A priori beliefs are beliefs without evidence. If a belief doesn’t respond to further evidence, this is also a property of a priori beliefs. Normal beliefs can behave like this, and be generally accepted. The concept of “faith”, to make it non-vacuous, needs to be opposed to normal human cognition (preferences). But this makes it similar to “insanity”, which is unlikely what the people who advocate the practice mean. Their concept of “faith” isn’t obvious.
A priori beliefs are beliefs without evidence. If a belief doesn’t respond to further evidence, this is also a property of a priori beliefs. Normal beliefs can behave like this, and be generally accepted.
I think the terms “a priori belief” and “faith” refer to the same concept. Can you provide an example of a “normal, generally accepted” a priori belief?
The concept of “faith”, to make it non-vacuous, needs to be opposed to normal human cognition (preferences).
I don’t see how this follows. What do preferences have to do with it?
Well, take the statement that there are green-skinned, blue-eyed, humanoid aliens living on a planet orbiting Betelgeuse. Since this is a very specific statement, it’s a priori very unlikely to be true, so that the belief that there are no such aliens is rational despite the lack of evidence for it; it’s not a faith-based belief.
Perhaps a better definition of faith would be, “intentionally self-deceiving belief”.
It’s certainly a more condescending definition, at least.
Has Occam’s Razor been semantically cleaved from the notion of “faith” in a convincing way here or elsewhere? How does your a priori reasoning differ from “faith in simpler explanations”?
The greater the specificity of a concept is, the less plausible it must be. For example, if I make two guesses, first that you have a sibling, and second that this hypothetical sibling is female, the second, more specific guess is necessarily less likely to be true than the first one.
Eliezer has written some stuff about this, but if you’re interested in a really rigorous argument, I recommend Paul Almond’s Occam’s Razor series (nine articles in total):
Thanks for the link. Some of Almond’s other stuff was already on my “to read” list, it looks like I’ll be sinking a weekend on his site some time soon.
I think I see the distinction you’re making… the things inherent in the structure and relationships of a set of concepts (like specificity) effectively function as a priori truths, even though the concepts may be “about” empirical matters. This is clearly different from “faith”, which has a more speculative nature. My vague intuition that Occam’s Razor is somehow “like faith” still isn’t discharged, but perhaps Almond will bludgeon that out of me with his series.
I think the terms “a priori belief” and “faith” refer to the same concept. Can you provide an example of a “normal, generally accepted” a priori belief?
I hope you mean “belief that doesn’t change in response to evidence”, because a priori beliefs is exactly what determines what you do with evidence and what you can come to believe later. For an example of unchanging belief: if you toss a normal coin 10 times and it lands “heads” each time, your belief about the probability of it coming “heads” should remain almost the same.
Epistemic beliefs are one side of the preference specification, the other in this formalism being utility function. If the behavior of “faith” beliefs is different from normal beliefs, it follows that they act contrary to the framework that expresses human preferences as probability+utility, breaking human preference as a result and leading to behavior that is insane (i.e. incorrect) according to human preferences (ethics/morality).
A priori beliefs are beliefs without evidence. If a belief doesn’t respond to further evidence, this is also a property of a priori beliefs. Normal beliefs can behave like this, and be generally accepted. The concept of “faith”, to make it non-vacuous, needs to be opposed to normal human cognition (preferences). But this makes it similar to “insanity”, which is unlikely what the people who advocate the practice mean. Their concept of “faith” isn’t obvious.
I think the terms “a priori belief” and “faith” refer to the same concept. Can you provide an example of a “normal, generally accepted” a priori belief?
I don’t see how this follows. What do preferences have to do with it?
Well, take the statement that there are green-skinned, blue-eyed, humanoid aliens living on a planet orbiting Betelgeuse. Since this is a very specific statement, it’s a priori very unlikely to be true, so that the belief that there are no such aliens is rational despite the lack of evidence for it; it’s not a faith-based belief.
Perhaps a better definition of faith would be, “intentionally self-deceiving belief”.
It’s certainly a more condescending definition, at least.
Has Occam’s Razor been semantically cleaved from the notion of “faith” in a convincing way here or elsewhere? How does your a priori reasoning differ from “faith in simpler explanations”?
The greater the specificity of a concept is, the less plausible it must be. For example, if I make two guesses, first that you have a sibling, and second that this hypothetical sibling is female, the second, more specific guess is necessarily less likely to be true than the first one.
Eliezer has written some stuff about this, but if you’re interested in a really rigorous argument, I recommend Paul Almond’s Occam’s Razor series (nine articles in total):
http://www.paul-almond.com/
Thanks for the link. Some of Almond’s other stuff was already on my “to read” list, it looks like I’ll be sinking a weekend on his site some time soon.
I think I see the distinction you’re making… the things inherent in the structure and relationships of a set of concepts (like specificity) effectively function as a priori truths, even though the concepts may be “about” empirical matters. This is clearly different from “faith”, which has a more speculative nature. My vague intuition that Occam’s Razor is somehow “like faith” still isn’t discharged, but perhaps Almond will bludgeon that out of me with his series.
I hope you mean “belief that doesn’t change in response to evidence”, because a priori beliefs is exactly what determines what you do with evidence and what you can come to believe later. For an example of unchanging belief: if you toss a normal coin 10 times and it lands “heads” each time, your belief about the probability of it coming “heads” should remain almost the same.
Epistemic beliefs are one side of the preference specification, the other in this formalism being utility function. If the behavior of “faith” beliefs is different from normal beliefs, it follows that they act contrary to the framework that expresses human preferences as probability+utility, breaking human preference as a result and leading to behavior that is insane (i.e. incorrect) according to human preferences (ethics/morality).