The Basic Battle Lines of Metaethics (with complaints about mistaken appropriations of philosophical language) by Christian Tarsney

After I posted an irritated Facebook status about debaters’ misuse of the term “truth-functional,” Adam suggested that I write something for VBD to explain the terminological issue. The mistake is one worth rectifying, I think, if nothing else so that debaters will make themselves a greater credit to the activity when they talk to their philosophy professors in college. But it seems a bit unfair to clutter the debate webs with my own minor grievances, without throwing in something more constructive as well, so I thought I’d pair the clarification of terms with a little elucidation of the cognitivist/non-cognitivist distinction in metaethics which debaters mistakenly take “truth-functionality” to trace. This latter is something which many debaters grasp fairly well, but for those just starting to learn metaethics, I’ve found that it can be a little obscure, and easily confused with the realism/anti-realism distinction. Hopefully, then, this article will shed a little light on these basic dividing lines in metaethics, along with urging a change of expressions.

So, first, the terminology: I’m not sure where the misuse of “truth-functional” came from in the debate context (although at least one instance of the same misuse has been pointed out to me in academic literature), but it’s been around now for quite a few years. Debaters tend to use “truth-functional” to mean either “capable of being true or false,” or else the stronger “at least sometimes true.” The latter use is most common with reference to skeptical arguments—I’ve seen several debaters characterize the view that there are no moral facts/truths as claiming that “morality is not truth-functional.”

The real meaning of the term is quite a bit more esoteric—“truth-functionality” is a logician’s term of art, which refers to property of certain logical operators like “and,” “or,” “not,” “if…then,” “if and only if” (and others which don’t neatly correspond to any expression of ordinary English). A connective of this sort is truth-functional to the extent that the truth value of a sentence it governs is a function of (i.e., is fully determined by) the truth values of the propositions it governs/connects. “And,” for instance, is (ordinarily, to a good first approximation) a truth-functional connective, in that to know the truth value of a sentence of the form “P and Q,” it is sufficient to know the truth values of P and Q (if they’re both true, the conjunction of them is true; if at least one of them is false, it’s false). “Because,” on the other hand, is not a truth-functional connective, in that the truth of a sentence “P because Q” depends on more than the truth values of P and Q (“I passed the exam because I studied” vs. “I passed the exam because Paris is the capital of France”—the component parts of each sentence might all be true, while the sentences themselves take different truth values). I can’t stress enough that this property of sentential connectives has no conceivable relevance to any debate round, ever.

Now, with respect to the distinction which debaters are trying to express: philosophers use the term “truth-apt” (hey, it’s more syllable-economic!) to describe the property of taking a true value—being, at least potentially, either true or false. A sentence like “The cat is on the mat” seems to be truth-apt because it has truth-makers—matters of fact which can determine its truth or falsity (it’s open to dispute whether any truth-apt sentence must have truth-makers, but it covers most ordinary cases). On the other hand, questions (“Is it going to rain today?”), commands (“Do rebuttal redos.”), and exclamations (“Hurray!”) seem not to be truth-apt—they carry meaning, but it would seem odd to describe them as true or false.

Cognitivism is the metaethical view that moral utterances are truth-apt; non-cognitivism is any view which denies this. Most ordinary first-order moral theories (deontology, utilitarianism, virtue ethics) are cognitivist—they hold that at least some moral utterances express propositions which can correspond to the way things are, or fail to correspond. Non-cognitivist views, on the other hand, tend to compare moral utterances to one of the more widely recognized forms of truth-inapt utterance: Emotivists and expressivists argue that moral utterances are like exclamations (so “You ought to tell the truth” just means something like “Truth-telling—hurray!”); prescriptivists argue that they’re like commands (so “You ought to tell the truth” is just a roundabout way of saying “Tell the truth!”).

It’s worth noting that the cognitivist/non-cognitivist distinction can get a little fuzzy around the edges, and that some versions of emotivism and expressivism will permit a certain kind of truth-aptness in moral discourse. It’s also worth noting that non-cognitivism is not a form of skepticism, from a philosophical standpoint—at least, most non-cognitivists don’t take themselves to be moral skeptics. In a debate context, however, if it’s accepted that the affirmative has to argue for the truth of the resolution, non-cognitivism can come out functionally equivalent to skepticism. The solution here, I think, is to adopt an interpretation of “affirming” and “negating” which is meta-ethic relative—if cognitivism is correct, then to affirm an ethical utterance is to declare it true; if non-cognitivism is correct, the affirming might mean something like “endorsing” or “enjoining.”

Finally, in contrast to the cognitivist/non-cognitivist distinction, the divide between realists and anti-realists has to do, not with the status of moral utterances, but with the sorts of moral phenomena that are out there in the world. A moral realist thinks that there is something like moral facts, moral properties, rightness, wrongness, objective value which moral language tries to get at. An anti-realist denies this. Someone who is a cognitivist and an anti-realist will (generally, but not with absolute necessity) be an error theorist—in other words, will say that moral utterances are never true (but are sometimes or always false), since they assert that an action has a moral property which, in fact, no action possesses. This is the second dividing line which debaters have sometimes mistakenly described in terms of truth-functionality. While there is no easy philosophical term to describe the propositions of a discourse being sometimes true, the view that they are is often referred to as a “success theory,” in contrast to an error theory.

Hopefully this brings a little bit of clarity to these two metaethical questions, and the language used to describe them.  Cognitivism/non-cognitivism and realism/anti-realism are generally seen as the two most fundamental divisions in value theory, so understanding and distinguishing them is crucial to navigating the terrain of metaethical views now entering into the debate world. Correct use of philosophical language, in addition, is valuable not just for aesthetic reasons, but also for one’s ability to read the philosophical literature which takes these terms for granted. So go forth, speak precisely, and best of luck to everyone gearing up for state tournaments, NFL qualifiers, and the TOC!

  • Excellent post

  • This was really helpful. I'm pretty sure I've used "truth-functional" incorrectly before in at least one debate round.