Libertarianism and Moral Foundations

George: Yes. And listen to this, listen to this; her uncle works for the Yankees and he’s gonna get me a job interview. A front office kind of thing. Assistant to the travelling secretary. A job with the New York Yankees! This has been the dream of my life ever since I was a child, and it’s all happening because I’m completely ignoring every urge towards common sense and good judgment I’ve ever had. This is no longer just some crazy notion. Jerry, this is my religion.

Seinfeld, The Opposite

This post is a brief rejoinder to Professor Andrew Cohen and Julian Sanchez that started with this post at BLH, Against Subjectivism.

First, let me outline my position a bit more formally than the one posted in a comment.

Libertarianism is a social theory, not a moral theory. By this I mean it should only be concerned with personal duties owed to each other but not with deriving what those duties should actually be “a priori.” In other words, it should be concerned with moral violations against one another but not with moral foundations. We can express this in a more technical manner: Libertarianism should be unconcerned with Meta-Ethics.

Recasting this as a Meta-Ethical problem allows for a bit more precise treatment. We then pose this question: Are moral judgements (1) descriptive Statements that are truth-evaluable or (2) Statements that are undecidable(neither true or false, no truth content).

(2) is the non-cognitivist camp. For example, David Hume or Anthony De Jasay. By default, non-cognitivism implies libertarianism should be unconcerned with metaethical foundations. So we scratch this camp from further consideration.

This leaves us with (1), the “cognitivist camp.” I realize that moral cognitivism is sub-divided(just as non-cognitivist camp is) into a litany of different branches/positions, but the basic position held by all is: moral judgements are descriptive Statements that are truth-evaluable.

Given (1), a cognitive meta-ethic framework can then be expressed as formal language, with each symbol of this language mappable to a Gödel number. This allows any formal language F to then be subject to the rules of arithmetic and deductive logic. And thus, subject to the famous Gödel’s incompleteness theorems:

(i) one can construct a sentence G which that is not decidable from F
(ii) any statement C which expresses the consistency of F will be undecidable.
Alternatively, F includes a statement C if and only if F is inconsistent.

(i) is more or less the “liars paradox.” (ii) can be thought of as the “Halting problem”

Essentially, Gödel’s theorems demonstrate that the consistency of F cannot be proved in F. Note: Gödel’s incompleteness theorems have nothing to say about “objective reality.” I quite realize this. Nor do they imply that G is not provable in another formal language, say, F1. What they do imply is this:

Principle I:
If moral judgements are truth-evaluable descriptive Statements, no single Moral Foundation, or cognitive meta-ethic Formal Language F, as a formal axiomatic basis to derive these descriptive Statements, can prove it’s own consistency.

Now, I’m going to establish another principle based on Löb’s Theorem:

Consider the below Sentence:

“If this Sentence is True, then Objective Morality Exists”

I’ve just logically demonstrated that Objective Morality Exists. If you are having a problem seeing this, refer to this explanation.


“If this Sentence is True, then Objective Morality Does Not Exist”

I’ve just logically demonstrated that Objective Morality does not exist.


“If this Sentence is True, then God Exists”

I’ve just logically demonstrated that God exists.


“If this Sentence is True, then Elvis is still alive”

I’ve just logically demonstrated that Elvis is not dead.

The point being demonstrated is that logical truth may having nothing do with “objective reality” and that “truth” should be differentiated from “proof.” Informal reasoning combined with self-reference can “prove” anything.

This leads to another principle:

Principle II:
Any Formal System F that asserts it’s own consistency can prove anything. In other words, no Formal System F can trust itself.

For a more thorough discussion regarding Löb’s Theorem, refer to this informative post by Eliezer Yudkowsky.

In a moment I will return to what I think the two principles above reasonably imply, but for now let me address Sanchez and Cohen.

Julian Sanchez

Is Subjectivism in Ethics Coherent?

Sanchez, in his piece–as an attempt to avoid the “incoherence” of moral non-cognitivism–makes the cases for a “coherent” ethical subjectivism that nonetheless allows at least one objective moral principle. Sanchez proposes the below as an objective principle:

It is wrong to act in ways that you believe to be wrong

The first thing I thought about when I read this was the Seinfeld episode, “The Opposite.” This episode presents us with a moral agent, George Costanza, that realizes he is own worst enemy. Thusly he concludes to use his own moral judgements as the basis of a principle of action to do the opposite. Yes, George Costanza is fictional character, but Larry David is real enough. We can easily see then that Sanchez’s principle is not necessarily an “objective principle.”

The second objection would refer to my second principle: using “self-reference” as the basis of a “truth” when you actually haven’t proved anything.

Andrew Cohen

I declare that I now believe in the “Purple Invisible Sky Spirit.” Part of the moral foundations that come with believing in PISS is that all believers are allowed to seize the property and kill anyone wearing the color purple because that color is allowed to only be donned by PISS.

Now does the introduction of PISS create a dilemma in “moral subjectivity,” or I have I invented PISS to logically demonstrate objective morality. After all, PISS objectively exists or it doesn’t exist. To quote Cohen, but substituting in PISS for God.

It is either the case that PISS exists or it is the case that PISS does not. One of those is the objective truth. If PISS exists, PISS-Deniers are wrong; if he does not, then PISS-Believers are wrong. There is an objective fact of the matter even if we do not know what that objective truth is. Our lack of knowledge of the truth has no effect on the truth whatsoever.

The obvious flaw here is conflating “truth” with “provable.” The proposition P=”God Exists” is not provable. It therefore, as a logical truth, is irrelevant as part of a logical deduction to establish “objective truth.” You are not proving anything. It is this constraint, of course, that also prevents my invention of PISS from proving anything.


The two principles I have outlined, of course, say nothing about “objective morality,” moral realism, moral subjectivism, etc. And they only apply to the cognitive assumption that moral judgements are descriptive Statements that are truth-evaluable. But what they do imply is that no single Formal Language F, as a formal axiomatic basis to derive these descriptive Statements, can prove it’s own consistency. And to claim such a consistency means that it can prove anything.

From those principles, therefore, a reasonable case can be made that if you are concerned with “liberty,” Libertarianism should not be a moral theory.

2 thoughts on “Libertarianism and Moral Foundations

  1. I think you’re misreading Cohen’s point a bit. My understanding was that he was saying the fact that one cannot prove the existence of PISS does not then mean that PISS is not real. In other words he’s saying that “truth” and “provable” or NOT related, as a preemptive hedge for being asked for proof of his assertion that morality is objective.

    A dialog might go as follows:

    Cohen: Morality is objective.

    Skeptic: Can you provide any proof of this?

    Cohen: No. But since something can be both true and unprovable at once, that lack of proof cannot be counted as an argument that my assertion is incorrect.

    1. Thanx for the comment…

      Actually, I thought my post demonstrated clearly the possible absurdities of when “truth” and “provable” are not differentiated. My interpretation is that he didn’t at all differentiate between the 2 concepts.

      Then again, given the extent of the commentary, I’m not sure anyone is actually sure what he is trying to logically demonstrate. In this sense, he should perhaps resort to recasting his theorem in symbolic logic.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

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

Google photo

You are commenting using your Google 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 )

Connecting to %s