KPMG LLP v Network Rail Infrastructure Ltd [2007] EWCA Civ 363

[2007] EWCA Civ 363Case No A3/2006/0281
IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM CHANCERY DIVISION
BLACKBURNE J
HC04C02193
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 27/04/2007
Before
LORD JUSTICE MUMMERYLORD JUSTICE CARNWATHSIR PAUL KENNEDY
Between
KPMG LLPAppellantNETWORK RAIL INFRASTRUCTURE LIMITEDRespondent
(Transcript of the Handed Down Judgment ofWordwave International LimitedA Merrill Communications Company190 Fleet Street, London EC4A 2AGTel No: 020 7404 1400, Fax No: 020 7831 8838Official Shorthand Writers to the Court)
Christopher Nugee QC & Timothy C. Dutton (instructed by Messrs. Linklaters) for AppellantMichael Driscoll QC, Thomas Leech & Adam Smith (instructed by Messrs. Thomas Eggar) for RespondentHearing Hearing dates : Tuesday 20 th& Wednesday 21 st March, 2007
Judgment

Introduction

[1]This is a dispute between KPMG and Network Rail (as I shall call them) over the effect of a break-clause in a reversionary sub-underlease (“the lease”). The lease was dated 11th July 1985. It followed generally the form of a draft lease attached to an agreement dated 13th December 1974 (“the 1974 agreement”). The predecessors of the parties during that period were, respectively, Peat Marwick Mitchell & Co (“PMM”) and British Railways Board (“BRB”). The other principal party at that time was King’s College, Cambridge (“King’s”).

The problem

[2]The relevant clause appears at paragraph 5 in Part II of the 5th Schedule of the lease. The schedule is headed:
“Provisions relating to the review of rent payable under this Sub-Underlease”
. Paragraph 1(1) introduces the lessor’s right to review:
“(1) The Landlords may give notice in writing not less than three months nor more than twelve months before the date of the expiration of the first year of the said term and of each successive period of five years thereafter (called "the date of review") throughout the said term requiring the rent payable from the date of review to be reviewed.”
The remainder of paragraphs 1 to 4 are concerned with the mechanics of review, and are not material to the present issue.[3]In its final form paragraph 5 reads as follows:
“5. If the Tenants shall be desirous of determining the said term hereby granted at any time after the expiration of the third thirteenth and twenty third years of the said term and of such desire shall give to the Landlords not less than six calendar months' notice in writing (such notice to expire on any quarter day and to be given within a period of six months following the determination of the reviewed rent to be payable from the expiration of the first eleventh and twenty-first years respectively of the said term [*] and thereafter during the year immediately preceding the eleventh and twenty-first years respectively of the said term) then on the expiration of such notice as aforesaid the said term shall cease and determine but without prejudice to the rights and remedies of either party against the other in respect of any antecedent claim or breach of covenant.”
(emphasis and asterisk added)[4]The effect of the paragraph, including only the first part of the parenthesis (that is, the emphasised passage up to the asterisk), would be reasonably clear: i) It gives the lessee three possible opportunities to determine the lease (at the end of the third, thirteenth and twenty-third years); ii) It defines the time at which the lessee’s notice to determine must be served; but iii) It allows for service of such a notice only following “determination of the reviewed rent”; in other words, if the lessor does not exercise the right to require a review, the lessee has no right to determine.[5]The problem arises from the second part of the parenthesis (after the asterisk). These words have no obvious place in the structure of the paragraph as a whole. Nor, read literally, do the words make any sense. Only in a looking-glass lease would the years “immediately preceding” the eleventh and twenty-first years be treated as coming “after” the expiration of those years.[6]When one compares the 1974 version of the same paragraph, it is apparent that some critical words have been omitted at the point marked [*]. In that version, the second part of the parenthesis reads:
“[but in the event only that such reviewed rent exceeds the yearly rent payable during the (first) I have substituted “the first year” for “the twenty-second year” and “the eleventh and twenty-first years” for “the thirty-second and forty-second years” (cf judgment para 12) to assist direct comparison. As explained below, the 52 year term agreed in 1974 was later split into two terms, of 21 years and 31 years, for stamp duty reasons, and the dates in paragraph 5 were adjusted accordingly. It is not suggested that this change is material to the issues in the appeal. year of the said term] and thereafter during the year immediately preceding the (eleventh) and (twenty-first) years respectively of the said term.”
I have added square brackets to indicate the passage (referred to by the judge as “the relevant words”) which has been omitted from the final version.[7]Even in the 1974 version, the meaning does not spring from the page. The drafting is generally inelegant. It is interesting to note that, in the draft lease attached to the 1974 agreement the page including this paragraph appears to be in a different type-face from the preceding page, and the drafting style is noticeably more elaborate, suggesting perhaps a different hand. Further, the contrast drawn between the rent payable “during the first year” and that payable “during the year immediately preceding the eleventh year” adds to the complexity of the paragraph for no obvious purpose. However, the general intention can reasonably be deduced, and is not in dispute. In its 1974 form, the second part of the parenthesis introduced a further limitation to the lessee’s right to determine: not only must there have been a rent review, but it must have resulted in an increase in the rent. (I shall refer to the second part of the parenthesis, which gave effect to this purpose, as the “increased rent condition”.)[8]The issues for the court are whether the omission can be corrected by any legally permissible technique, or, if not, what meaning is to be ascribed to the paragraph.[9]Before turning to the different ways in which each issue has been developed, it is necessary to set the two transactions in their chronological context. We have the advantage of a lengthy and detailed judgment delivered by Blackburne J, following a 7 day hearing. Reference should be made to that judgment for a full exposition and analysis of the evidence. In what follows, references to paragraphs in the judgment are given as (J:?).[10]The main surviving sources of documentary evidence were the files of the two firms of solicitors (J:63). PMM’s solicitors were Stephenson Harwood, led by Mr Amlot. Simmons & Simmons, led by Mr Butler, acted for BRB and King’s. Mr Amlot and Mr Butler both gave evidence. (J:114 - 125, 137 - 144) The PMM partners most directly concerned were Mr Benzie (J:90) and Mr Webber (J:128), both of whom were members of the PMM Accommodation Committee (J:90, 128). Mr Webber gave evidence. On the other side, evidence was also given by Mr Scutt, a development surveyor with BRB (J:134).[11]For present purposes, the relevant chronology can be shortly stated: i) On 13th December 1974 BRB, King’s, and PMM entered into an agreement for a lease of office accommodation in a new development proposed at Blackfriars, London. The agreement provided for the grant by BRB to PMM of a sub-underlease for a 52-year term commencing on the date of practical completion of stage 2 of the development. Subject to provision for modifications agreed in writing for specific purposes (cl 10), the sub-underlease was to be on the terms of a document annexed to the agreement (J:3, 67). ii) Part II of the 5th Schedule to the document contained provisions for rent review, including, in paragraph 5, the fore-runner of the break-clause which is now in dispute (J:11). iii) Practical completion of stage 2 took place on 25th July 1977, following which PMM went into possession (J:4, 67, 79). iv) During 1978 and 1979 various amendments to the form of the sub-underlease, as annexed to the agreement, were discussed between BRB and King’s, without reference to PMM, and the document was retyped on at least two occasions (J:(80) – (84)). v) In early March 1980 Mr Butler sent Mr Amlot a revised draft, for consideration by his client, noting that there were a number of variations, but “none… of any substance”. The draft was in effect a clean copy and did not show the extent of the changes. There were, as the judge found, over 40 changes, at least two on matters of substance (J:87-8). vi) The changes included the omission of the relevant words from the break-clause. This change, as the judge found, had not been intended by BRB or King’s (J:176). It was suggested in evidence (although the judge made no specific finding) that it was the result of a typist’s error (J:141). vii) On receipt by Mr Amlot’s office, the drafts were compared, and the changes were annotated in red on the new draft (a document which is still available). The omission of the relevant words from the breakclause was one of the changes noted (J:89). viii) On 21st March 1980, Mr Amlot wrote to Mr Benzie, a partner in PMM. The letter ran to over four pages, and commented in detail on the changes (J:90-94). Of the alteration to the break-clause he said: “The original draft provided that you only had the right to determine if there was an increase of rent. This now seems to have been dropped, which would appear to be in your favour”. (J:93) As the judge found, Mr Amlot acted in good faith, and assumed that this change, like the others, was deliberate (J:118, 178). ix) The changes were discussed at a meeting between Mr Amlot and Mr Webber between 25th June and 11th July 1980, but no record of the meeting survives (J:101, 124, 128). As the judge found, there was no specific discussion of the omission of the break-clause (J:187-8). x) On 11th July 1980 Mr Amlot returned to Mr Butler the revised draft, with his proposed amendments shown in red, and his comments. There was no reference to or comment upon the omission of the relevant words (J:102). Nor was this the subject of any further reference in discussions on the draft, either between the parties or internally on either side (J:113). xi) A further change of significance, agreed in 1984, was the division of the 52-year sub-underlease into two: an initial 21-year term, followed by a 31-year reversionary term. This was to save stamp duty for PMM. It necessitated an alteration of the dates in the break-clause, but no other change (J:108-111) xii) The initial and reversionary sub-underleases were finally executed on 11th July 1985 (J:112)

