Lecture I. Well-being, welfarism and hedonism
Lecture II. Desire-satisfaction and objective lists
Lecture III. Fitting attitude accounts of value
Lecture IV. Intrinsic value
Well-Being I: Well-Being; Welfarism; Hedonism
Part I: Well-being and welfarism
- Conceptual questions
In ordinary usage, one of the most important senses of the term ‘well-being’ has to do with health; it also brings to mind images of spas and holidays in the south of France. But the most common sense of the term well-being in 20th-century philosophical literature is ‘what is non-instrumentally or ultimately good for a person’; it is about how well someone’s life is going for them. Health (and indeed holidays in the south of France) might well turn out, on the correct theory of well-being, to be constituents of well-being but it would be odd to think they were the only things that mattered.
Well-being is sometimes used synonymously with the following terms: prudential good, prudential value, self-interest, welfare, happiness, flourishing. In philosophical usage, well-being can be both high or low (it’s not paradoxical to say someone has ‘negative’ well-being if their life is going very badly for them).
Well-being is about the good for, not the good simpliciter. As such, it is sometimes distinguished as being a prudential value, as opposed (say) to moral or aesthetic. The martyrdom of a soldier or the beauty of a painting are both good but they are not (or in the first case, not necessarily) good for the soldier/painting.
Questions about well-being are distinct from questions about morality: it is an open question whether (e.g.) my being morally good is good for me. Many theories have claimed that it is, but there are grounds for suspicion on this score and one needs to hear more of the relevant theory before one can endorse such a claim. It certainly isn’t an analytic truth.
- GE Moore’s argument
Moore famously rejected the very idea of the ‘good for’ as irredeemably obscure. He thought the idea of ‘my own good’ or ‘the good for me’ just means that my getting something is good (note: not good for me, just good). Nothing is added to this statement by saying that it is not just good but my good or good for me.
Moore’s argument as thus stated probably begs the question. It assumes that all well-being claims are really claims about impersonal good made in a misleading way. But it is not clear that this is more than rhetorical bluff. Why not say just the opposite: that nothing (no state of affairs, no possible world etc) is good unless there is someone for whom it is good. Most (how many?) statements about the impersonal good can be rephrased as claims about welfare. A counter-Moorean philosopher could claim that nothing is added to the claim that something is good for me by saying that my getting something is good.
No plausible moral theory could ignore well-being entirely. Joseph Raz, for instance, proposes a kind of constraint on moral theories when he says ‘the explanation and justification of the goodness or badness of anything derives ultimately from its contribution, actual or possible, to human [NK: why only human?] life and its quality’ (The Morality of Freedom, p. 194). This is one way of stating a view now called ‘welfarism’, the claim that any moral reason has justificatory force insofar as it involves some contribution to the well-being of human beings (and indeed other sentient organisms).
Act-utilitarianism is committed to a form of welfarism: the one thing that is non-derivatively good is well-being; every other moral claim (about duty or justice or rights or whatever) is admissible only inasmuch as it is ultimately a claim about well-being.
This is an ambiguous claim and this ambiguity is exposed in debates about (say) the value, if any, of equality. Recall the well-known ‘levelling-down objection’ to egalitarian positions in political philosophy: could an equal-but-poor society be better than a slightly-unequal-but-rich one even if there is no one for whom it is better?
- Arguments against welfarism
The usual strategy against welfarism is to show that there is some intuitively appealing value that cannot be accounted for in terms of welfare. E.g.
Autonomy. If welfare is the only value, on what grounds if any can one oppose paternalistic state intervention?
Knowledge. Recall the problem of adaptive preferences: could it be better to be a pig satisfied than Socrates dissatisfied?
Note that the plausibility of welfarism often turn out to depend on the particular theory of welfare one adopts; indeed, a commitment to welfarism might constrain what theory of welfare one endorses.
Part II: Hedonism as a theory of well-being
- Theories of Well-Being
Well-being: what is (non-instrumentally or ultimately) good for a person. As opposed to value, what is (non-instrumentally or ultimately) good simpliciter. The two notions are sometimes intertwined, as in utilitarianism, which regards the most valuable state of affairs as one in which well-being is maximised. The notion plays a crucial role in other areas of inquiry and life: e.g. political philosophy, medicine, (welfare) economics…
Theories of well-being are attempts to answer the question of what well-being consists in. Standardly, theorists distinguish between hedonist theories, desire theories, and objective list theories.
Psychological hedonism: The claim that human beings always pursue what they judge will give them the greatest pleasure, or the greatest balance of pleasure over pain; an explanatory thesis about human action.
Our topic in these lectures is evaluative or prudential hedonism: The claim that well-being consists in the greatest balance of pleasure over pain. Any other determinants of well-being are so only instrumental, i.e. insofar as they contribute to producing the greatest balance of pleasure over pain. More crudely, it is sometimes stated as the thesis that pleasure is the only or most basic good.
Hedonism of this sort has a long history. It is argued for (arguendo?) by Socrates in Plato’s dialogue the Protagoras (351bc). In the late 18th-century, the idea reappeared in the philosophy of the English utilitarians. The first line of Bentham’s Introduction to the Principles of Morals and Legislation reads, ‘Nature has placed mankind under the governance of two sovereign masters, pain and pleasure. It is for them alone to point out what we ought to do.’
- The Experience Machine
Suppose there was an experience machine that would give you any experience you desired. Super-duper neuropsychologists could stimulate your brain so that you would think and feel you were writing a great novel, or making a friend, or reading an interesting book. All the time you would be floating in a tank, with electrodes attached to your brain. Should you plug into this machine for life, preprogramming your life experiences? […] Of course, while in the tank you won’t know that you’re there; you’ll think that it’s all actually happening […] Would you plug in? (Robert Nozick, Anarchy, State, and Utopia, 1974: pp. 44-45)
What, honestly, do we feel about this? If we’re reluctant to ‘plug in’, why is this?
- Why not plug in?
We want to do certain things and not just have the experience of having done them. We want to be certain people – to plug in is to commit a form of “suicide” (613). We are limited in the experience machine to a human-created reality. A valuable life involves possessing certain character traits, the exercise of certain capacities, having relationships to others and the world, none of which seems obviously reducible to psychological states alone.
Nozick thinks our responses to the prospect of plugging in to the experience machine help us discover that there are things which matter to us more than simply having certain experiences, e.g. veridical experiences, real agency, risk…
The argument seems to be something like this: If all that mattered to us was pleasure, then we would want to plug into the experience machine. However, we do not want plug-in. Hence, there must be things which matter to us besides pleasure. Therefore, we are not (and maybe could not be?) hedonists.
- Hedonist responses?
What is the significance of the common reluctance to plug into the experience machine? Presumably, our intuitive or commonsensical judgments about what we value. But why endorse those judgments? Maybe, the hedonist could say, our having those intuitions is good for us precisely because they make a more hedonistic life possible.
Recall the ‘paradox of hedonism’: pleasure is best pursued by pursuing something other than pleasure itself. If we believe that pleasurable activities are valuable independently of the pleasure we gain from engaging in them, then we shall probably gain more pleasure overall. So it is, from a hedonist point of view, valuable for us not to believe in hedonism. But it doesn’t follow that the view couldn’t be true nevertheless.
Or, maybe our intuitions are the product of status quo bias: what if you were told you were already living in an experience machine: would you unplug? If you are actually a banker living in Monaco? If you’re a multi-millionaire? What is the significance of the fact that a lot of people (50% and more) who have been asked this question wouldn’t want to unplug?
Well-Being II: Desire satisfaction and objective-lists
- Two Components of Theories of Well-being
Substantive: Which things will make me well-off?
Explanatory: What makes the things said to make me well-off good for me?
A full hedonist holds two things: that pleasurable experiences contribute to my welfare insofar as they are pleasurable; further, it is their being pleasurable that makes them good for me. We considered some of the better-known objections to this kind of view (e.g. Nozick’s ‘experience machine’).
But we can imagine a theorist of welfare who endorses the hedonist’s substantive claim but rejects the explanatory addendum: ‘Welfare consists in pleasurable experiences but what makes these experiences valuable for a person is not their being pleasurable but their fulfilling her desires.’ Alternatively: ‘Welfare consists in the fulfillment of desires but what makes the fulfillment of desires valuable is that this fulfillment is pleasurable for the person.’
According to a full desire of account of welfare, welfare consists in the fulfillment of desire, and what makes the presence of desired items in a life good for one is their fulfilling one’s desires. (‘Desire’ here is a catch-all term that is supposed to cover similar pieces of terminology, such as ‘preferences’.)
What’s at stake here? Among other things: something like this kind of account has generally been assumed in much welfare economics of the twentieth century and has thereby informed a great deal of government policy. It’s important to know if the theory these interventions assume is philosophically robust.
- Present desire accounts of welfare
It’s helpful to start thinking about these accounts by starting with a deliberately crude statement of the view and working our way slowly to something more plausible. The simplest form of a full desire account of welfare is a present desire account: a person’s welfare consists in the fulfillment of her present desires.
This view raises obvious problems: Otherwise rational human beings are often irrational at particular times. It is strange to think someone who shoots themselves in a passing moment of anger or despair is advancing their own welfare insofar as they are fulfilling an (ephemeral) desire of theirs. Further, insofar as the desire in question is fleeting, accounts of this kind don’t seem to be accounts of welfare understood as pertaining to whole lives, as opposed to ‘time-slices’ of individuals. This might lead us to prefer…
- Comprehensive full-desire accounts of welfare
According to such accounts, welfare consists in the maximal fulfillment of a person’s desires over their life as a whole. But this immediately raises difficulties of interpretation. How do I use this account to evaluate how much welfare there was in a life?
Suppose we try a summative view: the best life for me will be that in which the largest number of (intense etc) desires is fulfilled. This is vulnerable to an obvious counterexample (paraphrased from Parfit’s Reasons and Persons, p. 497):
What if I offered you a life where I got you addicted to a drug that filled you with an intense desire for that drug every morning, which I would reliably fulfill. Surely you should be delighted by the offer; I’m giving you the chance to have a large number of very intense future desires fulfilled regularly! But you might understandably prefer not to get addicted to the drug in the first place.
Which leads us to…
- Global full-desire views of welfare
Maybe: my greatest welfare consists in the fulfillment of my desire to live in the way I most prefer. I am better off the higher up my ordering of preferences my actual way of life is. (Note how this dodges the objection above.) Again, this seems to fall foul of this kind of objection: what about someone who chooses a monastic life over that of a cook (say) only because he was unaware that he could have been a miner, which life he would have vastly preferred to the monastery had he only known about its existence. Maybe we need to introduce some kind of epistemic component as well…
A note on method: Note the incremental method we’ve been using in this inquiry; we are looking for an account of welfare that is consistent with the largest number of our considered or intuitive judgments on the question. Where our theories conflict with our intuitions or considered judgments, we either abandon/modify the theory or reconsider the intuition/judgment. For each of the views abandoned so far, we can reasonably ask whether we shouldn’t instead have reconsidered the intuition on which our rejection of the view was based.
- Informed desire accounts
Maybe: the life that would be best for a person is the one she would desire if fully informed of the facts about the various options available to her, and her greatest welfare would consist in the fulfillment of that…
But: even under conditions of full information, even one’s own judgements about one’s welfare may deviate from what one most desires. We often desire things we ourselves deny are good for us. (Fill your own example here.) What then?
- Counterfactual/idealised desire accounts
Maybe: welfare consists in whatever would fulfill the global desires I would have if I were rational and fully informed.
But, consider someone (example based on Rawls’s A Theory of Justice, p. 432) who, while fully rational, informed and capable of extensive achievement in other areas, strongly desires a life dedicated to counting blades of grass. Is a life of grass-counting the best life for her? Surely this is some kind of neurosis.
But why think that? Recall our earlier remarks on method: we’re looking for a theory that is consistent with, or sufficiently intuitive or explanatory or rationally authoritative to overrule, our intuitive or considered judgments about welfare. But this means that the selection of any particular desire account is shaped by a prior conception of what is good and bad for people.
But desire accounts haven’t been doing very well on that score so far… Why then are they so attractive to (e.g.) welfare economists?
- The appeal of desire accounts
Anti-paternalism. We like the sound of the idea that decisions about a person’s welfare should be left up to them; anything else sounds paternalistic and potentially authoritarian. But note that this runs together two questions: What is good for someone? And how should governments determine what is good for someone? A liberal could endorse a desire-based answer to the second without endorsing it for the first. The fact that my desires can be unreliable guides to what’s good for me doesn’t mean that governments have anything better they can use instead.
Interpersonal comparisons of welfare. Economists and policy-makers need to be able to make interpersonal comparisons of welfare and desire-based views offer a way of doing this. But how good a reason is this to endorse a desire-based view? Just because desire-fulfillment is easier to measure doesn’t mean it’s what’s actually good for us. Why assume that the true account of welfare will also be the easiest to quantify? (Note also that it’s not all that easily to quantify and rank desires –
Conflation of the substantive and explanatory. It’s often true that the satisfaction of people’s desires makes them better off, because desire is often for what is good for one. So desire accounts will often ‘pick out’ intuitively correct examples of things that advance our welfare. But this is a long way off from what desire accounts further assert: that what makes the desired things good for us is their being desired by us.
- The concept of desire as not-necessarily-good-making?
When someone desires (say) a saucer of mud, we naturally demand explanations, or conclude that these are not so much desires as irrational cravings. Why do these cases puzzle us? Why are we not able just to concede that the fulfillment of these cravings too would advance someone’s welfare?
Maybe because in normal cases, we desire something because we think it will be good in some way independently of its satisfying the desire. When we try to give an account of why we have the desires we do (the why of reason-giving, not of mere causal explanation), we advert to something beyond the satisfaction.
Consider the weirdness of this answer to the question of why you desire X: ‘I desire X because having X would satisfy my desire for X’. The problem is not that this is false but that it fails to give any kind of reason for desiring X. And when one is pushed to give a less empty answer to the question, we are compelled in answering to say something that rules out our accepting a desire-based account of welfare.
But if this is so, then it is part of our basic concept of desire that being desire-satisfying cannot itself be a good-making property. We’ll consider this objection in a different form and in greater depth next week when we think more generally about the nature of value and its relationship with reasons. For the moment, let’s consider instead an alternative to desire-based views of well-being, i.e. ‘objective list’ theories.
- 9. Characterising an objective list theory
Objective list theories are much harder to characterise than hedonist or desire-fulfillment theories: several theories go under that label, and the label itself is used imprecisely as a kind of residual or ‘dustbin’ category for any view that isn’t obviously hedonistic or desire-based.
Paradigmatic object-list theories endorse at least these two claims:
Attitude-independence: Things can be non-instrumentally good for some subject independently of whether that subject has some pro-attitude towards G.
Pluralism: There exists a plurality of (non-instrumental) prudential goods.
(NB. Prudential goods are things that are components of a subject’s well-being.)
Of course, there is room in logical space for possible objective list theories that don’t endorse both views; in particular, there could be monistic non-hedonistic attitude-independent objective list theories (e.g. ‘knowledge is the only good’).
- Some putative objective lists
Here are some list of basic prudential goods given by a few well-known objective list theorists.
Finnis. Life, Knowledge, Play, Aesthetic Experience, Sociability (friendship), Practical Reasonableness, ‘Religion’.
Murphy. Life, Knowledge, Aesthetic Experience, Excellence in Play and Work, Excellence in Agency, Inner Peace, Friendship and Community, Religion, Happiness.
