 the Parisian connection with the projectile multi-prolibrational properties and Cote-au-mache-sieau. Okay, thank you. So my talk is going to be about multi-prolibrational properties, and the aim of my talk is to show that it's to argue that multi-prolibrational properties are projectible against an argument which has been proposed by Jai Womkin. So very briefly, my talk has three parts. So first of all, I'm going to give you a bit of background about multi-prolibrational properties and each of the scheme's arguments. Then I'm going to go to actually one reply which has been already proposed by Ned Block, and then I'm going to tell you why Ned replied is not fully satisfactory and then I'm going to provide my two scheme's arguments. So first of all, so the multiple resability is the idea that some properties are realized by multiple different physical substrates. So first of all, for a property, for a high-level property to be realized, that means for the high-level property to obtain, in virtue of the fact that the low-level property, the realizer, obtains, right? So the classic example is that pain, which is a high-level property, it is realized in, for example, in humans by C fibers and by A delta fibers. These are fibers of the brains of the human brain. And whenever these fibers fire, then the subject feels pain. So that's what it is for a property to be realized by another low-level property. And multiple resability says that some high-level property, for example, pain, are realized by different physical substrates. For example, in human, pain is realized by C fiber and A delta fibers. But in other species, for example, octopuses, who have different narrow substrates from ours, they have different realizers of pain, let's say, octopus fibers, and then other species, they have different physical realizers of pain, for example, Martian, so species that we have not discovered yet. And so the claim of multiple resability was put forward by Larry Patnam in the 70s. And the claim it was supposed to be an argument against reductionism. So reductionism is the idea that all properties or scientific properties are actually physical properties or are at least explainable in physical terms, right? So basically all properties are actually physical. But if some high-level properties are multiply realizable, then reductionism cannot be true because in one case we have a high-level property which is realized by a given physical substrate and that physical substrate is supposed to explain the occurrence of the high-level property. But we have different cases in which there are different realizers. And so none of the realizers that you see down in the schema, which is from the classic paper from Fodder, none of the realizers alone would be sufficient in order to explain the occurrence of the high-level property, right? That's what multiple resability says. And so the argument is supposed to be reductionism is false. But Jagon Kim says that this is actually not true because so multiple resable properties cannot refute reductionism and to refute reductionism actually it's important to... How do I go back to the... Oh, left should probably... left arrow. So if reductionism is false, so since multiple resability is supposed to refute reductionism, it's also supposed to justify the autonomy of special sciences, right? So special sciences are sciences other than physics. So for example, psychology, biology, sociology, economics and so on. And the fact that reductionism is not true means that these science are actually autonomous from physics, right? But Kim says that that's not the case because actually multiple resable properties are not scientific and that's the argument he provides. So the argument goes as follows. So first of all, first premise is that scientific predicates, scientific properties and kinds are projectable kinds. So we are going to see what that means in a second. Secondly, multiple resable properties, they are not projectable and therefore multiple resable properties and predicates they are not scientific, right? And just to point out the importance of this kind of debate, if that's true that multiple resable properties are not scientific and if that's also the case that special sciences properties are multiple realizable, so if the properties studied by special sciences are multiple realizable, then you follow that special sciences are not science after all because they don't study scientific properties, right? So we have to find an answer to Kim's argument. So let's see what the premise is actually. Why they are supposed to hold the momentum they mean. So first of all, scientific kinds, we can characterize scientific kinds in the debate, they are characterized fairly minimally as the properties that appear in scientific laws. So just to be clear, here I say properties but actually I should have said predicates because scientific laws, they are formulated in linguistic terms. Actually there are predicates in scientific laws but if they are scientific, they are not properties, right? So predicates, scientific predicates, they are supposed to be not properties and if they are scientific again, they will be not scientific kinds so sometimes I will use these terms interchangeably but obviously they are not the same, right? So predicates, they are not always the not properties and even if they do, maybe these properties are not scientific but then if there is some misunderstanding, we try to make it clear. Otherwise they will use these terms sometimes interchangeably. So this is the first sense we have and then projectability. So projectability is, so I should say that Goodman talked a lot about projectability but I am not going to talk about Goodman's sense of projectability. This is how Kim characterizes projectability and this is the sense of projectability that I am interested in in this talk. So projectability is the characteristic of scientific, of low light generalization, sorry, of being confirmed by positive instances of a given type of phenomenon, right? The phenomenon that this generalization talked about and so for example the low light generalization metals are good conductors of electricity is projectable because it is confirmed if we find an example of a metal, right, which conducts electricity and that contrasts with for example the following generalization, all coins in my pocket are copper because if you find a coin which is in my pocket which is made out of copper that doesn't tell me anything about the material of other coins which are in my pocket. So projectability concerns first of all generalizations but we can talk by extension let's say of projectable predicates whenever this predicates figure in projectable generalizations, right? So predicates are projectable insofar as they appear in projectable generalizations but so the important point to keep in mind here is that the unprojectability, the fact that some generalizations are not projectable might be due to the nature of the predicates which appear in this generalization but if so the relation between the projectability of predicates and that of generalizations is kind of tricky but this is what's important for me that some generalization might be non-projectable because of the nature of the predicates. So this is the argument in support of the second premise namely that multivariable properties they are not projectable and that's from Kim. So the argument goes as follows. So first of all, multi-pluralizable properties and predicates actually are nomologically co-extensive and therefore equivalent to the disjunction of the realizers of the multi-pluralizable properties. So this is what Kim says. Now to be co-extensive, so first of all the extension of a predicate is the set of things to which the predicate applies to, right? So the extension of the predicate red is the set of red things and to be co-extensive means to have the same extension. Now what Kim says about multi-pluralizable properties, this is the scheme I just showed you, is that they are not co-extensive at a high level properties, multi-pluralizable properties, they are not co-extensive with any of the realizers alone, right? But they are actually co-extensive with the whole disjunction of the realizers because whenever the high level property occurs it must be the case that at least one of the disjuncts which is the realizer occurs too, right? Even though not all of them occur, but one of them occurs too but that's a disjunction. So the disjunction and the high level property they apply to the same set of things, right? Of the scheme to some extent. And they are co-extensive. The true predicate, the disjunction, low disjunction, the high disjunction, they are co-extensive as a matter of nomological necessity, let's say. So that depends on how the laws of nature of our universe are made. And because here we are talking about generalizations which are supposed to be laws and the realization-relation scheme says should hold as a matter of nomological necessity. So given this strong correlation between the two properties we can consider that the two properties, the high level properties, multiply the realizable and the low level disjunction of the realizers they are actually equivalent. As far as the extension goes. And this is the first premise. So the second premise is that actually disjunctive predicates heterogeneously disjunctive predicates they are unfit for projections. They are not projectible because they would functionally illegitimate, not legitimate confirmation procedure and that is because the reason for that is because disjunctions, they are satisfied even though only one of the disjunctions is satisfied. So if we try to project, if we try to confirm a disjunction the disjunction is confirmed even though only one of the disjunctions is confirmed, right? But that's illegitimate. We cannot confirm something without confirming all the things that it mentions, right? Intuitively. And therefore disjunctive predicates are not projectible because the disjunctions would be confirmed without being tested. And given the equivalence between multiple realizable properties and the disjunction of the realizers it follows that actually multiple realizable properties are not projectible, right? And so here is an example from Kim's paper. So Kim says, consider disjunalization. Namely, sharp pains administered at random intervals cause anxiety reactions, right? This is a putative law of psychology. And he says, suppose that disjunalization has been confirmed for humans should we expect on that basis that the same generalization will hold for species that we have not yet discovered or test the law upon? For example, Martians, right? And he says, no, we shouldn't expect that. We shouldn't expect that if we accept the physical realization of this, right? Which is the this that a high-level property are realized by physical entities. So this argument goes in support of the second premise. And therefore he concludes that multiple realizable properties are not scientific. So I'm going to skip on further supply, but if anyone is interested we can talk about that in the Q&A. So here is how block thinks about this problem. How block tries to solve the problem. So block distinguishes between two types of properties. What he calls D-proper... Oh, that's kind of... Question? Yeah. Okay, he distinguishes between D-properties and R-properties. So D-properties, D from design are properties which are... So that's bad, because design is selected. But so D-properties are properties which are selected by what he calls