Understanding in Science and Elsewhere

Interview by Ariane Tanner.


Interview with Catherine Z. Elgin 

What beliefs should we accept? How can it be reasonable that scientists use models and idealizations they know not to be true? And how do these questions bear on the epistemology of science and art? 

Catherine Z. Elgin, Professor of Philosophy of Education at Harvard University, started to investigate these far-reaching issues in a cooperation with Nelson Goodman more than thirty years ago. Since then she has developed an inventive and radical philosophical approach, which goes right to the core of epistemology.

The historian of science Ariane Tanner interviewed Catherine Z. Elgin two times in 2005 and 2018. Elgin talks on philosophy, her intellectual biography, how a woman became a philosopher and the new book “True Enough”(MIT Press, 2017).

Ariane Tanner: What is the place of “truth” in philosophy?

Catherine Z. Elgin: My view of truth is: it’s not nothing, but it’s not very important.Weneed truth, because we need logic. We have to know which inferences are truth-preserving. Sentences are true or false, and they have certain truth-related relations to other sentences. That’s correct. But mostly what we care about is not truth. We’re after interesting and important ways of characterizing things. Sometimes they’re true, sometimes they’re useful approximations, sometimes they’re metaphorically true but not literally true; sometimes a representation that is not sentential or propositional, such as a picture or a graph, gives you the information and the accuracy you need. The claim that we are after truth is really not true.

AT: So, we should have to accept that it’s all ‘work in progress’, and it will be ‘work in progress’forever.

CZE: That’s right! There are two questions: There’s a question about truth and there’s a question about known-truth. Even if we happened on the truth, because the universe is so vast, and humans are so fallible, we could never be confident that what we have found is the settled truth. We’d have to be open to the idea that maybe next week it’s going to turn out the little green man from outer space or the kids in Brooklyn or somebody does something that shows us that we weren’t quite right. We endorse fallibilism, assuming that this is a very good working hypothesis, that it is very highly confirmed, that we see very little reason to think we will ever utterly give up. But we might have to modify. That seems to me the responsible stance. And if you look at the history of science, you see that that really is what we ought to be saying. We can see that theories that have been very well confirmed later showed that they had to be modified, because something in the domain where you hadn’t thought to look or didn’t had the resources to look for, calls into question a hypothesis that worked well in a more limited domain.

We should accept only what we consider acceptable. If, for example, you’re going to do chemistry, one of the things you accept is the periodic table of the elements. You divide things into these units. That’s not a truth. You are buying into the correctness of using these categories to characterize the things you are going to be doing in chemistry. Other things would be: you accept standards, like statistical standards. How many data points do you have to have for your evidence to be statistically significant? That’s very important, but it’s not anything about truth. So there are more dimensions along which you want to ask about acceptance than just is it true or is it false.

AT: How does this hook up with the central role of reflective equilibrium in your thinking, especially as elaborated in your book “Considered Judgment” of 1996?

CZE: We start with what is accepted –initially tenable commitments are things we accept. That’s just a matter of psychology or sociology actually. For example, it could turn out that people think that asparagus is rich in vitamin C. Maybe they have learned it at their mothers’ knees, maybe they made it up, maybe came to them in a dream, maybe they heard it on television, anything: You start with the brute fact that people think these things. And then you notice that there are inconsistencies; they think things that disagree with this. Maybe they think that asparagus is full of vitamin C and also that only tropical fruits have vitamin C. They notice it can’t be both. So they’ve got to do something to bring these two discordant beliefs into accord. The idea is, you take what you initially believe or initially accept and you find that there are inconsistencies or incongruities that you have to resolve.

There are holes in your theory. If you’re going to have a theory of what vegetables have what vitamins, you need to do something about that; you integrate new data as well. You then try to bring these things into equilibrium, so that they are all reasonable in light of each other. Given that you believe these other things, it’s also reasonable to believe this thing, and it’s more reasonable to believe all of them together than to leave this other thing out. Then perhaps you’ve got a comprehensive theory about vegetables and vitamins. But of course a good fiction satisfies the equilibrium requirement. In a nineteenth century novel, all the parts are reasonable in light of each other. But you don’t think it’s true or approximately true or anything like that. You just think somebody made this up. So you also have to test your evolving theory against your initial beliefs, the things you started out with. Is the whole theory reasonable in light of them? Clearly it doesn’t incorporate all of them, but where you gave something up, it explains why it seemed reasonable until you gave it up. When the theory is internally in equilibrium and is reasonable when judged against your antecedent commitments, then you have good reason to accept it, and that’s what makes it acceptable. Roughly speaking you start with what you accept, and then do a lot of correcting and amending, and then you generate what is acceptable out of what you started out accepting.

AT: But, as I understand you, you don’t want to base everything on coherence either.

CZE: Yes, that’s right. Pure coherence is too easy. I keep citing 19thcentury novels, because they are incredibly coherent. And pretty much never you’re going to come up with a good scientific theory that beats “Barchester Towers” for coherence! You need some way of making sure that even the theories that contain devices that aren’t true are testable against the way the world is. This is sort of absurd, coming back to be sort of a closet verificationist, a total science verificationist. My God, I am a logical positivist! But the theories have to be testable. Suppose your are using a very simplified model of molecular bond. You’ve just got these little two things, kind of yoked together like that, and that’s all. If it’s not going to be idle, you have to see whether the theory that incorporates the model stands up in experimentation. Not the little fictional part itself, but the whole theory, has to be testable. And it has to be such that your test could fail, it could turn out to be overly simple. It could turn out to be emphasizing the wrong features or something like that, and you have to have some way that the evidence should show you that this model isn’t going to work.

AT: So, there is an interaction between what I internally accept and what I can know about the world. Isn’t there a tension between my opinions in equilibrium and what is intersubjectively accepted?

CZE:It looks like there is a problem if we start out as you suggest: “I’ve just got my own views and they’re all in equilibrium with each other, so that’s that." Among the things you have opinions about, are the facts that other people have opinions and that many of these opinions may disagree with yours. What should you make of that? If you believe that it’s raining in Boston, and I believe that it’s not raining in Boston, then both of those beliefs are initially tenable. They clash; we have to do something about that clash. What we would probably do is ask what your source of information is, what my source of information is. If it turns out we’re both just fantasizing then we throw them both out, because beliefs based on just fantasies don’t have any credibility. If it turns out that you were just listening to CNN and I just have the pessimistic view that is has been raining for a month and it probably still is, we would find that your grounds were better than mine. So we would discredit my belief, because you had a better grounded belief. We’re not isolated Cartesian egos, we interface with each others’opinions, and the equilibrium has to take that into account too. Now the important thing to worry about here is, you don’t want this to be a popularity contest. It can’t be that just because a lot more people think one thing than another, that makes it better.

A lot of people for example have pre-Newtonian beliefs about things like mass. It’s a very commonsensical view, so if you talk to ordinary folks, their views about mass and weight and so forth will be pre-Newtonian. That doesn’t mean that they’re going to prevail, because they don’t have the knowledge that the physicists have. When you start looking to their reasons, you see that their reasons are relatively weak. The physicists, although there are fewer of them, have much better grounded reasons, and in the end, the rest of us ought to be deferring to the physicists rather than saying, there is more of us then you, mass is pre-Newtonian. Among the things that we develop tenable belief about it, tenable commitments about, are the methods for assessing commitments. These improve with time. Evidence that once would have seemed to be okay no longer does. The idea that our methods improve over time by critical reflection on what they do and don’t do for us, that’s the kind of picture you get with reflective equilibrium. When we start noticing that there is another problem that we haven’t thought of, then we’ll do some more fine tuning.

AT: How can you integrate ethical, moral and social questions? How can we come to a minimal consensus?

CZE: Maybe you’re hoping for too much here. You have moral opinions, moral laws about human rights, or moral opinions about very mundane things like, it’s very rude to slam the door in somebody’s face. We have all of these, and we wonder how they fit together. We have people whose moral judgements we trust and people who we think are creeps. We’ll try and take these moral views and bring them into equilibrium with each other so that they seem reasonable in light of one another and in light of what we antecedently believed. This is going to involve some fine tuning. You might for example start out thinking “it’s always wrong to kill anybody”. Then you think about people at their end of life who are really suffering, and you can’t convince yourself it wouldn’t be better simply to pull the plug and let them die. You have to fine tune your principle “it’s always wrong”and say, no, there are some kinds of exceptions that I want to recognize. How can I mark out the class of exceptions so that even though it’s in general wrong to kill, I’m not giving this universal generalization anymore?

