Article Text

Download PDFPDF

Saviour Siblings: reply to critics
  1. Michelle Taylor-Sands
  1. Correspondence to Dr Michelle Taylor-Sands, Melbourne Law School, University of Melbourne, Parkville, Melbourne 3010, Australia; m.taylor-sands{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A key objective of Saviour Siblings was to begin a conversation about how we should approach the welfare of the child in the context of selective reproduction. I am very grateful to the journal and my critics for the opportunity to continue this conversation. By viewing the welfare of the child to be born through a relational lens, I challenge the prevailing orthodoxy surrounding questions of harm, benefit and moral obligation. I will address the questions raised by Stephen Wilkinson and Colin Gavaghan on these issues shortly.

I would first like to deal with the primary question raised by Emily Jackson about whether my relational model differs in practice from the ‘significant harm’ approach in the UK. The overarching purpose of my book is to propose a normative framework for regulating decision-making in the ethically complex and emotionally fraught area of selective reproduction. While I agree with Jackson that protecting the child's ‘basic interests’ within my relational model provides a similar threshold level of protection to the UK ‘significant harm’ test,i this is not the focus of my book.

My relational model is designed to accommodate the ‘conflict, confluence and confusion of interests’ within families seeking selective reproductionii more effectively than one centred on the interests of the child to be born (even one that recognises the child's interests are influenced by family outcomes). The aim of my devolved and collaborative decision-making model is to actively support patients to make the best decisions for their own families and according to their own values in an ethically complex area. I introduce the notion of collective family interests less as a test to determine whether selecting a saviour sibling is acceptable than to provide a clear ethical foundation on which to base the decision-making process.

For decisions such as saviour sibling selection, which are primarily motivated by the interests of the parents and existing child, it is particularly important to openly acknowledge relevant individual and collective family interests to facilitate frank and honest deliberation. Although a saviour sibling may well benefit from a successful donation, a failed donation can be traumatic for a sibling donor. Given the uncertainty of outcome, it is difficult to defend saviour sibling selection based solely on the interests of the child to be born. Gavaghan points out that this uncertainty may also impact negatively on the family unit. However, the overall benefit to the family involves less speculation as the potential benefits to the ill child (a chance at survival) and the parents (a chance at saving their child's life) are far more tangible than those attributed to the saviour sibling (a sense of self-worth based on retrospective altruism and indirect psychological benefit). Furthermore, the relational model I propose attempts to ameliorate any risk to existing family members through implications counselling.

Ultimately, I agree with Jackson that a relational model may not necessarily produce a different decision from the current UK harm-based model. My goal, however, is to improve the process by promoting sensitive and ethically responsible decision-making in this area. Furthermore, by articulating a clear ethical framework that acknowledges interconnected interests and the need for compromise within families, a relational model is more responsive to the needs and values of different families and may facilitate selection for a broader range of purposes.

In his critique, Wilkinson scrutinises my key objections to the argument that it is ‘better to be born than not’ when evaluating risk of harm to the child to be born. My first objection is aimed at distinguishing between two categories of risk associated with selective reproduction: (1) the risk of harm associated with the child's identity (for which the non-identity problem clearly arises) and (2) the risk of harm associated with the selection process itself, which involves biopsy of an otherwise apparently healthy embryo to facilitate preimplantation genetic diagnosis (PGD). My objection is a preliminary qualification to the application of the non-identity principle and is perhaps a more modest undertaking than Wilkinson perceives it to be. My example of the deafness case is used merely to illustrate the first category of risk and is not intended as a comparator for saviour sibling cases in this context.

I argue that the second category of risk associated with the embryo biopsy process used to select a human leucocyte antigen (HLA)-compatible sibling (or a deaf child for that matter) cannot be dismissed using the non-identity problem in the same way as the first category of risk because embryo biopsy is not a condition precedent to the saviour sibling's existence. It is true that PGD increased the saviour sibling's chance of being born by ‘weeding out’ the three out of every four embryos created through in vitro fertilisation (IVF) that were not HLA-matched to the existing sibling. It is also arguable that, had the parents not used IVF, the saviour sibling may not have been conceived in the first place. However, this ‘chance of being born’ argument is conceptually different to the non-identity problem and does not deliver the same knockout punch. Unlike the child born in Parfit's 14-year-old girl case, the saviour sibling could always argue that he or she would rather have taken his or her chances in the conception lottery than to have been subjected to the potential harms associated with PGD.

In response to my second objection, Wilkinson seems to agree it is plausible that a child can suffer harm from being born (even if his or her life is worth living) but challenges my argument that a different harm threshold should apply to selective reproduction from that applied in wrongful life cases. The harm threshold is, however, being used for very different purposes in these two scenarios. In the context of wrongful life cases, it is a pragmatic response to the need to perform the ‘logically impossible’iii task of quantifying the harm caused by an impaired life by reference to the harm caused by non-existence for the purposes of awarding damages. In selective reproduction, the threshold is used to justify the creation of a child who may face a lifetime of suffering. Courts can perhaps be forgiven for making a tenuous evaluation of ‘life’ versus ‘non-existence’ in order to reach a decision in wrongful life cases that does not devalue the life of the child who already exists. In my mind, it is too great a stretch to extend this reasoning to cases where the child is not yet in existence to justify creating a life of suffering.

Once we accept that a child can suffer harm as a result of his or her creation, it is rational and responsible to agree on a point at which the harm would be so unacceptable that the child should not be brought into existence. I argue that this point should be reached well before the child's life is so full of suffering that it is ‘not worth living’. This logic avoids potentially morally repugnant outcomes without the need to invoke the additional and problematic principle of procreative beneficence.

Gavaghan challenges the basis for family obligations in Saviour Siblings, quoting an extract of my discussion of favouritism within families. I use the notion of favouritism primarily to highlight the limitations of liberal theory in explaining family relations. The ‘additional steps’ Gavaghan requires can be found in the preceding section where I argue that membership alone forms the basis of family duty.iv

Drawing on the work of relational feminist and communitarian ethicists, I argue that family obligations arise not out of choice but as a result of membership, which takes place as soon as a child is born (or arguably even earlier, at the point of conception). Viewing the moral self as a self-in-relation, I argue that family members have obligations to one another as a natural consequence of being bound together in an intimate relationship. These obligations, which emanate from the relationship itself, are grounded in debt arising out of the receipt of benefits and reciprocity created by mutual reliance. Additional obligations to the common good are essential for maintaining the ongoing collective enterprise of family.

Membership provides the source of family obligation and defines its nature. The nature and extent of compromise required within a particular family will vary according to the individual circumstances of that family, including the character and values of the family and the needs and capabilities of individual members within it.v In contrast to universal moral obligations grounded in liberal theory, family obligations are greater than minimal and include cooperation, care and support. I argue that the instrumental and inherent value of intimate family to human flourishing justifies compromising some interests of individual members (even before they become aware of such compromise) for the benefit of the family as a Of course, there are limits on how much a child should be expected to compromise and I recommend state intervention in selective reproduction to protect the child to be born from exploitation, abuse and neglect.vii

Finally, I agree with Gavaghan that the notion of family duty is unlikely to make the decision-making process for selecting saviour siblings straightforward. However, by openly acknowledging the complex interests and dynamics that exist within families, my aim is to create a regulatory framework that attempts to accommodate potentially conflicting interests and support parents who are faced with difficult decisions. In particular, I hope to remove the need to rationalise decisions on the basis that they are in the best interests of the child to be born when this is neither true nor necessary.


  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

  • i My threshold extends beyond harm to address Kantian concerns about commodification.

  • ii Taylor-Sands M. Saviour Siblings: A Relational Approach to the Welfare of the Child in Selective Reproduction. Routledge, 78, 2013. Quote taken from Robert A Crouch and Carl Elliott, ‘Moral Agency and the Family: The Case of Living Related Organ Transplantation’ (1999) 8 Cambridge Quarterly of Healthcare Ethics 275, 283.

  • iii Gleitman v Cosgrove, 227 A 2d 689 (NJ, 1967).

  • iv Taylor-Sands, above n 2, 83–7.

  • v Ibid 87, 90.

  • vi Ibid 82.

  • vii Ibid 103.

Linked Articles