Parfit. Moral goodness, rational activity, development of abilities, having children and being a good parent, knowledge, awareness of true beauty.
- Considerations in favour of objective list theories
Pre-philosophical judgments. People tend to be committed, implicitly, to something like this view before they are persuaded on reflection to adopt an alternative. Our answers to questions about well-being tend to be ‘objective list’ in style. It might also be reflected in our naïve practice more than alternative theories. Of course, this evidence is weak: it could be that other theories ultimately explain unreflective practice better than objective list theories; further, they might offer a more robust explanatory basis for that practice. This remains to be seen.
Improvement on hedonism. Hedonism, insofar as it is vulnerable to something like the ‘experience machine objection’ recognises too few goods. Objective list views can accommodate some of what we feel is left out of hedonist theories.
Improvement on desire-based views. Desire-fulfillment views seem to yield too many goods: objective list theories can avoid some of the more implausible corollaries of desire views by asserting attitude-independence.
Piecemeal arguments for individual goods. By arguing for various distinct things being irreducibly, non-instrumentally good, we might find ourselves willy-nilly committed to an objective list view simply because the goods we want to recognise are either not compatible with hedonism/desire-based views or are not plausibly explained by those theories.
- Objections and replies
Explanatory impotence. Objective list theories, insofar as they are pluralistic, are vulnerable to the charge that they are explanatorily unsatisfying: why are these, and only these, items on the list? What unites them in one category?
Reply. Companions in guilt: ultimately, hedonist and desire-fulfillment views must also answer the explanatory challenge. As for the pluralism, maybe that’s just where the arguments lead and the apparent messiness of the objective list correctly reflects the messy nature of value. The argument for an objective list theory is as good or bad for the arguments in favour of each item on the list. (And a counter-challenge: why expect an account to be unified anyway?)
Elitist. Objective lists could be grounds for paternalistic coercion.
Reply. Not necessarily, or not just because they are objective lists; we need a separate argument for paternalism.
Insufficiently subject-sensitive and therefore alienating. ‘It would be an intolerably alienated conception of someone’s good to imagine that it might fail in any way to engage him’. [Peter Railton, ‘Facts and Values’, in Facts, Values and Norms (Cambridge: Cambridge University Press, 2003), 19]
Reply. Not all objective list theories: maybe the correct list of prudential goods is at least partly constituted by the subject’s affective states (e.g. Fletcher’s hypothetical four-good list: ‘pleasure, happiness, friendship, achievement’). Is this still an objective list? Or, alternatively, is it sufficiently subject-sensitive? What if a certain subject doesn’t want these things? But note that it would simply beg the question against an objective list view if we assume that any plausible theory of welfare must reject attitude independence. We might be at a dialectical impasse here. Is there more that we can say about this debate taken in isolation?
General reading on this topic: Guy Fletcher, The Philosophy of Well-Being: An Introduction (Abingdon: Routledge, 2016); Guy Fletcher (ed.), The Routledge Handbook of Philosophy of Well-Being (Abingdon: Routledge, 2016).
Next week: What is value? Recommended reading. TM Scanlon, ‘Values’, in What We Owe to Each Other (Cambridge, MA: Harvard University Press, 1998), 78–107.
Well-Being and Axiology III: What is value?
- Why care about value?
Pretty much everyone agrees that moral philosophy has something to do with goodness. So far, we’ve been discussing one part of this, i.e. what is good for someone? In other words, we’ve been asking about the nature of prudential value. But prudential value may not be the only kind of value; and the very idea of ‘kinds’ of value seems to presuppose a more basic idea of what it is for something to be good, or have value (most of the philosophers writing about this take these words/phrases to be interchangeable).
It seems obviously important to say what it is for something to be ‘good’. One way of doing this is by reflecting on the structure of the word we’re treating as a synonym for good, i.e. valuable. Note the structure: value-able. But like ‘desirable’, ‘detestable’, ‘damnable’ (but unlike ‘visible’ or ‘edible’ or ‘audible), the notion is not modal but normative. The valuable is not what can be valued, but what ought to be valued (what one has reason to value, what it is appropriate to value, or something else along these lines).
