Developing Intelligence

Suppose – rather reasonably – that soups which taste like garlic have garlic in them. You observe two people eating soup; one of them says to the other, “There is no garlic in this soup.” Do you think it’s likely that the soup taste like garlic?

If you said yes, then congratulations! You’ve just committed a logical fallacy (from the premise “if p then q” and “not q,” you have inferred p) so absurd that it’s only very recently been given a name. But don’t feel bad – this absurd inference, known as modus shmollens, can actually be elicited from a majority of adult human subjects when the situations are just right.

One such situation was demonstrated by Bonnefon & Villejoubert in 2007. They point out that, conversationally, human speakers are likely to make negative statements when they will correct the erroneous inference of a listener. That is, unless there is a good reason to believe (for example) that it might be snowing, there is little reason to state that it is not snowing.

In this example, why might a speaker believe that it might be snowing? One straightforward possibility is that both the speaker and listener have access to some other information – information we might call “p” – that supports the inference that it is snowing – which we might in turn call “q”. So, in a case where a speaker does bother to say that it is not snowing, or that a soup doesn’t taste like garlic (i.e., “not q”), one might intuitively guess that p is in fact true. Indeed, why else would the speaker bother to negate q?

Bonnefon & Villejoubert gave 60 young adults a series of situations just like this, which varied in whether the conditional “if p then q” premise was explicit in the situation or merely implicit, and whether the categorical “not q” premise was framed as an utterance by a human speaker or merely a fact of the world. In the situation where both the conditionals were explicit and the categorical premises were utterances, 55% of undergraduates actually endorsed the modus shmollens inference with high confidence. In a second experiment, the number was even higher – 75% of undergraduates endorsed the patently absurd modus shmollens inference.

To their credit, Bonnefon & Villejoubert do not tout this behavior as a new logical fallacy. Their view is much richer. They view their work as deriving from an infamous and often-criticized schism in psycholinguistics research, where “core” psycholinguistic phenomena are investigated independent of what are viewed as merely “pragmatic” phenomena which do not reflect a core language system. The obvious criticism of such an approach is that psycholinguistic theories which do not actually work in practice can be redefined so as to refer only to a small subset of situations where putatively “core” processes can be observed, and all other mere “pragmatic” phenomena swept under a rug. Bonnefon & Villejoubert suggest that for such an approach to be viable, we must take those pragmatic phenomena seriously as well, and begin to derive novel, falsifiable predictions based on them. As such, their demonstration of the problematic modus shmollens inference represents not merely a surprising and counterintuitive addition to the list of logical fallacies regularly committed by humans, nor merely insight into the context dependence of such fallacies, but also represents a more comprehensive approach to psycholinguistic theorizing.

Comments

#1 Burrows
November 16, 2011

Fantastic post I hope to see more, keep at it.
#2 Thad
November 16, 2011

Now let us imagine that Carole and Didier are eating a soup, and that Carole tells Didier there is no garlic in this soup. This negative utterance triggers the pragmatic inference that a reason exists in this context for Didier to believe there is garlic in the soup.

So far so good.

A taste of garlic would be a very good candidate.

No it isn’t a good candidate. 1. The taste of garlic cannot prompt Carole to utter “there is no garlic”. There is no inference that would allow it.

2. What if the soup recipe usually calls for garlic – if the soup is made in this style (reason p), then garlic (q) – if p then q – then saying not-q means the soup is not this style. That is more realistic context of pragmatic inference that some reason p to believe q exists.

I cannot imagine a situation where saying “there’s no garlic” implies the taste of garlic!

