This article is an appreciation of Steven Burns as an interpreter and teacher of Ludwig Wittgenstein. As a starting point to contribute to a conference in Burns's honour, we elected to reread his article “If a Lion Could Talk” (Burns, Reference Burns1994), which concerns Wittgenstein's opaque and enigmatic remark, “If a lion could talk, we wouldn't be able to understand him” (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, PPF §327).Footnote ¹ Burns situates the remark in its proper context of Wittgenstein's investigation of inaccessible (private) mental states, specifically, the role played by such states when posited in our convictions about meaning and understanding. This context lets Burns show us how the remark figures in Wittgenstein's practice of philosophy as a means of alleviating misunderstandings. Doing philosophy with this aim in mind, Burns brings a steadied guidance and a depth of insight to a philosophical text open to much misunderstanding. As we discussed the problems and arguments Burns raises, we were astonished to have chanced upon the very puzzles of misunderstanding about which he warns us and with which he helps us. We gratefully experienced anew the effect of Burns's teaching, and the present article is a testimony of that effect.

In what follows, we illustrate the type of misunderstanding that Burns identifies and alleviates with a short account of a discussion we had in preparation for writing this article. We then describe and interpret Burns's treatment of the lion remark, showing how he situates it in Wittgenstein's philosophical practice of resolving misunderstandings. We discuss Burns's teacher Peter Winch (1926–1997) and use features of Winch's and Burns's work to offer a description both of our specific misunderstanding, and of a source of philosophical confusion more generally. We conclude with some impressionistic remarks about Socrates in a short discussion of the difficulty of the philosophical technique and activity Burns demonstrates and recommends.

Burns begins his meditation on Wittgenstein's remark with a brief critical review of several interpretations. First, perhaps Wittgenstein is insufficiently imaginative, and a talking lion that addresses him directly would be perfectly understandable. Or, perhaps Wittgenstein is making a claim about the impossibility of communicating between species, declaring that no sort of linguistic transaction is possible between human beings and lions. Finally, perhaps he means that even if we imagine a lion so sophisticated in its linguistic ability, say, to be calling the play-by-play of a hockey game, his distinctive leonine take on the sport would be so inaccessible such that we couldn't appreciate it.

At any point so far, a reader may think it impossible to test the strength of these claims since no talking lions are forthcoming. But, if we found the talking lion, we could settle the matter and show whether Wittgenstein's claim is true or false. It is easy to incline toward taking the remark as a proffered and testable hypothesis, as a remark that prompts us to look for cases of confirmation or counterexamples. It is as if Wittgenstein's aphorism is challenging us to do the careful empirical work needed to reveal the lion's mental state that lies hidden behind its spoken words. The remark thus appears to be drawing our attention to something hidden about the lion's distinctive mental life.

In preparations for this article, we were among those readers who considered testing Wittgenstein's claim. We had occasion one evening to talk about our project with Daniel, a friend who wanted to know what we had been up to. We introduced a variation of one of Wittgenstein's remarks: a dog cannot expect his master home the day after tomorrow (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, PPF §1). Immediately, Daniel responded that his dog, a very smart dog, could be trained to have this expectation. Knowing that this was the improper response, we gainsaid Daniel's claim. We said Daniel's dog could not be trained to have this sort of expectation. Daniel stood his ground. Before we knew it, we found ourselves talking about animal psychology and the contested lines that separate human beings from other animals. Wittgenstein's assertion stood before us, guiding the discussion, as if it was a proffered and testable hypothesis, a quantified categorical truth claim about dogs: dogs are not the sort of creature that can expect something the day after tomorrow. By the end of our conversation, Daniel despaired that we strange Wittgensteinians must believe, wrongly, that dogs have no inner mental states. We denied this ascription of belief, but we were at a loss to explain ourselves differently. We authors later confessed to one another that each of us heard faintly Wittgenstein's exasperated remark: “And now it looks as if we had denied mental processes. And naturally we don't want to deny them” (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, §308).Footnote ²

