The Sceptic, The Outsider, and Other Minds

The usual way with scepticism is to formulate a problem in connection with the external world and then apply this to other minds. Drawing on work by Stanley Cavell and Richard Moran, I argue that the sceptic misses an important difference in our concepts of mind and of body, and that this is reflected in the sceptic’s formulation of a problem regarding other minds. I suggest that an understanding of this important conceptual difference is also missing from the work of those who attempt to reply to (dismiss, or ignore) the sceptic. In this connection I discuss both inferential and perceptual accounts of our knowledge of other minds. I identify an error in these accounts that may be thought to arise from a lack of understanding of the important conceptual difference here, and then develop an understanding of this error that draws on the work of Edith Stein and Stanley Cavell.


Introduction
Just as there are many varieties of anti-scepticism, there are many scepticisms one may be anti. 1 Furthermore, these scepticisms and anti-scepticisms may be considered in connection with different domains. Scepticism has been raised in connection with the world of bodies (or, as the sceptic would have it, the external world), in connection with minds (other minds), in connection with the past (the reliability of memory) and the future (the problem of induction). The usual way in the literature is to begin by considering scepticism with respect to the world and then play the same game out in connection with these other domains. 2 For the purposes of this paper, I will leave to one side scepticism about the past and the future, and simply consider its application to the world and to other minds. I want to concentrate on the tendency to consider scepticism, first in connection with the external world, and only secondarily in connection with minds. Proceeding in this way encourages us to think of other minds scepticism as nothing more than a special case of external world scepticism, obscuring important differences here. There is a lurking assumption that the Ur problem is the one formulated in connection with the world, and that both our formulation of the sceptical problem and our attitude towards it here should be the template for how we think about the problem and our attitude towards it in connection with other minds.
In Sect.
(2) I suggest some reasons for thinking that it is misguided to begin with a formulation of a sceptical problem in connection with the external world and then seek to apply this, without modification, to the case of other minds. This is not to deny that there are some similarities in the way we think about knowledge and the problems raised for it in these two domains, but the dissimilarities here are enough for us to consider our knowledge and the problems that may be raised for it separately. In Section (3) I examine two prominent suggestions in the literature for how we know the minds of others. I argue that both proposals can be shown to overlook an important difference between knowledge of bodies and knowledge of other minds identified in section (2). One idea that will run through both sections (2) and (3) is this: the differences that can be identified in the application of the sceptical argument to our claims to know the world and to know other minds reflect a deep difference in our concepts, on 1 For example, James Conant has identified two major varieties of scepticism, Cartesian and Kantian (see Conant 2004); Crispen Wright has distinguished a Cartesian and a Humean variety (see Wright 2004); Michael Williams has written of an Agrippean variety (see Williams 2004). 2 Induction does not quite fit the same pattern, hence Wright's identification of a Humean variety of scepticism (see Wright 2004).
1 3 the one hand, of body, and on the other, of mind. My interest throughout will ultimately be with questions concerning knowledge of other minds; my concern with external world scepticism is considered only by way of contrast.

Formulating the Problem
The first consideration I shall mention for thinking of radical scepticism differently in connection with these two domains is, perhaps, the most superficial. Consider the lengths that Descartes has to go to in order to introduce radical scepticism in connection with the external world. If we look at the First Meditation, we find that Descartes believes that the first two of his sceptical considerations are not up to the task of leading us to the radical scepticism he wants to convince us of. The mere possibility of error-and even the possibility of dreaming-does not lead Descartes to his moment of world loss. It takes something as radical as the possibility that God could be deceiving us-a suggestion he quickly rejects in favour of an evil demon doing the deceiving. The fact that it takes something with a God-like power to wrench us apart from our world is worth noting. When we turn to other minds, we see that we can more easily be led to doubt. Consider the vast range of non-human animals for which the question arises whether they have minds. While we are inclined to accept the existence of mind in some non-human animals, questions remain about how far down the phylogenetic scale mind extends. Furthermore, mankind has long wondered if any alien creatures that may inhabit other planets really have minds; and we continue to ponder whether we will ever build a robot that has a mind. Call this the different body question. Some might go so far as to question the mindedness of someone who may look like us but acts in ways that we cannot fathom. We can stretch the differentbody question to include these cases. Whatever reply we give, the different-bodies question seems a quite natural one to ask. If we accept this, then we might also think that it is rather a small step to extend this question even to those who look and behave much as I do. (Note that, in order to do this, there would have to be a move from a question that we ask to a question that I ask.) My point is that radical scepticism in connection with other minds may be thought to arise in everyday contexts -it does not require esoteric considerations or considerations that arise only in science fiction. 3 While I take this to be an interesting observation, I am not sure how deep it goes. The second consideration may be thought to build upon this initial observation and, in that sense, takes us a little deeper. This second consideration is taken from the work of Stanley Cavell. In The Claim of Reason Cavell weaves about, finding here a symmetry and here an asymmetry between external world and other minds scepticisms. 4 I want to emphasize the asymmetry and encourage reflection on the differences here (although I don't mean to ride roughshod over any similarities). In The Claim of Reason, Cavell explores scepticism as it is formulated in connection with the external world and with other minds, and he identifies the following asymmetry: "with respect to the external world, an initial sanity requires recognizing that I cannot live my skepticism, whereas with respect to others a final sanity requires recognizing that I can. I do". 5 Let us unpack why Cavell says this. Consider first the case of external world scepticism. In this connection Cavell distinguishes a reasonable scepticism from what he calls a "lunatic" one. 6 This distinction between what is reasonable and what lunatic may be thought of as attitudes towards what I have elsewhere referred to as a thin and a thick sceptical problem. 7 The thick sceptical problem is the radical one, the one posed by Descartes in his First Meditation. When Cavell labels this scepticism in connection with the external world a "lunatic" view, he means to associate it with a pathological human condition. It is a pathology to which Cavell believes we are all susceptible-although he points out that in the normal adult human this would be identified as a passing mood. Importantly, it is a mood, or condition, from which we can escape. Cavell reminds us of what Hume writes in his Treatise of Human Nature when, having given in to philosophical speculation doubting the existence of bodies, he admits that he is off to have his lunch and engage in a game of backgammon. 8 Cavell echoes Hume when he writes that, in connection with external world scepticism it is possible to "exit from the mood". 9 As both Hume and Cavell point out, it is possible to immerse oneself in the possibility of external world scepticism and at the same time envisage an alternative to it. One may, perhaps, here see a connection to something I said in connection with the first consideration-that it takes a God-like wrench to pull the world away from us.
