Journal of Applied Philosophy, Vol. 20, No. 3, 2003
Genetic Engineering and Autonomous Agency 223
Genetic Engineering and Autonomous Agency
LINDA BARCLAY
abstract In this paper I argue that the genetic manipulation of sexual orientation at the
embryo stage could have a detrimental effect on the subsequent person s later capacity for
autonomous agency. By focussing on an example of sexist oppression I show that the norms
and expectations expressed with this type of genetic manipulation can threaten the development
of autonomous agency and the kind of social environment that makes its exercise likely.
I
Some people claim that in the future it may be possible to manipulate DNA so as to
influence a person s sexual orientation. I have no idea how feasible this is. Nor do we
yet have conclusive evidence that sexual orientation is very much influenced by genes
in any case. But suppose there are so-called gay genes and suppose also that it will be
feasible at some later time to manipulate them. Would it be morally wrong to do so [1]?
I want to argue that the genetic manipulation of sexual orientation at the embryo
stage can potentially have a detrimental effect on that person s later capacity for
autonomous agency. The few philosophers that have considered the question have
explicitly rejected that this kind of genetic engineering could have any relevance to the
question of autonomous agency [2]. Most philosophers concerned with autonomous
agency focus almost exclusively on the individual capacities and attributes that an
agent must possess to be an autonomous agent and from this perspective how one
ultimately acquired one s genes for sexual orientation seems completely irrelevant, as
I will show in the next section. Very little attention, if any, is given to the conditions
required for the development of such capacities and the kind of social environment
that makes the effective exercise of them possible, or at least more likely. In the
following sections of the paper I will argue that more careful consideration of these
social preconditions for autonomous agency makes it quite plausible to suppose that
this kind of genetic engineering can impact detrimentally upon it.
II
One reason for thinking that the genetic manipulation of genes for sexual orientation
might be relevant to the issue of autonomous agency is based on a suspicion that some
choice properly belonging to the agent at some later stage of his life has been taken
away from him. But if we believe that parents selecting a non-gay gene means that the
later person will be heterosexual completely independently of any choices he makes,
then we must believe that what genes one has for sexual orientation determine one s
© Society for Applied Philosophy, 2003
© Society for Applied Philosophy, 2003, Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main
Street, Malden, MA 02148, USA.
224 Linda Barclay
sexual orientation completely independently of one s choices in any case: we don t
believe that sexual orientation is something over which we can exercise autonomous
agency. In that case, it is wholly mysterious why one would think genetic engineering
undermines autonomous agency, for sexual orientation is not something over which
one exercises autonomous agency in any case. From this perspective we can agree with
Nicolas Agar, who represents the commonly held view on this matter, when he says
 There doesn t seem to be much difference for me if a given psychological feature, be
it good or bad, is the result of the random juggling of genes in the formation of my
parent s sex cells which then come together to produce my genotype, or are the
consequences of the intervention of another person [3].
I am going to assume that even if there are genetic markers for sexual preference,
one s sexual orientation, and certainly one s sexual behaviour, is not entirely determined
by them. It would seem terribly implausible to suggest that education, socialisation,
culture and so on do not affect people s sexual orientation and sexual behaviour.
Our sexual behaviour and tendencies, and even our sexual desires, are a product of an
often inscrutable concatenation of all of these forces. As this suggests, we constantly
find ourselves as the bearers of desires, impulses, emotions, beliefs and even values that
have their ultimate origins outside of our choices  in socialisation, genes, education
and so on. From the perspective of most conceptions of autonomous agency, whether
or not we can be autonomous with respect to sexual orientation and behaviour turns
on how we can respond to those features of ourselves  desires, values, tendencies,
emotions and so on  that we find ourselves with. Some kind of self-reflection and
self-revision is central to such accounts, although what such ideal reflection is taken
to consist of varies widely. What threatens a person s autonomous agency is some
incapacity with respect to her ability to reflect on her will and revise it. The content of
the agent s will should be endorsed or authorised by her according to what she desires
or values most, where  most refers to what she desires at a higher-order [4], or values
that are derived from the non-appetitive part of the soul [5], or desires and values that
are a response to reasons [6], and so on. If she has this kind of control over her will,
and moreover, if her will determines her actions, then she is an autonomous agent.
Many, but not all, accounts of autonomous agency are procedural in nature. Such
accounts claim that freedom imposes no particular restrictions on the contents of a
person s will: as long as her desires and values are arrived at after a suitable process of
reflection and revision, then an agent who is able to govern her conduct in line with
such desires and values is thought to be autonomous, whatever their content. In other
words, there are no particular desires and values an autonomous person must have,
only that whatever desires and values she does have come about via a certain pro-
cedure of reflection and scrutiny [7].
