Slavery shows us “an example of how facts and reason can determine values and morals”, declares Michael Shermer. That’s starkly put, a direct challenge to Hume’s is/ought distinction between objective facts about the world and “oughts” that derive from values that we humans hold. I reject moral realism (the idea that moral injunctions have objective standing, independently of what humans think about them) and so reject Shermer’s claim.

I am not rejecting the idea that humans have an evolved nature, and that our feelings and values are very real and of the utmost importance to us. Nor am I attempting to dispense with morality, quite the converse. But moral injunctions must ultimately derive from us, from our values — our I like, I dislike, I laud or I abhor feelings — and cannot be derived from objective facts about the world.

Shermer argues for the latter, but I don’t think he succeeds. He argues by repeatedly translating one moral injunction into another moral injunction, giving the impression that eventually he has arrived at bed rock in a brute fact, when he has not. Slavery is an emotive example, so for clarity the discussion below is not about whether to reject slavery, it is about whether that rejection derives from our values, our repugnance and sympathy for other humans, or from facts that hold independently of our values.

“Slavery is morally wrong because it’s a clear-cut case of decreasing the survival and flourishing of sentient beings”, declares Shermer. But, as he then correctly asks: “Why is that wrong?”.

He answers: “It is wrong because it violates the natural law of personal autonomy and our evolved nature to survive and flourish; it prevents sentient beings from living to their full potential as they choose, and it does so in a manner that requires brute force or the threat thereof, which itself causes incalculable amounts of unnecessary suffering.”

That all seems very true. But (as he again asks): “How do we know that’s wrong?” He answers: “Because of what Steven Pinker calls the “interchangeability of perspectives,” which we might elevate to a principle of interchangeable perspectives: I would not want to be a slave, therefore I should not be a slave master.”

But that does not follow, at least not without additional premises. “Should” injunctions are instrumental, that is they pertain to desired goals. You only “should” do something if that gets you some goal you wish for. And it may well follow that: “I don’t want to be a slave; and there is least likelihood of me being a slave if society bans slavery entirely; and therefore it is in my interests to uphold that rejection and relinquish being a slave owner”. But that calculation is different from the above principle.

Let’s consider a thought experiment in which Daniel possesses a superpower such that he can enslave others, but with zero possibility of him being enslaved. If morality were objective and “slavery is wrong” were an objective fact then it would have to bind Daniel, and yet the logic leading up to “… therefore I should not be a slave master” doesn’t hold for him.

The injunction: “you would not want to be a slave, therefore you should not own slaves” might well work fine as an appeal to human sympathy, being a grounding of anti-slavery in human values, but it does not work as an attempt at objective logic.

Shermer then correctly brings in game theory, referring to: “… the evolutionary stable strategy of reciprocal altruism: “I will scratch your back instead of being your master, if you will scratch my back and not make me a slave.” It is the behavioral game theory strategy of tit-for-tat: “I won’t make you a slave if you don’t become my master.””

This may indeed be tactically astute, and adopting this attitude might be your best strategy, especially if you are in danger of being enslaved. But, again, this is a moral scheme adopted instrumentally based on ones values and self advantage (“I don’t want to be a slave”). That does not give you an objective morality. For that you need an argument for why Daniel, with his superpowers, should not own slaves. Indeed, game theory would say that, if you’re trying to maximise your winnings, and you know that you have the power to do so, then you should (instrumentally) exploit others by enslaving them.

Shermer continues: “The principle of interchangeable perspectives is also a restatement of John Rawls’ “original position” and “state of ignorance” arguments, which posit that in the original position of a society in which we are all ignorant of the state in which we will be born … we should favor laws that do not privilege any one class because we don’t know which category we will ultimately find ourselves in.”

Again, this is about tactics that would be in your interests if you indeed were in that “state of ignorance”. But, by the time we’re old enough to reason morally we’re not in that state. And, anyhow, are we really saying that the reason plantation owners in the American South should (morally) have freed their slaves was out of fear that they might one day be enslaved? Even if there were zero likelihood of that (as was pretty much the case), wouldn’t you still want them to abjure slavery? Is fear of being enslaved really your argument for why you consider slavery to be objectively wrong? This does not sound like moral realism (objective moral truths that hold always for everyone, even Daniel), it sounds like politics — the negotiations we make with each other to get along and to attempt to steer society to our liking. And that is indeed what it is!

Perhaps Shermer realises the problem since he then says: “Lincoln’s ultimate moral avowal was simple: “If slavery is not wrong, nothing is wrong”.”. This captures that rejection of slavery is, ultimately, a rejection rooted in human sympathy and human values. That’s all there is. It is indeed the case that “nothing is wrong” in the sense of objective moral injunctions that can be derived from facts and that are independent of human values, because that entire conception is misguided and untenable.

Shermer has produced an insightful descriptive account of human psychology. Game theory does indeed underpin how attitudes evolve in a species where social interactions are all important. Rawls’s “veil of ignorance” does indeed capture aspects of how humans think. It is entirely true that this sort of moral reasoning is in line with our evolved nature. But this still grounds morality in human values. The prescription still comes from us, from our evaluation of the sort of society we want to live in.

Nothing in Shermer’s reasoning makes the leap to moral prescriptions that hold independently of what humans think about them, and are thus objective. It is not true that “facts and reason can determine values and morals”; instead, values and morals derive from our evolved human nature. And yes, you can then rationally explain why we evolved to be like that; there is nothing here outside the realm of science, in that there is nothing here that science cannot explain.

Our human nature is not arbitrary, it is fully explained by our evolution as a social species. But Hume’s distinction holds. Moral values are not determined by “facts and reason”, they are instead part of our nature, part of us. That makes them subjective. To many, that label “subjective” seems akin to saying they are second rate or unimportant. But that’s utterly erroneous, in the end our subjective qualia are the only things that are important to us.