(no subject)
So, this is tangential to what I've been studying, but I've been thinking about it recently. If anyone has any thoughts or reading suggestions that would be cool.
So, I've been thinking that justification might not exhaust rationality. That is, it looks like it can be rational to hold some beliefs which are not justified. So, what if probability (say) tells the story of justification, rather than rational belief? If so, then the humean problem of induction would apply to justification, but not necessarily to rational belief. If so, and if some of these non-justified but rational inference rules or beliefs support believing induction is a guide to truth, then accepting inductive justification would be rational.
Plus, it seems that there are certain types of questions that a probability-based story for justification would leave us in the dark about: those questions the answers to which we don't even have a probability distribution for. That is, those questions for which we have no good evidence. If our options are believing p or notp, and there is no evidence for or against p, then (for instance) a dutch book cannot be constructed against you based upon your believing p or notp.
So, the above means that it (might be) rational to hold non-justified beliefs. Further, if rationality is taken as something that comes in degrees (as justification is), and as a property of beliefs, then it could even be the case that it is “more rational” to believe p or that it is “more rational” to believe not p. In any case, we would have beliefs in between "arational" and "justified."
What would make a belief rational, or more rational, but would not necessarily make it justified? That is something I have no idea about, but the answer would probably be a pragmatic and/or abductive one. If so, then it seems we would have at best abductive/pragmatic reasons for belief in metaphysical things, for instance (given we don't seem to have much evidence either way for the actual truth or falsity of metaphysical claims). If so, and if our non-justified, rational reasons for belief do give us reasons for believing induction to be somewhat veridical, then while we might not be able to independently defend accepting inductive justification, we could still regard it as a (strongly) rational process. If those non-justified rational reasons also gave us a reason to prefer “projectible” predicates over non-projectible predicates (or whatever), it might also help us out with the problem of the apparent inductive confirmation of gruesome things. And both of those might be a step up from where skepticism had left us before.
Now, when philosophyjeff suggested I post this here, he said he was worried that I was grasping at "prudential justification." This would be a non-epistemic rationale for believing certain propositions because they improve our chances of accomplishing a certain task. That would not be something I would want to do, although given the use of the word "pragmatic" it certainly would be a warranted suspicion. Also, I certainly didn't explicate what I thought those abductive/pragmatic reasons might be.
What I had in mind, instead, were simply those reasons-for-belief (if any) that did not revolve around our subjective probability assignments for the propositions expressed by our beliefs. Predictive success would be out, for instance, as would forms of "inference to the best explanation" that was spelled out in terms of the "likelihood" of the best explanation being higher than the "likelihood" for other explanations. I am thinking of epistemic virtues more generally: simplicity, explanatory power, and maybe something like "similarity to other rational beliefs." It is obviously quite a task to precisely describe those virtues. Further, as I said, I don't have any idea for how one could argue that these epistemic virtues give us a prima facia reason to think that a belief that had those virtues was "more rational" than one that didn't (and i'm not convinced, necessarily, that they do). In any case, these epistemic virtues shouldn't give us a reason to believe (at least solely) becuase they make it easier to accomplish some end: rather, I am wondering if these epistemic virtues by themselves are able to give us warrant for accepting some belief (that possesses those virtues) as rational.
My real question, though, is that assuming we can do this - assuming abduction can give us a rational reason to believe - could we then be in a position to judge inductive justification to be rational, and maybe to consider inferring projectible predicates rather than gruesome predicates to be rational (and, more interestingly, whether we could judge those things to be "more rational" then their negations). As a sketch of a sketch: it certainly seems simpler to posit a boring external world to positing any of the skeptical scenarios, and something like a uniformity of nature condition seems like a better explanation for our past predictive success and for our observations of constant conjunctions than does the alternate explanation that basically boils down to "stuff happens, and that is all you can say about it." If the simplicity or explanatory power of a belief then gives us some reason to think those beliefs are rational (or more rational than their alternatives), then it is rational (or more rational) to accept induction. At least maybe.
So, thoughts? Like I said, this is tangential to what I am actually reading and trying to write about, but I do think it is interesting: any recommendations for further reading on basically this topic?
