 Okay, welcome everybody to this fourth session in the Seminar on Metaphysics of Science. It's a pleasure for me to introduce Vanessa Seinfeld as a teaching fellow at the University of Athens. She's going to talk about chemical bounce and patterns. So Vanessa, you have roughly one hour for the talk and one hour for discussion, but you can manage the time as you please. And okay, if you want to share the screen, I find that. Okay, let me show you. Can you see it? Can you see my slides? Yep, perfectly. Yeah, perfect. Okay. I cannot see them. Wait, let me just. Sorry, people at the University of Luván can see the screen or? Well, let's see, it's complicated. I'm not very efficient here. But to do, I mean, we see anyway as whole screen, so it's fine. It starts, there's no reason to wait. I'll fix things here. Okay, okay, perfect. Okay, Vanessa. Perfect. So thank you very much for inviting me and especially Christian for this opportunity to give this talk. This has been a project that I started after finishing PhD. Can you hear me, everyone? Yes. Yes, okay. So let me give you a little bit of a context here. So I did my PhD at the University of Bristol under the supervision of James Lademan, where I worked on the relation between chemistry and quantum mechanics. During the PhD, I was a bit resistant going into thinking about real patterns and structural realism, despite the fact that my supervisor has done such an extensive work on this. I started working on it after I got my degree, and I presented it quite a lot this topic, but today, given the time that I have available, I will go into a lot of detail. The presentation is very much, well, it's completely based on the paper I've written, the chemical bond is a real pattern, which hopefully I expect is going to come out soon. It's in its final reviewing stage, so hopefully now I'll be able to upload it for everyone to read, I suppose. So there's going to be a lot of detail here. Let me start. So, as you can see from the title itself, the claim that I will propose and support today is very straightforward. I'm going to argue that the chemical bond is a real pattern. Now this is a twofold claim. It is a claim about the nature of chemical bonds, so what chemical bonds are chemical bonds are patterns. But it's also a claim, a realist claim. It's a claim saying that in virtue of being patterns, they are real, they exist, right? I'm going to discuss the claim in a lot more detail later on, but this is essentially what I'm going to try and argue today that chemical bonds are real patterns. Now, why should we be interested in the chemical bond? So the chemical bond is one of the most central concepts in chemistry, one of the most standard ways chemists use in order to identify the structure of molecules. And the structure of molecules in turn is also a very important concept in chemistry, but also in science more generally, because it is used in order to explain a variety of phenomena. So not only chemical phenomena, for example, how a certain molecule reacts with another, but also biological phenomena and physical phenomena. So for example, the boiling point of water is explained in terms of the bonds in the H2O molecule and between H2O molecules. The DNA structure is explained in terms of chemical bonds, even drug design, why certain drugs exhibit specific neurological properties, whereas others do not, is also explained in terms of the bonds that certain molecules have, right? So this is a very important concept in science, not only in chemistry. And this in its own, this kind of gives enough support of why we should be interested in what chemical bonds are. Now, you might think that chemical bonds and chemistry being quite of an established science would have already have a very concrete understanding of what chemical bonds are. But if you go into the literature, you will find that this is not the case, right? So what is the chemical bond? Michael Eisberg has famously said about this, that once one moves beyond introductory textbooks to more, oh, I cannot see the whole quote, because it's the images of people on the right, but you can read the quote in four, I suppose. He says, once one moves beyond introductory textbooks to more advanced treatments, I suppose one finds many theoretical approaches, but few, if any definitions or direct characterizations of the bond itself. While some might attribute this lack of definitional clarity to common background knowledge shared among chemists, I believe this reflects uncertainty, or maybe even ambivalence about the status of the chemical bond. Now, this is something that it's not only philosophers who have picked up this ambiguity. You can also find it when reading, you know, classic science papers, even very recent papers in chemistry journals, right? So for example, Zao and Al, who are chemists, have written in a paper where they study the chemical bond that later studies show that the nature of chemical bond is far more complicated than initially thought. And that the connection between the Lewis model and the physical nature of chemical bonding is quite in, what's the word there? I cannot see it. I cannot remember what the word is. Intricate? What does it say there? Intricate. I can only see the in. I couldn't see the rest of the word. Okay, good. So as you see, this is an actual problem. This is not a made up philosophical problem. It is a problem that chemists are also very much concerned with what chemical bonds are. And it's not obvious what chemical bonds are for neither scientists nor philosophers, right? Let me a little bit specify a bit more how this ambiguity arises. So what one standardly does, what one standardly does, when they try to find out what a chemical concept means, they go to UPAC, right? UPAC is the organization that sets kind of the standard definitions at all that the entire chemical community accepts worldwide. So UPAC has like a huge kind of repository of chemical concepts where they provide the definitions of different chemical concepts, including the chemical bond. And the definition of the chemical bond is quite interesting. It says that when forces acting between two atoms or group of atoms lead to the formation of a stable independent molecular entity, a chemical bond is considered to be