Behavioural Genetics 157© Society for Applied Philosophy, 2006© Society for Applied Philosophy, 2006, Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 MainStreet, Malden, MA 02148, USA.Journal of Applied Philosophy, Vol. 23, No. 2, 2006Behavioural Genetics: Why Eugenic Selection is Preferableto EnhancementJULIAN SAVULESCU, MELANIE HEMSLEY, AINSLEY NEWSON ANDBENNETT FODDY Criminal behaviour is but one behavioural tendency for which a genetic influencehas been suggested. Whilst this research certainly raises difficult ethical questions and is subjectto scientific criticism, one recent research project suggests that for some families, criminaltendency might be predicted by genetics. In this paper, supposing this research is valid, weconsider whether intervening in the criminal tendency of future children is ethically justifiable.We argue that, if avoidance of harm is a paramount consideration, such an intervention isacceptable when genetic selection is employed instead of genetic enhancement. Moreover, othermoral problems in avoiding having children with a tendency to criminal behaviour, such asthe prospect of social discrimination, can also be overcome.I. IntroductionRecent discoveries in human behavioural genetics indicate putative associationsbetween specific genetic markers and a range of complex traits, including criminaltendency.1 Unlike many such projects dogged by retractions, one subset of criminaltendency research has produced compelling results: the correlation of mutations in themonoamine oxidase A gene and criminal behaviour within a Dutch criminal kindred.If this research proves valid, questions will inevitably arise about the moral accept-ability of couples using reproductive technology to avoid having a child with criminaltendency.In this paper, we employ this Dutch criminal kindred research to discuss the moral-ity of selecting against criminality in future children. After briefly tracing the historyand controversy of eugenics and behavioural genetics research, we summarise thetheory that biochemical pathways involving monoamine oxidase can influence a per-son’s chances of engaging in criminal behaviour at some point in their lives. We thenutilise this example to argue (drawing on Parfit’s non-identity problem) that choosingchildren without mutations in their monoamine oxidase A genes is acceptable, particu-larly if genetic selection technology is employed over genetic enhancement. This isbecause genetic selection is more immune than genetic enhancement to argumentsdepending on concepts of harm to the child.However genetic selection remains subject to other objections not met by the use ofthis technique, which could affect its ‘immunity’ to arguments based on harm to thechild. Therefore, we consider potential problems such as the child’s right to an openfuture, privacy and parental expectations. We also consider arguments derived fromsocial harms, such as diversity and discrimination.158 J. Savulescu, M. Hemsley, A. Newson and B. Foddy© Society for Applied Philosophy, 2006We conclude that if used in a controlled and appropriate way with due con-cern for possible outcomes, genetic selection to avoid criminal tendency is morallyjustifiable.II. Behavioural Genetics and EugenicsResearch into genetics and criminality is, of course, haunted by the spectre of eugenics.From the late 19th century until after the Second World War, this movement in Europeand North America aimed to enhance the genetic pool. Proponents of eugenics desiredto eradicate ‘genes for’ those complex behaviours deemed undesirable, such ascriminality, psychiatric disease and mental retardation. They sought to encourage thosejudged to have a superior genetic constitution to reproduce, whilst discouraged the‘genetically unfit’ from so doing — sometimes involving involuntary sterilisation.2In the United States, the first sterilisation law was passed in Indiana in 1907. Overthe next ten years, fifteen more states passed legislation which empowered them tosterilise ‘habitual or confirmed criminals, or persons guilty of some particular offence,like rape.’3 A statute in Iowa went so far as to require the sterilisation of ‘twice-convicted sexual offenders, of thrice-convicted other felons, and of anyone convictedjust once of involvement in white slavery’.4Following the human rights abuses of the Second World War and, eugenic practicesof this nature rapidly and justifiably fell from favour. Not only was this movementbased on questionable normative assumptions, it was bad science and exemplifiedcrude genetic determinism.5 Heredity is clearly not the sole causal determinant forhuman behavioural and mental traits, and state fair charts declaring that ‘unfit humantraits such as feeblemindedness, epilepsy, criminality, insanity, alcoholism, pauperismand many others run in families and are inherited in exactly the same way as colour inguinea pigs’6 were rightly abandoned.Following the discrediting of determinism, statutes sanctioning sterilisation of vari-ous groups (including criminals) were gradually repealed. However, attempts to estab-lish a link between biology and criminal tendency continued during the 1960s and1970s. For example, researchers aimed to establish a link between criminal tendencyand the XYY karyotype; and criminal tendency and testicular size. These studies againwere liable to criticism from an epistemological perspective, providing more shadymilestones in the history of behavioural genetic research.7Given the bleak history of behavioural genetics, public concern with researchinto criminal tendency is certainly understandable. It is likely that genetic influenceson behaviour, if they exist at all, are so complex that any research undertaken willalways be prone to a low level of accuracy.8 This tendency to low accuracy may evenlead to results derived from researcher bias (such as racism) rather than scientificrigour.9Yet despite public concern, research into genetic influences on criminal tendencieshas failed to decline. In fact, recent research suggests that a link between genetics andcriminality is not only possible, but likely. And with the recent completion of theHuman Genome Project, it is possible that more genes will be discovered to signific-antly influence our behaviour. In one family at least, this already appears to be thecase.Behavioural Genetics 159© Society for Applied Philosophy, 2006III. The Dutch