Transcript of Edouard Machery on 'Experimental Philosophy'
Thomas Spiteri: Welcome to the HPS podcast, a podcast where we chat all things history, philosophy, and social studies of science. I'm your host today, Thomas Spiteri, and I'm thrilled to have with us Edouard Machery, a distinguished voice in the philosophy of mind, psychology, and cognitive science, and a leading figure in the experimental philosophy movement, which seeks to integrate empirical methods into philosophical inquiry – a focus of our conversation today.
Edouard is a distinguished professor at the University of Pittsburgh's History and Philosophy of Science department, and the director of The Center for Philosophy of Science. His work blends empirical research with philosophical investigation, offering new insights into how we understand concepts, human cognition, and the role of diversity and thought.
Edouard has published over 150 articles and chapters on these topics, so it is a privilege to be sitting down today with Edouard for a chat.
Edouard, thank you so much for joining us today.
Edouard Machery: Thank you very much for having me today. It's a wonderful opportunity to talk to you and your audience.
Thomas Spiteri: Edouard, you've made significant contributions to several fields, including philosophy of mind, cognitive science, psychology, epistemology, the list goes on and you've been a key figure in developing experimental philosophy, which is today's topic. But before we get into that, could you start by telling us about your journey into the philosophy of science and about the development of your interests?
Edouard Machery: So, as you might hear, and as your audience might hear as well, I'm French, so I was born in France, and I was educated in France. In fact, my PhD is from the Sorbonne, not from the US or the UK or Australia. In France, most people do mostly history of philosophy. So, as an undergraduate, I mostly learned history of philosophy. I really got tired of philosophy as history of philosophy, but I was lucky to meet, when I was a young man, the few philosophers in Paris, the few analytic philosophers, who were working either in the philosophy of language or in the philosophy of mind, with a very naturalistic bent. The philosophers of mind were really working closely with psychologists and that's how I got into the philosophy of science.
I wrote my dissertation on the notion of ‘concept’, which was a topic in the foundations of cognitive science back in the late 1990s, early 2000s. I was lucky that the field was already a bit saturated. Most people working on concepts were coming from the philosophy of mind, and I didn't want to write yet another epicycle on Jerry Fodor's views about concept or Chris Peacocke's view about concept. So, I looked much more closely at the psychology of concepts.
My dissertation did happen to be a dissertation in the philosophy of the special sciences, on the psychology of concepts, rather than philosophy of mind, which was a dissertation on philosophy of cognitive science. That's how I started as a philosopher of science, more as a philosopher of a special science, namely cognitive science. I was very lucky to get the job at Pitt right after my PhD, in the Department of History and Philosophy of Science. Being at Pitt, surrounded by some of the best philosophers of science and historians of science, really led me to broaden my interest toward philosophy of science. I wrote a bit in the philosophy of biology; I wrote about life; I wrote about human nature; I wrote a bit at the intersection of psychology and biology; evolutionary psychology, and these kind of questions. Of course, I got more and more interested in general issues of the field of science. These days, I'm working on a book on trust in science. It's really something which is about general questions in the field of science rather than the field of psychology more narrowly.
So, it's been a process of expansion, since I started as a graduate student. I think it's interesting how the environment has been shifting my interests. Being at the right place at the right time, having the right interlocutors, has really broadened the range of questions I've been working on over the last 30 years.
Thomas Spiteri: So today we're going to chat about experimental philosophy or "X-Phi" as it's been called. Can you explain to us what experimental philosophy is and how it differs from traditional approaches to dealing with philosophical issues?
Edouard Machery: Yes, with pleasure. I think it's useful to distinguish a narrow and a large understanding of what X- Phi – and I'll use "X- Phi" too – is.
So, the broad understanding is any use of any empirical method that bears, in some way, on a traditional philosophical question. Now, if you use that characterisation, this is very, very broad. You might think, for example, that work in HPS, or history and philosophy of science, is experimental philosophy, because you're going to use historical work, which is an empirical method, to support claims in philosophy of science. People who use, for example, the analysis of text using quantitative methods, are doing experimental philosophy by this definition, provided that their work bears on philosophical questions.
This is a very broad definition, and I'm very partial to the broad definition because I really want to suggest that the insight behind experimental philosophy is a need to broaden the tools we use to answer philosophical questions. So, I really like this broad way of thinking about experimental philosophy.
But it's also the case, and sometimes in the debates about experimental philosophy, we see a more narrow definition or characterisation of X-Phi at play. In this narrow characterisation, the issue is mostly the use of what philosophers call intuitions, and the study of intuitions mostly by experimental means when intuitions are relevant to a philosophical question. And that's a much narrower definition. In many ways, it limits the scope of experimental philosophy to the study of intuitions, and it limits the range of tools that are used. Instead of [including] any empirical methods – quantitative methods, qualitative methods, the use of text analytic methods – it's really limited to experimental methods, often drawn from social psychology or psychology more broadly.
From a historical point of view, this narrow definition was very influential. It's been debated a lot. It's been at the centre of the controversies in philosophy. But I do think we should actually embrace a broad definition, because I think it really captures the core insight behind what experimental philosophers are doing.
Thomas Spiteri: In much of traditional philosophy, intuitions have been treated as reliable guides to truth, forming the basis for many key arguments and theories. An obvious example here is the role intuitions play in Gettier cases, in reshaping certain theories of knowledge. However, you've been quite critical of this reliance, suggesting that intuitions are prone to variation, cultural bias, and cognitive errors.
How can, if at all, experimental philosophy help address some of the issues with intuitions?
Edouard Machery: Good. Let me, for your audience, explain a bit about what an intuition is and give an example.
So, the word intuition is itself controversial in philosophy; there's disagreement about how to understand intuitions. But roughly speaking, in this part of philosophy, an intuition is a judgment or a reaction to a thought experiment. And a thought experiment is usually a description of a fictional situation that is of philosophical relevance.
You mentioned the Gettier case. It is one of the most famous thought experiments. It goes roughly as follows - this is one version of the Gettier case - just imagine that John goes to the train station, John looks at the clock, the clock says 3 p. m, and so John comes to believe on the basis of what he sees that it's 3 p.m. And it happens to be 3 p. m. So, John forms a true belief. He said it's 3 p. m., what he said is true, his belief is true.
But it turns out that the clock is broken. It always says 3 p. m. So, if John had seen the clock 5 minutes earlier, he would have had a false belief. But because he said it by luck, at 3 p. m. exactly, he formed, by chance, a true belief.
So that's the thought experiment, the description of a fictional situation. The intuition is the reaction that you're forming in response to the thought experiment. Usually, the common intuition among philosophers is that while John has a true belief that it's 3 p. m., John doesn't know that it's 3 p. m. The conclusion that philosophers have drawn from Gettier cases is that knowledge requires more than having a true belief that has some justification. It requires a true belief, justification, and something else. Because in that situation, John has a true belief, he's got good reason for his belief, but still does not know that it's real.
All right, that's an example, that's a good example of a thought experiment and of an intuition. It's a judgment we make about knowledge and responses to it. Now, why have I been concerned with the use of intuitions in philosophy? I think it's fair to say that in many areas of philosophy, not maybe in philosophy of science, or less in philosophy of science these days, intuitions have been central to the growth of the field. Epistemology, the study of knowledge, is a wonderful example. But also, metaphysics, also ethics, and so on and so forth.
Philosophers when they use their intuitions in reaction to a thought experiment, are just using their intuitions in reaction to a thought experiment. At best, they're knocking on the doors of their colleagues in the philosophy department and asking, "Hey, here's a new thought experiment. What do you think?" But what they don't do is ask other people beyond their colleagues or a very narrow group of individuals.
One concern that got me started in experimental philosophy is the concern that this intuition that philosophers have might be very parochial. They might be the product either of their education – after all, philosophers have a very specific education – or they might be the outcome or the exposure to philosophical theories, or they might be as a result of their cultural background. After all, many philosophers in that tradition are Western; they've been educated in the West. There was a question about whether the judgment they make in response to such experiments, say, intuitions, would be universal.
