Can you trust your intuitions? (Michael Devitt)
September 5th 2006 11:08
A very paraphrased summary of a lecture by Michael Devitt of the City University of New York. This lecture was delivered to the Russellian Society on Thursday 17 August 2006.
Apologies for any misreadings, misrememberings, etc.
Firstly, what is an intuition?
Well, what’s basic to it is that it’s not obvious where it comes from. It’s something like a sudden insight. A thought that hits you from out the blue. You might have an intuition that I’m a liar, or that you’re in danger, or that murder is wrong.
Kant divided knowledge into two types: “a priori” and “a posteriori”. Knowledge whose authority depends on experience is a posteriori, and knowledge independent of experience is a priori.
Many people think their intuitions give them direct, a priori access to the truth. So, if a philosopher believed this, then, for any particular question, they might take it as their task to find a theory that fits all their various intuitions. And they might do “thought experiments”.
For instance, if one philosopher proposes that “mind” is identical to “brain”, another might reply with the thought experiment, “But hang on. If we had a Martian with completely different anatomy, wouldn’t our intuitions tell us that we could still say it has a ‘mind’? So minds can’t be identical to brains.”
A friend gives this example. When she was teaching philosophy 101, she presented a method of hypothesis and counterexample. “What is a chair?” You might answer, “Something to sit on.” But someone else might say, “Does a rock count as a chair then?” And you might consult your intuitions, think “No”, then have to add more to the definition to distinguish rocks from chairs.
The particular example Devitt uses is “reference”. Apparently, some philosophers of language employ intuitions when they ask what it means for a word to refer.
Chomsky is skeptical about this method of defining reference. Says Chomsky (at least according to Devitt), there can be no intuitions about reference, because it’s a technical term. There can be no intuitions about “angular velocity” or “protein” for the same reason.
But Chomsky is not similarly skeptical about the role of intuitions in linguistics. He subscribes, says Devitt, to the view that linguists are at the mercy of the population’s intuitions. If you want to find out whether a sentence is grammatical, consult the folk.
Devitt distrusts all intuitions, including linguistic ones, and all thought experiments. The task when explaining x should be to explain x, not to systematize all one’s intuitions about x.
Devitt also disbelieves in “a priori” knowledge: “The best reason we have for rejecting the a priori is that we don’t know what it is, we only know what it isn’t. It isn’t empirical knowledge.”
So everything is based on experience, including intuitions. Maybe the intuition seems to come from nowhere, but your brain has been working behind the scenes. Your brain has adopted some theory or other, has processed all the data at its disposal in the light of that theory, and has come up with a conclusion.
Devitt's definition of intuitions: “Judgments that are empirical [that is, justified by experience] theory-laden central-processor responses to phenomena, differing from many other such judgments only in being immediate and unreflective, not based on any conscious reasoning.”
If this is what intuitions are, then we can only trust the folk’s intuitions if we suppose the folk are right in their theory and their processing. But why should we suppose this? Why give folk intuitions a special status?
We trust the psychologist’s insights over those of the layman.
Devitt suggests that the linguist’s intuitions about grammaticality are better than the folk’s, because the linguist has a better theory. If you ask the folk, “Is this sentence grammatical?”, maybe they’d tell you no, but maybe they’d go on to use that sentence without giving it a second thought.
We don’t necessarily have conscious access to what we do. A good tennis player doesn’t necessarily make a good coach. A good typist mightn’t be able to draw an accurate picture of a keyboard.
Devitt distinguishes “competence” and “expertise”. You can be competent without being an expert. All that competence gives you is ready access to a large data set. It doesn’t mean you can verbalize what you’re doing. And it doesn’t guarantee that your concepts and conclusions and theories will be better than anyone else’s.
“Know how” is different from “know that”.
Devitt adds: Linguists believe that competent speakers have the principles of grammar stored in their heads. When the folk have intuitions about grammaticality, they’re accessing those principles. Philosophers of language similarly believe that competent speakers have theories of reference stored in their heads.
