Bright Morals

by Brian on July 27, 2003

Larry Solum has a typically insightful post responding to Matt Evans’s criticism of Richard Dawkins for proposing a naturalistic ethics. I think Larry’s criticisms are spot on, but for my money much too tentative.

Matt relies on a fairly crude no ought from an is principle. It’s notoriously difficult to get a statement of that principle that isn’t vulnerable to immediate and obvious objections. I won’t go through them all here, partially because I’d rather discuss the ethics than the logic of this point, and partially because Gillian Russell has a nice summary of the objections in her paper In Defense of Hume’s Law. As the title suggests, she thinks some version of the no ought from an is principle can be salvaged, but it’s clear that it won’t be easy.

But let’s set aside the technical concerns about the premises. The real problem is the conclusion that Matt reaches. Here’s the important passages.

bq. [T]here are no ethics in naturalism. Naturalism is an acceptance of _what is_, and ethics is the domain of _what should be_. There is no way to bridge the is/ought gap without referencing an extra-natural source. ..Other atheists and agnostics take naturalism seriously; they believe there are no moral absolutes, there are no ethics. .. To these people it isn’t wrong to kill Jews for being Jewish, it’s just that some people think it’s wrong to do so. Though I spoke with many atheists and agnostics in college and law school, I never found one who adopted this view. …Dawkins was wrong when he said his ethics are based on naturalism. His religion, like all others, ultimately rests on non-rational faith.

The core point seems to be that without some non-natural entity there are no ethics. Let’s spell out some conclusions of this position, because really it is just about the most absurd thing anyone could possibly say.

Say we know the following facts about the world. It contains creatures who are capable of feeling pleasure and pain, who have hopes and plans and fears and regrets, who are capable of great learning, and creating works of great beauty, who often love their children and parents and occasionally love each other, and who have emotional attachments to those people who they love so they are affected by the pleasures, pains, successes, failures etc of those they love. Now a naturalist could easily come to know all these things about the world.

The person who thinks naturalism can’t ground ethics thinks that we could know all those things about the world and still think it’s a wide open empirical question whether it is morally wrong to torture one of those creatures for one’s own amusement, or to kill all of these creatures to relieve a minor headache one has, and so on. Now I can imagine that some people really do think this is an open question, but only because some people are psychopaths. I really don’t think that anyone around here seriously thinks that in such a position we have to do extra work to find out whether it’s right or wrong to torture these creatures for fun. We already know enough to know full well that it isn’t. (Which is not to say that anyone who says that there’s no naturalistic ethics is psychopathic, but rather that they just aren’t being careful enough about following through the consequences of their own position.)

In case this isn’t entirely obvious (and frankly I can hardly think of a more secure premise in ethics, but just in case) try the following thought experiment. Imagine we find out tomorrow that all theistic theories are just wrong. (Everyone makes mistakes.) There’s really nothing around here but us baryons. Would anyone, I mean _anyone_, think that suddenly we had no ethical obligations whatsoever? That it was now OK to torture babies for fun? To put the point in Bayesian terms, anyone whose confidence in _any_ extra-natural hypothesis is as high as their confidence in the proposition that it’s wrong to torture babies for fun has a very odd worldview.

There’s an analogy here with an argument Jerry Fodor makes in one of his attacks on teleosemantics. On some teleosemantic theories, it would turn out that if Darwin was completely wrong and we really were all created by a divine being, none of our words or mental states would mean anything. But maybe Darwin was completely wrong and we really were all created by a divine being. It’s not very likely, but everyone makes mistakes. It isn’t at all plausible that if Darwin messed up then none of our words or thoughts mean anything at all. So evolutionary theories of content can’t be necessarily true.

The analogy here cuts reasonably deep. Since evolutionary theories are true, they may affect the precise nature of our semantic theories. In particular it’s probable that the evolutionary facts affect the boundaries of biological terms. And similarly if an extra-natural theory of some kind is true, it’s possible that it will affect the precise nature of our moral theories. (But note it’s not trivial to see how theistic theories could have such an effect, a fact I’ll come back to presently.)

Since naturalism and ethical obligation are clearly compossible, any argument that they aren’t must be mistaken. Still, it’s fun to run through (over?) some of the simpler arguments that they are not.

For example, some may argue that you can’t have ethics in a natural world because it’s impossible to see how to logically derive ethical claims from microphysical descriptions of reality. But this is probably just a fact about our inability to engage in complex reasoning. After all, it’s more or less impossible to see how to derive economic facts from microphysical descriptions of reality, but still I’m pretty sure that statements like “Inflation was higher in Britain in the 1970s than the 1990s” are true, and are true solely in virtue of the arrangement of microphysical particles around the world. Just how the connection between microphysics and economics is maintained is a mystery, but there must be one. Facts about inflation are not magical facts that need to be ‘added in’ to the physical world, even if it isn’t clear exactly where they are to be found in it.

Similarly, some will argue that naturalism and ethics are incompatible because it isn’t entirely clear in virtue of exactly what natural feature of the world it is wrong to torture babies for fun. Maybe, as Larry suggests, an explanation in terms of what makes for natural flourishing will be the start of a solution. But you know, this question stays hard even if we add in extra-natural entities. The arguments in the Euthyphro against the claim that something is good in virtue of being loved by the god(s) still look pretty good. The problem of saying what makes an action right or wrong is a hard problem for everyone. To conclude from that that the naturalists can’t answer it is a bad mistake.