Of significant concern for such proceduralist accounts is the fact that the process
of reflection and scrutiny can itself be vulnerable to manipulation and distortion. If a
person is brainwashed or manipulated, then it is mysterious how the process of reflec-
tion and revision supposedly secures her autonomy. The desires and values that result
from such reflection seem no better off, perhaps even worse off, than those she ori-
ginally found herself with [8]. Proceduralists have acknowledged this deep worry for
the theory, and have suggested a number of possible ways of distinguishing forms of
interference that defeat the autonomy-conferring nature of the reflective process from
those that do not. Thus, Gerald Dworkin acknowledges the need to specify more
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 225
carefully what constitutes  procedural independence , which would enable us to distin-
guish harmless education and socialisation on the one hand from brainwashing and
manipulation on the other [9], whilst Harry Frankfurt appeals to the agent s  whole-
hearted identification as legitimately ending further questions about the need for
further reflection and scrutiny [10]. It remains an open question whether strictly
proceduralist accounts of autonomous agency can properly handle these problems.
One significant response to the concern has been to develop a more substantive con-
ception of autonomy. For example, Susan Wolf argues that autonomy or freedom
consists in the ability to track the reasons that there are, or the True and the Good as
she loftily puts it [11]. It s not clear to me whether such a view is best described as a
proceduralist account which demands that the procedure of reflection track the truth,
or as a substantive account which demands that a genuinely autonomous agent hold
specific true beliefs and good values. Whatever the best description of the view, it
moves decisively away from the proceduralist insistence that a free person need not
have any particular desires or values as long as the desires and values she does have are
arrived at after some process of reflective scrutiny.
Despite the significant differences between these accounts, none of them would
seem to advance the idea that genetic engineering of sexual orientation could pose a
threat to autonomous agency. That one s genetic markers for sexual orientation have
been manipulated at the embryo stage would not appear to have any detrimental
impact, indeed any impact at all, on the relevant capacities for critical scrutiny and
revision. Clearly replacing someone s gay genes is not like brainwashing him, nor does
it prevent him from exercising his good reason just like most of the rest of us. While
one s genes may dispose one toward certain preferences or behaviours, this does not
preclude autonomy-conferring reflection and revision and for this reason precisely how
one came by one s genes in the first place is of no import to the question of one s
capacity for autonomy. In sum, whether we view our genes as wholly determining our
sexual orientation or not, genetic manipulation of these genes seems to have no bearing
at all on the question of autonomous agency.
III
To tell a different story about the possible relevance of the intentional manipulation of
our genetic markers for sexual orientation I want to begin by considering the story of
a nineteenth-century woman, loosely modelled on the life of feminist author Charlotte
Perkins Gilman, as discussed by Paul Benson [12]. In this story Charlotte s autonom-
ous agency has been undermined by the effects of oppressive norms about women in a
number of ways, some more obvious than others. Yet in other respects Charlotte also
manages to evade or overcome that oppressive socialisation. A consideration of the
ways in which oppressive socialisation can undermine autonomous agency in cases like
Charlotte s will throw light on the possible effects of genetic engineering to be discussed
in the next section.
Charlotte s story goes something like this. She so detests the conventional demands
made of her that she wants to leave her husband and young child to pursue an artistic
career. But so alien are her desires from norms of nineteenth-century morality and
femininity that Charlotte s family and friends fear for her sanity. A well-respected
© Society for Applied Philosophy, 2003
226 Linda Barclay
doctor diagnoses her as suffering from a nervous disorder and prescribes the rest cure,
which means Charlotte is to stay in bed, isolate herself from all friends and refrain
from any painting or writing. Importantly, Charlotte concurs with the doctor s judge-
ment: she thinks she must indeed be crazy to prefer a life of painting and writing over
a life spent caring for her husband and child. But despite concurring with the judge-
ment that she does have a serious nervous disorder, she just cannot tolerate the rest
cure. She abandons her husband and child to pursue her artistic career, all the while
convinced that she is crazy; indeed, taking her behaviour as evidence of her madness.
Failures of Reflection
For many (most?) women of the nineteenth-century subject to the same norms as
Charlotte, it is doubtful that serious ambitions to write and paint would have survived
at all. Surely one prominent effect of oppressive norms is the way they shape the
agent s beliefs, preferences, and values, working effectively in many cases to ensure
that those that conflict with powerful social norms don t flourish to begin with. More-
over, reflective scrutiny, if it is undertaken at all, takes place under the shadow of the
internalisation of norms that lead women to conclude that such desires are, selfish,
immoral or even perverse. In other words, even if they have begun to emerge, such
preferences don t usually survive reflective scrutiny because the scrutiny itself is tainted
by the internalisation of oppressive norms. Charlotte as we have described her is quite
exceptional, given that she retains her desire to paint and write despite the norms she
is subject to.
