Ethical Realism

January 31, 2009

Chapter 3.3 “Supervenience Revisited” by Simon Blackburn

Simply put, moral supervenience is the view that a description of nonmoral facts (such as physical and mental facts) is enough to determine whether an action is good or bad. If this is true, then that fact in itself seems like a problem for morality because we want to think that morality is about something more than just physical and mental facts. Let’s put that aside for a moment and take a look at the one of the important essays about moral supervenience:

According to many philosophers, moral supervenience is plausible. Simon Blackburn sees supervenience as a challenge to moral realism. In particular, he will discuss the fact that we ban “mixed worlds” and he argues that moral realists cannot explain this ban. (What does that mean? Read on to find out.)

Moral supervenience is simply the idea that there can be no moral difference between two situations without a nonmoral (natural) difference. A difference between moral facts require a change of nonmoral facts. However, there can be nonmoral differences and no moral differences.

If one action is right but another is wrong, then there must be some difference in the physical or mental facts. Cutting someone is (usually) wrong when you are trying to hurt them, but often right when you are doctor trying to help someone with surgery. We tend to explain that an action is wrong based on the intentions and effects of the action. Cutting someone is bad insofar as it harms someone, but it can be justified if it is intended to help someone. Being harmed tends to be based on physical damage, and suffering, which is mental phenomena. Intentions are mental phenomena. So, the facts we present in a moral judgment appear to be physical and mental rather than moral.

Moral supervenience is quite similar to “mental supervenience,” which means that difference in mental states is only possible if there is a difference in physical states.

Although moral facts may seem to depend on physical facts, it is possible for two different physical facts to produce the same moral fact (60). (Two different actions could be “unjust” or “wrong.” Also, consider mental supervenience: Two different brain states could produce the same mental state. Different parts of the brain light up when we think of the word “dog.”)

Blackburn also mentions that there can be apparent exceptions to supervenience, where a nonmoral state exists, but the expected moral state does not (61-62). Whenever we find a counterexample to supervenience, Blackburn believes it is because of an intervening factor, which he calls a “releasing property” (62). In other words, the nonmoral facts that seem to determine the moral facts could not include the releasing properties.

What exactly is an example of an releasing property? The example I gave before might be a good example of this. Usually harming someone is wrong, like cutting someone with a knife. However, there might be intervening factors, such as the doctor’s intention to help someone through surgery. The releasing property can easily be added to our description of nonmoral facts in order to discover how moral facts supervene on physical facts. In this case intentions are found to be relevant to deciding if an action is right or wrong.

Blackburn’s main challenge to the moral realist is: Why the ban on mixed worlds (64)? We can imagine a “possible world” in which different correlations of supervenience apply. In one world harming someone is bad, but in the other world it is good. A mixed world would say that either of these is possible in a third world: Sometimes it is wrong to harm someone, but sometimes it isn’t. (Ignore the fact that releasing properties can give us apparent exceptions.) We have decided that these mixed worlds can’t happen, but we need an explanation for why we we have done so. Blackburn argues that anti-realists can explain the ban on mixed worlds better than realists.

Why we reject mixed worlds: To hold that someone is evil “for enjoying the misery of others” is saying something true for all “possible worlds” (65-66). We cannot accept that a world with such people could be considered anything but evil. If a possible world was identified where some people who enjoy the misery of others is considered evil, but others aren’t, then we simply haven’t identified the natural facts that determine the moral fact.

But what is a “possible world?” There are three possibilities that are considered. In what sense is supervenience necessary (65)? Is it necessary by logic (analytically), by physical laws, or metaphysically? If it is necessary by definition or logic, then we have a different kind of “possible world” than if supervenience is necessary by metaphysical constraints. Consider each of these:

  1. Analytically necessary: An “analytically possible” world is any world of moral competence we can conceive of despite physical or metaphysical constraints. In other words, the world is “logically possible” (logically consistent). Blackburn also views analytically possible worlds as a kind of cultural relativism. A group of people who speak a language merely need to be competent in that language to realise if a word has a necessary meaning.
  2. Physically necessary: It is physically impossible for certain natural facts to be true, but for the certain corresponding moral facts to be false. The physical laws cause the moral facts. A physically possible world is one that is constrained by our actual physical laws.
  3. Metaphysically necessary: The very being of moral facts is dependent on the very being on certain natural facts. No “possible world” could have certain natural facts without having the moral facts. Here a “possible world” is any change in reality, such as different physical laws. So even with different physical laws, the natural facts will still guarantee the moral facts. It isn’t the physical laws that cause the moral facts.

