Ofra Magidor on Epistemicism and Moral Vagueness
29:57 Share

This is Philosophy Bites with me, David Edmonds. And me, Nigel Warburton. Philosophy Bites is available at www.philosophybites.com. If a man slowly loses his hair, at what point should we say he's bald? The tricky issue of vagueness is well known in philosophy, but it's not normally discussed in the context of morality. Othrar Magador of Oxford University says it applies here too.

This episode of Philosophy Bites is made in association with Vagueness and Ethics, a research project funded by the European Commission and based at Uppsala University in Sweden. Ofra Magidor, welcome to Philosophy Bites. Hi. The topic we're going to talk about is epistemicism and moral vagueness. Let's start with the vagueness. Epistemicism isn't obvious either, but what do you mean by vagueness in this context?

Okay, so let's take a word like tall. So we know that someone who's one meters in height is definitely not tall and someone who's two meters in height is definitely tall. But there are lots of heights in the middle. Like, you know, what about someone who's one meter 85? It's really hard to tell. They don't seem to be clearly tall. They don't seem to be clearly tall.

not tall. It doesn't seem like there's any further information we could find out about them to decide whether they're tall or not tall. So this is the kind of case that makes us think that words like tall are vague. They have borderline cases, cases in the middle where it's really hard to tell and it's not clear what would determine whether they're tall or not. Actually, even with tall, a metre could be tall if you were in Lilliput.

Or if you're in a, I don't know, in a kindergarten.

Yes, so that's an excellent point. So there is a bit of a complication here, right? So I was a little bit oversimplifying things when I was just talking about Tal. So maybe we need to hone down which particular meaning of Tal or reading of Tal we need. And it may be that it means something a bit different in different contexts. But what's interesting is that even if we focus on one context, right, even if we talk about

tall for someone in this particular kindergarten, we still don't know exactly what height you need to be to count as tall, even in that very special context. Yeah, so if we ask the tall people to line up on one side of the room and the short people on the other, we get it roughly right, but they're going to be some people who won't know which way to go, probably. Exactly. We won't even know how to instruct them to do it correctly.

So should we just accept that from a logical point of view, we just have these concepts which we call vague and they don't have a precise cut off where somebody's tall or short is a sort of grey area in the middle. We have all these notions of grey areas and similar phrases to denote that kind of vagueness.

Okay, so that's a really excellent question and it's sort of very helpful segue into the other word in our title, epistemicism. So certainly some views of vagueness think the lesson we should learn from these cases is that for all these difficult cases in the middle, there's just no fact of the matter. It's not that the person is neither tall nor not tall, okay?

The problem with that is that that conflict with everything we know from classical logic and mathematics. So epistemicism is a view of vagueness that wants to take a different route to handling these cases. It wants to say, even given this phenomena of vagueness, we shouldn't tinker with anything in our normal theories of logic and mathematics. If in general in logic, we know that for any statement P,

either P is true or not P is true, that's going to be the case even for these cases like Tallness and even for these borderline cases, even for that person who's sort of 1.85m and we're not sure if they're tall or not, we should still think in these cases that there's a fact of the matter.

The person is either tall or not tall. It's either true that they're tall or false that they're tall. And this is going to mean that there's a sharp cutoff point. In reality, there's a precise number such that that defines the boundary between the people who are not tall and the people who are tall.

Now, somebody who's not in the realm of logic might see this as a tail wagging the dog, because the world seems to have all kinds of vague concepts in it. And you're saying logic determines everything, that because we have this belief that every statement is either true or false...

then it must be the case that any statement about a vague idea must be true or false. Yes, so that is the view. It's maybe not the most intuitive view, but in other ways, if you think about it, it might be actually the default view. This is just what follows from the ordinary rules of logic and the ordinary rules of mathematics that exist.

Up until thinking about these vague concepts, we accepted fairly uncontroversially.