This is the way we design computers. We give them rules. But it’s unnecessary to suppose that human competence is like this. It’s unnecessary to suppose that the typist mentally represents a keyboard.
This applies also to competence in thinking. Just because we’re good at it, doesn’t mean that we can describe the rules of it. And it’s extravagant to suggest that we need to have stored those rules. The only plausible explanation, says Devitt, is that thinking is governed by rules you don’t represent.
So, Devitt has three main beefs. He dislikes: (1) the idea that intuitions give you privileged, a priori access to truth; he thinks they’re never completely trustworthy, and they’re not superior to other sorts of knowledge; (2) the idea that anyone should take it as their task to systematize intuitions (the less intuition-mongering, the better); and (3) the idea that we need to represent our actions in order to perform them.
He does think, however, that there is some sort of role for intuitions.
1. The intuitions of the folk often are correct. In particular, intuitions that distinguish kinds (this is an echidna, that is not) are more dependable than intuitions that try to tell us something about those kinds (for instance, that try to specify what is common or peculiar).
But even if intuitions sometimes count as good evidence, there’s usually better sources. For instance, the linguist can analyze the “corpus”, the data from language as actually used.
2. The intuitions of experts can be trustworthy to the extent that expertise is trustworthy. Their intuitions have been developed in the framework of a good theory. Devitt gives the example of a group of art experts who all agreed that a Greek statue was fake. They were unable to say why, but were later proved right.
3. And sometimes we have no reliable theory. Sometimes you’re at the stage of “proto-science”, and you’re forced to use intuitions, you’ve got nothing better to go on. So we need intuitions when starting from scratch, but thought experiments are no substitute for real experiments, and we should eventually abandon reliance on intuitions as our theory develops.
In the Q&A session, Devitt notes:
* The ideology of linguists is that all they do is systematize the folk’s intuitions. But in actual practice they’re more inclined to rely on corpus evidence.
* “Innate” intuitions, intuitions you’re born with -- are these empirical judgments also? Are they an exception to the idea that every judgment is based on experience? Well, Devitt is skeptical that any such intuitions exist. But if they do exist, he’d want to explore the idea that they’re based on your ancestors’ experience.
* What about logical intuitions? Are these based on experience, and are they unreliable? -- Devitt would try to argue that, yes, they also are based on experience.
* Ethical intuitions? Shouldn’t philosophy try to systematize these? Isn’t that the bulk of what ethicists actually do? -- construct thought experiments, and try to find a theory that fits their moral intuitions? -- Well, Devitt wants to say that there are moral facts, and so there is no need here either, to take your task to be intuition systematization.
* Devitt also notes that a particular target is Oxford philosophy in the tradition of JL Austin, where philosophy is about clarifying concepts, and where one clarifies concepts by systematizing intuitions.
This article is licensed under the GNU Free Documentation License. It uses material from Michael Devitt's website, from the Das Keyboard website, and from the Wikipedia articles Immanuel Kant and Noam Chomsky.
Apologies for any misreadings, misrememberings, etc.
***
Firstly, what is an intuition?
Well, what’s basic to it is that it’s not obvious where it comes from. It’s something like a sudden insight. A thought that hits you from out the blue. You might have an intuition that I’m a liar, or that you’re in danger, or that murder is wrong.
Sage of Konigsberg
Many people think their intuitions give them direct, a priori access to the truth. So, if a philosopher believed this, then, for any particular question, they might take it as their task to find a theory that fits all their various intuitions. And they might do “thought experiments”.
For instance, if one philosopher proposes that “mind” is identical to “brain”, another might reply with the thought experiment, “But hang on. If we had a Martian with completely different anatomy, wouldn’t our intuitions tell us that we could still say it has a ‘mind’? So minds can’t be identical to brains.”
A friend gives this example. When she was teaching philosophy 101, she presented a method of hypothesis and counterexample. “What is a chair?” You might answer, “Something to sit on.” But someone else might say, “Does a rock count as a chair then?” And you might consult your intuitions, think “No”, then have to add more to the definition to distinguish rocks from chairs.
