Note before we go on that Harris is replying to the issues in play as they are summarised by Jerry, rather than to the detail of the argument over in my review. That's fine as far as it goes, as long as we're all clear. Further, some of the points that he's responding to are more Jerry's than mine, or at least reflect Jerry's formulation.
That said, Jerry and I are, I think, largely in agreement in our criticisms of the Harris book. If we do disagree anywhere, it will probably be on something fairly subtle, or where Jerry raises a problem of his own that isn't part of my critique. E.g., Jerry raises an issue about trolley cases that I don't deal with directly in my review. I'm not saying I necessarily disagree with him here - I'd need to think about it, and the trolley cases may, for all I know, cause Harris difficulties - but it's simply not something that I relied on in pointing out that there are complexities that The Moral Landscape skates over, and sometimes gets wrong.
The metric for "well-being"
The first point that Harris deals with, however, is made by me as well as by Jerry. This is a problem with any talk that relates to maximising the well-being of conscious creatures or comparing the respective effects of two different strategies (two different actions, say, or two different systems of laws and customs) on their well-being. It's not the deepest problem with the book, but it's a genuine one that Harris needs to deal with somehow. It's this: We cannot make objectively binding comparisons of well-being unless we have a metric for well-being. We need to know what "well-being" actually is, and it needs to be something that comes in units of some kind, so comparisons can be made.
Harris just repeats a point that he makes in the book:
This is simply not a problem for my thesis (recall my “answers in practice vs. answers in principle” argument). There is a difference between how we verify the truth of a proposition and what makes a proposition true. How many breaths did I take last Tuesday? I don’t know, and there is no way to find out. But there is a correct, numerical answer to this question (and you can bet the farm that it falls between 5 and 5 million).I thought he might try to put an argument that there actually is a metric for "well-being", discoverable perhaps via research in neuroscience. That would be something of a leap of faith, though, especially without a very clear definition of "well-being". Instead, he gives an answer that misses the point of the criticism. Alas, it simply is a problem for his overall argument.
His example of how many breaths he took last Tuesday would do for my purposes, but let's first take another example that brings out some extra complexities. We can ask how many grains of sand there are on a particular beach. This might be difficult to answer precisely, partly because we need to know how far down the "beach" goes and partly because the boundaries of the beach on the map are somewhat vague: it will shade into whatever environment lies behind it, it will vary between high tide and low tide, and so on. All in all, we are going to have to agree on a thin line around the beach, thin enough so that each grain of sand falls on one side or the other, plus some kind of two dimensional boundary beneath. Only at that stage can we be sure what we are even measuring.
Once we've done that, however, we have a metric - probably. I'm assuming that we have a clear enough agreed understanding of what a "grain of sand" is to avoid disputes in particular cases. If that's so, our metric is "grain of sand" and there will be an objectively correct answer as to how many grains of sand there are on the beach (as defined by our boundaries) at a particular time. The answer might, theoretically, be, oh, let's say, 125 billion grains of sand. This figure can vary enormously among larger and smaller beaches.
Similarly, if we can agree on what the precise boundaries of "Tuesday" were (this will include working out which timezone we are going to use), and if we can agree on what is an inhalation strong enough to count as a breath, and on how we will count any incomplete breaths at the ends of the day, then theoretically there is an objectively correct answer, let's say 16,000, to the number of breaths that Harris took last Tuesday.
He is also correct that it is difficult - for all practical purposes impossible - to count the number of grains of sand on a particular beach or the number of breaths taken by a particular person on a particular day. That, however, is not because of the lack of metric. In these cases, we do have a metric: we know what counts as a "grain of sand" and what counts as a "breath". It's precisely because we have a metric in each case that there is an objectively correct answer, even if it's not possible for us to establish the answer in practice.
But the point about well-being was not that it is difficult - or, in practice, impossible - to count the number of units of "well-being". It's that we simply don't have a unit at all. If I decide to apply for a particular job, rather than deciding not to, we can't compare the respective number of units of well-being that are contributed to conscious creatures by these different causes of action, and it's not simply or only because of practical difficulties in counting, as with the number of grains of sand on the beach or the number of breaths taken by Sam Harris on a particular day. The problem, rather, is that we don't even know what we are supposed to count. The mere practical difficulties of counting might also cause a problem for Harris at some point in his discussion, but it is not what was raised. The problem is that we don't know what we're counting at all. We don't even have agreement on what "well-being" actually is.
Likewise, it's not good enough for Harris to say:
These are all interesting questions. Some might admit of clear answers, while others might be impossible to resolve. But this is not my problem. The case I make in the book is that morality entirely depends on the existence of conscious minds; minds are natural phenomena; and, therefore, moral truths exist (and can be determined by science in principle, if not always in practice). The fact that we can easily come up with questions that are hard or impossible to answer does not challenge my thesis.However, if some of the issues in question are impossible to resolve - and not just because of practical difficulties in counting - it certainly does challenge the thesis of moral objectivism, to which Harris appears to be committed. He could defuse the whole issue here if he simply abandoned moral objectivism, which I believe he could do quite readily. (Leave aside that the argument in this quote is not, as it stands, logically valid.)
One suspicion here (I don't put it beyond that at this stage) is that well-being is not simply one thing. It may be a combination of things, and reasonable people may disagree about the relative importance of those things. Harris seems clear that it is not simply pleasure, but if not what is it? Is it pleasure minus pain (and can these be placed on the same scale?)? Is it some set of goods that a person possesses, such the ability to do certain things? Is it some sort of summation of a number of things that people tend to care about, measurable in different units if they are measurable at all?
There will be no objectively correct answer - not even one that we can't discover in practice - as to what course of action produces more well-being units unless well-being actually has a metric. If it does, there will be deeper questions as to why we should feel bound to act so as to maximise these well-being units, whatever they are, but if it doesn't ... well, we can't even get that far.
Making judgments of merit
Note that we can still make judgments about the merits of moral systems, systems of laws, particular acts, people's characters, and so on. Moreover, those judgments need not be just arbitrary. In any actual discussion we may have, not just anything can count as a "good" act or a "good" system of customs and laws, or a "good" set of dispositions of character. Harris is quite right to attack people (none of whom seem to be found in philosophy departments, but it's rumoured that anthropology departments are well stocked with them) who see all this as just arbitrary. But he doesn't have to make claims to the effect that one course of action or whatever is objectively the correct one, or that one is always objectively superior to all others (except in the rare event of a precise tie). We don't think like that about other value judgments - not usually - so it's something of a psychological puzzle why so many people want to think like that about morality.
Consider how we judge the merits of motor-cars. I can't imagine that many people think that judgments of the respective merits of two similar cars are the sort of thing that can be just plain true or false, as if my 2009 Honda Civic is worth 1000 car-merit units while the price-equivalent Mazda is worth only 990 car-merit units. That is not how it works.
Rather, we can take account of data about such matters as performance (which can be broken up into many other sub-components, such as maximum speed, acceleration in a range of circumstances, towing power, stability when cornering, braking, and others), fuel economy, comfort, how easy the car is to drive (and maybe how "fun" it is to drive), various aspects of styling, reliability, and doubtless others. These are not arbitrary considerations for beings like us to take into account: they reflect stable desires that are widespread among human beings, given the kinds of beings that we are and the kinds of circumstances in which we find ourselves. Not just anything can count as a "good" car: we'd raise our eyes if someone recommended an uncomfortable, unreliable car that guzzles fuel while failing to accelerate, cornering with little stability, and braking dangerously.
And yet, I may judge my Honda to be of greater merit while you do likewise with the Mazda ... and neither of us is "just wrong" or being irrational. We may be placing different weights on different things. You might be putting more weight on some aspect of performance while I am putting more weight on reliability - or vice versa. My judgment of which is the better car will reflect my desires and values, while your judgment will reflect yours. Neither of us is going to be so naive as to insist that the Honda or the Mazda just is the better car of the two, as if one racked up more car-merit units on some scale that we are both bound objectively bound to apply.
Most of our judgments about the merits of things - knives, cars, friends, sunsets, or whatever you care to name - are like this. They are not simply arbitrary, so that just anything can count as a good knife (presumably we all want such things as sharpness, or at least the ability to be sharpened, and sturdiness), but nor must we all agree on pain of being just wrong ... objectively wrong like the person who produces the wrong figure for the number of grains of sand on the local beach.
Simple things and complex things
Once we see that judgments of the merits of relatively simple things, such as knives and cars, are like this, why do so many people struggle with the idea that judgments of the merits of much more complex things, such as people (with their complex psychologies) or systems of laws, are even less likely to be just right or just wrong? Admittedly, not just anyone will count as a "good" person, but we fool ourselves if we think that there are person-goodness units that we can use to decide as matter of objective fact that one person just is "better" than another. It's not because it's difficult in practice to conduct a count or do a sum, but because there are no such units.
But that doesn't entail that there are no non-arbitrary reasons to prefer a person who is honest, considerate, kind, non-violent, cooperative, and hard-working to someone with all the opposite qualities. Given widespread facts about human beings and our situation, these are not just arbitrary factors to take into account.
We can make perfectly rational judgments about the merits of systems of law, political platforms, individual actions, and people whom we meet or see in public life. In each case, we can use criteria that are not arbitrary for beings like us to use. In some cases, all the criteria will point in the same direction. But there remains room in a very large class of cases for people to disagree in their judgments of merit without either being just objectively wrong.
I doubt that Harris would lose much if he took all this into account. He could still make rational criticisms of religious morality and of crude, quietist forms of moral relativism, which seem to be his main targets, without being saddled with a naive kind of moral realism (and getting bogged down defending it).
He needs, however, to get a better sense of the points that are actually being made against him. In this example, we're not merely complaining about the practical difficulty in measuring well-being, like measuring breaths or grains of sand. The problem to be faced is whether well-being is the sort of thing that has a metric at all.
38 comments:
"The problem to be faced is whether well-being is the sort of thing that has a metric at all. "
But isn't that already conceding his main thesis, that morality is a scientific project? If merits, including moral ones, "reflect stable desires that are widespread among human beings" - then the science of morality is reduced to the task of finding out what are the stable and wide-spread desires (a question that falls within psychology), and moral engineering is how to advance them.
The argument between you two can then be reduced to the question of whether there are Normal desires - whether the distribution of desires and their weights is roughly Normal, so that the bulk of humanity mostly shares the same underlying values and can therefore be addresses as a single moral community. While it is fairly clear that is not the situation with desires in-practice, it is less clear that it can't be argued that Normativity is the case for final ends or with "unbiased" education (i.e. exposing people to truths such as the suffering meat-eating causes to animals, the lack of divine punishments, and so on).
To alter Harris' syllogism somewhat,
values entirely depend on the existence of conscious minds; minds are natural phenomena; [science can determine natural phenomena]; and, therefore, values can be determined by science in principle, if not always in practice.
As far as I understand your position and Harris' respons, he is equivocating.
When one asks a question like: "Why should I accept this conclusion?" it is a different question depending on context.
The first context is from an anti-science position from someone who doesn't think highly of scientific results. The real question here is: why should I value scientific results more than I do now.
The second context is from someone who does value science but is currently ignorant about the evidence and/or arguments leading to the conclusion. The real question here is: Show me the work.
As far as I understand, you are kind of asking the second question, while Sam Harris is answering as if you are asking the first.