The other thing I think you're pointing to, which is why some people might find this very troubling, is that it might sound like this view epistemicism isn't really a theory of vagueness, but a theory that does away with vagueness. Is that maybe part of what you're finding troubling here? Yeah, absolutely. Because it seems that the world has vague concepts and then you're resisting that because you're a logician.

Okay, good. So I think it's really important to understand that what the epistemicist is doing is not denying that there is vagueness or that there are borderline cases. They're giving a different account of what it takes to be vague. Okay, so if you had one theory of vagueness you could have is to say what is crucial to being one of these borderline cases is that there is no fact of the matter or there's no truth value.

If you had that view, then it may sound like the epistemicist is saying that there is no vagueness, but that's not the sort of interpretation of what vagueness is that the epistemicists want to accept. They want to say that the distinction between vague and non-vague words, it's not to do whether there is a sharp cutoff point or there is a fact of the matter. They want to draw the distinction in different terms.

So could you succinctly characterise that position on vagueness? What would an epistemicist say vagueness is?

I would say in one sentence, epistemicism says that vagueness is a distinctive kind of ignorance. Although words like tall, they in fact have a sharp cutoff point, the crucial other component of epistemicism is that we don't know what this cutoff point is, and in fact we can't know what it is.

So we're in principle ignorant of this sharp cutoff point. And that specific kind of ignorance about sharp cutoff points is the thing that characterizes vagueness. The central component of vagueness isn't about truth values. It's about knowledge. It's about what we know or don't know. Okay, so how does that play out for us?

something like a moral judgment moral judgments are controversial lots of different opinions usually about which things are right and wrong though on the whole we tend to think that murder is wrong most torture but there are lots of areas which we would talk about as gray areas

People who differ over whether abortion at a certain time after conception is morally acceptable, that would be a case where there'd be grey areas. So how does this concept of vagueness relate to those sorts of moral questions?

Good. So you've already pointed to one thing we're noticing here, which is it's not just words like tall or bald that seem to be vague, but it looks like also the kinds of words that we use in ethical discourse.

words like what you ought to do, what you're permitted to do, they also seem to have these characteristics of vagueness. So you mentioned the case of abortion, and this is a classic case to demonstrate this. So suppose you accept that at a very early stage, immediately after conception, it is permissible to abort a fetus. I know not

All of our listeners accept that. If they don't, we could just use another example, but just for the purposes of discussion, let's assume that. I think most of us think at a very late stage of pregnancy, unless they're very special circumstances, it's not permissible to abort a fetus. But just like our case with Tal, there seem to be a lot of cases in the middle where it's

very hard to know whether it's permissible and not permissible. So in a way, that's a contrast with the law, which tends to be very precise and somewhat arbitrary, it seems, as there is a precise cutoff point beyond which this is illegal. Yeah, it's interesting about the law, right? Because even the law might, I don't know if it specifies it up to a precision, right? The sharp cutoff points are

that we get from logic are going to say that there is a precise point up to the nanosecond. So even when you think of fairly arbitrary decisions of the law, they're not going to go as far as admitting sharp cutoff points as precise as this. I can just about make sense of the idea that there is an actual precise cutoff point

for whether somebody's tall or not, whether somebody's tall in the Netherlands is going to be different from whether someone's tall in China. But there's a precise cutoff point, I can just about accept that. How would it work for moral cases? That's really interesting, right? So there are some reasons you might think the moral case is actually better for this epistemicist view, works better with it. And some reasons you might think it's worse for the case of moral concepts.

So let's start with why you might think epistemicism actually works quite well for these moral cases. Well, let's assume we accept a fairly strong form of moral realism. So we think moral statements, they can be true and they're true not just because we think they're true or we say they're true. There's really fundamental facts, moral facts in the world

that make moral statements true or not, right? That kind of view

fits quite well with epistemicism. So take our fetus that is in this very borderline situation. For a moral realist, it feels very natural maybe to think it's not an arbitrary choice. This is a really serious, difficult question. If we abort this fetus, we might be doing something wrong. There has to be a fact of the matter whether what we did is right or wrong.

