Petropulos and Another V Dias 2020 (5) SA 63 (SCA)
Petropulos and Another V Dias 2020 (5) SA 63 (SCA)
JUDGMENT
Reportable
Case No: 1055/2018
and
Delivered: This judgment was handed down electronically by circulation to the parties’
representatives by email, and by publication on the Supreme Court of Appeal website
and release to SAFLII. The time and date for hand down is deemed to be 10h00 on
the 21st day of May 2020.
Summary: Neighbour law – duty of lateral support – owed to land and buildings on
it – English principle of lateral support, although influential, not part of our law – strict
liability – available in principle for breach of lateral support.
2
___________________________________________________________________
ORDER
___________________________________________________________________
On appeal from: Western Cape Division of the High Court, Cape Town (Bozalek J
sitting as court of first instance): judgment reported sub nom Dias v Petropulos and
Another [2018] ZAWCHC 93; 2018 (6) SA 149 (WCC); [2018] 4 All SA 153 (WCC).
The appeal is dismissed with costs, such costs to be paid by the appellants jointly and
severally, the one paying the other to be absolved.
___________________________________________________________________
JUDGMENT
___________________________________________________________________
[1] This appeal concerns the nature, scope and ambit of the duty of lateral support
owed in respect of contiguous properties. The court a quo, the Western Cape Division
of the High Court, Cape Town (Bozalek J), concluded that the duty of lateral support
is owed not only in respect of land but also buildings constructed on the land, save
where such land has been ‘unreasonably loaded so as to place a disproportionate or
unreasonable burden on the neighbouring land.’ The appeal is with leave of the court
a quo.
[2] The facts are comprehensively set out in the judgment of the court a quo, which
has been reported sub nom Dias v Petropulos and Another [2018] ZAWCHC 93;
2018 (6) SA 149 WCC; [2018] 4 All SA 153 (WCC). Briefly stated, the facts are: The
first appellant, Ms Petropulos, the respondent, Mr Dias, and Mr Dawid Venter (Mr
Venter), Mr Kenneth Wentzel (Mr Wentzel) and Mr Peter Babrow (Mr Babrow), owned
adjoining properties in Camps Bay, Cape Town, on a steeply sloping mountainside.
The land on which the properties are situated, is bound by Theresa Avenue, on the
upper end of the mountain, and Barbara Road, on the lower end. The respondent’s
property is situated in Theresa Avenue. It shares a boundary with the properties of the
3
first appellant and Mr Venter, both of which are situated downhill in Barbara Road. At
the relevant time, being March to August 2008, all of the properties, except for the first
appellant’s, which was still an undeveloped erf, had houses built on them.
The respondent’s house had been completed in 1994.
[3] During March 2008 the first appellant and Mr Venter each undertook
excavations on their respective properties, near the respective boundaries with the
respondent’s property. The excavation on the first appellant’s property was in
preparation for the building of a house, while Mr Venter was preparing to build an
additional garage. The building works on Mr Venter’s property were uneventful, and
were completed in April 2008. The excavation on the first appellant’s property, on the
other hand, involved fairly substantial excavations to produce three tiers, and for a lift
shaft. To provide lateral support, the three levels were each secured by a retaining
wall. Mr Naumann, the first appellant’s husband, an experienced builder, undertook
the building works on the property.
[4] From May 2008, problems became evident on the respondent’s property. A dip
appeared in the garden; furrows appeared in the garden between the respondent’s
property and the first appellant’s; the respondent’s terra-force wall, and the ground
under it, collapsed during the course of the construction of the top retaining wall.
Between 23 July and 1 August 2008, there was a major movement in the underlying
ground. The entire slope on which respondent’s property is situated, subsided. The
respondent’s property moved laterally and downwards towards the excavation on the
first appellant’s property, resulting in extensive structural damage to the property.
Cracks appeared in the walls, tiles, floor slabs, the boundary wall as well as the
driveway adjacent to Theresa Road. The pool rail detached from the house and a
hairline crack appeared in it. There were problems on Mr Venter’s property, too. The
property subsided and cracks appeared thereon. On 23 July 2008 Mr Venter, because
of safety concerns, was forced to abandon the property.
[5] The respondent attributed the damage to his property to the excavations
undertaken by the first appellant and Mr Venter on their respective properties. He
instituted a claim for damages against both, based on strict liability, for breach of the
4
duty to provide lateral support. It was alleged, among others, that the slope mobilised
through the mechanism of ‘a shallow slip circle with uplift at the toe, resulting in vertical
upward bulging of the ground surface between Barbara Road and the structures facing
onto it; and lateral movement towards Barbara Road.’ This allegation became the focal
point of the first appellant’s case during the trial, as will be clear later. The first
appellant and Mr Venter each defended the action and denied liability. The first
appellant also joined the second appellant, Nik Moroff & Associates, the project
engineer for the works on her property, as a third party to the proceedings.
[6] Before the trial commenced, an order was made, by a different judge, to
adjudicate the following issues separately in terms of Rule 33(4) of the Uniform Rules
of Court:
(a) Whether a common law duty to provide lateral support to the respondent’s property
was owed by each of the first appellant and Mr Venter properties;
(b) Whether the excavations carried out on each of the above properties in May or
June 2008 breached this duty to provide lateral support;
(c) If so, whether as a result of the respondent’s property being so deprived of such
lateral support by such excavations the scree slope on which respondent’s property
was situated mobilised and subsided in June 2008.
[7] The trial commenced before the court a quo on 21 November 2016. During the
course of the trial, Mr Venter reached an agreement with the respondent and ceased
participation in the action, hence he takes no part in this appeal. On 30 July 2018 the
court a quo delivered its judgment. It declared that: the first appellant and Mr Venter
owed the respondent a duty to provide lateral support to his property; the excavations
undertaken on their respective properties breached that duty, as a result of which the
slope on which the respondent’s property is situated, mobilised and subsided. No
substantive order was made against the second appellant, except that it was ordered
to pay the respondent’s costs, jointly and severally with the first appellant.
