Metaphysics abandoned, realism evaded

0:00 I'm sorry. I feel like I'm making my plan. So last time we talked about of Plum's paradigms and criticize the metaphysical attempts to solve it, to get out of it, specifically an encourse to anti-nominalism, to the idea that we have to postulate the structure on the side of nature in order to get out of this paradox. I'm going to present the paradox again because, as I pointed out in the fashion period, in fact by James Ladyham, I can't solution, because I have a problem too. The paradox also presents a problem for me. When I state my empiricist position in the philosophy of science, I make a distinction between empirically adequate and true. I say the theory empirically adequate is saved with phenomena, which means that it has to be true in part, but not necessarily entirely true, and that the aim of science is to provide us with empirically adequate theories. Not necessarily true ones. If Oppen's argument can be taken on face value and its conclusion accepts it, then it follows that all empirically adequate theories are true. So the distinction disappears from my position to be moving on. So I also have to be active with the argument. And I won't do that, but I put all my cards on the table. I've used three main arguments in the lecture so far. And I'm going to use basically those same arguments, all three of them, in this lecture.

2:30 So I hope to remind you exactly what they are. The first one I'll spend a few minutes on, the other two I'll more or less just remind you of them. So, the three arguments I try to give in mnemonic devices, so you can remember them. The first one I call, when I say the last step, that's the most important one. And it goes something like this. You remember that right in the first few lectures I talked about a theory or a model representing the phenomena or nature, and about the difficult question of what exactly it means to say that the theory or the model fits the phenomena or the domain of nature that it's meant to represent. because the model is a mathematical structure. And while we know how to compare mathematical structures, we talk about functions and relations, it's not clear what it means to say that there's something like that between a mathematical structure and something that is not a mathematical structure. And my response to it was to say that the theoretical model gets compared with the data model. And the data model is at the same time itself a phenomenon, because it's reduced by an experiment or by an observation procedure. And it is also a structure, it could be like a diagram or a list of numbers or something of that sort. So you have a mathematical structure and the theoretical model gets compared with the data model, so one structure compares Now, the question that the metaphysician won't ask is obvious. He won't say, well, how do you just push the problem one step back? Isn't that a last step that you are ignoring? Namely, the relation between the data model and the phenomena on nature are meant to be fit. Now, here's my response. Of course, he thinks that if I try to answer that question, I won't have to become an anti-nomelist, I won't have to postulate my physical structure in nature or something of that sort. But here's my response. If there's a problem there, it's got to be a problem really across the board

5:00 for any kind of representation, including, for example, simple observation reports, into our, like, pretty small data models, or just descriptions of what the things are like. Now, I will grant immediately that there's a big difference between asking the question, does the theory fit the description of the phenomenon? That's one question. And the question, does the theory fit the phenomenon? And those are two logically distinct questions, they are logically, practically distinct, simply. If that description is my description, then for me, the two questions amount to the same. I cannot possibly answer the two questions differently, we know that before then. So there is no further step. If that description is not my description, well then the question for me doesn't arise. And if it is my description of the phenomena, then the question of whether theory fits the phenomena, and whether it fits my description of the phenomena, among, for me, the same thing. Now, for me, this ends the argument, but I will come back because it's important that it will tie in the third argument as well. There's a kind of qualification to do that, the third argument will tie in two. The second one I'll just remind you of, by reading the map, it doesn't matter how much information you try to build into the map and how much you know objectively and in internal sentences about what's in the map and how it relates to the world. By itself, we don't have useful information on which we can base and break recommendations, unless we can also And this self-location is not the further rid of description, not the further rid of knowledge or belief, of the same sort. It is something instinctive, and it is actually crucial. That's what the indexes have come to me. The third argument, I won't say something more about. No arts are science. For me, that introduction comes up quite often. And I want to arrange for it too, which is innocence by association, what I have in mind

7:30 is this. Very often when a philosophical question comes up, it sounds a lot like an empirical question. And there is in fact a genuine empirical question in the neighborhood. But the philosophical question is thought of as distinct from that empirical question should try to find an answer to it. And if they confuse the two, they will find themselves doing armchair science. Armchair psychology, impossible of mine, armchair biology, armchair physics, right? And I say, no armchair science. And the innocence by association is that the the local question tends to sound in the synthetic world because of the nearby real questions. Now, that applies, I think, when we feel at the first argument that we have to be involved in other things, because there are certainly empirical questions in the neighborhood. If the description of the phenomena is your description of the phenomena, then the question whether you were reliable as a describer is certainly an important question for me. And it's an empirical question. And it shouldn't be answered for the philosophical answer. It's an empirical question, so it goes to the empirical scientists. And similarly, if I can ask myself about yesterday, when I was doing some observation reports, was I reliable then? I had a few drinks. I was a bit exhausted. Well, again, that's an empirical question. The question that the logical answer tries to answer is confused with these empirical questions which no philosopher should try to answer, because that would be one of those questions. So this is the little array of arguments that I have been using and that I'm going to use again today. So, here's the outline. I will first of all present Kotlin's argument again. I will not spend much attention on the validity of the argument structure and so on. I did that last last time. But after looking at this text, what he actually says, and commenting on it,

10:00 we will be able to dissolve the argument, show that it disappears of its own account. Of course, to people start an argument, we have to show that the people start to dissolve. That's the other thing to do. What it will lead to is another point about the crucial indexicality of the language of science, and I will have to back this up by actually talking about language and about proof, reference, and translation in order to show how it comes and stands on its own free. So that's the plan. Here is some of the text of Popham's argument. I'll quickly remind you of some things about it. The point of the argument is going to be that every ideal theory is true. The conditions for being ideal theory are extremely minimal. As a matter of fact, the central part of the argument that the theory is consistent, and if it's right about the size of the world, and it's finite or infinite, then the theory is true. And it's an incredible conclusion. So he begins with the theory, and he has assumed that the theory says the world is infinite and that in fact the world is infinite. Then he has a model of every infinite cardinality, including that means the size of the world itself. Pick a model with the same size as the world. Map, this is imperative, map the individual of this model, one-to-one, onto the world, and use, again, an imperative, the mapping to define relations of them directly in the world. So then you don't have an interpretation of the theory in the world, you have, by assumption, a model of the theory, which has the same size as the world. And I'll give you a picture back on that I was using. The theory is given an interpretation in two steps. The theory has a model, so it's interpreted in that model. The model has the same size as the world, so then there exists a correspondence,

12:30 correspondence between the two. In the modern world, in fact, many, many, very many correspondences. And the interpretation of the language in the model can be transferred to the world by using the images under such a correspondence. And then we have interpretation of the language of the theory, under which the theory is true. And he says, what more you want. See? We can now look on this theory as a true account of what the world is like. properly interpreted, of course. That's his argument. I want to draw attention to the very anthropomorphic language that he is using here. He's telling us to do things. The purpose is that we should end up with the interpretation of the language under which the theory is true. And apparently, here are the instructions about how to get that. But can we use things? What is the basis for his assumption that we can do the things he tells us to do? Only that, there exists a function that has the requisite features. That it has clearly the effect that under these conditions there will exist a model that has the same size of the world. And therefore that there will exist a one-to-one for its moments, a function that maps the one and the one to the other. The fact that this function exists, does that mean that we can interpret the language by mapping the model of the world? See, when mathematicians do their business, when they are giving proofs and so on, they do speak in an anthropomorphic way. And I don't have any attention to that in that context. But now, you know, we're doing a positive. And here, we should look at the literal meaning of the words. It is not true that we can't do these things. And if we cannot do them, then we cannot end up having an interpretation. Let me illustrate this one. I have a blind sheet of paper here. And it's very good paper. It's heavy palm paper. I got it in Allsorce College, so it's supposed to have any questions that you can find. Looking at it, I can see that it has little striations and pumps and, you know, it's heavy paper.

