Naturalism in Epistemology

I want to talk about naturalism in epistemology. Naturalism has a metaphysical aspect that I do not wish to address here. My effort is to be an empiricist, but one cannot be an empiricist. Many people ask me, “Are you an empiricist?” In response, I say that I am an empiricist in the same sense that Saint Paul said one can be a Christian, meaning that we strive to be Christian. Empiricism is an old, traditional philosophy. But it changes; it had misfortunes in the twentieth century, and I hope we have learned from these mistakes and tried to adopt an empiricist view that is defensible today. Here, I aim to draw attention to the distinction between empiricism and naturalism in epistemology. First, I will try to identify the common ground between these two philosophical traditions, then point to where their paths diverge and their differences. We will see the disagreement between empiricists and naturalists.

When I say I want to be an empiricist and see what empiricism can mean today, I must first point out that empiricism is not a theory or a thesis but a philosophical stance. By stance I mean a set of dispositions, methods, commitments, and beliefs that a person holds. Empiricism is a form of philosophizing. Naturalism has been presented in exactly the same way. In the 1940s, Columbia University in New York City was a center for naturalistic studies. At that time, naturalism was considered a new movement in American philosophy. The book Naturalism and the Human Spirit, which was the manifesto of that movement, was also published at that time. Some of the individuals who contributed to writing that book are no longer known to us, but some we remember well: Dewey, Hook, Nagel, and Randall. Each of these four philosophers emphasized that when naturalism is spoken of, it is not a theory about the description and nature of the world. By naturalism they did not mean a thesis, doctrine, or dogma. Rather, they referred to it as a philosophical approach, a method by which one ought to philosophize. Sidney Hook says naturalism is a commitment to a process, not a metaphysical theory. John Randall says naturalism is a temper and a disposition. These answers differ greatly from conventional philosophical positions. If I ask you about your position in philosophy, you are very likely to point to a specific theory. For example, in Platonism it is believed that abstract categories are real, whereas in nominalism the existence of universal realities is denied; or in the philosophy of science we have a position that believes in the reality of laws of nature, and at the opposite pole a position that says laws of nature do not exist. All of these are theories and make claims about reality and what exists. Naturalism was not put forward in this way because it is a philosophical approach or method. Penelope Maddy – professor at the University of California – is one of the main exponents of modern naturalism. Over the past fifteen years she has tried to explicate what it means to “be a naturalist.” She expresses exactly what was being said sixty years ago. She says naturalism is not a doctrine, but an approach; it is not a set of answers but a method for answering questions; she also emphasizes that naturalism is not a set of beliefs or a theory. I, too, point to these same matters when I want to say what empiricism really is. Although empiricism was put forward in the past as a factual claim with phrases like “experience is the only source of our knowledge of the world,” empiricism or naturalism should not be presented in this way; these are methods for philosophizing. A comparison between empiricism and naturalism reveals three points they have in common. First, both empiricism and naturalism are anti-metaphysical. But they do not critique metaphysics in the same way, and they also criticize different conceptions of metaphysics. Both approaches accord a central place to science. The third commonality is opposition to foundationalism. Empiricism in epistemology was associated with foundationalism – the idea that there is a firm foundation; a certain and fixed basis for knowledge and rational beliefs about how the world is. Eighteenth-century empiricism believed that our certain foundation is sense data. But in twentieth-century empiricism and naturalism it is accepted that no such foundation for knowledge or rational belief exists. We must learn to live without foundations. Let me speak about this a little to clarify the differences between my attitude as an empiricist and that of naturalists. The German empiricist philosopher Otto Neurath says that in confronting science we are like sailors at sea, sailors who are forced to rebuild their ship in the middle of the ocean and do not have the possibility of returning to shore. Quine repeatedly referred to this metaphor; Quine is someone who, in the mid-twentieth century, is spoken of as the father of naturalized epistemology. Quine explicates two directions for justifying the opposition to foundationalism. He also says that the Humean condition is also the human condition. This means that he accepts Hume’s critique of induction and consequently rejects any secure foundation and also denies any secure structure for justifying beliefs. In the twentieth century this was a major problem for empiricism. Naturalists in this era parasitically benefited from the ideas of empiricists.

