Questions about Some Uses of Genetic Engineering
There is a widespread view that any project for the genetic improvement of the human race ought to be ruled out: that there are fundamental objections of principle. The aim of this discussion is to sort out some of the main objections. It will be argued that our resistance is based on a complex of different values and reasons, none of which is, when examined, adequate to rule out in principle this use of genetic engineering. The debate on human genetic engineering should become like the debate on nuclear power: one in which large possible benefits have to be weighed against big problems and the risk of great disasters. The discussion has not reached this point, partly because the techniques have not yet been developed. But it is also partly because of the blurred vision which fuses together many separate risks and doubts into a fuzzy-outlined opposition in principle.
1. AVOIDING THE DEBATE ABOUT GENES
AND THE ENVIRONMENT
In discussing the question of genetic engineering, there is everything to be said for not muddling the issue up with the debate over the relative importance of genes and environment in the development of such characteristics as intelligence. One reason for avoiding that debate is that it arouses even stronger passions than genetic engineering, and so is filled with as much acrimony as argument. But, apart from this fastidiousness, there are other reasons.
The nature-nurture dispute is generally seen as an argument about the relative weight the two factors
Reprinted with permission from What Sort of People Should There Be? (London: Penguin Books, 1984). pp. 25-33, 33-36, 41-43,45-52. Copyright @ 1984 Jonathan Glover. Reprinted with permission of Penguin Books Ltd.
have in causing differences within the human species: "IQ is 80 per cent hereditary and 20 per cent environmental" versus "IQ is 80 per cent environmental and 20 per cent hereditary," No doubt there is some approximate truth of this type to be found if we consider variations within a given population at a particular time. But it is highly unlikely that there is any such statement which is simply true of human nature regardless of context. To take the extreme case, if we could iron out all environmental differences, any residual variations would be 100 per cent genetic. It is only if we make the highly artificial assumption that different groups at different times all have an identical spread of relevant environmental differences that we can expect to find statements of this kind applying to human nature in general. To say this is not to argue that studies on the question should not be conducted. or are bound to fail. It may well be possible, and useful, to find out the relative weights of the two kinds of factor for a given characteristic among a certain group at a particular time. The point is that any such conclusions lose relevance, not only when environmental differences are stretched out or compressed, but also when genetic differences are. And this last case is what we are considering.
W,e can avoid this dispute because of its irrelevance. Suppose the genetic engineering proposal were to try to make people less aggressive. On a superficial view, the proposal might be shown to be unrealistic if there were evidence to show that variation in aggressiveness is hardly genetic at all: that it is 95 per cent environmental. (Let us grant, most implausibly, that such a figure turned out to be true for the whole of humanity, regardless of social context.) But all this would show is that, within our species, the distribution of genes relevant to aggression is very uniform.
It would show nothing about the likely effects on aggression if we use genetic engineering to give people a different set of genes from those they now have.
In other words, to take genetic engineering seriously, we need take no stand on the relative importance or unimportance of genetic factors in the explanation of the present range of individual differences found in people. We need only the minimal assumption that different genes could give us different characteristics. To deny that assumption you need to be the sort of person who thinks it is only living in kennels which make dogs different from cats.
2. METIlODS OF CHANGING THE GENETIC
COMPOSmON OF FUTURE GENERATIONS
There are essentially three ways of altering the genetic composition of future generations. The first is by environmental changes. Discoveries in medicine, the institution of a National Health Service, schemes for poverty relief, agricultural changes, or alterations in the tax position of large families, all alter the selective pressure on genes. 1 It is hard to think of any social change which does not make some difference to who survives or who is born.
The second method is to use eugenic policies aimed
at altering breeding patterns or patterns of survival of people with different genes. Eugenic methods are "environmental" too: the difference is only that the genetic impact is intended. Possible strategies range from various kinds of compulsion (to have more children, fewer children, or no children, or even compulsion over the choice of sexual partner) to the completely voluntary (our present genetic counselling practice of giving prospective parents information about probabilities of their children having various abnormalities).
The third method is genetic engineering: using enzymes to add to or subtract from a stretch of DNA.
Most people are unworried by the fact that a sideeffect of an environmental change is to alter the gene pool, at least where the alteration is not for the worse. And even in cases where environmental factors increase the proportion of undesirable genes in the pool, we often accept this. Few people oppose the National
Health Service, although setting it up meant that some people with genetic defects, who would have died, have had treatment enabling them to survive and reproduce. On the whole, we accept without qualms that much of what we do has genetic impact. Controversy starts when we think of aiming deliberately at genetic changes, by eugenics or genetic engineering.