15:00 I can see a quite complex structure in this paper. It's almost invisible. This sheet of paper is a map of the city of Paris. There exists a 100 correspondence between most of the just discernible features in this paper and most of the important features of the city of Paris. As a matter of fact, this correspondence is a map that is pretty well continuous preserves spatial structure and so on. In fact, it's better than any map you're going to find at W-H-Smith or any border. It is much more extensive and much more detailed, much more precise, and I would like to sell it to you. Now, the point is this, that this is an accurate representation of the world in the very same sense that this is an app of affairs. Because in both cases, the only basis for the claim is the existence of a function that will do the requisite job. I want to give a different example, a second example, for more mathematically minded, because I think that the distinction between the easy and the morphine language and what is actually being said, you have to keep it in mind, especially if you're listening what I would say about the world. Let me use a mathematical example, the Euclidean sphere. I know that could be drawing programs, it was a little like an egg, but this represents the Euclidean sphere. And here is the set of numbers, and I want to ask a mathematical mathematically minded among you, consider a nucleon sphere and consider the problem of co-ordinizing the points in the sphere. Now, we know that that is a reasonable problem, and that in fact there are many many different ways to co-ordinize the sphere. Different standards for whether they do better coordination.

17:30 You can think of a really bad one like they're using geography, the latitude long achieving that a couple of points have to be treated specially, and false, right? Okay, so you won't say you can do it. And now I'm going to say, no, you cannot. The Euclidean sphere is an object that has perfect rotational symmetry. So you cannot identify a specific point on that sphere. There is nothing that you can say about one of the points that doesn't equal your five or the other points. If you wanted to co-organize it by putting in that to go with you and so on, I'd say, fine, go ahead, pick a point to use it more at all. You can't do that. You can't do it. What you can say is that there exist many functions that will matter between the real number triples and the point of the sphere. But not that can do it. And that's so even if, you know, yesterday evening we had this section called the Nassar and both Goebbels and Hitz at a mathematical intuition, even if you can think that we have an intuition of the real number continuum of the Dremelian type, that won't help with this object, it's a spiritual rotational symmetry. So here's the point. The fact that exist does not mean that we have an interpretation of the language. Now on what conditions would we have an interpretation of the language? Such interpretation is supposed to be a function that connects the words of the things. If you want to say that you can identify that function, you have to describe it. A function has a domain and a range. So you have to be able to describe the domain and describe the range. In this case, the domain is a mathematical objective model, and the range is the world. Can you identify such a function? Well, can you describe the world? You see now here, the conclusion of Potom's argument encounters a dilemma.

20:00 If we cannot describe the elements of the world, then we can't describe the range of the function. But in that case, we can't have an interpretation. And if we have no interpretation, the question of truth and falsity can't even arise. If we cannot describe the elements of the world, describe, or define, or identify any function that assigns extensions to our words in the world. But then the question of who is impossible does not mean to arise. Second part of the dilemma, if we can describe the relevance of the question of the world, then we can also distinguish between right and wrong assignments or extensions to our critics of the world. And then it doesn't follow that the theory is true because you can say, well, it isn't true on a good way, admissible way of assigning things to the good words. It is not true on any admissible interpretation. Let me illustrate with an example I was using last time. It was the first water theory. And it was supposed to be an example of a theory of the sort that Putnam is talking about. And if this argument works, it shows immediately that this theory is true. The theory has these actions. There are at least 10 things that are water, at least 10 that are not water. So really, the The theory says there's infinitely many bits that are water, and infinitely many bits that are not water. And our intuitive reaction is, well no, a bit of water consists of water molecules, so if there's only finitely many water molecules, that theory is false. And Pottenberg says, my theory, my argument establishes that the theory is true. Well, if I cannot describe the elements of the world, then of course I can't issue this, but then neither do I have any interpretation, and neither does the question of truth and false, but if I can describe the world, for example in terms of water and not water, then I can say an admissible interpretation of this language is one in which the technical term water has its extension to things that are water. And then the question whether the theory is true on an admissible interpretation

22:30 is a factual question. And there's nothing trivial about it. And his argument proves nothing about it. I also, if I use Milton interpretation, if I can distinguish between Milton and Milton, and I say in admissible interpretation, one makes the technical word water more. Funny. That's not what I use, but how I interpret it. But as soon as I have an interpretation, the question of truth and falsity is no longer doomed. It becomes a factual contingency. So, on that dilemma, the argument just, that's what I can see, dissolves. Now, I think you can also see that, as usual, you know, we don't get something for nothing. I had to introduce something just surreptitiously as I could, and I'll make the solution for the solution in this. I didn't introduce surreptitiously as a difference between the language we're talking about and our own language, the language in which we are talking about. And I said, you know, if you can describe your world in a certain way, and then in our own language, then this follows, and if you can, then that follows. Now, truth, when it comes to our own language, it's not elliptical, it's not validating, There's not a different interpretation, it's just things are true or false. I think in that position you'd like to challenge this by saying, all right, you know, what you've really shown is that maybe we have no interpretation of a theory like that, in our current theory, and so we can't regard it as true or false, but the argument stands, because truth is objective, and you're bringing in a reference to us and to our language, but there's something as objective accuracy of representation and objective truth, and you haven't really thought about that. And so if argument stands and they're still objectively true, then all your distinction isn't here at all. Well, he's talking about truth relative to some interpretation that And I should think that the map of Paris example shows you how to keep the country that

25:00 part is. But on the other hand, I can see why people would want to try and stick with that kind of objective plus independent way of thinking about the map. Because apparently what I have smuggled in by this heavy reliance on what is my language, this indexical reference to our language something like subjectivity or you know that truth is no longer just relation between the theory and the world that something like relativism has come in and there's a specter there so I have to respond to this by trying to convince you that any conception, any kind of conception of truth, accuracy of representation and reference is going to involve this kind of indexical reference to us. That will be how I have to dive into it. So I'm going to try to do that. And I'll try to do it in two parts. One is I'm going to talk very generally about representation, because after all, language in theory are a typical case of representation. And then I will apply that directly to the language of theories. And see if I can use it. So, representation in general, that's what I'll start with. Now, I think there has always been, and philosophers, wanted to talk about representation of hope that they couldn't keep any kind of subjective element out of them. We saw it very strongly in Karnam that we were struggling to keep any subjectivity out of the account. But it goes way back, because all discussions of representation seem to first hope for something like just a relation between the image and the thing that can stand on some feet and makes no reference to us. And the first thing that people then think of, of course, is resemblance, that the image must be something that resembles the thing it represents. And I'm sure that most of you are familiar with arguments to show that that wasn't, first of all, correct. In our time, Nelson Goodman gave number quite a dislike.

27:30 But it's an ancient subject, and it goes back to the videos' dialogues, and I'll use examples from Plato. just to quickly go through it. In the surface of the Crabbitos, Tapio picks up this question. The point that I want to take to the surface is not the remark, but I think whatever you count. The stranger in the surface asks the ideaist, so, under what condition is something an image or something else? And Tapio says, well, it's a likeness. and it resembles it. It's just one thing that resembles the other thing. That's a likeness, and so that's a copy of an image. And this painter says, you know, you're forgetting something. He asks him, if you were right about this, does that mean that a better likeness is a better image? And he said, yes, of course. He says, well, no. Because if a sculptor is making a sculpture that is going to be put on a very and people have to see it from below, then he can't make it in the right proportions. If he wants people to see an image of a man, then he can't make that sculpture an image of a man. So he has to distort the proportions. So it's not true that the better likeness, in the sense of more resemblance, more correct resemblance, is a better image. So being an image does not consist in fact-and-verse likeness. I think that's already a very nice point, because it means that representation can succeed sometimes only through distortion. And I think the scientific theory is what we always keep in mind that perhaps a good theory will be one that presents a picture of the world that is really false, distorts what is there, in order to be a good scientific theory. And the Cratulists, I think, gets more interesting, because the Cratulists, in the most entirely about this question, Socrates, it has a strange structure as I read it anyway. Socrates seems to be intent on convincing Cratulists that a certain kind of sophistical theory is wrong, that says that something other than these animals has defined, something like convention

30:00 or choice on the part of the users. And he takes up specifically representation in language. The result of this is that he ends up with a completely ludicrous theory of language. He says that sentences can represent the facts, but they have to resemble them. And just like You know, we have to have bits of pigment in it, whose colors are white, the colors of say the grass and the flowers and so on. So therefore, the sentence must have words that resemble the things. And by the same reasoning, the words resemble the things only because they have parts that resemble aspects of the things. So for instance, the word running naturally starts with the letter R, because R resembles swiftness. And the word language, it has to start with an L because it's an especially language sort of letter, right? Now, you can tell the, you know, it's a scientific thing, it's a very irony here. By the end of the dialogue, Socrates has brought in the element of the convention and has gotten himself and Crappless argued into the view that there must be something other than an example that's going on. that some element of conventional choice on our part. Now, the example I want to take from the dialogue this one. Right in the middle, Catholic apparently has just been thinking that, yes, Socrates must be right, and it's a perfect example, it makes for a perfect image. Socrates says to him, really? Now, suppose that a god where to make a likeness of you. And being a god, he could make it a perfect way, not a sculpture or a painting, but a living, breathing organism with the same physical structure as you have, even the same inarts, and the same feelings and thoughts. Would that be an image of you, or would that be another Crabillus? And Crabillus says, oh, that would be another Crabillus. Okay, so that makes the simple point of resemblance enough, but actually there's something more true here. And I always think that when Plato wrote these things, he wanted his students to go home and see what they were growing with the thing. What could and should Catholics have said to Socrates here? I think he should

32:30 have said, well, it doesn't depend on what God is doing. I mean, if you just want to make a clone, fine, he just made a clone. But if he wanted to display an example of Greek man, the other gods, and to do that he made a perfect copy of me, he would have made an image. You see how the element of intention and use comes in there, whether as an image, not an image. Let me, I'll just slightly update the example just a little bit. Suppose that I'm in Paris and I go to a gallery and there hanging on the wall is a famous photograph that I recognize, a famous photograph of the architect. And I have a camera with me and I drive the attendant to let me take a photo of the toilet. I go home and I have Now, what do I have at this time? Do I have a photo of a famous photo of the actual power or do I have a photo of the actual power? Well, that rather depends on what I do. If I just decide to use it as a postcard and on the back I write your address and I send it to you and I put a note saying and say, practically wonderful, as you can see, wish you were here, then it's a photo of the Eiffel Power. Or they might write a monograph about the historical development of topography, and I use that as one of the illustrations. That's a photo of the photo. Or maybe that's enough to make the point that representation does not consist in a two-place relationship, but a relationship that brings us in, that brings us in, how it's an intention, and so forth. Now the biblical language. And you'll see how that comes around to the metaphysical challenge of RTPM, this solution of the paradox. So, we're now going to talk about a great specific case of representation, namely representation and language. And then the concept is not accuracy or representation, but truth. And in the case of words, reference. And I won't make the same points about that. So, here are two statements that I think you will agree

35:00 with. Cat is a word for cat, and quote, all of you cannot, unquote, is to ignore your And you'll agree, I think, that you can't help but agree that you can only assent to these statements. That they could not be false sentences that I have put on the screen here. So, at first sight, they are topologies. But they're not ordinary topologies. Think of topologies like, a bachelor is unmarried. Well, just because of this epology, the statement, there could be a bachelor who is not married, is false. But these are not like that. Because the word calf could have denoted genats, rats, or bats. Because the language could have developed differently. And if the word calf can actually be the word for rats, then the sentence, quote, So what kind of thing are they in? I'm going to introduce a note from pragmatic ecology and asking why. What is it about these that makes them undeniable? Not necessity, as we just seen. It must be something else. But consider it is example. I was trying to choose a word that isn't, something that's not a word, really. Now, if any of you play Scrabble and you know that actually there is a word, carbo, then, you know, in Italian I won't change the example instead of the one that is not a word in our language. And if you do not acknowledge it as part of your own vocabulary, you will not ascend to that third first sentence. It's not the case, it's regarding that process. Now, I think that's the crucial difference. But by the first sentence, you don't even have to know what cats are to realize that it's true. That's not the point of knowing what cats are. But acknowledging that the word cat is part of our language is crucial. That's what makes the difference. So, there's There's something indexical again. That's why I don't say it's a pragmatical problem. And the nearest analogy in a written field that I know is Moore's Paradox.

37:30 Here we have examples that could be false, but are undeniable, right? In the case of Moore's Paradox, the sentence is like, I don't know if I have a condition. Yes, the sentence is like, it's Snowy and the King and I don't believe that it's Snowy and the King. I can't say that. Actually, it has often been true. I've never yet believed that it was snowing and you came in any given moment. And I'm sure it has. No, I'm not sure. I think probably it has snowed and you came in time to time. On those occasions, the sentence, what the sentence says, the proposition expresses true. But if I were to preserve it, I would actually display an incoherence in my thought processes, right? Incoherence in my state of opinion. It's not something that I can't believe. Okay. There's a pragmatic element here. The difference is, though, remarkably, usually you have to use an indexical, and that is, a word that has the function of playing the indexical role in order to produce these things. But in our first two examples, we have indexicality that is hidden. There is no indexical in these sentences that is doing it. But there is a hidden reference to our own language. That's what the statical sentences are. And that's why I say the in the second example, has this hidden index of challenging to it, that I don't mind, this is crucial to it. There are many pseudo-columns, metaphysical pseudo-columns that surround this topic. I'll just quickly talk about London, how it happens to it. A word can have many different extensions. The word captive, it has the extension, perhaps. The metaphysician asks, well, what determines that it has the extension of the ties? Of course, in my own language, I could start explaining something about what the word cat