[8] In this court, it was contended on behalf of the appellants that: First, the first
appellant did not owe a duty to provide lateral support to the respondent’s property,
inasmuch as the latter’s property was no longer in its natural state. Second, that the
5
excavations on the first appellant’s property did not breach the duty to provide lateral
support. Third, that the excavation on the first appellant’s property was not linked
sufficiently closely to the harm suffered by the respondent for legal liability to ensue
(causation). And, fourth, that on the facts of this case, it is inconceivable that the first
appellant should be held liable to the respond in the absence of a finding of fault. Each
of these contentions will be considered in turn.
Is the duty of support owed only in respect of land in its natural state?
[9] In answering this question, the learned judge undertook an extensive analysis
of the various authorities. This included East London Municipality v South African
Railways and Harbours 1951 (4) SA 466 (E). There, it was held that our law of lateral
support was the same as English law, in terms of which the right is confined to land in
its natural state and does not extend to constructions such as buildings on it.
[10] The court a quo also considered the decision of this court in Anglo Operations
Ltd v Sandhurst Estates (Pty) Ltd [2006] ZASCA 118; 2007 (2) SA 363 (SCA); [2007]
2 All SA 567 (SCA). The court a quo declined to follow East London Municipality, and
concluded (at para 59)1 that in our law, the duty of lateral support is owed to
neighbouring or contiguous pieces of land as well as the buildings on it. However, at
para 60,2 the court expressed the following caveat to that general principle:
‘However, too broad a formulation of the right or duty of lateral support could lead to conceptual
and equitable difficulties, particularly where the contiguous parcels of land are situated on a
slope. Where a property has been unduly or unreasonably loaded through the erection of
disproportionately large or heavy structures, it would seem unfair in my view that a
neighbouring piece of land should attract an equivalently onerous duty of lateral support’.
Later, at (para 63)3 the learned judge summarised the position as follows:
‘In the result, I consider that the appropriate approach is to hold that a duty of lateral support
extends not only to land but also to buildings, save where such land has been unreasonably
loaded so as to place a disproportionate or unreasonable burden on the neighbouring land’.
[11] The appellants submitted that the court a quo was wrong by not concluding that
the duty of lateral support in our law is similar to English law, in terms of which the duty
is owed to land only in its natural state, and does not extend to artificial structures such
as the buildings on it. Further, in English law, support for buildings can only be obtained
by means of a servitude, which is obtainable by a user of a building after at least 20
years or by agreement. This principle of English law was enunciated more than a
century ago in Dalton v Henry Angus & Co (1881) 6 App Cas 740 and is best
expressed in the oft-quoted passage of Lord Penzance’s speech at 804:
‘[I]t is the law, I believe I may say without question, that at any time within twenty years after
the house is built the owner of the adjacent soil may with perfect legality dig that soil away and
allow his neighbour’s house, if supported by it, to fall in ruins to the ground.’
[12] It is necessary to examine the development of our own law in this regard. The
duty of lateral support owed to an adjacent landowner corresponds with the
neighbour’s entitlement to such support. This means that the right to lateral support is
reciprocal between neighbouring landowners. That principle was first accepted into
South African law as a principle of neighbour law in London and SA Exploration Co v
Rouliot (1890-1891) 8 SC 74. There it was held (at 93) that the right of lateral support
is a ‘well established natural right’, incidental to the ownership of the property and not
servitudal in nature. Rouliot was followed, albeit on different grounds, in Johannesburg
Board of Executors and Trust Company Limited v Victoria Building Company
Limited (1894) 1 OR 43.
[13] However, the application of the principle to situations where land has been
improved with buildings or structures on it, and where excavation causes subsidence
and damage to buildings, has given rise to two contrasting views. Van der Walt4
explains the divergent underlying philosophies thus:
‘Milton argues that the right of lateral support is explained in terms of two theories. According
to the one theory, the right of lateral support is a servitude arising from the natural situation of
land (as opposed to servitudes created by grant or prescription). According to this theory, the
right would be restricted to the land in its natural state and would not apply to buildings on the
land. Furthermore, any infringement of the right would arise from the mere withdrawal of lateral
support and not only from damage caused by such withdrawal, with the implication that
prospective damages could be awarded. The second theory explains the right of lateral
support as a natural right of property that is based on the principle sic utere tuo alienum non
laedas and protected by nuisance law. Seen in this way, the right pertains to mutual respect
for normal use of land and there is no reason why it should not apply to buildings as well.
Furthermore, liability for infringements of the right would arise from actual damage and not
simply from withdrawal of the support, and consequently prospective damages could not be
claimed. Liability would be strict.’
As I have shown, Rouliot grounded the introduction of these principles in our law on
the second basis.
[14] In Victoria and in Phillips v South African Independent Order of Mechanics and
Fidelity Benefit Lodge and Brice 1916 CPD 61 it was held that Roman and Roman-
Dutch law recognised a right of lateral support for land and buildings. Consequently
the defendants were held liable for the collapsing of buildings caused by the
excavation of land on the boundary between two tenements. See also Demont v Akals’
Investments (Pty) Ltd and Another 1955 (2) SA 312 (N), where Selke J (at 316B-E)
said:
‘An owner of land is normally entitled to expect and to require from land contiguous to his own
such lateral support as would suffice to maintain his land in a condition of stability if it were in
its natural state. A landowner can, of course, alter the condition of his land, for example by
excavating or building on it, but he cannot normally, by the mere fact of doing that, acquire
greater or different rights to lateral support. His basic rights … remain the same whatever he
may choose to do with his land. They are rights ancillary to his ownership, and they are
enjoyed reciprocally by him and by all owners of contiguous land; and, while they exist
unimpaired, any infringement of them by the withdrawal or disturbance of lateral support
furnishes him with a cause of action’.
[15] However, a different path was followed in Douglas Colliery Ltd v Bothma and
Another 1947 (3) SA 602 (T) and in East London Municipality.5 Those cases relied
heavily on English law and consequently concluded that lateral support is owed only
5East London Municipality was uncritically followed in Gordon v Durban City Council 1955 (1) SA 634
(N) and in John Newmark & Co (Pty) Ltd v Durban City Council 1959 (1) SA 169 (N).
8
to land in its natural state, and not to artificial structures on it. Douglas concerned
mining law. Neser J held (at 612) that there is no natural right of support for that which
is artificially constructed on land. The learned judge relied on a passage in Halsbury
Laws of England (Hailsham ed, vol 22 under the title Mines) in which the following is
stated at para 1341:
‘There is no natural right of support for that which is artificially constructed on land: such a
right cannot exist ex jure naturae for the thing itself did not so exist. Therefore any right to the
support of such an artificial burden must in each case be acquired by grant, or by some means
equivalent in law to a grant. Thus it may be acquired by express grant, or implied grant, or by
prescription, or it may be created by statute.’
[16] In East London Municipality, a landowner had granted to the municipality and
the public in general a public road over his property. The municipality laid high tension
electric cables along the road. The defendant, in carrying out his quarrying operations,
removed the lateral support and caused a subsidence. Reynolds J held that in regard
to artificial constructions on land our law was the same as English law. Accordingly he
concluded, in line with English authorities, that the right of lateral support extends only
to land in its natural state and not to constructions such as buildings on it.
[17] Reynolds J (at 482H-484E) expressly declined to follow Victoria on the basis
that the Roman law authorities relied on by Morice J in Victoria were no authority for
the conclusion that lateral support was owed not only to neighbouring land but also to
buildings on it. The learned judge then referred to Halsbury Laws of England
(Hailsham ed (vol 11, para 640) in which the following is said:
‘The mere fact, however, that there are buildings on his land does not preclude an owner from
his right against a neighbour or subjacent owner who acts in such a manner as to deprive the
land of support, so long as the presence of the buildings does not materially affect the
question, or their additional weight did not cause the subsidence which followed the withdrawal
of the support.’
[18] Over 50 years after the decision in East London Municipality, this court in Anglo
Operations had to consider whether the principle of neighbour law should be extended
to govern the relationship between mineral rights holders and owners of the same
land. It was held that the principle should be restricted to the right of lateral support as
9
between neighbouring landowners, and that the relationship between the landowner
and the holder of mineral rights in the same land is regulated by the principle of
servitude. In the course of its judgment, the court considered the effect of Rouliot, in
respect of which it was pointed out (in para 8) that the gravamen of the decision was
that ‘a rule, similar in content to the English rule of lateral support, which provides
landowners, as an intrinsic element of their ownership, with the right of adjacent
support of their land, should be incorporated into our law’. The court took the view
that the origin of the principle was unimportant.
[19] After dealing with the conceptual differences between English law and our law,
Brand JA cautioned (at para 17) with reference to Rouliot:
‘Equally erroneous, in my view, is the statement that De Villiers CJ decided to incorporate the
English doctrine of lateral and subjacent support, with all its ramifications, into our law. On the
contrary, I agree with the statement by the Court a quo (at 366B) that what had happened in
Rouliot was that:
“De Villiers CJ and Smith J simply introduced, as Judge-made law, a rule which they regarded
as common to all civilised systems of law because, as they perceived it, a lacuna existed. The
Judges did not concern themselves with the exact pedigree of the rule. . . . The rule was
introduced because it was regarded as just and equitable.”’
[20] It would thus seem that one of the ‘ramifications’ of the English doctrine of
lateral support, which Brand JA cautioned against, is the slavish adoption of the
restriction of lateral support being owed to neighbouring land only, and not extending
the duty to buildings constructed thereon. This is surely understandable. English law
on this aspect is rigid, and results in anomalies, as demonstrated in the passage from
Dalton. Therefore, the significance of Anglo Operations is two-fold. First, it affirmed
Rouliot as the correct statement of our law on lateral support. Second, it qualified
Rouliot, and brought the principle of lateral support within the sphere of our neighbour
law.
[21] In our neighbour law, fairness and equity are important considerations. As
Hoexter JA explained in Regal v African Superslate (Pty) Ltd 1963 (1) SA 102 (A) at
114G those considerations are the basis of the law between neighbours. Furthermore,
in our constitutional context, the principle of lateral support must find expression in the
10
[22] It is significant that in at least one common law jurisdiction, Singapore, this
principle has been jettisoned. In Xpress Print Pte Ltd v Monocrafts Pte Ltd and Another
[2000] SGCA 37; [2000] 3 SLR 545, the appellant and the first respondent were
neighbouring landowners. As a result of excavation work done by the first respondent
on his land for the purposes of construction, the building on the appellant’s land
suffered massive damage. The appellant sued for, among others, wrongful
interference of support, which was dismissed by the trial judge on the basis of English
law as set out in para 11 above.
[23] On appeal to it, the Court of Appeal of Singapore held that the right of support
enjoyed by a neighbouring landowner extended beyond the land in its natural state to
the buildings erected thereon. In arriving at this conclusion, the court took the view
that the right of support must have its roots in ‘the principles of reciprocity and mutual
respect for each other’s property (at para 43). With regard to English law, the court
observed (at paras 33 and 37):
‘English law on the subject of the right of support … contains a number of curious propositions.
If my neighbour’s land is in its natural state, I may not remove the soil on my land without
providing alternative support for his land; but if my neighbour expends money and effort in
building a bungalow on his land, then I may excavate with impunity, even though his bungalow
may crumble to the ground. Yet, my liberty to ignore the support required by his house is not
perpetual, but lasts only for 20 years, at which time any indolence in pursuing my right to
remove my soil is transformed into a positive right of support in respect of his dwelling. . . .
Perhaps only lawyers can understand and appreciate how a simple issue such as this, through
the process of law, comes to be governed by a mass of convoluted and irreconcilable rules;
surely only the bravest among them would attempt to explain it to the average citizen. For our
part, we fail to see any legal principle capable of supporting the distinctions drawn by the
cases. Further, we are of the view that the proposition that a landowner may excavate his land
with impunity, sending his neighbour`s building and everything in it crashing to the ground, is
6Per Madala J in S v Makwanyane 1995 (3) SA 391 (CC); 1995 (2) SACR 1 (CC) para
236.
