Mark Andrew Schroeder interviewed by Richard Marshall.

Mark Andrew Schroederswims in the deep contested waters of meta-ethics thinking all along about why it ensnares all of us, about metaethical expressivism and noncognitivism, about its relationship to semantics, about its relationship to natural languages, about the importance of 'not', about technicality and depth, about slaves of the passions, about disagreeing with David Enoch, about moral realism, about what's at stake, about Bernard Williams, about hypotheticalism and about the Humean theory of reasons. Go figure.

3:AM:What made you become a philosopher?

Mark Andrew Schroeder:I was raised in a home where critical thinking – especially of received cultural ideas – was emphasized by my father and highly valued, by parents who are both two of the most compassionate, and two of the most principled, people I know. So when one of my first college courses – taught by Jennifer Manion at Carleton College – sought to apply critical thinking to the subject matter of moral philosophy, it was love at first sight. After that I was lucky to be in a supportive undergraduate environment with lots of good role models of other students who went on to do PhDs, and I always, at least early on in my career, believed that I could work harder than anyone else, so that gave me the confidence to see it through, even when the path looked dim.

3:AM:You work in meta-ethics – what are we to understand by that term and why do you find metaethics something we’re all likely to get snared up in at some point?

MS:The simplest definition is probably that metaethics is concerned with questions about ethics – paradigmatically, with questions in the philosophy of language, philosophy of mind, epistemology, and metaphysics, about moral or ethical language, thought, knowledge and reality. We’re lucky to be in a period, however, in which metaethics is often construed much more broadly – many questions start off as of interest in arguments about traditional metaethical questions, but end up having great interest in their own right, and are not obviously classifiable as just questions about moral language, thought, knowledge, or reality. If you attend the annual Wisconsin Metaethics Workshop, for example, which I think is a showcase for the range of issues that are of interest under the heading of ‘metaethics’ today, you’ll see that the range of issues covered is incredibly diverse – which I think is absolutely wonderful, and why I’m lucky to be working in this field today.

The easiest way of seeing why metaethics is something that can easily ensnare all of us is just to focus on simple cases in which real-world people disagree about important moral issues. In my introduction to Noncognitivism in Ethics, I focused on the case of a procedure called infibulation with excision, which is a particularly problematic variety of female genital mutilation. Most readers will find it obvious that it would be wrong to have this procedure – which comes with many well-documented drastic medical risks – performed on one’s daughter, but not only is there a large population that continues to perform this procedure, it is actually often seen as a duty, a matter of giving one’s daughter the best chances in life. These are very different perspectives. Because these issues are so important to us, it’s hard to avoid getting caught up in reflection over how such deep disagreements are possible, and to even begin that reflection is to get caught up in metaethics.

3:AM:Metaethical expressivism is what you call the heir to noncognitivist views in ethics. So what do you take noncognitivism to be in ethics, and briefly, what went before metaethical expressivism?

MS:‘Noncognitivism’ is a term of art that has been used differently by different people. When I was an undergraduate, a visiting speaker once chastised me for misusing this term when I followed Michael Smith’s use, from whom I had learned it. I guess that what I take to be interesting, here, is a tradition that starts with Axel Hagerstrom and Ogden and Richards, is popularized in the 1930’s, particularly by Ayer and Stevenson, and continues through Hare, Blackburn, Gibbard, and others today. There are many threads to this tradition, so we can define ‘noncognitivism’ in many plausible ways. But I think all of these theorists share the idea, at a minimum, that words like ‘good’ and ‘ought’ have a fundamentally different job description from words like ‘square’, ‘tall’, and ‘green’, in a way that plausibly makes them in some sense metaphysically lightweight, and also tells us something about what Stevenson called the ‘magnetic’ uses of moral language – ways in which moral language seems to exert a kind of pull or sway, or to be inherently practical in some way.

3:AM:Metaethical expressivism asks us to revise important issues in semantics. Can you say what’s at stake then in this theory?

MS:The basic turn, for the expressivist, is from characterizing the meanings of predicates in terms of what it takes for them to apply – their extensions – to instead characterizing them in terms of what it takes to think they apply. What is particularly fruitful about this turn is that in some sense everyone can make it, because if you can characterize meanings in the former way, you can characterize them in the latter way. But the advantage of this turn is that it allows for the possibility that for some predicates – perhaps moral terms among them – there is a fact of the matter about what it takes to think it applies, but no perspective-independent fact of the matter about what it takes for it to apply. Put differently, the expressivist framework is one that gives a particularly natural way of making sense of how moral predicates could have a kind of meaningfulness without determining any extensions solely in virtue of that meaning. Described in this way, it is an alternative to relativism and some kind of dynamic theories of meaning, as well as to some kinds of inferentialism.