We saw in the previous section that it remains an open question whether or
not proceduralists can adequately account for this defeat of autonomous agency by
oppressive norms that taint the reflective process. If they can successfully explain, on
proceduralist grounds, why oppressive norms undermine the integrity of the reflective
process then they can account for the loss of autonomous agency in such cases.
A number of feminists have been sceptical that a strictly procedural account can
adequately do this, a significant failure from a feminist perspective which considers the
day to day effects of oppressive socialisation a more real threat to autonomous agency
than bizarre scenarios involving mad scientists manipulating brain waves [13].
Non-proceduralists can more readily claim that autonomous agency is defeated simply
because of the agent s false beliefs and values, perhaps combined with a claim about
her inability, given oppressive social norms, to track the right ones. In any case,
whether or not such cases force us to abandon strictly procedural accounts, they point
to significant failures of reflection that uncontroversially undermine genuinely auto-
nomous agency.
Failure of Self Worth
Failures of reflection with respect to the desirability or value of certain options  such
as painting and writing as opposed to motherhood  represent only one significant
threat to autonomous agency. What Benson s discussion of Charlotte shows is that
Charlotte also suffers from a failure of self worth. She thinks she is crazy, or incompet-
ent to reflectively authorise her will. Benson argues that autonomous agency requires
not only various capacities and skills to reflectively authorize conduct but also crucially
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 227
requires a sense of worthiness to exercise those capacities and skills competently. He
suggests that Charlotte is disengaged or alienated from her actions because she doesn t
trust herself to use these powers competently, precisely because she thinks she is crazy.
Benson insists that the case  does not involve just another instance of an impediment
to an agent s powers to regulate her conduct reflectively. In this case, it is the agent s
very identification with the possession and exercise of those powers which has been
threatened by her revised view of her own conduct [14].
There is a significant question concerning whether failures of reflection regarding the
desirability or worthiness of certain options can occur independently of failures of self
worth. On the one hand it seems that they can. Because of the internalisation of certain
oppressive norms, one may well suffer a failure of reflection with regard to a certain
option. Let s suppose that Charlotte fully believes that her desire to paint and write is
deeply immoral and selfish. In choosing to write and paint anyway, Charlotte may
simply embrace her decision to do the  wrong thing. She may not doubt at all her
capacity for agency. She may even take delight in pursuing the course that she believes
is wrong. Or, alternatively, someone like Charlotte may simply conform to the norms
in question  give up her desire to paint and write  and believe that in doing so her
free and worthy agency has triumphed over the existence of immoral desires. She may
take pride in her ability to conquer her aberrant desires. In such cases Charlotte s
internalisation of oppressive norms has caused failures of reflection with regard to the
worthiness of a certain option, but not a failure of self worth as an agent. On the other
hand, failures of reflection regarding the worth of a certain option may be intimately
connected to failures of self worth. If a woman like Charlotte wishes to paint and write,
but believes that this is wrong, selfish or even perverse, this may well contribute to her
sense that she herself is crazy, or incompetent as an agent, given that she persists in
wanting to paint and write in any case.
There is an equally significant question at stake about whether, conversely, failures
of self worth can occur without leading to failures of reflection. Benson argues that
they can by suggesting that Charlotte may suffer a total loss of self worth as an agent,
and yet not suffer failures of reflection with regard to the worthiness of certain options.
She may carefully reflectively endorse her desire to write and paint, yet all the while
believe she is crazy. Despite Benson s suggestion, a failure of one s sense of worthiness
as an agent is overwhelmingly likely to lead to failures of reflection about the worthiness
of certain options. A woman who is made to believe she is crazy and thus doubts her
competence to make wise decisions is surely also likely to be a genuinely bad decision
or choice maker. Of course, the proceduralist and non-proceduralist accounts are likely
to identify such failures of reflection somewhat differently. She is likely to suffer severe
ambivalence and uncertainty in reflecting on and deciding what she most values or
desires and equal if not greater ambivalence and uncertainty in exerting those values
and desires in action. She is likely to be torn apart by inconsistencies, dissonance
and uncertainty, all of which undermine her autonomous agency even on very basic
proceduralist accounts [15]. Non-proceduralists will also emphasise that a woman
who thinks she is crazy is likely to be very bad at tracking good reasons. Moreover,
if a person suffers a complete loss of self worth as an agent, she is likely to suffer from
a severe lack of motivation to exercise agency, believing as she does that she will just
make a mess of it. A lack of motivation threatens autonomous agency on both pro-
cedural and non-procedural accounts.
© Society for Applied Philosophy, 2003
228 Linda Barclay
What particularly contributes to the loss of self worth as an agent, and thus further
contributes to these failures of reflection, is when oppressive social norms not only
mark out certain options as undesirable, but also mark out certain agents as less
capable of agency. I have indicated that just suffering a failure of reflection regarding
certain options may cause a failure of one s worth as an agent. Thus Charlotte, who
believes painting and writing are immoral, just by virtue of that fact alone may suffer a
failure of self worth, given that she wants to paint and write in any case. But what
makes it much more likely that this secondary effect on her self worth will follow
(rather than not) is when the oppressive norms that Charlotte has to contend with
don t merely identify painting and writing as undesirable for women, they also work to
cast doubt on women s capacity for agency in the first place. They do this by the
suggestion that women are less capable of autonomous agency than men. Thus, we
have a more complicated story about the way oppressive norms work. They may work
merely by singling out certain options, choices, activities as worthless, or they may
work by singling out certain people, or certain classes of people as less capable of
autonomous agency. Whilst singling out certain options as worthless may on its own
lead to a loss of self worth as an agent, it is much more likely to do so if such agents
are also, as a class, believed to be less capable of agency.
Failures of Competence
As should now be clear, oppressive social norms can significantly diminish the auto-
nomous agency of people subject to them. The internalisation of oppressive social
norms can cause serious failures of reflection and social worth. Yet even more directly,
oppressive social norms can hinder the very development of whatever reflective and
psychological capacities are necessary for autonomous agency. A number of feminists
have argued that the kind of socialisation that girls, as opposed to boys, are subject to
tends to hinder the full development of the skills and capacities required to engage in
reflective scrutiny and revision in the first place [16]. Thus, oppressive norms can not
only interfere with and distort an agent s attempt to exercise her skills and capacities of
autonomous agency; they can hinder their very development from a young age.
Failures of Action
A final obvious way that oppressive norms can undermine autonomous agency is by
undermining one s willingness or ability to act on one s reflectively authorised desires and
values. Many (most?) women like Charlotte who manage to retain their aberrant desires
or values, even after a process of critical reflection, and who retain their sense of worth
as agents, often fail to act on their desires and values, rightly fearing the kind of punish-
ment and censure they will be subject to. Failure to put one s will into effect, because
of weakness of will, fear, addiction or the coercion of others is another commonly iden-
tified source of autonomy failure. How significant such failures are likely to be depends
upon, among other things, the legal, economic and social forces that seek to impose
particular norms. Thus, quite apart from serious social censure, a woman like Charlotte
would have been likely to lose custody of her child and to have been thrown into eco-
nomic impoverishment, given that the norms of feminine socialisation were heavily
backed up with all kinds of legal and economic penalties for women who betrayed them.
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 229
What the discussion of failures of reflection, self worth, competence and action show
is that the prevalence and power of oppressive social norms is likely to have profound
effects on autonomous agency. But the way oppressive social norms contribute to the
diminishment of agency is a complex story. The existence of oppressive social norms
and expectations that single out certain options as undesirable, degraded or immoral
can cause failures of reflection  both about the options themselves, and about the
competence of an agent who pursues them. How profound these failures of reflection
are likely to be depends on the power of the norms and the degree to which they are
internalised. But oppressive social norms also work to more directly attack the agent
herself, casting doubt on her competence as an agent and further and more intensely
diminishing her autonomous agency. The likelihood of all these effects is greatly en-
hanced among those who suffer general failures of competence because the skills and
capacities for autonomous agency have never developed properly in the first place.
Despite the different ways that oppressive social norms can work to undermine
agency, such norms are characterised by the tendency to influence beliefs and beha-
viour in ways that are inimical to autonomous agency. In all cases it seems that a
person s potential for genuinely autonomous agency is valued less (if at all) to her
being a certain way or doing certain things. Such a person is not respected or admired
for making her own genuinely free choices about her life; she is not respected for who
she is or for what she might like to make of herself. The only way she is likely to
achieve a degree of regard from others is by living the life that they believe is appro-
priate for someone like her. Oppressive social norms usually provide only a very narrow
range of options that are likely to win such social regard and recognition.
IV
I want to argue that an appreciation of the way in which the norms and expectations
of others can undermine autonomous agency suggests the possibility that genetic engin-
eering may also have a detrimental impact on such agency. Whilst I don t wish to
make the strong claim that genetic engineering of the type we are concerned with is
akin to the systematic oppression of women, I do want to more modestly claim that in
structurally similar ways to cases like Charlotte s, the norms and practices of genetic
engineering may undermine the genetically engineered s autonomous agency.
To begin with, the norms and attitudes revealed by genetic engineering convey
implicit, and sometimes explicit views about the worthiness of certain choices and
decisions. The willingness to resort to genetic manipulation to rid an embryo of gay
genes reveals the attitude that certain attributes and behaviours are so undesirable as to
warrant genetic intervention to try and prevent them emerging in the adult person.