If we all had the same desires and beliefs we'd all presumably accept the same norms of conduct. Or so I suppose. Recall that Hume argued for something a bit like this.
Michael Smith has his argument about how we'd all converge on the same judgments if we were well enough informed, even if we started out with different desires. I doubt that, but I think we'd converge a lot. My evidence is just the way clearing up the facts can sometimes produce agreement on a course of action.
I.e., it looks from here as if Michael is right to the extent that a lot that divides us is actually our differing beliefs about non-moral matters.
Harris probably needs to come to grips with Smith's work, because it offers him at least some support from a modified Humean viewpoint. I don't think it'll get him all the way, but it's certainly an approach he should consider. And it does suggest that, at the least, a large degree of convergence is likely (which is something most of us would welcome).
Once again, Harris could adopt weaker metaethical premises and get most of what he wants.
The syllogism is still problematic, though. See the next post.
I'm not entirely sure that having an exact metric is necessary in order to make non-subjective pronouncements about morality. To take your grains of sand on the beach example, there is a wide variety of possible metrics. Each could give wildly different numbers (or not give exact numbers at all, but estimates) of the number of grains of sand on a beach. We could argue for years before we could agree on a best metric. But we don't need to agree on a metric to agree that if a dump truck dumps a load of sand on the middle of the beach, the number of grains will go up, with just about any metric or estimate that you could reasonably come up with.
Similarly, for many moral questions, we may not need to agree on a precise metric of well-being to agree that a particular course of action would increase or decrease well-being, because it would do so for a wide variety of possible metrics. Of course there will be (very interesting and relevant) cases where this can't be done, but I do want to point out two things:
(1) you usually don't need an absolute metric, you only need to know when there is an increase or decrease, and
(2) you often don't need to agree on a metric at all if you can agree what the effect would be for a variety of not unreasonable metrics or estimates - without knowing exactly what those metrics are.
Your example of choosing to take a job or not also raises some questions for me. Clearly, we do make decisions regularly where we weigh all sorts of different aspects (with different units) against each other. Is the higher pay worth the longer commute? How much extra well-being can I buy with that extra money, and how much well-being will the loss of free time cost me?
My questions then would be: Does the fact that I may use different weights than you do make our decisions subjective? I don't think it does. We probably both think we have good reasons to assign these weight factors as we did.
Which leads me to the next question: how do we measure (or estimate) our own well-being? Wouldn't answering this question also give a (partial) answer how to measure well-being in larger groups? And how else would we answer this question but through science?
"My judgment of which is the better car will reflect my desires and values, while your judgment will reflect yours. Neither of us is going to be so naive as to insist that the Honda or the Mazda just is the better car of the two, as if one racked up more car-merit units on some scale that we are both bound objectively bound to apply."
So the statements "The same car is better for both Russel and Brian," "The Mazda is better for Russel," "The Honda is better for Brian," are false. "Different cars are better for Russel and Brian,""The Mazda is better for Brian," "The Honda is better for Russel," are true. How does that situation have implications for Harris?
"We can ask how many grains of sand there are on a particular beach."
You did a good job showing that we can't simply ask that. Surely the problem is merely linguistic? Naming "a particular beach" isn't specific enough. Consider why you used the word "particular". Had you simply said "on a beach" or "in a heap", the sentence would have been obviously insufficient as evidence for what you are trying to show. I think that naming a beach makes the deficiency less obvious but no less present.
"These are not arbitrary considerations for beings like us to take into account: they reflect stable desires that are widespread among human beings, given the kinds of beings that we are and the kinds of circumstances in which we find ourselves."
Humans make moral decisions in which every aspect and emotion is fungible, as in your car example, based on their thoughts about the nature of reality and their personal desires. Desires are the metric we use to adjudicate moral questions within ourselves and there is no reason they couldn't in theory be applied to disputes among ourselves.
"It may, indeed, be one of the truths about the phenomenon of morality that moral norms are never objectively binding in the relevant sense. That is, roughly: they are never binding, on pain of irrationality, on the people concerned, irrespective of their actual desires."
Fulfilling desires is part of morality, so saying that moral realism fails without taking some desires into account isn't such a blow.
Russell,
Could you be more specific about where that argument is to be found? I'd be interesting in seeing it (although I tend to agree with your opinion, that's before I heard the counter-argument!).
Yair
"To alter Harris' syllogism somewhat,
values entirely depend on the existence of conscious minds; minds are natural phenomena; [science can determine natural phenomena]; and, therefore, values can be determined by science in principle, if not always in practice."
AFAIU, this is not Sam's view. SH is talking about how science can answer the question: what values should people have? Not how science can answer: what values do people have?
SH seems to tell that science can not just tell us what the values of specific people are, but that science can somehow tell us which of those values are "correct" and which are not. And that science may even come up with correct values that nobody is currently holding.
I, for one, am deeply grateful to you, Russell, for writing in such a clear and patient way about the naive philosophical reasoning Harris advances in his latest work. He is a powerful rhetorician and can seem very impressive to those not too familiar with the legitimate arguments against moral realism. Why he seems so hell bent on bullishly defending such a tenuous position, that you correctly point out his book really doesn't need, flabbergasts me. I'm sure I'm not the only one who feels that he has done more damage to the enterprise of truth-seeking than good, even though, as you say, his book is on the whole a force for most of the values enlightened people stand for. One huge positive from all of this is that the subject of metaethics is being discussed by a much wider audience. But it would be an intellectual tragedy for that philosophically untrained audience to think that Harris's 'The Moral Landscape' is to morality what his 'The End of Faith' is to religion. It's one thing exposing the absurdness of religious thinking; it's quite another dismissing the percipience of David Hume.
I love the analogy about comparing metrics of a vehicle's performance, etc., vs. the idea of "car merit-points". For me, it helped to clarify why I felt unable to reject Harris' thesis outright, even though it relied on assertions I just could not accept. What you present here salvages a non-arbitrary morality (which is a key point for me) without making unacceptable assertions.
The one thing that I wish you had gone one step further with on the car analogy is to state explicitly that if someone tried to say that some hypothetical lemon of a car (the one with poor acceleration, poor handling, poor safety, uncomfortable, ugly, etc.) was better than a Civic, then in that case we wouldn't just "raise our eye[brow]s" as you say, but we could safely assert that that person was objectively wrong -- even though we could say no such thing about the Mazda vs. the Civic.
You pretty much implied that, but you never stated it explicitly, and by my thinking that is an important point: In some cases, two people can disagree about what is moral and neither of them is objectively more correct than the other; in other cases, two people can disagree about what is moral and one of them is objectively wrong, in the sense that there is no way to reconcile the performance of that person's proposed morality in various objective metrics with any sort of defensible (within the context of H. sapiens) value system.
Well-being is no more measurable than the term 'normal' is measurable.
However, a very simple solution is to replace 'well-being' with 'sympathy' and then a science of morality is well within the grasp of psychology and neuroscience. I don't know why Harris is fixated on well-being, nor why he doesn't see that human emotion can form a measurable basis for moral judgements.
Your whole response could be leveled against our scientific understanding of health-as I'm sure you've read Harris mention in numerous occasions- but you didn't address this analogy.
What is the metric we currently use to decide whether someone is in good health? We really don't have a unit of measure, do we? However this doens't stop us from emitting prescriptive norms and practices that generally lead to better health.
Why couldn't the same case be raised in defence of morality?
(Caveat: I've not read Harris' book; I've only seen his TED talk and email to Jerry Coyne.)
Did Harris ever claim there *was* a singular, quantitative metric for "well-being"? My impression was "no" -- it's clearly a qualitative property. And those can still be evaluated and compared/contrasted scientifically, or even colloquially.
A simple exercise: Replace all instances of "well-being of creatures" with "redness of paint". Red is generally defined as "the range of light between 630 and 700 nm". We know that wavelengths outside that range aren't red. But there's no single, numeric value for "maximal red". Biology is no help: our L cone cells are most sensitive around 564 nm, which isn't within the range! This situation means that we can get by on comparison, saying that values closer to the mean of that range are "more red" and values further are "less red".
I find Harris' idea of a "moral landscape" to be similar. To paraphrase, he refers to it as having peaks and valleys, but with no single, obvious peak (no one-size-fits-all behavior to maximize well-being). And we know this intuitively: clearly beating a child is "less moral", whereas providing a healthy educational environment is "more moral". That doesn't tell you which exact school you should choose for your kid, but some information is better than none.
I definitely agree with your suspicion that Harris is using "well-being" as a catch-all for many properties. If I recall, he was explicit about that in the TED talk. This would make the moral landscape not a simple 2D graph of "goodness vs badness", but more like a 4- or 5- (or n-) dimensional topography. I'm not sure you can make a strict mathematical model like that without quantifiable and consistent units for the axes, but no analogy is perfect.
But nonetheless, what might matter more than the landscape analogy is the *results*. So exercise 2: stop thinking so much about humans, and start thinking about rats. I think Harris is suggesting that we can choose a handful of properties of organisms (well, mammals at least) which are quite nearly universally good, like perhaps Darwinian reproductive fitness or mental alertness. These properties let organisms solve problems (e.g. a novel maze), and pass traits that enhance problem-solving to the next generation.
We'll doubtless agree that rats need a baseline of food, water, oxygen, sleep, etc, and that depriving them is immoral. But beyond that are matters of genes, stress, affect, caretaking, socializing, play, and more which all may impact the maze completion time either subtly or profoundly (I'm pretty sure it's confirmed that genes and stress matter).
As you say, choosing to optimize a set of properties (e.g. problem-solving), both for individual and group, may seem arbitrary. But given the full circumstances of existence it's quite rational (issues like the Tragedy of the Commons would rule out blatant greed).
Overall I think you're wrong to characterize Harris as using simple "moral objectivity" -- he's acknowledged the qualitative nature of morality. Neither does the lack of quantitative values for "well-being" mean that science is totally mute -- our choices and behaviors can be informed by quantitative genetics, neuroscience, and cognitive science. And I agree with your suspicions about multiple variables and musings on moral judgment.
My suspicion regarding Sam Harris is that he clings to moral objectivism because he needs it for something else that he can't have: a sound claim of ultimate justification to impose his principles, and enforce them in practice, on people who don't accept them. I think that he would like for Western cultures with "enlightenment values" to be able to take a conscience-free skate over the objections of cultures with differing points of view in the same way that he does over every kind of philosophical and logical subtlety, so that we need not question ourselves or our chosen methods as long as we are promoting the correct values in the most expedient way, as determined by Science. He might even agree with me that this is an absolutely awful scenario, but it seems to be what he craves.
Excellent post!
The answer to your desperate question is of course: because many people want simple explanations and models for everything. Acknowledging complexity is difficult, and slows down decision processes.
Maybe this is off base but it makes sense to me now. On the issue of a metric for well-being.
Could there be a single metric for physical health? Several metrics? Maybe average life-span, chronic pain, # of diseases, aliments, disorders of all types, #'s of extremely painful events, risk of disease? If you could have such metrics for physical health, why couldn't you likewise have some for "well-being". For one thing, physical health clearly has a lot to do with overall well-being. But in addition it seems you could bring in a science of the mind. Can we not arrive at reasonable criteria for an open-ended conception of "well-being"?
Fear, hatred, shame, guilt, anxiety, etc.? Can anyone sensibly say that those states can't be generally considered "bad"? What would 'bad' mean if all of our negative emotions are not bad? While it would be of course, impossible to get all of the relevant data, objective answers would seem possible in principle, if not always in practice.
Again, does physical health have a single metric? If not, how does it make sense to ask for one in the case of health in a much broader sense? What are the units of physical health? I don't understand why, say, the metric of 'amount of anxiety experienced over a lifetime' would be an arbitrary 'well-being' metric. (Of course well-being is much more than not feeling anxious, but can one sensibly say that it is good to experience a lot of anxiety in one's life?)
The entire section 'making judgments of merit' doesn't seem to contain anything that would be a problem for Harris' thesis. When I read it saw 'many equivalent ways to thrive', 'many peaks on the moral landscape', and 'not one right food to eat'.
Hi! You said: It's this: We cannot make objectively binding comparisons of well-being unless we have a metric for well-being. We need to know what "well-being" actually is, and it needs to be something that comes in units of some kind, so comparisons can be made.
Umm .. I guess I disagree: Missing an arm is worse than not. Missing a leg is worse than not. Missing an arm may or may not be worse than missing a leg: Maybe different people can (reasonably) disagree. But I have no "units" for this: How about losing an arm is 73 and losing a leg is 82; for me of course!
For the concept of well-being, see:
* Crisp, Roger. "Well-Being." In Stanford Encyclopedia of Philosophy, 2008: http://plato.stanford.edu/entries/well-being
Theories of well-being:
According to one group of theorists, "well-being consists in the greatest balance of pleasure over pain." – What is good is good for me, you, or us; and what is good for me, you, or us is what feels good to me, you, or us.
So, in the (arguably utopian) state of perfect hedonistic well-being all people would only have pleasurable experiences and no painful ones. The essence of well-being thus understood is pleasure-experience.
According to a second group of theorists, "people's well-being [consists in] in the satisfaction of preferences or desires, the content of which could be revealed by their possessors." The essence of well-being thus understood is desire-satisfaction. Then, the (arguably utopian) state of perfect well-being is attained when all of my, your, or our desires are satisfied.
According to a third group of theorists, well-being consists "neither merely in pleasurable experience nor in desire-satisfaction. Such items might include, for example, knowledge or friendship."
I'm not sure whether I'm right, but it seems to me that the essence of well-being thus understood is some sort of value-actualization. For example, people who have many friends experience (partial) well-being because the value of friendship is actualized in their life. Then, the (arguably utopian) state of perfect well-being is attained when all of my, your, or our values are actualized.
I think the fact that Harris refers to it as a "landscape" implies that he's already thinking that moral judgement will have a multi-valued measurement rather than a single value. In his book, he also discusses how different reasonable people may disagree about how to weight different factors which is again a nod towards this multi-valued evaluation.
So rather than being challenged by your example of ranking cars or knives, I suspect that this may be exactly what he's thinking of. When you look at car guides, there are 4 star and 5 star cars, and there are 0 and 1 star cars. We may be at the point where we can distinguish between the 0 and the 5 stars without having detailed metrics, and what you're saying is that we'll have to design some sort of moral evaluation matrix before we can distinguish between the 4 and 5 stars (or even between two 5 stars). I agree but even without a comparison matrix, we can say that a rust-bucket car which breaks down every 20km and bursts into a fireball if you apply the brakes too hard is an objectively worse car than a brand new Honda. We may need more details later but for now, the moral world has plenty of rusty Ford Pintos amongst the new Hondas - how much quantification do we need to decide between them?
So in the end I agree that there's more work to be done, but I also agree with Harris that this need not detract from his work, nor does it necessarily need to be something that he needs to complete. Am I missing something or is there some reason why he needs to flesh out these details?
There is a problem with your analogy with car merit-points, which I did not realise until reading James Sweet's post. To paraphrase:
Person A believes the Honda Civic is the best car.
Person B believes the Mazda is the best car.
Person C beleives 'an uncomfortable, unreliable car that guzzles fuel while failing to accelerate, cornering with little stability, and braking dangerously' is the best car.
This value judgement is based on several factors weighed like this:
Factor A*importance+Factor B*importance... etc.
So, why can you say that person C is wrong? He merely puts a different importance to different factors than person A and person B. If A and B both have their best car by their own values, then C can have the best car by his own values. This makes the comparison loose ground, in that we still look objectively at the value of a car, and just have difficulty zooming in on the exact metric.
Back to talking about the beach, someone uses a certain method to determine it's 1.6 billion grains of sand, the other comes up with another method and determines it's 1.5 billion grains of sand. Because of uncertainty, one of them can't say the other is wrong, but they can laugh in someone's face if he says there are only 10 grains of sand on the beach. Clearly this person is doing something wrong in determining the amount of grains of sand.
Back to person C. Imagine that all he finds important is how a car looks. Then he is doing something wrong. Driving a safer car would reduce the risk of injury for him, a more fuel friendly car would open up money for him, a less loud car would reduce arguments with his neighbours. He does not realise his well-being could be increased by doing these things, or has a well-being that needs adjustment.
What we can basically conclude from this is that people ought to put a certain importance on things and deviating to far from the norm is socially wrong. If the well-being of others is factored standardly with factor 2, then someone is selfish if he uses factor < 1.5, and a bastard if he uses factor < 0.5
Having said all this, I still can't imagine how to set those numbers in a purely objective way. I just can't figure out whether that's because it is difficult, like Sam Harris says, or because it is impossible.
Okay, but note that I didn't actually say C is "wrong". I said we'd raise our eyes at C. We would, wouldn't we?
And we'd be perfectly rational in so doing.
