Unfortunately, I think what you’re clarifying isn’t what I was asking about :-). Let P be the proposition “as far as scientifically observable consequences go, evolution behaves as if it’s entirely natural and undesigned”. You and I agree that a theist can consistently believe P; call this thing that we believe Q. (Perhaps you also believe P, as I do, but that’s a separate question.) You made a remark about the RCC (which has spawned a discussion entirely out of proportion to the importance of that remark in anyone’s arguments, but no matter!) which I interpreted as saying that the RCC’s official position is P, whereas in fact perhaps you were saying that the RCC’s official position is Q (or perhaps the closely related Q’, which says that a good Catholic can consistently believe P).
So my question was: were you saying that the RCC’s position is (something like) P, or that the RCC’s position is (something like) Q?
Pr(...) = 1
I agree that HG can be read as saying faithful Catholics mustn’t take Pr(evolution)=1 but must take Pr(souls)=1 (where both “evolution” and “souls” are brief abbreviations for more complicated things, of course). I suppose what I was getting at is that saying “don’t take Pr(X)=1” effectively means quite different things depending on whether the community you’re addressing is in the habit of taking Pr(various things)=1 or scrupulously avoids it as e.g. LW tends to for good reason; and the fact that HG firmly endorses taking some probabilities to be 1 indicates that it’s in the former camp, which to me suggests that HG is saying not what an LWer would express by “don’t take Pr(evolution)=1″ but something more like “don’t treat evolution as definitely true in the same sort of way as you treat other ordinary things as definitely true”.
But it’s possible, as you say, that actually the position being sketched in HG would be, if written out with more care, something more like this: there are essential dogmas of RC faith, for which one must assign p=1; there are ordinary statements of fact, like (ha!) “the earth orbits the sun”, for which one would be ill-advised to assign p=1 but the RCC doesn’t take any particular position on that question; but for evolution the RCC specifically says not to take p=1 but leaves open the possibility of taking p=1-10^-20 or something.
I suspect these are questions that have no answers, in the following sense: senior RC clergy don’t generally think in terms that correspond so directly to the very quantitative probabilistic approach commonly taken here as to enable a clear distinction between p=1 and p=1-10^-20, etc., and any “translation” that makes them out to have been making statements in those quantitative probabilistic terms is liable to misrepresent their meaning. (E.g., p=1, as such, means that absolutely no possible evidence would change your mind, but I think Pius XII could probably have imagined possible happenings that would have convinced him that God doesn’t after all directly attach souls to human bodies.)
So my question was: were you saying that the RCC’s position is (something like) P, or that the RCC’s position is (something like) Q?
Something almost exactly like Q’.
But it’s possible, as you say, that actually the position being sketched in HG would be, if written out with more care, something more like this: there are essential dogmas of RC faith, for which one must assign p=1; there are ordinary statements of fact, like (ha!) “the earth orbits the sun”, for which one would be ill-advised to assign p=1 but the RCC doesn’t take any particular position on that question; but for evolution the RCC specifically says not to take p=1 but leaves open the possibility of taking p=1-10^-20 or something.
I get the impression they were hinting at something more like Pr(evolution)=0.9, which is a figure entirely unsupported by anything in the text and involves me taking a guess, but apart from that this is pretty much exactly how I read it, yes. (With the note that the only reason the RCC is specifically calling out evolution is because there’s been such a brouhaha over it from the protestant churches that staying silent on the matter would have been bad politics).
senior RC clergy don’t generally think in terms that correspond so directly to the very quantitative probabilistic approach commonly taken here as to enable a clear distinction between p=1 and p=1-10^-20, etc.
While there may very well be senior clergy who do think in such terms (I wouldn’t know, I’ve never met any seriously senior clergy) this is largely why I think that something like p=0.9 is probably closer to the intended reading. (Or p=0.95, or even p=0.99)
(E.g., p=1, as such, means that absolutely no possible evidence would change your mind, but I think Pius XII could probably have imagined possible happenings that would have convinced him that God doesn’t after all directly attach souls to human bodies.)
