(with whom Lord Clarke, Lord Sumption and Lord Carnwath agree)
A husband and wife each executed the will which had been prepared for the other owing to an oversight on the part of their solicitor; the question which arises is whether the will of the husband, who died after his wife, is valid.
The factual and procedural background
The factual background
On 17 May 1999, Alfred Rawlings and his wife, Maureen Rawlings, were visited by their solicitor to enable them to execute the wills which he had drafted on their instructions. The wills were short and, except for the differences required to reflect the identity of the maker, they were in identical terms. Each spouse left his or her entire estate to the other, but, if the other had already died or survived the deceased spouse for less than a month, the entire estate was left to the appellant, Terry Marley, who was not related to them but whom they treated as their son.
The will prepared for Mr Rawlings was in these terms:
This is the last will of me ALFRED THOMAS RAWLINGS of 15A Hillcrest Road Biggin Hill Kent TN16 3UA
IN WITNESS whereof I the said ALFRED THOMAS RAWLINGS have hereunto set my hand the .... day of .... 1999: ....
SIGNED by the testator in our presence and then by us in his:
Signature, name, address .... Signature, name, address .... of attesting solicitor: of attesting secretary: ....
The will prepared for Mrs Rawlings was in identical terms save that it was, of course, in her name instead of that of her husband, so that "ALFRED THOMAS RAWLINGS" was replaced by "MAUREEN CATHERINE RAWLINGS", and "my [said] wife MAUREEN CATHERINE RAWLINGS", "her", "his", and "testator" were respectively replaced by "my [said] husband ALFRED THOMAS RAWLINGS", "him", "her", and "testatrix".
By an oversight (which he candidly admitted in his witness statement in these proceedings), the solicitor gave each spouse the other's draft will, and nobody noticed. Accordingly, Mr Rawlings signed the will meant for Mrs Rawlings, and Mrs Rawlings signed that meant for Mr Rawlings, and the solicitor and his secretary attested the signature on each document, which was then dated 17 May 1999.
Mrs Rawlings died in 2003, and her estate passed to her husband without anyone noticing the mistake. However, when Mr Rawlings died in August 2006, the error came to light.
At the time of his death, Mr Rawlings was a joint tenant with the appellant of the house in which they both lived, so the tenancy passed to the appellant through the doctrine of survivorship. In addition, there was some £70,000 in Mr Rawlings's estate.
The respondents, Terry and Michael Rawlings, Mr and Mrs Rawlings' two sons, challenged the validity of the will which Mr Rawlings had signed. If it was valid, the appellant would inherit the £70,000 under its terms, whereas if it was invalid, Mr Rawlings would have died intestate, and the respondents would inherit the £70,000.
The procedural background
The appellant began probate proceedings, which came before Proudman J. She gave a judgment based on the understanding that his case was that Mr Rawlings's will ("the Will") should be rectified so as to record what he had intended, ie so as to contain what was in the will signed by his wife ("the wife's Will"), and that probate should be granted of the Will as so rectified.
The Judge dismissed Mr Marley's claim, on the grounds that (i) the Will did not satisfy the requirements of section 9 of the Wills Act 1837 ("the 1837 Act"), and (ii) even if it had done so, it was not open to her to rectify the Will under section 20 of the Administration of Justice Act 1982 ("the 1982 Act") -  1 WLR 2146.
The appellant appealed to the Court of Appeal, who upheld the decision of Proudman J on the first ground, namely that the Will did not satisfy section 9(b) of the 1837 Act (as well on at least one other ground), and they did not find it necessary to consider the second ground  Ch 271.
The appellant now appeals to this court.
The legal background
There are, unsurprisingly, a large number of cases in which courts have had to consider the validity of a will and the interpretation of a will, and a few cases where rectification of a will has been considered. The formalities have for a long time largely been laid down by the 1837 Act. By contrast, until very recently at any rate, the interpretation and possible rectification of wills was an issue which Parliament was content to leave to the judges.
The formal requirements of a will
So far as validity is concerned, the centrally important statutory provision, both in general terms and for present purposes, is section 9 of the 1837 Act ("section 9"). That section has been amended or re-enacted on a number of occasions. Most recently, it was re-enacted by section 17 of the 1982 Act, which is headed "Relaxation of formal requirements for making wills".
In its current form, section 9 is headed "Signing and attestation of wills", and it provides as follows:
No will shall be valid unless
In addition to these statutory requirements, as Chadwick LJ explained in Fuller v Strum  1 WLR 1097, para 59:
It is not, and cannot be, in dispute that, before admitting the document to probate, the judge needed to be satisfied that it did truly represent the testator's testamentary intentions; or, to use the traditional phrase, that the testator 'knew and approved' its contents. Nor is it in dispute that, if satisfied that the testator knew and approved of part only of the contents of the document, the judge was bound, before admitting the document to probate, to require that those parts with respect to which he was not so satisfied be struck out.
Interpretation of wills
Until relatively recently, there were no statutory provisions relating to the proper approach to the interpretation of wills. The interpretation of wills was a matter for the courts, who, as is so often the way, tended (at least until very recently) to approach the issue detached from, and potentially differently from, the approach adopted to the interpretation of other documents.
During the past forty years, the House of Lords and Supreme Court have laid down the correct approach to the interpretation, or construction, of commercial contracts in a number of cases starting with Prenn v Simmonds  1 WLR 1381 and culminating in Rainy Sky SA v Kookmin Bank  1 WLR 2900.
When interpreting a contract, the court is concerned to find the intention of the party or parties, and it does this by identifying the meaning of the relevant words, (a) in the light of (i) the natural and ordinary meaning of those words, (ii) the overall purpose of the document, (iii) any other provisions of the document, (iv) the facts known or assumed by the parties at the time that the document was executed, and (v) common sense, but (b) ignoring subjective evidence of any party's intentions. In this connection, see Prenn at 1384-1386 and Reardon Smith Line Ltd v Yngvar Hansen-Tangen 1 WLR 989, per Lord Wilberforce, Bank of Credit and Commerce International SA v Ali  1 AC 251, para 8, per Lord Bingham, and the survey of more recent authorities in Rainy Sky, per Lord Clarke at paras 21-30.
When it comes to interpreting wills, it seems to me that the approach should be the same. Whether the document in question is a commercial contract or a will, the aim is to identify the intention of the party or parties to the document by interpreting the words used in their documentary, factual and commercial context. As Lord Hoffmann said in Kirin-Amgen Inc v Hoechst Marion Roussel Ltd 1 All ER 667, para 64, "No one has ever made an acontextual statement. There is always some context to any utterance, however meagre." To the same effect, Sir Thomas Bingham MR said inArbuthnott v Fagan  CLC 1396, that "[c]ourts will never construe words in a vacuum".
Of course, a contract is agreed between a number of parties, whereas a will is made by a single party. However, that distinction is an unconvincing reason for adopting a different approach in principle to interpretation of wills: it is merely one of the contextual circumstances which has to be borne in mind when interpreting the document concerned. Thus, the court takes the same approach to interpretation of unilateral notices as it takes to interpretation of contracts see Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd  AC 749, per Lord Steyn at 770C-771D, and Lord Hoffmann at 779H-780F.
Another example of a unilateral document which is interpreted in the same way as a contract is a patent see the approach adopted by Lord Diplock in Catnic Components Ltd v Hill & Smith Ltd RPC 183, 243, cited with approval, expanded, and applied in Kirin-Amgen at paras 27-32 by Lord Hoffmann. A notice and a patent are both documents intended by its originator to convey information, and so, too, is a will.
In my view, at least subject to any statutory provision to the contrary, the approach to the interpretation of contracts as set out in the cases discussed in para 19 above is therefore just as appropriate for wills as it is for other unilateral documents. This may well not be a particularly revolutionary conclusion in the light of the currently understood approach to the interpretation of wills (see eg Theobald on Wills, 17th edition, chapter 15 and the recent supplement supports such an approach as indicated in RSPCA v Shoup  1 WLR 980 at paras 22 and 31). Indeed, the well known suggestion of James LJ in Boyes v Cook (1880) 14 Ch D 53, 56, that, when interpreting a will, the court should "place [itself] in [the testator's] arm-chair", is consistent with the approach of interpretation by reference to the factual context.
However, there is now a highly relevant statutory provision relating to the interpretation of wills, namely section 21 of the 1982 Act ("section 21"). Section 21 is headed "Interpretation of wills general rules as to evidence", and is in the following terms:
In my view, section 21(1) confirms that a will should be interpreted in the same way as a contract, a notice or a patent, namely as summarised in para 19 above. In particular, section 21(1)(c) shows that "evidence" is admissible when construing a will, and that that includes the "surrounding circumstances". However, section 21(2) goes rather further. It indicates that, if one or more of the three requirements set out in section 21(1) is satisfied, then direct evidence of the testator's intention is admissible, in order to interpret the will in question.
Accordingly, as I see it, save where section 21(1) applies, a will is to be interpreted in the same way as any other document, but, in addition, in relation to a will, or a provision in a will, to which section 21(1) applies, it is possible to assist its interpretation by reference to evidence of the testator's actual intention (eg by reference to what he told the drafter of the will, or another person, or by what was in any notes he made or earlier drafts of the will which he may have approved or caused to be prepared).
Rectification of wills
Rectification is a form of relief which involves "correcting a written instrument which, by a mistake in verbal expression, does not accurately reflect the [parties'] true agreement" The Nai Genova 1 Lloyd's Rep 353, 359. It is available not only to correct a bilateral or multilateral arrangement, such as a contract, but also a unilateral document, such as a settlement - see In re Butlin's Settlement Trusts  Ch 251. However, it has always been assumed that the courts had no such power to rectify a will see eg Harter v Harter (1873) LR 3 P&D 11 per Hannen P, and In re Reynette-James decd  1 WLR 161, per Templeman J.
As at present advised, I would none the less have been minded to hold that it was, as a matter of common law, open to a judge to rectify a will in the same way as any other document: no convincing reason for the absence of such a power has been advanced. However, it is unnecessary to consider that point further, as Parliament has legislated on the topic, in section 20 of the 1982 Act ("section 20").
Section 20 is headed "Rectification", and subsection (1) provides as follows:
If a court is satisfied that a will is so expressed that it fails to carry out the testator's intentions, in consequence -
it may order that the will shall be rectified so as to carry out his intentions.
Section 20(2) provides that, save with the court's permission, no application for rectification under subsection (1) can be made more than six months after the grant of probate. Section 20(3) protects executors who distribute in accordance with the terms of a will before it is rectified after the six-month period referred to in subsection (2).
Mr Ham QC, for the appellant, realistically accepted that it would be inappropriate for the court to hold that it had wider powers to rectify a will than those which were conferred by section 20. Given that Parliament decided to confer a limited power of rectification at a time when there was clear authority that the court had no inherent power to rectify, it would be wrong for any court to hold, at least in the absence of a compelling reason, that it actually had an inherent power which was wider than that which the legislature conferred.
The issues on this appeal
The appellant rested his case on three different contentions.
The first was that Mr Rawlings's Will, properly interpreted, should be read, in effect, as if it was the document signed by his wife on 17 May 1999.
The second contention was that the extent of Mr Rawlings's knowledge and approval of the contents of the Will was such that it could be validated, albeit with deletions.
The third contention was that the Will should be rectified so as to accord with Mr Rawlings's intentions.
I shall consider those contentions in turn.
Although Mr Ham primarily based his contention that the Will was valid on the ground of rectification (which was the sole basis on which the case was considered in the courts below), he accepted that the interpretation argument ought to be considered first, and the deletions argument second.
The appellant's contention on interpretation
The argument that the Will, properly interpreted, is valid and effective is based on two propositions. The first is that the Will can be read together with the wife's Will, given that it is clear from the face of the two documents that they were signed on the same date, by a cohabiting husband and wife, and were in very similar terms and in the same style, and had the same witnesses. While not mutual wills (ie separate wills entered into pursuant to an agreement between the two testators as to the terms of their wills), they were clearly closely related, and therefore each could properly be looked at when interpreting the other. The second proposition is that, when one looks at the two documents, it is obvious what has happened, and in particular it is obvious that Mr Rawlings intended the Will to be in the form of the wife's Will. Accordingly, runs the appellant's case, that is how the Will should be interpreted and read.
For the respondents, Mr Le Poidevin QC realistically does not challenge the basis of this argument, namely that the two documents can be read together, and that, on that basis, it is clear what happened and what was intended by Mr Rawlings. However, he contends that this exercise is not one of interpretation at all, but one of rectification.
This contention raises a point of some potential importance and difficulty. In Investors Compensation Scheme Ltd v West Bromwich Building Society  1 WLR 896, 912H-913E, Lord Hoffmann set out the principles which the court should apply when interpreting documents in five propositions. Most of the content of that passage is unexceptionable, although, in one or two places, the language in which the propositions are expressed may be a little extravagant; thus, the words "absolutely anything" in his second proposition required some qualification from Lord Hoffmann in Bank of Credit and Commerce, para 39.
However, the second sentence of Lord Hoffmann's fifth proposition in Investors Compensation is controversial. That sentence reads, so far as relevant, "....if one would nevertheless conclude from the background that something must have gone wrong with the language, the law does not require judges to attribute to the parties an intention which they plainly could not have had".
Lord Hoffmann took that approach a little further in Chartbrook Ltd v Persimmon Homes Ltd  1 AC 1101, paras 21-25. Having observed that the exercise of interpretation involves "decid[ing] what a reasonable person would have understood the parties to have meant by using the language which they did" and referring to the "correction of mistakes by construction", he said this:
[T]here 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 a forcefully expressed article, "Construction" and Rectification after Chartbrook  CLJ 253, Sir Richard Buxton has suggested that Lord Hoffmann's approach to interpretation in these two cases is inconsistent with previously established principles. Lewison on The Interpretation of Contracts (fifth ed (2011), para 9.03, footnote 67, in an illuminating chapter dealing with mistakes) suggests that Sir Richard has made out "a powerful case for the conclusion that the difference between construction and rectification has reduced almost to vanishing point", if Lord Hoffmann's analysis is correct.
At first sight, it might seem to be a rather dry question whether a particular approach is one of interpretation or rectification. However, it is by no means simply an academic issue of categorisation. If it is a question of interpretation, then the document in question has, and has always had, the meaning and effect as determined by the court, and that is the end of the matter. On the other hand, if it is a question of rectification, then the document, as rectified, has a different meaning from that which it appears to have on its face, and the court would have jurisdiction to refuse rectification or to grant it on terms (eg if there had been delay, change of position, or third party reliance). This point is made good in relation to wills by the provisions of section 20(2) and (3).
In my judgment, unless it is necessary to decide this difficult point, we should not do so on this appeal. Interpretation was not the basis upon which the courts below decided this case and it was not the ground upon which Mr Ham primarily relied. Furthermore, and no doubt because of those points, only limited argument was directed to the issue of whether the issue was one of interpretation or of rectification. For the reasons developed below, I consider that this appeal succeeds on the ground of rectification, so I shall proceed on the basis that it fails on interpretation.
It should be added that Mr Ham also relied on section 21(2). I do not think that it can take his case any further, although it would enable him to rely on Mr Rawlings's subjective intention, because his argument is still one based on interpretation. This point was made in In re Williams decd  1 WLR 905, 911G-H, where Nicholls J seems to have taken an orthodox view of interpretation. He said that "if, however liberal may be the approach of the court, the meaning is one which the word or phrase cannot bear, I do not see how .... the court can declare that meaning to be the meaning of the word or phrase", and "varying or contradicting the language used, would amount to re-writing", which is "to be achieved, if at all, under the rectification provisions in section 20".
The appellant's contention on deletions
The appellant's case under this head rests on two propositions. The first is that, in order to be a valid will, the testator must have known and approved of its contents see Fuller quoted in para 16 above. There is a rebuttable presumption that the testator knew and approved the contents of a regularly executed will with unexceptional provisions. However, that presumption may be rebutted by evidence of the circumstances in which the will was prepared or executed. It can also be rebutted where the will is so worded as to cast doubt on whether the testator can have known or approved of its contents. In the present case, the Will, as literally interpreted, plainly did not represent Mr Rawlings's intentions: accordingly, he cannot have known or approved of its contents, as it stood.
The second proposition invoked in the present connection is that, where the testator did not know or approve of only part of a will, that part can be notionally excised by the court, with the remainder being valid and admitted to probate as described in the last sentence quoted from Fuller in para 16 above. Examples of such cases are cited in Theobald, op cit, para 3-028.
On this basis, Mr Ham ingeniously argued that the Will can be validated by deleting (i) the opening sentence, (ii) clause 2, (iii) the first phrase of clause 3, and (iv) the reference to Mrs Rawlings at the end of the Will. If this were permissible, it would simply leave the Will as stating that the signatory, Mr Rawlings, revokes his previous wills and leaves his entire estate to the appellant.
In my view, this argument must be rejected. The most typical case where only part of a will is rejected on the ground that it was not known and approved by the testator, is where that part is self-contained eg a particular clause or subclause. One such example is in In the Goods of Oswald (1874) LR 3 P&D 162, 164, per Sir James Hannen P. However, it is also true that, in some cases, a simple word or expression can be deleted "if shewn to have been inserted by mistake" per Jeune J in In the Goods of Boehm  P 247, 250.
However, it is quite inappropriate to invoke this principle in order to justify selecting phrases and provisions for deletion from a will intended to be signed by someone else, to enable the will, effectively by happenstance, to comply with the testator's intentions. I note that Sir James Hannen P and Sir Gorell Barnes P took the same view in, respectively, In the Goods of Hunt (1875) LR 3 P&D 250, at 252, and In re Meyer  P 353, 354. Further, as Jeune J pointed out in Boehm at 251, there is obvious "difficulty [in] rejecting words where their rejection alters the sense of those which remain".
The appellant's proposed exercise in deletion summarised in para 45 above would involve converting what is a simple and beneficial principle of severance into what is almost a word game with haphazard outcomes. That is well illustrated by the fact that, in this case, the suggested deletions from the Will only achieve the intended result because Mrs Rawlings pre-deceased her husband, because clause 2 is deleted: therefore, if Mr Rawlings had pre-deceased his wife, this argument would not work.
I would accordingly reject the argument that the Will can be treated as a valid will by making the deletions suggested on behalf of the appellant.
The appellant's contention on rectification: introduction
The principal ground upon which the appellant contended that the Will should be held to be valid was that it should be rectified pursuant to section 20, so that it had the effect which Mr Rawlings intended, namely that it essentially stated what was in the wife's Will.