11
a proposition inimical to a society which respects each citizen`s property rights, and we cannot
assent to it’.
[24] These remarks are apposite, and accord with the principles of our own
neighbour law. So viewed, and in the light of this court’s exposition in Anglo
Operations, it is clear that the courts in Douglas and East London Municipality erred.
It follows that those decisions are not to be taken as correctly reflecting the position of
our law. The court a quo was accordingly correct in holding that the duty of lateral
support was not limited to land in its natural state, but extends to buildings on the land.
[25] However, as stated earlier, the court a quo articulated an exception to that
general principle. The court said that a duty of lateral support extends not only to land
but also to buildings, save where such land has been ‘unreasonably loaded so as to
place a disproportionate or unreasonable burden on the neighbouring land’. What
exactly the court a quo intended to convey by the quoted expression is unclear. The
exception is not without practical difficulties. A typical example is that of a landowner
who builds his or her home in full compliance with town planning and building
regulations and in accordance with architectural plans. In terms of the exception, such
an owner bears the onus to prove that the building had not ‘unduly or unreasonably
loaded’ the land, or that it is not ‘disproportionately large’ or ‘a heavy structure’. That
is untenable.
[26] Furthermore, the philosophical foundation of the exception seems, with respect,
doubtful. The learned judge relied heavily on the views of Professor Milton for the
conclusion that the English principle of lateral support is not part of our law. The
learned judge, said:
‘Professor Milton argues that the exception whereby the English law does not apply to all
artificial erections on land “so long as the presence of the buildings does not materially affect
the question, or the additional weight did not cause the subsidence which followed the
withdrawal of support” was doubtfully of any real value.’
However, in the same article, the learned author stated:
‘It is an inevitable tendency of modern life for more and more people to gravitate to cities. As
a result larger buildings must be erected to accommodate them and provide employment. The
larger the buildings, the greater the pressure on the soil and the less the duty of lateral support
12
owed by neighbouring land. This, it is submitted, is an illogical and unrealistic approach and,
on principle, it should not be preserved’.7
[27] The approach of the court a quo therefore appears incongruous. Furthermore,
it unwittingly introduces a feature of the English principle of lateral support, referred to
in East London Municipality, as set out in para 17 above. This is the very principle
which the court a quo had correctly declined to follow. It follows that the exception the
court a quo sought to introduce cannot be supported. As I demonstrate later in the
judgment, there are sufficient safeguards in our law to meet the concerns sought to be
addressed by this exception.
Did the excavations on the first appellant’s property breach the duty of lateral
support owed to the respondent?
[28] Seven witnesses testified on behalf of the respondent, two for the first appellant.
The second appellant did not call any witnesses. For purposes of this appeal, only the
evidence of the two geo-technical experts, Dr McStay and Dr Day is relevant. The
reason for this is that it is no longer in dispute that the respondent’s property was
damaged by the slope failure in July and August 2008. Both appellants have, in their
respective heads of argument in this court, conceded that aspect. Implicit in this, is the
acceptance that there was no prior damage or structural defects on the respondent’s
property before the slope failure. That issue is one in respect of which the respondent,
his wife, Mr Wentzel, Mr Babrow and Mr Naumann all testified. The other witness was
Ms Valentia Papanicolaou, whose evidence related to the measurements of the
ground movement from the end of July. Nothing turns on her evidence in the appeal.
7 Quoted in para 144 of the judgment of the court a quo (Footnote omitted.)
13
[30] The court a quo gave a commendably detailed exposition of their evidence. I
would therefore focus on what I consider the salient features of their respective
opinions. It was common cause between them that there was a slope failure which
caused ground movement on the affected properties. However, they differed on the
cause and mechanism of the slope failure. I find it convenient to commence with Dr
Day’s evidence.
[31] The defining theme of Dr Day’s evidence was his distinction between
mechanisms of slope failure – one as a result of the removal of lateral support, and
the other, slope instability. He went on to explain how each of them manifested. In
respect of lateral support failure, the primary cause of both ground movement and
failure is a reduction in the lateral (horizontal) pressure exerted on the face of the
excavation. Here, the ground movement is confined to the area of excavation.
Regarding the failure due to slope instability, Dr Day explained that it is normally
characterised by a rotational or translational movement on the ground above the
failure surface. In the event of a rotational failure, a scarp may develop at the top of
the failing mass and bulging may occur at the toe. Unlike in the failure caused by lateral
support, here the area of slope instability is generally not confined to a particular
property, but may pervade a general area.
[32] Applying these suppositions to this case, Dr Day testified that the failure was
caused by the removal of the weight of material from the toe of an already
compromised slope, the mechanism of which is a deep seated circular slip failure. On
this mechanism, according to Dr Day, the failure would not be through the removal of
lateral support, but attributable to the general instability of the hillslope, which, in turn,
was caused by a multiplicity of historical factors, including the earlier excavations and
loading of the affected properties when houses were built thereon, starting from the
early 1980s.
[33] According to Dr Day, the movement of the slope was triggered by a combination
of the excavation at the toe of the slope on the properties of the first appellant and Mr
Venter, and the added weight at the top of the slope on the properties of the
respondent and Mr Babrow. He explained further that the excavation at the toe of the
14
slope had two effects. Firstly, it reduced the weight of the soil at the toe. Secondly, it
reduced the shearing resistance of the soil over the part of the failure plane, below the
excavated area. When that happened, given the already compromised slope,
according to him, the excavation resulted in slope failure.
[34] In line with his mechanism distinction theory, Dr Day went on to explain that if
the ground movement was only as a result of the removal of lateral support, it would
have been confined to the area immediately above the retaining wall, ie it would have
a localised effect. As there was no sign of ground failure in the area immediately above
the retaining walls, Dr Day postulated that the ground movement was caused by
general instability of the slope rather than the removal of lateral support. This, as stated
earlier, was one of the ways in which failure due to slope instability manifested itself,
ie it generally pervades a general area, rather than confinement to a particular
property. Dr Day also thought it significant that when the slope mobilised, neither the
excavation itself nor the retaining walls built by Mr Naumann failed, but continued to
support the face of the excavation. This included the portion of the respondent’s land
that fell inside the failure zone. According to Dr Day, this further supported his view
that the lateral support afforded to the respondent’s property had not been
compromised.
[35] I turn now to the evidence of Dr McStay. The essence of his evidence was that
the deep-seated movement, which occurred under the properties of the first appellant
and the respondent, was a slope failure triggered by the removal of lateral support due
to the excavation on the first appellant’s property. According to him, the mechanism of
the failure was a progressive one, ie a series of smaller slip planes immediately above
the face of the excavation. In Dr McStay’s opinion, both his ‘progressive’ failure and
Dr Day’s deep- seated circular slip failure theories resulted from the removal of lateral
support because the mechanism in each case was the same, namely the excavation
on the first appellant’s property, which was the main triggering mechanism for the
slope instability. Thus, explained Dr McStay, it was largely irrelevant whether there
was a series of small progressive failures or the existence of a deep slip circle.
15
[36] Dr McStay further testified that the respondent’s house itself did not appear to
have undergone extreme lateral movement but rather relatively small scale vertical
settlement. This suggested that the original foundation of the house was largely below
the active slip circle causing the lateral movement. According to him, there was a
vertical down movement rather than just uplift, as suggested by Dr Day. To support
this view, he had regard to the crack in the paving between the respondent’s garage
and Theresa Avenue, which movement straddled two properties. Dr McStay also
explained why the excavations on the first appellant’s property stood for some time
before they affected the respondent’s property. According to him, this was not unusual,
as a slope failure normally occurred over a period of time, and not immediately,
especially on a deep-seated circle such as the one in the present case.
[37] That summarises the evidence of the two experts. To consider their competing
contentions, one has to bear in mind, the objective facts. Key among those is that the
respondent’s property was damaged when it moved laterally and downwards towards
the excavation on the first appellant’s property. This happened because lateral
support, previously provided by the first appellant’s property to the respondent’s
property, had been removed. Given these considerations, the exact mechanism which
caused the removal of lateral support is unimportant. The distinction by Dr Day in this
regard is artificial, has neither a factual nor legal basis, and is not borne out by the
objective facts. It was rightly rejected by the court a quo.
[38] A further string to the first appellant’s bow was this: as the respondent’s
property was contiguously situated on a slope with other properties, the weight of the
first appellant and Mr Venter’s properties was meant to support the entire slope, and
not only the respondent’s property. Accordingly, so went the argument, following the
slope mobilisation and damage to his property, the respondent does not, as a matter
of law, have a cause of action for breach of lateral support. The court a quo rejected
this submission as follows (at para 111):8
‘[O]ne reason is the inherent illogicality of the proposition that if an excavation is of such large
proportions that it causes not simply a localised subsidence or failure but also one which
undermines an entire slope comprising multiple properties, then the owner of a contiguous
property cannot sustain an action based on a breach of the duty of lateral support. To accept
this reasoning would mean that a landowner whose excavation or breach causes far-reaching
damage affecting a number of properties escapes liability whilst land owners, the
consequences of whose breach are much more modest, are saddled with strict liability’.
I cannot fault this reasoning.
[40] Counsel for the appellants made much of the averment in the respondent’s
particulars of claim that the slope mobilised through the mechanism of ‘a shallow slip
circle with uplift at the toe’ which had resulted in vertical upward bulging of the ground
surface at the bottom of the first appellant and Mr Venter’s properties. It was suggested
that there was evidence of such uplift and bulging. This, according to the first appellant,
was fatal to the respondent’s case because the pleaded mechanism fitted in with the
opinion of Dr Day that the slope mobilisation occurred when an uplift took place at the
toe of the excavation and the slip circle, thus excluding the removal of lateral support.
[41] There is no merit in this contention. In Gijzen v Verrinder 1965 (1) SA 806 (D)
at 810D-F, it was pointed out that, in most instances, the complaint of a plaintiff suing
for deprivation of lateral support arises from a subsidence that was caused by the
removal of such support. Nevertheless, such a subsidence (ie one caused by the
removal of lateral support) is not required for a successful plaintiff action. By way of
analogy, I conclude that is not required for a plaintiff in an action based on the removal
17
[42] The respondent’s averment as to the mechanism of the slope failure was thus
totally superfluous. Even in its absence, the thrust of his claim was clear: as a result
of the excavation on the first appellant’s property, lateral support owed to his property
was removed; the slope mobilised in the process of which extensive damage was
caused to his property. It is therefore patently opportunistic for the appellants to seek
to tie the respondent to a superfluous averment in his particulars of claim.
[43] In any event, the two mechanisms were fully explored during the trial and it
became clear that they overlapped; and that, in essence, as the court a quo correctly
observed, they were variations of the same mechanism. The position is analogous to
the converse situation, where an issue not pleaded is fully traversed during the trial.
As explained in Van Mentz v Provident Assurance Corporation of Africa Ltd 1961 (1)
SA 115 (A) at 122:
‘In a case where it is clear that the appellate tribunal has all the material before it on which to
form an opinion upon the real issue emerging during the course of the trial it will be proper to
treat the issues as enlarged (Collen v Rietfontein Engineering Works 1948 (1) SA 413 (AD)
at 433), where this can be done without prejudice to the party against whom the enlargement
is to be used (Robinson v Randfontein Estates, GM Co Ltd, 1925 AD 173 at 198).’
See also Marine & Trade Insurance Co Ltd v Van der Schyff 1972 (1) SA 26 (A) at
44H-45C).
[44] In the final analysis, the court a quo was faced with conflicting evidence of a
very technical nature. Where this is the case, the resolution of the dispute ‘must
depend on an analysis of the cogency of the underlying reasoning which led the
experts to their conflicting opinions’ (Buthelezi v Ndaba [2013] ZASCA 72; 2013 (5)
SA 437 (SCA) para 14). The court a quo preferred Dr McStay’s evidence to that of Dr
Day, and observed as follows (paras 127-128):9
‘The opinions which he [Dr McStay] expressed were rational and backed by consistent
reasons. What came through in his reports and evidence was a practical and common sense
approach which demonstrated his wide experience in the field…
As far as Dr Day is concerned there is no doubting his expertise as a geo-technical civil
engineer and his evidence was very helpful in understanding the geological aspects of what
took place on the site from March 2008 until the remedial measures were completed. Although
I do not doubt Dr Day’s sincerity or his professional integrity, I gained the distinct impression
that he became overly wedded to his client’s case, including the notion that the geological
event was not a failure of lateral support. Dr Day’s unwillingness to accept that the Dias
dwelling was in excellent condition prior to 2008, based on speculative or weak evidence
indicating the contrary, suggested that he fell into the trap of approaching some of the issues
in the matter in a less than balanced manner’.
[45] Having carefully considered the totality of the evidence of the two experts, the
court a quo cannot be faulted for preferring that of Dr McStay. Of the two experts, it is
Dr McStay’s evidence which provided the most reasoned and cogent explanation for
what had happened. His evidence closely matches the objective facts. It follows that
the respondent succeeded in establishing that the slope mobilisation had resulted from
a breach of the duty to provide lateral support due to the excavation on the first
appellant’s property. Given the objective facts in this case, it would indeed defy all
logic for a court to hold that the excavations by the first appellant did not destabilise
the respondent’s property and thus breached the duty to provide lateral support to it.
Causation
[46] I turn now to causation. As explained in Minister of Police v Skosana 1977 (1)
SA 31 (A) at 34E-35D, there are two distinct questions in the causation enquiry. The
first is a factual one and relates to the question whether the relevant conduct caused
or materially contributed to the harm giving rise to the claim. If it did not, then no legal
liability can arise. If it did, then the second question becomes relevant, namely whether
the conduct is linked to the harm sufficiently closely or directly for legal liability to
ensue, or stated differently, whether the harm is too remote from the conduct.
[47] The causa sine qua non (the ‘but for’ test) is ordinarily applied to determine
factual causation. The central theme of the first appellant’s case was that the slope
19
[48] Applying the above test to the facts of this case, it must be asked whether, but
for the excavation, the slope would have mobilised. In this regard, the excavation was
extensive, involving the removal of 5413m³ of earth, 57 blasting shots as well as the
removal of many large boulders. The lift shaft excavation was 13m in length, 5.5m in
width and 9.5m deep. It was excavated up to about 6m from the respondent’s property
and done without any bracing or support. It involved blasting at least one large boulder
and many others which needed to be broken and removed. In these circumstances, it
is hard not to accept Dr McStay’s opinion that there was a clear nexus between the
excavation and the slope failure.
[49] There must be a logical explanation as to why, after standing unaffected for 16
years, the respondent’s property mobilised shortly after the major excavation on the
first appellant’s property in 2008 and why the movement ceased when the remedial
measures were effected. During his testimony, Dr Day utilised a model to demonstrate
10Minister of Finance and Others v Gore NO [2006] ZASCA 98; 2007 (1) SA 111 (SCA); [2007] 1 All
SA 309 (SCA) para 33 (Citations omitted.)
20
the slip circle failure. After a demonstration with reference to four blocks, the court a
quo pointed out that in terms of the model he used, a necessary condition of the slip
circle was the removal of an excavation block, to which proposition Dr Day agreed. He
explained the role of excavation as follows:
‘It was a contributing element. There is no doubt about it. It’s no coincidence that this failure
occurred when excavation was formed. So the formation of the excavation contributed to the
instability of the slope, that is correct. But it contributed to the instability of the slope as
opposed to a lateral support failure’.
[50] It is also common cause that the excavation on Mr Venter’s property stopped
in April 2008. In answer to a direct question during cross-examination as to what event,
thereafter, could have caused the distress on the entire hill slope, Dr Day was
constrained to concede that ‘the major event was the removal of ground which then
set the process of slope instability in motion ….’ After suggesting that the rainfall was
a contributing factor, he conceded that the excavation was ‘a necessary condition’ for
the failure. The following excerpt from the evidence of Dr Day’s cross-examination is
illustrative of the centrality of the excavation to the slope failure and eventually the
damage to the respondent’s property:
‘MR BEY: So Dr, it is not clear that but for the Naumann [first appellant] excavation the land
on the Dias [respondent] property behind the [Mr] Venter property would not have failed? ---
M’Lord, if the excavations had not been formed we wouldn’t be here today’.
I take that as a yes--- Yes’
[51] The appellants emphasised that the role of the other factors such as the innate
instability of the slope, the excavation on Mr Venter’s property, and the winter rainfalls,
should not be discounted. Of course they should not. But, as shown above, given the
nature and extent thereof, the excavation was central to the slope mobilisation. As
pointed out in Van Duivenboden para 25 the respondent was not required to establish
the causal link between the excavation and the damage to his property with certainty.
All that was expected from him was to establish that the excavation was probably the
cause of the damage to his property.
21
[52] In Regal, Ogilvie Thompson JA (at 116A-C) referred with approval to the
American Restatement of the Law of Torts, vol IV at 277, where, dealing with factual
causation, the learned authors say:
‘In some cases the physical condition is not, of itself, harmful, but becomes so upon the
intervention of some other force – the act of another person, or force of nature. In such cases
the liability of the person whose activity created the physical condition depends upon the
determination that his activity was a substantial factor in causing the harm, and that the
intervening force was not a superseding cause.’
[53] Applying these tests to the facts of the present case, the excavation on the first
appellant’s property must be regarded as a ‘substantial factor’ or a proximate cause
of the slope mobilisation. In the circumstances, it is safe to conclude that but for the
excavation on the first appellant’s property, the slip circle failure would most probably
not have occurred. I thus find a direct and probable chain of causation between the
excavation and the slope mobilisation which caused damage to the respondent’s
property. Factual causation was accordingly established.
