The debate over intuitions is one of the hottest in philosophy today
But it—at least the “debate over intuitions” that I’m most familiar with—isn’t about whether intuitions are reliable, but rather over whether the critics have accurately specified the role they play in traditional philosophical methodology. That is, the standard response to experimentalist critics (at least, in my corner of philosophy) is not to argue that intuitions are “reliable evidence”, but rather to deny that we are using them as evidence at all. On this view, what we appeal to as evidence is not the psychological fact of my having an intuition, but rather the propositional content being judged.
The purpose of thought experiments, on this view, is to enable one to grasp new evidence (namely, the proposition in question) that they hadn’t considered before. Of course, this isn’t a “neutral” methodology because only those who intuit the true proposition thereby gain genuine evidence. But the foolishness of such a “neutrality” constraint (and the associated “psychological” view of evidence) is one of the major lessons of contemporary epistemology (see, esp., Williamson).
A quick review (for the benefit of others): Bugmaster asked: “Is there really any kind of a serious debate in modern philosophy circles regarding whether or not our personal intuitions can be generally trusted?” I replied: “Yes… the debate over intuitions is one of the hottest in philosophy today.” But Richard is right to say that most of the philosophical debate about intuitions “isn’t about whether intuitions are reliable, but rather over whether the critics have accurately specified the role they play in traditional philosophical methodology.” So, I apologize for my sloppy wording.
Now, a few words on intuitionist methodology. When I read the defenders of intuitionist methodology, I’m reminded of something John Doris said in my interview with him (slight paraphrase for clarity and succinctness; see the exact quote at the bottom of the transcript):
If experimentalists say that something is a mistake, then no one will admit they said it. And I’m no different: “Oh, is that false? Then I disagree. That’s not what I really meant.”
When experimentalists pointed out that our brains don’t store concepts as necessary and sufficient conditions, many philosophers rushed to say that philosophers had never been assuming this in the first place. But clearly, many philosophers were making such false assumptions about how concepts worked, since the “classical” view of concepts — concepts as mental representations captured by necessary and sufficient conditions — held sway for quite some time, even after Wittgenstein (1953). (For a review, see Murphy 2004.)
Or, given that experimentalists have raised worries about using intuitions as evidence in general, Ichikawa (forthcoming) now rushes to say that philosophers generally don’t rely on intuitions in a “central” way. (To narrow our discussion, I’ll focus on this, the first article you sent me.) What does Ichikawa mean by this? He distinguishes three metaphilosophical claims:
Intuited contents are (often) taken as important evidence/reasons/data/input in armchair philosophy.
Intuited contents are (often) taken as important evidence/reasons/data/input in armchair philosophy because they are intuited.
Intuition states, or facts about intuition states, are (often) taken as important evidence/reasons/data/input in armchair philosophy.
As far as I can tell, Ichikawa wants to argue that (1) represents philosophical practice better than (2) or (3), and that (1) is not particularly undermined by experimentalist critiques. Have I got that right? (I’ll hold my reply until I hear whether you agree with my interpretation. I found Ichikawa to be unclear on this, though not as unclear as Yudkowsky or Muehlhauser often are in their philosophical writings.)
Yes, that’s the idea. I mean, (2) is plausibly true if the “because” is meant in a purely causal, rather than rationalizing, sense. But we don’t take the fact that we stand in a certain psychological relation to this content (i.e., intuiting it) to play any essential justifying role.
Thanks for following up on this issue! I’m looking forward to hearing the rest of your thoughts.
In that case, I struggle to see why the “defeater critique” wouldn’t seriously undermine practice (1) in most cases. Philosophers can’t simply assume intuited contents p and then move from p to q. We want to know how likely p is to be true, and if our primary reason for thinking p is true is some unreliable cognitive algorithm (rather than, say, hard scientific data or a mathematical proof), then we are left without much reason to be confident that p is true.
Suppose a theist says he knows by Holy Spirit Communication (HSC) that Jesus is magic. An atheist replies, “HSC is not a reliable method. See all this experimental data on people making judgments based on the deliverances of (what they claim is) HSC.” The theist then says, “No, I’m not arguing from the HSC mental state to the conclusion that Jesus is magic. I’m arguing from the HSC contents (that is, from proposition p) to the conclusion that Jesus is magic.”
The atheist would be unimpressed, and correctly so.
In the case you describe, the “HSC content” is just that Jesus is magic. So there’s no argument being offered at all. Now, if they offer an actual argument, from some other p to the conclusion that Jesus is magic, then we can assess this argument like any other. How the arguer came to believe the original premise p is not particularly relevant. What you call the “defeater critique”, I call the genetic fallacy.
It’s true that an interlocutor is never going to be particularly moved by an argument that starts from premises he doesn’t accept. Such is life.
The more interesting question is whether the arguer herself should be led to abandon her intuited judgments. But unless you offer some positive evidence for an alternative rational credence to place in p, it’s not clear that a “debunking” explanation of her current level of credence should, by itself, make any difference.
Think of intuitied judgments as priors. Someone might say, “There’s no special reason to think that your priors are well-calibrated.” And that may be true, but it doesn’t change what our priors are. We can’t start from anywhere but where we start.
What you call the “defeater critique”, I call the genetic fallacy.
Thinking of things in terms of informal fallacies like the genetic fallacy throws away information. From a Bayesian viewpoint, the source of one’s belief is relevant to its likelihood of being true.
The more interesting question is whether the arguer herself should be led to abandon her intuited judgments. But unless you offer some positive evidence for an alternative rational credence to place in p, it’s not clear that a “debunking” explanation of her current level of credence should, by itself, make any difference.
Right; I mostly complain about arguments made solely from intuited contents when the claims are given with far more confidence than can be justified by the demonstrated reliability of human intuitions in that domain.
Thanks. I’m going to be extremely busy for the next few weeks but I will make sure to get back to you on this (and reply to your comment, so you get a notification) at a later time.
But it—at least the “debate over intuitions” that I’m most familiar with—isn’t about whether intuitions are reliable, but rather over whether the critics have accurately specified the role they play in traditional philosophical methodology. That is, the standard response to experimentalist critics (at least, in my corner of philosophy) is not to argue that intuitions are “reliable evidence”, but rather to deny that we are using them as evidence at all. On this view, what we appeal to as evidence is not the psychological fact of my having an intuition, but rather the propositional content being judged.
The purpose of thought experiments, on this view, is to enable one to grasp new evidence (namely, the proposition in question) that they hadn’t considered before. Of course, this isn’t a “neutral” methodology because only those who intuit the true proposition thereby gain genuine evidence. But the foolishness of such a “neutrality” constraint (and the associated “psychological” view of evidence) is one of the major lessons of contemporary epistemology (see, esp., Williamson).
A quick review (for the benefit of others): Bugmaster asked: “Is there really any kind of a serious debate in modern philosophy circles regarding whether or not our personal intuitions can be generally trusted?” I replied: “Yes… the debate over intuitions is one of the hottest in philosophy today.” But Richard is right to say that most of the philosophical debate about intuitions “isn’t about whether intuitions are reliable, but rather over whether the critics have accurately specified the role they play in traditional philosophical methodology.” So, I apologize for my sloppy wording.
Now, a few words on intuitionist methodology. When I read the defenders of intuitionist methodology, I’m reminded of something John Doris said in my interview with him (slight paraphrase for clarity and succinctness; see the exact quote at the bottom of the transcript):
When experimentalists pointed out that our brains don’t store concepts as necessary and sufficient conditions, many philosophers rushed to say that philosophers had never been assuming this in the first place. But clearly, many philosophers were making such false assumptions about how concepts worked, since the “classical” view of concepts — concepts as mental representations captured by necessary and sufficient conditions — held sway for quite some time, even after Wittgenstein (1953). (For a review, see Murphy 2004.)
Or, given that experimentalists have raised worries about using intuitions as evidence in general, Ichikawa (forthcoming) now rushes to say that philosophers generally don’t rely on intuitions in a “central” way. (To narrow our discussion, I’ll focus on this, the first article you sent me.) What does Ichikawa mean by this? He distinguishes three metaphilosophical claims:
Intuited contents are (often) taken as important evidence/reasons/data/input in armchair philosophy.
Intuited contents are (often) taken as important evidence/reasons/data/input in armchair philosophy because they are intuited.
Intuition states, or facts about intuition states, are (often) taken as important evidence/reasons/data/input in armchair philosophy.
As far as I can tell, Ichikawa wants to argue that (1) represents philosophical practice better than (2) or (3), and that (1) is not particularly undermined by experimentalist critiques. Have I got that right? (I’ll hold my reply until I hear whether you agree with my interpretation. I found Ichikawa to be unclear on this, though not as unclear as Yudkowsky or Muehlhauser often are in their philosophical writings.)
Yes, that’s the idea. I mean, (2) is plausibly true if the “because” is meant in a purely causal, rather than rationalizing, sense. But we don’t take the fact that we stand in a certain psychological relation to this content (i.e., intuiting it) to play any essential justifying role.
Thanks for following up on this issue! I’m looking forward to hearing the rest of your thoughts.
In that case, I struggle to see why the “defeater critique” wouldn’t seriously undermine practice (1) in most cases. Philosophers can’t simply assume intuited contents p and then move from p to q. We want to know how likely p is to be true, and if our primary reason for thinking p is true is some unreliable cognitive algorithm (rather than, say, hard scientific data or a mathematical proof), then we are left without much reason to be confident that p is true.
Suppose a theist says he knows by Holy Spirit Communication (HSC) that Jesus is magic. An atheist replies, “HSC is not a reliable method. See all this experimental data on people making judgments based on the deliverances of (what they claim is) HSC.” The theist then says, “No, I’m not arguing from the HSC mental state to the conclusion that Jesus is magic. I’m arguing from the HSC contents (that is, from proposition p) to the conclusion that Jesus is magic.”
The atheist would be unimpressed, and correctly so.
In the case you describe, the “HSC content” is just that Jesus is magic. So there’s no argument being offered at all. Now, if they offer an actual argument, from some other p to the conclusion that Jesus is magic, then we can assess this argument like any other. How the arguer came to believe the original premise p is not particularly relevant. What you call the “defeater critique”, I call the genetic fallacy.
It’s true that an interlocutor is never going to be particularly moved by an argument that starts from premises he doesn’t accept. Such is life.
The more interesting question is whether the arguer herself should be led to abandon her intuited judgments. But unless you offer some positive evidence for an alternative rational credence to place in p, it’s not clear that a “debunking” explanation of her current level of credence should, by itself, make any difference.
Think of intuitied judgments as priors. Someone might say, “There’s no special reason to think that your priors are well-calibrated.” And that may be true, but it doesn’t change what our priors are. We can’t start from anywhere but where we start.
Thinking of things in terms of informal fallacies like the genetic fallacy throws away information. From a Bayesian viewpoint, the source of one’s belief is relevant to its likelihood of being true.
(Edit 9/2/13: A good example of this is here.)
Right; I mostly complain about arguments made solely from intuited contents when the claims are given with far more confidence than can be justified by the demonstrated reliability of human intuitions in that domain.
Can you cite a specific paper on book chapter which makes the kind of argument you’re suggesting here?
Jonathan Ichikawa, ‘Who Needs Intuitions’
Elizabeth Harman, ’Is it Reasonable to “Rely on Intuitions” in Ethics?
Timothy Williamson, ‘Evidence in Philosophy’, chp 7 of The Philosophy of Philosophy.
Thanks. I’m going to be extremely busy for the next few weeks but I will make sure to get back to you on this (and reply to your comment, so you get a notification) at a later time.