formed, right? Now notice that this is not strictly speaking a definition. It states when we have a chemical bond, but it doesn't state what a chemical bond is, right? So even when we go to kind of a very standard and primary resources, we find that the ambiguity is already revealed around the nature of chemical bonds. And this becomes even more, what's the word, is even more revealed if we go into the different types of chemical bonds. So in chemistry, chemists pose different types of bonds. We have covalent bonds, ionic bonds, hydrogen bonds, multi-sender bonds, metallic bonds and so forth. Now here we see that the definitions kind of don't help out in clarifying what chemical bonds are because they specify a different referent each time for the chemical bond. So for example, the covalent bond refers to a region of relatively high electron density. The ionic bond refers to an electrostatic attraction. The hydrogen bond refers to a form of association, whereas the multi-sender bond refers to electron pairs, right? So we see that the definition of each type of chemical bonds has a different referent here, which is very strange because all of them seem to be referring to some type of chemical bond. So there should be some underlying definition of what chemical bonds are that encompasses all types of it, right? Now one might think, well, yes, okay, we have this, but doesn't quantum chemistry and quantum mechanics, haven't those sciences provided a way out of this? Haven't they given us a deeper understanding of the nature of chemical bonds? Well, unfortunately, well, to an extent, yes, but unfortunately, they haven't fully. And this is from one main reason that Robin Hendry first kind of identified. The reason has to do with the solution of the Schrodinger equation. So what is usually done when we apply quantum mechanics to describe molecular structure and the interactions that happen within a molecule is we solve the Schrodinger equation, right? So the Schrodinger equation provides the quantum mechanical description of molecules in terms of the interaction of its composing parts in terms of the interactions between the nuclei and the electrons, right? However, in practice, there is no single way of solving the molecular Schrodinger equation. Some argue that this is because it is in principle impossible to solve the Schrodinger equation from first principles and that is why we have to use approximations. This doesn't matter for the present debate. The fact is that there is no unique way of solving the Schrodinger equation. What in practice happens instead is that scientists have developed different methods of solving the equation. Each method seems to accommodate specific kind of explanatory or predictive purposes or seem to fare better with respect to certain types of molecules. Whereas other methods might be more accurate with respect to other types of molecules, right? Now, the problem with this is that the different computational methods that have been developed to solve the equation follow different mathematical structures and make different assumptions of the molecules that they describe. And this in turn has led to different methods kind of implying different understandings of what chemical bonds are. Now, in quantum chemistry, we can distinguish between two main approaches of solving the Schrodinger equation. The one is the valence bond approach and the other one is the molecular bond approach. Now, both approaches, these are kind of umbrella terms that refer to different kind of sub methods that are subsumed in those two main approaches, right? But if we are going to look to the main distinction, we see that the VB approach, the valence bond approach, implies or rather takes that chemical bonds referred to the change in elacone distribution and to the interjected stabilization that results from this change. Whereas the MO approach kind of implies that chemical bonds referred to phenomena, which due to localization are neither directional nor subatomic. So here we have seemingly contrasting understandings of the chemical bond. The VB approach kind of implies that chemical bonds are changes in electron distribution that are directional. They're identified with subatomic regions of high electron density and there are changes which imply they kind of involve also the participation of electrons in a bond, right? Now, this is not what this is not the kind of image that the MO approach produces regarding chemical bonds. In the MO approach, we have chemical bonds as a delocalized phenomenon that is not something that happens within a particular region of the molecule but is rather a molecular white phenomenon, right? Something that happens to the entire molecule and explains its entire energetic stabilization, right? So it's neither subatomic in that case nor directional and it doesn't happen between two sets of atoms, as is often assumed due to the VB approach. So here we see that we already have a problem. Quantum chemistry has in fact clarified the chemical bonds are. It has enhanced the ambiguity that exists on the nature of the chemical bond. Now here looking at the history of how these methods have developed is quite interesting and illuminating in an understanding why we have this contrast. So very briefly we see that initially the method that was initially developed was a VB approach, right? And the VB approach was very attractive because it was very attractive because there was especially by Linus Pauling who developed it in more detail because it seemed to provide a bridge between our standard understanding of chemical concepts and quantum mechanics, right? So the VB approach was taken up by Pauling who then set explicitly to show how through this approach quantum mechanics can become relevant to chemists. And this is why the VB was constructed in such a way so as to retain chemistry's image of the bonds, of the chemical bonds. Now what do I mean by chemistry's image of the bonds? Chemists were greatly influenced by Lewis's work on covalent bonds. So the work of Lewis implied that chemical bonds, especially covalent bonds, refer to electron pairs that connect atoms. And this is considered to be kind of the basis of how chemists understand chemical bonds. And here a quote by Lewis is very kind of illuminates