The fact that it would be wonderful and inspirational if rationality was always the winning strategy, doesn’t mean that rationality is always the winning strategy.
I would offer that rationality is not a winning strategy, it is a meta-strategy for identifying winning strategies.
I guess I was trying to make a side point by using the two as though they were synonymous. Maybe the precise way would be that “instrumental rationality” is the study of systematically winning strategies, just like “epistemic rationality” is the study of systematically accurate guessing.
I would offer that rationality is not a winning strategy, it is a meta-strategy for identifying winning strategies.
You have just passed the recursive buck. Identifying winning strategies and then using them is also a winning strategy, an adaptive one, which may make it stronger. It is this strength that matters, not the ritual.
That’s a virtuous pass, not a vicious pass. Deriving instrumental utility on the reflective level from instrumental utility on the object level is just what we want. Defining truth on the object level by invoking a definition of truth on the meta-level would be vicious.
Would you prefer “unboundedly recursive”? So on any given occasion it will only recurse to a finite depth, but there’s no bound on the depth of its recursion?
That’s fine, but I’ve never seen a “pass the recursive buck” scenario that actually did work by requiring only a finite recursive depth on any given occasion. It always degenerates into an infinite hierarchy of ordinals that you can’t describe without creating a new hierarchy on top.
Well, I mean yes there are programming exercises for computing the Fibonacci numbers; I’m referring to when this trick is tried in epistemology or logic.
I have in mind a scenario something like what Dennett describes in Consciousness Explained: we imagine that our awareness of our own thoughts is in some mysterious way infinitely recursive, because when we go looking for a bound on how many times we can repeat the step of becoming aware of the previous level of awareness, we don’t find one; but the bound arrives exactly whenever we care to stop looking. There’s no bound on how often we can reflect on the way we’re deciding a particular question and decide if that, in turn, is rational, but there will have to come a point at which you have to stop recursing if you want to actually decide the base question.
There again, it may be a mistake to look for a sensible meaning in Annoyance’s usual vague crap.
I would offer that rationality is not a winning strategy, it is a meta-strategy for identifying winning strategies.
I would agree but put “rationality” in quote marks, that is, it is the subject of the discipline named “rationality” to find rational strategies.
Did you mean “to find winning strategies”, or are you using those synonymously?
Either way, I agree with the reference/value distinction here.
I guess I was trying to make a side point by using the two as though they were synonymous. Maybe the precise way would be that “instrumental rationality” is the study of systematically winning strategies, just like “epistemic rationality” is the study of systematically accurate guessing.
You have just passed the recursive buck. Identifying winning strategies and then using them is also a winning strategy, an adaptive one, which may make it stronger. It is this strength that matters, not the ritual.
That’s a virtuous pass, not a vicious pass. Deriving instrumental utility on the reflective level from instrumental utility on the object level is just what we want. Defining truth on the object level by invoking a definition of truth on the meta-level would be vicious.
I would go even farther than you: rationality is an infinitely-recursive process of evaluation whose fundamental principle is consistency.
I am interested in how this infinite recursion manages to complete in finite time.
Would you prefer “unboundedly recursive”? So on any given occasion it will only recurse to a finite depth, but there’s no bound on the depth of its recursion?
That’s fine, but I’ve never seen a “pass the recursive buck” scenario that actually did work by requiring only a finite recursive depth on any given occasion. It always degenerates into an infinite hierarchy of ordinals that you can’t describe without creating a new hierarchy on top.
Well, I mean yes there are programming exercises for computing the Fibonacci numbers; I’m referring to when this trick is tried in epistemology or logic.
I have in mind a scenario something like what Dennett describes in Consciousness Explained: we imagine that our awareness of our own thoughts is in some mysterious way infinitely recursive, because when we go looking for a bound on how many times we can repeat the step of becoming aware of the previous level of awareness, we don’t find one; but the bound arrives exactly whenever we care to stop looking. There’s no bound on how often we can reflect on the way we’re deciding a particular question and decide if that, in turn, is rational, but there will have to come a point at which you have to stop recursing if you want to actually decide the base question.
There again, it may be a mistake to look for a sensible meaning in Annoyance’s usual vague crap.