I. Introduction
When a court is faced with a dispute as to whether a marriage ceremony complied with the requirements laid down by law, or even as to whether any ceremony took place at all, what should its starting point be? The orthodox answer is that certain presumptions operate in favour of marriage: accordingly to Halsbury's Laws, “where there is evidence of a ceremony of marriage having been conducted, followed by the cohabitation of the parties, everything necessary for the validity of the marriage will be presumed, in the absence of decisive evidence to the contrary”, while “where a man and woman have cohabited for such a length of time, and in such circumstances, as to have acquired the reputation of being man and wife, a lawful marriage between them will be presumed”.Footnote 1 Yet both the academic literature and judicial decisions display considerable confusion as to what is being presumed, when and why.
In thinking about why the presumptions in favour of marriage arise, it is important to remember that they are classified in the law of evidence as rebuttable presumptions of law and thus “true” presumptions.Footnote 2 As Swadling reminds us, such presumptions arise “because the existence of the secondary fact is the most likely inference to draw from proof by evidence of the basic fact”.Footnote 3 If the secondary fact being inferred is that the parties have gone through a valid ceremony of marriage, then there must be good reason to believe that this must be the most likely interpretation of the evidence. Nonetheless, it is not always clear that what is being presumed is that the parties have gone through a valid ceremony of marriage at some point. Some convey the impression that the role of the presumptions is in fact to validate doubtful ceremonies.Footnote 4 Other accounts focus on the evidence of cohabitation and reputation and fail to stress that what is being presumed is that there was a ceremony, even if no direct evidence of it can be found.Footnote 5 There is also a lack of clarity as to when the presumptions arise, in particular whether the beliefs and self-presentation of the parties are relevantFootnote 6 and whether a lengthy period of cohabitation is necessary for the validity of a known ceremony to be presumed.Footnote 7
A key reason for this confusion about the operation of the presumptions is the lack of understanding of how they have developed. This is understandable: one account goes so far as to describe the case law as “an impenetrable morass”.Footnote 8 Some untypical cases have assumed undue prominence,Footnote 9 while other key cases have been misunderstood.Footnote 10 No one has attempted to locate the presumptions within the broader context of how the validity of a marriage was determined at different times, or of what evidence of a marriage might be available in different contexts. Indeed none of the cases cited in the current edition of Halsbury are drawn from the ecclesiastical courts, despite the fact that these courts had exclusive jurisdiction over the validity of marriages until 1858. Many of the cases that are cited are in fact ones in which the validity of the marriage was not in issue at all and whose underlying rationale is no longer relevant.
Without an understanding of when and why certain inferences were drawn from cohabitation and reputation at certain times and in certain contexts, we risk elevating these evidential presumptions to rules of substance. In this article I will accordingly show why eighteenth-century courts needed to rely on presumptions in the first place, and when they did so. I will then go on to demonstrate how the presumptions became divorced from their original evidential assumptions over the course of the nineteenth century, with a decontextualised understanding of their operation crystallising in Halsbury in the early twentieth century, and how the Court of Appeal in Chief Adjudication Officer v Bath Footnote 11 further blurred the line between cases in which there was a known ceremony and cases relying on cohabitation and reputation. Drawing these strands together, I will show how a contextualised understanding of the way in which the presumptions have operated in the past points the way to how, when and why they should operate today.
II. The Evidential Problems Raised by Clandestine Marriages, Elopements and Untraced Marriages
In Hayatleh v Modfy,Footnote 12 McFarlane L.J. referred to the jurisprudence relating to the presumptions as going back “at least to Wilkinson v Payne”.Footnote 13 This illustrates the influence of Halsbury on perceptions of the presumptions: Wilkinson, decided in 1791, is indeed the earliest case Halsbury cites, but it was by no means the earliest case in which a presumption of marriage was raised and it was entirely untypical of the case law of its own time. Taking it as a starting point thus provides a very skewed view of the presumptions.
Instead, we need to start with the very basic question of why cohabitation and reputation was relied upon as evidence of a marriage in the early case law. This requires an understanding of the challenges facing eighteenth-century litigants trying to prove that a marriage had taken place after any length of time had elapsed. Before 1754, entry into marriage was governed by the canon law. Its stipulations were for the most part merely directory, the only essential requirements being that the parties had freely consented to marry each other before an Anglican clergyman.Footnote 14 It was thus entirely plausible that a marriage might have taken place without any formal record being made. Community knowledge of the wedding could remedy the lack of formal documentation, but if the marriage in dispute had taken many years earlier there might well be no one with direct knowledge of it. In such cases, the fact that the couple had lived together and been reputed to be married might be the only available evidence that there had been an actual ceremony of marriage at some point. After all, the rarity and perceived sinfulness of cohabitation, together with the community's financial interest in ascertaining the marital status of those residing in the parish, meant that it was unlikely that couples who had not gone through a ceremony of marriage would set up home together, and even more unlikely that they would be treated as married by the community if they did.Footnote 15
Moreover, even written records of marriage were not necessarily reliable: the clergymen who operated out of the area surrounding the Fleet Prison in the early eighteenth century made a lucrative living by marrying couples without asking too many questions and were not above inserting or deleting marriages from their personal registers.Footnote 16 In such cases the subsequent cohabitation and reputation of the parties might well be more reliable evidence than the written record.