The issues

[12]Against this confused background, Mr Nugee QC, for KMPG, contends that paragraph 5 has to be construed as it stands, without the relevant words. On this basis, he says, it provides the lessee with five opportunities to break the lease: three of them dependent on a rent review; the other two "free-standing”, and dependent only on service of the required notice during the tenth or twentieth year of the term.[13]For Network Rail, Mr Driscoll QC submits that there are only three opportunities to break the lease, all dependent on a rent review. To achieve this result, he argues that the lease should be rectified, on the basis of mutual mistake, so as to supply the missing words; or that the same, or a similar, result should be arrived at as a matter of construction. Alternatively, even if the lessee does indeed have two additional break-rights, those rights should be construed as also dependent upon a prior rent review.

The judge upheld Network Rail’s case for rectification based on mutual mistake (J:

[14]192). KPMG appeals. He rejected Network Rail’s alternative case based on unilateral mistake (J:178), but from that finding there is no appeal. He indicated that, if rectification had not been possible, he would have upheld KPMG’s interpretation (J:56).[15]The commercial significance of the issue is potentially great, particularly if, as is asserted by KPMG (although there is no evidence), the current rent is well above market rent. If the decision below stands, KPMG will be unable to break unless the lessor initiates a rent review in 2009 or 2019. On present rental values, there would be no incentive for the lessor to do this. Even if rental values increase enough to make a rent review worthwhile, the lessor can sidestep the break-clause by leaving the review until 2014 and 2024. KPMG could thus remain liable under the lease until its expiry in 2029. By contrast, on KPMG’s interpretation, it will be able to give notice to break in 2008.

Rectification

[16]Although the judge considered the issues of construction first, I find it more logical to reverse the order. The purpose of rectification for mutual mistake is to ensure that the terms of the written document accurately reflect the state of agreement between the parties. If rectification is ordered, it is the document as so rectified that has to construed.[17]There was no dispute as to the legal requirements. As the judge commented (para 146): “… it is necessary for the party seeking rectification and on whom the burden of proof lies, to demonstrate, on a balance of probabilities but by convincing proof, four matters: i) that the parties had a common continuing intention, whether or not amounting to an agreement, in respect of a particular matter in the instrument to be rectified; ii) that there was an outward expression of accord; iii) that the intention continued at the time of the execution of the instrument sought to be rectified; and iv) that by mistake the instrument did not reflect that common intention.” (See Swainland Builders Ltd v Freehold Properties Ltd [2002] 2 EGLR 71 para 33, per Peter Gibson LJ)[18]The only contentious issue was (iii). There was no doubt that paragraph 5, in the draft attached to the 1974 agreement, represented an “accord”, outwardly expressed. Equally it seemed clear that the omission of the relevant words from the 1985 lease was a mistake. Network Rail’s problem was to show “by convincing proof” that the intention reflected in the 1974 accord had remained unchanged on both sides. KPMG’s case, before the judge and before us, was that on the evidence summarised above such a conclusion is impossible. The only reasonable inference from the exchanges between March and July 1980 was that PMM (albeit mistakenly) thought that the paragraph had been changed deliberately in their favour, and intended to contract on that basis.[19]The judge rejected that case. He stated the question thus: “The question, to my mind…, is whether as a result of Mr Amlot's letter of 21 March 1980 and what followed PMM's intention in this regard underwent any and if so what change. It has rightly not been suggested that Mr Amlot's own understanding and views on the matter are to be imputed to PMM. His function was to advise the partnership, not to make decisions for it.” (J:184)[20]Having noted that Mr Webber was the channel for communications of decisions by PMM to its solicitors (J:185), he drew attention to three particular matters: i) Mr Amlot's letter of 11 July 1980 to Mr Butler following his meeting with Mr Webber, although making reference to many of the proposed amendments (including amendments to the fifth schedule), was silent on the omission of the relevant words from paragraph 5. ii) Mr Amlot did not recall any discussion of the paragraph with anyone either at PMM or at JLW, and stated that he did not believe any such discussion did take place. iii) Although Mr Webber had no particular recollection of what happened at his meeting with Mr Amlot, he nevertheless seemed clear (“…I can say with confidence …”) that he did not review the termination provisions at that meeting or at any other time. (J:187)[21]The judge decided accordingly that the omission was not discussed at that meeting. This view was reinforced by the considerations that Mr Amlot’s comment on this change in his letter of 21st March had been couched in less than certain terms (“this now seems to have been dropped, which would appear to be in your favour”); and that “… anyone giving the paragraph (without the relevant words) any degree of careful attention is almost bound to have asked himself exactly what the paragraph meant and would therefore surely have wanted to establish with the other side what was intended by the omission.” (J:188)[22]He concluded: “That leaves only the possibility that although Mr Webber was the means whereby PMM's views on the amendments to the draft sub-underlease were to be communicated to Mr Amlot, either Mr Benzie (or someone else at PMM) considered and agreed the omission from paragraph 5 and communicated this direct to Mr Amlot, or Mr Benzie (or others at PMM) considered and agreed the amendment but no one communicated this fact to Mr Amlot who, not having been instructed not to accept the amendment caused by the omission, simply allowed the matter to be approved, as it were, by default. There is no evidence beyond speculation to support either hypothesis. I can see no reason why Mr Benzie (or someone else at PMM other than Mr Webber) should have considered and agreed this particular amendment as distinct from the others that Mr Webber did consider with Mr Amlot at their meeting, and then communicated the fact to Mr Amlot. If, however, that is what did occur, I would have expected to see a letter on the file. There is none. As to the second hypothesis, this assumes a breakdown in communication between Mr Benzie (or whoever else was at PMM) and Mr Webber (PMM's channel of communication with Mr Amlot) of which there is no hint in the evidence. It follows therefore that on the balance of probabilities I am persuaded that PMM continued as much after 21 March 1980 as, by reason of the Agreement for Lease, it had before that date, to intend to enter into a sub-underlease which gave to it the same three opportunities to break as were conferred by paragraph 5 of the fifth schedule to the annexed sub-underlease (later adjusted only to reflect the split in the overall term effected by the initial and reversionary sub-underleases). Does Network Rail persuade me of this by convincing proof? Having rejected the claim based on unilateral mistake (and therefore having rejected any suggestion of dishonesty or sharp practice on the part of Mr Amlot and partners in PMM) I cannot but think that the obvious reason why the reversionary sub-underlease was executed containing a clause as ineptly drafted as the last seventeen words of the parenthesis in paragraph 5 is simply because when the matter eventually came to be considered by PMM, and despite Mr Amlot's letter of 21 March 1980, the omission was overlooked. The answer to my question is therefore ‘yes’.” (J:190-192)[23]Mr Nugee has given us a detailed and lengthy critique of these conclusions, by reference to the documents and the transcript of the oral evidence. However, his main points can be summarised quite briefly, and are in my view well-founded.[24]First, none of the witnesses had a clear recollection of the detail of events which had happened more than 25 years ago. The oral evidence was therefore a very fragile tool to undermine the natural inference from the documents. Secondly, it was wrong to place so much weight on the role of Mr Webber, and the lack of any specific discussion of this point at the meeting between him and Mr Amlot, or thereafter. Mr Webber was (in his own words) the “gofer” not the decision-maker; it was Mr Benzie who was in “the driving–seat” (Transcript: Day 4/430-1). In particular, Mr Webber regarded legal issues, such as the break-clause, as “outwith (his) responsibility” (Day 4/459). As to the meeting, the judge gave no reason for not accepting Mr Amlot’s evidence on why the point was not mentioned: “I did not see it necessary to refer to something, an amendment seemed to be made in our clients’ favour which did not seem… to be in dispute in any way.” (Day 3/336)[25]It is indeed surprising, as the judge said, that no-one gave the paragraph sufficient attention to work out precisely what it meant. It is perhaps even more surprising that it was again overlooked by both sides when, in 1984, it was revised, and the dates were altered, following the splitting of the term. However, that is to use the eyes of hindsight. There is no doubt that Mr Amlot did spot the change and advise PMM of its effect, as he believed it to be. Equally, there is no doubt that after his discussions with PMM, he did not find it necessary to refer to it again in his detailed response to Mr Butler. However, there was no reason to infer that Mr Benzie, who was the partner principally responsible, did not read Mr Amlot’s letter or understand its content. It is not speculation, but reasonable inference (consistent with Mr Amlot’s evidence), to assume that it was not referred to again, because neither Mr Amlot nor Mr Benziie saw any need to do so. In my view, the only reasonable inference from the evidence is that PMM accepted Mr Amlot’s view of the change, and were content for the transaction to proceed on that basis.[26]I would add that I also see force in Mr Nugee’s point that the judge’s interpretation was never put to the witnesses in that form. It is necessary to bear in mind KPMG’s original purpose in calling these witnesses. This was in the context of Network Rail’s case based on unilateral mistake. For that purpose it was necessary for them to establish that PMM were aware of the lessors’ mistake and deliberately did nothing to correct it (see J:163). This was the burden of Network Rail’s pleading. Accordingly, Mr Driscoll’s cross-examination of KPMG’s witnesses was directed to showing that there must have been discussion of the point within PMM, and that they consciously decided not to draw attention to it. The judge adopted the opposite interpretation, that far from being discussed the point was wholly overlooked by PMM. But as Mr Nugee submits:
“… this theory was never put to Mr Amlot. Nowhere in his cross-examination is it suggested that he approved the change to para 5 without having this authorised by Mr Webber and without instructions. The purpose of the cross-examination was the complete opposite: it was to try and get him to accept that he must have discussed it with Mr Webber, not that he acted off his own bat.”
[27]Although I would be reluctant to disagree with the judge on an issue of fact, the primary material on which he relied is clearly and fully set out in his judgment. In my view, it did not support his inference that PMM’s intentions remained unchanged in spite of the exchanges in 1980. The burden was on Network Rail to establish “by convincing proof” the facts necessary to support the rectification claim. In respectful disagreement with the judge, I would conclude that they failed to do so, and that the rectification claim should have been rejected.

Alternative case

[28]Network Rail seeks permission to argue in the alternative that the lease should be rectified to give effect at least to the common intention that there should be only three rights to break, all dependent on rent reviews. The lack of agreement on all the details should not be an obstacle. It relies principally on the approach of this court in Swainland Builders Ltd v. Freehold Properties Ltd (cited above).[29]In that case a vendor sold a number of freehold reversions to blocks of flats to a purchaser, at a price based on the common understanding that they were all subject to long leases. By mistake two flats were included in the sale which were not subject to long leases but were let on assured shorthold tenancies. The vendor sought rectification on the basis that the sale should have excluded the freehold of the two flats or that he should have been granted long leases before the sale. The judge ordered rectification of the transfer so as to provide for the grant to the vendor of leases of the two flats. The Court of Appeal upheld this decision. Peter Gibson LJ emphasised the need to distinguish between “the common intention to which effect has not been given by the unrectified document and the remedy which the Court can order”: “The court can grant relief by putting the parties in the same position vis-à-vis each other as they intended. No doubt in most cases it will be possible to see from the material that the parties have considered, and about which they have reached a common intention, the precise wording to be included in the document to be rectified. In the unusual case such as the present, Mr. Perry for the defendant accepted in cross-examination that he would not have objected to the inclusion in the transfer of a requirement to grant leases back to the claimant of the two flats. I see no reason, in principle, why equity should be prevented from giving relief merely because the parties had not agreed on the mechanics by which effect should be given to a clear and simple common intention.” (para 43-44)[30]So in this case, Mr Driscoll submits (in the words of his skeleton):
“(The parties) were agreed on the fundamentals: that there should be 3 rights and that those rights would be dependent on a decision by BRB to initiate a rent review determination (i.e. that BRB would have control over whether the rights would be triggered). The question is whether the removal of the additional qualification (which BRB continued to believe would still apply) that the right would be dependent on an increase in rent, so transformed PMM’s intentions that it is impossible to say that they continued to share a common intention with BRB. Network Rail submits that it did not do so. The fact that the parties might not have been agreed about the requirement for a rent review was no more fundamental than the absence of any agreement to grant long leases in Swainland.”
[31]Attractively as the argument is put, I am unable to accept it. In Swainland there was a common intention as to the commercial effect of the agreement. The only uncertainty was as to the legal mechanics for giving it effect. The grant back of the long leases was an admittedly acceptable means of achieving the common intention. In the present case, by contrast, the common intention extends only part of the way. The issue is not whether the additional qualification was “fundamental”, but whether it was a difference of substance. It was clearly so regarded by Mr Amlot, and there was no reason for the judge to disagree.

The judge rejected this argument on the ground that:

[32]“… it is not permissible, in my view, to order rectification of the paragraph so that it reflects those parts of the parties’ respective intentions about the matter which are the same but ignores those parts which are not.” (J:180)[33]I agree. I would grant permission to appeal so as not to shut out the argument if the matter proceeds further, but I would dismiss the appeal.

Construction

[34]Before considering the rival constructions in detail, it is convenient to deal with two general points:(i) the relevance of the 1974 draft lease to issues of construction;(ii) the court’s power to correct mistakes by construction, as opposed to rectification.

Relevance of 1974 draft lease

[35]In recording the submissions on construction, the judge summarised the well-known principles derived from Lord Hoffmann’s speech in Investors Compensation Scheme v West Bromwich Building Society [1998] 1WLR 896 (“the ICS principles”): “… a contract is to be construed objectively in the light of any relevant fact surrounding the transaction and known to the parties which a reasonable man would have in mind as shedding light on the parties' meaning of the words used. But those facts, as Lord Hoffmann pointed out (see the third of the five principles), exclude, for reasons of practical policy, evidence of negotiations or of the parties' subjective intentions which are only admissible in rectification claims. This is not least because it is a matter of speculation how far such negotiations or intentions were carried into effect.” (J:21)[36]As to whether a prior agreement should be taken into account under those principles, the judge referred to the judgment of Rix LJ in HIH Casualty and General Insurance Ltd v New Hampshire Insurance Co [2001] 2 Lloyds Rep 161 (paras 81 to 84): “83. In principle, it would seem to me that it is always admissible to look at a prior contract as part of the matrix or surrounding circumstances of a later contract. I do not see how the parol evidence rule can exclude prior contracts, as distinct from mere negotiations. The difficulty of course is that, where the later contract is intended to supersede the prior contract, it may in the generality of cases simply be useless to try to construe the later contract by reference to the earlier one. Ex hypothesi, the later contract replaces the earlier one and it is likely to be impossible to say that the parties have not wished to alter the terms of their earlier bargain. The earlier contract is unlikely therefore to be of much, if any, assistance. Where the later contract is identical, its construction can stand on its own feet, and in any event its construction should be undertaken primarily by reference to its own overall terms. Where the later contract differs from the earlier contract, prima facie the difference is a deliberate decision to depart from the earlier wording, which again provides no assistance. Therefore a cautious and sceptical approach to finding any assistance in the earlier contract seems to me to be a sound principle. What I doubt, however, is that such a principle can be elevated into a conclusive rule of law.” (para 83)[37]Applying that approach, Blackburne J concluded that the terms of the 1974 agreement were of no assistance in this case: “It may very well be that the fact that there was an antecedent agreement, in this case the Agreement for Lease, is an admissible background fact but, for the reasons explained by Rix LJ, that fact is of little if any assistance and the terms in which, in that earlier agreement, the parties have expressed themselves, of no assistance. Beyond that the only admissible background facts are(1) the nature of the property (commercial office premises) and(2) the fact that the reversionary sub-underlease follows immediately on from the expiry of the initial sub-underlease. Indeed, the initial sub-underlease is referred to in the reversionary sub-underlease for the purpose of identifying the initial rent payable under the latter.” (J:23)[38]Before us Mr Nugee has sought to support the judge’s narrow view of the relevance of the 1974 lease, but by a somewhat different argument. He accepts that, as a general principle, prior concluded agreements can be admissible to construe a contract (unlike mere drafts and negotiations). However, he argues that leases are in a special category. He starts from a passage of Lord Hoffmann’s speech in Homburg Houtimport BV v Agrosin Private Ltd[2004] 1 AC 715 (“The Starsin”): “The interpretation of a legal document involves ascertaining what meaning it would convey to a reasonable person having all the background knowledge which is reasonably available to the person or class of persons to whom the document is addressed. A written contract is addressed to the parties; a public document like a statute is addressed to the public at large; a patent specification is addressed to persons skilled in the relevant art, and so on.” (para 73)[39]Applying this logic, Mr Nugee submits that a lease, unlike an ordinary commercial contract, creates an interest in land, which may last many years and be owned in different forms by many different parties. It is addressed not merely to the original landlord and tenant, but also to their successors in title, their undertenants, their chargees and so on. Accordingly, what matters is the background material reasonably available to this disparate group of people. That, it is suggested, will include such matters as the physical location and layout of the property, and perhaps common form provisions found in a typical lease; but not a prior agreement for a lease “which is a private arrangement between the original parties and which successors have no right to know about, let alone see”.[40]I am unable with respect to agree with the judge’s reasoning, or the modified version pressed by Mr Nugee. Rix LJ’s words were directed to a different commercial context. There an insurance slip, in typically skeletal form, was expected to be superseded by a detailed policy. Even in that context he was unwilling to lay down a “conclusive rule” excluding reference to the earlier document. In the present case, by contrast, the draft lease attached to the 1974 agreement was on its face a detailed, carefully-negotiated document, to be departed from only for specified reasons. As Mr Nugee concedes, it was sufficient in itself to create an equitable lease, and, if nothing further had occurred, PMM’s occupation would have continued on that basis.[41]Similarly, I see nothing in Mr Nugee’s argument that the original agreement may no longer be available to successors. This is an issue of proof, not principle. In disputes about older documents of title it is a commonplace that the court may not have full information about the context in which the agreement was made. This is as true of questions about the physical state of the land, which Mr Nugee accepts as relevant, as it is about background transactions. Where relevant evidence is not available, the court has to do its best on what there is. I cannot see this as a principled reason for excluding such evidence in a case where it is available.[42]The argument is also inconsistent in my view with the decision of this court in Ladbroke Group plc v Bristol City Council [1988] 1 EGLR 126. That concerned the construction of a 125-year lease, entered into pursuant to a building agreement, which had included a draft lease. The issue related to a reference to a specific date in the rent-review clause, which looked at in isolation was “quite inexplicable”. However, the court considered that it was “permissible and relevant” to look at the draft lease “for the purpose of discovering the parties’ intentions in including the reference…” (per Slade LJ, p 128M-129A). Blackburne J did not find this case helpful: “In the Ladbroke Group case the court felt able to look at the earlier agreement because the lease was expressed to be granted pursuant to it and because the reference to a particular date in the lease was "quite inexplicable" (see page 128M) if read in isolation, i.e. without recourse to the earlier agreement.” (J:52) It is true that the lease in that case was, as the court noted, expressed to be granted “in pursuance of” the earlier agreement. However, I do not see that as a critical distinction. The matters to be taken into account under ICS principles are not limited to those specifically referred to in the contract.[43]In my view, the 1974 agreement, including the form and content of the draft lease attached to it, was an important part of the background and is a permissible aid in the construction of the lease in its final form.