Slight disclaimer: I know there are arguments more serious than a blog entry for the views I’ve been ridiculing here. A decent response to them all would also take longer than a blog entry. But essentially my response is going to be the same kind of Moorean response. (Kind of because I’m being a Moorean about epistemology to draw very anti-Moorean conclusions in ethics.) I’m more confident that torturing babies for fun is wrong, even if there is no super-natural force in the world, than I am in the premises of any complex philosophical argument, so if you have a complex philosophical argument for the conclusion that whether torturing babies for fun depends on the presence of super-natural entities, I’m just going to reject one or more of your premises. Given enough time, I’ll usually be able to figure out which premise I want to reject.

More serious disclaimer: This is my second pro-Bright post in two weeks, which is quite disturbing since I really don’t want to stand up for them particularly. But that doesn’t mean I’m going to endorse _bad_ anti-Bright arguments.

Final disclaimer: This has been edited slightly to moderate somewhat _some_ of the ranting, and edited to fix a typo.



Anyone 07.27.03 at 10:11 pm

The wrongness of harming sentient creatures may be obviously wrong to a rational sentient creature who assumes that ethical judgments exist to protect sentient creatures from harm.

But would it be wrong for a rational agent that isn’t sentient?


elina 07.27.03 at 10:47 pm

Perhaps I’m missing the point but I fail to see why people are so desperate to rescue ethical language, including the “brights”. Brian tells us that we’re all going to agree that it’s ‘wrong’ to torture babies even if our premises are proven mistaken.

Obviously I see what he’s getting at but why must it be ‘wrong’, why can’t it just be ‘not nice’. People are as universal in their agreement that torturing babies is not nice as they are in agreeing that being nice is positive. I don’t know what combination of particles makes me think being nice is positive rather than negative but I’m pretty sure everyone agrees.

Let’s cut the moral terms. They only serve to confuse things and you don’t need to labour on with these ‘oughts’ from ‘is’.


tew 07.27.03 at 11:36 pm

Perhaps I’m a psychopath, but I’m entirely able to imagine a robot fully programmed with all naturalistic facts about the world that still goes around poking babies with sharp sticks. You could tell the robot all kinds of facts like, “This is a baby. Babies feel pain. Babies dislike pain. Pain is caused by being poked with sharp sticks.”

Then you ask the robot, “What is this?” “ A baby.” “Would the baby like to be poked with a sharp stick?” “No the baby would not like that.

Okay, give the robot a sharp stick. Say, “Robot, poke this baby.” Is the robot going to do it? Sure it is, unless it’s been given some robotic laws. I can think of two or three offhand.


E. Naeher 07.27.03 at 11:38 pm

Your argument seems to rest on your idea that ‘everyone knows that torturing babies is wrong.’ Leaving aside in its entirety the fact that this is demonstrably wrong (why does your instinctual morality derive more directly from empirical observation than a psychopath’s? Sheer majority?), does this mean that moral questions on which there is no instinctual consensus obvious to anyone observing the natural world—abortion, capital punishment, even issues as seemingly clear-cut as slavery which an overwhelming majority of participants in certain societies saw as moral—are morally indeterminate? This seems about as flimsy to me as the ‘genetically modified babies shouldn’t be allowed because … ewwww!’ school of thought.


Elihu M. Gerson 07.27.03 at 11:56 pm

Perhaps organizations and institutions such as families, gossip networks, courts, churches, police, and other communal arrangements have something to do with defining and enforcing morality? I’ve always thought of these as part of the natural world.


Adam 07.28.03 at 12:02 am

You’ve made a typo in this post which renders the meaning rather unclear (to me, anyway):

“Similarly, some will argue that naturalism and ethics are incompatible it isn’t entirely clear in virtue of exactly what natural feature of the world it is wrong to torture babies for fun.”

Should there be an “if” in there, perhaps? Just requesting a clarification. Other than that, thanks! This is great.


Morat 07.28.03 at 12:12 am

Frankly, you can derive a fairly good code of ethics from just a straightforward look at social evolution. Us humans are social animals. We don’t live and die in a vacuum, but in a pack.

Social behavior, like any other form of behavior, can benefit or harm an organism.

Starting with even a crude version of evolution and crude social groups, you can move forward into a fairly reasonable (and familiar) set of ethics based just on that old favorite: enlighted self-interest.

As an atheist myself, I can assure you that my moral code is (as best I can tell) pretty much entirely based on enlightened self-interest, and that enlightened self-interest is built entirely on a rational look at human society and how humans react.

To address the baby issue: I’m likely to have babies some day. It’s against my self-interest to have a world where babies get poked by sharp sticks. That might be my baby getting the sharp end.

Same with rape, murder, etc. Even the intangibles like “freedom” or “property rights”. Thank you, but I like my stuff and my freedom. It’s certainly in my best interest to live in (and encourage) a society with maximal freedom for everyone.

Which does necessitate the odd restriction on my personal freedom.


Shai 07.28.03 at 2:03 am

Good post Brian. Whomever wrote the Bright page (probably not Dawkins) might only have meant that they reject ethical positions that rely on divine authority rather than arguments, but anyway…

In response to elina: to some philosophers, the problem is that ethics needs to be rescued from a position that submits that Hitler wasn’t wrong or evil, but merely “not nice”. This is the influence that some types of logical positivism, and radical behaviorism, et al theories have had.

For example, in Ayer’s emotivism ethical statements aren’t obviously true, like, for example, statements about fluid dynamics, so according to Ayer they only amount to saying “boo” or “i don’t approve”. So ethical statements are “non cognitive” (not amenable to truth i.e. being true or false).

There’s some wisdom in the emotivist position as a criticism of some absolutist grand metaphysical ethical positions, but more recent positions are more nuanced than the ethics that positivist emotivists were criticizing circa 1924. (interestingly correspondence theories and falsification criteria are strangely metaphysical themselves).