At the same time, there was little need for any presumption that the ceremony had been properly performed, since a marriage celebrated without banns or licence, at the wrong time of day, and not in any church would still be valid as long as the person solemnising it was authorised to do so. But after 25 March 1754, certain formalities became essential to the validity of any marriage (save those celebrated by special licence, according to the usages of Jews or Quakers, or involving members of the royal family). The plausibility of a couple having validly married other than in a church after the calling of banns or obtaining of a licence was therefore considerably reduced. The practice of marrying in the Fleet ceased, couples were much more likely to marry in the parish where at least one resided, and the quality of marriage registers was significantly improved. But registration was not essential to the validity of a marriage,Footnote 17 and so it was still possible for documentary proof of a valid marriage to be lacking. By contrast, the fact that certain formalities were now essential to the validity of a marriage meant that the courts were also faced with new questions as to precisely when a marriage should be upheld, particularly when young couples had eloped to evade the requirements of the 1753 Act.
Against this background, exactly how and when cohabitation and reputation were taken into account depended on the issue at stake. This is not to suggest any lack of coherence, but rather the importance of context. Within the ecclesiastical courts, whether the person bringing the suit was an irate parent trying to undo an unsuitable match, a party to the marriage attempting either to escape its bonds or to establish their claim in the event of the other's death, or a descendant needing to prove their legitimacy many years later, all inevitably affected both what evidence was available and the willingness of the courts to uphold the marriage. The secular courts, meanwhile, were precluded from deciding on the validity of a marriage or hearing direct evidence from the parties. Separating out these different strands of the early case law enables us to see more clearly what was variously being presumed when the case involved the parties themselves, when it occurred after the death of one or both, or when the validity of the marriage was only incidental to the dispute.
A. Challenges to the Validity of a Marriage while the Parties Were Alive
Even within the ecclesiastical courts, the import of cohabitation and reputation differed significantly depending on the action in question. Suits for jactitation of marriage were used by individuals to stop another person from falsely claiming to be married to them. Respondents could defend such a suit by showing that they had not claimed to be married, or that there had in fact been a marriage, or that the petitioner had represented that they had been married or authorised the respondent to do so. So even just holding out another person as one's spouse was sufficient to defeat a suit for jactitation.Footnote 18 The converse action was that of restitution of conjugal rights, whereby one spouse sought to enforce the duty of cohabitation. Respondents might seek to defend such actions on the basis that there had been no marriage at all. In either case, where the marriage was alleged to have taken place in the Fleet, the fact that the parties had lived together and been reputed to be married gave credibility to the claim that a valid ceremony had taken place.Footnote 19
Where the validity of the marriage was directly challenged by a suit of nullity on the basis of non-compliance with the formalities stipulated by the 1753 Act, how the courts interpreted the evidence depended very much on who was bringing the suit, and when. Where the suit was brought by a parent shortly after discovering their child's secret wedding, the courts did everything in their power to annul the marriage. Since advance parental consent was not necessary to the validity of a marriage by banns, the only way to challenge the validity of such a marriage was to show that the banns had not been properly called, and relatively minor mistakes assumed grave significance in this context.Footnote 20 Far from presuming the validity of these marriages, in Green, falsely called Dalton, v Dalton Footnote 21 the court went so far as to conclude against the bona fides of an addition to the bride's name in the absence of evidence to the contrary. The courts also took into account the broader public policy issues as to what should be necessary to the validity of a marriage,Footnote 22 justifying the need for the banns to be called in the correct names of the parties by reference to the public dimension of marriage and the need to notify the community of the intended union of these two persons.Footnote 23 Marriages were held to be void where this requirement had not been satisfied.Footnote 24
By contrast, where a nullity suit had been brought by one of the parties to the marriage, usually many years after the marriage had taken place, the courts were most reluctant to invalidate it and bastardise the children of the union.Footnote 25 In such cases weight was placed on the cohabitation of the parties and their mutual acknowledgement of each other as husband and wife not as evidence in favour of the validity of the marriage but rather as a reason for caution: the longer the relationship, the more reluctant the court would be to allow one of the spouses to negate its very existence by revealing an error of which both had been aware.Footnote 26 Ultimately, though, such reluctance could not override the facts of the case, and the time that had elapsed since the marriage could not justify upholding a marriage that was known not to have complied with the required formalities.Footnote 27 There were, therefore, a number of cases in which the courts were compelled to hold that the marriage was void regardless of the length of time that had elapsed since its celebration or the number of children that had been born.Footnote 28 There is little trace in this context of a general presumption in favour of marriage but rather a purposive approach that might require a more or less generous interpretation of what was required for a valid marriage depending on the context.
B. Questions about the Validity of a Marriage after the Death of One or Both Parties
The existence of a marriage might also be called into question after an individual had died if someone claimed to be their widow or widower and thus entitled to administration of the estate. The ecclesiastical courts expected any such claimant to provide at least some detail of where and when the marriage had taken place.Footnote 29 Evidence of cohabitation and reputation might need to be relied upon as supporting evidence where the marriage had been celebrated clandestinely, especially if the clergyman who had solemnised it had died or could no longer be found.Footnote 30 In an era of uncertain life expectancy this might be a problem even if a short time had elapsed since the marriage: in Plunkett v Sharp Footnote 31 the husband, parson and clerk had all died in the six years since the ceremony. The evidence of the one surviving witness was however supported by “cohabitation and a uniform owning by both parties, and their relations”,Footnote 32 and the widow's entitlement was duly established.
Proving a marriage after the death of both parties could be even more challenging, particularly when it was thought to have occurred many decades earlier or even overseas. In Lady Mayo v Brown,Footnote 33 the court recognised that it might be impossible for the daughter to discover where and when her parents had married, given they had lived in both London and Jamaica before her birth, while in Northey v Cook Footnote 34 it was deemed “absurd” to require any certificate of a marriage alleged to have taken place over 100 years earlier.Footnote 35 In such cases, the only evidence available was circumstantial. The public acknowledgement of the marriage during the parties’ lifetimes, the baptism of the children as legitimate, the surviving spouse being granted administration of the estate, and evidence as to declarations made by those “conversant in the families of the parties”Footnote 36 all raised a presumption that there had in fact been a valid marriage that those opposing the claim would have to rebut. (The validity of drawing inferences from such matters is, incidentally, fully vindicated by the fact that the marriages in question in Mayo and Northey can now be traced.Footnote 37)
C. Legal Outcomes Depending Incidentally on the Existence of the Marriage
The outcome of various actions in the secular courts might also depend on the existence of a marriage. In prosecutions for bigamy or actions for criminal conversation it was established that it was necessary to prove an actual marriage.Footnote 38 In other actions, if the validity of a marriage was crucial to a case, the issue had to be referred to the bishop for determination.Footnote 39 However, if it was incidental to the issue to be resolved, then a court could proceed on the basis that there was evidence of a marriage, without deciding on its validity.Footnote 40 An additional complication was that in civil proceedings neither the plaintiff nor the defendant, nor their respective spouses, were permitted to give evidence in their own behalf.Footnote 41 If the parties to the marriage could not give direct evidence of when and where they had married, and in the absence of admissible documentary evidence,Footnote 42 cohabitation and reputation would be the only available evidence that a marriage had taken place.
Exactly how convincing the evidence of such a marriage had to be depended on the issue at stake and who stood to benefit, and only a few examples can be given of the many different actions in which the existence of a marriage might arise. In ejectment cases, it was not necessary to prove the fact of the marriage: reputation was sufficient.Footnote 43 A husband's liability for his wife's debts could be established on relatively slight evidence: as one judge trenchantly noted, “[i]f a man chooses to hold out that a woman is his wife, it is his own fault”.Footnote 44 By contrast, if a man and a woman pleaded coverture to avoid a bond, the marriage had to be “well proved”.Footnote 45 Similarly, where a woman sought to escape a criminal conviction on the basis of marital coercion, the evidence of a single witness that she had lived with a man for two years and that they had passed as man and wife and been reputed to be married was deemed insufficient, especially since he admitted that he did not know when or where their marriage had taken place.Footnote 46 Slightly different considerations applied in settlement cases where the financial liability of contesting parishes was at stake: here the emphasis was on ascertaining the facts, even to the extent of allowing parents to give evidence bastardising their own children.Footnote 47
It is against this background that we should return to Wilkinson v Payne, the earliest case cited by Halsbury and referred to by McFarlane L.J. in Hayatleh v Modfy. The action was one for debt: Payne had promised to pay Wilkinson a sum of money if Wilkinson married Payne's daughter. The marriage ceremony took place and the couple duly set up home together. Sadly, the daughter then died and Payne defended the subsequent action on the promissory note by arguing that the marriage was not legal, Wilkinson having been underage at the time and no consent having been given by a parent or guardian. Since both of his parents were dead and no guardian had been appointed, this would have been rather difficult; as it was, the person whom his dying mother had asked to act as his guardian had consented. At the York Assizes, Grose J. viewed the father's claim with considerable disfavour, apparently commenting that it was “the hardest and most unreasonable defence that could possibly be made; because if it had been admitted, it would have made the children of that marriage all bastards, and the woman, who was his own daughter, a w___e”.Footnote 48
It was thus unsurprising that the jury should have found in favour of the plaintiff, and that the Court of King's Bench should subsequently have refused to grant a new trial on the basis that the verdict was consistent with “the justice, conscience and equity” of the case.Footnote 49 Buller J. doubted whether it was even necessary to prove a valid marriage in this case, suggesting that the fact of the marriage should suffice. Lord Kenyon, by contrast, suggested that it was possible that a subsequent valid ceremony had taken place, noting that the rule against a new trial being granted against the justice of the case was so strong that at least one instance was “bordering on the ridiculous”.Footnote 50
So Wilkinson was not directly concerned with the validity of the marriage, the judges were divided as to whether they were even presuming a valid marriage to have taken place, and it is the only case in which it was ever suggested that a second valid marriage might have taken place in the face of doubts about the validity of a known ceremony.
Why then, is it Wilkinson that is still cited? The reason, as the next section will show, is that over the course of the nineteenth century the case law of the ecclesiastical courts was effectively written out of the story and the fact that cohabitation and reputation had been relied on in different ways and for different purposes was overlooked.
III. Changing Ideas of Proof, Plausibility and Precedent
Again, we should start with the context. A number of factors should have reduced the need to rely on evidence of cohabitation and reputation as proof of a marriage over the course of the nineteenth century. Few nineteenth-century litigants needed to rely on a Fleet marriage,Footnote 51 record-keeping had improved in the wake of the 1753 Act, and a new centralised register of all marriages was introduced in 1837.Footnote 52 Elopements to Gretna Green posed new evidential challenges, but a “certificate” from the person officiating could be admissible as evidence of a declaration by the parties that they took one another as husband and wife.Footnote 53 After 1851, individuals were finally able to give evidence on their own behalf, and further legislation in 1853 made the spouses of parties both competent and compellable witnesses in civil proceedings.Footnote 54
While the growing number of cases involving marriages celebrated across the burgeoning British Empire had the potential to generate difficult questions, their validity was generally ensured by creative judicial decisions minimising what was needed to create a marriage.Footnote 55 Similarly, legislative changes meant that marriages celebrated in England and Wales were less susceptible to being declared void: after 1823 marriages would only be void if the parties “knowingly and wilfully” failed to comply with certain stipulated formalities.Footnote 56 In short, marriages should have been easier to prove and more difficult to challenge, thereby reducing the need to resort to any presumptions.