into understanding this image of the chemical bonds. He says, in the mind of the organic chemist, the chemical bond is no mere obstruction. It is a definite physical reality. A sum, what does it say there? Something that binds, essentially, something that binds atom to atom, right? So Lewis took that the chemical bond is a thing. It's an entity that exists within molecules. It is a connective that connects and relates pairs of atoms within molecules, right? So it's something that exists and it's there. It's part of what composes a molecule. And this was very influential and kind of underwrites the standard of understanding of chemistry of the chemical bond. And this is also kind of the spirit that the VB approach tried to also capture, right? However, then we had, unfortunately, and then we saw chemists saw that the VB approach wasn't very good at doing predictions, accurate predictions, and that it failed to describe accurately certain types of molecules. So quantum chemists started to develop an alternative methods of solving the Schrodinger equation, which was the molecular orbital approach. And this kind of gave you a completely different understanding of bonds. It kind of dissipated its reality. It made it a more abstract concept that referred to a phenomenon of the whole molecule rather than of a material part that constitutes a molecule, right? And here Coulson and Mulliken, who were two of the main scientists who developed this approach, were very influential. And they actually explicitly challenged the Louisiana understanding of chemical bonds and its reality as well. So for example, Coulson here said that the chemical bond is not a real thing. It does not exist. No one has ever seen it. No one ever can. It is a figment of our imagination, right? And this is based on his analysis of the MO approach or rather of his development of the MO approach. So we see here that quantum mechanics kind of gave us a good tool to describe molecules, but didn't help us out in understanding what chemical bonds are, it produced different understandings of the chemical bond, which what is important here for philosophy implied radically different metaphysical understandings, right? So we had a very anti-realist, eliminativist understanding of the chemical bond. And on the other hand, we had a very strong realist understanding of the chemical bond as a thing, as a material part within molecules. And we had this contrast in ideas. Now, what has philosophy done about this problem so far? And philosophy of chemistry in particular. There was one philosopher who kind of did extensive research on the nature of chemical bonds, namely Robin Hendry, who wrote this paper called The Structure on the energetic conceptions of the chemical bond. So what he did was essentially propose that there are two different understandings of the chemical bond, if we look at the scientific literature so far. The structural conception, which is in line with the VB approach and the standard chemical understanding of bonds, which states that bonds are material parts of the molecule that are responsible for for localized submolecular relationships between individual atomic centers. And then on the other hand, there's the energetic conception, which kind of corresponds to the work done within the MO approach, where we take chemical bonding to signify facts about energy changes of molecular states. And here it's interesting, and I quote him, he says that under the energetic conception chemical bonding is a theory of bonding rather than an explanation of what bonds are, right? So this is how essentially Robin Hendry tried to recapitulate what is happening in science and how the chemical bond is understood in the entire literature. And indeed, based on this distinction, what has been done so far is debating which of the two conceptions is the correct one. So Robin Hendry distinguished between the two, and then Michael Weisberg took this up and said, well, if we look at the structural conception, this doesn't seem to work out so much. So probably the chemical bond is something that looks like the energetic, what the energetic conception says it is, right? So so far we have had a competition between the two conceptions and a discussion of which one more accurately captures the nature of chemical bonds. However, what I argue in the paper is that neither conceptions provide actually a way out of the ambiguity around chemical bonds. It's not that they give us a solution, these two conceptions of what chemical bonds are, they explain the ambiguity that already exists around chemical bonds. And they summarize and highlight the main points of disagreements that exist in the scientific literature. And so that is why we have the structural conception, which kind of corresponds to the VP and the Louisiana understanding of bonds, and then the energetic conception which has been developed based on the, on our understanding of bonds in terms of the MO approach. Now the problem is that given this, neither of the two conceptions do justice to all of the results we have from chemistry and quantum chemistry, right? Exactly because they capture one of the two sites, neither of the two manages to capture all the facts and all the kind of modules and explanations we have around the chemical bond, right? So we are here a little bit in a standstill. And there are two strategies that one could follow in understanding the chemical bond philosophically. We could retain the discussion and the debate between the two conceptions and try to find evidence in support of either the one conception or the other. Or, as I want to do today, ditch the two conceptions completely and think of the chemical bonds in terms of completely different, in completely different terms. And this, as I will show, will also kind of help us understand the meaning of the two conceptions and their role in understanding the chemical bonds. Now, let me go to my proposal. My proposal is based on Dennett's classic paper, Real Patterns, right? Now, as you might know, Daniel Dennett wrote this paper about what is a real pattern and what is the method of identifying something as a real pattern. So let me explain a little bit this and then I will go to how this applies to chemical bonds. There is a standard image in this paper of six frames in Dennett's paper, right? Now, let's focus on one frame here. Let's say frame E. So each frame consists of black and white dots, right? Now, as you see, there are different ways that one can describe each frame. Let's say frame A. The one way is to identify the position of each and every black dot, right? Another way is to say that, well, this frame consists of six black boxes with, say, 5% noise in between those boxes, right? Now, what does Daniel Dennett say about this? He calls the first description, the bitmap description. So a description which specifies the position of each and every black and white dot that makes up a frame is a so-called bitmap, according to Dennett. And this, according to Dennett, is the least efficient description of the frame because what it does, it specifies the entities and properties of every single entity that comprises a frame, right? However, he says there might be alternative ways of describing that frame that don't require identifying the position of each and every black dot, right? If such descriptions can be offered, according to Dennett, then this means that there is a pattern in how the black and white dots are positioned in the frame. So if I say, for example, the frame consists of six black boxes, this constitutes a more efficient description of the frame than providing the bitmap. And therefore, this is sufficient evidence that my description reversed some pattern within the frame, right? So this is what Dennett actually, this is what Dennett says in this paper. And this is the famous quote where he kind of comes up with kind of encompasses his whole thesis. He says, a pattern exists and some data is real. Is it real? Yes. There is a description of the data more efficient than the bitmap, whether or not anyone concocted it. Now here, there are two important elements for Daniel Dennett. Not only is this a method of identifying the existence of a bitmap, of a pattern, sorry, but automatically if you find a pattern, that pattern is real. Now, of course, this has raised a lot of objections that I'm going to try and deal with some of them today. But this is essentially the account. If you have more efficient ways of describing a system that are more efficient than the bitmap, then that means that there is a pattern there and that pattern. In virtue of being a pattern, says Dennett, is automatically real. It exists, right? Now, let's see how all this can apply to the case of chemical bonds. What I essentially try to do in the paper is apply Dennett's argument to the case of chemical bonds. And indeed, I think you can see that how this goes. So let me explain. So Dennett's method of identifying real patterns applies to how subatomic interactions are described in terms of chemical bonds. First, we have a molecule and we describe it through the bitmap. So the bitmap here would correspond to the description of the quantum state, which is produced by the Schrodinger equation from first principles. So if I take the Schrodinger equation, make no idealization or approximation, the Schrodinger equation, what it would do, it would identify the properties and interactions of each and every entity that comprises the molecule, right? It would be a complete description in the sense that I wouldn't disregard any entity comprising the molecule and none of the interactions that happen between those entities. So in that sense, you could say that solving the Schrodinger equation from first principles corresponds to the so called the Dennettian bitmap. However, both chemistry and quantum chemistry have employed and developed alternative methods of describing a molecule. Quantum chemical methods, whether that is through the MO approach or the VB approach, are more efficient methods of describing the state of a molecule. The chemical description in terms of the postulation of types of bonds is also a more efficient way of describing the molecule. Therefore, this is sufficient to argue, in virtue of the existence of more efficient descriptions, that because such descriptions posit chemical bonds, chemical bonds are real patterns of subatomic interactions, right? So here, essentially, the argument is we take Dennett's framework, we apply it to chemical bonds, the framework works, and therefore we have sufficient evidence to conclude that his argument applies here as well, right? And let me give you a more specific example. Consider, for example, methane. So methane consists of one carbon atom and four hydrogen atoms. So here solving the Schrodinger equation from first principles would amount to identifying the interactions of each and every electron and nucleus that comprises methane, right? All of the atoms, all their nuclei and their electrons, how they interact with each other, this would amount to producing the bitmap. However, quantum chemists have developed different methods of identifying it alternatively. For example, Mendoza and his partners have developed a description of methane through the density functional theory, which is kind of a sub method within the MO approach, very crudely, I kind of describe it here. And this is a more efficient method than the bitmap, right? Also, I should note that it's in the paper, it's not here, that other scientists have developed other methods apart from the DFT. So for example, quantum theory, they have applied the quantum theory of atoms and molecules in order to describe methane. This also represents a different efficient, more efficient method than describing, than producing the bitmap. And then also if we go to chemistry, kind of standard chemistry. Chemistry describes this molecule by positing four covalent bonds, each bond kind of connects bonds, you know, the carbon atom with each of the hydrogen atom. And chemists usually also apart from positing for covalent bonds also identify the length of those bonds, the angles that are formed between them and so forth. And this is also a more efficient description than the bitmap. And this therefore is sufficient to say that Dennett's account is satisfied chemical bonds are real patterns, therefore, right? Now, here though, we need to kind of fill a lot of gaps. The major gap here is what makes a description more efficient than another. And this has been a major kind of point of debate and criticism against Dennett. I'm not going to discuss in detail the entire debate, but I'll try and kind of provide a definition of efficiency that seems to work well in the case of chemical bonds. OK, so Dennett first kind of specifies efficiency in terms of compressibility. He says that if the amount of bits that we use in one description are less than the amount of bits that we use in the bitmap, then this renders the former a more efficient description, right? Now, this is a very computational understanding of efficiency. And I don't want to criticize against it or kind of discuss the problems that this might have. But in any case, it seems a little bit weird to talk in terms of compressibility when you have like non mathematical chemical descriptions of molecules. So whether or not this is correct, it doesn't seem to be very intuitive. So let's just see if we can if we can clarify efficiency in a different way in different terms, right? Now, here there are two proposals. The first one is to just kind of remain naive about efficiency and say, well, efficient is a description that scientists regard as such, right? So scientists would take the MO approach as being more efficient than describing then solving the Schrodinger from first principles, or they would regard postulating for COVID and bonds as being more efficient. What efficient means, we don't care as long as scientists regarded as such, then that is enough for us, right? However, there are very there, as you can already imagine, I suppose there are some problems with this understanding of efficiency. One of them I'll just mention one is that this is a very anthropocentric understanding of efficiency, right? And then also kind of it is a very dynamic understanding of efficiency as well. It is kind of subjective because different scientists might regard something as more efficient than something else based on the specific epistemic goals that they might have. Or also more than that, it might turn out that, you know, once we develop our computational means even further, one might argue that, you know, solving the Schrodinger equation from first principles is no longer regarded as less efficient. So this doesn't seem to work out very well as an understanding of efficiency. So what is the second proposal here? The second proposal is to understand efficiency in terms of degrees of freedom, of minimizing degrees of freedom. And this idea is based mostly on Jessica Wilson's paper. I don't remember the entire title now, it was in BJPS 2017 on degrees of freedom, but also Ladyman and Ross have talked about an understanding of efficiency in terms of degrees of freedom. So let me explain a little bit what this means. So what Jessica Wilson has said is that degree of freedom is an independent parameter needed to characterize. Oh, I cannot see this. Well, let me say it in my own words, a certain description has certain degrees of freedom, the degrees of freedom correspond to the number of variables that we need to identify in order to kind of produce that description, right? Now, what Jessica Wilson says, or rather what the idea that kind of device from this is that description is more efficient than another if it restricts, eliminates or reduces the number of variables that are required to sufficiently describe a system. So for example, if one description requires has 10 degrees of freedom. And another one has five degrees of freedom, the one that has five degrees of freedom is a more efficient description, right? Put in a different way if a description can be formulated by specifying less variables and those required by the bitmap, then the former is more efficient than the latter, right? Now, I take that this is a good specification of efficiency for the present purposes, because it has two advantages. The first one is that this understanding of efficiency is no longer anthropocentric. It is a matter of fact of whether a certain description is going to have a specific number of degrees of freedom or not, right? Based, of course, on scientists' choices, but that those scientists' choices are restricted by empirical evidence and experimentation. And also the second advantage is that it kind of overwrites standard criticisms against then its understanding of efficiency. So the criticism was that then its understanding of efficiency associated that term with simplicity. And simplicity, again, seems to be a very ambiguous philosophical term, an anthropocentric one as well, one that could be subjectively kind of spelled out. So here we kind of dissociate efficiency with simplicity. It has a more robust and kind of matter of factly understanding, right, in terms of degrees of freedom. Now, what makes, though, the quantum chemical descriptions and the chemical description more efficient than the bitmap here? So if we apply our understanding of efficiency in terms of degrees of freedom, we see that it applies perfectly here, right? In the case of quantum chemistry, what we do essentially is apply idealizations and approximations. And by applying those approximations, what we do is disregard either certain entities within the molecule or certain interactions that occur within the molecule. And by doing so, we effectively reduce the degrees of freedom that are required in order to produce kind of a quantum mechanical description of the molecule. And a standard example of this kind of reduction of degrees of freedom is the BO approximation. So almost all quantum mechanical, quantum chemical methods, what they do first is make the BO approximation. And the BO essentially what it does, it disregards interactions between the nuclei in a molecule, right? And this is an example of minimizing the degrees of freedom here in solving the Schrodinger equation. Of course, each method employs other approximations as well, might disregard other additional things too. But we don't care about that here. As long as we can say, as long as we can provide an example of how quantum chemical descriptions have less degrees of freedom, this is sufficient to claim that they are more efficient descriptions than the bitmap. Now in chemistry, we can make a similar claim, though it's more qualitative than quantitative here. Again, chemists disregard certain interactions within a molecule when they produce their description of that molecule. Essentially, when they want to postulate the existence of a bond, what they do is they focus on the electrons that occupy the outer shell of the atoms. They don't have to identify which orbitals are occupied by all the electrons that make up a certain atom, right? So here again, we see that there's a certain disregard of certain interactions and entities, therefore a minimization of the degrees of freedom required. And this is enough to argue that in this sense, the chemical description is a more efficient one than the bitmap description. Than solving Schrodinger equation from first principles essentially, right? So this is a claim and based on all this now, I think one could argue that Dennett's account is successfully applied and therefore we can argue that chemical bonds are real patterns. Now, we have one main problem here. Well, Dennett's account has many problems. I admit that one could discuss many different problems, but there is one that is the most central perhaps. And that is a challenge from pluralism and instrumentalism. Now, what is this challenge? It has been argued, so this debate started primarily from Fodor, that Dennett's account seems to allow that all efficient descriptions identify patterns. And critics have taken this to imply either an instrumentalism about patterns or pluralism about them, right? So what does Dennett say? Dennett say that, you know, as long as we have a more efficient description than the bitmap, then this means we have a real pattern, right? But he doesn't deny that we can have countless efficient descriptions that are more efficient than the bitmap, because remember in the quote, he says whether or not one can cook such efficient descriptions. So strictly speaking, he allows for the possibility of many different efficient descriptions. And why does he allow many efficient descriptions? Because he admits that each description is allowed to make different sort of assumptions, approximations and idealizations. Each description can choose to disregard different things within a system when they describe that system. So in principle, it's possible that we can make up many different ways of describing the system, all of them being efficient, and all of them identifying real patterns, right? So which pattern is real? Which efficient description shall we take for as indicating the existence of a pattern, right? So Dennett doesn't really clarify that, and especially Don Ross has done extensive work in interpreting Dennett. He has pointed out that Dennett himself has been a little bit confused about that. In some writings, he seems to be more of an instrumentalist admitting that, yes, okay, there is kind of a functional role in having these efficient descriptions. There aren't really real patterns, but then he wanted to retain the realism as well behind his account. So there seems to be a bit of confusion here. Now, why is this relevant? Because we see that the efficient descriptions, even in the case of molecules and chemical bonds, suffer from the same problem. So let me return again to methane. So for example, if we describe methane in terms of the DFT method in quantum chemistry, the DFT would describe surfaces of electron charge distributions, not only between carbon and hydrogen atoms, but also between hydrogen atoms. And this would amount to saying that it kind of identifies or rather posits bonds in places where chemistry wouldn't posit bonds between methane, right? On the other hand, chemistry only identifies bonds between an interactions rather subatomic interactions between carbon and hydrogen. So given that both are efficient descriptions, then it would say we have to admit both of them as identifying real patterns. But the image that they produce is a little bit different because one of them identifies only interactions between certain pairs of atoms, whereas the other one identifies different ones. And the criticism goes like this. They say there are two options, therefore, either we admit all of those interactions as being chemical bonds. So any subatomic interaction that is identified by more efficient descriptions qualifies as a chemical bond. Or we are instrumentalist and we say, well, none of them exists. There are no chemical bonds, right? Everything has to do with the purposes we develop those methods. There is no metaphysical weight behind those descriptions, right? Now, this is a very difficult discussion that has taken up a lot of literature and I will not pretend that I can provide an answer against any of those positions. But what I can say is that at least for a scientific realist, none of these positions is appealing, right? Because both interpretations, both instrumentalists and the pluralists dismiss in different ways the special status that chemical bonds have in scientific practice, right? So there is, as a matter of fact, scientists do distinguish certain interactions as being chemical bonds and others as not. And we should try and make our metaphysical account kind of consonant to that fact, right? And in general, I'm here, I'm appealing to a kind of standard, what, intuition behind scientific realism that disregarding the success of science is not in line with the commitment of taking seriously the success of scientific concepts when evaluating the existence of the relevant entities, right? Because as a scientific realist, you want to take seriously into account how science classifies and uses scientific concepts in order to support the reality of the relevant entities. And this doesn't seem to be in line with either instrumentalism or pluralism. So what can we do? The point here is not to challenge any of these two views, but rather to find a way out of it. How can we find a way out of it? Well, what I argue here is that we can take up some ideas from structural realism. Now, structural realism is a very natural position to go to when you talk about real patterns because it is a realist thesis that has been very explicitly and largely based on then its account of real patterns. So the choice to go to structural realism isn't a random choice. It's based on the fact that structural realism have been, realists have been heavily based on the evaluation and use of then its real patterns. So this