No to finish reading the paper.
#3 Brian Utterback
November 16, 2011

I think you need a better example. It is kind of hard to see why anyone would think that a soup that we are told contains no garlic would be likely to taste like it does. Who would make such an inference?
#4 Kevembuangga
November 16, 2011

LOL, this is wrong, wrong, wrong!!!
Yes, it’s a logical fallacy in that the conclusion certainly does not follow from the premise but it’s a PRAGMATIC (yes…), VERY USEFUL SAFE BET.
Methinks that rather than “psycholinguistic difficulties” this highlights the stupidity of scientists who get drunk on their own trade (physics envy anyone?)
#5 Moopheus
November 16, 2011

In the real world, there’s a certain amount of ambiguity in your soup example. Which is to say, your supposition that soups that taste like garlic have garlic in them might not be reliably always true. Hence, it isn’t entirely illogical or unreasonable to think that someone remarking on the unusual–a garlicy soup with no garlic–might be correct. I’ve made such statements myself, but usually because I made the soup, and knew what the actual ingredients were. The circumstances we find ourselves in in our daily lives often have less certainty than can be described by simple logical relations.
#6 Thad
November 16, 2011

Now let us imagine that Carole and Didier are eating a soup, and that Carole tells Didier there is no garlic in this soup. This negative utterance triggers the pragmatic inference that a reason exists in this context for Didier to believe there is garlic in the soup.

So far so good.

A taste of garlic would be a very good candidate.

What? Who tasted the soup? Who knows what’s in it?

If the soup recipe usually calls for garlic – if the soup is made in some particular style (reason p), then garlic (q), then saying not-q means the soup is not this style, which is not a fallacy _without additional information_.

1. If Carole and Diddier are at a restaurant, they likely do not know the actual ingredients.
Thus the taste of garlic cannot prompt Carole to utter “there is no garlic”. There is no inference that would allow it.

1.A. If Carole tastes garlic, and then utters “there’s no garlic”, Didier would think Carole in error because, as the authors point out, “the taste of garlic is a cue to the presence of garlic.” But he couldn’t be sure, so his conclusion is a possible not-p until he verifies p for himself.

1.B. If Carole utters “there’s no garlic”, because she actually failed to detect its taste, Didier would be wise to believe her with a strong not-p.

1.C. If Carole doesn’t know what garlic tastes like, then Didier should dismiss her words.

2. In only one condition can yes-p be the inference. If Didier tasted garlic, and believed it to be of the correct style (if p then q), but Carole corrects his conclusion – it only tastes like garlic, but surprisingly doesn’t actually contain garlic (not-q), then yes indeed the soup tastes like garlic. THIS PRESUPPOSES THAT CAROLE KNOWS THE SOUP’S INGREDIENTS.

And following the other test questions, the other situations presupposes Geraldine knows the actual artist on the record, or that Emile knows the background of the professor, or that Alice knows the taxonomy of big cats.

In each of these cases, the real fallacy comes about when you presuppose the person who utters the categorical premise posses certain knowledge. If p then q, (and A has knowledge of q), A utters “not-q”: then yes-p.

But generally the test questions do not indicate that that person has that knowledge.

For a long time I couldn’t fathom the soup case. It seemed absurd, since I assumed that Carole said that because she didn’t taste garlic, not because she knew there was no garlic. An no time did the authors indicate that Carole possessed such knowledge.

This is why they find that…

The results of Experiment 1 support our prediction that reasoners will endorse the Modus Shmollens inference from an epistemic conditional ‘‘if p, then q’’ when the minor premise not-q is an utterance rather than a mere sentence.

It’s the utterance – it allows the test subject to infer the speaker has the requisite knowledge to make such a statement. But even then, only half the time do people infer that. Surprisingly, I can’t find the authors discuss this point.
#7 copdahl
November 16, 2011

I don’t buy “modus shmollens” as a new fallacy.

Instead of “If p then q; ~q, therefore p” (which is fallacious) the story about garlic should be read as “~q and reasons to believe [via Grice's conversational implicature] p; therefore ~(if p then q)”.

Or, in other words, if the statement ‘this soup has no garlic’ implies the speaker also believes that the soup tastes like garlic (otherwise why even mention it?), it is NOT a fallacious inference but instead a DENIAL of the original conditional.
#8 Lotharloo
November 16, 2011

Suppose – rather reasonably – that soups which taste like garlic have garlic in them.