Even if so, I’m pretty sure that it wouldn’t be the official Vatican position that that probability can ever be anything less than one. (Politics, again; if the Pope ever admits to that p being less than one and someone runs a headline based on it...)
And if he does see such evidence, he has to consider the possibility that he is hallucinating, or being intentionally tricked by someone; if p is high enough, then it may very well be the case that any sufficiently convincing evidence will merely convince him to report to the nearest psychologist with complaints of extraordinarily detailed and persistent hallucinations.
OK; then many of my earlier comments in this thread (which were essentially arguing that the RCC’s position is very different from P) have been entirely not-to-the-point and have wasted everyone’s time. I repent in dust and ashes.
p=0.9
Yeah, this is roughly my reading too. (Maybe more like p=0.7 or something back in 1950 with Humani Generis.)
it wouldn’t be the official Vatican position that that probability can ever be anything less than one.
So much the worse for the Vatican, then. (But I think you’re probably right.)
The Church has no official statements about probability one way or another. There are certainly Catholics who think that the probability of their beliefs is one (which is insane), but typically they do not hold this by saying that no possible evidence would convince them. They say, “such and such would convince me that Catholicism is false, but such and such is absolutely impossible, and I am absolutely certain that it is absolutely impossible.”
On the other hand there are many far more reasonable Catholics who admit that their probability is less than one, admit that there is evidence that would convince them their beliefs are false, and admit that they might later observe the evidence. The Church has never said anything against such opinions (or against the first kind of opinion).
One person I know, whom pretty much everyone considers to be a devout and orthodox Catholic, told me that he would be happy with a probability of 30% (that is, he would be happy to believe with that probability, based on Pascal’s wager type reasoning). I suspect that in practice his personal probability is around 50%.
The discussion about probabilities was all concerning the following specific question: How should we interpret the language in, e.g., Humani Generis about how faithful Catholics are required not to hold that some things are definitely true but the RCC teaches that some other things are definitely true? For sure, no translation into LW-style probabilityese is going to reproduce the meaning exactly, but one might reasonably hope to approximate what the RC documents say in terms that make some kind of sense to rationalists.
I think you need to understand that in terms of doxastic voluntarism. The Church holds that faith is voluntary, so that people can reasonably be praised or blamed for what they believe. They may not say it explicitly about other kinds of belief, but if they are right about religious beliefs, it would probably also be true that accepting or rejecting the theory of evolution (and any other similar thing) is also voluntary.
Given that account of belief, saying “you are obliged to say that this is definitely true and that this other thing is at least not definitely true,” is a statement about the choices you should make. You should choose to say (according to Pius XII), “Christianity comes from God,” and “Christianity definitely comes from God,” but not (according to Pius XII), “The theory of evolution applied to human beings is definitely true.”
Because it’s a question of choices, probability is not really relevant one way or another (except in the sense that the probability that something is true might be one reason that should affect whether or not you say something). That is why I gave the example of someone who estimates the chance of his beliefs being right, based on the evidence, to be about 50%. But despite that he chooses to say “this is simply true,” probably based on various moral considerations. Kind of like you might want to believe your brother about something rather than saying he is lying, not because of strong evidence for that, but because it’s hurtful to say “you’re lying.”
If this account is correct, the reason you can’t translate statements like that into statements resembling claims about probability is because that is simply not what they are about.
Rationalists say many things which have similar moral implications, and which don’t ultimately make a lot of sense apart from a similar doxastic voluntarism (e.g. “you ought to update on evidence,” which implies that you can and should choose to do so), but most people don’t accept an account like this explicitly. If you don’t accept doxastic voluntarism, you should simply say that the Catholic Church’s statements about such things are presuming an incorrect theory of belief, and just don’t make any sense apart from that theory.
(I agree with doxastic voluntarism, so such statements do make sense to me, whatever I happen to think about them on an object level.)