I want to make some brief remarks about eugenic policies, before suggesting that policies of deliberate intervention are best considered in the context of genetic engineering.
Scepticism has been expressed about whether eugenic policies have any practical chance of success. Medawar has pointed out the importance of genetic polymorphism: the persistence of genetically different types in a population? (Our different blood groups are a familiar example.) For many characteristics, people get a different gene from each parent. So children do not simply repeat parental characteristics. Any simple picture of producing an improved type of person, and then letting the improvement be passed on unchanged, collapses.
But, although polymorphism is a problem for this crudely utopian form of eugenics, it does not show that more modest schemes of improvement must fail.
Suppose the best individuals for some quality (say,.
colour vision) are heterozygous, so that they inherit a gene A from one parent, and a gene B from the other. These ABs will have AAs and BBs among their children, who will be less good than they are. But AAs and BBs may still be better than ACs or ADs, and perhaps much better than CCs or CDs. If this were so, overall improvement could still be brought about by encouraging people whose genes included an A or B to have more children than those who had only Cs or Ds. The point of taking a quality like colour vision is that it may be genetically fairly simple. Qualities like kindness or intelligence are more likely to depend on the interaction of many genes, but a similar point can be made at a higher level of complexity.
Polymorphism raises a doubt about whether the offspring of the three "exceptionally intelligent women" fertilized by Dr. Shockley or other Nobel prize-winners will have the same IQ as the parents, even apart from environmental variation. But it does not show the inevitable failure of any large-scale attempts to alter human characteristics by varying the relative numbers of children different kinds of people have. Yet any attempt, say, to raise the level of intelligence, would be a very slow affair, taking many generations to make much of an impact. This is no rea
son for preferring to discuss genetic engineering. For the genetic engineering of human improvements, if it becomes possible, will have an immediate effect, so we will not be guessing which qualities will be desirable dozens of generations later.
EUGENICS AND HUMAN GENETICS
There is the view that the genetic-engineering techniques requires will not become a practical possibility. Sir MacFarlane Burnet, writing in 1971 about using genetic engineering to cure disorders in people already born, dismissed the possibility of using a virus to carry a new gene to replace a faulty one in cells throughout the body: "I should be willing to state in any company that the chance of doing this will remain infinitely small to the last syllable of recorded time.,,3 Unless engineering at the stage of sperm cell and egg is easier, this seems a confident dismissal of the topic to be discussed here. More recent work casts doubt on this confidence.4 So, having mentioned this skepticism, I shall disregard it. We will assume that genetic engineering of people may become possible, and that it is worth discussing. (Sir MacFarlane Burnet's view has not yet been falsified as totally as Rutherford's view about atomic energy. But I hope that the last syllable of recorded time is still some way off.)
The main reason for casting the discussion in terms of genetic engineering rather than eugenics is not a practical one. Many eugenic policies are open to fairly straightforward moral objections, which hide the deeper theoretical issues. Such policies as compulsory sterilization, compulsory abortion, compelling people to pair off in certain ways, or compelling people to have more or fewer children than they would other
wise have, are all open to objection on grounds of overriding people's autonomy. Some are open to objection on grounds of damage to the institution of the family. And the use of discriminatory tax- and childbenefit policies is an intolerable step towards a society of different genetic castes.
Genetic engineering need not involve overriding anyone's autonomy. It need not be forced on parents against their wishes, and the future .person being engineered has no views to be overridden. (The view that despite this, it is still objectionable to have one's genetic characteristics decided by others, will be considered later.) Genetic engineering will not damage the family in the obvious ways that compulsory eugenic policies would. Nor need it be encouraged by incentives which create inequalities. Because it avoids these highly visible moral objections, genetic engineering allows us to focus more clearly on other values that are involved.
(To avoid a possible misunderstanding, one point should be added before leaving the topic of eugenics. Saying that some eugenic policies are open to obvious
moral objections does not commit me to disapproval of all eugenic policies. In particular, I do not want to be taken to be opposing two kinds of policy. One is genetic counselling: warning people of risks in having children, and perhaps advising them against having them. The other is the introduction of screeningprogrammes to detect foetal abnormalities, followed by giving the mother the option of abortion where serious defects emerge.)
Let us now turn to the question of what, if anything, we should do in the field of human genetic engineering.
3. THE POSITIVE-NEGATIVE DISTINCTION
We are not yet able to cure disorders by genetic engineering. But we do sometimes respond to disorders by adopting eugenic policies, at least in voluntary form. Genetic counselling is one instance, as applied to those thought likely to have such disorders as Huntington's chorea. This is a particularly appalling inherited disorder, involving brain degeneration, leading to mental decline and lack of control over movement. It does not normally come on until middle age, by which time many of its victims would in the normal course of things have had children. Huntington's chorea is caused by a dominant gene, so those who find that one of the parents has it have themselves a 50 per cent chance of developing it. If they do have it, each of their children will in turn have a 50 per cent chance of the disease. The risks are so high and the disorder so bad that the potential parents often decide not to have children, and are often given advice to this effect by doctors and others.
Another eugenic response to disorders is involved in screening-programmes for pregnant women. When tests pick up such defects as Down's syndrome (mongolism) or spina bifida, the mother is given the possibility of an abortion. The screening-programmes are eugenic because part of their point is to reduce the incidence of severe genetic abnormality in the population.
These two eugenic policies come in at different stages: before conception and during pregnancy. For this reason the screening-programme is more controversial, because it raises the issue of abortion. Those
who are sympathetic to abortion, and who think it would be good to eliminate these disorders will be sympathetic to the programme. Those who think abortion is no different from killing a fully developed human are obviously likely to oppose the programme. But they are likely to feel that elimination of the disorders would be a good thing, even if not an adequate
justification for killing. Unless they also disapprove of contraception, they are likely to support the genetic-counselling policy in the case of Huntington's chorea.
Few people object to the use of eugenic policies to eliminate disorders, unless those policies have additional features which are objectionable. Most of us are resistant to the use of compulsion, and those who oppose abortion will object to screening-programmes. But apart from these other moral objections, we do not object to the use of eugenic policies against disease. We do not object to advising those likely to have Huntington's chorea not to have children, as neither compulsion nor killing is involved. Those of us who take this view have no objection to altering the genetic composition of the next generation, where this alteration consists in reducing the incidence of defects.
If it were possible to use genetic engineering to correct defects, say at the foetal stage, it is hard to see how those of us who are prepared to use the eugenic measure just mentioned could object. In both cases, it would be pure gain. The couple, one of whom may develop Huntington's chorea, can have a child if they want, knowing that any abnormality will be eliminated. Those sympathetic to abortion will agree that cure is preferable. And those opposed to abortion prefer babies to be born without handicap. It is hard to think of any objection to using genetic engineering to eliminate defects, and there is a clear and strong case for its use.
But accepting the case for eliminating genetic mistakes does not entail accepting other uses of genetic engineering. The elimination of defects is often called "negative" genetic engineering. Going beyond this, to bring about improvements in normal people, is by contrast "positive" engineering. (The same distinction can be made for eugenics.)
The positive-negative distinction is not in all cases completely sharp. Some conditions are genetic disorders whose identification raises little problem. Huntington's chorea or spina bifida are genetic "mistakes" in a way that cannot seriously be disputed. But with other conditions, the boundary between a defective state and normality may be more blurred. If there is a genetic disposition towards depressive illness, this seems a defect, whose elimination would be part of negative genetic engineering. Suppose the genetic disposition to depression involves the production oflower levels of an enzyme than are produced in normal people. The negative programme is to correct the genetic
fault so that the enzyme level is within the range found in normal people. But suppose that within "normal" people also, there are variations in the enzyme level, which correlate with ordinary differences in [the] tendency to be cheerful or depressed. Is there a sharp boundary between "clinical" depression and the depression sometimes felt by those diagnosed as "normal"? Is it clear that a sharp distinction can be drawn between raising someone's enzyme level so that it falls within the normal range and raising someone else's level from the bottom of the normal range to the top?
The positive-negative distinction is sometimes a blurred one, but often we can at least roughly see where it should be drawn. If there is a rough and ready distinction, the question is: how important is it? Should we go on from accepting negative engineering to accepting positive programmes, or should we say that the line between the two is the limit of what is morally acceptable?
There is no doubt that positive programmes arouse the strongest feelings on both sides. On the one hand, many respond to positive genetic engineering or positive eugenics with Professor Tinbergen's though: "I find it morally reprehensible and presumptuous for anybody to put himself forward as a judge of the qualities for which we should breed" [Guardian, 5 March, 1980].
But other people have held just as strongly that positive policies are the way to make the future of mankind better than the past. Many years ago H. J. Muller expressed this hope:
And so we foresee the history of life divided into three main phases. In the long preparatory phase it was the helpless creature of its environment, and natural selection gradually ground it into human shape. In the second-our own short transitional phase-it reaches out at the immediate environment, shaking, shaping and grinding to suit the form, the requirements, the wishes, and the whims of man. And in the long third phase, it will reach down into the secret places of the great universe of its own nature, and by aid of its ever growing intelligence and cooperation, shape itself into an increasingly sublime creation-a being beside which the mythical divinities of the past will seem more and more ridiculous, and which setting its own marvellous inner powers against the brute Goliath of the suns and the planets, challenges them to contest.:5
The case for positive engineering is not helped by adopting the tones of the mad scientist in a horror
EUGENICS AND HUMAN GENETICS
film. But behind the rhetoric is a serious point. If we decide on a positive programme to change our nature, this will be a central moment in our history, and the transformation might be beneficial to a degree we can now scarcely imagine. The question is: how are we to weigh this possibility against Tinbergen's objection, and against other objections and doubts?
For the rest of this discussion, I shall assume that, subject to adequate safeguards against things going wrong, negative genetic engineering is acceptable. The issue is positive engineering. I shall also assume that we can ignore problems about whether positive engineering will be technically possible. Suppose we have the power to choose people's genetic characteristics. Once we have eliminated genetic defects, what, if anything, should we do with this power? . . .
4. TIlE VIEW TIlAT OVERALL IMPROVEMENT
IS UNLIKELY OR IMPOSSIBLE
There is one doubt about the workability of schemes of genetic improvement which is so widespread that it would be perverse to ignore it. This is the view that, in any genetic alteration, there are no gains without compensating losses. On this view, if we bring about a genetically based improvement, such as higher intelligence, we are bound to pay a price somewhere else: perhaps the more intelligent people will have less resistance to disease, or will be less physically agile. If correct, this might so undermine the practicability of applying eugenics or genetic engineering that it would be hardly worth discussing the values involved in such programmes.
This view perhaps depends on some idea that natural selection is so efficient that, in terms of gene survival, we must already be as efficient as it is possible to be. If it were possible to push up intelligence without weakening some other part of the system, natural selection would already have done so. But this is a naive version of evolutionary theory. In real evolutionary theory, far from the genetic status quo always being the best possible for a given environment, some mutations turn out to be advantageous, and this is the origin of evolutionary progress. If natural mutations can be beneficial without a compensating loss, why should artificially induced ones not be so too?
It should also be noticed that there are two different ideas of what counts as a gain or a loss. From the point of view of evolutionary progress, gains and losses are simply advantages and disadvantages from
the point of view of gene survival. But we are not compelled to take this view. If we could engineer a genetic change in some people which would have the effect of making them musical prodigies but also sterile, this would be a hopeless gene in terms of survival, but this need not force us, or the musical prodigies themselves, to think of the changes as for the worse. It depends on how we rate musical ability as against having children, and evolutionary survival does not dictate priorities here.
The view that gains and losses are tied up with each other need not depend on the dogma that natural selection must have created the best of all possible sets of genes. A more cautiously empirical version of the claim says there is a tendency for gains to be accompanied by losses. John Maynard Smith, in his paper on "Eugenics and Utopia,,,6 takes this kind of "broad balance" view and runs it the other way, suggesting, as an argument in defence of medicine, that any loss of genetic resistance to disease is likely to be a good thing: "The reason for this is that in evolution, as in other fields, one seldom gets something for nothing. Genes which confer disease-resistance are likely to have hannful effects in other ways: this is certainly true of the gene for sickle-cell anaemia and may be a general rule. If so, absence of selection in favour of disease-resistance may be eugenic."
It is important that different characteristics may turn out to be genetically linked in ways we do not yet realize. In our present state of knowledge, engineering for some improvement might easily bring some unpredicted but genetically linked disadvantage. But we do not have to accept that there will in general be a broad balance, so that there is a presumption that any gain will be accompanied by a compensating loss (or Maynard Smith's version that we can expect a compensating gain for any loss). The reason is that what counts as a gain or loss varies in different contexts. Take Maynard Smith's example of sickle-cell anaemia. The reason why sickle-cell anaemia is widespread in Africa is that it is genetically linked with resistance to malaria. Those who are heterozygous (who inherit one sickle-cell gene and one normal gene) are resistant to malaria, while those who are homozygous (whose
genes are both sickle-cell) get sickle-cell anaemia. If
we use genetic engineering to knock out sickle-cell anaemia where malaria is common, we will pay the price of having more malaria. But when we eradicate malaria, the gain will not involve this loss. Because losses are relative to context, any generalization about the impossibility of overall improvements is dubious.
