England and Wales High Court (Queen's Bench Division) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
COMMERCIAL COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		Date: Thursday 3 May 2001

Lord Justice Rix :

1. Every day of the legal year courts of higher jurisdiction are reversing the judgments and orders of lower courts. In this case, a losing party in a lengthy, difficult and complex dispute, after a six month trial, and a reserved judgment running to 655 paragraphs, seeks reconsideration of that judgment, not on appeal, but by the trial judge himself on the ground that he is in error. The applicant prays in aid what has been called the Barrell jurisdiction (see In Re Barrell Enterprises [1973] 1 WLR 19) or what is submitted to be a still wider jurisdiction called the "reconsideration jurisdiction". In sum, the applicant asks the court to reverse its judgment so that a claim for $100 million under an alleged settlement agreement should be held to have succeeded rather than failed. The question is whether this application falls properly within the appropriate judicial exercise of such a jurisdiction.
2. I am the trial judge and the application has been made to me. For the purposes of this application I shall have to set the scene by saying something about the trial and the reserved judgment which I have written, but nothing I say by way of what is necessarily a compressed gloss of that judgment is intended to affect it or qualify it in any way.
3. In these proceedings in the Commercial Court in London, two actions have been brought by a Geneva company, which I shall call Noga (Compagnie Noga d'Importation et d'Exportation SA) against, among others, a group of individuals and companies who have throughout been known, after their original, but not current, solicitors, as the SJ Berwin defendants. Those defendants are essentially the estate, family and business associates of the late General Abacha, a former ruler of Nigeria, as well as certain of their companies. The subject matter of those actions was a transaction by which Nigeria bought back the debt instruments which it had guaranteed in respect of the construction of a steel plant at Ajaokuta in Nigeria. Noga claims to have interests in those instruments, the Ajaokuta bills. Subsequently, the Federal Government of Nigeria (the FGN) itself commenced a third action against the SJ Berwin defendants in respect of the same debt buy-back transaction, claiming that it was a corrupt transaction, engineered between General Abacha himself and his family interests, under which the SJ Berwin defendants had obtained a corrupt profit.
4. Subsequently the three parties, by which I intend to refer to Noga, the SJ Berwin defendants and the FGN, met in Abuja in Nigeria in July and August 1999 for the purpose of negotiations designed to settle their disputes. As a result, three agreements were made, the first involving all three parties, and the second and third involving the FGN and the SJ Berwin defendants alone. The first agreement has been called the tripartite agreement. It is dated 11 August 1999. The other two agreements are known after their dates as the 13 August agreement and the 16 August agreement. The 16 August agreement states that it supersedes the 13 August agreement.
5. Under the tripartite agreement both the FGN and Noga agreed to drop the claims in their London actions and the SJ Berwin defendants agreed to pay "a settlement sum" to Noga. Under the 13 August agreement, the SJ Berwin defendants agreed to pay DM 300 million to the FGN in return for the settlement of all claims that the FGN might have against them, not only in respect of the Ajaokuta transaction but also in respect of anything else. This was known as the "global waiver". The FGN does in fact advance against the SJ Berwin defendants a raft of other claims, known generally as the "looting claims", under which the complaint is made that General Abacha with the aid of the SJ Berwin defendants looted his own country and its treasury. Under the 16 August agreement, the 13 August agreement was superseded and the DM 300 million payment was restricted to the settlement of the FGN's Ajaokuta claim.
6. The tripartite agreement mentioned a "settlement sum" but did not identify it. In clause 2 it is said that "NOGA, having agreed to receive a settlement amount…agrees to withdraw its claims…" and in clause 3 it is said that the SJ Berwin defendants agree "to pay a settlement sum to NOGA in consideration of the above".
7. The parties soon fell again into dispute. Was the tripartite agreement a binding contract at all, or a mere agreement to agree? Had a settlement amount or sum ever been agreed between the SJ Berwin defendants and Noga? Noga said that $100 million had been agreed, but the SJ Berwin defendants said that nothing had been agreed. Was the 13 August agreement valid and effective? The FGN said that it had been orally stipulated to be subject to the approval of the President and the Vice President, which had been refused, and that in any event it had been induced by a fraudulent misrepresentation that the global waiver had been previously agreed with the Vice President, and that in consequence it had been validly avoided. Had the 13 August agreement in any event been superseded by the 16 August agreement, or was the latter agreement ineffective because of want of consideration?
8. On 30 September 1999 Mr Justice Moore-Bick gave directions for preliminary issues to be tried in the three actions as to whether "any claims in them have been settled and, if so, which of the claims have been settled and on what terms". The trial of these preliminary issues began in December 1999. It continued, with some interruptions, into May 2000. Judgment was reserved. The completion of the judgment was delayed for a while by an application by the SJ Berwin defendants to introduce further evidence. That application was resolved on 19 December 2000. The judgment was distributed in draft to the parties' counsel on 13 February 2001. In between the time of such distribution and the appointed time for its formal handing down in court on 26 February, Noga's solicitors gave notice to me by fax dated 19 February 2001 and by an application issued on the same day that Noga wished to apply to me to reconsider that section of my judgment where I had concluded that Noga had failed to prove that the tripartite agreement was binding. In particular a section of my judgment headed "Was the $100 million agreement contractually binding? ", running between paras 579 and 601, was said to have been in error. Para 601 concludes:

"Be all that as it may, it still remains the position that at the end of the day, it is Noga that bears the burden of proving a binding agreement with the SJ Berwin defendants through Mr Bagudu. I am prepared to accept that the figure of $100 million was agreed, but not unconditionally and not by way of a contractually binding contract to be found in the tri-partite agreement."

9. The Mr Bagudu there referred to is one of the SJ Berwin defendants. He is the close business associate of General Abacha's son, Mohammed Abacha, another of the SJ Berwin defendants. It was Mr Bagudu who had the burden of the Abuja negotiations on behalf of the SJ Berwin defendants.
10. Thus it was that Noga ultimately failed to win the case which it had been presenting at the trial of the preliminary issues, which was that there was a binding settlement agreement involving the payment to it by the SJ Berwin defendants of $100 million. Its defeat may be described as a narrow defeat. By far the largest single factual issue at the trial was whether there had been agreement on $100 million with Mr Bagudu on behalf of the SJ Berwin defendants. (Mr Bagudu denied that $100 million had even been discussed let alone agreed.) Noga won that issue, but I held that it had failed to prove that that agreement was unconditional.
11. In due course, when informed of my judgment, Noga authorised the application, foreshadowed by its lawyers, that I should reconsider my decision. That application has now been heard. For the purpose of making it, Mr Steven Gee QC and Ms Vasanti Selvaratnam on behalf of Noga have produced written submissions running to 388 paragraphs, and reply submissions of 182 paragraphs. Those submissions referred to some 70 authorities, encompassing authorities both on the reconsideration jurisdiction itself and on the underlying merits of the application, the substantial majority being of the latter category. I made it clear, in giving directions for the hearing of the application, that I had first to be satisfied of the jurisprudential basis of the reconsideration jurisdiction. Mr Gee felt that he could not do justice to his application or to my direction without explaining in detail how I had fallen into error. As a result of his and his junior's written submissions, and the court's pre-reading, he was able to complete his oral submissions in a little over a day. The hearing of the application as a whole was completed in two days. Compared to the length of the trial and the importance of the interests involved, and the range of argument deployed, that was not perhaps a disproportionate amount of time. But there is no doubting that the application has involved a substantial commitment of resources, and that greater delay to the process of completing the handing down of a final judgment has been restricted only by very considerable efforts having to be made by all involved, including court listing but in particular counsel, for which I am grateful.
12. It is common ground that until an order has been perfected, the court retains control over its judgment and its decision, and can permit argument to be reopened. As a result it may modify or even reverse a decision to which it has already come, and which it has communicated to the litigants. As Jenkins LJ said, giving the judgment of the court, in In re Harrison's Share under a Settlement [1955] Ch 260 at 276:

"We think that an order pronounced by the judge can always be withdrawn, or altered, or modified by him until it is drawn up, passed and entered. In the meantime it is provisionally effective…"

13. In that case a judge made an order in chambers approving a variation of trust and a few days later the House of Lords gave a decision which showed that the judge had lacked jurisdiction to do what he had done. He recalled his order before it had been drawn up, and the court of appeal said that he was entitled to do so. The same jurisdiction was considered in In re Barrell Enterprises [1973] 1 WLR 19, where it was sought to reopen an appeal on the ground of fresh evidence before the court of appeal's order had been entered. The court of appeal declined to permit the appeal to be reopened. Russell LJ gave the judgment of the court. He said (at 23/24):

"When oral judgments have been given, either in a court of first instance or on appeal, the successful party ought save in most exceptional circumstances to be able to assume that the judgment is a valid and effective one. The cases to which we were referred in which judgments in civil courts have been varied after delivery (apart from the correction of slips) were all cases in which some most unusual element was present…

"…In all these cases there were circumstances of a wholly exceptional character.

"It is clearly not permissible for a party to ask for a further hearing merely because he has thought of a possible ground of appeal that he originally overlooked. The discovery of fresh evidence has never been suggested as a ground for reopening the argument before the Court of Appeal. If fresh evidence comes to light, of such a character as to call for further consideration of the issues, the right way to deal with the situation is by applying for leave to appeal to the House of Lords: see Murphy v. Stone-Wallwork (Charlton) Ltd [1961] 1 WLR 1023; or, if such appeal be not available in a contempt case, by application for release."

14. Since the entry into force of the CPR regime, there is now the recent authority of Stewart v. Engel [2000] 1 WLR 2268, where the court of appeal had to consider whether what it described as the Barrell jurisdiction had survived the introduction of the CPR and if so in what terms. That was a case where the judge at first instance had reopened his decision, before the drawing up of his order, to strike out the claimant's case. After the claimant had changed her mind as to whether she wished to replead her case in another way the judge permitted her to do so and thus changed his own order and declined to dismiss the claim. The court of appeal held, by a majority, that the judge was wrong, as a matter of discretion, to reopen the matter. Sir Christopher Slade said this (at 2274E/H):

"I accept that it is possible that the Barrell jurisdiction falls to be regarded as a rule of practice rather than law and was capable of being abrogated by the introduction of the CPR. Nevertheless, I am satisfied that there is nothing in the CPR which obliges us to hold that it was so abrogated and that we should not reach any conclusion, On the contrary, the jurisdiction, if very cautiously and very sparingly exercised, in my judgment serves a useful purpose, fully in accord with the overriding objective of enabling the court to deal with cases 'justly, as particularised in r 1.1.

"Neuberger J in Re Blenheim Leisure (Restaurants) Ltd (No 3) (1999 Times, 9 November gave some helpful examples of cases where the jurisdiction might justifiably be invoked before the order in question was drawn up:

'...a plain mistake on the part of the court; a failure of the parties to draw to the court's attention a fact or point of law that was plainly relevant; or discovery of new facts subsequent to the judgment being given. Another good reason was if the applicant could argue that he was taken by surprise by a particular application from which the court ruled adversely to him and that he did not have a fair opportunity to consider.'

"It is to be observed that in all these instances, if the court had no power to reconsider its order before it was drawn up, the only remedy open to the party prejudiced would be by way of appeal from the order. Though on such hypothetical facts an appeal would itself have a good chance of success, common sense suggests that in such cases the judge who made the order should himself have the power to vary it before the appeal procedure has to be set in motion, with the likelihood of exposing all parties to far greater expense and delay than an application to the court of first instance."

15. Later on, Sir Christopher Slade made these additional comments (at 2275G/2276D):

"Since there must be some finality in litigation and litigants cannot be allowed unlimited bites at the cherry, it is not surprising that, according to the authorities, there are stringent limits to the exercise of the discretion conferred on the court by the Barrell jurisdiction.

16. He then quoted from the judgment in the Barrell case, and continued:

"This principle must apply a fortiori where the judgment is a formal written judgment in final form, handed down after the parties have been given the opportunity to consider it in draft and make representations on the draft. The principle recognises that the doing of justice requires justice to both parties in litigation, not merely one.

"At least until the coming into force of the CPR, the Barrell decision would have been clear authority, binding on this court, for the proposition that only in exceptional circumstances can it be proper for a judge to exercise his discretion under the relevant jurisdiction to vary a previous order of his once such order has been made. It may be that now, having regard to the CPR and what was said as to their effect in Biguzzi's case, the Barrell decision is no longer strictly binding authority. Nevertheless, all the considerations which led the court to decide as it did in that case in my judgment still apply. They are in my judgment not merely consistent with but also a proper application of the overriding objective of enabling the court to deal with cases justly as stated in CPR 1.1(1), having regard to all the various factors that fall to be taken into account, by virtue of r.1.1(2), in dealing with cases justly."

17. Roch LJ put the matter in this way (at 2291G/2292A):

"The power of a court to reopen, whether by revocation or variation, its judgment or order must be exercised sparingly in my judgment, if it is to be exercised in accordance with the overriding objective of the CPR. The overriding objective of the new code is to enable a court to deal with a case justly; see CPR 1.1(1). In dealing with a case justly, as in discharging its duty to manage a case, the court must bear in mind expense, the financial position of each party, the desirability of ensuring the parties are on an equal footing and that cases should be dealt with quickly and efficiently, which includes dealing with as many aspects of the case as the court can on the same occasion; see CPR 1.1(2)(a)(b) and (d), and 1.4(2)(i) and (1).

"The judge in his judgment accepted that the jurisdiction of a court to reopen a judgment or order which it has delivered but which has not yet been drawn up was discretionary; a discretion to be used in special cases. The judge said: 'It is clear that where the court has heard argument on a point and made a decision, it will be exceptional that it will allow it to be reopened'."

At 2294A he said:

"The court should require the party seeking to reopen the full and final judgment to demonstrate that it is an exceptional case or that there are strong reasons for doing so."

18. Clarke LJ, dissenting, thought that the overriding principle of CPR 1.1 had overtaken the Barrell test, and that the critical question was not whether the situation was exceptional but what the interests of justice required: see at 2282D/2283B and 2284C.
19. Mr Gee submitted that the Barrell case was concerned with the issue of fresh evidence and that Stewart v. Engel was concerned with the issue of the amendment of pleadings, and that both such contexts partook of their own nature and were not a good guide to all possible cases in which the reconsideration jurisdiction might arise. In particular they were not a good guide to the present case, where his submission was that I had simply erred in the application of classic principle and in disregard of leading authorities binding on me. Over the wider jurisdiction as a whole, he submitted that there was no need of exceptional circumstances, and that other decisions of the court of appeal concerned with the reconsideration jurisdiction had not imposed the requirement of exceptionality. If necessary, however, he would submit that this case met such a requirement.
20. I shall have to revert to these submissions, but first I must seek to say something further about the error which Mr Gee submits I have made in my judgment.
21. I will seek very much to simplify a complex range of submission. In essence, as it seems to me, Mr Gee complains that I have failed to apply the parol evidence rules as they should be applied, and in doing so have reversed the burden of proof. Having found that Mr Bagudu had agreed a sum of $100 million, I should have simply concluded that the tripartite agreement was, with the insertion of $100 million for the "settlement amount" in clause 2 and the "settlement sum" in clause 3, a binding contract. Instead, I have gone on to consider whether the $100 million was agreed unconditionally, and having concluded that it was not proved that it was, I have on that ground alone wrongly declined to give effect to the tripartite agreement. I was wrong to do so because while it is always permissible to have regard to parol evidence to give content to (Mr Gee would say to "interpret" or give meaning to) a descriptive phrase such as "a settlement amount" (clause 2) or "a settlement sum" (clause 3), it is, however, impermissible to look at such parol evidence for the purpose of modifying the written contract. The only ground on which it might be permissible to consider parol evidence for that purpose is if a case of a collateral condition or arrangement had been pleaded by the party who sought to undermine the effectiveness of the tripartite agreement, namely the SJ Berwin defendants. If such a case had been pleaded, then it would have to have been proved. However, such a case had never been pleaded, and there is in any event no finding in my judgment that it had been proved. Since I had in every other respect held that the tripartite agreement on its true construction was intended to bind its parties (para 586 of my judgment) and had rejected various cases that the legal effectiveness of the agreement was upset by collateral representation, stipulation or agreement (para 587), I was wrong to do other than conclude that the tripartite agreement was binding and effective: binding inter alia on Noga to withdraw its claims against the SJ Berwin defendants in its two actions and on the SJ Berwin defendants to pay a settlement sum of $100 million to Noga in return.
22. In this connection Mr Gee himself at one point of his submissions boiled down the essence of his submissions to two or at most three authorities. He relied on Prenn v. Simonds [1971] 1 WLR 1381 (a) for Lord Wilberforce's statement of the legitimate and illegitimate uses of parol evidence, what Mr Gee described as the "inclusionary rule" and the "exclusionary rule" (at 1384/5), and (b) for Lord Wilberforce's approval, as part of the inclusionary rule, of Macdonald v. Longbottom (1859) 1 E & E 977, where a contract of sale for "your wool" was upheld with the assistance of parol evidence to identify the wool in question. Thus Lord Wilberforce said (at 1384):

"Moreover, at any rate since 1859 (Macdonald v. Longbottom, 1 E & E 977) it has been clear enough that evidence of mutually known facts may be admitted to identify the meaning of a descriptive term."

23. His second authority was Hillas & Co Ltd v. Arcos Ltd (1932) 38 Com Cas 23, where the doctrine of certum est quod certum reddi potest was affirmed in order to save a contract which was otherwise said to be too uncertain to stand as other than a mere agreement to agree. And his third authority was Heilbut, Symons & Co v. Buckleton [1913] AC 30, where Lord Moulton, in a classic passage at 47, said that collateral contracts must be strictly proved, otherwise the authority of written contracts would be undermined.
24. This being the error of which Mr Gee complains, he submits that I should rectify it by rewriting those parts of my judgment at paras 579/601 which are inconsistent with the authorities which he has cited and the analysis which he has based on those authorities. He submits that there is no need to alter or add a single finding of fact. The conclusion for which he contends is inevitable upon the findings that I have made, if only the correct analysis were adopted and the binding authorities applied.
25. In this connection, Mr Gee recognises that it is not in every case that a judge can properly be asked to reconsider and reverse his judgment before it is perfected by order. He therefore relies on the following considerations deriving from the instant case to support his submission that either exceptional circumstances or strong reasons are present here, and/or that the overriding principle and the interests of justice demand that I reconsider my judgment. He submits that I should do so ex debito justiciae.
26. I trust that I give proper effect to the extensive range of his submissions in the following passages of this judgment.
27. Mr Gee submits that my judgment as it stands is inconsistent with numerous binding decisions or decisions of great persuasive importance, none of which were cited at trial, and which ought to have been cited, as a matter of the duty of counsel. My judgment is inherently inconsistent. It is therefore per incuriam. A short point of law is involved, which has been overlooked, which is determinative of Noga's case and which should now be determined. In the critical part of my judgment I overlooked clause 2 of the tripartite agreement (see paras 579 and 584, where I mention inter alia clause 3 but not clause 2). In closing oral submissions, Mr Pollock QC, counsel for the SJ Berwin defendants, misrepresented Noga's case, which was simply based on the written contract contained in the tripartite agreement, as being a case based on a contract partly written and partly oral: Noga's case was therefore "swapped". As a result Noga's case at trial had been overlooked. It was not a case based on a contract partly oral and partly written, but a case based on a written contract which referred to the figure of $100 million by means of the descriptive phrase "settlement amount" or "settlement sum". In the circumstances, my judgment simply fails to articulate reasons, or sufficiently transparent reasons, for its conclusion. In any event, Noga was not given a fair opportunity to deal with the manner in which its case has been dealt with. Finally, in the light of my finding that $100 million was agreed, Noga has won and is entitled to judgment.
28. In order to put these submissions into context, it is necessary for me to say something about the trial. It was a complex and lengthy one. With three parties involved, the permutations of interests, cases and possibilities multiplied. At the end of the evidence, there was an adjournment of some four weeks for counsel to write and serve final submissions and for me to read them. The request was for a longer than normal period, both because of the weight of the oral evidence, and also because counsel wished to put as much as possible into writing. Following the service of final written submissions, which ran into many hundreds of pages, there was a period of four (long) days for closing speeches. The period available was by agreement between the parties and by direction of the court divided up so as to give all three parties a fair allocation of time. The end of the spring term in May 2000 represented a deadline for completion of the argument of which everyone had long had notice. I was due to take up my responsibilities in the court of appeal in the summer term. Everyone had been working to that deadline for some time. The directions for the timing of final written submissions and closing speeches took this timing into account. Mr Gee, representing the claimant Noga, had two closing speeches, one in reply. Everyone was under a lot of pressure. The pressure on Mr Gee and his team was increased by the late realisation, initiated by enquiries made by the SJ Berwin defendants, that the timings of telephone records had been misunderstood and needed to be re-evaluated. Noga had to respond, during the run-up to and during final speeches themselves, to a new case made by the SJ Berwin defendants about the telephone record timings (see eg at paras 85 and 237 of my judgment). These telephone timings were relevant to the events of 17 and 29 September 1999 and thus important to the whole structure of the argument relating to the $100 million question (see paras 71/85). Mr Gee was under such pressure that at one point he had to absent himself from court for a period during final speeches. Nevertheless, I am satisfied that he was able to follow the argument, either through the transcript and/or from reports from other members of his team. To the extent that any point of importance was made in his absence, he had an opportunity to reply, and did in fact reply to it.
29. In their final written submissions and in their final speeches, all three parties concentrated on the facts rather than on legal analysis or authorities. Outside the issues of want of consideration and hearsay, practically no authorities were cited, other than a few passages from CHITTY on Contracts.
30. Despite the pressures involved in bringing a large and complex trial to completion, I am satisfied that Noga suffered no unfairness in the process. If Mr Gee had thought that there was any matter that he had not been able to address as he would have wished, he would have been able to request permission to deal with it in writing following the end of the trial.
31. On the subject of legal authorities in particular, it seems to me that the parties had many months to consider what particular points of legal analysis they might wish to emphasise with the use of citation. They chose not to do so, save as I have indicated above. In the course of writing my judgment, I referred on my own initiative to two authorities: one was Prenn v. Simonds (see para 586) and the other was Hillas v. Arcos (see para 607), as it happens two out of the three authorities which Mr Gee now submits encapsulate the core of his case on his application for reconsideration of the judgment.
32. It is in these circumstances that I have to consider the validity of Noga's application. When it was first foreshadowed in Stephenson Harwood's fax of 19 February 2001, it was said that further findings of fact would be requested. This was a matter of particular concern to me, for it was important that I had made all necessary findings of fact. When the parties appeared before me on the occasion which had been fixed for the formal handing down of my judgment, I was told by Mr Gee that no further findings of fact were requested after all. I enquired of the other parties whether they were concerned, in the light of Noga's application, about the adequacy of my findings of fact, and I directed that if any request for further findings was to be made, it should be done without delay. No party has made any request for further findings.
33. On that same occasion, Mr Gee made two further submissions or assertions. One was that the reconsideration of my judgment would be requested on the ground of new points which had not previously been argued, supported by new authorities which had not previously been cited. The other was that, apart from the discretionary jurisdiction to reconsider a judgment which had not yet been perfected by the drawing up of an order in respect of it, there was also a mandatory obligation to reconsider a judgment under the provisions of article 6 of the European Convention of Human Rights. In the event, Mr Gee has persevered in neither of these submissions or assertions. It has become clear in the course of Mr Gee's written and oral submissions that the points which Mr Gee seeks to raise are those which were covered by competing submissions at trial, albeit without the aid of the authorities which Mr Gee now brings into play, and without the refinement of analysis which he now seeks to deploy. As for article 6, Mr Gee has made no separate submission in respect of it. He is prepared to accept that, where no order has yet been drawn up, the jurisdiction to reconsider a judgment in the interests of justice, even if an exceptional one, covers the same ground as article 6.
34. I revert to Mr Gee's submission that the jurisdiction in question does not require exceptional circumstances. For these purposes, he relies on three decisions of the court of appeal other than Barrell itself, of which one, Millensted v. Grosvenor House (Park Lane) Limited [1937] 1 KB 717, preceded Barrell but was not cited in it. Of the other two, the earlier, In re Harrison's Share [1955] Ch 260, preceded Barrell and was cited in it, and the other, Pittalis v. Sherefettin [1986] QB 868, postdated Barrell and referred to it.
35. In Millensted Hilbery J gave an oral judgment in a personal injury claim, finding liability and awarding damages of £50 with costs. He was then informed that there had been a payment of £20 into court. Overnight, he changed his mind of his own motion about the level of damages. The next day, he told the parties that he was revising the level of damages down to £35, and his order was drawn up in that form. His costs order was not changed. The plaintiff appealed. The essence of the appeal was in the submission that the judge was no longer competent to revise his decision on damages after being informed of the sum paid into court. This submission was based on a rule of court stating that a judge should not be informed of a payment in "until all questions of liability and amount of debt or damages have been decided". The court of appeal held that the rule was directory, not compulsory, and left the judge with a discretion to continue. It was common ground that in all other respects a judge was at liberty to reconsider his decision until his order had been drawn up (at 722). It is true that there is no statement that the jurisdiction to reconsider a decision before an order has been drawn up is only exercisable in exceptional circumstances. That, however, is because there is simply no discussion at all in the judgments either as to the circumstances in which the general jurisdiction to reconsider is to be exercised or as to the specific discretion to continue with a case after the court has been informed of a payment in. The reason is that there was no appeal based upon a wrongful exercise of the court's discretion, only on the court's absence of jurisdiction by reason of the special rule of court regarding payment in. In my judgment Millensted is only of interest as an example of the general jurisdiction and as a ruling on the specific rule regarding payment in, but not otherwise. It tells one nothing about the judicial exercise of the jurisdiction to reconsider.
36. I have already referred to In re Harrison's Share above. That was the case where a House of Lords decision showed that the judge had acted without jurisdiction. The only submission of present relevance is that "if and so far as there is any discretionary power the judge should not have exercised it as he did" (at 275). That submission was dealt with by the court at 281/4. The facts were that the House of Lords decision was absolutely plain: the judge was prepared to offer the litigants a further hearing, but they did not want it, because, as the judge himself said "the result is a foregone conclusion" (at 283). In the circumstances, the only argument which was presented upon appeal on the question of discretion was that there was inequality in the treatment of litigants as between those cases of variations of trust affected by the House of Lords decision where a judge's decision had already been drawn up in an order before the House of Lords decision became known to him, and those, such as the instant cases, where the order had not yet been drawn up. But, as the court of appeal pointed out, in either case the variations of trust were doomed, either in the first court or in the court of appeal, and it was wrong to put the parties to the cost of going to the court of appeal to reverse them. In the circumstances there was no need for the court of appeal to discuss the question of whether the jurisdiction to reconsider could only be exercised in exceptional cases. In Barrell, however, Russell LJ pointed out that this decision, like others also reviewed, involved "circumstances of a wholly exceptional nature" (at 24).
37. Mr Gee's third case was Pittalis v. Sherefettin. That concerned a rent review arbitration. The tenant was late in claiming arbitration, and applied to extend time under section 27 of the Arbitration Act 1950. The judge decided that he should not extend time. He had second thoughts and on the very next day he wrote to the parties to say that he was minded to grant the extension of time applied for. At a further hearing, he confirmed his later view. In his second judgment he said that "he was sure" that his first decision was wrong "within minutes of delivering his first judgment" (at 879). The factor which seemed to have driven his ultimate decision was that he thought that the new rent demanded by the landlord was "manifestly grossly inflated". A number of points were argued on appeal, and in the event the appeal succeeded, for the court of appeal pointed out that the judge had had no adequate material upon which he could come to the conclusion that the new rent demanded was grossly inflated. However, despite the success of the appeal, the court of appeal held that the judge had been entitled to reconsider his judgment. Fox LJ said (at 879):