[54] With regard to legal causation, the court a quo expressed doubt whether it was
necessary to enquire into legal causation, since liability was strict in the present case
and ‘the question of reasonable foreseeability does not arise’. With respect, the court
a quo overlooked the fact that there has to be a measure by which it is determined
whether the conduct that factually caused the harm suffered, is too remote from the
harm. The test provided by the law for this part of the enquiry is a flexible one, in which
reasonable foreseeability is but only one factor, among several. Other factors include
directness, the absence or presence of a novus actus interveniens, legal policy,
reasonableness, fairness and justice, as explained in S v Mokgethi and Others 1990
(1) SA 32 (A); [1990] 1 All SA 320 (A) at 40I-41D . It could well be that in a particular
case, such as the present, one or more or all of reasonable foreseeability, directness,
or the absence or presence of a novus actus interveniens, play a subsidiary role, or
no role at all. But it is difficult to imagine a case where legal policy, reasonableness,
fairness and justice would play no role at all.
22
[55] Viewed in this light, legal causation is necessary, irrespective of whether liability
is strict or not. As explained by the Constitutional Court in Mashongwa:11
‘No legal system permits liability without bounds. It is universally accepted that a way
must be found to impose limitations on the wrongdoer’s liability. The imputation of
liability to the wrongdoer depends on whether the harmful conduct is too remotely
connected to the harm caused or closely connected to it. When proximity has been
established, then liability ought to be imputed to the wrongdoer provided policy
considerations based on the norms and values of our Constitution and justice also
point to the reasonableness of imputing liability to the defendant.’
[56] In International Shipping Co (Pty) v Bentley (Pty) Ltd 1990 (1) SA 680 (A) at
700H-J Corbett CJ neatly summed up the position with regard to legal causation as
follows:
‘[D]emonstration that the wrongful act was a causa sine qua non of the loss does not
necessarily result in legal liability. The second enquiry then arises, viz whether the wrongful
act is linked sufficiently closely or directly to the loss for legal liability to ensue or whether, as
it said, the loss is too remote. This is basically a juridical problem in the solution of which
considerations of policy may play a part. This is sometimes called “legal causation”.’
[57] In determining the presence of legal causation, the question is whether, having
regard to the considerations alluded to, the harm is too remote from the conduct or
whether, it is fair, reasonable and just that the first appellant be burdened with liability.
In my view, the question should be answered against the first appellant.
No fault liability
[58] As stated already, none of the affected properties were in their natural state.
They had all been developed for the building of houses. It was submitted on behalf of
the appellants that for that reason, our law does not permit a claim under strict liability
for breach of the duty of lateral support. The respondent’s claim, it was submitted,
should have been brought as an Aquilian action, so that negligence and wrongfulness
11Mashongwa v Passenger Rail Agency of South Africa [2015] ZACC 36; 2016 (3) SA 528 (CC);
2016 (2) BCLR 204 (CC) para 68. (Citations omitted.)
23
on the part of the first appellant could be established. In their heads of argument,
counsel submitted:
‘The imposition of strict liability can only be justified in principle where prospect of direct harm
is so obvious that there can be no question of a lack of foreseeability and where there is clear
and obvious single cause. By contrast, where potential for harm, as in this instance where
mechanism of failure is more complicated, or obscure, and hence not readily foreseeable, or
involves more than one cause, including a contribution by the claimant, and the conduct may
be neither negligent nor unlawful, the entire blame for the earth movement should not be
visited on one neighbour by virtue of a rule of strict liability.’
[59] Broadly stated, every landowner has a right to the lateral support and where
subsidence or other destabilisation occurs, as a result of excavations on an adjacent
property, the owner of the adjacent property will be liable in an action for damages
irrespective of whether she was negligent or not. That is not to suggest that an
adjacent property owner is not entitled to excavate. His or her entitlement to do so, is
limited by the duty not to withdraw the lateral support which is afforded to the adjacent
property. The right is reciprocal. Neither culpa nor dolus is a requirement for liability
for damage caused by the withdrawal of lateral support. Of course, if an aggrieved
property owner can prove that he or she suffered pecuniary loss through dolus or
culpa, she can likewise sue in delict by virtue of the lex Aquilia.
[60] It is now settled that liability in subsidence cases is strict. In D&D Deliveries
(Pty) Ltd v Pinetown Borough 1991 (3) SA 250 (D) it was explained (at 253H-I) that:
‘In subsidence cases it is unnecessary to prove an unlawful act or negligence; the cause of
action is simply damage following upon deprivation of lateral support. The action lies only
against the owner of the adjoining property, and each successive subsidence gives rise to a
fresh cause of action’.
See also Gijzen at 811E.
[61] Prof JC van Der Walt12 offers the following justification for strict liability:
12 JC van Der Walt ‘Strict liability in the South African law of delict’ (1968) 1 CILSA at 63.
24
‘Liability based on risk is usually created - either by legislation or by the courts - in cases where
a particular activity normally entails an extra· ordinary increase in the risk of harm to the
community. Fleming states it thus: “Certain types of activity which involve extraordinary risks
to others, either in the seriousness of the harm threatened or, more often, in its high degree
of probability, are charged with responsibility for ensuing harm, even if the most diligent care
has been exercised to obviate its occurrence. In these situations, it is widely felt that he for
whose benefit the risk is created should bear the loss unavoidably entailed rather than the
random victim.”
…The most common defences at the disposal of a defendant in cases of strict liability are “act
of God” (vis maior) and fault on the part of the injured party.’
[62] There is sufficient safeguard in our law to meet the appellants’ concerns, in the
form of legal causation, which, inter alia, rests on policy considerations. The elastic
approach to legal causation adopted by this court in Mokgethi is ‘sensitive to public
policy considerations and aims to keep liability within the bounds of reasonableness,
fairness, and justice’. (See De Klerk v Minister of Police [2019] ZACC 32; 2020 (1)
SACR 1 (CC) para 19, referring to Mokgethi at 40I-41D).
[63] Also, a cause of action based on strict liability in cases such as this, serves to
ensure that those who suffer damage are not non-suited because of the absence of
fault or because of their inability to prove the presence of fault. As is evident in this
case, the respondent simply did not know exactly what was happening on the first
appellant’s property, other than that his property was damaged. Importantly, there is
no attack on the strict liability action as being contra bonos mores, or unconstitutional.