5. THE FAMILY AND OUR DESCENDANTS
Unlike various compulsory eugenic policies, genetic engineering need not involve any interference with decision by couples to have children together, or with their decisions about how many children to have. And let us suppose that genetically engineered babies grow in the mother's womb in the normal way, so that her relationship to the child is not threatened in the way it might be if the laboratory or the hospital were substituted for the womb. The cruder threats to family relationships are eliminated.
It may be suggested that there is a more subtle threat. Parents like to identify with their children. We are often pleased to see some of our own characteris
tics in our children. Perhaps this is partly a kind of vanity, and no doubt sometimes we project on to our children similarities that are not really there. But, when the similarities do exist, they help the parents and children to understand and sympathize with each other. If genetic engineering resulted in children fairly different from their parents, this might make their relationship have problems.
There is something to this objection, but it is easy to exaggerate. Obviously, children who were like Midwich cuckoos, or comic-book Martians, would not be easy to identify with. But genetic engineering need not move in such sudden jerks. The changes would have to be detectable to be worth bringing about, but there seems no reason why large changes in appearance, or an unbridgeable psychological gulf, should be created in anyone generation. We bring about environmental changes which make children different
from their parents, as when the first generation of children in a remote place are given schooling and made literate. This may cause some problems in fam
ilies, but it is not usually thought a decisive objection. It is not clear that genetically induced changes of similar magnitude are any more objectionable.
A related objection concerns our attitude to our remoter descendants. We like to think of our descendants stretching on for many generations. Perhaps this is in part an immortality substitute. We hope they will to some extent be like us, and that, if they think of us, they will do so with sympathy and approval. Perhaps these hopes about the future of mankind are relatively unimportant to us. But, even if we mind about them a lot, they are unrealistic in the very long term. Genetic engineering would make our descendants less like us, but this would only speed up the natural rate of change. Natural mutations and selective pressures make it un
likely that in a few million years our descendants will be physically or mentally much like us. So what genetic engineering threatens here is probably doomed anyway. . . .
[6.] RISKS AND MISTAKES
Although mixing different species and cloning are often prominent in people's thoughts about genetic engineering, they are relatively marginal issues. This is partly because there may be no strong reasons in favour of either. Our purposes might be realized more readily by improvements to a single species, whether another or our own, or by the creation of quite new types of organism, than by mixing different species. And it is not clear what advantage cloning batches of people might have, to outweigh the drawbacks. This is not to be dogmatic that species mixing and cloning
could never be useful, but to say that the likelihood of other techniques being much more prominent makes it a pity to become fixated on the issues raised by these ones. And some of the most serious objections to positive genetic engineering have wider application than to these rather special cases. One of these wider objections is that serious risks may be involved.
Some of the risks are already part of the public debate because of current work on recombinant DNA. The danger is of producing harmful organisms that would escape from our control. The work obviously should take place, if at all, only with adequate safeguards against such a disaster. The problem is deciding what we should count as adequate safeguards. I have nothing to contribute to this problem here. If it can be dealt with satisfactorily, we will perhaps move on to genetic engineering of people. And this introduces another dimension of risk. We may produce unintended results, either because our techniques turn
out to be less finely tuned than we thought, or because different characteristics are found to be genetically linked in unexpected ways.
If we produce a group of people who turn out worse than expected, we will have to live with them. Per
haps we would aim for producing people who were especially imaginative and creative, and only too late find we had produced people who were also very violent and aggressive. This kind of mistake might not only be disastrous, but also very hard to "correct" in subsequent generations. For when we suggested sterilization to the people we had produced, or else corrective genetic engineering for their offspring, we
EUGENICS AND HUMAN GENETICS
might find them hard to persuade. They might like the way they were, and reject, in characteristically violent fashion, our explanation that they were a mistake.
The possibility of an irreversible disaster is a strong deterrent. It is enough to make some people think we should rule out genetic engineering altogether, and to make others think that, while negative engineering is perhaps acceptable, we should rule out positive engineering. The thought behind this second position is that the benefits from negative engineering are clearer, and that, because its aims are more modest, disastrous mistakes are less likely.
The risk of disasters provides at least a reason for saying that, if we do adopt a policy of human genetic engineering, we ought to do so with extreme caution. We should alter genes only where we have strong reasons for thinking the risk of disaster is very small, and where the benefit is great enough to justify the risk. (The problems of deciding when this is so are familiar from the nuclear power debate.) This "principle of caution" is less strong than one ruling out all positive engineering, and allows room for the possibility that the dangers may turn out to be very remote, or that greater risks of a different kind are involved in not using positive engineering. These possibilities correspond to one view of the facts in the nuclear power debate. Unless with genetic engineering we think we can already rule out such possibilities, the argument from risk provides more justification for the principle of caution than for the stronger ban on all positive engineering. . . .
