I've been meaning for some time to say a few things about Sam Harris's TED Talk. It's been fairly heavily criticized. I have to say, though, that I liked it quite a bit. Yes, there are more sophisticated issues and questions than Harris could get into in the limited amount of time available to him; the time to bring up those sorts of objections will be if he doesn't address them in his book. I don't actually have a problem with the bare idea of taking ethics as a science in a sense closely analogous to the sense in which physics, chemistry, and biology are sciences; in fact, the term originally started out that broadly. When Whewell in the nineteenth century popularized the term and originated history and philosophy of science as we know it, his term included the moral sciences. The debates between Whewell and Mill on philosophy of science were due chiefly to their different moral views: they were arguing over which view of morality held greater promise of a moral science. There is simply no a priori reason to shrink from the idea of a science, or a cluster of sciences, of morality. Despite Sayre-McCord's piece at the SEP (I mean, honestly, who thinks that an encyclopedia article on a philosophical position should spend virtually its entire space laying out arguments against it, while not spending much space at all looking at the different varieties that have been developed?), moral realism is popular and philosophically respectable. And there are good reasons for this.
So why was Harris's talk so controversial? There are certainly aspects of Harris's argument that seem problematic; his answers to his critics have been less than stellar, for instance (for example, here and here), and that shows some weaknesses in Harris's overall views. And Harris is using 'science' in a broad sense of the term, which leaves open questions about what, precisely, we're considering here. But many of the arguments raised against him have been less than stellar as well.
Harris's argument is that there clearly are facts about human wellbeing that are scientifically discoverable and that these are moral facts because, in fact, we value human wellbeing. Many of the objections simply don't address this argument in an effective way. For instance, as Harris points out, it's simply question-begging to point to disagreement about what constitutes morality or wellbeing on particular points as a reason for rejecting this line of argument. There is disagreement across the entire spectrum of human thought. If disagreement were on its own a reason to reject that something fits into the realm of facts, we would have to be claim that there are no facts at all. If disagreement is even relevant it must be the type of disagreement that makes morality so different. But the usual reason for identifying them as different is simply a denial of Harris's thesis.
One of the problems, and it is a problem from which Harris doesn't fully extricate himself, is that oughts are regularly being confused with values here. They clearly are different; we don't even talk about them the same way -- values are described in terms of noun phrases, while oughts are described in terms of verb phrases, for instance. By not recognizing the distinction himself, Harris makes matters for himself more difficult than they need be. It can be just a bare, primitive fact about human beings that, in the main, we like health and that we like poetry; saying this doesn't smuggle in any 'oughts'. It is useless and silly to respond to such a claim by saying it doesn't tell us 'whether we ought to prefer poetry over health or vice versa'; it wouldn't change the original fact, if fact it is. But given the value there are optimal and suboptimal strategies to achieving it, and this is a factual matter, and it is at least closely tied to the question of what 'ought' means. And thus Harris is right that there need be no problem in regarding the structure of all of the following as more or less the same structure:
"We ought to go by Route A rather than Route B" if we want to get home in the quickest way we can
"We ought to avoid cigarette smoke" if we value lung health
"We ought to encourage poetry" if we value social cohesion
"We ought to advocate for property rights" if we value prosperous society
It will simply be a matter of fact, investigatable like any matter of fact, whether the strategy in quotation marks is a successful way of achieving the result in the if-clause. Nor is any 'ought' smuggled in when we do this. You do need an account of what 'ought' means in order to do this, but it suffices on its own, and an account of what 'ought' means is not itself an 'ought'. (This is one of the problems one faces in talking about the 'no ought from an is' slogan; what it really seems to mean for most people is that 'ought' is an ineffable primitive. But this is a view with some rather substantial commitments, and a controvertible one at that; it simply should not be taken as being as obvious as the slogan often is taken to be.)
And note as well that something like this can be true, and moral facts capable of scientific investigation, even if there are moral conundrums that will never be cracked by scientific investigation. The fact that some moral problems may be too tough to be conclusively resolved by any science we can imagine tells us nothing about whether science can resolve smaller moral problems. The inability to come up with definitive answers for some things may result from any number of factors; it's not in itself an argument that no definitive answers can be given for anything.
This does, however, raise an issue that Harris doesn't treat in enough detail, namely, that unless values are hierarchically arranged, and provably arranged in this way rather than that, adjudication becomes impossible. How can one tell whether Rimbaud ought to have deliberately deranged his senses? Only by establishing whether poetry is better than health. The 'oughts' are the easy part of Harris's project. It's the values that are the troublemakers. Chris Schoen raises this point, although I think he spoils it a bit by connecting it to what seems to be a very different argument in the quote of Ophelia Benson. To Benson's point Harris can simply shrug and argue from parity again, saying, "And what of the fact that human beings often lie and fudge the details? You're not going to get rid of that problem by rejecting my view, and similar problems arise all along the border between fact and action without causing any fundamental problem." And he'd be exactly right. But he does need a real account of how to adjudicate the values that constitute wellbeing, and this he has not given. Again, it's possible he addresses it in his book. If he doesn't, it's a point critics should press.
I think part of the reason why Harris's talk was so controversial is that it made him run afoul of a large group of atheists whose atheism depends at least in part on the sharp separation of fact and value. One does run across them, the people who think that such separation provides a reason for rejecting religious beliefs (which often do not recognize a sharp separation between the two). And they certainly have been showing up in combox discussions of Harris's talks. That he also thinks all facts fall under the domain of 'science', however precisely he understands the term, means that he can be accused of scientism from one side and of giving away the store from the other. Whether he is in fact guilty of scientism depends on what, precisely, he means by science (since he is never very precise about it), and whether he is giving away the store depends on whether he is right in his larger view that putting morals under the science column takes it away from religion. However, while there are plenty of particular points on which I disagree with Harris, it's not an intrinsically unreasonable position, at least as he has presented it so far, and too many of the discussions have treated it as if it were. And in its general structure there is much to be said for it.