Anyhow, there are many good counter-arguments of the emotivist position. For example, I find Hilary Putnam’s deflationist/neo-pragmatist ethics pretty convincing, even if I’m not as enthusiastic about deflationism/anti-realism as some people:

on “nonscientific knowledge” here (in mp3); here Putnam does cite a book titled “Fieldwork In Familiar Places” and more specifically the Egyptian Book of the Dead to show that there’s more regularity in human “ethics” than most people presume, so I suppose his understanding is somewhat naturalistic.


“ethics without ontology and “the collapse of the fact/value dichotomy” here The second part of the collapse book is a pretty damning assessment of the emotivist position. (also see a review of the book, here )

For more direct criticisms of emotivism, see CL Stevenson “ethics and language”, and Blanshard “The New Subjectivism in Ethics”


John Landon 07.28.03 at 2:09 am

Moore’s naturalisitic fallacy is very well known, less well know are a ton of dissenters–even in the analystic field. Can we get a bibliography.
What should naturalism be defined in separation from ethics? Why?


Shai 07.28.03 at 3:42 am

Nevermind on my first point. I see that Dawkins did indeed write that.

I think Bernard Williams once said that the naturalistic fallacy is strange because it’s neither a fallacy or about naturalism. (I may be mistaken on that). But I did attend a course on European history and there were some peculiar ideas about “natural order” and “goodness” from the Enlightenment period that philosophers such as Hume and Moore may have been reacting to. Plato, Aristotle, Montesquieu, Rousseau, Nietzsche etc all have a different understanding of ‘natural’. And today we tend to think of ‘natural’ being associated with ideas from evolutionary psychology, or anthroplogy, etc. But I’m certainly not an expert on any of these points.


Brian Weatherson 07.28.03 at 3:54 am

In response to John’s question, naturalism shouldn’t be defined in separation from ethics. Indeed, since it’s meant to be a global theory, I think it shouldn’t be defined in separation from anything at all. Here’s the definition of naturalism that I think might work. (It’s a slightly weaker version of the claim made by Frank Jackson in From Metaphysics to Ethics.)

bq. There’s no possible world which has the following three properties:
(1) Every microphysical fact is exactly as it is in this world.
(2) The microphysical facts in that world provide a complete desription of that world.
(3) There is a (non-indexical) proposition that is true in this world and false in that world.

The idea is that the microphysical facts determine all the facts – no proposition could differ in truth value without a change in the microphysical base.

Dave Chalmers uses a similar definition in his arguments against what he calls physicalism. (He still takes his view to be naturalist, because he thinks the only propositions that make (3) true are propositions about consciousness, and consciousness is part of the natural world.)

I tend to agree with Dave that propositions about consciousness are the most likely to generate counterexamples to (3). The next most likely would seem to me to be propositions about meaning or propositions about necessity. I’m inclined to think it’s very unlikely that morality will create counterexamples here – even if there are super-natural features of the universe, I’m not sure how they could affect what is right and wrong.

This isn’t a perfect definition of naturalism (naturalism per se isn’t committed to all the facts being fixed by microphysics) but it’s the closest I know. I really don’t know if it’s what the self-described Brights believe. But that’s because their gospel isn’t particularly clear.

(By the way, the reason I include non-indexical in (3) is that it’s possible that there are propositions of the form “This world is w” where w names this world that are true here but not in other worlds microphysically just like this one. They don’t make naturalism false I think. But if “Brian is happy” is true here and not true over at the other world, then naturalism about happiness is at least in trouble, and physicalism is almost dead.)


Chun the Unavoidable 07.28.03 at 4:08 am

I believe that leftist thinkers such as Alan Dershowitz and Eric Alterman have argued that torturing babies is wrong but might be necessary in a “ticking time bomb” situation to save American lives and thus should be made available to the appropriate authorities.


dsquared 07.28.03 at 7:11 am

Brian, I don’t understand your 1-3 above. Surely someone who was opposed to naturalism would say that there are no possible worlds where (2) is true?


Shai 07.28.03 at 8:12 am

Here’s a long post; I should start a weblog :/

It’s one thing to argue that every substance is physical and quite another to ground ethics in physical facts. If physical facts unequivocably fixed moral statements like “torturing babies is wrong” then there would be no reason to have that ethical statement, because no one would torture babies. But that begs the question: what exactly is ethics then? I’m not sure myself.

Steven Pinker is more of a Hobbesian than an Aristotelian. He thinks that we tend to be nice to our family and close friends, but by default tend to be indifferent or nasty to others from a different tribe or nation. We can modify that default setting to include other nations, all of humanity, and even other animals (Singer’s expanding circle). I don’t have “How the Mind Works” in front of me right now, but if I recall correctly, he says that more rational (or at least, better) parts of the brain can win out over others, so I guess he’s more optimistic than an average Hobbesian (He’s more eloquent and precise in his book, I’m sure).

Dennett believes that consciousness is like “fame”. Different parts, modules, processes, whatever compete for control. There is no Cartesian theater and he thinks we are less conscious than we normally think we are. In a chapter of Freedom Evolves, he argues that memetics may inform ethical and policy decisions (I suppose in a way similar to how the heuristics and biases program in social psychology can make us aware when we use stereotypes or how emotion affects our memory). “It is only we human beings who have the long-range knowledge capable of identifying and then avoiding the pitfalls on the paths projected by our foresightless genes” (p. 166, Freedom Evolves). Dennett’s fame metaphor is more neutral than what Pinker says, so one could imagine both pessimistic and optimistic implications even if Dennett himself seems to be somewhat optimistic. Students of Plato’s Republic will be reminded of his tripartite soul, and in particular the problem of “weakness of the will” even where there is knowledge. But I’m sure Pinker and Dennett will think that possibility will turn on the facts. The metaphors are there just to keep you interested.

