LORD HOPE OF CRAIGHEAD
1. I have had the privilege of reading in draft the opinion of my noble and learned friend, Lord Hoffmann. Like my noble and learned friend, Lord Walker of Gestingthorpe, whose opinion I have also had the privilege of reading, I agree with all his reasoning and I share Lord Walker’s admiration for the way it has been expressed. For the reasons they give I would allow the appeal.
2. I agree that Persimmon’s argument that the House should take account of the pre-contractual negotiations raises an important issue. Every so often the rule that prior negotiations are inadmissible comes under scrutiny. That is as it should be. One of the strengths of the common law is that it can take a fresh look at itself so that it can keep pace with changing circumstances. But for the reasons that have been set out by Lord Hoffmann I think that the arguments for retaining the rule have lost none of their force since Prenn v Simmonds  1 WLR 1381 demonstrated, as Lord Wilberforce put it at p 1384, the disadvantages and danger of departing from established doctrine.
3. In the Court of Appeal Lawrence Collins LJ said that the policy reasons for the rule have not been fully articulated:  EWCA Civ 183, para 106. I am not sure, with respect, that everyone would agree with him. Lord Gifford did his best to explain what they are in his dissenting opinion in Inglis v Buttery (1877) 5 R 58, 69-70. When that case came before this House Lord Blackburn said that they set out exactly what he himself thought: (1878) 3 App Cas 552, 577. As Lord Gifford explained, the very purpose of a formal contract is to put an end to the disputes which would inevitably arise if the matter were left upon what the parties said or wrote to each other during the period of their negotiations. It is the formal contract that records their bargain, however different it may be from what they may have stipulated for previously
4. Lord Blackburn clearly saw no conflict between the exclusionary rule and Lord Justice Clerk Moncreiff’s proposition that the Court was entitled to be put in the position that the parties stood before they signed: (1877) 5 R 58, 64. In River Wear Commissioners v Adamson (1877) 2 App Cas 743, 763 he had already acknowledged that the court should look beyond the language of the contract and see what the circumstances were with reference to which the words were used. As he put it, the meaning of words varies according to the circumstances with respect to which they are used. It was the reasons that Lord Gifford articulated in Inglis v Buttery (1877) 5 R 58, 69-70, that persuaded him that to admit evidence of prior negotiations would be a step too far. I think that what appealed to Lord Blackburn still holds true today. If more is needed, Lord Hoffmann’s analysis provides it. As he has indicated, it would only be if your Lordships were confident that the rule was impeding the proper development of the law or contrary to public policy that it would be right for it to be departed from. That this is so has not, as I see it, been demonstrated.
5. On 16 October 2001 Chartbrook Ltd (“Chartbrook”) entered into an agreement with Persimmon Homes Ltd (“Persimmon”), a well-known house-builder, for the development of a site in Wandsworth which Chartbrook had recently acquired. The structure of the agreement was that Persimmon would obtain planning permission and then, pursuant to a licence from Chartbrook, enter into possession, construct a mixed residential and commercial development (commercial premises below, flats above, parking in the basement) and sell the properties on long leases. Chartbrook would grant the leases at the direction of Persimmon, which would receive the proceeds for its own account and pay Chartbrook an agreed price for the land. Planning permission was duly granted and the development was built, but there is a dispute over the price which became payable.
6. Schedule 6 contained the relevant provisions. The price was defined as the aggregate of the Total Land Value and the Balancing Payment. The Total Land Value was made up of three parts: Total Residential Land Value, Total Commercial Land Value and Total Residential Cark Parking Land Value. Total Residential Land Value was to be Â£76.34 per square foot multiplied by the area for which planning permission for flats was granted. Total Commercial Land Value was Â£38.80 per square foot multiplied by the area for which planning permission for shops and other commercial uses was granted. And Total Residential Cark Parking Land Value was Â£3,024 multiplied by the number of spaces for which planning permission was granted. The Schedule set out the dates upon which the Total Land Value was to be paid. In principle, payment would fall due as each flat, shop or parking space was sold. But there was also a backstop provision for payment of specified percentages of the Total Land Value (so far as not already paid) by dates commencing about two and a half years after the grant of planning permission and ending about two years later, by which time the whole sum was due, whether the properties had been sold or not.
7. The provisions about Total Land Value are all quite straightforward and only require the insertion of the appropriate figures from the planning permission (which are not in dispute) into the formulae provided. The other element in the price is the Balancing Payment. For reasons concerned with its drafting history which need not be explored, the Schedule defines the Balancing Payment as the Additional Residential Payment (“ARP”) and then goes on to define the latter expression. So when I refer to the ARP, that means the Balancing Payment.
8. The definition of the ARP, over which the whole dispute turns, is outwardly uncomplicated:
“23.4% of the price achieved for each Residential Unit in excess of the Minimum Guaranteed Residential Unit Value less the Costs and Incentives.”
9. This contains three more defined concepts. Residential Unit means a flat. The Minimum Guaranteed Residential Unit Value (“MGRUV”) means the Total Residential Land Value divided by the number of flats. And Costs and Incentives (“C and I”) mean the additional expense which Persimmon might have to incur to induce someone to buy a flat; for example, by providing fittings better than specification or paying legal expenses. Such payments are economically equivalent to a reduction in the price achieved.
10. Chartbrook says that the meaning of the definition is perfectly simple. You take the price achieved, deduct the MGRUV and the C and I and calculate 23.4% of the result. That gives you a figure for an individual flat which, together the figures for similar calculations on all the other flats, makes up the ARP or Balancing Payment. That and the Total Land Value is the price. On the agreed figures, that produces a Total Land Value of Â£4,683,565 and an ARP of Â£4,484,862, making Â£9,168,427 in all. The judge (Briggs J)  EWHC 409 (Ch) and a majority of the Court of Appeal (Tuckey and Rimer LJJ)  EWCA Civ 183 agreed.
11. This construction is certainly in accordance with conventional syntax, at any rate, up to the point at which one decides when C and I should be deducted. As Briggs J said (at para 53) —
“ARP means 23.4% of something. To the question ‘23.4% of what?’ the clear answer is the excess of the price achieved for each Residential Unit over the MGRUV, less the Costs and Incentives.”
12. I do not think that the syntax helps one to decide whether C and I should be deducted before or after calculating the 23.4%, that is to say, whether there is a notional pause for breath after “MGRUV", represented in the passage I have quoted from the judgment by a comma which does not appear in the contract. That is a grammatical ambiguity which must be resolved by considering the business purpose of providing for a deduction of C and I. But the judge was clearly right about the effect of the syntax employed in the first part of the definition.
13. Persimmon, on the other hand, says that the purpose of dividing the price into Total Land Value and ARP was to give Chartbrook a minimum price for its land, calculated on current market assumptions, and to allow for the possibility of an increase if the market rose and the flats sold for more than expected. It is agreed that, at the time of the agreement, the parties expected that a 700 square foot flat would sell for about Â£200,000 or so, maybe slightly more. The MGRUV at Â£76.34 a square foot for such a flat was Â£53,438 or 26.7% of a price of Â£200,000. If the realised price was Â£228,000, it would represent 23.4%. The purpose of the ARP was to provide that if the flats sold for more than Â£228,000, Chartbrook would be entitled to the amount by which 23.4% of the higher price exceeded the Â£53,438 MGRUV. What the definition therefore means is that you deduct C and I from the realised price to arrive at the net price received by Persimmon, then calculate 23.4% of that price, and the ARP is the excess of that figure over MGRUV. On this calculation, ARP is Â£897,051, compared with Chartbrook’s claim for Â£4,484,862. In the Court of Appeal Lawrence Collins LJ, dissenting, held that Persimmon’s construction was correct.
14. There is no dispute that the principles on which a contract (or any other instrument or utterance) should be interpreted are those summarised by the House of Lords in Investors Compensation Scheme Ltd v West Bromwich Building Society  1 WLR 896, 912-913. They are well known and need not be repeated. It is agreed that the question is what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean. The House emphasised that “we do not easily accept that people have made linguistic mistakes, particularly in formal documents” (similar statements will be found in Bank of Credit and Commerce International SA v Ali  1 AC 251, 269, Kirin-Amgen Inc v Hoechst Marion Roussel Ltd  RPC 169, 186 and Jumbo King Ltd v Faithful Properties Ltd (1999) 2 HKCFAR 279, 296) but said that in some cases the context and background drove a court to the conclusion that “something must have gone wrong with the language". In such a case, the law did not require a court to attribute to the parties an intention which a reasonable person would not have understood them to have had.
15. It clearly requires a strong case to persuade the court that something must have gone wrong with the language and the judge and the majority of the Court of Appeal did not think that such a case had been made out. On the other hand, Lawrence Collins LJ thought it had. It is, I am afraid, not unusual that an interpretation which does not strike one person as sufficiently irrational to justify a conclusion that there has been a linguistic mistake will seem commercially absurd to another: compare the Kirin-Amgen case  RPC 169 at pp. 189-190. Such a division of opinion occurred in the Investors Compensation Scheme case itself. The subtleties of language are such that no judicial guidelines or statements of principle can prevent it from sometimes happening. It is fortunately rare because most draftsmen of formal documents think about what they are saying and use language with care. But this appears to be an exceptional case in which the drafting was careless and no one noticed.
