“And if [cannibals existed], we’d be morally obligated to kill or imprison them.” That is a possible conclusion if you can find objective moral truths, but you cannot use that conclusion to help you prove those truths.

Right. That is a proposition to be proved. It follows from what we want (not to be eaten or have people we care about eaten) and what we have to do to get what we want (police cannibals).

It seems to me that your method is something like this:
1) Find out what people really “should want”, given perfect knowledge and reasoning.

More exactly, what they will want when.

In other words, you can avoid the circularity of normative language (“should”). The normative (“should”) is simply a restatement of “what you will do when rational and informed.”

The meta-question exists as to why you “should” do now what you will do when rational and informed (and thus of what you will do when irrational and uninformed, but have true beliefs at least of what you would do if you weren’t irrational and uninformed), but that resolves the same way (hence note 36, pp. 426-27 in TEC). We simply define as insane people who prefer to be irrational and uninformed (and deal with them accordingly). And if we aren’t them, what they would want is irrelevant to what we should do; what we should do derives from what we want. And if it’s not to act irrationally and uninformedly, it’s to act rationally and informedly. QED.

2) Hypothesize that everyone or nearly everyone would want the same things.

Not just that. Any more than in any other science. A mere hypothesis is just a hypothesis. It needs to be tested to become a scientifically confirmed theory.

And this hypothesis is testable; and testing it will be a component of any fully developed moral science. As explained in TEC, pp. 351-56.

3) But if they don’t, declare them to be insane if at all possible, so that you can ignore them.

No. Moral science could find a coherent system of moral relativism true (as laid out in TEC, pp. 347-51). So the options are not “universal morality” or “insanity.” Empirically it may be, but that would have to be discovered. It can’t be known in advance (for humans anyway).

Insanity only gets declared when someone is implacably preferring to be irrational and misinformed, even to the danger of themselves and others around them. And that’s not me saying this. That’s the entire definition of insanity in the diagnostic standards of psychological science.

4) And if that is not possible (eating animals or not?), declare the issue to not be a moral issue.

Rather, if we empirically look, and empirically find no imperative is true in the matter, then by definition no imperative is true in the matter (to whatever probability of certainty our method entails). That is a possible outcome. But it can’t be declared from the armchair. It has to follow from evidence. In other words, it has to actually be true. And known (more or less) to be true.

I hypothesize that everyone will want different things…

Then you need to engage with my fact-based case to the contrary in TEC: pp. 351-56.

You can’t know this in advance any more than you can “know” faeries exist. If you don’t look, and look by a methodologically sound way, you don’t know. I present the scientific evidence that already makes a prima facie case against you. And a proper moral science could shore that up (or refute it) simply by engaging a more rigorous factual investigation than I already have. But if you want to gainsay the prima facie case, you have to actually read it, and address it. With facts. Because only facts can counter facts.

But I also see another problem, which is that I’m not sure how your hypothesis could be falsified.

A definition can’t be falsified. Definitions are tautologies. They are always necessarily true.

You are confusing a definition of what we are looking for (“what you will do above else when not in error”), with what we will find when we look (“what actually is it that you will do above all else when not in error?”)

The latter entails only falsifiable hypotheses. That’s what makes it science.

If we did our best to find out people’s “true” desires, but we still find that they diverge, we can always resort to the possibility that if we had just done a better job at finding the relevant knowledge or reasoning, we would get a convergent morality.

If that’s true, it’s true.

This is already the case. For example, in Young Earth Creationism vs. the facts of the earth’s age. We can do our best to find out what people will believe when reasoning informedly and rationally (that the earth is billions of years old). And that is what we decide to call the scientific truth of the matter. But that doesn’t make everyone believe that. “We still find that [YECs] diverge” and we are forced to conclude that “if we had just done a better job at finding the relevant knowledge or reasoning” they’d change their mind. Does that make them right? Does that make it not true that the earth is billions of years old? Does what they think have anything to do with what’s true?

Moral science is no different than the rest of science here.

So if this is a problem for moral science, it’s a problem for all scientific truth of any kind whatever.

A third problem is that even examining this seems to be practically impossible without introducing certain moral assumptions into the process.

No, it does not.

You are doing that, by making assumptions about what the results will be (“surely we can’t approve of cannibalism!”). And then using those unevidenced assumptions as “evidence” against what I’m saying may be morally true. This is a common philosophical error: always philosophers will say some metaethical theory must be false because it gets results they don’t like and are thus “sure” can’t be moral. But how do they know? Maybe they are the ones who are wrong about what is and isn’t moral. If you don’t look, you don’t know. Hence see TEC, pp. 343-47, “The Moral Worry.”

I’m the one who isn’t making assumptions about what science will find. I have evidence-based guesses, my own hypotheses I can make prima facie empirical cases for, but I admit the facts may turn out differently on a rigorous inquiry than they appear on present inquiry. And that is precisely in fact why we need morality to become a science.

So this is a problem not with the moral science I propose. To the contrary, it is precisely the problem my proposed moral science would solve.

Sociopathy/psychopathy are not classified as mental illnesses…

Yes they are.

…they are colloquial terms used to describe certain character traits (lack of remorse, lack of empathy etc)…

Actually, sociopathy’s primary pathology is diminished fear response. Which causes all the other symptoms (and those symptoms are used to diagnose the disorder; among them are symptoms exhibiting an absent fear response). I cite the science on this in SAG, p. 344. More has been published since.

A moral system that assumes that we can be or would want to be machines would be useless, because we are not and never could be.

We actually are machines. So you seem to be sneaking some other assumption in here (some particular “kind” of machine that you are arbitrarily deeming to be “better” than some other).

I want to make this empirical, and you keep sneaking in unempirical, unexamined, undefended assumptions—can you not see why you have the wrong end of the stick here? Your entire methodology is not defensible.

The only way to talk about what is true, is to talk about the facts. There is no more reliable access to the facts than science.

So why would you be against more science on the facts that determine what’s true about morality?

Broadly, relativist morality is not morality at all in any meaningful sense…

All philosophy disagrees with you.

And here you seem to have some talismanic superstitious belief about the word “moral.”

And I’d like to take that talisman away so we can talk about what’s true instead.

So I heretofore ban you from ever using the word “moral” or its cognates anywhere further in this conversation. I will do the same.

We will heretofore only talk about “zoral” and cognates, which consists entirely and solely of what you ought to do above all else.

That’s the only thing that matters. Whatever you think the word “moral” means, it is either that, or it is not that. And if it is not that, then “we ought to do above all else” what isn’t moral. In which case what is or isn’t moral is irrelevant to this entire conversation and to the entire science I propose.

(Hopefully this will start to jog your mind into realizing what’s actually going on here; and what we have actually been talking about from beginning to end.)

Richard Carrier wrote:
“It’s called moral relativism.”

Aha! That is why we have in some ways been talking past each other. That is of course, as you know, a path that most philosophers have rejected as hopeless, which is not to say you cannot try, but that is why I did not expect you to take that position.

“And if [cannibals existed], we’d be morally obligated to kill or imprison them.”

That is a possible conclusion if you can find objective moral truths, but you cannot use that conclusion to help you prove those truths.

It seems to me that your method is something like this:
1) Find out what people really “should want”, given perfect knowledge and reasoning.
2) Hypothesize that everyone or nearly everyone would want the same things.
3) But if they don’t, declare them to be insane if at all possible, so that you can ignore them.
4) And if that is not possible (eating animals or not?), declare the issue to not be a moral issue.

I hypothesize that everyone will want different things, even given perfect knowledge and reasoning, and that even though these differences will often be subtle, there will be no way to find a pareto efficient morality that can cover even a majority.

But I also see another problem, which is that I’m not sure how your hypothesis could be falsified. If we did our best to find out people’s “true” desires, but we still find that they diverge, we can always resort to the possibility that if we had just done a better job at finding the relevant knowledge or reasoning, we would get a convergent morality.

A third problem is that even examining this seems to be practically impossible without introducing certain moral assumptions into the process. That could still lead to useful results, but then we no longer have a system based on objective moral truth, but rather on some moral axioms (which is a common and useful approach).

“You argued sociopathy was adaptive.”
I did not, I argued it was not *necessarily* maladaptive. Sociopathy/psychopathy are not classified as mental illnesses, they are colloquial terms used to describe certain character traits (lack of remorse, lack of empathy etc), which in extreme cases are associated with diagnosed mental illnesses. However, this is not a fundamental problem for a system based on moral relativism (although that brings on other, probably fatal problems).

“Risk management is a function of moral reasoning.”
It is certainly not only a function of reasoning, but also of temperament. A moral system that assumes that we can be or would want to be machines would be useless, because we are not and never could be.

“The only way to get a moral system that is relevantly true, is to get one that you indeed will follow, when you aren’t reasoning fallaciously or misinformedly. There is no other way to get anything to be true about morality.

Can you think of any other truth conditions by which you could say a moral system is true, and should (in actual fact) be obeyed?”

No, and I do not believe there is such a thing as objective moral truth. Broadly, relativist morality is not morality at all in any meaningful sense, it is not possible to find a universal morality based on human traits that would also be objective, and there is no possible method to examine morality that does not involve humans.

Rather than objective truth, I believe we can only strive to find something as close as possible to a consensus. It cannot ever be objectively true or final, but if the consensus-making process is based on openness of discussion and a spirit of free inquiry, we can hopefully avoid it becoming a “might makes right” or “tyranny of the majority”. Hopefully as in what I personally prefer, and what I am hopeful can be widely agreed on.

“If racism increases violence[..]”
Then that is a drawback for anyone who desires decreased violence, but for some individuals that drawback could (and, per my hypothesis, probably would) be overcome by other advantages.

Yep.

Sociology and psychology and political science have all been doing things like that, too.

But they haven’t developed much in the way of methods yet, for finding out things like what an ideal agent would want once they are cured of irrationality and ignorance. For instance, an economist doesn’t need to know that; they only need to know what people actually want, even if it’s totally irrational and ill-informed.

Thus, it’s a descriptive science (it can make predictions about systems) but it’s not a prescriptive science (it doesn’t even ask what people should want).

The prescriptive science would be more difficult, and requires somewhat different methods. But it certainly can benefit from interdisciplinary knowledge (methodological and factual) already gained in these other sciences.

No, because I do not believe that it is possible to link individual “ideal” preferences to objective moral truth.

That doesn’t make sense, though. If you don’t think you ought most to do what you would do when not fallacious or misinformed, then you are saying you believe you should act illogically and uninformedly. And I doubt that’s what you really believe. And if you don’t think what you ought most to do when not fallacious or misinformed is moral, then you are not using “moral” in any sense capable of being relevantly true.

Remember, we are trying to get at what is true. And that limits your options. We can invent countless moral systems. And then fail to show anyone should follow them. We cannot then claim any of them are true.

The only way to get a moral system that is relevantly true, is to get one that you indeed will follow, when you aren’t reasoning fallaciously or misinformedly. There is no other way to get anything to be true about morality.

Can you think of any other truth conditions by which you could say a moral system is true, and should (in actual fact) be obeyed?

(This is a trick question. Because it’s logically impossible to answer. But maybe you have to try, in order to discover why.)

That depends on their preferences.

The effect of racism on the social system we all have to live in is not a function of preferences. You can’t “desire” the world be a certain way, and expect it to be that way. If racism increases violence, then that increases your risk of being a victim or increasing costs on you to police that violence (for example; I’m not saying this is the only effect of racism on everyone, I’m just picking one). What you “prefer” can have no effect on that fact of the world. If it’s a fact, it’s a fact.

Thus the only way “preferences” can matter here is if you are seriously arguing that there is someone who, even when completely rational (reasoning without a single fallacy) and fully informed (knowing all true things and being ignorant of nothing and having no false beliefs about anything), prefers a society that is more expensive and dangerous for them.

I would propose that empirically you will never find such a person.

That’s our hypothesis.

And even if you do find such a person, we will be morally obligated to restrain them to save the rest of us from them. (Which fact being a fact makes it hard to continue to think they’d still agree they should persist in that preference—because, as stipulated, they know this will be the outcome.)

