PER / PELJ - Pioneer in peer-reviewed, open access online law publications
Author James T Faber
Affiliation University of the Free State South Africa
Date Submission 29 October 2021
Date Revised 14 February 2022
Date Accepted 14 February 2022
Date published 9 March 2022
Editor Dr N Kilian
How to cite this article
Faber JT "Uncertainty About the Condonation of Formally Non-Compliant Wills, and the Rectification of Cross-Signed Mirror Wills: Is an Act-Based Model the Solution?" PER / PELJ 2022(25) - DOI http://dx.doi.org/10.17159/1727-3781/2022/v25i0a12691
A recent contribution proposed a processual act
Online ISSN 1727-3781
Act of testation; act-based model; condonation; law of succession; rectification; testator's intention.
A recent contribution proposed a processual act-based approach to conceptualising wills in South African law. In brief, this approach regards a will as the product of a will-making process which sees various parties perform specific acts with specific associated forms of intention to establish a will. The act-based model also paves the way for the introduction of an intent doctrine into South African law, in terms of which the testator's intention is viewed as a compound, multilateral concept that encompasses multiple forms (or facets) of intention. In terms of this intent doctrine, testator's intention is best understood by linking an appropriate form of intention to each act performed by the parties to the will-making process to create a will. The aforesaid act-based approach runs counter to the traditional view that a will is the result of once-off compliance with several set requirements (for instance, the presence of testamentary capacity, the existence of animus testandi, and compliance with a range of statutory formality requirements), or the so-called "requirements model".1
* James Thomas Faber. BProc LLB LLM PhD. Lecturer, Faculty of Law, University of the Free State, South Africa. Email: email@example.com. ORCID: https://orcid.org 0000-0001-7379-5965. I am grateful to Prof François Du Toit (University of the Western Cape) for his valuable contribution and guidance with this article".
1 See Faber 2021a TSAR 504; Faber 2021b TSAR 740.
2 In Ex parte Estate Davies 1957 3 SA 471 (N) 474A-C and Oosthuizen v Die Weesheer 1974 2 SA 434 (O) 436C-D, three essential elements were identified for a testamentary disposition to be established, namely a bequest of assets, the extent of the interest being bequeathed, and the identity of the beneficiaries.
The act-based model identifies the act of testation as the written manifestation of the testator's dispositive intention and animus testandi. The dispositive intention is expressed in a dispositive act which is primarily aimed at disposing of assets (in simple terms, it prescribes who inherits what). To qualify as an act of testation, the dispositive act must be complete: all the elements of a testamentary disposition must be present.2 Animus testandi in turn represents the intention for the testamentary dispositions to be given legal effect upon the testator's death. Therefore, a will is the documentary expression of this act of testation, but could, at the same time, also embody other acts with associated intentions (such as the act of revocation) and govern other matters (such as the nomination of an executor and the appointment of a guardian). Finally, the document must also be validly executed, which requires compliance with all the statutory
formality prescripts in terms of the Wills Act,3 for the document to have legal force as a will.4
3 Wills Act 7 of 1953 (hereafter the Wills Act or the Act). The execution of a will entails a process that involves a number of parties (the testator and witnesses, for example), each of whom performs a specific execution act in order to comply with the formality requirements. To execute the document, each party needs to sign it, which act must occur with the necessary intention. This intention is known as animus signandi. See Faber 2021a TSAR 519-520; De Waal and Schoeman-Malan Law of Succession 60; Jamneck et al Law of Succession 65, 72.
4 See Faber 2021a TSAR 504; Faber 2021b TSAR 740; Schoeman-Malan et al 2014 Acta Juridica 80.
5 See Faber 2021a TSAR 504 and Faber 2021b TSAR 740 for a discussion on the intent doctrine.
The schematic representation below of a simple will illustrates how the act-based model, with its focus on the various phases of the will-making process, and the intent doctrine interlink to conceptualise a will in South African law:5
In this contribution the functioning (and viability) of the proposed processual approach and act-based model is tested and illustrated by applying it to two scenarios: the condonation of formally non-compliant wills in terms of section 2(3) of the Wills Act and the rectification of cross-signed mirror wills in terms of the common law. These particular two scenarios were selected because both continue to be plagued by uncertainty as a direct consequence of the lack of a proper definition, explanation and contextualisation of testator's intention in South African law. As a result,
testator's intention is not expressly linked to the act of testation, which in turn is not clearly distinguished from the other acts in the will-making process. Therefore, this contribution attempts to address and resolve the current uncertainty by applying the will-making process and act-based model to these two scenarios.