Moral Supervenience is Analytically Necessary

Consider if moral supervenience is definitionally necessary (logically necessary). In this case Blackburn suggests that we would understand morality by understanding moral practices to “choose, commend, rank, approve, forbid, things on the basis of their natural properties” (66). He believes that we can indeed understand morality in this way. If you fully understand how competent users of moral language use the words (or behave appropriately), then you too can use the words appropriately (and act appropriately).

“[T]he explanation [that competence of moral practice] depends crucially upon the role of moralising being to guide desires and choices amongst the natural features of the world. If, as a realist ought to say, its role is to describe further, moral aspects of reality, there is no explanation at all of why it is constitutive of competence as a moralist to obey the constraint” of moral supervenience (67).

In other words, if behaving how society deems to be appropriate can be understood through competence of moral superveneience, then it isn’t clear why we should be a moral realist.

It should be noted that somehow Blackburn thinks he has explained the ban on mixed analytically possible worlds through the fact that supervenience is based on competence of a moral practice, but it isn’t clear how this fact can be used to explain the ban on mixed worlds. Perhaps a moral practice would be impractical to say that sometimes something is wrong and sometimes it isn’t without any reason to say so. What I think Blackburn would say: It is our moral practice in which we ban mixed worlds. A different moral practice might not ban mixed worlds.

My Objection: If supervenience is analytically necessary, then I would expect that any consistent moral practice should be acceptable. I don’t know why we should be restricted to accepting a “moral practice.” Although we must accept a moral vocabulary, any set of coherent moral beliefs is “logically possible.” In other words, “moral facts” simply won’t exist and the notion of “competence” is irrelevant. Real moral competence would merely require logical consistency.

Blackburn’s Reply to an Objection

One problem with saying that a moral realist can’t account for our ban of mixed worlds is the fact that there might be supervenience in other areas of philosophy which don’t require us to be anti-realists. Blackburn considers why many of these areas of philosophy can involve supervenience without requiring anti-realism.

Example 1: Mental Supervenience

Many people believe that mental states are supervenient on brain states (68). Why do we reject that there can be mixed worlds involving mental supervenience? We believe that one brain state can give someone a headache in one world, but not give them a headache in another. However, there couldn’t be a possible world in which the brain state would sometimes give a headache and sometimes worldn’t.

Is headaches that supervene on brain states analytically necessary? This is implausible because plenty of people have completely denied such a supervenience and such views were “perfectly coherent” (69).

Is it metaphysically necessary? If so, we can easily ban mixed worlds because we can just deny that it is possible to have one (69).  However, he then considers an argument given by Donald Davidson that the supervenience of the mental on the physical is not lawlike (69-70). If superveniene isn’t lawlike, then we might be living in a mixed world right now! Sometimes a brain state will give us a headache, and sometimes it won’t. He responds to this problem by saying that if supervenience isn’t lawlike, then it isn’t supervience after all. Additionally, such a view might end up being anti-realist by “convincing ourselves that the physical reality is at bottom the only one” (70). He finds this to be much like his view about moral supervenience, in which moral facts are really just natural ones. This view could be seen as a practical concern about “how we have to relate this particular vocabulary to the underlying reality” and “it is derived from constraints on the way in which we must react to a non-mental, physical world” (70).

My objection: Yes, we could “just deny that mixed worlds are possible,” but could we justify that denial? If supervenience by definition requires us to have a lawlike relation, then it will always imply a ban on mixed worlds because a mixed world is precisely a non-lawlike relation: Sometimes the supervenience holds and sometimes it doesn’t. However, if supervenience doesn’t require a lawlike relation, then mixed worlds can only be banned with a justification. I can conceive a metaphyiscally possible world in which sometimes the mental supervenes on the physical and sometimes it doesn’t. Why should I deny this possibility?

Note: I agree that if mental facts supervene on physical facts, then it isn’t lawlike. We would like to think our mind can move our body. If the mind depends solely on our body, then the mind couldn’t do anything. (It would be epiphenomenal.) It might be that the mind does supervene on the brain, but we we have some reason to believe in mental causation as well. (No wonder Blackburn suggests that the physical reality might be the only real one.)

