclusion in the second may very well be true, but it doesn’t follow from the premises. The logical form is invalid, and it’s an inductive (probabilistic) argument.

    What you may be thinking of is formal vs. informal reasoning instead of deductive vs. inductive. Deductive and inductive reasoning are both formal reasoning processes. Informal reasoning is more often associated with logical fallacies.

    Also, Yo, you said ‘to name it ‘fallacy of examples’ seems to deny ‘examples’ their power in specific cases of argument.’

    Um, at best that’s quibbling. At worst, that’s wrong. An appeal to authority is an informal fallacy, but we don’t think that calling it that makes genuine authorities any less authoritative.

    It’s called that because we sometimes inappropriately say, ‘You should believe X because Bill believes X , and he’s an expert’ instead of ‘You should believe X for these reasons, all of which are detailed by Bill, who’s an expert in the subject.’

    Calling this ‘The Fallacy of Examples’ doesn’t entail or even necessarily imply that the use of examples is always bad. The post says quite the opposite.

    And for what it’s worth, I vote for this being an extremely well done post.

  7. Jay Kelly, Where did I say there was a definitive list of fallacies? I never did. I asked for an explanation of the etymology of a term that seems to be an unnecessary renaming of a fallacy, with an ambiguous title, that already existed.

    I think you don’t understand the concept of validity in logical argument. Your examples make no sense at all and shows you are seriously confused about logical argument.

    The examples you gave are if-then conditionals based on on Modus Ponens (a type of deductive argument), which the structure is as follows:

    (1) If A, Then B.
    (2) A.
    (3) Therefore B

    Your example of an invalid inductive argument is also completely wrong and bordering on a red herring.

    Being a mortician has nothing to do with the ampliative conclusion that someone lied. If it were an inductive argument that premise can be ignored, as it neither strengthens the conclusion, nor makes the argument defeasible.

    Given that the mortician premise is irrelevant to your example, then that leaves us with the if-then conditional. The If-then conditional is asserting a causal relationship between being a politician and probably lying. If-then conditionals by themselves aren’t arguments, they are part of premises and demand further support.

    Saying the following:

    (1) If you’re a politician (A), you probably lie well (B).
    (2) Therefore you probably lie well.

    begs the question, as you need further reasons to support (B) i.e. further observations for the probabilities. So, really it isn’t an argument at all. It’s an assertion.

    Your comparison to arguments from authority to my problem with the use of the word ‘examples’ is irrelevant. Arguments from authority are fallacies of relevance. The fallacy of examples has nothing to do with relevance, if it did the original post would be about how ‘the fallacy of examples’ is completely irrelevant (rather it is a fallacy because it relates to an inductive fallacy insofar as it doesn’t give us enough information).

  8. Jay Kelly, perhaps, to make my last paragraph more clear, as I think you made a strawman out of my argument when you stated “Calling this ‘The Fallacy of Examples’ doesn’t entail or even necessarily imply that the use of examples is always bad. The post says quite the opposite.” I never said they did imply that. Examples are an important part of inductive arguments (and also important in refuting deductive arguments based on universals). What I have a problem with is calling it ‘the fallacy of examples’ comes across as ambiguous. The majority of the more notable fallacies are named in a clear manner that when people quote them they can have an idea of what the fallacy entails e.g. slippery slope, poisoning the well, wishful thinking, biased sample etc. etc.

    ‘The fallacy of examples’ doesn’t do this. I dunno about you, but I’m left wondering what the ‘fallacy of examples’ is by the title alone, and secondly, when it is defined as an invalid argument it should be a deductively based fallacy, it isn’t. It’s based on the hasty generalisation, a form of inductive argument. So the whole re-wording of ‘the fallacy of examples’ and then adding a stipulative definition makes the whole thing muddled headed.

  9. […] I know! I had wanted to comment on David Weinberger’s and Ethan Zuckerman’s observations about the Kristof “Donate a Goat” article. David calls it […]

  10. @ Yo,

    Can you email me your CV so I can tell if you’re someone worth communicating with, or if you’re copying sentences out of a logic text?

    You’ve misunderstood me, and perhaps I’ve misunderstood you. Would love to clear that up, but if you’re a 3rd year undergrad with just enough logic under your belt to think you understand it, I’ll pass.

    If it turns out you have advanced degrees, then in all likelihood we’re just talking past each other, and I’d love to get some clarification on where you’re coming from and likely learn a bit along the way.



    jaymichaelkelly AT gmail DOT com

  11. […] – bookmarked by 2 members originally found by desorganisations on 2008-08-02 The fallacy of examples – bookmarked by 1 members […]

  12. [email protected] kelly being thoroughly pwned and then running away with his tail btw his legs by asking for CVs.

    Man up, child. If you can’t hang with people who know what they are talking about, then don’t get involved.

  13. […] So writes David Weinberger. […]

  14. your website didn’t answer well my question.

  15. yeah right , because you not involve in that so you nust stay away from the matter

Web Joho only

Comments (RSS).  RSS icon