40:00 is going to refer to, but then we just push the problem once that happens, because then the question would arise for the other words that are used to do that. So what is it? There must be things like truth makers and reference fixers and metaphysical glue and what knows what all in order to do the job. But I ask myself, if this is a question that we ask ourselves in our language, what's the worry? Is it the worry that there are actually green things that are not in the extension of green? What's the word green? That the word green doesn't have the extension it's supposed to have? So that means that there are some things that are green, but the word green doesn't apply. What that would imply is that sometimes the sentence, x is green, is not true, even though x is green. But now we just deny the pragmatic pathology. So we just end up in an absurdity, a pragmatic absurdity. So it seems to be that depression isn't here. but it has a certain grip on the imagination because of this innocence by association because these metaphysical questions are surrounded by real empirical questions and that's where I think we get their grip their force and I will say that I will say that by talking about about translation In the translation you can very clearly see how different questions separate themselves from each other. So I'm going to talk about a Dutch linguist, his name is Steve, and he has the job, an empirical job, in empirical linguistics, of investigating the dialects on Europe's offshore islands. So he decides to focus on the city of Oxford, and he has a question, an empirical question that he has to find the answer to. And although some of you don't know Dutch, I have the feeling that you understand. Right? Okay, now, do you think, do you think you understand that? Do you think that this is the question that he asked? But look, the question he has is an empirical question.

42:30 And the question is, can the word for cancer not be part of the question at all? We're dealing with a pragmatic topology here. And, you know, he might even have learned so much about English that you would know already that any native informant would have to say yes to this. But it will not answer this and this will not answer. Now, you see, what happens is of course that these metaphysical puzzles about what makes number two true and what determines the reference relations and the truth-making and so on, gets a kind of sense of legitimacy from the fact that there is a real empirical question in But the real empirical question has to go to the real empirical scientist. And it's not supposed to get the metaphysical answer. So what we have here, we have usually a kind of amount of a logical question, an all sense question, an empirical question, and then finally the consultative questions are kind of confused amount of the truth. All right. Well, again, I think that you probably realize that. I've now tried to start physically to keep something a little bit out of the picture that you might want to challenge. And so, again, I can imagine a mathematician coming to me at this point and saying, once again you are ignoring real challenges. You are relying so much on this idea of your own language. And what is your own language? And you are not even critically examining this. But if you think that you are going to rely on your own language in this way, you must have a very strong belief about your own language. You must believe that your language is up to the job. That your language is adequate in certain respects. is specifically to the job of describing the world, but that is a substantial question because the language can be very well adapted to that. And if you were to try to answer what it means for your language to be well adapted to this kind of job of describing the world, and the reasons you would have to believe it, you

45:00 will find yourself a metaphysical realist again, you will have to postulate structures from the side of nature, and so on and so forth. And, you know, I can just go through the whole series of arguments again, trying to show how exactly that kind of thing is dissolved by asking him, can we ask these questions about our language and our language for a moment, please? And not act as if you're talking about an alien language on another planet? Or is that what you want to do? If you want to ask expressions about our language and our language, we are just going to be stuck in pragmatic pathology. And if you want to ask about an alien language, well that's not the point. But, there is a real question that I think is being pointed to here. I am relying so strongly on this idea that the indexical element in truth and accuracy and representation and that the indexical reference is to what is my language, then the definition is to ask, well, what is your language? And among conditions, is something your language? What is the relationship? What is trying to say, this is my language? I think it is a serious question there. So I think that's the last thing that I have to take out in order to give the vacuum from the investigation. So I won't try to do that now. That's the last part. I'll begin just by giving you two simple examples of cases So that's what the language is like. I'll start with myself, and I'll just give you a little bit of information about my language. My language is English. And, as you know, recent converts are always much more passionate than all the big ways. So I'm quite passionate about the purity of English language. And in an airplane, I ask for a cream for my coffee. and the stewardess brings me an edible petroleum product, and I get upset. And I get upset not just about what she's giving me, but about the word, because I do not think that the

47:30 word cream should be used for anything that doesn't come from cows, or sheep, or a goat perhaps, but not from the petroleum industry. Now, what am I expressing here? I am expressing a certain kind of usage. And partly, my intention to never refer to the screen to refer to that kind of style, right? That does something about what my language is wrong. In the case of philosophy, I'm a little bit more liberal, but there are still things like qua, that I've banished automatically and completely. Even if it's translated in the English word as, and you would like to go over a quad, I think there's something hyper-intentional that probably won't pass at all. But okay, mostly I'm fairly liberal about it. There are examples of philosophers who are not very liberal. So in my second example I will take the least liberal, and that is Klein and Word and Object. In the book Word and Object, he actually tells us precisely what is the canonical language for analytic philosophy. And it is it has no modality whatsoever, and it has a number of other pictures. So there's a case where a person really tells you what his language is. And the third example is another philosopher who also tells you what his language is, but he has a much bigger, richer language than one, and that's David Lewis. In some respects, it's like one language, but he certainly has no doubt about this, and he can express modal statements. And not only that, he can express many modal statements that other people would shy away from, including, as he himself says, that he can quantify other people in different worlds. And here's a good quote that he has about it. He says, I can. Some say they can. I have no doubt to offer these unfortunates. Now, here we get something that's demonstrative. Since the expressive power of this language is greater than what language they have, you see, once you get to the point where you can actually describe your language with this kind of precision, you can also prove things about expressive power, right? And that this language has greater expressive power than

50:00 very good examples of the state languages, and people say, that is my language. But the fact that they say it doesn't excuse us from the question of under what conditions is some language a personal language. If these people are unilingual, I guess that's not much of a question to be asked. We should ask about multilinguals. What would it be like if a And we wanted to ask about which one is his language. So, I think Quine would have insisted on a behavioral criterion. And we have a nice example of a behavioral criterion for things like that, and that's the Turing test. Remember that Turing proposed a behavioral test for the question whether a computer is conscious. And he suggests that you put a computer in one room and a human being in the other room and then you have people interrogate you without seeing them, I suppose on email. And if you have this interrogation, you cannot tell which is the human and which is the machine, that the machine is conscious. That's his test. Well, alright, let's do the same thing here. Suppose that Klein, arch-enemy of modal fidelity, during his lifetime actually learned David Lewis's language very, very well. You learn the syntax, you learn the semantics, you learn the ex-evidentity rules at this age, or pragmatics. You learn so well that anybody giving an argument in that language trying to correct the validity, you know, correct the perspective of the validity of the argument. People teach a seminar in which students at least know that he wasn't advocating David Lewis's philosophy. At what point we still want to say, yeah, but that's not his own language. Well, you know, we could try a Turing test, right, for a behavioral criterion. So we could find in one room, and David Lewis in another, and then interrogate the email, right, and if the students, if no student can tell which one is fine, which one is Lewis, then would we have to say that Lewis's Well, I don't think these behavioral tests work very well. For one thing, even for the Turing concept of tests, you have to get a computer to be quite exactly the seed for.

52:30 I mean, if you ask, are you animal, mineral, or vegetable, it's not supposed to answer you truly. And if you ask, what is the number pi, it's not supposed to just give you 500 of inches in a row in the decimal expansion of pi, which it would normally do. So the test will only work if you first of all instruct the computer to be deceitful. And similarly, you know, with Quine and Lewis, if we sent the message to the question, would you use this kind of discourse if you were giving a lecture expressing your own philosophical then finally you're not supposed to answer this truly, right? So the whole thing really works with a very artificial kind of constraint on what they can say. Now, notice the kind of question that you would like to ask for them. Are you willing to use this kind of discourse for the following purpose? And that doesn't sound behavioral. Karnat suggested a very different kind of test in his article about The Monomy and Animaticity and it isn't quite behavioral. In fact, it asks that kind of question. So Karnat said if you have a criminal subject and you want to find out what he means by the word fear you don't just point out forces to him and cows, but you also show him pictures like unicorns And you ask if you would be willing to apply the word to something like that. Now, notice these words that I've been using and I've been trying to express this. In the case of a queen, I talk about commitment and intention, right? Here I'm talking about willingness and purpose. And I think here's the fact that what my language is will not consist in any objective knowledge or belief that I have, but only in my having certain attitudes, commitment, intention, and so on, which I can express. Now, can I also describe them? Well, I can describe anything,

55:00 if you don't mind your prescription, not very informative. Can I describe it? Well, if I describe something, I can describe it in my own language. In a talk yesterday by Gödel, he also handed out a poem by Gödel, one of the verses says, you can describe your own language in your own language, but not quite. Meat guns. The answer is that, yeah, I do want to describe it to some extent in my own language. But pretty soon I get to play my own pathologies, and then again what I'm doing is I'm expressing by linguistic commitments, rather than describing facts about myself. So, that's my argument for saying that, first of all, all the way through, there has to be this reliance on language, on my own language, on what my own language is, that this is indexical, that it cannot be equated with a bit of objective fact, known or believed, that it consists in being able to do something such that having a description of it is not sufficient for having it, just like even that. And that the kind of challenges to this somewhat subjective account, my physical challenges, are actually cases of pseudo-questions that simply amalgamate different real classes. So, what about Putnam's Paradox? Well, Putnam's argument only became a paradox in the context of the kind of analytic metaphysics that we were trying to let happen. And it reminds me of Pont-Barré's book about logicism when he heard about the Russell Paradox. He said, oh, the Lord just said, it's not stout, it doesn't have the crown on it. Thank you.

57:30 Thank you. If you view that was a spiritual problem, why not? Yeah, mostly. I think that it's... and also there's a kind of huge thing out there, I think, in some ways quite well. If there is an empirical practice, right? In an empirical past, you have to examine the dialects of the offshore islands and the uncritical fractions, which you answer in your own language. And there's nothing philosophical about this. The standards for what's an acceptable hypothesis are the same as they are in biology and in chemistry. If you then start thinking of your own language as just being one of these arbitrary languages, then you get this descendant to implore for relativity and so on. That's the mistake to lose your anchor the moment you stop thinking that, oh, but any discussion I carry on about this, I do it in my own language. And then what he tries to do next was to, well, of course, the brain is going to try different things, but one thing he tries to do is something like the position of saying, oh, well, my language is less than the home language is, to kind of establish, I call it the home language, and so on and so forth. Well, that doesn't get the other down, unless by that he just means when I discuss it when I'm speaking my own language. If you take the Dutch linguist who's investigating the offshore islands and he's trying to find out what the adaptation of the word cats

1:00:00 and we were puzzling over this and saying what principles would you use to find the right answer to the question what the English word cat denotes? I mean, not the specific answer to the specific question, but we want to know the methodological thing. How do you know when you've got the right answer to the question, what the one path denotes? Is that a suitable problem? I think there's a real question there. What are the methods that the scientists can use and how do they engage in theory choice and what are the standards for management of scientific opinion in terms of visual opinion? There are a number of logical and historical questions about this, but not the helpful ones. So, I don't think you're right, but there's nothing specific to linguistics, I mean nothing specific to acquiring a language about this. There are quite general theoretical principles. Well, I'm not saying they're general theoretical. Well. I'm saying it's the same as science. Do you have anything to say to the puzzlement that that would arouse, it seems like a fairly sensible question, how would I know when I got the right answer, and something like when you got the right theory, we wanted something a bit more specific than that, about how you'd know when you had the right answer to the question, what is this, what do you know? specifically about principles trying to not take a question? Well, there's something specifically said in the sense that the linguists are bringing out their methods perhaps depending on what they say about how they do it. You can start doing empirical linguistics, but there are no general questions that are different from one science to another. I think the problem is that there seems to be genuine scientific progress made by considering

1:02:30 these issues are structural restrictions. But the progress has come about through not falling back on the desk of this or some pragmatic elements of thought, or as it were resisting the problem, but rather trying actually to solve the problem on the term of images pose. I'll put it in a very conflict point for you, which is in as it is, it's an attempt to take The whole algorithm is one of which you don't seem to change the elements of general activity, which actually seems to involve the question.