 Bonjour, j'espère que tout le monde a bien déjeuner. Tout est à leur place. On va commencer la séance de l'après-midi avec professeur Arthur Fein, de l'Université de Washington. Arthur Fein, de l'Université de Washington, va parler du réalisme structurel maintenant. Arthur? Je suis froid. Bonjour, monsieur Adam. Je suis heureux d'être ici. Je veux remercier l'organisation du comité pour cette opportunité et pour la conférence qu'ils ont installée. Je voudrais aussi remercier les staff qui ont fait des gens comme moi qui ne sont pas habitués à cet environnement, qui sont très confortables et qui ont été utiles de toute façon. Nous savons, bien sûr, que Poincaré est un grand mathématicien, qui a fait beaucoup de contributions seminales pour les mathématiques, c'était l'originateur des programmes que nous avons écoutés dans les derniers jours. Et un grand mathématique, ou un théorique, plus généralement. Mais il y avait un troisième arrêt pour Poincaré, et c'est son travail comme philosophie. Et c'est l'aspect que je veux parler de aujourd'hui. Poincaré, en philosophie, s'agit de son travail scientifique. Il est un philosophiste, dans un sens très particulier, pas un thème systématique, pas quelqu'un comme Kant. Ou Aristotle, mais quelqu'un qui a une pensée philosophique et qui a des réflexions sur son travail scientifique. Et en ce sens, il est un philosophiste très contemporain, car beaucoup de philosophies contemporaines viennent de la réflexion des pratiques de l'un ou de l'autre. Et il est contemporain dans un autre sens, et c'est que beaucoup de personnes qui travaillent en philosophie aujourd'hui sont aussi travaillées sur Poincaré. J'ai, pour l'exemple, supervisé deux dissertations, une sur Poincaré's philosophy of mathematics, et une sur Poincaré's work on geometry versus empirical science. Et il y a d'autres que je sais dans les works, maintenant. Quand les philosophistes parlent de Poincaré, ils généralement parlent d'eux, de sa philosophie de mathématiques, qui est, bien sûr, très appropriée, et aussi très intéressant de concentrer, peut-être, sur son intuition, ou sur son important, mais très difficile de comprendre la notion de prédicité en mathématiques. Ou d'autres, si ils sont intéressés par Poincaré as a philosopher of empirical science, ils sont très intéressés dans sa notion de conventionalisme et en essayant de faire des distinctions entre son geometrie de conventionalisme et son conventionalisme, en respectant la science, ou en essayant de sortir le tango qu'il lui-même a créé en distinguant la notion de hypotheses dans diverses et overlapisées. Et je pense que, dans l'hier, j'ai parlé de Jeremy Gray. Les deux sont tous des topics touchés sur, et si j'ai compris, demain, j'ai parlé de Gerard Heitzman. On va probablement concentrer plus sur l'aspect de la hypothesis. Aujourd'hui, je vais faire quelque chose complètement différent. Je peux commencer par vous dire ce que je ne vais pas faire, ce n'est rien de plus. Ce que je vais essayer c'est de donner un parler qui est très similaire à un nombre de paroles techniques que nous avons déjà écoutées dans cette conférence, c'est de commencer avec des réflexions de Poincaré, des suggestions intéressantes qui s'adressent à un problème particulier, et puis de montrer ce qui s'est passé en cours de temps, à l'heure très présente avec ces suggestions. Donc, le problème que je vais être concerné est du réalisme structurel et je vais commencer avec un brief summary de comment le parler va. Commençons avec cette belle quotation. Poincaré a peur que la nature fémérale de la théorie scientifique pourrait être vue comme une montagne à ce qu'il s'appelle la bankruptie de la science. Il y a une histoire intéressante pour être dit à cette phrase que j'ai originally pensé que c'était un de Poincaré's many beautiful expressions but turns out to be an expression common in the latter 19th century and reflecting actually a quite anti-science, religiously motivated to tack on science. That's not Poincaré's concern as I will explain. So, what I want to do is then to trace out a little bit of the history and kind of response that Poincaré begins beginning with Poincaré and then contrasting him with Duhem, with Dewey, with Russell and with Vile. And I will do that rather briefly. And then I will try to give some interpretation of certain confounding of conceptions that I think is involved both in Poincaré and in part in some of the transition. Then I want to look very briefly, sketch out briefly a modern view of structural realism and then I want to look much more seriously and critically at the contemporary structural realism debate and I will end with some pictures which suggest that there's some serious criticism to be made of this whole program. So, let's begin with Poincaré and this wonderful quote that comes out of Foundations of Science in 1905. At first blush it seems to us that theories last only a day that ruins upon ruins accumulate today theories are born tomorrow they are fashioned the day after tomorrow they are classic. The fourth day they are super annuated and the fifth day they are forgotten. Now Ok, so in the contemporary literature this idea that the sciences come and the sciences go is sometimes referred to as the disastrous sometimes as the pessimistic meta induction meta because it's an induction over the history of science Ah, I've got it twisted Yeah, ok Thank you Anyway, so so that these theories come and go and that it's a meta induction because it's over the history of science and it's a disastrous pessimistic meta induction because if you carried it out as a kind of straight line induction then of course you would conclude that today's theories however the evidence however well confirmed they are also going to come and go and with them the various objects that we think of as the objects of scientific interest and the properties that we think of them as having So Poincaré's solution to this problem of the passing phase of science is to focus on relations and so what he tells us is that if if theory taught us and here I want to emphasize a true relation then this relation is definitely acquired and it will be found again under a new guise in other theories which will successively come to reign in the place of the old so the interesting suggestion is that if we want to deal with the problem of theories that come and go with the pessimistic induction then what we want to do is we want to focus on relations ok because relations are going to be part of the permanent furniture provided he says that they are true relations now the thing to pay attention to here is the question of relations among or between what are we talking about relations in the data are we talking about relations in the data models that is the data as process the phenomena are we talking about relations about whatever it is in reality that underlines the phenomena ok these remarks here are certainly ambiguous about that I want to compare Poincaré's remarks here with those of Pierre Duhem on a related topic now like Poincaré Duhem is often considered used Jeremy Gray's word from yesterday a kind of skeptic ought to have a skeptical attitude very often classified as an instrumentalist and if you read remarks like this then of course it looks that way I mean the scientific hypotheses are only devised in order to save the phenomena that don't reflect what's really going on in the world on the other hand that's from his book 1908 he also tells us the physical theory confers on us knowledge of an external world which is irreducible to merely empirical knowledge and then he goes on to say in a very famous passage there's not merely an artificial system suitable today and useless tomorrow making reference to the pessimistic induction here but an increasingly natural classification which is to say a system of relations so it seems that Duhem and Poincaré here are trying to focus on the same problem the same solution to the same problem namely look at relations I say this is very famous passage of Duhem it's really an infamous passage because everyone who interprets Duhem interprets this passage differently there's a very nice article if you can read it at the end there by Karen Merica King as Darling in philosophy of science who draws out these contrasting instrumentalist realist aspects of Duhem now what I actually want to do here is to digress very briefly I can't help it coming from the continent that I come from and mention a little bit about John Dewey who also struggled with this problem of the so-called pessimistic induction and here's what Dewey says we can get it he talks about in a longer passage how scientists are wrestling with a problem and he describes what goes on there is evolution of new techniques of control and inquiry there is search for new facts there is an institution of new types of experimentation there is gain in the methodical control of experience and all of this is progress so in the course of scientific investigation all of these things are going on and then Dewey goes on to comment it is only the worn out cynic the divided sensualist the fanatical dogmatist who interprets the continuous change of science is proving that since each successive statement is wrong the whole record is simply error and folly and that the present truth is only the error not yet found out now I want to contrast Poincaré and Dewey's approach to this issue with Dewey's just really briefly what Poincaré and Dewey both suggest in slightly different ways is a solution to a problem problem is what do we do about the fact that theories come and go well we focus on relations which are more stable what Dewey does with the problem is to make a characteristic pragmatic move what Dewey says is there isn't really a problem if you pay attention to what scientists actually do you'll see that in the course of every investigation some progress is made not progress with respect to relations not progress with respect to this and that but progress all over the place it's a very different way of thinking about it but to return to Poincaré this is what he asked to tell us about Maxwell's equations they express relations and the equations remain true it's because the relations preserve their reality reality where? reality about real things or reality about the phenomena or the data they teach us now as they did then but there is such and such a relation between this thing and that only the something which we then called motion we now call electric current but these are merely names for the images we substitute it for the real objects which nature will forever hide from our eyes so the problem is as he poses it here we can't know what the real things are between which relations hold ok? and that's why we can't focus on the things but we can focus on the relations that hold between the things so again the true relations between these real objects are the only reality we can attain and the sole condition is that the same relation celles exist between these objects as between the images we are forced to put in their place so here is a picture of the phenomenal world that's the world of images and the language that we use and the real world and looking at relations in the phenomenal world we project as it were isomorphically to what must be the case in the real world although we can't really know anything about it ok? not anything directly so Poincaré is pretty serious about this so this is what he has to say about the principle of conservation of energy if we wish to enunciate the principle and all its generality and apply it to the universe we see advantage so to speak and nothing is left but this this is the principle of conservation of energy something is constant ok? that's what we can actually learn that something is constant this may be a reduxio of the view if not it seems to me it's very close to a reduxio of the view but let's persist for a bit now that's when Poincaré is talking about this coming and going of scientific theories but he carries on another dialogue which is not about that but which is about the topic of objectivity what he says about objectivity is very important he says that what is objective is what must be common to many minds ok? so now we're not talking about real or what's real or what's true but what's common to many minds and consequently transmissible from one to the other quality is not so transmissible but relations are transmissible from this point of view what is objective are only pure relations so the focus on relations here we get a very different angle we're not trying to find out what's real we're trying to find out what's objective in the sense of can be transmitted from one person to another or intersubjectively transmissible so I'm gonna flip to Bertrand Russell a little bit later who will see expresses very much the same ideas but most that can be known and that only in the most hopeful view is that there are certain relations in the physical world etc notice again the epistemological twists the most that can be known right? this is very much like what Poincaré had to say ok I have a lot of quotations here I don't think I will read them through take for example this is one of my favorite ones take for example what's common between a gramophone record I think some of you in the audience don't know perhaps what a gramophone record is think of a disc and the music it plays the two share certain structural properties which can be expressed in abstract terms etc in virtual of the structural similarity one can cause the other I think this is one of the most puzzling remarks in the whole Russellian corpus we have two structural relations not two events not two things that happen but two structural relations and one structure can cause the other structure I've never been able to understand what Russell means but this quote from Russell is actually what brought me many many years ago now to do philosophy ordinary language is totally unsuitable for expressing what physics really asserts since the words of everyday life are not sufficiently abstract only mathematics and mathematical logic can say as little as the physicist means so I think that's a wonderful wonderful attitude it's an attitude that's basically shared by Hermann Weill again science can only determine its domain of investigations up to an isomorphic mapping the particulars remains quite indifference to the essence of the objects the idea of isomorphism demarcates the boundary of cognition here the word is your kentness of what we can know but even if we do not know that is canon or not acquainted with the things in themselves that is with real reality we have just as much cognition knowledge about them as we do about the phenomena through the disclosure of these isomorphic relations so here too is this idea that we can establish in the data sets of relations which could be very well confirmed empirically and then we simply project those relations to that which we can't know anything else about right to the real stuff so it's the same idea we see repeated once in Poincaré, another time in Russell another time in Weill but I remind you not an idea that Dewey picks up in his response to the same problem well here I want to suggest that there are two problems of knowledge I've been suggesting it all along one problem of knowledge is to find out what's real or if you like what's true if you take truth as correspondence to reality then these are basically the same question and the picture here is to go back to Kant and Kant who says that we have to assume the existence of the numiner, the numiner of the things in themselves the objects concerning which we cannot know by perception or ordinary means anything but we have to assume that there are such objects or else the world is not intelligible for us ok and in the 19th century a great many scientists reflecting on what they were doing in doing the scientific work