Bergstedt v Rhode (Appeal) (A245/2005) [2026] ZAWCHC 165 (13 April 2026)

60 Reportability

Brief Summary

Delict — Dog bite — Legal causation — Appeal concerning whether dog bite caused fracture to respondent's ankle — Respondent's case based on dog bite causing injury, but during cross-examination conceded that misstep caused injury — Trial court found dog bite initiated chain of events leading to injury — Appeal upheld, finding no causal link between dog bite and ankle fracture, as intervening act of respondent's negligence broke chain of causation.

THE REPUBLIC OF SOUTH AFRICA

IN THE HIGH COURT OF SOUTH AFRICA
WESTERN CAPE DIVISION, CAPE TOWN
Case No.: A245 / 2005
In the matter between:
SHAUN ERROL BERGSTEDT Appellant
and
WILLIAM HENRY RHODE Respondent
Summary: Delict - Dog Bite - Novus Actus Interveniens - Appeal Upheld.
Coram: Wille, J et Gxashe, AJ
Heard: 13 March 2026
Further submissions: 20 March 2026
Delivered: 13 April 2026

JUDGMENT

THE COURT:
INTRODUCTION
[1] This is an appeal from the lower court about a dog bite. That the appellant’s dog bit
the respondent is conceded. The issue for determination on appeal is simply whether the
dog bite caused a fracture to the respondent`s ankle.1
[2] Thus, this appeal concerns legal causation. On appeal, the appellant seeks a
substitution order to the effect that the dog bite did not cause the fracture to the respondent’s
ankle.2
[3] It was the respondent’s case (as defined in the pleadings) that the dog bite caused
the fracture to his ankle. This is significant because it was not (and is not) the respondent’s
case that his ankle fracture was caused by the circumstances that eventuated because of
the dog escaping from the appellant`s yard.3
THE GROUNDS OF APPEAL
[4] The grounds of appeal are three. In summary, they are:
(a) Did the court a quo venture beyond the scope of the respondent`s pleaded case on a
cause of action not pleaded by him and only introduced during the closing argument?
(b) Did the evidence adduced at the hearing support the conclusion that no novus actus
interveniens was established?
(c) Did the respondent advance any medical evidence in support of the causal
connection between the dog bite and the fracture to his ankle?4

1 This is the narrow ambit of the appeal.
2 The issue is one pf legal causation.
3 The respondent’s pleaded case is that the dog bite caused his ankle injury.
4 No such medical evidence was tendered.

THE COMMON CAUSE FACTS
[5] The common cause facts are that : (a) the dog belonged to the appellant; (b) the dog
escaped from the appellant’s premises; (c) the appellant’s dog bit the respondent; (d) the
respondent suffered two bite marks on his ankle; (e) the appellant did not see h is dog bite
the respondent but accepted that it did so bite the respondent, and (f) the respondent
pleaded that the dog bite caused his ankle to be fractured.5
THE ISSUES ON APPEAL
[6] The core issues on appeal are these:
(a) Did the court of first instance impermissibly allow the respondent to establish a
different case (than the pleaded case) at the trial, which new case then formed the
basis of its judgment?
(b) Was there a novus actus interveniens in the form of the respondent`s own
negligence, which broke the chain of causation between the dog’s bite and the
respondent`s subsequent alleged ankle injury?
(c) Did the respondent prove that his ankle injury occurred at all?6
THE PLEADED CASE
[7] The respondent pleaded that it was the dog bite that caused the fracture to his ankle.
This was aligned with the evidence tendered by the respondent during the trial. Most
importantly, this was the basis on which the appellant was cross -examined and the case
which the appellant had to meet.7

5 This is the respondent’s pleaded case.
6 The appellant advances that this did not occur.
7 The appellant says it was the respondent’s case the alleged fracture was caused by the dog bite.

[8] During cross-examination, the respondent was driven to concede. that it was not the
dog bite that caused the fracture to his ankle. It was the misstep that he took from the
sidewalk that caused the injury to his ankle. The court of first instance found that it was the
respondent’s fall that caused the fracture to his ankle, not the dog bite.8
[9] The judicial officer in the lower court found in favour of the respondent (on this issue)
on the basis that the dog's attack initiated a chain of events culminating in the respondent’s
fall and ankle fracture.9
[10] In addition, the finding was made that even if there was a further misstep by the
respondent, the initiating event was caused by the attack on the respondent by the
appellant’s dog. This, however, was never the respondent’s pleaded case.10
[11] It is trite that it is impermissible in our adversarial system for a p arty to plead a
particular case and seek to establish a different case at the trial.11
[12] The issue of judicial restraint was eloquently and definitively formulated by Wallis JA,
in the following way:
‘…Turning then to the nature of civil litigation in our adversarial system , it is for the parties, either in
the pleadings or affidavits, which serve the function of both pleadings and evidence, to set out and
define the nature of their dispute, and it is for the court to adjudicate upon those issues…’
and
‘…If they wish to stand by the issues they have formulated, the court may not raise new ones or
compel them to deal with matters other than those they have formulated in the pleadings or
affidavits…’12

8 The appellant avers that this was not the respondent’s pleaded case.
9 The appellant says this finding was wrong and impermissible.
10 A new cause of action was advanced during argument.
11 Jowell v Bramwell-Jones and Others 1998 (1) SA 836 (W) at 902 - H.
12 Fischer v Ramahlele and Others 2014 (4) SA 614 (SCA) at para [14].

[13] It is self -evident that the new case advanced by the respondent during the closing
argument was not canvassed during the trial in the court of first instance, yet the court a quo
pronounced on this non -pleaded issue, which formed the basis of the order against the
appellant.13
A NOVUS ACTUS INTERVENIENS
[14] No credibility findings were touched upon in the judgment by the trial court. Th is is
important because the appellant testified that his dog was once again secured within his
property when the respondent injured his ankle.14
[15] The trial court found that the respondent’s ankle was fractured because of the dog
bite. The factual findings were t hat the dog ’s bite caused two bite marks /punctures on the
respondent’s ankle.15
[16] The respondent’s particulars of claim confirmed t hat the respondent’s pleaded case
was that the actual physical bite (by the dog) caused the fracture, not the circumstances that
eventuated because of the dog escaping from the appellant’s property.16
[17] This was the case that the appellant was called upon to meet , and the respondent
had to stand or fall in accordance with this pleaded case. What was striking from the record
is that no medical evidence was tendered on behalf of the respondent.17
[18] The ‘but for ’ test was applied a quo , and the reasoning was that but for the dog
escaping from the appellant`s property, the incident would not have occurred , which was
ultimately the test upon which the finding against the appellant was made.18

13 This is then the misdirection contended for by the appellant.
14 This evidence was accepted by the judicial officer in the court a quo.
15 This is what the appellant also pleaded in his plea.
16 The respondent’s case shifted during the legal argument advanced by the respondent.
17 The matter was postponed for several months for this specific purpose.
18 The appellant says that this was not the respondent’s pleaded case.

[19] The ‘but for’ scenario was never the case pleaded by the respondent. The appellant
was never alerted to the ‘but for ’ case chartered on behalf of the respondent. Had the
appellant been alerted to this new case by the respondent, the appellant might have
prepared differently. The appellant may have submitted expert medical evidence.19
[20] Regrettably, the trial court also made a finding that the respondent’s ankle was never
injured prior to the dog bite and that the respondent’s ankle was fractured because of the
dog bite incident. These aspects were evaluated as common cause aspects.20
[21] No expert medical evidence was tendered by the respondent. No medical evidence
was tendered in support of the fractured ankle. Thus, it was incorrect to make a finding that
(in the absence of expert medical evidence) the respondent indeed fractured his ankle.21
[22] In the analysis by the trial court of whether a novus actus interveniens was proved, it
must have (by way of implication) rejected the appellant’s evidence that the appellant’s dog
was safely secured in the appellant’s premises when the respondent allegedly dislocated his
ankle.22
[23] The respondent (at one stage) averred that the appellant’s dog bit him on his ankle
and then threw him the length of ‘ twenty-one centimetres’ onto the sidewalk. No doubt this
is highly improbable. The respondent also advanced the view that his misstep caused the
alleged ankle injury, not the dog bite.23
[24] The initial affidavit deposed to (shortly after the incident) by the respondent also
differed markedly from the versions given in his testimony in the trial court.24

19 The appellant says he was subjected to “ambush” litigation by the respondent.
20 The record reflects that these were not common cause issues.
21 This was never demonstrated save of the evidence by the respondent.
22 This it could not have done without making a credibility finding against the appellant.

22 This it could not have done without making a credibility finding against the appellant.
23 The respondent’s version of events also shifted from time to time.
24 Dated 21 April 2020.

