Philosophy is littered with realisms and skepticisms. Meta-ethics is no different. Moral skepticisms abound and there are many moral realisms. It’s a risk of excess to add one more to the pile, but hopefully the risk is worth it. This essay charts a middle course that could just as easily be called Quasi-realism, a more glass half-full name. That is, if Simon Blackburn hadn’t already coined the term to explain his own midway position (the title of this essay is a blatant ripoff of Blackburn’s coinage). In the end, though, the view outlined here should be considered a form of moral skepticism. Whether it will depends on the community’s ultimate judgement on where the lines should be drawn and what names should go on which side.
One matter of housekeeping to begin: the terms “skepticism” and “anti-realism” are sometimes treated slightly differently and sometimes interchangeably. This piece uses the terms as closely synonymous, even interchangeable. The Stanford Encyclopedia of Philosophy (SEP) uses moral anti-realism to capture the family of specific challenges to moral realism like expressivism, error theory, etc. Moral skepticism, says the SEP, is a more general term and so includes multiple reasons why philosophers might doubt moral knowledge, including but not limited to objections lodged by moral anti-realists proper, like expressivists and error theorists. Finally, “Quasi-Moral Anti-Realism” has less of a ring to it, so Quasi-Moral Skepticism will do for now.
The idea is that an important part of moral discourse cannot be preserved by scientific, rational, or philosophical treatment. The “quasi” part consists in the view’s defense of only one significant point of contention skeptics make, while granting the rest to the realists. The significant point of contention is skepticism toward “categoricity,” a concept touched on in a previous piece. Categoricity is the intrinsic bindingness moral claims come with, a kind of ultimate obligation from which there is no escape, save a small bucket of exceptions like mental incompetence, childhood, etc.
Other factors realists want to preserve include the irreducibility of rational and normative action guidance in general. The hypothetical imperative involves action guidance through reasoning familiar in “IF you want to become a champion chess player, THEN you should practice chess.” Some philosophers have pointed out that the means-end fit in this kind of action guidance isn’t warranted by raw observation, so involves a normative interaction with the world. The action guidance is an additional take-away or projection, more than the objects of the world alone compel.
The same is true, and then some, for intellectual values like parsimony, elegance, and epistemic justification. These intellectual considerations provide a kind of normative action guidance that isn’t fully accounted for by the sum of its parts. If epistemic justification (which supports belief in scientific truth), prudential advice and moral realism all come in a package, then we should hesitate before rejecting moral realism. Not only would we lose our connection to the real world and render our otherwise ordered lives incoherent, but it would be a kind of performative contradiction. We’d be using intellectual reasons to reject moral realism while undermining all support for confidence in intellectual reasons in the first place.
So, if hypothetical (prudential) and intellectual reasons are irreducible, and the objection to moral reasons is that they can’t be reduced to something more fundamental, then we’re in trouble. But this piece denies that means-end “oughts” or intellectual values require ultimately overriding reasons. In other words, a kind of normative action-guidance can be preserved even if categorical reasons are discarded – a normative but not a moral realism, if you will.
Irreducibility and categoricity are often treated as the same concept in the blunderbuss of meta-ethical argument, but they are severable. To make this kind of pulling apart explicit, it’s helpful to borrow an insight from 20th century philosophers of language. In the course of describing key linguistic functions, they realized that previous attempts to characterize language as referring to things in the world had left out promises, exclamations, commands, and much more. Similarly, we don’t have to treat all of ethical language as doing the same thing.
Maybe your neighbor truly is a kind person, a thick moral assessment answering to whether events in the world justify the description. Furthermore, without the larger concept of “good” to give it substance, it would be an unmotivated and empty statement (it would make no sense to say your neighbor is a deeply kind and bad person). Perhaps statements like this can survive without categoricity, the super strong bind that requires a morally good reason to virtually always override a worse or non-moral reason, as if the selfish hedonist is irrational or unhappy, by definition.
The rational but selfish hedonist is a good example for pushing back against attempts to treat categoricity as real. The point regarding thick moral concepts was made by Hilary Putnam – “kindness” grounded by real world events but still requiring the larger concept of goodness to make it sensical. Putnam was a virtuosic defender of a moral realist approach without strong metaphysical commitments. The cause has been taken up by more ham-fisted advocates, like Sam Harris in The Moral Landscape. Harris uses horrific examples of abuse which only serve to elicit a normal emotional response in his readers, challenging anyone to say that the examples aren’t cases of genuine evil.
The problem with that approach is, only a very small slice of people are actual psychopaths. Most of us don’t easily relate to horrid, direct cruelty. But we can at least find the selfish hedonist intelligible, from a motivational standpoint. This is a person who takes what most of us apply in smaller doses in our own lives, occasionally, and turns it up to 11. This person would never torture a puppy for fun, but might skip town on the debt he owes you, only for you to spot him years later at Lake Tahoe buying drinks for prospective sexual partners.
We comfort ourselves that the lives of people like this are inevitably shallow or guilt-ridden. They probably have some guilt, but apparently aren’t too burdened by it. It’s true that they don’t get the same edification and enrichment that a deeply moral life can offer, but again what they choose is intelligible. Their lives can be fun and free. At least, we can’t rule out the possibility by fiat. They hurt people, but not like a psychopath would. We see examples of this, at least in part, all around us.