Some of the strongest objections to positive engineering are not about specialized applications or about risks. They are about the decisions involved. The centralline of thought is that we should not start playing God by redesigning the human race. The suggestion is that there is no group (such as scientists, doctors, public officials, or politicians) who can be entrusted with decisions about what sort of people there should be. And it is also doubted whether we could have any adequate grounds for basing such decisions on one set of values rather than another. . . .
1. NCYf PLAYING GOD
Suppose we could use genetic engineering to raise the average IQ by fifteen points. (I mention, only to ignore, the boring objection that the average IQ is al
ways by definition 100.) Should we do this? Objectors to positive engineering say we should not. This is not because the present average is preferable to a higher one. We do not think that, if it were naturally fifteen points higher, we ought to bring it down to the present level. The objection is to our playing God by deciding what the level should be.
On one view of the world, the objection is relatively straightforward. On this view, there really is a God, who has a plan for the world which will be disrupted if we stray outside the boundaries assigned to us. (It is relatively straightforward: there would still be the problem of knowing where the boundaries came. If genetic engineering disrupts the programme, how do we know that medicine and education do not?)
The objection to pl~ying God has a much wider appeal than to those who literally believe in a divine plan. But, outside such a context, it is unclear what the objection comes to. If we have a Darwinian view, according to which features of our nature have been selected for their contribution to gene survival, it is not blasphemous, or obviously disastrous, to start to control the process in the light of our own values. We may value other qualities in people, in preference to those which have been most conducive to gene survival.
The prohibition on playing God is obscure. If it tells us not to interfere with natural selection at all, this rules out medicine, and most other environmental and social changes. If it only forbids interference with natural selection by the direct alteration of genes, this rules out negative as well as positive genetic engineering. If these interpretations are too restrictive, the ban on positive engineering seems to need some explanation. If we can make positive changes at the environmental level, and negative changes at the genetic level, why should we not make positive changes at the genetic level? What makes this policy, but not the others, objectionably God-like?
Perhaps the most plausible reply to these questions rests on a general objection to any group of people trying to plan too closely what human life should be like. Even if it is hard to distinguish in principle between the use of genetic and environmental means, genetic changes are likely to differ in degree from most environmental ones. Genetic alterations may be more drastic or less reversible, and so they can be seen as the extreme case of an objectionably God-like policy by which some people set out to plan the lives of others.
This objection can be reinforced by imagining the possible results of a programme of positive engineer
ing, where the decisions about the desired improvements were taken by scientists. Judging by the literature written by scientists on this topic, great prominence would be given to intelligence. But can we be sure that enough weight would be given to other desirable qualities? And do things seem better if for scientists we substitute doctors, politicians or civil servants? Or some committee containing businessmen, trade unionists, academics, lawyers and a clergyman?
What seems worrying here is the circumscribing of potential human development. The present genetic lottery throws up a vast range of characteristics, good and bad, in all sorts of combinations. The group of people controlling a positive engineering policy would inevitably have limited horizons, and we are right to worry that the limitations of their outlook might become the boundaries of human variety. The drawbacks would be like those of town-planning or dogbreeding, but with more important consequences.
When the objection to playing God is separated from the idea that intervening in this aspect of the natural world is a kind of blasphemy, it is a protest against a particular group of people, necessarily fallible and limited, taking decisions so important to our future. This protest may be on grounds of the bad consequences, such as loss of variety of people, that would come from the imaginative limits of those taking the decisions. Or it may be an expression of opposition to such concentration of power, perhaps with the thought: 'What right have they to decide what kinds of people there should be?' Can these problems be side-stepped?
2. TIlE GENETIC SUPERMARKET
Robert Nozick is critical of the assumption that positive engineering has to involve any centralized decision about desirable qualities: "Many biologists tend to think the problem is one of design, of specifying the best types of persons so that biologists can proceed to produce them. Thus they worry over what sort(s) of person there is to be and who will control this process. They do not tend to think, perhaps because it diminishes the importance of their role, of a system in which they run a "genetic supermarket," meeting the individual specifications (within certain moral limits ) of prospective parents. Nor do they think of seeing what limited number of types of persons people's choices would converge upon, if indeed there would be any such convergence. This supermarket system has the great virtue that it involves no centralized decision fixing the future human type(s).'.?
This idea of letting parents choose their children's
characteristics is in many ways an improvement on decisions being taken by some centralized body. It seems less likely to reduce human variety, and could even increase it, if genetic engineering makes new combinations of characteristics available. (But we should be cautious here. Parental choice is not a guarantee of genetic variety, as the influence of fashion or of shared values might make for a small number of types on which choices would converge.)
To those sympathetic to one kind of liberalism, Nozick's proposal will seem more attractive than centralized decisions. On this approach to politics, it is wrong for the authorities to institutionalize any religious or other outlook as the official one of the society. To a liberal of this kind, a good society is one which tolerates and encourages a wide diversity of ideals of the good life. Anyone with these sympathies will be suspicious of centralized decisions about what sort of people should form the next generation. But some parental decisions would be disturbing. If parents chose characteristics likely to make their children unhappy, or likely to reduce their abilities, we might feel that the children should be protected against this. (Imagine parents belonging to some extreme religious sect, who wanted their children to have a religious symbol as a physical mark on their face, and who wanted them to be unable to read, as a protection against their faith being corrupted.) Those of us who support restrictions protecting children from parental harm after birth (laws against cruelty, and compulsion on parents to allow their children to be educated and to have necessary medical treatment) are likely to support protecting children from being harmed by their parents' genetic choices.
No doubt the boundaries here will be difficult to draw. We already find it difficult to strike a satisfactory balance between protection of children and parental freedom to choose the kind of upbringing their children should have. But it is hard to accept that society should set no limits to the genetic choices parents can make for their children. Nozick recognizes this when he says the genetic supermarket should meet the specifications of parents "within certain morallirnits." So, if the supermarket came into existence, some centralized policy, even if only the restrictive one of ruling out certain choices harmful to the children, should exist. It would be a political decision where the limits should be set.
There may also be a case for other centralized
EUGENICS AND HUMAN GENETICS
restrictions on parental choice, as well as those aimed at preventing harm to the individual people being designed. The genetic supermarket might have more oblique bad effects. An imbalance in the ratio between the sexes could result. Or parents might think their children would be more successful if they were more thrusting, competitive and selfish. If enough parents acted on this thought, other parents with different values might feel forced into making similar choices to prevent their own children being too greatly disadvantaged. Unregulated individual decisions could lead to shifts of this kind, with outcomes unwanted by most of those who contribute to them. If a majority favour a roughly equal ratio between the sexes, or a population of relatively uncompetitive people, they may feel justified in supporting restrictions on what parents can choose. (This is an application to the case of genetic engineering of a point familiar in other contexts, that unrestricted individual choices can add up to a total outcome which most people think worse than what would result from some regulation.)
Nozick recognizes that there may be cases of this sort. He considers the case of avoiding a sexual imbalance and says that "a government could require that genetic manipulation be carried on so as to fit a certain ratio."s He clearly prefers to avoid governmental intervention of this kind, and, while admitting that the desired result would be harder to obtain in a purely libertarian system, suggests possible strategies for doing so. He says: "Either parents would subscribe to an information service monitoring the recent births and so know which sex was in shorter supply (and hence would be more in demand in later life), thus adjusting their activities, or interested individuals would contribute to a charity that offers bonuses to maintain the ratios, or the ratio would leave 1:1, with new family and social patterns developing." The proposals for avoiding the sexual imbalance without central regulation are not reassuring. Information about likely prospects for marriage or sexual partnership might not be decisive for parents' choices. And, since those most likely to be "interested individuals" would be in the age group being genetically engineered, it is not clear that the charity would be given donations adequate for its job.9
If the libertarian methods failed, we would have the choice between allowing a sexual imbalance or imposing some system of social regulation. Those
who dislike central decisions favouring one sort of
person over others might accept regulation here, on the grounds that neither sex is being given preference: the aim is rough equality of numbers.
But what about the other sort of case, where the working of the genetic supermarket leads to a general change unwelcome to those who contribute to it? Can we defend regulation to prevent a shift towards a more selfish and competitive population as merely being the preservation of a certain ratio between characteristics? Or have we crossed the boundary, and allowed a centralized decision favouring some characteristics over others? The location of the boundary is obscure. One view would be that the sex-ratio case is acceptable because the desired ratio is equality of numbers. On another view, the acceptability derives from the fact that the present ratio is to be preserved. (In this second view, preserving altruism would be acceptable, so long as no attempt was made to raise the proportion of altruistic people in the population. But is this boundary an easy one to defend?)
If positive genetic engineering does become a reality, we may be unable to avoid some of the decisions being taken at a social level. Or rather, we could avoid this, but only at what seems an unacceptable cost, either to the particular people being designed, or to their generation as a whole. And, even if the social decisions are only restrictive, it is implausible to claim that they are all quite free of any taint of preference for some characteristics over others. But, although this suggests that we should not be doctrinaire in our support of the liberal view, it does not show that the view has to be abandoned altogether. We may still think that social decisions in favour of one type of person rather than another should be few, even if the consequences of excluding them altogether are unacceptable. A genetic supermarket, modified by some central regulation, may still be better than a system of purely central decisions. The liberal value is not obliterated because it may sometimes be compromised for the sake of other things we care about.
3. A MIXED SYSTEM
The genetic supermarket provides a partial answer to the objection about the limited outlook of those who would take the decisions. The choices need not be concentrated in the hands of a small number of people. The genetic supermarket should not operate in a completely unregulated way, and so some centralized decisions would have to be taken about the restrictions that should be imposed. One system that would answer many of the anxieties about centralized
decision-making would be to limit the power of the decision-makers to one of veto. They would then only check departures from the natural genetic lottery, and so the power to bring about changes would not be given to them, but spread through the whole population of potential parents. Let us call this combination of parental initiative and central veto a "mixed system." If positive genetic engineering does come about, we can imagine the argument between supporters of a mixed system and supporters of other decisionmaking systems being central to the political theory of the twenty-first century, parallel to the place occupied in the nineteenth and twentieth centuries by the debate over control of the economy. 10
My own sympathies are with the view that, if positive genetic engineering is introduced, this mixed system is in general likely to be the best one for making decisions. I do not want to argue for an absolutely inviolable commitment to this, as it could be that some centralized decision for genetic change was the only way of securing a huge benefit or avoiding a great catastrophe. But, subject to this reservation, the dangers of concentrating the decision-making create a strong presumption in favour of a mixed system rather than one in which initiatives come from the centre. And, if a mixed system was introduced, there would have to be a great deal of political argument over what kinds of restrictions on the supermarket should be imposed. Twenty-first-century elections may be about issues rather deeper than economics.
If this mixed system eliminates the anxiety about genetic changes being introduced by a few powerful people with limited horizons, there is a more general unease which it does not remove. May not the limitations of one generation of parents also prove disastrous? And, underlying this, is the problem of what values parents should appeal to in making their choices. How can we be confident that it is better for one sort of person to be born than another?
The dangers of such decisions, even spread through all prospective parents, seem to me very real. We are swayed by fashion. We do not know the limitations of our own outlook. There are human qualities whose value we may not appreciate. A generation of parents might opt heavily for their children having physical or intellectual abilities and skills. We might leave out a sense of humour. Or we might not notice how important to us is some other quality, such as emotional warmth. So we might not be disturbed in advance by
the possible impact of the genetic changes on such a quality. And, without really wanting to do so, we might stumble into producing people with a deep coldness. This possibility seems one of the worst imaginable. It is just one of the many horrors that could be blundered into by our lack of foresight in operating the mixed system. Because such disasters are a real danger, there is a case against positive genetic engineering, even when the changes do not result from centralized decisions. But this case, resting as it does on the risk of disaster, supports a principle of caution rather than a total ban. We have to ask the question whether there are benefits sufficiently great and sufficiently probable to outweigh the risks.
But perhaps the deepest resistance, even to a mixed system, is not based on risks, but on a more general problem about values. Could the parents ever be justified in choosing, according to some set of values, to create one sort of person rather than another?
Is it sometimes better for us to create one sort of person rather than another? We say "yes" when it is a question of eliminating genetic defects. And we say "yes" if we think that encouraging some qualities rather than others should be an aim of the upbringing and education we give our children. Any inclination to say "no" in the context of positive genetic engineering must lay great stress on the two relevant boundaries. The positive-negative boundary is needed to mark off the supposedly unacceptable positive policies from the acceptable elimination of defects. And the genes-environment boundary is needed to mark off positive engineering from acceptable positive aims of educational policies. But it is not clear that confidence in the importance of these boundaries is justified... .
1. Chris Graham has suggested to me that it is misleading to say this without emphasizing the painful slowness of this way of changing gene frequencies.
2. The Future of Man (The Reith Lectures, 1959), London, 1960, chapter 3; and in 'The Genetic Improvement of Man," in The
Hope of Progress. London, 1972.
3. Genes. Dreams and Realities, London, 1971, p. 81.
4. "Already they have pushed Cline's results further, obtain
ing transfer between rabbit and mouse, for example, and good expression of the foreign gene in its new host. Some, by transferring the genes into the developing eggs, have managed to get the new genes into every cell in the mouse, including the sex cells; those mice have fathered offspring who also contain the foreign gene." Jeremy Cherfas: Man Made life. Oxford, 1982, pp. 229-30.