The following day, we began to analyze our disagreement with Daniel. We were struck by how quickly he had challenged Wittgenstein's claim, as if Wittgenstein (and we) were making an assertion about his dog's substandard intelligence. We knew this wasn't the case. We also knew we weren't denying the validity and importance of the work to be done in animal psychology. We wondered how the discussion got away from us so quickly. If our disagreement had carried on into the night, we might have found ourselves at Daniel's apartment closely examining his dog and wondering whether some instance of her behaviour counted as being in an expectant state. We might have established criteria for the dog's expecting Daniel's arrival home to contrast with criteria for her being generally excited. We may have spent time debating whether it is possible for the dog to be excited if she's not excited about some particular future event. And we would have thought: if only Daniel's dog could talk. She would tell us about the nature of her expectations. And maybe we'd be a little better positioned to assess the truth of Wittgenstein's claim about talking lions.

In Burns's article, we find a remarkable account of what happened in our disagreement. He shows us how to go slowly, to attend to the context of our target point, to let the context help reveal its aim, and to let the aim of the remark further invite reflection on ourselves in philosophical discussion. First, Burns displays the curious remarks about animals in the wider context of Philosophical Investigations Part II, §xi. Wittgenstein is there examining the concept of subjective privacy and its role in our concepts of thinking and meaning. His remarks about dogs and lions come in the context of examining an inclination to treat the meaning of our words as depending on inner states and processes. Next, Burns draws our attention to the immediate context of the lion, specifically, a remark showing Wittgenstein considering the meaning of the claim “What is internal is hidden from us” (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, PPF §323). Then, the following remarks appear leading up to the lion:

Let's take each of these in turn. With §324, Wittgenstein calls to mind an instance in which he (and, he thinks, humans more broadly) are not liable to scepticism about someone's inner state. When someone is writhing in pain before us, we may be inclined to a variety of responses, but it would be quite a callous or dissociated reaction to doubt that the person is in pain. This is not to say it is impossible to summon that doubt. Rather, the remark invites consideration of a case in which we do not doubt. Readers’ sensibilities may vary.

With §325, Wittgenstein reminds us of an expression he uses and thinks others use, namely, that a person can be transparent to us. A parent, for example, may say they know the motives and intentions of their own child better than the child knows themselves. Wittgenstein immediately contrasts this kind of case with another case he says is important to his considerations, namely, that sometimes one human being can be an enigma to another. His model here is cultural misunderstanding, as when we understand the language spoken in a culture foreign to us, but nevertheless find ourselves at a loss to understand what a person or people in this culture mean with every expression and gesture. He adds that this misunderstanding is “not because of not knowing what they are saying to themselves,” which is to say, what we are misunderstanding or missing here is not found in some inner, hidden subjective space.

In §326, Wittgenstein introduces the term “picture,” which he clarifies as “the convincing expression of a conviction,” and which as expressed “does not give reasons for the conviction.” His example of a picture is “I can't know what is going on in him.” This recalls §324, in which we are invited to consider the case of someone's writhing in pain. In that case, we would not say “I can't know what is going on in him.” We are not in the grip of that picture, that conviction so expressed. Wittgenstein mentions this picture here in the context of examining a philosophical conviction that misunderstanding another person is explained by our ignorance of what goes on inside their private subjectivity. This conviction is one source of the thought that lions and dogs could make themselves understood if only they could express their subjectivities in language we understand. This conviction also inclines us to challenge the truth of an assertion about the limits of non-human animal cognition. The assertions are thus cast as hypotheses that could be confirmed or disconfirmed if only we had more data on the inner life of some lions and dogs. The assertion “I can't know what is going on in him” need not be proffered with these convictions and inclinations in the background. But taking it standing alone, the reasons for its assertion are “not obvious.”

These remarks frame Burns's treatment of the talking lion at §327. Burns believes this material in PPF Part XI “represents a later development and reworking of what is standardly-known from the ‘anti-private-language-argument’ of Part One. Any understanding of the lion will be inadequate which depends on an unrevised distinction between inner and outer” (Burns, Reference Burns1994, p. 4). Burns thus presents Wittgenstein's discussions of seeing aspects, “seeing as,” and the experience of meaning as a new development of Wittgenstein's re-engagement with the philosophical picture that thought and meaning are essentially private.

