Dan Brock: Quality of Life Measures in Health Care and Medical Ethics
Dan Brock: Quality of Life Measures in Health Care and Medical Ethics
Abstract and Keywords
Griffin expresses doubts regarding Brock's thesis that one can extract from medical ethics literature the outline of a general account of a good life. Griffin contends that while we need prudential values as the broad conception of the quality of life for our reasoning about the good life, there are moral and practical reasons to work with a narrower conception of the quality of life in making certain social decisions. If it is the case that we need several different conceptions of the quality of life for different kinds of social decision, then we need both the broad conception and an understanding of the various considerations at work in generating narrower conceptions.
Professor Brock asks, What does the literature of medical ethics have to tell us about the quality of life? He answers, Not much. That is not surprising. What makes up the quality of a life? It is a tough question. Doctors' dilemmas make it urgent, but not easier. And medical practice cannot do without settling at least a bit of theory. Medical administrators need to know to what extent the values that go to make up the quality of life are commensurable, and how commensuration in their case works, and how interpersonal comparisons work too.
Can these administrators get by with only a fairly narrow conception of well‐being? Well, the aim of medicine is health, and it would be wrong, I think, to let the notion of ‘health’ swell to include everything that bears on the quality of life. Which patients should get scarce treatment? Should we do fewer expensive heart transplants and more inexpensive hip replacements? When should we pull the plug? Medical decisions — at any rate, decisions about when, and how far, and in what way, and for whom, to pursue health — need to use much wider normative notions than that of health alone. Brock says: ‘my concern will be with the broadest conception of . . . “what makes a life go best” ’. To my mind, he is right to be concerned with the broadest conception. But perhaps we should also be concerned with narrower conceptions. Perhaps the notion of the quality of life fragments into several notions appropriate to different sorts of social decision, even decisions about health. It is wrong of us, I think, to expect one notion of the quality of life to be up to answering all the questions that we try to answer with it. I shall come back to this shortly.
Brock thinks that we can extract from the literature on medical ethics ‘the main outlines of a general account of a good life’. I think that the literature suggests his account no more than it suggests any one of many others. Still, his account is an interesting possibility. What I want to do is to express doubts about it.
Brock divides theories of a good life into three: (1) hedonist, (2) preference satisfaction or desire fulfilment, and (3) ideal. He does not think that we should force a choice between them but, rather, give independent place to items from each, as (he says) ideal theories can do. So Brock's is a kind of ideal theory. It can also be seen as a ‘vector view’ of the sort that Amartya Sen anatomized in his article ‘Plural Utility’.1 Brock goes on to identify four ‘components’ of a good life: (p.134)
1. primary functions (such as mobility and communication, which are used in carrying out just about any life plan);
2. agent‐specific functions (ones that are used in carrying out one agent's life plan in particular);
3. desire fulfilment, and
Then, in a particularly fundamental place, there will be autonomy.
Now these items are very different in category. Happiness and autonomy are substantive prudential values. Functions, as Brock defines them, however, are not; they are characteristic means to realizing such values. And desire fulfilment, I should say, is not a substantive prudential value at all but rather a meta‐ethical feature of those values. So, Brock's scheme does not touch rock bottom (that is, intrinsic prudential values), and I think, for reasons that I shall come to, that we ought to touch ground before taking off again.
Take desire fulfilment. Brock seems to me to treat it as a prudential value. At least, he speaks at one point of autonomy's being perhaps only instrumentally valuable, valuable because it leads either to happiness or to desire fulfilment. But I think that desire fulfilment is at some remove from prudential values. The desire fulfilment account of value, as Brock points out, has to be concerned with desires that are corrected in some way, because the fulfilment of actual desires comes nowhere near being a plausible candidate as a prudential value. But once we start talking about corrected desires, once we ask how stiff a demand ‘corrected’ represents, we move into territory where old maps do not help a lot. Now, how stiff a requirement should it represent? The rough idea of a corrected desire is one formed by an appropriate appreciation of the nature of its object, so it has to include whatever is necessary for that. ‘Corrected’ might be taken (as, for example, Richard Brandt takes it2) to require only that desires survive criticism by facts and logic, where it is meant to be a question of fact — largely of psychological fact — whether a desire is corrected. But a particularly irrational desire — say, one planted deep when one was young — might well survive criticism by facts and logic, and its mere persistence does not seem to guarantee that its fulfilment will make one better off. For instance, I might wish to hog the limelight on all occasions. I might have learned from long experience that succeeding does not work to my advantage, but still want to. I might not react appropriately, or strongly enough, to this important piece of self‐knowledge. So ‘corrected’ cannot mean just ‘formed while possessed of factual knowledge and correct logic’; it must mean something stronger such as ‘formed in proper appreciation of the nature of the object’.
There are appropriate responses to certain pieces of knowledge, and ‘appropriate’ cannot here be just ‘most common’ because most of us sometimes go on wanting certain things — self‐assertion, say — too much even when our desires (p.135) are formed while possessed of factual knowledge and correct logic. ‘Appropriate’ has to mean something close to ‘correct’. But this shifts importance away from the mere presence of a desire to the proper appreciation of its object. It gives us, I should say, some sort of ideal theory of prudential value. In that I agree with Brock. Where I disagree is with his regarding desire fulfilment as itself one of the values. Desire fulfilment seems to me best seen not as a substantive value, but as a formal notion of what it is for something to be prudentially valuable: it is for it to be (not entirely trivially) the object of a sufficiently well informed desire. A person can move from a desire's being unfulfilled to its being fulfilled and become a lot worse off as a result. He can make the same move, even with a fully enlightened desire, and be no better off. It is not the state of a desire's being fulfilled that is valuable. What are valuable are certain objects of desire. True, one can see various things that might make it seem that desire fulfilment is itself a prudential value — the frustration, for instance, that sometimes comes from non‐fulfilment, the boost that can, although need not, come with fulfilment, and the autonomy that we respect in respecting the fulfilment of a person's actual (not corrected) desires. And frustrations, boosts, and autonomy all have to do with substantive values. Desire fulfilment counts towards the quality of life, not as such, but because when suitably corrected and restricted, desires are linked to objects subsumable under some desirability characteristic (autonomy, enjoyment, deep personal relations, accomplishing something with one's life, etc.). This, of course, changes the face of desire accounts a lot. But this is what they have to become for there to be a plausible link between fulfilment of desire and prudential value. I must apologize for whizzing through, and dogmatizing about, immensely difficult matters. I have, however, discussed much of this elsewhere.3
Take another item from Brock's list of the ‘components’ of a good life: functions. What Brock refers to as ‘functions’ are not, I take it, the same as what Sen calls ‘functionings’ or, as Sen also puts it, achievement of ‘beings’ and ‘doings’.4 Brock, I think, is interested in the possession of certain pretty basic abilities, while Sen, though his list includes those, wants to include much more besides — for example, the exercise of decidedly non‐basic abilities. So Sen's category is so wide that it includes both means to substantive goods and the substantive goods themselves.
Now, what one includes in an account of a good life will, I think, depend upon what the notion of a good life is used for. We need a full notion in taking decisions about how we want to live our own lives, and a doctor probably needs the same full notion in taking decisions about what the best trade‐offs are for a particular patient. It is, I think, a different, perhaps narrower, notion of a good life that comes into play in many of a government's decisions (p.136) about how it ought to allocate resources. And it is not a notion of a good life at all, but some accessible, tolerably reliable indicator of it, that we often have to fall back on. What Brock calls primary functions, for instance, are so central to most people's lives that it is entirely reasonable for them to feature in ‘Sickness Impact Profiles’ (Table 1). Even though they are just means to prudential values, they are accessible and measurable, and the link is close. So it makes sense to give them prominence in measures of quality of life that we must carry out in everyday settings. Still, they are at some remove from actual prudential values. The same handicap — say, the loss of a finger — can devastate some lives but have only minor cosmetic disadvantages in others. Medical decisions have often to be tailored to a particular patient; often it is only having seen the values at stake in that life that we can decide what to sacrifice for what. We have often got to get behind functions to the real prudential values that they are means to.