I do not claim that there’s a unique equilibrium. You could get something like what Rawls finds in political liberalism: a possibility of overlapping consensus. Maybe on a bunch of very important moral rules you can get everybody to agree, but you can’t get them to agree about why they are the moral rules.

AT: I have the impression that although you took the idea of reflective equilibrium from Goodman, you do not follow his footsteps in “Ways of Worldmaking”.

CZE: I find the talk about world making unhelpful. It’s not that it’s exactly false. I think there is actually a problem at the base of it, in the idea. Goodman says “there are many worlds if any”, and they’re made. I don’t think he can say that, because individuation is always within a world. So you can’t say there’re many worlds, because there is no perspective from which you could stand to say that. You have to count, and to count you’ve already got to be in a world that presupposes one way of individuating or another. So in a fact, what Goodman is trying to say, by his own standards he can’t say. One of the things Goodman did, (he did this purposely), is always put things in a most obnoxious way possible. Because he wanted people to recognize that it’s not that easy, that things really are very different from what you might think. So he says that by constructing versions you construct worlds, and I mean real worlds, hence there are lots of real worlds. Nobody knows what he is talking about! There are lots of versions, and versions are in part constituted of the things that are versions of. That’s all you can say, that’s all reasonably intelligible. But “we live in two different worlds”–I mean literally –it’s very hard to know what that means.

Some of the concerns in “Ways of Worldmaking” are just not central to my way of looking at things. It might be though that there isn’t much difference in approach between “reflective equilibrium”and “Ways of Worldmaking”. Because the notion of reflective equilibrium really starts in “Fact, Fiction and Forecast”.

AT: …it starts in a footnote, doesn’t it?

CZE: Yes. It starts in a footnote in “Fact, Fiction and Forecast”, and it’s an utterly brilliant idea, and I think it really is behind what Goodman is doing –it’s behind what he was doing even in “Structure of Appearance”, but certainly in “Ways of Worldmaking”. You could simply say that each of the worlds constitutes an equilibrium. So I think that really would make sense of what he is doing. And then he is focussing on the metaphysics of it, and I’m back home focussing on the epistemology and saying, how do we understand things if we understand them this way? It’s a difference in emphasis.

I don’t at all think that there’s anything alien about thinking you could do a Ways of Worldmaking account that would be adequate to ethics. Sometimes people would ask Goodman “why don’t you do ethics?” Partly the answer was that it just didn’t interest him. And he’d done enough other things…how much do you want from this guy? Which I think actually is a sound answer. But at one point he said: “You know, maybe Rawls did my ethics.” Maybe. Not that Goodman would have come up with the two principles, the two rules and the priority principle and all that, but what he thought about ethics was exemplified in the way Rawls constructed “Theory of Justice”. I don’t know what he thought about the content of the principles, but the mode of going about it, I think he thought, yes, Rawls approached the problem correctly.

AT: How did your career as a philosopher develop and how did you get involved with Goodman’s thinking?

CZE:  I went to Vassar College, where I studied philosophy. Vassar College was a women’s college at that point and they very much discouraged students from studying philosophy, because women couldn’t get into graduate programs. And if you had a degree in philosophy and didn’t go to graduate school you couldn’t get a job, because everybody thought you were too intellectual. Although they did not want people to study philosophy, I got permission to do to. I was a philosophy major, I had a great time. Then I went to Brandeis University for my doctorate, and I studied the “Tractatus”. There is a way in which the program was absolutely awful, because the professors were always fighting with each other. They neglected their students. But, we were self-motivated; we just went off and did a lot of interesting things on our own. I and a couple of other people did quite well by being ignored by our professors. Brandeis was not a first rate doctoral program in philosophy which meant that it was very hard to get a job afterwards. I realized, when I finished my dissertation, that although the theory of reference in the “Tractatus” is absolutely beautiful, you could not learn language if what it says about language is right. That set me to wondering of what kind of philosophy of language would satisfy me. I read around a lot and I thought about a lot of things, and my conclusion was that a good theory of language would be Quine without the physicalism. Quine has a holistic theory of language that is grounded in the conviction that physics underlies everything. I never thought he justified the point about physics. But the picture of language struck me as very good. Then I realized that Quine without the physicalism is Goodman. Goodman had this very nice theory, but he didn’t have a philosophy of language. I wanted to come up with what Goodman’s philosophy of language should be given his other theories. That’s what my first book is about. The book is called “With Reference to Reference” and it basically asks –given, what Goodman did in his other works, what he says about art, what he says about logic, what he says about science –what ought somebody who says these things be thinking about language. A rather nice theory of language emerges.

