Tomatoes
post by hangedman · 2010-10-23T18:50:14.100Z · LW · GW · Legacy · 6 commentsAre tomatoes fruits or vegetables?
I've been reading Eliezer's criticisms of Aristotelian classes as a model for the meaning of words. It occurred to me that this little chestnut is a good illustration of the problem. The best part about this example is that almost everyone has argued either on one side or the other at some point in their lives. One would think that the English speaking world could come to some consensus on such a simple, trivial problem, but still the argument rages on. Fruit or vegetable?
In my experience, the argument is usually started by the fruit advocate (we'll call him Lemon). "It's the fruiting body of the plant," he says. "It contains the seeds." He argues that the tomato is, by definition, a fruit.
Bean has never thought of tomatoes as fruits, but when her belief is challenged by Lemon, she's not entirely sure how to respond. She hesitates, then starts slowly -- "All the things I call fruits are sweet," she says. "Not that tomatoes are bitter, but they're certainly not sweet enough to be fruits." Bean is proposing a stricter definition -- fruits are sweet fruiting bodies of plants. But does Bean really think that's the difference between a fruit and a vegetable?
Not really. Bean learned what these words mean by talking to other people about fruits, vegetables, and tomatoes, and through her cooking and eating. There was never any moment when she said to herself, "Aha! a tomato is not a fruit!" This belief is a result of countless minute inferences made over the course of Bean's gustatory life. The definition she proposes is an ad hoc defense of her belief that tomatoes are not fruits, not a real reason.
Bean's real mistake was to think that she needed to defend her belief that tomatoes are not fruits. Tomatoes are what they are regardless of how they're classified, and most people classify them as fruit or vegetable long before they learn anything about Aristotelian classes or membership tests. The classification is made as the result of a long history of silent inferences from the way parents and peers use those words. The first English dictionary was written in 1604, several hundred years after both "fruit" and "vegetable" had entered the English vocabulary (right about the same time as "tomato," actually). Before that, Lemon couldn't point to a definition to make his case. He could only rely on his experiences with usage just as Bean does in her rebuttal, and it's not clear why we should privilege one's experience other the other. The meaning of the word is prior to the definition.
There is a simple solution to the tomato problem, by the way. "Vegetable" is any edible plant matter and "fruit" is a subclass of "vegetable." All fruits are vegetables, and so tomatoes are both. In the same way, wheat is both a grain and a vegetable. The distinction is made only for convenience -- consider the fact that before electric guitars were invented, there were no "acoustic guitars" -- only "guitars." It's not false to describe an acoustic guitar as a guitar, merely imprecise. This points to what I think is a common phenomenon in spoken language which leads to errors in reasoning: a distinction is made between a subclass B and superclass A with the understanding that x in B -> x in A; later, the distinction is maintained but the understanding of the interconnection is lost so that A and B are considered distinct categories -- x in A xor x in B. Can anyone think of any other examples of this kind of error?
6 comments
Comments sorted by top scores.
comment by Relsqui · 2010-10-23T21:06:08.450Z · LW(p) · GW(p)
it's not clear why we should privilege one's experience other the other
Because in the argument as described, Lemon does have more than his experience to go on--he also has the agricultural definition. That's why we invented definitions. To resolve these arguments.
In the same way, wheat is both a grain and a vegetable.
Isn't this another clear example of prioritizing the technical definition (and, in fact, an arbitrary one of several overlapping technical definitions) over the common usage, which you appeared to be arguing against? If you tell someone you're cooking vegetables, and they come over to find you with a bowl of cream of wheat, you've been misleading, and I don't see how that's a productive use of language.
As a rule of thumb, I prioritize the precise definition unless it's misleading in the given context. When I say "the performance was so bad, they threw rotten fruit," no one will be confused if some of the projectiles were tomatoes. But I wouldn't say "I'm serving fruit and cheese" when I plan to put out tomato slices with mozzarella. It's true, but it creates a false expectation.
Replies from: hangedman↑ comment by hangedman · 2010-10-24T22:21:14.424Z · LW(p) · GW(p)
Because in the argument as described, Lemon does have more than his experience to go on--he also has the agricultural definition. That's why we invented definitions. To resolve these arguments.
It seems to me that introducing the definition is what starts this argument, not what resolves it. But playing along, are eggplants fruit? Green peppers? String beans? If my hypothesis is that "fruit" means "fruiting body of a plant" and my experiment is to ask people whether things that fit the definition are fruit, does the hypothesis anticipate the results of the experiment?
When I say "the performance was so bad, they threw rotten fruit," no one will be confused if some of the projectiles were tomatoes.
Do you think anyone would be confused if some of them were turnips?
If you tell someone you're cooking vegetables, and they come over to find you with a bowl of cream of wheat, you've been misleading, and I don't see how that's a productive use of language.
But I wouldn't say "I'm serving fruit and cheese" when I plan to put out tomato slices with mozzarella. It's true, but it creates a false expectation.
These are much more clever ways of making my point. There are exceptions, I think -- I'd prefer if people stopped using "theory" to mean "guess" -- but in many cases, it would only be confusing if people used a particular definition of a word to decide their usage.
Replies from: Relsqui↑ comment by Relsqui · 2010-10-25T00:32:17.275Z · LW(p) · GW(p)
Was your point, then, that you should use words in the way which is most likely to be clear and useful? If so, I agree, and would rather address it directly.
The clear and useful definition varies, of course, with context--a plant biologist and a pastry chef use the word "fruit" differently, and that's okay! The trouble arises when you have two people in conversation trying to use a word for which they have different contexts, as is the case when lay people interpret "theory" as "guess," since that's what it often means in prose. In such cases, more or different words are needed to ensure the right meaning is being interpreted.
Replies from: hangedman↑ comment by hangedman · 2010-10-25T03:28:56.997Z · LW(p) · GW(p)
My point was that "fruit" is the flip-side of Eliezer's bird example. A bird is a feathered flying thing. What about an ostrich? What about a penguin? A fruit is the seed pod of a plant. What about eggplants? What about cucumbers? My intention wasn't to give any advice about how to pick and choose definitions or interpret words generally or in a particular context.
The reason I started the discussion is because I think examples like "fruit" where the common usage of the word deviates from any strict definition can help us to understand language and language acquisition better. But this means that we should be asking people why they don't think cucumbers are fruit, not insisting that cucumbers are fruit by definition of the word fruit.
Replies from: erratio↑ comment by erratio · 2010-10-25T19:33:01.575Z · LW(p) · GW(p)
The reason I started the discussion is because I think examples like "fruit" where the common usage of the word deviates from any strict definition can help us to understand language and language acquisition better.
I suggest reading up on Eleanor Rosch's prototype theory, it explains the whole categorisation thing very clearly. The basic theory is that our categories aren't Aristotelian classes, but are fuzzy, and formed from seeing exemplars of a class and extracting the most common features. It's a well-understood property of cognition and has plenty of experimental backing.
In the case of cucumbers, I'd say it fails the 'fruit' test because it lacks almost all the features I associate with fruit: it's not red/yellow, it's not sweet, I don't eat it for a snack or dessert, and it doesn't have obvious seeds or pits inside. Therefore I would look at you very oddly if you told me that it was.
Replies from: hangedman