Thursday, April 1, 2010

Is Morality Objective?

Sam Harris, author of The End of Faith and Letter to a Christian Nation, gave a recent TED talk in which he argues, against what seems to be a common consensus, that morality and moral values can be put on a firm "scientific" - meaning fact-based - ground.

Many bloggers have responded, including one of my favorite physics bloggers, Sean Carroll. Sean points out the saw, going all the way back to Hume, that one can't derive moral truths from factual statements about the world. This is often shortened to, "You can't get 'ought' from 'is'."

Harris responded to Carroll and other critics in a post at Project Reason. He quite reasonably points out that he only had 18 minutes to present his case (per TED requirements), and lays out his view in more detail. (He also points to a forthcoming book in which he will make the case in greater length.) Harris says Hume is by no means the last word on "is" and "ought", and not all philosophers accept Hume's conclusion. Luke Muelhauser makes a similar point on Common Sense Atheism.

Carroll answered Harris again here. Russell Blackford came down on Carroll's side, too, citing Peter Singer.

I have no idea who, if anyone, is right in all this. I know I would like to have an objective definition of morality,  both because I would like a rational way of ordering - and defending - my own actions and because I would like a rationally defensible way of requiring others to do the same. If it's all just a matter of personal preference, then no progress can ever be made in the moral arena. But I'm skeptical of Harris's approach precisely because of that desire: my desire for an answer might mislead me into thinking that there is an answer.

At any rate it is a fascinating, and, I think, valuable debate.


  1. Was my post on the subject of any help?

  2. Hi, Luke. Yes, I found your post helpful, but to be honest I did not think it was up to your usual standard of clarity and excellence.

    Thanks for checking out the new blog!

  3. Hi Robert. I just came across this old post of yours, and couldn't resist commenting.

    The word "ought" is problematic, since there can be both moral and non-moral oughts. Suppose I say, "If you want to catch the train, you should leave now." I'm not imputing to you any moral obligation, or indeed any sort of obligation. I'm just giving you advice on how to achieve your goals. This is not a moral ought. Now suppose I say, "You ought to give money to charity [regardless of whether that contributes to your goals]." Then I'm imputing to you a moral obligation to give to charity. Context is key here.

    To avoid any such ambiguity, I prefer to state the dictum as follows: you can't infer a moral fact from a set of facts that don't include a moral fact. Or, better, you can't infer a moral conclusion from a set of premises that don't include a moral premise.

    To make sure that we don't conflate moral with non-moral premises, it's a good idea to include an explicit term of moral judgement in the premise, such as "morally wrong", or "have a moral obligation". Then it's clear that you can't directly infer a conclusion that contains such a term from a set of premises that contain no such terms. You are going to need to give some sort of account of how you can justify introducing such a term. That account will need to make some claim about what terms of moral judgement signify.

    Moral naturalists--like Harris and Luke--claim that certain moral facts are equivalent to some apparently non-moral facts. For example, they may claim that "X is morally wrong" is equivalent to "X is contrary to well-being". Now, I think most people's competence with language enables them to see intuitively that this is not so. Perhaps (for the sake of argument) it's the case that those things which are morally wrong are those things which are contrary to well-being. In philosophical terms, "morally wrong" and "contrary to well-being" would then have the same "referents" or "extension". But those terms do not mean the same thing. They don't have the same sense. To say that something is morally wrong is to impute a moral obligation not to do it. But to say that it is contrary to well-being is just a neutral statement about its consequences.

    The problem becomes much clearer when we talk directly about moral obligation, and my experience is that moral naturalists generally avoid that term. Harris at one point explicitly says he's going to avoid that term. That enables him to continue using the ambiguous term "ought", over whose meaning he can (unwittingly) equivocate. He also equivocates between moral and non-moral good. The arguments he uses to support his basic claim commit fallacies of equivocation.

  4. Oops. I messed up my first example a bit by substituting "should" for "ought to". Of course, they're near enough equivalent. But it makes the passage read a little oddly.

  5. Thanks for checking out the old posts, Richard. I agree with your comments. I posted my own analysis of Harris's error here:

    I'd be curious to know if you agree.