Olga Cazalet v Walid Abu-Zalaf

This appeal concerns the approach the court should take when faced with an application by a petitioner to rescind a decree nisi under s.31F(6) Matrimonial and Family Proceedings Act 1984 (‘MFPA 1984’) and a cross-application by a respondent for the decree to be made absolute under s.9(2) Matrimonial Causes Act 1973 (‘MCA 1973’).

The appeal is brought by Olga Cazalet (‘the wife’) against the order of Mr Justice Mostyn (‘the judge’) dated 17 October 2022, whereby he dismissed her application to rescind a decree nisi which had been granted on 15 November 2013 and instead granted the application of Walid Abu-Zalaf (‘the husband’) to make the decree absolute. The wife challenges both the judge’s formulation of the test to be applied and his application of that test to the facts of the case.

By a Respondent’s Notice dated 16 December 2022 and amended on 16 March 2023, the husband seeks to uphold the judge’s decision on the basis of a different formulation of the test to be applied for rescinding a decree nisi or making a decree absolute.

The wife is 49 and the husband is 65. They have two biological children: G, aged 17, and A, now 8, who was born after decree nisi. A further child, J, aged 7, was adopted under Russian law by the wife in July 2015 some 20 months after the decree nisi. The husband agrees that J is a child of the family within the meaning of s.52 MCA 1973.

The parties began their relationship in July 2001 but did not marry until 1 June 2012. They entered into a prenuptial agreement on 30 May 2012. That agreement provided for the wife to receive increasing levels of financial provision upon divorce depending on the length of the marriage, which would be measured in full years from the date of the marriage ceremony to the date of separation.

The parties separated in August 2013. The wife filed a petition for divorce on 12 September 2013 on the grounds that the husband had behaved in such a way that she could not reasonably be expected to live with him and that the marriage had irretrievably broken down. The petition included allegations of physical abuse.

The petition was not defended by the husband. Accordingly, on 17 October 2013, the court certified pursuant to r.7.20(2)(a) of the Family Procedure Rules 2010 (‘FPR 2010’) that it was satisfied that the wife was entitled to a decree, on the basis that the husband had behaved in such a way that she could not reasonably be expected to live with him and that the marriage had irretrievably broken down. Decree nisi was pronounced on 15 November 2013. The decree stated on its face that the marriage should be dissolved unless ‘sufficient cause’ be shown to the court within six weeks.

The wife issued financial proceedings simultaneously with the divorce petition. The husband in turn issued an application for Notice to Show Cause, seeking to uphold the terms of the prenuptial agreement.

On 5 June 2014, the parties’ financial claims against one another were concluded on the basis of the prenuptial agreement. The financial remedy order therefore made provision for the wife on the basis of a marriage of under two years.

Neither the husband nor the wife thereafter sought to make the decree absolute. The wife says that the parties reconciled in or around November 2014 for a period which lasted until March 2020. The husband denies that there was a reconciliation and says that, whilst the relationship was ‘rekindled’, it was not a marital reconciliation but remained the same type of unhealthy relationship which had led to the granting of decree nisi.

Between 2014 and 2020 and particularly from 2017 onwards the husband and wife discussed the terms of a proposed postnuptial agreement, although no concluded agreement was ultimately reached between them.

The wife made an application on 21 November 2021 to rescind the decree nisi, dismiss the divorce petition and set aside the financial remedy order, her intention being to file a fresh divorce application if successful. This would lead the marriage to be treated for the purposes of the prenuptial agreement as lasting for eight years, thus significantly increasing the level of financial provision to which she would be entitled.

In response to the wife’s application, the husband on 14 January 2022 filed an application for the decree nisi to be made absolute.

Before moving on to consider the judge’s judgment and his approach to the applications before him, it is helpful to consider that part of the law which is uncontentious and provides a useful starting point.

It should first be noted that the issues arising in this appeal concern the provisions of the MCA 1973 before the reforms made by the Divorce, Dissolution and Separation Act 2020 (‘DDSA 2020’), which reforms include the introduction of ‘no-fault divorce’ with effect from 5 April 2022. As such, all references to the MCA 1973 are to its provisions in its previous form.

The divorce petition was issued by the wife pursuant to s.1(1) MCA 1973 ‘on the ground that the marriage has broken down irretrievably’. In order to establish that the marriage had broken down irretrievably, the wife relied on the fact found in s.1(2)(b) MCA 1973, namely that ‘the respondent has behaved in such a way that the petitioner cannot reasonably be expected to live with the respondent’. As provided by s.1(4) MCA 1973, it is this fact which establishes that the marriage has irretrievably broken down unless the court ‘is satisfied on all the evidence that the marriage has not broken down irretrievably’.

In Owens v Owens [2018] UKSC 41, [2018] AC 899 (‘Owens’),Lord Wilson at [37] reminded family lawyers that the commonly used shorthand of ‘unreasonable behaviour’ as a descriptor of this fact is wrong and that ‘The subsection requires not that the behaviour should have been unreasonable, but that the expectation of continued life together should be unreasonable’.

In the context of this case therefore, when decree nisi was pronounced pursuant to s.1(4)-(5) MCA 1973 and r.7.20(2)(a) FPR 2010 on 15 November 2013, it was on the basis that the husband’s behaviour meant that the wife could no longer reasonably be expected to live with him.

Having obtained the decree nisi, the wife could, under r.7.32(1) FPR 2010, have given notice that she wanted the decree to be made absolute after six weeks. She not having done so, the husband, as the party against whom the decree had been obtained, could pursuant to s.9(2) MCA 1973 himself have made an application under r.7.33(2)(c) for the decree to be made absolute three months after those six weeks had passed. As related above, neither the husband nor the wife did either of the above.

On 5 June 2014, the judge made the final financial order. The terms of the order were not implemented. Although a financial remedies order is not enforceable until the granting of decree absolute (s.23(5) MCA 1973), that is not why it was not implemented; rather, it was because the husband continued in every respect to maintain the wife and children at a standard significantly above that provided under the terms of the order.

Although there had been ‘on and off’ negotiations between the parties in order to agree the terms of a postnuptial agreement, there matters stood for eight years from the granting of the decree nisi in November 2013 until 22 November 2021, when the wife made her application to rescind the decree nisi, which application was met with a cross-application by the husband on 14 January 2022 for the decree nisi to be made absolute.

The wife’s application to set aside the decree nisi was made pursuant to s.31F(6) MFPA 1984 which provides as follows:

“The family court has power to vary, suspend, rescind or revive any order made by it, including—

(a)

power to rescind an order and re-list the application on which it was made,

(b)

power to replace an order which for any reason appears to be invalid by another which the court has power to make, and

(c)

power to vary an order with effect from when it was originally made.”

In NP v TP (Divorce) [2022] EWFC 78, [2023] 1 FLR 270 (‘NP v TP’), Cobb J considered how a court should exercise its discretion under s.31F(6) MFPA 1984 by reference to his earlier decision in Re A and B (Rescission of Order: Change of Circumstances) [2021] EWFC 76, [2022] 1 FLR 1143 (‘Re A and B’). Cobb J summarised the ways in which, since its introduction, the power under that section, which on the face of it is extremely wide, had been circumscribed in various first instance decisions. He set out at [22] principles by which the court will determine whether to exercise its power to rescind as follows:

“i)

Litigants should not be permitted to have 'two bites at the cherry' by applying again before the same court in relation to the same matter; there is an important public policy in achieving finality of litigation;

ii)

It is equally important for the court not to subvert the role of the Court of Appeal; if the litigants assert that the trial judge was wrong, the route for them to follow is an appellate one;

iii)The first point of reference should be whether one of the 'traditional grounds' for proposed review has been established:

a)

Fraud, mistake, innocent (or otherwise) misstatement of the facts on which the original decision was made;

b)

Material non-disclosure;

c)A new event or material change of circumstances which invalidates the basis, or fundamental assumption, upon which the order was made;

d)

If the order contains undertakings;

e)

If the terms of the order remain executory.”

Cobb J concluded at [23] that although in Re A and B he had been considering s.31F(6) MFPA 1984 in the context of children proceedings, the principles adumbrated there are of equal application to an application in other forms of family proceedings. I agree.

The judge, however, disagreed with Cobb J’s approach. Between [17] and [31] of this otherwise short judgment, the judge conducted a detailed historical review of the court’s power to set aside a decree from ‘the very dawn of the era of secular divorce’ through to the passing of s.31F(6) MFPA 1984 before concluding at [32] that as the power stretches back ‘unbroken to the Divorce Rules of 1865’, the old authorities apply to an application where a party applies to set aside a decree.

It was this historical analysis, culminating in the 1964 Divisional Court decision (Sir Jocelyn Simon P and Scarman J) of Owen v Owen [1964] P 277 (‘Owen’) at 284,rather than the modern analysis found in Cobb J’s analysis in NP v TP, that the judge applied when considering an application under s.31F(6) MFPA 1984 and which he then imported into a global test applicable to both applications before him.

The passage in Owen upon which the judge relied for the second element of his test is at p284 where Scarman J (who gave the judgment of the court) referred to the court’s power to order a rehearing in the following terms:

“We think that today the justification for the existence of the court's power to order a rehearing is the public interest and that its exercise should be governed primarily by that consideration. The true nature of the public interest is, as Pilcher J. remarked in Tucker v. Tucker, to see that in matrimonial matters, where questions of status are involved, any order made by the court is made upon the true facts. Certainty is not within the power of the court to achieve; but it must be satisfied the court to achieve; but it must be satisfied that there are substantial grounds for the belief that a decree has been obtained contrary to the justice of the case before it takes the serious step of setting aside an order of the court obtained by due process of law.”

With respect to the judge’s undoubted academic knowledge and analysis, I do not agree that the test to be applied on an application to set aside a decree nisi under s.31F(6) MFPA 1984 is governed primarily, or substantially, by the public interest; moreover, I am clear for my own part that the test in respect of an application to rescind a decree nisi under s.31F(6), at least for the purposes of the circumstances with which this appeal is concerned, fits within the general test as set out by Cobb J in NP v TP.

In my view, the wife’s application falls comfortably within (iii)(c) of Cobb J’s principles at [22] of NP v TP: the decree nisi should be set aside if it has been established by reference to the alleged reconciliation that there has been ‘a new event or material change of circumstances which invalidates the basis, or fundamental assumption, upon which the order was made’. I would add that, in my view, the important public policy imperative of finality of litigation is woven into the principles set out in NP v TP.

Returning then to the usual procedure where decree nisi has been granted. In the ordinary course of events, the spouse who obtained the decree nisigives notice under r.7.32(1)(a) FPR 2010 that he or she wishes the decree nisi to be made absolute. The court then proceeds to make the order, if it is satisfied in relation to the various matters which are set out in r.7.32(2)(a)-(h). These include at (a) that no application for the recission of the decree nisi is pending and at (e) that no application to prevent the decree nisi being made absolute is pending.

Where the application is, however, made more than 12 months after the making of the decree nisi, the application for a decree absolute must, by r.7.32(3) FPR 2010, be accompanied ‘by an explanation in writing, stating’:

why the application has not been made earlier;

Rule 7.32(4) FPR 2010 empowers the court (a) to require the applicant to file an affidavit verifying the explanation or to verify the explanation with a statement of truth; and (b) to ‘make such order on the application as it thinks fit’.

