 So first of all, sorry everybody for being late. It was a very stressful morning with lots of practical problems. Is that your present to you? You may know me as one of your colleagues. So yeah, we are going to present some, or I'm going to present some recent results on ground theoretic semantics. Maybe you've all heard me say something about it already, about ground theoretic semantics, but it has been, I mean there has been some progress and I wanted to present this to you. There's very much work in progress in the sense that I'm currently very frenetically working on a paper that should have been handed in almost two months ago, but it's finally getting there and in the process of writing the paper that was supposed to be very programmatic became much more substantial and it's these extra results that I will present to you now. So I was in the middle of writing this and I thought when the opening came in the whip, or there was still a whip to be opened for this month, I thought well I'll be able to say something about it, given that I'm working on it all the time and then I did have something to say about it, but I didn't prepare this talk very much in detail. So the substance is well prepared but we'll see about the presentation. I hope to keep it not too technical because I think the basic ideas can be perfectly captured without the formal stuff around it. I can be of general, hopefully general, philosophical purpose beyond logic, let's say. So what we'll do, we'll propose to characterize the semantics of the meaning of sentences by means of the reasons there may be for their truth and falsity should be added to that. That's key elements, the notion of reasons will be all the time here. We don't mean reasons in an epistemic sense. It's not reasons an agent has for something, but a reason why something is true, independent of who that might be. Now you could give this a more epistemic reading by saying the reason rational agents may ideally have to, or something like that, but I prefer to keep it just purely unlogical if you want. And so a reason is something out there and not in the minds of people. It's not a mental state or something like that, or an aspect of a mental state. And we will practically carry out a semantics based on reasons by means of a theory of conceptual grounding. I'll explain what conceptual grounding is. We've been talking a lot about grounding in the context of the MIS project and several talks in this center. But here it's a specific kind of grounding that only cares about concepts and meaning and not about any metaphysics or something like that. So that will be a big difference. I just hope to say this upfront, do not give any confusion about this. We'll only talk about meaning here. And we claim that this gives the right amount of brain, focusing on the reasons of center. There might be for sentences in the true or false, gives the right amount of brain to introduce many interesting philosophical notions of consequence and propositions that are not part of the orthodoxy, at least. And even some of them are completely original, so very much open for discretion. But it's this misdemeanor proposal of ways you could do things rather than a complete endorsement of it. In that sense, it still is programmatic. It could set up, or I have the secret hope that it could set up a whole research program rather than that this is like an ultimate defense of a fixed and stabilized system. So, we use reasons to... Can you see the full list? I found out the whiteboard a little bit in front of it. If you can not hear it, then it should be fine. So, we want to characterize sentences, not by their truth conditions as it's usually done, but by the reasons there might be, there may be for their being true or false. Again, I won't write true, but it should be false, too. And this is rather unusual. We usually think in possible world semantics, which is the most well-known framework for thinking of such things. We are talking about where and when sentences are true, in which circumstances sentences are true, and possible variations thereof. That's a possible world. You could read this in a sort of temporal way, like possible is possible in the future and in the past. A possible is true in some real possible worlds. Like you always would say, you might see this possible world as some sort of imagined things, theoretical objects, that allow us to think about modality but don't really exist. In any case, there are some points in a frame, a famous set of possible worlds, some points in a frame at which a sentence is true. That's what is of interest. We are able to answer questions about where and when a sentence is true. But a lot of questions that can be asked about the true values of sentences cannot be answered in terms of possible worlds. The why question usually cannot be answered by means of a possible world. It doesn't provide you with reasons, it just provides you with points at which it is true. While these questions are perfectly legitimate, same with how questions you cannot answer them. You need something more fundamental, and given that you can get different answers, and we will give examples immediately, give that you can get different answers to why questions for sentences that are true in exactly the same possible worlds, at exactly the same points, means that maybe those sentences don't mean the same thing. If you can get a different answer to a question with respect to a sentence, one would say that they don't mean the same thing in a deep way. Equating meaning with sets of possible worlds is a weird, but weirdly became a quite standard approach to meaning. This dates back to ideas by Karnab, who considered an intention as a set of descriptions. But then there's Kripke, and a lot of people who work on modal logics have seen it in this way, usually calling it an intentional approach. Although the term is less obvious than one would think, to call it intentional. That's not the issue here. In any case, let's give examples. So if we have the sentence all squares in which one of the four angles is 89 degrees are natural numbers. That's a true sentence. It might be a bit surprising, but it's a true sentence. It's a necessarily true sentence, simply because there are no such squares. So all these no fakes are natural numbers. And I also have all squares are geometrical objects. We have two sentences about squares, but they are true for completely different reasons. This one is true because there are no such squares. This one is true because of the nature of square quad geometric objects. The nature of square, I might say. So there are different reasons why they are true. Why they are true in exactly the same circumstances, namely all possible circumstances. Consider if one supposes that all possible worlds are mathematically well structured at least. Or mathematically obey basic principles of geometry and natural numbers. Which is a fair assumption, I guess. Another example would be, Hillary is a banker, but not a married bachelor. Supposed for a moment that bachelors are just non-married men. I mean, there might be some discussion about that, but bachelors are non-married men, by definition. And we have this dead sentence, and then the sentence, Hillary is a banker, is either male or not male, and either married or not married. Both of these sentences are contingently true. They affirm something about Hillary, it can be true or false. I mean, they are not contingently true, they are contingent, and in some points they might be true, in some points they might be false. But whatever is wherever they are true or wherever they are false, they will both be or neither of them will be. If Hillary is a banker, the first will be true, if he is not, then the first will be false. And same for the second. Given that everything that comes after it is a tautology, or it's necessarily true. Assuming that it's necessarily true that all bachelors are married men, non-married men. So they are true in exactly the same possible worlds, but again there are different reasons for it. And somehow overlapping, like the fact that Hillary is a banker is a good reason to accept these sentences. But the sentences may also be true, because Hillary is a banker, he is male, and he is, well, the first is not. So there are components of these sentences, the big conjunction of three things, a big conjunction of three conjunctions, and this disjunction itself, about this conjunction itself, sorry, is a disjunction. So if we pick this, a reason for this, a reason for either this or this, and a reason for either this or this, that should be a reason enough for the sentence. And here we have also a secret disjunction, so not a married bachelor, means that he's not married, or he's not a bachelor. So a reason should make this true, and then either not being married or not being a bachelor. And so if you analyze those sentences, there are different reasons for these sentences, even if there is also a reason they have in common. So one might accept those for exactly the same reason, but there are other reasons, that are perfectly good reasons for those sentences that distinguish those sentences, which means that they do not mean exactly the same. Although, again, they have a reasoning tone. This meal is a beef stew without beef. That is, it all points false, but it's false because of the beef, the necessary ingredient of beef in a beef stew. This one, this meal is a beef stew cooked without liquid, is false for a whole different reason, namely that every stew is cooked in liquid, assume that this is the case. So it's not about the concrete examples, of course every aspect of it can be discussed. It is about this idea that we can indeed make very good sense of sentences that come with different reasons for their truth or different reasons for their falsity, although they are true in the same circumstances. Wherever we look. Another example is just, if you are not convinced by these cases, because you only think that logical necessity is proper necessity. Just take P or not P and P or not Q. P or not P can be true because of P and because P is being false because of Q. P or not P and Q are not Q. These are both necessary, logically necessary sentences. They are true at all points in all possible, logically possible worlds. And they are both true because of nothing at all. A good reason for them to be true is in the theory I will propose. And I'm quite convinced that they are true for no reason at all. They are just, or no substantial reason, let's say no reason in the world. You don't need any information about the world to know that they are true. So they are true for the zero reason, I'd say zero substantial reason. While this might be true because P is true, this might be true because P is false, this might be true because Q is true, this might be true because Q is false. So while they share the empty reason, this one shares a reason, it gives a reason already like that that the notation will become clear later. So there is three reasons for it, the fact that P is false, the fact that P is true, three possible reasons, the fact that P is true or the empty reason. And here you have also three reasons. Q, false, sorry for my early hang event writing, maybe called now a little bit. So you see that they are different, that they give their truth for different reasons, even though they share one of their reasons. This is a case where they are really the same in all logically possible worlds. All the examples I gave are true in all metaphysically possible worlds or conceptually possible worlds or all mathematically possible worlds. There is a more precise sense of necessity necessary to build those examples. Here I have a basic question, but can you substitute variable? I mean, maybe you could have written a P instead of Q. So is it allowed in your theory at that stage to make a substitution of variable? No, take for this piece concrete sentences about worlds. Yeah, but even concrete sentences, we can abstract the vertical structure of the sentence. Yeah, but then you're making another claim, then it's a formal claim. This would exclude immediately the whole project. We're talking about sentences just as they come and not yet specific properties of them. So does it imply some semantic, for instance, when you speak about passion, it implies the way we define the world, so maybe that's why we can substitute? Oh, yes, yes, yes, yes, sure. But in this case it doesn't depend on bachelors or anything. Take for P, when L is present at the seminar, and for Q, Charles is present at the seminar. I mean, you cannot just replace Charles being present at the seminar or you being present at the seminar before having good grounds to do that. I mean, this might be a consequence of the logic you're using, but if you just look at the sentences and what they mean, it is not obvious that we can do such substitution. See? Yeah, for me, it's obvious. No, but I know that it's obvious, but I would tend to defend the idea that what is interesting in this logic is the structural reasoning behind the sentence. So I'm not very interested about is it Q this person exactly or not? I'm interested in the fact that we are using the same word, and it's contrary with the conjunction, and so that's why our description, that's why the same reason applies for the truth, or wrongness of the sentence. Well, so I think we might capture that same intuition by the idea that they are true for formal reasons on its own, on their own. That is one aspect of these sentences, and that's why they have the anti-substantial reason as one of the reasons. But they might also apply for other reasons than the pure formal nature of them being tautologies, because it's a logical principle as well that from B you can derive, and you're not B. That's what you mean. I mean, in a nice level, I won't see the interest of not allowing substitution at that stage. Well, I don't say that it's in principle forbidden, or that this cannot be an interesting theorem of a certain account of meaning in certain contexts that you can substitute, but you cannot, like, a priorly, or pre-theoretically claim this as a disidirata or something. It comes as a result, or not, at least that's how I would say it. I see your intuition, of course. No, no, no, but I understand. So I will just wonder what we'll be next. Okay. I will say what comes next. Another interesting example for assuming that we are physicalists about mental states, which I completely not, but assume that we are, then my throat aches is true in exactly the same metaphysically possible worlds, as my brain, seen as something very physiological, is in a type of state that perfectly correlates with my throat aches. And necessarily correlates with it. A physicalist would say that every mental state corresponds to a type of physical state. So also these two things will be true together or false together. Fortunately, they're right now false together. So you can, when you ask when questions or where questions or that kind of questions, or in whose questions, like who is me in this case, you will not see any difference between those two, but they are true for different reasons. This is true for reasons of mental states, this is true for reasons of physical states. And then of course those two kind of reasons maybe on a deeper level come to the same thing or grounded in each other, but on a pure conceptual level, they still differ. It's a substantial fact about the world that physicalism is true, or it's for me not true, but assuming that it's true, this is a deep affirmation of the world. And it does not directly on its own give you reasons to go from here to here or from there to there. The reasons, some properties of mental states in this case and properties of physical states in this case are not related. Sorry, I will say this again. So somebody who is a true physicalist will say that this reason for this one gives also a reason for this one that I want to accept. The difference is that these reasons are not exactly the same given that here the reasons are just about mental states and here the full reason is about mental states and the metaphysical facts or metaphysical facts or not that mental states can be reduced to physical states. So this needs a bigger reason than this one, which makes them different, having different meanings even though they are true in exactly the same situations. I mean this is a little bit more tricky of course. You need a fine rate approach of reasons and you need the full reasons sometimes to see the difference. But I think there is the affirmation that mental states can be reduced