Straight from the horses mouth:

“I’ve come to recognize a pattern. My fundamental perspective is naturalism, the idea that philosophical investigations are not superior to, or prior to, investigations in the natural sciences, but in partership with those truth-seeking enterprises, and that the proper job for philosophers here is to clarify and unify the often warring perspectives into a single vision of the universe. That means welcoming the bounty of well-won scientific discoveries and theories as raw material for philosophical theorizing, so that informed, constructive criticism of both science and philosophy is possible. As I present the fruits of my naturalism, my materialist theory of consciousness […], and my account of the mindless, purposeless Darwinian algorithms that created the biosphere and all its derivative products […], I encounter pockets of uneasiness, a prevailing wind of disapproval or anxiety quite distinct from mere skepticism. Usually this discomfort is muffled, like a faint rumble of distant thunder, a matter of wishful thinking almost subliminally distorting the agenda. […]” etc. (p. 15, Freedom Evolves)

Pinker and Dawkins are apt to have a different definition, but I think that all three would agree that evolutionary theory, anthropology, neuroscience, cognitive psychology, and science in general (in addition to great literature, and everyday knowledge) can inform our ethical positions by teaching us something about ourselves. A foundation for ethics, even everyday ethics, I don’t know. Here’s Putnam quoting Peirce: “we’re on swampy ground, but that’s what keeps us moving”.


raj 07.28.03 at 9:37 am

I’m amazed. You actually take Evans seriously?

He says

“There is no way to bridge the is/ought gap without referencing an extra-natural source.”

My response:

I”m floored. I really am.

OK, when you provide some evidence that an “extra-natural source” exists, people might believe you.

You know, evidence. You (presumably) went to law school. So you should know what evidence is.

Note: Someone actually pays people to come up with crap like that?


dsquared 07.28.03 at 11:23 am

Raj: You’re assuming that it’s possible to bridge the “is/ought gap” at all.


dsquared 07.28.03 at 11:53 am

For what it’s worth, I don’t like Brian’s Moorean argument. I think it just shows that we’re sure it’s definitely *disgusting* to torture babies, not that we’re sure it’s *morally wrong*. I can think of at least one example in the Bible of someone (Abraham IIRC) who was all set to execute his own son, despite feeling the exact same things as the rest of us, because he thought it was the right thing to do. So while’ you’re entitled to conclude from introspection that it (intersubjectively and universally) *feels* wrong to tortue babies, I don’t think that you can then move on to say that it’s a “fact” that it’s wrong, unless you’re going to take a position that intersubjectively and universally feeling wrong is the same as being wrong. Naturalism about what’s disgusting seems much more defensible to me than naturalism about what’s wrong.


Chris Bertram 07.28.03 at 12:54 pm

I liked the post very much, Brian, especially the Bayesian move. But I’m not so happy with your clarification of what naturalism is in this context in terms of all the facts being fixed by the microphysical ones. That might be right, depending on what counts as “a fact”, but not be very helpful to naturalism in ethics (or to naturalistic accounts of normativity more generally – e.g. in epistemology).

Given the way the world contingently is, it may be true that all the ought-to-be-dones, and ought-to-be-judgeds and all the ought-to-bes quite generally are fixed. But I take it that’s because of the way in which the world contingently is, * coupled with general normative principles * fixes the ought-tos, where those general normative principles (in ethics, logic, epistemology, rationality) are what they are quite independently of the way this world is, microphysically speaking.

Not that there aren’t derivative principles that may be local-fact dependent, but that at some ur-level the guiding normative principles have to be clear of the contingent features of any particular world.

The fact that there can’t be a naturalistic explanation of such ur principles (indeed such principles would be needed to judge what would even count as a good explanation) doesn’t, of course, mean that we have to have recourse to some supernatural entity or deity or whatever. I take it that’s a completely different Q.

[This is meant to be similar to/derived from a Jerry Cohen argument against Rawlsian constructivism, so if you know that argument and I’ve fluffed my lines somewhere, go with that instead]


Chris Bertram 07.28.03 at 12:58 pm

Just a brief addendum. When I’ve made the 2-worlds-alike-in-all-microphysical-respects move in arguments in the past, people have often questioned the coherence of the thought. After all, what individuates these worlds? It isn’t as if you can point and say “this one over here as opposed to that one over there”. I’m not sure what I think about that objection – passing it on anyway.


Jack 07.28.03 at 2:45 pm

Isn’t the real problem with Evans argument that he seems to assume that everything is on a par once some extra natural assumptions are involved.

This is not the case, instead a slightly different mode of argument is applicable. The situation is then one like the argument againt creationism -A god might have created the world in seven days complete with already ancient fossils (differing distributions of carbon isotopes or whatever) but not only would it be strange thing to assume, making the assumption does violence to normal ways of resolving situations which even the creationist would use where things unclear. As such there is the possibility of a naturalism that isn’t logically necessary, may be incomplete, may be lacking in positive assertions but is intelligible and more robust than alternatives.


Brian Weatherson 07.28.03 at 3:06 pm

(Warning: Long post ahead)

_On Microphysically Similar Worlds_

I think dsquared is right that _some_ naturalists won’t think the definition of naturalism is coherent. Whether you think the definition of naturalism is coherent depends on what kind of anti-naturalist you are. If you are an anti-naturalist because you believe in ghosts, for example, then it seems you could at least conceive of what the world would be like with no ghosts. Similarly, I think, if you are an anti-naturalist because you believe in souls. You can just imagine the souls away. Now this might have dramatic effects – maybe no one would have conscious experiences. I doubt it, but no one really understands consciousness so it’s wrong to rule out anything at this stage. But I don’t think it’s incoherent to talk about what the world would have been like had there been these collections of flesh and bones walking around without souls attached to them.