Where the validity of a marriage was directly in issue before the ecclesiastical courts (or, after 1857, the secular courts), there were indeed fewer cases invoking the presumptions. The courts continued to take a purposive approach to the legislation and a correspondingly strict approach to unsuitable marriages entered into without parental consent.Footnote 57 In such cases, no reference was made to any presumption in favour of validity. Yet in other contexts the presumptions in favour of marriage actually grew stronger, largely due to the particular circumstances – and a crucial omission from the report – of the 1849 decision of the House of Lords in Piers v Piers,Footnote 58 which therefore merits careful consideration.
A. Piers v Piers
The issue was whether a valid marriage had been celebrated in May 1815 between Sir John Bennett Piers and Elizabeth Denny, by then mother to five of his children and pregnant with a sixth. There was ample evidence that a ceremony had been conducted by an ordained clergyman at their home on the Isle of Man, and that they believed this to be a valid marriage. The only issue was whether a special licence had been granted, since legislation passed by the Tynwald in 1757 had invalidated all marriages celebrated other than in a church, unless by special licence.Footnote 59
Reading the version of the case that appears in the English Reports, it is easy to come away with the impression that the House of Lords presumed the grant of a special licence despite the evidence suggesting that no such licence had been granted.Footnote 60 The report gave considerable prominence to the evidence of the former Bishop of Sodor and Man that he had never granted any such licence and would not have done so had he been asked. What has been overlooked is that the contemporary report of Piers in The Times provides the telling additional detail that under cross-examination “it was established that the memory of his Lordship was very defective”.Footnote 61 He had, for example, forgotten holding a special court to permit one of his relatives to register his marriage, and his insistence that affidavits for licences were always taken before him personally was directly contradicted by the episcopal registrar. It was also clearly established that record-keeping had been poor at the time: an examination of the register of one local chapel revealed that of the 59 marriages noted as being by special licence, records of only two such licences could be found. In any case, there was a distinct possibility that the licence had been granted by his predecessor, since the marriage had been contemplated for some time.
On the basis of the evidence as a whole, then, it seems entirely plausible that a special licence had in fact been granted. But given the episcopal eminence of the witness whose evidence cast doubt on this, it is perhaps unsurprising that their Lordships should have wished to rest their decision on something more than the mere balance of probabilities. Their earlier decision in Morris v Davies Footnote 62 provided a useful precedent and justification. In that case, a child born to a married woman had been presumed to be the child of the husband on the basis that the evidence to repel the presumption of legitimacy must be “strong, distinct, satisfactory and conclusive”.Footnote 63 Applying this to the context of marriage in the Piers case, the Lord Chancellor, Lord Cottenham, echoed Morris in requiring the evidence to repel the presumption to be “strong, distinct, satisfactory and conclusive”.Footnote 64 Lord Brougham preferred “clear, distinct and satisfactory”,Footnote 65 having some qualms about stipulating that it be “conclusive”, but Lord Campbell felt that the presumption “can only be negatived by disproving every reasonable possibility”.Footnote 66
The significance of Piers lies both in what it decided and what it was subsequently thought to have decided. First, it very clearly raised the bar for rebutting the presumptions in favour of marriage. Second, the version in the English Reports implied that a marriage could be presumed to be valid in the face of evidence that the necessary formalities had not been completed, namely that the grant of a special licence had been presumed not because this was the most likely explanation of the evidence but because this is the only way of validating a marriage that would otherwise have been void. The impact of both of these developments was then magnified by the greater weight being attached to precedent and by the institutional changes that occurred over the second half of the nineteenth century.
B. The Proof Required to Rebut the Presumptions
In the wake of Piers in 1849, the presumptions began to be referred to in stronger terms. Where the issue before the court was the validity of a known ceremony, courts began to invoke the maxim omnia præsumuntur pro matrimonio or omnia rité acta præsumuntur to justify presuming that everything necessary for the validity of the marriage had taken place.Footnote 67 This represented a small but significant shift from the phrase semper præsumitur pro matrimonio – always presume in favour of a marriage – used in the earlier case law.Footnote 68 As the Earl of Selbourne noted in The Lauderdale Peerage,Footnote 69 it was the duty of the court “to presume… that whatever was necessary was actually done”.Footnote 70 This presumption was said to apply with “particular force” to cases involving marriage and legitimacy,Footnote 71 with counsel in Gompertz v Kensit Footnote 72 going so far as to suggest that “the presumption in favour of a valid marriage is the strongest known to the law”,Footnote 73 and Sir James Bacon V.C. agreeing that it was “very strong”.Footnote 74
Where the issue was whether a ceremony had taken place at all, the courts were initially more tentative. In the 1859 case of Goodman v Goodman Footnote 75 it was held that the evidence rebutting the presumption had to be at least as strong as the evidence raising the presumption. Two decades later, however, Piers was being cited to the effect that the presumption in favour of marriage “must prevail unless it is most satisfactorily repelled by the evidence in the cause appearing conclusive to those who have to decide the question”.Footnote 76 Having to show conclusively that there was no marriage at all, as opposed to showing that a known ceremony did not comply with the law, was a hard task and a significant departure from the earlier cases.
