So as I understand it, normative ethics asked questions like, "what should I do?" whereas metaethics asks... "what makes something good or bad."

What I'm confused about is... if someone's metaethics is moral antirealism... ie: they say, "there are no moral facts". Then how can they make normative ethical claims, like "torture is wrong" etc? Seeing claims by moral antirealists that they can still make these claims while denying there are moral facts. I can't make sense of this.

I mean if there are no moral facts, I would conclude there is nothing in particular that I should do or should not do. How can I make the claim that there are no moral facts and at the same time say, "A person should do this or that". What does it mean to say a person "should" or "should not" do something if there are no moral facts?

  • 4
    There is no "moral antirealist" position. Moral antirealism covers a collection of incompatible views that share the position that moral statements aren't objective statements of the universe. As such there is no single answer to this question. Commented Apr 8 at 8:02
  • 3
    That there are no objective normative facts does not imply that there is nothing in particular that you should do. You should not stick your hand into fire, objective normative facts are not needed for that. Moral anti-realists hold that all moral claims are of such sort, they are implicitly conditioned on some idea of human well-being, or emotions, or desires, or threats of consequences, or all of the above. One does not even have to be anti-realist to deny objective moral facts, they can think them real but socially constructed.
    – Conifold
    Commented Apr 8 at 9:21
  • @Conifold So for the moral antirealist, all 'should' statements are conditionals? Commented Apr 8 at 23:43
  • 1
    That is one form of it, error theory. In another form, noncognitivism, they are not even statements, just speech acts expressing feelings, approval/disapproval, attitudes, veiled commands, etc. Of course, almost all of those are, at least implicitly, conditional on context:"don't do that (in such and such circumstances)".
    – Conifold
    Commented Apr 8 at 23:53

4 Answers 4


How can a moral antirealist make moral normative claims?

When I say or hear "X is wrong" or "people shouldn't do Y", etc., I generally understand these as abbreviations for "I feel X is wrong, etc.", "we believe people shouldn't do Y", "I have an intuition that X is wrong", "our communal convention is that people shouldn't do Y", that is, really as expressions of personal or communal preferences, broadly construed, and not really as claims of existence of moral truths. [ I've never ever found any example of these whatsoever, by the way, but if you have one, please share :) ]

What does it mean to say a person "should" or "should not" do something if there are no moral facts?

Generally it means something along the lines "If you want to avoid going to prison/being hit over the head then you shouldn't do X"

[tangential edit: maybe part of the difficulties/confusion comes from our use of words as 'right' and 'wrong', which are also used in descriptive contexts, or maybe their use in normative contexts is imported from the descriptive use, via 'naïve moral realism'... it's complicated]

  • but does it also mean "X is wrong"? i don't think non-cognitivism makes any difference to that, but i could be mistaken. i think that icecream tastes good and my wife is pretty, even supposing i cannot justify that without reports of my internal states
    – andrós
    Commented Apr 8 at 1:58
  • 1
    well, i'm saying this is what "what 'X is wrong' mean" means
    – ac15
    Commented Apr 8 at 2:08

There is no inconsistency there. You can deny the existence of absolute moral principles, yet recognise that man-made moral rules are useful and should be respected. Indeed, I could make an argument that the absence of absolute moral facts means that people can reasonably take conflicting positions about moral questions. Consider your example of torture. Many people might take the view that torture is wrong, presumably because they find it repugnant and would not like to be subjected to it. However, there are people who argue that torture can be justified by circumstances. Neither position requires any appeal to some absolute moral standard, and whether one is more preferable than the other is ultimately decided by society. It should also be clear to you that what society deems to be morally acceptable is as variable as society itself.

  • 4
    "that man-made moral rules are useful and should be respected"... But if it 'should be respected' that's a moral fact right? If it's not a moral fact that it 'should be respected'... then should it be respected? Commented Apr 8 at 9:40
  • @AmeetSharma not really- it's just what people might think. For example, in the UK we think homosexuality is fine, whereas in some countries it's punishable by the death penalty. You can't really say that the acceptability of homosexuality is a moral fact in the UK but not a moral fact in Saudi Arabia. Commented Apr 8 at 11:04
  • @AmeetSharma no, that's an Axiom. Aka something assumed to be true in further considerations (in that scope of reason). It's a manufactured starting point of reason.
    – Hobbamok
    Commented Apr 8 at 14:02

One way is to distinguish between, "It is a fact that..." and, "It is true that..." or then to differentiate between forms of truth (one might invoke the notion of multiple axiomatic theories of truth and claim that the truth operator obeys certain rules with respect to moral discourse/language that it doesn't obey with respect to non-moral talk). This discrepancy would run parallel to e.g. the substance-attribute realism vs. truth-value realism dialectic in the philosophy of mathematics. So the phrase "moral antirealism" is ambiguous enough that some examples of the general thesis are consistent with making normative claims and some examples aren't.

Or, then, from a fictionalist point of view, moral antirealism can be construed along the lines of being willing to say e.g. "Linus waits for the Great Pumpkin," without believing in Linus or the Great Pumpkin as factual objects. Likewise, that is, one might think that the moral narrative in which one participates is such that, "It is good to be kind," is true, or we might say quasi-true, in that narrative,K without being committed thereby to goodness as a property of an object, or an object itself so much, or whatever. (Granted, the concept of an object per se is itself dramatically ambiguous, and fact-talk can be abstract enough to not be much different from abstract truth-talk, so at some point we might find that a so-called moral antirealist is so only on account of peculiarly limited definitions of facts or truth.)

KKant, however, was what we might call a transcendental fictionalist insofar as we can interpret talk of the pure will with talk of a pure story being told by, and through, this will. A transcendental fictionalist about ethics thinks that there are universal, necessary preconditions for the possibility of moral stories, or stories that we use the word "moral" for, such that actions expressive of those preconditions are the ones categorically recommended by the form of the moral story regardless of how that story's content has been cashed out. (For such cashings-out, see Kant's Religion within the Limits of Reason Alone, for example.) One author, Tamar Schapiro, conceives of such a thing in terms of a "generic practice" that the linked-to SEP article goes on to call by the name of "Intendo." Now, I don't know enough about Schapiro's full moral outlook to claim that she is herself strictly a transcendental fictionalist, so I cite her, here, for illustrative more than demonstrative purposes, but I should like to have conveyed the spirit of transcendental fictionalism on terms such as hers, since the interplay between the concept of a story and the concept of a game can be in the form of games that involve the telling of stories (e.g. as with video games) and so we can find another mathematical parallel (between game-theoretic formalism in the philosophy of mathematics, and fictionalism both there and in the moral realm, too).


words don't have meanings, only effects. the words 'x is wrong' don't correspond to some moral fact even if moral facts do indeed exist. i say 'x is wrong' about some x that i personally dislike because of the effect it has on people i'm saying it to, ie. that they might be less likely to do x, and on myself, because we tend to enjoy feeling that we're expressing our feelings. if anti-realism is true then this is the case for everyone, anti-realists are just self-aware about it.

  • Interesting ideas - could you neaten up the capitalization though? Thank you ~ Commented Apr 10 at 3:47
  • @JuliusHamilton i actually write in lowercase on purpose (having to do with my somewhat fraught relationship with language)
    – Silver
    Commented Apr 10 at 14:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .