Editors note: Because of a major snowfall in Savannah, GA during the recent Eastern APA, the Blog of the APA will be hosting papers by people who missed the opportunity to present or who would like a larger audience for their work. If you were on the schedule for the Eastern and would like to take advantage of this opportunity, please fill out the submission form.
Thanks to Juan for a great paper! I think a lot of what he says is correct, and I also think that PKP is wrong, but I have my doubts about the way the anti-luminosity argument proceeds.
The way that Juan goes about denying PKP is this: he sees the similarity between PKP and a luminosity principle (luminosity being something like self-intimation, but in Williamson’s terms); he then argues that there are cases with relevant gradations between your Φing and not Φing, and uses a modal condition on knowledge—what he has called MARA, which is something like a reliability principle—to motivate a reductio of PKP.
I have two main points and a broad structural question to ask Juan.
First Main Point
The first point is about this margin of error principle. As Juan notes in his paper, some have argued that a constitutive connection—or an otherwise tight necessary connection—between your believing something, and its being true, threatens the plausibility of this kind of reliability principle. The intuition is something like this: it seems that if your belief that p could easily have been false in a very similar situation to the one you are actually in, then your belief isn’t non-accidentally true—and so it doesn’t have the epistemic merit to count as knowledge. But if your believing something actually implies that it is true, then the modal closeness of various possible worlds in which that belief is false doesn’t seem like a threat to the knowledge status of that belief. It’s seemed plausible to people to raise this kind of worry against anti-luminosity arguments in the case of self-ascription of feelings and so forth, because believing that you feel cold may constitute, or partly constitute, your actually feeling cold.
Here I think that this kind of worry arises in a similar way. Now, as Juan is careful to state, it’s not the case that believing you are Φing implies that you are Φing. The world has to cooperate for you to do all sorts of things, things as simple as lifting a finger or as complex as directing a musical. But there’s a plausible constitutive claim in the vicinity here which I think deserves further attention: believing that you are trying to Φ implies that you really are trying to Φ. If that’s true, then a margin of error principle for tryings wouldn’t hold. The relevant closeness of possible worlds in which you are not trying to Φ does not seem like any threat to your knowledge that you are trying to Φ in this world.
The interesting necessitated conditional that I am here proposing as plausible is this:
If an agent is trying to Φ (under that description), then she knows that she is trying to Φ (under that description).
I raise this explicitly because I don’t think what Juan says about MARI obviously applies here. So I’d like to hear what Juan thinks about this related principle about tryings.
Second Main Point
Here’s a related (but I think distinct) worry. PKP as stated isn’t exactly in the form of a luminosity principle—though Williamson himself slides between several formulations of the luminosity principle. The luminosity principle is most plausible, I think, when it’s phrased in terms of your being in a position to know some condition C holds, not in terms of your actually knowing it. (Perhaps you’re inattentive, and so you don’t even consider the question of whether you feel cold, and so you don’t form any belief about the matter, and so you can’t be properly said to know you’re cold, and so forth.) Now, I understand that traditionally, the principle Juan is picking up is not formulated merely in terms of being a position to know something; that’s important to the point as made by Anscombe, for example. But the anti-luminosity argument, as it’s run, seems to rely on the fact that there is very little cognitive distance between being in a position to know that some condition C holds and actually knowing it holds. At the very least, this kind of anti-luminosity argument relies on the fact that doing whatever it is you do to go from being in a position to know something, to actually knowing it, doesn’t have a substantive impact on the holding of the target fact. That is: your finding it out doesn’t affect whether or not it is true.
Now here’s the question: does this structure also apply in the case of practical knowledge?
Let’s consider a weakening of PKP, call it “PKP-“:
If an agent is Φing (intentionally and under that description), she is in a position to know that she is Φing (intentionally and under that description).
PKP- seems more plausible to me than PKP itself. And it also takes more to run a reductio on this principle than on PKP. Importantly, as in the original anti-luminosity cases, the move between being in a position to know something and actually knowing it needs to be relatively trivial in terms of cogntive achievement, and it also needs to have no effect on the actual holding of the target fact to be known.
But there’s a real question whether that’s so in cases of practical knowledge. Paying attention to what you’re doing is something you get told to do in order to do it better, or to have a greater chance of success. Especially in cases of skill-building, attending to what you are doing, or focusing on your goal, or thinking explicitly about your intention, can have a real practical effect on whether you actually accomplish your goal—whether you really are Φing. The question of whether you are Φing, when you raise it to your conscious, directed attention, is not one that leaves the fact of your Φing untouched; it is entangled with it in a meaningful way.
It’s not clear to me that this is an objection to Juan’s argument per se. I wanted to raise it because I think that the appeal of Juan’s paper comes in part from his applying plausible principles concerning empirical knowledge to cases involving practical knowledge. So the general kind of point I’m making here is: here’s this difference between empirical and practical knowledge that concerns what gets done when you actually come to know something you were previously in a position to know (but didn’t yet know). I’d like to hear more about what Juan has to say on this too.
Final Question
In closing, I’d like to ask Juan a broad question. There’s an intuitive sense in which doing something intentionally involves having something in mind. Asking someone “What did you have in mind, when you did that?” can be a way of asking someone about what they were aiming at when they did something. But if Juan’s argument is correct, then there aren’t any plausible necessary connections between various doxastic and epistemic states (on the one hand) and your doing something intentionally (on the other hand). Doing something intentionally, on his view, doesn’t involve knowing what you’re doing, knowing what you’re trying to do, knowing what you intend to do, knowing what you’re doing under at least some description, or having a belief about what you are doing that matches the content of your intention.