(By the way, this is why in the cases I was thinking of there isn’t a problem of distinguishing the worlds in question. If one of them contains Casper the Friendly Ghost and the other doesn’t, they’re different! I’m assuming here that naturalists think Casper is possible, just not actual. I know some naturalists who reject this, but I really don’t understand their position.)

The real problem arises if you (a) believe in a monotheistic God, (b) take that God to necessarily exist, as some religions do and ( c) interpret that necessity as meaning existing in all possible worlds. This is a possible view, though it isn’t an easy one to make sense of. It’s committed to saying that it’s simply incoherent to talk about what the world would be like were there only material objects in it, which is strange because it sounds like a perfectly coherent supposition. (I’ve always thought theists who accept (b) should interpret the kind of necessity being attributed to God to be a different kind of necessity to that talked about by modal logicians. I was convinced of this by Bob Meyer, who’s a theist and a modal logician, and knows a lot more theology and logic than I could hope to know, so I’m not trying to be unfair to _all_ theists here.)

Perhaps it would be better to just strengthen the naturalist claim to include a clause that there is a possible world which is microphysically just like this one and contains nothing else. The naturalist agrees with this – she thinks this world is it. And if the anti-naturalist disagrees that’s OK, because she’s an anti-naturalist.

_On Torturing Babies for Fun_

I was making some fairly sweeping statements about this, and it might have been better to say just what I needed to say to make the point. (Not that I’m backing away from the stronger claims, but it’s worth knowing where my last line of defense is.) The kind of anti-naturalist I have in mind is committed to

(1) Super-natural entities are required for there to be anything that’s right or wrong.

(I think most of the targets believe something more subtle than that, but let’s start with the extreme cases. The skirmishing over the details can be put off.) From that it follows

(2) If there are no super-natural entities, then torturing babies for fun is not immoral.

I wanted to reject (2), and had a bunch of ‘arguments’ against it: appeals to intuition, thought experiments, name-calling, etc.

Now I think that (2) is pretty definitely false. But I don’t really need to prove that. My opponent requires that (2) is definitely true. If I can show that (2) is probably false, then I’ve shown that my opponents position is probably false. Even if I can show that it’s simply not supported by any evidence, then I’ve shown that my opponents position is, at a crucial point, not supported by any evidence. So I don’t really have to prove the universal claims I was making, even though I do think they’re true.

One might quite reasonably object that I haven’t even shown (2) is improbable, because after all, all I’ve done is appeal to intuitions and thought-experiments and engaged in name-calling. I think there are hard problems around here about the epistemology of ethics. How does anyone prove anything in ethics? At the end of the day, I think appeals to intuition are almost unavoidable. If that doesn’t look like enough to ground knowledge of moral facts (and I acknowledge that it looks like a pretty shaky platform) then I don’t see how a pretty thorough-going moral scepticism can be avoided.

Two final qualifications.

For this argument I don’t really care about whether torturing babies for fun is not wrong in wildly improbable situations. I care about whether it’s wrong in situations almost exactly like this one, but minus whatever super-natural entities one believe in. What I really care about is whether someone who’s having a crisis of faith – not an unheard of occurrence round here – should also start to wonder whether torturing babies for fun is really wrong. I say no, or at least probably no. If that’s the weakest premise I have to rely on in philosophical argument, I’m having a lucky day.

Second, I don’t mean by this to say that moral facts just are facts about intuition. I take it the relation between intuition and moral facts is similar to the relation between perception and physical facts. In the early 20th century it was fashionable to reduce talk about the physical world to talk about phenomenology, because that was the only way we knew about the physical world. This project was, I take it, a giant mistake. This is (slightly) more contentious, but I think it would be a similar mistake to reduce ethical talk to talk about intuitions. (Not that anyone here was doing that, but some people had the impression I might have been, so I thought I should clarify.)

One of the nice things about being brought up anti-positivist is that I always feel just as comfortable using premises like “Torturing babies for fun is wrong” as the foundation of arguments as I feel using premises like “Needle n on machine m pointed to reading r at time t” or whatever positivists thought should be the foundations of all reasoning. This does have the downside that occasionally other people don’t agree with the premises though, and it’s a little harder to get them to change their mind.

_On Extra Moral Principles_

I think Chris’s challenge to the definition of naturaism is very hard to reply to. I can feel the force of the objection, but at the end of the day I don’t have any direct response to it, just a feeling that it’s an objection we have to live with. I’ll try saying a little about why I think we can live with it, then say a little about why what I just said wasn’t very convincing, then go away and think about it a bit more.

One might think microphysicalists are committed to a simple contradiction. Microphysicalists think (a) microphysics is all there is, (b) there are such things as football teams and rates of inflation and (c ) football teams and rates of inflation are not microphysical entities. Isn’t this bad?

Well, no, as long as there’s some special relationship that holds between the microphysical facts and the facts about football teams or inflation rates, so the latter aren’t in some way ‘extra’ facts. I think the special relation is metaphysical entailment, but we can be flexible about this point to a degree. My ultimate line of defense here is that if the very same special relation holds between the microphysical facts and the moral facts, then the microphysicalist doesn’t need to worry about moral facts being extra facts. (Unless she really does have a problem with football teams and inflation rates, in which case microphysicalism really is in a bad way.)

Now I’m inclined to think that the very same relation does hold, but I can see why someone might think otherwise. Here’s how I’d put the core problem. (I hope this comes to the same kind of thing as Chris’s worry, or I’ve missed the point a bit. Wouldn’t be the first time.)