As you say, there's a problem with setting the numbers in a purely objective way, even with a case as simple as that of judging the merits of motor-cars. But there are things that we can say to appeal to person C and at least get to this person to think about their real values. Do they really only care about styling? Do they really expect other people to take a judgment based only on styling seriously, and so on. In fact, in a case as simple as this, where the thing whose merit we are discussing has been designed for a particular purpose, we can probably say quite a lot this person. Person C may not be simply and strictly "wrong" but certainly seems unreasonable.
(Mind you, there's no straightforward metric for unreasonableness either. However, we can recognise people who are so uncompromising about their own very idiosyncratic values that the rest of us would rather avoid them, not give them power, not want them around if we can help it, and so on.)
In a sufficiently extreme case, we might be able to say that the person apparently doesn't even understand what a car is for.
But my point is partly that even in these straightforward cases involving things like knives and cars we can end up with room for perfectly legitimate disagreements, while at the same time we have perfectly good, non-arbitrary reasons to judge some knives and some cars (and some laws, political systems, etc.) as "good" ones even if it's theoretically possible for someone to disagree without, strictly speaking, making an error.
We're not stuck with a forced choice between totally arbitrary judgments and purely objective ones. It's more complicated than that, but it seems to me (I won't insist that it's clear) that we are actually quite fluent in recognising and negotiating the complications in most cases. We only seem to lose sight of this with claims that we consider moral ones.
"Person C may not be simply and strictly "wrong" but certainly seems unreasonable."
Person C may be wrong by not having considered all relevant aspects. For instance, as was mentioned above, person C may not be aware of how unsafe the car is. We may also be wrong in our judgement of person C's decision if we haven't thought of all aspects that are relevant to C. For instance, we may not be aware that person C simply can't afford anything better.
All of these are factual matters, though, and should fall within the boundaries of science, at least in principle. The fact that we may never have full information on all aspects of a choice is a practical problem, for sure, but it's also a problem that sciecne in general faces. This has not stopped scientific progress, though.
"In a sufficiently extreme case, we might be able to say that the person apparently doesn't even understand what a car is for."
Actually, having different purposes for a car is probably quite common. For most, a car is primarily a means to get from A to B in some comfort and in a reasonably economical way. For others (say, those who are buying a second car), it may be more primarily a means to signal status.
However, I don't know whether ignoring the purpose of a car as status symbol would count as a failure to take a relevant aspect into consideration, or just assigning a weight of essentially zero to this aspect.
I think there's a reason why we have more issues with that sort of analysis for morals than for things like cars or knives: importance.
Presuming that we all have at least a loose agreement on what we want a car for, we can certainly come to a general agreement on what would make a car better or worse for that purpose. But at the end of the day, there will be disagreements over the rankings of those factors -- as already stated -- that will lead to disagreements. But we're willing to accept that because it really doesn't matter to either of us whether you buy a Honda Civic or I buy a Ford Ranger; we're willing to let that be settled by personal opinion with at most a comment of "That's not what I would have done".
We don't think that for morality, and I think rightly so. But those sorts of differences come into play. As we both know, you and I have a completely different idea of what considerations one should use in determining moral actions. However, you and I will agree on an awful lot of even the really big questions, for different reasons (see the Phoenix abortion case as an example; there's a long post on that on my blog which agrees with what was done, but for radically different reasons than you used). So, for the most part, our different moral values lead to the same behaviour. We, basically, buy the same or similar cars, most of the time.
The issue, though, is that in at least some cases it's clear that you and I will disagree, and disagree strongly. And it's clear that in some of those cases you'll consider my moral decisions to be as "amoral" as those of the psychopath, say, and that I'll think the same about some of yours. Neither of us think the morality of psychopath is tolerable, but both of us think that, in some instances, the other's moral decisions are just as intolerable.
At this point, there are major problems. If you want to try to tolerate my moral judgement in those areas of strong disagreement, then you open up the door to accepting the psychopath's as well. And if you want to be intolerant of my moral judgement there, you need something other than values particular to you to back it up.
And this is because morality guides our behaviour in stronger ways than, say, car selection generally does. And because we think morality important -- or, at least, I do -- we don't just want "acceptable", but we want right. Because if we're going to make decisions that impact the lives of people and may even cause their deaths, we definitely want to be doing the right thing. That cries out for something beyond mere preference, which is what generally clashes in the car and knife cases when disagreements can't be settled.
That importance, I think, is rightly why we're far less willing to accept anything that looks like "It's a matter of personal opinion" for morality. Whether that is right or wrong I cannot say.
I agree with DEEN that we can make comparative judgements without needing a specific metric. I don't have a metric for health, and it is a somewhat fuzzy concept, but surely a person in the final stages of terminal cancer is in a poorer state of health than a typical person. Well-being is an even more fuzzy concept than health, but I still think some people undeniably have more of it than others. We can make unimpeachable comparisons in extreme cases, even if we can't do so in most cases.
We can reasonably vary in the criteria by which we judge health or well-being, but there are limits to this variation. If someone insists on judging well-being by the number of letters in a person's name, we can reasonably say that he's no longer talking about well-being at all.
The situation with (non-moral) goodness is rather different, because the context tends to give or imply some purpose. If we say a car is good, we usually mean it's good for the usual purposes of a car. But in another context we might mean good for propping up a signboard, or good for dropping off a cliff! And I would say that this sort of descriptive meaning may be combined with a non-descriptive meaning which merely expresses the speaker's approval.
"Could there be a single metric for physical health?"
I think so. We decide between health outcomes, such as between having a relatively mangled limb or a prosthesis, or having a lean body suited for rock climbing or a stronger one for roller hockey. The fact that within one person non-arbitrary decisions are made implies that it can be done among multiple people.
"Person A believes the Honda Civic is the best car."
This statement is literally meaningless without saying for whom. If he believes the Civic is the best car for everyone, he is objectively wrong. If he believes it is best for him, he is either objectively right or objectively wrong.
