Why an “is” can’t entail an “ought”, except why maybe it can
A lot of natural rights libertarian reasoning depends on the possibility of deriving oughts from ises. Rothbard, in The Ethics of Liberty, plays fast and loose with the distinction, making arguments for libertarian moral principles that jump from ises to oughts and back again, with no apparent awareness of the invalidity of the system he builds. Later, Rothbard did notice the issue when he credited Hoppe’s “argumentation ethics” argument with transcending the is/ought distinction, acting as though the distinction was a problem to be cracked, rather than a fundamental feature of the world.
There’s a good reason why ises ought not to entail oughts. The truth or falsity of ises depend on what situation holds in the actual world, while the truth or falsity of oughts depends on what holds in possible worlds. Since the set of possible worlds is infinite, or at least large, and since ought describes a situation that holds universally in the relevant set of possible worlds, the facts of the actual world are unable to constrain the set of possible worlds enough to force the universal statement to be true. Without this forcing, there can’t be entailment.
However, there may be a way in which the actual world constrains possible worlds enough to make an is imply an ought. If a mind is constitutionally compelled to believe a moral proposition, or constitutionally incapable of believing it, then the actual features of that mind — the is — entail that for that mind, the moral proposition is correct or incorrect — the ought.
Why an “is” cannot entail an “ought”
When people talk about oughts, they are usually talking about moral theory. But morality is just one form of modality, that feature of understanding, reasoning and language that deals with possible worlds. English words like should, must, have to, and many others express modality, and are found together with ought, across the divide from is.
The use of these modal words and concepts in the domain of morality is just a proper subset of their use in the larger domain of deontology, meaning rules. Other than morality, other sources of deontological rules include state laws, tribal customs, family rules, general norms of politeness, and the rules of institutional settings such a courtroom or a fraternal lodge. If I tell you that you ought to greet a person when she walks into a room, it probably derives from norms of politeness, but it could also be a formal or informal rule drawn from any of these other domains.
Deontology, in turn, is just one among many domains of modality. Other domains include knowledge and belief (epistemic or doxastic modality), goals and plans (teleological modality), and desires (bouletic modality).
The argument that an is cannot entail an ought comes from the nature of modality. Modality involves possibilities and necessities. Possibility and necessity deal, by their nature, with situations that are not true or false in the real world, but rather in possible worlds bearing some resemblance to the real world. Understanding the meaning of sentences involving modality requires us to hypothesize about classes of non-actual worlds.
Ises, on the other hand, deal with the actual world only. “John is 6 ft tall” is evaluated for truth or falsity by comparing it to John’s actual height in the actual world. “John would be 6 ft tall if he had not been malnourished in his childhood” is a modal statement that cannot be evaluated by reference to the actual world. We evaluate its truth by imagining non-actual worlds which are similar to the actual world, but different in that John had not been malnourished in them. If we believe that John would be 6 ft tall in all such worlds (simplifying), then we believe that the statement is true.
Similarly with moral oughts. “People ought to respect the private property of others” means roughly that in all non-actual worlds consistent with the correct moral principles, people respect the property rights of others. Whether people in the actual world respect the property rights of others does not determine whether they ought to.
Since oughts are universal statements about non-actual worlds, it is not possible to derive truths about them from premises that involve only the actual world. It is the equivalent, in the modal domain, of reasoning from a particular to a universal. That is, it is like reasoning from “John is 6 ft tall” to “everybody is 6 ft tall.”
It is possible, however, to derive a might from an is. From “people have walked on the moon,” it follows logically that “people might have walked on the moon,” leaving aside pragmatic awkwardness. That is because these sentences involve existential quantification and are true if they are true in any world — and this one is true in the actual world. It’s not just the modality that makes the is > ought entailment bad. It’s the strength of the quantification.
Why an “is” might entail an “ought”, after all
Notwithstanding the above, there may be a way for an is to entail an ought. It has to do with the fact that truth statements are relativized to persons, or minds.
A truth is a correspondence between a proposition — the logical form of a linguistic utterance — and a model of the world. A proposition is true if it corresponds to the model. This presupposes the existence of a mind that is capable of comparing the proposition and the model. No truth can be calculated or held outside of a mind that calculates it and holds it.
This could be formalized by adding a parameter to the representation of any proposition, representing the mind that holds the proposition to be true. So instead of “John is 6 feet tall” as believed by Henry, we write “John is 6 feet tall (Henry).”
The mind exists in the actual world, the realm of is. But the mind can believe truths about oughts and other modals that transcend the actual world and engage nonactual possibilities and necessities. The mind’s ability to entertain nonactual worlds while existing in the actual world might be the bridge — or portal? — that allows the derivation of an ought from an is.
It is my view that not much is known about minds. But suppose that through advances in cognitive science, we confirmed what seems right, that the mind is determined or constrained by the physical brain. Suppose further that we came to understand the human brain well enough to be able to know, given the configuration of a physical brain, many of the features of the mind that it is associated with.
Using this knowledge, suppose we came to understand the following about a particular person, who we’ll call Marcie, and her mind, which we’ll call M. M emerges from the structural configuration of Marcie’s brain — we’ll call that configuration C. Suppose we came to understand that due to C, M is constrained to reject, as a matter of moral truth, the proposition “Nobody may initiate force against another,” which we’ll call the NAP.
Thus Marcie’s brain, which is in the actual world, causes M, which is in the actual world, to reject the NAP, a value proposition. From C, which is in the realm of is, we can derive, for Marcie and her mind, a denial of the NAP, which is in the realm of ought.
The following summarizes the argument:
(1) Marcie’s brain is in configuration C.
(2) If a brain is in configuration C, its corresponding mind rejects the NAP.
(3) Marcie’s brain’s corresponding mind is M.
(4) It is false that nobody may initiate force against another (M).
From premises (1) — (3), which express facts about the real world, we can deduce (4), which is about possible worlds.
Nothing about this example limits the is-ought portal to particular or unusual minds. There could be species features of the human brain that restrict it to certain configurations, with the result that humans universally accept or universally reject moral or other modal propositions.