Rationalists say many things which have similar moral implications [...] e.g. “you ought to update on evidence”
When I say “you ought to update on evidence” I’m fairly sure I am neither endorsing doxastic voluntarism nor making a claim about the morality of updating (or not) on evidence. Rather, I mean that if you update appropriately on evidence then your beliefs will, over time, tend to grow more accurate compared with the beliefs you would hold if you didn’t update appropriately on evidence, and that this is likely to benefit whatever goals you may have. (I might add that it benefits other people for your beliefs to be accurate, so there is some moral import to whether you update on evidence, but that’s not what I would mean by “ought”.) Is it a pointless thing to say if we don’t get to choose what we believe? Not if we have more choice about what overall strategy to use when adjusting our beliefs than we have about individual beliefs, which I think may well be the case.
(As to the actual question of doxastic voluntarism, I think it’s clear that it comes in degrees and that the truth isn’t right at either extreme. I don’t believe either of us could, right now, decide to believe that grass is pink and forthwith start doing so; but I’m pretty sure either of us could incline ourselves more toward believing (say) that Charles Dickens was born in 1834 simply by repeating “Charles Dickens was born in 1834” in a confident tone of voice fifty times. Choose another example if you happen already to have a confident opinion about whether Dickens was born in 1834.)
I am fairly sure that many rationalists do in fact make claims about updating on evidence with moral implications, even if you do not intend those implications yourself. But in any case, even your response implies a certain degree of voluntariness, if it is possible to adopt an overall strategy of adjusting our beliefs; if it were totally involuntary, we could not affect the strategy (and in fact you agree in the second part that it is not totally involuntary.)
I agree that in the way people ordinarily mean it, I could not start to believe that grass is pink. But I also could not go and kill myself right now. That doesn’t make not killing myself involuntary, since the reason is that I think it would be bad to kill myself. The case of the grass might be different, and it might be impossible to start to believe that in a stronger sense. In other discussion of this issue someone compared it to holding your hand in a fire until it is burned off; that might well be a physical impossibility, not just a question of thinking that it is bad.
But even in the case of holding your hand in a fire, I would see that as a certain kind of desire (to pull your hand away), even if it is one that we cannot resist with our conscious desires, and the case of belief seems pretty similar. This might just be a question of how much you are willing to strain an analogy.
It’s not as if “belief” is a name for one objective thing in the world which is either there or not. There are a whole bunch of things, words, actions, thoughts, and feelings, and we call various patterns of these things a “belief.” Some of these things are voluntary and some are not. Simply for consistency I choose to call the voluntary parts of that pattern “belief” and exclude the other parts, at least when there is competition between them. According to this way of speaking, I could choose to believe that Dickens was born in 1834, even right now, if I had a motive to do so. But that would not affect the involuntary parts of my assessment, and you might prefer to call it something else (e.g. “belief in belief”).
Unfortunately, I think what you’re clarifying isn’t what I was asking about :-). Let P be the proposition “as far as scientifically observable consequences go, evolution behaves as if it’s entirely natural and undesigned”. You and I agree that a theist can consistently believe P; call this thing that we believe Q. (Perhaps you also believe P, as I do, but that’s a separate question.) You made a remark about the RCC (which has spawned a discussion entirely out of proportion to the importance of that remark in anyone’s arguments, but no matter!) which I interpreted as saying that the RCC’s official position is P, whereas in fact perhaps you were saying that the RCC’s official position is Q (or perhaps the closely related Q’, which says that a good Catholic can consistently believe P).
So my question was: were you saying that the RCC’s position is (something like) P, or that the RCC’s position is (something like) Q?
I agree that HG can be read as saying faithful Catholics mustn’t take Pr(evolution)=1 but must take Pr(souls)=1 (where both “evolution” and “souls” are brief abbreviations for more complicated things, of course). I suppose what I was getting at is that saying “don’t take Pr(X)=1” effectively means quite different things depending on whether the community you’re addressing is in the habit of taking Pr(various things)=1 or scrupulously avoids it as e.g. LW tends to for good reason; and the fact that HG firmly endorses taking some probabilities to be 1 indicates that it’s in the former camp, which to me suggests that HG is saying not what an LWer would express by “don’t take Pr(evolution)=1″ but something more like “don’t treat evolution as definitely true in the same sort of way as you treat other ordinary things as definitely true”.
But it’s possible, as you say, that actually the position being sketched in HG would be, if written out with more care, something more like this: there are essential dogmas of RC faith, for which one must assign p=1; there are ordinary statements of fact, like (ha!) “the earth orbits the sun”, for which one would be ill-advised to assign p=1 but the RCC doesn’t take any particular position on that question; but for evolution the RCC specifically says not to take p=1 but leaves open the possibility of taking p=1-10^-20 or something.
I suspect these are questions that have no answers, in the following sense: senior RC clergy don’t generally think in terms that correspond so directly to the very quantitative probabilistic approach commonly taken here as to enable a clear distinction between p=1 and p=1-10^-20, etc., and any “translation” that makes them out to have been making statements in those quantitative probabilistic terms is liable to misrepresent their meaning. (E.g., p=1, as such, means that absolutely no possible evidence would change your mind, but I think Pius XII could probably have imagined possible happenings that would have convinced him that God doesn’t after all directly attach souls to human bodies.)
Something almost exactly like Q’.
I get the impression they were hinting at something more like Pr(evolution)=0.9, which is a figure entirely unsupported by anything in the text and involves me taking a guess, but apart from that this is pretty much exactly how I read it, yes. (With the note that the only reason the RCC is specifically calling out evolution is because there’s been such a brouhaha over it from the protestant churches that staying silent on the matter would have been bad politics).
While there may very well be senior clergy who do think in such terms (I wouldn’t know, I’ve never met any seriously senior clergy) this is largely why I think that something like p=0.9 is probably closer to the intended reading. (Or p=0.95, or even p=0.99)
Even if so, I’m pretty sure that it wouldn’t be the official Vatican position that that probability can ever be anything less than one. (Politics, again; if the Pope ever admits to that p being less than one and someone runs a headline based on it...)
And if he does see such evidence, he has to consider the possibility that he is hallucinating, or being intentionally tricked by someone; if p is high enough, then it may very well be the case that any sufficiently convincing evidence will merely convince him to report to the nearest psychologist with complaints of extraordinarily detailed and persistent hallucinations.
OK; then many of my earlier comments in this thread (which were essentially arguing that the RCC’s position is very different from P) have been entirely not-to-the-point and have wasted everyone’s time. I repent in dust and ashes.
Yeah, this is roughly my reading too. (Maybe more like p=0.7 or something back in 1950 with Humani Generis.)
So much the worse for the Vatican, then. (But I think you’re probably right.)
Something in that general order of magnitude is probably more-or-less right.
...I think we’ve pretty much come to agreement on these points, then.
The Church has no official statements about probability one way or another. There are certainly Catholics who think that the probability of their beliefs is one (which is insane), but typically they do not hold this by saying that no possible evidence would convince them. They say, “such and such would convince me that Catholicism is false, but such and such is absolutely impossible, and I am absolutely certain that it is absolutely impossible.”
On the other hand there are many far more reasonable Catholics who admit that their probability is less than one, admit that there is evidence that would convince them their beliefs are false, and admit that they might later observe the evidence. The Church has never said anything against such opinions (or against the first kind of opinion).
One person I know, whom pretty much everyone considers to be a devout and orthodox Catholic, told me that he would be happy with a probability of 30% (that is, he would be happy to believe with that probability, based on Pascal’s wager type reasoning). I suspect that in practice his personal probability is around 50%.
The discussion about probabilities was all concerning the following specific question: How should we interpret the language in, e.g., Humani Generis about how faithful Catholics are required not to hold that some things are definitely true but the RCC teaches that some other things are definitely true? For sure, no translation into LW-style probabilityese is going to reproduce the meaning exactly, but one might reasonably hope to approximate what the RC documents say in terms that make some kind of sense to rationalists.
I think you need to understand that in terms of doxastic voluntarism. The Church holds that faith is voluntary, so that people can reasonably be praised or blamed for what they believe. They may not say it explicitly about other kinds of belief, but if they are right about religious beliefs, it would probably also be true that accepting or rejecting the theory of evolution (and any other similar thing) is also voluntary.
Given that account of belief, saying “you are obliged to say that this is definitely true and that this other thing is at least not definitely true,” is a statement about the choices you should make. You should choose to say (according to Pius XII), “Christianity comes from God,” and “Christianity definitely comes from God,” but not (according to Pius XII), “The theory of evolution applied to human beings is definitely true.”
Because it’s a question of choices, probability is not really relevant one way or another (except in the sense that the probability that something is true might be one reason that should affect whether or not you say something). That is why I gave the example of someone who estimates the chance of his beliefs being right, based on the evidence, to be about 50%. But despite that he chooses to say “this is simply true,” probably based on various moral considerations. Kind of like you might want to believe your brother about something rather than saying he is lying, not because of strong evidence for that, but because it’s hurtful to say “you’re lying.”
If this account is correct, the reason you can’t translate statements like that into statements resembling claims about probability is because that is simply not what they are about.
Rationalists say many things which have similar moral implications, and which don’t ultimately make a lot of sense apart from a similar doxastic voluntarism (e.g. “you ought to update on evidence,” which implies that you can and should choose to do so), but most people don’t accept an account like this explicitly. If you don’t accept doxastic voluntarism, you should simply say that the Catholic Church’s statements about such things are presuming an incorrect theory of belief, and just don’t make any sense apart from that theory.
(I agree with doxastic voluntarism, so such statements do make sense to me, whatever I happen to think about them on an object level.)
When I say “you ought to update on evidence” I’m fairly sure I am neither endorsing doxastic voluntarism nor making a claim about the morality of updating (or not) on evidence. Rather, I mean that if you update appropriately on evidence then your beliefs will, over time, tend to grow more accurate compared with the beliefs you would hold if you didn’t update appropriately on evidence, and that this is likely to benefit whatever goals you may have. (I might add that it benefits other people for your beliefs to be accurate, so there is some moral import to whether you update on evidence, but that’s not what I would mean by “ought”.) Is it a pointless thing to say if we don’t get to choose what we believe? Not if we have more choice about what overall strategy to use when adjusting our beliefs than we have about individual beliefs, which I think may well be the case.
(As to the actual question of doxastic voluntarism, I think it’s clear that it comes in degrees and that the truth isn’t right at either extreme. I don’t believe either of us could, right now, decide to believe that grass is pink and forthwith start doing so; but I’m pretty sure either of us could incline ourselves more toward believing (say) that Charles Dickens was born in 1834 simply by repeating “Charles Dickens was born in 1834” in a confident tone of voice fifty times. Choose another example if you happen already to have a confident opinion about whether Dickens was born in 1834.)
I am fairly sure that many rationalists do in fact make claims about updating on evidence with moral implications, even if you do not intend those implications yourself. But in any case, even your response implies a certain degree of voluntariness, if it is possible to adopt an overall strategy of adjusting our beliefs; if it were totally involuntary, we could not affect the strategy (and in fact you agree in the second part that it is not totally involuntary.)
I agree that in the way people ordinarily mean it, I could not start to believe that grass is pink. But I also could not go and kill myself right now. That doesn’t make not killing myself involuntary, since the reason is that I think it would be bad to kill myself. The case of the grass might be different, and it might be impossible to start to believe that in a stronger sense. In other discussion of this issue someone compared it to holding your hand in a fire until it is burned off; that might well be a physical impossibility, not just a question of thinking that it is bad.
But even in the case of holding your hand in a fire, I would see that as a certain kind of desire (to pull your hand away), even if it is one that we cannot resist with our conscious desires, and the case of belief seems pretty similar. This might just be a question of how much you are willing to strain an analogy.
It’s not as if “belief” is a name for one objective thing in the world which is either there or not. There are a whole bunch of things, words, actions, thoughts, and feelings, and we call various patterns of these things a “belief.” Some of these things are voluntary and some are not. Simply for consistency I choose to call the voluntary parts of that pattern “belief” and exclude the other parts, at least when there is competition between them. According to this way of speaking, I could choose to believe that Dickens was born in 1834, even right now, if I had a motive to do so. But that would not affect the involuntary parts of my assessment, and you might prefer to call it something else (e.g. “belief in belief”).