Construction of mistakes by correction

[44]There is no dispute that as part of the process of construing a contract the court can correct obvious errors. Lewison The Interpretation of Contracts (2004), in a section headed “correction of mistakes by construction” (cap 9 para 9.01) states:
“As part of the process of construction the court has power to correct obvious mistakes in the written expression of the intention of the parties. Once corrected, the contract is interpreted in its corrected form.”
[45]In support there are citations from Wilson v Wilson (1854) 5 HLCas 40, 66 (per Lord St Leonards) and, more recently, East v Pantiles Plant Hire Ltd [1982] 2 EGLR 111. In the latter (at p 112), Brightman LJ summarised the requirements:
“Two conditions must be satisfied: first there must be a clear mistake on the face of the instrument; secondly it must be clear what correction ought to be made in order to cure the mistake. If those conditions are satisfied, then the correction is made as a matter of construction. If they are not satisfied then either the claimant must pursue an action for rectification or he must leave it to a court of construction to reach what answer it can on the basis that the uncorrected wording represents the manner in which the parties decided to express the manner in which the parties decided to express their intention.”
[46]So much is uncontentious, in so far as one is dealing with errors “on the face of the instrument”. However, as Lewison observes, the cases are not confined to such errors:
“… in order to decide whether there is such a mistake, the court may take into account such evidence of background facts as is admissible in order to interpret the contract.”
The principal cases cited in support of that proposition are Holding & Barnes plc v Hill House Hammond Ltd [2002] L&TR 7; [2001] EWCA Civ 1334, where the Court of Appeal looked at six other leases executed contemporaneously as part of the same overall transaction; and The Starsin, where an obvious gap was filled by words imported from a standard model clause.[47]I will need to look at the latter case in more detail, since, as the judge acknowledged, it provided the closest analogy to the present case. At this stage I would make one general comment. In the passage cited above Brightman LJ might be read as distinguishing “correction as a matter of construction”, as he called it, from on the one hand “rectification”, and on the other construing the uncorrected wording. However, as I am sure he would have recognised, correction of mistakes under this head is not a separate process, but is simply one facet of the task of “the court of construction”.[48]This point was emphasised by Sir Martin Nourse in Holding & Barnes (above). He referred to the principle stated by Brightman LJ, and concluded that in the instant case not only was there a clear clerical error, but it was also clear what the correction should be. He added:
“I arrive at this conclusion by seeking to ascertain the common intention of the parties from the words they have used in the Barking lease in the light of the material provisions of the other six leases. This is the classical process of construction […] It enables the court to correct an obvious clerical error in a document that it may conform with the obvious intention of the parties. Although in a loose sense the document is rectified, indeed the process is sometimes referred to as common law rectification, it is not rectification in the correct sense. It remains an exercise in construction.” (para 47) I agree respectfully with that passage. I confess, however, to having omitted his comment that the process “owed nothing” to the wider approach adopted in recent cases such as ICS (which had been cited in the preceding judgment of Clarke LJ). It is true that the principle stated by Brightman LJ pre-dated the more flexible approach to interpretation adopted in the recent authorities. It was in effect confirmed in the fifth of the ICS principles: “The “rule” that words should be given their `natural and ordinary meaning' reflects the common sense proposition that we do not easily accept that people have made linguistic mistakes, particularly in formal documents. On the other hand, if one would nevertheless conclude from the background that something must have gone wrong with the language, the law does not require judges to attribute to the parties an intention which they plainly could not have had.”
[49]However, without the wider approach allowed by cases such as ICS,the legitimacy of having regard to the other lease might have been in doubt. As Clarke LJ observed, “Those cases are to my mind of particular assistance here because they show that the question is what a reasonable person would understand the parties to mean by the words of the contract to be construed. It is important to note that that reasonable person must be taken to have knowledge of the surrounding circumstances or factual matrix…” (para 18) It was in reliance on the new approach that he felt able to look beyond the particular lease to others in the same group.[50]The relevance of this point in the present case is that, having decided that the 1974 draft lease is a legitimate aid to construction, it is wrong to compartmentalise the process of interpretation. Both in the judgment, and in the arguments before us, there was a tendency to deal separately with correction of mistakes and construing the paragraph “as it stands”, as though they were distinct exercises. In my view, they are simply aspects of the single task of interpreting the agreement in its context, in order to get as close as possible to the meaning which the parties intended.

