The Error Theorist’s Dilemma

I mentioned in my previous post that I was once a full-blown error theorist. I remember going through JL Mackie’s arguments in his book Ethics: Inventing Right And Wrong and thinking “Yes… Yes, that’s exactly right!”


I’ve moved on from there, especially since reading Philippa Foot’s Natural Goodness (to which I had pretty much the exact same reaction). I’m a moral realist of sorts now. But I still have a soft corner for error theory, so I want to briefly examine it and some of its implications in this post.

To recap, moral error theory claims that our first-order moral statements are objective claims about the world. These claims presuppose the existence of certain moral properties, but none of these properties actually exist, so most of our first-order moral statements are false. The error theorist claims that moral talk is like talk about witchcraft, phlogiston or unicorns: systematically mistaken.

More specifically, I was attracted to a projectivist form of error theory, which can arguably be traced back to David Hume. Hume wrote in A Treatise of Human Nature that…

…taste has a productive faculty, and gliding and staining all natural objects with the colours, borrowed from internal sentiment, raises in a manner a new creation.

What exactly is Hume saying here? Here’s how the Stanford Encyclopedia of Philosophy explains it:

Consider a straightforward, observation-based moral judgment: Jane sees two youths hurting a cat and thinks “That is impermissible.” The causal story begins with a real event in the world: two youth performing actions, a suffering cat, etc. Then there is Jane’s sensory perception of this event (she sees the youths, hears the cat’s howls, etc.). Jane may form certain inferential beliefs concerning, say, the youths’ intentions, the cats’ pain, etc. All this prompts in Jane an emotion: She disapproves (say). She then “projects” this emotion onto her experience of the world, which results in her judging the action to be impermissible.

This is one way the error theorist can explain why moral propositions and moral disagreement have all the features of propositions and disagreements about factual matters. We mistakenly assume that there are certain moral properties such as impermissibility and permissibility that exist in the real world, whereas they’re actually only projections of our own emotions. This is why our moral discourse resembles witchcraft-related discourse: it is entirely about non-existent entities.

Eliminativism or fictionalism?

Now that we’ve got that out of the way, let’s move on to the really hard question: what exactly is the error theorist supposed to do at this point? Moral deliberation, moral language and moral debate are an integral part of our everyday life. We can’t help but think in terms of rightness and wrongness, permissibility and impermissibility. Whether we’re contemplating the nature of an action ourselves or trying to persuade someone else to change his/her mind, we inevitably end up making moral judgements. Even when I was a staunch error theorist, I couldn’t help but think “That’s just wrong!” when reading about things like FGM. Ethicist Daniel Fincke has a nice post here about the difficulty with moral skepticism that I recommend looking up.

So the error theorist faces a dilemma here. If he wants to be rationally consistent, he will have to stop using words like “right,” “wrong,” “immoral,” “permissible” and “obligatory.” To keep using them or thinking in those terms is like having long debates on whether someone is a witch or not when you don’t believe in witches in the first place. The position I have just described is called moral eliminativism. Moral eliminativism is the view that in order to be rationally consistent, we must eliminate moral talk altogether – we must give up on moral discourse.

An alternate view is called moral fictionalism. Moral fictionalists argue that it might be more rational and practical to keep using moral language even when we know our statements are false. So moral claims become useful fictions of a sort. I don’t know how convincing people are going to find this, but it’s certainly got something going for it.

Richard Joyce asks us in a short but well-written article on moral fictionalism to put ourselves in the position of an error theorist, David. Says Joyce:

David doesn’t believe that punching babies is morally wrong, but we can imagine various situations in which he’ll have good reason to utter the sentence “Punching babies is morally wrong.” Imagine that David is surrounded by a population who do believe in moral wrongness and believe that baby-punching has it. We should remind ourselves that David is no fan of baby-punching. In fact, the thought of it sickens him. He believes that baby-punching ought to be prevented and perpetrators severely dealt with. He thinks all this on non-moral grounds. So when a moral believer asks him his opinion of baby-punching, David could embark on a long and likely-to-be-horribly-misunderstood explanation of his non-moral grounds for opposing the action. But would we really accuse him of any great transgression if he simply says “Baby-punching?! Oh, that’s just morally wrong!”

By “great transgression,” Joyce doesn’t mean moral transgression, of course. He’s just pointing out that there may not be anything necessarily irrational about David’s behavior. Joyce spends some time describing how fictions and narratives play a very important role in human lives, and how we often find stories powerfully moving. The error theorist, he concludes:

takes advantage of these quirky aspects of his psychology. He cultivates a habit of bringing moral concepts to bear on practical problems; he allows moral emotions like disgust, anger, and guilt to wash through him; he is acutely familiar with classifying his social environment in moral terms. All going according to plan, this moral fiction doesn’t encourage him to do anything that he wouldn’t upon reflection choose to do anyhow on non-moral grounds.

I hope you’re beginning to feel the pull of both these positions. On a straightforward analysis, moral eliminativism seems the most rational option. But it’s just too bloody hard. Moral fictionalism might be the only way for creatures like us to adapt to a situation in which we’ve recognized that our moral discourse is systematically mistaken. So should the error theorist be a fictionalist?

A third way?

I’m going to try and propose something that’s halfway between fictionalism and eliminativism here. Echoing the sentiments of my previous post, I recommend that the error theorist stop using words like “immoral” or “impermissible.” So he cleans up most of his moral vocabulary. However, I suggest that he keep using terminology that involves virtues and vices.

What we commonly think of as virtues (compassion, generosity, temperance, prudence) and vices (cruelty, callousness, greed) are called “morally thick” concepts. This is because although they can be defined in a straightforwardly factual manner, it’s not entirely clear that they are purely descriptive terms. For example, “compassion” can be described descriptively as “a concern for the suffering of others,” but it seems to have (like all the other virtue and vice words) an evaluative component as well. These are morally loaded terms. If you think about the way they’re used, you’ll see that they sort of have a normative element built into them. When someone says “That’s a cruel thing to do,” we almost instinctively conclude “I shouldn’t do that.” Similarly, we usually hear “That’s very compassionate” as “That is the right thing to do” or at least “That’s a very good thing to do.”

I also want to point out how virtue and vice terms can be used to describe both people and actions in interesting ways. Consider a man living in an ancient tribal society plagued by droughts. He believes that the best way to rid his village of its troubles and bring the rains is to sacrifice his daughter to the gods. Now, this man may be an extremely compassionate person. But he believes that sacrificing his daughter will send her to a better place. He believes it will eventually bring prosperity and joy to his village. It’s not immediately obvious that we should label him a cruel person even if he does sacrifice his daughter. There is a clear, morally relevant difference between him and a psychopath who kills a child for the satisfaction of watching her suffer. But we also have a very strong intuition that human sacrifice is a terrible thing. It’s a cruel thing. So here we can say that the father’s act of sacrificing is cruel without being committed to the view that the father himself is cruel. Killing an innocent child has all the characteristics of an act that a cruel person would engage in.

I think that most of the time, when we engage in moral debate, we are trying to persuade our opponent to change his or her beliefs. But how is an error theorist supposed to do this? Being a complete eliminativist about moral discourse and merely reporting his preferences won’t help. And you might think there’s something dodgy about being a fictionalist. This is where speaking in terms of virtues and vices might be useful. By adopting this mode of discourse, the error theorist can engage in robust moral/factual debate without being a complete eliminativist or fictionalist. He can try and reason about whether a certain act is compassionate or cruel. He can argue about whether XYZ is the kind of behavior we would expect from a courageous person or a cowardly one. So if, for example, he’s debating the ethics of meat-eating and convinces his opponent that eating meat is not the compassionate thing to do, he’s almost certain to make his opponent revise his moral beliefs.

This error theorist is a partial eliminativist because he has given up on a lot of moral discourse. He’s a partial fictionalist because as I said earlier, the words he’s using are probably not purely descriptive in the minds of most people who use them (including himself). But this seems to me to be the best he can do. What can I say, error theory is disturbing that way.

I actually think the kind of moral discourse I’ve outlined may be a good thing generally, regardless of the metaethical positions of those involved. Issues of subjectivity and objectivity, error and success theories, all become beside the point. Do you value compassion? Then the rational choice for you would be to behave in such and such way. Do you want to be courageous? Then it is in your interest to take my argument that XYZ is a cowardly thing to do seriously.

So where does all of this leave us? I’m not too sure. I’ll leave that for the error theorists among you to discuss. As I said earlier, I’m not an error theorist anymore – I’m a moral realist of sorts. But mostly I just think our moral language is far too messy to parcel neatly into different categories like this. In closing, I recommend reading this very thoughtful piece by Russell Blackford on error theory and the complexity of moral language. He seems to lean towards fictionalism.

This entry was posted in ethics, metaethics, skepticism and tagged , , , , , , . Bookmark the permalink.

6 Responses to The Error Theorist’s Dilemma

  1. I liked this very much.

    But where’s it all start? I agree with Philippa Foot in “Natural Goodness” that it starts from what Kant called “the good will”. Without that, there are no judgments. Foot criticises Kant for trying to develop a purely abstract universal system of morality from these beginnings. She opts for a virtue based response (I think) though she appears to be uncomfortable with it. But what are virtues but collected commonly agreed moral judgments? We need to do practical things, like administer justice and make laws. We need some agreement – your moral fictionalism?

    Well done, anyway!


    Liked by 1 person

    • Thank you, and thanks for your comment.

      I’m not sure Foot would agree that virtues are just commonly agreed moral judgements. She’d probably say that they are dispositions to feel certain things and act in certain ways that humans need to flourish (“Men need virtues as bees need stings.”). The question is if there’s one set of right virtues, and only one way of flourishing, given that different cultures have had different concepts of both.