AT:But you just said that you were ignored by some professors. Who motivated you to go into theoretical philosophy?

CZE: No one. I think I am stubborn; basically people kept telling me not to. In college they said: Don’t do it. My parents kept telling me to drop out of Graduate School because it wasn’t appropriate for a girl. The professors, first of all, were ignoring everybody and secondly they certainly didn’t want to devote any time to women, because they were just totally sexist. So, I just thought philosophy was neat, and the fact that everybody was saying “no” to me…I just didn’t stop.

AT: But were you supported by your parents, financially?

CZE: No no. I got grants, but we were fairly poor in Graduate School. My husband was also going to Graduate School. He got grants, and I got grants. We had enough money to get by, but his parents had no money, and my parents would not support me because they totally disapproved of all of this. So even though they probably could have come through with some money, it was not in the cards. My mother was just keeping me calling, saying you should really drop out.

AT: And do what?

CZE: Be a housewife.

AT: Ah, not another topic?

CZE: No, no! No no! I was supposed to stay at home and take care of my husband, who as far as I can tell doesn’t need taking care of.


AT: From the point of view of the topics, would you say that there were some progressive professors who had a wide horizon, who were working on interesting new things?

CZE: Certainly there were at other places: Brandeis is quite near Harvard and you had people like Quine and Putnam and Goodman doing really exciting things. And it is near MIT and you had people like George Boolos doing wonderful things in logic. So there was a lot of exciting philosophy sort of around. And you could go to talks, you could join discussion groups. I don’t like to make this sound bleak, I actually had a very good time, but it was more or less in spite of the people who I should have been working with rather than because of them. I really did have a good time, I had lots of good friends. One of the guys I went to Graduate School with is still one of my closest friends. It was really quite good, but the sorts of things you would think would have made it good, were not what made it good.

AT: Do you have the impression that political changes had some impact on your biography?

CZE: Oh, dead sure, if it were not for the civil rights act in 1973, I would never have gotten a job. It used to be flat out accepted to discriminate against women in academia. The civil rights act made that illegal. Now, there’s a lot of discrimination, and there certainly was as I was going through, but things became slightly more open. And it’s very clear, that were it not for that law…it wouldn’t have mattered how good I was, I would never have gone anywhere.

CZE: Around 1980 you finished that work.

AT: I got my degree in ´75 and probably for about the first five years after I got my Ph.D. I was still kind of working on the “Tractatus”- type topics and maybe in ´79 or ´80 I kind of moved over and started doing Goodman stuff.

AT: ‘Goodman stuff’ that means what lead you to the book “With Reference to Reference”?

CZE: Yes.

AT: And how did he get interested in your work?

CZE: This is an astonishing story. I just told you the story of how I decided that what I was interested in was Quine without the physicalism. Well, I was at a conference, and I had breakfast with a publisher who had been Goodman’s student. I told him what I was interested in. And he said, “Goodman would be very interested in this. Why don’t you call him?” And the answer which struck me as perfectly obvious is: People like me do not phone up people like Nelson Goodman. That’s completely out of the question. So of course I didn’t. But I thought: Wouldn’t it be nice? I told my best friend from Graduate School, who also loved Goodman’s work –he loved Goodman’s work on induction mainly –but we both were serious Goodman’s fanciers. “Jon, wouldn’t it be really nice?”And he said: “Yeah, it would be very nice.” But it was like: Wouldn’t it be nice to be tall and blond? You know it’s never going to happen, it’s just something in your imagination, so forget it! I was in my office one day and I got this phone call. The voice on the phone said: “This is Nelson Goodman, and I understand you are interested in my work, and I think we should get together and talk about it. ”Whereupon I replied: “Jon, this isn’t funny. You know I would love to work with Goodman and this is not my idea of a good joke.” So, there was a dead silence. The voice on the other end of the phone said: “Well I don’t know how I can convince you over the phone that I’m really Nelson Goodman. But if you call the main Harvard number and ask them to connect you to my extension, if I answer the phone, will you believe me?”… I said “yes” and hung up and thought: “Oh, this is awful, I just ruined that one chance in my life! I have just ruined it by making a fool of myself!”