There’s a difference between what it takes to possess a moral concept and what it takes to possess an economic or sporting concept. If someone sees prices rising, but still denies there’s any inflation, that’s prima facie evidence that she doesn’t have the concept INFLATION. If she sees the whole game and sees Liverpool score more goals than Arsenal, but still is unsure that Liverpool have won, that’s prima facie evidence that she doesn’t have the concept WINNING THE GAME.

Compare this to what happens in ethics. Imagine someone who knows all the physical facts relevant to the institution of capital punishment as practiced in America, but who still finds it morally acceptable. I take it that our first reaction is not to say this is evidence that she lacks the concept MORALLY ACCEPTABLE. She’s morally confused, not conceptually confused. (If you think she isn’t even morally confused, change the example. I don’t care about the details for now.) Steve Yablo puts this point by saying moral concepts are essentially contestable. I take it part of Chris’s point is that who wins this concept isn’t the person who really possesses the concept, but the person whose concept matches up with the special moral principles.

There’s three responses available here for the microphysicalist. First, she could say that even though there’s a disanalogy here it doesn’t matter because there’s some other microphysical account of moral facts. This way lies hard philosophical work, and I have no idea how to do it all. But can’t rule it out. Second, she could just bite the bullet and say the supporter of capital punishment really is conceptually confused. Some days this is what I feel, but it’s not the best position to hold. Third, she could say that these conceptual facts are not relevant to the special relationship between microphysical facts and economic or sporting facts. The special relationship just is a metaphysical relationship, not a conceptual relationship. I think that’s the best way to go, but again there’s a lot of details to fill in before it’s more than a hand-waving solution.


Paul Orwin 07.28.03 at 5:55 pm

This may be your typical blinkered scientist viewpoint, but it seems to me that if you take naturalism, as I understand it, seriously, then the is/ought question sort of goes away. Here’s the way it seems to me that it could work.

1) Here is how the world works, and how people behave.
2) Here is our set of philosophical principles
3) Here is an explanation of how we arrive at those principles through reason, regardless of whether we invoke “extra-natural” beings or sources.
4) Here is an evolutionary explanation, or a physical mapping of the brain even, that demonstrates why we reason that way, and where our concepts of “extra-natural” beings comes from

In other words, it seems to me that the rules of social interaction that we follow (writ large, of course) derive from our evolutionary path, which led us to start forming groups, perhaps a few million years ago. For whatever reasons, our ancestors brains developed to harbor certain abilities and conceptions, including gods and supernatural phenomena, in attempts to explain and justify the world around us and our actions in it. This is not an argument that God is an evolutionary artifact, far from it. It seems to me that there is room in a naturalistic philosophical viewpoint for people needing to be able to invoke “extra-natural” beings, without those beings existing, in order to create and maintain the valued set of ethical precepts.

I apologize in advance if 1) this makes no sense or 2) this has been written before in philosophical language that I can’t make head or tail of.


Russell L. Carter 07.28.03 at 9:43 pm

Perhaps the question is formally undecidable from within every rational ethical system. The theist gleefully concludes that what should be is what he has been told to believe by the flavor of his faith. The naturalist shrugs his shoulders and concludes that what should be is whatever his group(s) have decided over the reach of time. Inevitably some of these groups will err, as inevitably do theistic prescriptions. The naturalistic explanation is the simplest possible. Given the theistic requirement of enormous elaboration beyond the observable, which historically has produced ethical systems at least as repugnant as those of any secular ideology, the theists have quite the burden of proof.


Shuan Rose 07.28.03 at 9:55 pm

Did I miss something? Sixty years ago, Nazis tortured and killed Jews by the millions, man, woman, or child. they thoought they were doing good, indeed ridding the world of evil.
Nine years ago, Hutus slaughtered Tutsis in Rwanda. The armies of mass murderers who worked for Hitler, Stalin, and Pol Pot did’nt think they were doing anything wrong-no, they thought they were making the world safe for Communism and that was a Good Thing.Many of those people were quite regular folks, not Hannibal Lector types.
Given the facts of history, I do not think it is at all obvious that torturing babies is wrong.


Shai 07.28.03 at 11:02 pm

Nevermind what the Nazis thought, I think it’s pretty obvious that torturing babies for fun is wrong course of action even if it isn’t a naturalistic fact. There is a social consensus that functions as if it were a fact, and further, anyone living in a society with remotely similar values and social institutions will probably agree it isn’t the best course of action. But I think there’s something about being human that makes most people sick just thinking about the possibility. Even if there were a society where this is the thing to do, the person arguing for it would still have to convince me that it’s a good idea.

In my previous post I wasn’t descending into ethical relativism! I agree with Brian.


russell l. carter 07.28.03 at 11:58 pm

“Sixty years ago, Nazis tortured and killed Jews by the millions, man, woman, or child. they thoought they were doing good, indeed ridding the world of evil.”

But the Nazis didn’t get efficient until they removed responsibility from ordinary soldiers onto a somewhat self-selected very few:

“Use of gas vans began after Einsatzgruppe members complained of battle fatigue and mental anguish caused by shooting large numbers of women and children.”

But of course a variety of “communists” were in sheer numbers considerably more lethal. However! You don’t say it explicitly, but if you mean to imply that history does not provide examples of morally equivalent atrocities committed at the behest of all manner of theists, then that’s not true. Most of the truly egregious ones in terms of proportional devastion occurred in earlier times when theists controlled state power. Now that they don’t, the worst they can do is devastate their own local flock.

Considerably simplifying, I’d offer that the death of stalinism occurred at least in part because enough of the world irregardless of whether they were theist or not eventually recognized the moral bankruptcy of it as an ethical system. And that includes a large number of its adherents as well.