It may be thought that something of Cavell's observation here is captured in a more recent position in epistemology, contextualism. According to the contextualist, we must always be alive to the context in which one asks one's questions about knowledge. It is one thing to ask a question about how we know while sitting in our study, or in the context of a philosophy class, it is quite another to ask it when we are out and about in the world. We can, the contextualist might say, "exit from the mood".
What Cavell goes on to point out is that, in the case of other minds scepticism, it is not so easy to "exit from the mood". Indeed, he goes so far as to venture that radical scepticism with respect to other minds is not such a pathological or lunatic position. That such scepticism is neither lunatic nor easy to escape are related. One can appreciate this if one reflects on the reason that Cavell identifies for why both are the case: "there is no human alternative to the possibility of tragedy". 10 In this connection, Cavell asks us to recall Shakespeare's Othello. But we don't need to get quite so dramatic to appreciate the point. Cavell also reminds us of the ways in which scepticism regarding others is woven into the texture of our everyday dealings with them; it is revealed in "problems of trust and betrayal, of false isolation and false company, of the desire and the fear of both privacy and of union." 11 One might say that it would be "lunatic" to ignore such problems. Of course, the very young child ignores them. But it is nearer the mark to say that she doesn't yet understand these problems and, as a consequence, behaves without thought of them. In the young child this is not to be thought of as lunatic, so much as necessary. It is just such childish trust that helps the child to learn from others; eventually the child learns when not to trust. 12 Cavell considers that someone may try to re-assert a symmetry here by offering the following by way of a counterpart to Hume's game of backgammon: your child enters your study and all sceptical thoughts and fears vanish. But Cavell insists that trying to re-assert a symmetry with this thought experiment should be taken to be the result of a prejudice that external world and other minds scepticisms must be seen to be parallel. He confirms his claim of an asymmetry here by suggesting an alternative reading of what happens when my child enters the philosophical study: seeing the child does not bring "relief from my isolation" but rather brings me "some solace for it"-a solace "not because the child frees me from my speculations about others, but because he is not, as I conceive him now, relevantly other; he merely extends the content of my narcissism". 13 The narcissist is someone who suffers from extreme self-involvement. Cavell is suggesting that this self-involvement may prevent someone from taking the presence of their child as "bringing relief from [their] isolation". It is more likely that the narcissist will end up drawing the other-in this case, the child-into their own self-involved world. 14 Cavell's point is that it is not so easy to snap out of other minds scepticism as it may be to snap out of external world scepticism.
Contextualists do not spend much time considering the case of other minds, but we might reflect on how the position could be thought to be brought to bear on this sceptical problem. The contextualist might point out that, while it might make sense to ask if one's child has a mind in the context of a philosophy seminar, it is not a question we ask in the context where we are busy raising our children. But Cavell has given us another way of viewing the situation. I don't escape scepticism in the context where I am dealing with my children, but, rather, I bring my children into my own narcissistic orbit. If Cavell is right to point out that scepticism with respect to others is more deeply woven into our relations with others than it is with respect to the external world, this could also be raised as a difficulty for accepting contextualism in connection with our knowledge of other minds.
This second observation, like the first, concerns our attitudes towards the different sceptical scenarios. While I believe that the second observation goes deeper than the first, I do not think either of these observations is sufficient to warrant that we approach epistemological issues in these two domains differently. While these observations may not be sufficient in this regard, however, they-and especially the second-may be thought to lead us to consider whether we should be thinking of these two domains differently. I suggest that the following observation-the third and final one-does point to some significant differences between external world and other mind scepticism, which differences should prompt us to think differently about both the formulation of and the response to these two scepticisms.
I now turn to consider a third, and final, reason for thinking of radical scepticism differently in connection with the external world and with other minds. I take this from the work of Richard Moran-although the point is originally due, once again, to Cavell. 15 Following Cavell, Moran draws 1 3 our attention to a very important difference in the formulation of the radical sceptical problem in connection with the external world and in connection with other minds.
If we return to the work of the father of radical scepticism, Descartes, it is clear that the scepticism that most concerns him is one concerning the external world. Descartes does not explicitly raise the sceptical problem in connection with other minds. 16 When considering our knowledge of the external world, the Cartesian sceptic begins by assuming that each of us is, as Moran puts it, "sealed within [our] own circle of experience", and that from this position we strive to know whether our knowledge is experience is 'of' the world. 17 Descartes uses the devise of an evil demon to help us to appreciate that our grip on the world may not be what we take it to be. Thompson Clarke introduced the term "the arch outsider" to refer to Descartes' evil demon. 18 There may be many different formulations of the outsider idea, but the general idea is to imagine a figure whose capacity for knowledge is not limited in the way my (or our human) capacity is limited. 19 We are urged to think of the imagined Outsider as someone who stands outside my experience, someone not limited in the ways that I am thought to be; someone who is able to know what I don't seem to be able to know. The Outsider can stand in a position orthogonal to my ideas about the world and compare them with the things they are 'of'. 20 While I am not in a position to know if everything I experience is but a dream, the Outsider is someone who is able to know this. The idea is a familiar one in philosophy. What Cavell notices, however, is that something odd happens if we try to apply this idea of an Outsider to knowing another mind. Moran sums up the difference by saying that, when we consider the idea of an Outsider in connection with another mind, what we find is that the Outsider we imagine is both too far outside and at the same time not far enough outside. Let us consider these two thoughts in turn.
Consider first the thought that the sceptic is too far outside. We are helped to understand this thought if we return to Cavell's original introduction of the Outsider. Cavell tells us that the application of the Outsider idea to other minds is quite a different matter from its application in connection with the external world because "what I have to imagine [the Outsider] to know is not merely whether a given other is a being or not, but to know something I do not know about how to tell, about what the difference is between human beings and non-human beings or human non-being". 21 We tend to think of the Outsider as just having knowledge of whether something exists -a knowledge the sceptic claims we lack. What Cavell notices is that, in the case of the external world, the Outsider not only knows whether something exists over and above my ideas but she also knows what reality is in contradistinction to my ideas; in contrast to the Outsider, I am not in a position to know either of these things.
In connection with knowledge of other minds, however, we find that the Outsider is only pictured as being in a position to know whether the other has a mind -a knowledge thought to be denied to me. In connection with another mind, there is no analogue for the second piece of knowledge that the Outsider is acknowledged to have in connection with the external world. I don't expect that the Outsider should know something that I don't know about the difference between being minded and not being minded. I do not want to know something about the other that is a complete mystery to me. Indeed, in the case of another, what I want to know is if she has what I have. It is in this sense that the Outsider we image with respect to the external world is "too far outside" when it comes to other minds. While I may resort to imagining an Outsider in order to have a conception of what it would be to know external objects, I recognize myself as an instance of what it is to be the sort of thing in question when it comes to others. I have thoughts and feelings; I am an example of what I am questioning in the case of the other. There is a sense in which, with respect to mind, I am in Insider as well as an Outsider. 22 Let us now consider the other difference that Moran identifies here: the imagined Outsider is not far enough outside. The thought here is that, in so far as the Outsider is in a position to know of others whether they are minded, she is not intended to raise this question in connection with me. But this is precisely what we would have to do if we were able to let our imaginary Outsider be fully outside; we would have to countenance the Outsider questioning what sort of thing she thinks I am. But this, Cavell observes, is a step too far. The Outsider, he concludes, "is not really an Outsider to me. If he exists, he is in me". 23 Cavell considers an analogy with seeing red. He considers that the Outsider discovers that when I say I see red and you say the same, what we see is different. The question now arises, which of us -if either 16 While that problem is not explicitly formulated in Descartes' work, it is still believed that the problem here can be traced back to his work. For a discussion of Descartes on other minds, see Avramides (2001). 17 See Moran (2011), p. 240. 18 See Clarke (1972). Moran reminds us that Cavell and Clarke gave seminars together at Berkeley in the lead up to the publication of The Claim of Reason. 19 Moran (2011) writes, "The point of imagining such a figure is that it tells us something, by contrast, about my own epistemic position", p.241. 20 Cavell (1979), p. 417. 21 Ibid. 22 Moran (2011) writes concerning this, "any restriction on his knowledge is imposed by me, which suggests that I am not allowing any genuinely skeptical fantasy to be released" (p. 244). 23 Cavell (1979), p. 418.
-is right? This question "makes sense". However, when we turn to consider the case of pain, things are quite different. Cavell writes, "But if I and the other do not feel the same when we sincerely exhibit pain, I cannot tolerate the idea that the other might be right and I not. What I feel when I feel pain, is pain". 24 Following Cavell, Moran writes: "I seem forced to place restrictions on what the Outsider can discern, can pronounce upon, even though the whole point of constructing him was to imagine a position for knowledge that was unrestricted, or at least not restricted the way my own position is." 25 It is in this sense that the Outsider is "not far enough outside". Or, perhaps we should say, the Outsider considered in connection with another mind cannot be far enough outside to be an Outsider in the same sense as the Outsider we imagine in connection with the external world. According to Moran, the Outsider that we are trying to imagine is, in fact, the other who we are trying to know. 26 In this section I have outlined three considerations which indicate an asymmetry in the formulation of radical scepticism in connection, on the one hand, with the external world, and, on the other, the mind of another. Moran uses the third of these considerations to argue that the application of the sceptical problem cannot simply be transferred from consideration of the external world to consideration of another mind. The first two considerations may be used to reinforce Moran's point here. 27 But while Moran emphasizes the difference in the formulation of the sceptical problem in connection with these two domains, we can also push the point here further-we can also say why the sceptical problem needs to be formulated differently in connection with these two domains. The asymmetries observed between external world and other minds scepticism may be thought to point to a difference in the way we think about the world, on the one hand, and mind, on the other. While all three considerations outlined above may draw us to appreciate differences in the application of sceptical considerations to different domains, it is the third observation that goes the deepest. It points to a profound difference, not just in our knowledge whether the external world or other minds exist, but in our conception of world and of mind.