to physical states even if you accept it. It's a substantial thing. It's not some conceptual notion or something. It's not about language. It's a fact about metaphysical words. Metaphysical universe however you want to call it. That's how I like to see this. So now we're going to make that idea into a semantic approach. How do we do it by grounding claims? These grounding claims are not metaphysical grounding claims because then we cannot say anything about semantics. We're talking about conceptual grounding claims. So what are the conceptual grounds for accepting a certain sentence? It will be a set of analytic predicates about sentences and an analytic predicate about sentences that this set grounds this. So just an example. We use the truth predicates as superscripts. Do we get a bit shorter? We could also use normal truth and falsity predicates. But the fact that it's false, the sentence John is married is false. And the fact that it's true, that the sentence John is a man is true, grounds the fact that John is a bachelor, the sentence that John is true. This is a grounding claim, a conceptual grounding claim. It doesn't say anything about John. It says something about the meaning of bachelor. So one might have deep grounds to say that John is a man. One might have proper grounds to say that John is married, like the existing of a certain certificate or something, or a certain marriage speech act having happened. There might be deep reasons for all these things. But just if we assume being a non-married man to be in the meaning of the concept bachelor, then this holds for conceptual reasons. Purely conceptual reasons. That's why it's conceptual grounding. So simple things, semantically simple things that are prior in sort of semantics of the virtual semantic order of priority, ground, semantically more complex things. And we're not talking about the sentences themselves, but them having certain true value. If you talk about the sentences themselves, you have this real idea that the sentences, they usually express propositions. And you would have that the proposition that John is married and the proposition that John is a man, grounds a proposition that John is a bachelor, which I think is a bit weird because we don't want to ground the things that are expressed by those sentences. We want to ground the words in the sentences, linguistic objects. This grounding relation is all about language and how words in language get their meaning from other words in language. So this is a sentence and we are talking about its falsity, this is a sentence, we are talking about its truth, and this is a sentence we are also talking about its truth. So the concept that is being explained here conceptually is a bachelor and it's explained in terms of manhood and marriedness. Qua words, qua concepts, not qua things in the world. Because I'm not sure that there is a grounding relation between the thing that is represented by a bachelor, the thing that is represented by man and the thing that is represented by marriage. If anything, it seems to be just identical, the conjunction of those two and this qua, their reference, the explaining is there only as you think about semantics, about what means what. Yes. So just about this grounding relation, so you take it to be reflexive. Yes, yes. So we go back to simple or properly simpler stuff and we'll say immediately that that's a bunch of grounding statements like that together will form what we call a constitution, a constitutional language, and that will have to be well found in the sense that there are no circles and there is no infinite chains going back to like always reducing the meaning but never coming to an end where there's a simplest sort of concept that just receives its reasons from the world, let's say. It's reasons to be true. So to define this, but the details here don't matter so much, but the constitution, so this big, big scenario so to say for a language is defined by first defining a proto-constitution. This is a set of schema tough for such grounding claims like the basic grounding claims, not one that are derived or something, but one that are principled. And we add these square brackets because we can, that grounding can happen in a context. For example, if John is Belgian then he's unmarried and if John is Belgian then he's a man. This will still ground that if he's Belgian then he is a bachelor. So there is some context, some conditions that might be added to these elements that made that you can also ground stuff or reduce the meaning of stuff in context, not just when absolutely put John as a bachelor even in a complicated sentence you should be able to reduce this if conditionally asserted. And it's not, in the constitution we will not give the concrete cases of grounding claims, you will give schema tough for grounding claims. Because what you want to stipulate is the meaning of bachelor, not of John. So you take away John as a concrete case has nothing to do with the definition of bachelor and you replace it by a meta variable. And so each constitutional article, as we will say it, we will call it for language, will have such a scheme for the concepts used in the language. The truth of those concepts when applied to an object and falsity of those concepts when applied to an object. But they are for now completely put apart the truth and the falsity and that has important advantages. So we call a constitution, a proto-constitution if indeed it is well founded as I just said. So at some point you reach a bottom and it's indeed non-reflexive. It's really about the reduction of meaning or building meaning on something. You cannot build on yourself. That's like a circular definition. We can apply this to semantic concepts in natural language, but also for logic we can do this. For example, the truth of A and the truth of B semantically grounds the truth of A conjunction B and so on. Maybe this one is interesting. The truth of A and the falsity of B grounds the falsity of A implies B, materially implies B. Nothing like this is very surprising. And I said it in order of docs. But this is maybe a bit more surprising and it's a very interesting thing to have for any coherent concept or complete concept that either the concept is true or false. Our sentences formed with the concept are true or false. This R is an intentional disjunction but that doesn't concern us here too much. And this is zero grounds. So this is not something you have to base on all our semantic elements. This is just an aspect of the meaning of sentences. I think I proposed in the paper as an alternative that you have a sort of predicate, a meta predicate for being byvalent, so that byvalence of go would be the reason, the semantic reason for accepting this principle. That's an option but it has other disadvantages. For now, if we just take this, then in fact we can get full classical logic out of here by just these basic constitutional schemas, constitutional articles as I call them. So this proves that this is actually an adequate semantics for classical logic, besides for many other things but also for classical logic, which gives some reliability to the specific schemas we used here. Maybe not the only thing that is possible but it's one of them. Okay, then to make sense of propositions, given that we have now a theory of constitutions and the constitutional articles, we add a constitutional article to every constitution and we call the resulting augmented constitution. It has this form, this is a meaningless marker that is added to the language and it stands for the proposition expressed by Ips or something like that. So Ips is like A is true or something, or John is a bachelor is true. Well, that's not a good example because this is only introduced for itself ungrounded sentences. For example, if being green is itself ungrounded, the truth of being green is ungrounded in something else, it's just something that we cannot further define or further analyze, it's just either true or false in virtue of acts in the world. Then we need a kind of t-schema at the bottom of this that says A is green is true and the conceptual reason for this is just the proposition that A is green. So this is just nothing else but say that once you reach bottom, once you have itself ungrounded that the analytic principles, then