Genetically engineering someone to have non-gay genes sends a clear message to that
person regarding his parent s attitudes to the worth of homosexuality. It sends a clear
message that certain possible choices and decisions by that person in his adult life will
be deplored, or at least regretted. In so far as such attitudes are revealed by the practice
of genetic engineering, then it is plausible to suggest that independent agency is some-
what discouraged. It is vividly clear from cases like Charlotte s that our willingness and
ability to exercise independently reflective choices can be significantly affected by the
norms propagated by others. By so strongly revealing attitudes about the worth of
© Society for Applied Philosophy, 2003
230 Linda Barclay
certain options, genetic engineering increases the likelihood of failures of reflection. It
distorts the formation of desires and values as well as the reflective process itself, such
that critical scrutiny of one s desires and values is likely to be of little help in detecting
or preventing such failures of reflection.
That it is one s parents that reveal these attitudes is particularly likely to exacerbate
such failures of reflection. It is our parents who have the first and strongest impact on
our emerging desires and values, both because it is they who transmit our earliest
experiences of norms and expectations, but also because it is their recognition and
praise we most need, particularly at a younger age. Children are very likely to absorb
parental norms and expectations regarding the worth of homosexuality or homosexual
behaviour, with all the attendant effects this can have on their capacity to exercise
autonomy-conferring reflection.
It could be argued that such failures of reflection need not be particularly pro-
nounced if they are counteracted by the expression of norms and expectations that do
not represent homosexuality in a negative light. It might be said that in their behaviour
and explicit dealings with the child, parents who have used genetic manipulation could
make it quite clear that homosexuality is not deplorable. They could even encourage
their child to reflect carefully about his sexuality and make it clear to him that they
will support whatever he chooses. But how are we to make sense of this possibility? If
parents genuinely do not believe that homosexuality is deplorable, if they think that
first and foremost it is important that their child autonomously reflects on his sexuality
and chooses what is best for him, why would they have resorted to genetic engineering
in the first place? Even if it is conceivable that there is an answer to this question, it is
unlikely to be one that the child of such parents could fully appreciate. What he knows
is that his parents had his genes manipulated to try to prevent any genetic homosexual
influence. The message in that seems pretty clear.
It is important to consider how pronounced negative effects on agency would be if
the genetic manipulation of genes for sexual orientation became a widespread practice.
I specifically have in mind a scenario where many people seek to replace a gay gene
and where there are very few people (if any) choosing to replace a non-gay gene
with a gay one. When such practices are widespread, and endorsed both by law and
convention, they represent a graphic public expression of homophobic attitudes not
too dissimilar to some forms of sexist attitudes. In such a scenario not only will a given
individual genetically engineered to have non-gay genes suffer failures of reflection,
even those who have not been engineered are likely to do so as well.
Such wide-scale practices increase the likelihood that failures of autonomy will also
include failures of self worth and competence, in addition to failures of reflection. In
the previous section I argued that failures of self worth could occur just as a result of
the disparagement of certain options. Thus a person who turns out to have persistent
homosexual tendencies despite his professed desires and values may come to doubt his
competence as an agent. But I also suggested that attacks on the agent himself are
likely to contribute to failures of self worth. Widespread resort to genetic manipulation
can constitute such an attack as it encourages increasing acceptance of the view that by
the manipulation of genes we can profoundly influence people, perhaps even deter-
mine how they turn out, thus leaving individuals with little sense that it is their own
agency which is primarily responsible for their significant desires and values. Of course,
most geneticists will say it is a misconception that all genes are necessitating: most of
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 231
our genes will play no more of a necessitating role than do most forms of environmen-
tal influence [17]. But while it might be a mistake to think of our genes as necessitat-
ing, it doesn t follow that we won t. It is already easy to discern in common discourse
an assumption that genes have far greater determining power than do environmental
influences: indeed, the sheer amount of concern that the possibility of genetic engin-
eering raises for most people is testimony to this. Those who would want to use
genetic manipulation  and I m sure there would be thousands  presuppose that
manipulating someone s genes can be very efficacious in determining what he or she is
like. If they didn t think this, why would they want to do it?
It is quite possible that these kinds of beliefs will flourish just by our ever-increasing
knowledge of the human genome, with the further development of genetic engineering
having little impact [18]. Nonetheless, a society that endorses the resort to genetic
engineering either through law or custom, or both, is a society that potentially encour-
ages or at least feeds this view about the power of our genes. In any case, to make sense
of the idea that certain views about the power of genes will flourish, we do not have to
suppose that people will come to believe that genes alone are fully necessitating.
Genetic engineering gives us some control over a domain that we previously had little
control over. Gaining this kind of control, in conjunction with the kind of influence
and control we can already exert in non-genetic realms, may, jointly, convince each
individual that his own agency has little influence on how he turns out.