      And yes, moral judgements are a practical necessity. So I suppose we either keep our moral language (like “ought” and “impermissible”) and become fictionalists, or give up as much of it as we can and stick to speaking exclusively in terms of virtues/vices. It’s not easy to decide between the two.


  2. Perhaps,I could,explain what I think about Foot’s view of virtue.

    You are right that her account does not say anything about commonly agreed moral judgments. She quotes Geach when she talks about the necessity for bees to have stings. But she doesn’t, I think, mean that virtues are a sort of biological necessity.

    Immediately before the Geach quotation she talks about co-operation in societies and the fitness for purpose of an individual animal or human in their society. She seems to want to say that both animals and human can be considered moral if they are able to play their part in their society, and I say this must surely be a virtue they acquire over time. According to Wittgenstein, everything in the inner (the mind) must come from the outer (be learned), so I reckon that what we pick up, in order to play our proper part in society, are virtues, or moral judgments. I’m inclined to think they are commonly agreed, and it takes evolutionary timescales for some of them to form. They change over time. And some of them may be mistaken.

    None of this detracts from your excellent thesis.


    • Natural Goodness was hugely influential for me, but I read it a while ago, so I’m a bit fuzzy on the details. But you’re right – I don’t think Foot claims you can derive the virtues from a purely biological standpoint. And she does think virtues play an important role in co-operation in societies; I agree with you there.

      Liked by 1 person

  3. SamL says:

    Ahoy there!

    As I said on Twitter, great post—I particularly liked the discussion of eliminativism and fictionalism. I think that the tensions between these two versions of error theory hint at the tensions inherent in other forms of moral scepticism, and indeed in any area where where common linguistic practices are suspected by some to rely on spooky metaphysics.

    I have a question about the third way. Seems that rather a lot is going to hinge on the notion of ‘moral thickness’, so I was wondering about how you might flesh that out? I totally agree that in practice, if you can convince someone that x-ing is compassionate then that will be sufficient to get them to admit that they should x.

    But there’s a traditional Humean analysis which would say that this not because “x-ing is compassionate” provides a complete reason to x, but because the premise that completes the inference—“I want to be compassionate”—can be assumed in most cases, on empirical grounds. (You seem to get on board with this in your second to last paragraph? Correct me if wrong.) But if this is right then there’s nothing special about virtue concepts—the scheme that makes the practical inference above good is the same as the one which makes the following practical inference good: juggling cacti would make me look like a boss; I want to look like a boss; therefore I shall juggle cacti. Compassion may well be a concept that can be defined descriptively—in terms of dispositions to act, for instance—but if the Humean analysis is right then there’s nothing morally thick about it, and reasoning with it is just another case of instrumental reasoning.

    So aren’t we really interested in a concept of compassion thick enough to allow “x-ing is compassionate” to act as a sufficient premise in practical inferences, irrelevant of whether someone values compassion or not? But then why can’t the committed eliminativist just eliminate these morally thick virtue concepts as well, or at least use the Humean analysis to eliminate them in their thick senses? In fact I think this is all equally applicable to action concepts—‘murder’, for example, can be used in a morally thick or in a morally thin way. What’s to stop an eliminativist playing this sort of move: insofar as ‘murder’ is morally thick there is no such thing as murder, though there is a whole lot of premeditated, unmitigated killing? If they can do that with murder then why not with compassion?

    Throwing some thoughts out there—I’m not at all invested in this objection since I’m not an eliminativist about moral language (though perhaps for quite different reasons)—was just wondering how you might reply to this line of attack!



    • Hi Sam, thanks a lot.

      You’re right that the “third way” presupposes that most people want to be compassionate. It’s also true that the scheme that makes the inference is similar to the one used in the juggling cacti example. But I still think there’s something special about virtue/vice concepts. Because they have morally loaded connotations, they have (I think) a special psychological effect when used in arguments.

      Consider: “You don’t want to cause unnecessary suffering, giving up meat might help reduce unnecessary suffering, so you should probably give up meat.” An argument like this might work, but if you can get your interlocutor to think that he’s not just doing what he wants to do – he’s doing the *right thing* (even though you don’t believe in rights and wrongs yourself), it will have that extra motivating force that ordinary practical reasoning might not have. This is where words like compassion/cruelty help. What the error theorist is essentially trying to do is give up using explicitly moral language as far as possible, while retaining the power that leftover moral concepts can have in changing people’s minds. And I think this is why virtue terms can be useful for the error theorist: on the one hand they are descriptive, but there’s this moral baggage they implicitly carry around.
      Your point about murder is interesting. I think the same thing (descriptive word + moral undertone) applies with the word “murder.” I suppose this is why animal rights activists typically say things like “meat is murder.”

      I hope this makes sense. I’ve tried to articulate what exactly I have in mind, but there’s still a lot of issues I need to think about and sort out. Thanks again for your comment. 🙂


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s