As I see it, three possible objections may be raised to this contention. The first is that the correction which needs to be made to validate the Will is too extreme to amount to rectification. The second is that section 20 only applies to a "will", and, because the Will, as executed, does not satisfy section 9 and/or because it was not executed with Mr Rawlings's knowledge and approval of its contents, it was not a "will", and therefore section 20 cannot be invoked. The third argument is that the rectification cannot be justified under either para (a) or para (b) of section 20(1). I shall consider those arguments in turn.
The appellant's contention on rectification: is it rectification?
The first argument did not really figure in the reasoning of the courts below or, unless it was impliedly subsumed in the third argument, in Mr Le Poidevin's submissions. Either way, without calling into question the third argument for the moment, I consider that the first argument should be rejected. The fact that it can be said that the claimed correction would effectively involve transposing the whole text of the wife's Will into the Will does not prevent it from being "rectification" of each of the Wills.
As a general proposition, there may be force in the point that the greater the extent of the correction sought, the steeper the task for a claimant who is seeking rectification. However, I can see no reason in principle why a wholesale correction should be ruled out as a permissible exercise of the court's power to rectify, as a matter of principle. On the contrary: to impose such a restriction on the power of rectification would be unprincipled - and it would also lead to uncertainty.
Subject to the other two points, the present circumstances seem to give rise to a classic claim for rectification. As Black LJ, who gave the leading judgment in the Court of Appeal, observed in para 7, "[t]here can be no doubt as to what Mr and Mrs Rawlings wanted to achieve when they made their wills and that was that [the appellant] should have the entirety of their estate and that [the respondents] should have nothing" (subject, of course, to the survivor enjoying the entirety of their property until his or her death). Thus, there is certainty as to what Mr Rawlings wanted, and there is certainty as to how he would have expressed himself (as there can be no doubt that he would have signed the will prepared for him if he had appreciated the mistake). Accordingly, this is a very clear case for rectification subject always to the two other points raised by the respondents.
The appellant's contention on rectification: is the document a "will"?
That brings me to Mr Le Poidevin's second argument, which impressed both Proudman J and the Court of Appeal. Black LJ, with whom Sir John Thomas P and Kitchin LJ agreed, considered that the Will was not a "will" for the purposes of section 20, because (i) (at least arguably) it failed to satisfy section 9(a), (ii) it failed to satisfy section 9(b), and (iii) it was not made with the knowledge and approval of Mr Rawlings; and that therefore it could not be rectified.
As already indicated in para 43 above, I accept that, on the basis that it must be interpreted at face value, the Will was plainly not executed by Mr Rawlings with his full knowledge and approval. However, I have been persuaded by Mr Ham that it did not fall foul of section 9(a) or (b).
While it is clear, even on a cursory reading of the Will, that something has gone seriously wrong, it is unchallengeable that Mr Rawlings signed it, and that he did so, both on the face of the document, and as a matter of fact, with the intention of it being his last will and testament. Thus, whatever else may be said about the document, it is, on its face (and was in fact according to the evidence), unambiguously intended to be a formal will, and it was, on its face (and was in fact according to the evidence), signed by Mr Rawlings, in the presence of two witnesses, on the basis that it was indeed his will.
It is important to bear in mind that section 9 is concerned with formalities. The fact that it is pretty clear from the provisions which it contains that a will may well face problems in terms of interpretation or even validity does not mean that it cannot satisfy the formality requirements. In that connection, it is worth referring to what Lord Wilberforce said in In re Resch's Will Trusts  1 AC 514, 547E, where (approving what had been said by Luxmoore J in In re Hawksley's Settlement  Ch 384, 395-396) he discussed the difference between the function of the court when considering whether to admit a will to probate and the function of the court if it subsequently interprets the will. As he explained, "[t]he fact that a document has been admitted to probate .... does not prevent a court of construction from coming to the conclusion that this document has no operative effect".
It is true that the Will purports in its opening words to be the will of Mrs Rawlings, but there is no doubt that it cannot be hers, as she did not sign it; as it was Mr Rawlings who signed it, it can only have been his will, and it is he who is claimed in these proceedings to be the testator for the purposes of section 9. Accordingly, section 9(a) appears to me to be satisfied. It is true that the Will does not make sense, at least if taken at face value, but that is a matter for "a court of construction", as Lord Wilberforce explained. There can be no doubt, however, from the face of the Will (as well as from the evidence) that it was Mr Rawlings's intention at the time he signed the Will that it should have effect, and so it seems to me that section 9(b) was also satisfied in this case.