If they're not universal, then there's a question about why we should give so much importance to philosopher’s reactions to thought experiments. Just to give an example, if it turned out that Western epistemologists agree that in the Gettier case we described a few minutes ago, John does not know that it's 3 p. m, but if people in other cultures – maybe in China, as we've done some work there, or maybe in a small scale society in the middle of the Amazon – people actually have a very different understanding of knowledge and say, 'Well, of course, John knows that it's 3 p. m.', then there's a question about whose intuition should we really follow to do the theory of knowledge?
This is the kind of question that really got me started. First, just out of curiosity, and then it was out of concern as reliance on the type of intuitions that philosophers happen to be relying on. So that has been the original concern for me.
In addition to this concern, other philosophers, including myself, we've been looking at possible biases that might influence these intuitions. So, for example, if you give two cases, but you change the order of cases, does that impact your intuitions? In principle, it should not. It does not really matter whether you start with case A and then go to case B, or with case B and then go to case A. You should have the same intuition. But if it does matter, if the order of cases has an influence on the intuition you are expressing or adding, then one should worry that these intuitions are not fully rational. They are influenced by factors that should not matter.
We've been examining, using empirical tools, whether these kinds of biases do actually influence intuition. We found that there is actually quite a bit of variation, not for all thought experiments, not for all intuitions, but for many thought experiments and we've also found that there are many biases that influence people's intuitions.
Now you ask the question, how can experimental philosophy help? I'm not sure I'm the best person to answer that question, because my view is that we should actually avoid relying on intuitions. So, I tend to be what experimental philosophers call a restrictionist. I want to restrict the use of intuition. Some of my colleagues have a different view on the matter. They think we can use experimental philosophy more positively, not just to criticize the use of intuition, but to try to identify intuitions that are better, in some sense, that don't vary, or that are not influenced by various biases.
I think this is possible sometimes. I don't think we've been extremely successful at that; I think we've been mostly successful at documenting the challenges that the use of intuition faces in philosophy.
Thomas Spiteri: Can you maybe share with us an example where experimental philosophy has reshaped a traditional philosophical problem, or has reshaped the philosophical problem for science?
Edouard Machery: Yes, I'd be happy to give you a couple of examples.
One, which is drawn from my recent work, where it hasn't yet shaped the way we do philosophy, but I hope it will. Also, it bears on philosophy of science, which is also of relevance to your podcast.
So, my next book, once I'm done with the book on trust and science I'm finishing right now, is probably going to be about the notion of neural representations. So, what are representations in the brain? This will be in the HPS tradition, where I hope to look at the history of the notion of a neural representation and how it bears on philosophical questions.
Now, I got into this topic three years ago. I've always been interested in the philosophy of representation, but my interest was rekindled a few years ago when I had this empirical project on representation. What I wanted to know was, what do scientists mean, particularly neuroscientists and psychologists, when they use the word "representation"? Here, the goal was to try to understand the meaning of a scientific notion that is used throughout the neurosciences.
What I did with a philosopher at Indiana University, in HPS as well, was design four different surveys that describe realistic neuroscientific experiments and asked participants how they would describe the results of this experiment. Whether they would say, for example, that the brain activity that is elicited in this experiment is about something in the world; or whether they would say it's caused by something in the world. So, we had a few ways of describing the reaction of the brain to stimuli. Participants saw a long vignette that described a realistic neuroscientific experiment, and then they were asked how they would use various types of words, including the word "representation" and various ways of describing brain activity. The project was to get a sense of how the word representation is used by neuroscience.
What we found is that the word “representation” in neuroscience does not have a very clear sense. By which we mean, that neuroscientists don't really know how to use that word; they were very ambivalent and uncertain about whether it was okay to describe the brain activity by means of this concept. By contrast, when they say "causing" or "processing", they were totally happy to say that the brain is processing some kind of stimuli. However, when we ask, 'Is the brain representing the stimuli?', they were very uncertain about what to say. They were also not sure about which scale in the brain representations were to be found, it did not really matter whether the brain state had function in the brain, and so on and so forth.
So, what we found is a great level of uncertainty in the concept of representation. That has led me to ask the following questions. It starts with a very precise question, here's an important concept in neuroscience, what does it mean? So, the way traditional philosophers would have answered that is to develop a theory of representation. Nick Shea, for example, has a wonderful book on representation. Jerry Fodor back in the 1990s, Ruth Millikan has written about that. So, they have a theory of representation.
My take is very different; my method is very different. Instead of developing a theory of representation, I first want to know how neuroscientists understand the concept of representation, how they're using that word. Now, I found that this provided evidence that they don't really understand, they don't have a very clear notion of representation. That leads me to ask the question: but why do they use that notion? What is the concept? What is the function of the concept of representation?
It's used all over the place, but it does not have a very precise meaning. So, maybe it should be eliminated? But maybe not. Maybe the right reaction is to say, 'Not all scientific concepts have to be clear. Sometimes unclear scientific concepts are very useful.'
The question of course is what's the utility here of the concept of representation?
I hope to be changing the conversation about neural representations. Instead of having a theory of representation, like nearly all philosophers of mind and of cognitive science before me, I want to have a theory of the concept of neural representation. And, I want to do that by a very historical manner to understand why the concept has been developed in neuroscience, why it has come to play the role that it has played.
It's an example. You asked me, how does it change by philosophical practice? It changes because it has allowed me to redefine the philosophical question, move away from a theory of representation to a theory of the concept of representation. So that's one nice example of something that might happen.
An example of something that has happened, a very small one, is in epistemology. There are these cases called “fake barn cases”. The fake barn cases are also about knowledge and roughly go as follows: just imagine you are driving in the countryside, and unbeknownst to [you or] John, there was a tornado in that countryside, and the tornado destroyed all the barns. The farmers didn't want to rebuild the barns, it was too much work, too expensive, so they built fake barn facades. There are just the fronts of the barns that are standing and there's nothing behind the barn. Except for one barn that has not been destroyed. So, all the facades that you see, all the fronts, are not real barns, they're just the fronts, but there is one that is really intact. When you or John happen to be driving in front of this barn, you say, 'Oh, this is really a beautiful red barn'. Now you form a true judgment because that's the only one that's a true red barn. You have good reason to form that judgment, you're seeing a true red barn, but of course, if you had said the same thing about any of the other facades, you would have formed a false belief. So, you can only get it right by luck. It's a bit like a Gettier case that we saw a bit earlier.
Now, until 10 years ago, the majority of epistemologists, the people working on knowledge in philosophy, thought that it was like a Gettier case, and that the intuition was that no, of course, John doesn't know that it's a red barn, because he got it right by luck, right? But we did some work about 10 years ago, with David Colaço as the first author, one of my former students, and what we found is that most people, when you give them a fake barn case roughly similar to the one I've described to you, actually are ascribing knowledge. In that kind of situation, they say 'Of course, the person knows. The person, they see a red barn, their eyes are functioning very well, and they form a true belief. What more would you want to have knowledge?'
So that is a place where the dominant intuition in philosophy and the dominant intuition among lay people are quite different from one another. And, I think our paper has changed the discussion in that respect. I think it has led more and more epistemologists to rethink a little bit the claim that fake barn cases are like Gettier cases, and that the right thing to say in response to a fake barn case is, 'There's no knowledge there.'
When you realise that really everyone outside your normal circle seems to have a different view, you might be lead maybe to rethink the conclusion or the claim you had made about the case.
I think that's what has been happening in epistemology, so that's another example.
Thomas Spiteri: Experimental philosophy seems to overlap with some methods from social psychology. On the relation between those two fields, people have said that psychology tends to focus on how people actually think and behave, while philosophy tends to deal with how people should think and behave. How does experimental philosophy bridge this gap between descriptive and normative approaches? In other words, how does X-Phi integrate the "is" of psychology with the "ought" of philosophy.
Edouard Machery: Yes, this is a very good question. A very difficult one too. I want to say a few things there. The first thing I want to say is, I don't think philosophy is just in the business of answering normative questions. It is part of what philosophy is, after all, normative ethics, as its name suggests, is in the business of telling us how we should act or what's a good life.
But I think there's more to philosophy than just normative questions. I think part of philosophy is straightforward and descriptive. If you work in the conceptual foundations of some science, your work is going to be, I think, descriptive. If you try to understand scientific theories, as many philosophers of science do, your work is going to be in part descriptive. Okay, this is the first answer. I think philosophy is just broader than the study of normative questions, but it is also partly normative, right? I think it is true that philosophers are not just in the business of describing things.
I think here's the right answer: while you can never derive an ought from an is – so you can't straightforwardly conclude that something ought to be the case just because it is the case - it is also the case that answers to normative questions (so “ought” answers) are always informed by claims about what really is.
In fact, it's very easy to find in the history of philosophy. When Aristotle theorises about a good life, his work is surely influenced by his empirical beliefs about the world. Even in Kant himself. Kant of course writes about what is a moral judgment and how to recognize what makes somebody right or wrong, to give us a categorical imperative. But he also writes about the good life, and when he writes about the good life, he pays attention to questions of fact, you know, questions in anthropology. Hume, of course, constantly blends claims about what we ought to do from his objective claims.
So, I think again and again, while we cannot derive simply an “ought” from an “is”, we can't really have detailed views about what we ought to do or what or how we ought to think, and so on and so forth, without paying attention to the way we are.
I do think people have been too impressed in many ways by the maxim: “is does not imply ought”, which is true, but it's also true that “there's no ought without an is.”
One way people have come there, but it's only one way, is to notice that if you can't do something, which is an empirical claim, then you ought not to do it. Your capacities provide in some sense a limitation to what you ought to do. So that's one way to connect the descriptive and normative. But there are also other ways.
I think, more broadly, normative views are always, more or less, obviously influenced by empirical commitments. So, when one changes one's empirical commitments, one is bound to change one's normative commitments. That's why I tend to actually really reject the idea that we can easily disentangle the descriptive and the empirical from the normative.
Thomas Spiteri: Looking ahead, what new directions in experimental philosophy are you most excited by?
Edouard Machery: So experimental philosophy has grown a lot and in ways that are both exciting but a little bit worrying.
Most experimental philosophy has become really quite technical. Experimental philosophers now publish mostly in scientific journals. Sometimes excellent scientific journals like PNAS and Nature, Nature of Human Behaviour. I think the worry is that it's disconnecting experimental philosophy from philosophy. I actually am slightly worried about the future where experimental philosophers turn themselves into scientists, and kind of forget that experimental philosophy is meant to bear on philosophical questions, and not just on interesting questions. So, that's a worry. You asked me what excites me most, I wanted to mention what worries me most.
What excites me most is the diversification of the methods we are using. We used to be mostly drawing on social media, psychology, whereas nowadays we are using all possible methods. We are using quantitative analysis of texts, we're using corpus analysis, we're using sometimes ethnographic studies. It has become a really diverse and exciting field.
I think it's very good for philosophers, so they can see, 'Oh, here's an interesting tool that they could use for that specific project'. So that's really what excites me about experimental philosophy.
Thomas Spiteri: Edouard, it's been a fascinating conversation today, thank you so much. To finish on a slightly different note, you recently set up your own podcast, Conversations at the Center. Could you tell us a bit about it and where we might find it?
Edouard Machery: Thank you for putting a plug for that new podcast. We only have had one year on that podcast. It started from the Center for Philosophy of Science that, as you mentioned earlier, I direct.
The reason is that we've got so many great guests at the Center, wonderful philosophers of science, and I wanted to take advantage of their skills and expertise and friendliness to talk about issues in the philosophy of science of interest. Not only to specialists but also, just like your podcast, to the broader public.
Part of the goal is indeed to bring the philosophy of science, just as you've been successfully doing, to the larger public. I think we've got a lot of things to tell people and it's very unfortunate that sometimes our audience might be limited to our colleagues. The podcast can be found on YouTube for now and also on Spotify.
Thomas Spiteri: Thank you so much for your time today and for helping us reconsider the ways in which we go about generating knowledge.
Edouard Machery: Thank you very much for having me today.
Comments