forces of convergence which are forces such as evolutionary force or consciousness design or learning. So this force converge on some high-level say properties which are selected because they have a certain function. They do something which is important for these forces. So D-properties that contrast with R-properties which are properties or actually phenomena so here what blocks us in mind is actually generalization, not properties. But properties are so to the extent that generalizations are so. Let's say. But we are going to see an example in a second. So D-properties they contrast with R-properties which are properties which depend on how a high-level property is realized in a certain in a certain realizer. So these are properties which are due to the peculiarity of the realization base and not to the function that the high-level property has for does. So here are two examples. So as an example of a D-property we can actually put that in flow which is sharp pains that we insert random intervals cause anxiety reaction and that can be considered as a D-property it's a D-generalization involving D-property D-properties cause that's plausibly, let's say that's plausibly evolutionary useful to have this kind to show this kind of pattern so that whenever we have pain we are anxious because we want to avoid the pain, right? And so the two properties which are mentioned in the generalization they are D-properties because they appear in a generalization which is let's say we can consider, we can assume which is shaped by a force of convergence which is an evolutionary force in this case and that contrasts with the second example I'm providing which is a phenomenon which is called aridontalgia so that's basically stimulating nasal mucosa that recreates previous dental pain previous dental pain so basically you touch the nasal mucosa and the subject has some previous dental pain you have at some point of his life and this this thing, this phenomenon this is really due to how dental pains are realized in humans because the realization is connected to some how to nasal mucosa, right? So all of the distinction is kind of clear and now what Bloch says Bloch uses this distinction to say that actually the questions of whether a property is projectable and whether it is a scientific kind these are relative questions, these are relative because these are relative to the property with respect to the first property is being projected, right? So we cannot say that a property is a scientific kind or not, we should ask the question in relation with the property with respect to which this property is being projected because actually what we project are generalizations not properties and therefore the question of whether a property is a scientific kind or not is graded because we have in difference so depending on the type of property there are different degrees of kindness or sensitivity so for example going back we have that pain is more of a scientific kind in the first case because the generalization is more projectable than the second and the second is less scientific kind because that's due to the realization base and so Bloch says that actually Kim's apparatus, conceptual apparatus is just wrong and these questions cannot be answered in absolute terms they have to be relativized to the property with respect to which the property we are considering is being projected and that's the sensitivity is graded is graded because also in the case of the R property there is some reason to think that other, so in the first case, in the generalization case, Bloch says that we have a lot of reasons to think that actually other species will obey the law because the law the generalization is highly projectable because it's a big generalization because it is selected and in the second case we have very low reasons to think so but actually it says that we do have still some reason to think so because there are some constraints which are put which are that the realizers have so a thinker cannot be made of just whatever because there are some constraints that the realizer of the thinker realization of the thinker must obey because for example a thinker cannot be made out of water because that's not solid enough and so and therefore there are some reason to think also that to some extent lower extent than the first generalization other species will obey the second law and that's what he calls Disney principle but well I will not tell you why but maybe you are interested so that's what he says the forces that create complex functions can only move into certain channels the ones provided by the restrictions mentioned in Disney principles these are the constraints and then he says when I can expect different strengths of productivity with respect to different sorts of properties in the Disney principle there is a non-zero probability of similarity even in realization of properties and he goes on to say is this gentleman kind he says no because they are disjunctive and they are not projectable and Bloch says the answer is yes and no in various degrees these judgments are kind of relative to psychological design properties but to a lesser extent relative to physical chemical properties however there are some problems with how satisfactory is Bloch's reply and the problem is that Kim's point actually remains and what I want to point out is that there seems to be an inductively relevant difference between two cases the first case is when we project a generalization whatever whichever type of generalization on the same type of entities with respect to which it had already been tested before so we test we have a generalization you take just whichever generalization a de-generalization or a generalization you test it on humans and then we have some high degree of expectancy that new instances of humans not tested yet but of the same type of entities will obey the law so this is the first case so the same type of entities when the generalization is projected on the same type of entities and on the other hand when we consider whether a generalization that we test on humans is projectable to other species and that is true it seems to me regardless from whether the generalization is a de-generalization or an air generalization so what I say is that the difference between a degree of certainty that other humans by contrast to autopuses of marching will obey both their dental jaw law and their anxiety law so in both cases we have a lot of more reasons to think that humans will obey the law than other species even though in one case we think that the deep property will hold and so what I say is that Blocks Reply is not able to differentiate between two cases but this is what is needed to meet Kim's challenge and so why is this so? So why is Blocks Reply so unsatisfactory? seems so unsatisfactory and the reason is that actually Kim's points concern the structure of disjunctive predicates which is an a priori fact it seems to be an a priori fact the structure of disjunctive predicates whereas Blocks Reply it's about empirical a posteriori fact so the fact that our universe happens to be such that there are some forces which have made some laws more believable for other species than others so these are two different ways of approaching the question but that doesn't meet Kim's point it really changes the framework obviously Blocks is aware of that but it doesn't seem to be fully satisfactory and so what Blocks says is what is required for a high degree of projectability is that processes like selection and design have connected properties that would not otherwise be connected but what Kim says in his paper is that disjunction is implicated in this failure and the only thing is the following way namely what I've just said and so there seems to be two ways what I call two dimensions two ways the projectability of a property can vary two dimensions let's say but this is really terminological that's nothing substantial about dimensions right so that's really to say that there are two ways the projectability of a property can vary and the first way is relative to the other property with respect to which it is being projected namely whether the generalization is a de-journalization that we have high reasons to believe that other business will obey or not as our generalization so this is the first way projectability of a property can vary and the second is this kind of vertical dimension which is whether we are projecting the property to the same type of entities with respect to which we have already tested the properties of the property to different types of entities heterogeneously different and so this is my what I am going to say about this problem and on the same track of my previous remark I want to distinguish two types yet two types of projectability in Kim's sense so again absolutely here it just contrasts with the relative that just means what Kim means by projectability so we have to distinguish two types of projectability restricted when we project the generalization on the same type of entities that was I just said and unrestrict the projectability when we project the generalization on different types of entities so here is the point Kim's argument so Kim assumes that low multi-realizable laws that involve multi-realizable properties they should be projectable in the unrestricted sense because Kim says should we believe that the generalization will hold for Martians for example that's the unrestricted sense but is the assumption warranted and I am going to argue that it is not so here is some science I hope so this example I have taken so this is the winner of France law and the winner of France law says that very simply that metals whenever they conduct electricity they also heat up and that's because in most metals the particles which conduct electricity they also bring heat they also bring heat right sorry but the same particle do the same job let's say however so this is what the winner of France law says however this isn't finding from Lee and others that Van Aden Dioxide which is a metal violates the winner of France law because it conducts electricity without heating up and that's because in that metal the two jobs are made by different particles and so Van Aden Dioxide conducts electricity without heating up and so the question I want to ask is does the winner of France law need to be projectable to all types of metals so every type of metal whatsoever in order to be scientific and the intuitive answer is that no it does not right so to some extent we would have expected that Van Aden Dioxide would have obeyed the law but before testing Van Aden Dioxide if one tells you hey do you think that the law is is not scientific if that doesn't give us the reason to think that Van Aden Dioxide will not obey the law I think that we don't have to require that and so I say that his assumption is not warranted because no scientific property or law is projectable in the unrestricted sense, right? so why should we require that multivirasable properties have to be projectable in that sense and so what I want to say is that the semantic import actually the extension let's say of multivirasable scientific predicates has to be restricted to some extent at least as far as projectability is concerned so what I want to say is that in the well known France law before testing it on Van Aden Dioxide what metals has to refer to are just only the types the types of metals on which the law has already been tested so here is an objection that I am going to skip because I don't have enough time and here is a more substantive objection which is actually you, let me, me I'm using physical law but actually multivirasable properties and laws they are different from specific physical laws because they are multivirasable, right? because the properties are multivirasable and that is the reason actually why multivirasable properties should be unrestrictedly projectable, a scheme requires right? and so that would go against me, obviously but what I want to answer to this reply is that so first of all one might think one might try to argue that actually even metals they are multivirasable, so the high-level property of metal is multivirasable, right? because we have different physical realizers of metal, we have Van Aden Dioxide Copper and so on but independently from this issue which can be can be contested independently from this issue actually I want to propose that the realasable sentry be kinds they can be individuated without requiring them to pick out all of the realizers so we have a way of conceiving those properties without requiring them to be about all the possible realizers all the possible realizers and this is how I propose to do so this is what I call top-down individuation and approximately, right? so the idea is that actually we define a high-level multivirasable properties in high-level terms without referring to the details of any realizer, right? so we define a high-level property, so paying is the functional property, or whatever high-level property then we fix a realizer type once we have the high-level property we fix the realizer a given type of realizers on which we are going to test the property, the high-level property, right? so so we fix the realizer, so for example we have pain and we say that the realizer of pain at first is human brain, human's brain and thereby the high-level property becomes absolutely projectable in Kim's sense absolutely projectable on new instances of humans once we have fixed the realization base, right? and if that is confirmed then that will give us reason to think that new humans will obey the same law the law within which the high-level property is embedded, right? and that alone that already ensures us that the property is projectable in Kim's sense and then we can add new realizer types by using pragmatic heuristics, right? so we see that that holds on humans but maybe we think that that holds on for other animals because of some heuristics and then we add new realizer types to the initial realization basis, right? and so we gradually add the realization basis without requiring the high-level property to pick out all possible realizers and so that's how I propose to individuate multivariate properties and what I really think is that Brock's sense of relative projection should really be conceived as a pragmatic heuristic to extend multivariate scientific kinds to heterogeneous to new types, different types of entities so this is all I have to say this is a conclusion so what I conclude is that multivariate properties they are as projectable as any other scientific property which is to say they are projectable in the absolute restricted sense restricted on the same type of entities they are relatively in Brock's sense projectable to whatever type of entities because everything is because that's just pragmatic reasoning but they are not absolutely projectable absolutely in his sense projectable in the unrestricted sense which is what Kim assumes in his argument and therefore multivariate properties they can be considered scientific as far as the projectability test is concerned Questions and minutes guys no wrap over so thank you for this really interesting I did not refuse it's not a metal it's an oxide of a metal it's an oxide of a metal it's an oxide of a metal so I I'm not sure whether that's important but it's all about metals it should be able to be naked but not in the oxide and I want to resist the thing that metals are relatively realizable because I think the typical behavior that you will see in metals is uniquely realized by one or the same mechanism so surely different metals but those differences are not relevant to the way they actually exhibit the metal behavior in the same way that for example hydrochloric acid and nitric acid are clearly different chemical substances but the way they exhibit their acidic behavior is by nomating protons and they both use the same behavior so this is in a sense uniquely realized the type of question I was hoping for I'm not a physician but this is precisely the reason why I'm saying that my reply does not hinge on considering metal multivariable right and that's why I propose this way of inhibiting multivariable properties which holds regardless from whether metals are multivariable physical properties are multivariable or not and so as concerned the first question I don't know you should argue with Lee and others because what they say is that we report an order of magnitude breakdown of the weather and France law and so if the weather and France law is not a metal that's hard to understand why that would be a violation of the weather and France law right because the weather and France law concerns metal so maybe you can go and argue with that but so I should have said that actually this is really an example so the weather and France law an example because actually Kim uses the example of metals right this is Kim's example to say that some generalizations are projectable and this is what Kim says about metals and he just googled it and found this violation of the oxide and thought okay that's a super nice example but you can find a lot of examples you can take a law which is thought to be a scientific law now and then you can think so is this law going to hold for other for now untested types of entities and the answer I want to give is that we should not require them if we stick to Kim's sense of projectability right then if we go to to block sense the obvious whatever is projectable is that that's that I actually almost have the same question as you it's despite what we say they're just wrong it's not a head on so I think that's the end of the story but but I think you could find other examples yeah I think I think you could find other examples that said I wonder whether or not this is the best type of law to illustrate your point which I think is correct I think your point is correct the reason why is that this is like an empirical law that is known it's the only sense in which it's correct is if you don't measure things too precisely and most metals violate it to some degree yeah or other if you can measure them precisely enough so the thing is that baron dioxide is particular because it violates the law in so let's say high temperature which are not as high as one would think because they are 2013 to 2014 Kelvin which is really really normal temperature and that's why baron dioxide really violates that law and it's not you know so Lee and others they say I think you could do a search you'll find other examples some of them so some people have already suggested me some examples and that one Janis suggested me concerning some of them some of them some of them some of them some of them some of them some of them Janis suggested me concerned the right handedness or left handedness of some atoms but then I don't know anything about that I should leave more details but yeah thanks anyway for this this is obviously important because I cannot make the case if that's not right okay yeah thanks I was wondering whether your reply could not still be regarded as somewhat reductionist in spirit because when you have these laws like the Wiedemann Franz law as the example you gave that are not in the unrestricted sense projectable you would normally have I guess replacement law maybe that is explicitly restricted to those kinds of realises for which it works which is still projectable in the unrestricted sense maybe and if we have such a replacement law that at least first looks like better law and better candidate for like being ontologically committed to the kind of entity that appears there so I guess that reductivists could just say well this speaks in favor of our position because the real laws the laws that are projectable in the unrestricted sense they only contain the non-multiple realisable properties yeah right because you have like made all the distinctions that you end up with probably non-multiple realisable laws or properties that are projectable in the unrestricted sense so I just thank for the question but notice that I'm not trying to defend anti-reductionism this time so that's not my even if multi-realizable properties they happen to be reduced after all still they are scientific right that's what I'm defending here but I see the point but what I say to that is that just that they are not unrestricted projectable but if we find a high-level property which is so if we define high-level property and fix a realizer right and we see the high-level property holds for the realizer and then so we think the property holds for that but then we have another realizer type then the law is really multi-realizable it's hard to see how that can be reduced right because these are two different orders of sides so you say if metal is multi-realizable then that does not guarantee the anti-reductionism right but I want to what I can say is that that does not imply reductionism either and that's maybe any pretty good question whether we can reduce them it depends on whether we can find scientific generalizations which hold only so which holds for the high-level properties and that we cannot you know study by studying the realizers yeah but I may be with that okay thanks but if you want to go back to I want to make it back I was just wondering about the pragmatic heuristic you were talking about when widening the number if you just go back to this slide so when you widen the number of realizers I was wondering wouldn't this like introduce some human element into some human or human human element something human yeah yeah because that's pragmatic yeah of course but then the worry is wouldn't introducing this pragmatic element somehow reintroduce or introduce the idea that it somehow not I don't know how to put it like the idea that the idea that like real I was searching for like no the idea that the science, the scientific theory it's not so if we introduce the pragmatic realizers and the science that uses them is not as real but super nice because that's what I say to block right that's actually your thing is just pragmatic and we need something more semantic to meet Kim's challenge and this is why using his way of conceiving projectivity as a pragmatic heuristic okay so you just taking one step back and saying like taking a stock like holding the pragmatic thing and saying okay I'm I'm assuming a pragmatic stance so you cannot fault me anymore no no no no no I provide my semantic stance right that we should restrict projectivity in Kim's sense and that's what's needed to meet Kim's challenge we want to make something out of block's proposal is that we should conceive it as a pragmatic heuristic right because that depends on a posteriori parts that depends on the fact that we have more reasons to believe that the anxiety law will hold for Martians more than the realizer law and that's why I think it's important to provide the semantic story right but that's obviously that's right but that goes in my direction okay yeah I sympathize with this just that many some people might have some I mean block surely you won't agree okay thank you for this so let's thank our