they should be true in virtue of effect in the world. So this is like a way to to... Sorry, I should get to the... Do you want or do you explicitly not want that that's the only good reason you could have for believing that That is an excellent question and it makes a paper double as long. Is there a 5 second answer or are you sure? No, no, I can give it. It's rather technical. So this is like a way that we can already nicely discriminate propositions and define the right kind of consequence relations that we're interested in. So it's like a good first syntactical step but this doesn't really give you a model theoretic semantics that can account for reasons. For now this is just a theory of grounding and the talk of reasons, if you want to make that work properly then you have to give the truth maker semantics in the sense of you define a state space, a mere logical state space so you can add part root in these states and these states they represent possible reasons and then you're going to associate to such things states they are supported by states and these states are reasons so the reasons of complex elliptic statements like this one is not complex but like bachelor, jump is true, this is a complex one those reasons that one gets the reasons by the constitutional article from the reasons for being man and the reasons for being false that is merits and then so you make, you always reduce the reasons to simpler to simpler claims until you reach a basic one like this and that just gets the reasons straight from an assignment function in the model that gives a bunch of states namely a bunch of reasons to a primitive sentence a primitive elliptic claim about that sentence it should always be a bit more it's always about true and false it's never just sentences so that's the way to make the talk of reasons properly work but I was going to like set that aside to avoid the technicalities although it's not very technical I mean it's just basically what I said is the whole semantics but then you have to prove that it does what you say it does you have to you have to set it up and there's a bit of technicality involved it's minimal but this already doesn't, I think it's rather I mean doesn't say anything about reasons because the talk of reasons you need to formalize that you need the state semantics I think but this is sort of T-scheme right in the first sense once you so the conjunction of A you can say A and B is true if and only if A is true and B is true but then at some point you read primitive sentences and they are just true because of the model you have like the basic T-scheme for primitive sentences and this is what that represents and I think that's also a conceptual grounding it's not yet the grounding of the deep reasons the grounding for example of of mental states facts in physical state facts it's not like that well if this linguistic object is true then the truth of A degree is a fact this could be said as a linguistic fact and this is a real fact so I don't think this augmented constitution is will be absurd at all but it's anyway necessary to develop the kind of propositions we are interested in so it's very natural in this approach to give two kinds to define two kinds of very fine grained propositions this one being I would say overly fine grained but interesting for applications and this being the right kind of level of grain to express frequent sense if that's what we want to express so this is simply the set of substantial and I've seen that's non-factive conceptual grounds so the the set of possible grounds one can have conceptual grounds one can have for it but the substantial ones are the ones preceded by this symbol the ones where you really talk about the worlds if those reasons are the same for two sentences then they express then they have the same meaning well I should say for both the truth and for the falsity the same substantial reasons are available then these propositions express or have the same sense and these sentences have the same sense so in that sense they are sense propositions again this is much more fine grained than seeing a proposition as the orthodoxy says it as a set of possible worlds because I mean there can be different reasons while the same while they were the same the same to write exactly the same things in frame this is much more fine grained the structured proposition and the structured proposition just looks at all conceptual grounds so not just the substantial ones but also the superficial ones the for example if we have John as a bachelor then one of the conceptual reasons for it is that John is a man John is a man and not married but this is not this is kind of a step towards the world but it's not yet a fact about the world it's only at this level that you're at the world so it says more about the structure in the way the semantics is obtained for that sentence and that's why it's a structured proposition it doesn't just look at the deep reasons for it but just at the way the sentence is composed it might be a bit weird to have such propositions but structured propositions have been around in philosophy since the mentions of propositions basically I mean since the 20th century I start people in the center of the century started using propositions and they might have some some advantages maybe if you want to represent beliefs like maybe a belief that A and B and C different from a belief that A and B and C even though they just for exactly the same reasons but here it's structured in a different way they're just structured in a different way here the fact that B and C is true is a reason or is a ground let's say here this is not a ground for this one this is the ground for that one and it's still a ground for this one so you they have a very fine-grained kind of propositions that I think doesn't reflect the meaning of a sentence but as I said it's been around in philosophy and I think it has some interesting you can also define just possible world propositions in here but that's not interesting so in that sense it's kind of a conservative approach but then we can and then I will stop define some exact some consequence relations that are much more fine-grained than just classical consequence which you can also define if the constitutions is set up correctly so we will there is something like exact consequence which is a notion that get fine as introduced and this is just inclusion of senses to get the sense all the reasons for the truth of A are reasons for the truth of B one says that it's exact consequence the reasons why he calls it an exact consequence but it's also surprisingly richer than the way he defines it so you would get out of this approach with the constitution that I gave for the classical connectives A exact A exactly A or B is an exact consequence of K you have that kind of things you don't have A and B XA well why not because here the reason for this is the fusion of a reason for A and a reason for B and that is not a reason for A in itself a fusion is supposed to be a unbreakable operation I won't go into the details but it's a consequence relation where you construct bigger things and it might mean something like conceptually explains A conceptually explains A or B but it also holds when it's reflexive so in that sense it's sort of a weak explanatory relation but what I want to say is it's different from the way it's defined introduces it because we have this surprising so A A has as an exact consequence B and not B or A if we look at the reasons for A well whatever they are it will also be reasons full reasons for this one because B doesn't matter what the true value of B is I don't know it changes that I'm very sorry for that so obviously this other one shouldn't hold but the reason for A should also be a reason for B or not B and A given that you don't need reasons for B I mean you can give reasons for B or not B I mean this a reason for B will also be a good enough reason here a reason for not B and a reason for A together will be enough reason for this one but also just a reason for A enough you need to have in the world to get to the truth of this one so in that sense I think any notion of exact consequence shoots maybe that's a too strong point but it makes sense to say that also these things are okay and this is impossible and the existing formalizations