thought that they were overcoming the Kantian problem of the numiner insofar as they could have well established scientific knowledge then they could project that knowledge not just to the empirical or the phenomenal realm but they could project it from the phenomenal realm down to the realm of the numiner and that's how you find out how things really and truly are that's the picture here but there's another problem of knowledge and it's a problem about what's objective when do we consider something to be a part and parcel of objective knowledge and one conception of the scientific enterprise is that we're after reality or truth or something along those lines but a quite different conception is that what we're after is objective knowledge and these I think are two very different issues so I think what we find in Poincaré and Vile and also very clearly in Arthur Eddington and I didn't have time to find quotes from is this equation that what counts as objective or objective knowledge is something that's intersubjectively communicable ok we saw that very clearly in one of the quotes that I had earlier from Poincaré what's intersubjectively communicable is what's invariant sorry under changes in points of view what's non-perspectival what's non-perspectival are structures, relational structures that are preserved by a set of quote-un-quote admissible transformations this is an idea pursued by Robert Nozick in his last book which was called invariance and it's the idea that I've tried to highlight a little bit here, I don't know how well we can see it with Arthur Eddington who described this as the point of view of no one in particular phrase that I think is very appropriate but this whole idea of objectivity as intersubjective communicability which leads to this conception of what's objective as what's invariant has to do with admissible or preferred transformations and that all has to do with us and so this pulls away from the idea of getting at the numerator getting at the things in itself the things that are unknowable to something that's socially and humanly centered and so it pulls away from the idea of objective as what's really real to the idea of something that's suitable for what I call here a constructivist or a conventionalist picture and of course that's a picture that's very often associated with Poincaré so that's the background and that's the suggestion and now I want to look very quickly at the modern revival beginning with Ernest McMullen in a very widely read piece says that the basic claim of scientific realism is that the long-term success of science gives us reasons to believe in something like the entities and structures postulated by the theory so again it's this picture from empirical structures to structures in the real world but here I want to want you to note that when McMullen first announces this idea he talks about entities and structure scientists construct theories which explain the observed features of the physical world by postulating models of the hidden structures of the entities being studied so that's the picture we're trying to model in our science the hidden structures the numinal structures later John Warrell revives a kind of position that we call structural realism the rule seems to be that whenever a theory replaces a predecessor which has enjoyed genuine predictive success the mathematical equations of the old reemerge as limiting cases of the mathematical equations of the new this is precisely a version of Poincaré's idea that structure is preserved so the structural realist then simply asserts that in view of the theory's empirical success the structure of the universe is probably something like what the theory says repeating the words of McMullen pretty pretty closely here notice there's an inference involved here let's talk about it in a moment it's the inference from the success of science to the reality of the scientific claims it's a very particular kind of inference and a very odd one finally the most latest version of structural realism is so-called ontic structural realism a book that came out in 2007 called everything must go by James Ladiman Don Ross is the bible for this particular kind of structural realism I'll give you just a quick flavor of it in the title everything must go the emphasis really needs to be on things it's everything that must go because the basic axiom here is that there are no things structure is all there is so it's like turtles all the way down but in this case it's structure all the way down there are no things what do we do with objects well objects are sort of pragmatic devices that agents use agents means you and me to orient ourselves in the world ok so they're sort of provisional tools Plunkeray sometimes uses the phrase useful fictions so one might think of them that way what's very striking to me about this is that objects I mean we have to provide some accounts of objects so the promatic devices used by agents used by agents that's you and me who are also objects who are presumably also merely pragmatic devices used by agents but what agents are using us and for what purposes I think we see here a real problem for ontic structure realism namely can you have just relations all the way down there's a long history here that includes people like Carnap and Carnap so called out problem and we could talk about it in the question period but I want to move on ok so I want to give you a really quick really quick run on how the realism debate goes for realism that is the view that science tells us what's really going on out there one standard argument is the so called no miracle success argument and that's the argument we've been looking at to a certain extent which simply projects from the success of the scientific enterprise look would be a miracle if science were as successful as it is unless it had latched on to the truth of things ok so that's the basic form of the argument and it has sophisticated versions science gives us the best explanation and so on I'm going to skip all that the other interesting argument for a realist perspective is an argument that curiously philosophers at least in the 20th early 21st century have stayed away from by and large it's a kind of Kantian argument namely that how on earth could you understand what's going on in science unless you thought that science was exploring the world around us and giving us reliable and good information about it right so it's a necessary condition for understanding scientific practice an argument philosophers have not paid nearly enough attention to on the anti-realist side we have the pessimistic meta induction theories come and go and with it the ontologies of those theories and we have a very well known standard argument well known to every scientist certainly the argument from under determination which is simply to say that I don't care how much data you collect I don't care what the quality is of it and so forth it always under determines the hypothesis there are always many ways of accounting there are many ways for accounting for the data why I believe this one as opposed to that one that again pulls on the anti-realist side now I'm not saying either for the realist or the anti-realist that any of these arguments are compelling or convincing all of these arguments needed to need to be looked at carefully and hedged about carefully but these are the central kinds of arguments that one finds there is a third kind of anti-realist argument associated with constructivism sometimes social constructivism would be a better phrase the constructivist project is very straightforward the constructivist project is to explain what scientists do that is to explain scientific practice without presupposing the truth of the theories that scientists employ now if the constructivist project could be successful then it would undercut the canteen argument that we have to assume the reality of the world in order to understand science but constructivist project is precisely to understand science in all of its detail without presupposing the truth of the scientific theories that we're interested in so that's the playing field for the debate the problem is to as it's come to be formulated is to latch on to the no miracle success line which pulls you towards a realist idea science is successful would be a miracle if it didn't really get that good part of the truth and in a way that avoids the pessimistic meta induction and there is a really generic form of a solution here namely find an X such that in success of scientific theories X is preserved okay so it avoids the pessimistic meta induction and that X is a ground for our scientific success right so there's the problem and there's the generic form of solution okay so in Poincaré's hands and in the hands of Vile and in the hands of Russell the X that we're looking for to provide the solution here is structure so I want to devote the next five minutes to asking what's special about structure why have we focused on structure is there anything special about structure I want to offer you a set of incomplete arguments I was about to say poor arguments I hope they're not really very poor but they are certainly incomplete and what I'm not trying to do is to provide a knock down argument that one should not be interested in structure or that structure is irrelevant but I'm trying to disabuse you of the notion that structure is itself very interesting okay so that's the function of the next five minutes so suppose we're looking at theories which come and go but in which structure is preserved okay then what does it follow that structure holds really truly okay so I want to give you some examples so the question is does the preservation of structure lead us to reality so here's here's my first example it's very simple pretty amusing look at classical mechanics and look at a corollary to the second law okay F equals MA so for a fixed mass force is a function of acceleration and for a fixed acceleration force is a function of mass that's pretty straightforward now instead of looking at classical mechanics look at classical economics and the law of supply and demand okay well that tells us that for a fixed supply the price is going to be a function of demand and it also tells us that for a fixed demand the price is going to be a function of supply okay all that's pretty trivial thus classical mechanics and classical economics share a rather interesting structure namely there are three variables such that one of them is a function of each of the others provided the third variable is held fixed alright so we have as between classical mechanics and classical economics a common structure alright and if you thought that there was an inference from common structure structure preserved to structure is real then you really need to be asking yourself at this point but in what domain is the structure real what are we talking about that's reality here okay well obviously this is a kind of category mistake that I've concocted all on my own I can't blame anybody looking from economics on the one hand mechanics on the other hand so maybe the solution is to fix the domain alright and then look at what happens if structure is preserved and as it were roughly speaking the same domain that we're working in so we're not working here in economics and here in mechanics so do we now want to approve of the inference that the preservation of structure tells us that we've discovered something true about the domain well so look at Newton's verse law again for big bodies in slow motion well this is preserved at least approximately between Newtonian mechanics and spatial relativity so according to the inference since it's preserved between that great scientific revolution the inference is real in this domain but of course here's another structure that's preserved the global space-time structure in classical mechanics and the global space-time structure in special relativity is flat we work in flat Minkowski space and so there's a structure that's preserved between the same two theories but of course the global structure in general relativity is not flat we suggest that we are a little hasty in saying well this is preserved by evolution so therefore this is the mark of something real ok well in addition to fixing the domain since that doesn't work what we could do is to add that the structure should be preserved not just across a couple of theories classical mechanics to relativistic mechanics to general relativity should be preserved across a whole bunch of theories so maybe if something is just there across a whole bunch of theories then that will show that it's a mark of what's real ok here I want to give up the game of example and counter-example although I can't quite tease myself away from it so I suggest you make up your own examples probably everyone in this audience knows quite a few examples that would work here's one of my favorites ok look at the structure that binds together state variables that we refer to as determinism ok there's a long history of deterministic theories across many different domains in the sciences and there's a long history of the termination of those deterministic theories and the intrusion of indeterministic theories so this is a very nice example in which you have structure preserved a lot of things but it doesn't follow that the structure is as it were a mark of the real ok so those considerations ought to lead us to think that there isn't a good inference of the form structure preserved implies blah blah is real right there is no simple inference of that kind I don't think what I want to show you now is that actually this whole business of preserving structures is in a certain sense trivial that give me any bunch of scientific theories say roughly in the same domain there's always going to be structure preserved and if there's always going to be structure preserved then it can't be of any particular epistemological significance that this structure preserved because the reason there's always structure preserved is purely logical so I begin with the conjecture which is that where several theories succeed one another they'll be they'll let me read it succeed one another there'll be areas where the theories do not disagree up to a certain order of approximation so the theories might be radically different but there'll be areas, some areas of application which they basically agree ok that is in fact the way science works right we talk about scientific revolutions but they always change but not everything changes and between one theory and another there's always overlap between that theory and a third there'll be overlap with the first and so on so I take that as a reasonable hypothesis about the history of science so based on the conjecture we can formulate the following theorem given any succession of theories whether they're true theories or false theories I don't care there'll be a relational structure that they all share at least approximately applications so structure is trivial in the sense that it's always going to be shared what's the proof well the proof is really pretty straight forward the conjecture says that if we have a bunch of theories, successive theories there'll be some domains of application in which they're consistent with one another ok if there's a domain in which they're consistent then we can formulate a first order theory of that domain that first order theory will be consistent by assumption therefore by the completeness theorem it will have a model that model is of course nothing other than a relational structure that's what we mean by a model so the preservation of structure is pretty trivial ok so far we've got this far structure preservation across theories is a mark of the real I want to say that doesn't work but we haven't got that structural preservation that under right success is a mark of the real so I've been doing a little conjures trick here directing you to one aspect of what the solution was supposed to be with my right hand while deflecting what my left hand is doing so I've been looking at the significance of structure and whether structure or preservation structure alone is a mark of the real without saying well not just preserving structure we want to preserve structure that grounds the success of the theories that we're talking about so we have to come back to this so the question is structural preservation that under right success is that going to be a mark of the real again I would invite you to construct your own counterexamples again I'll prime the pump by giving you one that you've already seen classical mechanics to special relativity with respect to global space-time structure right that's a very nice and that space-time structure grounds the success of use flattening Kowski space grounds the success of special relativity but of course that isn't going to be a mark of the real it won't generalize up to general relativity ok so this is an area where if I had half an hour more I would try to construct some more examples but I think you could construct your own instead I want to end with a kind of deconstruction of the project rather than arguments about the details of the project so here's what I think is the usual heuristic that goes on in science we have an old theory and we have a new theory and the new theory has some kind of successful prediction and the new and the old theories as I said before generally overlap so from the new theory if we're lucky we get a novel successful prediction of some sort or some kind of success we can solve some problem we can solve before ok we take the success of the new theory as an indication that the new stuff that we've introduced is good stuff to introduce into the theory that's how science usually works we make a scientific shift that shift is successful and a predictive and empirical sense and we say well that redounds to the credit of the shift that we've made the heuristic in the program of scientific realism is exactly backwards we introduce a new theory the theory is successful but the success of the new theory doesn't accrue to some new features in here the success of the new theory accrues to the old stuff that the new theory shares with the old theory namely in Conqueray's idea or in Viles or in Russell's to the relations that are shared between the old theory and the new theory that tends to confirm that those relations really refer to reality ok what I want to suggest is that this is really a reactionary heuristic and I know I only have two minutes so I will do this quickly in two minutes I want to contrast classical quantum mechanics orthodox quantum mechanics with de Broglie Bohm program de Broglie in 1927 introduces un orthodoxe way of thinking about quantum mechanics Bohm in the 1950s revives that and I was going to say improves on it I'm not sure he improves on it he adds a theory of measurement to it and it's a lively program well I work in foundations and as some of you know if you've touched on foundations it's a shambles right I mean there are a dozen competing foundational interpretations each one trying to beat up on the other and one of them is the de Broglie Bohm program what's really nice and interesting about de Broglie Bohm is just this without going into the details in orthodoxe quantum mechanics we don't have a space time theory in general things don't happen in a particular place at a particular time processes don't have trajectories that's a radically different kind of theory in the Bohm theory everything that happens happens at a particular place at a particular time and particles or whatever are moving around in determinant spatial temporal trajectories so the Bohm theory shares with classical mechanics this enormously important feature right it's a space time theory it's a theory with processes that actually go on with trajectories and so on well if we adopted the heuristic that's built in to this structural realist program we would say that the Bohmian theory simply wins out because all of the new successes of quantum mechanics were redound to the Bohmian theory because it shares that structure with classical mechanics and my view is that's insane so I will end by saying that what's happened here is we're playing the same old game that we've done in the realism debate for about 25 years we have a circle of question begging assumptions what wants to infer from some feature of scientific practice namely the preservation of structure to the truth in the correspondence sense of some aspect of scientific theories here that structure is real but the inference has as a premise that utility or success is a mark of truth that's a premise that's disputed by various antirealisms by instrumentalist pictures in particular and so the inference itself is of course question begging so what to do in these circumstances shall we simply say to my realist friends that they're irrational so I simply say as I tried once and got beat over the head that you know your realism is just a leap of faith it's kind of like a religious faith or political ideology something like that that didn't work very well so what I like is to try and follow this line of thought that the only true criterion is to see what's possible and proper to do this is very pragmatic doing in there are no rigid criteria ever you have to consider the contingencies of the period in question and those of today and then you have to see what works that's all and although this may sound very doing and very pragmatic and perhaps very American this is actually Pierre Boulez in his thoughts on conducting so I leave you with Boulez's thoughts about how to deal with structural realism thank you very much we have time for a few questions I see one in the back Professor Heinzmann I agree with you if one accept your definition of structural realism what is and you defined it with the continuity problem but there are two sides naturally there are the side of continuity and progress and the other side it of the ontological problem and if you take the realist, anti-realist discussion and structuralism with respect to the ontological problem I think Poincaré is not structural realist why because as Hilbert he he he doesn't agree that's the elements that's the first elements of geometry are propositions Hilbert says they are propositions of schemata and Poincaré says they are apparent hypothesis not true, not false now Hilbert to justify this axon schemata he needs non-contradiction proof Poincaré says no, no, this is not the right way because of the unpredictability problem he says we should give a system that exemplifies the structure and then he in the foundations of geometry he gives his reconstruction psychological reconstruction of geometry and this gives this shows that he is not under-rem the structure is not under-rem reality but it is not an inner ray structure it is not true but it is between self-reflection about the structure can itself be exemplified itself so a little bit in this sense from this point of view I would say Poincaré is not defender of of structural realism but in your sense of continuity you are right so Professor Heitzman I think points us in the direction I mentioned very early on of there being a great tangle in the way Poincaré sorts out the different kinds of hypotheses or in this case propositions and I actually in very considerable agreement with your own way of trying to work through and sort it out but I think we of us who read Poincaré have got to admit that it's a tangled that his texts here are tangled and it takes careful interpretive work to try to disentangle him and here the under-determination problem rears its head in scholarship just as it does in science namely there's not just one way of trying to untangle them although your way I think is a very excellent one for the purpose of this talk I thought it would be a good idea to take a suggestion from Poincaré himself and to treat Poincaré's structural realism as a useful fiction my aim of course is not to construct an argument against Poincaré as a structural realist my aim is really to construct an argument against the structural realist program per se whether one attaches Poincaré in the ontological sense or not I think that's a long way of saying we're in a considerable agreement concernant la philosophie de Poincaré je pense que le mot philosophie il connait le langage conscient mais sans moque ce serait mieux de parler de la pensée scientifique concernant I need a translator concernant le réalisme scientifique le réalisme structural je pense que pour Poincaré réalisme est complètement inadéquat et structural est un peu un peu trop fort c'est très difficile de saisir Poincaré il est sur un équilibre instable que lui seul peut occuper on tombe à gauche en conventionnalisme ou à droite dans le réalisme je pense que le mieux pour lui ça serait de parler de stabilité structurelle mais stabilité relationnelle serait peut-être mieux mais c'est pas de la variante ni de la rigidité j'ai du mal parce que je ne trouve pas une question là-dedans c'est une remarque sur le c'est une remarque la remarque c'est si j'ai compris la partie centrale comme pour les Heidzmann ma construe de Poincaré structurelle est trop forte oui ? le réalisme est structurel un peu oui Poincaré la picture Poincaré est focussée sur la structure mais il n'est pas focussé sur la structure comme un réaliste et je c'est une question très sensible et nous devons regarder des textes différents des textes qui vont exactement dans votre façon et des textes qui vont différemment et c'est pourquoi j'ai essayé de différencier le problème du réalisme de l'objectivité il est un structuraliste réaliste c'est une question beaucoup plus debateable il parle de cette façon Poincaré a-t-il essayé de définir la relation ou avez-vous essayé de définir la structure ? je pense que c'est une bonne question et j'espère que j'ai une bonne réponse ce que je prends comme relations sont, premièrement les équations mathématiques c'est assez clair par exemple, on parle de Maxwell's Equations quand vous avez une bien définie équation mathématique cela établit une structure relationnelle si il y avait que je fais dans le milieu de mes paroles vous pouvez utiliser la théorie de completité d'une forme consistant et le modèle est une structure relationnelle si il reconnait en général je pense que c'est quelque chose que nous devons regarder très bien j'espère que la réponse est non c'est une bonne question si nous avons plus de questions merci prof.