Where, however, no application is made by the party who obtained the decree nisi, an application can be made by the respondent (here the husband) in the divorce proceedings under s.9(2) MCA 1973. Section 9 provides (in its form as enacted and applicable to these proceedings):

“Proceedings after decree nisi: general powers of court

(1)

Where a decree of divorce has been granted but not made absolute, then, without prejudice to section 8 above, any person (excluding a party to the proceedings other than the Queen's Proctor) may show cause why the decree should not be made absolute by reason of material facts not having been brought before the court; and in such a case the court may—

(a)

notwithstanding anything in section 1(5) above (but subject to sections 10(2) to (4) and 41 below) make the decree absolute; or

(b)

rescind the decree; or

(c)

require further inquiry; or

(d)

otherwise deal with the case as it thinks

fit.

(2)

Where a decree of divorce has been granted and no application for it to be made absolute has been made by the party to whom it was granted, then, at any time after the expiration of three months from the earliest date on which that party could have made such an application, the party against whom it was granted may make an application to the court, and on that application the court may exercise any of the powers mentioned in paragraphs (a) to (d) of subsection (1) above.”

Any application under s.9(2) MCA 1973 is made under r.7.33 FPR 2010 which states at r.7.33(1)(a) that:

“(1)

an application must be made

(a)

in matrimonial proceedings for the decree nisi to be made absolute; or

(b)

……..

Paragraph 2 says:

“(2)

The conditions referred to in paragraph (1) are—

(a)

the Queen’s Proctor gives notice to the court under rule 7.31(6)(a) and has not withdrawn that notice;

(b)

there are other circumstances which ought to be brought to the attention of the court before the application is granted; or

(c)

the application is made—

(i)

in matrimonial proceedings, by the spouse against whom the decree nisi was made; ….”

It is common ground that when considering an application under r.7.33(2)(c) FPR 2010 the court needs to be satisfied of the matters in r.7.32(2) (no application for recission etc of the decree nisi outstanding). In my judgment, it must equally be the case for the ‘other circumstances’ referred to at r.7.33(2)(b). It follows that as the application is made more than 12 months after the making of decree nisi, regardless of whether the court is concerned with r.7.33(2)(b), r.7.33(2)(c) or both, the court will refer back to the factors in r.7.32(3) when considering whether to grant the application for a decree absolute made by the party against whom the decree nisi was made.

In this case, therefore, regardless of whether an application is made by the wife to rescind the decree nisi or by the husband for the grant of a decree absolute, the court will be ‘seeking an explanation stating’ per r.7.32(3)(b) FPR 2010 ‘whether the applicant and respondent have lived together since the decree nisi… and if so, between what dates’.

For the purposes of r.7.32(3)(d) FPR 2010, the respondent gave birth to A after the decree nisi was granted on 13 November 2013. It should be noted, however, that the wife was already pregnant with A at the date of the original separation. Of more significance, on the facts of this case, is that J was not adopted by the wife until two years after decree nisi, but he was nonetheless wholly accepted by the husband as a child of the family and is maintained by him. I cannot agree with the judge at [57], who regarded it as ‘virtuous’ of the husband to have maintained J and to have treated him as a child of the family and went on to say that such maintenance and treatment did not lead him to conclude that ‘the resumed relationship constituted a functioning marital reconciliation’. Whilst such ‘virtue’, if it can be described as such, may have led the husband to maintain J even though he had no obligation to do so, in my judgment, that J became a child of the family jointly with the father’s two biological children is a significant piece of evidence when considering whether the five years for which the parties were in a relationship after the granting of the decree nisi was a ‘reconciliation’ or, as the judge put it, a ‘resumption’ of an unhealthy relationship.

Pausing to pull together the convoluted procedural threads, the position is this: the husband was entitled to make an application for decree absolute, the wife having failed to do so. The court in considering that application will take into account that the application is opposed and will need to consider the explanations required by r.7.32(3) FPR 2010 and if necessary to find as a fact whether the parties have ‘lived together’ since decree nisi and, if so, when. A judge will also need to consider any relevant ‘other circumstances’ when deciding how the discretion should be exercised. He or she may make such order as they think fit, which can include rescinding the decree nisi.

It follows that the court has a wide discretion under s.9(2) MCA 1973 as to whether to grant a decree absolute, but I agree with the judge at [40] that it needs to be a ‘structured form of discretion’. Both parties agree that the proper test in respect of an application under s.9(2) is that found in the case of Savage v Savage [1982] Fam 100, (1983) 4 FLR 126 (‘Savage’), a case where a wife applied for a decree absolute following a failed reconciliation which had taken place after decree nisi.

In Savage, as in the present case, the fact relied on was that in s.1(2)(b) MCA 1973, namely that ‘the respondent has behaved in such a way that the petitioner cannot reasonably be expected to live with the respondent’. Following the grant of a decree nisi in May 1977, the parties reconciled in July 1977 until February 1981, when the husband left the matrimonial home. The wife applied for the decree to be made absolute. This application was rejected and the decree was rescinded.

Mr Holman (later Holman J), acting on behalf of the Queen’s Proctor as amicus curiae, proposed that the test should be whether the inference drawn by the court originally from the facts that ‘the petitioner cannot reasonably be expected to live with the respondent’ was still justified in the light of subsequent events. Wood J accepted that general approach and said at 103G:

“I am quite satisfied that at the present time this marriage has irretrievably broken down and that the husband has behaved in such a way that his wife cannot reasonably be expected to live with him, but one of the main issues in the exercise of this discretionary jurisdiction is whether the original decree nisi was pronounced upon sound evidence and upon sound inferences to be drawn from such evidence. The final phrase of section 1(2)(b) is too often overlooked. It is an essential factor.”

Wood J explained that there had been a ‘reconciliation which had extended for a very long time, some three and a half years’. He went on at 104B:

“In looking at the period of cohabitation it was argued that the quality of the cohabitation should be examined in each case to see how long the reconciliation continued. I am not convinced that that is the correct approach in view of the wording of many parts of section 2 of the Matrimonial Causes Act 1973. It is also extremely difficult to assess such a test and although cohabitation will always be with the hope of reconciliation, it is the living together which is the period which must be examined, in my judgment. All the factors which I have mentioned above lead me to the inevitable conclusion that the inference originally drawn under the special procedure, that the wife could not reasonably be expected to live with the husband, was the wrong inference, looked at in the light of all the circumstances now known.

To approach the problem in this way is not to undermine attempts at reconciliation. There is the period of 12 months referred to in rule 65 of the Matrimonial Causes Rules 1977 to which I have already referred and the periods of time outlined in section 2 of the Act of 1973 are within that span, thereafter the court has a discretion. It is perhaps surprising that the substantive law does not direct that a decree nisi shall lapse after a given period - possibly two years. This might help to cement any reconciliation which had taken place within that period and to encourage finality where the condition of the marriage was in reality hopeless.”

The approach in Savage, in my judgment, dovetails with the approach set out by Cobb J in NP v TP at (iii)(c) of his principles at [22] in relation to the application of the wife to rescind the decree nisi pursuant to s.31F(6) MFPA 1984; that is to say that one of the first points of reference should be whether there has been a ‘new event or material change of circumstances which invalidates the basis, or fundamental assumption, upon which the order was made’.

The judge in this case summarised his approach to the discretionary exercise as follows at [43]. Route A (s.31F(6) MFPA 1984) is where a party applies to set aside a decree and be granted a rehearing (ordinarily the respondent but here the petitioner), and Route C (s.9(2) MCA 1973) is where a respondent applies to make a decree nisi absolute (Route B, unmentioned here, is where either party applies to rescind the decree by consent after reconciliation). He concluded, at [44]:

“In my judgment, however, there is (or should be) no substantive difference between the test under Route A and the test under Route C. It would be illogical and irrational if it were otherwise. Under each route, in a “structured” discretionary exercise, the court will need to be satisfied of the following:

i)

that material facts existed at the time of the making of decree nisi but which were not placed before the trial court (“Category 1 facts”), and/or that subsequent events occurred (“Category 2 facts”), which furnish the clear conclusion that the findings made, or inferences drawn, by the trial court when making decree nisi were not justified and therefore wrong; and

ii)

that the degree of error is such that to allow the decree to stand would be so contrary to the justice of the case that the serious step of setting aside an order made by due process of law is justified. 

Although the exercise is said to be discretionary it is more realistically to be regarded as evaluative. The evaluation of the materiality and weight of the new facts will drive the decision. It would be an error of law if a judge decided a rescission case by reference to factors outside this discipline.”

The judge therefore held that there should be a single approach which essentially combined the two tests as he had identified them, namely: (i) the Owen ‘contrary to justice’ test in relation to the wife’s s.31F(6) MFPA 1984application; and (ii) the Savage ‘is the evaluative exercise carried out upon the granting of decree nisi still valid in the light of subsequent events?’ test in relation to the husband’s s.9(2) MCA 1973 application.

Ms Carew Pole KC, on behalf of the wife, submitted that the court should divide the discretionary assessment into Lord Wilson’s three stage test as set out in Owens. In that case, Lord Wilson reviewed the judgments of the lower courts when consideration had been given to what the law requires a petitioner to establish when seeking a decree nisi based on the fact of the alleged unreasonable behaviour of their spouse. Having considered those cases, Lord Wilson concluded at [28] that the inquiry has three stages:

“…first (a), by reference to the allegations of behaviour in the petition, to determine what the respondent did or did not do; second (b), to assess the effect which the behaviour had upon this particular petitioner in the light of the latter’s personality and disposition and of all the circumstances in which it occurred; and third (c), to make an evaluation whether, as a result of the respondent’s behaviour and in the light of its effect on the petitioner, an expectation that the petitioner should continue to live with the respondent would be unreasonable.”

In my judgment, an application of the Owens three stage test would not be of assistance in applications such as these. Reverse-engineering the Owens test would only serve to add a layer of complexity to what should be a simple test, easily applied.

Mr Molyneux KC, on behalf of the husband, argued that Ms Carew Pole should not be permitted to argue her Owens point as it is an entirely new approach from that which was argued at first instance. Given that I do not think that it has any application to this appeal, it is therefore unnecessary for me to address the rival submissions in relation to the admissibility of the arguments in its favour.

Mr Molyneux submitted that the submission on behalf of the wife that the judge erred in adding a second ‘contrary to the interests of justice’ limb is misplaced, since the test is supported by a long line of authorities unchanged since the Divorce Reform Act 1969 and the judge’s ‘contrary to justice’ limb is simply no more than a reformulation of the dicta from those cases.

I do not agree with the judge that there is a second limb to the discretionary analysis, whereby the decree is only set aside if it would be so contrary to natural justice that the serious step of setting aside the order is justified. As I said at [28], in my judgment, the judge was in error in deciding that Owen rather than NP v TP was the proper approach to the case and therefore it was wrong to import that test as part of a universal test to be applied in disputes such as this.

