Provisional Knowledge in Physics and Elsewhere

0:00 Thank you for making this all possible and really treating us like kings. So, I don't know about all of you, but I've been working very hard so far, and I'm reminded of a cartoon which has a studious bear sitting cross-legged with an open book popped up in his legs with this kind of befuddled look on his face. And the caption reads, there has been an alarming increase of the things that I know nothing about. So my talk, I hope will not tax your brains as much, nearly as much as some of the other wonderful and really informative talks that have done. And I'm hoping that it will, at the same time, manage to actually increase the amount that we know. So, how do we get it started again? Just put your machine went to sleep. No, no, it's just machine that we should just hear you. Your machine is in sleep. No, that's where we're going. I don't know. Yeah, yeah, we're going to. There we go. That's Ben. That's two now. I'll get it all set up well in advance because I'm nervous it's not going to work. I think that lower left button there, that lower left little thing. The lower left? Oh, I'm on the other screen. Here we go. That's it. All right.

2:30 So, let me begin as one often does and give you a kind of idea of what I'm going to try to do. I'm going to be talking in a general way about problems with foundational attitudes in physics. Now, the prelude, I won't be taxing your brain, is in part in recognition of what I'm quite sure is the fact that everything I'm going to say in the stuff about foundationalism is pretty clear to most, if not all of you. But I hope that pulling it together and seeing how bad things are is going to be a worthwhile enterprise. And then this is going to serve as a springboard for rethinking the way we think about both truth and knowledge. We're all familiar with the problems with foundationalism in physics, but I think that people have not yet begun to appreciate the extent and the ways that this has repercussions much more broadly for how we think about truth and knowledge. So let me begin with examining a rationale for foundationalism in physics. For the moment, we'll just take foundationalism to be the idea that foundational physics provides the ultimate basis for all human knowledge. You ask people, why should you believe this? Often it's hard to get much of an answer. I take it that what people usually have in mind is, well, the world is a physical thing, and physical things in the world have parts, and if you know how the parts behave, then you're going to know how everything behaves. So it's a kind of atomism. Now, let's go and look just for a moment historically on where this way of thinking, of thinking arises. I take it to come from the beginning of the scientific revolution. It may well have earlier roots, but it shows up very prominently at the birth of modern science with the clock metaphor, analogizing the world to a clock and understanding the world as understanding the way

5:00 the mechanism of a clock works. So to put it in a phrase, it's understanding understanding the world is understanding shape, matter, and motion, and so constitutes a basic mechanistic attitude towards understanding the world. Well, since the 16th and 17th century, the notion of what counts as shape, matter, and motion has been enormously broadened. That's why I'm putting quotes around the shape, matter, and motion. I'm going to use that as a catchphrase. And now what we take to be the physical parts is enormously different from what the early modern scientists had in mind. And we've not only changed our ways of thinking about the physical parts, but we've expanded the kind of parts, the kind of things that we take to be parts, for example, in thinking of what I'll call analytic parts, basic waveforms and analyzing waveform, analyzing anything that can superimpose. So, the notion of mechanism has been accordingly broad. So, let's now look in more detail at what I think maybe people have in mind gripped by this picture of foundationalism in physics. So I've sketched a little argument here. Alright? First step, our only source of information about what is external to us is by a perception of shaped matter in motion. So it's a very empiricist idea. We see ourselves as affected by the world, and for there to be any difference in the knowledge I have, there certainly has got physical difference out there in the world. And the next step is that, well, those are the things that immediately affect our senses, but they give us information about other things. Shape, matter, and motion can only carry information about shape, matter, and motion. Or another way to put the thought, anything that causes the things that directly cause us ought to count is physical. So we put these together and we get the conclusion that all knowledge

7:30 of the external world is knowledge of shape, matter, and motion. Now put that together with the thought that fundamental physics, what we mean by fundamental physics is the ultimate account of shape, matter, and motion. And then putting all these things together we get as our conclusion, physics, fundamentalism, that fundamental physics is the ultimate source for understanding all natural phenomena. Now, I've gone over this because I think this is a terrible argument. And it's, again, I think the reasons will be well known, but let's review them quickly. Why these, or just a few of the ways this is a bad argument. The thing I want to start with is that it's not that we receive knowledge passively. We contribute an enormous amount. You don't have to be a diehard Kantian to appreciate this. For example, the physiological understanding of color perception shows that our color receptors, it's not like we have a complete band of receptor over the visible spectrum. receptors where what I guess called trichromats gives a very complicated relation between what we experience by way of color and what's going on out there in the real world. It's not a one-one correlation. It's a many, many connections. So another example, when we think of the world as populated by discrete physical objects so this infernal device that I'm holding in my hand or this glass it's a particular thing as soon as I give it a name object number one I raise questions about identity but then I start worrying if I start taking little chips out of the glass it's the same glass philosophers wrestled mightily with this problem seems to me the natural So, populating the world with discrete physical objects is a simplification, it's an idealization that we make probably hard, hardwired into us. Well, these examples can be multiplied. The conclusion is that what we see is certainly no mirror of nature.

10:00 Next, we extend this to the fact that we contribute enormously to the theoretical phenomenon. So bearing in mind what Jean reminded of us a couple of days ago, that phenomena have to be, phenomena that we're concerned with, have to be accessible, we start with the immediately perceptible phenomena and we build on those, building bridges, of course nature slapping us in the hand very frequently and we then try again. And so altogether, I take it that we're driven to a kind of Kantian constructivism. It's not that we passively receive things, but we have to bring perceptual concepts. I'm going to understand the word concept broadly to include the frames, the mechanisms, the way that our perceptions get constituted by our physiological mechanism. conceptual frameworks in framing and developing our theories of course always chastised by things independent of us, but we do a lot of construction and the Kantian constructivism it's not, it completely lets go of the rigid part of Kant it's not Kant's rigid formulation, it's the dynamical, relativized conception of the that Michael and others have been talking about in great detail. So if we find any plausibility to this Kantian constructivism, I take the positive rationale for physics, fundamentalism, really to have collapsed. Well, fundamentalists still are very far from conceding. They think that science can give us, let me interject something else before I continue. One of the things that is an issue for many people is that when we look at things other than foundational physics, of hydrodynamics, aerodynamics, many-bodied physics,

12:30 the theories are always highly idealized. And people like Larry Sklar, for example, look at this and say, you know, these theories aren't true. They're useful fictions. But not to worry, foundational physics is going to make them honest. and there is hope that we can get exact truths even though many, many of our theories in physics really are egregious idealizations and so very much strictly speaking false. So here's a contemporary reason for thinking that despite the ubiquitous use of idealizations in constructing theories that science can give us truth. We think of the world as having things and there being properties, and so the things out there, things independent of us, have various properties, kinds of things, have properties, qualities, and relations. So we can attach labels to these things, words that refer to the things and properties, and then put those words together. The glass has water in it, electrons have unit negative charge, to formulate exact truths. Now, this doesn't work. Let me illustrate with an example. Example of the quantity mass. Is there a completely determinate thing to which the word mass has been attached? Well, the notion was thought to be determinate, one nice thing according to the Newtonians, but then when we get to relativity, the notion of mass splits between rest and relativistic mass. Relativistically, there's interconvertibility between mass and energy. in general relativity the special case of gravitational energy it's not even strictly localizable and then to complete the mess mass is a renormalization parameter in quantum field theory now I submit that nowhere along the line has the term mass got nicely attached to some completely determinant thing that hasn't happened yet it's not about to happen in the very near future. Well, fundamentalists still have another ground for thinking that,

15:00 for arguing that fundamental physics really does in some sense give a foundation for all the facts about the world. And so at least in principle, anything about the natural world that we can know. They say, well, this is again the idea that all these idealizations like hydrodynamics, thermodynamics, theories in solid state physics, that these are idealizations, so they're strictly speaking false, they're useful fictions. and in fact the claim is that we've gone a long way to understanding those idealizations by connecting, by getting the higher level theories out of the lower level theories. Well, again, this I think has massively failed. I think everyone here probably is very on board with the idea that reductionism doesn't work, but let's again go over just how bad the problem is. To begin with, I want to make a point which is going to be important here and is going to be important really tacitly throughout the whole talk. That idealized accounts can stand on their own. What do I mean by that? Well, go back to the 18th, 19th century treatment of hydrodynamics with the idealization of water as fluids as continuous media. Suppose we'd never learned better. That account still would have really provided us with real understanding, real knowledge. If we'd been stuck with the Tony mechanics and nobody had ever taken us beyond that, the Tony mechanics, although unbeknownst in this thought experiments, unbeknownst to the people in the thought experiments that it's wrong, still would have stood on its own and given a very important, visible understanding of the way the world works. So, first of all, idealized accounts can stand on their own. And there's a good thing because reduction to fundamental theory requires further idealization of the things that we're trying to reduce.

17:30 Again, the example of hydrodynamics provides a nice example that when you try to understand a higher level theory on the basis of a lower level theory, usually, if not always, there's what I call exogenous supplementation. That you have to, for example, in getting thermodynamics out of statistical mechanics, you have to import assumptions about the initial probability distribution. No way that comes out of the fundamental theory. And finally, perhaps most importantly, at the bottom of the theories that we have today, there's idealization all over the place. So again, a lot of this is familiar to it. very quickly. Quanta only work, the quanta of quantum field theory only work when a flat or regularly curved spacetime, so that's an idealization that we have Hague's problem in quantum field theory. You have a description of particles at a distance coming in and interacting, but the part of the theory that describes them coming in and the part of the theory that describes them interacting They can't, in the current applicable theory that we have today, can't be consistently put together. As we've been talking about for the last four days, there's the worry about combining quantization and gravity. Altogether, we need further idealization at the top. Stuff stuck in trying to get the top from the bottom. And in any way, the bottom already involves very heavy idealization. So what we have is that throughout physics, and indeed throughout a lot of science, everything involves theories which are faulty in one way or another. They're looked at too carefully. They're not, very broadly, if not always, they're not exactly correct. All right, so, well, maybe we could still bail out by admitting that the theories that we have today are all faulty, but we should still think of science as trying to put together an exact way

20:00 of understanding things, a completely exact and comprehensive theory, what in philosophy you refer to as the Persean limit of inquiry. that should act as a motivation to try to make our theories better and better well fine one should take one's motivation for where one will but in so far the problem is to understand the science that we have today or any science that's remotely like it with respect to the shortcomings that it has I submit that what goes on in the pursuing limit is just quite irrelevant So, that's the problems with fundamentals. Now, I want to turn to the repercussions for truth and knowledge. As philosophers, at least, try to analyze and understand the notion of knowledge, knowledge is stated to require truth. However, we have much less of truth, very little exact truth in physics, and so if knowledge requires truth and given the ubiquitous use of idealizations in physics, it looks like we know much less in physics than we thought we did. physics doesn't after all give us knowledge that's nuts, that's crazy physics, the sort of understanding of the world both practical and theoretical and intellectual ought to be a paradigm of what it is to know about the world so let's turn things around and start from the conviction that we do have knowledge does give us knowledge. Of course, if we do that, then we've got to let go of the idea that knowledge requires at least the philosopher's notion of exact truth. So what I'm going to be doing now is looking at the process of letting go of this idea or transforming the idea of knowledge and truth in a way that enables us to understand the kind of idealized we have as nonetheless providing us with real knowledge. So let's start with a quick review of how, at least in philosophy,

22:30 we understand the idea of knowledge. The framework that we use almost universally is that knowledge is true, justified belief. So I'm going to try to put that neutrally. Instead of talking about truth, I'm I'm going to say knowledge has a content component. A knowledge statement represents things as they are. A successful piece of knowledge, a real piece of knowledge, represents things as they are. And then to have knowledge, you also have to have justification for the claim that things aren't represented as they are. And then in traditional usage in philosophy, representing the way things they are is traditionally understood as the notion of truth. thing I want to call here, whoops, sorry, I'm going to use the philosopher's notion of truth with a big T, okay, and what is this traditional conception of truth, there's an enormous amount of discussion about this in philosophy, but what I'm going to need is that whatever it is, it's something that's not graded, it's not context dependent, it's not qualified in any way, something, when we say something is true, it's just plain truth full stop. You don't have to qualify it or say only relative to a context or anything of that sort. Now, how are we going to reconstrue that? Let's again go back to the idea that our representations of the world may be successful, may be unsuccessful. Sorry, what's involved when they're successful is that they represent things the way they are, alright and when a statement represents the way things things the way they are again traditionally the idea that the label that's put on this kind of epistemic success is truth but of course there are all kinds of other representations for which this idea of capital T truth doesn't apply, so the analogy of a map I'm going to use the map analogy broadly here, but this really is a stand-in for all kinds of representations, figures, graphs, pictures, and many more technical things that go on in science.

25:00 Success in representing the way things are for a map is very different from what appears to be the requirements of success notion of truth. A map is successful if it's good enough, if it enables us, if it's accurate enough to enable us to find our way around to the kinds of things that we're trying to find with the accuracy that suits our current needs. So we start with this opposition between truth as success for statements and accuracy as success for maps, and now I can put the kind of thing I want to do really as a one-liner. We should try to think of truth as a kind of idealization, that the way our statements succeed in representing the world is a kind of success that is more like maps than we usually appreciate that of course when we idealize well what kind of an idealization we have a statement uh the statement if we work carefully enough is not going to correspond not going to represent the world exactly the way it is but for many purposes and we have a successful statement, we can put that aside. We can treat it as if it were, in the philosopher's sense, just plain true. Like so many idealizations, this is an enormously useful idealization. I'm not suggesting that we should give up the notion of truth for an enormous amount of users, just like the idealizations in science. It's an idealization that's basically that we couldn't live without. But we do have to remember that it's an idealization. And a current problem, I want to suggest, the problem of understanding how it is, or what kind of knowledge, or the way in which physics can give us knowledge, why that appears to be a problem, I suggest, is really an artifact

27:30 of neglecting the point that truth is an idealization. So instead of thinking about truth as corresponding to the world flawlessly, let's go back knowledge as a justification component and a content component. Let's understand the notion of the content component of knowledge in the spirit of that the content is supposed to be advanced as adequate I hope these informal phrases will get across the sort of thing that I have in mind. Where this sort of success for a statement is just like the success for representational success for something like a map is going to be something which is understood in the sense of precise and accurate enough to meet current standards. So it's going to be relativized to what our current needs. It's very much going to be a contextual item. So in sum up to this point, we need to understand statement success truth as really working like success for a map. By doing that, if our standard of success is like success for a map, And our idealized theories aren't perfect. If what we require in knowledge is statements that achieve the kind of success that a map has, this immediately reinstates knowledge throughout physics. And this is the sort of thing, I'm going to be talking a little bit more about this, but this is the sort of thing that I have in mind when I talk about provisional knowledge. Now, again, to emphasize the comparison, traditionally, knowledge is thought to require capital key truth. We can think of that, and I think this is the way people in philosophy, and probably many people generally think, the world has determined objects, they're determined properties, you've got words for these things, when the word applies to the property, applies to the thing, you put the words together, it's just plain truth. It's not thinking of statements like a map. In provisional knowledge, instead, we think of evaluation in a more map-like way. We ask, which levels of precision and accuracy do we have?

30:00 We call it knowledge when they are good enough broadly for our present interests, purposes, and standards. Well, okay, this is a sketch, and all I'm going to be able to do really is to give you a sketch of this idea. Let's look at it by circling around it. Let's do that now by looking at some problems that may arise in thinking about knowledge as proficient. If knowledge, if a sort of statement of success is just being advanced is adequate or being true enough, does that really give us something that counts as knowledge of an independent reality? There's an awful lot to be said here. There's a low option, and just again to go back to the map analogy. Yes, a map gives us objective knowledge of independent reality, and whatever that method, whatever that way is, it's going to work as well for this notion of provisional knowledge. There's a lot more that can be, this is a kind of low option. We can look at higher options. If I have time at the end, and I don't think I'm going to, but if I have time at the end, I'll try to say a little bit more about this. Then a worry that I get from my philosophical colleagues is you can't use language to describe the world without the traditional notion of truth. Don't you see? To say that snow is white is to say that it's true that snow is white. If you're not, conversely, if you're not taking that to be just plain true, you aren't saying anything at all. At least, I don't know whether people outside of philosophy find this problematic, but I find that most of my colleagues are very distressed for this kind of reason. There's no help for it. The world is very complex, and the best that we have been able to do so far, and the best that we clearly are going to be able to do in the very near future, is the kind of knowledge which I'm describing as provisional,