The private language argument of Philosophical Investigations interrogates how shared natural expressions of sensory experience figure in human learning and use of certain sensory concepts, such as pain, sound, and colour. The material in PPF Part XI, however, moves to consider experiences connected with techniques that can be mastered, such as whistling a musical theme with correct expression. For experiences attendant on technical mastery, we find experts who notice “fine shades of behaviour” to distinguish the mimic from the master. Burns notes that our concepts of meaning and understanding share this feature with our concept of technical mastery: “A thought, too, is connected with the possibility of its expression, and normally that will not be natural but will depend on a learned practice” (Burns, Reference Burns1994, p. 6). Someone may show understanding of a subtle point of detail by offering a nuanced and aptly timed expression of that point. The understanding here is not explained by appeal to an especially refined mental state, but rather in virtue of the circumstances of the expression:

With this analogy in place, we can consider how expressing a thought involves mastery of a technique (or indeed many techniques). The technical mastery is an achievement in language or other signs, a highly social activity of communication involving conventions, expectations, and norms of correction. When we stipulate that the lion talks, we “begin to need to refer to other members of his linguistic community” (Burns, Reference Burns1994, p. 11) to make sense of what we are imagining. We thus add a new dimension to the lion, a dimension “which is inevitably social [… and] it is here, in what is neither private experience nor public language” (Burns, Reference Burns1994, p. 11), that we must look to make sense of the talking lion.

Wittgenstein's talking lion is a master of a technique. His mastery is acknowledged as such by participation in what Burns calls “social space.” For Burns, this social dimension of our lion's imagined life is both a condition of talking with him and the source of our misunderstanding him. Burns writes:

So, even if we understand the language of the lion, a further dimension is lost to us. Burns reminds us that Wittgenstein's discussion of perception plays a role here. Just as a perceptual aspect of a gestalt image can be unavailable to us, so too can the lion's meaning be unavailable to us. Perceptual unavailability is a widely shared experience. Sometimes we don't see, or taste, or hear a nuanced quality someone else reports to have detected. This person reports and describes it in myriad terms and expressions. We may use such cases to compare with cases in which we say someone else's thoughts and speech are opaque or incomprehensible. With the difference cast in terms of technical mastery of a practice, explaining these unavailable experiences does not depend on positing a private, hidden space where we'll find the source of our misunderstanding.

We may further consider cases of perceptual unavailability with cases of misunderstanding people of a culture foreign to our own. This comparison invites us to consider that if a lion could talk, we'd have at least as much difficulty understanding him as we'd have understanding someone from another culture. This is not a claim about the inaccessible mental state of the lion, nor of the person whose culture is foreign to us. When we are unfamiliar with the norms of a people, we don't know what their gestures invite and solicit from us. When musicians don't know the expression proper to a piece of music, the written score can only help with some of the unknowns. They may look for patterns in the score, but they will also call to mind analogical cases (looking at how similar pieces of music are played), draw on other modalities of experience (emotional states), and perhaps read some relevant fiction and history of the culture to gauge audience expectations. Much of that is not in the printed score, just as much of the cultural background of a sense of humour is not in the sentences of the joke. If musicians don't have these aids to interpretation already of as a matter of course, it will take a lot of immersive research to do justice to the technical mastery needed to understand the piece.

When we misunderstand the meaning of signs, written or spoken, it can be helpful to think of what we lack in terms of what Wittgenstein calls a “perspicuous” or “surveyable” representation (übersichtlichen Darstellung) (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, §122). Wittgenstein introduces this term in Philosophical Investigations in the context of its relevance to his treatment of misunderstandings:

Wittgenstein's terms here convey a sense of an engaged activity: “surveyable,” “seeing connections,” “finding and inventing.” Wittgenstein is describing the goal (a surveyable representation) of a technique (finding and inventing intermediate links) for clarifying misunderstandings. These intermediate or “mediating” links connect what we understand with what seems unavailable to us. The job is to “render surveyable” (übersehbar zu machen) the situation and circumstances of the misunderstanding (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, §125). And Wittgenstein describes this representation less in terms of an inner mental state and more as a competency with techniques to assist in noticing connections.

