Monday, June 01, 2009

On Debunking IV: Non-Selective Debunking

(by guest blogger Tamler Sommers)

So far I have considered whether evolutionary explanations undermine love and whether they can be used to debunk non-consequentialist moral intuitions while leaving the consequentialist one intact. In this post, I want to bring these thoughts together to examine a debunking strategy in metaethics I’ve defended in the past: the attempt explain away objective moral values in general.

Here’s a rough outline of the strategy. The explanandum, the thing to be explained, is our moral intuitions—intuitions like “burning cats is wrong!” Moral skeptics and moral realists offer competing explanations for the explanandum, and the debate hangs on which of the explanations is more plausible. The objectivist claims that this intuition is picking up on real moral properties, out there in the world—the wrongness of burning a cat. But the skeptic points out that our biological/cultural evolutionary processes account for these intuitions, and so we would have them whether or not they referred to anything real. So with a clean slice from Occam’s razor we can banish objective moral values from our ontology.

As I said, this has always sounded plausible to me. But consider this strategy when applied to love for one’s children. The explanandum is my deep feelings of attachment for my daughter Eliza. Kin selection theory shows that I would have these feelings whether or not I really loved her. So with a clean slice from Occam’s razor we can banish love from our ontology.

Now the strategy seems completely misguided! Why? Because as Manuel and other commentators point out, my love for Eliza is constituted, at least in part, by the feelings of attachment.

The skeptic will object that unlike love, moral values are not supposed to be constituted by feelings or intuitions that arise from an evolutionary process. Love is subjective. Morality is objective. Fair enough. But what about colors? We don’t say that it’s false that snow is white because evolution designed us to view snow in this fashion.

At this point, the skeptic can respond in two quite different (and perhaps incompatible) ways. The first is to say that there is universal agreement about the whiteness of snow. But there is no universal agreement about morality. And that is why we should reject moral realism.

The second is to say that morality has essential features that are incompatible with these naturalistic explanations, features like its categorical nature or “bindingness.” Since these features cannot fit within a naturalistic ontology, even if there were universal agreement under normal conditions about certain moral judgments—perhaps due to our common evolutionary history—it would still not vindicate moral realism. (These two replies, of course, parallel Mackie’s arguments from relativity and queerness.)

I’ll talk about both responses in more detail in my next post. But for now, let me conclude with an observation about the latter reply. Ashley, in my first post, thought that real love was essentially incompatible with an evolutionary/neuroscientific account of its origin. As some commentators pointed out, one option available to Ashley upon learning of this account is to revise her concept of love accordingly. She could say: love doesn’t quite have the status and history that I thought it had, but it’s still real love, I still love my son. Would anyone begrudge her this revision? Would anyone accuse her of “changing the subject” about love and putting something bogus in its place? Similarly, even if we thought morality had certain features that we now realize are inconsistent with a naturalistic account of its sources, why couldn’t we just revise our concept of morality accordingly? If we allow that Ashley truly loves her son, why can’t we say it’s truly wrong to burn that poor cat?


Gary Williams said...

Great post. Charles Taylor makes a similar argument in his "Human Agency and Language," although his explanandum is subjective agency itself. He uses examples like honor and shame to point out that, like you put it, humans are partially constituted by subjectivity and subsequently, can't be fully captured in a crudely naturalistic vocabulary.

Richard said...

Sometimes revisionism is clearly fine. Other times it is clearly better described as 'eliminitivism'. (Imagine a naturalist who says, "God is love. I still believe in love. Therefore, I still believe in God!" This is clearly to change the subject.) It all depends whether the eliminated features are the ones that we consider to be of central importance.

Personally, I would consider your proposed 'revision' of morality to be eliminitivist in nature. The eliminated feature of normativity (or 'categorical bindingness') is precisely the feature I'm concerned about. You could, of course, start using the word 'morality' to talk about some mere response-dependent property, but then I would no longer have the slightest interest in what you call 'morality'. (Others' mileage may vary.)

P.S. You write: "The explanandum, the thing to be explained, is our moral intuitions"

I don't think any moral realist worth their salt would accept this. (Positing moral realism as an explanation for psychological facts would be downright dopey.) Rather, the moral realist has a more expansive conception of what philosophical data need to be explained. In effect, we take it as a datum that murder is wrong, etc.

Compare: qualia eliminativists like Dennett claim that all that needs to be explained is the third-personal phenomena ("heterophenomenology"), like the fact that we claim to be conscious. Clearly that can be explained without appeal to any essentially first-personal qualities, hence (by Occam's Razor) we should be eliminativists.

The rest of us don't dispute Dennett's claim that third personal phenomena suffice to explain why we claim to be conscious. We just think Dennett has an impoverished view of what needs to be explained. It's a datum, not just that we claim to be conscious, but that we are conscious.

The dialectic is exactly the same in the moral case.

Tamler said...

Richard, I think there's a spectrum in these revision/eliminitivst decisions--and your 'God is love' example is towards the end of the eliminitivist side. Morality without bindingness seems closer to the middle. I'm not sure that there's a determinate answer in this case.

I don't understand your second point. How can these salt-worthy realists, in their debates with moral skeptics, begin with the datum "murder is wrong"? Why isn't that a clear case of begging the question?

Compare: "Any theist worth his salt would not accept religious experience as the explanandum. Rather, they have a more expansive idea of what needs to be explained. In effect, they take it as a datum that some people have been touched by God's power."

Where's the disanalogy?

Richard said...

Tamler - there's a difference between what we take as data (which constrains what arguments can convincingly be used against us) and what we can use in a non-question-begging positive argument against our opponents.

My point is just that there isn't any neutral (non-question-begging) conception of 'the data to be explained'. This makes positive arguments difficult, but then I was making a purely defensive point, so I don't see how that's any objection to what I said.

I think the real work is done in these debates by rough attempts to undermine (or support) the perceived need for a more expansive conception of what is to be explained. Skeptics can beat the drum of empiricism, draw troubling analogies to theism, etc. On the other hand, moral realists can point to how moral normativity isn't any more mysterious than epistemic normativity, and we can't even make sense of the skeptic's claim to be making an argument (which commands our assent) without the latter.

Note that the dialectic is in fact similar in the religious case. (If I understand correctly, Plantinga and many others consider belief in god to have "basic" warrant, rather than being an explanatory hypothesis.) But it seems -- to me at least -- that skeptics have the upper hand in this case since the theist's attempted expansion of the data lacks motivation. (They'd probably disagree. Again, I don't claim there's any univerally recognizable "neutral" position of judgment here. Dialectical stalemates are ubiquitous in philosophy, after all.)

Tamler said...

Richard, that makes sense. I agree that the analogy with epistemic normativity makes things tough on the skeptic. But how does that allow the realist to support a more expansive position on what needs to be explained? And what precisely makes the moral realist's explansive position on what needs to be explained better motivated than the theist's?

Josh Weisberg said...

Interesting discussion!


Here's an attempt to argue against a more expansive conception of the data to be explained in this debate.

1. If the data is expansively conceived, there is no natural property answering to the data.

2. Non-natural properties are problematic--metaphysically, but especially epistemically.

3. So, moral realism is false, or at least problematic.

Instead, if we take a more restrictive conception, there are natural properties answering to the data. So we don't have to say that there are no moral properties.

I'm not sure this moves the debate about moral realism much, except we naturalists don't have to sound as odd when we present our views.

PS I don't think Dennett denies that we are conscious. Consciousness to him is "fame in the brain." He does deny that there are nonfunctional intrinsic qualities of experience, though he attempts to explain why some folk believe there are these spooky things!

PPS Do you know Jordan DeLange? If so, tell him Josh at U of H says hi!

Robin Hanson said...

So what is this strange concept of epistemic normativity, which licenses such expansive concepts of data regarding qualia, morality, and God?

Anibal said...

Moral skepticism and its evolutionary debunking of morality seems to me a perverse reversion of reasoning against the objectiveness of moral intuitions and values.

There is a coevolution between our niche (enviroment)and our CNS (central nervous systems)or if we prefer our adapted mind.

Our minds response to events and stimuli of the enviroment in a manner as faithfully as possible in order to navigate the enviroment succesfully.

Arguing that love or our morals are the consequence of evolutionary processes does not preclude us to say that love and morals are real.

Indeed, it is for this reason that they are real. They were selected to serve a function.

In the case of morals, to sort out the social world into rights and wrongs, in the case of love, to bond and affiliate with others.

James said...

Thanks for a very interesting post. I wonder,

Is there not an important disanalogy?

Ashley does truly love her offspring, I am inclined to agree with you here. Surely the analogous claim to make about morality is that: I truly believe that burning cats is wrong ... (not 'burning cats is truly wrong').

According to my line of thought, you have is a position opposed to, for example, the expressivist who believes that there is not such thing as truly having moral beliefs. But only regarding that point! You do not have a position opposed to the notion that there are no moral facts.

The notion driving the objection is that: explaining moral speak, is not the same as saying what the moral facts are. Error theorists seem to make exactly this point saying, for example, that the folk's moralising practise invokes such-and-such a property, which does not exist.

If I understand correctly, you escape the debunking in the case of Love, because you identify Love with 'the feelings that you have' whether or not you would have felt them in the absence of a certain evolutionary past.

I think that you can do the same with Moral Speak, potentially, but not with Moral Facts or Properties.

kvond said...

Eric: "We don’t say that it’s false that snow is white because evolution designed us to view snow in this fashion."

Kvond: But what we might say is that it is trivally true (unless we are teaching someone the meaning of the word "white"). Is it trivally true that you love your daughter? In a certain sense it is, that is, the truth of it really doesn't matter until there is a dispute over the matter (a visitation dispute, etc.)

The question arises when "You should call that color 'white'" collides with "You should love your daughter". Just how much do you want to be part of a community of users?

Tamler said...

Thanks for the comments everyone. Real quick (for now):

Josh, I like that argument--it seems plausible to me. The only issue is whether it might be equally effective against epistemic values like simplicity and explanatory success. If so, then... well, I don't know. But the Mackie/Joyce argument against moral values relies on those values.

Robin, I think the above remarks can answer your question as well.

Anibal, in some moods, I'm inclined to agree with you. In others, I'm with Richard and James in thinking that morality has to be something more than just a means for social coordination or the expression of values. It needs that bindingness...

James, it's interesting that you and Anibal had polar opposite responses to the post. It might speak to the idea of the indeterminateness of identifying what the essential features of real moral facts or properties are.

Kvond, Eric is not to blame for this post. I (Tamler) am. Could you elaborate on your last question? I'm not sure I understand it, as it stands.

Gary Williams said...

It seems to me that philosophers have seriously engaged with moral skepticism largely because we are still working out the consequences of accepting a Lockean division of reality into primary and secondary qualities. By dividing our ontology into a neutral world "out there" and a subjective world constructed internally, whatever that means, out of the objectively neutral "sense-data," we have robbed ourselves of the philosophical vocabulary necessary to state the claims of moral realism within a plausible framework concerning "mind."

As long as we are still working with an antiquated notion of the human experience as being literally built out of neutral sense-data, we will always be stuck wondering if we can ever "get out of our heads" and into the shared world we all live in. We need to re-shape the philosophical vocabulary we use to talk about human consciousness from the ground up. Without a more adequate conception of how we most basically interact with a meaningful environment, I don't think we will ever escape from these "dialectical stalemates" as Richard put it.

Tamler said...

Gary, thanks. Our copy of "Human Agency..." is checked out. Is there a paper of his that expreses the view you're referring to?

Gary Williams said...

Tamler, I can't recommend any particular articles because I have only read his books, which are excellent. Maybe a shortcut would be to read something like the cambridge companion. However, it would be much better to get your hands on Taylor's own words, because part of his greatness is a lucid and down-to-earth writing style. Any one of his collections like "Philosophical Papers I & II" or "Philosophical Arguments" is sure to have great essays. Most of the essays in these collections have been published elsewhere, so you might be able to find them individually, but that might be hard.

Go here:

and then use search function to find articles that come from his Human Language and Agency book. The most relevant to you are:

"What Is Human Agency?"
"Language and Human Nature,"
"Self-interpreting Animals,"