Next I want to distinguish between two kinds of induction. One kind of induction is something we all do; in these cases we go beyond our evidence in what we believe. For example, Saint Augustine says I believe the ocean is real even though he has never seen the ocean; the existence of the ocean goes beyond the evidence available to him. In this sense we are always performing induction. Nevertheless, people who are superstitious also go beyond their evidence. Wise people do so as well. But there is another kind of induction that philosophers have always debated; it is in fact a rule that a wise person must follow when going beyond the evidence. This rule can be learned and taught. One can learn it and go beyond the evidence. It is also objective; that is, it does not involve your emotions and feelings, it is rationally binding—meaning if you do not follow it you have not acted rationally—and finally it provides you with a reliable method for arriving at the truth you are investigating. Now the question is whether such an induction is possible. When Hume answered this question, he lived in an era whose intellectual hero was Newton. Newton claimed that he had carried out his work on the basis of induction and therefore believed that such a method existed. Hume, in his theories, is criticizing this position. In the twentieth century, Reichenbach—I should say Reichenbach is one of my philosophical heroes—was developing the empiricism of his time. When he became president of the American Philosophical Association, his first address to the association was a critique of traditional rationalists and empiricists. Reichenbach wanted to be an empiricist, but he criticized eighteenth-century empiricism. He said he accepted Hume's critique to this extent: one can imagine that the world is designed in such a way that no matter what method you employ, you cannot access the truth, but you can preserve a conception of induction that works under certain conditions, when other rules are also operative. He developed a theory on this matter which, in brief, states that you can use probabilities that existed in the past to predict the future. If the available statistics show a strong relationship between smoking and lung cancer, it seems rational to say that on the basis of this statistical probability we make predictions about the future. The point is that if you state a rule this simple, it will be very trivial and unsatisfactory. Suppose my new job is weather forecasting and every day I must estimate the probability of snowfall the next day. I look at the snowfall statistics for the past fifteen years and see that it snowed on 5% of the days of the year. I want my predictions to turn out correct, so every day in my forecast report I say there is a 5% probability of snowfall; in the long run these results will be correct. But no one likes this kind of weather forecast; telling people in summer that the probability of snowfall is 5% and having them hear the same thing in winter. We want to say that whatever rule you follow, you encounter specific questions about specific conditions. For example, we want to estimate the probability of snowfall on the third day of the winter season. We can call questions of this kind challenges. Naturally, it is desirable for rules to answer challenges. The question of whether a rule can be found that answers all challenges has been examined by statisticians and philosophers. In fact, the first person to realize that Reichenbach's idea was wrong was his student Hilary Putnam. There is no rule that holds whenever any other rule holds. This was exactly contrary to Reichenbach's theory. Such an induction is an unattainable ideal. I have emphasized this point, just as Quine and empiricists must agree on this issue. Because there is no such thing as "the method of induction," and empiricists had been philosophizing on the basis of this fact for about two hundred years. This finding is part of anti-foundationalism; we must get rid of the idea that we can build our rational beliefs on a firm foundation. But it should be noted that this does not mean "anything is possible and acceptable." Just as we still have criteria for rationality, we maintain that these criteria are not built upon foundationalism. It is a common mistake to say that with anti-foundationalism everyone must become a skeptic.

On the contrary, we must say that if foundationalism is impossible, we are then entitled to set skepticism aside and choose a foundation. We must start from where we are and proceed rationally. In a life without foundations, evading the responsibility of choosing a foundation is not courage; it is the condition in which we humans live.

So far, I have pointed to points of agreement between naturalists and empiricists. Now I want to talk about their disagreements. When I say we must start from where we are, the question is: where do we start? Where exactly is our starting point? When Quine was writing about this in the 60s and 70s, it is quite clear that his starting point was what science had affirmed to exist. In his view, existing science is the only perspective that can provide the starting point. Through the natural sciences, however fallible and imprecise, I will investigate what exists. When Penelope Maddy refers to this approach of Quine's, she explains that this is the naturalistic response. There is a tension in this claim. When we look at naturalism as it was presented in the 1940s, we find that naturalism was not a thesis, a true proposition, or a belief, but an approach to questions. Yet we see that naturalism is currently presented as a belief, a belief with which they begin: science tells us what the world is like. I do not mean to say this is irrational, but it is very strange that what was an approach in philosophy has suddenly turned into a belief.

Penelope Maddy, in her book, refers to her ideal philosopher. When I read this book, I thought to myself that Maddy's ideal person does not seem like a philosopher at all, but resembles the naturalists' ideal. When she describes this ideal person, we find that they look at subjects outside of science only from psychological, sociological, and anthropological perspectives. Maddy's ideal philosopher, when encountering art, literature, or religion, is only interested in the psychological examination of the artist and their work, or the sociology of literature, or examining the psychological illnesses of the religious; in other words, they look at non-scientific subjects only in this way. This is Maddy's ideal philosopher. In my opinion, this ideal person is not a philosopher; one can imagine a person doing exactly these things without ever having studied philosophy or engaged in philosophizing. We must see what the reaction of Maddy's ideal philosopher is to the questions present in the philosophical tradition. I must say that Maddy reminds me of the heyday of logical positivism. Alfred Ayer was the person who brought positivism from Vienna and Prague to the English-speaking world. When he spoke about realms outside of science, he would say that if a sentence matters, it is scientifically examinable, and if it is not scientific, it does not matter. Maddy holds the same view regarding philosophical questions. Even Carnap, who was a member of the Vienna Circle, had extensive discussions about ontology, realism, induction, semantics, and so on, but in Maddy's view, these discussions are useless and meaningless. Maddy says that the second-order analyses conducted about science are not of interest to naturalists. In her view, philosophers are people who take a step back from science, look at it with a critical eye, and examine it. In her view, this level of analysis and explanation is beyond science and we must shut it down. We can ask ourselves, what remains of philosophy in Maddy's attitude? That is why I say I am reminded of Ayer; Ayer believed scientific questions are meaningful, anything else is cognitively unimportant and no effort should be made to understand it. When I try to diagnose the problem with this naturalistic approach, I think the main issue is the problem of interpretation. Philosophers deal with questions that are interpretive, for example, the question of what science is and what it does. How do we understand existing sciences? For the naturalist favored by Maddy, these questions cannot even be posed. The only appropriate word to describe this approach is fundamentalism. What makes a person a fundamentalist is not a particular set of beliefs; we have Marxist fundamentalists, Hindu fundamentalists, and Christian fundamentalists. The common point among all of them is that they have a problem with interpretation, and to any interpretive question you pose, they will tell you: what are you talking about? This is completely clear. Naturalism, too, says we are inside science, like a fish in water; what meaning could there be in questioning the meaning of science? What question can be asked about our relationship with science? This is what we believe; the world is this way. Science describes the world, and that's it! This resistance to interpretation is the core of the naturalistic attitude that Maddy represents.

In the philosophy of science there are two opposing approaches, and in Maddy’s view the debates of both groups are meaningless because one cannot ask what science is. The empiricist approach to science holds that science aims at success in empirical tests, and that this is the goal of science. Scientific realism believes that the criterion of success is truth; that is, everything in science, whether observable or not, can only be measured by truth. The confrontation between these two views was very heated in the 80s and 90s and still continues. How does Maddy react to this philosophical debate? She can say that insofar as this question is scientific it is meaningful, but if it is not a scientific question it will not be meaningful either. The empiricist says that in science, only the truth about what is observable matters, and the scientific realist says that the truth about what is unobservable matters. For example, atoms are unobservable. What Maddy does is to transform interpretive questions such as what science is, what the goal of science is, and what success in science is into a scientific question, and asks: Are atoms real? In her view, this is a completely specific question about a matter of fact. Whereas an empiricist would say the main question is the validity, status, and superiority of atomic theory. In fact, when confronting philosophical questions, Maddy poses the closest scientific question to that philosophical question and answers it. My question, however, is this: Are naturalists committed only to science, and do they have no commitment to the philosophy of science and their own philosophical presuppositions, which they do not articulate or of which they are unaware? If I want to say that a person is not only concerned with facts but also engages in interpretation, I must offer an alternative interpretation to describe the facts. In this way, my interpretation stands against the naturalist’s interpretation; both of these are interpretations, and rivals to one another. An example that Penelope Maddy examines several times is the empirical tests carried out at the beginning of the twentieth century that prove atoms and molecules are real. This story is always presented by scientific realists in the following way: In the nineteenth century, several physical models were developing; one was atomic theory or kinetic theory (the theory that states bodies and gases are composed of small moving particles), and another was a theory that did not accept the existence of atoms and intended to proceed solely by means of force and energy. These two theories were rivals. This story is presented by realists in two parts. The first part is that philosophical prejudice against unobservable phenomena prevented the victory of atomic theory; this was the negative influence of philosophers on science, as they forbade scientists from considering unobservable phenomena. The second part of the story is that in the early twentieth century, all scientists agreed that atomic theory was empirically acceptable, but they could not believe in its truth because they could not see and determine some of the properties of molecules, and therefore they were unable to declare that it was true. Nevertheless, in 1905 and 1908, Jean Baptiste Perrin, in a series of empirical tests, clearly showed that the explanation of the empirical success of these theories is proof that molecules and atoms are real. That was the realists’ narrative of this affair. Let me explain a little about these tests. At the beginning of the nineteenth century, Robert Brown observed dust particles in liquids in his experiments. When you look at dust particles in sunlight, you can discern the direction of their movement and, for example, explain this movement by saying that one part of the room is warm and another part is not. Also, if there is some dust in a bottle of water and you see their movement, this movement could have different causes. Brown conducted several tests and examined an isolated container of water. The particles in the water continued to move rapidly. Brown said the reason these particles are moving in the water is that there are molecules in the water that are striking the particles in the water and causing the particles in the water to move. Well, now let us pose a few questions.

First, let us examine the scientific realist narrative of this event. Was a test really carried out to demonstrate the reality of molecules? It does not appear so. In 1910, Perrin noted that the existence of molecules and atoms was accepted; therefore, Perrin was not seeking to prove the existence of molecules and atoms with his experimental tests. Nor is it the case that scientists at the time thought the theory was empirically successful but were unsure of its truth. Was the resistance to atomic theory really philosophical prejudice? The rivals of atomic theory also spoke of unobservable categories, so the possibility that they objected to atomic theory for this reason is very slim. This historical narrative of science presented by scientific realists has many problems, a few of which I have mentioned. Now the question is, what is my interpretation of what Jean Baptiste Perrin saw? I want to say that what they recount about Perrin is not history, but their interpretation of historical events, and they themselves have become engaged in philosophical debate. I must offer an alternative and say that this alternative interpretation is a rival to your interpretation.

I look at this episode in an empiricist way. Perrin was doing something very important. He was examining the elementary requirements that any scientific theory must satisfy and dealing with empirical success. This point is far more complex than correct predictions. We know well that correct prediction is not sufficient for the empirical success of scientific theories. A subtle point lies here. To clarify this point, I will first recount an old example, then note that the same thing occurred in the Perrin episode. The old example is this: in the seventeenth century, before Newton, Descartes proposed a theory in physics. Descartes had many followers, and for about a hundred years after the theory was proposed, Cartesian physicists were numerous in Europe. The important point in Cartesian physics was that only categories that are directly measurable were recognized. This point stemmed from Descartes' metaphysics; Descartes believed that matter is extended, therefore matter is spread out in space and time and can be measured. Thus, length and duration can be measured directly. Cartesian physics was built upon kinematic quantities. Cartesian physics had problems—every scientific theory has problems that must be resolved, so this was not a problem in itself. But Newton took a stand against Cartesian physics and introduced two quantities that are not directly measurable: force and mass. Cartesian physicists criticized Newton for destroying science and wanting to bring back the mysterious qualities of the Middle Ages to science; they criticized Newton for assuming non-empirical categories in physics that were irrelevant to physics. What was the response of Newton and the Newtonians? They said you are mistaken; mass and force are measurable, not directly, but they are still measurable. George Atwood was one of the individuals who participated in these debates; he invented a machine to demonstrate the effects of Newton's first law of motion. But this did not convince the Cartesians, because the Newtonians measured mass based on Newton's theory, and they did not accept Newton's theory. However, the Cartesians' stance was wrong. Because what was done is precisely what a theory must do; it must be shown that the quantities posited in the theory deal with measurement in the way the theory claims. If a theory lacks this ability, it is empirically unsuccessful and incomplete. Ernst Mach also examined this debate; he explained that George Atwood carried out the measurements according to Newton's theory, and in "science," measurement is done in this way; measurement must be carried out within a theory, and raising this issue cannot be considered a criticism. This demonstrates the subtle criteria of empirical success that must be applied.

Now let us return to Perrin. What did Perrin do? He did the same thing. Perrin saw that at the level of the gases and liquids he dealt with, atomic theory worked well. Perrin says, like the physicist before me, I add one more thing to this theory. The behavior of molecules, which we cannot see, is exactly like the behavior of dust particles that we can see. He was aware that this was an assumption he was adding to the theory. With the addition of this hypothesis, he could carry out measurement, and this was his success. This is a different interpretation of what Perrin did, an empiricist interpretation of what happens in science. Scientists are cultivating empirically successful theories; scientists are not seeking to find the truth behind what we can see and touch. Finally, let me state my own critique of naturalism. Naturalists are offering a metaphysical interpretation of science and adding a metaphysical narrative to science that is not present in it. Whereas for empiricists, the issue is one of interpretation, an issue that the naturalist does not even allow to arise.

.

---

.

Author: Bas van Fraassen

Translator: Zohair Bagheri Noaparast

.

---

.