The rival interpretations

[51]As already noted, three possible interpretations were advanced before us: i) (For KPMG) interpreting the paragraph “as it stands”, it provides the lessee with five opportunities to break the lease: three of them dependent on a rent review; the other two "free-standing”; ii) (For Network Rail) after reinserting the relevant words, it provides only three opportunities to break, each dependent on a rent review followed by an increase in rent; iii) (For Network Rail) interpreting it as it stands, even if the lessee has two additional break-rights, those rights should be construed as also dependent upon a prior rent review. Before the judge, Network Rail advanced a further interpretation, based on simply ignoring the last part of the parenthesis (“and thereafter… said term”) as meaningless. The practical effect of this would be to retain the paragraph in its original form, but without the increased rent condition. In other words it would have the same effect as the alternative rectification case, which I have already rejected. Mr Driscoll’s attempt to advance this version as a “fall-back” argument as part of the construction case was dismissed by the judge, and there is no appeal by either party. However, it remains relevant for reasons I will explain.[52]In choosing between these alternatives, my starting point would be to consider the structure and grammar of the paragraph as a whole. The difficulty comes not in the main part of the paragraph, but in a parenthesis, indicated by brackets. A parenthesis (in the words of the Concise Oxford Dictionary) is:
“… inserted as an explanation or afterthought into a passage which is grammatically complete without it…”
[53]Without the part in brackets, paragraph 5 reads as follows:
“If the Tenants shall be desirous of determining the said term hereby granted at any time after the expiration of the third thirteenth and twenty third years of the said term and of such desire shall give to the Landlords not less than six calendar months' notice in writing… then on the expiration of such notice as aforesaid the said term shall cease and determine but without prejudice to the rights and remedies of either party against the other in respect of any antecedent claim or breach of covenant.”
Read in this way, the general effect of the paragraph is clear; it is to give the tenant three opportunities to determine the lease at specific dates. Grammatically, the parenthesis does not purport to alter that general effect, but appears as a qualification, or explanation, of the requirements for the notice (“such notice…”).[54]That analysis to my mind is a powerful obstacle in the way of KPMG’s interpretation, unless there is no other permissible alternative. The paragraph was clearly designed to give three rights to break, not five. That is confirmed by comparison with the 1974 draft. If the draftsman had been intending in 1985 to introduce a significant change in that respect, it is almost inconceivable that he would have done it in this way. Even inelegant drafting may have its own consistency. The change in the number of break opportunities would have been signalled in the substantive part of the paragraph, not the parenthesis.[55]The judge was driven to accept KPMG’s preferred interpretation without enthusiasm, because having rejected the possibility of correction, he thought it the best available of the paragraph “as it stands” (J:56).[56]His reasons for rejecting the possibility of correction had been given in paragraphs 48 to 52 of the judgment. The main points can be summarised as follows: i) It was wrong to have regard to the 1974 draft, without also taking account of the fact that changes were made to it, including changes to paragraph 5: “The court cannot be clear that there has been an erroneous omission of the relevant words. For all the court knows, the parties' intention may have been to omit all of the words in the parenthesis from and after the words "but in the event only…"…, alternatively the word "thereafter" may have been mistranscribed from an intermediate draft, alternatively some other words might have been intended but, in the course of typing, erroneously left out.” (J:48) ii) The “plain and obvious” intention of the parties was that the 1985 lease should set out the terms of their leasehold relationship to the exclusion of the terms of the 1974 draft: “Once that position is reached, the remarks of Rix LJ in paragraph 83 of HIH Casualty and General Insurance v New Hampshire Insurance … are in point, namely, that the parties' contract must be found exclusively in the terms of the initial and reversionary sub-underleases and that the Agreement for Lease, in particular the annexed sub-underlease, cannot be used to contradict those terms.” (J:50) iii) The cases relating to correction of obvious mistakes were of no direct assistance. In particular, in The Starsin the House of Lords was able to supply the missing words from a standard form “in very wide use”; “That, in my view, is a far cry from this case where the paragraph in question is, or would seem to be, one specially drafted for use in the sub-underlease.” (J:52)[57]I do not, with respect, find any of these points convincing, or sufficient to justify adopting a construction which pays no regard to the structure of the paragraph as a whole. On the second point, I have already given my reason for disagreeing with the judge’s view, relying on Rix LJ’s comments in HIH, that the 1974 draft lease should be given no weight. As to the previous point, the fact that some alterations had been made to that draft does not justify rejecting the assistance to be gained from those parts which had not changed. The basic structure of paragraph 5 was identical in both drafts. It is true that there is room for argument as to how the mistake should be corrected (to which I shall return), but that is no reason for adopting a construction which no-one can have intended.[58]Finally, contrary to the judge’s view, I regard The Starsin as providing direct support for Mr Driscoll’s preferred approach. To explain why, it is necessary to refer to the case in more detail. It concerned a so-called “Himalaya” clause in a bill of lading. The particular clause read as follows: "(2) Without prejudice to the generality of the provisions in this bill of lading, every exemption limitation, condition and liberty herein contained and every right exemption from liability, defence and immunity of whatsoever nature applicable to the carrier or to which the carrier is entitled hereunder shall also be available to and shall extend to protect every such servant or agent of the carrier [*] is or shall be deemed to be acting on behalf of and for the benefit of all persons who are or might be his servants or agents …" (asterisk added by Lord Bingham) It was accepted that something was clearly missing at the point indicated by the asterisk. The problem was how to correct it. The defendants’ case was that the omitted words could be found in a standard bill of lading form (the “Conline” form), on which the clause appeared to have been closely modelled, “although with some additions, deletions and changes of language”. On that basis the omitted words should be taken as: “… acting as aforesaid and for the purpose of all the foregoing provisions of this clause the carrier…” It was suggested that the omission was to be explicable – “… by the phenomenon, technically known as homoeoteleuton, where one sentence contains a word which closed the preceding sentence and the transcriber's eye has wandered from one to the other, leading to the entire omission of the whole passage lying between them.” (see para 22, per Lord Bingham)[59]The claimants accepted that there was an omission, but submitted that it should be corrected by the insertion of the word “who”. That approach had been accepted in the lower courts. Lord Bingham summarised the argument in support: “While acknowledging that there was no justification for inserting "who" other than a need to correct an obvious grammatical solecism, Mr Milligan urged that, since it was not the function of the courts to make or re-write the parties' contracts, the interpolation to be made should be the least intrusive reasonably possible to make sense of the clause.” (para 22)[60]Lord Bingham, with the agreement of the rest of the House, preferred the defendants’ approach: “I take it to be clear in principle that the court should not interpolate words into a written instrument, of whatever nature, unless it is clear both that words have been omitted and what those omitted words were … In the present case there is agreed to be an omission. It is also plain, in my opinion for the reasons which Mr Gee gave, what words were omitted and how they came to be omitted. I would accordingly construe the clause as if the words 'acting as aforesaid and for the purpose of all the foregoing provisions of this clause the carrier' appeared in place of the square brackets I have inserted.” (para 23)[61]To similar effect, Lord Millett said: "192 … The clause does not make grammatical sense as it stands, and it is obvious that words have been omitted. The court must, therefore, supply the omission by implying at least the minimum necessary for the clause to make grammatical sense. This is what all the judges below did. But the authorities show that in a proper case the court will go further. Where it can see, not only that words have been omitted, but what those words are, then it is its duty to supply them. It is not necessary that the court should be certain precisely what words have been omitted; it is sufficient that it knows their gist. The process is one of construction, not rectification; this is evident from the fact that the Court of Chancery not infrequently supplied omissions in wills at a time when it had no jurisdiction to rectify them. … 194. It is obvious that the clause is not an original work of legal draftmanship but is taken from a precedent. Several versions of the clause are in circulation, and it is impossible to identify the particular precedent from which the defective clause in the present case was taken. But they all employ the same mechanism of agency to give legal efficacy to the clause; they all do so by identical or nearly identical words; and they all incorporate the mechanism at precisely that part of the present clause where words have been omitted. In my opinion this is a clear case where the court can and should supply the missing words.” It is noteworthy that the cases referred to in the House did not include those principally relied on in Lewison, such as East v Pantiles (Plant Hire) Ltd or Holding & Barnes.They were, as Lord Millettt indicated, principally concerned with the power of the Court of Chancery to correct errors in wills (in the absence of any power to rectify).[62]Blackburne J cited these passages. However, as already noted, he regarded the case as distinguishable because in that case the missing words could be identified without difficulty from a standard form. Whatever differences existed between the various versions of the precedent, they all used the same formula at this point. By contrast, the 1974 draft lease was “specially drafted”.[63]I would make three comments. First, it should not matter in principle that the draft in this case was tailor-made, rather than “off-the-peg”, as in The Starsin. What matters is the help it can give in practice, as a matter of common sense rather than law, as to the nature of the mistake and how it should be corrected. Secondly, it is to be noted that the House was untroubled by the facts that the clause appeared in a number of versions, and that there were changes of detail. What mattered was its relevance to the particular passage in issue. So here, the changes made to other parts of the lease, or even to this paragraph, do not detract from the value of the 1974 draft in explaining the place of the relevant words in the parenthesis.[64]Thirdly, and perhaps more contentiously, I think it would be wrong to apply too literally Lord Bingham’s reference to the need for clarity both as to the omission of words and “what those relevant words were”. As Lord Millett said, it is sufficient if the court is able to ascertain “the gist” of what has been omitted. I would go further. Once the court has identified an obvious omission, and has found in admissible background materials an obvious precedent for filling it, it should not be fatal that there may be more than one possible version of the replacement, or more than one explanation of the change. Thus, in The Starsin, even if the various versions of the standard clause had contained some material variations of detail in the relevant passage, I do not think that the court would have been forced for that reason alone to adopt the construction proposed by the claimants. Of course such variations may be sufficient to throw doubt on the precedent as providing an explanation for the error. But, if not, the court is simply faced with the ordinary task of choosing between the competing interpretations, using the ordinary techniques of construction.[65]In the present case, for the reasons I have given, it is obvious from a reading of the 1985 version on its own that something has gone wrong. Comparison with the 1974 draft leaves no doubt as to the parameters of the error. It lies in the transposition of the second part of the parenthesis. I agree, however, with the judge, that comparison of the two versions, by itself, does not yield a definitive answer. In my view it leaves only two realistic possibilities: either too much was left out, or not enough. Either(a) the relevant words should not have been left out, and the parenthesis was intended to remain as in the 1974 draft; or(b) it was intended to leave out not just the relevant words, but the whole of the second part of the parenthesis, that is, the increased rent condition.[66]For my part, if the latter alternative had been still in play, I would have seen some attractions in it. It would have the great advantage of simplifying the over-elaborate language of the paragraph. Furthermore, it is doubtful if it would have made much difference in substance. Since the lessor is unlikely to serve a rent review notice unless he expects to achieve at least some increase, there may be little practical purpose in a specific condition requiring an actual increase. Such considerations might well have led those advising the lessor on the revisions to the draft lease in 1980 to decide that the condition should be omitted altogether, as Mr Amlot understood it to do. Such considerations might also explain why he did not find it a particularly surprising change. Furthermore, in choosing between the competing versions (a) and (b), there would be a principled case for choosing the version which least favoured the originator of the document (the contra proferentem rule).[67]As it is, we do not have to make that choice. However, the existence of two plausible alternatives does not undermine the case for correction, or force the court to adopt a solution which has no plausibility at all. I have no doubt that Mr Driscoll’s proposed interpretation, based on the 1974 draft, is a better reflection of the parties’ intentions than Mr Nugee’s. Unlike the judge, I can see no legal obstacle to its adoption.

Conclusion

[68]Thus, in the end and by a different route, I arrive at the same conclusion as the judge. Accordingly, I would dismiss the appeal.

Sir Paul Kennedy :

[69]I agree.[70]I also agree.

Cited in 27 later judgments

and 18 more of the most senior