"We are dealing with a case where the judge, practically as soon as he gave the judgment, decided that he was wrong. As a matter of the sensible administration of justice and fairness between parties, it seems to me proper in the circumstances that the judge should be at liberty to recall his order. The position can properly be called exceptional."

38. Dillon LJ agreed that the case was exceptional, but nevertheless appears to have thought that the jurisdiction did not require exceptional circumstances. He said (at 882):

"It is submitted for the landlords that the decision of this court in In re Barrell Enterprises [1973] 1 W.L.R. 19 is to be regarded as a comprehensive exposition of the exceptional circumstances in which it is proper for a court or a judge to recall an order which has been pronounced orally but has not yet been drawn up, registered, or otherwise perfected. But I cannot regard In re Barrell Enterprises as overruling or qualifying the earlier decision of this court in Millensted v. Grosvenor House (Park Lane) Ltd. [1937] 1 K.B. 717, which apparently was not cited in In re Barrell Enterprises. In the Millensted case this court approved statements of the law in earlier authorities to the effect that a judge can always reconsider his decision until his order has been drawn up or perfected, and more importantly, this court upheld the action of a High Court judge, who had in an oral judgment awarded a certain sum by way of damages, but then withdrew that judgment and substituted judgment for a lower figure, because before his order was drawn up, he was satisfied after serious further consideration that the sum he had originally awarded was excessive…

"The Millensted case is, in my judgment, a close parallel to the present case. It is indeed exceptional for a judge who has pronounced an order in court to be completely satisfied, before the order has been drawn up, registered or perfected, that the order was wrong. That happened, however, in the present case, and accordingly the judge was entitled, taking the view he did, to recall his earlier order. I see nothing unfair in the procedure he followed to do so."

39. Neill LJ agreed with both judgments (at 888) and only added further comments on other aspects of the appeal.
40. Mr Gee points out that in Pittalis Dillon LJ remarked on the fact that Millensted was not cited in Barrell; and that neither Millensted nor Pittalis were cited in Stewart v. Engel.
41. Nevertheless, in my judgment, I am bound by the decision in Stewart v. Engel, following the spirit, if not the letter, of the decision in Barrell in the light now of the requirements of the overriding principle, to regard the need for exceptional circumstances as a requirement for the proper exercise of the jurisdiction to reconsider a decision. If in Pittalis Dillon LJ is to be understood as saying by reference to Millensted that the discretion is a wide open one, unrestricted by the requirement of exceptional circumstances, then I would with respect feel bound to disagree. In my judgment the width or narrowness of the discretion was simply not in issue in Millensted. As for Pittalis, both Fox LJ and Dillon LJ accepted that the circumstances in that case were exceptional.
42. Of course, the reference to exceptional circumstances is not a statutory definition and the ultimate interests involved, whether before or after the introduction of the CPR, are the interests of justice. On the one hand the court is concerned with finality, and the very proper consideration that too wide a discretion would open the floodgates to attempts to ask the court to reconsider its decision in a large number and variety of cases, rather than to take the course of appealing to a higher court. On the other hand, there is a proper concern that courts should not be held by their own decisions in a straitjacket pending the formality of the drawing up of an order. As Jenkins LJ said in In re Harrison's Share (at 276):

"Few judgments are reserved, and it would be unfortunate if once the words of a judgment were pronounced there were no locus poenitentiae."

43. Provided that the formula of "exceptional circumstances" is not turned into a straitjacket of its own, and the interests of justice and its constituents as laid down in the overriding principle are held closely to mind, I do not think that the proper balance will be lost. Clearly, it cannot be in every case that a litigant should be entitled to ask the judge to think again. Therefore, on one ground or another, the case must raise considerations, in the interests of justice, which are out of the ordinary, extraordinary, or exceptional. An exceptional case does not have to be uniquely special. "Strong reasons" is perhaps an acceptable alternative to "exceptional circumstances". It will necessarily be in an exceptional case that strong reasons are shown for reconsideration.
44. In the present case Noga asks the court to reconsider its judgment because of the submission that it has got the answer wrong. In every case where an appeal is allowed, the court below has, by definition, got it wrong. The solution is to appeal. What is special, what is exceptional about this case? What are the strong reasons? It is not a case of an ex tempore oral judgment. The judgment here, whatever its defects, has been reserved and is the product of substantial reflection. It is not a case where a new binding precedent has immediately reversed the previous law so as to make a judgment simply unsustainable, as in In re Harrison's Share. It is not a case where a judge has of his own motion immediately come to the conclusion that he is wrong, as in Millensted, or Pittalis (not perhaps a good example of judicial second thoughts), or In Re Australian Direct Steam Navigation Company (1876) 3 Ch D 661, where Sir George Jessel MR realised, after giving an oral judgment, that he had not had his attention directed to the crucial article in the company's articles of incorporation. It is not a case where, even before judgment, a court has realised that it has not had its attention drawn to the critical section in a statute, as in Glebe Sugar Refining Company Ltd v. Trustees of Port and Harbours of Greenock [1921] 2 AC 66, and has itself required a renewed hearing. It is not a case of new evidence, or of amendment. It is not a case of new thoughts. Nor is it a case such as occurred recently in Spice Girls Ltd v. Aprilia World Service BV (unreported, 20 July 2000). There, following the handing down of a judgment on quantum, Arden J received a submission that permission to appeal should be granted because her judgment was inconsistent with a concession of fact that had been made. Arden J sought clarification, and it was then submitted on the other side that no concession had been intended. In such circumstances, Arden J concluded that the appropriate course was to reopen argument, because she was satisfied after a further hearing both that the concession had been made, and that it had been made in such circumstances that it could not be withdrawn. In the event, therefore, she was obliged to reconsider her judgment in the light of the concession which had been overlooked. It was in that context that she said (at para 9):

"In my judgment, an appeal is not the appropriate course where there are errors in judgments which can be corrected by the court which conducted the trial. To leave such matters to an appeal means further delay, uncertainty and costs, which is not in the interests of litigants. The trial judge is in a strong position to consider the effect of the error in the context of the entire case."

45. If this case is like none of those, what is it then? It is a case where it is said that the judge has got it wrong, on points which have been argued. The very issue for reconsideration is in dispute. On behalf of the SJ Berwin defendants Mr Lawrence Cohen QC submits that I have got it right. I am satisfied that I have considered Noga's case and its submissions. That is I think clear from para 584 of my judgment where I specifically consider Mr Gee's submission that it is sufficient to point to the tripartite agreement and say that "with $100 million thought of as being inserted into clause 3 in place of or alongside the words 'a settlement sum', that documentary agreement is complete and binding". Mr Gee complains that I should have mentioned clause 2 as well, but I am not impressed by that submission. I am also satisfied that I had in mind the leading authorities which Mr Gee has mentioned in his application: see para 586 where I mention Prenn v. Simmonds and para 607 where I mention Hillas v. Arcos, even though neither of them were cited at trial. As for Mr Gee's analysis, it seems to me that I had it sufficiently in mind at para 590 where I state:

"Perhaps the true analysis is that the tripartite agreement itself was the final (and binding) outcome of the agreement with Mr Bagudu whereby he was prepared on behalf of the SJ Berwin defendants to agree that Noga should receive $100 million, provided only that the FGN agreed to accept nothing for Ajaokuta – which under the tripartite agreement the FGN did agree to do. In such a case, the conditionality of Mr Bagudu's agreement would have been satisfied as long as the tripartite agreement was binding on the FGN."

46. That is not perhaps a wholly complete statement of Mr Gee's present submissions, but it is very close to being at their centre. Mr Gee would perhaps add that, provided that the tripartite agreement was binding on the FGN, it would not matter what other conditions were stipulated by Mr Bagudu for his agreement to pay Noga $100 million, unless those other conditions were either to be found in the tripartite agreement itself, or were pleaded and proved by the SJ Berwin defendants.
47. I do not wish to say anything against the usefulness of the reconsideration jurisdiction, within its proper limits. I have made use of it myself. However, it is in the nature of the legal process that, once judgment has been rendered, analysis thereafter becomes clarified and refined, and citation of authority is applied to the findings made at first instance so as to illuminate that clarification and refinement of analysis of which I speak. But that is the function of the appeal process. In my judgment, to grant this application that I reconsider my judgment would subvert the appeal process itself. In doing so, it would not answer the interests of justice, but would be the antithesis of justice according to law. There are of course cases where an error of fact or law may be too clear for argument. The best test of that is perhaps – but not necessarily – where the judge himself identifies the error which concerns him. In such a case, it is better that the error is corrected without imposing on the parties the need for an appeal. But no parallel to Noga's application has been cited to me. It is in my judgment wrong for a judge to be treated to an exposition such as would be presented to a court of appeal. If in such circumstances a judge should be tempted to open up reconsideration of his judgment, an appeal would not be avoided, it would be made inevitable. Every case would become subject to an unending process of reconsideration, followed by appeal, both on the issue of reconsideration and on the merits.
48. I therefore reject Noga's application to reconsider my judgment. But I allow its application for permission to appeal. The grounds of such an appeal concern only a limited section of my judgment, do not challenge findings of fact, could be argued in two days, concern a claim for $100 million, and have a realistic prospect of success.
49. I have also been asked for permission to appeal by the SJ Berwin defendants and the FGN.
50. The SJ Berwin defendants want permission to appeal in connection with the point of want of consideration relating to the 16 August agreement (paras 639/651). The point is again a limited one, entirely one of law, worth (potentially at any rate) a great deal both to them and to the FGN, and would not take a great deal of court time. There is a realistic prospect of success. I have granted them permission to appeal. If the FGN needs it, I have also in this context granted it permission to appeal in respect of the Williams v. Roffey point discussed at paras 648/649 of my judgment.
51. The SJ Berwin defendants also want permission to appeal with respect to the issue whether the $50 million promissory note given by Mohammed Abacha to the FGN on 11 May 1999 was given in full and final settlement of the FGN's Ajaokuta claim: see paras 2/50, 423/430, 462/471 and 541/546. I have not granted permission to appeal on that issue. It is essentially a question of fact, or to the extent that it involves any question of law it is entirely dependent upon findings of fact. Those findings cover a large area of evidence, and my ultimate conclusions depend on preferring a witness with credibility over one who lacks credibility (see para 545). Moreover, it is not clear to me that even if the SJ Berwin defendants succeeded on the point, they would be able to overcome the fact that ultimately the 16 August agreement (which required the payment of DM 300 million in settlement of the FGN's Ajaokuta claim) superseded the 13 August agreement; and the 13 August and the 16 August agreements both stated that they discharged the $50 million promissory note. In such circumstances it seems to me that if permission to appeal is to be granted at all it should be done by the court of appeal, who will have the advantages both of bringing to that application a fresh insight and of doing so in the knowledge that their jurisdiction on appeal is primarily one of review and not one of rehearing (see CPR 52.11.1).
52. The FGN applies for permission to appeal in respect of the issue as to the validity or effectiveness of the 13 August agreement. Under my judgment, it failed in its attack on that agreement (see paras 609/638 under the heading – "Was the 13 August agreement induced by misrepresentation or subject to approval?"). Mr David Railton QC submits, however, that he does not need permission to appeal since he can raise such issues under a respondent's notice in response to the SJ Berwin defendants' appeal in respect to the 16 August agreement. Nevertheless, he asks for permission out of an abundance of caution, in case he needs it. I am by no means sure that the FGN does not need permission to appeal, but that is a matter which is still subject to further submissions. If it does need permission to appeal, I am not minded to grant it. The issues are entirely matters of fact. Those factual matters again cover a wide range of evidence and ultimately turn on the credibility of the FGN's critical witness, the attorney general. In such circumstances, I am again of the view that if permission to appeal is to be granted at all it should be done by the court of appeal. There is in any event no point in such an appeal unless the 16 August agreement is void for want of consideration.
53. Finally, I mention a contingent argument raised by the SJ Berwin defendants in their written submissions on Noga's application to reconsider, but not mentioned in their oral submissions. The SJ Berwin defendants submit that if Noga were right in its submission that the tripartite agreement is a binding contract, then under it the FGN would be obliged to withdraw its Ajaokuta claim against the SJ Berwin defendants. It goes on to submit that the 16 August agreement does not expressly supersede the tripartite agreement, as distinct from the 13 August agreement; and that in any event the 16 August agreement could not be accorded any validity where it was itself in breach of the FGN's promise in the tripartite agreement to withdraw its Ajaokuta claim. These were points contingently raised in the SJ Berwin defendants' pleadings but not pursued in final submissions. If I had reconsidered my judgment on Noga's application and concluded that I should change my mind as to the effectiveness of the tripartite agreement, then the SJ Berwin defendants would have wanted also to reconsider their arguments as to the relationship of the tripartite agreement and the 16 August agreement. That is an example of where the reopening of argument by way of reconsideration of a judgment might lead. I have not, however, taken any account of those contingent ramifications in concluding that Noga's application fails. If it had succeeded, and I had reversed my judgment as to the tripartite agreement, then I would have had to consider as an entirely separate matter whether the SJ Berwin defendants' contingent application should succeed.