32:30 the kind of knowledge that we do get for a man. So in our complex world, we have no choice but to work with provisional knowledge. And indeed, I don't see why this should be seen as a problem. Indeed, when you go back and look at these very same philosophers who are so upset at my trying to wrench, capital T, truth from them, these very same people, they agonize about how to understand truth. And one attack that they take is, don't ask too many questions. Just use the notion of truth. don't try to make sense of it in terms of correspondence to things out there in the world and properties and this picture I gave of objective properties attaching to objective things just use the notion of truth and you'll get along just fine well we can take as a low option again and it should be adequate enough we can use provisional knowledge on its own terms kind of spirit that we use the knowledge that we have for math. But let me try to say a little bit more about what it means to use provisional knowledge in its own terms. Here's what I have in mind. It's back to this idea that use of truth as an idealized notion is, in almost all contexts, just the right thing to do. And the way I like to describe things, we have a range of ways of thinking about the world that we take as basic for the moment. Now, if we stand back, we know they're not exactly correct. But for most purposes, theoretical, practical, we don't have to worry about that. And so we treat this little microcosm. You can think of this as a kind of microscopic Kuhnian paradigm. we treat it as if the world were just like that and we make mistakes because we don't apply it right we make mistakes in observation but by and large we get along very well sometimes however we run into trouble and one of our options is to say maybe there's something wrong with this platform so then we construct a new platform that from the point of view of which we can handle the problems we couldn't handle before

35:00 I don't know if universally, but generally, we want to be able to look back from the new platform to the older one and understand why things worked as well as they did. A straightforward model for what I have in mind here is understanding the success and continuing to make use of the successes of Newtonian mechanics seen from the point of view of both relativistic theory and quantum mechanics. But at any one time, there's going to be a... I'm clicking the wrong button here. At any one time, there's going to be a bottom-level platform at the moment. How is it judged? Well, tacitly, what I've been saying is it's being judged in a practical way. If it serves our purposes, not only practical, but if it serves our intellectual purposes in understanding the world, then we treat the world as being that way, and we're very satisfied with the situation. But since it is impassively being evaluated as pragmatical, this proposal clearly has something to do with pragmatism. And now, again, the philosophers stand up and say, look, we know that pragmatism is rubbish, so this whole proposal must be rubbish. So the next thing I need to do is very briefly explain what the pragmatism is supposed to be and what the difficulties are supposed to be and why they don't apply to the kind of system that I'm trying to schedule. The idea of pragmatism is very simply that you shouldn't understand truth in this traditional correspondence way. You should somehow interpret truth, again, as working. To be true is to work. To be true is to be known to work. The notion is formulated in a number of different ways. The first way that I'll talk about is that to be true is to be found to work. This is usually put with the phrase, truth is assertibility with want.

37:30 Now, lots of things for which you have very good evidence, the best possible evidence, So truth can't be assertable with Warren. It can't be true working in the sense of having been found to work so far. And my account certainly isn't pragmatism, this sort of epistemic account of truth. In the way I'm thinking of things, the justification component doesn't itself measure the success. it gives you evidence for that you'll be able to count on the content component to continue being successful. So the other way of understanding pragmatism is the phrase, to be true is to work, where working here now is to understand not having been found so far to work, but whether we are aware of it or not, we may be mistaken, whether it in fact works. Let me illustrate with a concrete example. design, a new kind of widget, maybe this widget is supposed to make my car run better, and I may have evidence that this widget is going to work well, I may not, maybe I'm just guessing, maybe I'm just hoping that it will work, but there's a question, will it work or will it not? It's this objective sense of whether in fact it is going to perform in the relevant and endeavors up to the standards that I have in mind. So what's supposed to be problematic with this second way of understanding pragmatism? Well, the objectors say, look, lots of things that work aren't true. Suppose I've got another five. Yeah, go ahead. suppose I have the false belief that my size 9 feet are size 10 then the belief that these size shoe are size 10 shoes is going to work just fine if that's mistaken, if they really are size 9 shoes There isn't any way that that would count as being true. Well, this kind of objection to pragmatism, I agree, is fatal if you take the truth to be an all-or-nothing affair.

40:00 A statement can work, can work extremely well, as well as they ever do, but still be, in the strict sense of the word, in the sense of capital T truth, it can be false. But you see, this complaint just is completely wide of the mark if we're thinking of success for statements more like success for maths. If what we're demanding is not complete accuracy, but accuracy and precision enough for current purposes, then there's always some latitude some slack and if it turns out that it wasn't exactly correct then we haven't shown that relative to the standards that were in play at the time that it shouldn't count as true. Well, now wait. Do even I want to say that in the little scenario I gave that the statement that my, these, what are in fact shoes at size 10, do I want to count that as true? No, no, that's not a sensible proposal. What we need to do is to extend the map analogy. Surely the notion of truth is going to be an applicable term. We want to use the word true, little t true, for success in an amply broad range of cases. Well, just how is that going to be worked out? I haven't begun to do it. What I hope to have shown with the argument that I've given, is not, well, sorry, let me put it differently. What I'm not claiming is, to be true is to work, is to work well enough, therefore we understand truth, job done. This is not what I've suggested. What I'm suggesting is that pragmatism is a viable program. I'm arguing for pragmatism as a program rather than a slogan. The problem that was cited when you were thinking about capital T truth, that problem was, sorry, the problem that was raised is fatal if you're thinking about capital T truth. If you're thinking about capital T truth, then the program really is dead on arrival.

42:30 But when we move to this much more accessible way of understanding truth and knowledge, we have at least a program which is worth moving forward. Now, I've talked for 40 minutes, and I'm at a good stopping point. I could talk for another five minutes. I could talk for longer, but I won't. You can talk for two minutes. I will talk for up to another four minutes. You know, yeah. We're inflation. Let me begin. There's still a hankering of a worry here. How do we get a sense? If I say that we think of truth as being true enough, as being like a map, and going back to the point about Kantianism, that this is something we can construct, you know, the objectivity is wise, there's no problem with that in the fact that we construct these representations, but they get slapped down very, very frequently in a way that we can't control. So that would seem to be objectivity enough. But in what sense is this still reflecting a world independent of us? Of course, it's the Kantian attitude that that question, in a sense, doesn't make sense. How do we, we've got a hankering after it, and I'd like to suggest a way that we can think about it. I'm done here, so let me turn that off. I want to follow up on a suggestion that came up briefly in Matteo's talk yesterday, that we can think of what is independent of us as a kind of potentiality that we can think that, and now as a metaphor, not literally as a metaphor, we can think of the world independent of us as a kind of grand Aristotelian prime matter. Is that to say that what hasn't been conceptualized is only potential but not actual?

45:00 No, that's not the proposal. That's not what I'm suggesting. I'm not suggesting that things that in places that are very removed from any conceptualizing creature, very removed in time and space, that those things are somehow not real. I'm not urging that. I'm also not asserting that. It's part of the contention here of the framework that I find so very attractive that it's a question that literally doesn't make sense. And if that comes up in discussion, I can say a little bit more about why I think it doesn't make sense. But we can still think of there being something independent us in the fact that we probe the world this way and we get one reaction, we get probed things with another set of concepts, we get another reaction. You know, the world may be an incredibly complex place. We may only be able to evince a tiny fraction of the reactions of the activated dispositions which in principle could come about. What it is to experience the world is to tickle it and to get a response, to tickle it with our conceptions and get a response. This is what it is for things to be real and I want to suggest this is really all the kind of independent real world that we ought to ask for and makes any sense to me. So, good. I stopped half a minute short, 45 minutes. Thank you. Questions? Oh dear. Brigitte, first. We were fast again. Okay. Two points. I remember Yes, yes, Tuhlman did. This analogy has come up a number of times. Tuhlman is a prominent example. Tuhlman's book fell on completely deaf ears.

47:30 The world of philosophy is gripped by this idea of capital to be truth. It was rejected. And what is different about the situation now than I think was then, I don't know, 30 years ago, 40 years ago. I think it was a good bit longer than 20 years ago. At that time, many, many more of us were in the grip of this ideal from the Newtonian period of having, of getting, thinking, making, we had in parts and we were constructing a uniquely, absolutely correct, absolutely comprehensive theory. 30, 40 years later our consciousness has been raised that you know God knows what will happen in the prosely limit. That's not the knowledge we have now and it's not the knowledge we're going to have in the next 10 years in the next 50 or 100 years and you know quite possibly possibly never. Again if our task as interpreters of science is to understand the nature of knowledge that the kind of science that we have now Having our consciousness raised, we can go back to Tuleman's analogy and reinvestigate it and appreciate it really how important it was. I have a second short question. There was this conception of getting closer and closer of truth and this conception of problems finding a distance to truth and problems like that. is you related to that or would you let me tell you very briefly what I think about this idea of thinking well our theories are not true but they're approximately true there's been an enormous amount of criticism of the idea of approximate truth and I view that criticism as absolutely correct if the notion of approximate truth is supposed to be context independent, an absolute notion but here's a better way to think about approximate truth what is it for a statement of theory to be approximately true it's for the world described by the theory to be similar to the way the world actually is ah wait a minute similarity is not an absolute notion two things are similar in some respects they're dissimilar in others I've cast approximate truth with the word similarity

50:00 exactly the same reservations apply. Similarity is a relevant notion. And so we do have a notion of approximate truth by specifying the standards, the respects in which we call for accuracy and precision. But that's not to have any reason to think that we're on a convergent path of things which are absolutely, in an absolute sense, in a context-independent sense, true, for which there be some measures so we can see ourselves approaching the Persean idea. Chris, then Howard. Thanks. That's a really, I enjoyed the talk, very provocative, very nice, but I don't think your argument in the first half establishes your position. It establishes you need to introduce some looseness somewhere, but then there's a choice about where you introduce the looseness. So you introduce the looseness between what is said and the world. You have this sort of similarity relation. The standard view says you have a little bit of looseness What exactly is us if? But then the relation to the world is type. And you've given us no reason to suppose we should do the former rather than the latter. Now, here's some reasons why we shouldn't do the former. Which was the former? So the former is your one. The latter is the standard one. And, okay, give us the latter, the standard one again? Yeah, the standard one is where you say you allow a little bit of looseness around exactly what's expressed by a given sentence, but then you have a type relation to what happens to the world. And the reason why one might opt for the latter one, for example, is that a deflationary account of the representation of truth relations is going to be available in the latter case, I think this is kind of what some of your colleagues may be getting at. And the trouble is, I mean, it looks like you're going to need both a tight relation and a loose relation in yours, so you've got more kinds of resources than in the previous one. And also, you probably owe us an account of meaning, because you're going to be severing the relationship, the standard kinds of relationships between meaning and truth condition. So what's your account of meaning that's going to underwrite this picture? Okay, I'll give you a really quick answer, and then and gesticulate it a bit longer in this way. Snow is white, the sentence snow is white is true if and only if snow is white. Now if we're going to understand snow is white statement that snow is white as a not completely precise statement, then the notion of truth on the other side is going to have to be deflated from capital T truth to little t truth. Add the comment about approximate truth being in context relevant in the sense, and I think we end up

52:30 at the same place. That's the short answer. The long answer is that, and I'm not going to try to give you this, I just, you know, I'll tell you what I'm trying to do. Yes, I've gotten this exactly your comment many, many times. The first time I got it, I had no idea what to do with it. I've got a body of work in which I argue that there's a between false exact statements in true vague ones. And that these are different ways of accomplishing the same representational aim. So where you saw two things going on, I see two facets of the same thing. But it would take me too long to try to fill in the details later or I can send you stuff to read. Okay. Howard and John. So this may be something you can clear up quite quickly, but when you are offering an account of truth as working, I take it there has to be some representational component. In fact, you just mentioned representation again that has to be added to that. Because the widget could work, but you wouldn't say, well, the widget is true. So there is some notion of representation, and I wonder how you get that going, and maybe related to the question about... Guilty is charged, and this is the point of my having included that line. What I'm claiming is not that we figured out the notion of truth by appealing to this slogan, to be true is to work. This is a program. Words by themselves don't do any work. What does it mean to say that a statement works? There's an enormous amount that has to be filled in here. I'm claiming is that this prima facie objection, which has the consequence, don't even bother trying to figure it out, which I think was a corrected objection, if you understand truth is capital T truth, that no longer applies. And so this, to try to answer your question now is a sensible thing to try to do it. But I have, that's enormous, this is an enormous project and that's a very important part let me amplify let me amplify that in just one very quick thought in many ways I'm doing things that other people have been doing

55:00 there's one element that I take to be new I feel very strongly that people have not been paying attention to the fact that all our representations fail of complete precision they're all vague, indefinite They don't give us any, no representations, exceptions in mathematics aside, are not completely precise. You take that into account, a lot of issues in the Kantian circle, a lot of issues of the sort I've been talking about, you take on a very importantly different cast. John, then, Pallop. Is there a relation between provisional knowledge and the differences produced by Gilbert Hyde, Labicard, and others between, on the one hand, to know that as propositional declarative knowledge, and on the other hand, to know how as propositional knowledge. Is, in your propositional knowledge, a sort of shift to know that, science as to know that, to what science is conceived of as a way of knowing how using sophisticated models approximately true? Good question. I don't know the answer to that question. It's a sensible, I think a very good question to ask. very closely related to Howard's question. What is the sense of working? Can we get at it by trying to reinterpret it, at least in some aspects, knowing that as or as a reflection of knowing how? This is a big research program, and there's another part of it. I'm not prepared to give you a good answer, but I welcome the question. We have five more minutes. Paolo, then Alexei. Page 66, the last page. Let's see if I can get this thing started again. There we go. So, whoops. Now what do I have to get? I'm going to get this. So 67, you are. Wrong one. It's the stuff that I didn't get to. Okay, yes. He presents his ideas in the context of the reconfiguration of the idea of knowledge truth.

57:30 This is a very good idea, I agree completely with his approach to the problem. But I have a question, the first, to be true is to work, how can we not be able to admit some imprecision and inaccuracy? My question is, why speak of the possibility that truth can admit some imprecision because we know that this idea is against the usual notion of truth you might speak because I ask for every configuration but in this case it seems to me So, it is possible to say we have the usual notion of truth and we must input the precision, the accuracy to the model we use and we say, I do not claim that these laws, these statements are absolutely true, these statements, these models and so on are true within some limits. what is the difference if I explain it? Yes, yes. Another way to cast this and this is a one way to try to start to work on Jean's question is to say that the traditional notion of truth does apply within a model. I'm not in any way arguing that claiming that the traditional notion doesn't make sense It's a perfectly coherent notion. We know that because we can model it. And we use this notion inside of our models. And then we take our models and say, oh, our models work in this way and fail to work in this other way. So it's really exploiting that idea with which I I try to develop, I have a program for developing the thing that I mentioned in response to an earlier question,

1:00:00 that a false but idealization, which however functions as a truth on the one hand, and a corresponding imprecise statement, which we can unproblematically regard as true, that these are really just two different sides of the same coin. it really is just briefly it's just the point that you made we can take the idealization is it true inside the idealization does this idealization I hope to be able to say more about this someday does this idealization work predictively, intellectually in all kinds of ways if it does then we have a corresponding statement which we can which is imprecise but we can say without any hesitation or reservation that it's true the hydro Euler's equations are strictly speaking false but water does behave approximately as if it were governed by Euler's equations that I take to be an unproblematically unproblematically true statement We have just three minutes. We would say that the platform of the moment can only be evaluated pragmatic. I would like to know better what you mean with pragmatic. Yes, it works. but what is with this evaluation the role of the methodological use because it means that the platform is a good platform because of his simplicity of his power because the only thing I can say and this is I think really John's question you take these idealizations you treat the world as if it were like that and you don't run into trouble that's the best i can really do at this point by saying what i mean by working okay let's go yeah yeah alexa then catherine then okay i just wanted to tell a short story now there is this uh work in quantum foundations who's pretty

1:02:30 really well known uh to to many of us in this room whose name is chris fuchs Thank you.