Notwithstanding the fact that the contents of the Will, unless rectified, did not satisfy the requirement that they had the full knowledge and approval of Mr Rawlings, and even if the Court of Appeal had been right in their view that the Will did not satisfy the requirements of section 9(b) or (possibly) section 9(a), I consider that it would still be open for the appellant to invoke section 20. In other words, it does not appear to me that a document has to satisfy the formal requirements of section 9, or of having the testator's knowledge and approval, before it can be treated as a "will" which is capable of being rectified pursuant to section 20.
Black LJ said at para 39 that "the logical place to start indeed, it seems to me the only place to start is with the question of the formal validity of the will", and, only if it was formally valid would it be open to the court to consider whether to rectify it. In terms of academic linguistic logic, I see the force of that point, but it appears to me to be wrong for a number of reasons.
First, the approach adopted by the Court of Appeal takes away much of the beneficial value of section 20. If it could not be invoked to rectify a document which was currently formally invalid into a formally valid will, that would cut down its operation for no apparently sensible reason.
Secondly, it seems to me to be equally logical, but plainly more consistent with the evident purpose of the amendments made to the law of wills by sections 17 (which contains the new section 9) and 21 of the 1982 Act, to deal with the validity and rectification issues together, at least in a case such as this, where the two issues are so closely related.
Thirdly, the observation of Lord Wilberforce, quoted in para 58 above, demonstrates that a document which subsequently turns out to be invalid as a will can be, and no doubt frequently is, admitted to probate. Thus, even in the context of an entirely traditional approach, there is no objection to treating a document which purports to be a will as a will, even though it may subsequently turn out to be invalid.
Fourthly, while it would be wrong to express this as an exclusive definition (although it may be), it appears to me that the reference to a will in section 20 means any document which is on its face bona fide intended to be a will, and is not to be limited to a will which complies with the formalities. Indeed, the opening words of section 9 itself seem to use the word "will" to include a purported will which does not comply with the requirements of section 9(a) to (d). It provides that "no will" shall be valid unless it so complies, which clearly carries with it the irresistible implication that a document that does not so comply is none the less a "will" for the purposes of the section, but not a valid will.
Even if that were not right, as a matter of statutory interpretation I can see no reason why the word "will" in section 20(1) could not be read as meaning a document which, once it is rectified, is a valid will. After all, rectification operates retrospectively see eg per Lord Sterndale MR and Warrington LJ in Craddock Brothers v Hunt  2 Ch 136, 151 and 160.
Fifthly, in another area of the law where formalities are required for validity, land contracts, rectification was permitted even where it had had the effect of converting an ineffective (albeit not an invalid) contract into an enforceable contract: see Domb v Isoz  Ch 548, 559A-C per Buckley LJ, with whose reasoning Bridge and Templeman LJJ agreed. (That case was concerned with section 40 of the Law of Property Act 1925, which has now been replaced by section 2(1) of the Law of Property (Miscellaneous Provisions) Act 1989, which, in subsection (4) specifically envisages a contract which does not comply with subsection (1) being rectified so that it does.)
The appellant's contention on rectification: is it a clerical error?
The final issue raised by the appellant's rectification claim is whether it is within the ambit of section 20(1). It is not suggested that the claim falls within para (b), "a failure to understand [the testator's] instructions", but Mr Ham argued that it is within para (a), "a clerical error". There is no doubt that there was an error. The question is whether it can be said to be "clerical". Proudman J concluded that it could not, and the Court of Appeal did not determine the point.
It is clear that, owing to the solicitor's error in muddling the two draft wills, the contents of the Will except for three signatures and details of the witnesses, that is the opening words, the three operative clauses and the declaration at the end, were wrongly included in the document signed by Mr Rawlings, as they were intended for Mrs Rawlings's will. Accordingly, if they are to comply with Mr Rawlings's intention, they should be replaced by the equivalent provisions of the wife's Will. The question is whether this can properly be achieved under section 20(1)(a).
The meaning and ambit of section 20(1)(a) has been considered in a number of cases at first instance, which are helpfully discussed in Hodge on Rectification (first ed (2010), paras 7-37 to 7-46). Those cases, like the present case, require one to consider what sort of error constitutes "a clerical error" for the purposes of section 20(1)(a). However none of those cases involves the sort of error which arose in this case, although they do provide some insights into the problem raised here.
The best judicial summary of the effect of the cases so far decided on section 20(1)(a) was given by Blackburne J in Bell v Georgiou  EWHC 1080 (Ch) (quoted in para 7-42 of Hodge op cit):
The essence of the matter is that a clerical error occurs when someone, who may be the testator himself, or his solicitor, or a clerk or a typist, writes something which he did not intend to insert or omits something which he intended to insert. .... The remedy is only available if it can be established not only that the will fails to carry out the testator's instructions but also what those instructions were.