"Having said all this, I still can't imagine how to set those numbers in a purely objective way."
Look at what conscious creatures want, and what would fulfill those desires. I think the answer involves that.
There are a number of good points here.
In defending Harris, several posters have suggested that there is a landscape with multiple peaks. Sensible, but that also implies that there may be multiple moral determinations depending on which peak one prefers. This is a bigger problem than just informing people, with the assumption that they will all converge on a standard set of values. How many would agree that 'global' well being is more important than their family's well being, accepting that any money they make above the global average should be taken to redistribute to strangers? 'Global' well being may suggest that, but most people would want to give high priority to their family first which is exactly how we evolved to behave.
Person C beleives 'an uncomfortable, unreliable car that guzzles fuel while failing to accelerate, cornering with little stability, and braking dangerously' is the best car.
This is essentially a straw man argument, few people would consciously do that. But some will fall far outside the conventional parameters. Let me illustrate that with a personal example. My vehicle (my favorite ever) is basically an offroad biased vehicle that rides like a brick, is noisy inside, and not great on gas. But it has a feeling of solidity and no no-fluff functionality that I really enjoy. Additionally when I travel miles deep into back country trails, I pick up additonal dents and scrapes (from rocks etc) which pretty much stay there as 'scars of honor'. To most people, this makes no sense whatever (why are you abusing your car?). To me it does.
Without going too far of track here, though, reasonable peole can come to different preferences without being wrong. Evolution guarantees also that there will always be significant diversity in fundamental behaviors because that is how species adapt to varying circumstances. There are risk takers and risk averse, there are leaders and followers. How much should a legal system limit activities 'protect people from themselves'? There will be strong disagreement on that between the risk takers and the risk averse.
This is one area that frightens me when people try to come up with an overarching 'scientific' framework. There is a tendency to standardize people into a group rather than to allow for individual autonomy. We have government trying to 'guide' people (with varying degrees of coercion involved) in their living habits, eating habits, gambling habits, sex habits, health habits, etc. Once these guidelines become merged with 'moral imperatives' I can see a frightening increase in activity that the government feels empowered to enforce.
(I'm sorry, I'm illiterate in Hebrew) said: "The argument between you two can then be reduced to the question of whether there are Normal desires - whether the distribution of desires and their weights is roughly Normal, so that the bulk of humanity mostly shares the same underlying values and can therefore be addresses as a single moral community."
It seems to me that human psychology bears the same relation to cultural norms that a lumberyard does to a neighborhood of houses. You can find 2x4's, plywood, floor coverings and so on in the houses, but looking at a pile of dimension lumber won't tell you how to build a sound stud wall, let alone suggest a floor plan or zoning regulations. We don't know what are the elements of human cognition, we don't understand how they work to form a stable personality, and we don't understand how personalities interact to form stable societies. All of this is very much fit topic for inquiry (and we seem to be making progress). But knowing the distribution of available lumber is of only very limited use in understanding architecture or city planning. There are a lot of alternatives at every stage.
Russell: "But again, I do think that it's possible to make rational, non-arbitrary criticisms of, say, the treatment of women under Shariah law without relying on anything like a strict moral objectivism. In fact, that's my take-home message at the moment."
You don't need "strict moral objectivism", you don't need to give up cultural relativism, you need a willingness to root for the home team. Our home team is the common post-enlightenment morality. It's a political question, not a scientific one, but if handed the ball I am clear on what goalpost to go for.
Actually, women are biologically not the same as men, different hormone cycles and so on; it would not be "objective" to treat them as if they were men, not that I personally think that sharia is a good response. (For emphasis I would like to repeat this argument saying "Men are different and should be treated differently; it would not be objective to treat them as if they were women.")
(... I do think it's a bit evasive of Sam to respond to Russell over at Jerry's place rather than directly. But never mind.)
James Sweet: "The one thing that I wish you had gone one step further with on the car analogy is to state explicitly that if someone tried to say that some hypothetical lemon of a car (the one with poor acceleration, poor handling, poor safety, uncomfortable, ugly, etc.) was better than a Civic, then in that case we wouldn't just "raise our eye[brow]s" as you say, but we could safely assert that that person was objectively wrong
What if that person is functioning as a Judge at the Monterey (California) Concours d'Elegance? He's comparing a pre-war Silver Shadow with Russell's Honda???
Nurse! More conditionals, Stat!
A response Harris could have made to your objection concerning a metric of well-being is that because conscious beings are natural phenomena that arise from the activity of a brain, what constitutes well-being could be determined from a thorough understanding of the neurology that underlies our mental states.
If one ties well-being to certain "positive" brain states, then one would have a metric. Whatever produces those brain states maximally across a given population constitutes the morally correct course of action. That fails to answer the question of why maximising well-being is what should be done in the first place, but it does seem to provide a metric, although, as is pointed out, an exceedingly difficult one to delimit; akin to counting grains of sand perhaps, but theoretically feasible.
Russell, I fathom that wide reflective subjectivism underpins any objective morality as Googling covenant morality for humanity- the presumption of humanism illustrates.
This idea stems from John Beversluis's commentary on the Hobbes- Hume's subjectivism against what C.S. Lewis maintains in "C.S.Lewis and the Search for Rational Religion." Beversluis notes two points about this form of subjectivism that are objective.
By objective I mean by the consequence of actions on people, other animals and the enviornment.
Ignostic Morgan[ aka Skeptic Griggsy,Naturalist Griggsy, Skeptic Griggsy and Inquiring Lynn- Google any one of these names to find I mean business!
I'm no philosopher but...