Suppose, then, we were to take not ‘functions’, as Brock uses the term, but the much more capacious class of ‘functionings’, as Sen uses the term. A large part of the interest in both functions and functionings stems from the belief that the value space cannot be identified with either material or social goods (e.g. Rawls's primary goods) or subjective responses to them (e.g. the classical utilitarians' utility), but is to be found somewhere between them. That seems to me right, but it is nothing new. Though classical utilitarians often wrote as if they saw utility in terms of mental states, at other times they suggested something quite different (recall Mill on qualitative differences in utility: the higher one is that to which ‘all or almost all who have experience of both give a decided preference’;5 and recall the example of the informed preference for Socrates dissatisfied — short on pleasures, in the ordinary sense of the term, and short on desire fulfilment — over the Fool satisfied; the values referred to here are far from subjective responses and close to, perhaps identical to, beings and doings). And certain preference utilitarians long ago moved utility into the realm of beings and doings.6 With their notion of a ‘rational’ desire, they took the desire fulfilment account of utility well down the road towards, if (p.137) not quite all the way to, the list of prudential goods that seems to me the inevitable destination of a fully worked out desire account.7 Then there was G.E. Moore's ideal version of utilitarianism as early as 1903. Prudential values can indeed be seen as beings and doings. Even ‘happiness’, in the sense most relevant to the quality of life, is probably best understood, for reasons that Aristotle gives, as falling into the category of activity, and so is caught by ‘doings’. The talk of beings and doings leaves us well short of a delineation of value space, because those terms include vastly more than just values. We need a differentia. Without it, the job of locating the value space is virtually undone.
For Sen, nourishment and health are paradigmatic functionings.8 But neither is itself a prudential value. It would make perfectly good sense to choose to be, at least for a while, undernourished through fasting, if one's spiritual life were thereby enhanced. And I do not think that it is just the presence of choice here that gives undernourishment through fasting a different moral status from undernourishment from starvation; to think so would be to overlook what makes sense of the choice. What makes sense of it is that fasting (and the discomforts of hunger that go with it) is to be seen as part of a process of reaching a larger good — say, spiritual growth. That is why the case of a child in a family that fasts who does not himself choose to fast is morally different from starvation too. Discomfort is small stuff compared to a spiritual life. We understand these cases best, I think, in terms of the various substantive values in play. It might also make good sense to accept some avoidable forms of ill health. It would make good sense not to spend long periods exercising one's legs to combat a wasting disease, if ease of walking had little relevance to the quality of one's particular life, while using the exercise time instead for one's work did have. Nourishment and health are both at some remove from prudential values. Unless we acknowledge that, we cannot explain the rationality of much prudential deliberation.
Why does this remoteness from real prudential values matter? Why does it matter whether we deliberate in terms of functions, desire fulfilment, and subjective reactions or in terms of what seem to me to be prudential values: enjoyment, accomplishment, deep personal relations, the elements of human dignity (autonomy, liberty), and so on? It seems to me to matter because we have to get a sense of the whole deliberative project that we are engaged in. We have to see how to compare one value with another, and one person's values with another's, and how far commensurability and comparability go. And we shall not find out — we shall not get even our crudest bearings — until we see what is prudentially valuable and what makes it valuable. Brock speaks of the vector view. But choosing vectors that are at some remove from actual prudential values does not allow us to get at how the vectoring actually works. For example, how important is this primary function (say, legs that work) to (p.138) this person (say, Itzhak Perlman)? What is the significance of this person's contentment with a disability? Contentment with one's lot, as Brock notes, can variously be both a desirable constituent of the good life and an undesirable obstacle to a better. We cannot proceed, I think, until we have got the right materials to work with, and that means going deeper to the prudential values at stake.
This is not an objection to the vector view. On the contrary, as a view about the broad notion of the quality of life (as opposed to a moral view, which is an entirely different matter), some form of it must be right. There may be several irreducibly different prudential values. They may have different weights in different cases; they may vector into a single direction. But all of this just mentions possibilities. We need to see how it actually works. Well, suppose we use the vectors ‘desire fulfilment’ and ‘subjective reaction’. Let me take an example of vectoring from Sen's paper ‘Plural Utility’.9 Suppose someone prefers bitter truth to comforting delusion but is palmed off with the latter. We feel sorry for him. Then later he learns that he was deceived. We feel sorrier for him. Sen says that bare desire fulfilment and experienced desire fulfilment (that is, subjective reaction) are both relevant, and that to insist on a choice between them seems ‘arbitrary and uncalled for’.10 But the trouble with leaving it at that is that at least sometimes we have to supply some weighting for various vectors, and the weighting ought to be neither arbitrary nor left to haphazard intuition. I doubt that the categories ‘bare desire fulfilment’ and ‘experienced desire fulfilment’ will help us get at the weightings that we are after. We are after a comparison of values, and neither ‘bare desire fulfilment’ not ‘experienced desire fulfilment’ even names a value.
Suppose, then, that we use instead actual prudential values as our vectors. One range of prudential values, I think (and shall have just to go on dogmatizing), comprises the elements of human dignity — that is, such things as our being able autonomously to choose a course through life and being at liberty to follow it. We also value pleasure and the avoidance of its opposites. With these in focus, with some understanding of their bounds and grounds, we can hope to make progress towards weightings. We value our liberty to carry out our plans, greatly in the case of our most central plans and less when the plans are more peripheral. Following these lines, we might hope eventually to explain how much weight to attach to the desire to live out the most central features of one's life plan, and how this weight compares with the upset and distress it might cause others. We can hope also to get an explanation of why liberty has different weights on different occasions — though the desire to carry out the central parts of our life plan generally outweigh our neighbours' huffiness over our doing so, the desire to swim nude, being not all that central, might not outweight the upset it might cause.
In arguing for the need to identify actual prudential values, I am not claiming (p.139) that their identification is exclusively, or even supremely, important. We need the broad conception of the quality of life (that is, the list of prudential values) for our reasoning about how to make our own lives go best. Doctors need it for decisions about certain patients. But there are many reasons, both moral and practical, to work with a narrower conception of the quality of life in taking certain social decisions. And it may be (I strongly suspect it is) that we need several different conceptions of the quality of life for different sorts of social decision. That is, we need both the broad conception and also an understanding of the various considerations at work in generating narrower conceptions.
Brandt, R. B. (1979). A Theory of the Good and the Right. Oxford: Clarendon Press.
Griffin, James (1986). Well‐Being. Oxford: Clarendon Press.
——(forthcoming a). ‘Against the Taste Model’, in J. Elster and J. Roemer (eds.), Interpersonal Comparisons of Well‐Being. Cambridge: Cambridge University Press.
——(forthcoming b). ‘Value: Reduction, Supervenience, and Explanation by Ascent’, in K. Lennon and D. Charles (eds.), Reduction, Explanation, and Realism. Oxford: Clarendon Press.
Mill, J. S. (1863). Utilitarianism. London.
Sen, Amartya (1980–1). ‘Plural Utility’, Proceedings of the Aristotelian Society, 81.
——(1985). Commodities and Capabilities. Amsterdam: North‐Holland.
——(1987a). On Ethics and Economics. Oxford: Blackwell.
(6) This is obscured in Sen's writings by his generally construing ‘utility’ in quite a narrow way; he sees ‘utility’ as committed to a ‘metric of happiness or desire‐fulfilment’, which (given his narrow interpretation of these two notions) he rightly says has obvious limitations as an explanation of human well‐being (Sen, 1987a: 45). It is not that Sen is unaware of the great variety of possible, or actual past, uses of ‘utility’ — that ‘versatile name’, as he remarks (Sen, 1985: 17); far from it. But he thinks that happiness and desire fulfilment are ‘the traditional meanings of utility’ (Sen, 1985: 3), and he generally uses the term in that traditional sense. Still, his objections to utility as an interpretation of well‐being show just how narrowly he uses ‘happiness’ and ‘desire fulfilment’. People often become reconciled to deprivation, he rightly observes; they stop minding or hoping. So using either happiness or desire fulfilment as a metric may distort how well off a person actually is. That seems to me right and important, but the point I want to make here is that this works as an objection only if one uses especially narrow interpretations. Sen even seems to confine desires to actual desires, ignoring the tradition of corrected or rational or informed desires, and certainly ignoring any especially strong standard for a desire's being informed. See Sen, 1985: 20–2, 52–3; Sen, 1987a: 45–6; Sen 1987b: 7–12.
(10) Ibid., 203.