One way of thinking about this, from a metaethical point of view, is as making room for the idea that in some sense there are no moral facts – since ‘wrong’, for example, does not determine any extension in virtue of its meaning, there is no property in the ordinary sense that we are responsible to, when we use the word ‘wrong’, because it is ordinarily assumed that properties determine extensions. But the big idea of expressivism is important far outside of metaethics. It may even be that it gives us a much more fruitful perspective on semantic theorizing in general that is less prone to paradox.

The expressivist turn offers something more promising – a way in which we might actually be able to describe the meanings of the words in the language that we are actually speaking, without courting paradox. So I think its interest extends far beyond metaethics.

3:AM:You argue that you can’t understand natural languages if this type of expressivism is true – and the word ‘not’ seems to have a big role in your thinking here. What’s the problem and is this a knock down reason for saying metaethical expressivism cannot be true?

MS: That’s a great question – I have written a great deal about a problem sometimes called the ‘negation problem’ for expressivism. But this is a little bit misleading. In fact, there isn’t anything special about the negation problem; it’s just the simplest case of a much more general problem about structure that any adequate expressivist theory needs to respect. To get a sense for the problem about structure, note that on a naïve gloss, an expressivist might tell us that to think that stealing is wrong is just to disapprove of stealing. So now compare the following three sentences:

N1) Caroline does not think that stealing is wrong.
N2) Caroline thinks that stealing is not wrong.
N3) Caroline thinks that not stealing is wrong.

According to the expressivist, to know the meaning of ‘stealing is not wrong’, we have to know what it takes for N2 to be true. But unfortunately, our naïve expressivist view doesn’t allow for enough distinctions. We can distinguish:

1*) Caroline does not disapprove of stealing.
3*) Caroline disapproves of not stealing.

And some simple reasoning shows that 1* has to correspond to N1, and 3* has to correspond to N3. But there is nothing left to correspond to N2, which is what we really want to understand. So there is a problem understanding the meaning of the logically complex sentence, ‘stealing is not wrong’. The problem gets some more constraints when we observe that it’s not enough just for the expressivist to tell us some story about what it is to think that stealing is not wrong; she also needs to tell a story that explains why ‘stealing is not wrong’ has all of the right logical and semantic properties. Note that none of this is a special problem about negation; we could easily have started by distinguishing sentences like the following:

v1) Caroline thinks that stealing is wrong or Caroline thinks that killing is wrong.
v2) Caroline thinks that stealing is wrong or killing is wrong.
v3) Caroline thinks that stealing or killing is wrong.

Again, there is an important distinction between the first and second sentence, and another important distinction between the second and third. And again, to understand the complex sentence, ‘stealing is wrong or killing is wrong’, according to the expressivist, we need to know what makes v2 true. But again, there are not enough distinctions in ‘Caroline disapproves of stealing’ to go around.

To solve this problem, the expressivist needs a robust notion of content. As I would put it now, they need a surprising theory of the nature of propositions. There are different ways of going about constructing such a notion, and the one I developed and then criticized in Being For is just one of them. I don’t know of one that I find wholly satisfactory; the one I developed in Being For runs into troubles with certain natural language expressions, and I worry about others for different reasons. I don’t know that this is a knock-down argument against expressivism, but I think it tells us a great deal about what a successful expressivist view would have to look like, and I think that the vast majority of people who continue to write and publish about this and related problems are still focused only on special cases of the problems, without thinking about them in full generality.

3:AM:Why get so technical – why not just say it fails for ‘deep’ reasons like many metaethicists do?

MS:I’m extremely skeptical of high-level attempts to say why a view like expressivismmust fail, without getting our hands dirt with the details. The basic reason is that expressivism constitutes such a different perspective from that occupied by many of its high-level critics, that it’s very easy to take for granted things that look very different from an expressivist perspective, or to misunderstand how an expressivist could actually treat some phenomenon. More generally, the fact of the matter is that we don’t have a birds-eye perspective on our own language and minds – we understand them from the perspective of our language and minds, using our own minds to formulate ideas in our own language. That gives the expressivist – in principle – enormous resources for making sense of things that puzzle people about naïve statements of expressivism, because how our minds and language work are themselves some of the free variables, for the expressivist. It’s an ongoing question, I think, the extent to which those resources can be utilized.