I have to say I don't quite get the comparison of measuring morality to the measuring of if someone is healthy. I would say the measurement of a person's health is simply how close they are to death. If you have a sucking chest wound we know from facts that you will die within minutes or hours. Your health is poor. If you have cancer it is similar. If you have no discernable faults in your body that indicate imminent near term death then you are 'healthy'. Using such a measure you can make statements about things like being over-weight as unhealthy because we know (facts again) that life expectancy is shortened and from biology about its impact on your body's ability to sustain your life. Health clearly has a measure or metric that well-being seems to lack.
As for measuring well-being in terms of emotions like shame or anxiety it still seems to me you would have problems. For instance how would compare the shame of a person who has sex before marriage because their religion told them it was a sin to the shame of a person who is stealing money from an elderly person who doesn't know any better? Person A can feel shame in a situation where person B does not. What is going to tell us then whether that feeling of shame is appropriate to the situation and hence counts in the well-being calculus? It still seems that we are just talking about measuring 'what we currently think is right as a majority' which, if I understand Russell correctly, isn't really in dispute. But that is something entirely different than having something that tells us, outside of our own prejudices, that action A is objectively morally superior to action B.
To me the moral landscape seems almost like a tautology. Basically, the moral landscape will tell us what actions are 'right' based on our current inclinations and the reason we know this is because we used our current inclinations to define what the moral landscape should measure (via our choice in the way we define 'well-being' or whatever word you want to use).
Health does actually turn out to be very difficult, though. Fortunately, there are many clear cases - someone has a debilitating bacterial infection, so we try to cure it, or someone has a broken leg so we try to fix it. Most everyday medical practice is like that. But there are lots of less clear-cut cases that give conniptions to bioethicists and health policy folks. Distance from death won't do because there are many things we want from our bodies apart from that, and some may be in competition with it (e.g. what if I will live longer by adopting a way of life that actually makes me physically weaker, mentally less astute, and less happy?).
This issue about health opens up the whole thorny "therapy" versus "enhancement" debate, on which there's now a vast literature. Do we want biomedicine to enhance our bodies and lives or something more modest and specific, like curing diseases and treating injuries?
Maybe I am misunderstanding what you are saying but I think where I am coming from is this. There are kind of two concepts of 'healthy' that we are talking about here. One is a narrow, medical kind of healthy and the other is a wide more vague notion of healthy that could include things like emotional and mental health. If you take a common bioethical situation like if we should keep someone in a coma alive or not that is a question for the wide definition and morality in general but really doesn't impact our narrow evaluation at all which is just a question of, is their body functioning correctly or not, and in the case of the coma it is clearly not. What should be done about that situation is not a question for the narrow medical definition.
I think when people are proposing the similarity to 'well-being' they are basically saying 'well-being' is vaguely defined like the wide definition of healthy is and yet we have medicine so science is obviously working without good metrics. But doesn't this just equivocate and pivot on the definition of healthy? Subtly switching from the wide definition to the narrow? End of life questions are, as you said, thorny hotly debated issues just like what constitutes 'well-being' is, but questions of if my heart is properly pumping blood isn't open to much interpretation. The problem as I see it is that we don't have a true apple to apples comparison of the narrow definition of healthy as it relates to medical science and some equivalent definition of 'well-being'. Does that make any sense? It is a bit hard for me to articulate how I was looking at the issue.
So I think distance from death works just fine for the narrow view of health. Certainly it is unsatisfactory in the wide view but I don't think we would claim that questions of bioethics are a hard science like medicine anyways. If people want to compare the measure of 'well-being' to bioethics I could get on board with that. I would just object when trying to compare it to the more well defined science of medicine solely on the use of imprecise words like 'healthy.
Jeremiah, some of what you're saying sounds sensible to me, but I'm losing track of who is defending what positions by now. Are you saying that Harris-style morality is not strictly objective but is (often) workable because it uses a concept that, while not a strict metric, roughly tracks what we care about? Or have I misconstrued your underlying position? If you're saying the above, I can see your point and would not necessarily disagree, or at least not very much.
But as I say, I'm losing track. :(
Russell,
I would say that is probably pretty reasonable although I think others were advocating that position more than I was. I (think) I am mostly in agreement with your position towards Harris's moral landscape. (Reading your book review was quite enjoyable) Mostly my comment was directed toward some of the sentiments expressed by others above (like the "Could there be a single metric for physical health?" comment) that I interpreted (perhaps incorrectly) to sound like:
1) Medicine is an objective science.
2) Medicine deals with health.
3) 'Healthy' can be a fuzzy, hard to pin down term just like 'well-being' is.
4) Therefore we should expect that 'well-being' can have an objective science just like health has medicine thereby validating Harris's position.
That line of reasoning seemed fallicious to me because IMHO 'healthy' means different things at different places in the argument and written out like that it seems to sort of be affirming the consequent too.
Sorry if I didn't formulate my thoughts very well, I'm not trained in philosophy and just discovered your blog and was intrigued by the conversations. :)
No worries - I should be following the discussion a bit better than I'm actually managing.
It's a very good discussion, and I especially like the fact that there are a few viewpoints here ... and yet we're all discussing it in a constructive way.
I don't think health has a metric either, but we have knowledge of both what improves and detiorates health and we strive to maintain a certain level of it, whatever that may be.
Similarly, Harris argues for well-being, which I think of as long-range happiness or minimizing suffering, to entail similar knowledge based on a growing understanding of the brain and its interaction with society. I think the problem is that the science required to understand well-being is either infant or barely existent, and to Harris' worry, it may never be an area of focus because it's taboo for science to have a say on values.
Aristotle and Epicurus' flourishing are what Sam means.
Sure, but Aristotle never thought that "flourishing" in that sense had a metric. My objection is not to using such concepts; it is to the passage in The Moral Landscape that only make sense if he thinks there is a metric to be applied. I.e. he continually speaks of quantifying various outcomes.
Post a Comment