Wittgenstein first wrote about the idea of a “surveyable representation” in the early 1930s in connection with his reading parts of The Golden Bough, James George Frazer's anthropological study of religions, rites, and magic. In his notes on Frazer's text, Wittgenstein objects over and over to what he perceives to be Frazer's failure to understand many of the tribal rites and practices of the people he describes. His general complaint is that Frazer interprets and presents his subjects exclusively in the terms of his own contemporary, British, and scientistic account of nature and culture. The result is that

The priest-king of Nemi was ceremonially killed at his physical prime, which is to say at a relatively young age. Frazer explains this by appealing to the Nemi tribal peoples’ belief that the soul of this king would transfer in its full robustness to the next king, and so it was important not to let the king become decrepit in mind or body before transferring the soul. This explanation is supposed to reveal the practice as terrible, horrific, and to that extent, irrational. Wittgenstein agrees that the practice is terrible and horrific. But his complaint is that Frazer's analysis makes it seem as if the explanation is what reveals it to be terrible and horrific. Frazer's strategy thus lends itself to a dangerous kind of thinking about ourselves, which is that the practices of people in the past were “pieces of stupidity,” and thus not exercises of rationality, not expressions of impulses still manifest and multiform across human experience. Frazer's explanation distances us from the tribe at Nemi. To that extent, it refuses understanding of tribal practices of the human past and thereby of ourselves in the present (and, possibly, of projections of ourselves in the future). What Wittgenstein believes is more honest, less misleading, is for us to take account of such practices and let them teach us about ourselves. One lesson certainly may be “such an act is stupid, insofar as it violates so many of our contemporary sacred beliefs.” But another lesson may be, “look for the impulse in us, find an analogue in our lives, for where this sort of practice shows itself in us.” Do we find it? If yes, what is it? If no, what would have to change about our life to find such a practice as this? What are we inclined to say about such a change, and what does it show about the role of this sort of practice in our conceptual and social life?

We learn from Burns that this attitude is germane to appreciating Wittgenstein's lion remark. Burns goes on to show that he learned from his own teacher, Peter Winch, to pay special attention to noticing unwitting affinities and unnoticed convictions, and to look for meaning where it can be found. Winch's work in philosophy may be broadly characterized by its distinctive focus on problems of understanding and misunderstanding, centred often on the methods of the social sciences in their approach to and analysis of world cultures (Hertzberg, Reference Hertzberg and Edelman2009). Burns discusses one such example from Winch's (Reference Winch1964) own work, namely, an exchange between Winch and Alasdair MacIntyre on the subject of understanding cultures other than one's own. MacIntyre considered a tribal culture whose members believe their souls are embodied in a small piece of wood that each person carries with them. The tribespeople believe that each person's soul is in the wood, “and if they lose it they anoint themselves as though for burial” (Burns, Reference Burns1994, p. 10; see also MacIntyre, Reference MacIntyre and Hick1964; Winch, Reference Winch1964). MacIntyre understands enough of their vocabulary to ask them what motivates their practices and ways of life. Nevertheless, such people couldn't be thought to have MacIntyre's own conceptions of evidence, of what's reasonable, of what's an obvious fact, and so on. Western scientific conceptions commit him to a particular way of looking at the matter, and from that vantage point, there can be no soul in the wood. The belief has great significance in its own terms — which is to say, in the terms of the practicing community; but these are terms that do not survive a conceptual reorientation into MacIntyre's (and modern social scientific) interpretive analysis. So, even though MacIntyre understands enough of the language to describe the account of the soul in the wooden totem, he occupies a conceptual position that makes impossible anything like his coming to understanding what they believe.

Winch was concerned that MacIntyre's conclusion amounts to situating the tribespeople's belief as “thoroughly incoherent,” and thus motivating the conviction that “we confront a blank wall here, so far as meaning is concerned” (MacIntyre, as cited in Winch, Reference Winch1964, p. 323). It's not that MacIntyre treats the tribespeople as if their belief is “a piece of stupidity,” as Wittgenstein believed Frazer to have done in the case of the tribe at Nemi. Rather, Winch's concern is that arriving at “a blank wall” in our search for meaning should be an invitation to turn around and look somewhere else, as Wittgenstein wrote in PPF §325, for how to find our feet with these people.