Example 2: “Natural kinds” superven on the physical.

For example, some philosophers argue that being “water” supervenes on H2O (71). Why would we ban mixed worlds in which sometimes water is H2O and sometimes H2O is something else.

Is the supervenience of water on H2O analytically necessary? Do we only require their competence? Many people never found out that H2O is water, but they still use the word appropriately. However, “analytically necessary” might involve some scientific competence (72). For example, we expect everyone to agree that competent people will believe that no two things can be “identical physically without also forming the same stuff” (73). In order to deal with analytically possible words we merely have to know who is competent, so the supervenience of H2O on water can be analytic.

In other words water might only supervene on H2O for some cultures, or if this is necessary for us to be coherent, which we believe is necessary for any culture that does know what H2O is. We are then antirealists about this kind of supervenience. “Natural kinds” are merely a part of language and not part of the metaphysical reality. “Water” is not part of the “true reality.”

Third Example: Color supervenes on refractive properties

Many philosophers beleive that Color is supervenient on refractive properties of surfaces. We see objects as having colors based on the wavelength of light that reflects off of it.

Blackburn agrees that color supervenience isn’t analytically necessary, or one “constitutive of competence with a colour vocabulary” (74). Such supervenience requires the knowledge of a specialist, so it has to be a kind of scientific fact. Therefore, the supervenience of color is based on physical necessity. So far it seems necessary for the physical world we live in to have the supervenience of colors. (We could find out we are wrong about this fact, but it is a conclusion that scientists currently agree with.)

Note: I don’t understand why water’s supervenience can be analytically true, but color’s supervenience can’t. Blackburn already admitted that some scientific knowledge can become mixed with analytical necessity. Additionally, it isn’t clear to me how the “color realist” accepts supervenience in a way that can’t be done by a moral realist. He says that the fact that color supervenes on refractive properties is plausibly physically necessary, (and therefore it is physically impossible for it to be otherwise based on current scientific observations) (74).

A Final Objection

Why overcomplecate the issue with our “ban on mixed worlds?” Because realists don’t see supervenience as a problem in areas, such as “color’s supervenience on refractive properties.” The “ban on possible worlds” merely seems to mean that we want to say that the supervenience is always in effect. It can’t be otherwise.

Blackburn tries to take a look at why realists accept superveience in the philosophy of mind, in color, and in natural kinds. If we can be a realist about one of these things and simultaneously accept supervenience, then the moral realist could also accept supervenience. Most important: He seems to agree that a “color realist” can agree to supervenience. If this is so, why can’t the moral realist also accept supervenience in the same way?

Why a moral realist bans mixed worlds: First of all, I don’t think moral supervenience is true. If it is true, then it isn’t “complete.” There are moral facts that don’t supervene on nonmoral facts.

Why moral supervenience seems to be false: Just like other anti-realists, Blackburn doesn’t discuss the fact that pain feels bad. Of course morality partially supervenes on “nonmoral facts” because the mental state “pain” is considered to be “bad.” In other words, some so-called “nonmoral facts” (mental or physical) might really be good or bad. What else would a moral fact look like? (The judgment “such and such is bad” will supervene on things like the mental state “pain.”

Perhaps there are some moral facts that don’t supervene on the physical. The fact that “pain is bad,” for example. What about this statement supervenes on the nonmoral? Or another example: We should do good things. (We know that because we experience pain and pleasure and so forth.)

I am sure moral realists could answer the problem of bannng mixed worlds in other ways. For example, we might ban mixed moral worlds for the same reason that scientists do. If pain in and of itself is bad, then we will argue that is merely a fact about all pain that occurs. Badness will supervene on pain. Moral states can be caused by mental states, just like mental states can be caused by physical states. We don’t need to justify the fact that pain is always bad (in some sense) any more than the fact that one refractive surface always gives a certain color in the right conditions. It seems true so far, so we assume that it is always true.

1 Comment »

  1. […] for the moral facts to be different and for the world to be the same. More examples can be found in Simon Blackburn’s Supervenience Revisited. It is well-accepted that physical and psychological facts seem to determine moral facts. (Many […]

    Pingback by “Moral Explanations” by Nicholas L Sturgeon « Ethical Realism — June 21, 2009 @ 11:32 pm | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

%d bloggers like this: