Cross-post from the Uehiro Centre's Practical Ethics blog

Last week saw the launch of a campaign (run by the group Vegetarian For Life) that seeks to ensure that older people in care who have ethical commitments to a particular diet are not given food that violates those commitments. This is, as the campaign makes clear, a particularly pressing issue for those who have some form of dementia who may not be capable of expressing their commitment.

Those behind campaign is quite right to note that people’s ethical beliefs should not be ignored simply because they are in care, or have a cognitive impairment (see a Twitter thread where I discuss this with a backer of the campaign). But the idea that one’s dietary ethics must be ‘for life’ got me thinking about a more well-established debate about Advance Directives. (I should stress that what I say here should not be taken to be imputing any particular motivation or philosophical commitments to those behind the campaign itself.)

Briefly, one way of thinking about advance directives is that they are a way for people to, while we retain the relevant cognitive capacities, direct what happens to them at a future point where they lack capacity. Ronald Dworkin (in Life’s Dominion) draws a distinction between two kinds of interests. Some interests are purely experiential – it is in my interest to enjoy this piece of cake, for instance, merely because of how it tastes in the moment. But other interests are critical: they lend coherence and shape to our lives (he acknowledges that people may have such interests to greater or lesser degree).

When it comes to certain kinds of advance directives, these interests may come into conflict. For instance, someone may write an AD which says that if they are suffering from severe dementia, and have lost the capacity to do various things that currently give their life meaning (e.g. writing; certain kinds of interactions with loved ones), they should not be resuscitated if they fall ill. Their critical interest in shaping their life (exercised initially while still possessing capacity) may come into conflict with their later experiential interests if their quality of life is still good at the time when doctors are deciding whether to resuscitate. While dementia can of course negatively impact quality of life, many individuals with dementia lead good, enjoyable lives, full of valuable relationships, pastimes and experiences.

In Dworkin’s view, it is the critical interests that should win out. Up until now I have been convinced by a response to Dworkin from Rebecca Dresser. Dresser argues that it is far from clear that critical interests should always take precedence. Dresser raises a number of important issues, but a particularly compelling example of hers, to my mind, is that of pain relief: we can imagine a person whose religious or philosophical commitments led them to value the experience of physical pain, perhaps seeing it as more authentic than using pain relief. But imagine such an individual later on, as a patient suffering dementia and having forgotten all about these commitments. It is hard to see the value in nonetheless respecting her past critical interests, and refusing to give a confused, frightened patient pain relief when doing so would benefit her immensely.

In such a case, we allow a person’s current experiential interests to override our respect for her past critical interests. And it seems to me that we do so quite defensibly; indeed, I think it would be indefensible to insist on respecting her past commitments which have no resonance with her now. To my mind, the case of life-prolonging care is of the same kind; if a patient with dementia no longer has any sense of the commitments that led her to request that her life not be extended if they were to suffer from such a condition, and her continuing life is of a good quality, then our moral obligation is to save her life.

The Vegetarian for Life campaign has, I admit, given me pause for thought. As someone with ethical dietary commitments, I am certainly unsettled by the thought that I might one day unwittingly violate those commitments, even if I no longer explicitly hold them. So I will end by offering a couple of thoughts about what might justify a difference in approach between diet on the one hand, and pain relief and the extension of good quality life on the other.

1. No harm done: If I am denied pain relief, I am directly harmed. Although philosophically more controversial, many of us believe that we can be harmed by being deprived of good quality life. In many cases, though, there is no harm done to someone by feeding them, say, a vegan diet. As such, the experiential costs of respecting the critical interest is lower. 

Such a defence will, of course, be limited. If a patient’s tastes change so that they will only eat previously abhorred foods, then there would be a serious cost to maintaining their past commitment.
 
2. Evidence: We typically get fairly direct evidence that someone is in pain, and that the patient in pain wants it to stop. Judgements of whether an individual life is overall good are more difficult, but if a patient seems happy on a day-to-day basis, we have little reason not to judge that they would benefit from continuing to live. It is perhaps more controversial to say that we can attribute to a patient with dementia a desire to go on living; but we can typically attribute a desire to carry on various pleasurable activities, for which living is of course a requirement. In both these cases then, we have evidence of the patient’s individual experiential preferences. 

Contrast this with dietary preferences. In general, the evidence we could get that a patient’s preferences had changed would not come about organically (as with the preference against pain, and for the continuation of life), but through a conscious decision by carers to change the patient’s diet. But there is usually no good reason to make that change.
 
3. Moral wrong vs. ideal lifeEven amongst people with particular dietary values, reasons vary. My own reasons are other-regarding moral ones of a fairly standard kind (animals have rights, and it’s thus wrong to kill them for food). What’s more, I think that it would still be wrong of me to eat meat even if I changed my mind.

This contrasts to some extent with at least one way of viewing Dworkin’s example of critical interests. Imagine someone who values abstract philosophy, who knows he will have no interest in that sort of thing once he has dementia. He currently sees his life as valuable in a very specific way. But he may not (and, I would suggest, should not) see other kinds of life as without value, at least for other kinds of people. What’s more, even if he does see a life without philosophy as containing some kind of mistake, it would be odd to see it as a moral mistake. Is there anything to be said for the idea that a person’s deep moral commitments have a greater claim to being respected (so long as they are reasonable) than other kinds of evaluative commitments? I expect many people will want to say no!

​Of course, my varying intuitions may just be due to personal value prejudice. While I have a strong commitment to my dietary values, I have no similar commitments that would make me want to prematurely end my life if I were to suffer from dementia (it is not exactly that I welcome dementia as Charles Foster does, but that I recognise that my life may well have significant personal value, albeit of a different kind than it does now). And I certainly don’t have any commitments against avoiding pain (authenticity schmauthenticity, I say). So it may be that I’m just being parochial, and insisting that while the things I care about in a critical sense are deeply important and should be respected, the things others care about aren’t, and shouldn’t.

This is a cross-post from the Uehiro Centre's Practical Ethics blog: 

We recently saw a legal challenge to the current UK law that compels fertility clinics to destroy frozen eggs after a decade. According to campaigners, the ten-year limit may have had a rationale when it was instituted, but advances in freezing technology have rendered the limit “arbitrary”. Appeals to arbitrariness often form the basis of moral and political criticisms of policy. Still, we need to be careful in relying on appeals to arbitrariness; it is not clear that arbitrariness is always a moral ‘deal-breaker’.

On the face of it, it seems clear why arbitrary policies are ethically unacceptable. To be arbitrary is to lack basis in good reasons. An appeal against arbitrariness is an appeal to consistency, to the principle that like cases should be treated alike. Arbitrariness may therefore seem to cut against the very root of fairness.

However, there are at least two ways a policy’s arbitrariness is not a knock-down argument for scrapping or changing it. The first of these is when a policy is in fact based on arbitrary grounds, but where there are good alternative grounds for it. Consider, for instance, the furore around a decade ago over the firing of government health advisor David Nutt. Nutt was sacked for contradicting government guidelines over drug safety. Nutt publicly appealed to the idea of arbitrariness in his criticism of government policy. In particular, he said, it was arbitrary to support the continued legality of alcohol and tobacco while opposing the legalisation of less harmful drugs such as cannabis, LSD and ecstasy.

It was certainly worrying for the government to sack an advisor simply for disagreeing with them. And Nutt was right to say that government policy should be based on evidence, and probably wasn’t. So, government policy was arbitrary. But he was wrong to imply that because alcohol is more harmful than LSD, it is essentially arbitrary and thus wrong to keep the former legal while criminalising the other.

The reason is that this arbitrariness is ingrained in society. What we want to know when facing the question of whether to legalise or criminalise a particular drug is not only how it compares to other legal and illegal drugs, but how much good or harm will come from changing its legal status. It is possible, then, that although alcohol is worse than ecstasy, we should keep alcohol legal and ecstasy illegal because:

• criminalising alcohol would be more dangerous than keeping it legal, and
• legalising ecstasy would be more dangerous than keeping it illegal.*

That the existing legal framework is arbitrary is therefore relevant but not decisive, because that arbitrariness may have given rise to settled expectations that are also relevant (e.g. the expectation of being able to drink alcohol freely). While this argument appeals to consequences, I should note that it’s not supposed to be a purely consequentialist argument. For instance, the fact that an ultimately arbitrary distinction is ingrained in society is not a good argument for keeping it when it involves denial of basic rights.

Here’s the other way a policy can be arbitrary and yet justified. A particular limit (e.g. a speed limit of 20mph outside schools; or an age limit of 18 to vote) can be arbitrary in the sense that any particular limit would be arbitrary, and yet a limit is still required. In this case, the question of whether something is arbitrary is comparison-sensitive. If we ask why the speed limit should be 20mph rather than 100, we can give good reason. But if we ask why it should be 20 rather than 21 or 19, the answer looks less clear. Similarly, we can explain why we should allow people to vote at 18 rather than 5, but may struggle when presented with articulate, intelligent and politically motivated 16-year-olds to say why the limit should not be dropped. Nonetheless, it seems plausible that in both cases we do need some limit. And any limit that attempts to cover a general population will either face exceptions (dropping the voting age to 16 will face some intelligent, articulate 15-year-olds), or be open to symmetrical challenges (just as there is no reason to prefer 20mph to 19mph, so too is there no reason to prefer 19 to 20).

These cases come in two kinds. One is where an easy-to-measure feature tracks, to some degree, a feature that is far more difficult to get a handle on. Age tracks emotional maturity, politically-relevant knowledge and independence, but it does so very imperfectly (some teenagers outdo some adults on all of these). The other is where we have good reason for preferring something in a particular range, but not for preferring any individual within that range. It really is best to have a speed limit of around 20 near schools, but it really does seem impossible to justify 20 precisely.

Faced with such cases, we have three choices. We can scrap the limit altogether as unavoidably arbitrary. In the first kind of case, we can attempt to impose limits based on non-arbitrary features (e.g. replacing a voting age with epistocracy, where voting rights are based on knowledge and intelligence). Or we can accept that while an arbitrary limit is imperfect, it’s the best we’ll get. Like the drugs debate, this latter view requires attention to the pragmatic consequences of adopting any particular policy. For instance, if the idea of epistocracy looks less attractive than moderate arbitrariness (and I think it looks much, much less attractive), we will need to accept that our voting limit is always going to exclude some people unfairly. Nonetheless, we might think that a particular age limit (not necessarily 18) will do the best available job of excluding as few people as possible, without opening us up to a mass of uninformed, clueless voters (as might be the case if we simply dropped an age limit altogether.

These cases point to a further feature that’s required for arbitrariness to be justified (the first being that basic rights are not violated or denied). It is that the area in which an arbitrary limit is in place is something that requires a limit, and that no better criteria for a limit are available.

Strikingly, these three features map onto the case being made against the ten-year storage limit. Campaigners have argued that the limit is not only arbitrary, but arbitrary in a way that violates the right to a family life. And while the individual bringing the challenge accepts the need for a limit of some kind, she argues that this should be in line with the age at which most fertility clinics would not accept patients in any case: 55. Such a limit might itself be arbitrary, if some of those over 55 could have children. But importantly, it might be less arbitrary than the status quo.

* Of course, these claims may be false. It may be safer to legalise ecstasy. But the truth of this claim has nothing to do with a comparison with alcohol.