Any belief that genes, or genes in combination with other influences, really do have
this kind of power is likely to have a detrimental impact on a person s self worth as an
agent. What this scenario suggests is not only that agency is threatened by the attitudes
shown toward certain options, but that it is threatened by beliefs about the agent s
competence as an agent. In this case the attack on the agent s competence comes not
so much from norms and expectations regarding this particular agent, or class of
agents, but about shared norms and attitudes towards the possibility of agency itself. A
person who believes he has no capacity for real autonomy suffers a loss of self-worth
understood as the sense of competence to exercise capacities for agency. From the
perspective of the various theories of autonomy such a person may well have all the
powers needed to reflectively authorise his will: the problem is that he has become
dissociated from these capacities because he lacks a sense that such reflection and
revision will be ultimately efficacious in determining who he becomes. He believes he
is what he is and that what he is is not, and cannot be, a result of his own agency.
Thus, while a person with such beliefs about his genes can reflect just as you or I can,
he is dissociated from his reflective capacities in a way that we are not because of a
profound loss of self worth as an agent [19].
I suggested in the previous section that if one suffers this kind of loss of one s status
as an agent, then this contributes to further failures of reflection. A person who be-
lieves he is incompetent to authorise his will would almost certainly lose (or fail to
acquire) the actual ability to autonomously authorise his conduct. If a person believes
that his reflections on his will are inefficacious in determining who he is, then it is likely
that he will cease to use his reflective capacities well. A society in which these failures
of self worth are widespread is one in which actual autonomy competence is less likely
to develop in the first place, thus contributing to failures of competence. If the view
that our own autonomous agency is relatively impotent in determining who we are
becomes widespread, it is plausible to suppose that socialisation and education will be
© Society for Applied Philosophy, 2003
232 Linda Barclay
decreasingly likely to instil the reflective and psychological capacities required for
effective autonomous agency.
Genetic engineering of the type we are discussing, and the attitudes and norms it
reveals and encourages, can lead to failures of reflection, self worth, competence and
action and thus can contribute to the significant diminishment of autonomous agency.
Like the oppressive norms that Charlotte was exposed to, the norms and expectations
that are revealed in the resort to genetic manipulation of sexual orientation are more
concerned with ensuring that a person be a certain way than that he learn to exercise
his autonomous agency. Just like Charlotte, it seems that his potential for autonomous
agency is valued less than his being a certain way  in this case, not gay.
V
It should be clear from my argument that I think the problem with genetic engineering
is not specific to it. Norms and expectations that discourage the development of
autonomous agency can be revealed in various non-genetic ways of trying to influence
behaviour as well, as is obvious in Charlotte s case, which has nothing to do with
genetic engineering. That there is this similarity between genetic and non-genetic cases
is also revealed very clearly in the sexual orientation example. There can be no doubt
that the chances of failures of reflection, action and self worth are significantly raised
not only by legal punishment or discrimination but by the myriad of discriminatory
attitudes towards homosexuality betrayed by parents and society at large, all of which
takes place now, without the aid of genetic manipulation. Nothing in my argument is
meant to suggest that genetic engineering presents some wholly new and unique threat
to autonomous agency. I rather see it as potentially a new technique that will add to
the bad ways we already have of discouraging agency.
Acknowledging this similarity invites an important objection to my argument that
needs to be addressed. We express our views about the worth or otherwise of various
activities all the time, and, in so doing, affect people s perceptions and potential
agency. Why is genetic engineering bad just because it reveals and encourages certain
norms and expectations when much of our non-genetic behaviour does that already?
Moreover, we very often think it is a good thing to express such attitudes and to try and
influence peoples preferences and actions. For example, criminal and immoral beha-
viour are discouraged both by legal censure and by the norms that govern our interac-
tion with others. The cost of choosing such options is usually high, and therefore we
are steered away from them, hopefully in the formation of preferences and values, but
at the very least in our actions. There are very many choices we don t want people to
make and we discourage them from doing so. Thus the objection to my argument
raises the more general question of why the expression of norms and expectations is
sometimes acceptable and sometimes not, and this question can equally arise in situ-
ations that have nothing at all to do with genetic engineering.
One important response to this objection is to distinguish the different forms that the
expression of norms can take. Attention to the way in which something is discouraged
is crucial to understanding the difference between harmless socialisation and moral
education on the one hand, and oppressive, autonomy-undermining socialisation on
the other [20]. Ideal education and socialisation do not obstruct the agent s ability to
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 233
reflect on the worthiness or desirability of various options, even if some of those
options have been stigmatised by negative norms and expectations. So, for example,
while immoral or criminal conduct draw condemnation, ideal education and socialisation
also draw the agent s attention to the reasons for the condemnation and encourage the
agent to reflect on those reasons for herself. Ideal moral education, for example,
doesn t merely make assertions about right and wrong, backed up by censure and
punishment. It also, ideally, encourages agents to understand why certain things are
thought right and wrong, and encourages them to autonomously choose the right over
the wrong. In other words, an ideally moral person is not so just because of the way
she acts: she also has a good understanding of why it is worthwhile to act in a certain
way, and she acts in that way because of her understanding.
