Simulating the World & Remodeling Philosophy (Session 11)

Hello and welcome to the 11th session of Simulating the World and Remodeling Philosophy with Reza Negrestani. I'm going to pass the mic to him now. Thank you very much, Theo, and my apologies. So we have, if I remember correctly, we have next session and two free sessions which I promised. So we have three sessions total, Is that correct? So during these three sessions, we are going to talk a little bit more this session and the next session. We are going to talk about toy models. Small toy models, namely simplified and idealized toy models and big toy models, which are essentially

of model pluralism. We will talk accordingly about their specific modes of understanding which they afford scientists or systems of scientific theories. we'll get um so this session we'll talk about canonical toy models namely toy models in in the sense that there are simply models whose details are truncated they are essentially collapsed models, collapsed models. So a small in this sense simply means extreme idealization,

kind of like, for example, you know, Schilling's model of segregation. That is actually a small toy model, okay? Then next session we will talk about big toy models. And we are going to make some examples from today's perspective of physics, theoretical physics, particularly with regard to the irreconciliation or the clash between theory of relativity and Newtonian mechanics. To show that big toy models are essentially toy models which are not going

to give us an aspect of a specific phenomenon in the world what simply decomenciate different perspective or pictures of world stories of the world, kind of like relativity and Newtonian mechanics, or a special relativity and general relativity, for that matter. So, so far as we are a little bit behind, let me start our discussion today and then of course we'll have a break as usual and also questions and answer sessions maybe after

I go a little bit into our discussion and then also toward the end. So, we know that across the natural and social sciences, researchers construct very simple, highly idealized models, which the experts in particular field of inquiry can cognitively grasp with ease. following common terminology from sciences, particularly in physics. Actually, the term toy model is coming from physics, from theoretical physics.

We call such highly simplified, highly idealized models toy models. Of course, as I mentioned, this is not purely an accurate description of what a toy model is, precisely because, as I mentioned, toy models also come in different shapes, small and big. A small simply means what I just said, highly simplified, highly idealized. big toy models on the other hand yes they are still idealized models but they can accommodate different models which can be either small or big

they can be toy or non-toy models so they are kind of like as I mentioned as I have mentioned in the past big toy models from a philosophical perspective are essentially metatheoretical models in the sense of metatheory of Carnap or metalogic, that they are capable of accommodating different possible theories which can be brought under them. Okay? So, today when I say toy model, I simply mean a small toy model, the canonical toy model.

Paradigmatic examples of toy models include, for example, Ising model in physics, Lotka-Volterra model in population ecology, and Schelling model of segregation in social sciences. A useful characterization of toy models appeals to three essential features. features. One, models of this type are strongly idealized in that they often include both Aristotelian and Galilean idealization. I will talk about this further on. Toy models are extremely simple

in that they represent a very small number of causal factors. And those of you who took my class in the previous course, by causal factor here I mean exactly what Himpal talks about explanatory factors. So, toy models are not going to give you a full scope of scientific explanation of a set phenomenon or a set model. They are in fact tailored to a set of very limited explanatory factors or causal factors or mechanistic explanations.

3. Toy models refer to a target phenomenon, usually in the sense of small toy models, as opposed for instance models of data. So here again number three can be also understood as one of the reasons why small toy models are different from big toy models. So small toy models actually deal with a real target system, a real phenomenon in the world.

whereas big toy models not only pertain possibly to real phenomena but also and more importantly they pertain to those kinds of systematic theorization or models which cover a specific target system or a specific real phenomenon differently, like a special and general relativity, or relativity and Newtonian mechanics. So, from this point, these three items, we already get the idea that there is in fact

no sharp boundary between small toy models, or toy models in this session, and other models. with the understanding that every model requires a quite good deal of simplification and idealization of both causal factors, explanatory factors, and parameters and variables. So instead of a sharp distinction, there seems to be a continuum of models with respect to the degree of simplicity and independently the degree of idealization.

If one compares toy models with more complex models representing a large number of causal factors responsible for target phenomena, such as complex models in climate sciences, or for that matter in biology, particularly molecular biology, then toy models are located at the simple end of the spectrum of this continuum. If one contrasts toy models with less idealized models, that is models involving fewer idealized and more approximately true assumptions, then toy models are located at a strongly idealized end of this continuum.

Characterization of toy model as simple and highly idealized does of course permit the existence of one, simple and less idealized, two, complex and highly idealized, and three, complex and less idealized models. Remember, so far we have talked about this, the simplification and idealization in accurate scientific modeling sense are two different kinds of procedures. Just because something is more simplified doesn't mean that it can also be more idealized. No, so you have different variations.

And hence these three, again, results. results. Idealization and models have been major topics in recent philosophy of science. For example, you can check the work of 2012, work of Frigg and Hartman. It is however a curious fact that philosophers of science have not devoted sufficient attention to toy models, despite their apparently central role in many scientific enterprises. Toy models, first of all, are deeply positing because they're strongly idealized and simple nature raises hard questions.

And what are these questions? Well, to what end do scientists construct toy models? And to what end are they deepening or escalating the procedures of simplification and idealization? Why should one have any confidence in the claim that a strongly idealized and simple models can be used for modeling any real social or natural phenomenon?

Or even more notoriously, why should one believe that toy models are anything more than so-called purely mathematized scientific fictions, giving us no more clues about real world phenomena than, in fact, non-mathematized false models? Any question? Is all of this clear so far? Okay, good.

So if we were going to be pessimists, we could respond or at least tempted to respond that toy models are not in fact very useful because they cannot represent real or actual physical phenomena. To motivate this sort of a skepticism, suppose that explanation and prediction are two central goals of modeling in science, or for that matter, modern scientific project, the scientific method. We know as a matter of fact that toy models as idealized models are literally false of

their intended targeted systems. put in more semantic terms, toy models clearly do not accurately map onto their target systems. For instance, because a toy model is not isomorphic to its target, isomorphism, insofar as isomorphism is required for representation hence this truncation of both explanatory factors, causal factors and variables which what you might call to be MAIM, the possible isomorphism between

And the model and the physical phenomenon actually translate that to this dictum that such models are in fact not representational. They don't represent anything useful, even if they could represent. So, being false is a feature that at least prima facie, and according to a standard accounts of explanation, undermines the explanatory character of a model.

as the explainance, those factors which explain the explanando of a certain phenomenon. So, sorry, being false is a feature that at least prima facie, and according to a standard accounts of explanation, undermines the explanatory character of a model. as the explainance of an explanation is required to be approximately, in representational sense, to be true. Similar worries, of course, emerge also with regard to the predictive powers or the predictive use of toy models.