Why? That’s an unreasonable assumption. At least to me, the only reason that “there is no garlic in this soup” might imply “the soup tastes like garlic” is the possibility of existence of soups that taste like garlic but have no garlic.
#9 Rev.Enki
November 16, 2011

So, the question is actually: what sort of information might lead someone to think garlic might be in a soup, decide there isn’t any, and (I think this is the important part) find this interesting enough to comment on. To me, the most likely scenario seems to be that they thought they tasted garlic, but couldn’t see any confirmation of the fact (or that garlic wasn’t listed in the ingredients, or something equivalent) so believed that their failure to confirm the garlic meant that that there really wasn’t any. So, maybe we’re just deciding which is more likely, a false positive for garlic flavor, or a false negative for the confirming evidence?
#10 Rev.Enki
November 16, 2011

I think I can say this more succinctly. I don’t think we’re applying purely deductive reasoning to to the obvious evidence, and it’s a mistake to pretend that’s what we’re doing and label it a “fallacy.” What we’re *actually* doing is applying a mix of inductive and deductive reasoning to what we *infer* to be the experiences of our companion.
#11 Insufferable Pedant
November 17, 2011

I think most of you are overthinking this.

You know how, when you tell a child not to do something, it will invariably do exactly that which you told it not to? Same here: Quick, don’t think of a pink elephant!

The mention of garlic in the soup, or the lack of it, prompts the thought of garlic soup. And how would we know if the speaker meant it as an assertion (she’s the cook) or a question?

Of course we think: “Someone mentioned garlic in the soup, therefore there must be soup that tastes like there is garlic in it.” Only when asked why we think that, we rationalize that: “Why else would they mention it?” Clearly, something unexpected must have occurred to trigger that statement: Maybe it does taste like garlic when it shouldn’t. Or maybe it doesn’t when it should. 50-50. (Or, apparently, more like 55-45.) As usual, the truth is much simpler: They only mentioned it to mess with us.

And it works: Wouldn’t you rather have the soup that is clearly labeled as being asbestos-free than the one that isn’t?
#12 Sam C
November 17, 2011

I think this post demonstrates the underlying fallacy of logicians: that human speech involves logic. Human speech is partly about communication, largely about the speaker influencing the listener. It is NOT about logic assertions and deductions. Major fail by poster!

If I say “I don’t have a cup of coffee”, that can indicate many things, but it is rarely intended as a simple assertion of a boring fact. It is reasonably unlikely as a statement anyway, because something will have happened to prompt the utterance, and that something will probably lead to the topic of the sentence being understood and replaced by a pronoun or elided completely.

So to come up with this drivel about a diner saying “There is no garlic in this soup.” is asking the hapless victims of these experimenters to speculate on what went before that prompted this utterance, the style of speech of the speaker and their motivation.

The statement CANNOT be a response to the question “Is there any garlic in your soup?”. Why? Because it’s completely idiotic style of speech, which would itself pose the question in the other person’s mind “why on earth is my companion talking so weirdly?” So it is not an answer to a specific question about the soup or garlic.

OK, the circumstances under which someone would say this when the soup has garlic in it are very small; the two most obvious are that the speaker is exaggerating slightly, saying “no garlic” to mean a disappointingly small amount, or perhaps they have a garlic-hating companion and want to tease them into trying something that will revolt them for the lolz.

Speech is NOT about logic!! Silly examples of implausible utterances might delight logicians but provide more insight into the mind of the logician than into the working of normal humans’ brains.
#13 Becca
November 17, 2011

It was the dishwasher who never cleaned out the pot of the previous garlicky contents of Tuscan white bean and garlic soup!
#14 Rick
November 17, 2011

An interesting article on pragmatics.

However, there is no logical fallacy; the mistaken belief that there is arises from a confusion between a general statement that allows exceptions (e.g. “Generally speaking, if a soup tastes like garlic, then there is garlic in the soup”) and a universal one (“(In every case) if a soup tastes like garlic, then there is garlic in the soup”). Modus tollens requires a universal premise in order to be valid, modus tollens is not a valid deduction from a generalization that allows exceptions.

For example, “Generally speaking, if it is a bird, it can fly. This creature cannot fly. Therefor this creature is not a bird” is not a valid deduction; it could be a kiwi, for example.