This way of putting it makes it natural to think that this is (or can yield) an analysis of value. What it is for something be good is for it to have value; what it is for something to ‘have value’ is for it to be such as ought to be valued; and ‘valuing’ (the verb) is the name of an attitude. This gives us what’s commonly called a ‘fitting attitude (FA) account’ of value.
In this lecture, I’ll discuss this account, or rather set of accounts, of value. I’ll contrast it with its most appealing rivals and try to motivate it out of its rivals’ failings. In particular, I’ll be discussing a form of this view developed by TM Scanlon. Then I’ll discuss two of the most serious arguments against Scanlon’s version of the view and ask whether they can be satisfactorily answered.
- Robust value realism vs FA accounts
A robust realist about value holds that values are, or are like, ‘primary qualities’: they exist independently of human responses to them. This means that evaluative attitudes (deeming desirable, appreciating, applauding…) are understood as being sensitive to values that exist independently of those attitudes. Roughly: valuing answers to value, not the other way around.
As opposed to this, FA theorists generally (though not always) hold two things:
(1) Normative reduction: The evaluative can be reduced to the deontic. (Examples of deontic terms: reason, duty, obligation, ought, must…
(2) Response-dependence: Value is at least partly constituted by human attitudes and responses.
- Buck-passing accounts of value
The idea of ‘passing the buck’ about the good/value is most prominently defended in TM Scanlon’s What We Owe to Each Other. We can understand Scanlon’s account as a marginal member of the family of FA accounts (it shares some important elements with other members of that family, though it is importantly distinctive). The crucial passage is this:
[B]eing good, or valuable, is not a property that itself provides a reason to respond to a thing in certain ways. Rather, to be good or valuable is to have other [natural?] properties that constitute such reasons. Since the claim that some property constitutes a reason is a normative claim, this account also takes goodness and value to be non-natural properties, namely the purely formal, higher-order properties of having some lower-order properties that provide reasons of the relevant kind. It differs from the first alternative simply in holding that it is not goodness or value itself that provides reasons but rather other properties that do so. For this reason I call it a buck-passing account. [What We Owe, 97]
The central idea here is that ‘goodness provides no reasons of its own’; it is a sort of proxy for other things which are valuable. It would be a mistake on Scanlon’s view to say ‘I value this because it is good’. Further, his view involves what we’re calling the ‘normative reduction’ of the evaluative: we understand value in terms of reasons to respond in certain ways (what ways?)
Understanding the value of something is not just a matter of knowing how valuable it is, but rather a matter of knowing how to value it—knowing what kinds of actions and attitudes are called for. [What We Owe, 99]
- The Wrong Kind of Reason?
Consider this case:
a malicious demon will impose a severe punishment unless we desire a saucer of mud. Or, to vary the example, the demon will punish us unless we make him an object of our admiration. His threat of punishment gives us reasons for the relevant pro-attitudes, but it does not make the demon admirable or the saucer of mud desirable. [Wlodek Rabinowicz and Toni Rønnow-Rasmussen, ‘Buck-Passing and the Right Kind of Reasons’; example originally from Roger Crisp]
The basic idea is this: if we’re trying to analyse one thing in terms of another, the analysans shouldn’t have a different extension from the analysandum. The trouble is that if we analyse value in terms of (any kind of reasons), we have to count as valuable all sorts of things which intuitively aren’t valuable. A natural way to put the point is this: in the cases with the structure of the example above, we have the wrong kind of reason to take pro-attitudes to something; that’s what explains our not wanting to accept that those things could be valuable.
What we need is a principled way of distinguishing the right kind of reason from the wrong kind. The most obvious way to do this so that we’ll get the right extension is this: we have the right kind of reason to take a pro-attitude to something when that thing is valuable. But this gets the right extension at the cost of being trivial, or worse, viciously circular. Either we’re saying – boringly – that in all cases of genuine value, one has the right kind of reason to take a pro-attitude to it – in which case we’re not really given an analysis. Or, we’re trying to give an analysis and saying that we should understand the right kind of reason in terms of value, but we were trying to understand value in terms of the right kind of reason in the first place. We haven’t really achieved an analytic reduction. (Maybe this isn’t so bad? Circular claims could still get us to see important facts about some conceptual structure; not all illuminating accounts need be reductive.)
- State-given reasons vs object-given reasons?
There’s another, more promising, way of making a principled distinction between the right and wrong kind of reason. The right kind of reason is ‘object-given’ – it is grounded in features of the object of the attitude (e.g. evil demon, saucer of mud). The wrong kind of reason is ‘state-given’ or ‘attitude-given’ – it is grounded in features of the mental state or attitude. (Terminology from Derek Parfit, ‘Rationality and Reasons’.)
The distinction is roughly this: the pro-attitude you have reason to take towards the saucer of mud or demon is one you have reason to take because of something about taking that attitude (e.g. it preventing pain or making you rich). But in a case of genuine value, the reason you have to take a pro-attitude to something is grounded in something about the object of the pro-attitude.
Can this be made to work? It may depend on a more general metaphysical question about the nature of properties/attributes/features. It’s not very hard, with only a little verbal ingenuity, to redescribe all attitude-given reasons as object-given reasons:
it is easily seen that properties of the attitudes may be recast as properties of the objects; if, e.g., the attitude of preferring the saucer of mud has the property of preventing our suffering severe pain, then the saucer of mud has the corresponding property of being such that preferring it would prevent our suffering severe pain. [Jonas Olson, ‘Buck-Passing and the Wrong Kind of Reason’]
Do we have a good account of properties that can help us to get around this? If not, maybe we either need to give up on buck-passing accounts of value, or try to argue that the circularity mentioned above isn’t of a vicious kind.
Next week: What, if anything, is intrinsic value?
Recommended reading: Christine Korsgaard, ‘Two Distinctions in Goodness’, Philosophical Review, 92 (1983): 169-95.
Well-Being and Axiology IV: Intrinsic value
- Intrinsic value: the ordinary concept
The phrases ‘intrinsic value/goodness’ and ‘intrinsically valuable/good’ appears often enough in ordinary discourse. Sometimes the word ‘intrinsic’ is replaced by phrases such as ‘in itself’, ‘in its own right’, or ‘for its own sake’. Let’s consider the latter phrase first; the phrase ‘for its own sake’ appears more naturally with the verbal constructions (valued, valuable). We can get at the idea by starting with something we judge to be valuable and asking why it’s valuable, and then to ask again…
Why is X valuable? Because it brings about Y (or for the sake of the Y it produces). Why is Y valuable? Because it leads to Z. Why is Z valuable? It just is; it’s valuable for its own sake.
‘Intrinsic value’ is sometimes taken to mean whatever appears at the end of the explanatory/justificatory chain; everything else is only extrinsically, instrumentally, derivatively good. This is how it’s conceived in much Greek philosophy, e.g. Glaucon’s division of goods in Book II of Plato’s Republic, and the argument for happiness being the final end and ultimate good in I.2 of Aristotle’s Nicomachean Ethics. But the adverbs above aren’t obviously synonymous and seem to make more than one distinction.
- Korsgaard’s Two Distinctions
Christine Korsgaard (‘Two Distinctions in Goodness’) thinks it’s confused to contrast ‘intrinsic value’ with ‘instrumental value’. She thinks the two notions are orthogonal, i.e. form part of different contrasts. Instead she proposes that make two separate distinctions:
Intrinsic Value: An object x is intrinsically valuable iff it is valuable in a way that supervenes only on the intrinsic properties of x.
Extrinsic Value: An object x is extrinsically valuable iff it is valuable in a way that supervenes – at least to some extent – on the extrinsic properties of x.
Final Value: An object x is finally valuable iff it is valuable as an end.
Instrumental Value: An object x is instrumentally valuable iff it is valuable as a means.
As Rae Langton (‘Objective and Unconditioned Value’) puts it in her comment on Korsgaard’s paper: ‘“intrinsic” comes from the distinction in ways things have value, whereas “instrumental” comes from the distinction ways in which we value things.’ 
- Moore vs. Kant
On Korsgaard’s view, there are two excesses to which one might be drawn depending on whether the ‘valuable’ in ‘intrinsically valuable’ is read as ‘can be valued’ or ‘ought to be valued’.
(1) Subjectivism: To have intrinsic value is just to be valued as an end.
(2) Realism: One ought to treat as ends those things that have intrinsic value; intrinsic value is attitude-independent. (Roughly, GE Moore’s view.)
(1) is too subjective, (2) too objective. But Korsgaard thinks there is a sensible via media, viz. Kant. Kant has a view of the value of the rational will such that it can ‘confer’ value. This gets us to a kind of objectivity, but without Moorean attitude-independence.
As she puts it, on a Kantian view, ‘the good end is the object of a rational choice. The things that we want, need, care for, are good so long as certain conditions of rational choice are met. Thus, the reasons that things are good bear a definite relation to the reasons we have for caring about them.’ 
The attraction of this view is that ‘it gives an account of the “objectivity” of goodness that does not involve assigning some sort of property to all good things. Good things are good […] because of attitudes taken up towards them or because of other physical or psychological conditions that make them important to us.’ 
This means that there is a special human power, a rational power, to confer value onto the world. As Korsgaard puts it:
If human beings have an intrinsic value by virtue of the capacity for valuing things, then human beings bring goodness into the world. The distinction between a thing that is intrinsically good and a thing that is extrinsically good yet valuable as an end allows for the possibility that the things that are important to us have an objective value, yet have that value because they are important to us. 
- Euthyphro dilemma
Is this objectivity enough? Langton urges a version of the old Euthyphro dilemma against Korsgaard, and notes that Korsgaard is evidently embracing one of the horns of that dilemma in claiming that things are valuable because we, rational beings, value them. But why do we get to confer value on things?
A possible answer: Because we (possessors of rational wills) are intrinsically valuable. But why would that give us the power to confer value on other things? As Langton notes, the painter of blue things doesn’t have be blue; the creator of babies doesn’t have to be a baby.
At the heart of the disagreement between Langton and Korsgaard is that between ‘realist’ and ‘constructivist’ interpreters of Kant. Insofar as some of these interpreters are also Kantians, the interpretative debate feeds a philosophical one: about whether we should be constructivists or realists about morality, and such central moral notions as value.
- Constructivism about intrinsic value: still intrinsic?
On a constructivist view, claims about intrinsic value turn out to be claims about the rationality of valuing things intrinsically. Recall the earlier definition: ‘An object x is intrinsically valuable iff it is valuable in a way that supervenes only on the intrinsic [non-evaluative] properties of x.’
The question for a constructivist is when it makes sense to value things in this way, i.e. when the ‘conditions of rational choice are met’. This is, as Langton would have it, a kind of ersatz objectivity, a second-rate kind of intrinsic. But Korsgaard would presumably reply that this is all we’re going to get.
Korsgaard’s Kant (as opposed to the historical, realist, Kant?) may still turn out to have a number of allies in the history of philosophy.
– Knowledge as intrinsically valuable (think of Edward Craig’s ‘genealogy’ of the concept of nature in Knowledge and the State of Nature)
– Justice as intrinsically valuable (think of the argument in Plato’s Republic that we must value justice intrinsically in order to get some of its extrinsic benefits; also Hume’s account of justice as an ‘artificial virtue’)
– Truthfulness as intrinsic virtue (think of Nietzsche’s argument in The Gay Science and elsewhere for this element of the ‘old metaphysical faith’; or of a latter-day Nietzschean Bernard Williams’s genealogical account of the intrinsic value of sincerity and accuracy in Truth and Truthfulness).