3:AM:You’re a moral realist aren’t you and take a naturalist reductionism position don’t you? What’s this and why is supervenience important?

MS:I would characterize myself as a reductive realist, yes. I don’t get too worried about the title of ‘naturalism’, because that’s not something I ever set out to defend, although I think that by most ways of counting, the view I defended in Slaves of the Passionswould qualify. I don’t have a lot to say about why to be a realist in the first place outside of what I think goes wrong with particular irrealist views, metaethical expressivism among them. But I’m attracted to reduction in metaethics, largely because I think it promises to explain what to me is the most puzzling feature of moral facts: their necessary relationship with the non-moral facts, as exhibited by the fact, generally referred to as the supervenience of the moral on the non-moral, that there can’t be any moral difference, without some underlying difference in the non-moral facts. Necessary connections between distinct existences strike me the kind of fact that can’t be just a brute, unexplained fact about the world – I see them as in need of some kind of explanation, and a reduction would provide that explanation.

3:AM:Why is it better than other kinds of realism for you, such as Moorean non-naturalism defended by David Enoch, for instance?

MS:I think that Davidand I part ways over the importance we attach to explanatory power in theory selection. I start doing philosophy with things that puzzle me, and looking for ways to explain them. David, I think, sees things he finds deeply right or deeply wrong. He and I share a holistic perspective in the assessment of philosophical theories, which he often characterizes by considering how many ‘plausibility points’ a view receives for this or that feature. But in addition to disagreeing with him, in some cases, about the measuring of these ‘plausibility points’, I guess that I don’t think the relevant points are all exhausted by plausibility considerations. The reason is that it’s merely a conceit to think that the philosophical theories that we are comparing are full-formed theories. In many cases, they are more like research programs.

As an investigator who is making decisions not about which research program to bet on before some divine oracle reveals the answer tomorrow, but about which research program to invest my own time and efforts in over the long haul, the upside potential of a view is of much more significance to me than the work required to realize that potential. I’m not sure if this answers the question, exactly.

3:AM:If realism wasn’t true – and many philosophers don’t think it is – then what would you believe instead?

MS:The short answer is that I would still believe it but be wrong! But if you talked me out of realism, it would probably be into some form of expressivism. I don’t know what that form would be, but over the last 5-6 years I’ve become increasingly sympathetic to the idea that we need expressivism for more pressing applications outside of metaethics – particularly, as an account of truth. And I don’t really see how to give an expressivist account of truth that won’t make room for metaethical expressivism. So that has warmed me up to that idea a little bit.

Depending on how we define ‘realism’, a contextualist theory like that of my colleague Steve Finlay might also qualify as a retreat from realism, and in that case I would be cautiously sympathetic to some variant of a view like his End-Relational Theory, which I am most impressed about for his treatment of deontic modals. And if you talked me out of reductionism but not out of realism, I think I’d be split between adopting the views of my colleague Ralph Wedgwood whole cloth, on the one hand, and trying to develop some of what I think are the most interesting threads in Ralph Cudworth, Richard Price, and H.A. Prichard, on the other. I’d probably fight a valiant fight on Cudworth and Price’s behalf before conceding that Ralph is right.

3:AM:The Humean theory that reasons for action are instrumental, ‘slaves of the passions’ , is an unpopular theory these days but you’re drawn to it and defend it don’t you. What’s at stake with this theory?

MS:The biggest thing at stake over this theory is the universality and prescriptivity of morality. On a standard interpretation, Mackie assumed that nothing could be wrong, unless it was based on rules that apply to absolutely everyone, no matter what they are like, and are genuinely normative, in the sense of giving them reasons to act. But since he assumed that whether a person has a reason depends on her aims, it depends on what she is like. So, he inferred, there can’t be rules that both apply to absolutely everyone and give them reasons, and hence nothing is wrong.

Gilbert Harman argued, around the same time, that morality doesn’t really require such universal rules, on the grounds that moral judgments of the right kind do require reasons, but for Humean reasons, no reasons can be universal in this way. And Philippa Foot famously argued, again around the same time, that morality can maintain a kind of absoluteness by not providing reasons – again for the same Humean reasons. To me, these three authors in the mid to late 1970’s, and Bernard Williams, in ‘Internal and External Reasons’, which followed just a few years later, set the agenda for much of the most interesting work about morality and reasons over the next twenty or more years, by people like Christine Korsgaard, David Velleman, Michael Smith, Jean Hampton, Warren Quinn, and many others – particularly including Foot herself. The theory is unpopular today because these three authors made vivid its natural consequences.