[25] The respondent was unable to explain why there were no other injuries to his left leg
depicted on the doctor`s report, other than the two dog bite puncture wounds.25
[26] Most importantly, the respondent made no mention of his alleged fractured ankle in
the affidavit that he deposed to shortly after the dog bite. This failure by the respondent to
refer to his broken ankle (a mere eight days after the incident) is a factor that weighs heavily
in favour of the appellant’s version that the dog bite did not cause the fracture to the
respondent’s ankle.26
[27] The first time the alleged fractured ankle was formally raised by the respondent was
about 3 years after the incident, when the respondent brought a claim for damages against
the appellant in a not insignificant amount.27
[28] The appellant testified that: (a) the respondent was well known to him; (b) he had just
removed his vehicle from the garage as his wife wanted to clean the area; (c) he noticed the
respondent in the road ; (d) he forewarned the respondent that there was a roadblock down
the road and suggested that he turn around to avoid a fine in contravention of the lockdown
regulations; (e) he heard a scream from the respondent; (f) he noticed his dog in front of the
respondent; (g) the r espondent was standing with his hands in the ai r, holding his cap and
his mobile phone; (h) he chased his dog away towards his house, and (g) he did not see his
dog bite the respondent.28
[29] Once the appellant’s dog was no longer in the vicinity , the appellant made his way
over to the respondent with a view to enquiring whether he had been bitten by his dog. This
is when the respondent stepped awkwardly onto the edge of the drain behind him and fell.29

25 This was left unexplained.
26 This is the most probable version.
27 In the sum of about R350 000,00.
28 This was not materially challenged.
29 This is the more probable version.

[30] Factually, the a ppellant`s dog was no longer in the vi cinity of the respondent when
the misstep occurred, and thus the misstep was a novus actus interveniens which broke the
legal chain of causation . Put another way, to say that the a ppellant`s dog caused the
misstep was too remote.30
THE FRACTURED ANKLE
[31] There was no medical or expert evidence tendered to demonstrate (on a balance of
probabilities) that the dog bite fractured the respondent’s ankle. This factual position was
supported in part by the legal conclusion drawn by the judicial officer in the trial court, to the
effect that the alleged fracture was caused by the respondent’s fall and not by the dog bite.31
CONCLUSION
[32] The negligent act complained of is the appellant’s failure to ensure that his dog did
not escape the yard, and, because of that act, the dog ultimately bit the respondent. But, in
this case, there was an intervening external force that came in the form of the appellant, who
chased his dog back into the yard. Thus, the respondent`s action in ensuring that his do g
was no longer a threat to the respondent constituted an intervening force , and the act of
falling thereafter was due to the respondent`s own negligence in failing to keep a proper
lookout.32
[33] What was not explained was why the court rejected the appellant`s version. We say
this because the respondent approached the matter on the basis that the dog`s bite
allegedly fractured his ankle, only to concede during cross -examination that it was not th e
case. There were material and unexplained discrepancies between the respondent's
version, given eight days after the incident, and his oral testimony at trial.33

30 The facts support this legal conclusion.
31 This factual determination was correct, but the legal conclusion was incorrect.
32 This is how the respondent fractured his ankle.
33 Thus, the respondent`s version is inherently more improbable.

[34] For all these reasons, the appeal falls to be upheld. The trial court’s finding of liability
in relation to the alleged fracture of the respondent`s ankle falls to be amended by replacing
and substituting it with the finding that the dog bite only caused two bite marks to the
appellant`s ankle.34
COSTS
[35] The costs occasioned (and those that may be occasioned) by issues relating to
quantum are, with this, ordered to stand over for determination, together with the ultimate
determination on quantum to be made by the trial court. We say this because the trial court
will be best placed to decide on the issue of the ultimate successful party to this litigation.
Regarding the costs of the appeal, there is no reason they should not follow the outcome, as
the appellant has been the successful party. Thus, the respondent should be held lia ble for
the costs of the appeal.35
ORDER
[36] The following order is granted:
1. The appeal is upheld.
2. The order by the trial court to the effect that the defendant (the appellant) is to be held
liable for the damages arising from the plaintiff’s (the respondent’s) fractured ‘distal
fibula’ is set aside and replaced with the following orders:
(a) The dog bite by the appellant’s dog only caused two bite marks/punctures to
the respondent’s ankle.
(b) The costs of and incidental to the trial shall be determined by the trial court
unless otherwise agreed.

34 The quantum is to be decided with reference to these two bite marks.
35 This appeal court is best suited to make this determination.

2. The respondent (plaintiff) shall be liable for the costs of this appeal on the scale as
between party and party (as taxed or agreed ), including the costs of counsel on Scale
B.
3. The matter is remitted to the trial court to determine all r emaining issues relating to
quantum.

_______
WILLE J

I agree:
___________
GXASHE, AJ