2 Scenario 1: The condonation of formally non-compliant documents
2.1 Condonation and the intention requirement in terms of section 2(3)
In accordance with section 2(3) of the Wills Act, a court may condone a document that fails to comply with (all) the formality requirements for a valid will by ordering the master to accept the formally non-compliant document for purposes of administering the deceased's estate. Condonation is possible only if the requirements in section 2(3) are satisfied, namely:
In many respects, the intention requirement is the key consideration in determining condonation in terms of section 2(3).6 Therefore, one is again confronted with the testator's intention. Judging from the jurisprudence, there is no absolute certainty as to the exact scope of this intention and its connection with animus testandi.7 Consequently, an additional form of intention has started cropping up in legal literature – the so-called "section 2(3) intention", which appears to be something completely different from the testator's intention referred to in the introduction to this contribution.8
6 Ex parte Williams: In re Williams's Estate 2000 4 SA 168 (T) 179A.
7 See for instance Osman v Nana (37220/2018)  ZAGPJHC 161 (3 May 2019) para 23, where a single judge did in fact make a connection between animus testandi and the s 2(3) intention requirement. On appeal, however, the same connection was not clearly established by a full bench in Osman v Nana 2021 JOL 50242 (GJ).
8 See Jamneck 2008 THRHR 603. This article will show that the creation of such artificial forms of intention is unnecessary and not ideal.
The intention requirement in section 2(3) is traditionally viewed from two points of view:
9 Ex parte Maurice 1995 2 SA 713 (CC) 716I-J, 717A-B; De Waal and Schoeman-Malan Law of Succession 80, 89.
10 Ex parte Williams: In re Williams's Estate 2000 4 SA 168 (T) 179D-G.
11 Van Wetten v Bosch 2004 1 SA 348 (SCA) para 16.
12 Webster v The Master 1996 1 SA 34 (D).
13 See Webster v The Master 1996 1 SA 34 (D) 42B-G. Also see Jamneck et al Law of Succession 79.
14 Dryden v Harrison (WCC) (unreported) case number 11912/17 of 20 May 2019.
15 Van der Merwe v The Master 2010 6 SA 544 (SCA).
In my view, however, the real question to be addressed at this stage is not what the document means, but whether the deceased intended it to be his will at all. That enquiry of necessity entails an examination of the document itself and also of the document in the context of the surrounding circumstances.
The first perspective potentially diminishes the application of the power of condonation as it seems to be based on the view that only formally non-compliant wills – in other words, documents that resemble a will in both appearance and format – can be condoned. In fact, in Webster v The Master12 the court narrowed the application of section 2(3) even further when it found that the provision was restricted to irregularly executed wills, leaving no room for unexecuted wills. From the court's reasoning in this matter, it would further seem that the restriction of section 2(3) to formally irregular wills inevitably also means that informal documents – in other words, documents not cast into the testamentary mould – cannot be condoned either.13 Therefore, in Webster the court appears to have ruled that the intention requirement of section 2(3) can be satisfied only in respect of a document that is a will per se, but simply has not been properly executed. More than two decades later the court also refused condonation in Dryden v Harrison,14 seemingly because of the informal way in which an e-mail "will" had been created. The court drew a distinction between this informal document and the unexecuted will transmitted by e-mail in Van der Merwe v The Master,15 which the Supreme Court of Appeal in Van der Merwe indeed condoned. The court in Dryden reached this conclusion even
though the subject line of the e-mail concerned contained the words "Final will", and that the initial sentence of the e-mail was: "Hi, this serves as my final will and testament."16
16 Dryden v Harrison (WCC) (unreported) case number 11912/17 of 20 May 2019 paras 17-18.
17 Smith v Parsons 2010 4 SA 378 (SCA).
18 De Waal and Schoeman-Malan Law of Succession 81.
19 Van Wetten v Bosch 2004 1 SA 348 (SCA) para 16.
20 Also see De Waal 2016 Annu Surv SA L 968.
The approach in the decided cases of Webster and Dryden is diametrically opposed to that in the second instance mentioned above, namely to give effect to the deceased's dispositive intention, even outside the context of a formally non-compliant will. In Smith v Parsons,17 for example, the Supreme Court of Appeal condoned a suicide note that outlined certain amendments to the deceased's existing will. According to De Waal and Schoeman-Malan,18 while the letter as such clearly had not been intended as a will, it nevertheless contained a disposition of assets (a bequest). Therefore, the Smith matter is perfectly aligned with Van Wetten v Bosch,19 particularly in terms of the court's emphasis on the document’s being a carrier of the deceased's intention (to amend), instead of being primarily a suicide note.
In the light of the Van Wetten judgment, it would appear that despite the phrasing of section 2(3) the power of condonation is not restricted to formally non-compliant wills (in the sense of documents that resemble wills in both appearance and purport, but have not been duly (or at all) executed in terms of the prescribed formality requirements, but also applies to informal documents that embody acts of testation – and, therefore, testator's intention – but which cannot necessarily be labelled "wills".20
2.2 The act-based model as a potential solution to condonation uncertainty
2.2.1 The distinction between formally non-compliant wills and informal documents
It is suggested that the tension between Webster and Dryden on the one hand and Smith on the other regarding which document types lend themselves to condonation in South African law may be attributed to a lack of consistent focus on the act of testation – being the carrier of both testator's dispositive intention and animus testandi. Therefore, a solution would be rather to view the condonation requirements against the backdrop of the act-based model in the context of the will-making process. The focus
should be on the act of testation, as discussed earlier. If this is found to be present – in whatever shape or form – the formally non-compliant document embodying the act of testation should be condoned (provided that the other requirements of section 2(3) have been satisfied).21
21 However, it is argued below that the s 2(3) "drafted or executed" requirement is problematic and probably needs to be reconsidered. See para 2.2.2 below.
22 Also see Wood-Bodley 2011 SALJ 612.
23 Consequently, it appears unnecessary to bring a separate s 2A application for the condonation of the revocation act, should the document contain a revocation clause as well. Although this contribution does not attempt to address the uncertainty regarding the interaction between ss 2(3) and 2A of the Act, the act-based model – which clearly suggests a distinction between the act of testation and the act of
Viewed from the perspective of the act-based model and intent doctrine suggested in the previous contribution and referred to in the introduction to this article it would seem that the intention requirement upon condonation implies two particular forms of intention, namely the dispositive intention and animus testandi. In terms of the former, it has been firmly established that the document serving before the court for condonation in terms of section 2(3) needs to deal with the disposition of assets. The act-based model clearly states that the dispositive intention – and, therefore, a dispositive act – must be clearly manifested in such a document.22 In addition, the act-based model requires the dispositive act to be a true act of testation. In terms of the act-based model, a dispositive act qualifies as an act of testation only if it is accompanied by animus testandi. It follows, then, that the actual question to be answered by the court in considering an application for condonation in terms of section 2(3) is whether the deceased, in drafting or executing the particular document, formed the necessary animus testandi in respect of the disposition of assets upon his/her death.
Should the court find that the required animus testandi was indeed formed, the dispositive act contained in the document would qualify as an act of testation. If the rest of the section 2(3) requirements are satisfied, the court would then excuse the fact that the document embodying the act of testation fails to comply with (all) the formality prescripts of the Wills Act, irrespective of the format of the document, for the purposes of giving effect to the deceased's intention. The act of testation may therefore be contained in either a formally non-compliant or an informal document. The latter may not necessarily resemble a will in appearance or content and thus may technically not be a will but rather be considered a letter, a note or an entry in a diary. Where a formally non-compliant will is at stake, therefore, the court would excuse the formal non-compliance and give effect to the entire document as a will.23 In respect of an informal document, on the other hand,
revocation, being two separate, independent legal acts – should be able to clarify any potential ambiguities in this regard as well. See De Waal and Schoeman-Malan Law of Succession 95-100.
24 Similar to condoning an informal revocation document embodying an act of revocation performed with the necessary animus revocandi in terms of s 2A of the Act.
25 See for example the case of Taylor v Taylor 2012 3 SA 219 (ECP), where the testator compiled a so-called "wish list" before his death. This wish list was presented to the court for condonation. The court found that the testator intended the wish list to be merely a guideline in the distribution of his assets rather than an amendment of his existing will. This intention that the document in question should merely be a guideline necessarily excludes the distribution intention and animus testandi and thus disqualifies the document as one embodying the act of testation. See De Waal and Schoeman-Malan Law of Succession 78; Schoeman-Malan et al 2014 Acta Juridica 88-89.
26 Bekker v Naude 2003 5 SA 173 (SCA).
27 See Du Toit 2010 De Jure 156.
28 See, for instance, Giles v Henriques 2008 4 SA 558 (CC) for an example. See below for a discussion of the case.
29 See Faber and Rabie 2004 TRW 202.
the court would instead give effect to the act of testation embodied in the informal document.24 Granted, it is often more challenging to determine the presence of animus testandi in the latter instance;25 yet the forms of intention involved in the two document types are exactly the same and exist independently of the format of the document. Therefore, the act-based model offers a seamless explanation of why both formally non-compliant wills and informal documents are fully condonable.
2.2.2 Wills drafted by a third party
It is further proposed that the act-based model also lays the groundwork for addressing the issue of certain documents drafted by a third party being non-condonable. The ruling in Bekker v Naude26 seems to support the view that a testator can appropriate a document drafted by a third party (in the sense of the testator’s making known his or her animus testandi in respect of the document) only by executing the document concerned (as contemplated in section 2(3)).27 Therefore, judging by the Bekker interpretation of the section 2(3) requirements, an unexecuted will drafted by a third party cannot be condoned. Thus, in terms of documents drafted by third parties it would seem that our courts prefer to focus on who performed the drafting act instead of considering whether the testator in fact adopted the drafted document as his or her own by claiming ownership of the dispositive act(s) contained therein and forming the necessary animus testandi in respect thereof.28 Consequently testator's intention appears to play second fiddle to the drafting act per se in so far as the condonation of documents drafted by third parties is concerned.29 The outcome in Van der
Merwe v The Master,30 for instance, would have been different if it had come to light that the will in that matter had been drafted by a third party on the testator's behalf, even though the deceased clearly took ownership of the document and, as required by section 2(3), intended for it to be his will (having e-mailed it to his friend as such).
