By Professor R.A. Hughes,
Head of School of Law,
University of the South Pacific
The wills legislation of the countries of the South Pacific have largely adopted the principles relating to revocation of wills which appeared in section 19 of the Wills Act 1837 (U.K.) The legislation provides that will can be revoked by a voluntary act of a certain sort or is automatically revoked by the occurrence of certain events. Thus it provides for voluntary or involuntary revocation, the latter often being termed ‘revocation by operation of law’.
The methods for voluntary revocation are another will or codicil, another revoking document executed in the manner required for a will or destruction by burning tearing or otherwise destroying. In the South Pacific jurisdictions the only revocation by operation of law provided for is subsequent marriage of the testator. (See for example, section 15 Wills Act, Cap. 61, Fiji; section 7 Wills Act 1975 Samoa; section 10(1) Wills, Probate and Administration Act 1987 Solomon Islands; sections 7 and 8 Wills Act Vanuatu Cap. 55) In other jurisdictions, such as some of the Australian States, dissolution of marriage is also provided as a ground. (See, for example, Wills, Probate and Administration Act 1898 (NSW) s. 17)
On the face of it, the acknowledgement of two categories of revocation, voluntary and involuntary, appear to make sense. One involves a direct act of the testator done with animus revocandi (an intention to revoke). The other does not seem to require any act of free will or specific intention on the part of the testator. In the latter case, revocation is deemed to take place as soon as a certain event takes place, an event which in the South Pacific jurisdictions, is marriage. Both categories would seem to rest on different justifications. In the first case, if a person has freedom of testation, then they should also have freedom to revoke on the proviso that the methods of revocation prescribed are appropriate to the gravity of the event. The second is based on a policy of requiring a person to reconsider his or her social obligations given a radical change in circumstances. Given that marriage is the only current ground for such reconsideration in the South Pacific jurisdictions it may be that the breadth of this assertion is in need of refinement.
Much has been made of this distinction particularly when the effect of contracts not to revoke a will are concerned. It has been held that where there is a contract not to revoke which is enforceable on the face of it, there is a breach of that contract only where the testator undertakes a voluntary revocation. There will be no breach where the testator subsequently marries and therefore, by operation of law, the original will is revoked. This was decided in Re Marsland; Lloyds Bank Ltd. v Marsland  Ch 820 on the apparent basis that a breach of contract could be imputed to a party only where he or she had undertaken a voluntary act. But why, we might ask, was the decision and the act of marriage itself, not a voluntary act capable of involving a breach of contract in an appropriate sense. Surely the legal consequences of the act cannot be so clearly and categorically distinguished from the act itself.
However, if we look at another context, which is not too dissimilar in many ways, a reverse assumption as to the nature of the difference seems to have been operating. This is in cases where there are mutual wills made in such a way as to give rise to equitable obligations on the party who survives. (The leading case in this area is Birmingham v Renfrew (1937) 57 CLR 666) The obligation is one not to revoke a will and thereby to transgress the agreement which is entered into at the time the mutual wills were made. Where one party has died the other will be obliged not to revoke his or her will subsequently. Where the surviving party purports to do so a constructive trust will be imposed to remedy the injustice which might arise from breach of the original agreement.
In such a case it has been held that it makes no difference whether the revocation by the surviving party was a voluntary or an involuntary revocation. In one of the few authorities in this area, Re Green  Ch 148, it was held that this was the case because the constructive trust is deemed to arise at the time when the first party died. The fact of revocation might have been import, but just how it was brought about was not. According to this decision, the rights of the beneficiaries under the constructive trust arise prior to the revocation. Hence it would make little difference whether the revocation which amounts to a breach of the equitable obligation was one which arose from a voluntary act of the second party, or arose by operation of law.
But the reasoning here is also questionable. In the case of imputation of constructive trusts it is quite clear that the trust can be imposed retrospectively. It need not be treated as arising only when the breach of the obligation has occurred. Were the position otherwise then it may be that a party could avoid accounting for all of the gains which the court would want to treat as flowing through to the beneficiaries. However, this does not mean that the nature of the revocation can be simply side-stepped by say that the trust arose before revocation. The correct position is to say that the trust arises because there has been a breach of an equitable obligation but it is then deemed to exist retrospectively to a point prior to the date of the breach. It is still the case that, in the circumstances at hand there must have been a breach of the obligation involved. The obligation in Re Green, as with other mutual trust situations, is the agreement not to revoke coupled with the reliance on that agreement by one of the parties. The death of one of the parties is taken as sufficient to constitute an intractable reliance on that agreement. It is the revocation which constitutes an attempt by the other party to renege on the situation. It is this factor which constitutes the potential injustice in allowing the other party to depart from the terms of the mutual wills.
Thus the position in Re Green merely purports to ignore what it should have addressed. It was inappropriate simply to assert that the nature of the revocation need not be considered because the trust had already arisen. The trust could not have arisen unless there was a breach of the obligation in the sense mentioned. Whether the court then seeks to make the operation of the constructive trust retrospective or not is a matter to be considered thereafter. It is suggested therefore that the decision puts the cart squarely before the horse. It needed to consider whether the subsequent marriage, as a revocation by operation of law, in fact constituted a repudiation by the other party of the obligation placed on him as a result of the death.
The court in Re Green could have adopted the approach in Re Marsland (above). Both in a sense involved potential breaches of current obligations, the one in contract and the other in general equity. Both involved acts of involuntary revocation. If so, the court in the later case could have decided that only an act of voluntary revocation would have constituted a breach of an obligation sufficient to give rise to a constructive trust. However, in my view, that would have been an unsustainable conclusion because the approach in Re Marsland was flawed. In fact, I will suggest that the notion of a categorical differentiation between voluntary and involuntary revocations is ultimately mystifying and ought to be ignored, except perhaps, as a matter of convenience to writers of academic texts.