Consider the following argument for a claim X:

  1. Present proof of statement X', which is similar to but different from X (and doesn‘t imply it).
  2. Conclude that X holds.


  1. Polls show that 80% of all people prefer strawberry over vanilla ice cream.
  2. Conclusion: Nobody likes vanilla ice cream.

This is a fallacy, since it is perfectly possible to prefer something over something else, while liking both of these arbitrarily much.

Does this fallacy have a name?

Can this be classified as a straw man fallacy (even if you are not misrepresenting someone else's arguments)?

  • 7
    It's not so much a strawman fallacy as a strawberry fallacy.
    – Alex
    Dec 19, 2018 at 17:49
  • 3
    It is not a single given fallacy, it just a rhetorical technique, often known as misdirection or legerdemaine.
    – user9166
    Dec 20, 2018 at 11:36

4 Answers 4


Here is Wikipedia's description of the straw man fallacy:

A straw man is a common form of argument and is an informal fallacy based on giving the impression of refuting an opponent's argument, while actually refuting an argument that was not presented by that opponent.

The example provided is

  1. Nobody likes Vanilla Ice Cream.
  2. Indeed polls show that 80% of all people prefer strawberry over vanilla ice cream.

Assuming the second statement is true, which seems doubtful, the problem with the first statement is that it might be too broad. If the second statement is true, then 20% do prefer vanilla ice cream and so claiming "nobody" does is inaccurate.

Bo Bennett has a couple pseudo-logical fallacies that might describe this situation better than "straw man argument":

Too Broad: The definition includes items which should not be included. This is more of an error of fact than reason.

Weasel Wording: Using ambiguous words in order to mislead or conceal a truth: “Save up to 50% or more!” This is more of a marketing gimmick than a fallacy.

The question is:

Can this be classified as a straw man fallacy (even if you are not misrepresenting someone else's arguments)?

Wikipedia further states:

This technique has been used throughout history in polemical debate, particularly in arguments about highly charged emotional issues where a fiery "battle" and the defeat of an "enemy" may be more valued than critical thinking or an understanding of both sides of the issue.

Without having an opponent to defeat there does not seem to be any point in calling the argument a straw man when other terms, such as "too broad" or "weasel wording", would be more descriptive even if they aren't considered more than "pseudo"-logical fallacies.

Bennett, B. "Pseudo-Logical Fallacies" https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/6/Pseudo-Logical-Fallacies

"Strawman" Wikipedia https://en.wikipedia.org/wiki/Straw_man

  • The example given may not be the best; I just gave it to be concrete. The fallacy I have in mind is really this: You claim to prove X by proving a different statement X‘ (which at first glance might seem to imply statement X, but which really doesn’t). Dec 19, 2018 at 22:32
  • The closest item from Bennets list to the above fallacy seems to be Snow Job, but it‘s not precisely the same. Dec 19, 2018 at 22:39
  • 1
    @BlenderBender. It's beginning to look pretty clear that the fallacy embodied in your example does not have a standard name. Since this is what you are looking for, I have deleted my answer because what interests me is the analysis of what is going wrong in the example - and a number of things certainly are - and this analysis leads me to conclude that the search for a standard label is unlikely to be fruitful. But you have definitely given us something to think about and puzzle over. I welcomed the stimulation. Best - GLT.
    – Geoffrey Thomas
    Dec 20, 2018 at 9:49

This would appear to be just an invalid argument, as it takes the premises of a proof of X' (or the simple assertion of X') and asserts that they imply X, when it appears to be a premise of this question that they do not. One might call it a non-sequitur - a formal, logical or deductive fallacy - but to identify it as a more specific specific named fallacy, one would have to look into the reasons why its proponent claims X' implies X.

  • I would like to maintain that it‘s a proof of X' being presented, as opposed to the mere statement of X', what should be characteristic of the fallacy I have in mind. Dec 20, 2018 at 18:51
  • Also, it seems that this fallacy is similar to a motte-bailey-fallacy. Dec 20, 2018 at 18:55
  • @sdenham. I agree, see my comment at foot of page.
    – Geoffrey Thomas
    Dec 20, 2018 at 19:11
  • @BlenderBender I see the similarity with motte-and-bailey, but there the person switches between different conclusions (typically, one that would be significant and one that is unremarkable) according to circumstances, and so is being inconsistent. With regard to there being a (sound) proof for X', that might be regarded as a lemma within the (fallacious) argument for X. I'm with Geoffrey Thomas in thinking that this probably is not generally recognized as one specific type of fallacy.
    – sdenham
    Dec 20, 2018 at 20:01

This is a textbook example of "faulty generalization" where a specific case is illegitimately taken as representative of a larger group. The most familiar form is "hasty generalization" where a larger conclusion is drawn from a very small dataset (as when a racist generalizes about an entire group of people from the one bad person of that race he or she has met).

This case is particularly egregious, since the conclusion actually contradicts the evidence, which is most naturally taken as implying that 20% of people prefer vanilla to strawberry.

There is an additional illegitimacy in the line of argument, in as much as preferring strawberry is not the same as disliking vanilla. This is perhaps a disguised version of the equivocal fallacy, where a term with one meaning in the given context is treated as if it had a different (related but distinguishable) meaning.

  • I am posting this as community wiki because this answer was given twice before, and both times deleted by moderators. Thanks to @metaquest and metaquizz for providing this correct answer. Jan 4, 2019 at 17:30

Prefer and like aren't the same thing.

It's quite plausible that everybody likes vanilla ice cream (after all, it doesn't taste of anything so no reason to dislike it) but they all prefer something else (which varies according to what actual taste they like).

It looks like ignoratio elenchi to me.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .