Phil Atkins
Hey gang. I thought I'd kick off with some thoughts on epistemology. Let me know if you guys have thoughts of your own.
The notion of a relevant alternative is at center-stage in contemporary epistemology (equivalently, the notions of relevant scenarios, relevant situations, etc.). It is suggested by Goldman, for instance, that in order to know p one must be able to distinguish between the actual situation in which p is true and certain relevant situations in which p is false. In the same spirit, Dretske suggests that knowing p requires the ability to rule out relevant not-p alternatives. These approaches to knowledge are collectively known as “Relevant Alternatives” theories (or RA theories, for short).
Classic skeptical scenarios--such as those in which I am a bodiless brain-in-a-vat who is merely being deceived into thinking that I have a body--have been standardly dismissed as epistemically irrelevant scenarios. So, even though I can’t rule out the skeptical scenario, I can know that I have a body because the skeptical scenario is an irrelevant alternative to the actual situation. I’m not epistemically obligated to rule out such scenarios.
Although there are few precise characterizations of epistemic relevance (if anybody knows of any interesting papers on this subject, then please let me know), I think that it is widely assumed that relevance is closely connected to modal proximity. On this assumption, the relevance of an alternative scenario has something to do with the closeness of the possible world in which the scenario obtains. If the possible world is nearby, then the scenario is more relevant; if it is faraway, the scenario is less relevant.
I want to argue, however, that modal closeness is neither necessary nor sufficient for an alternative scenario to be epistemically relevant. I want to show, first, that there are cases in which an alternative scenario obtains on a nearby possible world but is still not relevant; thus, modal closeness is not a sufficient condition for epistemic relevance. Then I want to show that there are cases in which alternative scenarios obtain only on faraway possible worlds but are still relevant; thus, modal closeness is not a necessary condition for epistemic relevance. Now I'm going to assume that some version of RA theory is true. I understand that there are various ways of cashing out an RA theory, but I don’t think that details will bear upon my cases.
In the first case, I ask you to imagine that last night one of your neighbors had their brain placed in a vat. We can imagine that this is due to the activity of a certain mad scientist (who else?). In any case, your neighbor woke up this morning believing that he was embodied, but in fact he was a bodiless brains-in-vat who was being deceived into thinking that he had a body. For him, the skeptical scenario is a reality. Now, your brain was supposed to have been placed in a vat instead of his, but you were spared due to a clerical error on the part of the mad scientist. Now imagine yourself waking up this morning. You look at your body and come to form the belief, say, that you have hands. Suppose that you’ve done this every morning for years. On this particular morning nothing is different--other than the fact that your neighbor was recently envatted. You, of course, are blissfully unaware of his circumstances. Things appear to be the status quo. Now it seems clear that on every previous morning you have had a strong claim to knowing that you have hands: sure, you were unable to rule out the skeptical scenario that you were a handless brain-in-a-vat, but so what? The scenario was taken to be irrelevant. Now I ask: on this particular morning, do you know--based on your perceptual faculties, which are being exercised in the same way as on every other morning--that you have hands? I think the answer is clearly ‘yes’. Still, you are unable to rule out the alternative skeptical scenario that you are a handless brain-in-a-vat. And, what’s important, this skeptical scenario is one which obtains on a very nearby possible world. Still, it is not one which seems to threaten your claim to knowledge. In this case, at least, you know that you have hands despite the fact that you can’t rule out the skeptical scenario and despite the fact that you were almost envatted on the night before. I submit that you are capable of such knowledge because the skeptical scenario is still irrelevant, despite its modal proximity. So this is a case in which an alternative scenario is epistemically irrelevant but modally nearby. Thus, modal closeness is not a sufficient condition for epistemic relevance.
In my second case, I ask you to imagine that you and your friend are hiking through the woods. Your friend spots a hawk in the sky and comes to form the belief that there is a hawk flying above. Now your friend knows very little about ornithology. His belief is based on the following grounds: there, in the sky, is something with wings and feathers. You ask your friend how he knows that the bird is hawk. He answers, “Well, because I see that the bird has wings and feathers.” You slap your head, appropriately in this case. For it is obvious that on such grounds your friend has no claim to knowledge. His belief is true but it just so happens to be true. His evidence fails to rule out various alternative scenarios: that there is a pigeon in the sky, or a blue-jay, or an eagle. Without being able to rule out such possibilities, your friend is hardly in a position to know that he sees a hawk (rather than a pigeon, or a blue-jay, or an eagle). But suppose that, months before, every other bird in the country was eradicated by a deadly virus. Hawks, for whatever reason, are immune to the virus. They are the only birds left within the United States. Now the powers that be have kept this bird-epidemic under wraps. Thus, you and your friend are unaware of the virus and its effects on the bird population. Still, the fact remains that the only birds within miles are hawks, and we can assume that there are no pesky bird-facsimiles in the area. Now I ask you: in the situation I’ve described to you, in which the only birds within countless miles are hawks, does your friend know that there is a hawk above him (rather than some other bird)? The answer here is ‘no’. His evidence is underdetermining: as I stressed before, it can’t rule out the pigeon-alternative, or the eagle-alternative, or any number of other alternatives. But note: due to the effects of the virus, these alternatives obtain only on faraway possible worlds. Even still, your friend has no claim to knowledge. For, despite the distance between those worlds and the actual world, your friend is still epistemically obligated to distinguish between them and the actual situation. He is still obligated to rule out the alternatives that obtain on those worlds. What we have, then, is a case in which someone fails to know something because he is unable to rule out alternative scenarios--despite the fact that the relevant scenarios obtain only on faraway possible worlds. Thus, modal closeness is not a necessary condition for epistemic relevance.
There's more to be said about these cases, but I've written a lot already.
Hey gang. I thought I'd kick off with some thoughts on epistemology. Let me know if you guys have thoughts of your own.
The notion of a relevant alternative is at center-stage in contemporary epistemology (equivalently, the notions of relevant scenarios, relevant situations, etc.). It is suggested by Goldman, for instance, that in order to know p one must be able to distinguish between the actual situation in which p is true and certain relevant situations in which p is false. In the same spirit, Dretske suggests that knowing p requires the ability to rule out relevant not-p alternatives. These approaches to knowledge are collectively known as “Relevant Alternatives” theories (or RA theories, for short).
Classic skeptical scenarios--such as those in which I am a bodiless brain-in-a-vat who is merely being deceived into thinking that I have a body--have been standardly dismissed as epistemically irrelevant scenarios. So, even though I can’t rule out the skeptical scenario, I can know that I have a body because the skeptical scenario is an irrelevant alternative to the actual situation. I’m not epistemically obligated to rule out such scenarios.
Although there are few precise characterizations of epistemic relevance (if anybody knows of any interesting papers on this subject, then please let me know), I think that it is widely assumed that relevance is closely connected to modal proximity. On this assumption, the relevance of an alternative scenario has something to do with the closeness of the possible world in which the scenario obtains. If the possible world is nearby, then the scenario is more relevant; if it is faraway, the scenario is less relevant.
I want to argue, however, that modal closeness is neither necessary nor sufficient for an alternative scenario to be epistemically relevant. I want to show, first, that there are cases in which an alternative scenario obtains on a nearby possible world but is still not relevant; thus, modal closeness is not a sufficient condition for epistemic relevance. Then I want to show that there are cases in which alternative scenarios obtain only on faraway possible worlds but are still relevant; thus, modal closeness is not a necessary condition for epistemic relevance. Now I'm going to assume that some version of RA theory is true. I understand that there are various ways of cashing out an RA theory, but I don’t think that details will bear upon my cases.
In the first case, I ask you to imagine that last night one of your neighbors had their brain placed in a vat. We can imagine that this is due to the activity of a certain mad scientist (who else?). In any case, your neighbor woke up this morning believing that he was embodied, but in fact he was a bodiless brains-in-vat who was being deceived into thinking that he had a body. For him, the skeptical scenario is a reality. Now, your brain was supposed to have been placed in a vat instead of his, but you were spared due to a clerical error on the part of the mad scientist. Now imagine yourself waking up this morning. You look at your body and come to form the belief, say, that you have hands. Suppose that you’ve done this every morning for years. On this particular morning nothing is different--other than the fact that your neighbor was recently envatted. You, of course, are blissfully unaware of his circumstances. Things appear to be the status quo. Now it seems clear that on every previous morning you have had a strong claim to knowing that you have hands: sure, you were unable to rule out the skeptical scenario that you were a handless brain-in-a-vat, but so what? The scenario was taken to be irrelevant. Now I ask: on this particular morning, do you know--based on your perceptual faculties, which are being exercised in the same way as on every other morning--that you have hands? I think the answer is clearly ‘yes’. Still, you are unable to rule out the alternative skeptical scenario that you are a handless brain-in-a-vat. And, what’s important, this skeptical scenario is one which obtains on a very nearby possible world. Still, it is not one which seems to threaten your claim to knowledge. In this case, at least, you know that you have hands despite the fact that you can’t rule out the skeptical scenario and despite the fact that you were almost envatted on the night before. I submit that you are capable of such knowledge because the skeptical scenario is still irrelevant, despite its modal proximity. So this is a case in which an alternative scenario is epistemically irrelevant but modally nearby. Thus, modal closeness is not a sufficient condition for epistemic relevance.
In my second case, I ask you to imagine that you and your friend are hiking through the woods. Your friend spots a hawk in the sky and comes to form the belief that there is a hawk flying above. Now your friend knows very little about ornithology. His belief is based on the following grounds: there, in the sky, is something with wings and feathers. You ask your friend how he knows that the bird is hawk. He answers, “Well, because I see that the bird has wings and feathers.” You slap your head, appropriately in this case. For it is obvious that on such grounds your friend has no claim to knowledge. His belief is true but it just so happens to be true. His evidence fails to rule out various alternative scenarios: that there is a pigeon in the sky, or a blue-jay, or an eagle. Without being able to rule out such possibilities, your friend is hardly in a position to know that he sees a hawk (rather than a pigeon, or a blue-jay, or an eagle). But suppose that, months before, every other bird in the country was eradicated by a deadly virus. Hawks, for whatever reason, are immune to the virus. They are the only birds left within the United States. Now the powers that be have kept this bird-epidemic under wraps. Thus, you and your friend are unaware of the virus and its effects on the bird population. Still, the fact remains that the only birds within miles are hawks, and we can assume that there are no pesky bird-facsimiles in the area. Now I ask you: in the situation I’ve described to you, in which the only birds within countless miles are hawks, does your friend know that there is a hawk above him (rather than some other bird)? The answer here is ‘no’. His evidence is underdetermining: as I stressed before, it can’t rule out the pigeon-alternative, or the eagle-alternative, or any number of other alternatives. But note: due to the effects of the virus, these alternatives obtain only on faraway possible worlds. Even still, your friend has no claim to knowledge. For, despite the distance between those worlds and the actual world, your friend is still epistemically obligated to distinguish between them and the actual situation. He is still obligated to rule out the alternatives that obtain on those worlds. What we have, then, is a case in which someone fails to know something because he is unable to rule out alternative scenarios--despite the fact that the relevant scenarios obtain only on faraway possible worlds. Thus, modal closeness is not a necessary condition for epistemic relevance.
There's more to be said about these cases, but I've written a lot already.