30 Van der Merwe v The Master 2010 6 SA 544 (SCA).
31 De Waal "Testamentary Formalities" 398.
In following Bekker in their approach to the (non)condonation of documents drafted by third parties, however, our courts lose sight of the fact that a testator can legally make a will without any physical involvement in the will-making process. Consequently, all the relevant forms of intention can arise independently of the physical act of drafting the document or the testator's signing of the will. For example, a testator may issue a verbal instruction to a third party to draft his or her will along with clear instructions regarding his or her dispositive intention. Once the will has been drafted and the testator has taken ownership of the dispositive act and formed the necessary animus testandi in respect thereof, the testator may have the will signed by an amanuensis (on the testator's behalf) and two competent witnesses. The result would be a valid will, even though the testator was never physically involved in making it and never personally executed (signed) it. The act-based model and intent doctrine would suggest that such a document, should it be formally non-compliant in some respect (for instance, if one witness were to sign by making a mark instead of a signature), should indeed be condonable, since the relevant forms of intention are present, even though the deceased did not physically take part in the drafting and execution processes.
In the light of the foregoing analysis, one could regard the Bekker ruling as correct in terms of its interpretation and application of section 2(3) and its alignment with what the lawmakers probably had in mind with the power of condonation.31 However, the unsatisfactory state of affairs that has arisen post-Bekker with regard to certain documents drafted by third parties, particularly viewed against the act-based model, is no longer tenable. In effect, Bekker causes testator's intention to fail, even though the deceased clearly had both dispositive intention and animus testandi, for the single reason that the unexecuted document had been drafted by a third party. This outcome clearly clashes with upholding testator's intention in South African law. It is proposed, therefore, that the Supreme Court of Appeal overturn Bekker at some point in the future and instead ratify an approach
similar to that in Back v The Master of the Supreme Court.32 In Back, Van Zyl J correctly stated:33
32 Back v The Master of the Supreme Court 1996 2 All SA 161 (CC).
33 Back v The Master of the Supreme Court 1996 2 All SA 161 (CC) 174A-C.
34 See para 2.2.1 above.
35 See for example Du Toit 2020 OUCLJ 139 who, in analysing the testamentary condonation dispensations in the Canadian provinces of Manitoba and British Columbia, points out that neither of these jurisdictions' relevant laws contain any drafting or execution requirement such as that in s 2(3) of the Wills Act. The statutory provisions in these two Canadian jurisdictions merely require the "testamentary intentions of a deceased" as the key requirement for the condonation of formally non-compliant wills or informal documents.
Another reality is that many would-be testators give full instructions as to their final wishes to their attorneys or bankers and the attorneys or bankers have draft wills prepared in accordance with such instructions. If a draft will is subsequently perused and approved in every detail by a testator, he then ... associates himself with and adopts it as his own. On a flexible interpretation of s 2(3), it may be regarded as having been drafted by him personally. As long as it is incontrovertible that the testator intended the draft will to be his will, it should be totally irrelevant whether he personally or physically drafted it.
In addition, it is suggested that the entire judgment in Back be viewed from the perspective of the act-based model: not in the way of a flexible interpretation of section 2(3), but by shifting the focus to the fact that the unexecuted document drafted by a third party still contains the act of testation, even though the testator was neither involved in drafting it, nor was it executed. The act of testation marks the presence of the testator's dispositive intention and animus testandi, and, as argued earlier,34 a document embodying these forms of intention should be condonable. Alternatively, the Wills Act may need to be amended to turn the current narrow and formalistic application of section 2(3) into an intention-oriented application. Section 2(3) may be redrafted to stress the act of testation, as the carrier of both dispositive intention and animus testandi, in accordance with the suggested act-based model. Such an amended provision would also be in line with statutes governing testamentary condonation in other jurisdictions.35
3 Scenario 2: The rectification of cross-signed mirror wills
3.1 The rectification issue in South African law
Rectification, or the correction of an error in a will, is required when the testator's expression of intention is incorrectly or inadequately captured in the will. Therefore, rectification comes into play when the will is not an
accurate expression of the testator's true wishes.36 In this regard, Corbett et al37 have stated that "[b]efore a court will rectify a will, it will require proof … that the alleged discrepancy between intention and expression was due to a mistake". Therefore, the mistake or error results in a variation between the testator's true intention on the one hand, and the intention as expressed in the will on the other. A court may correct the error that has given rise to the discrepancy by applying rectification to give effect to the testator's true intention.38
36 Henriques v Giles 2010 6 SA 51 (SCA) para 16.
37 Corbett, Hofmeyr and Kahn Law of Succession 497.
38 Du Toit 2008 Obiter 330. Say for instance the testator's true wish (intention) is for A and B to be her heirs, but the will erroneously indicates only A as her heir, or nominates A, B and C as heirs. In the former instance the will would be rectified by inserting B, and in the latter by removing C. See Henriques v Giles 2010 6 SA 51 (SCA) para 15.
39 Du Toit 2014 APLJ 64.
40 Henriques v Giles 2010 6 SA 51 (SCA) para 24.
41 In the Estate of Hennekam (dec'd) 2009 104 SASR 289.
42 The court found that it was not appropriate to utilise rectification in the "mirror wills" cases, because rectification is used to "enable the court to correct a document which does not accurately reflect the testator's intentions. It is generally concerned with rectifying mistakes as to the meaning or the contents of the will." This is in contrast to the courts' dispensing power, which is concerned with remedying documents that do not comply with the formality requirements. The court concludes: "In my view, to delete the portions of the will of the deceased's wife which the deceased actually signed, so that the document complies with the known intentions of the deceased, is of greater artificiality than to admit to probate the actual will of the deceased, despite its lack of appropriate execution." (In the Estate of Hennekam (dec'd) 2009 104 SASR 289 paras 36-37).
43 Kerridge Parry and Kerridge: Law of Succession 56, 82-83, 263.
44 Marley v Rawlings 2015 AC 129.
In Henriques v Giles the Supreme Court of Appeal ruled that so-called "cross-signed mirror wills",39 where parties have signed each other's drafted wills in error, could indeed be rectified.40 Yet the way in which instances of cross-signed wills should be dealt with is fairly contentious. In Australian law, for instance, the power of condonation or "dispensing power" is used for this purpose. In In the Estate of Hennekam (dec'd)41 the Supreme Court of South Australia concluded that rectification was not the appropriate solution to the problem and labelled it an "artificial remedy".42 In English law, on the other hand, rectification is indeed applied to cross-signed wills. English law does not currently contain a power of condonation, and until fairly recently cross-signed wills were simply regarded as invalid for lack of compliance with the formality requirements of the Wills Act.43 However, the English Supreme Court44 ruled that such wills could indeed be rectified in
terms of a broad interpretation of section 20 of the Administration of Justice Act45 in respect of a "clerical error".46
45 Administration of Justice Act, 1982.
46 Kerridge Parry and Kerridge: Law of Succession 263; Martyn et al Theobald on Wills 335. As early as in 1875, in the case of In the Goods of Hunt 1875 LR 3 P & D 250, the English courts had to rule on a case where a woman prepared two almost identical wills for herself and her sister but unfortunately executed the will that she drew up for her sister by mistake. The court refused to accept the will, because the woman did not have the required knowledge and approval regarding the document that she had executed. The court held: "[T]he lady signed as her will something which in fact was not her will" and later "if she had known of the contents she would not have signed it" (In the Goods of Hunt 1875 LR 3 P & D 250 252). Kerridge Parry and Kerridge: Law of Succession 82-83 points out that this is a typical case that can now be rectified in view of the Supreme Court's ruling in Marley v Rawlings 2015 AC 129.
47 See para 2.2.2 above.
48 Bekker v Naude 2003 5 SA 173 (SCA).
49 Henriques v Giles 2010 6 SA 51 (SCA) paras 20-21.
50 Giles v Henriques 2008 4 SA 558 (CC) para 40. It seems that condonation would definitely have been possible had the parties drafted the wills themselves.
51 Henriques v Giles 2010 6 SA 51 (SCA) paras 1-3.
52 Henriques v Giles 2010 6 SA 51 (SCA) para 21.
In South African law rectification is applied perforce to cross-signed wills drafted by third parties because, as illustrated above,47 condonation of such wills is no longer an option since the Supreme Court of Appeal's judgment in Bekker.48 This is also why in Henriques v Giles49 the court a quo stated that "[u]nfortunately the applicants in this matter could not rely on the remedial provision of section 2(3)".50 In the latter case a man and a woman signed each other's wills in error, having issued instructions to a third party to draft their respective wills, and each having read and approved their drafted will.51 The court rightly ruled that the error that had crept in was the "erroneous 'cross-signing' of the wills" (in other words, that the wills were formally non-compliant), but then proceeded to question whether the signed will could be rectified.52 Some commentators regard the court's reasoning in this regard as problematic and confusing. Viewed objectively the erroneous execution process in Henriques yielded two invalid wills: the relevant testator did not sign the relevant will as required by section 2(1)(a) of the Wills Act. From the testator's perspective, therefore, he executed the wrong document or, put differently, he did not execute his will. Yet the Supreme Court of Appeal continued to rectify the erroneously executed document to include the content of the testator's will. Granted, the court's modus operandi in this regard was made somewhat easier by the fact that the two wills at stake in Henriques were largely similar in content (in the sense that they contained virtually the same stipulations) and that rectification was therefore used to correct the few discrepancies between the two wills so that the will signed by the testator indeed represented a full
and accurate reflection of his wishes. Nevertheless, both Jamneck and Jacobs regard the court's approach in Henriques as dubious.53 Their critique relates principally to the question whether rectification is possible at all if the document that needs to be rectified is not a valid will (for not complying with the formality requirements), and whether rectification is the appropriate remedy in general when a mistake manifests itself in the erroneous signing of the document. Jacobs states that rectification should be used to remedy "an error in the document itself, brought about by erroneous formulation or scribing". He reiterates that rectification "relates to the very substance [content] of the document and not the form or formalities in respect of its execution".54
53 Jamneck 2009 THRHR 506; Jacobs 2011 De Rebus 34.
54 Jacobs 2011 De Rebus 36.
55 See Faber 2021a TSAR 504; Faber 2021b TSAR 740.
56 In Marais v The Master 1984 4 SA 288 (D) 291G, the court clearly stated that a validly executed document without a testamentary bequest – and therefore without an act of testation – does not qualify as a will. Also see De Waal and Schoeman-Malan Law of Succession 96.
57 Sonnekus 1990 TSAR 120.
58 Wiechers Testamente 26.
3.2 The act-based model as a potential solution to rectification uncertainty
It appears that whether one regards a ruling such as that in Henriques as correct or incorrect depends largely on exactly how a will is conceptualised, as well as the connection between a will and the statutory formality requirements. Here too it is suggested that the act-based model and intent doctrine provide the basis for resolving this issue in South African law. It is submitted that, when dealing with cross-signed wills (including those drafted by a third party), the focus should always be on the content of the document or, put differently, on the act of testation embodied in the document. After all, a document is typified as a will because of its content. The execution of a will in accordance with the prescripts of the Wills Act is a formality that does not in itself turn a document into a will: the document's status is determined by its testamentary content.55 Therefore, if the content of the document does not embody the act of testation – as a legal act – the document would not qualify as a will, irrespective of compliance with the statutory formality requirements.56 For this reason Sonnekus57 correctly states that on their own the formality requirements of the Wills Act have no right of existence whatsoever, but merely serve to establish the testator's last wishes following his/her death with a high degree of certainty. Wiechers58 too confirms that the witnesses' acts – in respect of compliance
with the formality requirements – do not contribute to the testator's legal act as such, but simply to the formal validity thereof.
This then highlights that, when dealing with cross-signed wills, the testator's dispositive intention and animus testandi do not pertain to the particular erroneously executed document as such but rather to the content of the checked and approved (and, therefore, adopted) document, which was ultimately wrongly executed. In this regard it is imperative to distinguish between, on the one hand a case where a document was executed after its content was approved and adopted and in respect of which the necessary animus testandi was indeed formed, and on the other a document that was executed without having been adopted, and in respect of which the required animus testandi was never formed. The effect of the forming of animus testandi in the former case is that it transforms the dispositive act into an act of testation, which affords the document embodying this act the status of a will, which in turn gains legality once executed. In the case of a cross-signed will, the testator's will – being the document that embodies the act of testation and in respect of which the necessary animus testandi was formed – was not executed in a legally valid manner, as the testator did not sign the document in terms of section 2(1)(a) of the Wills Act.
Clearly, therefore, in the light of the explanation above, rectification in the traditional, narrow sense – as the remedy to correct an error in a testator's will – is vastly different from its application to cross-signed wills, namely to replace the content of a document that was incorrectly executed as a will, for it to be turned into a particular testator's will. Rectification in the traditional, narrow sense deals with an actual will that already embodies the testator's act of testation and was validly executed. It merely involves an error in respect of the act of testation stipulated in the will, which error is corrected through rectification.59 In the event of cross-signed wills, however,
59 Although not expressly stated, it does seem as if the courts rectify wills where the act of testation is flawed. The error in the will in respect of the disposition of assets and the testator's animus testandi are closely linked. See for instance Aubrey-Smith v Hofmeyr 1973 1 SA 655 (C) 659E-H, where the court said: "There would have been no animus testandi in respect of the words or clauses inserted in error." Even more importantly, see Botha v The Master 1976 3 SA 597 (EC) 603A-C, where the court, with reference to Aubrey-Smith, rightly not only links the intention to the words in general, but specifically to words in relation to the "bequest" and "disposition recorded in the will". Also see De Waal and Schoeman-Malan Law of Succession 233. This position is different from an error in a will with regard to the act of revocation, for instance. In Ex Parte Lutchman 1951 1 SA 125 (T), a revocation clause (as act of revocation) was erroneously included in a will. The court chose not to deal with the matter by way of the rectification of wills, as rectification is aimed at deleting a section from the will that was inserted without the necessary animus testandi. In this case, the revocation clause that had been inserted without the
required animus revocandi was simply declared invalid. Also see De Waal and Schoeman-Malan Law of Succession 92, 96, 244
60 In the context of the act-based model, the fact that the documents in question are either very similar or completely different in content would seem to be legally irrelevant. The reason is that, although the documents are almost exactly the same, they are undoubtedly distinguishable insofar as whose act of testation is embodied in each of them concerned. Only the document that contains the dispositive intention of the testator, coupled with his/her animus testandi – and therefore his/her act of testation – renders that specific document the will of that specific testator.
rectification is applied to cause a document that was erroneously executed as the testator's will to become such a testator's will. The latter approach loses sight of the fact that the executed document does not represent the testator's will, for the simple reason that it does not embody the testator's act of testation. In my view this holds true even for cross-signed wills that are largely similar in content, as was the case in Henriques.60
The act-based model draws a distinction between the acts included in the content of a will, particularly the act of testation as a legal act, and the execution acts in compliance with the statutory formality requirements. Nevertheless, the act-based model does acknowledge the undeniable nexus between the act of testation as articulated in a will on the one hand, and compliance with the formality requirements by execution of a will on the other. Just as a testator who is not physically present to execute his/her will would not be able to execute a blank page as a "symbolic" gesture, having listened to and approved the will telephonically or via Skype, the effect of cross-signed wills is that the testator did not personally execute his/her will. Therefore, testator's intention and the act of testation on the one hand, and compliance with the formality requirements on the other, are technically not linked. For this reason, in accordance with the intent doctrine condonation is the appropriate remedy to excuse the formal non-compliance in respect of the execution of the will as the carrier of the act of testation; not rectification, which is used to correct a flawed act of testation where the expression of intention is an inaccurate reflection of the testator's wishes.
It is nevertheless acknowledged that the court's hands were tied in Henriques insofar as it could not invoke the Wills Act's condonation provision to preserve the testator's intention. It is moreover acknowledged that, insofar as rectification concerns the correction of errors in the substance or content of a testamentary document and, moreover, the act-based model's overall focus on the preservation of testator's intention, the court's reliance on rectification in Henriques attained an entirely supportable outcome, even if the Henriques decision may well be situated on rectification's outer limits. It nevertheless bears repetition that, for the
reasons mentioned above, condonation is legally speaking a far more elegant and indeed palatable way of dealing with cross-signed mirror wills. This view also presents yet another reason why the "drafted or executed" requirement of section 2(3) of the Wills Act – the requirement that stood in the way of condonation in Henriques – is problematic, and why the lawmakers should strongly consider removing or amending this requirement to allow for the condonation of wills drafted by third parties. Such removal or amendment would open the door to the condonation of cross-signed wills, thereby dealing with the matter appropriately in accordance with the specific tenets of the act-based model, and obviate the need for unnecessarily pushing the boundaries of rectification in these instances.
This contribution clearly shows that, in terms of the act of testation as embodied in a will, a processual view premised on an act-based model (as opposed to the traditional requirements model) presents the appropriate methodology to address and resolve issues pertaining to testator's intention in South African law of succession.
As far as section 2(3) of the Wills Act is concerned, the proposed act-based model requires the focus to be on the act of testation – and specifically on the dispositive intention and animus testandi underlying the act of testation. Due to the absence of an act-based model in South African law, the courts are currently left speculating on or guessing as to the testator's intention, which is not always clearly identified, defined or contextualised. The inevitable consequence is that the courts in certain instances overemphasise other aspects, such as the form of the document, in an attempt to find the evasive intention for the purposes of section 2(3). In addition, an act-based model could serve as a departure point for changing the legal position established in Bekker v Naude, ensuring that the decision to condone (or not to condone) does not rely on who drafted the document, but on whether the document embodies the deceased's intention. The focus should be on the act of testation, and if found to be present (no matter in which shape or form or by whom it was drafted), the document embodying the act of testation should be condoned.
Regarding rectification, the act-based model highlights the important distinction between content and formality – the act of testation as opposed to compliance with the statutory formality requirements through the execution of a will. Rectification is appropriate where an error has caused a discrepancy between the testator's true intention and the intention as
expressed in the act of testation embodied in the will. Rectification seems less appropriate when dealing with cross-signed wills, which do not involve a discrepancy between the testator's intention and the act of testation on account of an error but instead are the result of a flawed execution process in respect of the wills concerned. Condonation appears to be much better suited for correcting the formal non-compliance of cross-signed wills – a claim that is fully supported by the act-based model.
In conclusion, it is proposed that an approach to testator's intention that is based on the act of testation, and which clearly distinguishes such an act from the other acts in the will-making process (particularly the acts of execution), would ensure that testator's intention is done justice in South African law of succession. For this reason, the processual view of a will as the product of a will-making process, the proposed act-based model as well as the introduction of an intent doctrine in South African law of succession are all strongly supported.
