 Hi there and welcome back to history and philosophy of science and medicine. I'm Matt Brown and today we're talking about scientific realism and anti-realism. So our topic today concerns, what is real? How do we know? What sort of attitude should we have towards the claims that science makes when those claims are about things that we can't just see in our ordinary everyday lives, right? All the claims that science makes about what will be, the laws that science posits, the sorts of entities that it concerns. Are these things real? Should we believe in them? That's what we're talking about here today. We talk about scientific realism. I think it's important first to start with just the term realism. What does it mean to be a realist or to accept realism about anything? It means different things when we're talking about different kinds of realism. So I could be realist about consciousness, my phenomenal consciousness, the objects as I experience them. That's not a very heavy weight commitment, although some people would deny that the way I think of my conscious experiences is real. On the other hand, if I'm a realist about the objects of my everyday perceptions, the tables, chairs, plants, and puppy dogs, and so forth, then I'm committed to the existence of those objects, and perhaps to the truth of certain claims about those objects, or at least their approximate truth, or some slightly more fallibilistic notion than truth simplicity. When we talk about scientific realism, we're making a kind of realist commitment to something that science tells us about. So we're making a realist commitment to the content of scientific theories, or the objects that the scientific theories are about, to the truth or reality of those theories or entities, or at least some subset thereof. Now in our discussions of scientific realism, there are kind of three major dimensions. On the one hand, there is a kind of metaphysical commitment in scientific realism, a commitment to the existence of a mind-independent reality that science purports to be about. The second major dimension, or claim of scientific realism, or type of scientific realism, is a semantic dimension. This is a claim about the interpretation of scientific theories or scientific claims. A scientific realist is committed to interpreting those literally as claims about objects in the same way that our everyday discourse is to be interpreted literally as making claims about the world around us. And then the epistemological dimension of scientific realism tells us something about the record of scientific achievement. Should we believe that it is accurately approximating the truth about the world? The epistemological dimension says that science is actually delivering us knowledge above and beyond what sort of everyday knowledge tells us about. So we think about these three dimensions, and the denial of any one of these dimensions then gives us a type of anti-realism. So the denial of the metaphysical commitment of realism is a denial of a mind-independent reality. So one way of denying that there's a mind-independent reality is to be a certain kind of idealist. So you might say, look, the only thing that exists is the mind, or everything that exists is dependent on the mind, or on some kind of pre-existing mental or conceptual structure. So various kinds of idealism are anti-realisms in this sense, scientific anti-realisms in this sense. Many contemporary anti-realists are not so much idealists, but they might be, for example, they might have a kind of Kantian view that tells us that whatever exists in and of itself is inaccessible to us. And the world of our experience and the world of our scientific practice is one that is kind of co-constituted by the mind, by our theories, our concepts, and so on. So that is a kind of denial of a mind-independent reality. The denial of the semantic form of realism involves the reinterpretation of theoretical claims as non-literal claims, not literally claims about unseen objects, or powers, or laws, or whatever, but rather as kind of compact descriptions of something else, or covert descriptions of something else. So there's a kind of empiricist version of this that tells us that scientific theoretical claims are just shorthand for more complicated empirical descriptions that can be reduced to claims about observable objects. What we call instrumentalism is the view that theoretical claims should be interpreted, not as literal claims, but only as instruments for prediction and control. So observational prediction and experimental control of the world, and there are other versions of that as well. And then finally, the denial of the epistemological claim involves a denial that the theoretical claims are true, or that the unobservable entities posited in those claims. So it's a denial or at least a withholding of belief about the truth of the theoretical claims, or the reality of the unobservable entities posited by science. And you can see that one could be a realist in one respect, but a anti-realist in another respect. For example, you could acknowledge the metaphysical thesis and say that there is a mind-independent reality. You can endorse the semantic thesis and say, look, science does make literal claims about the nature of the world, including claims about things that we cannot observe. But we're not confident that science has the power to deliver knowledge about those things, right? So we are agnostic about those claims. That would be an epistemological form of