There are therefore no policy considerations for our law to, a priori, set itself against
an action based on strict liability for breach of lateral support, as cases always turn on
their own facts.
[64] In sum, the answer to counsel’s submission is, first, that culpa or dolus is not
required for liability because the right of support is a natural right of ownership.
Second, there are sufficient safeguards and flexibility in our law so as to ensure that
one is not unjustifiably punished at the expense of others. Third, liability without fault
here is usually restricted to damage to life, limb and property. On the facts the court a
quo correctly held that the first appellant is liable to the respondent. Although there is
25
no unanimity among scholars on a theoretical justification for strict liability, the authors
of Neethling-Potgieter-Visser Law of Delict13 observe:
‘[w]here a person’s activities create a considerable increase in the risk or danger of causing
damage, that is, an increased potential for harm, there is sufficient justification for holding him
liable for damage even in the absence of fault . . . Van der Walt, however, points out that the
question whether or not the potential of risk has been increased enough, will depend largely
on the legal convictions of the community, as reflected in legislation or case law. This theory
[the risk or danger theory] provides a satisfactory explanation for most of the instances of strict
liability which are recognised in our law.
Nonetheless, a satisfactory and universally accepted scientific basis for every instance of
liability without fault has not yet been found, and will probably never be found. A flexible
approach is therefore necessary so that each specific case may be valued on its own merits
and judged accordingly.’
[65] It remains to sum up the position of our law on the right of lateral support owed
between contiguous properties. First, it is a natural right incidental to the ownership of
the property and not servitudal in nature, as enunciated in Rouliot. Second, it is a
principle of neighbour law as explained in Anglo Operations, which rests on justice
and fairness, as articulated in Regal. Lastly, it is owed to land not only in its natural
state, but extends to buildings upon it. Although influential in the acceptance of the
right of lateral support into our law, English law was not slavishly implanted into our
law.
[66] Before I conclude, something needs to be said about the manner in which this
litigation has been conducted. The order of separation followed on an application by
the appellants, which was opposed by the respondent. The costs of that application
were reserved. The trial of the separated issues was lengthy, taking place over a total
of 27 days. In this court, the record spans 4248 pages, which includes the court a
quo’s judgment 129 pages. It is thus disquieting that despite this circuitous journey, in
terms of the separation order, the judgment of this court would not result in a final
determination of the dispute between the parties. It was with this in mind that it was
enquired of counsel during the hearing of the appeal whether the parties would be
prepared to accept the order of this court as a final word on the liability dispute between
the parties. Counsel for the parties accepted that this judgment will finally dispose of
all the disputes between the parties, as far as liability is concerned.
[67] It is regrettable that this court has, once again, to express disquiet on how rule
33(4) is often not properly considered.14 As it was stated in Denel (Edms) Bpk v Vorster
2004 (4) SA 481 (SCA) para 3:
‘Rule 33(4) of the Uniform Rules ─ which entitles a Court to try issues separately in appropriate
circumstances ─ is aimed at facilitating the convenient and expeditious disposal of litigation.
It should not be assumed that that result is always achieved by separating the issues. In many
cases, once properly considered, the issues will be found to be inextricably linked, even
though, at first sight, they might appear to be discrete. And even where the issues are discrete,
the expeditious disposal of the litigation is often best served by ventilating all the issues at one
hearing, particularly where there is more than one issue that might be readily dispositive of
the matter. It is only after careful thought has been given to the anticipated course of the
litigation as a whole that it will be possible properly to determine whether it is convenient to try
an issue separately. But, where the trial Court is satisfied that it is proper to make such an
order ─ and, in all cases, it must be so satisfied before it does so ─ it is the duty of that Court
to ensure that the issues to be tried are clearly circumscribed in its order so as to avoid
confusion.’
See also ABSA Bank Ltd v Bernert [2010] ZASCA 36; 2011 (3) SA 74 (SCA) para 21.
[68] In Consolidated News Agencies (Pty) Ltd (in liquidation) v Mobile Telephone
Networks and Another [2009] ZASCA 130; 2010 (3) SA 382 (SCA), this court
cautioned against piece-meal litigation:
‘Piece-meal litigation is not to be encouraged. Sometimes it is desirable to have a single issue
decided separately either by way of a stated case or otherwise. If a decision on a discrete
issue disposes of a major part of a case, or will in some way lead to expedition it might well
be desirable to have that issue decided first.
This court has warned that in many cases, once properly considered, issues initially thought
to be discrete are found to be inextricably linked. And even where the issues are discrete, the
expeditious disposal of the litigation is often best served by ventilating all the issues at one
14 See, for example, Firstrand Bank Ltd v Clear Creek Trading 12 (Pty) Ltd and Another [2015] ZASCA
6; 2018 (5) SA 300 (SCA) paras 9-10; Feedpro Animal Nutrition (Pty) Ltd v Nienaber NO and Another
[2016] ZASCA 32 para 15; Cilliers NO and Others v Ellis and Another [2017] ZASCA 13 paras 12-14;
and Transalloys (Pty) Ltd v Mineral-Loy (Pty) Ltd [2017] ZASCA 95 para 6.
27
hearing. A trial court must be satisfied that it is convenient and proper to try an issue
separately.’15
[69] It is by no means clear that these principles informed the decision to separate
issues in this matter. In my view, the issues raised in the separated order are
inextricably linked to the rest of the issues in the pleadings. They could conveniently
have been ventilated in one hearing. This should have been clear to the parties and
the judge who granted the separation order.
[70] In all the circumstances the appeal has to fail. The following order is made:
The appeal is dismissed with costs, such costs to be paid by the appellants jointly and
severally, the one paying the other to be absolved.
____________________
T M Makgoka
Judge of Appeal
15Consolidated News Agencies (Pty) Ltd (in liquidation) v Mobile Telephone Networks and Another
[2009] ZASCA 130; 2010 (3) SA 382 (SCA) paras 89-90. (Citations omitted.)
28
APPEARANCES: