JASWINDER KAUR SANGHA v THE ESTATE OF DILJIT KAUR SANGHA & Ors

This is an appeal against one part of the order of Deputy Master Bowles dated 21 September 2021 in this matter.

When Hartar Singh Sangha ('Hartar') died on 3 September 2016, he left behind him a complex family life and a large portfolio of property and other assets in both the UK and India. He was a prolific writer of wills, and the interaction of these instruments has produced significant disputes amongst his family members. These disputes were largely resolved by Deputy Master Bowles in a 9-day hearing in late 2020, which resulted in a 59-page closely-written judgement addressing almost every aspect of Hartar’s life and activities.

In order to understand what follows, it is necessary to briefly overview the testamentary and family picture with which Deputy Master Bowles was presented. In summary, Hartar at some points in his life regarded himself as married to the first respondent, (Diljit). At other points, he regarded himself as married to the appellant (Jaswinder). The question of who he was married to, and when, is to be determined at a subsequent hearing.

Deputy Master Bowles had before him four wills executed at various times by Hartar:

The 2016 will contained a clause in the form “this is my last and final WILL and all such previous documents stand cancelled”. It is this phrase which is the basis of this appeal.

The Deputy Master held that that these words were effective to revoke all of the previous wills, such that the 2016 will was the only valid existing testamentary instrument. Since it dealt exclusively with property in India, it could not be admitted to probate in the UK, but could be and was declared to be authentic and valid. This finding had the necessary consequence that the English property must therefore be dealt with as an intestate estate.

The appellant contends that the intended effect of the clause cited was not to revoke the 2007 will in its entirety, but only to the extent that it applied to the property specifically bequeathed by the 2016 will, and that the Deputy Master was wrong to conclude otherwise. She therefore argues that the two wills can and should be read as coexisting.

The respondents support the Deputy Master’s decision on the interpretation of the revocation clause. However, they also present a conditional cross-appeal. They say that the Deputy Master was wrong to conclude that the 2007 will was validly executed as a matter of English law, and that the English assets should therefore be dealt with as an intestate estate. If I were to decide against the appellant on her appeal this would be a moot point, since that is the effect of the Deputy Master’s decision in any event. However, the respondents say that even if I decide in favour of the appellant on her appeal, the invalidity of the 2007 will under English law means that the English assets should in any event be dealt with as an intestate estate.

As noted above, the appeal turns on the construction of the provision of the 2016 will

“this is my last and final WILL and all such previous documents stand cancelled”

The Deputy Master expressed the view in his judgement that the correct interpretation of this phrase was "All other wills are cancelled" (para 203).

The appellant’s case is that this is not a correct interpretation of the provision. Specifically, the appellant’s case is that the 2016 will is clearly confined to assets in India, and the “such previous documents” referred to are any previous document which deal specifically with Indian assets. She says that as a matter of construction, if the phrase is construed in context, it does not have the meaning which the Deputy Master attributed to it.

The appellant also argues that the Deputy Master’s approach cannot be right because it contravenes the presumption against intestacy (Theobald on Wills (17th ed) (22-057)). She argues – in effect – that, the more complex and difficult the conduct of an intestacy would be, the more strongly the presumption should be applied. In this case an intestacy would be particularly unfortunate because it would raise issues as to what law is applicable to it (since domicile is in dispute) and who is the “widow” and so principal beneficiary. It would also potentially have the effect of conferring benefits on a person – Mandi – who is not identified as a beneficiary in any of Hartar’s wills made after 1979. The essence of this argument is that an intestacy would produce consequences so wildly at variance with any outcome actually expressed to be desired by Hartar that the court, in construing the will, should strain against it.

Following a strongly fought contest during the trial, the Deputy Master found against the appellant’s contention that the 2016 will was a forgery (at paragraph [191] of his Judgment). He then proceeded to consider the construction of the 2016 will and the question of whether it revoked all previous wills. He found as follows:

It should be noted that the Deputy Master had heard a very great deal of evidence on this matter. He considered both the possibility of the admission of extrinsic evidence of the testator’s intention, pursuant to Lamothe v Lamothe, and the possibility of the admission of evidence under s.21 of the Administration of Justice Act 1982. He concluded that neither issue arose, as there was no such evidence. That is significant for the purposes of this appeal, since it means that the issue before me is simply one of construction.

At various points in the hearing I was invited to interpret the will in the light of speculations as to what the testator might or might not have intended. I do not think that that approach is relevant or – indeed – permissible. In particular, it is not open to me to consider whether or not Hartar would have intended to provide for Jaswinder or not, or what his intentions might have been as regards intestacy. I agree with Mr East that the second ground of the appellants challenge must fail.

However, I was also asked to imply the wishes of the testator by applying the presumption against intestacy. I agree that the presumption against intestacy is a partial route to inferring a testator’s intentions, and it is to this that I now turn.

The Deputy Master considered the question of the application of the presumption against intestacy in his judgement. He concluded (at paragraph 209) that

“In this case, I am not persuaded that the fact that the literal effect of the revocation clause in the 2016 will gives rise to an intestacy in regard to Hartar's English estate is, in the context of this case, sufficient to establish the clear and unequivocal intention that the revocation clause must have been intended to be confined to the revocation of those wills, or parts of wills, affecting, only, Hartar's Indian estate.”

The presumption against intestacy is an old-established rule, whose primary effect is that the court should err in favour of giving effect to a testators attempt to make a will, even if those effects are misdirected or ineffective. Lord Esher set out the basic position in Re Harrison (1885) 30 Ch D 398 at 393:

This principle is clearly applicable where the testator has attempted to make a will but has failed to satisfy one or more formal or other requirements, or has given incoherent or contradictory instructions. I do not think that that it is a principle of construction. Its application was recently considered in Almond v Goff 2021] EWHC 1703 (Ch) by Ms. Clare Ambrose, sitting as a Judge of the High Court. She explained that

“The presumption against intestacy was of limited weight for a number of reasons. As explained in Howell v Howell , it will have less weight if it conflicts with the principle of treating beneficiaries with equality. More significantly, it is well established that it should generally not be applied to do violence to the clear meaning of a will and this approach has been reinforced by reliance on the natural and ordinary meaning of the wording. As counsel for Rafferty explained, the previous authorities on construction are to be regarded with some caution. The presumption against intestacy is one that has already long been regarded with some scepticism”.

In this respect she cited the words of Lord Greene in Abbott [1944] 2 All ER 457:

I do not think that on the facts of this case the presumption against intestacy is of any assistance to the appellant. It is possible to suggest reasons as to why Hartar might have positively intended to leave the position as to his English estate unresolved – specifically, that it avoided him having to engage with the difficult issues caused by his marital situation. What is not possible, on these facts, is to construct an argument to the effect that his intentions were so clear that it would be illegitimate to apply a construction to the will concerned that defeated them. That is simply not the position on the facts as found in the Deputy Master’s judgement.

Mr Blackett-Ord, for the appellant, advanced the proposition that the natural implication from a will dealing very specifically with the Indian estate but containing a general revocation clause, is likely to be that what is to be revoked are the previous testamentary dispositions which relate to the Indian estate, rather than the parts which do not. He says that any phrase in a will must be construed in context.

As regards construction, the starting point is as stated by Lord Neuberger in Marley v Rawlings (2015) AC 129 at 19:

So in Dempsey v Lawson (1877), 2 PD 98, Sir James Hannen P said at p.107:

Mr Blackett-Ord submits that because the 2016 will has been carefully composed to deal with only the Indian estate, but the testator had already made a will in favour of his wife relating to both his English and his Indian estates, it would be perverse if the later will, which left no English property to anyone, was intended to revoke the existing English estate dispositions.

He accepts that the position would be different if the revocation clause had specifically referred to the testator’s property "wherever situated" or there was evidence that the testator had considered the matter and specifically intended a general revocation. That was the case in Lamothe v Lamothe (2006) WTLR 1431 at para 51, and the evidence of intention in that case was held to be conclusive. However, he argues that that case is distinguishable from the present facts on exactly that ground.

The Deputy Master held that the ordinary, natural meaning of the words “all such previous documents stand cancelled” was that all previous wills stand cancelled. We are therefore construing the word “such” in this context. I think the starting point here is that the word could operate in one of two ways. One is that it could refer to the word “will” in the same sentence – thus, the reading could be expanded as “this is my last and final will and all [previous wills] stand cancelled”. The other is that it could refer to the document in which it was contained – thus it could be expanded as “this is my last and final will [in respect of my Indian property], and all [wills in relation to that property] stand cancelled”.

The question before me is as to whether the Deputy Master was correct to find that the intended meaning was the first, and that this was the ordinary natural meaning of the words.

I do not find it at all easy to say which of these is the correct interpretation – there are arguments for both. However, there are two material facts from which I think assistance can be drawn. It seems to me that if it had been the testator’s intention to revoke the will dealing with his English property he would have mentioned that fact, or at least alluded to it. I think some significance must be attached to his decision not to do so. The 2016 will is an entirely Indian document, in that it was drafted in India, with the assistance of an Indian lawyer, and relates exclusively to property in India. The terms of the document itself provide no indication at all that the document was intended to have any application at all outside the specific estate with which it dealt.

The other is that there is an entirely plausible alternative explanation for the chosen wording. In 2003 the testator had made an earlier will which also related exclusively to his Indian property. It seems that some of the property covered by the 2003 will was also covered by the 2016 will. There is a clearly identifiable document which could have been intended by the term “such documents”, that being the 2003 will. The idea that the purpose of this language was to ensure that the earlier Indian will in relation to Indian property was revoked seems to me to be more likely than the idea that it was an indirect attempt to revoke by implication the provisions of a different will on a different continent dealing (in good part) with different property.

As a result, I think that the Deputy Master was wrong to construe the words used as being intended to, or having the effect of, wholly revoking the 2007 will. The two documents must therefore be taken together to constitute the will of the testator.

There are two authorities which bear on the issues in dispute here. One of them is Benjamin. v Bennett [2007] All ER (D) 243 (Feb), where the court held that a Bajan will, dealing solely with Bajan property, headed 'Barbados' and containing a wide ranging revocation clause did not revoke a prior English will dealing solely with the testator's English estate and containing, as the Barbados will did not, provisions as to the testator's burial and funeral services. The Deputy Master found that, unlike in the Barbados will, there was nothing in Hartar's 2016 Will to give rise to any suggestion that the entire document, including the revocation clause, was limited to Harter's Indian estate. Benjamin v Bennett is inadequately reported, and does not appear to have been fully argued. However, it appears to be an assertion of a relatively uncontroversial point – that a will which is clearly intended to deal with only specific, identified part of a testators property should be read as applying only to that property unless there is some good reason to do otherwise. The Deputy Master seems to have concluded on this point that this will had no equivalent of the heading “Barbados” in the 2016 will to indicate that its scope was intended to be geographically confined. However, that conclusion is hard to square with the contents of the document itself, which is explicitly confined to Indian property.

The other potentially applicable authority is Re Wayland [1951] 2 All ER 1041. In that case, the deceased had made a Belgian will dealing with his Belgian assets, and later made an English will. The court held that the later English will did not revoke the Belgian will. However in that case, unlike in this one, there was evidence of intention on the part of the testator which contradicted the plain meaning of the revocation clause. This evidence was derived both from the conduct of the testator and from the words of the document. As regards conduct, when he was giving instructions for his English will, the deceased was asked by his English solicitor what should be done about the Belgian estate and was told not to worry as his Belgian lawyer would take care of it as he had made a Belgian will. He also wrote to his Belgian lawyer shortly after execution of the English will stating that he noted the lawyer would 'take charge' of the Belgian will. As regards documentation, in one of its clauses the English will expressly stated that it was intended to deal only with the deceased's English estate: 'this will is intended to deal only with my estate in England.'. If the revocation clause had been construed in a manner which resulted in the Belgian will being revoked, then this would not be consistent with that clause. The Deputy Master held – I think correctly - that Wayland provided no assistance in this case.