16. I agree with the dissenting opinion of Lawrence Collins LJ because I think that to interpret the definition of ARP in accordance with ordinary rules of syntax makes no commercial sense. The term “Minimum Guaranteed Residential Unit Value", defined by reference to Total Residential Land Value, strongly suggests that this was to be a guaranteed minimum payment for the land value in respect of an individual flat. A guaranteed minimum payment connotes the possibility of a larger payment which, depending upon some contingency, may or may not fall due. Hence the term “Additional Residential Payment". The element of contingency is reinforced by paragraph 3.3 of the Sixth Schedule, which speaks of the “date of payment if any of the Balancing Payment.” (My emphasis).
17. The judge declined to regard the terms Total Land Value and Minimum Guaranteed Residential Unit Value as indicative of an intention that MGRUV was to be the minimum Chartbrook would receive as the land value of a flat because both terms were defined expressions. They might just as well have been algebraic symbols. Indeed they might, and I strongly suspect that if they had been, they would have made it clear that the parties were intending to give effect to Persimmon’s construction. But the contract does not use algebraic symbols. It uses labels. The words used as labels are seldom arbitrary. They are usually chosen as a distillation of the meaning or purpose of a concept intended to be more precisely stated in the definition. In such cases the language of the defined expression may help to elucidate ambiguities in the definition or other parts of the agreement: compare Birmingham City Council v Walker  2 AC 262, 268. I therefore consider that Lawrence Collins LJ was right to take into account the connotations of contingency to be derived from the defined terms.
18. On Chartbrook’s construction, there is virtually no element of contingency at all. ARP is payable in every case in which the flat sells for more than Â£53,438. Chartbrook submits that is still a contingency. Who could tell whether or not the market for flats in Wandsworth might not collapse? In the Court of Appeal, Rimer LJ accepted that submission. He said that the “relevant language", i.e. the language of contingency, was “strictly consistent also with Chartbrook’s construction.”
19. My Lords, I cannot believe that any rational parties who wished to make provision for such a catastrophic fall in the housing market (itself an unlikely assumption) would have adopted so precise a sum to represent their estimate of what might happen. Why Â£53,438? That was the agreed minimum figure for that part of the value of a flat attributable to the land which Chartbrook was selling. It was clearly based upon a careful and precise estimate of current market prices and building costs. But how could this figure have been appropriate as a minimum expected sale price of the entire flat at some future date? If the parties were wanting to guess at some extraordinary fall in the market against which Chartbrook was to be protected, why Â£53,438? Why not Â£50,000 or Â£60,000, or Â£100,000? A figure chosen to represent someone’s fears about a possible collapse in the market could only have been based upon wild speculation, not the kind of calculation which produces a figure like Â£53,438. That figure cannot have been meant to play the part in the calculation which Chartbrook’s construction assigns to it. It must have been intended to function as a minimum land value, not a minimum sale price. To compare it with the realised sale price would not be comparing like with like.
20. It is of course true that the fact that a contract may appear to be unduly favourable to one of the parties is not a sufficient reason for supposing that it does not mean what it says. The reasonable addressee of the instrument has not been privy to the negotiations and cannot tell whether a provision favourable to one side was not in exchange for some concession elsewhere or simply a bad bargain. But the striking feature of this case is not merely that the provisions as interpreted by the judge and the Court of Appeal are favourable to Chartbrook. It is that they make the structure and language of the various provisions of Schedule 6 appear arbitrary and irrational, when it is possible for the concepts employed by the parties (MGRUV, C and I etc) to be combined in a rational way.
21. I therefore think that Lawrence Collins LJ was right in saying that ARP must mean the amount by which 23.4% of the achieved price exceeds the MGRUV. I do not think that it is necessary to undertake the exercise of comparing this language with that of the definition in order to see how much use of red ink is involved. When the language used in an instrument gives rise to difficulties of construction, the process of interpretation does not require one to formulate some alternative form of words which approximates as closely as possible to that of the parties. It is to decide what a reasonable person would have understood the parties to have meant by using the language which they did. The fact that the court might have to express that meaning in language quite different from that used by the parties (“12th January” instead of “13th January” in Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd  AC 749; “any claim sounding in rescission (whether for undue influence or otherwise)” instead of “any claim (whether sounding in rescission for undue influence or otherwise)” in Investors Compensation Scheme Ltd v West Bromwich Building Society  1 WLR 896) is no reason for not giving effect to what they appear to have meant.
22. In East v Pantiles (Plant Hire) Ltd (1981) 263 EG 61 Brightman J stated the conditions for what he called “correction of mistakes by construction":
“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.”
23. Subject to two qualifications, both of which are explained by Carnwath LJ in his admirable judgment in KPMG LLP v Network Rail Infrastructure Ltd  Bus LR 1336, I would accept this statement, which is in my opinion no more than an expression of the common sense view that we do not readily accept that people have made mistakes in formal documents. The first qualification is that “correction of mistakes by construction” is not a separate branch of the law, a summary version of an action for rectification. As Carnwath LJ said (at p. 1351, para 50):
“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.”
24. The second qualification concerns the words “on the face of the instrument". I agree with Carnwath LJ (at pp 1350-1351) that in deciding whether there is a clear mistake, the court is not confined to reading the document without regard to its background or context. As the exercise is part of the single task of interpretation, the background and context must always be taken into consideration.
25. What is clear from these cases is that there is not, so to speak, a limit to the amount of red ink or verbal rearrangement or correction which the court is allowed. All that is required is that it should be clear that something has gone wrong with the language and that it should be clear what a reasonable person would have understood the parties to have meant. In my opinion, both of these requirements are satisfied.
26. That leaves the question of the deduction of C and I, which the judge and the majority of the Court of Appeal regarded as an insuperable obstacle to Persimmon’s construction. I cannot see why this should be so. Everyone agrees that the only sum from which C and I can rationally be deducted is the headline price achieved on the sale, so as to arrive at the net amount received by Persimmon. That is accordingly what the parties must have meant. You deduct the C and I from the nominal price achieved and the ARP is the excess, if any, of 23.4% of that net sum over the MGRUV. Giving this meaning to the provision about C and I does not in any way weaken or affect the argument for interpreting the rest of the definition in a way which gives ARP a rational meaning. To say, as Rimer LJ said, that it requires “rewriting", or that it “distorts the meaning and arithmetic of the definition” is only to say that it requires one to conclude that something has gone wrong with the language - not, in this case, with the meanings of words, but with the syntactical arrangement of those words. If however the context drives one to the conclusion that this must have happened, it is no answer that the interpretation does not reflect what the words would conventionally have been understood to mean.
27. If your Lordships agree with this conclusion about the construction of the contract, the appeal must be allowed. There is no need to say anything more. But Persimmon advanced two alternative arguments of very considerable general importance and I think it is appropriate that your Lordships should deal with them. The first was that (contrary to the unanimous opinion of the judge and the Court of Appeal) the House should take into account the pre-contractual negotiations, which in the opinion of Lawrence Collins LJ (at paragraph 132), were determinative confirmation of Persimmon’s argument on construction. The second was that the judge and the Court of Appeal had misunderstood the principles upon which rectification may be decreed and that if Persimmon had failed on construction, the agreement should have been rectified.
28. The rule that pre-contractual negotiations are inadmissible was clearly reaffirmed by this House in Prenn v Simmonds  1 WLR 1381, where Lord Wilberforce said (at p. 1384) that earlier authorities “contain little to encourage, and much to discourage, evidence of negotiation or of the parties’ subjective intentions.” It is clear that the rule of inadmissibility has been established for a very long time. In Inglis v John Buttery and Co (1878) 3 App Cas 552, 577 Lord Blackburn said that Lord Justice Clerk Moncreiff (at (1877) 4 R 58, 64) had laid down a principle which was nearly accurate but not quite when he said that in all mercantile contracts “whether they be clear and distinct or the reverse, the Court is entitled to be placed in the position in which the parties stood before they signed". The only qualification Lord Blackburn made was to reject Lord Moncreiff’s view that the Court was entitled to look at the pre-contractual negotiations because unless one did so, one could not be fully in the position in which the parties had been.
29. Instead, Lord Blackburn preferred (at p. 577) the opinion of Lord Gifford ((1877) 4 R 58, 69-70):
“Now, I think it is quite fixed - and no more wholesome or salutary rule relative to written contracts can be devised - that where parties agree to embody, and do actually embody, their contract in a formal written deed, then in determining what the contract really was and really meant, a Court must look to the formal deed and to that deed alone. This is only carrying out the will of the parties. The only meaning of adjusting a formal contract is, that the formal contract shall supersede all loose and preliminary negotiations - that there shall be no room for misunderstandings which may often arise, and which do constantly arise, in the course of long, and it may be desultory conversations, or in the course of correspondence or negotiations during which the parties are often widely at issue as to what they will insist on and what they will concede. The very purpose of a formal contract is to put an end to the disputes which would inevitably arise if the matter were left upon verbal negotiations or upon mixed communings partly consisting of letters and partly of conversations. The written contract is that which is to be appealed to by both parties, however different it may be from their previous demands or stipulations, whether contained in letters or in verbal conversation. There can be no doubt that this is the general rule, and I think the general rule, strictly and with peculiar appropriateness applies to the present case.”
30. To allow evidence of pre-contractual negotiations to be used in aid of construction would therefore require the House to depart from a long and consistent line of authority, the binding force of which has frequently been acknowledged: see Bank of Scotland v Dunedin Property Investment Co Ltd 1998 SC 657, 665 (“well-established and salutary", per Lord President Rodger; Alexiou v Campbell  UKPC 11 (“vouched by…compelling authorities", per Lord Bingham of Cornhill.) The House is nevertheless invited to do so, on the ground that the rule is illogical and prevents a court from, as the Lord Justice Clerk in Inglis v John Buttery and Co (1878) 3 App Cas 552 said, putting itself in the position of the parties and ascertaining their true intent.
31. In Prenn v Simmonds  1 WLR 1381, 1384 Lord Wilberforce said by way of justification of the rule:
“The reason for not admitting evidence of these exchanges is not a technical one or even mainly one of convenience, (though the attempt to admit it did greatly prolong the case and add to its expense). It is simply that such evidence is unhelpful. By the nature of things, where negotiations are difficult, the parties’ positions, with each passing letter, are changing and until the final agreement, though converging, still divergent. It is only the final document which records a consensus. If the previous documents use different expressions, how does construction of those expressions, itself a doubtful process, help on the construction of the contractual words? If the same expressions are used, nothing is gained by looking back: indeed, something may be lost since the relevant surrounding circumstances may be different. And at this stage there is no consensus of the parties to appeal to. It may be said that previous documents may be looked at to explain the aims of the parties. In a limited sense this is true: the commercial, or business object, of the transaction, objectively ascertained, may be a surrounding fact. Cardozo J. thought so in the Utica Bank case. And if it can be shown that one interpretation completely frustrates that object, to the extent of rendering the contract futile, that may be a strong argument for an alternative interpretation, if that can reasonably be found. But beyond that it may be difficult to go: it may be a matter of degree, or of judgment, how far one interpretation, or another, gives effect to a common intention: the parties, indeed, may be pursuing that intention with differing emphasis, and hoping to achieve it to an extent which may differ, and in different ways. The words used may, and often do, represent a formula which means different things to each side, yet may be accepted because that is the only way to get ‘agreement’ and in the hope that disputes will not arise. The only course then can be to try to ascertain the ‘natural’ meaning. Far more, and indeed totally, dangerous is it to admit evidence of one party’s objective - even if this is known to the other party. However strongly pursued this may be, the other party may only be willing to give it partial recognition, and in a world of give and take, men often have to be satisfied with less than they want. So, again, it would be a matter of speculation how far the common intention was that the particular objective should be realised.”
32. Critics of the rule, such as Thomas J in New Zealand (Yoshimoto v Canterbury Golf International Ltd  1 NZLR 523, 538-549) Professor David McLauchlan (“Contract Interpretation: What is it About?” (2009) 31:5 Sydney Law Review 5-51) and Lord Nicholls of Birkenhead (“My Kingdom for a Horse: The Meaning of Words” (2005) 121 LQR 577-591) point out that although all this may usually be true, in some cases it will not. Among the dirt of aspirations, proposals and counter-proposals there may gleam the gold of a genuine consensus on some aspect of the transaction expressed in terms which would influence an objective observer in construing the language used by the parties in their final agreement. Why should court deny itself the assistance of this material in deciding what the parties must be taken to have meant? Mr Christopher Nugee QC, who appeared for Persimmon, went so far as to say that in saying that such evidence was unhelpful, Lord Wilberforce was not only providing a justification for the rule but delimiting its extent. It should apply only in cases in which the pre-contractual negotiations are actually irrelevant. If they do assist a court in deciding what an objective observer would have construed the contract to mean, they should be admitted. I cannot accept this submission. It is clear from what Lord Wilberforce said and the authorities upon which he relied that the exclusionary rule is not qualified in this way. There is no need for a special rule to exclude irrelevant evidence.
33. I do however accept that it would not be inconsistent with the English objective theory of contractual interpretation to admit evidence of previous communications between the parties as part of the background which may throw light upon what they meant by the language they used. The general rule, as I said in Bank of Credit and Commerce International SA v Ali  1 AC 251, 269, is that there are no conceptual limits to what can properly be regarded as background. Prima facie, therefore, the negotiations are potentially relevant background. They may be inadmissible simply because they are irrelevant to the question which the court has to decide, namely, what the parties would reasonably be taken to have meant by the language which they finally adopted to express their agreement. For the reasons given by Lord Wilberforce, that will usually be the case. But not always. In exceptional cases, as Lord Nicholls has forcibly argued, a rule that prior negotiations are always inadmissible will prevent the court from giving effect to what a reasonable man in the position of the parties would have taken them to have meant. Of course judges may disagree over whether in a particular case such evidence is helpful or not. In Yoshimoto v Canterbury Golf International Ltd  1 NZLR 523. Thomas J thought he had found gold in the negotiations but the Privy Council said it was only dirt. As I have said, there is nothing unusual or surprising about such differences of opinion. In principle, however, I would accept that previous negotiations may be relevant.
34. It therefore follows that while it is true that, as Lord Wilberforce said, inadmissibility is normally based in irrelevance, there will be cases in which it can be justified only on pragmatic grounds. I must consider these grounds, which have been explored in detail in the literature and on the whole rejected by academic writers but supported by some practitioners.
35. The first is that the admission of pre-contractual negotiations would create greater uncertainty of outcome in disputes over interpretation and add to the cost of advice, litigation or arbitration. Everyone engaged in the exercise would have to read the correspondence and statements would have to be taken from those who took part in oral negotiations. Not only would this be time-consuming and expensive but the scope for disagreement over whether the material affected the construction of the agreement (as in the Yoshimoto case) would be considerably increased. As against this, it is said that when a dispute over construction is litigated, evidence of the pre-contractual negotiations is almost invariably tendered in support of an alternative claim for rectification (as in Prenn v Simmonds and in this case) or an argument based on estoppel by convention or some alleged exception to the exclusionary rule. Even if such an alternative claim does not succeed, the judge will have read and possibly been influenced by the evidence. The rule therefore achieves little in saving costs and its abolition would restore some intellectual honesty to the judicial approach to interpretation.
36. There is certainly a view in the profession that the less one has to resort to any form of background in aid of interpretation, the better. The document should so far as possible speak for itself. As Popham CJ said in the Countess of Rutland’s Case (1604) 5 Co Rep 25b, 26a:
“it would be inconvenient, that matters in writing made by advice and on consideration, and which finally import the certain truth of the agreement of the parties should be controlled by averment of the parties to be proved by the uncertain testimony of slippery memory.”
37. I do not think that these opinions can be dismissed as merely based upon the fallacy that words have inherent or “available” meanings, rather than being used by people to express meanings, although some of the arguments advanced in support might suggest this. It reflects what may be a sound practical intuition that the law of contract is an institution designed to enforce promises with a high degree of predictability and that the more one allows conventional meanings or syntax to be displaced by inferences drawn from background, the less predictable the outcome is likely to be. In this respect, it is interesting to consider the reaction to the statement of principle in Investors Compensation Scheme Ltd v West Bromwich Building Society  1 WLR 896,912-913, which was viewed with alarm by some distinguished commercial lawyers as having greatly increased the quantity of background material which courts or arbitrators would be invited to consider: see Lord Bingham’s recent paper (“A New Thing Under the Sun: The Interpretation of Contract and the ICS Decision” (2008) 12 Edinburgh LR 374-390) and Spigelmann CJ, “From Text to Contract: Contemporary Contractual Interpretation” (2007) 81 ALJ 322. As Lord Bingham pointed out, there was little in that statement of principle which could not be found in earlier authorities. The only points it decided that might have been thought in the least controversial were, first, that it was not necessary to find an “ambiguity” before one could have any regard to background and, secondly, that the meaning which the parties would reasonably be taken to have intended could be given effect despite the fact that it was not, according to conventional usage, an “available” meaning of the words or syntax which they had actually used.
38. Like Lord Bingham, I rather doubt whether the ICS case produced a dramatic increase in the amount of material produced by way of background for the purposes of contractual interpretation. But pre-contractual negotiations seem to me capable of raising practical questions different from those created by other forms of background. Whereas the surrounding circumstances are, by definition, objective facts, which will usually be uncontroversial, statements in the course of pre-contractual negotiations will be drenched in subjectivity and may, if oral, be very much in dispute. It is often not easy to distinguish between those statements which (if they were made at all) merely reflect the aspirations of one or other of the parties and those which embody at least a provisional consensus which may throw light on the meaning of the contract which was eventually concluded. But the imprecision of the line between negotiation and provisional agreement is the very reason why in every case of dispute over interpretation, one or other of the parties is likely to require a court or arbitrator to take the course of negotiations into account. Your Lordships’ experience in the analogous case of resort to statements in Hansard under the rule in Pepper v Hart  AC 593 suggests that such evidence will be produced in any case in which there is the remotest chance that it may be accepted and that even these cases will be only the tip of a mountain of discarded but expensive investigation. Pepper v Hart has also encouraged ministers and others to make statements in the hope of influencing the construction which the courts will give to a statute and it is possible that negotiating parties will be encouraged to improve the bundle of correspondence with similar statements.
39. Supporters of the admissibility of pre-contractual negotiations draw attention to the fact that Continental legal systems seem to have little difficulty in taking them into account. Both the Unidroit Principles of International Commercial Contracts (1994 and 2004 revision) and the Principles of European Contract Law (1999) provide that in ascertaining the “common intention of the parties", regard shall be had to prior negotiations: articles 4.3 and 5.102 respectively. The same is true of the United Nations Convention on Contracts for the International Sale of Goods (1980). But these instruments reflect the French philosophy of contractual interpretation, which is altogether different from that of English law. As Professor Catherine Valcke explains in an illuminating article (“On Comparing French and English Contract Law: Insights from Social Contract Theory”) (16 January 2009), French law regards the intentions of the parties as a pure question of subjective fact, their volontÃ© psychologique, uninfluenced by any rules of law. It follows that any evidence of what they said or did, whether to each other or to third parties, may be relevant to establishing what their intentions actually were. There is in French law a sharp distinction between the ascertainment of their intentions and the application of legal rules which may, in the interests of fairness to other parties or otherwise, limit the extent to which those intentions are given effect. English law, on the other hand, mixes up the ascertainment of intention with the rules of law by depersonalising the contracting parties and asking, not what their intentions actually were, but what a reasonable outside observer would have taken them to be. One cannot in my opinion simply transpose rules based on one philosophy of contractual interpretation to another, or assume that the practical effect of admitting such evidence under the English system of civil procedure will be the same as that under a Continental system.
40. In his judgment in the present case, Briggs J thought that the most powerful argument against admitting evidence of pre-contractual negotiations was that it would be unfair to a third party who took an assignment of the contract or advanced money on its security. Such a person would not have been privy to the negotiations and may have taken the terms of the contract at face value. There is clearly strength in this argument, but it is fair to say that the same point can be made (and has been made, notably by Saville LJ in National Bank of Sharjah v Dellborg  EWCA Civ 2070, which is unreported, but the relevant passage is cited in Lord Bingham’s paper in the Edinburgh Law Review) in respect of the admissibility of any form of background. The law sometimes deals with the problem by restricting the admissible background to that which would be available not merely to the contracting parties but also to others to whom the document is treated as having been addressed. Thus in Bratton Seymour Service Co Ltd v Oxborough  BCLC 693 the Court of Appeal decided that in construing the articles of association of the management company of a building divided into flats, background facts which would have been known to all the signatories were inadmissible because the articles should be regarded as addressed to anyone who read the register of companies, including persons who would have known nothing of the facts in question. In The Starsin (Homburg Houtimport BV v Agrosin Private Ltd  1 AC 715) the House of Lords construed words which identified the carrier on the front of a bill of lading without reference to what it said on the back, on the ground that the bankers to whom the bill would be tendered could not be expected to read the small print. Ordinarily, however, a contract is treated as addressed to the parties alone and an assignee must either inquire as to any relevant background or take his chance on how that might affect the meaning a court will give to the document. The law has sometimes to compromise between protecting the interests of the contracting parties and those of third parties. But an extension of the admissible background will, at any rate in theory, increase the risk that a third party will find that the contract does not mean what he thought. How often this is likely to be a practical problem is hard to say. In the present case, the construction of the agreement does not involve reliance upon any background which would not have been equally available to any prospective assignee or lender.
41. The conclusion I would reach is that there is no clearly established case for departing from the exclusionary rule. The rule may well mean, as Lord Nicholls has argued, that parties are sometimes held bound by a contract in terms which, upon a full investigation of the course of negotiations, a reasonable observer would not have taken them to have intended. But a system which sometimes allows this to happen may be justified in the more general interest of economy and predictability in obtaining advice and adjudicating disputes. It is, after all, usually possible to avoid surprises by carefully reading the documents before signing them and there are the safety nets of rectification and estoppel by convention. Your Lordships do not have the material on which to form a view. It is possible that empirical study (for example, by the Law Commission) may show that the alleged disadvantages of admissibility are not in practice very significant or that they are outweighed by the advantages of doing more precise justice in exceptional cases or falling into line with international conventions. But the determination of where the balance of advantage lies is not in my opinion suitable for judicial decision. Your Lordships are being asked to depart from a rule which has been in existence for many years and several times affirmed by the House. There is power to do so under the Practice Statement (Judicial Precedent)  1 WLR 1234. But that power was intended, as Lord Reid said in R v National Insurance Comrs, Ex p Hudson  AC 944, 966, to be applied only in a small number of cases in which previous decisions of the House were “thought to be impeding the proper development of the law or to have led to results which were unjust or contrary to public policy". I do not think that anyone can be confident that this is true of the exclusionary rule.
42. The rule excludes evidence of what was said or done during the course of negotiating the agreement for the purpose of drawing inferences about what the contract meant. It does not exclude the use of such evidence for other purposes: for example, to establish that a fact which may be relevant as background was known to the parties, or to support a claim for rectification or estoppel. These are not exceptions to the rule. They operate outside it.
43. There is however a group of cases in which judges have found an exception to the exclusionary rule and your Lordships will have to decide whether such an exception can be justified. The leading case is the decision of Kerr J the Karen Oltmann (Partenreederei MS Karen Oltmann v Scarsdale Shipping Co Ltd  2 Lloyd’s Rep 708. This concerned a time charter for 2 years (14 days more or less in charterers’option) which contained a break clause:
“Charterers to have the option to redeliver the vessel after 12 months’ trading subject to giving three months’ notice".
44. The issue was whether “after 12 months’ trading” meant that the break clause could be operated only at the end of the first year or at any time during the second year. The judge said that he was entitled to look at telexes by which the fixture was negotiated in which the parties discussed various lengths of break clauses and were clearly using the word “after” to mean “on the expiry of” rather than “at any time after the expiry of". He justified the admissibility of this evidence on the following principle (p 712):
“If a contract contains words which, in their context, are fairly capable of bearing more than one meaning, and if it is alleged that the parties have in effect negotiated on an agreed basis that the words bore only one of the two possible meanings, then it is permissible for the court to examine the extrinsic evidence relied upon to see whether the parties have in fact used the words in question in one sense only, so that they have in effect given their own dictionary meaning to the words as the result of their common intention. Such cases would not support a claim for rectification of the contract, because the choice of words in the contract did not result from any mistake. The words used in the contract would ex hypothesi reflect the meaning which both parties intended.”
45. In his judgment in this case, Lawrence Collins LJ said of this principle (in paragraph 121) that he doubted whether it differed in any material respect from admitting evidence of prior negotiations in construing a contract. Indeed, the case is frequently cited as an example of an exception which undermines the rule: see for example Professor McLauchlan, “Contract Interpretation: What is It About?” (2009) 31:5 Sydney Law Review 5-51. It is true that evidence may always be adduced that the parties habitually used words in an unconventional sense in order to support an argument that words in a contract should bear a similar unconventional meaning. This is the “private dictionary” principle, which is akin to the principle by which a linguistic usage in a trade or among a religious sect may be proved: compare Shore v Wilson (1842) 9 Cl and F 355. For this purpose it does not matter whether the evidence of usage by the parties was in the course of negotiations or on any other occasion. It is simply evidence of the linguistic usage which they had in common. But the telexes in the Karen Oltmann did not evidence any unconventional usage. There was no private dictionary. The case involved a choice between two perfectly conventional meanings of the word “after” in a particular context. In my opinion Lawrence Collins LJ was right in saying that the admission of the evidence infringed the exclusionary rule. It is perhaps significant that the evidence merely confirmed the meaning which Kerr J, as an experienced commercial judge, would in any case have given to the clause.
46. What would have been the position if Kerr J had thought that, without the evidence of the telexes, he would have construed the clause in the opposite sense? He said that rectification would not be available because “The words used in the contract would ex hypothesi reflect the meaning which both parties intended.” I do not understand this, because, on this hypothesis, the telexes would show that the words (as construed by the judge) did not reflect the meaning which both parties intended. And it is generally accepted that Brightman J was right in Re Butlin’s Settlement Trusts  Ch 251 in holding that rectification is available not only when the parties intended to use different words but also when they mistakenly thought their words bore a different meaning.
47. On its facts, the Karen Oltmann was in my opinion an illegitimate extension of the “private dictionary” principle which, taken to its logical conclusion, would destroy the exclusionary rule and any practical advantages which it may have. There are two legitimate safety devices which will in most cases prevent the exclusionary rule from causing injustice. But they have to be specifically pleaded and clearly established. One is rectification. The other is estoppel by convention, which has been developed since the decision in the Karen Oltmann: see Amalgamated Investment and Property Co. Ltd. v. Texas Commerce International Bank Ltd.  QB 84. If the parties have negotiated an agreement upon some common assumption, which may include an assumption that certain words will bear a certain meaning, they may be estopped from contending that the words should be given a different meaning. Both of these remedies lie outside the exclusionary rule, since they start from the premise that, as a matter of construction, the agreement does not have the meaning for which the party seeking rectification or raising an estoppel contends.
48. The last point is whether, if Chartbrook’s interpretation of the agreement had been correct, it should have been rectified to accord with Persimmon’s interpretation. The requirements for rectification were succinctly summarized by Peter Gibson LJ in Swainland Builders Ltd v Freehold Properties Ltd  2 EGLR 71, 74, para 33:
“The party seeking rectification must show that:
(1) 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;
(2) there was an outward expression of accord;
(3) the intention continued at the time of the execution of the instrument sought to be rectified;
(4) by mistake, the instrument did not reflect that common intention.”
49. To explain how the claim for rectification arose, I must summarise the relevant pre-contractual exchanges between the parties. They began by discussing a proposal for an outright sale of the land by Chartbrook to Persimmon at a price calculated by reference to such planning permission as Chartbrook might obtain. In early 2001 this structure was abandoned and Persimmon in a letter dated 1 February 2001 proposed the building licence arrangement eventually agreed. The letter included the following passages:
“we would be prepared to pay you 29.8% of the net sales proceeds generated from the private sale residential element of the scheme and a further 45% of the net sales revenue generated from the disposal of the commercial element of the site. We would pay you this proportion of the income regardless of the development costs incurred by my Company and the quantum of accommodation that we ultimately obtain planning permission for…By tying your land value to a percentage of the income, you will also automatically share in any sales uplift that we experience.”
50. This offer of a straightforward sharing of the proceeds was modified in a letter dated 6 February 2001 by the addition of what were described as “guaranteed backstop dates and minimum payments":
“Upon receipt of the purchase monies, the revenue will be apportioned to Chartbrook on the basis of 29.8% of the net revenue achieved from the disposal of the private sale residential units and 45% of the net revenue from the disposal of the commercial units. In addition, we are prepared to provide you with guaranteed backstop dates and minimum payments that will be made regardless of the actual performance of the project both in terms of timescales and costs. I set out on the attached schedule our proposals concerning this element of the deal.
Based on the current scheme for 80 units, and 9020 sq ft of commercial floor space, the minimum land value we are prepared to pay to Chartbrook on the disposal of each residential unit is Â£67,000, together with a further minimum payment of Â£400,000 on the disposal of the commercial unit. If as a result of improvements in the market, Chartbrook are entitled to more than the minimum payments I suggest an equalisation calculation takes place following the disposal of the last unit…
Within the contract, I…suggest that a formula is included whereby the land value is calculated using the following inputs:
Private Sale Residential Accommodation…..94.96/sq ft…
Once the total land value has been calculated, a simple formula can then be applied to divide the land values by the number of units, in order for us to calculate the guaranteed payments that you will receive on the sale of each plot…”
51. On 12 February there was a further modification to make separate provision for the sales of car parking spaces, but the overall offer for land value remained the same. The judge found (paragraph 110) that Chartbrook accepted this offer in principle and Persimmon’s solicitors were instructed to draft an agreement. Their draft was attached to an e-mail dated 1 March 2001 and contained essentially the same formulae for calculating the price as those in the final agreement. The definition of “Additional Residential Payment” was (save for the percentage figure) in precisely the same words as those of the final agreement.
52. Between March and May Chartbrook acquired some additional adjoining land and Persimmon revised its cost estimates. The result was a change in the figures but not in the formulae. In a letter dated 24 May 2001 Persimmon offered a new total land value of Â£7,191,947. The letter contained a table setting out —
“the minimum guaranteed land values that you will receive for the respective elements of the scheme, together with the percentage of sales revenue that you will also be entitled to if the project performs better than is currently anticipated”
53. The figures in the table were 23.4% for “percentage of sales revenue” and Â£53,333 for “minimum value per plot.” The judge found that this offer was also accepted in principle and the new figures were inserted into the final contract. The words of the definition of ARP in the final draft remained (subject to the change in the percentage figure) exactly the same as in the first draft.
54. It is I think clear that a reasonable person who read the February and May letters in the light of the background known to the parties would have taken them to have been intending that Chartbrook should receive an ARP if, but only if, “the project performs better than is currently anticipated".
55. Persimmon’s case on rectification at the trial was that the letter of 24 May 2001 was an outward expression of the common and continuing intention of the parties and (if Chartbrook was right about its true construction) the definition had been drafted in the mistaken belief that it gave effect to that common intention. On the other hand, the evidence of the two principals of Chartbrook, Mr Vantreen and Mr Reeve, was that they had made no mistake. The definition accorded exactly with what they had thought they were being offered in the letters of February and May 2001. Indeed, they said they would not have done the deal for any less. It was put to them in cross-examination that no rational person could have understood the letters in the sense which they claimed and Mr Vantreen was caused some little difficulty by the fact that, on his copy of the May 2001 letter, he had calculated the amount which (on Persimmon’s construction of the definition) the sale price of a 700 sq ft flat would have to exceed before any ARP became payable (Â£228,000). This calculation would have been irrelevant on his own construction of the definition and he was unable to explain why he had made it. Nevertheless the judge accepted the evidence of Mr Reeve and Mr Vantreen that they had honestly believed that the definition (as they claimed to have understood it) was what had been agreed and they were not been mistaken. The judge therefore held that the mistake was not common to both parties and dismissed the claim for rectification.
56. The case was argued at trial on the assumption that rectification required both parties to be mistaken about whether the written agreement reflected what they believed their prior consensus to have been. In the Court of Appeal, Persimmon challenged the finding of fact about what Mr Reeve and Mr Vantreen had believed, but not the underlying proposition of law. The Court of Appeal unanimously dismissed this part of the appeal on the ground that it could not disturb the findings of fact. There are accordingly concurrent findings of fact about the states of mind of Mr Reeve and Mr Vantreen. Your Lordships indicated at the hearing that in accordance with the usual practice, you would not re-examine them: see Smith New Court Securities Ltd v Scrimgeour Vickers (Asset Management) Ltd  AC 254, 274-275.
57. In the printed case, however, Persimmon (encouraged by articles in the Law Quarterly Review by Marcus Smith (“Rectification of Contracts for Common Mistake, Joscelyne v Nissen and Subjective States of Mind” (2007) 123 LQR 116-132 and Professor McLauchlan (“The ‘Drastic’ Remedy of Rectification for Unilateral Mistake” (2008) 124 LQR 608-640)) asked for leave to challenge, for first time, the proposition of law. Mr Nugee submitted that the judge and the Court of Appeal had been wrong in their assumption about what a party had to be mistaken about. Rectification required a mistake about whether the written instrument correctly reflected the prior consensus, not whether it accorded with what the party in question believed that consensus to have been. In accordance with the general approach of English law, the terms of the prior consensus were what a reasonable observer would have understood them to be and not what one or even both of the parties believed them to be. In the present case, submitted Mr Nugee, the prior consensus was contained in the May letter, which made it clear that the terms were to be as contended for by Persimmon. If the definition in the final agreement did not have that meaning, it was not in accordance with the prior consensus and if Mr Reeve and Mr Vantreen believed that it was, then they, like the representatives of Persimmon, were mistaken.
58. Mr Robert Miles QC, for Chartbrook, objected to Persimmon being given leave to advance this argument. He said that if the point had been taken at the trial, the evidence might have taken a different shape. I rather doubt this, but as I understand that the Committee shares my view that Persimmon is entitled to succeed without rectification, the question is academic. Nevertheless, as it has been very well and fully argued, I propose to express an opinion about it.
59. Until the decision of the Court of Appeal in Joscelyne v Nissen  2 QB 86 there was a view, based upon dicta in nineteenth and early twentieth century cases, that rectification was available only if there had been a concluded antecedent contract with which the instrument did not conform. In Lovell and Christmas Ltd v Wall (1911) 104 LT 85, 88 Sir Herbert Cozens-Hardy MR said that rectification “may be regarded as a branch of the doctrine of specific performance". It presupposed a prior contract and required proof that, by a common mistake, the final completed agreement as executed failed to give proper effect to the prior contract. In Joscelyne‘s case the Court of Appeal declared itself puzzled by the reference to specific performance, but I think it is clear enough that the Master of the Rolls had in mind a contractual obligation to execute a lease, conveyance, settlement or similar instrument, giving rise to a specifically enforceable obligation to do so. A failure to execute a document giving effect to the terms of the agreement would be a breach of that obligation and the court, in rectifying the instrument, would be specifically performing the agreement. Since the decision in Joscelyne’s case extended the availability of rectification to cases in which there had been no enforceable prior agreement, specific performance is plainly an inadequate explanation of the doctrine. But for present purposes the significance of cases like Lovell and Christmas Ltd v Wall (1911) 104 LT 85 is that the terms of the contract to which the subsequent instrument must conform must be objectively determined in the same way as any other contract. Thus the common mistake must necessarily be as to whether the instrument conformed to those terms and not to what one or other of the parties believed those terms to have been.
60. Now that it has been established that rectification is also available when there was no binding antecedent agreement but the parties had a common continuing intention in respect of a particular matter in the instrument to be rectified, it would be anomalous if the “common continuing intention” were to be an objective fact if it amounted to an enforceable contract but a subjective belief if it did not. On the contrary, the authorities suggest that in both cases the question is what an objective observer would have thought the intentions of the parties to be. Perhaps the clearest statement is by Denning LJ in Frederick E Rose (London) Ltd v William H Pim Jnr and Co Ld  2 QB 450, 461:
“Rectification is concerned with contracts and documents, not with intentions. In order to get rectification it is necessary to show that the parties were in complete agreement on the terms of their contract, but by an error wrote them down wrongly; and in this regard, in order to ascertain the terms of their contract, you do not look into the inner minds of the parties - into their intentions - any more than you do in the formation of any other contract. You look at their outward acts, that is, at what they said or wrote to one another in coming to their agreement, and then compare it with the document which they have signed. If you can predicate with certainty what their contract was, and that it is, by a common mistake, wrongly expressed in the document, then you rectify the document; but nothing less will suffice.”
61. Likewise in the Olympic Pride (Etablissements Georges et Paul Levy v Adderley Navigation Co Panama SA  2 Lloyd’s Rep 67, 72, Mustill J said:
“The prior transaction may consist either of a concluded agreement or of a continuing common intention. In the latter event, the intention must have been objectively manifested. It is the words and acts of the parties demonstrating their intention, not the inward thoughts of the parties, which matter.”
62. An example of the application of this objective ascertainment of the terms of the prior transaction is George Cohen Sons and Co Ltd v Docks and Inland Waterways Executive (1950) 84 Lloyd’s Rep 97 in which a landlord negotiating a new lease proposed to the tenant that “the terms and conditions contained in the present lease to be embodied in the new lease where applicable.” The tenant accepted this offer, but the new lease as executed made the tenant liable for repairs which under the old lease had been the responsibility of the landlord. In answer to a claim for rectification, the landlord said that the new lease was in accordance with what he had understood to be the effect of his offer. The Court of Appeal said that this was irrelevant. What mattered was the objective meaning of what the landlord had written. Sir Raymond Evershed MR said, at p 107:
“If the defendants…did misconstrue [the letter] that is unfortunate for them, but at least they cannot be heard to say that their letter was intended to mean anything other than that which the words convey to the reader as a piece of ordinary English.”
63. As against these authorities, there are two cases upon which Mr Miles relied. The first is Britoil plc v Hunt Overseas Oil Inc  CLC 561, in which the Court of Appeal by a majority (Glidewell and Hobhouse LJJ, Hoffmann LJ dissenting) refused to rectify an agreement which was alleged not to be in accordance with what had previously been agreed in summary heads of agreement. Hobhouse LJ, who gave the majority judgment, affirmed the decision of Saville J, who said that the defendants had failed to establish that there was a prior common agreement or intention in terms that the court could ascertain or (which is probably another way of saying the same thing) that the definitive agreement failed to reflect that prior agreement. In other words, the language of the heads of agreement was too uncertain to satisfy the requirement stated by Denning LJ in Rose’s case that one should be able to “predicate with certainty what their contract was". Hobhouse LJ noted that Saville J “did not base himself upon any consideration of the evidence as to the actual state of mind of the parties” and in my opinion the case lends no support to the view that a party must be mistaken as to whether the document reflects what he subjectively believes the agreement to have been.
64. The other case is the decision of Laddie J in Cambridge Antibody Technology Ltd v Abbott Biotechnology Ltd  FSR 590, in which he rejected a submission that evidence of the subjective state of mind of one of the parties contained in statements which had not been communicated to the other party (“crossed the line”) was inadmissible. In my opinion, Laddie J was quite right not to exclude such evidence, but that is not inconsistent with an objective approach to what the terms of the prior consensus were. Unless itself a binding contract, the prior consensus is, by definition, not contained in a document which the parties have agreed is to be the sole memorial of their agreement. It may be oral or in writing and, even if the latter, subject to later variation. In such a case, if I may quote what I said in Carmichael v National Power plc  1 WLR 2042, 2050 - 2051:
“The evidence of a party as to what terms he understood to have been agreed is some evidence tending to show that those terms, in an objective sense, were agreed. Of course the tribunal may reject such evidence and conclude that the party misunderstood the effect of what was being said and done.”
65. In a case in which the prior consensus was based wholly or in part on oral exchanges or conduct, such evidence may be significant. A party may have had a clear understanding of what was agreed without necessarily being able to remember the precise conversation or action which gave rise to that belief. Evidence of subsequent conduct may also have some evidential value. On the other hand, where the prior consensus is expressed entirely in writing, (as in George Cohen Sons and Co Ltd v Docks and Inland Waterways Executive (1950) 84 Lloyd’s Rep 97) such evidence is likely to carry very little weight. But I do not think that it is inadmissible.
66. In this case there was no suggestion that the prior consensus was based on anything other than the May letter. It is agreed that the terms of that letter were accepted by Chartbrook and no one gave evidence of any subsequent discussions which might have suggested an intention to depart from them. It follows that (on the assumption that the judge was right in his construction of the ARP definition) both parties were mistaken in thinking that it reflected their prior consensus and Persimmon was entitled to rectification.
67. Since, however, I think that the judge and the majority of the Court of Appeal were wrong on the question of construction, I would allow the appeal on that ground.
LORD RODGER OF EARLSFERRY
68. I have had the privilege of considering the speeches of my noble and learned friends, Lord Hoffmann and Lord Walker of Gestingthorpe, in draft. For the reasons which they give, I consider that the construction favoured by Persimmon is appropriate. In particular, it seems to me that once you grasp the general structure of schedule 6 of the agreement, as described by Lord Walker in para 79 of his speech, the appropriate interpretation becomes clear.
69. Like Lord Hoffmann, I would decline counsel’s elegant but, in the event, unnecessary invitation to revisit the rule in Prenn v Simmonds  1 WLR 1381. No-one could possibly say that the rule is based on some error of law or misconception. On the contrary, the main pros and cons of having regard to prior negotiations when interpreting a formal contract have been known and discussed for centuries. The present law represents a choice which was already second nature to the Earl of Eldon LC as long ago as Millers v Miller (1822) 1 Sh App 309. When interpreting a clause in a marriage contract which had been preceded by “a vast deal of correspondence", the Lord Chancellor assured the House that he did not recollect a case to which he had given more earnest attention, but still gave the correspondence short shrift, at p 317:
“My Lords, all the previous correspondence I lay entirely out of the case, because I cannot conceive that any thing can be more dangerous than the construing deeds by the effect of letters and correspondence previous to the execution of them.”
Subsequently, at p 319, he described the possibility of looking at the effect of the correspondence as “a very singular thing". Some sixty years later, with rather more deliberation, the House affirmed that approach in Inglis v John Buttery (1873) 3 App Cas 552 and, a century after that, reaffirmed it in Prenn v Simmonds  1 WLR 1381. The rule could scarcely be more firmly embedded in our law.
70. Of course, in Miliangos v George Frank (Textiles) Ltd  AC 443 the House departed from a rule, of which Lord Denning had said some fifteen years previously, “if there is one thing clear in our law, it is that the claim must be made in sterling and the judgment given in sterling": In re United Railways of Havana and Regla Warehouses Ltd  AC 1007, 1068-1069. But not only was that rule essentially procedural: in addition, the House could point to a change of circumstances which seemed to cry out for intervention. Here, by contrast, the rule about prior negotiations forms part of the law of evidence and there are no particular pressing circumstances which call for a change. The House is simply being asked to make a fresh policy decision and, in effect, to legislate to provide for a different rule. The wisdom of the proposed change is, however, debatable. So, if there is to be a change, it should be on the basis of a fully informed debate in a forum where the competing policies can be properly investigated and evaluated. Although counsel presented the rival arguments with conspicuous skill, your Lordships’ House in its judicial capacity is not that forum.
71. Like Lord Walker, I see no reason to differ from what Lord Hoffmann has said on rectification.
LORD WALKER OF GESTINGTHORPE
72. I shall first address, on a traditional approach, the issue of construction raised in this appeal. That approach requires the court to consider the disputed definition of “Additional Residential Payment” in the context of the agreement as a whole, and the parties’ shared understanding of the general situation and the aim of the transaction they were entering into.
73. The owner (Chartbrook) had assembled a site off Wandsworth High Street, London SW18 (Numbers 1,3,5,7 and 9 Hardwicks Way), with valuable potential for development. Under the agreement dated 16 October 2001 the developer (Persimmon Homes, a subsidiary of a well-known quoted company which guaranteed the developer’s obligations) had the responsibility of applying for planning permission for a mixed commercial and residential development in a form approved by the owner. The agreement was conditional on the developer obtaining planning permission in a satisfactory form within 15 months (subject to extension in certain circumstances). If planning permission in a satisfactory form was obtained the owner would continue as registered owner of the site, but would execute a charge of the property as security for its obligations to the developer. The developer would occupy the site as a licensee and carry out the development at its own expense, including responsibility for insurance. The developer’s obligations in carrying out the residential development (by the construction of flats) were not prescribed in detail. Its obligations in carrying out the commercial development were limited to what were described as “core and shell works".
74. As the flats were developed they were to be marketed by the developer, at its own expense, and sold (together with parking spaces) on 125-year leases at escalating ground rents. The commercial development, when the core and shell works were completed, was to be sold to a nominee of the owner on a 125-year lease at a peppercorn rent. There was to be a premium calculated at the rate of Â£110 per square foot of the net internal area of the commercial premises (plus VAT). The developer was also to negotiate the eventual sale of the freehold subject to all these leases. The owner was under an obligation to grant all the necessary leases and to make the eventual transfer of the freehold.
75. As I have mentioned, the agreement did not provide in detail for the specification or cost of the construction by the developer of the residential part of the development - that is, the flats. The owner had to approve the planning application, and the developer had to meet NHBC standards, but that was all. In particular, the developer did not commit itself to any particular level of expenditure, with two minor exceptions: the developer undertook to spend a sum of at least Â£250,000 on planning gain through an agreement under section 106 of the Town and Country Planning Act 1990, and to pay at least Â£25,000 in compensation to adjoining owners for the loss of rights to light. There was also an unquantified contingency sum for dealing with possible pollution in the sub-structure. These three items were to be deducted in computing the Total Residential Land Value (“TRLV”) for the purposes of schedule 6 (the Price) but there was nothing at all in the agreement providing for the developers’ overall profit as such to be computed and brought into the bargain.
76. All this is background, but to my mind relevant background, to the problem at the heart of this appeal, that is the correct construction of the definition of the Additional Residential Payment (“ARP”) set out in para.1 of schedule 6. The ARP (also pointlessly relabelled as the Balancing Payment) is one of two components which had to be aggregated to make up the Price—that is, the total consideration payable by the developer to the owner. The other component was the Total Land Value (“TLV”), that is the aggregate of (i) the TRLV already mentioned; (ii) the Total Commercial Land Value (“TCLV”) and (iii) the Total Residential Car Parking Land Value (“TRCPLV”). Under para 3 of schedule 6 the TLV was payable by instalments over 52 months, starting nine months after the grant of planning permission, and the ARP was payable on completion of the last residential sale or six months after the completion of the development, whichever was the earlier.
77. Each of the three components of the TLV was defined by a formula. The TRLV was to be computed by reference to the net internal area of the residential units for which planning permission was obtained at the rate of Â£76.34 per square foot (“less the section 106 money and less the rights of light money and less the sub-structure assumptions additional cost”). The TCLV was to be computed by reference to the net internal area of the commercial premises for which planning permission was obtained, at the rate of Â£38.80 per square foot. The TRCPLV was to be Â£3,024 multiplied by the number of residential parking places for which planning permission was obtained.
78. Both parties were experienced in the property world—the owner as a land dealer, the developer as a developer—and they shared the knowledge that the site (including units 1 and 3 Hardwicks Way which the owner acquired during the negotiations), with the benefit of planning permission on favourable terms, would have a market value in the general region of Â£5m. They hoped that planning permission for residential development would be granted for up to 100 flats with an aggregate internal area of 50,000 to 60,000 square feet. The background facts known to the parties included the recent takeover by the developer’s group of Beazer Homes Ltd, as a result of which the developer decided that it could not afford an outright purchase of the Wandsworth site. Instead the purchase was to be funded out of the proceeds of the disposal of the development, and the owner would expect to be compensated for the deferment of its consideration.
79. The definitions that I have already mentioned, and others, such as Costs and Incentives (“CandI”), that I have yet to come to, can be quite confusing. It is important, I think, to discern and keep in mind the general structure of schedule 6 of the agreement. The components of the price referable to the commercial development (TCLV) and the residential parking (TRCPLV) were to be calculated under the simple formulae already mentioned (together they eventually amounted to about Â£1.727m). By contrast, the price for the residential development was to consist of two elements, the TRLV and the ARP. The TRLV is agreed to be approximately Â£4.684m, reflecting the approved residential internal area of rather over 61,000 square feet at Â£76.34 per square foot. The question is how much this sum has to be increased by the ARP to make up the owner’s total consideration for the residential development.
80. The ARP is defined as follows:
“23.4% of the price achieved for each Residential Unit in excess of the Minimum Guaranteed Residential Unit Value ['MGRUV'] less the [CandI]”
The amount of the CandI is agreed to have been relatively trivial - a little less than Â£250,000 for all 100 flats - and I put it aside for the moment, while recognising that it plays an important part in the technicalities of the argument on construction. If this item is disregarded for the moment the disputed text can be set out in a simplified form, using “RP” (for residential price) where the judge and the Court of Appeal referred to Unit Price:
“23.4% of the RP in excess of the MGRUV".
81. The MGRUV was defined as meaning:
“for each Residential Unit [ie each flat] the [TRLV] divided by the number of Residential Units for which Planning Permission is granted".
Under the planning permission eventually granted on 23 August 2002 there were to be exactly 100 flats, which simplifies the arithmetic. Nevertheless it may be unfortunate that the draftsman chose to define the ARP by a formula referring to the MGRUV rather than by referring directly to the TRLV (to which the MGRUV is directly linked, being, as events turned out, one per cent of it). The use of the two linked formulae rather than one, and the fact that the formulae are not set out in mathematical notation, make it harder to keep clearly in mind the structure of the arrangements contained in schedule 6.
82. Briggs J expressed the issue in paras 20 and 21 of his judgment:
“Leaving aside for the moment the point at which the CandI are deducted, the broad commercial effect of each of the parties’ rival submissions may be summarised as follows. Chartbrook’s case was that it was entitled to a 23.4% share of the net proceeds of sale of each Residential Unit in excess of a minimum guaranteed amount (being the unitised Total Residential Land Value of Â£76.34 per square foot of Residential Net Internal Area). Put another way, its stake in the residential part of the development was to be the whole of the first Â£76.34 per square foot of net sales value, and 23.4% of the surplus.
By contrast, Persimmon’s case was that Chartbrook was to receive an additional payment only if 23.4% of the net sales price amounted to more than the Minimum Guaranteed Residential Unit Value. Put more broadly, Chartbrook’s stake in the residential part of the development was whichever was the greater of :
(i) 23.4% of the net residential sales price; and,
(ii) the guaranteed minimum of Â£76.34 per square foot of Residential Net Internal Area.”
83. That is, with great respect to the judge, a confusing way of putting it, because it fails to distinguish between the two elements of the price for the residential development and to make clear whether it is addressing both elements, or only the ARP. The owner was to get the TRLV in any event, as the most important component of the TLV (a point that may be reflected in the expression “in excess of” in the definition of the ARP). Indeed paras 20 and 21 of the judgment are to my mind two ways of describing the same result, unless the judge intended para 20 to state the owner’s view of the ARP alone, and para 21 to state the developer’s view of the TRLV and the ARP operating in conjunction.
84. In his brief judgment Rimer LJ quoted the definition of the ARP and observed in para 183:
“There is nothing unclear, uncertain or ambiguous about that. It is clear, certain and unambiguous and its arithmetic is straightforward".
Tuckey LJ agreed. With profound respect to both of them, I totally disagree. The definition is obviously defective as a piece of drafting. To start with defects that can be spotted and remedied fairly easily, the draftsman could not decide whether he was dealing with the flats (“each Residential Unit”) collectively or individually. The MGRUV was one-hundredth of the TRLV, but the CandI was plainly defined as a single aggregate figure.
85. Much more significantly and problematically, the draftsman has failed to notice the ambiguity of the formula “x per cent of the RP in excess of the MGRUV less the CandI". The ambiguity could be resolved by the use of mathematical notation, as the judge observed in paras 18 and 19 of his judgment (though he did not mention that there was also a choice to be made as to putting another set of brackets round MGRUV - CandI, so producing four possibilities rather than two).
86. Treated acontextually, the formula “x per cent of A in excess of B” is undoubtedly ambiguous. It can mean (x/100 x A)-B or x/100(A-B). If required to guess I would opt for the latter meaning, because the expression “in excess of” has been used rather than “less", and to my mind “in excess of” suggests a focus on B as an integer and distances it from the percentage. But I readily accept that this would be little more than guesswork.
87. In a contract negotiated between businessmen there always is a commercial context. If a contracting party agrees to pay the whole of some budgeted cost (B, say Â£100,000) and also agrees to pay 25% of the eventual actual cost (A, say Â£140,000) in excess of the budgeted costs, he would expect to pay a total of Â£110,000. Both elements of the obligation are, as it were, in the same currency, that is, cost, and the wording of the second element naturally translates into the formula Â¼ (A-B), as (Â¼A)-B would be commercial nonsense. The owner can therefore give plausible examples in which Â¼(A-B) would obviously be the right answer. But the present appeal is not such a case.
88. In this case the very significant difference between the results produced by the competing formulae is demonstrated in the figures set out in para 14 of the judgment of Lawrence Collins LJ in the Court of Appeal. The difference between the two bottom lines (Â£5,580,616 and Â£9,168,427) is Â£3,587,811. That difference figure is 76.6% of the TRLV (Â£4,683,565). On the owner’s case it gets one hundred times the MGRUV as the TRLV, but in effect has to give credit for only 23.4% of it in the calculation of the ARP. That seems to be a fairly surprising bargain for commercial men to make. It becomes not merely surprising but totally incredible if one takes account of the fact that although schedule 6 does not in terms state that the ARP is to have no value unless the actual receipts from sales of flats exceed the TRLV, that is implicit in its structure. On the owner’s construction any such limitation is contradicted, and the ARP has a substantial value even if the sales do not reach the “trigger point” of Â£20.015m (23.4% of which is Â£4.684m).
89. This can be illustrated by considering the effect of the competing formulae (set out in paras 18 and 19 of the judge’s judgment, but continuing to exclude CandI for the present) for assumed residential sale price totals (RP) of Â£18m, Â£20m, Â£20.015m (the trigger point), Â£22m and Â£23.849m (the actual result):
Owner’s Construction Developer’s Construction
23.4%(RP-MG) = ARP (23.4%RP)-MG = ARP
18.000 23.4%(18.000-4.684) = 3.116 (23.4% x 18.000)-4.684 = Negative
20.000 23.4% (20.000-4.684) = 3.584 (23.4% x 20.000)-4.684 = Negative
20.015 23.4%(20.015-4.684) = 3.587 (23.4%x 20.015)-4.684 = Zero
22.000 23.4%(22.000-4.684) = 4.052 (23.4% x 22.000)-4.684 = 0.464
23.849 Actual Result 23.4%(23.849-4.684) = 4.485 (23.4% x 23.849)- 4.684 = 0.897
RP = Total actual receipts from residential sales
MG = Total Residential Land Value (100 x MGRUV)
All amounts in Â£m
The figures are, on the owner’s construction, commercial nonsense. They would bring the owner a total of Â£7.8m for the residential development (Â£4.684m + Â£3.116m) even if sales of flats were disastrously low at Â£18m. The idea of the TRLV (linked as it is to the MGRUV) as a guaranteed minimum would be totally subverted.
90. The judge accepted that there was some force in the developer’s reliance on the words “if any” which occur in para 3.3 of schedule 6 with reference to the Balancing Payment (alias the ARP). He saw less force in the submission that he should give weight to the natural meaning of the expressions “minimum", “guaranteed” and “additional” in the definitions of the MGRUV and the ARP, on the ground that by the use of a special definition “[t]he word or phrase is stripped of its natural meaning"(para 61). In preferring the owner’s submissions the judge attached particular weight to the difficulty (for the developer) of explaining the words “less the CandI". The judge observed (paras 55-57):
“An equally serious problem with Persimmon’s construction is what to do with the subtraction of the CandI. It is common ground that the Costs and Incentives have a linear relationship with the amount of the price achieved for each Residential Unit. For example, Persimmon may agree the sale of a flat for Â£250,000 after incurring Costs and Incentives of say Â£50,000 or, with the same commercial consequence, sell the same flat for Â£200,000 but incur no Costs and Incentives. Typical Incentives would include payment of the purchaser’s legal fees or stamp duty, or the installation of special features such as wooden floors, over and above the standard fit-out specification.
One would expect Chartbrook’s profit share to be unaffected, one way or the other, by the decision of Persimmon to sell a particular flat by one or other of those methods (high price plus Incentives or low price without Incentives). Chartbrook’s construction, under which the Costs and Incentives are deducted from the price achieved for each Residential Unit before the application of the 23.4% share, fulfils precisely that expectation.
By contrast, Persimmon’s construction deducts the CandI from the 23.4% of the price achieved for the Residential Unit before the net amount is compared with the MGRUV, to ascertain whether there is any excess. By comparison with Chartbrook’s construction, that calculation magnifies the negative effect of CandI by a factor of more than 3 in comparison with the positive effect of the increase in the Residential Unit Price attributable to the CandI.”
91. Lawrence Collins LJ took a different view, and I unhesitatingly prefer his view. His reasons are set out clearly in paras 90 to 94 of his judgment, and I cannot usefully add much to them. But I would make a few further comments.
92. The first is as to the perceived problem about CandI. It is true that from the developer’s point of view it made little or no difference (except perhaps for timing and tax) whether or not, in relation to a particular flat, it spent an extra Â£2,000 on parquet flooring or granite worktops and managed to sell the flat for Â£2,000 more as a result. But it did make a marginal difference to the owner, as its ARP was calculated (in some way or other, on any view) by reference to the total achieved by residential sales: that is, by reference to the developer’s turnover and not by reference to the developer’s profit (the judge’s reference to “Chartbrook’s profit share” was not therefore entirely apposite). So (to adopt the expression used in the courts below) CandI had a linear relationship for the developer, but not for the owner. It would therefore have made sense for the parties to have agreed that CandI expenditure and allowances should be deducted from the total obtained for residential sales for the purposes of these computations (rather as the section 106 money and the rights of light money were to be disregarded in computing the TRLV). But it would not make sense to deduct the whole of the CandI from 23.4 per cent of the total obtained for residential sales.
93. Rimer LJ gave an example (para 185) to reinforce his view about the CandI. His example produces very odd results but that is partly because the figures taken are unrealistic. If one takes more realistic figures (such as a normal average sale price of Â£200,000 with CandI of Â£2,500 and an MGRUV of Â£47,500) the result is much less surprising. But it is still anomalous and makes no commercial sense, as Lawrence Collins LJ observed. An ARP of (23.4% of [RP less CandI]) less MGRUV does make commercial sense, and in my opinion it is well within the principles in Antaios Cia Naviera v Salen Rederiana AB  AC 191, 201 and Investors Compensation Scheme Ltd v West Bromwich Building Society  1 WLR 896, 912-913, to read the agreement in that way.
94. I am sure that Lawrence Collins LJ was right to give a lot of weight to the terms “minimum", “guaranteed” and “additional” in the relevant definitions. There is a good deal of authority, if authority is needed, to give weight to the natural meaning of words in a definition. In relation to statutory definitions there are the observations of my noble and learned friend, Lord Hoffmann, in Macdonald v Dextra Accessories Ltd  4 All ER 107, para 18 and Birmingham City Council v Walker  2AC 262, para 11 and Lord Scott of Foscote in Oxfordshire County Council v Oxford City Council  2 AC 674, paras 82-83. I would apply the same principle to a definition in a commercial contract.
95. That brings me back to what I said earlier about the need, in the midst of a thicket of rather confusing definitions, to keep in mind the general structure of the bargain. As part of the TLV the owner was to receive the TRLV, the total residential land value, representing the estimated value attributable to land which would (on the agreement becoming unconditional) have the benefit of a favourable planning permission for residential development (but on which development had not yet taken place). The developer was to bear all the costs of the development. The owner was also to have the prospect of an additional payment, the ARP. As regards the residential development the bargain could have been expressed between businessmen as “a guaranteed minimum of the first Â£76.34 per square foot of residential internal area from the total proceeds of the flats, and 23.4% of the excess” (indeed the judge, in para 20 of his judgment, summarised the owner’s case in very similar terms, except that he used “surplus” rather than “excess”). If one approaches it in that way, the developer’s case as to the meaning of the definition is not merely linguistically possible, but is linguistically (as well as commercially) compelling. The owner’s case becomes plausible only if one concentrates on the ARP, forgetting the TRLV (to which the MGRUV is directly linked). Lawrence Collins LJ dealt with this point quite briefly, in paras 81 and 93 of his judgment. He must have thought it unnecessary to spell it out more fully. But as he ended in the minority I have dealt with the point more fully.
96. Since preparing this opinion I have had the privilege of reading in draft the opinion of my noble and learned friend, Lord Hoffmann. In paras 1 to 22 of his opinion Lord Hoffmann reaches precisely the same conclusion as I have reached in regard to the correct construction, by traditional methods, of the agreement. I agree with all his reasoning, which is essentially the same as my own, but more trenchantly expressed. I have however thought it worthwhile setting out my own more pedestrian route to the conclusion that this House should, without having to depart from established principles of construction, allow the appeal and dismiss Chartbrook’s claim.
97. I have also read with interest and admiration Lord Hoffmann’s observations, in the remaining part of his opinion, on the important questions that we do not have to decide. I would not differ from any of these views. In particular, I agree that Karen Oltmann is a questionable application of the “private dictionary” principle, since the meaning of the English word “after” can hardly be equated to the use of a technical or trade term.
BARONESS HALE OF RICHMOND
98. I too have had the privilege of reading in draft the opinions of my noble and learned friends, Lord Hoffmann and Lord Walker of Gestingthorpe. For the reasons they give, together with those of Lawrence Collins LJ in the Court of Appeal, I agree that Persimmon’s construction of this contract is correct and that this appeal should be allowed.
99. But I have to confess that I would not have found it quite so easy to reach this conclusion had we not been made aware of the agreement which the parties had reached on this aspect of their bargain during the negotiations which led up to the formal contract. On any objective view, that made the matter crystal clear. This, to me, increased the attractions of accepting counsel’s eloquent invitation to reconsider the rule in Prenn v Simmonds  1 WLR 1381, the pot so gently but effectively stirred by Lord Nicholls of Birkenhead in his Chancery Bar Association lecture of 2005 ( 121 LQR 577). My experience at the Law Commission has shown me how difficult it is to achieve flexible and nuanced reform to a rule of the common law by way of legislation. In the end abolition may be the only workable legislative solution, as eventually happened with the hearsay rule (Law Com No 216 (1993), The Hearsay Rule in Civil Proceedings). Even that can prove difficult if, on analysis, the view is taken that the rule has no real content, as with the parol evidence rule (Law Com No 154 (1986), The Parol Evidence Rule). The courts, on the other hand, are able to achieve step-by-step changes which can distinguish cases in which such evidence is “helpful” from cases in which it is not.
100. However, the approach to rectification adopted by Lord Hoffmann would go a long way towards providing a solution. If the test of the parties’ continuing common intentions is an objective one, then the court is looking to see whether there was such a prior consensus and if so what it was. Negotiations where there was no such consensus are indeed “unhelpful". But negotiations where consensus was reached are very helpful indeed. If the language in the eventual contract does not reflect that consensus, then unless there has been a later variation of it, the formal contract should be rectified to reflect it. It makes little sense if the test for construing their prior consensus is different from the objective test for construing their eventual contract. This situation is, and should be, quite different from the situation where one party is mistaken as to its meaning and the other party knows this - the latter should not be permitted to take advantage of the former.
101. For those reasons, I would respectfully associate myself, as does Lord Walker, with the views of Lord Hoffmann on the issues which we do not have to decide. In particular, I would like to express my admiration for the skill and charm with which both issues were argued by counsel on each side. It is perhaps surprising that questions of such practical and theoretical importance in the law of contract should still be open to debate and development. But that is also the great strength of the common law.