So I could see if somebody thought, I don't believe this, but if there were a God who gave us moral rules, then that God might give us very precise moral rules. It's just we haven't quite worked out at the right level of precision where the cutoff points are. But there is a truth in the matter then. It's written in stone somewhere.

Exactly. That's a kind of view that you might think works well with this component of epistemicism that says there really is a fact of the matter. And you don't have to think it's God that gave us the moral rules, right? You might think there's just moral reality, there are moral facts, which is a view I think that is attractive to a lot of people. Now, I know some people who accept the

epistemic vagueness as an approach to vagueness generally have a problem with it in the area of morality specifically. Could you say something about that? Yes. So there are two very different kinds of reasons I think people have thought, even if you like epistemicism in general, it has some trouble when we try to apply to the case of moral concept. Okay, so one kind of problem that people have

raised, it has to do with the fact that according to epistemicism, these sharp cutoff points, there are truths that we don't and can't know. And some people have had the following worry. Moral truths are things that are supposed to guide our action. And if it turns out, take this fetus in this sort of borderline area, suppose they're just over the cutoff. Okay, so in fact, we

We ought not to abort this fetus, but this is so close to the cutoff. This is one of these borderline areas where, according to the epistemicist, we can't know that we're just over the cutoff. You might think this is a really weird case, right? The moral truth is,

tells us we ought not to abort the fetus, but we also can't know that this is what the moral truth tells us to do. You can think of it as a version of the "ought implies can" problem, right? The thought is if there's something we ought to do, then we can do it. And you might think part of being able to do it is being able to know that you ought to do it.

How would you reply to that? Yes, so one thing to say is that this actually isn't a specific problem just for epistemicism. In a way it's a problem for any of you that thinks there's moral vagueness.

because pretty much on any theory of vagueness on these borderline cases, you can't know. You can't know that you ought not to abort the fetus and you can't know that it's permissible to abort the fetus. The only difference with epistemicism is that there's also this fact of the matter. But just saying there isn't a fact of the matter doesn't really solve the problem in terms of being action guiding, right? If I'm trying to decide what to do and...

the theory of vagueness tells me, well, we're not saying it's permissible for you to do this. We're also not saying it's not permissible for you to do it. Then I'm still stuck with the theory not really helping to guide my action. So it could be quite reassuring because you're going to be okay either way.

Yeah, so I think it's tempting to think that this is what these views say. But in fact, I think that's not what they say, right? So it's not the case that you ought not to abort the fetus. And you might think, oh, great, this means I'm permitted to abort the fetus. But of course, on these views, it's also not true that you're permitted to abort the fetus, right? So it's not that the view tells you that you're okay either way. It's

just doesn't really tell you anything. It's not saying that you're okay, and it's not saying that you're not okay. What about the second objection that you mentioned there is to this approach to understanding moral concepts, vague moral concepts? Okay, so the second issue is quite interesting, and to understand it, we have to delve a little bit deeper into the fine details of how epistemicism works.

So I said that epistemicists say that in cases of vagueness, you are ignorant of these sharp cutoff points and there's a very specific reason or kind of ignorance that is special in the cases of vagueness. So the bit we need to delve a bit deeper into is what is this explanation? Why are you ignorant? And what is this special kind of ignorance?

So to do that, let's talk about the one most prominent answer to this question by the most prominent defender of epistemicism, which is Timothy Williamson. And Williamson has roughly the following idea. It starts with the thought that

How we use words is going to determine what they mean. If you think about the word tall, there's nothing special about this specific set of sounds or this specific thing we write. The very same sound, right? If we had a community that used the word tall in a completely different way and used it, the sound is tall. Whenever people saw cats, then is tall might have picked out the property of being a cat.

and not the property of being tall. Okay, so that's true of all words. However, here's the interesting thing. According to this theory, there's a difference between how sensitive what we mean is to our use in cases of non-vague words and in cases of vague words. So the thought is, take a precise word like "is prime", okay, the mathematical concept of being a prime number,

The thought is, of course, the fact that we mean by that this particular mathematical property depends on how we use it doesn't depend on exactly how we use it. We can think of lots of communities that used

is prime roughly the way we do and it's really going to pick up the mathematical property of being prime, right? If a few more children mistakenly in their math class thought that nine is prime and uttered the sentence nine is prime, the word prime would still pick out the same familiar mathematical property.

