(back to preceding text)

    38. Given this latter circumstance, intervention by the courts would be peculiarly inappropriate when the change being sought in the law raises issues such as the following.

    39. First, much uncertainty surrounds the circumstances in which gender reassignment should be recognised for the purposes of marriage. The present case concerns one individual and her particular condition and circumstances. Although some of the evidence produced is of a general nature, the evidence before the House is focused on the facts of this case. So were the arguments. In particular, Miss Scriven QC submitted that wherever the line marking the transition from one sex to the other is to be drawn, Mrs Bellinger is on the reassigned gender side of the line.

    40. I do not consider this would be a proper or, indeed, a responsible basis on which to change the law. Surgical intervention takes many forms and, for a variety of reasons, is undertaken by different people to different extents. For men it may mean castration or inversion of the penis to create a false vagina. For women it may mean a mastectomy, hysterectomy, or creation of a false penis by phalloplasty. There seems to be no 'standard' operation or recognised definition of the outcome of completed surgery. Today the case before the House concerns Mrs Bellinger. Tomorrow's case in the High Court will relate to a transsexual person who has been able to undergo a less extensive course of surgery. The following week will be the case of a transsexual person who has undergone hormonal treatment but who, for medical reasons, has not been able to undergo any surgery. Then there will be a transsexual person who is medically able to undergo all or part of the surgery but who does not wish to do so. By what criteria are cases such as these to be decided?

    41. But the problem is more fundamental than this. It is questionable whether the successful completion of some sort of surgical intervention should be an essential prerequisite to the recognition of gender reassignment. If it were, individuals may find themselves coerced into major surgical operations they otherwise would not have. But the aim of the surgery is to make the individual feel more comfortable with his or her body, not to 'turn a man into a woman' or vice versa. As one medical report has expressed it, a male to female transsexual person is no less a woman for not having had surgery, or any more a woman for having had it: see Secretary, Department of Social Security v SRA (1993) 118 ALR 467, 477.

    42. These are deep waters. Plainly, there must be some objective, publicly available criteria by which gender reassignment is to be assessed. If possible the criteria should be capable of being applied readily so as to produce a reasonably clear answer. Parties proposing to enter into a marriage relationship need to know whether their marriage will be valid. Other people need to know whether a marriage was valid. Marriage has legal consequences in many directions: for instance, housing and residential security of tenure, social security benefits, citizenship and immigration, taxation, pensions, inheritance, life insurance policies, criminal law (bigamy). There must be an adequate degree of certainty. Otherwise, as the majority of the Court of Appeal observed, the applicability of the law to an individual suffering from gender identity disorder would be in a state of complete confusion: see [2002] 2 WLR 411, 435, para 104.

    43. Your Lordships' House is not in a position to decide where the demarcation line could sensibly or reasonably be drawn. Where this line should be drawn is far from self-evident. The antipodean decisions of Attorney-General v Otahuhu Family Court [1995] 1 NZLR 603 and Re Kevin (validity of marriage of transsexual) [2001] Fam CA 1074 and App. EA 97/2001 have not identified any clear, persuasive principle in this regard. Nor has the dissenting judgment of Thorpe LJ in the present case. Nor has the decision of the European Court of Human Rights in Goodwin v United Kingdom (2002) 35 EHRR 18. Nor is there uniformity among the thirteen member states of the European Union which afford legal recognition to a transsexual person's acquired gender. The pre-conditions for recognition vary considerably.

    44. Further, the House is not in a position to give guidance on what other pre-conditions should be satisfied before legal recognition is given to a transsexual person's acquired gender. Some member states of the European Union insist on the applicant being single or on existing marriages being dissolved. Some insist on the applicant being sterile. Questions arise about the practical mechanisms and procedures for obtaining recognition of acquired gender, and about the problem of people who 'revert' to their original gender after a period in their new gender role.

    45. Secondly, the recognition of gender reassignment for the purposes of marriage is part of a wider problem which should be considered as a whole and not dealt with in a piecemeal fashion. There should be a clear, coherent policy. The decision regarding recognition of gender reassignment for the purpose of marriage cannot sensibly be made in isolation from a decision on the like problem in other areas where a distinction is drawn between people on the basis of gender. These areas include education, child care, occupational qualifications, criminal law (gender-specific offences), prison regulations, sport, the needs of decency, and birth certificates. Birth certificates, indeed, are one of the matters of most concern to transsexual people, because birth certificates are frequently required as proof of identity or age or place of birth. When, and in what circumstances, should these certificates be capable of being reissued in a revised form which does not disclose that the person has undergone gender reassignment?

    46. Thirdly, even in the context of marriage, the present question raises wider issues. Marriage is an institution, or relationship, deeply embedded in the religious and social culture of this country. It is deeply embedded as a relationship between two persons of the opposite sex. There was a time when the reproductive functions of male and female were regarded as the primary raison d'être of marriage. The Church of England Book of Common Prayer of 1662 declared that the first cause for which matrimony was ordained was the 'procreation of children'. For centuries this was proclaimed at innumerable marriage services. For a long time now the emphasis has been different. Variously expressed, there is much more emphasis now on the 'mutual society, help and comfort that the one ought to have of the other'.

    47. Against this background there are those who urge that the special relationship of marriage should not now be confined to persons of the opposite sex. It should be possible for persons of the same sex to marry. This, it is said, is the appropriate way to resolve problems such as those confronting Mrs Bellinger.

    48. It hardly needs saying that this approach would involve a fundamental change in the traditional concept of marriage. Here again, this raises a question which ought to be considered as part of an overall review of the most appropriate way to deal with the difficulties confronting transsexual people.

    49. For these reasons I would not make a declaration that the marriage celebrated between Mr and Mrs Bellinger in 1981 was valid. A change in the law as sought by Mrs Bellinger must be a matter for deliberation and decision by Parliament when the forthcoming Bill is introduced.

    50. Mrs Bellinger advanced a further, alternative claim for a declaration that in so far as section 11(c) of the Matrimonial Causes Act 1973 makes no provision for the recognition of gender reassignment it is incompatible with articles 8 and 12 of the Convention. Her claim is advanced on the footing that, although she and Mr Bellinger celebrated their marriage long before the Human Rights Act 1998 came into force, and although the Goodwin decision dealt with the human rights position as at the date of the judgment (July 2002), the non-recognition of their ability to marry continues to have adverse practical effects. The statute continues to prevent them marrying each other.

    51. Mr Sales advanced several arguments on why such a declaration should not be made. There is, he submitted, no present incompatibility between the statute and the Convention. The European Court of Human Rights, in its decision in Goodwin, envisaged that the government should have a reasonable period in which to amend domestic law on a principled and coherent basis. The court said it 'will be for the United Kingdom Government in due course to implement such measures as it considers appropriate to fulfil its obligations': see (2002) 35 EHRR 18, 33, paragraph 120 (emphasis added).

    52. I cannot accept this submission. It may be that, echoing the language of the European Court of Human Rights in Marckx v Belgium (1979) 2 EHRR 330, 353, para 58, the principle of legal certainty dispenses the United Kingdom government from re-opening legal acts or situations which antedate the judgment in Goodwin. But that is not the present case. In the present case section 11(c) of the Matrimonial Causes Act 1973 remains a continuing obstacle to Mr and Mrs Bellinger marrying each other.

    53. It may also be that there are circumstances where maintaining an offending law in operation for a reasonable period pending enactment of corrective legislation is justifiable. An individual may not then be able, during the transitional period, to complain that his rights have been violated. The admissibility decision of the court in Walden v Liechtenstein (App no 33916/96) is an example of this pragmatic approach to the practicalities of government. But the question now under consideration is different. It is more general. The question is whether non-recognition of gender reassignment for the purposes of marriage is compatible with articles 8 and 12. The answer to this question is clear: it is not compatible. The European Court of Human Rights so found in July 2002 in Goodwin, and the government has so accepted. What was held to be incompatible in July 2002 has not now, for the purposes of section 4, become compatible. The government's announcement of forthcoming legislation has not had that effect, nor could it. That would make no sense.

    54. Then Mr Sales submitted that a declaration of incompatibility would serve no useful purpose. A declaration of incompatibility triggers the ministerial powers to amend the offending legislation under the 'fast track' procedures set out in section 10 and Schedule 2 of the Human Rights Act 1998. But the minister's powers have already been triggered in the present case under section 10(1)(b), by reason of the decisions of the European Court of Human Rights in the Goodwin case and the associated case of I v United Kingdom (App no. 25680/94). Further, the government has already announced its intention to bring forward primary legislation on this subject. For this reason also, counsel submitted, making a declaration of incompatibility would serve no useful purpose.

    55. I am not persuaded by these submissions. If a provision of primary legislation is shown to be incompatible with a Convention right the court, in the exercise of its discretion, may make a declaration of incompatibility under section 4 of the Human Rights Act 1998. In exercising this discretion the court will have regard to all the circumstances. In the present case the government has not sought to question the decision of the European Court of Human Rights in Goodwin. Indeed, it is committed to giving effect to that decision. Nevertheless, when proceedings are already before the House, it is desirable that in a case of such sensitivity this House, as the court of final appeal in this country, should formally record that the present state of statute law is incompatible with the Convention. I would therefore make a declaration of incompatibility as sought. I would otherwise dismiss this appeal.

LORD HOPE OF CRAIGHEAD

My Lords,

    56. My noble and learned friend Lord Nicholls of Birkenhead has explained nature of the condition from which Mrs Bellinger has been suffering from as long as she can remember and the profound changes which she has undergone, both physically and socially, to give effect to her wish to live her life as a woman rather than as a man. Her courage and that of Mr Bellinger, who has supported her constantly throughout their marriage, deserve our respect and admiration. If there was a legitimate way of solving their problem and making the declaration which Mrs Bellinger seeks, I would of course wish to take it. But I agree with my noble and learned friend that the expressions "male" and "female" in section 11(c) of the Matrimonial Causes Act 1973 are not capable of being given the extended meaning that would be needed to accommodate her case, and that we have no option but to dismiss this appeal.

    57. The essence of the problem, as I see it, lies in the impossibility of changing completely the sex which individuals acquire when they are born. A great deal can be done to remove the physical features of the sex from which the transsexual wishes to escape and to reproduce those of the sex which he or she wishes to acquire. The body can be altered to produce all the characteristics that the individual needs to feel comfortable, and there are no steps that cannot be taken to adopt a way of life that will enable him or her to enter into a satisfactory and loving heterosexual relationship. But medical science is unable, in its present state, to complete the process. It cannot turn a man into a woman or turn a woman into a man. That is not what the treatment seeks to do after all, although it is described as gender reassignment surgery. It is not just that the chromosomes that are present at birth are incapable of being changed. The surgery, however extensive and elaborate, cannot supply all the equipment that would be needed for the patient to play the part which the sex to which he or she wishes to belong normally plays in having children. At best, what is provided is no more than an imitation of the more obvious parts of that equipment. Although it is often described as a sex change, the process is inevitably incomplete. A complete change of sex is, strictly speaking, unachievable.

    58. It is tempting to regard the fact that a complete sex change is unachievable as a mere technicality when this is compared with everything else that can be achieved in the case of post-operative transsexuals. But the law of marriage exists in order to define the circumstances in which the public status that follows from a valid marriage may be acquired. There is much to be said for the view that the words "male" and "female" should each be given a single, clear meaning that can be applied uniformly in all cases. That was achieved by the decision in Corbett v Corbett [1971] P 83, which pre-dated the re-enactment of section 1(c) of the Nullity of Marriage Act 1971 in section 11(c) of the 1973 Act. Any enlargement of the meaning of those words to accommodate the problems faced by transsexuals would raise questions of fact and degree which are avoided by the use of the words chosen by Parliament.

    59. I do not overlook the fact that Mrs Bellinger's consultant urologist, Michael Royle, declared in a letter dated 5 January 1999 that she underwent gender reassignment surgery on 21 February 1981 and that "she is physically female." But it seems to me that this is an incomplete statement of the facts. The wording of section 11(c) demands that they be subjected to a more rigorous assessment. In Secretary, Department of Social Security v SRA (1993) 118 ALR 467 it was held that the respondent, who was a pre-operative male to female transsexual, did not fall within the ordinary meaning of the word "female" as her anatomical sex and her psychological sex had not been harmonised. One of the medical reports referred to by Lockhart J in the Federal Court of Australia, at p 477, explained very clearly what the surgery seeks to achieve, and what it cannot do:

"Genetically, and anatomically she is a 'male', however, she dresses and behaves as a woman. She considers herself as a woman. It is not for me to decide what the court or the Department of Social Security chooses to consider someone - but I do not think of, and treat the respondent as a woman. The fact that she has not had surgery to me is irrelevant. The aim of the surgery is to make somebody feel more comfortable with their body, not to 'turn them into a woman'. The surgery does not supply the patient with a uterus, nor with ovaries. It is purely and simply an attempt to allow the person's body to approximate to how they feel within themselves."

    60. Lockhart J said in the SRA case at p 480 that the common understanding of the words "woman" and female" and the phrase "opposite sex", which were ordinary English words, was a question of fact and that the crucial question was whether different conclusions were reasonably possible as to whether the facts or circumstances fell within their ordinary meaning. In Re Kevin (validity of marriage of transsexual) [2001] FamCA 1074 Chisholm J held that the ordinary contemporary meaning of the word "man" according to its Australian usage included post-operative female to male transsexuals, and that no good reasons had been shown why the ordinary meaning of the word should not apply in the context of marriage law: para 327. He went on to say that there was no formulaic solution for determining the sex of an individual for this purpose, that all relevant factors had to be considered including the person's biological and physical characteristics at birth, the person's life experiences, the extent to which the person has functioned in society as a man or woman, any hormonal, surgical or other medical sex reassignment treatments the person has undergone and the consequences of such treatment and that it was clear from the Australian authorities that post-operative transsexuals will normally be members of their reassigned sex: paras 328-329. He held that a marriage which "Kevin" had entered into with "Jennifer" on 21 August 1999 was a valid marriage under Australian law.

    61. In Re Kevin (validity of marriage of transsexual) (unreported) Appeal No EA 97/2001, 21 February 2003, the Full Court of the Family Court of Australia, after a comprehensive review of the authorities including the decision of the Court of Appeal in this case (see [2002] 2 WLR 411), agreed with the approach of Chisholm J. The essence of that decision is to be found in the following paragraphs:

"110. The definition of 'marriage' is essentially connected with the term 'man'. In these circumstances, for the reasons stated by the trial judge as amplified by our reasons that appear subsequently, we take the view that the words 'marriage' and 'man' are not technical terms and should be given their ordinary contemporary meaning in the context of the Marriage Act.

111. In our view, it thus becomes a question of fact as to what the contemporary, everyday meanings of the words 'marriage' and 'man' are respectively.

112. It is then a question of law for this court to determine whether, on the facts found by the trial judge, it was open to him to reach the conclusion that he did, namely that at the relevant time, Kevin was a man and that the marriage was therefore valid. As it was in SRA (supra) so, too, it is that the answer to that question is 'at the heart of the present case' ."

    62. I need hardly say that I entirely agree with the Australian judges that the words "male" and "female" in section 11(c)of the 1973 Act, which is the provision with which we are faced in this case, are not technical terms and that they must be given their ordinary, everyday meaning in the English language. But no evidence was placed before us to suggest that in contemporary usage in this country, on whichever date one might wish to select - 23 May 1973 when the 1973 Act was enacted, 2 May 1981 when Mr and Mrs Bellinger entered into their marriage ceremony or the date of this judgment, these words can be taken to include post-operative transsexual persons. The definition of "male" in the New Shorter Oxford English Dictionary (1993) tells us that its primary meaning when used as an adjective is "of, pertaining to, or designating the sex which can beget offspring". No mention is made anywhere in the extended definition of the word of transsexual persons. The word "transsexual" is defined as "having the physical characteristics of one sex but a strong and persistent desire to belong to the other." I see no escape from the conclusion that these definitions, with which the decision in Corbett v Corbett [1971] P 83 and the views of the majority in the Court of Appeal in this case are consistent, are both complete and accurate. The fact is that the ordinary meaning of the word "male" is incapable, without more, of accommodating the transsexual person within its scope. The Australian cases show that a distinction has to be drawn, even according to the contemporary usage of the word in Australia, between pre-operative and post-operative transsexuals. Distinctions of that kind raise questions of fact and degree which are absent from the ordinary meaning of the word "male" in this country. Any attempt to enlarge its meaning would be bound to lead to difficulty, as there is no single agreed criterion by which it could be determined whether or not a transsexual was sufficiently "male" for the purpose of entering into a valid marriage ceremony.

    63. In Goodwin v United Kingdom (2002) 35 EHRR 18, 24, paras 82-83 the European Court of Human Rights noted that it remains the case, as the court held in Sheffield and Horsham v United Kingdom (1998) 27 EHRR 163, that a transsexual cannot acquire all the biological characteristics of the assigned sex. It went on to say that it was not apparent in the light of increasingly sophisticated surgery and hormonal techniques that the chromosomal element, which is the principal unchanging biological aspect of gender identity, must inevitably take on decisive significance for the purpose of legal attribution of gender identity for post-operative transsexuals. So it was not persuaded that the state of medical science or scientific knowledge provided any determining argument as regards the legal recognition of transsexuals on grounds of social and legal policy. But this approach is not at all inconsistent with the view which I would take of the facts. The question which the court was asking itself was not whether the applicant, who was of the male sex when she was born, was now female. Post-operative transsexuals were assumed to fall into a distinct category. The question was whether it was a breach of their Convention rights for legal recognition to be denied to their new sexual identity.

    64. Of course, it is not given to every man or every woman to have, or to want to have, children. But the ability to reproduce one's own kind lies at the heart of all creation, and the single characteristic which invariably distinguishes the adult male from the adult female throughout the animal kingdom is the part which each sex plays in the act of reproduction. When Parliament used the words "male" and female" in section 11(c) of the 1973 Act it must be taken to have used those words in the sense which they normally have when they are used to describe a person's sex, even though they are plainly capable of including men and women who happen to be infertile or are past the age of child bearing. I think that section 5(4)(e) of the Marriage (Scotland) Act 1977, which provides there is a legal impediment to a marriage in Scots law where the parties "are of the same sex", has to be read and understood in the same way. I do not see how, on the ordinary methods of interpretation, the words "male" and "female" in section 11(c) of the 1973 Act can be interpreted as including female to male and male to female transsexuals.

    65. What then are we to make, in this case, of the decision in Goodwin v United Kingdom (2002) 35 EHRR 18? If it could be said that the use of the words "male" and "female" in section 11(c) of the 1973 Act was ambiguous, it would have been possible to have regard to that decision in seeking to resolve the ambiguity. But, for the reasons which I have given, I do not think that there is any such ambiguity. Then there is section 3(1) of the Human Rights Act 1998, which places a duty on the courts to read and give effect to legislation in a way that is compatible with the Convention rights if it is possible to do so. But we are being asked in this case to make a declaration about the validity of a marriage ceremony which was entered into on 2 May 1981, and section 3(1) of the 1998 Act is not retrospective: R v Lambert [2001] UKHL 37, [2002] 2 AC 545; R v Kansal (No 2) [2001] UKHL 62; [2002] 2 AC 69; R v Lyons [2002] UKHL 44; [2002] 3 WLR 1562, 1580D, para 45 per Lord Hoffmann and 1586B-C, para 63 per Lord Hutton. The interpretative obligation which section 3(1) provides is not available.

    66. But I do not think that it would be right to leave the issue there. If, as I would hold, the 1981 ceremony cannot be held to be a valid marriage ceremony, that is not an end of the matter. It would be open to Mrs Bellinger to try again some other day. It must be emphasised that this is not what she wants to do, as she regards herself as having been happily married since 1981. But we have been asked to say whether the provisions of section 11(c) are incompatible with her Convention rights and, if we find that they are incompatible, to make a declaration of incompatibility. I agree that it is proper that we should undertake this exercise, although neither of these steps can have any effect on the validity or otherwise of the 1981 ceremony.

    67. We cannot proceed to the making of a declaration of incompatibility under section 4(2) of the Human Rights Act 1998 without examining the question which section 3(1) of the Act treats as the logically prior question, which is whether the legislation can be read and given effect in a way which is compatible with the Convention rights. As Lord Steyn put it in R v A (No 2) [2001] UKHL 25; [2002] 1 AC 45, 68D-E, para 44, a declaration of incompatibility is a measure of last resort. But the word "must" which section 3(1) uses is qualified by the phrase "so far as it is possible to do so". As I said in R v Lambert [2002] 2 AC 545, 585B-D, para 79, the obligation, powerful though it is, is not to be performed without regard to its limitations. The obligation applies to the interpretation of legislation, which is the judges' function. It does not give them power to legislate: see also In re S (Minors) (Care Order: Implementation of Care Plan) [2002] UKHL 10; [2002] 2 WLR 720, 731B-E, paras 38-39, per Lord Nicholls of Birkenhead.

    68. If the only problem of interpretation had been one of timing, on the view that section 11(c) regards "male" and "female" as something that cannot be changed after birth whereas other provisions in the same section such as section 11(b) relate to the position at the time the marriage is entered into, I would have been prepared to read the words "at the time of the marriage" in to section 11(c) so as to give that provision a meaning which was compatible with the article 12 Convention right. If the only obstacle was that the parties' sex at the time when they were born had been assumed wrongly to be immutable, it could be overcome by disregarding the niceties of language and finding a compatible construction by reading these words in. But that would only have solved the problem for the future if it could indeed be said that Mrs Bellinger had completely changed her sex since birth and that she was now female. That, for the reasons I have sought to explain, is not a possible view of the facts.