After gathering up my courage I called the main Harvard number and got connected to his office and it was Nelson Goodman and he still wanted to talk to me. I told him in general terms what sort of thing I was interested in. He thought it would be a really good idea. He had some ideas about how this might go. He was hugely supportive: He read what I wrote, he made some criticisms, he gave me access to unpublished materials. He actually talked the publisher into publishing this book when the book wasn’t written! It is completely unthinkable for an assistant professor with no standing whatsoever to get a book contract on the basis of not one word on paper. But Goodman convinced him that this was going to be great, and that they had to sign up immediately or else they will lose the opportunity to publish this book. He was a kind of demanding, picky person. A lot of people found him hard to get along with. But if he liked you, which mainly meant, if he thought you were doing good work and worked hard and didn’t waste time, he’d be enormously supportive. He really was terrific to me. Basically that’s how I wrote “With Reference to Reference”.

After a few years, maybe two or three years, I was still writing the book, he started showing me things he was writing and asking for my advice. He said: “I’ve got all these papers I’m writing on, now what do you think about them?” I would read them and suggest ways in which things might be different or better. At that point it sort of became a more reciprocal relationship, while I was helping him as well as he was helping me.

AT: And this was maybe on the way to the common work, “Reconceptions”, I suppose?

CZE: Initially what happened is this: Goodman was writing the paper that’s in that book on the interpretation of literary works. He showed it to me and I said: “There is a problem, there’s something you need to discuss, you haven’t discussed. You need a section about the syntax or something.” I do not remember exactly what my objection was. He listened to what I had to say and said: “Yes, you’re right.” He went off, over the weekend he rewrote the section and gave it back to me. I read it again and said: “I still don’t think you got it right, you’ve got to do something about this.”He said, “yes, you’re right”, went off and revised some more. Third time: “I still don’t think it’s right. It just seems that there’s this problem that you’re going to run into, and you haven’t figured out a way to avoid it.” Whereupon he got extremely frustrated and slammed the thing down on my desk and said: “If you’re so smart, you do it.” Well, instead of saying “Don’t be ridiculous, it’s fine, I made a mistake”, I thought, okay, I’ll see if I can. Maybe it’s not so easy to fix this.

When I showed him my revision, he said: “Yes, you’re right. You’re co-author of this paper, because this was a major point.” Once he saw what I had done, he saw that it did need to be done, and it really strengthened the paper. So, instead of just putting a grateful footnote to Catherine Elgin, for “inspiring me for these three paragraphs” or something, he said, this makes you co-author.

Then he was thinking about putting together a collection. He showed me what he had in mind, and he asked me what I thought. I looked at it, and again I thought there are holes, you really ought to have something on this, and something on this, the thing would hang together a lot better, if it weren’t so spotty. And he said: “But you’ve already done all that, why don’t we just put your papers and mine together and make it one thing?” So the book “Reconceptions” ends up having chapters that are mainly done by me, chapters that are mainly done by Nelson, and chapters that we’ve done together. At the end of each chapter, there’s a little parenthetical notation –a kind of sign of responsibility. Once we set that, we didn’t just stop there. We each went through the whole thing completely. The papers that I had already done went through revisions that he was very involved in, and ones that he had already done, went through revisions that I was very involved in.

AT: It seems that the way this book was written reflects the topic of the book itself. It’s not only a reconception, but it was repeatedly revised by two people.

CZE: Yes, I think that’s right. Also it is a reconception in another way. Because there are books that are just co-authored. Where both parties were intimately involved with the whole thing. And there are books, as I said, where you have one person writing the introduction and the other one does the body. But this is a much more dialectical type of thing, sharing responsibility, a little bit more he a little bit more I. That itself is a rethinking of how one makes a philosophy book.

AT: Do you imagine ‘the reader’ during writing? And does it have an influence on what and how you write?

CZE: I do think a lot about readers. It’s a major reason why I try to use a lot of examples. Philosophy can be very very abstract, and it’s helpful to have examples to tie it down. Examples are very good test cases. Bas van Fraassen and I had this argument, where he said an example counts for nothing, because one instance is never statistically significant. Well one instance is never statistically significant, but my view is that the difference between one and zero is enormous. If you can’t even come up with one example of your theory, then you can be reasonably confident, it’s no good. If you can come up with one, then there probably are more. And I think once you’ve come up with it, it’s only polite to offer it to the reader, so they don’t have to do the work of coming up with it themselves. Examples are extremely illuminating.