If, as a result of a slip of the pen or mistyping, a solicitor (or a clerk or indeed the testator himself) inserts the wrong word, figure or name into a clause of a will, and it is clear what word, figure or name the testator had intended, that would undoubtedly be a clerical error which could be rectified under section 20(1)(a). It is hard to see why there should be a different outcome where the mistake is, say, the insertion of a wrong clause because the solicitor cut and pasted a different provision from that which he intended. Equally, if the solicitor had cut and pasted a series of clauses from a different standard form from that which he had intended, I do not see why that should not give rise to a right to rectify under section 20(1)(a), provided of course the testator's intention was clear.
Accordingly, the notion that a wholesale replacement of the provisions of a will is permissible under section 20(1)(a) is demonstrated by the fact that it is difficult both as a matter in principle, and also in practice, to see where the line should otherwise be drawn.
However, Mr Le Poidevin contended that, even if a slip of the pen, a mistyping, or a failure to cut and paste correctly, which extend to virtually the whole of the document, can all be characterised as "clerical errors", giving the testator the wrong will is a mistake of a rather different character, which cannot naturally be referred to as a clerical error.
I accept that the expression "clerical error" can have a narrow meaning, which would be limited to mistakes involved in copying or writing out a document, and would not include a mistake of the type that occurred in this case. However, the expression is not one with a precise or well-established, let alone a technical, meaning. The expression also can carry a wider meaning, namely a mistake arising out of office work of a relatively routine nature, such as preparing, filing, sending, organising the execution of, a document (save, possibly, to the extent that the activity involves some special expertise). Those are activities which are properly be described as "clerical", and a mistake in connection with those activities, such as wrongly filing a document or putting the wrong document in an envelope, can properly be called "a clerical error".
For present purposes, of course, "clerical error" is an expression which has to be interpreted in its context, and, in particular on the assumption that section 20 is intended to represent a rational and coherent basis for rectifying wills. While I appreciate that there is an argument for saying that it does nothing to discourage carelessness, it seems to me that the expression "clerical error" in section 20(1)(a) should be given a wide, rather than a narrow, meaning.
First, rectification of other documents (including unilateral documents) is not limited to cases of clerical error, however wide a meaning that expression is given. Accordingly, given that there is no apparent reason for a different rule for wills, it would appear appropriate that the grounds for rectification is as wide for wills as the words of section 20(1) can properly allow.
Secondly, there is no apparent limit on the applicability of section 20(1)(b), which supports the notion that section 20(1)(a) should not be treated as being of limited application. However, section 20(1)(b) also has a potential limiting effect on the ambit of section 20(1)(a), in the sense that section 20(1)(a) should not be given a meaning which significantly overlaps with, let alone subsumes, that of section 20(1)(b).
Thirdly, sections 17 to 21 of the 1982 Act are, as I see it, all aimed at making the law on wills more flexible and rendering it easier to validate or "save" a will than previously. Section 17, which re-enacts section 9, is concerned with the "relaxation" of formalities (see para 14 above); sections 18 and 19 introduce greater flexibility in relation to the effect of the testator's marriage and death of his issue; section 20 introduces rectification for the first time for wills, and section 21 permits the testator's subjective intention to be taken into account for the first time. The whole thrust of the provisions is therefore in favour of a broad interpretation of a provision such as section 20(1)(a).
Fourthly, I consider that the law would be somewhat incoherent if subtle distinctions led to very different results in cases where the ultimate nature of the mistake is the same. If a solicitor is drafting two wills, and accidentally cuts and pastes the contents of B's draft will onto what he thinks is A's draft will, and hands it to A, who then executes it as his will, that will would be rectifiable under section 20(1)(a), as the solicitor's mistake would, on any view, be a clerical error see paras 72 and 73 above. On the other hand, if the solicitor accidentally gives B's will to A to execute, and A executes it, that would not, on the respondents' case, be a clerical error and therefore rectification would not be available.
While I accept that fine distinctions can often lead to different outcomes where one is near the limits of the scope of some statutory provisions, a distinction of this sort seems to me to be capricious or arbitrary. The position is essentially the same in the two cases. In each case, it was because his solicitor accidentally handed A a document which contained B's will rather than A's will, that A executed B's will thinking that it was his will. In each case, the reason that the will which A executed did not represent his intentions was a silly mistake by the solicitor in the mechanics of faithfully carrying out his instructions. In neither case did the mistake involve the solicitor misunderstanding or mischaracterising the testator's intention or instructions, or making any error of law or other expertise, so the error may fairly be characterised as "clerical" and there is no question of trespassing into section 20(1)(b) territory.
As explained in para 75 above, the term "clerical error" can, as a matter of ordinary language, quite properly encompass the error involved in this case. There was an error, and it can be fairly characterised as clerical, because it arose in connection with office work of a routine nature. Accordingly, given that the present type of case can, as a matter of ordinary language, be said to involve a clerical error, it seems to me to follow that it is susceptible to rectification.
I accept that the error in this case is not within the narrower meaning of "clerical error", as is reflected by the approach to the expression summarised by Blackburne J in Bell as representing the effect of the first instance authorities. However, for the reasons given in paras 75-82 above, I have concluded that, the expression can, and, in the context of section 20(1)(a) should, be given its wider meaning, which covers the mistake made in this case.
For completeness, I should make two further points. First, in the course of argument, we were taken to parts of the Law Reform Committee's 19th Report (Interpretation of Wills) Cmnd 5301 (1973). It seems clear that much of Part IV of the 1982 Act stems from the 1973 Report. In my view, however, the Report does not help, because, while it gives an example of a clerical error, it does not spell out the intended limits of the expression. Further, it seems that, in enacting Part IV of the 1982 Act, Parliament did not give effect to the recommendations of the Report in their entirety.
Secondly, during our deliberations, we wondered what Scots law would make of the problem thrown up by this appeal. In that connection, it is instructive to read Lord Hodge's judgment. As frequently happens, the law north and south of the border each appear to have something to learn from the other, and to involve slightly different ways of arriving at the same outcome.
I would therefore allow this appeal, and hold that the Will should be rectified so that it contains the typed parts of the will signed by the late Mrs Rawlings in place of the typed parts of the will signed by Mr Rawlings.
I agree and confine myself to some observations on how Scots law might have dealt with the problem if it were the governing law.
The Law Reform (Miscellaneous Provisions) (Scotland) Act 1985 introduced the remedy of rectification of legal documents into Scots law. The 1985 Act implemented the recommendations of the Scottish Law Commission's report on rectification of contractual and other documents (Scot Law Com no. 79), which was published in 1983 shortly after the introduction of the rectification of wills into English law by section 20 of the Administration of Justice Act 1982. Section 8(6) of the 1985 Act excludes from the scope of the statutory remedy any "document of a testamentary nature". The reason for the exclusion was a postponement of a policy decision rather than a rejection of the policy of extending rectification to such documents. The Commission had concluded (report para 3.11) that a policy decision on the rectification of such documents should be made in the context of a review of the law of succession when problems and policies relating to the interpretation of wills had been resolved after due consultation. That work began in 1986 but has taken a long time to bear fruit.
The Commission's consultation did not disclose good reasons for denying the remedy of rectification to testamentary documents. In 1990 the Commission recommended in its report on succession (Scot Law Com no. 124) that the remedy be extended to such documents. The Commission repeated that recommendation in its 2009 report on succession (Scot Law Com no.215, BAILII:  SLC 215 (Report) ) and set out a draft statutory provision in section 27 of the draft Bill appended to the report. It proposed to confine the remedy to the rectification of a will prepared by someone other than the testator because of the very great difficulties in obtaining evidence to satisfy the court of the need to rectify a home-made will.
Until 1985 the only remedy for defective expression which Scots law provided was the rather cumbersome device of partial reduction of the document and a declarator of the terms which were to be treated as having always had effect. This remedy was available for both multilateral and unilateral documents. I have not found any case in which the remedy was applied to a will; but the case law is sparse. As this court has not been addressed on the issue of Scots law, my views do not have the benefit of counsel's researches, and must be treated accordingly.
I see no reason in principle why the remedy of partial reduction and declarator should not be available to cure defective expression in a will. In Hudson v St John 1977 SC 255 Lord Maxwell used the remedy to correct errors in an irrevocable inter vivos deed of trust. A trust of that nature may have attributes similar to a will, and in particular beneficiaries who are not parties to the document.
In Scots law the remedy of partial reduction and declarator is not confined to errors of expression caused by the person who prepared the document. In the leading Scottish case on reduction as a method of correcting defective expression, Anderson v Lambie 1954 SC (HL) 43, Lord Reid stated (at 59) that the court could remedy an error on the part of the professional who instructed the preparation of a document as well as an error by the person who prepared it. He thought that the phrase "clerical error", which had been used in the case law, did not prevent the remedy from being available where a solicitor who had two old contracts gave his clerk the wrong one to copy as the style for a new contract. That is a circumstance not far removed from the facts in this appeal.
In this case both the testator's intention and the solicitor's mistake are clear. I see no reason why in Scots law there would not be a remedy of partial reduction and declarator or, in principle, a rectification if the Scottish Law Commission's proposals are enacted.
Robert Ham QC & Teresa Rosen Peacocke (instructed by Hugh Cartwright & Amin) for appellant.
Nicholas Le Poidevin QC & Alexander Learmonth (instructed by Gillan & Co) for respondent.
all rights reserved