anti-realism. Actually, probably the most popular or well-regarded form of anti-realism in the contemporary scene, which is known as constructive empiricism, it does precisely that. It is metaphysically and semantically realist, but epistemologically anti-realist. So why should we not be realists? What are some of the reasons that we might consider for denying realism? I mean, you might think that realism is a kind of default position. Science is very sophisticated form of knowledge production, and we should be realists about what it tells us. I mean, what else are we going to believe, right? But there are some pretty significant arguments, some of which builds on what we've talked about in previous weeks that at least raise some doubts about realism, right? So one of these is the theory ladenness of observation. Remember that a number of thinkers, including Kuhn and Fireaban, have told us that the theories that we hold actually, to a large extent, influence our interpretation of observation, the way in which we translate our experience into descriptions of the world in such a way that you might think the current theories that we believe are infecting our data in a certain way. So there's a certain kind of potential circularity at work, right? So in other words, you might think that it is a powerful way in which science predicts and explains our observations that gives us a reason to be realist, but if those observations themselves are not kind of neutral arbiter or external criteria, but are themselves influenced by or created by the theories that we have come up with in science, it raises serious questions about whether that's a good ground for realist commitment. A similar issue is raised by what we've called the undetermination of theory by data, right? This is the idea that multiple theories might equally well explain a certain body of data, right? As far as we, as far as the history of science goes, we have some reason to think that there are multiple possible options for developing our knowledge. And even if we don't see undetermination that often in the history is a logical matter, we can imagine that different theories might explain that we have not considered might explain the data equally well, right? And if there are multiple logically incompatible theories for any set of data, again, that raises the question of whether the data is adequate to sort of justify belief in one particular theory. The kind of historical stories that Kuhn and Feierab and tell us in which there are radical revolutionary changes and incommensurability between different theoretical perspectives or paradigms also raises concerns about scientific realism. And again, similarly to theory ladenness and undetermination, the issue is, you know, can we find a neutral or sort of independent criterion for what makes a theory a good theory and what makes it worthy of belief? And if theories aren't really comparable, if progress is not, we can't sort of track progress over long scales of time, then it raises questions about whether we should be committed to a form of realism. One very powerful argument against realism, it comes from the history of science, and it is sometimes called the pessimistic induction or the pessimistic meta induction. And the basic idea here is, if we look at the history of science, we see a number of examples of very sort of well-established ideas, theories, claims, entities. In the scientific literature, they are sort of objects to a large degree of scientific consensus. They are powerful in helping us predict and explain the world around us. And yet, subsequent changes in science, sort of over a historical time, have led us to reject those theories, to deny those theoretical claims, to remove those entities from our lexicon or our sort of our understanding of the world, right? So if we look at chemistry, right, there's a period in which the concept of flogiston as a sort of chemical element responsible for combustion is a fairly well established part of a fairly sophisticated set of chemical theories. And yet, if we look at the sort of modern chemical revolution, there is no place for any entity like flogiston, right? If we think about changes in the history of physics from Aristotle to Newton or better, from Newton to Einstein and from Newton to quantum mechanics, we see major revisions to our understanding of the basic sort of constituents of the world and the basic laws of nature. This suggests that we should have a pretty skeptical attitude towards the commitments of our current science, which after all, in the fullness of history might be equally well rejected in time, right? And replaced by something better. So if those are the arguments against scientific realism, what kind of arguments can we muster in favor of realism? And I think there are still a number of sophisticated arguments in its favor. So one of them is sort of the common sort of pairing with the pessimistic induction. And that's the no miracles argument or sometimes simply the miracles argument. And the basic claim here is if we look at much of the science we have today, if we look at our most sort of mature and successful sciences, they're very successful. The high degree of prediction and control, very convincing explanations of various elements of the world around us. It would be a miracle if all of that success happened. And yet this sort of underlying sort of a structure or explanatory system that makes them those successes possible weren't true, right? To make that a little more concrete, we have a high degree of predictive precision and accuracy in our contemporary physics. It would be a miracle that we would be so consistently successful if the physical laws and the kind of physical entities that are posited by those theories were not real, right? That would be a really surprising thing. Another reason that we might be realists, at least about certain kinds of very successful science is that we have what is sometimes called conciliant evidence or what is sometimes called robust evidence or what we might call multiple corroborations, right? And the idea here is, okay, we have, suppose we have one technique for gathering evidence on a certain kind of phenomenon. We can see it in a scanning electron microscope, but not with any kind of ordinary observational techniques, but we can detect it with a scanning electron microscope. Now, suppose there is a second kind of observational device works on completely different physical principles using completely different physical mechanisms, and yet it also allows us to detect the same phenomenon. Multiple lines of evidence that are mutually independent from one another, the idea goes, give us a good reason to believe that the thing that they're both sort of providing evidence for is real, right? This points not to just the success of sort of one line of evidence, the strength of one line of evidence, but actually the sort of a process of triangulation from very different directions. This gives us reasons to be realist because the best explanation for both of these techniques working is that the thing is really there, right? That's the idea. Of course, for this kind of argument to go through, you have to be sure that your different lines of evidence really are independent, and that can be an issue. In response to arguments like the pessimistic induction, some scientific realists have become more selective about their realist commitment. They've agreed that you do see components of successful theories over time being rejected and replaced with very different things, but they want to still insist that there is something that those older theories are getting right, something that is preserved in new theories, and so the goal of these realists is to identify what is the thing that is preserved across theory change, right? Or across paradigm shifts, you might say. One kind of view is that for the most part entities don't, the sort of entities that the science is about don't change even though our theories about those entities do change, right? So once we've discovered the existence of atoms, our theories about the nature and constitution of atoms changes a lot over time, but our commitment to atoms doesn't change. Our commitment to electrons doesn't change even though we understand them very differently than the scientists who first posited the existence of electrons did. And so that's the idea of entity realism, a kind of selective realism focused on entities. Another kind of selective realism is called structural realism, and so it emphasizes not the entities that the science is about, but the mathematical structures that make the predictions work, right? For example, they might trace the existence of fairly similar sort of mathematical equations or mathematical structures through changes in different stages of electromagnetic theory and say, look, from one theory to the other we change a lot of our basic understanding of what's going on, but the mathematical structure is preserved, right? So it's really, it's really structure that science is, we should be realists about in science. Another kind of realism, which we might call a selective realism is Ian Hacking's experimental realism, right? And experimental realism can be summed up with the bumper sticker slogan. If you can spray it, it's real, right? Hacking's idea is that we may not be convinced to have a realist commitment to something that we have somehow detected. We've built a detector that gives us a positive indication that that thing exists, that that we made sort of for theory ladenness and under determination and other kinds of reasons be skeptical about those kind of detection claims. But what Hacking says, we can't really be skeptical about is when we then use those entities to affect something else, right? The electron is again a good example for Hacking, right? So there's an earlier stage of science where we sort of indirect, there's a stage at which we posit the existence of electrons but have no means to detect them as a separate entity. Here he thinks it's merely, it's sort of merely speculative explanation. Then there's a point where we build detectors and we think we have detected the existence of electrons and we can even measure their charge. But still, we might not be ready to be realists about electrons. But once we can do things like build a scanning electron microscope and use electrons to detect the features of other things or we can build a certain other kinds of devices that depend on the properties of electrons to alter other things, then we should be committed to the reality of electrons. That's experimental realism. So those are some basic arguments and ideas about scientific realism and anti-realism. I look forward to discussing the way that those ideas play out in the readings by Godfrey Smith, Loudon, and Hacking that we did for today, for our class this week. I look forward to hearing what you think here in the comments or on Discord or in class. So otherwise, I will see you next week. Have a great week!