What's special, according to this theory about vague words like "is tall" is that they are very highly sensitive to how we use them. They have this property that we can call semantic plasticity. And what semantic plasticity means is that if we look at a bunch of different communities that use the word only very slightly differently, they're going to pick up a slightly different property.

So if it turns out that the way we use the word ends up picking up a property with a sharp caught off point, say,

1.813 meters, maybe a different community that used it a little bit differently would mean by it 1.812 meters. And the final building block we need is that this kind of semantic plasticity is the key to explaining why we can't know these sharp cutoff points.

And very roughly the thought is we're not sensitive enough to these differences. We don't have beliefs that can track the exact way in which these used facts determine these very specific sharp cut-off points. Because of these plasticities, we lack the knowledge for the sharp cut-off points. And then is the claim that moral terms have a lot of this plasticity about them.

Exactly. So now we finally get to what the crux of the second problem is. The second problem has to do with this claim that moral terms have this extreme plasticity. So just to recap where we got to so far.

the epistemicist wants to explain vagueness in terms of semantic plasticity. So if you want to apply epistemicism to the case of moral terms, then you're also going to have to say that these moral terms have this extreme semantic plasticity.

Let's think of our community. So we go around, we make all these different utterances using the word permissible, for example. Okay, and let's suppose we end up picking out a particular property.

And then there's this other community that goes around using a word that sounds just like ours and has very, very similar function. They use it in a very similar but not quite exactly the same way. And it turns out they pick out a slightly different property. And we can see the difference in going back to our interesting case of the fetus. Suppose that

as we use the term permissible, it turns out, we don't know this, but it turns out that the exact cutoff points were permitted to abort the fetus exactly up to when the fetus is n minutes old. And if we have a fetus that

And minus one minutes old, it's fine. We're permitted to abort it. So if we let's take our fetus that is and minus one minutes old, I go around saying that I'm permitted to abort it and I'm saying something true.

Now look at the person in that other community that uses the word permissible slightly differently and say they pick up a slightly different property, one where you're only permitted to abort the fetus up to n minus two minutes. So here's a really weird thing that's going on.

When they say that you are permitted to abort this fetus, they're saying something false. So just because we're using the word permissible slightly differently, it turns out that when I say it's permissible to abort the fetus, I'm saying something true. And when they utter the sentence, it's permissible to abort the fetus, they're saying something false.

This, I think, has been a thought that people have found very troubling, especially people who accept a kind of moral realism. So is the conclusion that we should draw from all this a kind of reductio of moral realism, that you have to do mental gymnastics to survive as a moral realist?

There are, I think, several things to say here. But here's one. It's not even though I made this speech about the two communities and the fact that it looks troubling. When you think about it, it's not at all clear this scenario actually conflicts with moral realism.

Remember in my scenario I said that when I utter the sentence "it's permitted to abort this fetus" I'm saying something true and when someone in this other community utters it they're saying something false. But remember on this view we don't mean exactly the same thing by the word "permitted". So this isn't exactly a conflict with moral realism, right? It might look like it, might look like

what's morally true depends on how our community is using the terms. But they're not really disagreeing with me, right? They're not making the same claim. When they say you're permitted to abort the fetus, that sentence in their mouth means something a little bit different. But that then sounds like relativism, where you're saying that, well, it's true for them. And this is true for me. It's permissible for them, but impermissible for me or vice versa.

Good, so this is a really common misunderstanding. So it might sound like this is a form of moral relativism, but it isn't. So moral relativism would think that there's actually a claim, a proposition such that it's true for me, but false for you. This isn't the case here, right? The case here is that there's a sentence, there's a certain set of sounds that...

say something true for me and say something false for you. So you can imagine a sentence where the same sounds in English, they mean that two plus two equals four, but maybe the very same sounds just

accidentally turn out to also be a sentence in French, where in French that sentence says that 2 plus 2 equals 5, right? So that would be a sentence that when an English speaker says it, they're expressing the claim that 2 plus 2 equals 4 and they're saying something true. And when a French speaker says those very same sounds, they're expressing the claim that 2 plus 2 equals 5 and they're saying something false. That

wouldn't be relativism about mathematics, right? That would just be a case where the words are being used in our two languages to mean something slightly different. So why does that present a problem for the approach to vagueness in the moral realm?

So I think there is a bit of a problem for how to apply this to moral terms. Okay, so think of our two communities that use permissible slightly differently. And we saw one of them judges that.

It's permissible to abort the fetus. The other judges that it's not permissible to abort the fetus. And each of them in their own language is saying something true. Okay, so they're speaking slightly different languages. It's not, as we said before, that we get some kind of moral relativism. But you might think for a very serious moral realist, that's still a problem, right? Because both

Both of the terms they're using are a kind of moral judgment. They're using permissible to make pronouncements that we might think of as moral.

So what we're getting here is that there are two different properties. There's the permissibility as it's used in our language, and there's this other property, you can call it permissibility star, that's used in their community. And now we have two moral properties, permissibility and permissibility star.

And you might think that doesn't directly clash with moral realism, but you might think that's not a nice thing for a moral realist to think that there are many, many different moral or moral-like properties that we could be using, right? If you're a serious moral realist, you should think there's just permissibility. There are no other

similar properties in the vicinity that play a very similar role in our society. Maybe this is just a general problem with moral realism, but across history, a concept like permissibility does obviously seem to have that kind of feature. 1960s California permissibility is very different from 1890s London permissibility. Yeah, so I think serious moral realists are going to think that's not correct. They're going to think that

Many, many societies along history just were wrong about what was permissible and what wasn't. It wasn't that permissibility meant something different in their language. It meant the same thing all along. It's just that some people had beliefs about permissibility that were wrong and some people had beliefs about permissibility that are correct. So what would it mean to reject moral vagueness?

Okay, so earlier in our discussion, we just assumed that moral terms are vague because we noticed cases like this case of the abortion, where when the fetus was one minute old, we were permitted to abort it. And when it was 39 weeks old, we were not permitted. But there were all these cases in the middle where it seemed like we have no idea and it's

It's not even clear what further facts we could gather to find out whether we can abort it. But we need to be a little bit careful here because although these cases of ignorance, they look similar to other cases of vagueness, not all cases where we don't know the answer to a question are cases of vagueness.

One way to see that is in the case of mathematics, the one domain where everyone thinks there is no vagueness at all. We use very, very precise language. But even in mathematics where we use very precise language, there are open questions in mathematics. There are questions where we really don't know the answer to them.

Maybe we will never find out the answers. Maybe we don't know what further facts that are out there to find out in order to resolve this question, right? So there are cases of questions that are just very, very difficult to

to resolve. We have no idea how to answer them. And the reason we don't know how to answer them has nothing at all to do with the phenomena of vagueness. So one option about these moral cases is to just think, we don't know the answer to these questions, but that's not because they are vague, but just because they are very hard questions. So do you agree with that last way of saying things? Do you think

that it's at least possible that there are moral facts, it's just that we don't know them and it's not due to vagueness at all. Yes, I think there's something very attractive about that way of resolving this trilemma, or at least this is one direction that we should very seriously explore and I think hasn't been sufficiently explored in the literature up to now. Ofra Magidul, thank you very much. Thank you.

For more Philosophy Bites, go to www.philosophybites.com. You can also find details there of Philosophy Bites books and how to support us.