Corbett, Hofmeyr and Kahn Law of Succession
Corbett MM, Hofmeyr G and Kahn E The Law of Succession in South Africa 2nd ed (Juta Cape Town 2001)
De Waal "Testamentary Formalities"
De Waal MJ "Testamentary Formalities in South Africa" in Reid KGC, De Waal MJ and Zimmermann R (eds) Comparative Succession Law: Testamentary Formalities (Oxford University Press Oxford 2011) 381-403
De Waal 2016 Annu Surv SA L
De Waal MJ "The Law of Succession (Including Administration of Estates) and Trusts" 2016 Annu Surv SA L 959-975
De Waal and Schoeman-Malan Law of Succession
De Waal MJ and Schoeman-Malan MC Law of Succession 5th ed (Juta Cape Town 2015)
Du Toit 2008 Obiter
Du Toit F "Rektifikasie (en Kondonasie?) van Verkeerdelik-Ondertekende Testamente" 2008 Obiter 329-338
Du Toit 2010 De Jure
Du Toit F "Is Rektifikasie van 'n Dokument Gekondoneer Ingevolge Artikel 2(3) van die Wet op Testamente Moontlik?" 2010 De Jure 149-157
Du Toit 2014 APLJ
Du Toit F "Testamentary Rescue: An Analysis of the Intention Requirement in Australia and South Africa" 2014 APLJ 56-82
Du Toit 2020 OUCLJ
Du Toit F "Remedying Formal Irregularities in Wills: A Comparative Analysis of Testamentary Rescue in Canada and South Africa" 2020 OUCLJ 139-162
Faber 2021a TSAR
Faber JT "A Conceptual View of the Act of Testation to Elucidate a Testator's Intention in the South African Law of Succession: A Proposed 'Act-Based Model' as Opposed to the Traditional 'Requirements Model' (Part 1)" 2021 TSAR 504-520
Faber 2021b TSAR
Faber JT "A Conceptual View of the Act of Testation to Elucidate a Testator's Intention in the South African Law of Succession: A Proposed 'Act-Based Model' as Opposed to the Traditional 'Requirements Model' (Part 2)" 2021 TSAR 740-753
Faber and Rabie 2004 TRW
Faber JT and Rabie PJ "Die Behoefte aan ŉ Wyer Artikel 2(3) van die Wet op Testamente 7 van 1953 (Soos Gewysig): ŉ Kritiese Beskouing" 2004 TRW 198-203
Jacobs 2011 De Rebus
Jacobs J "Cross-Signed Wills" 2011 De Rebus 34-37
Jamneck et al Law of Succession
Jamneck J et al The Law of Succession in South Africa 3rd ed (Oxford University Press Cape Town 2017)
Jamneck 2008 THRHR
Jamneck J "Testeerbevoegdheid, Animus Testandi, Testeervryheid, Bedoeling en die 'Bedoeling' Ingevolge Artikel 2(3) van die Wet op Testamente (1)" 2008 THRHR 603-612
Jamneck 2009 THRHR
Jamneck J "Die Anomalie Veroorsaak deur die Uitleg van Artikel 2(3) van die Wet op Testamente in Bekker v Naude 2003 5 SA 173 (HHA): Giles v Henriques 2008 4 SA 558 (K)" 2009 THRHR 502-507
Kerridge Parry and Kerridge: Law of Succession
Kerridge R Parry and Kerridge: The Law of Succession 13th ed (Sweet & Maxwell London 2016)
Martyn et al Theobald on Wills
Martyn JGR et al Theobald on Wills 8th ed (Thomson Reuters London 1927)
Schoeman-Malan et al 2014 Acta Juridica
Schoeman-Malan et al "Section 2(3) of the Wills Act 7 of 1953: A Retrospective and Critical Appraisal of Some Unresolved Issues" 2014 Acta Juridica 78-103
Sonnekus 1990 TSAR
Sonnekus JC "Videotestamente Naas Skriftelike Testamente" 1990 TSAR 114-133
Wiechers NJ Testamente: 'n Kortbegrip (Juta Cape Town 1988)
Wood-Bodley 2011 SALJ
Wood-Bodley MC "Suicide Notes, Wills, Testamentary Capacity, and s 2(3) of the Wills Act 7 of 1953: Smith v Parsons NO; Henriques v Giles NO" 2011 SALJ 612-620
Aubrey-Smith v Hofmeyr 1973 1 SA 655 (C)
Back v The Master of the Supreme Court 1996 2 All SA 161 (CC)
Bekker v Naude 2003 5 SA 173 (SCA)
Botha v The Master 1976 3 SA 597 (EC)
Dryden v Harrison (WCC) (unreported) case number 11912/17 of 20 May 2019
Ex parte Estate Davies 1957 3 SA 471 (N)
Ex Parte Lutchman 1951 1 SA 125 (T)
Ex parte Maurice 1995 2 SA 713 (CC)
Ex parte Williams: In re Williams's Estate 2000 4 SA 168 (T)
Giles v Henriques 2008 4 SA 558 (CC)
Henriques v Giles 2010 6 SA 51 (SCA)
In the Estate of Hennekam (dec'd) 2009 104 SASR 289
In the Goods of Hunt 1875 LR 3 P & D 250
Marais v The Master 1984 4 SA 288 (D)
Marley v Rawlings 2015 AC 129
Oosthuizen v Die Weesheer 1974 2 SA 434 (O)
Osman v Nana (37220/2018)  ZAGPJHC 161 (3 May 2019)
Osman v Nana 2021 JOL 50242 (GJ)
Smith v Parsons 2010 4 SA 378 (SCA)
Taylor v Taylor 2012 3 SA 219 (ECP)
Van der Merwe v The Master 2010 6 SA 544 (SCA)
Van Wetten v Bosch 2004 1 SA 348 (SCA)
Webster v The Master 1996 1 SA 34 (D)
Administration of Justice Act, 1982
Wills Act 7 of 1953
List of Abbreviations
Annu Surv SA L
Annual Survey of South African Law
Australian Property Law Journal
Oxford University Commonwealth Law Journal
South African Law Journal
Tydskrif vir Hedendaagse Romeins-Hollandse Reg
Tydskrif vir Regswetenskap
Tydskrif vir die Suid-Afrikaanse Reg