Equally, I do not think that Almond v Goff [2021] EWHC 1703 (Ch) is of any assistance here. This case was decided after closing submissions were heard by the Deputy Master in this case, but I do not think that there is anything in it which has any direct bearing on the issues that were before him. In Almond v Goff the effect of a clear and unambiguous revocation clause - "I revoke all former wills made by me and declare this to be my final will and testament" was held to constitute a revocation of an earlier will relating to the Spanish property of the testator, leaving him intestate as regards that property. Ms Clare Ambrose sitting as a Deputy Judge of the High Court, held (at para 68) that in order to hold such a clear general revocation to be ineffective, the 'test is that there must be clear evidence of an intention that the earlier will has not been revoked.' However, the point at issue in Almond was precisely that the revocation clause was clear and unambiguous - it does not help in the present case, where the issue is precisely to elucidate the challenge posed by the ambiguous words actually used.

It seems to me that where a will is expressed to apply to specific, identified property in a particular jurisdiction, is made in that jurisdiction with the assistance of lawyers established and qualified in that jurisdiction, and has no other connecting factor with any other jurisdiction, the starting point should be an assumption that the will as a whole is only intended to apply to that property in that jurisdiction unless there is some good reason to believe otherwise. In this case there was none.

Mr East, on behalf of the respondents, challenges this construction. He says that this is not a case like Wayland where there were two separate wills, one dealing with the property in one jurisdiction and one dealing with property in another, such that it can be said that it could not have been intended for the later will to revoke the earlier. He says that it was open to Hartar to restrict the revocation clause to specifically Indian wills, and that the fact that he did not do so shows that he could not have intended to do so. He argues that Hartar took a deliberate decision to say that 'all such previous documents stand cancelled' rather than 'all the parts of such previous documents which deal with my Indian estate but not otherwise stand cancelled' or similar.

This approach seems to me to reverse the natural approach. If the will had been drafted by a legal professional, interpreting Hartar’s intentions in legal language, it would have more force. However the evidence at trial was that the will was written out, initially in Punjabi, by an Indian advocate Mr Bedi, at Hartar's dictation, before then being translated into English and then executed (paragraph [150] of the Judgment). This was in common with Hartar having provided substantial input (or wholly written) the 2007 and 2003 Wills himself. If the revocation clause was dictated by Hartar, then I do not see that any very great reliance can be placed on technical drafting. In particular, Mr East argues that Hartar cannot have intended to allude to the 2003 will when he used the phrase “such previous documents”, since he must have known that that will was revoked by the 2007 will. I have no way of knowing whether this actually was the case or not – as regards the Indian properties, it is a matter of Indian law. What I am reasonably certain of is that Hartar, as a lay testator, is unlikely to have had that degree of certainty. I am highly confident that, as a frequent testator, he had an idea that it was necessary to revoke previous wills. I also think that he knew that he had made an earlier will relating to some of the property disposed of in the will he was now dictating, and that that earlier will had been registered with the Indian government. However, I do not think that it is possible to ascribe to him anything like the certainty that Mr East suggests that he must have had as to the fact of revocation of that earlier document.

Mr East also argues that this construction departs substantially from the natural and ordinary meaning of the words used, that it would therefore require “clear and unequivocal evidence” that Hartar's intention was different from the meaning of the words which he dictated to give them the meaning that Jaswinder contends for. However, I do not agree that the construction which the Deputy Master placed on these words is in fact their natural and ordinary meaning, so I do not think this point has any force.

Finally, Mr East argued that Hartar’s intention to revoke the 2007 will could be inferred from the steps which he had taken in respect of his English assets. Deputy Master Bowles found that Hartar had taken steps to transfer the bulk of his English properties into the joint names of himself and Jaswinder, such that they would pass to her on his death outside the estate. He concluded from this that it was “not at all implausible” that Hartar might therefore have taken the view that his English estate was already dealt with, and that the earlier will relating to it could therefore safely be revoked. The problem with this argument is that by no means all of Hartar’s English estate had been put into joint names or was in the form of real estate, and Hartar himself was well aware of that – as Mr East points out, in the 2007 will, which Hartar himself wrote out, he refers to his “bank accounts, moveable and immoveable property in India and the U.K.”. My understanding from Counsel was that the element of the English estate that did not pass by survivorship, although as yet undetermined, may well have been in excess of £1m, and it is hard to imagine that Hartar was unaware of that fact. I agree that the construction put forward is not implausible, but I do not find it compelling.