The majority of toy models are not suited for precise quantitative predictions. So why should anyone trust the predictions generated by toy models, given that so much of the variables, parameters, and causal factors have been truncated? If one knows that these predictions rest not only with regard to the toy models, not only on truncated assumptions, massively truncated assumptions, but also on false assumptions. In the sense that these assumptions should at one point or another have some sort of representational or isomorphic fidelity.

question? what kind of isomorphic fidelity are you talking about? well that is the whole point isomorphic fidelity simply from a standpoint of modeling in the sense that what we have been talking about at least for a map to be regarded as a model of a territory it should have within the scope of the map's structure

it should cover all the possible nodes of a territory like the states the main roads, a little bit of the, you know, off the beaten track roads, so on and so forth. Essentially, and that's the legend of the map. So, by that I simply mean not in the sense that there is, in fact, a strong representationalism involved. simply I am pointing that even when we are taking representationism or representational fidelity or isomorphism in the sense of a map which is already distorted

with regard to toy models we are in fact talking about a whole new different level of compressing the correspondence or isomorphism in the sense that you don't even show maps, show roads, or tag the name of the cities and the basically nodes or the network. But you simply give a very sketchy, in the technical sense of a sketch, really, an outline of the set phenomenon. and of course this comes back to this

question that any kind of represent there is no such a thing as a representational fidelity outside of a theoretical system when we are talking about representation here we are not talking about the kind of Szilardian picturing or Kantian representational, sensory representational system. We are talking essentially representation from the perspective of scientific theories. That every theory, now you see, representation from a sensory perspective or the concept of observation

from a sensory perspective in the Szilard's and picturing scenario, or Kant in fact, is about experiential observation. But in science, we don't really just have experiential observation. In fact, the notion of the observable in science is quite a very complex concept. It is not about that if you see some dot moving, you can make a rudimentary representational report out of it. No, no, no, no. It's absolutely not a scientific idea of observable or observation.

In science, the notion of observability is completely tied up with the notion of unobservables, usually fictions. Things that are beyond our sensory reaches, as simple as that. I mean, Ian Hacking, I would really suggest his work with regard to notion of observability. Also, Poincaré. Also, Humboldt. These are, I think, really interesting works and fundamentally philosophical also.

to completely get rid of this idea of this kind of a very rudimentary idea that we as philosophers usually have with regard to the concept of observation or representation. No, scientific representation is very, very different from sensory representation, just as the concept of observability is very different from the rudimentary or ordinary sense observation. In the sense that, as I mentioned to you, they are entangled with these theoretical constructs which quite a great deal of them are counterfactual, are modal. They are not sensory.

And of course, this is actually a point that both Carnap and also recently Van Frassen make. These stuff that, from an ordinary perspective, when someone says that, well, electron or an atom, are they real entities? you know yes they are real but they are not real in the sense that you learn it from Kant or from any kind of philosopher this reality is not any kind of sensory metaphysical reality it's a theoretical

reality and that's why Carnap in fact thinks that such questions are essentially pseudo problematic, pseudo questions and they shouldn't be posed within the realm of science. So there is a whole genre of literature around this idea of what actually scientific representation means, what scientific observation means, and whether what is observed from a scientific theoretical point of view can be said to be real in an ordinary mundane sense of how we as philosophers call something real. This is a realism and anti-realism debate that you're alluding to.

I would say that it's actually even more. It is modal realism, realism, anti-realism, and irrealism and fictionalism all bunched together. For a good, so let me give you a kind of like a little bit of this stuff. So you get a structural realist like Lady Man, who is against the kind of big metaphysical question. Nevertheless, when you listen to some of his interviews or read some parts of Everything Must Go, you actually think that, well, he's not that different

from a kind of, you know, a Roth mill realist in a naive sense. He actually does share a little bit of that kind of realist naivete, so to speak, even though he's on a different field altogether. So you have something like that. you have realists in the really naive sense of realism. Then you have good realists who are kind of like coming from a philosophical, you know, basically trajectory that is unfolding post-Cant, so to speak, epistemological ones.

Then you have irrealists like Nelson Goodman, Then you have fictionalists or modal realists. Of course, these can also be fundamentally two different trajectories. For the sake of brevity, I'm just bunching them together. Like people who are like conventionalists. or you can even move further into the extreme modal realism of the kind that Lewis talks about in counterfactuals or possible worlds. So there are all these

genres and essentially one thing is quite evident that after the fall of naive realism or the big metaphysical questions in science after the wake of our power or Vienna circle more generally we know understand that even though our methodologies might differ, our conclusions might differ, essentially scientific representation always, always require the ingression of some modal component, some kind of factual components, unobservables, fictions, literally, so to speak. So there is actually a really fantastic argument

that James Oshie puts forward. But I'm not convinced now. And the thing is that he thinks that Sellars basically war against empiricism, even the constructive empiricism, ironic empiricism, ultimately boils down that Sellars actually defends these fictional modal components as part of the scientific representation. Whereas the empiricists think that, no, it's just really sensory representation all the way down. But I don't think that this is really tenable in the wake of all these new factions that we have in philosophy of science.

In fact, the majority, a great deal of empiricists today also believe in those fictional entities, those modal components or counterfactuals, unobservables. Joven, any input? Oh yeah, sorry. So, dependent on, I guess this is a more general question because I'm a little... So, dependent on all of these different kinds of realism, let's say if we talk about modal realism and fictionalism,

How do we think about being, I guess? Like, what part does being have to play? It's a really general question, but... Yes, well, I don't think that this is... This is a purely philosophical question, I would say. Sure. In the sense that science cannot answer this by any means possible. However, looking into systematization of scientific theories and theory construction, scientific theory construction, we can borrow certain kinds of methods that science does in order to enrich the question of being from a philosophical standpoint.

I, myself, actually take side with deflationaries. Not probably global deflationary, but local deflationaries. In the sense that the designation of being simply is a concept. Of course, this concept, we don't simply mean it in a Kantian sense. sense. It can be just a system of concepts, namely a theory. And this, of course, comes back to Parmenides. Being and thinking are one. I used to think like Ray that I thought that this

Eliotic doctrine actually is trying to elide the distinction between thinking and being. But I think it's far more fundamental, far more profound in the sense that thought is a designation of being. Literally, if something that cannot be designated by thinking, it is not being. It is not being in a true philosophical sense. Of course, thinking, that's why I said that we have to bring some of the scientific methodologies. The idea of thinking can go all sorts of wacky ways. Are we talking about cogitation, or are we talking about cogitato?

Are we talking about objects of thought, or are we talking about the process of thinking? Well, that's, I think that is a question that Erli Hossel put his entire life to answer this question. And of course, he was in a very, very fruitful conversation with the scientists of his time. I'm not sure if he fundamentally answers this question, but I think that it is a worthwhile project. Because to me, at least to me personally, there is no other way to talk about being. And hence, this is the idea of what deflation is. deflationary stance toward the question of being. If you approach the question of being as a B, with capital B,

metaphysical question, then I don't think that in the wake of speculative realism, so on and so forth, you can deny that it can backfire. It becomes the worst kind of mystical question, a sophisticated question in fact. And the only way that I'm seeing it, that we can coherently approach it, is by saying that being is nothing other than existence quantified by thoughts, existential quantification.

And I know that this probably, knowing where you are coming from, this will probably put us in a very good debate down the line with regard to your Laruelian heritage and my deflationary rationalist heritage. Because I think that there is actually from the surface there is a clash here. But I don't think that this clash is fundamental. What clash are you talking about? The clash between a Larwellian. No, but what?

Like where do you locate this clash? I would say that it is simply about that whether thought makes being or being makes thought in the last instance. Yeah, yeah. Yeah. You see? This is, this is complicated. I have no interest in going to any of this, but this is just complicated because there's just a lot in La Role that is actually kind of close to what you're doing. Yes, yes. No, I see, I see it.

Particularly after I have been reading Francois, I think that the connections are actually becoming more clear. So I can actually think that, okay, some of these apparent clashes between these two views are actually not clashes. It's just that we are trying to work on the same problem from different perspectives. Exactly. So like for Enfance, for us, the question of integration. I think for you, it's crucial that integration is an integration of being in some sense, but I don't think it's the same thing for her. Yes, yes. The question is, under what premises can you claim this, and then under what conditions of being can you, in the end, try to justify via the last instance, et cetera?

But those are open-ended questions. Yes, yes, yes. Lenka, there is no such thing as an observable thought, unobservable thought, unobservables are in fact the products of systematic thinking namely theorization so when we are talking about unobservables we are simply talking about counterfactual fictional or modal entities or components they can be also systems which are purely the products of theorization and you when you look into the history of physics you can see so

many of these in fact you know starting from the time of Galileo Newton to you know Carnot Boltzmann, Einstein, essentially when a scientist tries to observe a phenomenon, that observation is only an observation to the extent that it can be supported by certain kinds of modal and fictional components within that theory.

So, what I just simply wanted to say that the question of observation in science is really complex question. And it is something that we cannot talk about it right now, but you should perhaps all of you look into it further. It is not a sensory observation in the common sense idea of observing something. Okay? Scientific observation is not an ideal sensory observation. Okay.

Give me one minute. I will come back. Why don't we take a just like a five minute break. Okay, that's superb. Fantastic. Excellent. All right. Thank you.

Thank you. Thank you.

Thank you. Thank you.

Thank you. Well, with regard to Theo's question, well, you have to say exactly where the circularity happens in either of these two debates, either with regard to deflationary realism or what

you might call to be, you know, kind of inflated, inflationary realism. But more importantly, when you say that toy models can be considered in a simplistic way a fundamental epistemological construction of science, well, what do you mean by fundamental here? As if the theory bottoms up, bottoms out at the level of toy models, no, that doesn't really happen. In fact, toy models are derivatives. They can only come from the perspective of theorization. There are simple products, most fundamental products, not

most fundamental components, most fundamental products of theorization. What here do you mean by theorization? Essentially a system, a system of theory in which, as I mentioned to you, you not only deal with sensory experiential component, a la Helmholtz, not only deal with pure rules, a la conventionalism and not only dealing with pure hypotheticals a la Poincaré you are dealing

with all the three of these and how they can be systematized is exactly how a theory is being constructed a scientific theory a modern scientific theory requires three dimensions Sensorial aspect, the empirical aspect, in a naive sense of empiricism, not in the more, what you might call to be, you know, complicated one. Two, conceptualization, rule systems, inferences, and also introduction of hypotheticals, whether there are fictional entities, modal scenarios, counterfactual, so on and so forth.

You can think about all of this. So the dynamic between these three dimensions and how they relate to each other at different phases of a theory development is fundamental to any theory. I mean, I challenge any of you, and this is my challenge to you, come up with a scientific theory that doesn't have these three dimensions, which are not interacting with one another in a dynamic fashion. You can, of course, make such an example in biology.

for example, or in higher order stuff. But of course, when we are talking about biology, you are also talking about molecular biology and chemistry. So yes, on a surface, a specific higher-level theoretical system might look like as if it's just dealing with one of these dimensions, most probably the observation part, or the observation part and the conceptualization part. But if you actually see this theory as nested within more fundamental theories, within chemistry, within the molecular biology or physics, then you see that that kind of illusion evaporates.

Adam, would you say something before I start? I think that was the territory I just wanted to retrace. Just the previous conversation just jumped fairly quickly from toy models to through observability and then we're in thinking and being. And I just wanted to… Oh yeah, yeah. Well, that is unfortunately… that is the natural order of the day for any kind of philosophical discussion no no it's so good i just sort of wanted to replay it a bit so i i think that answer already sort of covered it and and traced it through so that's cool okay no there is nothing

to be sorry essentially as i mentioned you know we are not simply doing uh scientific modeling essentially we want to see these you know scientific methods scientific dimensions in the light of the deep philosophical, corinial philosophical questions. So there is absolutely nothing to apologize for the so-called digression. So I provided a series of what you might call to be cynical or pessimistic or skeptical questions with regard to the question of toy models.

Now opposing a pessimistic attitude towards toy models, some philosophers have claimed that the epistemic goal of toy model is to obtain understanding of natural and social phenomena. For example, Hartmann, the Richt, so on and so forth. By virtue of their simplicity, toy models enable scientists to retain a sort of epistemic access, a minimal but yet necessary epistemic access, to scientific models and the mathematical procedures for solving these models. The simplicity of toy models distinguishes them from other kinds of models.

for instance, from complex models that can only be solved via massive computer simulation, like what I said earlier, like climate models. In the claim that toy models being idealized models yield understanding really warranted, Are toy models appropriate for achieving understanding? And if so, what kind of understanding do scientists get from toy models? So, our goal here is to understand the nature of these questions and properly answer them.

One, by distinguishing two kinds of toy models. And two, developing an account of understanding that is adequate for toy models, such that we don't actually fall into the pessimistic domain that say that, well, toy models are just not bad representations or inadequate representations but they can also be misrepresentations. Well that is absolutely outside out of the question precisely because toy models even when they are small, idealized or simplified fall under a different mode of understanding

and we're going to talk about it and perhaps three determining whether different kinds of toy models are apt for different kinds of understanding, which links it back to two. According to people like Weisberg, the more recent literature on modeling, simple and idealized models are portrayed as minimal models. However, although we are convinced that

some toy models may be interpreted as minimal models, we can argue that other toy models are not best understood in terms of minimal models. So, minimalist interpretation of toy models actually kind of pigeonhole the wider scope of why scientists actually deal, use, construct toy models. So first, in order to analyze whether and how toy models yield scientific understanding,

is perhaps helpful to introduce a distinction between two kinds of toy models. By that I don't mean small and big, but a different kind of distinction between embedded, so called embedded and so called autonomous toy models. So what are embedded toy models? If you remember earlier on I talked about this idea that small toy models are models of phenomena.

So this distinction that I just mentioned with regard to embedded and autonomous toy models perhaps only can be applied to small toy models, canonical toy models and not big toy models. They perhaps can be applied also to victory models but unfortunately I haven't thought about it. But let's assume that we are still in the business of merely canonical or small toy models where the model is actually a model of phenomena, of course highly idealized or highly simplified. However, some toy models, even in this realm,

are also models in a different sense of the term model, in the sense that there are also models of a theory. This observation will permit us to introduce a central distinction between embedded and autonomous toy models. as a kind of a link that will lead us in the next session when we are introducing big toy models. So, some toy models are embedded. These are called embedded toy models in the sense that they are embedded into an empirically,

well-confirmed theory. So this is what the meaning of an embedded toy model is. Embedded into an empirically well-confirmed theory. More precisely, embedded toy models are models of an empirically well-confirmed framework theory within which you can do certain kinds of itself have certain kinds of observation and not others. Essentially you have to abide by the enframing of the theoretical edifice. Embedded toy models, as I mentioned, are models

of an empirically well-confirmed framework theory, whereas autonomous toy models are not. We'll get back to autonomous toy models later. Now this characterization of an embedded toy model relies on a familiar distinction the philosophical literature on models and model theory in mathematics. Which is what? It is the distinction between one, a framework theory and two, models of the framework theory. Framework theory, models of the framework theory.

In model theory, theory is a set of uninterpreted sentences. When model theory is used to express a framework theory, this set of sentences includes, most prominently, the framework theory's abstract calculus and its generative laws, laws of transition transformation. Models of a framework theory are taken to be structures in which the sentences of the framework theory, such as the theory's abstract calculus and generative laws, are true. Or to a state the same point in a less confusing manner,

models of a framework theory consist of a domain of objects and an interpretation of the theories, abstract calculus, and the laws over the domain. Now, in that sense, examples of empirically confirmed framework theories include Newtonian mechanics and quantum mechanics. Well-known examples of models of framework theories are the models of a pendulum and models of planetary motion, being examples of models of classical mechanics. And of course, the standard model of motion being example of model of classical, sorry, a standard model of particle physics, which is a standard model or simply a model of quantum mechanics.

Now models of a theory are constructed within a framework. Constructing this kind of model in order to represent a target phenomenon often requires moving beyond the resources of the framework theory. It consists in making a number of specific assumptions about the target. And of course, usually such assumptions are not prima facie afforded or supported by the framework theory in which the model operates. Sorry if this is getting a little bit too tortuous.

This is what I'm trying to say. You see, here, you know, when... Has any of you read Jay Rosenberg, Fusing the Images? Particularly the last chapter on convergent realism. It's a top-notch, very, very difficult book, but really one of the best masterpiece of contemporary philosophy. Jay Rosenberg, Fusing the Images. I can send it to you if you don't have it.

So, this is the point here. That when we are in the business of modeling, sometimes we actually do weasley moves. In the sense that, so basically every model that we are supposed to make should respond to the framework theory. Right? But for a great deal, when we make models in science, we make certain kinds of assumptions which are not in fact supported by our framework theory. So where are they coming from?

They're coming from different kinds of theories, sometimes rival theories in fact. So you see, here, we are in a kind of a Kuhnian scenario. Why I'm saying a Kuhnian scenario? In the sense that the enterprise of scientific progress, or in our small domain, model making, is reflective. of scientific history as a whole, that we are never in fact in the business of one single theory or framework.

a great deal of scientific achievements just exactly like philosophy can only be achieved when we start to instill some possible components of other theories which are not well founded which are not well founded in our current theoretical framework in which we are making the model. Exactly like a philosophical question. So to do philosophy is always to do philosophy

in the context of the history of philosophy. There is no way around it. Even if you think that you are answering a question under a certain kind of creed of philosophy, there are certain kinds of moves that you usually do. And of course, that's the distinction between a great philosopher and okay philosopher. A great philosopher knows what he or she is bringing from other kinds of doctrines, philosophical doctrines, and reintroduce them to the fold of the existing framework, philosophical framework. Whereas an okay philosopher usually does this blindly.

The same thing about scientists or model makers. A good model maker is the one who knows exactly what she's smuggling from other theories to this specific theory which does not allow to have those kind of extraterritorial assumptions which are nevertheless necessary to make a model. So,

So with these conceptual tools or movements in mind, Now, we are somehow in a position to characterize embeddatory models more precisely. Embeddatory models are one, models of well-confirmed framework theory, and two, they are simple and idealized models of phenomena.

Before we turn into a more detailed example of an embedded toy model, let's add a note on terminology. As I mentioned, we use the term idealization as an umbrella term for at least two general kinds of idealizations that are usually distinguished in the literature, Aristotelian and Galilean idealizations. A model involving an Aristotelian idealization completely strips away some features that the target system of the model in fact possesses. For instance, a model of pendulum strips away the color of the pendulum.

Or remember our egg example back a few sessions where we talked about that clock thingy. Comes with the shape of an egg made of a certain kind of plastic egg timer. Sorry, clock thingy. Egg timer. Yeah, so the egg timer made of a certain kind of plastic, and you introduce it to the water with your eggs. You really don't care about the shape of this egg, whether it's a perfect egg or not, whether it's pink, blue, or this and that. So that's an Aristotelian idealization. Okay? That's an Aristotelian idealization.

Usually when we talk about... I've noticed that listening to some of you, when we talk, we've been talking about idealization, mostly you think of idealization in that sense, in the Aristotelian sense, okay? But there is also a Galilean idealization. So as I mentioned, a model involving an Aristotelian idealization strips away some features that the target system of the model in fact possesses. Or the model rests on the assumption that some causal factor actually influencing the target system is absent, ignored, or neutralized.

Okay? Aristotelian idealizations are discussed by people like Catwright or Huttman in terms of abstraction and isolation, in a very philosophical sense of abstraction, taking away the process of apharosis, to take away from something. Of course, this taking away can either pertain to the causal factors, the exponential factors, or simply what you might call to be the features,

or a set of features of a possible physical phenomenon. And of course, Gilles Chatelet begins his Magna Moccus, the stake of the mobile, the enjou de mobile, precisely as an assault on Aristotelian abstraction. To what extent can we ever isolate? To what extent can we take away from something that is physical?

You should understand that the greatest goal of Aristotelian abstraction or idealization is to achieve pure form in a platonic sense. And what is that pure form? Does anyone know that? When you purely abstract from a physical component, a limb, step by step, you strip and isolate all of its physical messy problems, what will it give you, what will afford you at the end of the day. Well, it gives you mathematics, Platonic

Mathematism. So that was a little bit about Aristotelian idealization or idealizations. By contrast, Galilean idealizations deliberately distort the target system. For instance, by making the assumption that agents are perfectly rational, that the number of animals in a population or the number of molecules in a gas goes to infinity, and so

on. So you can in fact from this perspective you can think of Aristotelian and Galilean idealizations as a difference between two forms of abstraction in the deep philosophical history. The passive-aggressive one and the active-aggressive one. In the sense that the passive-aggressive one simply doesn't want, it simply thinks that, you see, when Aristotle talks about it in the second book of Metaphysics, it is clear as a day. He thinks that physics

or physical phenomena are just messy. So after all, he's a bad student, perhaps, of Plato. Plato wouldn't have actually approved of this move, but nevertheless. Aristotle thinks that he's quite platonic, and he's so proud of himself, by the way. So essentially, what he wants to do is that okay you get a physical phenomenon and the way that he talks about physical phenomenon is that usually he identifies a physical phenomenon involving motion or mobility of one way or another you can think about different kinds of motions in terms of optics

in terms of projectability you know in terms of a physical motion so on so forth So these he calls corruptions. These are corruptions, physical corruptions. And the only way that we can actually, along with his platonic reading, look into a physical phenomenon is by isolate these goddamn messy problems. aphirosis, take away, abstract, the passive aggressive one. Whereas for Galileo, it's a different kind. And of course, we are called Galilean precisely because it unfolds as a systematic method

after the Galilean term. But of course, even in the Middle Ages, with people like Nicole O'Rean, you know, a few of the Merton College, Oxford School people, this was already a well-founded method of abstraction. In the sense that, of course, Auguste Comte also talks about this, in the sense that sometimes we should think of abstraction not as what you might call to be that Aristotelian idea of purifying, getting rid of the concrete messy problems,

but actually inject something into the concrete system. Distorts the vision of the phenomenon, the physical phenomenon, which we are studying. By saying that the molecules of gas are infinite, Boltzmann. You see, that is a different form of abstraction altogether. is a Galilean idealization. And quite a great deal of scientific discoveries have been made simply through the Galilean idealization. The only way to unveil a black box

is by playing with it. Questions before I proceed? OK. I take it as a no. OK. So, well, Theo, I don't think that we can really.

Essentially, I don't think that is possible. I mean, maybe it is possible, but all I am saying is that from the historical perspective of the progress of science, after Renaissance, after the advent of modern sciences, virtually every theory that has been constructed, every scientific model that has been constructed, in one way or another had to implement or employ both of these forms of idealization. Of course, with various degrees, with various prioritization,

whether you want to emphasize more on the Galilean idealization or the Aristotelian idealization. But I thought that this is actually a really good, outside of our modeling course, it's actually a really great understanding for us in the realm of philosophy, precisely because we usually think of abstraction in the naive Aristotelian sense to take away to isolate and hence we fall victim of to to this parochial opposition within the abstract and the concrete

right the abstract either whether you are more on the side of concrete or not becomes with respect to your approach to the concrete and abstract in this naive Aristotelian sense either less genuine or more genuine essentially you reduce the question of abstraction to an ontological difference between the abstract and the concrete the Galilean idealization the Galilean abstraction changes this from an ontological plane to a purely a methodological plane

that there is no an ontological difference between the concrete and the abstract They are co-constituted epistemologically. Is there another example that you could share of Galilean idealization injecting another element? talking about something like when you like in the heliocentric sort of theory then it brings with the complexity of science in its entirety is actually an example of Galilean

idealization introduction of something like hidden states introduction of something like trajectories or tendencies into the system. Introduction of infinite time via Lyapunov global exponent. These are all examples of Galilean abstraction, particularly global Lyapunov exponents, as the unfolding of tendencies of a complex dynamic system in an infinite time, in an infinite time. OK.

Thanks. I will actually think about a really like a miniaturized, you know, so-called concrete example for you. But I thought that, yes, you remember that we were talking about this stuff in the complexity computation class, I think that yes, really I would say that Poincaré's legacy on complexity science is exactly why it is important from a kind of abstract dynamic point of view, is that he managed to shape a new science which was called ugly. Ugly precisely because it didn't basically abide by the Aristotelian purism or

purification because he actually tried to add stuff to distort physical phenomena in order to understand them better. Either through the mathematical introductions, mathematics bestial, the beastly mathematics, which is the mathematics of entropy and thermodynamics, or through other kinds of hypothetical realizations. Yeah, I said that the black box thing is a very good example, you know, in the sense that the abstraction doesn't become that you can ever unveil a black box, but to unveil

a black box, you have to introduce something to it, which is not inside the black box, to disturb it. And only through this disturbance, this manipulation, this distortion, you can create a set of possible scenarios which you can weed out through observation and theorization. Right. No, no, it sounds really powerful to me. does it connects in a really nice way with some other thoughts that I have. But I just want to, if there's some examples, that would be good because it would be a way of testing my own understanding of it, I guess. Sure, sure.

Let me think about some OK example. Cool. OK. OK. So I merely wanted to highlight that when we are talking about idealization, the term idealization is not a matter of all or nothing. Nevertheless, we have spent a lot of time on this, even though briefly arguing about what an Arsotelian idealization is and what a Galilean idealization is.

We are not going to talk about this anymore. Essentially, from now on, when we are talking, as we are talking about idealization, we are simply treating idealization as an umbrella term for both the Aristotelian and the Galilean methods. Similarly, we are also going to talk about de-idealization, de-idealization in the sense

of denoting both cases which pertain to de-isolation or in an Aristotelian sense and de-manipulation or de-idealization in the Galilean sense. What matters for our concerns is that modeling assumptions involving Aristotelian and Galilean idealizations assert something that is literally false of the target of system. Or put differently, models involving Aristotelian or Galilean idealizations,

primophagy do not accurately represent their targets. For instance, because they are not isomorphic to their targets, if that is what the theory of representation requires, namely simple isomorphism. So let's get back to concrete example of an embedded toy model. Consider Newtonian mechanics, okay, consider Newtonian mechanics as a framework theory. This theory lays out a small number of general laws which are called Newton's equations of

motions or laws of motion and it provides the scientists with guidelines for the construction of concrete models for a specific systems or physical phenomena. To study, for example, the motion of a single planet around the Sun in our solar system, a number of model assumptions have to be made. In the simplest case, one might want to study a system consisting of only the Sun and the planet under consideration. call this simple model the Sun plus one planet model this model is a model of

Newtonian mechanics the Sun plus one planet model is a structure in which the sentences of Newtonian mechanics the framework such as of course the The sentences can be either about the abstract calculus of the theory or its general laws are true. Furthermore, if one analyzes the Sun plus one planet model as the model of the phenomenon, for instance the Earth orbiting around the Sun, this model involves idealizations because the modeler disregards the other planets, the moon, the other stellar objects, celestial objects that are known to exist.

Moreover, the model refers only to the gravitational in the Newtonian sense, the gravitational interactions between the sun and the planet. From Newton's laws of motion and the model assumptions, one can then derive the orbit of the planet in a simple calculation, which can be found in any school textbook on physics. One obtains that the orbit of the planet is approximately an ellipse with the sun in one of the two foci. So the Sun plus one planet model is an embedded toy model because, one, it is a model of a framework theory, Newtonian mechanics in this case, that is well confirmed, at least in a particular domain of application.

2. It is simple, as it describes few causal or explanatory factors, i.e. a physical system of only two interacting bodies. 3. It is idealized, as it deliberately disregards the gravitational influence of other planets and refers only to the gravitational interaction. Four, it is a model of phenomenon, a target phenomenon such as the Earth orbiting around the Sun. Clear? Okay.

Now, there are numerous examples of embedding toy models in physics, including the Ising model of non-relativistic quantum mechanics and Phi exponent 4 theory of quantum field theory with quantum field theory as the embedding framework theory. So we are not going to talk about these. Essentially I just simply wanted to talk about what

what embedatory models are. And of course if you notice that I have only made so far examples derived from physics properly. But you can extend these models. You can actually think to other disciplines. You can think about Fisher's famous sex ratio model as an embedded toy model which is embedded into the Darwinian evolutionary theory of life. So, the only reason that I talked about toy models, precisely because mostly with physical

example was because, as I mentioned to you earlier on, is that toy model is coming from philosophy of physics, physics proper. But of course you can think about so many other examples in the history of other disciplines, biochemistry, molecular biology, even engineering. So, now that we talked a little bit about embedded toy models, we can move to autonomous toy models.

Some actually well-knowing toy models are not embedded ones. Which is to say, they are not models of a well-confirmed framework theory. We call toy models of this sort autonomous toy models in the sense that they are not embedded within a well-confirmed empirical theory. Okay. Autonomous toy models share the simple and idealist character with their embedded cousins.

One example of autonomous toy models is what we have been already talking about quite extensively. Schelling's model of segregation and Lotka-Volterra model of predatory prey population dynamics. So a paradigmatic autonomous toy model as I just mentioned is Thomas Schelling's model of segregation. Schelling developed a famous toy model of the phenomenon of racial and other kinds of segregation.

Racial segregation is a general kind of phenomenon that is contingently instantiated in actual or real-world cities such as Chicago, Detroit, so on and so forth. Schilling's model works with a small number of simple assumptions. One, two source of agents, for instance, black and white agents, simply, again, an idealization. Namely, regardless of what black and white means to us, regardless of all such explanatory factors, they are simply tokens, in a game-theoretic sense.

One, two sorts of agents, for instance black and white agents, live in a very sparse environment, a two-dimensional grid. Two agents are assumed to be initially randomly distributed on the grid. Sorry. And we said this earlier in previous sessions, that simply the random distribution of the black and white coins or go pieces on this grid is really an implicit counterport, an implicit counterport.

You see, there is a difference between an explicit counterport, namely embedded toy model, and implicit counterport, of the theory of the second law of thermodynamics. random initial random distribution represents or is a counterpart of far from equilibrium molecules of a gas of course Schelling is not making no connection here, explicit connection, with thermodynamics. But you should understand that, that the initial random distribution of

of these agents on the grid is simply the initial condition of a system of gas, far from equilibrium. So three, the agents interact in accord with a simple behavioral rule. For instance, each agent moves to an empty spot in her or his neighborhood on the grid, if less than about 30% of his or her neighbors do not have his or her color, the so-called utility function.

So utility function also has a thermodynamic interpretation. But of course, Schelling doesn't want to talk about that. In fact, this is not even part of the model. He just made the model, but the model really is not embedded explicitly in the theory of thermodynamics. The utility function is exactly what Boltzmann calls a disturbance. And do you know what the disturbance is? They can only call Maxwell demon. So you have a bottle, you have a cylinder. The cylinder is compartmentalized to two sections.

In one, you have molecules of a gas, right? Far from equilibrium. It is not far from equilibrium yet, though. It's actually... Now, that's an interesting thing. That's why I think the Thomas Schelling model from a thermodynamic perspective is wrong. Capital W. Wrong. So, the initial condition of any thermodynamic system, in fact, does not begin with far from equilibrium. The hypothesis already speculated in the first law of thermodynamics. is that every system of gas, in fact, starts from pure equilibrium.

The random distribution is, in fact, an additional statement that he has introduced to the thermodynamic system. So, coming back to our cylinder. So you have molecules of gas in one section. Okay, there is a compartment. This compartment is governed by a demon. His name is Maxwell Demon. So the demon occasionally at its whim opened this trap door, leading to the second compartment of the cylinder.

So that's when you get far from equilibrium behaviors, random distribution. From the perspective of the macroscopic laws, namely observational interpretation of second law, is that equilibrium, so usually this is how people read thermodynamics, which is totally wrong. Far from equilibrium to equilibrium. Low entropy to higher entropy. No, actually, thermodynamics experiments

and thermodynamic laws actually talk about this idea that we are essentially working with three phases. One, equilibrium, a disturbance, utility function, represented by the demon, creates a disturbance in the overall system, two compartments of the cylinder, opens the trapdoor, then you get far from equilibrium, random distribution, and this random distribution again tends toward complete equilibrium in isolated closed system. So here two things, a little bit of digression.

Two fundamental, what you might call to be forms of idealization in the Schillingian model. one random distribution can never be from a thermodynamic perspective can be taken as the initial condition of a system okay the movement from equilibrium to far equilibrium to equilibrium again only arises in an isolated environment, in a closed system,

a cylinder, a bottle, so on and so forth. From a physical standpoint, in an open system, you in fact don't get the same kind of phenomenon. Hence, when Shalim talks about the model of segregation, of such behavior, not only he idealizes the second law, distorts it in the Galilean sense, okay? But also, he idealizes the boundaries of the model. It is not a city, really. It is a ghetto.

Model of segregation is a ghetto. literally a closed system from a social perspective in Schellingian sense any questions here if I was unclear or anything I had a quick question absolutely absolutely Justin I was just trying to understand what you were just saying about thermodynamics in relationship to maybe my more general which is really an undeveloped

understanding in general but are you sort of using equilibrium in the sense as like high entropy or how are you I don't see how you're relating entropy and equilibrium in what you Well, you see, that's both from a thermal and a statistical perspective, of course, you see, so here we are. You see, we have two concepts of entropy and equilibrium, a statistical and thermal. So, a statistical equilibrium and a statistical entropy should abide simply by the micro-mechanical laws of physics, where time is symmetrical, right?

whereas the thermal thermodynamics the more common sense thermodynamics is not abiding by the mechanical laws of course it should be explained by the mechanical laws but follows the phenomenological observation of a time arrow in the sense that when you unseal an isolated gas within a cylinder, the gas escapes. Now here, the question of increase in entropy, or decrease in the quantity H from Boltzmannian perspective,

is completely tantamount to attainment or tendency, tendency toward equilibrium. Think of it in terms of Brownian motion, right? The Brownian motion is that all these particles are bumping to each other as you heat it up. They bump to each other faster, collide with each other faster. They evaporate. And that would be the equilibrium. So it's a tendency. High entropy is a tendency for equilibrium.

There are not exactly equal concepts where one leads to the other. From, at the very least, from the perspective of irreversible dynamics observed at the macroscopic level, namely thermal thermodynamics, from a statistical point of view, it's a different matter altogether. And of course, for any scientist, since the time of Boltzmann, essentially one thing is clear, that if we are going to observe, to explain what we have just observed, when a gas escapes through its tendency for higher entropy toward equilibrium,

from a gas chamber, we can only explain it not by what we have observed with our naked eye, but by a recourse to a more what you might call to be causal level, explanatory level. And that explanatory level is a statistical account of thermodynamics. Of course, this is like a, sorry to say this, this is a total mindfuck for Boltzmann and

later scientists. In the sense that, okay, so you say as I mentioned to you guys before, the problem becomes really you know, a headache in the sense that, okay, so, well, okay, wait a minute. So you say that only in the thermal realm, in the macroscopic observable phenomenologicalistic realm, we can see that there is an irreversibility that follows the time arrow. Okay? But then, in the realm of the explanatory factors, which are the micro laws of mechanics, of Newtonian mechanics, or the micro laws of mechanical physics, time is symmetric.

Then how on earth can we actually say that the universe began with low entropy? You see, that's the point here. It basically, it creates an anomaly, an explanatory anomaly. So you wanted to explain what you just observed by moving to a more fundamental physical level, a causal level, and explain what you have just observed. But at that level, of course, the causal factors obey different kinds of laws than the phenomenological laws of observation.

They are the laws of microphysics. They are symmetrical laws. Essentially, there is no such thing as a time arrow. So if there is such a thing as a time arrow, you can also defend the contradictory counterpart of what you have just observed in the sense that you can, in fact, defend the possibility that when the gas escape from the bottle, it can miraculously come to the bottle. Or in fact, more sinisterly, universe began from a very high entropy state.

And it is moving toward low entropy, complexification, so on and so forth. And of course, that creates an explanatory anomaly, Precisely because now the initial position of your system becomes the target of explanation, not what arose, not what came out of that initial condition. So what you in fact, again coming back to Schilling, what you in fact to explain from now on is not the behavior of these coins and tokens on the grid from an arbitrary distribution to a uniform segregated equilibrial state distribution.

But in fact, why is that you chose a random distribution to begin with as the initial condition of your system? Anything here? Any heckling stuff? Okay. So, coming back again to Schelling's model, if we want to start with randomly distributed

agents, then running the simple model by reiterating the behavioral rules according to, of course, utility function, the preference bracketing of these agents. Then running this simple model by reiterating the behavioral rules leads to the emergence of a segregation after a small number of steps. Schelling took the model to explain that racial segregation can occur even if the agents do not have strongly and explicitly racist attitudes but merely conform to

the utility function the 30% rule and agents would actually prefer to live in non-segregated cities the model also allows us to consider the consequence of varying initial conditions and the rules. The Schelling model has racial segregation as its target phenomena. As mentioned, racial segregation is a general kind of phenomenon that is contingently instantiated, for instance in Chicago. If one takes the Schelling model to apply to particular instantiations of racial segregation, for instance the racial segregation in Chicago in the 1960s, then the rules and other modeling assumptions are simplified and idealized

to such an extent that they do not accurately represent, say, the preference of the actual inhabitants of Chicago, Chicago's highly segregated neighborhoods in the 1960s, or, for example, in 2017 or 16. The model is simple in assuming a very sparse environment, a grid, and agents that are characterized by very few properties, most importantly by their color and a behavioral rule. The model is idealized in the following manner. For instance, one, each agent is assumed to know

how many agents of each color live in her or his environment. Two, every agent is assumed to be able to move whenever she or he is dissatisfied with the color of her or his behavior. His neighbor, sorry. Three, social and economic factors such as education and income are taken not to make a difference at all in this model. for the inhabitants of say Chicago and Detroit never randomly distributed and so on. In fact, Schilling actually talks about this when he proposes the model.