Evidently there is something more fundamental at the root of this evolution in thought than theistic sermonizing.


Stuart Buck 07.29.03 at 2:25 am

I respond here.


Lawrence Solum 07.29.03 at 2:39 am

More on this here:


Mike Huben 07.29.03 at 2:41 am

The “no ought from is” principle works because “ought” really implies somebody or something to which duty or obligation is owed. Otherwise “ought” would have to be some sort of reification, some abstraction of duty. Philosophers either tend to make the duty owed to an imaginary deity, or they choose a naturalistic approach and base the duty on the desires of mere mortals.

Much of philosophy seems to be based on incomplete statements: once you complete the statement (as in this case youfind whome the duty is owed to), the situation is greatly simplified and much more concrete and honest.


Anarch 07.29.03 at 6:41 am

Otherwise “ought” would have to be some sort of reification, some abstraction of duty. Philosophers either tend to make the duty owed to an imaginary deity, or they choose a naturalistic approach and base the duty on the desires of mere mortals.

Not necessarily; you could have a notion of a “natural order”, where moral deviation is measured by deviation from that order. One can look at the Tao in that light, for example.


Keith DeRose 07.29.03 at 4:37 pm

I reject the basic form of one of the key arguments used in the post. It looks like: If we came to know that P, we’d still believe that Q; so, P and Q are compatible (or “compossible,” as Brian’s post has it). Perhaps more charitably, it’s something more like: Even if you came to know that P, you’d still believe that Q; so, you’re committed to the compatibility of P & Q. Or something like that. No matter: I reject all such forms of argument.

I’m far from sure about the matter, but I’m inclined to think free action and determinism are incompatible: in no possible world is it both the case that determinism is true and free actions are performed. I also believe, fairly tentatively, that determinism is false and, fairly firmly, that I do perform free actions. If some compatibilist were to ask me whether I’d hang on to my belief that I perform free actions even if I came to know that determinism is true, I’d definitely admit the answer is “yes.” If she then declared: “Aha! So you are committed the compatibility of free action and determinism,” I’d reject that as a fallacious argument. I answer the three relevant questions as follows:

1. Do you sometimes act freely? Yes [fairly firm]
2. Is determinism true? No [fairly tentative]
3. Is free action compatible with determinism? No [fairly tentative]

The compatibilist is asking me to imagine that I came to know that the answer to 2 is yes, and asking how I’d revise my relevant beliefs if I did come to know that. My response is that I’d then hang on to my positive answer to 1, and change my answer to 3 from “no” to “yes”: Show me that determinism is true, and, even though I’m an incompatibilist, you won’t get me to give up my belief in free action (though perhaps some others have belief sets where this would be their reaction, and it would be a rational reaction). Rather, you’ll then make me convert from incompatibilism to compatibilism. But what I don’t see is why that should make me a compatibilist now, without having seen the demonstration of determinism. [Here I take myself to be adopting the same basic position that Peter van Inwagen adopted in his great book, AN ESSAY ON FREE WILL (Oxford UP, 1983); for a very sensible discussion of this way of reacting to the possible news that determinism is true, see esp. pp. 219-221.]

Let N be some naturalist thesis, and change our questions to:

1. Are some actions morally wrong?
2. Is N true?
3. Is N compatible with the moral wrongness of some actions?

And I think we get the same story. Some type of anti-naturalist may well answer: 1: Yes [very firm]; 2: No [less firm]; 3: No [less firm]. They can admit that if they came to know that they were wrong about 2, they’d then change their mind about 3, but, as far as I can see, have been given no reason yet to change their mind about 3 now. Similarly for the teleosemanticist who thinks that meaningfulness of mental states is incompatible with a particular creationist story, C, but is willing to admit that if she actually came to know that C is true, she’d then revise her claim of incompatibility.

Here’s a less controversial example that might convince some of the invalidity of the form of argument in question. Let A be the anti-Goldbach property: a number has property A if it is *not* the sum of two primes. You are a mathematician considering a very large natural number which we’ll call “Charlie.” You’re sure that Charlie is an even number and is greater than 2. You haven’t checked whether Charlie has property A, and you’re not certain as to whether it does, but you’re strongly inclined to think it doesn’t. You think, and have good reason to believe, that Goldbach’s Conjecture [every even number greater than 2 is the sum of two primes – in our terminology: no even number greater than 2 has property A] is probably true. Suppose that some mathematicians you know and rationally trust claim that they’re closing in on a proof of Goldbach’s Conjecture. They’re not sure it will work, but it needs just a few holes to be plugged, and they think they will probably be able to plug those holes successfully – though they admit that there is some chance that the proof can’t be completed, and that maybe GC isn’t even true after all. Still, they think that GC is probably true, and, based on their reasons, you rationally believe that GC is probably true….. So, you’re an “incompatibilist”: you think that, probably, Charlie’s being an even number greater than 2 is incompatible with it having property A (and you think your friends will probably soon prove this incompatibility). Still, if some “compatibilist” were to ask what you would conclude if you determined that Charlie had property A, you’d have to admit that you’d continue to believe that Charlie is an even number greater than 2 and would revise your belief about the incompatibility of that with Charlie’s having property A. What you would do is inform your friends that they turned out to be wrong: GC turns out to be false, and Charlie’s being even (and greater than 2) is compatible with its having property A after all.

That’s what you would do *if* you came to know that Charlie has property A. But, as I hope everyone can sense, admitting that doesn’t commit you now to the compatibility of Charlie’s being even (and greater than 2) with its having property A. That’s no way to prove the compatibility of being an even number greater than 2 with having property A! If that were a good way to show compatibility, we’d be in a position to very quickly and all-too-easily lay to rest a lot of problems in math!


Brian Weatherson 07.29.03 at 5:01 pm

Well, OK, there’s a gap in the particular argument I gave. I was certainly assuming that around here conceivability and possibility are pretty closely tied together, and maybe that’s wrong. Let’s look at what the lie of the land is without that assumption.

All the thought experiment shows is that we (or at lest those of us who aren’t convinced beyond all doubt that naturalism is false) take it to be an _epistemic_ possibility that there could be ethical obligations in a perfectly natural world. Similarly in the Goldbach case, there’s an epistemic possibility that the Goldbach conjecture is false. And we know in general that this kind of epistemic possibility (epistemic possibility without perfect understanding of the concepts involved, in this case numbers) doesn’t imply metaphysical possibility, which might be what’s important here. (Or it might not, as I’ll get to presently.)

What does this show? Well, a few things:

(a) If conceivability entails possibility _in the ethical case_ then it’s a metaphysical possibility that there are ethical obligations in a perfectly natural world. For what it’s worth, I’m inclined to think a restricted conceivability-possibility connection will go through here.

(b) It shows that there’s no proven incoherence in the position of the naturalist who believes in ethical obligations, because as we’ve agreed, it’s an epistemic possibility that there are ethical obligations in a perfectly natural world, but if this incoherence had been proven that would no longer be an epistemic possibility.

In general, I think there’s an important disanalogy between reasoning here about mathematics, where all the relevant propositions are necesssary and a priori, and reasoning about ethics. In maths incoherence and impossibility really do go together, and both come apart from known falsity. In ethics/meta-ethics, it’s not clear how these concepts connect.

If you want to hold a position that says, “It could have turned out that ethical obligations apply in a natural world, but in fact as it actually turns out ethical obligations are necessarily linked to some extra-natural being”, the arguments here were meant to challenge that position. (I’d think there are other arguments against it, particularly Euthyphro arguments, but that’s a different post. I may as well add here that I wasn’t even arguing against the view that naturalists have to take some unattractive but coherent meta-ethical position, like non-cognitivism. Again, I doubt it, but nothing here even _starts_ to be a response to those arguments, which are serious arguments.)

The anti-naturalists I had in mind were arguing (or at least seemed to be arguing) that we can infer from commonly accepted premises that there are no ethical obligations in a perfectly natural world. As far as I can tell, your position agrees that that is false. It might be impossible that there are ethical obligations in a natural world, but any argument for that involves a premise that is in dispute, like the actual falsity of naturalism.

Remember I wasn’t trying to give some knock-down proof of naturalism. It’s certainly (epistemically) possible that there’s an argument from naturalist premises plus some other things that are true to the conclusion that there are no ethical obligations. But that’s just because if naturalism is false there are arguments from naturalist premises plus true premises to any old conclusion whatsoever, because our premises will now be inconsistent. (I assume here that we also don’t have conclusive proof that naturalism is true – anyone who disagrees should ignore this last paragraph, and indeed probably this entire comment.)


Keith DeRose 07.29.03 at 5:09 pm

The anti-naturalists I had in mind were arguing (or at least seemed to be arguing) that we can infer from commonly accepted premises that there are no ethical obligations in a perfectly natural world. As far as I can tell, your position agrees that that is false. It might be impossible that there are ethical obligations in a natural world, but any argument for that involves a premise that is in dispute, like the actual falsity of naturalism.

Right. We’re agreed here.
(Sorry: I haven’t found how to itallicize in comments.)


Keith DeRose 07.29.03 at 5:33 pm

In general, I think there’s an important disanalogy between reasoning here about mathematics, where all the relevant propositions are necesssary and a priori, and reasoning about ethics.

And you seem to think in the realm of ethics, the argument form, *You’d continue to believe that P even if you came to know that Q; so, you’re committed to the compatibility of P & Q*, or something like that, holds promise, even though that argument form fails in math.

I wonder, then, what you think of the compatibilist argument I reject in the non-mathematical area of free action & determinism? I’m guessing you back that argument, or at least think it holds some promise.

I think the van Inwagen-style incompatibilist stance is quite sensible (and in fact take that stance myself): I’m an incompatibilist, but show me that determinism is true, and I’ll become a compatibilist rather than give up my belief in free action. But that involves rejecting the compaitibilist style argument in the realm of ethics-not-math.


Brian Weatherson 07.29.03 at 7:18 pm

I don’t find that kind of position on free will very attractive, but I don’t have any idea how to start building an argument against it. But my gut feeling is that my position is going to end up resembling Dave Chalmers’s views on consciousness. As in

bq. Whether someone is free in a particular situation is independent of what the actual world is like. This makes ‘being free’ like ‘having experiences’ and unlike ‘being water’. Whether a particular substance is water depends on what the actual world is like. But not the others. (Insert sound of table being hit hard just about her.)

Er, if I had a better argument than that, I’d be running it here. Basically I think conceivability and possibility only come apart for natural kind terms, and ‘free will’ isn’t a natural kind. But both premises here are at best highly debatable.


Matt Evans 07.30.03 at 5:44 am

I’ve issued a challenge to those who think they can deduce a moral premise from nature at The Buck Stops Here. Please stop by!

– Matt Evans


gry java 01.13.04 at 9:55 pm

Mein Hobby ist es Gästebücher zu besuchen. Das ist immer ganz interessant und widerspiegelt so, was die Leute im Internet wirklich denken. War auch interessant bei Dir ! Bis zum nächsten Mal. All The Best OfNew Year. Sorry for my english i’am from Germany.

Comments on this entry are closed.