It is one thing to say that there are observations that point to differences in the way we think about these two domains-the domain of objects or bodies, and the domain of mind-but it is quite another to explain just what the conceptual differences here amount to. In what follows I aim to say more about these differences. To this end I draw on the work of both Edith Stein and Stanley Cavell. While there are significant differences in the work of these two philosophers, I find in the latter a way of understanding a very important point brought to the fore in the former. The work of both Stein and Cavell can help us to see in just what ways our concepts of body and of mind differ and how this difference leads to a difference in the way we should think about our knowledge in connection with these two domains.

Responding to the Problem
In Section (2) I outlined three considerations designed to alert us to a problem in transferring without modification a sceptical problem formulated in connection with consideration of the world to the domain of minds. Unsurprisingly, these considerations may also be thought to be relevant when evaluating the responses we give to scepticism in these two domains. In this section I look at two prominent responses to scepticism in connection with other minds that can be found in the philosophical literature and suggest that they both suffer from a similar problem-a problem that may be thought to be traceable back to the fact that the responses are to a scepticism that has proceeded first by formulating a problem in connection with the external world and then applying this to other minds without regard to important differences here. Two important points to note before I proceed. Firstly, it should be clear from the outset that by "responses to the problem" I don't always mean "answers"-although some philosophers I consider do present their response as answers. When I write of "responses", I shall simply mean ways of reacting to the problem-and that can include everything from trying to give an answer to trying to dissolve, or even respectably ignoring-the problem. And secondly, while I claim that both responses I consider are in danger of missing the important differences that the lack of parallel in setting up the sceptical question calls to our attention, I allow that one kind of response -or one version of it-may have more scope to accommodate the difference here. Before explaining the error, I briefly outline the two lines of response that I claim exhibit it.

The Inferential Response
The first response I want to consider is sometimes known as Inferentialism. Inferentialism is any view that holds that what I (directly) see is the body of the other and that I know that another has a mind by reasoning from what I (directly) see. Knowledge on this view is necessarily indirect; it is also necessarily only probable. While inferentialism comes in several guises, I shall only mention the two most prominent forms: the argument from analogy and the argument 1 3 from best explanation. 28 Inferentialists accept a traditional formulation of the radical sceptical problem and offer their proposals by way of a solution to that traditionally formulated problem. They account for our knowledge of another mind by appealing to what one can know directly or noninferentially-that is, that the other is moving in certain ways. They take this non-inferential knowledge and combine it with what can be known from one's own case-for example, that I feel pain. The inference that is made on the basis of this information results in an hypothesis to the effect that another also has a mind. In one case the hypothesis is arrived at by invoking analogy (I see the behaviour of the other and I observe in my own case that I am not the cause of this behaviour, so I argue by analogy that the cause of that behaviour must be another mind.) In the other case the hypothesis is arrived at by invoking an argument from best explanation (I see the other's behaviour and, using what I know from my own case, I hypothesise that the best explanation of the behaviour I see is that the other is minded).
The inferentialist accepts the sceptical challenge and aims to respond to it by showing that it is possible to achieve knowledge (or at least probable knowledge) of another's mind by application of a form of reasoning from what can be known directly. As the inferentialist takes knowledge here to be indirect, it is tempting to see this knowledge as a kind of 'second best'. What this means can then be understood by contrast with an imagined 'first-best' case of knowledge. One could think of first-best knowledge as knowledge that is direct, rather than the result of reasoning from what can be known directly. When one combines the idea of first-best knowledge with the idea of an Outsider in application to other minds, one ends up with the idea that, but for human limitation it would be possible to achieve a better position from which to know the mind of another. While I cannot do more than achieve hypothetical or probable knowledge concerning your state of mind, the Outsider can do better; the Outsider is in the imagined position of (somehow) 'reaching into your mind' and coming to know for certain what you think and feel. This gives us a picture of the other's mind as laid out for the Outsider to know-as it would be laid out for me to know, were it not for the limitation in my capacity for knowledge here. 29 What this picture of mind gets wrong is something I shall explain once I outline the second line of response to the sceptical problem.

The Perceptual Response
Perceptualism is a response to the other minds problem which has been explicitly rejected by many philosophers throughout the history of philosophy-at least up until the fork in the philosophical road that produced the two paths of analytic philosophy and phenomenology. It is fair to say that analytic philosophy has largely tended to proceed in line with its Cartesian history when it comes to considerations concerning other minds, while phenomenology has strongly rejected that history. Perceptual accounts can be found in the work of philosophers from Husserl to Merleau-Ponty. For example, the latter writes: I perceive the grief or the anger of the other in his conduct, in his face or his hands, without recourse to any 'inner' experience of suffering or anger. (2003 [1945]: 415) It should be noted that many analytic philosophers of mind also reject this Cartesian history, and one can also find attempts among some analytic philosophers to defend a perceptual account of our knowledge of another's mental life. The jumping off point for a perceptual account of our knowledge of other minds is sometimes traced back to work by John McDowell. In one place he writes: We should not jib at, or interpret away, the commonsense thought that, on those occasions which are paradigmatically suitable for training in the assertoric use of the relevant part of language, one can literally perceive, in another's facial expression or his behaviour, that he is [for instance] in pain, and not just infer that he is in pain from what one perceives. (1998 [1978]: 305) It is important to note that McDowell is not replying to the sceptic when he writes that we can see the pain in another's facial expression. Rather, he sees his work as providing a picture of our relation to the world and to others that, in McDowell's words, makes it respectable to ignore the sceptic. 30 It is natural to ask how it is that we know about the world around us, and McDowell points out that nothing stands in the way of responding that in the case of another's mental states we know them through perception. 31 McDowell believes that a problem would arise for 29 The idea here is that human beings must come by knowledge of another mind by inference because they lack the capacity to know in a more direct manner. Some have considered telepathy as an example of such a direct way of knowing that is imagined to bypass the need for an inference here. For a good discussion of this possibility see A.J. Ayer (1954), p.196. 30 McDowell (1994 writes: "The aim here is not to answer the sceptical questions, but to begin to see how it might be intellectually respectable to ignore them, to treat them as unreal, in the way that common sense has always wanted to" (113). 31  a perceptual account if one introduces a logical or conceptual divide between mind and body-something a Cartesian does and that McDowell is keen to deny. However, while McDowell concentrates on closing the logical gap between mind and body, he pays less attention to the difference between our concepts of mind and of body that may be taken to emerge from considerations such as those outlined in section (2). This is not to say that McDowell does not appreciate these differences, but they are not to the fore in his writing. It may be thought that it is hardly surprising that McDowell does not appreciate the differences pointed up in the considerations of section (2) as he is not concerned to reply to the sceptic. Nonetheless, the differences observed in our concepts of mind and of body may be thought to be relevant even outside of the sceptical challenge. McDowell does not pay enough attention to these conceptual differences when proposing his perceptual account of our knowledge of another's mental states; he does not take care to explain just how perception of another mind differs from perception of bodies. This is not to say that his perceptual account cannot accommodate this difference; it is just to point out that there is work to be done in this connection. 32 While McDowell is often cited as an early proponent of a perceptual account of our knowledge of another mind, it is sometimes not appreciated that Fred Dretske proposed a similar idea 17 years earlier. 33 Like McDowell, Dretske does not aim in his work to respond to the sceptic. 34 Scepticism aside, in order to know there must be, according to Dretske, a defensible account of how we know. In connection with other minds, Dretske gives what for the time in analytic philosophy was a rather surprising reply: we know other minds by direct perception. 35 And Dretske wants to defend the idea that we know about others with minds in the same way we know about bodies in the world. He begins his paper Perception and Other Minds thus: The point I mean to be emphasizing in my attempt to demote the problem of other minds is that, as commonly conceived, our ways of knowing (or one of our way) about other minds (e.g. that his finger hurts) is exactly the same as our way (or one of our ways) of knowing about other bodies (e.g. that his finger is in his mouth). (35) A few pages later in the same paper, Dretske writes: "If there is a problem of knowledge, if there is a problem about how we can see what we commonly purport to see, then it would seem that these difficulties affect both our knowledge of objects and our knowledge of other minds." (35) Note that Dretske in the first quotation says he wants to "demote" the problem of other minds, and that he insists in the second on affirming a symmetry between how we know about bodies and how we know about minds. While Dretske is not explicit about this, the problem he is keen to demote is the traditional sceptical problem about other minds. The only problem that Dretske recognizes here is what Anil Gomes has referred to as "the problem of sources". 36 The problem is to identify the source of our knowledge (in Dretske's words to identify "how we know") and to show that there are no obstacles or challenges to this as a source of knowledge in a particular domain. The source of knowledge that Dretske defends in connection with other minds is exactly the same as the one we use in connection with knowledge of bodies. Knowledge in both domains is directly available through (direct) perception.
While Dretske, like McDowell, may not interested in the sceptical problem, he-again, like McDowell-may be accused of overlooking the conceptual difference that is thrown up by Cavell's consideration of the sceptical challenge. While Dretske sees no difference in the way we think about knowledge in connection with bodies and with other minds, the considerations of sec. (2) lead us to think that there are conceptual differences here that should be reflected in the way we think about knowledge with respect to these two domains. However, while I have suggested that it may be possible to adapt McDowell's perceptual account to accommodate these differences, I am not as sanguine that this is the case with Dretske's perceptual account. The reason is that Dretske, unlike McDowell, does not question any of the sceptic's metaphysical or conceptual assumptions. What Dretske does question, along with other knowledge externalists, is the connection between rational justification and knowledge. 37 Rational justification aside, we can say how we have knowledge-both of bodies and of other minds. But this way with the sceptic leaves the core of his way of thinking about body and about mind in place. It is important to note that, while the perceptual accounts that one finds in the work of McDowell and Dretske are superficially similar, 32 Nor is it to say that it can be done. My point is simply that it has not yet been done. 33 See Dretske (1969) and(1973). It should be acknowledged that, some 50 years earlier, Nathalie Duddingham defended the idea that our knowledge of other minds is "as direct and immediate as our knowledge of physical things"-although I think it is fair to say that the idea did not catch on at the time. See Duddingham (1919), p. 147. 34 But it should be clear that the reasons for not responding to the sceptic are very different for each of these philosophers. I say a little more about this towards the end of this section. 35 For a discussion of Dretske's work here see Avramides (2019b). 36 See Gomes (2011). While Gomes does not consider Dretske in this connection, I think it is clear that this way of characterizing the problem also fits Dretske's work. 37 For Dretske, knowledge is to be thought of in terms of reliability, not reason. See Dretske (1991). 1 3 they emerge from very different philosophical positions. One aspect of this difference is reflected in the fact that, while neither Dretske nor McDowell put forward their positions as responses to scepticism, the reasons each has for this are very different. McDowell believes that the sceptic can be ignored if one understands aright the relationship of mind and world. Dretske allows that the sceptical position makes sense, but deems it to be irrelevant to knowledge if one drops the connection between knowledge and rational justification. It is because Dretske inherits the sceptic's way of thinking about the mind in relation to the world that I am pessimistic about whether his perceptual account can be adapted to acknowledge the conceptual differences here.
Thus far I have outlined two responses to the problem of other minds and suggested that both embody an error. While I have indicated that neither response fully appreciates the conceptual differences between body and mind that the considerations of sec (2) reveal, I have yet to identify why this matters. I now turn to identify the error that one may be thought to fall into if one does not appreciate the conceptual differences here.

Identifying the Error
I want to begin the work of this section by drawing attention to the work of Edith Stein. Stein wrote her PhD dissertation under the supervision of Husserl, and it was published under the title On the Problem of Empathy. Stein, like her teacher, rejected the idea that our knowledge of others comes about as the result of an inference from observed behaviour. And like many in the tradition in which she worked, Stein was drawn to a perceptual account of our knowledge here. But Stein observed an important difference in the application of perception to the domain of bodies and to the domain of minds. In so far as our knowledge of others comes about through perception we must, insists Stein, appreciate that it is "a kind of perceiving sui generis". 38 It is like perception in so far as it is direct; it is unlike perception in the way in which it gives us its object. In order that this difference is not lost, Stein prefers to write of empathy in connection with our knowledge of others. What Stein is keen to emphasize is the way in which our knowledge of the other is always imperfect: there is always a distance between what I am aware of when I empathize with the other and what the other is experiencing; I experience the other as other, as a center of intentionality or as a perspective on the world that is different from mine. 39 Empathy (or a kind of perceiving sui generis) is the way in which we know a foreign consciousness. Stein takes the distance introduced by an imperfect knowledge here to capture the separateness that exists between persons.
With this insistence on an imperfect knowledge of another, Stein is attempting to draw our attention to a very important fact about other minds: that they are other. This inability fully to know another is one of the things to which Cavell also draws attention. I take it to be an appreciation of this that lies behind his remark (quoted above) that "there is no human alternative to the possibility of tragedy". 40 What Cavell means here is not entirely clear. Marie McGinn takes Cavell to hold, as a result of this observation, that radical scepticism is endemic to our lives and cannot be eradicated. I disagree with this understanding of what Cavell is teaching here. 41 Rather, I see Cavell as holding that the retreat into scepticism is a misunderstanding of our condition in relation to others. What the possibility of human tragedy reveals is the otherness of the other. To draw on what Stein teaches, we could say that we are unable entirely to know the other who is separate from me. But while Stein writes of what we are unable to know, Cavell aims to accommodate this separateness without appeal to knowledge. I understand Cavell as saying that radical scepticism is our reaction to the understanding that we are other. 42 We retreat into scepticism and take our condition here to be one of an "intellectual shortcoming". But Cavell takes this scepticism to be a misunderstanding of our condition. According to Cavell, the shortcoming here should not be thought of as an intellectual one but as a shortcoming in how we react to others, how we treat them.
We can, perhaps, come to appreciate why Cavell thinks that the shortcoming here should not be taken to be an intellectual one if we return to the thought that, when one tries to use the Outsider idea in connection with the minds of others, one discovers that the Outsider cannot get far enough outside. Recall that the reason why the Outsider cannot get far enough outside in connection with another mind is that she cannot question or deny my mentality. I am an example of what I am questioning in connection with the other. In this respect, I am an authority with respect to my mentality. And Cavell points out that, in so far as others exist, they also have an authority over their mentality. It follows from this that anything that I say about the other, in so far as she is minded, she is in a position to accept or to reject. With respect to other minds, the Outsider is the other. And Cavell thinks that it follows from this that, just as I can raise questions about another mind, so the other is in a position to 38 See Stein (1917), p.5. 39 Cf. Zahavi (2014), p. 192. 40 Cavell (1979 For similar criticism of McGinn's reading of Cavell here, see Minar (2004). 42 The point comes out clearly in Cavell's discussion of Othello towards the end of Cavell (1979), especially pp, 492-3. raise questions regarding my mind. As Moran writes, "So everyone is Outsider and Insider, privileged with respect to one mind, and for that reason dispriviledged with respect to all others." 43 It is these observations that lead Cavell to write of a duality in the problem of other minds. 44 This idea of Cavell's, that there is a duality in the problem of other minds, is, if true, a very powerful observation. It is also a complex idea. I want to suggest that we can tease out one thread in Cavell's idea here and use it to further understand Stein's observation that perception, as a source of knowledge of other minds, must be understood to be a "kind of perception sui generis". 45 Cavell considers perception as a source of knowledge of other minds, and about it he writes that seeing, in the case of another mind, is not enough. 46 At first glance, it can seem that Cavell's observation stands in opposition to Stein's. But on closer inspection, we find that both are motivated by the same thought. Both Stein and Cavell recognize the need to understand the other's thoughts and feelings as the thoughts and feelings of another-of an individual separate from oneself. Stein does this by insisting that, in so far as we know about others through perception, that perception is limited in the way that our knowledge of the world is not. Cavell does this by insisting that, when we consider perception as a source of our knowledge of another, we must appreciate that there is something more going on here than just seeing. It is in filling out what this something more is that we can appreciate that Cavell, like Stein, is drawing our attention to the otherness of the other.
Both Stein and Cavell appreciate that our knowledge of other minds needs to be understood in a manner that differs from our knowledge of the external world. Stein points out the way in which our knowledge here can only ever be partial. We can begin to get a better idea of Stein's observation here if we contrast our perception of other minds with that of bodies. Stein's idea is that, while I can perceive a body in its entirety, I cannot perceive another's mind in its entirety. In response to this someone may point out that I also cannot perceive a body in its entirety-something also always eludes me here. Consider my perception of a table: in my usual perception of a table, the underside eludes me; of course, I can turn the table upside down and now perceive the underside of the table-but now the upper side eludes me. What this response may be thought to miss, however, is the fact that, in the case of bodies, what eludes me at one time may be brought into view at another. By contrast, in the case of another mind, something always and necessarily eludes me. That which now eludes me cannot be brought into view in the way that the underside of this table can be brought into view; there is no way I can rectify things here. 47 Cavell is less clear about what he wants to say about the way in which the other eludes me. What he is clear about, however, is that the problem here is not an intellectual one. I want to suggest that understanding what the problem is, if it is not an intellectual one, can help us to see (a) why it is not possible fully to know the other, and (b) how we can know what we know about the other while still respecting the separateness or otherness of the other. In other words, I want to suggest that we can use Cavell's work to further develop Stein's.
Let us begin with a rejection of the sceptic's way of understanding the problem of other minds-with a rejection of the idea that the problem here is an intellectual one. In the place of an intellectual shortcoming, Cavell suggests that the shortcoming here is a failure to acknowledge something. According to Cavell, it is not just a matter of what I see-for example, that you are jumping about and clutching your arm-but it is also a matter of what I do in response to what I see. How do I respond to you? Do I reach out to help you? Do I turn away? Do I leave it for someone else to respond? All of these ways of responding fall under the heading of acknowledgment. According to Cavell, acknowledgement plays a role in our knowledge of others. When I encounter another, not only do I need to recognize that she is unhappy-as opposed, say, to being angry, but I also need to acknowledge her pain-acknowledge her. The need for acknowledgement of the other plays an important role in Cavell's work. If I see what you are doing-jumping about and clutching your arm-and I question, as the sceptic does, whether or not you really are in pain, this shows that you are not appreciating the way in which knowing in the case of other minds is not the same as knowing in the case of bodies. In the latter case, seeing is enough; in the former case, seeing is not enough. According to Cavell, as well as seeing what you are doing, I must respond to you, I must acknowledge you. Cavell writes: "my ignorance of the existence of others is not the fate of my natural condition as a human knower, but of my ways of inhabiting that condition". 48 It is at this point that the duality in the problem of other minds needs to be remembered: just as I must acknowledge you, so you must acknowledge me. In this case, acknowledgement by you will take the form of responding to my 43 Moran (2011), p. 241. 44 Cavell (1979 In order to bring out the connection with perceptual accounts of other minds I shall continue to refer to Stein as favouring "a kind of perception sui generis" rather than moving to talk of empathy in this connection. 46 Cf. Cavell (1979), p. 421. 47 This contrast between our knowledge of bodies and of another mind may be thought to be connected to the idea of subjectivity, in contrast to objectivity, that one finds in the work of Tom Nagel. 48 Cavell (1979), p. 432. acknowledgment-say, by thanking me for my help, pushing me away, or by trying to withhold the expression of your pain. In all these activities there is acknowledgement of the other as someone who is minded and, importantly, as someone who is separate. What separates us, writes Cavell, is "a particular way in which we relate, or are related … to one another…". 49 Acknowledgement is not an intellectual business, but an activity. It is embodied in what I do, in how I respond to what I see. This is the 'something more' that perception requires. We might also say, considering that Stein writes, that it is what make perception of mind sui generis. It might be considered to be sui generis for the very reason that it needs to be supplemented by acknowledgment. We can tie up this idea with an understanding that the problem of other minds has a dual nature, thus: just as I can ask about the mind of another, so the other can ask about my mind; and just as I must acknowledge what I see in the case of the other, so the other must acknowledge what she sees in me. In other words, acknowledgement is reciprocal. This reciprocation gives us all the reassurance we need-and all that we can have-that this is the behaviour of a minded being. It is in connection with acknowledgement that Cavell introduces the idea of human tragedy. Where McGinn takes Cavell to cling to some truth in the idea of radical scepticism, I read Cavell as pointing out potential shortcoming in our acknowledgement of others. When Cavell writes that our relations to others involve problems of trust and betrayal (and more) he is identifying some of the difficulties inherent in the acknowledgement which is involved in our relationships to one another; these are the difficulties that the sceptic wrongly takes to be the result of an intellectual shortcoming on behalf of one individual trying to know another.
According to Cavell, the sceptic mistakes an active and reciprocal business for an intellectual one. And Cavell traces this mistake back to the fact that the sceptic sees a symmetry in the way a problem arises with respect to other minds and the external world. By bringing us to understand that there is an asymmetry in the formulation of the problems here, Cavell brings us to see that the sceptic has misidentified the problem we face in connection with others. It is a problem not of knowledge, but of acknowledgement. I may know what pain is, but I must also acknowledge your pain by responding to it in some way-I show that I understand that you really are in pain. And just as we can acknowledge each other's thoughts and feelings, so we can fail to do so. This is where Cavell sees problems of betrayal and, in some cases, tragedy. In failing to acknowledge one another's thoughts and feelings we fail to acknowledge them-we fail to acknowledge the humanity of the other. And, crucially, acknowledgement of the humanity of the other, is acknowledgment of the separateness of the other. 50 Stein observed that, in so far as our knowledge of the other comes about through perception, it is a kind of perceiving sui generis. In saying this Stein wants to capture the fact that I can only ever know the other partially and, in capturing this, Stein aims to capture the fact that the other is other to (or separate from) me. I am suggesting that we can use Cavell's work in this connection to help us to better understand what Stein is trying to capture here. We can think of the sui generis nature of the perception here to be a perception of the other that is accompanied by an acknowledgement of the other. 51 And this acknowledgement is plagued by "problems of trust and betrayal, of false isolation and false company, of the desire and the fear of both privacy and union"-all of which show my separation from you. My separation from you is evidenced in the ways in which our acknowledgement of each other can fail. Cavell writes, "The truth here is that we are separate, but not necessarily separated (by something)…". 52 We can now return to the two responses to scepticism that I outlined earlier and say how the error Cavell identifies in the sceptic's work has an echo in these responses to it. Consider first the Inferentialist. The Inferentialist accepts the sceptic's formulation of the problem, and offers various forms of reasoning to help us overcome it. But the Inferentialist, like the sceptic, fails to appreciate the dual nature of the problem here. The Inferentialist, like the sceptic, pictures an Outsider who is able to see directly the thoughts and feelings of others; these thoughts and feelings are only available to us mortal human beings through some process of reasoning. But if Cavell is right and the Outsider idea, as it is formulated in connection with the external world, does not have application to other minds, then the Inferentialist picture is mistaken. It is not a matter of anyone's seeing into the mind of another. 53 The assurance the sceptic seeks for the existence of another mind is to be found in my response 49 Cavell (1979), p. 369. 50 In this connection Cavell also writes of human finitude (cf., for example, p. 493). 51 It is important to flag up a potential misunderstanding of what I am saying about Cavell on acknowledgment. I have written of acknowledgment as a "something more", as something that "accompanies" perception, and the like. The relationship here between what I know and acknowledgment is a delicate one and requires much more attention than I can give it in this paper. In the hopes of steering the reader in the right direction, I offer this quote from Cavell: "acknowledgment 'goes beyond' knowledge, not in the order, or as a feat, of cognition, but in the call upon me to express the knowledge at its core…." (CR, p. 428). I am indebted to Samuel Williams for pointing out the need to avoid misunderstanding in this connection. 52 Cavell (1979), p. 369. 53 While I speak here of "seeing directly" and "seeing into" the mind of another, I mean to distinguish this imagined idea from the Perceptualist position outlined in sec. (3.2). The "seeing into" that I am sug-to the other-a response that is, in turn, reciprocated by the other. This reciprocated activity embodies the assurance that we seek. The activity which yields up this assurance is what Cavell labels "acknowledgment". It is the failure of acknowledgment, not of reason, that accounts for the problem of other minds. It is not that I need to reason that the other is like me in some way; rather, I need to acknowledge that I stand in a very particular kind of relationship to the other, and that the other stands in this relationship to me. It is this relationship that we need to understand.
The Perceptualist response to the problem of other minds is complicated by the fact that (as I explained above) different perceptualists formulate their account here against very different ways of thinking about mind in relation to world. On the one hand, there is the Perceptualism developed by Dretske which, in its failure to question the logical gap accepted by the sceptic, may be more easily seen to be in error. It is hard not to see Dretske as falling into the same error as the Inferentialist. The only real difference between Dretske and the Inferentialist is that Dretske believes that perception can reach across the gap (from the body of the other to the mind of the other) without the aid of reason. 54 On the other hand, there is the Perceptualism of McDowell which, because of its rejection of the Cartesian metaphysics that underlies the sceptic's stance, is harder to assess with respect to the identified error. What we can say is that McDowell wants to be able to say that we know -through perception -that another really is, for example, in pain. He insists that this is something I can see. In the words of Cavell, it would appear that McDowell thinks that perception is enough. 55 That seeing is enough for knowledge is a thought that both (McDowell's and Dretske's) Perceptualist accounts may be thought to share. If Cavell is correct, however, seeing is not enough; if Stein is correct, if we are to think of knowledge here in terms of perception, we must think of this as a "kind of perception sui generis". It may be that perception plays some part in what we know about another, but without appreciating the relationship we bear to one another this perception lacks the all-important dimension of humanity. Cavell writes: "my ignorance of the existence of others is not the fate of my natural condition as a human knower, but of my ways of inhabiting that condition; that I cannot close my eyes to my doubts of others and to their doubts and denials of me, that my relations with others are restricted, that I cannot trust them blindly….my position here is not one of a generalized intellectual shortcoming." 56 To understand the relationship here is to understand that the other is other to me. We are separate beings, united in our humanity.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.