Roger Crisp writes…
In his new book The Moral Landscape, Sam Harris claims that science 'reveals' values to us. Kwame Anthony Appiah is one of the many who have pointed out that Harris makes the common mistake of seeking to derive an 'ought' from a series of mere 'is' statements, a mistake pointed out by David Hume centuries ago.
But the relationship between natural science and normative ethics does raise interesting questions. Let's assume that Harris is correct in thinking that the right action is that which maximizes overall well-being, and assume also that well-being consists in the greatest balance of pleasure over pain. Pleasantness and painfulness are properties that scientists — psychologists in particular, but others too — can talk about, and if we assume that these properties can be measured in some way or other, then we find that the property that makes actions right is a property that can be studied by natural science.
But what is the relation between rightness and the property of maximizing well-being? In recent years, several naturalistically inclined philosophers have been inclined towards identifying them. So, just as heat turns out to be the same as molecular kinetic energy, so rightness turns out to be the same as maximizing well-being.
In a book which should be published in the new year, Derek Parfit argues that moral properties and 'natural' properties (the ones studied by science) are far too different to be identified. The claim of the naturalists about rightness in the previous paragraph is, he thinks, rather like the claim that heat is a cabbage.
But there is another option — a broader form of naturalism, to which Parfit is less hostile. This might allow that moral properties and natural properties are indeed different, but insist that moral properties are 'anchored' (to use Robert Audi's term) in natural properties in the same sort of way that mental properties are anchored in physical properties — by supervening on them.
If we continue with our assumption (made just to simplify the argument) that rightness in this world supervenes on the maximization of well-being, it is hard not to think that rightness in all possible worlds will supervene on such maximization. That gives us a necessity claim linking moral and natural properties which those on both sides of the metaphysical debate between naturalism and non-naturalism can agree on. Indeed it may be that there's enough common ground here for that debate to be put to one side, so that we can engage in another genuine philosophical question — how it is that we can understand these necessities. We — those who've understood Hume, anyway — know that it's not through application of the scientific method as that is usually understood. But how, then, do we do it?
If only there was a better thought through argument on how rightness, or “value” at least, could be anchored in and even identified with some natural property. Oh wait! I believe there is one!
It seems to me that the non-naturalist argument from Hume forward underestimates what natural properties can do. For one thing, you need nothing but naturalist properties to account for both heat and cabbages. If we only describe carefully enough what sort of importance and impact moral properties are supposed to have, it might be easier to find a role for “rightness” to play, and to see how some natural property might come to play that role.
If I understand it correctly, the notion of supervenience implies that a given state of “maximisation of well-being” with certain defined characteristics will always produce the same concept of rightness.
Either this means that the same “maximisation of well-being” always produces in any one individual the same moral judgement, which seems fairly non-controversial but probably tautologous.
Or it claims that the same situation, described in exactly the same way, will produce a consensus of moral judgement, which seems empirically false. (Even if I agree that many moral disagreements are caused by different perceptions of the facts, and that there would be more consensus if we agreed more on the facts.)
But let us assume with Roger that supervenience bridges the gap between naturalism and non-naturalism, and address your question of how the link (if I understand correctly your use of the term “necessity claim”) between natural events and moral judgement arises :
I think that there are some scientific approaches such as history and anthropology which help us address this question, and I would suggest that by shedding light on the history of moral judgements they can influence our current views. As a general example, the fact that philosophic liberals of the 18th Century could own shares in slave-trading companies may well give us a certain humility to counterbalance our own deeply-rooted, but often socially-constructed, convictions.
A more specific example :
The award of the Nobel prize to Robert Edwards reminds us that attitudes towards IVF have significantly changed since 1978, at which time the idea of “playing god” with fertility aroused so much controversy, which has largely disappeared.
I’m sure that this does not constitute a genuine philosophic answer to your question, but would suggest that this approach could help explain the necessities you describe.
Thanks, David and Anthony.
David: I’m inclined to think this could never work. Let me quote Sidgwick (p. 2 of *The Methods*, 7th edn.): ‘it seems clear that an attempt to ascertain the general laws or uniformities by which the varieties of human conduct, and of men’s sentiments and judgments respecting conduct, may be *explained*, is essentially different from an attempt to determine which among these varieties of conduct is *right* and which of these divergent judgments *valid*’. But go ahead and try — Hume, Sidgwick, Parfit, and I may be wrong.
Anthony: I take supervenience to be a relation between the properties themselves — judgements don’t have to come into it, though they may be constrained by requirements grounded in supervenience. So the essential idea here is that if, in this world, certain natural properties ‘make’ actions right, those same properties must make actions right in all possible worlds. And if this is the case in all possible worlds, then it’s necessary.
I do agree that we should adopt an attitude of humility towards our own moral views, many of which must be mistaken. I also agree that history and other ‘descriptive’ disciplines can help us make progress in ethics. Let me quote Sidgwick again (from the same page): ‘On any theory, our view of what ought to be must be largely derived, in details, from our apprehension of what is; the means of realising our ideal can only be thoroughly learnt. by a careful study of actual phenomena and to any individual asking himself `What ought I to do or aim at?’ it is important to examine the answers which his fellow-men have actually given to similar questions’.
The idea that moral properties “supervene” on natural ones in the same way that mental properties supervene on physical properties is an interesting one, but I think it raises a lot of question. In the first place, I’m not sure to what extent “natural properties” and “physical properties” are supposed to be different things here (except that one is Latin and the other is Greek). Similarly, I would suggest that moral “properties”, if such things can be said to exist at all, are in fact a subcategory of mental properties.
To understand this idea, let us consider that our experience of life convinces us, unambiguously, that both physical properties and mental properties exist. We have mental, subjective experiences, and observe that our fellow human beings report similar experiences. In the mean time those (shared) experiences clearly indicate the existence of a real, physical world, which includes our bodies and brains, on which our ability to have mental experiences depend.
To what extent is it the case that moral properties exist in the same way? One thing that clearly does seem to exist is the phenomenon of choice. (I prefer “choice” to “free will” since the latter has many connotations which may or may not be relevant or appropriate here.) Furthermore, our choices have consequences, and sometimes what we may feel like doing or want to do might conflict with the welfare of others, or with some socially-received sense of obligation. So moral *dilemmas* exist, and in response we have also developed moral principles, to which we then look for guidance to resolve our moral dilemmas. It seems to me that both are in fact sub-categories of mental objects.
But here’s the thing: it’s one thing to say that moral principles exist – as mental objects – and another thing to say that one or other moral principle is “correct”. Yet this seems to be what is being claim, even by Parfit. But why should this be? The idea that rightness supervenes on maximisation of well-being is nothing other than the utilitarian moral framework, but while I (unlike some other contributors to this blog) personally find this framework attractive, I do not understand in what sense this can be said to be “correct”, other than in the arbitrary sense that we have decided it is.
So the conclusion has to remain that science indeed *cannot* reveal “values” to us, any more than it can decide for us what to do. But if we have decided, for example, that maximisation of well-being (however that is defined precisely) is what we’re after, then certainly science can then tell us (in principle at least) which actions we need to take. (In practice, trial and error is often a more promising strategy.)
Hmm reminds of that Isaac Asimov story where robot/A.I. follwing his rules didn’t allow humanity to do anything because they weren’t allowed to be harmed.
BTW can we even agree on what a maximisation of well-being would entail?
Hi Simon.
From my perspective the answer to your question is, “it will be difficult, but perhaps we can pull it off”. Obviously we’ve been discussing at length issues surrounding who’s well-being we should be interested in (persons? humans? sentient beings?…..), and also teleological considerations that are relevant in defining what we mean by “well-being”. But the very fact that we are discussing these things shows that we are interested, at some level, in exploring the utilitarian-consequentialist position that our moral positions should be guided by the idea that we are trying to maximise well-being, and it is a position that resonates strongly with me.
In relation to the Asimov story, the lesson I would draw from that is a lesson also that has emerged within positive psychology and is reflected in various examples of ancient wisdom (see Jonathan Haidt’s excellent The Happiness Hypothesis in this regard). Namely, happiness is generally not achieved by focusing on it directly. Better to clarify one’s ideas about what kind of person one wants to be, what kind of things one wants to do, and generally how one wants to behave, and then set goals with those ideas (i.e. values) in mind. (Russ Harris’s The Happiness Trap covers this ground excellently in my opinion, from a self-help perspective.) The last thing we want is for some superintelligence to be telling us what to do, however benevolent its intentions towards us.
Btw this reminds me of an article I read a month or so ago quoting the Google CEO as saying that people don’t want google to answer their questions, they want google to tell them what to do next. I’m sure there’s some truth in this; the crucial difference, however, is that we still want to be free to ignore google’s advice.
The important point here is that none of this provides any kind of “reductio ad absurdum” against utilitarianism, as has been suggested; what it does do is to shed light about how well-being can best be maximised in practice, and conversely how our efforts to do so can sometimes go wrong.
Several days ago, Sam Harris appeared as a guest on the Daily show, which provides a humorous attack on the news and some occasionally interesting interviews. Harris said (not in so many words) that he was interested in human flourishing. This Aristotelian turn seems to allow one to come up with some model of human flourishing and then find the kind of ethical regime that furthers it. With this approach, one can determine empirically what furthers flourishing so long as ones model for flourishing is accepted as fact. I suspect that this won’t go all the way — a Chinese or Senegalese or Azeri model of human flourishing will not be a universally accepted one.
Nevertheless, I think the concept of human flourishing is extremely promising as an expression of what, when we manage to strip away various false beliefs and assumptions, think about what we want and are honest with ourselves, resonates with many of us.
A related point is that, much as I have been defending utilitarianism (and “sentientism”), I wonder how many of us really want to be that good. An alternative moral framework, which I have also been exploring recently, would involve some kind of concentric “circles of concern” (phrase borrowed from Covey’s “Seven Habits of Highly Effective People”), in which it is seen as perfectly legitimate to value one’s own welfare and that of a few loved ones above that of strangers in a remote country, without of course any implication that we are in any way “superior” to such people. This may well be far more psychologically realistic than more purely altruistic ethical systems.
Hi Roger. I have two general questions about your thoughts, and they are on the problem of science and ethics and Hume’s Law. The first one is this. Appiah said, commenting on Harris’ new book, The moral landscape, that Harris’ view is that “truths about morality and meaning must ‘relate to facts about the well-being of conscious creatures,’ and science alone — especially neuroscience, his field — can uncover those facts”. Appiah also add that, according to Harris, “values, too, can be uncovered by science — the right values being ones that promote well-being. ‘Just as it is possible for individuals and groups to be wrong about how best to maintain their physical health,’ he writes, ‘it is possible for them to be wrong about how to maximize their personal and social well-being.’” Well, that view, in his general frame, is not an original insight of Harris, since the view that human goodness and well-being can be objectively described was previously sustained also, though in a different Aristotelian and in a non consequentialist guise, by Philippa Foot (specially in “Natural goodness”). Foot´s view is that there is a human natural normativity that can be ascertained not only by introspection, but mainly by observation (including scientific). Facts concerning human goodness can be taken as premises in agent’s practical reasonings; hence, we certainly can, following Foot, derive (not deductively, anyway) an “ought” from an “is”. Judith Jarvis Thomson also stated a similar view (although without references to Aristotle; her references are at most to Peter Geach – but the general view, I think, is almost the same). I’m must confess that I am very sympathetic to those naturalistic and/or objectivistic approaches. Why don’t see normativity in a naturalistic guise? What do you think about Foot’s approach?
The second is on Hume’s Law. I’m not sure that Hume sustained the so-called Hume’s Law. The well-known passage of the Treatise is actually the final paragraph of the first section of the first part of the third book. He begins saying that he couldn’t forbear adding to his previous reasonings “an observation” that he thought may be, “perhaps”, of “some importance”. The observation was that all other theories of morality made statements or propositions “connected with oughts” after established “the being of a God” or “observations concerning human affairs”, without any clear explanation. Hume said that “a reason should be given, for what seems altogether inconceivable”, that is, “how this new relation can be a deduction from others, which are entirely different from it”. His view seems to be that those new affirmations cannot be derived directly from the observations, that is, they cannot be ENTAILED by them. Then he advances his thesis that “the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceiv’d by reason”. That virtue and vice pronouncements are not founded MERELY on the relation of objects means that the mind, from the statements of facts, needs a natural principle that leads or conveys the mind from perceptions to some “new creations”. Farther Hume will say that “sympathy is the chief source of moral distinctions” and that “a sense of morals is a principle inherent in the soul”. The public good, for example, would be indifferent to us, “except so far as sympathy interests us in it”. He will conclude that, when the mind reflects on itself, sympathy again makes the mind reflectively pride of their own principles (including virtue, utility, justice and benevolence, all natural and artificial “virtues”). Hence, his theory is naturalistic and his “subjectivism” on morals is very distinct from a mere subjectivism (David Wiggins has called it a “sensible subjectivism”).
” The claim of the naturalists about rightness in the previous paragraph is, he thinks, rather like the claim that heat is a cabbage.”
I can’t resist to transform that:
But there is heat _in_ the cabbage, if you eat it.
Peter IMO anyway, well-being is constructed personally, culturally and at least some apsects, scientifically as well.But since there are large personal and cultural components I don’t think you could just rely on the science based aspects and then use a broad brush to tell eveyone this is what a good life is.
Thanks for all the interesting points and discussion. A few largely unrelated and quick points:
1) Some natural properties are physical; some aren’t — e.g. biological or social ones.
2) The debate between naturalists and non-naturalists I had in mind is premised on the assumption that there is some kind of mind-independent truth in ethics. Simon Blackburn has suggested that supervenience poses a problem for such views. Myself, I think it’s more of a problem for subjectivism — doesn’t it seem odd that supervenience relations, in this domain, could vary across possible worlds? What might be an example? A world in which causing pointless suffering to non-rational beings is morally required?
3) Note that I employed the claim that the maximization of well-being is what makes actions right only to simplify my argument. With almost all moral views, there’ll be some underlying base for rightness to supervene on, and that is enough to get my position going. Nor am I relying on some particular conception of well-being. Whatever your view of that, there will be some naturalistic, non-evaluative base for well-being (I’m taking ‘well-being’ to be an evaluative term).
4) Circles of concern. See Peter Singer’s *The Expanding Circle* on this. My own view is that it’s hard to draw a non-arbitrary, morally acceptable line, which will distinguish e.g. giving priority to one’s family from e.g. racism or sexism.
5) Foot is indeed a kind of naturalist. But I think she’s not going as far as Harris. The first premises of her position — i.e. certain claims about virtue and human flourishing — come not only from natural science, but also from elsewhere — that is, philosophical ethics. I do think there are some quite serious problems with this naturalist position, brought out well by e.g. Tom Hurka and Brad Hooker.
6) I agree that Hume’s position is open to various interpretations, and is certainly not a version of any kind of crude subjectivism. My hunch is that his attitude to contemporary normative realism would have been quite similar to Bernard Williams’s attitude to claims about external reasons — he would have been mystified.
7) Heat and cabbages. Great point. Indeed Dan Robinson pointed out to me that in a sense everything supervenes on energy. So very weird claims might turn out to be true. But surely we still have to assume that it counts against some claim that it sounds weird?
On this last point Roger, another view in science (which has been gaining ground in recent years) is that everything supervenes on information.
But my main points are the following:
(i) on the circles of concern: I take your point (it’s one that I’ve been discussing at length with SimonJM!) but it may help to draw a sharp distinction between (a) thinking that one’s own group (be it family, race, gender, whatever) is actually morally superior, and (b) to give priority to one’s own group in practice. The latter is basically what most of us do, one way or another: does this make us bad?
(ii) in response to your second point, I think my own position is that moral views are nothing more than mental objects, and therefore only “supervene” on the physical world only in that specific sense. We can no more be “morally required”, in some absolute sense, to cause pointless suffering to non-rational beings (is bull-fighting an example of this?) than we can be “morally required” to avoid suffering. I don’t know if this makes me a “crude subjectivist”: I just haven’t yet observed anything or come across any argument that convinces me that moral views can be considered correct or incorrect in any absolute sense.
Thx Peter
(i) Well, if you’re discriminating without having an acceptable reason, I suppose the answer is ‘Yes’! (Consider someone who discriminates against blacks or women just because they feel like it.)
(ii) What do you think about normativity in logic and mathematics? If there can be requirements there which aren’t just ‘mental objects’, why not in ethics?
On the second point, mathematics = axioms + logic. Given the right choice of axioms, two plus two can equal five. Similarly, logic is basically the application of a certain set of rules (of the type “p implies q is equivalent to not q implies not p” and so on). Presumably one could also choose different rules. What seems to set certain logical rules and axioms from others is that they lead to particularly rich and beautiful conceptual structures, in a way that the others don’t (they just lead to a mess). It’s somehow like finding pearls among the sand.
But this does not quite mean that one mathematical structure is “right” and others are “wrong”. Hilbert spaces are no more or less right than Abelian groups. They’re just different structures. The amazing thing is that some of those structures (but not all, even though they are incredibly beautiful – see Penrose’s “Road to Reality” for a discussion of the relationship between beauty and truth) seem to resonate strongly with the workings of the physical universe at a very fundamental level. It only really becomes a matter of “true vs false” once we make claims about the physical universe, which can then be tested through experimentation.
The way I would characterise the analogy between maths and ethics is thus as follows: as one mathematical system is no more “true” than another, nor is one ethical system more “true” than another. Once you’ve chosen the system, however, including the rules of logic to be applied, you have to follow them correctly: it’s possible then to talk of mistakes in ethical reasoning just as it is in mathematic reasoning. What seems to be lacking in the case of ethics, however, is an analogue of experimentation to sort out which systems correspond to “reality”.
So applying this to racism/sexism, or other forms of arbitrary discrimination (e.g. my family over yours), whether this is “bad” really does depend on which ethical system you happen to be applying. That being said, from a historical point of view racism and sexism have both tended to be associated with false beliefs and half-truths about physical (and biological) reality, that is to say with scientific errors. There will always be bad apples – psychopaths, or people who identify so strongly with a narrow social group that they are blind to the welfare of those outside – unless, that is to say, we manage to genetically-engineer them away, but we should at least be able to avoid the scientific errors that have historically reinforced these phenomena.
Thx Peter. Well, if ethics can have the same sort of status as mathematics, that’s fine with me. I’m not so happy with: ‘ It only really becomes a matter of “true vs false” once we make claims about the physical universe, which can then be tested through experimentation.’ Is this sentence of yours then neither true nor false? If you can know it to be true, it must be through some intellectual capacity or method independent of experimentation. And that capacity, I think, is what enables us to grasp some (not especially contentful) ethical principles.
I take your point here Roger. If I make a claim (such as the one you quote), then implicitly I’m saying that this claim is true, and yet it’s difficult to see how you test it through experimentation. So perhaps what we really have is (i) statements that can be tested experimentally, and (ii) those that can’t. The question then is to what extent it makes sense to regard statements falling into the second category as true or false. A related, but not quite identical, question is what kind of statements we should make and/or believe.
The thing about statements in the first category is that that they are claims about the physical world, which is why they can (at least in principle) be tested experimentally. Mathematical “truths”, by contrast, are basically applications of agreed rules of logic to chosen axioms. In fact the same could arguably be said for ethics, which raises the intriguing prospect that ethics should be regarded as a branch of mathematics! The “intellectual capacity or method independent of experimentation” would then presumably be the ability to define and decide on such axioms, to define and decide on the rules of logic to be applied, and to apply them successfully. In other words, it’s the ability to reason.
There still seems to me to be a difference between a mathematician and an ethicist, however, in the sense that the mathematician does not claim to be saying anything about the world, or indeed about how we should live our lives. The mathematician only claims to be saying something about the world when he applies his or her mathematical reasoning to scientific theories, which are then testable. It’s this last step that seems to be lacking in the context of ethics.
That’s a very nice point about maths, Peter — that we might judge one mathematical system better than another by whether a positive science that uses it does better or worse. So consider one system including an addition rule that says that 7+5 = 12, and another that says that 7 + 5 = 11. The success of a science based on the former speaks in favour of that mathematical account. And in fact we don’t need to go through science. If some mathematician tells me that 7 + 5 = 12, then I think she is telling me that, in any world, if I put 7 objects with 5 objects, there’ll be a set of 12. Likewise, if some ethicist tells me that it always counts against some action that it causes suffering to some non-rational being, she is telling me that this is so in any world. And one can ‘test’ this claim by trying to find circumstances in which this isn’t so. (I don’t think there are any, and so for me this principle is just about as secure as 7 + 5 = 12 — though not quite, because of the possibility that normative egoism might be true.)
But surely there is still a difference between the empirical observation that if I put 7 objects with 5 objects I’ll get 12, and the statement that it counts against an action that it causes suffering. The former we can observe directly, with our eyes. The latter is something that we feel, because of some combination of innate social instinct and upbringing. In the former case we are perceiving something real about the world. In the latter case we are making a value judgement. This value judgement is not a property of the external world, nor does it describe such a property: it is simply a value judgement, which (depending on the integrity of the person holding that judgement) will be mirrored by that person’s behaviour. It is a property of the person, not of the world.
But you don’t observe that 7 + 5 = 12. And, just as you can grasp with your intellect that it counts against an action that it causes suffering to a non-rational being, so you can see that someone actually causing such suffering is doing something they have a reason not to do. At least, that’s how I see it!
I’m not sure that we really *do* “grasp with our intellect” that it counts against an action that it causes suffering to a non-rational being. We can certainly grasp the *idea* that it does it, but what actually determines whether we believe it?
In the case of 7+5=12 I agree that we don’t directly observe the equation itself; what we do do is to learn, normally at school or from some other adult source, the general idea, and then we find that it accords with what we do observe. In the case of actions causing suffering, I think it’s more that we find that the idea accords with what we *feel*. Hence my belief that such ideas are statements about the person holding them, rather than about the (external) world.
(Whether this counts as “grasping with our intellect” depends on how we define “intellect” – does it include emotions or not? – but either way I think the two cases are qualitatively different.)
Comments are closed.