3:AM:Bernard Williamshas been an important philosopher in how the theory has been formulated for contemporaries – is it your view that he and others haven’t been careful enough in formulating it?

MS:I don’t think Williamslacked care in formulating his version of the theory – though there are many difficult features of the article, ‘Internal and External Reasons’, which began this literature. I think, rather, that by taking care over different things than Williams did, it’s possible to formulate a general view that has much weaker commitments, and that by standing on his shoulders, we can worry about different things, now, than he had to be worried about when he originally wrote. Sometimes we can recognize the achievements of an article that set an agenda without constraining ourselves to think within the space of possibilities set out by that article.

3:AM:You’re interested in background assumptions that moral philosophers import into their arguments – why are these so important to your approach?

MS:I don’t think they’re important only to my approach – I think we should all care about when it is that we find our arguments persuasive only because we are working with some implicit picture. Interesting philosophical arguments are complex and often resistant to straightforward formalization, and philosophical prose that insists on formalizing every important argument is not only stilted, but sparse on important arguments and ineffective at communicating its most important points.

So some of the key pieces are in place for there to be gaps in our arguments that we don’t, ourselves, recognize. We know these gaps are real, and coming to grips with them is at least one important first step toward trying to imaginatively inhabit the views to which we are unsympathetic. If good philosophy ultimately involves the holistic comparison of packages of ideas, as I think it does, you can’t do it well while importing assumptions that are tacitly rejected by proponents of the package of views that you are considering.

3:AM:You defend a version of the theory called Hypotheticalism don’t you? So how do you formulate the Humean theory of reasons and why do you call it a ‘parity’ thesis?

MS:My formulation of the Humean Theory of Reasons starts with what I take to be a paradigmatic example of a desire-based reason. The Humean Theory of Reasons is the theory, essentially, that all reasons are relevantly like the reason given in the example. This is why I call it a ‘parity’ thesis, because at bottom it says that two things are the same, without building in a lot of theory about what makes them the same – though you can arrive at more specific versions of the Humean Theory by adding more about what makes the reasons the same, or, put differently, by adopting more specific views about what is going on in the paradigmatic example.

3:AM:So how does your approach handle those arguments against the Humean thesis?

MS:There are a lot of different arguments against the Humean thesis, but the argumentative task of Slaves of the Passions, my first book, is to show that each of them is really only an argument against some versions of the Humean Theory, because each turns on assumptions about what is going on in the paradigmatic example. I also argue, for each of these assumptions, that the best version of the Humean Theory of Reasons has independent grounds to reject the assumption, and hence has independent grounds to take a form that is immune to the objection. That, at any rate, is the organizational conceit of the book.

3:AM:And finally, are there five books you could recommend to readers here at 3:AM, to help us explore further your philosophical world?

MS:Aside from some of the historical canon, my own engagement with many of the issues I think of as broadly metaethical derives from two books published in the mid-’90s that I think have been formative for everyone of my generation: Christine Korsgaard’s The Sources of Normativityand Michael Smith’s The Moral Problem. I was lucky enough to have teachers who steered me toward both of these as an undergraduate, I spent a lot of time with each early in my philosophical development, and both have deeply informed my work, in different ways. The best source, I think, for insight into what metaethical expressivism might be and what you might do with it is Allan Gibbard’s Thinking How to Live, which I was lucky enough to have a chance to start thinking about while in graduate school, while it was still in manuscript. The hardest material in the book is quite difficult, but Gibbard writes with a lucid and charming style that makes it a great pleasure to read, and its scope is stunning. David Lewis has always been a role model for me, and his On the Plurality of Worldsis, in addition to being justly one of the most widely cited works of contemporary philosophy, a model of constructive theorizing. Finally, my ‘philosophical world’ is grounded in the history of philosophy, and particularly in the history of moral philosophy. I’m not sure if it’s the right thing to mention, here, but I’m currently working through Terry Irwin’s The Development of Ethics, which is a massive but wonderful take on one perspective on the history of ethics from Socrates through Rawls, and has the virtue of including a lot of discussion of figures from the tradition who are slighted by sparse conceptions of the canon. There are many other books worthy of mention, here, but those are five that illustrate different things of importance to me.

Richard Marshallis still biding his time.