Piers was also relied upon in a number of Scottish appeals in which the question before the House of Lords was whether the parties had given the necessary consent to marry each other. In the Breadalbane Case,Footnote 77 Lord Cranworth declared in sweeping terms that “By the law of England, and, I presume, of all other Christian countries, where a man and woman have long lived together as man and wife, and have been so treated by their friends and neighbours, there is a primâ facie presumption that they really are and have been what they profess to be”.Footnote 78 Similarly, in De Thoren v AG Footnote 79 Lord Cairns L.C. specifically invoked Piers to establish that the presumption in favour of marriage was something “much stronger”Footnote 80 than other presumptions. The message as to the difficulty of rebutting the presumptions in favour of marriage was reinforced as these decisions were in turn cited alongside Piers in subsequent English cases.Footnote 81
C. The Plausibility of the Marriage Having Taken Place
Given that the official reporting of Piers also implied that a marriage could be presumed to be valid in the face of evidence that the necessary formalities had not been completed, it is unsurprising that the courts also began to accord less weight to the plausibility of the alleged marriage having occurred, and more to the desirability of it having taken place.Footnote 82 Thus in Collins v Bishop Footnote 83 the judge articulated: “the broad principle, so essential to the interests of society, that when two persons have lived together for many years as husband and wife, have been universally received as such, and had their children baptised, they should be considered as married people.”Footnote 84
Increased emphasis was placed on the hardship that would result if there were no marriage.Footnote 85 In this context, more weight began to be attached to the duration of the marriage, or at least to the time that had elapsed since it was supposed to have taken place, since the longer it had been assumed that a particular couple were husband and wife, the more hardship might be caused by a decision that they were not.Footnote 86
In such cases, the presumption was moving decisively away from its evidential moorings. The fact that cohabitation and reputation had been relied on in earlier cases because other evidence was unavailable seems to have been overlooked, with little consideration being given to the reasons why proper proof of the disputed marriage was not forthcoming. This was no doubt partly because many of the cases coming before the courts continued to involve marriages that were alleged to have taken place prior to the introduction of civil registration in 1837.Footnote 87 Yet this was not always the case: in Elliott v Totnes Union,Footnote 88 the child whose legitimacy was in question had been born just eight years earlier, in 1884, and the father swore that he was not married to the mother, but the court still relied on cohabitation and reputation to find in favour of the marriage. In this case the inference does not seem to have been warranted, as searches for the marriage have proved fruitless.
Nonetheless, judges did attempt to maintain a line between marital and non-marital relationships, reflecting the importance attached to reputation as well as cohabitation in raising the presumption. Where individuals had acted in a way that was regarded as incompatible with them being married, and where the wider family was unaware of any marriage, any reputation that did exist would be much weaker, and correspondingly easier to rebut.Footnote 89
D. Giving Weight to Precedents Independently of Their Context
The influence of Piers was further magnified by the significant institutional changes that took place in the second half of the nineteenth century. The ecclesiastical courts lost their jurisdiction over both marriage and the administration of estates, with these functions being transferred to new courts in 1858.Footnote 90 There was at the same time a conscious move on the part of other courts to distance themselves from the approach of the ecclesiastical courts and develop new doctrines.Footnote 91 The circumstances in which the ecclesiastical courts had relied on cohabitation and reputation were largely forgotten. Moreover, with the more fundamental reorganisation of the courts in the early 1870s, attempts were made to find commonalities between different doctrines. The fact that different courts had relied on cohabitation and reputation may have suggested a common doctrine, even if a closer examination of the context would have revealed the differences in application.
But this was not a period in which much attention was paid to context. The move to more formal and abstracted methods of legal reasoningFootnote 92 resulted in both a greater reliance on precedent and a reduced sensitivity to the contexts of earlier cases. One late nineteenth-century case illustrates all of these trends: in Lyle v Ellwood Footnote 93 reliance was placed upon a mix of Scottish and English authorities, with the latter emanating variously from disputes over inheritance, ejectment and jactitation of marriage. The simpler, abstracted rules derived from nineteenth-century cases were then relied on in turn to the exclusion of earlier cases.Footnote 94
E. In re Shephard and Halsbury's Laws
All of these developments – the high level of proof needed to rebut the presumption, the desirability of the marriage trumping the plausibility of it having taken place, and the focus on precedents regardless of their applicability – culminated in 1904 in In re Shephard.Footnote 95 It was claimed that the couple had travelled to France and married in a registry office there. Neither party was able to provide much corroborating evidence of this, as the husband had subsequently suffered an accident resulting in a loss of memory, while the wife had apparently been too excited at the time to pay much attention. Kekewich J. noted that the evidence that a marriage had actually taken place was “doubtful” and “romantic”.Footnote 96 He also explicitly assumed that “it was impossible according to the law of France”Footnote 97 to contract a valid marriage in the manner alleged. Nonetheless, he held that the children of the couple were entitled to share in the estate, on the basis that he would be “going against the authorities” if “after thirty years” he held otherwise simply because “the marriage has not been proved and cannot be proved”.Footnote 98
Despite Kekewich's assumption that he was simply following the existing case law, his willingness to presume the existence of a marriage on the basis of long cohabitation once it was accepted that the original ceremony was invalid went further than any of the preceding English cases. Unfortunately, he had the opportunity to entrench his interpretation in a number of subsequent decisions.Footnote 99 Even so, it might not have gained the prominence it did had it not coincided with a new initiative, Halsbury's Laws of England, first published in 1907. This completed the process of decontextualisation. Two presumptions were stated in Halsbury, one “from cohabitation without ceremony” and one “from cohabitation after ceremony”.Footnote 100 Shephard headed the list of supporting authorities, with the earliest authority cited being the atypical Wilkinson v Payne. No reference was made to the varied contexts in which evidence of cohabitation and reputation had been accepted: a context-specific approach to the evidence had effectively become two distinct presumptions. Even more surprisingly, none of the cases mentioned involved nullity suits – namely the only cases in which the validity of a marriage was directly in issue. Since such nullity cases continued to follow the approach established by the ecclesiastical courts rather than Piers,Footnote 101 this gave a highly misleading impression of when a marriage would be upheld. The overall effect was as if a mutant fly had been trapped in amber to be viewed for posterity as a definitive example of its species.