If reasons can guide action, or even have a dose of (overriding) categoricity, it’s question-begging for the realist to deny this privilege to the selfish hedonist. The existence of moral realism is supposed to forestall immoral realism, so defenses of moral realism based on mere objective action guidance are over-inclusive. “Bad” reasons are still reasons. Moral realism can’t simply establish morality as one item on the menu. If moral realism is true, at least in its strong forms, the overriding authority of moral reasons has to not only guide the action of the morally motivated, by somehow deny any of the same authority to people motivated differently.
Terence Cuneo, in The Normative Web, pushes the analogy with epistemic reasons. Epistemic reasons have the power to override other reasons for belief. This much is true. We give extra weight to certain considerations over others in belief formation – to expert consensus, personal experience, logical coherence, etc. More formally speaking, philosophers tend to speak of epistemic warrant, which includes empirical evidence, parsimony, elegance, and the like. These reasons carry authority within the intellectual lens, and granted, they can override other considerations. But they are not ultimately overriding.
To illustrate the point, imagine someone came to us and said they decided to undergo a deprogramming process so they would be free to form beliefs however the mood strikes them, without the burden of trudging through all the intellectually responsible considerations (considerations they currently admit are weighty and real). We could plead with them that such a life isn’t honest or principled. They could even agree with our assessment, but confess that they’d rather be unburdened than honest and principled.
In the next 48 hours, this person will have completed the deprogramming process. We scramble for reasons, but find that none of the considerations mentioned above – evidence, parsimony, expert consensus – can even speak to the goal of the prospective irrationalist. What people perceive in ultimately overriding intellectual considerations is actually categoricity manifesting itself in intellectual matters. We need some categoricity!
And we will inevitably call upon it. Some realists say the requirement for categoricity is satisfied by simply referring to the selfish hedonist as immoral (like the irrationalist is irrational). But this label should presumably be more than a self-congratulatory sticker we slap onto something just to distinguish it from items that aren’t morally relevant at all, like a sneeze or a table. The terms should pick out real differences in normative weight in the world, and these differences should be greater than any arbitrary labels we decide to apply to them.
That the selfish hedonist is not morally good is agreed upon, and of course it’s fine to use whatever terminology we like to refer to this. But when we look under the hood and kick the tires, what is the distinction in deeper action-guiding weight between immorality and moral irrelevance, between being a hedonist or a bird-watcher? This is the question categoricity-based moral realists cannot answer. The question is important because the concept “immorality” carries more weight in our discourse than simply referring, with a label, to something that isn’t moral (again, like a chair or a sneeze). Immorality then, in our theories, has to be treated as special in its lack of goodness, not as merely labeled differently than goodness.
The insight that we can pull categoricity apart from other key meta-ethical considerations (like irreducibility) was made explicit by François Jaquet in his essay “Prudential Parity Objections to the Moral Error Theory,” providing most of the impetus behind what I’m now presenting as “categoricity-only” moral skepticism (“categoricity-based” is from Jaquet). He calls the view he defends a “local” error theory. Jaquet tell us that common objections against error theory fail “when targeted at the categoricity-based error theory, and they fail because they conflate it with the irreducibility-based error theory.” He points out that philosophers who adhere to the irreducibility-based error theory also hold the categoricity-based objection, though what we’re learning from insights like Jaquet’s is that the reverse isn’t necessarily true.
Some deny that moral realists should be tasked with preserving categoricity. Any version of moral realism this weak could focus, for example, on the irreducibility piece while leaving categoricity-only error theory alone. The disagreement between this kind of moral realism and what I’ve dubbed quasi-moral skepticism would not be found at the object level, but the meta level. That is to say, the disagreement would be over the parameters of the conversation, (what’s required of moral realism or perhaps what moral discourse is committed to) rather than about the expected subject matter of realism vs. skepticism.
One way of handling this kind of disagreement could be found in naming these different versions of realism and skepticism in ways that advertise the levels of strength they aspired to, like weak or “minimal” moral realism, “robust” realism, quasi-skepticism, and so on. “Weak moral realism,” then, would signal a non-categoricity based (or irreducibility-based) realism, while “quasi-moral skepticism” would signal a categoricity-only skepticism, and so on.
But this would not clear up the entire disagreement between realists and skeptics even just inside cognitivism (realism vs error theory). Many disagreements are still found down on the object level, with realists agreeing to take on the burden of accommodating categoricity and claiming to do so successfully. Many of these allegedly successful attempts involve characterizing immoral behavior as necessarily irrational, or as some sort of failure of mature personhood.
Critics often scoff at moral skeptics who say moral reasons can still motivate them, or who identify with moral values in their everyday lives. The absurdity is supposed to come from still being influenced by considerations that aren’t supposed to exist in the judgement of the moral skeptics themselves. Regardless of which side might prevail in such a dispute, it’s not a concern for categoricity-only error theory. Categoricity-only error theory can co-exist with a kind of normative realism that accepts irreducibility, and it can accommodate reasons for action motivated by moral considerations, just as easily as it can accommodate any other kind of action-guidance.
Leave a Reply