is what we're trying to, what I'm trying to do here as well. Now, what's important here? The important thing is that structural realism, especially Leitman and Ross have proposed an amendment to then its understanding of real patterns. What they say is, is that real patterns are not only those things which we describe by more efficient descriptions. This is not sufficient, but they are those that indispensably figure in generalizations that allow us to predict and explain the behavior of the world, right? So special science descriptions they claim identify real patterns, not only because they are more efficient than the big month, but because they also make counterfactual and homological generalizations that are very successful when we explain and predict scientific phenomena. And this is what qualifies and this is what we can use as a criterion to distinguish between actual real patterns and not real patterns. Because then this is what qualifies only some efficient descriptions as correctly identifying those patterns and stands as a criterion to discern those, right? So how can we kind of, how can we rephrase this in terms of the chemical bond? Well, if we follow this amendment to the account of real patterns, then we can say that it's only specific interactions, which in the scientific descriptions of bonds of molecules, rather, figuring the counterfactual and homological generalizations, because it is those that play the largest role in explaining a molecule's behavior. Of course, this doesn't mean that the other interactions do not occur, they do not exist within the molecule, or that they don't have some effect on a molecule's behavior, those other interactions that we might kind of dismiss with an efficient description. But we can take those as being negligible, nomologically and counterfactually right. And therefore, it is correct that scientists do not place them and they do not figure in the generalizations that they make when they describe a certain molecule. Now, applied to chemical bonds, this of course explains why only some patterns are identified as chemical bonds in science and why some are not. So, that aromatic compounds, for example, are unusually stable is explained by the fact that bonds are formed between the carbon atoms, right, that water boils at a high temperature is explained by the hydrogen bonds that are formed between H2O molecules, that metals conduct electricity is explained by the ionic bonds. So, here we see a postulating chemical bonds or postulating that certain interactions correspond to chemical bonds and labeling them as such, has a very important role in explanatory and predictive generalizations. And this is what explains why only those interactions are identified as bonds, whereas others are not, right. Now, a criticism that can be raised against this is that the usefulness of a specific methods of identifying certain interactions as bonds and others are not. It's not an objective thing. You know, it's subjective whether a postulating a pattern is going to be useful or not, depending on the cognitive and practical aims and goals and abilities of the person who employs that method, right. And of course, in a sense, to a certain degree, this is correct. The usefulness of positive chemical bonds and the choice of which subatomic interactions are going to be identified as such has a certain degree of is to a certain degree goal oriented. But it is also very largely based on the explanatory and predictive success of the method that they use. And the explanatory and predictive success is evaluated in terms of the empirical evidence we have and we draw from the experimental manipulation and the measurement of the system. And it is this that kind of retains or rather minimizes the challenge from subjectivity, because even if we accept that there is subjectivity about which method we choose and what assumptions we make in that method, the usefulness of depositing chemical bonds within that method that we have subjected chosen is always evaluated on empirical evidence. And this is objective, right. The evaluation of it in terms of empirical evidence, you cannot make up the experimental results you get. So that concludes the presentation of the account. Let me go a little bit into the advantages of this proposal. I'm going to mention four advantages here. The first one, my favorite one is that this understanding of bonds kind of incorporates something that has been missed by previous understanding is conceptual conceptions of the bond, namely that chemical bonds are dynamic things, right. They are interactions between moving electrons. There are things that are dynamic and time dependent, right. They are not a frozen fixed thing in space. There's something that happens a process that happens within a molecule. And this is very much this kind of feature of chemical bonding is very nicely accommodated by its understanding in terms of real patterns, because philosophers who have talked about real patterns themselves have argued that real patterns are dynamic and dynamical things, right. So this fits very well with this feature of chemical bonding because patterns allow for this at diachronicity and dynamicness. No, this sounds wrong. Anyway, this dynamic and diachronic feature of bonds to come up and emerge in the account of real patterns. So that's the first advantage of this proposal. The second one is that it resolves the ambiguity we had around chemical bonds without undermining or dismissing any of the classifications or methods of that we have in science of describing those bonds. So all methods of solving the Schrodinger equation and positing chemical bonds, all types of chemical bonds that are positive in chemistry, all of them are admissible in the sense that all of them identify patterns, right. And now, why do we have so many different methods one might ask well this is accommodated by the account because you can say that you know the proliferation of methods is partially due to the specific aims scientists have, like the predictive aims the accuracy they want from their methods, but also by the specific characteristics of the molecules that they examine. So what becomes very, what is revealed especially if you look closely at the different quantum methods is the main reason why we have so many different quantum methods of solving the Schrodinger equation is because each type of molecule, the rather the inner workings of each molecule are substantially different or rather unique in a way, all of them come down to interactions due to clonforces, but nevertheless, the final pattern the final result of those instructions are very much influenced by the number of electrons and nuclei that we have in the molecule and this changes the image a lot right and this is why we have different methods because some methods manage to identify, for example, very accurately the interactions of very large molecules, where we have many interactions between many subatomic entities, others are more successful for smaller molecules, right. What about the structure and energetic conception? What about them? Well, in one sense, those two conceptions now become obsolete. And in what sense do I mean that? In the sense of what they imply metaphysically. If you take the structural conception as requiring chemical bonds to be material parts of molecules, then this is wrong. And it has been already pointed out so by people such as Weisberg, for example, because then certain molecules are left out by this conception, right. So you cannot say that the structural conception is correct in terms of implying the chemical bond as being a material thing. The energetic conception is also incomplete, metaphysically speaking, if you take it to imply that the chemical bond is a molecular wide phenomenon or property because then you would disregard important features that chemists use when identifying chemical bonds. So the fact that indeed in certain cases of molecules, it is a submolecular phenomenon and this explains, for example, the reaction mechanisms that certain molecules undergo. This is kind of missed out by the energetic conception, right. So metaphysically those two become obsolete. Nevertheless, they are so useful. They might not be metaphysically accurate, but they do identify each one of them highlights different important features of chemical bonds. Chemical bonds as patterns, right. So in certain cases, it is correct to say that those patterns have large energetic effects and like molecular wide effects for certain molecules as per the energetic conception. For other molecules, it is right to say that this pattern kind of is more enclosed in a more specific subatomic region as per the structural conception. But in all cases, what underwrites both of them metaphysically at least is that all that all this refers to patterns of interactions within the molecule. And those patterns are the result of clone forces, right. And that's what the real pattern gives you here. It essentially specifies the nature of the chemical bond and also kind of supports its reality as a pattern. Without without displacing that the two conceptions are epistemically useful in identifying certain distinct features of chemical bondings under specific context, right. So the other thing, the last thing that I want to mention is also suits very well our understanding of the use of approximations and idealizations in chemistry and quantum mechanics. So now under this understanding of chemical bonds, the use of idealization and approximations is no longer a threat towards the reality of chemical bonds. It is no longer, it doesn't need to imply some sort of instrumentalism for descriptions. Instead, the role of approximations and idealizations is kind of to, they correspond rather and then its account to what he calls noise, right. So each efficient description has a certain amount of noise, namely things that it disregards, right. Each efficient description might have a higher or a lower level of noise. This corresponds to having more or less approximations, more or less idealizations in the descriptions. Nevertheless, they all identify in virtue of having such descriptions, we can identify patterns. Now, does that mean that every description with any amount of noise can be accepted as correctly identifying real patterns? For Dennett, yes. For structural realists, no. This is why we needed the amendment to his account of real patterns because what we say here is that science decides on empirical and theoretical grounds whether a specific approximation or idealization is acceptable, whether a specific amount of noise. Let me put it in Dennettian terms is acceptable and therefore whether the description adequately agrees with empirical evidence. So everything comes down to judging the description in terms of empirical evidence and this is going to be the final stage, the deciding step into admitting that description as correctly identifying a real pattern or not, right. So that's the last one. So in conclusion, I argued that chemical bonds are real patterns of several different fractions. This however, in order to circumvent the main challenge from instrumentalism and pluralism, I also adopted an amendment to Dennett's understanding of real patterns, one that includes ideas of patterns from structural realism. Essentially, we take here the fact that such generalizations, that these patterns have to figure in counterfactual and normal logical generalizations. So we kind of admit all of them, right. Now this proposal I argues was supported by scientific evidence fits well with how we describe chemical bonds and how we use them. It does not undermine the special science status of chemical bonds in science. And it captures very important features of bonds, namely the fact that they are dynamic and diachronic. And I take the advantage here is that we essentially hit two birds with one stone. We correctly identify the physical nature of chemical bonds, while also proposing an understanding of the chemical quantum chemical descriptions of those bonds, right. We don't dismiss any of the things that scientists do with respect to chemical bonds. We incorporate them in a way that is consistent, and also provides us a robust metaphysical understanding of what chemical bonds are. I think, yes, I've talked too much. Thank you very much. I'd like to thank. So I've written here all the people that have supported me and helped me out with this project. Thank you. I'm looking forward to your comments. Okay. Thank you very much, Vanessa, for this great talk.