Another and more literal translation of Wittgenstein's expression in PPF §325, “Wir können uns nicht in sie finden,” is “We cannot find ourselves in them.” Burns describes how Winch intervenes to reconnect MacIntyre with these other human beings. “Winch helps the person in MacIntyre's position by examining something appropriately similar which does make sense in MacIntyre's own form of life” (Burns, Reference Burns1994, p. 10), namely, an engagement ring that is also a family heirloom. “The loss of such a ring,” says Burns, “betraying apparent carelessness, may well provoke remorse and gestures of mortification” (Burns, Reference Burns1994, pp. 10–11). Winch writes: “there might be a parallel here to the aboriginal's practice of anointing himself when he ‘loses his soul.’ And is there necessarily anything irrational about either of these practices?” (Winch, Reference Winch1964, p. 323). Winch finds an intermediating case between MacIntyre's life and the life of the tribespeople. Thinking about cases of his own attachments to material objects may serve MacIntyre as helpful objects of comparison, as shared prototypes of human life that overlap with a practice he finds alienating. Thus, attending to such cases may eliminate MacIntyre's puzzlement about what the tribal gesture could possibly be. Noticing the similarities between the cases dislodges a conviction that commits us to a particular way of looking at the matter, namely, that what explains this cultural practice is a set of beliefs that does not in any way correspond with features of our own. MacIntyre is eager to correct the misunderstandings and biases inherent to bringing one's own conceptual training to the study and understanding of another way of life. Winch offers a path to understanding by acknowledging a practice in MacIntyre's own life that may well allow him to find himself in the tribespeople, a sense of his own soul in a sacred object.

In recounting Winch's discussion of this example, Burns presents MacIntyre as being in the same position as those of us who attempt to understand the talking lion. Facing the people of a culture he finds to be alienating, MacIntyre understands the words of their language but he can't understand the people. Just so, we understand the words of our talking lion, but we cannot understand him. The suggestion is that we might helpfully characterize the misunderstanding of both situations in the same way: we occupy such different ways of life that we cannot understand the lion in the same way that MacIntyre cannot understand the tribal custom. As Burns puts it, “the talking lion illustrates our finding someone else unfathomable” (Burns, Reference Burns1994, p. 9). Wittgenstein offers it as an exaggerated case of a more ordinary and familiar experience of misunderstanding one another.

Burns writes that the sort of misunderstanding Wittgenstein has in mind is “a very important kind of failure” (Burns, Reference Burns1994, p. 11), and decidedly more common than we might imagine. “It extends,” says Burns, “to misunderstandings between sexes, as when a man is unable sufficiently to understand what pregnancy is like for his wife. It extends to intergenerational ‘communication gaps,’ […] it includes much political [and] many moral and religious misunderstandings” (Burns, Reference Burns1994, p. 11). MacIntyre cannot find his feet with the people who carry their souls around their necks, and Winch's suggestions offer him a foothold. This exchange, Burns says, “exemplifies the sort of misunderstanding and aid to understanding which are of central importance to Wittgenstein's view of philosophy of language. If a lion could talk it is this sort of problem, many times multiplied, which we should have in trying to understand what he meant” (Burns, Reference Burns1994, p. 11).

Wittgenstein's remark is an imagined case meant to serve as an object of comparison for convictions and tendencies in our thinking.Footnote ³ What is the lion remark compared with? Cases in which we talk of misunderstanding someone by appealing to a hidden mental state or some otherwise-inaccessible subjectivity. What is Wittgenstein doing with this comparison? Reminding us of cases of not understanding someone's meaning, imagining a limiting case with a non-human animal, and revealing the similarity of that imagined case to our available and usual misunderstandings of our fellow human beings. What is the goal? In the context that Burns brings us through so carefully, the goal is to alleviate the temptation to explain these misunderstandings by appeal to a hidden subjectivity that just cannot get across in the language. The danger of such an explanation is its tendency to block communication, to shift the burden of understanding one another to a new vocabulary that will one day help us figure out the problem. We puzzle about misunderstanding others, and we may be possessed of a conviction to posit a hidden, subjective mental state to locate the source of the misunderstanding. Wittgenstein's talking lion is an invented intermediate link between our conviction to talk of private subjective feeling and our misunderstanding someone's meaning in a language we understand. Presented with this link, we may find ourselves reconciled to a possibility that had eluded us, namely, that we may be missing something that is available and on display from a different approach.