This takes us to the cross-appeal. The respondents argue – in effect - that the Deputy Master was wrong to reject their submissions at first instance that the 2007 will was invalid at English law for want of compliance with s.9 of the Wills Act 1837. This will was, again, executed in India. However, it does purport to dispose of property in the UK, so its validity as a matter of English law is of critical significance. If it is not valid under English law, then the English estate must go on intestacy.

The 2007 Will itself contained an attestation clause stating that it had been 'signed by the testator/ testatrix in our presence and by us in his/ hers'. However, one of the witnesses to the will, Mr Kulbir Singh Khaira, gave oral evidence that he was not in fact present when Hartar signed the Will. As per para [75] of the Judgment, 'he had been summoned, by Hartar, to the living room of the Chandigarh property, from his nearby office, and by the time he had arrived the will had already been signed by Hartar and by Mr Balraj Singh, as witness. Hartar had requested Mr Khaira to sign as the second witness and he had signed.'

As a result of the evidence from Mr Khaira, which differed from the form of the attestation clause, the Deputy Master correctly held that the presumption of due execution had no application, and instead, the court had to consider whether there had been due execution in accordance with section 9 of the Wills Act 1837 based on the actual evidence before it [105-107].

Section 9 of the Wills Act 1837 provides as follows (so far as relevant)

No will shall be valid unless—

At trial, the first to third respondents submitted that section 9 (d) could not be satisfied in light of the evidence given by those present at the signing of the 2007 will. The Deputy Master dealt with the Respondents' submissions on this point from para [117] onwards. He found that the 2007 will was duly executed on two bases:

Mr East, for the Respondents, argues that the Deputy Master erred in law on his first conclusion. He says that although the law has changed following the amendment of section 9 of the Wills Act 1837 by the Administration of Justice Act 1982, it clearly is still a requirement that each witness either attests and signs the will or acknowledges his signature after the testator has signed the will or acknowledged his signature in the presence of the two witnesses.

This argument draws on the sequence of events relating to the execution that was found at trial. Hartar and Mr Balraj Singh were together. Hartar produced the will and signed it. Balraj Singh signed it as a witness. Mr. Khaira was then summoned to act as second witness. Hartar produced the will to him to sign. It seems clear that by producing a document with his own signature upon it, Hartar thereby acknowledged that signature in the presence of both witnesses, thereby complying with s.9(c). This is not disputed. However, Mr East’s argument is that the requirements of s.9(b) were only satisfied when Hartar produced the will to Mr Khaira, since that was the first time that both witnesses were together. Since Mr Balraj Singh had signed the will before that moment, Mr East argues that that signature was ineffective under s.9(d). In order to validate the will, he says, Mr Balraj Singh was required to do something else after that moment – either to re-sign it or to acknowledge it – in the presence of Mr Kaira. Since he clearly did not do the first, and there is no evidence that he did the second, Mr East argues that the will must be treated as invalid.

The position for which Mr East argues was clearly the case prior to the amendments of s.9 made by the Administration of Justice Act 1982. The position under the previous legislation was that there was clear authority that where either or both witnesses signed before the testator had signed or acknowledged the will, the will was invalid (see Williams, Mortimer & Sunnucks at 9-24). However, the basis for this conclusion was that the old wording of the act was that the testators signature “shall be made or acknowledged… in the presence of two or more witnesses… and such witnesses shall attest and shall subscribe the will…. ” The word “shall” in this context was held to denote a time sequence, so that it was read as “shall then” (Re Allen (1839) 2 Curt 331). The amendments made to s.9 in 1982 removed the word “shall” from this section and reworded it. The editors of Mortimer, Williams and Sunnucks (at 9-25) express the view that this rewording removed the concept of time sequencing. The express intention seems to have been that three things are required – the testator must sign or acknowledge his signature in the presence of two witnesses, and each witness must either sign or attest to the testator. If that is correct, the fact that Mr Balraj Singh signed the will before the testator acknowledged the will before two witnesses does not invalidate the will, and the fact that he did not reacknowledge his own signature after that event is neither here nor there. This is exactly the conclusion which the Deputy Master came to in paras 119-121 of his judgement, and I have nothing to say about his conclusion beyond the fact that I agree with it.

Mr East also argues that the only way that the Deputy Master could have reached this conclusion was by illegitimate speculation, conducted in the absence of any evidence on the point. His starting point is that the burden of proof in a probate case lies on the person who is seeking to propound the will, here the appellant: Theobald on Wills at 14-012. He therefore argues that since there was no evidence upon which to make the finding of fact as to acknowledgment that burden cannot have been discharged. Alternatively, he argues that the finding that there was an acknowledgment was plainly wrong as there was insufficient evidence to support such a finding on the basis of the surrounding circumstances referred to by the Deputy Master.

On the question as to whether there was evidence of any actual acknowledgement, the Deputy Master recognised that the decision was 'marginal (para [127]) and that there was a 'lacuna in the evidence' (para [123]). Nevertheless, he concluded that there would have been 'some conversation, or discussion, and that that discussion would have included an assertion, or indication, by Mr. Balraj Singh that he had already signed the will, as witness. ' (para [126]). The Deputy Master goes on to conclude in the same paragraph that the greater probability was that this would have occurred after Hartar had acknowledged his own signature in the presence of Mr Khaira and Mr Singh.

What does not help here is that although there is (almost) no room for debate as to what constitutes a signature, there is almost no limit to the question of what might constitute an acknowledgement. The authorities concerning acknowledgment are summarised in Williams, Mortimer & Sunnucks at 9-21: some words must have been spoken by the witness or some act must have been done by him (or by some other person on his behalf) as may properly be regarded as an acknowledgement of his signature. There is no particular formality required for an acknowledgement - in Couser v Couser [1996] 1 W.L.R. 1301, the relevant witness acknowledged her signature to the testator after he had acknowledged it to her and her husband (the other witness), by protesting to the testator about the manner in which the will had been signed, and advising him to go to the bank and have it done properly. Accordingly, protesting in this manner amounted to a positive act 'acknowledging that something had happened and that she had indeed signed the document' (see the report at 1306). Thus the factual question before the deputy master was as to whether anything at all had been said or done that might constitute an acknowledgement. He concluded that it had.

This is difficult ground for any court exercising an appellate jurisdiction. The true basis for Mr East’s argument is, I think, the observation of Lord Simon in Thomas v Thomas 1947 SC (HL) 45, 61; [1947] AC 484 (cited with approval by the Supreme Court in McGraddie v McGraddie [2013] 1 WLR 2477 at [27]) to the effect that where the court below has reached a conclusion on a matter of fact, “If there is no evidence to support a particular conclusion (and this is really a question of law), the appellate court will not hesitate so to decide.”. Mr East argues that that is the case here.

The starting point here is that the Deputy Master found on the facts that the terms of the attestation clause did not reflect what had, actually, occurred and, held that the attestation clause being demonstrably and admittedly incorrect, no presumption of due execution could arise therefrom. This is clearly correct as a matter of law – see Kayll v Rawlinson [2010] WTLR 1443, applied in Wilson V Lassman [2017] EWHC 85. However, it is important to emphasise that the failure of the presumption of due execution does not create a presumption of ineffective execution which requires to be rebutted. The position where the presumption fails is simply that the Judge, having regard to all the circumstances, must decide what probably happened. True it is that the burden of proof falls on the person propounding the will. But that absolutely does not mean that, if the presumption of due execution fails, the will itself fails unless the proponent produces direct positive evidence of that execution. The question of whether the actions of the testator and the witnesses were such as to constitute due execution in accordance with s.9 is a question of fact for the court. The court may decide that point on the basis of direct evidence, but it is also entitled to proceed by inference. The Deputy Master clearly took the view that, on the balance of probabilities, the relevant facts were proved to his satisfaction. I have very little doubt that he was right to reach the conclusion that he did – the idea that three men, gathered together to create a will, having signed it, would not have behaved in such a way as to acknowledge to each other the signatures which they had applied to the document, seems to me to be fanciful. However, even if that were not the case, I do not think that I would have any business challenging a finding of fact by the tribunal which actually heard the witnesses. The deputy master proceeded by inferring facts from the evidence before him. I think he was entirely within his rights to draw those inferences from that evidence, and I am in no position to dispute his findings.

I would therefore allow the appeal and reject the cross-appeal.