of it so this is a sort of constructive consequence it constructs more complex sentences from simpler ones another kind of constructive consequence relation would be constitutive consequence where we take inclusion not in the fundamental reasons the fundamental grounds but we take inclusion in the structural propositions so any reason for this should be a reason for that which means that it creates a sort of constituting consequence so for example if we have A or B this will be a constitutive it will have as a constitutive consequence A or B or C but it will not have as a constitutive consequence B or C because this one is not constituted out of this it's semantically constructed in a different way so if we care about the construction of conceptual reasons why sentences are true this make a quite big difference that's just a really cool relation is that relation played with elsewhere in logic because that just seems neat no it doesn't but it's very close to the notion of grounding because it's almost like so if you had this lego piece it's really a very classic and it's just really cool I'm kind of surprised that I've never thought about that as an interesting logical relation I don't think it exists anywhere it's very simple to define super cool as far as I know 6 5 3 just to give you no excuse we'll quickly stop so one of the things that I didn't say about the structural proposition is that I think it gives if you look at the sentence that expressed the same structure proposition I think this gives an interesting account of synonymy so synonymy is more than arguably is more than just having the same meaning having the same sense it means that it's like built up in exactly the same way too it's just that you don't need a deep analysis to say you kind of need a deep analysis to say that some things are synonymous and this works actually pretty nice in the sense that if we had if single single man and bachelor would be synonyms that's the stupidity but this could be given a formal approach here because the grounds were exactly the same and so these like very close concepts they would get to be completely synonymous while things like well I cannot think of an example but if the deep grounds are the same but it's completely different statistically constructed you cannot say that they are synonymous synonymy is a linguistic notion that also takes into account the way things are linguistically built up so having the same structure proposition expressing the same proposition I think is a defendable account of synonymy that's a major application I would say even though I don't think it expresses meaning it's a good way to formalize meaning so the last thing I wanted to say is that you can give counterparts of these consequences that are constructive, namely the destructive ones so for example going from A and B to A and there is again F2 you can look at inclusion of the senses but now not the truth of the senses corresponding to the truth but the senses corresponding to the falsity now of course we need to discuss B and A so the falsity of B, A and B has grounds for that will be also grounds for the falsity of A right so just by turning this around and getting T and instead of F you get a notion of analytic consequence and analytic implication is also that something that goes way back in literature but again we have an alternative notion here namely the counterpart of this which is the zero-pistologism so not A and A or B gives you B you have this analytically or as an analytic consequence because B reasons for the falsity of this if you reasons for the falsity of this yeah and you don't need anything else this is false just by pure logic too so it also has other reasons of course by B not A and not A but this is already I mean B is really enough to make this whole thing false so you have the same thing that comes back and it's arguable I think as a full notion but in literature this kind of extensions of analytic implication are nowhere taken seriously nowhere even considered and I'm finally and I think this is the nicest outcome of the whole project because I'm kind of surprised by it maybe it will go down my feeling that it's nice so we call it essential consequence and I don't want to say anything about the physical essence I don't understand that literature I mean I have some sympathy for it as a sort of exercise in how to reason research concepts but this is conceptual essence I don't think it's absurd to say that if you accept there is too daily in definition for a human being a man he says as a rational animal I guess he doesn't say man literally it's an intranslation but let's say human being is defined as a rational animal of course again we can discuss whether this is a good characterization of human being human being but then it is in the essence of John's being human being that he is a man he's an animal and that he is also rational and this is not because of the properties of John this is just because of the nature of the concepts of human being and we do this all the time if we wonder what's essentially in a concept like we are thinking about species and biology and we are wondering whether the fact that they can reproduce that they cannot reproduce with you know better this property of reproduction that people associate with species we can wonder is this an essential component of being a species or is just some consequence of it or something that this modify as some laws of nature correlated with it but we like in conceptual research we all of the time wonder whether something is an essential part of a predicate and this gives you a few lines to sentences, essential consequence so it's in the very nature of of A of the meaning of A of the sentence A that that B is also true and so this is simply this destructive relation but now the structural structurally fine-grained counter-parter so like we said here from A or B to A or B or C we can also have from A and B to A and B or and no from this one to this one otherwise it's another consequence this is destructive we break it down and if we're talking about concepts being a bachelor John's being a bachelor S in its essence that he's a man for example or that he's not married so we break things down we build nothing up and we respect the internal structure of the concept which also means that this will always respect semantic priority and that's why I would call it essence so this one says something about reasons in the world it doesn't say something about concept itself but given that the way we define the structured propositions and the way the grounding relation works you get to superficial reasons for the falsity of A by superficial reasons for the falsity of B so we are not allowed in this case to just apply other effects but really the way we're interested in the way A is defined what is in its essence of the concept is about how it is grounded conceptually grounded in what prior concept is it grounded like this one is grounded in this simpler concept and being a bachelor is grounded in being a man for example sorry the falsity of being a bachelor is grounded in the falsity or can be grounded in the falsity of being a man so essence requires a priority of this you in the conceptual essence I think it's quite essential for conceptual essence that you reduce to something more simple and this is automatically guaranteed by the way we set up a grounding relation and I don't know of any point in the literature where people have tried to define conceptual essence and certainly not in a way that is as simple as this and it is reflexive more over it is in the essence of being a bachelor that you're a bachelor given that we have this which also I think makes sense grounding is non-reflexive but this one for questions immediately everything is allowed I'm actually quite thirsty so too nice yeah I'll send an email but you don't have to I should the youtube channel is still on because Pierre is following I know for a fact still on for questions as well feel free to questions in the chat or totally fine I just switched it back on we're good I'll start Peter is in charge oh sorry Peter no no no no that you should that you should ask a question or preside ah yes the whip must be very formal not just to talk yeah now I think it's good to make someone decide who can speak but we're not so many Peter it is this is a clarificatory question I'm confused about the notion of augmented grounds I understand what it is for a truth or a falsity to conceptually ground a truth or a falsity but what does an augmented ground express exactly if not a truth or a falsity is it apt to be a conceptual ground brackets open my impression is that those augmented grounds look like a syntactic proxy for a semantic notion I guess that's okay for a technical reasons but I'm curious whether what philosophical meaning you give to those augmented grounded claims if you want to see it on text no no no did you parse it because yes yes yes no it was very clear it was very clear also well well presented well read so that's an important question so I first thought it was just like a technical trick because after all we want to say that two basic athletic propositions that are completely different express different propositions this doesn't roll out from it automatically given that they don't have grounds at all they are basic so as having zero grounds they would all be the same proposition I mean not zero grounds that's something else you can have empty ground as a ground like in tautologies but having no grounds at all being ungrounded is just a typical property of when you cannot reduce any further so you're like stuck at bottom and you say everything means the same at that level that of course would be awful and completely detrimental for the whole thing I mean you don't have that consequence in semantics in the truth maker semantics because I think all of this is about semantics but I then later it does say something and it does say something you say augmented grounds but I don't think this is a right way to put it and I hope I didn't use that word it's an augmented constitution in the sense that we add articles to a normal constitution where the normal constitution was about reducing a complex concept to simpler concepts at the basic level we think that or I think that the it's here so that may be the table is brown is true as a conceptual grounds simply the fact that the table is brown is brown and this is what that is supposed to express and I think this is still about language and not about something with a physical or something like that that is what a primitive sentence means it just means the fact that corresponds to it in the world there is no nothing else that semantically grounds it but I do think it's about semantics or conceptual I do think maybe the confusion is that I call it conceptual grounding so it's semantic grounding to include such cases but I think it's the same relation it's just at that point for the meaning of a sentence to understand it to look at its grounds you have to look at the world at the facts that corresponds to it and that's what it expresses it's as I said a basic t-scheme t-scheme at the bottom true schema I'm not sure what that clarifies so there's a constitution that we augment with these weird beings but the relation is just the same semantic grounding relation we had before in my opinion maybe this is life chat the last by augmented grounds I mean yes whatever dagger exists substantial ground I've called it yeah so I don't know if that's something about you being convinced or not convinced you might be right that this is a syntactic proxy for something semantic but given that I present these constitutions as fully about semantics and about people define clauses in a truth conditional semantics you can read all these constitutional articles as mere truth conditions so I mean this whole thing is a syntactic proxy of the semantics so I don't see this as a very deep objection and I'm it's related my mind is not he is typing something I believe it's related the question if the project is conceptual grounding why it should always be well founded if it's conceptual grounding isn't it some part of language that is not well founded or circular grounding exist in language because it's related to this it's not exactly the same but here you have your way to get to the bottom it's this variation to get to the bottom all the time in all part of language so this constitution is supposed to be a rational reconstruction of the way concepts are built up it's not supposed to be the meaning somebody has in their head or the meaning in colloquial language this is but then there's something wrong but the way these objects are constructed the concepts are constructed no if they if you define but just maybe an example in Newton's mechanic force defined is defined in terms of mass and mass is defined in terms of force it's purely circular it does not forbid you to apply it in a certain sense but you cannot the concept of force is not well defined and the concept of mass is not grounded something completely independent it's purely circular but when you apply both of them at the same time you get to the world you cannot apply it so it seems that in certain cases not in all cases because I hope it's not the case all the time of course you're right if it was all all circular but sometimes in science it's not well covered but still grounded I can follow a grounding relation between force is defined in terms of mass and blah blah blah and mass is defined in terms of blah blah blah so I can have a local grounding grounding relation but it seems to go in round case in the project we've come up with alternatives for for well founded mass that is more liberal and accepts a sort of circularity I won't go into the details of that but I thought I didn't need it here because you can always as anyway these constitutions are rational reconstructions also in that case it seems that even if it's not it is underdetermined by the claims Newton made you can still I guess take one of the two as fundamental and then ground which one were forced and forced by mass they were ended maybe inertia too right inertia is defined mass inertia is defined in terms of force and force is explicitly defined as variation of quantity of movement which is dependent of mass and if you read it seems that Newton noticed it so at one point in this column he says I have to define mass now and he says its density multiplied by volume which is the recipe is mass volume and later in the book he says mass is quantity of matter which is absolutely no way to operate to be operable so it's a concept that is completely useless if it's defined like that because there is no way to measure an abstract thing like quantity of matter but in turn but what is surprising is that it's not made the theory unattainable so there is still a connection to the world if you worry about that I can understand pure syntax system seems to define their terms in a circular way but they are not talking about the world directly but here you can apply but it's an interesting case but I would say either these these concepts are both like measurable verifiable in the world by separately yeah so that that's I don't know what I didn't understand but and then neither of the two I would say ground each other maybe on a deeper metaphysical level or something but not conceptually then it's just like a theorem of subs like a law of nature rather than a definition so that would be one option but I should have no knowledge of the system so then I would say nothing grounds conceptually want among force and and mass or there is indeed some circularity in definitions and only one of the two is like objectively measurable and then I would say that that one that is objectively measurable is the fundamental unit and the other one is defined in terms of it or neither of the two are verifiable and then still it seems arbitrary which one you choose as the fundamental one all properties have to be observable and your answer is that this apparent coupling of grounding is because we don't have the right concept and we have to find which one is the main one or neither of them and it's something else that would be my conservative answer and that would be for language not the word language so you really convince that language meaning is well found not in a deep way I mean it's always a rational reconstruction right it's for me that I try to conceive of a coherent meaning I guess it's not a deep claim it's like a normative statement yes exactly just for me so I had a similar objection for natural language so I would call it the dictionary paradox when you open a dictionary you always need to read new words into the dictionary and eventually everything is certain and I won't say personally but I won't say this is a problem this is wrong I would rather say this is the way human intuition works you know and a reformer system that when you try even to explain intuitively a reformer system trying to explain some concept like mass, matter like space mass or like vector and number you in your natural language you eventually come up to some circularity and I would rather defend that this is good, this is always to understand the actual things and maybe it's related to infinity because we have a finite concept to speak about infinite possible objects and even when we speak about a tree there is an infinite possible trees and so we cannot see them as an actual infinity but we can see them as a potential one and so the concept of potential infinity and because we can conceive potential infinity we can have concepts but if we were powerful enough to see actual objects we wouldn't need any more the notion of concepts and the reverse is true if we weren't powerful enough to have potential infinity we wouldn't be able to think in terms of concepts so I would say because we are finite we are circular and of course it gives sense to formalize our knowledge and when we formalize we need the bonding to be well ordered and so we have minimal elements but it's also not well ordered, it's well found it's a big difference we have a straight way down there can be several grounds I didn't see that so that was my so what I would like to say is maybe it's more common than Alexan say in natural language yeah no it's definitely more common and I do think there is a lot circularity in natural language so yeah I guess all ends are the same thing like you can do a rational reconstruction and maybe we'll have the same results as the way the circular definitions are used or then I'm being a bit, I go out of this norm of well foundedness what if we try to defend before and what makes a lot of sense for a more metaphysical notion of grounding and I think actually should be every notion of well foundedness should be replaced by this is that in the end there are every sentence is grounded in ungrounded sentences or propositions and this is much more liberal this allows circles because we have over determination I mean you can have several grounds several reasons let's say for a sentence to be true so one of the branches might show a loop if there is still another way to to find a connection to the world and that's perfectly fine that there is another branch that should not just move with your house stuff yes so you should see a grounding relation as a sort of a tree like structure where ideally you have well foundedness so it stops but you can I said by making it more liberal and allowing that there are infinite branches I'm not like that but this thing stops this thing stops this thing stops this thing stops and there also it continues like that like that so here we have an infinite tree of grounding I mean the tree is a bit too simple this is just a representation to make the point clear so you can continue here infinitely but still every single fact will have a fundamental grounds and I think in cases where there is operation of mobile circularity like you have that both of you it probably works like that there is always even though you further define it there is also a measure way to operationalize it and have points inside the circle so maybe man is defined as I don't find an example that's probably a lot of examples but each of these concepts have their different ways to verify them you can say if it's not an instance but you can't say if it is so you need an infinite number of operations that would be a perfect example every question you can say if it's not you cannot answer if it is and so you have an infinite number of applications in your tree and I think there is quite a lot of object like this I have to think about it but also it's possible here it's like a loop on this level because it doesn't mean if you look at this fact it's still grounded so I should specify now the order it makes more sense so every point in here is grounded in some in the worlds or the world is not even though there is this and there is this infinite chain and then the whole thing works the same way except that the augmented constitution should then only deal with those real fundamental elements which anyway is the case that doesn't change so that's a way to maybe include some things and for now this is not at all my final answer it's an interesting point I would say that either there is something wrong with the circular definition if we do it in math or something and it's really viciously circular you never know what they are talking about that can exist that people define concepts like that and then you don't have to develop a good constitution for it because it's not a coherent meaning theory but if you do this circular definition but at each point in the circle or at several points in the circle you find ways to verify them other than by the circle like you have a way to measure math and a way to measure force for example then you have ways into the circle and that might be enough there can be circular definitions but as soon as there is some way into the circle it's not harmful anymore and then you get such structures so that would be maybe an answer but I'm personally I like your opinion because I think you need mathematics for instance the halting program an easy example would be you have a real number and you would like to know if it's stopped, if it's fine, find it or not so you start counting the digits at every step you can say if it's stopped but you cannot say if it doesn't stop so I think you will need that kind of infinite grounding in mathematics and since mathematics is quite useful for a lot of science we need both social physics at some point I think it can also to the objection I'm not sure about the example it seems like not a case of conceptual grounding but a case of computation maybe that's quite different we don't say that it halts because in the next step it halts that's what halting means the way to verify it is different from the way halting is defined right I'm not absolutely sure about the example we can talk about it it might be variance but in general I think in mathematics you might need cases like this I think it may be it depends on your concept of grounding I mean because written on the blackboard I had no idea if it's computable but I assume it is but of course maybe it's not just like that case by example but we didn't speak about that aspect yeah it should be computable and the basic form I presented it I don't see a reason why it wouldn't be but like more interesting cases like the conceptual grounding of quantifiers have more complex kind of constitutions and they will not be computable anymore in any circumstance but maybe in that case even more will be useful today yeah but even non-computable functions are well defined you see there's a difference between that thing we understand very well what they mean we know what they're about but we can't get to the answers we wouldn't care so much if we thought that if the meanings didn't make sense in the same way they probably wouldn't inspire so much research we cannot use it in the definition of math but I think we have to be more careful about structure of meaning and structure of the world we have no reason to believe the world is well founded even the mathematical one but it seems that you have a different reason to really go to well founded structure of meaning which I understand it's in a normative way and we should try to avoid that the question is that your difficulties probably not with this theory is that should we have is there a chance that the language the meaning is well founded useful to describe this messy world that is probably no, that was not the point that point is that do we really believe that human mind the way we give meaning need the minimal things that's a normative claim that's a normative claim that's a normative claim that we should do that because since it's normative we should do that according to Peter yeah but it's normative with a I mean it's normative in the sense that and I also want to claim that it's a useful explication in the Carnapian sense of the way meaning is ideally working in good circumstances so it's not like my arm chair notion of normative like me imposing this on if there is indeed perfectly coherent usage of concepts where it is circular in a deep way then I would be felt compelled to include this if it's dominant coherent it gives coherent language and it cannot be modeled in such a way but I don't really see a very compelling example in such a natural language I mean if you try to program a computer let's say but here I'm making an assumption using this approach instead of statistics and I'm not sure the computer will be able to use words the way it is nowadays when we use words like the table it's always circular at some point but we can understand each other so that's why I think my argument was better for natural language than formal language I agree that we can have this normative statement I saw that I'll think about it, thanks anyway so fine if this is too long of a question you can just immediately pump to beer and that's fine because I just want to ask you to step back and do a comparison just because I'm interested so I predict the answer will have to do with your consequence operators but normally people who get motivated by the kinds of examples that motivate you start to play with impossible world stuff and so I just wonder why like what was the what was other than like you were already sort of you were building these tools in the shop anyway and you saw a cool place you could use them like I think to be clear like I think this is super awesome but like yeah like what what motivated you to try a different approach to those kinds of cases yeah that's an excellent question so this is supposed to be completely conservative approach which already makes it unpleasant to go to impossible worlds um maybe a more provocative way to put it do you think you could implement like translation theorems where everything that you're doing is something that a motivated person could do in an impossible worlds framework if they wanted to yeah so these states they I mean you could interpret them as you want it's just like a state space with the meteorological relation that's all so they could be reasons but you could also call them worlds like very ill-behaved worlds because they like the reason why this table is brown has just to do with the table itself it's not doesn't say anything about rocks on Venus or even about the beer in the beer bar like you have a part of the world and of course an impossible an impossible world can be impossible for two reasons namely that it's inconsistent or that's incomplete so this is a this reason is a radically incomplete worlds namely one where there's nothing at all sets about anything except like the the reasons why why the world the reasons why the table is brown so yeah I think the whole thing can be completely reinterpreted as a bunch of worlds instead of a bunch of states and then they will be impossible worlds does this have an advantage I don't think so I mean you you need to I'm quite algebraic in such approaches you just need the basic structure the most simple structure you can use to to create the thing you want to create the meaning you want to create in this case and if you want to capture them as possible worlds I'm afraid you're going to need some extra technical baggage that is not per se doing anything a structure like a structure as basic as it can be in my opinion and this is very basic but there is no fundamental I mean you can re-translate it as you suggest it's cool I have a very similar impression but is it really a detail somewhere to create a logic or something that seems to be something that our reasons for thinking would be true yes yes yes no of course so that's why the talk comes from and all these consequence relations are relevant consequence relations I mean you need the premise to get to the conclusion they're just relevant consequence relation that like traditional philosophers have taken seriously at some point but you can also, I mean if you just like for propositional logic let the constitution do its work you get a relevant version of classical logic dropping out of it, not the full one to get the full one you have to take into account not just the reasons themselves but also like redundant reasons or redundant grounds like you had too much in a ground than the real ground itself in exact grounds as I would scold them so then you can get full classical logic so I was a bit quick in saying with this constitution you get classical logic you get a relevant variant in fact so the whole thing is deeply relevant at all like the grounding relation is relevant all these consequence relations are relevant these propositional are in some sense relevant but there are two but as a sort of garbage extra for those who don't are not interested in the fine grain of the system so yeah thanks for the question that's indeed where it comes from what's next? you mentioned that you were hoping that this was maybe to start with a much larger project so what we partially have been doing is design constitutions for many logics not just classical logic but also a trististic logic many valued logics and so on in any case this is about true and false always but you can imagine more quantifiers is a very urgent thing to do but it's less evident you need some tweaking in concept of a constitution some pila has been working on some intentional logics and providing what they call logics with intentional connectives and providing constitution for that so there's a whole logical work to be done but just the seeing what these consequences actually come to comparing them with the literature on a philosophical level seems to be quite necessary I just propose them as like a tool and it seems different than the literature like now figure out what to do with it also the concept of propositions so I claim that it's a pretty good way to capture fragrance sense even in cases of like morning star and evening star I won't be able to go into that you know that there's like all these links to different dispositions can also quite easily be captured in such a framework because you have a sort of conditional relation that is relevant in nature and dispositions are always conditional it's always the trigger and the thing that is manifested if you I mean the relevant features of it make that it's quite easy to capture that those should be sticking together in a deep way and you can't just make the trigger bigger or something and still make this position to a word so there's all this kind of philosophical applications that should be worked out or could be worked out if the whole thing is rational in itself yeah so in that sense it's I'd say it's a new way to look at many philosophical concepts maybe mainly philosophy of language and logic but maybe also outside of it a little bit mature another thing and I will stop there is the essence relation the essence relation that was in the conceptual essence it makes kind of sense to to define metaphysical essence if it's essence of properties also in this way with them not with grounding principles that are about meaning but grounding principles that are actually metaphysical grounding principles then you can also define these propositions but they are deep propositions not just linguistic propositions and we find the same relation and maybe this could be an interesting way to conceptualize metaphysical essence but that is extremely speculative and not at all worked out and maybe a final thing I should also say that this is all the technical concepts here are all developed with the whole team which are Daniel, Pilar and Pierre but seeing it as a way to capture meaning is my own position so I don't want to at all like a plain Pierre for this while he contributed to it Pierre prefers to see for now as of course I understand the grounding relation even in the case of logic as metaphysical not conceptual and we wrote a paper of 100 pages almost together recently where we do defend it in that way so I'm a bit incoherent in that respect I for now accepted this notion given that it's more with the literature and so on but there's a story to be told calling it metaphysical too but here everything is about semantics that's important Cavalier type