So, I've been thinking that justification might not exhaust rationality. That is, it looks like it can be rational to hold some beliefs which are not justified. So, what if probability (say) tells the story of justification, rather than rational belief? If so, then the humean problem of induction would apply to justification, but not necessarily to rational belief. If so, and if some of these non-justified but rational inference rules or beliefs support believing induction is a guide to truth, then accepting inductive justification would be rational.
Plus, it seems that there are certain types of questions that a probability-based story for justification would leave us in the dark about: those questions the answers to which we don't even have a probability distribution for. That is, those questions for which we have no good evidence. If our options are believing p or notp, and there is no evidence for or against p, then (for instance) a dutch book cannot be constructed against you based upon your believing p or notp.
So, the above means that it (might be) rational to hold non-justified beliefs. Further, if rationality is taken as something that comes in degrees (as justification is), and as a property of beliefs, then it could even be the case that it is “more rational” to believe p or that it is “more rational” to believe not p. In any case, we would have beliefs in between "arational" and "justified."
What would make a belief rational, or more rational, but would not necessarily make it justified? That is something I have no idea about, but the answer would probably be a pragmatic and/or abductive one. If so, then it seems we would have at best abductive/pragmatic reasons for belief in metaphysical things, for instance (given we don't seem to have much evidence either way for the actual truth or falsity of metaphysical claims). If so, and if our non-justified, rational reasons for belief do give us reasons for believing induction to be somewhat veridical, then while we might not be able to independently defend accepting inductive justification, we could still regard it as a (strongly) rational process. If those non-justified rational reasons also gave us a reason to prefer “projectible” predicates over non-projectible predicates (or whatever), it might also help us out with the problem of the apparent inductive confirmation of gruesome things. And both of those might be a step up from where skepticism had left us before.
Now, when philosophyjeff suggested I post this here, he said he was worried that I was grasping at "prudential justification." This would be a non-epistemic rationale for believing certain propositions because they improve our chances of accomplishing a certain task. That would not be something I would want to do, although given the use of the word "pragmatic" it certainly would be a warranted suspicion. Also, I certainly didn't explicate what I thought those abductive/pragmatic reasons might be.
What I had in mind, instead, were simply those reasons-for-belief (if any) that did not revolve around our subjective probability assignments for the propositions expressed by our beliefs. Predictive success would be out, for instance, as would forms of "inference to the best explanation" that was spelled out in terms of the "likelihood" of the best explanation being higher than the "likelihood" for other explanations. I am thinking of epistemic virtues more generally: simplicity, explanatory power, and maybe something like "similarity to other rational beliefs." It is obviously quite a task to precisely describe those virtues. Further, as I said, I don't have any idea for how one could argue that these epistemic virtues give us a prima facia reason to think that a belief that had those virtues was "more rational" than one that didn't (and i'm not convinced, necessarily, that they do). In any case, these epistemic virtues shouldn't give us a reason to believe (at least solely) becuase they make it easier to accomplish some end: rather, I am wondering if these epistemic virtues by themselves are able to give us warrant for accepting some belief (that possesses those virtues) as rational.
My real question, though, is that assuming we can do this - assuming abduction can give us a rational reason to believe - could we then be in a position to judge inductive justification to be rational, and maybe to consider inferring projectible predicates rather than gruesome predicates to be rational (and, more interestingly, whether we could judge those things to be "more rational" then their negations). As a sketch of a sketch: it certainly seems simpler to posit a boring external world to positing any of the skeptical scenarios, and something like a uniformity of nature condition seems like a better explanation for our past predictive success and for our observations of constant conjunctions than does the alternate explanation that basically boils down to "stuff happens, and that is all you can say about it." If the simplicity or explanatory power of a belief then gives us some reason to think those beliefs are rational (or more rational than their alternatives), then it is rational (or more rational) to accept induction. At least maybe.
So, thoughts? Like I said, this is tangential to what I am actually reading and trying to write about, but I do think it is interesting: any recommendations for further reading on basically this topic?