Oh no. I meant founder: as a ship, to strike bottom, sink, fail.

Inconsistent? What’s inconsistent about moral facts being true because of the desired outcomes?

You aren’t making any sense.

If what is morally true is true because of the consequences of the agent’s actions, that’s consequentialism. No sham formalization can get you out of that fact.

Since all choice is driven by a desire for an outcome, all outcome-driven decisions are consequentialism.

No, inconsistent moral reasoning is not ethics. Consequentialism must be formalizable and consistent if you want it to be an ethics, otherwise it’s vacuous as I said.

“Outcome-driven decision” is not a meaningful formalization.

If the “moral facts” would have to be true only for a certain individual, and could be different for each individual, most philosophers would probably deny that they would be “moral” facts at all.

To the contrary. It’s called moral relativism. A well established theory in moral philosophy. (As is ethical egoism.)

Facts are facts. Anytime it is factually true you ought to do something, that’s a fact. And if it’s true you ought most to do a thing (above all other things), that’s by definition what morality is true for you. (Your opinion cannot trump it. You can have a false belief about what you ought to do.)

That said, note, it’s unlikely there will be a different morality for every individual. That’s simply a logical possibility. It’s even unlikely that there will be a different morality for different human sub-groups (as I explain over several pages in TEC). But “unlikely” is not “impossible.” Thus, this is still an empirical question.

Perhaps, someone has an abnormal desire to eat human flesh, a desire so strong that he is rationally justified in risking a life sentence to satisfy that desire.

Logical possibility doesn’t get us anywhere. If it turns out to be true, it’s true. What you don’t like has nothing to do with it. I have a whole section in TEC titled “Cave Man Say Science Scary” that explains why this cannot be an objection. For all you know, cannibalism is moral. You don’t know otherwise, until you check.

You can’t decide from the armchair what’s true about the world. You have to go out and actually look and find out what’s true about the world.

This is as true of morality as of any other fact of life.

You need to read the chapter in TEC. Peer review already covered objections like yours. They are already answered there. With answers that passed the peer review of multiple professors of philosophy.

(But just FYI, even apart from the semantics, almost certainly no such person could exist. And if they did, we’d be morally obligated to kill or imprison them. This is just another iteration of the sociopath problem.)

The mortality rate for exceptionally tall people is also extreme…

You argued sociopathy was adaptive. I showed you the evidence is to the contrary. Yes, extreme height is maladaptive. So is sociopathy. Both are classified as a disorder. For that very reason.

But what classifies sociopathy as a mental illness is that it is maladaptive even for the individual’s reasoning (they are not happy, act self-defeatingly, and cannot change their behavior even after admitting both facts: that is what defines a mental illness). Height is not a mental property and therefore cannot be a mental illness.

(And BTW, yes, early death is by definition maladaptive. It’s a standard function in differential reproductive success calculations in evolution theory.)

Researchers who have considered the broader aspects of psychopathy have theorized it as a parasitic behavior to game the surrounding moral system. It should also be noted that as an evolutionary strategy it must be considered as high risk/high reward.

Risk management is a function of moral reasoning. If doing X gains the same satisfaction output as Y but at a higher risk of death, you ought to do Y.

Of course. But we also have lots of documentation of non-sociopaths engaging in clearly inhumane and cruel behavior, evidently caused by the desire to follow the prevailing morality of the surrounding society.

Not relevant to your point.

Harris and I fully accept the existence of irrational and ignorant people, who thereby follow false moral systems.

The influence of morality on an individual is not in any way conditioned on that morality being in any sense objectively true.

That’s true of all knowledge. Even beliefs about the age and shape of the earth. And that being the case, has nothing to do with what is nevertheless true.

Even a morality that no one believes to be objectively true has these effects. Therefore, the existence of such an effect cannot be used to demonstrate that there is such a thing as objective moral truth.

I don’t know what you are talking about with this statement. You seem to have lost track of the argument. You said X proves Y. I showed X does not prove Y. That in no way means I said either X or ~X proves Z.