In any event, it hardly needs saying that a decision to rescind a decree nisi is of the utmost importance and will not be taken lightly by a court. That that is the case does not necessarily mean that recission should only be permitted where not to do so would be ‘so contrary to the justice of the case’. That is a much higher threshold than that found in Savage. In any event, Owen was addressing an entirely different issue. There the wife wished to set aside a decree nisi in circumstances where she wished she had defended a divorce but had taken a deliberate decision not to defend the suit after having taken advice. The court, perhaps unsurprisingly, was not convinced that an injustice had been done or that it was reasonable to suppose that it had been done.

I agree with the judge that, regardless of whether an application is made under s.31F(6) MFPA 1984 (rescission of decree nisi) or s.9(2) MCA 1973 (granting of decree absolute), there should be no substantive difference between the manner in which the discretion is applied. The endorsement of Cobb J’s approach to applications to rescind ensures that that is the case and, in my judgment, no further gloss or second limb is required or appropriate.

The test to be applied to both applications is therefore simply this:

“Is the evaluative exercise carried out upon the granting of decree nisi which led to the conclusion that it was unreasonable to expect the applicant to live with the respondent still valid in the light of subsequent events?”

I have adopted the test as phrased in Savage, but the test applies to both elements of the decree nisi, namely the decision that the wife could not reasonably be expected to live with the husband and that the marriage had irretrievably broken down.

I should add that, in common with the approach of Wood J in Savage at 104B, I am firmly of the view that there should be no examination of the quality of the marriage when applying the test. For it to be otherwise would require the court to conduct an analysis of the nature of the marriage throughout the entire period both before and after the granting of the decree nisi. It would also risk importing personal judicial mores and standards into the decision-making process. As Wood J said, what should be examined is ‘whether the original decree nisi was pronounced upon sound evidence and upon sound inferences to be drawn from such evidence’. In my judgement in so far as the judge imported a qualitative assessment of the parties’ relationship as a means of determining whether there had or had not been a reconciliation, he was in error.

The judge started his analysis with his assessment of the parties. Having heard oral evidence from both he found at [46] that:

“The wife was by far the better witness. Her evidence was generally clear and given in reasonable tones. She generally answered questions directly. In contrast the quality of the evidence of the husband was poor. He was combative, evasive, rhetorical, strident and in some respects obviously untruthful. ”

Notwithstanding this clear finding by a very experienced, specialist judge, he went on to disregard the view he had formed of the witnesses in the witness box saying at [48]:

“However, this case is a good example of the perils of placing emphasis on the demeanour of a witness, or placing too great a reliance on a witness’s irrelevant lies or other low conduct, when finding facts or exercising a discretion. In my judgment, the demeanour of a witness when giving evidence is unlikely to be a reliable aid either to finding facts, or exercising a discretion on uncontested facts. It is not just that a dishonest witness may have a very persuasive demeanour - that is of course, the first trick in a conman’s repertoire. But the opposite side of the coin is equally problematic in that a truthful witness may unfortunately have a classically dishonest demeanour. It is obvious to me that over-reliance on the “quality” of the evidence of a witness, good or bad, can lead to facts being found, or discretion exercised, by reference to influences that are irrelevant.”

In support of this, the judge, in a footnote, referred the reader to a speech made by Lord Leggatt in 2022 where he had questioned whether demeanour is properly an aid to the assessment of witness credibility. In that lecture, Lord Leggatt was building upon his judgments in the commercial cases of Gestmin SGPS SA v Credit Suisse (UK) Ltd [2013] EWHC 3560 (Comm) (‘Gestmin’) at [15] to [22] and Blue v Ashley [2017] EWHC 1928 (Comm) at [68] to [69].

The judge did not, however, refer to the judgment of Floyd LJ in the case of Kogan v Nicholas Martin and Others [2019] EWCA Civ 1645, [2020] FSR 3 (‘Kogan’),in which, as a consequence of his understanding of Gestmin, the first instance judge had regarded Leggatt J’s observations as ‘an admonition that the best approach for a judge is to place little if any reliance at all on the witnesses' recollections of what was said in meetings and conversations and instead base factual findings on inferences drawn from documentary evidence and known or probable facts’. The judge in Koganconsidered that approach to be appropriate to apply to the case with which he was concerned. Mostyn J’s approach in the present case mirrors that of the judge in Kogan.

In finding that the judge in Kogan had wrongly and unfairly ignored Mrs Kogan’s evidence, this Court said:

“88….We start by recalling that the judge read Leggatt J's statements in Gestmin v Credit Suisse and Blue v Ashley as an "admonition" against placing any reliance at all on the recollections of witnesses. We consider that to have been a serious error in the present case for a number of reasons. First, as has very recently been noted by HHJ Gore QC in CBX v North West Anglia NHS Trust [2019] 7 WLUK 57, Gestmin is not to be taken as laying down any general principle for the assessment of evidence. It is one of a line of distinguished judicial observations that emphasise the fallibility of human memory and the need to assess witness evidence in its proper place alongside contemporaneous documentary evidence and evidence upon which undoubted or probable reliance can be placed. Earlier statements of this kind are discussed by Lord Bingham in his well-known essay The Judge as Juror: The Judicial Determination of Factual Issues (from The Business of Judging, Oxford 2000). But a proper awareness of the fallibility of memory does not relieve judges of the task of making findings of fact based upon all of the evidence. Heuristics or mental short cuts are no substitute for this essential judicial function. In particular, where a party's sworn evidence is disbelieved, the court must say why that is; it cannot simply ignore the evidence.

89.