If that’s right—if Juan’s argument works—then is there sense to be made of intentional action as involving a distinctive kind of understanding of what you are doing—that is, having something in mind?
Thanks again!
Antonia Peacocke
Antonia Peacocke is finishing her PhD at the University of California, Berkeley. She works on issues in self-knowledge and the philosophy of literature. You can find more about her work at her website, https://antoniapeacocke.com/.
Let me start by thanking Antonia for her excellent comments. They have helped me to see new avenues of inquiry, and deepened my understanding of what’s at stake in these debates. All three comments get at important issues, and I shall try to address them in turn.
In her first comment, Antonia proposes a principle, closely related to (PKP) that might avoid the problems associated with the anti-luminosity argument, namely:
If S is trying to Φ, then she knows that she is trying to Φ.
The reason this principle looks better is that it seems plausible to suppose that if a person believes that she is trying to Φ, then she is trying to Φ. If so, the beliefs that a person holds about what she is trying to do seem quite reliable—absolutely so, in fact, since she couldn’t be mistaken about them! Hence, the anti-luminosity argument cannot get off the ground, since, as Antonia notes, it presupposes that an agent can be epistemically unreliable with respect to the subject matter in question.
Let me start by acknowledging that the argument against the principle in terms of trying is weaker than the argument against my primary target, (PKP), for the reasons that Antonia presents. Nevertheless, I think the argument is quite strong against this other principle. First, it seems to me that it is possible to believe that you’re trying to do something when in fact you are not. For instance, a person might believe that he was trying to help his friend by telling him certain things, and later realize (e.g. through psychoanalysis) that he was instead trying to hurt him. Second, trying to Φ —as distinct from merely aspiring to Φ —seems to require commitment above a certain threshold. As commitment is a gradable notion, and this opens up the door for another a sort of unreliability to creep in, as is evident in the party case: the person might find herself in a situation where it is hard to tell whether she is trying or merely aspiring to do something, and when she is in such a situation, her beliefs are not reliable with respect to whether she is trying to so act, since she could too easily have confused her attempts with mere aspirations. Finally, I should note that a different sort of anti-luminosity argument, targeting degrees of confidence instead of full-out beliefs, is still available to the anti-luminist. Getting into the details of it would get us too far afield, though; and, in the end, I would rest content if the anti-luminosity argument is as strong against tryings as it is against appearances or pains, supposing these to be mental states necessitated by one’s belief that one has them.
This takes me to Antonia’s second point: perhaps when we are dealing with practical entities the anti-luminosity argument is weaker, because in the practical case the kinds of conditions that make possible the transition from being in a position to know to knowing add epistemic value. For, as she notes, it is plausible to think that when one concentrates on what one is doing, it is more likely that one will bring about what one is doing.
I think this is true, and important; but I don’t think the point can be used to call the argument into question. I have two minor worries about building a response to the argument pursuing this line of thought: the first is that it is equally plausible to think that in certain cases, concentrating on what one is doing makes it less likely that one will do what one intends or tries to do. Such might be the case when, e.g. one attempts to not act awkwardly. Concentrating on this action might make it more likely that one acts awkwardly. Second, for the strategy to work, we would need to ensure that the contribution that concentration makes is of the right sort: after all, there are cases of wishful thinking that might increase one’s chances of doing something, as in William James’s famous climber case, that don’t make an epistemic contribution, because they fail to give us specifically epistemic reasons. Setting these two points aside, though, my main response is that we can set up the cases so that the agents are paying full attention to what they are doing and still be able to run the anti-luminosity argument. As such, even if concentration increases one’s epistemic standing, it cannot guarantee knowledge.
Finally, the third point gets at what I think is the deepest motivation for (PKP), central at least to Anscombe’s understanding of the principle: as Antonia puts it, to do something intentionally is to do something with something in mind. One way to make sense of this idea is to claim that intentional action is necessarily accompanied and constituted by a certain sort of knowledge. Since knowledge is a mental state it can do the intentional (in the sense of ‘aboutness’) work: an action will be connected to what one has in mind via its epistemic component. But I’m rather attracted to a more radical idea, one according to which we let actions themselves do this intentional work. This would require us to conceive of actions as themselves intentional entities (entities with content), which, like beliefs, might or might not provide a good basis for knowledge. Thinking of actions in this way may seem somewhat foreign to us, but I think this idea, which goes back also to Aristotle as I read him, has much to be said for it, and is I think the view that Anscombe ultimately defends in Intention.
However, one could take a much less radical alternative without holding anything like (PKP); for one could hold that intentions are intentional entities, so for an action to count as having something in mind would be for that action to be appropriately connected with an intention (in my preferred way of understanding the appropriate connection, this comes out as the radical view outlined—but that is just one way of cashing out that view).
These last remarks are suggestive at best. Moreover, I’ve had to run rather quickly over material that requires much slower thinking. Hopefully what I said is enough to give a sense of where I stand on these issues to get a conversation started (I’m happy to follow up on any of this). To close, then, thank you again, Antonia, for your thoughtful and stimulating comments.