In all three of the examples in the paper, the situations are treated as generalizations that allow exceptions rather than universals. Thus modus tollens does not apply, and there is no logical fallacy.
#15 Alberto Veleiro
November 18, 2011

I think this is just a case in which the utterance is understood as a bi-conditional statement, such as ‘if and only if p, ten q’. It is a common source of error in deductive reasoning and I think that Mental Models Theory of Reasoning (from Phil Johnson-Laird) accounts for this results and other known fallacies
#16 Sean
November 21, 2011

My favorite fallacy is proof by tradition.
My father p,
My grandfather p,
My great-grandfather p,
and I’ll be damned if I q!
#17 Houston Chiropractor
November 26, 2011

I think it is like telling the child not to go out on the back porch when you are gone. As soon as this statement leaves your mouth, your child is already thinking about being on the back porch.
Don’t say what you don’t want.
#18 balderesi
November 26, 2011

It was the dishwasher who never cleaned out the pot of the previous garlicky contents of Tuscan white bean and garlic soup!
#19 best films
November 28, 2011

A really enjoyable read Chris. I’ve been casally studying pragmatics for a while. I look forward to reading the rest of your articles.

Darren Low
#20 v-pills
January 19, 2012

However, there is no logical fallacy; the mistaken belief that there is arises from a confusion between a general statement that allows exceptions
#21 afrika mangosu
February 7, 2012

However, there is no logical fallacy; the mistaken belief that there is arises from a confusion between a general statement that allows exceptions (e.g. “Generally speaking,
#22 Sesli Chat
February 12, 2012

And following the other test questions, the other situations presupposes Geraldine knows the actual artist on the record, or that Emile knows the background of the professor, or that Alice knows the taxonomy of big cats.
#23 Sesli Chat
February 13, 2012

1.B. If Carole utters “there’s no garlic”, because she actually failed to detect its taste, Didier would be wise to believe her with a strong not-p.
#24 Sesli Chat
February 20, 2012

And following the other test questions, the other situations presupposes Geraldine knows the actual artist on the record, or that Emile knows the background of the professor, or that Alice knows the taxonomy of big cats.
#25 Sesli Chat
February 25, 2012

So, maybe we’re just deciding which is more likely, a false positive for garlic flavor, or a false negative for the confirming evidence?
#26 Jay Howard
March 4, 2012

It’s clear that a bunch of students of logic have tried to ply their craft on this situation in order to codify the questionable linguistic behavior into a formalized hypothesis–complete with sentential variables and the like.

What the authors bring to the fore is that there is indeed an opportunity to make poor judgment calls about the world based on some conventions of communication of this particular period in time, in this particular culture. And this observation, if not necessarily a “fallacy” proper, is most undeniably a repeatable, testable behavior pattern. What accounts for it, and what can we make of it?

What we can say is that we live in a consumer society–regardless of whether you are insulated by academia or making your way in the real world. And in our consumer society, someone is ALWAYS trying to sell us something. The assumption which arises from the statement that “there’s no garlic in this soup!” is that it tastes like garlic because of some artificial flavoring, or that there is a definite reason it DOES indeed taste like garlic despite the absence of the potent bulb.

It’s a natural assumption. Call it a fallacy or don’t, but either way, the assumption will (and does) continue to be made by a majority of people. And it is my opinion that this is so by virtue of the fact of our consumer society whose job it is to sell us shit we don’t need, and our natural defense mechanisms that develop as a result of the need to protect our resources in a hostile environment.
#27 afrika mangosu
March 5, 2012

Thus modus tollens does not apply, and there is no logical fallacy.
#28 Niel Carol
March 12, 2012

Kinda interesting topic but full of informative for that educative thoughts indeed. You use the title “Modus Tollens, Modus Shmollens!” is very touching and enjoyable sounds for me. Thanks!

Jason @ best pizza Houston.
#29 afrika mangosu
March 19, 2012

It was the dishwasher who never cleaned out the pot of the previous garlicky contents of Tuscan white bean and garlic soup!