Secondly, the judge in the present case did not remark that the observations in Gestmin were expressly addressed to commercial cases. For a paradigm example of such a case, in which a careful examination of the abundant documentation ought to have been at the heart of an inquiry into commercial fraud, see Simetra Global Assets Ltd & Anor v Ikon Finance Ltd & Ors [2019] EWCA Civ 1413 and the apposite remarks of Males LJ at paras. 48-49. Here, by contrast, the two parties were private individuals living together for much of the relevant time. That fact made it inherently improbable that details of all their interactions over the creation of the screenplay would be fully recorded in documents. Ms Kogan's case was that they were bouncing ideas off each other at speed, whereas Mr Martin regarded their interactions as his use of Ms Kogan as a sounding board. Which of these was, objectively, a correct description of their interaction was not likely to be resolved by documents alone, but was a fundamental issue which required to be resolved.”

(My emphasis)

There are echoes of Kogan in this case: in Kogan, Ms Kogan’s case was that ‘they were bouncing ideas off each other at speed’ and Mr Kogan’s case was that she was merely ‘a sounding board’. In this case, it was common ground that this was a volatile relationship marked by disagreements which featured offensive and unattractive behaviour both before and after decree nisi. Ultimately the question for the judge was whether this was, as the wife said, a lengthy reconciliation or, as the husband said, a rekindling that did not amount to a reconciliation.

In my judgement, the judge’s assessment of the parties’ credit was an important feature which should have fed into the judge’s determination, alongside objective findings of fact, of whether the parties had reconciled following the making of the decree nisi.

The judge found at [56]:

“My finding on the evidence is that this was always a highly defective marriage. The husband was rightly found to have behaved in such a way that the wife could not reasonably be expected to live with him. On 17 October 2013 (the date of the certificate, which was formalised on 15 November 2013 by pronouncement of decree nisi) the court rightly found that the marriage had irretrievably broken down.  The parties were drawn back together about 12 months after the making of the decree nisi, but it would be an abuse of language to describe their resumed relationship as a marital reconciliation. While they may have referred to each other, and to the world, as husband-and-wife there was no enjoyment of each other’s society, and no mutual comfort and assistance. They did not, in the words of Lord Stair, derive any solace or satisfaction from their relationship. The treatment by the husband of J as a child of the family, with the consequential acceptance of financial liability, is very virtuous, but does not, in my judgment, lead me to conclude that the resumed relationship constituted a functioning marital reconciliation.”

The judge concluded by saying that the evidence in the case came ‘nowhere near to demonstrating that the findings made on the making of the decree nisi were wrong, let alone so wrong that to allow the decree to stand would be demonstrably unjust’. The evidence, he said, shows that the parties had a ‘highly defective marriage which was rightly put out of its misery by the making of the decree nisi’.

The Grounds of Appeal can be summarised as follows:

The Respondent’s Notice filed on behalf of the husband said that the judge’s two-part test was correct, but if he was wrong then the correct alternative test should centre around the need for there to be ‘special circumstances’ justifying a refusal to grant decree absolute.

As already discussed, in my judgment, the three stage test in Owens has no part to play in the resolution of the present applications. Ground 2 is therefore dismissed.

Sometimes in a case a ground of appeal effectively falls away once the matter is aired orally. The Ground 4 public policy ground is one such ground, and Ms Carew Pole, whilst not specifically abandoning the ground, did not press it and I accordingly also dismiss Ground 4.

That then leaves the real issues raised in Grounds 1 and 3 of this appeal, namely:

In my judgment, the appeal must be allowed on both Grounds 1 and 3.

So far as Ground 1 is concerned:

Ground 3 relates to the judge’s application of Savage,namely whether, in the light of subsequent events, the evaluative exercise carried out at the time of the granting of the decree nisi, which had led to the conclusions that it was unreasonable to expect the wife to live with the husband and that the marriage had irretrievably broken down, remained valid. In order to apply the test, in summary, the judge had to hear evidence and decide whether or not there had been a reconciliation between the parties subsequent to the grant of decree nisi. It seems to be common ground that if the judge was wrong and should, on the facts, have held that there had been a reconciliation, then the proper course in this case would have been for him to have rescinded the decree nisi.

In my judgment, the judge’s evaluation was undermined by:

The judge also, in my judgment, preoccupied as he was by what he regarded as the abject quality of the relationship, failed to put into the equation his own factual findings which included (i) that the parties had resumed their relationship ‘as before’ which included their preferred way of living over many years namely of living substantially in two separate houses; (ii) that J was adopted two years after decree nisi and was thereafter treated by both parties as a child of the family; (iii) that neither party applied for the decree absolute for a period of eight years; (iv) that they had a sexual relationship and holidayed together; (v) that both to the world at large and to each other they referred to each other as husband and wife, attending many high profile social functions in those roles; (vi) that the financial remedy order was not put into effect by the husband, but rather he maintained the wife and the children of the family at a level far in excess of that to which she was entitled under the order; and (vii) that between 2017 and 2020 the parties were actively negotiating the terms of a postnuptial agreement designed to replace the pre-existing prenuptial agreement which had formed the basis of the financial remedy order.

The judge drew an adverse inference against the wife because she neither applied for a decree absolute before the reconciliation nor applied for the recission of the decree nisi during the reconciliation. In my judgment, he was wrong to do so. Either party could have rectified matters during the eight years they continued in a relationship after the decree nisi; neither chose to do so.

In my judgement, a proper analysis of the circumstances of this case should have led to the conclusion that there had been a reconciliation between November 2014 and March 2020.

Given that this Court is always reluctant to go behind a finding of fact made by a judge, Mr Molyneux understandably and appropriately submitted that this Court should hesitate before going behind the judge’s finding that there was no reconciliation. I agree. This is however one of those rare occasions when, in my judgment and with the greatest of respect to the judge, this Court can set aside that key finding. Properly directed, the evidence, in my view, led inexorably to the conclusion that there had been a reconciliation and that even if the judge was right to categorise the relationship as ‘toxic, damaging and unhealthy’ with none of what he regarded to be the ‘qualities of marriage’, it was nevertheless the manner in which this couple chose to ‘live together’ both before decree nisi and for over five years after they had reconciled.

