About Me

My photo
Australian philosopher, literary critic, legal scholar, and professional writer. Based in Newcastle, NSW. My latest books are THE TYRANNY OF OPINION: CONFORMITY AND THE FUTURE OF LIBERALISM (2019); AT THE DAWN OF A GREAT TRANSITION: THE QUESTION OF RADICAL ENHANCEMENT (2021); and HOW WE BECAME POST-LIBERAL: THE RISE AND FALL OF TOLERATION (2024).

Sunday, January 30, 2011

Sean Carroll on The Moral Landscape

Sean Carroll has a fairly brief post over here. He and I are pretty much on the same page, although I should repeat that I don't see the metric problem as the deepest problem. Even if we had a metric and had only practical problems with measuring well-being, there would still be the question of whether it would be a requirement of reason or some such thing that we must act so as to maximise universal well-being.

Or, alternatively, is the claim "You must do X as it maximises universal well-being" just a fact about the world analogous to empirical claims such as "There is currently no rhinoceros in this room"? It doesn't seem to be a logically true claim or an empirical claim, so what is it?

If you say, "Well it's a moral claim" that just seems to show that moral claims are not claims about matters of fact or reason, so how is it binding on us that we must actually act in accordance with them? In very many circumstances it may be a good idea to do so, even by our own lights, but that will depend on just what our own desires, values, etc., actually are.

I don't believe that someone (Alice, say) who acts, on some occasion, in a way that fails to maximise well-being is necessarily mistaken about any matter of fact or acting in a way that is self-defeating or irrational in any other sense that Alice need bother about. We can cheat and redefine rationality in a way that is already moralised, but that will get us nowhere because Alice can just refuse to accept our concept of rationality (why should she accept it if she is not already committed to our idea of morality?).

Sean is correct to insist that the metric problem is not merely a practical one, but one of principle. But the crucial issue goes even deeper than this and would remain even if the metric problem could be solved somehow (by some super-duper neuroscience, or whatever).

I've seen many attempts to get around the problem with morality - that it cannot possibly be everything that it is widely thought to be (action guiding, intrinsically prescriptive, objectively correct, etc.). In my experience, this is like trying to get rid of the bump in the carpet. The bump always pops up somewhere else. We have to live with it.

(Harris has since published a reply to his critics, which is on my "to-read" list. I'm sure it'll be interesting.)


Tom Clark said...

Harris writes at Huffpost in reply to his critics:

“For those unfamiliar with my book, here is my argument in brief: Morality and values depend on the existence of conscious minds -- and specifically on the fact that such minds can experience various forms of well-being and suffering in this universe. Conscious minds and their states are natural phenomenon, of course, fully constrained by the laws of Nature (whatever these turn out to be in the end). Therefore, there must be right and wrong answers to questions of morality and values that potentially fall within the purview of science. On this view, some people and cultures will be right (to a greater or lesser degree), and some will be wrong, with respect to what they deem important in life.”

Harris’s argument is valid and sound if read this way:

~ Science can discover truths about natural phenomena.
~ Morality is a natural phenomenon.
~ Therefore, science can discover truths about morality.

However, Harris’s over-reaching conclusion is that science can confirm *moral truths* (as distinct from truths about morality), normative claims about what we *should* value. But science, a strictly descriptive enterprise, can’t do this.

Values, as Harris points out, come from what we fear, need or want – our human motives. Science can describe as natural phenomena what we value, but it can’t say what we *should* value, since the normative force of that should can itself only derive from another motive; it can’t derive from a scientific description. Descriptions don’t carry emotional valence, the consciously felt moral imperative of motive-driven values; descriptions are only more or less accurate, more or less true according to standards of evidence.

Values and what we deem important are only right or wrong according to standards set by human motives, and these values vary since cultures and worldviews differ in emphasizing certain motives over others - see Jonathan Haidt’s work on the 5 moral foundations and liberal vs. conservative morality. These value structures and their differences can be described by science, but not morally adjudicated. Achieving consensus on values is the work of moral discourse itself, http://www.naturalism.org/normativity.htm

Harris's reply to his critics is at http://www.huffingtonpost.com/sam-harris/a-response-to-critics_b_815742.html

Sean Carroll said...

I don't think there is any disagreement between us. When I am harping on the non-existence of a metric on well-being, what matters to me is that there is not (in principle) any way to derive the "correct" metric from purely empirical considerations. It's not an experimental result.

In other words: you could imagine all sorts of functions of mental states. But the decision to label one as "human flourishing" (or what have you) always comes down to an extra-scientific judgment call.

Brian said...

"...the question of whether it would be a requirement of reason or some such thing that we must act so as to maximise universal well-being."

Sentences impart information.

In your opinion, morality as it must be understood by moral realists, entails that there will *always* be a combination of sentences that makes any given person act to maximize well-being (or do whatever action is uniquely prescribed for that person at that time by the moral system) once they acquire a sufficiently more true understanding of reality, *regardless* of their desires?

So, "moral realists" means "novice mind controllers". How can we describe the actions of someone whose desires adhere to the golden rule, i.e. who does not arbitrarily discriminate among desires; someone who has (had) added to his desires a sufficiently substantial desire that the desires of others be fulfilled?

"...(action guiding, intrinsically prescriptive, objectively correct, etc.)"

The idea that beliefs can be intrinsically prescriptive regardless of desires is silly. You think that idea is inseparable from moral realism? Is there a handy label for moral realism minus that idea?

If Al must choose between himself or many others, if his desires for personal gain outweigh his other desires, he will choose selfishly.

Were ideal observer Bill able to direct this choice for Al, Bill would choose based on Bill's unbiased desire to fulfill everyone's desire.

When people say someone else "should" do something, they mean one or more of several things.

Some mean "according to your desires as they are." Others mean "according to your desires as they would be if you didn't desire well-being for others." Others mean "according to your desires as they would be if you did sufficiently desire well-being for others." Others mean "to fulfill my desires." Others are just making noise.

In the context of meta-descriptive morality, they mean according to the golden rule. Universal morality entails undifferentiated treatment of desires except to the extent that there is rational reason to do otherwise (e.g. such as due to ethical egoism).

Saying that we cannot have true claims that map very well onto what most mean by "morality" because "should" can mean different things is committing the fallacy of equivocation.

Sean Carroll enthusiastically does this when he says:

"Taken at face value, this implies that truths about the best TV shows or most delicious flavors of ice cream also exist."

It either doesn't mean anything to say those things or it means something (true or false) according to only one cogent meaning of each term. "Delicious" means "pleasurable to someone" or implicitly/contextually it can mean "pleasurable to me" or "pleasurable to many" so the statement "This is the most delicious ice cream" collapses down to meaningful statements that are true or false.

"Chocolate chip cookie dough is the most delicious ice cream to someone/me/each person in this room" are all clearly true or false. "Chocolate chip cookie dough is the most delicious ice cream to people in general" becomes either meaningful or nonsensical depending on how we unpack "most". "Most delicious" could plausibly mean a conclusion of any of several ways of evaluating, such as "most commonly ruled pleasing" (i.e. more people like the flavor than they do any other), "most commonly preferred", and "most preferred in aggregate under a BCS-style scheme. The phrase's appearance in a sentence could mean the speaker is selectively picking baggage from several distinct meanings of the word.

What one can't reasonably do is see that objective systems use words like "most" and "delicious", imitate that with a context-free sentence in which those words are ambiguous, and claim victory.

Jason Streitfeld said...

I read Harris' response today and it wasn't so interesting, but it did prompt me to blog again about him--mainly about the problems with his talk about health. But there is one point of interest in his response: he singles you out as being particularly worthy of his response time. Also curious: He seems to simultaneously accept and reject the fact/value distinction. Our oughts cannot come from science, he says; but then, science itself relies on oughts which do not come from science. Why won't he just come out and admit that he was wrong, and that the fact/value distinction is sound?

What he should be doing is trying to substantiate his claim that universal well-being is, or should be, our primary concern, as well as his claim that all moral values are reducible to concerns about well-being. Maybe he's waiting for somebody to publicly challenge his "worst-case scenario" argument.

Russell Blackford said...

Yes, science (or, better, rational inquiry, but I'm not so much interested in harping on that distinction) can find out truths about morality. But as Tom says it's overreaching to conflate that with moral truths in the sense that Harris seems to have in mind. Relativists, moral sceptics, non-cognitivists, etc., all think that there are truths about morality that can be found through science and reason. As I said in another post, one of those truths may be that moral judgments are not objectively binding on all comers. Another may be that they are made relative to social standards. And so on. Even if the argument can be put in a form that is logically sound, it is not an argument for moral realism, merely for the practice of a scientifically-informed moral philosophy. That, of course, is something I'm happy to agree about in any event.

Russell Blackford said...

This is probably just an idle thought, but I wonder how one gets published in the Huffington Post. Have any of y'all been published there?

Brian said...

"Values and what we deem important are only right or wrong according to standards set by human motives, and these values vary since cultures and worldviews differ in emphasizing certain motives over others...These value structures and their differences can be described by science, but not morally adjudicated...http://www.naturalism.org/normativity.htm"

How, if we are naturalists, do we decide which tendencies to encourage, which to discourage, and in what proportion? By...remembering that we’re all persons, with very much the same needs and dispositions.

...basic moral presumption that’s characteristic of the current reflective equilibrium in liberal democracies: there are no privileged persons – we are presumptively equal in the eyes of morality and the law.

Acting morally in any culture means refraining from privileging people's desires more than there is sound reason to. That's what it means to be moral.

There is merely disagreement about what constitutes a good reason to distinguish.

The question of "what is reasonable" is given different answers. How can some responses not be wrong? How is this not an empirical question?

A: "I am campaigning to take the franchise away from gingers."
B: "Why?"
A: "It's moral to forbid gingers from voting."
B: "Why?"
A: "Because the desires of redheads should not be considered at all."
B: "Why?"
A: "Because none of my close relatives are gingers."
B: "So?"
A: "It's reasonable for my family to live under a tribal moral code."
B: "How so?"
A: "It'd be profitable for fulfilling our desires."
B: "What of the desires of others?"
A: "I'm not concerned with them."
B: "Why?"
A: "My genetics and cultural upbringing."
B: "True. That is a reason for you to do it. But in that case you reach your conclusion from the presumption that it is moral to do whatever your genetics and cultural upbringing have led you to do."
A: "The alternative to assuming only some desires are valid is to assume all are valid. While my argument may circuitously begin with the presumption that only some desires' fulfillment is reasonable simply because I have been so conditioned, the only alternative is to assume that all desires' fulfillment is presumptively equally reasonable because another has been so conditioned to believe! That's just as good and just as bad!"
B: "No. Conversations about beliefs in which we take anyone's weighting of desires seriously are conversations about descriptive ethics. Conversations in which we presume desires are all valid except to the extent there is valid reason not to fulfill them are conversations about normative ethics. (This includes e.g. Leviticus, which differentiates according to priestly birth, gender, disability, age, menstruation, etc. It implies that those are *reasonable* reasons to distinguish.) To say 'Let us speak of universal, objective morality,' is to presume that hominid biases hold no weight in the discussion (except as objective features of the world) and exclude statements assuming their truth, just as 'Let us speak of universal, objective physics,' is to presume that hominid biases hold no weight in the discussion (e.g. it's a particle *or* a wave, not both!) and exclude statements that presume their truth (e.g. apparently most animals can survive without food for 40 days and nights).

Modern liberal morality is only slightly different than other systems because it has discovered that some differentiations are baseless. Nonetheless it contains a great many. Discerning this makes it clear that other systems are *also* based on being "presumptively equal in the eyes of morality and the law".

Russell Blackford said...

I don't think A has to assume anything about desires being "valid". A desire is not the sort of thing that can be true. Nor is it the sort of thing that can be valid. A can dismiss that whole line of thinking.

A doesn't have to say any of those things about her desires. All she has to say is that she sees nothing self-defeating or otherwise irrational about acting on her desire to further the interests of her family. In fact, she doesn't even have to say that. She can simply go ahead and act on her desire and not even bother answering B's questions, which are presupposing ideas that she doesn't accept. She's not thereby making a mistake about the world: she may be fully informed of all the facts. Depending on the circumstances, she may not be doing anything self-defeating. What she does may be exactly what she needs to do to fulfil her goals.

Again, she doesn't need to argue about whether her goals are somehow "valid". That's introducing a concept that probably isn't meaningful and is certainly not one that she is forced to accept at this stage of the argument.

Sean, yeah I agree: whatever metric is adopted needs to be advocated on some basis or other without circularity. Otherwise, somebody can simply say that she doesn't feel bound to use that metric.

Brian said...

"A desire is not the sort of thing that can be true. Nor is it the sort of thing that can be valid."

Several times I slipped and spoke of desires being valid. I didn't mean to, and hope that it's clear that when I spoke that way I meant as in "Because the desires...should not be considered at all," "...only some desires' fulfillment is reasonable," and "...all desires' fulfillment is presumptively equally reasonable." Apparently it's not clear; I never mentally made that mistake and where I miscommunicated the argument can be simply amended by assuming I am talking about the fulfillment of desires being a valid project, not the holding of them being somehow "valid".

"She can simply go ahead and act on her desire and not even bother answering B's questions..."

Obviously it is possible to act without answering questions, that doesn't mean one has a good answer for them.

"Depending on the circumstances, she may not be doing anything self-defeating."

B never challenged A by saying A was doing something self defeating.

"What she does may be exactly what she needs to do to fulfil her goals."

Of course. B argues that A is baselessly discriminating among which desires A acts to fulfill, such that A acts to fulfill his or her family's desires for no better reason than that A is predisposed to do that. B claims that objective morality is a meaningful concept rooted in evaluating actions based not on any person's desires, but from a neutral point of view that doesn't arbitrarily disregard anyone's desires.

I think in English it's fair to establish a sophisticated Golden Rule as corresponding to meta-descriptive human morality. It's not impossible or irrational to act only for one's self (or family), just immoral by a reasonable definition because such an actor would be totally (and therefore excessively) discounting all other beings' desires.

The irrational things would be to claim that someone's random product of prejudices is moral (ethical subjectivism), that something has intrinsic value (Sam Harris, most non-nihilism), that no actor acts morally, even when he or she rationally takes into account all information and every being's desires (error theory), expressivism etc.

Ritchie the Bear said...

Not surprising to see Sam Harris overtly endorse benevolent totalitarianism:

"Try as he might, Kim Jong Il just couldn't shake the feeling that his cognac didn't taste as sweet without millions of people starving beyond his palace gates. Given our advances in science, however, we were able to alter preferences of this kind. In fact, we painlessly delivered a firmware update to everyone. Now the entirety of the species is fit to live in a global civilization that is as safe, and as fun, and as interesting, and as filled with love as it can be."


Of course, his whole response to Blackford is totally inadequate. He once again tries to use the stupid comparisons between the "science of morality" and other sciences, like health sciences, complaining of a double standard where no double standard exists.