The implication, then, is that we may engage in this clarifying work to help us understand the talking lion. Humans and lions do not inhabit completely incommensurable ways of life. We share basic necessities for finding food and shelter, protecting our young, enjoying the warmth of the sun. Surely, with enough time and effort, we could understand, even if, as Burns says, “the shared practices and common agreements needed for normal understanding […] would take years to build up” (Burns, Reference Burns1994, p. 8). Burns, following Winch's and Wittgenstein's ways of handling misunderstanding, provides a model for how we could proceed.

Our misunderstanding with Daniel is an example of the sorts of misunderstandings discussed and treated by Burns, Winch, and Wittgenstein. We set out to demonstrate a treatment of linguistic confusions and fell into the very misunderstanding the treatment is proffered to resolve. Dominant or otherwise usual habits of uptake ran interference with our attending to cases that changed the role of the remark. Winch has a note on this:

The philosophical picture, the conviction, that misunderstanding is governed by as-yet ungained access to a subjective space, prescribes or legislates a kind of approach to the remarks we make. It is a picture that motivates a tendency to postulate and theorize about subjectivity, locks the discussants into the adversarial attitude of debate and criticism typical of academic philosophy, and primes us to receive Wittgenstein's claim as inviting our adherence or dissent. When we challenge its veracity, we cannot at the same time allow the remark to guide our thinking to familiar or even remote cases and conditions of misunderstanding.

It is perfectly sensible to take Wittgenstein's remarks about the talking lion and expectant dog as a testable hypothesis about a hidden mental state. But, when we do this, we miss Wittgenstein's point. We miss the sense in which it is a remark inviting reflection on our criteria for the meaning of expecting and the conditions of talking. It presents a case for testing our inclinations to speak about inner processes and meaning, intention, and expectation. Crossing these two ways of taking the remark is what characterizes our misunderstanding with Daniel. We began with an instance of what Wittgenstein calls the “first step” that altogether escapes notice. He writes:

The first step which escaped notice was for us to have taken the dog remark as a claim for which counterexamples and requests for inductive justification are appropriate next moves in the game. One conception of what it means to learn to know a process better is to go inside a thing and examine a hidden mechanism. Treating “expecting” as a mental state invites this sort of inquiry, perhaps in neurobiological terms. Once that conviction seizes us, we're now beholden to a different set of commitments given by the empirical sciences. Just like that, we're on the hunt for a neurologically granular account of “expecting” and an even finer-grained account of the mental state “the day after tomorrow.” What was the decisive moment in the conjuring trick here? We told Daniel, “No, your dog could not expect you home the day after tomorrow.” It seemed quite innocent, as we intended to direct Daniel away from an empirical investigation so as not to lose sight of the point of Wittgenstein's remark. But, in denying Daniel's dog could expect, it looked as though we were denying something about her mental life, something hidden from only the most acute scientific observation. Daniel's defence of his dog had changed the rules of our conversation, and we hadn't noticed.

Here is a way to think about the remark “a dog cannot expect his master home the day after tomorrow”: we have rough and ready criteria for describing a dog as expectant. We also have criteria for describing a dog as expecting her master home. What would show a dog to be expecting her master home the day after tomorrow? The question invites us to reflect on the rough and ready criteria we have for saying, of one another, that we expect an event the day after tomorrow. We notice how facility in language use overlaps with criteria such as knowing the days of the week, referring to a calendar, and these too connect to broader customs with the concepts of time and measurement. We may be disinclined to call these criteria “mental” or “private.” And this may loosen the hold of the picture of thought as an inner mental state. Wittgenstein's remarks are intended as cases that may help disincline us to speak about “expecting the day after tomorrow” as an inner mental state. Thus disinclined, we may survey any remaining reasons for thinking that Daniel's dog expects him home the day after tomorrow. Without positing a hidden mental state doing the work here, the expression stands in need of different reasons. We may well go in search of these, or, we may incline to think that our only reasons have vanished. Such is the effect of the investigation Wittgenstein recommends, and achieving this effect is the purpose of the remark about dogs and expectation.

This practice helps us to notice how and when we should have changed tactics. When Daniel challenged the truth of the claim about his dog, we should not have contradicted him. We missed the first step in the conjuring trick. Winch offers a good way to describe the unnoticed trick, an account of the misunderstanding that derailed us:

The remark “If a lion could talk, we wouldn't be able to understand him” is an elucidation about the nature of reality. One way we speak about reality is with factual assertions the truth of which can be challenged by counterexamples. But elucidating can also be revelatory of something real we already know or are capable of noticing, namely, that for all humans share with lions and dogs, there are situations in which communication, mutual participation in one another's goals and lives, reaches an end. Our tendency to prioritize the first kind of assertion interrupts our availability to the remark as an elucidatory reminder of the second kind of case.

In a careful handling of a strange remark, Burns has shown us something central to Wittgenstein's practice of philosophy, namely, that Wittgenstein's remarks alleviate misunderstandings of the sorts we've discussed throughout. The lion remark is one example of many designed to facilitate this. Moreover, there is, evidently, no rote and failsafe procedure for this activity of finding and inventing intermediate links. If anything, it involves intense self-scrutiny with an aim to noticing inclinations and tendencies in thinking and speaking.

In closing our appreciation of Burns's treatment of the lion remark, we offer a somewhat impressionistic notion: Burns's account shows how Wittgenstein's procedures are deeply Socratic. Whether or not Burns would incline to this comparison, we think he would agree that philosophy, for both Socrates and Wittgenstein, is primarily, as Wittgenstein wrote, work on oneself:

Socrates, at least in the early dialogues, patiently worked to recall his interlocutors to what they mean. What is courage? What is virtue? Of course, Socrates presses for a particular kind of definition that Wittgenstein eschews. But the practical effect of Socrates's interventions is not all that different from the effects we've described here in cases from Wittgenstein, Winch, and Burns. Socrates's interlocutors are brought, usually with painful difficulty, to realize that they misunderstood what they believed they understood. As with Wittgenstein, these misunderstandings are not to be resolved by further and more refined empirical investigations; they are resolved, if they're resolved, by paying attention to the successes and failures of our language. Socrates and Wittgenstein are always drawing us back to what we're inclined to say, what we're tempted to say, what we want to say. It's as if, in this work, our only data is always ourselves, the concepts we have at our disposal. Each is involved in a therapeutic investigation that proceeds by paying attention to those pictures that control how we engage in philosophical discussion. We've seen how Wittgenstein characterizes these pictures as convictions. Our delusion, as we've seen, is to suppose that these convictions express an insight, such as that I cannot know whether someone is really in pain. But Wittgenstein reminds us how strange it would be to insist that this is in fact an insight. The kind of examination we find on display in Socrates, Wittgenstein, and Burns, is so nicely characterized by Stanley Cavell. It is an examination

One such prejudice may well be that we can, given the proper access to a subjective mental state, understand a talking lion, or be confident that my dog can expect me home the day after tomorrow. Part of determining how a remark is taken up is understanding the person (often ourselves), their habits and presumptions of conversation and inquiry. Like Socrates handling his interlocutors, the methods and examples a teacher uses, as well as which of these are effective and which fail, are determined only partly by the teacher's sensitivity and availability to the many possible meanings of our words and experiences of our lives. The rest, in large measure, is determined by the character of the student. Lars Hertzberg, writing on Winch, philosophy, and disagreement, tells us:

We are moved by Hertzberg's invocation of those powerful “models” without whom we'd be left without an example of what it means “to be seriously engaged with the issues.” Of course, he means teachers. Wittgenstein wrote: “Language is a labyrinth of paths. You approach from one side and you know your way about; you approach the same place from another side and no longer know your way about” (Wittgenstein, Reference Wittgenstein, Anscombe, Hacker and Schulte1953/2009, §203). This is also true of work in philosophy. We are always liable to assimilate one another's remarks into different and competing logical spaces, based on our temperament, our training and our teaching, and our understandings of the goals of philosophical discussion. Steven Burns has been a steady, trustworthy guide through the labyrinth. We're grateful for his example and we believe his patience comes in part from his own expressed conviction: “There is no single place from which all philosophy begins, and likewise no single place where all questions come to their natural end” (Burns, Reference Burns1994, p. 5).