Once there was a finding of a reconciliation, and for such a lengthy duration, then, on an application of the Savage test, the proper outcome would have been to hold that the conclusions reached at the granting of decree nisi that it was unreasonable to expect the wife to live with the husband and that the marriage had irretrievably broken down were no longer valid in the light of their subsequent reconciliation.

In these circumstances the appeal will be allowed, the decree nisi will be rescinded and the petition dismissed.

I agree with King LJ that the appeal should be allowed. I also agree with her reasons but I propose to explain in my own words why I agree with her proposed outcome.

As King LJ explains, the legal landscape with which this appeal is concerned no longer applies with the introduction of “no fault” divorce and the removal of the requirement to establish any facts by the DDSA 2020. The changes effected by that Act also reflect the extent to which the public interest in and concerns about the divorce process have changed. I would add, for the avoidance of doubt, that status is still a matter of public interest (see, for example, Akhter v Khan (Attorney General and others intervening) [2020] 2 WLR 1183, at [29]-[30] and Shahzad v Mazher [2021] 2 FLR 707, at [33]) but the extent of the change in the public interest in the divorce process can be seen from the current form of s.1(3) of the MCA 1973 which provides:

“(3)

The court dealing with an application [for a divorce] under subsection (1) must—

(a)

take the statement to be conclusive evidence that the marriage has broken down irretrievably, and

(b)

make a divorce order.”

I return to this below, but it is clear that what was said in Owen, as relied on by the judge, is very far removed from this changed landscape.

As King LJ explains, the only ground on which a divorce can be obtained is “that the marriage has broken down irretrievably”: s.1(1) MCA 1973. A court cannot “hold the marriage to have broken down irretrievably unless the petitioner satisfies the court of one or more of the following facts”, namely those set out in s.1(2). In the present case, the wife relied on s.1(2)(b). In addition, by s.1(4) MCA 1973:

“If the court is satisfied on the evidence of any such fact as is mentioned in subsection (2) above, then, unless it is satisfied on all the evidence that the marriage has not broken down irretrievably, it shall …. grant a decree of divorce”.

The “fact” proved does not have to be the, or even a, cause of the irretrievable breakdown as “they are separate requirements”: Buffery v Buffery [1988] 2 FLR 365 at 366, May LJ, and Owens, Lord Wilson at [27]. However, both elements have to be satisfied so that, even when a relevant fact is proved, a decree will not be granted if the court is satisfied that the marriage has not irretrievably broken down and vice-versa.

The special procedure, first introduced in 1973, reduced the determination of whether a petitioner was entitled to a decree in undefended cases to a paper exercise. The process became increasingly light touch but, nevertheless, as a matter of law, the court had to be satisfied that the petitioner was entitled to a decree. Accordingly, r.7.19(4) FPR 2010 required the application for a decree nisi to be accompanied by a statement confirming that the contents of the petition were true. Rule 7.20(2)(a) provided that “the court must .. if satisfied that the applicant is entitled to … a decree nisi … so certify”. A certificate of entitlement to decree nisi would be issued and the case would be listed in open court for the pronouncement of the decree. If the court was “not so satisfied” then, by r.7.20(2)(b), it was required to give further directions.

As is demonstrated by the decree nisi in the present case, the entitlement to a decree is based on the court holding both that the husband had behaved in such a way that the wife could not reasonably be expected to live with him and that the marriage had irretrievably broken down. Both elements appear in the decree.

A decree of divorce is one decree (s.1(5) MCA 1973) and it takes effect from the date on which it is made absolute. Until that date, the marriage remains legally in existence. The decree nisi is, as its title indicates, a conditional order (as it is now described). After six weeks, the petitioner can give notice under r.7.32(1) FPR 2010 “to the court that he or she wishes the decree nisi to be made absolute”. This, in the normal case, is an administrative exercise as provided by r.7.32(2): Manchanda v Manchanda [1995] 2 FLR 590, at 595.

However, the fact that the decree is initially only a decree nisi explains why additional information is required if the application for the decree to be made absolute is made more than 12 months after the date of the decree nisi. King LJ has set out the provisions of r.7.32(3) FPR 2010 (paragraph 31 above) and of r.7.32(4) FPR 2010 (paragraph 32 above). As referred to, r.7.32(4)(b) provides that the court may “make such order on the application as it thinks fit”.

It is clear from these provisions that the court is being provided with this information so that it can decide whether something has occurred since the decree nisi which might justify the court refusing to make the decree absolute and, indeed, deciding to set the decree nisi aside. I mention this in part because this is another route by which a decree nisi might be set aside in addition to those mentioned by the judge and in part because it demonstrates, as I have said, that the court may have to determine whether the statutory requirements for a decree remain satisfied.

The judge’s decision as to the legal approach he should take was significantly influenced by Owen. However, that case was considering whether a decree nisi should be set aside to enable a spouse who had previously chosen not to defend a divorce to do so. Further, there are a number of authorities, for example Nash v Nash [1968] P 597, which show that the court’s approach would depend on whether the spouse was unaware of the divorce; was aware but had chosen not to defend before changing their mind; or was aware and wanted to defend but had failed, through ignorance or lack of full advice, to take a step required to defend. Owen was, as I have said, concerned with the second scenario.

Accordingly, as King LJ has said, the context for the decision in Owen is a long way from the circumstances of the present case, which is concerned with the effect of events subsequent to the decree nisi and not with whether a respondent should be permitted to obtain a rehearing of a decree nisi and to defend proceedings when, initially, he or she had chosen not to do so. It is understandable that, in that situation, the court will be considering whether there are “substantial grounds” for concluding that the decree had been obtained “contrary to the justice of the case” so as to justify “the serious step of setting aside an order of the court obtained by due process of law”. I acknowledge, of course, as King LJ does, that refusing to make a decree absolute and setting aside a decree nisi are serious steps but, as always when applying what has been said in other cases, context is important.

I do not consider that the judge was right simply to take what was said in Owen and apply it to the circumstances of this case. In addition to the very different context, as also referred to above, the public interest in the divorce process no longer reflects the concerns expressed by Scarman J in Owen. Further, in any event, the judge’s formulation of the second part of his test does not follow what was said in Owen. Scarman J used the words “contrary to the justice of the case” not, as the judge put it, “so contrary to the justice of the case” (emphasis added).

Accordingly, I consider that the judge’s inclusion of sub-paragraph (ii) of the formulation of the test, at [44], and his other formulation, at [59], were flawed. I set them out here for ease of reference:

“… the court will need to be satisfied of the following:

i)

that material facts existed at the time of the making of decree nisi but which were not placed before the trial court (“Category 1 facts”), and/or that subsequent events occurred (“Category 2 facts”), which furnish the clear conclusion that the findings made, or inferences drawn, by the trial court when making decree nisi were not justified and therefore wrong; and

ii)

that the degree of error is such that to allow the decree to stand would be so contrary to the justice of the case that the serious step of setting aside an order made by due process of law is justified.”

And, at [59], as follows:

“For there to be a rescission of the decree under [an application to rescind the decree] the law requires not merely that it is proved that the original court’s findings were erroneous but that the making of the decree was contrary to the justice of the case justifying the serious step of setting aside an order made by due process of law. This requires the error to be of such a degree that it would be demonstrably unjust to allow the decree to stand.”

As referred to by King LJ, the decision in Savage was, by contrast, dealing with the situation with which this appeal is concerned. The petition had relied on s.1(2)(b) MCA 1973, namely that “the respondent has behaved in such a way that the petitioner cannot reasonably be expected to live with the respondent”. Following the grant of a decree nisi in May 1977, the parties reconciled in July 1977 until February 1981, when the husband left the matrimonial home. The wife applied for the decree to be made absolute. This application was rejected and the decree was rescinded.

It is a first instance decision but it has remained unquestioned for 40 years, and, indeed, was applied in Kim v Morris [2013] 2 FLR 1197. It was also decided after full argument had been advanced on behalf of the Queen’s Proctor by Mr Holman (as he then was), which argument Wood J accepted. There is, in my view, nothing to suggest that it does not accurately set out the approach which should be adopted by the court now. It also, as King LJ has said, reflects paragraph [22(iii)(c)] of the approach set out by Cobb J in NP v TP. It, therefore, provides an approach which can be applied in the circumstances of the present case whether the court is considering an application under s.31(F)(6) MFPA 1984 or under s.9(2) MCA 1973 or is applying what was r.7.32(4) FPR 2010 (and is now r.7.19(5)). Having a uniform approach is what the judge, in my view rightly, sought to achieve and, indeed, is what Mr Molyneux submitted should be achieved.

The issue in the present case, therefore, is whether the inferences, or conclusions, that it was unreasonable to expect the wife to live with the husband and that the marriage had irretrievably broken down have been shown to be wrong by subsequent events. As referred to above, this is not a decision which a court will lightly make. To be fair to the judge, the first part of his formulation, at [44(i)], included “that subsequent events occurred … which furnish the clear conclusion that the findings made, or inferences drawn, by the trial court when making decree nisi were not justified and therefore wrong”. However, his formulation, at [59], required “the error to be of such a degree that it would be demonstrably unjust to allow the decree to stand”. It may be that this is another reason why the judge’s evaluation was flawed.

I consider that it was flawed because, in agreement with King LJ, I consider that the judge was wrong to introduce a qualitative assessment into his analysis of the marriage, an approach which had been rejected in Savage. Putting it simply, the issue was whether this marriage continued such that it had been wrong to conclude in 2013 that it had irretrievably broken down (I focus on this element of the statutory requirements). If the marriage continued for a substantial period of time, it clearly would become increasingly likely that the conclusion that it had irretrievably broken down as at the date of the decree nisi in 2013 was wrong.

The judge’s clear conclusion was that the parties had resumed their relationship. Indeed, Mr Molyneux submitted during the course of the appeal hearing that the parties’ relationship “continued as it had before”. It was his case, as referred to by King LJ, that the relationship had to be “materially better”. I also do not agree. I repeat, the issue is whether this marriage had irretrievably broken down. The judge found, at [60], that the parties “resumed” their relationship which “endured until March 2020”. He decided, however, that this did not undermine the previous conclusions because, at [49], of “the dismal quality of their relationship from the moment of its formation”; at [57], it “was always a highly defective marriage”; and, at [60], the parties “resumed a toxic, damaging and unhealthy relationship which had none of the qualities of marriage”.

If the marriage, with all its flaws and defects, continued then self-evidently it had not irretrievably broken down. The judge so found in that he found that the parties had “resumed” their relationship in November 2014 and that it continued until March 2020. Indeed, one of the conundrums in the judge’s judgment is that, on his analysis, the parties’ relationship before the decree nisi was not a marriage because it did not have, what he considered to be, the necessary qualities for it to constitute a “functioning marriage”. Accordingly, it is clear to me that the judge’s ultimate conclusion should have been that the basis on which the decree nisi had been granted in 2013 had been shown, and I would say clearly shown, to be wrong and that it should be rescinded.

In conclusion, I agree with King LJ, that the appeal must be allowed, the decree nisi rescinded and the petition dismissed.

I agree with both judgments.