String Theory — Influence on Philosophy of Physics & Science

0:00 Thank you. Thank you. Thank you. I'm sorry about the late start, I wanted to give Dr. David a cup of tea before, and you were coming from another seminar, in philosophy of physics, no less, so welcome, and this is the last seminar of terms, thank you, and it's a great pleasure to ask Richard, who's actually flown yesterday, all the way from Vienna, to tell you what string theory has to say to the philosophy of science. So, we thought I'd like you to come on. We look forward to talk.

2:30 Yeah, it's a great pleasure for me to be here. Thank you for the invitation. So, today I want to talk about that possible philosophical implications of string theory. And I'm afraid that David will probably know something of it, because the book of York is, in a sense, included in what I would say today, what it would be really well. So let me just start. I want to say... I will first say just a few words about string theory and about the stages of string theory today, about the general situation, what's somewhat strange about it. Then I will talk about a concept that will be important for my argument, which is the the concept of scientific under-determination, and I will try to define it and humiliate it from other kinds of under-determination. And I then come to the question, how string theory may weaken this concept, this principle of scientific under-determination, which will basically be the main argument that this might lead for this kind of different understanding of scientific progress in the context of strange viewing, and then I will have to ask you a few minutes. So, let me just say a few words. It's not only very basic points about what string theory is about. The goal of string theory is to provide unified descriptions, description of all natural forces, also the current elementary part of physics and gravity.

5:00 And the fundamental principle, basic idea is very simple. So, it is to replace a point-like character of the elementary particle, as we are used to them in particle physics, by extended objects, wonderful objects, which are called strings. The hope is, or the assumption is, that this replacement can solve the minimization problems, which are usually there in quantum gravity, and lead to a finite theory. So, just to give a short history of Swim Theory, Swim Theory is not really a very new theory anymore actually, the concept was already proposed in 1968, however in a quite different context in order to describe the strong interactions. And when it turned out that in this context it had considerable problems but in chromodynamics probably was the more promising way to go it was suggested that string theory instead could be used as a theory of all known interactions. This was in 1974 and after that there was basically a rather smaller group of people working on scientific theory and it was considered a rather speculative theory quite beyond the mainstream of elementary mathematical physics and it had quite considerable theoretical problems. In particular, it wasn't clear whether it was impossible at all to formulate a theory that could include Perkins. And ten years later, in 1984, there occurred a so-called First Spring Revolution, where people for the first time were able to formulate a coherent theory, string theory,

7:30 which also includes permanent, so-called, string theory. And this was the start of the big evolution of string theory when an awful lot of people went into string theory and lots of developments happened. String theory is a very influential and, in a sense, central subject in the context of elementary particle physics. In the beginning 90s, once again there was a kind of crisis in the context of string theory. people didn't know how to handle the strongly interacting sector, and in 1995, the second string revolution once again boosted the enterprise string theory by understanding the so-called string duality relations, which offered a first glimpse into how one might be able to handle the small couple sectors which led to many different developments in the 1990s. Right now we are in a situation where no grand progress is happening and where people are once again waiting for the next crucial step to be taken. So it might be an example of misplaced meta-induction to itself. A year later, around 2005, we will enter the third string revolution, but that's probably what string theory is. A few words about the alternatives to string theory. one must distinguish so one has to distinguish between question question how to unify gravity

10:00 and quantum mechanics fundamental principles and in this context the answer whether there is yes. There are several approaches in canonical quantum gravity. In particular, there is the approach of quantum loop gravity that tries to deal with these problems. In the context, if you ask the question how to unify gravity and elementary particles in its current state, The answer one has to give is no. There is no alternative theory that really got off the ground and really can represent a serious contender today. to shrink theory. So we are facing a rather unusual situation, because shrink theory is considered in terms Because of the traditional ways to evaluate this theory has a very important background. It is not all automatically confirmed, as we say. which means that the scale will be the characteristic prediction of the screen theory, the extendedness of the object and the existence of additional dimensions and stuff like that, where they could be tested are according to conventional models very far beyond our today's experimental capabilities. actually there is there doesn't exist any idea even in theory how one could be very much tested but even beyond the string theory is not even possible to reproduce the empirical data of physics which we which we know today so in this

12:30 For example it must be distinguished from concepts like brand unified theories where we know that brand unified theories have a, some of them do have a low energy limit which can be able to fit today's data. In the case of string theory, we do not know that, we do not know whether there exists a low energy solution to string theory that actually can reproduce what we seem to bear. And the second point, which is related to what I have just said, is that string theory is not a full reflection theory in a way, it's by no means fully understood, and many of the crucial and most important questions are still open, and it's not here, or in the possible future wouldn't want to predict that they will be solved and this second point is actually related to the first point because string theory might be a highly predictive theory for low energy physics so it might have only a very narrow regime of possible low energy solutions which means that it might well very well be wrong in low energies but we have no we don't know the yet don't have the means to calculate the low energy predictions so these are the problems of string theory so to say on a sociological level string theory is This is very strong. It is highly influential today, and in a certain sense, well established. Influential because it, you know, some of the adjacent peaks to string theory, for example, you know, to a part of the modern building or cosmology are very strongly influenced by string physics. Many, a large share of their models is actually based on some stringy established. And on the other side, there is quite a considerable share of theoretical physicists working in string theory since the West Wing Revolution 1984.

15:00 And probably, well, at least if one looks at the citation indices, when we find out that in theoretical physics, string papers for a very long period by now are the most cited papers in theoretical physics. So in this sense, it is a very strong theory, and it is also believed quite strongly by the majority of string physicists, I would say. So if you would, the majority of string theorists would argue that they would be quite confident that string theory in some way represents a very important step for a better understanding of nature, even though it is still a speculative theory. So, all these advantages, positive points, and also the negative points about specific theory seem to be fairly stable and are not likely to go away any time soon. So, we are witnessing a situation for 20 years by now where the theory seems to defy what one the characteristics one would expect in relation to a certain theory and the question is why whether there is anything to learn from this situation and if we face such a situation and it seems to be stable it might be interesting to look for for possible modifications of some concepts in philosophy of science which actually account for this new situation. And what I will try to do now is to suggest one possible way to do so. So, well, the concept that will be crucial is the concept of what I call scientific under-determination.

17:30 And this concept should be clearly distinguished from other kinds of under-determination which which are very prominent in the literature, for example, well, maybe one should explain this diagram. One could distinguish two possible ways of distinction between different kinds of underdetermination. First, one could distinguish between underdetermination of theory-leading by all possible evidence and underdetermination by available evidence. And underdetermination by all possible evidence would come up to empirical equivalence in the way this, for example, discussed by Quine or in the arguments of Van Fraassen. While we are interested here in underdetermination by the evasible evidence, so the scientists had a certain set of empirical evidence today and the question is how far this determines the theory of evidence. The second distinction one can make is the distinction between underdetermination logically and uncreatively in the sense of Mary Lord which means logically one would say that one would ask the question whether it is logically possible to have a different theory or a different prediction for the future than the one our present theory would give us. One example for that would be human induction problem, or another example would be the main point thesis. While we are not interested in these questions, we are interested in the question whether there exist other theories which are, in a sense, scientifically acceptable. So, this of course is a difficult notion, but one would assume that scientists would have a certain intuitive notion of what can count as a scientific theory, and if they

20:00 are faced with a certain set of empirical data, they can ask the question whether they would believe that there exists an alternative scientifically satisfactory theory that could describe this set of data and this specific kind of underdetermination I want to call a scientific underdetermination and it is important to know that any statements about scientific under-determination do not carry any implication for human induction. So those are two separate segments. Okay, so the principle of scientific under-determination to begin with is implied by our notion of theory succession meaning theory succession where where the one theory is not where theoretical progress does not just constitute in adding to the existing theory but constitutes in replacing one theory by a conceptually If that is so, I must assume that my present theory and potential future theory, that one day will replace the present theory, are both compatible with present data, and therefore I must expect that scientific underdetermination is there. So, scientific evidence will be the foundation of this understanding of theory succession. On the other side, we would say that there seem to be some limits to science under determination, which we find first because we see that science is often capable of predicting new phenomena correctly. And we wouldn't have any explanation why this happens if we would assume that we have at each stage in the scientific process

22:30 We have an entirely unlimited spectrum of possible alternative theories which might produce different predictions concerning real-time color. And the other point here would be that there seems to be a certain tendency for consensus in science concerning the most appropriate theory in a certain state, which will also speak against the unlimited nature of under-determination. So we see that there is a certain tension here between the demand for underdetermination and the limitedness of underdetermination that characterizes the scientific process in general. And one would have to draw the conclusion that there is, of course it's not possible to do this explicitly but there is a kind of balance between the two of them and one might try to ask on a scale from from one to infinity where our expectation of other determination in specific stage of the scientific process would be and therefore it would be a natural consequence to expect, or at least to accept the possibility, that this expectation might change within the scientific process longer. So, one more point is that this principle of scientific underdetermination is the foundation for for science's focus on empirical confirmation because if we would not some scientific animation we would have a we would assume that there is a very high possibility that the theory we have adapted to our present empirical data will also be able to predict all the future phenomena which are implied by theory.

25:00 And therefore, in that situation, we might assume that theoretical confirmation, arguments of theoretical consistency, similar arguments, might play an equally important role in the context of theory confirmation as empirical confirmation. This obviously is not the case and the reason for the fact that it is not the case is that we have this assumption of scientific undetermination. Now the next step would be that if we observe in the context of string theory that string theorists seem to give more credit to purely theoretical arguments from within their theory than one would usually allow a scientist to do. This might be a sign of the fact that string theorists actually apply a different interpretation or a different assessment of scientific underdetermination. And this will be the hypothesis that will be tested. Of course, if I just deduce this weakening of scientific underdetermination from the way string theorists behave, It's not an awfully interesting statement because it doesn't say anything about the theory and it doesn't explain in any way what is going on. So what one has to do is one has to look at the theory itself and try to see whether one finds in the context of the theory itself any arguments which support this weakening of the plan to begin with the imagination. So, okay, she raises the question, how scientific other determinations might be? So, in this picture, I just, I just showed the basic scenario where I have particle physics

27:30 on a small sample of relativity. Each of the theories are based on empirical evidence. apparently there is no way to unify those three theories directly and what spring theory suggests to do is to introduce those assumptions of extended elementary objects and this triggers the spring theoretical research program. So, we have this arrow going towards something which no one is able to understand today and no one is able to tell how this theory will eventually look like and that is string theory on top. And now there are two different possibilities how the weakening of scientific underdetermination might happen. It can happen at the level of the string theoretical research program. And it can be related to the way string theoretical research proceeds. It can be related to the arguments which arise from this scientific progress which suggests to the string theories. that the theory might be viable even without any empirical evidence. And on the other side, at the second level, there might be arguments which come genuinely spring here, which emanate from the theory itself, or from what we believe to know about the theory already today, which shift the understanding of the role of scientific and determination. And if you turn out that on both levels, one can find candidates, which might actually do the job of becoming a determination.

30:00 First, have a look at the first level. So there are the principle that there are three possible ways to read it. I know there are actually three lines of argument in the context of strings. The first line of argument would be the argument of no chance. But string theorists often explicitly say that from how they understand the overall context and the way scientific progress moved towards string theory, they believe they don't have any choice, other choice than any unique string theory. So, we already talked a little bit about the possible question of alternatives to scientific theory, and I mentioned that at the level of full unification of all mature forces, today there is no viable alternative. One might add that this wasn't always the case, people actually have tried out several alternatives. One alternative that was quite powerful for a while was the idea to try to get towards a finite quantum theory that involves also gravity based on specific internal symmetries, based on higher degree supersymmetry. And eventually it turned out that this strategy couldn't work. And from And in continuing this line of argument, people today assume that based on very general physical assumptions, the extendedness of objects would be the only viable way to allow for a

32:30 I mean, I will find a solution to this problem. Now, of course, there are those very fundamental assumptions, and one might imagine that one could also attack some of these assumptions. And people have tried to do so without any real success, I'd say. to this question in this nature. And now, for the time being, I just want to say that people also tried to find very general arguments in favor of the assumption that string theory is the only possible choice. And in several contexts, these arguments actually were For example, there is an argument by Kuczynski, where he starts off from a very different fundamental assumption, kind of position-position uncertainty, and it turns out that if you start with this uncertainty and you try to create a coherent model out of that, you eventually are forced, so to say, back into string theory. So there is this idea among string theories that if you try something different and try to make it coherent eventually you will end up back in a distinct theoretical context. Now of course one must say that these arguments of no choice are not invented in the context of string theory. There have been arguments of this kind before. For example, there is the case of deduction from the phenomena, which has recently been discussed quite intensely. One might just highlight one difference that might exist between this deduction from the phenomena way of arguing and the string theory of arguing. And this is that in the case of string theory, the argument is in no way any more related to any human kind of human intuition. It's an entirely mathematical consistency argument. And therefore might carry more force than in various stages of science.

35:00 Now, one second line of argument actually in a way merges with the first one. This is based on the success of the Stadler model. And which means it is in a sense based on the longer history, on the earlier history of elementary particle physics. And it puts string theory in a more general context and argues that the mechanisms which lead one to have more trust in string theory than one actually might want to base just of the reason of considerations of empirical confirmation, that these arguments are embedded in a more general context of scientific evolution. The main point here is that the standard model, which is the physical theory that today is able to describe all elementary particle phenomena without any deviation. So the standard model can reproduce all the existing data in the three particles. And this standard model actually has been created along the same line, the string theory has been created later. It was created in order to remove a purely theoretical problem without having any empirical evidence that at the moment of its creation that the fundamental assumption of the standards are correct. So this at that time was the problem of the re-anormalizability of the electrical strong interactions. And the introduction of gauge field theory in order to cure these

37:30 problems and the introduction of dynamical of uh electric symmetry breaking in order to make this really massive they were just based on purely theoretical considerations and after the the model had been set up it could be um confirmed by by theoretical evidence So I would say most of, most aspects of the theory were confronted with computer science. And so what Springfield says that there is one apparently very successful research program going on. And in one case, we have seen that this principle that you, so to say, you knock on all possible doors and you find precisely one opens and there you enter that this principle is very successful and it has been in the case of the standard model successful for many decades actually. So it might be reasonable to assume that it is successful even in today's context. So what one has to add here is that of course in the case of the standard model, in the beginning 70s, did not mean that you would exclude that there are other possibilities to solve the problem of randomizability. For example, one might have thought about using the idea of extended objects already back at that period in time. So what the example just shows is that, and actually what people already also were aiming at that time, was that a simple and very convincing solution to a certain theoretical problem in a certain context will be able to to predict future empirical phenomena correctly. And if there arises at the later stage the necessity to replace other more fundamental physical principles,

40:00 this should be expected to imply the predictions of these, of the standard model of, let's say, general terms of this first theoretical concept in the end. And this is also what is proven by the standard model and nothing more. Therefore, if one goes back to string theory, the fact that one works on string theory or puts trust into string theory today truly does not mean that more fundamental physical principles won't be changed in the future. What the string theorists would just say is that if one would try to modify more fundamental physical principles, in the end, this modification is carried out should imply string theory, not together. So this is the question of success of the standard model. And then there is a third argument, which I'm going to go into. And this is the point in terms of string theorists argue that, And actually one finds that often in string theory, arguments which seemingly have been, seemingly unconnected before, join into one coherent whole. and this might be understood in the same terms on similar terms as empirical evidence which has not been let's say the prediction of empirical phenomena which had not been expected when constructing the theory but once I you can say if you construct a theory and theory unexpectedly predicts new phenomena and eventually you find those phenomena this is the strongest way to confirm theory in a similar way you could argue that if a theory is constructed to solve a certain problem and then it it has a certain amount of

42:30 collateral advantages, so theoretical advantages, situations you cannot understand better in this new framework than in the old one, without having constructed the theory to do so. This might be also some kind of theory combination. In the case of string theory, this is the case in the context of gravity, for example, where you introduce string theory to produce a framework calculate gravity, and at the end it turns out that actually spring theory enforces gravity. And spring theory also enforces gate theory. So, when you started out, you didn't want to deduce those two theoretic concepts, but you just wanted to connect them. And then it turns out that you can and actually deduce it from this principle. This would be one example. Other examples would be related to supersymmetry, different ways to understand supersymmetry, and then adjoining to a coherent whole in the context of space. And then there is a question of black hole physics, and a question of the dimensionality of the background space, and things like that. But I think I should better go to the next point. Level 2. So, the generally string theoretic arguments, which might affect certainly under determination. So here we have one important property of string theory, which I want to call the property of structural uniqueness. And structural uniqueness, I want to call it structural uniqueness, and actually it's two different statements. First one, that string theory has no free

45:00 properties, which means that you don't have anything in the theory that you can, in the fundamental theory that you can tune freely in order to fit the empirical data. There is only one sort of initial constant, which is the string length, but since everything else can be derived from the string length, this is just an example of the division. And the second point is that string theory has no freedom of model being. So if you insert the concept of extended objects into a quantum system, and you try to build a coherent theory, the only one possibility to do so. Initially, it was thought that there are five different coherent models in theoretical models, but then, during the second string revolution, it turned out that those five models are related by the so-called string dualities, so eventually there is a conjecture that these five models are just different combinations of one general theory. So put this together we have a remarkably rigid structure of string theory which might translate into a very strong predictive force of the theory. And now one must say that it is not entirely clear that this is the case because what I was talking about so far is only the fundamental theory, the structure of the fundamental theory, while what is responsible for the low energy phenomenology of think theory is the ground state of the theory. Now there arises the question whether how many ground states this theory might have. So So since we speak about a quantum theory, we have no way, if there are many equivalent ground states, we have no way to calculate which ground state we feel we can be in.

47:30 And therefore, if the spectrum of ground states is very large, this might actually destroy the low-energy coefficient of spin theory altogether. So there is a whole new spectrum of possibilities and the understanding of string theory today is by only sufficient to judge how the structure of string theory will turn out to be. So there are those people who expect that there should be some making selection mechanism that strongly restricts the spectrum of allowed ground states, which would recover the predictive force, if you will. And there are arguments that, based on statistical considerations, it would be possible to retain some level of predictive force, even if there is a huge number of ground states, and there are other positions which suggest that it might happen that actually the string theory is not a low energy predictive theory at all. So what I will need for the following argument is the assumption that a string theory is in a certain sense highly predictive, which does not mean that it just has one or a few states but that it has that the number of ground states is sufficiently limited to to be to make the theory more predictive than a theory that that has continuous free parameters because in those those Because, um, waking states are all considered to be discrete states, certain points of gravity so to say. So high energy, and that whole theory high energy predictive, what I mean is that some of these points, which constitute the allowed way pure, still does not come up to the freedom you have if you, if you, have concluded brief parameter.

50:00 This, of course, is the natural goal of the theory, though at the moment, we are not in a position to say that actually string theory will at any time be able to achieve that goal. Now, the argument goes as follows. If I assume that string theory is a highly predictive theory. Then I can compare the situation of string theory with the situation of any other physical theory that is not structurally unique, highly predictive structurally unique. In a conventional theory, My expectation would be that whatever empirical evidence comes up, the scientist is able to have a theory that will account for this empirical evidence. So the scientist does not ask, is there a scientific theory to deal with this regularity, but the scientist asks, which theory is there? The natural extension of this assumption is that if there is one theory that can do the job, there will be several, there may be several of them, so we are back with a scientific under-determination. Now in the case of highly predictive structural theory, we have no reason to assume that there are too many of them. We only know it at all, we know one such a unique theory, which is strength theory. Of course we cannot exclude that there are others, and of course we cannot exclude that there are many of them but we have no reason to expect that we will have so many of these theories that they actually cover the parameter space if they are not sufficient to cover the parameter space we also have no reason to assume that if I find one structure unique theory then search an empirical limit that is able to describe a certain set of the theory of nature within certain limits.

52:30 If that is the case, I have no reason to assume that there will be alternative to such a unique theory within this unit. And therefore, I have no reason to assume that I can, at some later period, replace this structurally in theory by a different one. So, this is where I wrote down the argument here, and the implication is that the principle of underdetermination, of scientific underdetermination, is not valid if you just apply it to the That's the first part of the argument. The second part concerns the question whether I should expect that I can replace, or will replace at the late time, highly predictive structure of the unique theory. By analogy, it does not have this problem. And the problem here, if I assume that this would happen, is that the later theory will not be able to explain, the later theory which of course is structurally significantly less predictive than the earlier theory, it will not explain why the earlier theory has worked well within certain units. theory, seen from the later theory, will become a miracle, which means that it may be not satisfactory to expect that this process could happen. The consequence in this case is that under-determination, even, scientific under-determination, even if it applies in this context, may be not that relevant. So to put those two lines of argument together, it means that I find this evaluation of the

55:00 scientific underdetermination principle, and I end up with something I might call a final theory claim. Maybe I shouldn't expect my theory to be replaced by a final predictive structure unique theory. And I also shouldn't expect it to be replaced by a non-structure unique theory. So therefore, I might expect that I have a final theory in my hand. So this is, would be a, would contradict the pessimistic meta-induction. However, it would not contradict it on the basis that I can in any way exclude the possibility that there will be a different theory that it places my present one. But it works at the same level as pessimistic mutant induction works itself, mainly at the level of expectation, what is leading to exactly what we try to do. So, I should say once again, this cannot be deduced from the present state of spring theory, because we don't know whether spring theory is highly predictive. However, it is interesting that the theory contains this perspective so far. So it might show a perspective on evaluations or assessments of scientific progress that differ from the way we understand scientific progress. So I just, I don't say that much about the other final, the other versions of Final theory claims, they want to say that the way to derive a final theory claim here is not the only one that occurs in the common system theory. And actually, they occur different kinds of final theory claims which do not rely on this

57:30 assumption of high predictive force. and the most significant is the second one is the question of theoryality theoryality is a reality relation that relates theories you have so to call compactified dimensions which means these are dimensions which go back into spatial directions, which go back into themselves after a very small translation, like a similar surface. and this is important to construct a coherent strength degree and it turns out there that there consists of relative relation which means a relation that that identifies a certain type of theory that has a certain radius of compacted vitamin with another theory that has an inverse radius and this eventually can be expanded into the notion that that there exists an overall a limit to, a lower limit to distance scales. So whatever we can understand in terms of very small distances, distances that are smaller than a strict scale, we can translate into to something that happens at larger distance scales. So in this sense, this constitutes a final theory claim in the sense that you have a termination of the possibility to look empirically at every smaller distance scales, test every smaller distance scales. So we have these two new approaches, the final theory approach, and the structure of uniqueness.

1:00:00 And one might conclude that this could imply a claim of an end of science, which is not really an expectation of what I'm going to do. So if one looks at the status of string theory today, and what we observe today, if you look theory is a situation where two things happen, where on one side, on one hand, the principle of scientific underdetermination is weakening, and on the other hand, the expectation that But if you start building a scientific theory, you will complete it within a certain amount of time. It does not apply anymore. So it would be a more interesting perspective, I think, a more interesting or more suitable interpretation of what's going on in string theory, If we say we are not approaching any kind of fundamental science, but we enter a different regime of scientific evolution, in my country, where scientific progress happens based on working on very complex consistency questions within a certain theory. And these considerations might lead to new concepts, to new ideas, which, however, are all motivated by making this certain system consistent. And if we look at string theory that way, we might say it might be a natural conclusion that in such a context, theoretical confirmation, theoretical arguments for theory evolution become more important and may be nearly as important

1:02:30 as arguments of empirical argument. Now, to conclude, it is important to say what is not implied by this statement. So this is not meant to be a philosophical proof of string theory in any way, of course. And it is not meant to be a strict reputation, or it cannot be a strict reputation of the traditional understanding of theory confirmation. However, what it might be is circumstantial evidence that a certain shift of some philosophical conception is going on in the context of particle physics. And that it might be reasonable to account for these shifts. One important point here, I think, is that one should bear in mind that the scientific, the concept of science, of what science means and how scientific process works, itself is based on possibility arguments, the considerations which try to make sense of what we actually observe if we look at the scientific process. So if one would argue for a shift of some scientific concepts of science, it might be justified to argue for them, also at the level of circumstantial evidence, without having any smoking gum, so to say, that's actually a good point. Now, one thing is important to note is that, of course, all those things are very specific characteristics of one specific, very fundamental theory, and they cannot be translated directly into general considerations about science. So it rather shows that there is a growing gap between the metaphysics and maybe the scientific principles

1:05:00 which guide from the metaphysics, and the more applied fields of science where the guiding principles may be different. and one last point here would be that even to to think about little things does not imply that one has to take string theory seriously instead one has to assume that string theory is a valid or correct theory one might also look at the whole story in terms of the kind of test if you look at string theory as one One idea that shows how science might evolve and to demonstrate that it might be possible that concepts of scientific progress change just based on the characteristics of certain scientific theories. So even if string theory might not be one that eventually turns out to solve the riddle, one might still think about the question whether scientific theory being in the context of those very fundamental, very unifying, in a sense, final theoretical considerations might be like a spread. So let's thank you. What sensitivity is the general relativity? Sorry? General relativity? Yeah, what sensitivity is the general relativity? Well, I mean, it depends on what level one is discussing. General relativity, let's... First one has to make one very general part.

1:07:30 There may be theories which don't have dimensional three parameters. Those do exist. However, in a sense, our work with those theories still implies, let's say, assumption of a certain pre-parameter because the theories are not entirely unified so we have to in a sense we have to to insert them into our worldview and therefore we have to to interpret, to test their meaningful parameters. So in the case of general relativity, for example, you have to measure the gravitational constant. Of course, in general relativity, there is even another argument that that we should reject the user. It makes a few reasons that much. OK, so please, please, if you have a suggestion. Well, then I'm very angry, because I find it surprisingly much as finance and engineering, which has been the prime steps in it as well. So that actually we don't do it. Actually, it does not. the string theory has one correct characteristic scale which is the string scale and the planck scale enemies from that however the difference is the difference between string theory and for example general activity without the cosmological concept would be that the characteristic scale of string theory actually we have to reduce everything derive everything from the scale those can be free whenever we measure any relation within our world which is the way we measure we have to measure them based on this assumption so so the the definition of the string yet is just a complete one

1:10:00 Yeah, it is not because our measurements from a summer based on to say, a hypothetical terminology because of the matter would have, in which you don't want to say it's true, because then you've got several subventionists with them in the final iterations of the publication of the same GR. So it would be a kind of subventionist report in a subventionist. I wonder if it would be unfair to describe the talk as suggesting the following. Here we have string theory, a beautiful mathematical theoretical structure. Unfortunately, it's untested, hitherto. Therefore, we ought to amend the hitherto demarcation between science and non-science in terms of testability. I would say that this would characterize my talk up to the point where I, so to say, make this sociological statement that spring theorists believe in their theory in spite of the fact that the theory is not very contestable. However, what I tried to do afterwards is I tried to find aspects of the theory itself which suggests this shift. So I tried to say it's not just the case that string theory is not a theory could be testable, and it's not just the case that people who might want interpretation of scientists things they want to believe in their theory, and if they don't have any sufficient reason to do so, they just shift their conceptual science. However, what I try to argue for is that this is not the whole story,

1:12:30 that one actually defines within the theory arguments which might be able to justify their children. Well, I found all those arguments rather weak, I must say. I mean, I can go through them, but I don't see any, I mean, they're all sort of, for example, internal coherence. I mean, that suggests that you go back to the idea that logic alone can lead to reality. I think it's contrary to what most businesses can accept. For sure, it is not what the classical understanding of science would tell. Of course, I don't want to claim, as I say in the last transparency, I don't want to claim that I want to abolish the difference between theoretical confirmation and empirical confirmation all I want to say is that what we observe and what we may be able to argue for within the theoretical context is that the weight of theoretical confirmation becomes stronger in the overall future. Can I have one more point? I also care to see how string theory could have any implication for the concept of under-determination, given that it is untested. Under-determination suggests that here's the set of data, and there's two competing theories so long and we can't empirically distinguish between this. Now how, if string theory doesn't actually relate to data, can it weaken or strengthen on that discrimination? I think it wouldn't be fair to say that string theory doesn't relate to data, because the problems that arise and the problem string theory tries to solve are problems which emerge from theories we have big based on empirical tests so we

1:15:00 have we have empirical tests of elementary particle physics which have led us to our elementary particle theories and we have tests of gravity which will lead us to a series of gravity. So in this sense, screen theory definitely is based on empirical data. The question just is, how far we are, so to say, allowed to move on on the basis just based on our theory. Are you saying that in the limit, or in some limit, you can regain, say, the standard model from screen theory? That's what the claim of string theory would be. We cannot calculate it now. So this is not a matter of calculation, but theoretically, can you obtain that? In the living in some limits. You can do that. The theory tells us that if we were able to do these calculations, we would find a certain low energy limit, and then it could be tested against even empirical data. But it isn't true, surely, that the theory also tells us that from the fundamental spring length, we will get the standard model parameters correct. We don't know enough about it. We don't know. We don't know whether this takes... So, I mean, the relation to data is conditional on some things going well, especially when you put in, you know, structurally unique, only one continuous parameter. I mean, that's, you're hoping there, I can't, or there, I can't, sorry, David, and I'm not there. So just the comment is in the way I read your talk, and I think what I'm doing, is in the research programme, it owes us the fault, firstly it owes us the reduction of the number of actions that we've done to test the data, and secondly it does allow the test the benefit to best be that we get the sound of a lot back out.

1:17:30 And your discussion of how do you structure these theories in case is all conditional on the work. If it doesn't work, it's strictly the wrong answer. Okay, so, sorry, I'm just happy to try that. So, let's go to the other side. It doesn't work. It doesn't work. It doesn't work. I'm still not clear why that reminds the pessimism of the problem. I don't see why we turn to that sort of situation, or even the full situation, So this is about the slide that has to be applied to the final theory. I think this is part of the exact idea also, that there are assumptions on that slide which seem to be, well, it's the rabbit that's been pulled up at the hat, you know, by the bottom of the slide, by the conjurer. I mean, as Harvey said, we thought we had a theory that was impossible to test and hardly looked like science. And the rabbit is, don't worry, it is science, it's just a new kind. Sorry, there's a trillion points on this time, the premise of the assumption that we I've got a problem later. I have a comment about why we shouldn't still find the string theory was actually the consummated to another subject. I can see how you are going through to the weakest link, or in the weakest link, which is that having once got to the structurally unique theory, we should expect that our future theories are also structurally unique. The argument is that it's a matter of expectation, so this argument is not able to exclude that we will find another theory that is also highly predicted. It's weaker than that. But all it says is that if we compare the situation in, let's say, traditional science and the situation in this scenario, the difference would be that in traditional science we have very good reasons to expect that an alternative theory exists within the limits that can also

1:20:00 described our present data while in this context we have no good reason to expect that to be the case so the point in the pessimistic meta-induction is not to prove that actually there is a new theory around the next corner that we replace our theory today based on very different conceptual assumptions but the I would argue that actually if we look at the evolution of science and the structure of our theories, we have reason to assume, to expect that this will happen here. And what I want to say here is that in this context, this expectation is not any more justified. That's all that is. I'm not clear why that's any more racist than the general situation of theory. We have a very general way of course counter argument to the best of the spectrum action, and I'm not clear about that strongly. Secondly, I'm not sure why this sort of underdetermination is to do with parameter freedom leads into the possibility of theory change in the sense that, surely if we've got parameter freedom we can speak in... We can move to a structure-based theory, which has a sense of what comes to us. Yeah, sure, if we change theories, we find a different theory, different parameters which also fits the situation. The point here is that in this context, the, let's say, the specter of possible structure unique theory, you might expect, is far smaller than the context, in a general context. So if we, of course, without the second part, this is probably a very strong argument. All you're saying is that we find, let's say, a subclass of all possible scientific theories, and for this, if we just go with the subclass, the pessimistic, let's say, the scientific underdetermination principle concerning the subclass, is far weaker.

1:22:30 The belief in scientific underdetermination is not justified within this subclass. Then I need the second step to argue that the return from this subclass back towards the general class of possible scientific theories is not, let's say, would be difficult to understand either. And these two arguments together lead to the conclusion that scientific underdetermination definitely is in a weaker situation here than this, in a general scientific context. Well, we could come back perhaps. Robert. Okay. Typically when we talk about underdetermination theories the usual context is given a set of theories that are empirically equivalent but make different theoretical and now you're talking about a situation as I think it's troubled several people in the room where we have a particular theory we can't talk about its empirical commitments yet because we haven't been able to, say, complete this project of finding what that low energy limit actually looks like at the end of the day once we crunch through whatever sort of mathematics is required to get there. So it seems to me, now you can tell me if I'm mischaracterizing your talk, but I'm sort of understanding what you're doing in the following sort of way. It's like you're wanting to, at this point in time, given the situation we're at the moment of the evidence available to us as you put it on a very early slide as far as you're wanting to draw a distinction between in some sense experimental confirmation and theoretical confirmation theoretical confirmation being looking at the kinds of theoretical reasons we don't have access to any sort of empirical data that you know in terms of a kind of experiment in a laboratory that's in the right energy regime to connect with what what we've got from string theory. But we have certain sorts of theoretical considerations

1:25:00 that we seem to hold valuable. So how can we deploy all of those to see what it says about whether or not there's only, our expectation, I just put it a couple of minutes ago, should be that there is sort of only one kind of string theory, or are there lots of theories out there in this kind of theory space, and parameter space? And I think one of the reasons why Harvey thinks you gave are kind of weak is because if you're playing that kind of game without you know at the end of the day touching base with some sort of experimental evidence in the laboratory there's always a kind of worry right that the considerations the theoretical considerations you can deduce aren't going to be enough to really limit the field well enough so given this kind of That's sort of the way that I understand your talk except that you want to take the line that at least to explain perhaps or diagnose perhaps that why it is that many, many physicists who are string theorists or high energy particle physicists who maybe look fondly upon what string theorists are doing, here's a way of trying to understand why it is that people are at least want to put quite a bit of belief in string theory at the moment, right? through your the arguments you gave us so we should just rehearse there for David in terms of the of trying to blunt perhaps the pessimistic meta-introduction so that's my kind of sketch of what it seems to come away with there's a kind of diagnosis as to why it is the physicists lots of physicists might be very very convinced at the moment upstream or i think it's it's really worth riding that horse at the moment then if that sort of understanding is correct I think there's still a way to to sort of rescue all these string theorists at the end of the day that they still really are perhaps not closeted part games where they're just ultra rationalist but at the end of the day they really do have an empirical kind of worry because supposing that we But if we are able to do this and get into the right sort of energy regime, the limit of the stream theory system, we now can at least run some tests on our mark, considering

1:27:30 when we get to the case where we have a kind of reproduction of something like the standard model, where we go out and look and we see that, for example, it gets that on the wrong. It's quite, I mean, it's quite conceivable to me that there might be multiple paths for trying to make this limit work, right? And so perhaps one of the problems, this is something that Klein or perhaps to him might say, well, it's just a matter of, you know, pick the way you win.