IV. Irregular Ceremonies and Long Cohabitation
Over the course of the twentieth century, judges almost invariably turned to Halsbury's Laws, and to the authorities there cited when invoking the presumptions.Footnote 102 Since it was decreasingly likely that a case would turn on the validity of a marriage alleged to have been celebrated before the advent of civil registration in 1837, there were very few cases in which it was necessary to presume a marriage on the basis of cohabitation and reputation alone,Footnote 103 and only one in which a marriage was presumed to have taken place in England without any direct evidence of a ceremony.Footnote 104 Changes to the position of married women also meant that many of the situations in which the validity of the marriage had to be determined as an incidental matter no longer arose.Footnote 105 The presumption in favour of the validity of a ceremony of marriage continued to be invoked in cases involving the legitimacy of children and, increasingly, where one spouse was seeking a divorce or decree of nullity in relation to a marriage that had taken place outside the jurisdiction.Footnote 106
Meanwhile, as England and Wales became more religiously and ethnically diverse, new questions emerged about the validity of marriage ceremonies conducted outside the statutory framework, and the assumptions made in In re Shephard and enshrined in Halsbury suddenly assumed crucial importance. In Chief Adjudication Officer v Bath Footnote 107 the question was whether Mrs. Bath was entitled to a widow's pension. Both parties had regarded themselves as married, the union had lasted 37 years, and Mr. Bath had paid tax and social security contributions on the basis that he was a married man. But the pension was refused on the basis that there was no evidence of a valid marriage ceremony. Her appeal was initially disallowed on the basis that the marriage had been celebrated in a Sikh temple that had not been registered for marriage, but a further appeal to the Social Security Commissioners was allowed on the basis that the marriage was validated by the common law presumption from long cohabitation.
The Court of Appeal then dismissed the appeal of the Department of Social Security. Evans L.J. correctly noted the absence of any provision rendering the marriage void on the basis of non-registration of the venue.Footnote 108 He also held that the evidence relied upon to show that the temple was not registered was not sufficiently positive to displace the presumption that it was, and held that in the absence of evidence that other formalities had not been observed, it had to be presumed that they had been.Footnote 109 Robert Walker L.J., on the other hand, referred to “manifold non-compliance”Footnote 110 and doubted whether the marriage could have been upheld had the parties separated immediately afterwards. Nonetheless he held that “[w]here there is an irregular ceremony which is followed by long cohabitation, it would be contrary to the general policy of the law to refuse to extend to the parties the benefit of a presumption which would apply to them if there were no evidence of any ceremony at all”.Footnote 111 He concluded, however, by agreeing with Evans L.J. that the evidence was not sufficient to rebut the presumption in favour of validity. The third judge, Schiemann L.J., expressed a preference for the reasoning of Robert Walker L.J.
The result was clearly correct on the basis that the non-registration of the temple would not have rendered the marriage void, and that no other defects in the marriage were alleged. But the reasoning employed in the case was nonetheless highly problematic,Footnote 112 and it is small wonder that most commentators have regarded the decision in Bath as upholding a marriage that was known not to comply with the formalities. As in Shephard, the key problem here is the blurring of the two different presumptions – that which presumes a marriage on the basis of long cohabitation and reputation; and that which presumes the validity of a known ceremony.
A. The Unnecessary Requirement of Long Cohabitation
The importance attached to the duration of the parties’ relationship in Bath has had one particularly unfortunate consequence, namely the assumption in subsequent cases that the presumption in favour of the validity of a known ceremony only arises if the couple have cohabited for a lengthy period after that ceremony.Footnote 113 While this has not yet had an impact on the outcome of any of the reported cases,Footnote 114 the illogicality of such a requirement needs to be pointed out.Footnote 115 A complete absence of cohabitation might raise the suspicion that the parties did not regard the ceremony as valid for some reason, but the duration of the parties’ subsequent cohabitation can have no logical bearing on the likelihood of whether or not the ceremony was valid. Of course, the duration of the relationship might make the courts more reluctant to find that the original ceremony did not comply with the law, as in the earlier decisions of the ecclesiastical courts. The lapse of time might also make it more difficult to establish non-compliance. But this is not the same as requiring there to have been a lengthy period of cohabitation.
Given the assumption in Bath that a marriage would have been presumed in the absence of any ceremony, it is worth noting that the courts have not been willing to presume a couple to be married in the face of positive, if slight, evidence that there was no ceremony at all. Martin v Myers Footnote 116 involved a relationship that was just as long as the one in Bath and a couple who had been thought to be married by all who knew them. But the fact that Mrs. Myers had told one of her daughters that she was not married, coupled with the absence of a marriage certificate, was held to rebut the presumption in that case.
B. The Uncertainty as to How to Approach Ceremonies that Are Known to Be Invalid
The requirement of long cohabitation was clearly linked to the assumption that the presumption could be used to validate irregular ceremonies. In the cases decided in the wake of Bath, there has been some ambiguity as to the appropriate approach where the only known ceremony demonstrably fails to comply with what is required for a valid marriage but where the parties have in fact cohabited for a lengthy period of time and been reputed to be husband and wife.
One approach has been to presume that a second, valid ceremony has taken place. In A.-M. v A.-M.,Footnote 117 where the original ceremony was both polygamous and made no attempt to comply with the Marriage Act, Hughes J. presumed that a second ceremony had taken place in a country permitting polygamy at a time when both parties were still domiciled in a country that permitted polygamy. Even though there was evidence that the parties had realised the original ceremony did not create a valid marriage, were contemplating how to redress this, and had had the opportunity to do so, this approach was treating the presumption as akin to a fiction.
A second approach has been that cohabitation and reputation raise a presumption that there was a marriage, but that this can swiftly be rebutted by proof that the ceremony was invalid.Footnote 118 And a third is that the presumption only arises where there was a ceremony that is prima facie capable of establishing a valid marriage. In a number of recent cases where the ceremony was decided to be a non-marriage – namely where it complied with none of the requirements for a valid marriageFootnote 119 – the courts have either explicitly held that there is “no room”Footnote 120 for the presumption to apply or made no reference to it at all.Footnote 121
Where then does this leave the presumptions in favour of marriage today? Despite the occasional calls for the wholesale abolition of marriage or its replacement by a range of alternatives, it remains the case that the law as a whole still values marriage and that those who are married have obligations to each other that those who are not married do not. In other words, there are still good policy reasons for the law to have a preference in favour of finding couples to be married, although it should be couched in terms that pay due respect to other family forms. But we need to appreciate that they now operate in a very different context.
V. Rethinking the Presumptions
Once we understand the way in which the presumptions have developed, it is clear that certain authorities should no longer carry any weight. The institutional divisions that prevented the common law courts from deciding on the validity of a marriage have long gone. Individuals are no longer barred from giving evidence in their own cause. The actions for jactitation of marriage and restitution of conjugal rights survived for longer, but were rare even before they were abolished almost 50 years ago, and there are no contexts in which proving the fact or reputation of marriage, as opposed to its validity, will suffice. In thinking about how the presumptions should operate, then, let us return to our starting point – that evidential presumptions are based on the most likely inference to be drawn from the evidence.
A. Is It Still Valid to Draw Inferences from Cohabitation and Reputation?
As the early case law shows, what was being presumed from cohabitation and reputation was that the parties had gone through a valid ceremony of marriage. The reason why this was presumed was both because of the lack of other evidence being available and because this was the most likely interpretation of the evidence. There were no cases in which an individual claiming to be married was able to plead cohabitation and reputation alone and throw the burden of proof on the other to show that there was no marriage.
Is it still legitimate to draw any inferences from cohabitation and reputation? The rise of cohabitation outside marriage over the past few decades means that no inferences about marital status can be drawn from living under the same roof. Neighbours and acquaintances may assume cohabiting couples to be married, or married couples to be cohabiting, depending on their own preconceptions as to what is most likely. Nor, at least outside those communities that still place considerable emphasis on marriage, are cohabiting couples likely to be treated any differently from married ones. At the same time, very few cohabitants now pretend to be married, and a couple's general reputation in the community was always less important than the evidence of family and friends, who would be more likely to know whether in fact a couple were married. Cohabitation and reputation together thus still provide good reason to presume that a couple have gone through a ceremony of marriage at some point. So, the presumption is still valid, as long as the basic facts that give rise to it are clearly understood.
However, the circumstances in which such a presumption would need to operate are today limited. After all, it would be bizarre if modern litigants were able to establish a marriage on slighter evidence than their eighteenth-century counterparts just because the context of earlier cases had been misunderstood. Anyone who has married in England and Wales should be able to produce a marriage certificate, as should their descendants. As long ago as 1867, Lord Cranworth explicitly commented that while cohabitation and reputation might raise a good prima facie case that there had been a marriage, if this was disputed, “inasmuch as in England all marriages are solemnized in public, and publicly recorded, it is reasonable to require the claimant to give positive evidence of its celebration, or else to explain why he is unable to do so”.Footnote 122
Individuals claiming to be married elsewhere should similarly be able, and so expected, to provide at least some evidence of how, when and where that ceremony took place. It is only really where the marriage was thought to have taken place in a jurisdiction without reliable registration, and the case is brought by the descendants of the couple whose marriage is in question, that the presumption might play a determinative role. Given the limited relevance to their legal rights and status of whether a child's parents were married, such cases will be few.
Lest it be thought that requiring positive evidence of a ceremony is a radical suggestion overturning long-established authorities, it is worth briefly reviewing the authorities cited in Halsbury in support of the statement that a marriage may be presumed where a couple “have cohabited for such a length of time, and in such circumstances, as to have acquired the reputation of being man and wife”.Footnote 123 The apparently authoritative list in fact includes a number of cases in which the issue was the validity of a known ceremony,Footnote 124 and several more where the marriage was celebrated in a jurisdiction that required no ceremony at all.Footnote 125 Other cases were ones in which it was only necessary to establish that the parties were reputed to be married.Footnote 126 The latest case cited in Halsbury in which a marriage was presumed to have taken place in England and Wales on the basis of cohabitation and reputation alone was Re Taylor in 1961. There, the court does not seem to have been particularly convinced that any marriage had in fact taken place, and searches of the electronic index of marriages for any record of it have proved fruitless, in telling contrast to the ease with which the eighteenth-century marriages were traced. In other words, limiting the presumption in the way suggested above would have little impact, and be more consistent with the earlier authorities and with the inferences that can be drawn from different forms of evidence – or their lack – today.
B. When Should the Law Presume in Favour of the Validity of a Known Ceremony?
It was well established prior to Bath that the presumption in favour of the validity of a marriage only arises if the ceremony was one that was capable of creating a valid marriage under English law. Otherwise, as Ormrod J. pointed out in Kalinowska v Kalinowski (Balletine intervening),Footnote 127 proof of any ceremony, whatever its nature, coupled with cohabitation, “would give rise to an almost irrebuttable presumption that there was a valid marriage”.Footnote 128
The point is that the presumption applies to all ceremonies of marriage that are prima facie conducted within the statutory framework. The most likely inference to be made in this situation is that all of the relevant formalities were observed, on the basis that couples would wish to do what was necessary to be validly married and most do comply with what is required. While the case law inevitably deals with those cases where some question as to the validity of the marriage has been raised, this is the context and not the rationale for the presumption. The presumption is that the ceremony was regular, and as a matter of logic it cannot arise at all where the ceremony is known to be irregular.
It is also clear that it is the fact of the ceremony that raises the presumption, not the subsequent conduct of the parties. Prior to Bath, long cohabitation was never a prerequisite for a presumption to arise in favour of the validity of a known ceremony,Footnote 129 and there is no reason to impose this additional requirement now.
On this basis, when should a marriage be presumed to be valid, and when will it be a non-marriage? If the parties have complied with none of the formalities required for a valid marriage, the result must inevitably be a non-marriage. Religious-only ceremonies outside any place of worship, along with ceremonies conducted by non-religious belief organisations, are thus not sufficient to raise the presumption of marriage. If it can be shown that the ceremony complied with at least some of the requirements for a valid marriage, then the presumption will apply and the onus is on the person challenging the marriage to show that the marriage was void. For marriages that took place overseas, it will need to be shown that the local law was not observed.Footnote 130 For marriages that took place in England and Wales, this requires it to be shown that both parties knowingly and wilfully failed to comply with the conditions set out in either section 25 or section 49 of the Marriage Act. If it cannot be shown that the formalities were not observed, or that the parties had the requisite knowledge that they were flouting the law, then the presumption will not have been rebutted and the marriage will be valid.Footnote 131
Two apparently long-standing principles to the contrary need to be dealt with. The first is that a second, valid ceremony can be presumed once the first has been shown to be invalid. The only support for this is Lord Kenyon's obiter suggestion in Wilkinson, which was subsequently relied upon in A.-M. Against this there is a huge body of case law in which marriages were found to be void without any hint that presuming a second marriage was a possibility. As Bodey J. noted in Al-Saedy v Musawi,Footnote 132 distinguishing A.-M. on the basis that there was no evidential foundation for a second marriage ceremony having taken place, to hold otherwise “would be tantamount to elevating a presumption born of common-sense into the status of a rule of substance, whereby long cohabitation plus a reputation of marriage would establish marriage, even when all the identified evidence showed that no valid or even void marriage ever took place”.Footnote 133
The second principle is the statement in Halsbury that the presumption “cannot be rebutted merely by showing that [the ceremony] could not be valid according to the law of the place of celebration”.Footnote 134 Two of the three authorities cited in support of this – Goldstone and Spivack – were in fact explicitly decided on the basis that the question was whether the marriage was valid according to the local law, and the third authority is Shephard. Against this weak corpus of authority needs to be set the considerable body of case law in which the presumption has been rebutted precisely by showing that the marriage was not valid.
Once a presumption does arise – whether on the basis of cohabitation and reputation or of a known ceremony – what standard of proof is necessary to rebut it? As I have shown, the elevated standard of proof in Piers may have been no more than a strategic response to the facts of that particular case. In any case, it was explicitly linked to the presumption in favour of legitimacy, and it is now well established that the normal civil standard applies in that context.Footnote 135 Similarly, the higher standard of proof once required in various contexts has now been replaced by the normal civil standard of probabilities, and it has long been established that the presumption in favour of the capacity of the parties to marry can be rebutted on the balance of probabilities. There is no good reason nowadays for holding that the evidence impugning the marriage must be so strong as to show beyond all reasonable doubt that the marriage was invalid, or that no ceremony ever took place.
VI. Conclusion
The evolution of the law in this area has seen the reliance on cohabitation and reputation in certain specific contexts morphing into two distinct presumptions in favour of marriage and then a further blurring of what is required to raise these presumptions. Disentangling its history illustrates the importance of not simply reading the history of a particular rule backwards from the present day, since the danger is that past cases will be interpreted through the lens of the present and their context will be lost. Reading the history of the presumptions forwards allows us to see more clearly when changes in the way they operated occurred, either by design or (more usually) by accident.
But the significance of understanding how the presumptions operated in the past lies in the implications for the present. My suggestion that the presumptions have a much more limited operation than most accounts suggest is in fact consistent with the results of the case law. Accounts of the presumptions thus need to be updated to reflect the case law as a whole, and the cases that are out of line should be removed or explained as such. The consequence of failing to do so is to hold out false promise to those who have lived together and been reputed to be married but have no chance of establishing that their marriage was valid under English law. The fact that there have been a number of recent cases where unsuccessful attempts were made to do so is an indictment of the confusion in this area.
While it might be desirable as a matter of policy to recognise long-standing relationships, whether the parties have gone through a valid ceremony of marriage or not, evidential presumptions are not the best mechanism to achieve this. After all, formalities should not be seen as mere matters of detail or as inconvenient hurdles.Footnote 136 They alert the parties to the significance of their act and protect them against inadvertently acquiring legal obligations. They reflect the law's policy on what makes a marriage and what checks are needed to ensure that the parties are eligible to marry and freely consent to do so. Their importance is recognised by international conventions, by Parliament and by the courts.Footnote 137 It should thus be recognised that the presumptions in favour of marriage are based upon the importance of these formal requirements, rather than overriding them. If the law were to presume – or pretend – that ceremonies conducted altogether outside the legal framework were in fact legally valid, there would be little point in having any system of formalities at all.