AT: Do you understand your work in general as one step of the development of analytic philosophy? Or is there a term that would be an adequate labeling or description of your work?

CZE: I’ve never come up with a term for what I do. It’s sort of nice if you can say “such and such ‘ism”, but I don’t have that. It is analytic, it has a pragmatic element in it in the sense that it seems to me that justification is not just a matter of the sources from which you build up a theory, but where the theory takes you. There is a forward looking element which resonates with James and Dewey as well as with Goodman and Quine. And there is a constructionist element in it. I don’t think –and this is very much like Goodman –that there is a determinate way the world is that is prior to a way of conceptualizing it. I think that we conceptualize things and impose order on reality. This might be kind of a contingency, I am not sure. But you’re not free to make things up as you go along, because the world resists certain things. There’s a constructivist element in this, but I don’t have any good words for this.

AT: As I see your work and the development of your philosophy, I would say that there is a continuous widening of philosophy’s range.

CZE: You have to figure out what you are saying very carefully, because philosophy always said it had universal scope. So you can’t widen something with universal scope, it’s already covering everything. What I’m doing is not taking in more things, but showing the legitimacy of more approaches to things.

It’s very traditional to have theories of art. But the account that I give, for example, of cognitive functions of art links it very closely to cognitive functions in science, instead of each having its own little separate realm, its own separate methodologies, its own separate interests. I’m saying there’s a lot more interanimation. In a way it’s not expanding the scope, but it seems that there is more texture in what’s in the scope. There isn’t a nice division of labor. There really isn’t a sharp division between, say, philosophy of literature and literally criticism or even between theoretical physics and philosophy of science. That is what we have learned from the history of philosophy up to now, that the sharp divides that maybe our ancestors hoped we would find just aren’t there. It’s not my fault! It’s not that I’m doing anything awful by saying, well, there is no figuring out whether this is theoretical physics or philosophy of science. It’s just the way things work out.

AT: In your book, “True Enough” from 2017, the philosophical idea that science and art can be treated in an analogous way is still prominent. At the same time the reader becomes aware from the very first pages that you emphasize science. You offer an approach to “accommodate science in epistemology”. Your primary aim, as I read it, is to describe how science works and also how science should work, given that it is not based in truths. Instead of truth, you introduce “true enough”. What are the characteristics of “true enough”?

CZE:Okay, a little bit of background. The problem for connecting science with epistemology is that standard epistemology treats truth as mandatory. Something is epistemically credible only if it is true. But science uses models and idealizations that are known not to be true and are not considered defective on that account. I should add science does not expect to eliminate the practice of using models. It is not a temporary expedient. So the question is how can models –and idealizations and thought experiments – which are known not to be true actually advance understanding. The way I characterize it as I say that they are true enough which is to say that the respects in which they are inaccurate are respects that do not matter to their cognitive functioning. You could use Michael Strevens’s term and say the respects in which they are inaccurate are not difference makers. The idea is that some of the features of a model are functioning to illuminate the phenomena they are dealing with, and others aren’t. Part of what you need to do in properly interpreting a model is knowing which are which. The reason for moving to “true enough” is mainly to accommodate the cognitive value of science.

AT: How does “true enough” relate to “reflective equilibrium” ? What does it add?

CZE: A system is in reflective equilibrium just in case its components are reasonable in light of one another and it is as a whole, as it is reasonable as any available alternative in light of what we antecedently are committed to; that’s the standard. What I’m doing with true enough is highlighting the fact that those components need not to be true –even the propositional components. Achieving reflective equilibrium isn’t just a sort of weaving together of individual truths to get a systematic system of truths. It is weaving together representations, and I would also say other sorts of commitments to get that equilibrium. The components are not defective if they are not true as long as the goal of reflective equilibrium is reached. If you understood for example gas dynamics using the ideal gas law, it would be a holistic account in reflective equilibrium. But the ideal gas law has these molecules that can’t exist. So the system in reflective equilibrium can contain representations that are not actually true and don’t purport to be true.

AT: As you mentioned, science relies on models and idealizations and thought experiments which are obviously not true. They are, as you coined it, “felicitous falsehoods”. This is an intriguing expression. Why should we endorse felicitous falsehoods in our scientific practices?

CZE: The term “falsehood”means they are not strictly accurate. The “felicity” means that despite not being strictly accurate, they’re performing an important epistemic function. The ideal gas law, strictly treats of putative molecules that are said to be dimensionless spheres which exhibit no mutual attraction and don’t collide with each other. No gas is like that. However, if you use that model you can highlight features of gases that are very important –the relationship between temperature, pressure and volume, which are there even in real gases. But if you added things like the different weights of the different molecules, the different geometries of the different molecules, the effects of the different collisions and everything, the interactions would be so complicated that you wouldn’t see the dominant features that are underlying them. By distancing yourself from irrelevant details that are always present but you get to see what’s really of interest. That’s why you want to do it. This is a common feature throughout science. You want a way of marginalizing features that you think are irrelevant in order to examine the features you think are important.

AT: This means “felicitous falsehoods”are not goals in themselves, but rather tools for increasing understanding.

CZE: Good, yes, but I would also say that truths are not goals in themselves. Basically we want understanding, and irrelevant truths are no better than irrelevant felicitous falsehoods. A system in reflective equilibrium is an abstract structure that consists of a variety of commitments that hang together. If you understand something then you can exploit it, use it, draw interferences, perform actions maybe, when your goals are cognitive. Understanding is a cognitive achievement. It is the achievement you get when you are in a position to properly use a system in reflective equilibrium.

AT: So I have to reflect by myself whether my system is good or not. But how can I know that I am working with a felicitous falsehood and not with a simple falsehood?

CZE: There are two elements here. One is you don’t have to reflect by yourself necessarily; there is a whole communitarian element. There is a community of inquirers who are working on this together. So there’s going to be various sources of input into your deliberation that might be very helpful. But you can’t know that you’ve got a felicitous falsehood. You can reasonably suspect that you do, if for example the experiments work out. If I think something is a felicitous falsehood, I might be wrong; it might be that it was just something that worked for accidental reasons.

AT: I find this very interesting, because I think it also has a strong political aspect. I perfectly realize that it is not possible to reach “ultimate truth”, in the sense of an absolute certainty, and I also I realize that it would be an advantage if scientists would not even pretend to have attained absolute truth. Because insisting on unquestionable truth amounts to the permanent confession that science has not yet reached its goal –and this makes it attackable and vulnerable. Do you agree with this view?

CZE: With a couple of slight exceptions. I have not found scientists claiming to have achieved absolute truth. I think scientists are very circumspect. Journalists and maybe science teachers in the lower grades make these strong truth claims, but the scientists I know, (and some of them are very arrogant) don’t say “I’ve got the truth”. So, I think we should be nice to scientists here. I think science’s goal is a moving target. Putnam and Boyd and Sellars talk about the end of science –a framework where all the answers will be in hand. That’s when science has reached its goal. Physicists know everything there is to know about physics or whatever. But another perspective comes from John Dewey. That’s the idea that every good answer leads to good questions. So, science is open ended. My favorite example of this is a sentence toward the end of the Watson and Crick paper where they’re describing the structure of DNA. And they say “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material.” This means that you haven’t just settled a question, you’ve opened up a whole new field of inquiry. And that’s the Dewey-perspective, instead of kind of converging on a conclusion, what you have is an opening of new lines of investigation. I think that’s a much better…it’s just a nicer view, but is also strikes me as a much more plausible view than that there is such a thing as “this is fixed and final and now we can go home”.

AT: As a new cornerstone of your epistemology, you introduce the idea that we are autonomous epistemic agents, although not independent from the scientific community. Do you think we should replace the idea that reaching truth is the ultimate goal of epistemic activities by the idea that epistemic agents have to establish their epistemic goals and standards themselves, but –that’s crucial –in a responsible way?

CZE: It’s not that I am throwing out truth all together, that’s not so. But distancing from truth means that you can’t say that what vindicates methods, procedures, standards of justification and so forth is truth-conduciveness. That is, as I said in the book, a consequentialist goal. “The end justifies the means” is the standard picture in epistemology, and it has all the problems that consequentialism has plus the problems it inherited when settling on truth as the goal. I need some other standard for epistemic rightness. Basically what I do is elaborate on ideas I find in Kant on ethical rightness. The kingdom of ends formulation of the Categorical Imperative is: You should act only on a maximum that you could reflectively endorse as a legislating member of a realm of ends. I say, let’s make the ends epistemic ends. Then we get epistemic responsibility.

One of the brilliant things in Kant that I think we should pay more attention to, is that Kant does not say that you should act on what you could reflectively endorse as a philosopher king. He says “legislating member” of a realm of ends. Legislators work together. In order to enact legislation they have to convince each other of the rightness of the laws they want. And that means, if I’m going to vindicate my own view, I have to convince you. But I have to therefore show that it is reasonable from your point of view, not just from my point of view. I evade subjectivity by the requirement that I’ve got to be able to get the other members of my realm of ends from their own points of view to say that “yes, from my point of view too, I could reflectively endorse this”. We start developing a collective understanding. That’s where the reasonability of a view comes from.

AT: And this is also what you mean by “epistemic responsibilism”?

CZE: Yes, I’m basically, you could either say stealing from Kant or elaborating it, to present an epistemic imperative. Obviously a realm of epistemic ends is not an ordinary community of inquiry. You’ve got to ask about the political shape of a community that would make its verdicts reasonable. I suggest there that the members of the community must construe each other as from a political point of view in that community being free and equal. Free means: Whoever you are in the community you can propose any hypothesis you can think is reasonable to take seriously. Equal means, you have an equal right to be heard, so power and prestige cannot dominate discussions in such a community.

This is obviously an idealization. It’s not saying this is the way actual communities of inquiry operate. But it is saying that it is an ideal by their own standards communities of inquiry ought to endorse. Because if we let power and prestige in, and if we let certain voices be silenced, it is not only unjust, it also sacrifices information. Even if you want to leave moral considerations aside, you don’t want to throw out information. And if some isolated individual has an outsider’s idea that’s really worth pursuing, and you’ve silenced her, you’ve lost that idea.

AT: As you’ve said, this is an ideal. But with respect to current politics, one has the impression that there are two camps: On the one hand, there are people who, for example, deny climate change, and they have power and influence. On the other hand, there is responsible science with its standards, techniques, peer-review and so forth. What should be the answer from a scientist’s view considering denialism and skepticism?

CZE: Okay there is a long-term answer: that’s better science education. I think one reason why the claim that human activity has nothing to do with climate change gets foothold, is that the public is not well enough educated in science. To report on science in the popular press you have simplify and simplify and simplify. But if the public was aware of how scientific models function, you could explain what the models are actually showing. The public would recognize that the models are felicitous falsehoods, not lies. And the students –or the ex-students, the grown-ups on their own –would have reason to reflectively endorse the reports, if the journalism is well done. Now, for those of us who were not scientifically well educated, who are already grown-ups and are not going to go back to school, a lot more depends on science journalism.

The other part, which is more flat out political, is get special interests out of the picture. Don’t let the oil companies dominate the debate, telling the politicians that there is no problem, and of course we need bigger and better and more polluting cars. We need to get special interests to be less dominant in political decisions. But there is also an educative element.

Something I should add is that if this picture is right, if the picture of epistemic responsibility being a matter of people working together for shared common goals, there is a political element already in standard epistemology. I mean the kind of Cartesian picture is: it’s you and the phenomena; and that’s it. But I am saying: “No, it isn’t. It’s you and others and the phenomena.”Once you start getting others in, then you have the seeds for some kind of political relationship.

This interview was edited from two conversations held at Erfurt (little Synagogue, June 11, 2005) and Berne (Institute for Philosophy, September 16, 2018). Special thanks to GeorgBrun, who accompanied all steps of the project.

Mentioned books of Catherine Z. Elgin:

Between the Absolute and the Arbitrary (1997) Cornell University Press

With Reference to Reference (1983), Hackett.

Reconceptions in Philosophy and Other Arts and Sciences (1988), together with Nelson Goodman, Routledge.

Considered Judgment (1996), Princeton University Press.

True Enough (2017), MIT Press.


Ariane Tanner is an independent researcher and writer. She studied history and philosophy at the University of Zurich; for her PhD at the Swiss Federal Institute of Technology Zurich she specialized in the history of science (published in German “The mathematization of life”, Mohr Siebeck: 2017). Currently she works as a lecturer and prepares a book about the history of plankton research and the concept of biomass. In (online) journals she publishes media critiques and writes about ecological issues. She is involved in performance art.