The particular example Devitt uses is “reference”. Apparently, some philosophers of language employ intuitions when they ask what it means for a word to refer.
***
"Big Noam"
But Chomsky is not similarly skeptical about the role of intuitions in linguistics. He subscribes, says Devitt, to the view that linguists are at the mercy of the population’s intuitions. If you want to find out whether a sentence is grammatical, consult the folk.
***
Devitt distrusts all intuitions, including linguistic ones, and all thought experiments. The task when explaining x should be to explain x, not to systematize all one’s intuitions about x.
Devitt also disbelieves in “a priori” knowledge: “The best reason we have for rejecting the a priori is that we don’t know what it is, we only know what it isn’t. It isn’t empirical knowledge.”
So everything is based on experience, including intuitions. Maybe the intuition seems to come from nowhere, but your brain has been working behind the scenes. Your brain has adopted some theory or other, has processed all the data at its disposal in the light of that theory, and has come up with a conclusion.
Devitt's definition of intuitions: “Judgments that are empirical [that is, justified by experience] theory-laden central-processor responses to phenomena, differing from many other such judgments only in being immediate and unreflective, not based on any conscious reasoning.”
***
If this is what intuitions are, then we can only trust the folk’s intuitions if we suppose the folk are right in their theory and their processing. But why should we suppose this? Why give folk intuitions a special status?
We trust the psychologist’s insights over those of the layman.
Devitt suggests that the linguist’s intuitions about grammaticality are better than the folk’s, because the linguist has a better theory. If you ask the folk, “Is this sentence grammatical?”, maybe they’d tell you no, but maybe they’d go on to use that sentence without giving it a second thought.
A blank keyboard for the style conscious. Are you l33t enough to fill in the blanks?
Devitt distinguishes “competence” and “expertise”. You can be competent without being an expert. All that competence gives you is ready access to a large data set. It doesn’t mean you can verbalize what you’re doing. And it doesn’t guarantee that your concepts and conclusions and theories will be better than anyone else’s.
“Know how” is different from “know that”.
***
Devitt adds: Linguists believe that competent speakers have the principles of grammar stored in their heads. When the folk have intuitions about grammaticality, they’re accessing those principles. Philosophers of language similarly believe that competent speakers have theories of reference stored in their heads.
This is the way we design computers. We give them rules. But it’s unnecessary to suppose that human competence is like this. It’s unnecessary to suppose that the typist mentally represents a keyboard.
This applies also to competence in thinking. Just because we’re good at it, doesn’t mean that we can describe the rules of it. And it’s extravagant to suggest that we need to have stored those rules. The only plausible explanation, says Devitt, is that thinking is governed by rules you don’t represent.
***
So, Devitt has three main beefs. He dislikes: (1) the idea that intuitions give you privileged, a priori access to truth; he thinks they’re never completely trustworthy, and they’re not superior to other sorts of knowledge; (2) the idea that anyone should take it as their task to systematize intuitions (the less intuition-mongering, the better); and (3) the idea that we need to represent our actions in order to perform them.
He does think, however, that there is some sort of role for intuitions.
1. The intuitions of the folk often are correct. In particular, intuitions that distinguish kinds (this is an echidna, that is not) are more dependable than intuitions that try to tell us something about those kinds (for instance, that try to specify what is common or peculiar).
But even if intuitions sometimes count as good evidence, there’s usually better sources. For instance, the linguist can analyze the “corpus”, the data from language as actually used.
2. The intuitions of experts can be trustworthy to the extent that expertise is trustworthy. Their intuitions have been developed in the framework of a good theory. Devitt gives the example of a group of art experts who all agreed that a Greek statue was fake. They were unable to say why, but were later proved right.
3. And sometimes we have no reliable theory. Sometimes you’re at the stage of “proto-science”, and you’re forced to use intuitions, you’ve got nothing better to go on. So we need intuitions when starting from scratch, but thought experiments are no substitute for real experiments, and we should eventually abandon reliance on intuitions as our theory develops.
***
In the Q&A session, Devitt notes:
* The ideology of linguists is that all they do is systematize the folk’s intuitions. But in actual practice they’re more inclined to rely on corpus evidence.
* “Innate” intuitions, intuitions you’re born with -- are these empirical judgments also? Are they an exception to the idea that every judgment is based on experience? Well, Devitt is skeptical that any such intuitions exist. But if they do exist, he’d want to explore the idea that they’re based on your ancestors’ experience.
* What about logical intuitions? Are these based on experience, and are they unreliable? -- Devitt would try to argue that, yes, they also are based on experience.
* Ethical intuitions? Shouldn’t philosophy try to systematize these? Isn’t that the bulk of what ethicists actually do? -- construct thought experiments, and try to find a theory that fits their moral intuitions? -- Well, Devitt wants to say that there are moral facts, and so there is no need here either, to take your task to be intuition systematization.
* Devitt also notes that a particular target is Oxford philosophy in the tradition of JL Austin, where philosophy is about clarifying concepts, and where one clarifies concepts by systematizing intuitions.
***
This article is licensed under the GNU Free Documentation License. It uses material from Michael Devitt's website, from the Das Keyboard website, and from the Wikipedia articles Immanuel Kant and Noam Chomsky.
| 138 |
| Vote |
Add To: 
subscribe to this blog
Comment by Damo
Even before I read it I had a gut feeling that would be good.
After reading it I had the evidence.
But what if I didn't like the blog?...
I think my head just exploded.
Personally as someone who has spent 90% of my career testing technical faults I am very reluctant to trust intuition.
Comment by Adrian
Philosophy Blog
There's this quote in Under Siege 2: the bad guy says, "Assumption is the mother of all fuck-ups. Go and make sure that Steven Segal is dead."
My sympathies, with some reservations, tend to lie with you, and with Devitt. I think that intuitions are inherently suspicious. They present to us as coming from nowhere, so were we Christian mystics we wouldn't know whether it was god speaking or the devil.
Incidentally, have you been clicking "Create a new blog" instead of "Create a new post"? You seem to have so many different blogs!
Comment by Harmony Rocks
head for threads
Harmony's Forum For You
Harmony's Forum For You
Comment by Adrian
Philosophy Blog
Well, I think he'd distinguish between good ideas and intuitions. And I think he'd reply that we should do without intuitions as much as possible, because they're unreliable. (You can use the spark to get your essay going; but you can't rely on the spark as evidence for a claim.)
But maybe the basic answer is "Yes". Yes, he does think all knowledge is a posteriori, so that would mean that all philosophical knowledge is empirical as well.
Whether the method is empirical is harder...
On the one hand, Devitt's own argument above presumably was not based on empirical research, but on logic and a good idea (he might be engaging in the very vice he condemns).
On the other hand, Devitt is suggesting that instead of trying to systematize intuitions form the armchair, philosophers should get their hands dirty and engage in empirical research as much as possible. And there is in fact an increasing trend towards just this. For instance, an ethics journal was recently created devoted to the empirical research of happiness.
Comment by Harmony Rocks
head for threads
Harmony's Forum For You
Harmony's Forum For You
Ok. Funk & Wagnall's 3rd definition of philosophy: the calm judgment and equable temper resulting from study of causes and laws; practical wisdom, fortitude, as in enduring reverses and suffering.
So maybe it could be stated that intuition begets experience begets intuition?
HR
Comment by Adrian
Philosophy Blog
Intuition begets experience: there is no such thing as raw sense data. Any information we receive is filtered through an already-existing framework. What one person and another person sees might be different.
Experience begets intuition: Devitt thinks all intuitions (beliefs that pop into your head from nowhere) are unconscious judgements your brain makes that are based on experience.
Comment by Kris
Quotable Quips
Banana and Mango
He says that intuition is no better than other types of knowledge - I think it is better, in terms being competent at something, like he said. In terms of actual knowledge, well, he seems to be a big fan of empirical knowledge, and I don't think that empirical knowledge leads to anything other than competence at something (eg, science, makes us competent at technology).