And that's why I mentioned that it's not that Schilling thinks that this is actually a social model, a populational dynamics of behavior which is accurately faithful to thermodynamics he actually himself says that there is no such a thing as an initial random distribution and in that sense you can I think about this aspect of Schelling's modeling as a very good example of a Galilean idealization,

In the sense that he assumes that thermodynamic dissipation, even in contradiction to the second law, begins with a random distribution. Nevertheless, the entire system of the model corresponds implicitly, as I mentioned to you, with three laws of thermodynamic.

So, and of course, lastly, Schelling model is not embedded into, that is, it is not a model of an empirically confirmed framework theory. You see, this was an important sentence, what I just said. An embedded model is a model of a theoretical framework. in the sense that its idealization with regard to its affording framework is more like Aristotelian idealization. Simply simplification of the theories, but not manipulation of the theoretical framework.

Whereas this one, Schelling's model, is not actually a model of the theory itself. It is a model that actually, even though has thermodynamic undergirdings, manipulates certain laws of its theoretical framework. certain assumptions of its theoretical framework, namely thermodynamic. And hence, it is what? It is more on the side of the Galilean idealization.

Any questions? Are these stuff clear or are they becoming a little bit too much? I had a quick question about the embedded nature of that example. Shelling's model is not an embedded model. No, I guess we're sort of talking about it as autonomous, but I guess what I'm saying is like, there seems to be, and maybe I'm misunderstanding you, but there's like certain factors, like it does seem to be embedded in a notion of like decision making agents, like a sort of psychological model. So there are certain factors where he's not embedded, but then other factors.

Yes, yes, yes. Now, this is something that actually leads us to what I was talking about earlier on. And of course, ultimately, to the question of the big toy models. You remember that I said that there is no such a theory without isolated from other theories, right? So, yes, Schelling actually made this model in response to thermodynamical behaviors. So, from the perspective of thermodynamic behaviors, it is not an embedded model. It is an autonomous model. but from the theory

that he himself develops out of his goddamn model which is canonical game theory it is an embedded model and not an autonomous model and of course this brings back what I said that the big toy models that's what big toy models are so useful precisely because yes a model within a certain theoretical construct can be seen as embedded or autonomous. From another perspective, exactly like Schelling's game theory, game theory, not thermodynamics, can be thought as embedded and no longer autonomous model.

But you see that there is all these kinds of weird dynamics unfolding here that So the initial point of Schillingian game theory is thermodynamics. So thermodynamic, so this is how it goes. The story goes like this. So Schilling makes an autonomous model, segregation, of thermodynamic principles. principles. To make these models he has to make a great deal of manipulation in the Galilean abstraction sense or idealization sense. Then this model

creates a certain kinds of what you might call to be idealized behaviors within a populational dynamics. Then Schelling's works on these populational dynamic principles. It creates a new theory, which is in a really weird way actually is connected, can be connected back to thermodynamics. It's actually a new theory. The game theory, decision choice theory, Rational choice theory. And rational choice theory actually is something that comes after the model making business.

Now from the perspective of this new theory, the model can be retroactively seen as embedded. but from the perspective of predecessor theoretical framework it is autonomous it is not representative of the theory of thermodynamics and this is not just about shelling you can see these kinds of weird connections

and this kind of basically moving back and forth from one context to another moving from one model which might be embedded to an autonomous but nevertheless might be embedded in a different context as a kind of, as I mentioned to Theo earlier on, as a kind of what you might call to be a very dynamic view of the Koenian progress of science. Well, of course, we can attack this position, this historical view from the standpoint of epistemological questions. But nevertheless, we should recognize

that I mentioned, as I mentioned, the progress of science is not a straightforward path. It creates convergences, divergences, interactions between incompatible theories. They might. And within those theories, you can make models, models that can manipulate theories, models that can represent theories, so on and so forth, and through that you might actually come across a new theory. Theo, you have to elaborate this a little bit.

I don't really know how to elaborate it further just but just read it just read it for me okay yeah you should understand this by now that I'm practically blind I cannot see anything on the sidebar okay all right I just said can I'm not sure if this is a naive question so I suspect it is but can mechanism deal with time asymmetry it seems like at some level a mechanistic framework is wholly atemporal or symmetrical and if it's if it's consistently mechanistic um and i guess that for me it has questions about like the possibility of

what it means for mechanistic theory generating in general no no you are you are right well this is this is exactly the bane the bane at the heart of physics a curse a curse that is a curse only as reichenbach said it once it's only a curse to people who are trying to understand reality simply by means of philosophy But it is a bliss for those philosophers who equip themselves with mathematical physics. Yes, mechanistic explanation is incommensurable to the phenomenal realm.

But, of course, the phenomenal realm since the time of Kant should be explained by some more fundamental level of explanation. Obviously, just because we see a stuff, we experience such and such orders, doesn't mean anything. Really, it's just simply an aperture, an aperture to a wider realm. And that wider realm should explain why and how we see things like that. and yes absolutely this is there is a fundamental historical lesson here beginning with what Hugh Price calls

Boltzmann's time bomb a time bomb the time bomb is simply the metaphor of this asymmetry between the time observed time asymmetry, or what you might call to be the phenomenal observations, and their order, and the mechanistic realm. And how can we ever, well, Boltzmann thought that he actually managed to bridge them together by creating a kind of an objectivist, objectivist, account of the of mechanisms, responsible mechanisms for the observed phenomenon. And that objectivist account is the statistical physics

or a statistical analysis. But then he noticed that, no, he never actually managed to bridge the gap between the two. And literally, all of science, all of the good amount of physics that has been done post-Moltzmann is centered on this gap, the phenomenal and the mechanistic. To the point that some physicists believe that you don't want to suture this gap. You don't want to patch it up. You want to make it bleed as much as it can.

Because that is the source of how we can undermine not only our assumptions about mechanisms, but also our phenomenal observational assumptions or dogmas. But no, this is a serious, serious, serious problem. I actually, I can tell you that if you are interested in this, read the word of Ian Hacking, which I mentioned earlier this session.

You know, definitely Joe's Ofnick, which I have mentioned before, and a couple of other. Also I mentioned, you know, the greatest works in past couple of decades has been written on these issues, fundamental issues in physics, particularly statistical physics, is The Road to Maxwell Demon, published by Oxford, written by Meyer Heimel and Orly Schenker.

h-e-m-m-o and s-h-a-n-k-e-r just to follow up on it it's just a question that kind of confused me ever since like reading kant more deeply that it seems like so much of the description of the way that people are talking about mechanisms seems pretty critical to me. No, no, no. I think that you are making a mistake here. That's why I said, I actually, precisely because I knew that you are going to advance this question by way of Kant's

critical project, I mentioned that quote by Reichenbach, that philosophers, no matter what they are going to do, they are going to always fall victim for dogmas of their own making, right? Philosophical dogmas of their own making. You should understand that the realm of probability and statistics is a fundamentally different realm. It is not philosophical by any stretch of the word philosophical in a Kantian sense. Statistics, of course, is philosophical as a concept. And there is, in fact, some really great work done to, you know, unpack the philosophical concepts of statistics and probability in a modern sense.

you should understand that the statistics is not an observation-based procedure. It is purely objectivist. Yeah. That's why I think that really, that's why I said it if you saw it on Facebook, that one of the greatest questions that has opened up after the advent of modern physics, neuroscience, AI, all of this stuff, is really what probability can teach us. Not just about perception, in the sense that, you know, Hosserlian adumbrations are perceptual awarenesses or perceptual representations.

are essentially probabilistic through and through, but also that we can actually reforge scientific theories from purely a statistical point of view, with the understanding that statistics ultimately is a logical concept. Not in a formal logical concept, but in a transcendental sense. in which the question, these kinds of Kantian stratification between concepts and sense, or categories and sense, are actually become far more, far more complicated.

I think you see this is one of my thoughts is that I think Kant did a great basically thing for philosophy he already he has already paid his dues so no disrespect to him but and what What is this gift that he gave to philosophy? It's just a stratification of a perception. Every perception is an apperception and a perception is in a stratified hierarchical

realm. So you get sense, you get intuitions, you get organized intuitions, you get basically categories of pure understanding, concepts, rules, so many of... and also all different interactions between them. It's just that Kant simply, one, didn't know anything about the concept of probability. His concept of probability was precritical, in fact more rudimentary than Hume. One. Two is that a stratification is great in terms of it shows us the kind of set of problems that we have to deal with.

But if a stratification is taken to be something more than a simple epistemological tactics or strategy, and handled as almost an ontological hierarchization, it totally screws up the whole problems that we are trying to solve, including the preneal epistemological problem, transcendental deduction, so on and so forth.

And of course, in opposition to Kant, you see these new scientific theories like predictive processing brain, you know, called Friston Free Energy Principle, if you have followed his work. by the way, I highly, highly recommend him. He's, after all, the most cited scientists in the world, literally the most cited scientists in the world. All of these are great, fantastic, but you see there is still something missing, and that's why I'm saying that these are...

So you get a scientific bipolar opposite of these Kantian philosophical problems. Of course, even the Kantian, even the scientific oppositions are in an implicit way, can be phrased back to Kant. But there is no what you might call to be philosophical skepticism or philosophical investigation between what science does, actually, with regard to the question of perception, observation, so on and so forth, and the philosophical free plays with such questions with no whatsoever with no scientific methodology

so they should be coordinated and that's the only healthy way to move forward because either way you think that you are doing science, but in fact you are doing pre-critical philosophy, and you are thinking that you are doing critical philosophy, but you are doing really antiquated science. Any more questions? I think we I don't want to overload you with any more

information we will continue next sessions and we have three sessions so we are fine but if any questions Jovan Lenka you know anyone absolutely please yes for so long I'm through my phone so I don't know if something is written on the side but in the beginning I think Theo mentioned about the failure of the models. So I would like to ask if we claim that toy models or models in general are fake models because, I mean, not fake, what is the word? False models.

Because they fail to represent in an absolute way the real. is that because I I don't understand why we say that there are false right maybe I lost yes a big part no no when I said false models I didn't say it's false was not a belittling adjective essentially there are false only in a certain kind of context in the sense of representational fidelity to physical phenomena. Now, there are not really failed models. Yes, from the perspective of

isomorphism, namely representational fidelity, they are failed. But that is exactly the misunderstanding of what toy models are, why they are made of, and why they should be actually defended as the ultimate models precisely because they are not made in order to tell us how actually a physical phenomenon works They are not made to tell us what are the mechanism responsible for such and such phenomenon, physical phenomenon or target system.

They are actually made for a fundamentally different purpose, which is even more radical. How possibly a physical phenomenon can work, not how actually. You see, the failure here is actually the result of misunderstanding the range of applications and the purpose of toy models. Yeah, they reveal secret aspects of reality, let's say. Yes, hypotheticals, yes, so on and so forth. Yeah, kind of like that, yes. how possibly, how possibly the order of things

works. Not how actually it works. And that's something that we will cover next session. Essentially, two understanding with regard to models. How actually understanding and how possibly so understanding. And remember, this is all so goddamn Carnapian here. this is

remember that I told you that the task of enlightenment was not to uncover the order of is but the infinite possibilities of what can be or what might be and that's where the true power of science shines forth not by simply uncovering the facts of reality. Facts of reality can be, well, goddamn rigid dogmas of human experience, but simply to uncover what you might call to be gaps, anomalies, possible scenarios,

which can actually lead us to something else. to expand the order of is. Any more questions? If no questions, then I think that we can finish our class today. and as I mentioned to you, whatever time I owe you, I will just add it to my debit, I will do it.

You see, here now, just when we think that we are done, some people, some people, let's say not who they are, they say that it only seems vaguely dismal. To think of the order of might be seems like a bit of a step down from ought. Well, you see, the order of ought. So there is a whole thing that I can give you on this, in the sense that ought can be interpreted differently. In the Wittgensteinian sense, where ought is simply a rule of established game.

Ought is simply the rule of established game. Established game. This is in boldface. Or ought can actually be a kind of free play, such that we might actually become skeptical of certain kind of game that we are playing. And late Wittgenstein actually believed this, that the language games right are the oughts. But what if that language game, like natural language, like natural language, actually is tethered, chained,

yoked to certain kind of experiential or subjective is dogmas? Then what are you going to do to break from those chains? So you see, you have two different odds if you think of odds as rules. rules of pre-established games and rules of play. With the understanding that the game and play are two different things. One is establishing advance. And two requires a certain kind of manipulation, disturbance, controlled environment, of course. And that's where ought becomes might be.

and engineering, a fundamentally engineering idea, as Karnap said it. The engineering idea is the ultimate idea of the environment, where philosophy rules and possibilities convert. In a sense, exactly like a real concrete engineer. An engineer sometimes, so Mark Wilson actually talks about this story at the beginning of physics avoidance,

his new book. That, okay, so, you know, the engineer, so there is this river, there is this logging plantation on one side of the river, and then there is, they have to carry these logs to the other side of the river, and so they can export them, right? Now the thing is that an engineer can follow what other people have been doing, rules of the games. But an engineer can come up with a certain new kind of model, a model that even though it has been couched in the principle of theoretical physics, but still has give us a little bit

of leeway such that that engineer can make a certain kind of torrential currents or choose This is torrential currents where instead of carrying these across the river, you just simply put the logs into the water on a certain torrential current to carry them across. That's it. Simply sneaking, basically conning established order of things. Conning the established order of things. And that reveals not only a new space of possibility,

but also, as Mark Wilson really goes into detail, so I don't want to give you a headache, it actually gives the engineer a new theory of turbulent fluid dynamics where it can be exploited. You don't need to apologize for being cynic. Well, you see, the idea of objectivity for ought

since the time of place is about fuses. But really, the idea of fuzis is not a reality out there. Fuzis is essentially what Sellars calls the nomos. First, it arises through intersubjectivity in the sense of coordination of reports, experiential reports on a specific phenomenon. And such coordination goes not only through communication of experiential reports, but also goes through the overarching platform of this exchange, namely language and its rules.

OK. It was great to be with you. We'll get back to our discussion next week. Have a great day. Thanks, Reza. Absolutely. Take care.