Much oppressive socialisation does not primarily strive to present the agent with the
best reasons that there are for being or behaving a certain way. Discouraging women
from painting and writing, or men from homosexual partnerships, does not typically
take the form of conscientiously enabling the agent to see the relevant reasons for her
or himself. To the contrary, a key part of much oppressive socialisation is to attack
the agent herself, such that victims of oppressive socialisation also typically suffer
failures of self worth and failures of competence. Women like Charlotte, recall, are not
only subject to failures of reflection due to the way oppressive norms shape their
preferences and values, they also suffer failures of self worth and competence because
of the way such norms cast doubt on their very capacity for agency and discourage its
development in the first place. This means that victims of oppressive socialisation are
much less likely to be able to challenge the stigmatisation of various options and are
thus much more likely to suffer failures of reflection. Ideal education, in contrast,
encourages the agent to see herself as increasingly responsible for her choices and
provides her with many opportunities for exercising and developing her actual com-
petence to make autonomous choices. Indeed, that she is considered a competent and
responsible agent is revealed precisely to the degree that we hold her responsible for
her conduct, both morally and legally, rather than, like Charlotte, declare her crazy.
This equips her with the self worth and the competence to consider for herself the
views  negative and positive  that she is exposed to, and thus minimises serious
failures of reflection.
One crucial response, then, to the question of why the expression of norms some-
times undermines agency and sometimes does not, is to draw attention to the very
different form the expression of such norms can take. In cases of oppressive socialisation,
but not in cases of harmless socialisation, the expression of norms and expectations is
likely to discourage agency not only because of the way it discourages certain options,
but because its attack on the agent s self worth and competence is likely to leave her
ill-equipped to thoroughly consider for herself those negative views she is exposed to.
In short, it is likely to further exacerbate significant failures of reflection and action.
If my argument about the possible effects of genetic engineering is plausible, then
genetic engineering bears more similarity to cases of oppressive socialisation than
harmless socialisation. I suggested that in resorting to genetic manipulation, parents
were very unlikely to encourage their child to conscientiously consider for himself
the value and desirability of different sexual choices. In short, it is unlikely that his
own reflective agency would be encouraged. I also argued that genetic engineering is
also likely to be associated with failures of self worth and competence because of its
© Society for Applied Philosophy, 2003
234 Linda Barclay
connection to beliefs about the impotence of agency, thus compounding failures of
reflection and action.
Attention to the form of the expression of norms therefore largely explains why
sometimes expressing norms and expectations is acceptable and sometimes not. I am
also inclined to think that in both the environmental and the genetic cases, objection to
the negative norms and expectations surrounding certain options is based on the belief
that those particular options do not in fact deserve to be so disparaged. If there is in
fact nothing at all wrong with homosexuality, and nothing at all wrong with women
pursuing artistic careers, why should these options be discouraged, even if the form the
discouragement takes is not necessarily much worse than relatively benign forms of
socialisation?
We have good reason to be extremely cautious about discouraging certain options,
either through environmental or genetic means. A healthy scepticism or at least open-
mindedness with respect to many of our values cautions us to be careful. Of course,
many things, for our own sake and for the sake of all of us, we do want to discourage.
But consider, even something like criminality we don t want to make extremely difficult
to choose: for the laws may be seriously unjust, in which case a willingness to choose
criminality might be a good thing. We can only exercise judgement here. But note that
there does seem to be at least one difference between the environmental and the
genetic cases. We simply cannot avoid, and in some cases don t want to avoid, creating
some environmental incentives and disincentives for choosing certain options. But
we certainly can avoid the use of genetic engineering, and, given the need to exercise
care with respect to the incentives or disincentives we attach to the choice of certain
options, we might think that our powers to exert non-genetic influence are more than
enough.
Philosophers who like to make much of the fact that genes may be no more deter-
mining than environment constantly overlook this combination effect. For example,
Jonathan Glover argues that genetic engineering can t be wrong just because it is a way
of influencing people, for we already influence people all the time through socialisa-
tion, education and the like [21]. He thus claims that one would have to be a genetic
determinist to suppose that genetic engineering is somehow worse than socialisation. But
this overlooks the possibility that we may already believe that some ways of influencing
people are indeed bad and that we may fear genetic engineering because it will be a
new way of exerting such undesirable influence. Relatedly, we may fear genetic engin-
eering simply because it adds to our power of being able to influence how people
are, and more power in this regard might be unnecessary and potentially dangerous.
Neither of these two fears rests on the particular error that Glover identifies.
VI
I have tried to focus on the relevance of the social environment for the development
and exercise of autonomous agency in order to suggest possible ways in which the
norms and expectations surrounding genetic engineering and even the mapping of the
human genome may contribute to the erosion of such agency. Nothing in my argument
suggests that these effects will necessarily follow from our ever-increasing knowledge
and potential control over our genes. What effects such knowledge and control will
© Society for Applied Philosophy, 2003
Genetic Engineering and Autonomous Agency 235
ultimately have, and how much we will be disturbed by any such effects, will depend
upon how widespread a practice genetic manipulation becomes, how such knowledge
and control interact with our beliefs about ourselves and our own agency, and on our
capacity to resist the opinions of others, among other things. If we do accept that the
norms and expectations salient in a given social environment can erode our abilities to
be autonomous agents then there is a plausible case to be made that some of the norms
surrounding the mapping and manipulation of the human genome could have a detri-
mental impact on autonomous agency.
Linda Barclay, Department of Philosophy, University of Aarhus, Aarhus C, 8000, Den-
mark. [e-mail]
NOTES
[1] I have nothing to say in this paper about the further question of whether or not genetic engineering
should be legally prohibited or controlled.
[2] See, for example, P. S. Greenspan (1993) Free will and the human genome project, Philosophy and
Public Affairs, 22,1.
[3] Nicolas Agar (1995) Designing babies: morally permissible ways to modify the human genome Bioethics,
9,1. p. 10.
[4] See, for example, G. Dworkin (1988) The Theory and Practice of Autonomy (Cambridge, Cambridge
University Press); H. Frankfurt (1971) Freedom of the will and the concept of a person, The Journal of
Philosophy, 68.
[5] See, for example, G. Watson (1975) Free agency, The Journal of Philosophy, 72.
[6] See, for example, S. Wolf (1990) Freedom Within Reason, (Oxford, Oxford University Press).
[7] Hierarchical accounts such as those proposed by Harry Frankfurt and Gerald Dworkin are proceduralist.
If a person s first-order desires are subject to second-order critical evaluation, then those second-order
volitions are hers freely, whatever they may be desires for.
[8] This criticism is made by, among others, M. Friedman (1986) Autonomy and the split-level self,
Southern Journal of Philosophy, 24.
[9] The Theory and Practice of Autonomy, ch.1.
[10] H. Frankfurt (1993) Identification and wholeheartedness in John Martin Fischer and Mark Ravizza
(eds.) Perspectives on Moral Responsibility (Ithaca, Cornell University Press).
[11] See note 6.
[12] P. Benson (2000) Feeling crazy: self-worth and the social character of responsibility in Catriona
Mackenzie and Natalie Stoljar (eds.) Relational Autonomy. Feminist Perspectives on Autonomy, Agency
and the Social Self (Oxford, Oxford University Press).
[13] See, for example, N. Stoljar (2000) Autonomy and the feminist intuition in Relational Autonomy.
Feminist Perspectives on Autonomy, Agency and the Social Self.
[14] P. Benson (1994) Free agency and self-worth, The Journal of Philosophy, 91. p. 657.
[15] Benson only wishes to argue that Charlotte could pass procedural tests for free agency. He asks us to
imagine that a woman like Charlotte is able to  commit herself to effective desires through higher-order
endorsement of them and that she need not be  deeply ambivalent at the same time that she thinks she
is crazy, and quite unable to use her powers of reflective endorsement sanely. As I ve just suggested, if
Charlotte thinks she is crazy, then she is likely to be deeply ambivalent and inconsistent, factors which
even the most undemanding proceduralist accounts identify as undermining agency. It seems to me that
the cases where one can suffer a failure of status as an agent without this causing other significant failures
of reflection are rare indeed. Benson is right to identify the crucial significance of self worth for free
agency, and he is also right to note that it has been all but ignored by most standard theories of free
agency. But the best analysis of its overwhelming significance is that in suffering a failure of self worth,
an agent is likely to suffer all kinds of failures of reflection, thus profoundly compounding the diminish-
ment of free agency.
[16] See in particular, D. T. Meyers (1989) Self, Society and Personal Choice (New York, Columbia Univer-
sity Press).
© Society for Applied Philosophy, 2003
236 Linda Barclay
[17] I suggested this at the beginning of Section II.
[18] John Campbell (La Trobe) brought this to my attention.
[19] If this is the way a genetically engineered person comes to think about himself, then a further blow to his
self-worth might be expected. Given that he fails to take responsibility for those attributes or traits which
are genetically engineered, believing, as he does, that they are not products of his agency, then they
cannot be a basis for self-respect. An important aspect of respecting a person is to respect those parts of
his character that arise from his own choices and decisions. Stephen Darwall calls this appraisal respect
in (1977) Two kinds of respect, Ethics, 88,1. pp. 36 49.
[20] Recall that the proceduralist accounts of autonomy seek to draw such a distinction.
[21] J. Glover (1984) What Sort of People Should There Be? Genetic Engineering, Brain Control and Their Impact
on Our Future World (USA, Penguin).
© Society for Applied Philosophy, 2003