1. Introduction
If there is a case for our moral enhancement – if such enhancement is thought desirable – it is presumably based on the idea that, while we humans are not as bad as we might be, neither are we as good as we could be. And if we are not as good as we might be this is, again presumably, because we do things that are (i) various shades of bad, and (ii) at least to some extent avoidable. Human history is littered with bad but seemingly avoidable episodes. And so too, barring the occasional moral saint or savant, are the lives of each one of us. The purpose of moral enhancement would be to get us to stop doing these bad, avoidable things. But if this is the purpose of moral enhancement then the concept of moral enhancement is likely to be as broad and mottled as the reasons why people do bad but avoidable things. The messy, mottled character of the concept of moral enhancement, I shall argue, is not merely of academic interest. On the contrary, it has implications for the practical feasibility of such enhancement.
2. Dumb and Dumber: Enhancing Doxastic Health
A fact so obvious that it is frequently overlooked is that many – perhaps most – of the bad, avoidable things we humans do are the result of us holding stupid, stupid beliefs.Footnote 1 The belief that one belongs to a master race; that one is the victim of a worldwide Jewish conspiracy; that one can attain heaven, not to mention the ministrations of seventy-two virgins, simply by murdering a few non-believers; that there is no such thing as climate change, or if there is it does not have an anthropogenic source; and so on and so forth. These days, it might not be politically correct to characterise the sincerely held worldviews of people as stupid. But the belief that we can't or shouldn't do this is, as far as I can see, just another stupid belief from which misery ensues. And, anyway, these examples are just off the top of my head. We all have our favourite examples of gross human stupidity and the misery that results. Without taking a stand on just how stupid each belief is and just how much misery results from it, one thing is clear: there are more than enough stupid beliefs to go around, and more than enough misery that results from them. This misery is entirely avoidable … if only we could find a way of foregoing the stupid beliefs.
This type of moral failing, then, yields a corresponding form of moral enhancement: enhance people so they are less likely to hold stupid beliefs. How does one do this? Is there a pill that will do this? Can we achieve it by snipping a few genes here and there? Perhaps we can, or one day will. But there is no obvious reason why such a strategy would be preferable to, for example, a better education system – one that, say, more effectively enhances the ability to think critically about issues, and emphasises the importance of such trifles as logic and evidence and even that old chestnut: truth.
I wouldn't want to unfairly rule out a potential role for cognitive enhancing drugs. Stupidity is, after all, a cognitive deficit. So, why not try to mitigate it through the use of cognitive enhancers – nootropics? We can overlook the fact that the current generation of such enhancers still looks rather primitive – with effects that are highly variable between individuals, often highly dependent on baseline cognitive functioning, and which seem to unevenly target long term memory rather than intelligence per se. I have in mind something a little better than this. Not a stimulant – that won't be much use. And not something that primarily targets memory – that probably won't be much use either. But perhaps some sort of positive allosteric modulator of AMPA receptors – some suitably developed version of the racetam family, for example – might do the trick. Suppose, in short, we did have a proven, reliable cognitive enhancer at our disposal. If we did, it would likely make us intelligent enough to realise that this is not what we need. There is something about the appeal to cognitive enhancement as a route to moral enhancement that misses the mark – by a rather spectacularly wide margin. We might put matters thus: it's not rocket science! It's not rocket science that views that have no evidence supporting them, or are illogical, should not be trusted – not enough to base a life and a consistent policy upon them. This is not an idea lying just beyond the limits of our cognitive powers – limits that we might be able to transcend with the right chemical stimulation. We don't need a racetam to work that out. It is a fact that belongs to the realm of the bleeding obvious.
This suggests that the stupidity that is at the root of many moral problems is not, in fact, a cognitive phenomenon – certainly not a purely cognitive phenomenon. The sort of stupid beliefs that can engender moral problems arise from an exercise of the will as much as of a lack of exercise of the intellect. When a belief is really, really, stupid, and one nevertheless has it, this can only be because one really, really wants to have it. This might be because one has a significant stake in the belief being true. Perhaps the status of the group with which one identifies depends on the belief being true. Perhaps one's status within that group depends on it. There can be a variety of reasons for one having a stake in the truth of such a belief. But none of these involves a deficit of the intellect. We are talking about affect, not intellect. This brings us to the next possible form of moral enhancement.
3. Enhance Critical Scrutiny
There is a tradition of thinking of moral action as dependent on our scrutiny, or at least the ability to scrutinise, our motivations. The tradition in question is, of course, the Kantian tradition. Christine Korsgaard puts the idea very nicely:
As rational beings we are conscious of the principles on which we are inclined to act. Because of this, we have the ability to ask ourselves whether we should act in the way we are instinctively inclined to. We can say to ourselves: “I am inclined to do act-A for the sake of end-E. But should I?”Footnote 2
It is this ability, according to Kant, that separates us from the rest of creation: this is what elevates us to the kingdom of ends rather than that of mere means. We do not simply have motivations; in addition we can scrutinise them. That is, we can interrogate our motives and try to work out if they are motives we should endorse or ones we should reject. I have motive M, I can think to myself. Is this a motive I should embrace or one I should resist? We can ask ourselves these sorts of questions. Nothing else can. Kant's view of how we should go about answering such questions is well known. In a nutshell, we must identify if our motivation is one such that we can consistently will that it be adopted by all members of the kingdom of ends – i.e., rational beings. If we can consistently will this, then the motivation is one that we should embrace.
We can, however, take this general Kantian picture of moral motivation and divorce it from Kant's specific conception of in what the critical scrutiny of motivations consists. Instead of asking whether our motivation is universalisable, we might instead ask whether it maximises utility; or whether it is a motivation that would be agreed to by rational contractors under appropriate conditions of ignorance; or whether it is a motivation that would be embraced by a virtuous person, and so on. These are all possible forms that critical scrutiny of one's motivations might take – and there are, of course, others. My interest here is with the general idea of critical scrutiny rather than the specific forms such scrutiny might take.
If being a moral actor requires the ability to critically scrutinise one's motivations, then the corresponding form of enhancement seems clear: enhance the ability of people to critically scrutinise their motivations. Simple! Yet the sheen of simplicity quickly dissipates – fragmenting into several quite different ideas. A failure to critically scrutinise one's motivations can be the result of more than one type of cause. It might simply be that one can't be bothered. One is perfectly capable of scrutinising one's motivations but, for one reason or another, one would rather not. One is unwilling to scrutinise one's motivations. Perhaps it seems like a lot of trouble? Perhaps one has other things one would rather be doing? Whatever the reason, this is an affective deficit. One has the ability to critically scrutinise one's motivations. But one does not care enough to exercise this ability. If this is the problem, then the corresponding form of moral enhancement will have to target an affective deficit of this sort. It will have to instill a concern for, and desire to engage in, the scrutinising of one's motivations.
On the other hand, it may be that a person's failure to scrutinise her motivations does not stem from a lack of desire but from lack of ability. She does not critically scrutinise her motivations because she is unable, rather than unwilling, to do so. This would be a cognitive rather than affective deficit. The corresponding form of moral enhancement would, therefore, involve augmenting one's cognitive – specifically one's meta-cognitive – powers. This, however, is only the beginning of the story. There are at least two distinct cognitive deficiencies that might be at the root of her problem.
First, a person might be hopeless at identifying her motivations. We know this is a problem for all of us to some degree. We commonly, for example, attribute to ourselves judgements that we never really made. In a classic study, Nisbett and Wilson convinced various test subjects that they were taking part in a market survey. The unwitting subjects were presented with four pairs of identical pantyhose and asked which they preferred.Footnote 3 The majority of subjects strongly preferred the pantyhose placed on the right. When asked to explain this, they confidently – but wrongly – declared that their preferred pantyhose were superior because they were softer, had the nicest colour, etc. In fact, they chose the pantyhose on the right because of a quirk of our brain: we have a strong right-preference. It was the location of the pantyhose that was the basis of their judgement.
A common refrain emerging from the “new unconscious” research is that we have far from infallible access to the workings of our minds and the causes of our behaviour.Footnote 4 Confabulation is the norm, not the exception. Even if this overstates matters, a healthy scepticism regarding one's ability to identify one's real motivation in any given circumstance is certainly justified. One form moral enhancement might take, therefore, is augmenting our ability to identify our motivations – to make confabulation a rare exception rather than depressingly common.
On the other hand, even if a person is exceptionally good at identifying his motivations, he might suffer from another deficiency: he is absolutely hopeless at bringing moral principles – universalisability, utility maximisation, impartiality, etc. – to bear on them. He has a motivation, and has correctly identified this motivation, but is hopeless at working out whether it will maximise utility, or whether it is universalisable, or would be agreed to by rational contractors under appropriate conditions of ignorance. Like the person who is useless at identifying her motivations, this person suffers from a cognitive deficiency. But it is a different cognitive deficiency. Each deficiency would require a corresponding form of enhancement. The enhancement would aim to instill or improve a certain meta-cognitive skill, but this would be a different skill in each case.
Even if we could work out which meta-cognitive skill we must try to enhance in any given case of scrutiny-deficiency, there is a further problem. Someone may be superb at identifying their motivations, and faultless in the bringing to bear of ethical or meta-ethical principles on them. But this would hardly qualify as moral enhancement if the principles thus brought to bear were hopelessly warped – for example, murderous or xenophobic. Enhancing a person's ability to critically scrutinise his motivations only counts as moral enhancement if the ethical/meta-ethical principles brought to bear in such scrutiny are morally good or right ones. Now we have some real problems, of course, since what counts as morally good or right will vary between theories. Whether universalisability, and the good will that seeks and acts on this, is the paramount moral consideration depends on whether you are a Kantian or a utilitarian. So, it seems, to properly pursue the idea of moral enhancement as enhancement of the ability to critically scrutinise one's motivations, we would have to enhance the ability to identify correct moral theories. This, as millennia of internecine strife amongst moral philosophers seems to demonstrate, is a bit of a tall order (to put it mildly).
4. Enhance Virtue
Never mind. Never mind. We've crossed off critical scrutiny, but never mind. There are plenty of other traditions in moral psychology we might work with. There is the Aristotelian, virtue-ethical tradition, for example. Perhaps moral enhancement should be understood as enhancement of one's moral virtues. Of course, there are disagreements as to what counts as a moral virtue. As Michael Slote has pointed out, kindness and compassion did not feature highly on Aristotle's list of virtues, but contemporary theorists such as Philippa Foot, John McDowell, and Rosalind Hursthouse assume they are virtues.Footnote 5 But that doesn't really matter. We can decide which things are moral virtues, and then go about enhancing them – or so the idea goes.
The first problem with this is that virtues do not exist in a vacuum: they exist only in a surrounding cognitive and emotional milieu. This is so for several reasons. First, consider this famous passage from the Nicomachean Ethics. Aristotle writes:
But for actions in accord with the virtues to be done transparently or justly it does not suffice that they themselves have the right qualities. Rather, the agent must also be in the right state when he does them. First he must know that he is doing virtuous actions; second, he must decide on them, and decide on them for themselves; and, third, he must also do them from a firm and unchanging state.Footnote 6
The expression, ‘decide on them and decide on them for themselves’ is commonly taken to mean that the agent must perform an action because it is the virtuous thing to do. Thus, Aristotle's first two conditions place what we might call a reflection condition on virtuous action. Roughly:
For any action φ, performed by agent A, for φ to be an instance of virtuous behaviour, it is necessary that A (1) understands that φ exemplifies virtue V, and (2) performs φ because he wishes to be virtuous.
Therefore, contained in this account of what is required to engage in virtuous behaviour are two possible forms of moral enhancement. First, one might enhance: the ability to understand which actions exemplify or instantiate virtues. Second, one might enhance the desire to be virtuous. Once again, we find the enhancement proposal straddling both cognitive and affective elements. The notion of enhancement breaks down into two kinds – one cognitive and one affective – just as it did with the scrutiny-based model of enhancement.
There is, however, more to the complexity of the virtue-based model of enhancement than this. The third condition Aristotle places on an action counting as virtuous invokes the notion of a ‘firm and unchanging state’. This firm and unchanging state is far more than a mere behavioural disposition. The following definition of a virtue, inspired by Hursthouse, is typical:
A virtue is (i) a good, admirable, or otherwise praiseworthy character trait, where (ii) a character trait consists in a relatively stable set of behavioural dispositions that are embedded in an appropriate surrounding milieu of judgments and emotions, broadly understood.Footnote 7
To possess a virtue is not just to act but also to judge and feel. To have the virtue of honesty, for example, is not just to behave in an honest way. A person might behave honestly because he was terrified of being caught and punished. To have the virtue of honesty – to be an honest person – one must also judge dishonesty to be wrong and to deplore cases of dishonesty whenever one encounters them, whether this dishonesty is exhibited by others or yourself, and so on.
Therefore, a virtue-based programme of moral enhancement would require us not only to modify behaviour, but also enhance dispositions to feel and judge. As in the case of the Kantian approach to moral enhancement, any moral augmentation pursued here would have to be multifactorial: targeting a variety of quite different psychological systems, some of which are cognitive, and some of which are affective. It is not immediately clear what sort of genetic manipulation or chemical modifier would be able to hit these sorts of diverse targets.
Judgement enters the virtue-ethical picture in another, even more important way. Virtue-ethical accounts typically accord a central role to what is known, following Aristotle, as phronesis or practical wisdom. In another famous passage from the Nicomachean Ethics, Aristotle writes:
Getting angry, or giving and spending money, is easy and everyone can do it; but doing it to the right person, in the right amount, at the right time, for the right end, and in the right way is no longer easy, nor can everyone do it. Hence doing these things well is rare, praiseworthy, and fine.Footnote 8
To have a virtue requires the ability to judge what a particular situation calls for. To lie in order to spare a person's feelings may be required in some situations. But in other situations the truth, though painful, may be something that a person really needs to hear. According to Aristotle, this sort of wisdom – the ability to judge what a particular situation does and does not call for – is acquired only through years of experience and consistent practice. It is not immediately clear how genetic tweaking or chemical infiltration is going to be able to supply this or otherwise make up for it.
5. Enhancing Sentiment
It might be thought that, if moral enhancement is our game, the most promising ethical theory would be some or other form of sentimentalism. The guiding idea is a simple one: if only we could get people to be a little nicer. If only we could get them to commiserate more with the sufferings of others. If only we could get them to rejoice more in others’ happiness. If only, that is, we could get them to be more sympathetic.
The centrality of sympathy to morality is an idea strongly associated with David Hume, Adam Smith, and others.Footnote 9 Sympathy is, on this view, the primary moral sentiment, and is essentially connected to moral good and bad. There are, however, two distinct ways in which the centrality of sympathy to moral good and bad might be developed. According to the first, emotional reactions, such as sympathy, allow us to detect the good- and bad-making features of situations. This is an objective form of sentimentalism. Moral good and bad exist independently of our emotional reactions. It is just that these reactions give us an especially reliable – perhaps uniquely reliable – insight into which things are good and which are bad.Footnote 10 Our emotional reactions allow us to identify the features of situations that make them either good or bad. They provide us with sensitivity to such features.
It is more usual, however, for moral sentimentalism to be understood in another way. According to this more usual understanding, moral good and bad are constituted by our emotional reactions. This is the subjective version of sentimentalism – the form of sentimentalism that Hume and Smith had in mind. It is not clear where, precisely, this subjective form of sentimentalism leaves the idea of moral enhancement. The desirability of enhancement is based on the idea that human behaviour is not as good as it might be. But if good and bad are simply constituted by our emotional reactions then this idea of our behaviour not being as good as it might be is quite difficult to understand. By what standard of good does our behaviour fall short? There seems to be no independent standard with which to measure the alterations to emotional response that are enhancing and those that are not. This is, probably, not fatal to the idea that we can use the subjective version of sentimentalism to explain the idea of moral enhancement. Subjectivist ethical theories have long been faced with this problem and developed a variety of ways of dealing with it. But it is worth pointing out that we can't just assume that subjective versions of sentimentalism are compatible with moral enhancement. A not inconsiderable amount of work will have to be done to make them compatible. I shall ignore this complication, and assume that either objective or subjective versions of sentimentalism are available for understanding the idea of moral enhancement.
The idea that we can understand moral enhancement in sentimentalist terms also fragments into several distinct possible views. When discussing the moral sentiments, it is orthodox to draw a distinction between empathy and sympathy. Very roughly – it is sometimes said that there are as many definitions of empathy as there are people working on empathy – empathy is the ability to understand what another is feeling or thinking. Sympathy, on the other hand, is the ability, indeed willingness, to care about what you find when you do this. Empathy is the ability to understand something. Sympathy is the ability to have an affective response to this understanding. So, the first question arises for the moral enhancement project understood in sentimentalist terms: do we enhance empathy (a form of understanding) or sympathy (an affective response)?
Suppose we decide to try to augment empathy. The problem is that the notion of empathy breaks down into two different sorts.Footnote 11 On the one hand there is projective empathy. This is the ability to put oneself in the shoes of the other and imagine what they must be feeling in their current circumstances. Projective empathy is an imaginative ability. On the other hand, there is receptive empathy. This is the ability to respond to the mental condition of another by producing, in oneself, a condition of the same sort, or of a similar sort. There is nothing imaginative about this. It is a form of sensitivity. Imagination will involve significant contribution from the cortex. But the amygdala will be centrally involved in receptive empathy. Genetic manipulation or chemical intervention will, somehow, have to target two distinct brain regions – sometimes simultaneously, and always in the right way. It is not clear how we are going to get it to do this.
Even if we can sort out the issue of what, precisely, we are trying to enhance, there is a further problem. We can enhance neither empathy nor sympathy in a vacuum. Enhancement of empathy and/or sympathy will be useful only if it is located in a complex cognitive/intellectual milieu. We can't avoid the necessity of phronesis here any more than we could when our focus was on virtue. Following Nussbaum, we might distinguish between thick and thin versions of sympathy.Footnote 12 A sympathetic response is thick if it is embedded in a surrounding milieu of judgements of the sort constituted by phronesis. A sympathetic response is thin, on the other hand, if it is not thus embedded.
In the absence of phronesis our sympathetic responses are likely to go awry in one of three different ways. First they are susceptible to problems of size. Without phronesis, we are likely to misrepresent the size or significance of another's suffering. Let us suppose that a friend, a multimillionaire best-selling author, is devastated by the critical reaction to his latest offering, and hasn't left his huge mansion in the Hamptons for days, so upset is he. If we had a little practical wisdom we would realise that he should just “cop on” and that, all things considered, he is one of the most fortunate of people. Without this wisdom, and the perspective it provides, we are likely to be swayed by his outpouring of grief and, consequently, to be sucked in to his orgy of despair.
Another friend insists on making the same mistake over and over again. You've told her, you really have. But still she persists. Practical wisdom tells you that any sympathy you feel for this person should at least be tempered by their persistent recidivism and refusal to take your advice. Without it, you may fall victim to their overt displays of grief.
Sometimes the truth can be painful, but at the same time something a person really needs to hear. Sparing them the truth may temporarily save them unhappiness, but will only increase their unhappiness in the long term. The long term/short term distinction is the most obvious version of the distinction between local and global interests. A person needs to give up smoking – for the sake of their health, a global interest. Doing so will cause them anguish now – a local interest. Local interests typically receive more vivid experiential expressions than global interests. Without practical wisdom we are likely to unacceptably emphasise local over global interests.
The sympathy we feel is one thing. The sympathy we should feel is quite another. Practical wisdom is what allows us to identify, and appropriately emphasise, the latter.
6. Moral Education and Moral Enhancement
Nothing I have said entails moral enhancement cannot work. Indeed, moral enhancement clearly does work. Traditional moral education is a form of moral enhancement, and moral education works – not all of the time, of course, but at least sometimes. There are, however, significant differences between traditional moral education and technological approaches to moral enhancement. If the arguments developed in this essay for the messy, variegated character of moral enhancement are correct, these differences are likely to be crucial.
Certain environmental contingencies, with which I shall not bore you, have caused me to think it instructive to think of moral education by analogy with football (or, as the locals on this side of the pond insist on calling it, soccer) education. That is, it is instructive to think of the development of moral skills as analogous, in some ways, with the development of football skills. If you are coaching an Under-11 football team – don't ever do it, you're essentially wrangling cats for hours every week – the most important thing to realise is that, with regard to skills, different players need different things. Some will have developed bad habits – trying to control the ball with their preferred foot, when the situation requires they use the other foot, for example – and these will have to be expunged. So you devise drills to target that. Others make poor decisions – for example, they run when they should pass, or pass when they should run – and this sort of decision-making is also something you can target with the appropriate drills. Others are highly skilled, but not sufficiently aggressive. There are drills you can use to target that. The development of soccer skills requires this sort of targeted programme and approach. There is no one method for the improvement of soccer skills. There are different methods that are sensitive to what the developing player already has and, crucially, what he/she does not have.
It is this directed and flexible nature of targeting that is also a crucial feature of traditional moral education. A developing moral person can go awry in many ways – some cognitive, some meta-cognitive, some affective – and which specific aspect requires targeting depends on the kind of moral truancy in question. Perhaps the most obvious component of traditional moral education involves simply talking, and in particular questioning. But this sort of questioning can target several quite different brain structures and aspects of functioning – sometimes individually, at other times all at once. We can talk to a child in a variety of ways – depending on what we, their moral educator, think they need. “How do you think Johnny felt when you did that?” This sort of question targets projective empathy: the child is being asked to imaginatively put herself in the shoes of another. “How do you think you would feel if someone did that to you?” This functions in a similar way: the child is being asked to imaginatively project itself into a counterfactual situation. The same is true of “Remember when so-and-so did that to you?” Other questions target different systems. For example, the target of “You have a toy but Johnny doesn't. Does that seem fair?” seems to be the child's sense of fairness. This is also true of the far more advanced, Kantian inspired, question: “What would happen if everyone did that?” Notice that nothing in this sort of approach assumes that the deficit in question is a persisting one. It is not necessary that a child suffer from, for example, a persisting projective empathy deficit. Rather, all that is required for the above approach to be legitimate is that, in the judgement of the moral educator, this sort of deficit is evident at the present time.
By the time the question-and-answer form of moral education is likely to be efficacious, much stage setting – to use an expression of Wittgenstein's – must already have been put in place. Prior to that one must not overlook the importance of a distinctly Aristotelian component of moral education – habit. One becomes virtuous by doing virtuous things – repeatedly, habitually. Excellence – moral excellence included – is not an act, but a habit. Moral habits can be established in children in several ways but, again, the analogy with football education is quite useful. In the latter, you first demonstrate a skill you wish the players to acquire, explain why this skill is a desirable one to have, and then you let them practice it – at length – taking care to immediately correct mistakes. In moral education, a virtue, and its desirability, is often explained by way of a story in which a character exhibits it (or, sometimes, the lack thereof). Then the child's deviations from this virtue are corrected as expeditiously as possible.
Even habit, however, is not the earliest stage of traditional moral education. Before habit, there is the setting of an example. The tears of a sibling, for example, are met with the hugs and soothing voice of a parent. This is, to the other sibling, both a lesson in behaviour and an axiological lesson – a lesson about what has value or what should be valued. If those tears were met by the parent's indifference or annoyance, then a very different lesson would have been imparted.
This, of course, only scratches the surface of the forms traditional moral education may take. The crucial point, however, is that traditional moral education can, and often does, target very different capacities and deficiencies of a developing moral person – both cognitive and affective. The targeting is not indiscriminate but intelligent and based on an assessment of what a particular individual, at a particular time, needs, and with no assumption that these needs will stay the same over time. In a nutshell, traditional moral education addresses the person as a whole, and not some neurochemical or genetic aspect of that person. That is why it has been so successful.
A vehicle-content confusion is the mistake of supposing that structures and distinctions appropriate to one level of analysis will transpose seamlessly to another level, merely because the former supervenes or is otherwise dependent on the latter.Footnote 13 It may be that the idea of moral enhancement, as a neurochemical or genetic enterprise is guilty of this sort of confusion. We are moral beings, ultimately, because of our brains and our genes. However, being a moral person does not depend on this or that specific feature, whether neurochemical or genetic. It depends, rather, on a multiplicity of distinct features, each interacting with and dependent on the others. Traditional moral education – being capable of intelligent, flexible targeting of deficiencies as and when the need arises, is well suited to dealing with such an arrangement of factors. But the idea that neurochemical or genetic targeting will achieve the same sort of – or even better – outcome rests, I suspect, on a grossly oversimplified way of understanding what it is to be a moral being. This is not to say that such technological intervention could never be successful. But it is to say that, as yet, we have no clear idea of how it is supposed to work.
