Lever v Purdy (472/1991) [1993] ZASCA 48 (31 March 1993)


Case No 472/91

NORMAN ALLEN LEVER Appellant

ARTHUR HARRY PURDY Respondent

IN THE SUPREME COURT OF SOUTH AFRICA APPELLATE DIVISION

In the matter between:

NORMAN ALLEN LEVER Appellant

and

ARTHUR HARRY PURDY Respondent

Coram: JOUBERT A C J et SMALBERGER KUMLEBEN NIENABER J J A et HOWIE A J A

Heard: 22 February 1993 Delivered: 31 March 1993

J U D G M E N T JOUBERT, A C J:

2

This is an appeal against a judgment of ELOFF J P (SPOELSTRA and MYNHARDT J J concurring) in the Transvaal Provincial Division, allowing an appeal against a judgment granted by ESSELEN J in an action instituted by the present respondent ("Purdy") against the appellant ("Lever"). The appeal comes to us with special leave in terms of sec 20(4)(a) of Act No 59 of 1959. The judgment of the Court a quo has been reported : see Purdy v Lever, 1991 (3) SA 431 (T).

The factual background to this appeal may be broadly outlined as follows:

1. Lever was at all relevant times the owner of a vicious dog which he kept at his residence in Johannesburg. While he was temporarily overseas during September 1986 a certain Cohen was in occupation and control of his residence and was also taking care of the dog. The nature of the legal relationship between Lever and Cohen

3

is not apparent from the pleadings or the evidence. Suffice it to say that the question of any vicarious liability on the part of Lever does not arise in the present matter.

  1. On 3 September 1986 Purdy was bitten by Lever's dog at his residence in circumstances to which reference will hereinafter be made in greater detail.

  2. In June 1987 Purdy instituted an action in the Witwatersrand Local Division against Lever as first defendant and Cohen as second defendant claiming damages from them jointly and severally for the personal injuries sustained by him from the bite of Lever's dog. The claim against Lever was based on the actio de pauperie in that he was the owner of the dog which acted contrary to the nature of its class and from inward vice when it injured Purdy. Several defences were raised by Lever in his plea but the defence relevant for

4

purposes of this appeal was that Purdy was bitten by the dog "as a result of the second defendant's negligence". The claim against Cohen was founded on the actio leqis Aquiliae for his negligence in failing "to lock up the dog in the backyard of the residence, or to otherwise safely secure the dog, and in permitting the dog to be in the vicinity of the gates at the entrance to the said residence, unattended at about 6.30 p m on the day in question" i.e. upon Purdy's pre-arranged arrival at the residence.

  1. The first trial came before MORRIS A J in the Witwatersrand Local Division in March 1989. Evidence was led and after judgment was reserved, but before delivery thereof, the Judge passed away.

  2. Subsequently in terms of a settlement with Cohen, Purdy decided not to proceed with his claim against Cohen who was exonerated from the obligation of making any payment to him.

5

6. The second trial commenced de novo before ESSELEN J on 5 March 1990. At that stage by agreement between the parties the quantum of the damages had been fixed at R21 382-71. Purdy was the only witness who testified. ESSELEN J found that Purdy established the essentials of the actio de pauperie against Lever, but he held that Lever "proved the defence which in casu is available to him, namely culpable conduct on behalf of Cohen, the second defendant". Such defence, it was held absolved Lever completely from pauperien liability to Purdy. He accordingly granted an order of absolution from the instance with costs which included the costs of the previous hearing before MORRIS A J.

  1. On 12 April 1990 ESSELEN J dismissed the application of Purdy for leave to appeal.

  2. On 13 August 1990 Purdy was in terms of sec 20(4)(b), read with (2)(a) and (3)(c), of Act No 59 of 1959

6

granted leave to appeal to the Full Bench of either the Transvaal Provincial Division or the Witwatersrand Local Division as directed by the Judge President. 9. The Full Bench of the Transvaal Provincial Division, the Court a quo, allowed Purdy's appeal by setting aside the judgment of ESSELEN J and by substituting for it judgment in favour of Purdy against Lever in the agreed amount of R21 382-71 with costs as set out in its reported judgment (p 438 F-G).

I now turn to consider the relevant circumstances regarding the dog's attack on Purdy. According to the latter's uncontroverted evidence he was at the relevant time in business in connection with the sales and repairs of television sets and video recorders. He knew Cohen with whom he had had previous business dealings. At approximately lunch time on 3 September 1986 they were in telephonic communication with each other. Cohen requested him to visit him at Lever's house in order to adjust his video

7

recorder and television set. He informed Purdy that he was looking after Lever's house during his absence overseas. He also mentioned that there was a vicious dog on the premises. Purdy requested him to lock it away before his arrival at the house. Their arrangement was that Purdy would be calling on him between 6.30 p m and 7 p m. No instructions were given to Purdy as to how he should proceed onto the premises. When he arrived as agreed it was dusk. After parking his motor car near the front wrought iron gate which was partly open he hooted to announce his arrival. He alighted from his vehicle and proceeded to walk through the wrought iron gate along a fenced-in passage towards a closed wire mesh gate consisting of two sections where he stopped on its outside. He leant over one of the sections and placed his right hand on it whereupon he shouted for Cohen. Until then there was no sign of any dogs on the premises. Nor were any dogs barking at him. Suddenly a dog appeared on the opposite side of the wire mesh gate and

8

seized his right hand in its teeth. A tug of war ensued between them. The dog succeeded in pulling him through the two sections of the wire mesh gate which had opened. Purdy managed to hit the dog on its nose with his left hand whereupon it let go of his hand but bit him on his left thigh. Purdy closed one section of the wire mesh gate on the dog's nose. That action resulted in the dog's letting go of his left thigh. Purdy succeeded in moving to the outside of the two sections of the wire mesh gate while the dog continued to face him across the gate. At that stage Cohen arrived at the scene and invited Purdy into the house. The dog behaved perfectly in a natural manner in the presence of Cohen without causing any further trouble to Purdy. It was common cause in this Court that ESSELEN J correctly found that Purdy established the essentials of the actio de pauperie against Lever. The latter, however, challenges in this appeal the correctness of the finding by the Court a quo that Lever could only escape pauperien liability if the

9

negligence of Cohen could be said to have excited or provoked the animal, or to have caused it to bite Purdy (p 438 D-E).

Justinian's Roman law relating to pauperien liability of the owner of a domesticated animal, which by acting contrary to the nature of its class (contra naturam sui generis) and from inward vice (fera mota) caused damage, as subsequently developed in the Middle Ages until the second half of the 15th century, was received in the Netherlands. See De Blecourt-Fischer, Kort Begrip van het Oud-Vaderlands Burgerlijk Recht, 7th ed p 19. As part of Roman-Dutch law the law of pauperien liability regarding the damage done by domesticated animals was introduced into South Africa. According to the law of South Africa two important modifications were effected. The first related to the principle of noxal surrender (noxae deditio) which enabled an owner to avoid liability by surrendering the animal to the injured party while the second concerned that of noxa caput

10

sequitur according to which the owner at the time of litis contestatio was liable, not necessarily the owner at the time of the injury. These two principles were held by this Court to be obsolete. See O'Callaqhan N O v Chaplin, 1927 A D 310 and South African Railways and Harbours v Edwards, 1930 A D 3. Reliance on the negligence of a third party as a defence to the actio de pauperie is res nova as far as our case law is concerned. It now remains to consider our common law on this particular issue.

Justinian's law of pauperien liability is treated rather cursorily in Inst 4.9 pr but in somewhat greater detail in D 9.1. The latter source deals inter alia with those instances in which the culpable conduct of a third party causes a domesticated animal to act contrary to the nature of its class in injuring the injured victim. In such instances the owner of the animal was exonerated from pauperien liability to the victim. The latter could, however, claim damages from the third party under the lex

11

Aquilia. The texts in question may be analysed and

classified in two categories, namely:

First Category

This category comprises those instances in which a

third party, as a mere outsider through his culpable conduct

caused the animal to inflict the injury upon the victim, e.g.

where the animal was provoked by him (D 9.1.1.6); or where he hit or wounded the animal (D 9.1.1.7).

This first category of texts may be supplemented by texts dealing with Aquilian liability of such a third party, e.g. where the third party scared a horse which a slave was riding with the result that the slave was thrown into a river and died (D 9.2.9.3); and where someone annoyed a dog and accordingly caused it to bite the victim (D 9.2.11.5).

The distinguishing feature of this category is that the culpable conduct of the third party consisted of some positive act such as provoking, striking, wounding, scaring or annoying the animal.

12

Second Category

This category relates to those instances in

which a third party in charge or control of the animal by his

negligent conduct failed to prevent the animal from injuring

the victim. The relevant texts are the following:

D 9.1.1.4 (Ulpianus) : - - - quod si propter loci iniquitatem

aut propter culpam mulionis, aut si plus iusto onerata quadrupes in aliquem onus everterit, haec actio cessabit damnique iniuriae agetur. Translation by Watson et alii:

"On the other hand, if an animal should upset its load onto someone because of the roughness of the ground or a mule driver's negligence or because it was overloaded, this action will not lie and proceedings should be brought for wrongful damage." I underlined the word injuriae because the medieval

Glossator AZO (± 1150-1230 A D) wrote the following gloss on

this text to elucidate the word in its context, viz. that

an actio in factum or actio leqis Aquiliae utilis lies

against the muleteer or the person whose negligence caused

the damage, since he who provided the opportunity for the

13

damage to be done is deemed to have caused it. The gloss

reads as follows:

Iniuriae. id est in factum, vel utili Aquilia contra mulionem, vel eum, cuius culpa damnum est datum, qui enim occasionem damni dans, damnum dedisse videtur. (I may add in parenthesis that the lex Aquilia applied where

the damage was directly caused by the body of the wrongdoer

to the body of the injured person or to the damaged thing

(damnum corpore corpori datum). Where the injury was,

however, indirectly caused by the wrongdoer to the body of

the injured person or thing (damnum corpori non corpore

datum) the Praetor granted an actio utilis or in factum in

order to extend Aquilian liability to such instances (ad

exemplum legis Aguiliae, D9.2.53)). See Buckland,

A Text-Book of Roman Law from Augustus to Justinian, 3rd ed.

p 589.

D 9.1.1.5 (Ulpianus) : Sed et si canis, cum duceretur ab

aliquo, asperitate sua evaserit et alicui damnum dederit : si contineri firmius ab alio poterit vel si per eum locum induci non

14

debuit, haec actio cessabit et tenebitur qui canem tenebat. Translation by Watson et alii:

"Take the case of a dog which, while being

taken out on a lead by someone, breaks loose

on account of its wildness and does some harm

to someone else : If it could have been better

restrained by someone else or if it should

never have been taken to that particular

place, this action will not lie and the

person who had the dog on the lead will be

liable."

It is the actio de pauperis which will not lie against the

owner, while the person who was in control of the dog and

whose negligence caused the injury will be liable under the

actio leqis Aquiliae utilis.

The second category of texts clearly establishes

the principle of law that the owner of a domesticated animal,

which contra naturam sui generis harmed a victim, may

15

successfully avoid pauperien liability by proving as a

defence that the harm was caused by the controller's

negligence in his control of the animal.

These texts must, however, be read in conjunction

with those texts which shed light on the Aquilian liability

of a controller of a domesticated animal. The texts in

question are the following:

D 9.2.8.1 (Gaius): Mulionem guoque, si per imperitiam

impetum mularum retinere non potuerit, si eae alienum hominem obtriverint, vulgo dicitur culpae nomine teneri, idem dicitur et si propterinfirmitatem sustinere mularum impetum non potuerit : nee videtur iniguum, si infirmitas culpae adnumeretur, cum affectare quisque non debeat, in quo vel intellegit vel intellegere debet infirmitatem suam alii periculosam futuram. idem iuris est in persona eius, qui impetum equi, quo vehebatur, propter imperitiam vel infirmitatem retinere non potuerit. Translation by Watson et alii: "Furthermore, if a mule driver cannot control his mules because he is inexperienced and as a result they run down somebody's slave, he is

16

generally said to be liable on grounds of negligence. It is the same if it is because of weakness that he cannot hold back his mules - and it does not seem unreasonable that weakness should be deemed negligence; for no one should undertake a task in which he knows or ought to know that his weakness may be a danger to others. The legal position is just the same for a person who through inexperience or weakness cannot control a horse he is riding." D 9.2.11.5 (Ulpianus) : Item cum eo, qui canem irritaverat

et effecerat, ut aliquem morderet, quamvis eum non tenuit, Proculus respondit Aquiliae actionem esse : sed lulianus eum demum Aquilia teneri ait, qui tenuit et effecit ut aliquem morderet : ceterum si non tenuit, in factum agendum.

Translation by Watson et alii : "Again, Proculus gave an opinion that the Aquilian action lies against him who, though he was not in charge of the dog, annoyed it and thus caused it to bite someone; but Julian says the lex Aquilia only applies to this extent that it applies to him who had the dog on a lead and caused it to bite someone; otherwise, if he were not holding it, an actio in factum must be brought."

17

Inst 4.3.8 : Impetu quoque mularum, quas mulio propter

imperitiam retinere non potuerit, si servus tuus oppressus fuerit, culpae reus est mulio. sed et si propter infirmitatem retinere eas non potuerit, cum alius firmior retinere potuisset, aeque culpae tenetur, eadem placuerunt de eo quoque, qui, cum equo veheretur, impetum eius aut propter infirmitatem aut propter imperitiam suam retinere non potuerit. Moyle's translation:

"- - and similarly, if your slave is run over by a team of mules, which the driver has not enough skill to hold, the latter is suable for carelessness; and the case is the same if he was simply not strong enough to hold them, provided they could have been held by a stronger man." I indicated supra in discussing the first category

of texts that the culpable conduct of the third party

consisted of some positive act on his part, such as

provoking, striking, wounding, scaring or annoying the

animal causing it to act contra naturam sui generis and to

injure the victim.

18

What is the causative position concerning the negligent conduct of the third party in the second category of texts ? Here, the third party happens to be in charge or control of the animal. Take the instance of a muleteer who is in control of his team of mules which he, on account of his inexperience or weakness, cannot restrain from running away (impetus) and injuring the victim. His failure in exercising proper, i.e. reasonable, control over the mules provided them with the opportunity to continue their flight and run over the victim. He is guilty of negligent conduct which resulted in the injury to the victim. He will incur Aquilian liability whereas the owner of the mules will be exonerated from pauperien liability. The muleteer did not by any positive act cause the mules to run away. The question of causality in regard to the conduct of the controller or handler of a dog is determined in the same manner by application of the same legal principles. By his negligent conduct he fails to exercise proper, i.e.

19

reasonable, control over the dog in his care. He accordingly provides the dog with the opportunity to injure the victim. As a result of his negligent conduct he fails to prevent the dog from biting the victim. He did not by any positive act cause the dog to bite. His negligent conduct likewise renders him liable under the Lex Aquilia, whereas the owner of the dog will be exonerated from pauperien liability.

I now turn to consider the Roman-Dutch law. Unfortunately there is a dearth of Roman-Dutch authority on the nature of the culpable conduct of a third party in control of the owner's domesticated animal which injured the victim. The Dutch jurists, without any significant discussion or original contribution of their own, adopted the principles of Roman law as discussed supra.

Damhouder (1507-1581) in his Practycke in Criminele Saken, 1650, 142 Capital nr 3 concisely states the following:

Van ghelijcken, indien dat yemandt bevolen

20

waer te bewaren eenen Hondt, oft ander Beeste, ende dat hy sulcken Hondt, oft ander Beeste, van selfs ontbonde, ende uyt quaetheydt liet loopen, oft dat sy van selfs ontliepen, door de qroote neqligentie, ende roeckeloosheydt van den bewaerder, ende indien dat sulcken Hondt of Beeste, alsoo yemandt quetste, in dit cas, soude den bewaerder oft Knape in de schuld vallen, ende te punieren zijn, ende niet den Meester. (My underlining). Paul Voet (1619-1667), the father of the famous

Johannes Voet, ad Inst 4.9.5 affirms the approach by Roman

law to the Aquilian liability of a third party as controller

or handler of a dog. The conduct of the controller amounts

to negligence because he failed to restrain the dog from

biting, or led it to a place where he should not have taken

it. His passage reads as follows:

An non ergo si canis pauperiem dederit, actioni directae locus erit? Id videtur velle Zoes. D hoc tit. verum contra expressum juris textum. Nam in D 9.1.1.5 statuitur hanc actionem, scil, de pauperie directam cessare; & nihilominus eum qui canem tenebat

21

obligari. An Pesulonia vel Pesulania, vel ut alii consent Solonia lege ? Cujac. ad Paul 1 sent. tit. 15. An potius lex Aquilia ? Quae postrema sententia probabilior, cum damnum ejus culpa acciderit, qui canem vel non retinuit satis, vel per eum locum duxit, per quern duci non oportuit. (My underlining). Van Leeuwen (1626-1682) in his C.F. 1.5.31.3 (as translated by Margaret Hewett, 1991) makes the following concise observations regarding pauperien liability, which are based on the principles of Roman law as expounded in the two categories of texts (supra) :

"I said aut culpa hominis (or negligence on the part of a human being) because if there is negligence on the part of the owner or of anyone else, this action lapses and a suit is brought under the Lex Aquilia, for example if a mule does damage because of the uneven-ness of the road, or the negligence of the muleteer, or because it was too heavily loaded or was provoked by someone, or if the animal acted in some way on account of human inexperience or negligence or when aroused by pain (D 9.1.1.4, D 9.1.1.5, D 9.1.1.6, D 9.1.1.7, D 9.2.2 pr et seqq, D 9.2.27.5). (My underlining).

22

Johannes Voet (1647-1713) in his Commentaries ad

Pandectas book 9 title 1 founds his entire comment in

respect of pauperien liability on the principles of Roman

law, as elaborated on supra. In 9.1.6 he states

inter alia the following (which is almost verbatim derived

from D 9.1.1.5 (supra)) in relation to injury inflicted by a

dog in the control of a third party as controller or handler:

Sed &, si canis, cum duceretur ab abliquo,

asperitate sua evaserit, & alicui damnum

dederit, vel oves, gallinas, anseres

alienas occiderit, si contineri firmius ab alio potuerit, vel per eum locum duci non debuerit, cessante hac actione de pauperie, contra ducentem utili Aquiliae locus est, D 9.1.1.5, D 9.2.11.5. Gane's translation:

"Then again, if a dog, when he was being led by someone, escaped through his own rough temper and did damage to somebody, or killed another person's sheep, hens or geese, and if he could have been more firmly held in by another or ought not to have been led over such a spot, this action on pauperies falls

23

away but there is room for a beneficial

Aquilian action against the leader." (My underlining). In this passage Voet endorses the views expressed by Ulpianus

in D 9.1.1.5, read with D 9.2.11.5, viz. that a third

party in control of a dog, which owing to his negligent

conduct injures a victim, is liable under the Lex Aquilia,

whereas the owner of the dog will be exonerated from

pauperien liability. These texts, as I indicated supra,

fall in the second category of texts, and the remarks I made

supra concerning the causality of the negligent conduct of

the controller of the dog and the injury or harm to the

victim are also applicable here. The negligent conduct

amounts to a failure on the part of the controller to

exercise proper, i.e. reasonable, control over the dog in

his care. That negligent conduct provided the dog with the

opportunity to injure the victim. The controller did not

by any positive act cause the dog to bite or harm the victim.

Kersteman (1728 - ± 1793) in his Aanhangzel tot

24

het Hollandsch Rechtsgeleerdt Woordenboek, 1772, vol 1 p 292 seqq. s v Damnum ab Animalibus Datum furnishes us with a translation in Dutch of Voet 9.1, which includes Voet 9.1.6, without any original comment or contribution of his own. We are concerned with the application of these principles to the facts of the present matter. It is not necessary to consider whether they have any wider application. Cohen was at all relevant times in charge of and had control over the dog. He knew the propensities and nature of the dog. He even mentioned them to Purdy whom he led to believe that he would lock the dog away when Purdy called at the premises. Moreover, he knew when Purdy would arrive as pre-arranged by them. In the circumstances he owed Purdy a legal duty to take reasonable precautionary measures to contain or restrain the dog from biting him. Cohen's failure to adopt any reasonable precautionary measures in the circumstances amounted to negligent conduct. He accordingly provided the dog with the opportunity to

25

injure Purdy and failed to prevent it from doing so. According to the second category of texts (supra), which Voet 9.1.6 endorses, Cohen as controller of the dog was in the circumstances guilty of negligent conduct which resulted in the injury to Purdy despite the fact that he did not by any positive act cause the dog to bite Purdy. Cohen's Aquilian liability to Purdy afforded Lever, the owner of the dog, a defence which exonerated him from pauperien liability to Purdy. See D 9.1.1.4, D 9.1.1.5, Voet 9.1.6, Damhouder loc.cit., Paul Voet loc.cit.

The Court a quo, in my judgment, erred in deciding that only the first category of texts afforded an owner exemption from pauperien liability.

Lever in my judgment therefore succeeded in establishing his defence.

In the result the appeal must succeed.

The appellant disavowed any order for costs

in this Court and in any of the courts below.

26

The following orders are granted : 1.. The appeal is upheld. 2. The following order is to replace the order of the

Court a quo:

The appeal is dismissed.

C. P. JOUBERT A C J. SMALBERGER J A Concur. HOWIE A J A

JUDGMENT

KUMLEBEN JA/...

1.

I have had the privilege of reading the judgment of my Brother Joubert (the "other judgment"). In it the facts giving rise to this dispute are fully set out. I agree that the appeal should be allowed. I reach this conclusion, however, on a somewhat different approach and for reasons which in certain respects differ from those relied upon by my learned colleague. Hence this separate judgment.

The fact that a third party has incited or provoked an animal to behave contra naturam sui generis for instance, by striking, wounding, scaring or annoying it, constitutes a valid defence to a claim based on the actio de pauperie. In such a case the conduct of the third party is the causa causans of the harm suffered and not any vicious propensity inherent in the animal concerned. This (the "established exception") is an acknowledged ground exempting the

2/...

2. owner from pauperien liability (D 9.1.1.6 and 7; O'Callaqhan, N.O. v Chaplin 1927 AD 310 at 329; and

the judgment of the court a, quo at 438D).

A further qualification (the "wider

exception") is at this stage controversial. 1 refer to

the question whether fault on the part of a third

party, short of conduct falling within the established

exception, which causatively contributes to injury

arising from an animal acting contra naturam sui

generis, is similarly an answer to a pauperien claim.

For instance, the negligence of a visitor to premises

who leaves a gate open giving access to a yard within

which a vicious dog is confined. There is authority

favouring or pointing towards the recognition of this

exception. (See van Leeuwen C.F. 1.5.31.3; Le Roux

and Others v Pick 1879 Buch 29 at 37; and Joubert The

Law of South Africa Vol 1 para 378 page 225.) However,

3/...

3. this question as far as this court is concerned, remains an open one and for the purposes of this appeal need not be decided.

The pertinent and only question calling for decision in this case is whether the culpable conduct of a person to whom the owner entrusts custody and control of an animal relieves the owner of liability which would otherwise by virtue of the actio de pauperis have existed (the "exception in issue"). This exception has thus two components: the delegation of control to some person and causative negligence on his part.

The Justinian rubric (D 9.1.) in which the actio de pauperie is discussed, after relating the history and explaining the nature of this action, recites a miscellany of instances giving rise to such an action, interspersed with illustrations of

4/...

4. situations in which the injured party is restricted to the Aquilian remedy. From these examples one must attempt by deduction to discern the nature and extent of any qualification of pauperien liability recognised in Roman Law. As regards the exception in issue, D 9.1 .1 .5 is to my mind significant. It is cited in the other judgment, but for ease of reference I repeat Watson's translation of this text:

"Take the case of a dog which, while being taken out on a lead by someone, breaks loose on account of its wildness and does some harm to someone else: If it could have been better restrained by someone else or if it should never have been taken to that particular place, this action will not lie and the person who had the dog on the lead will be liable."

It is true that this example does not expressly state

that the owner of the dog entrusted it to the person

leading it though this would seem to be a justifiable

inference. But, even if it is to be given the wider

5/...

5. connotation, it is the negligence of the person who is in control of the dog which results in the pauperien action not being available to the injured person. Thus on either interpretation I consider that Mr Cameron, who appeared on behalf of the appellant, was justified

in relying on this example in support of the exception

in issue. In regard to D 9.1.1.4, if it can be taken to justify or confirm the wider exception it would a fortiori serve as further authority in favour of the exception in issue. I, however, cannot thus construe it. It reads - again Watson's translation - thus:

"Therefore, as Servius writes, this action lies when a four-footed animal does harm because its wild nature has been excited, for example, when a horse given to kicking actually kicks someone or an ox likely to gore tosses someone or mules cause damage cm account of some unusual vice. On the other hand, if an animal should upset its load onto someone because of the roughness of the ground or a mule driver's negligence or because it was overloaded, this action will not lie and

6/...

6.

proceedings should be brought for wrongful damage."

The first sentence of this text plainly deals with

pauperien liability. The second, as clearly and in

stated contrast, gives an illustration of Aquilian

liability. One cannot deduce that the animal or mule in

this second illustration acted contra naturam sui

generis: in fact the opposite is the more natural

inference. That being the case the pauperien action

will not lie, not because the Aquilian remedy is

available, but because the animals behaved secundum

naturam sui generis. I cannot accept in reference to

this text that the mere fact that an injured party has

an independent cause of action (the Aquilian one)

against a wrongdoer ipso jure excludes another

remedy (the pauperien action) to which an injured

person is in law entitled. After all, in the parallel

7/...

7. situation of vicarious liability, the actions against both the master and the servant co-exist.

As regards our common law writers, I endorse the observation of Eloff DJP in the judgment of the court a quo at 437 A - B that they, following the precedent of the Justinian texts on which they comment, are casuistic in their approach. No explicit statement of exceptions to pauperien liability is laid down by them nor is one manifestly discernible. As pointed out in the other judgment, there is a general dearth of Roman Dutch authority dealing with the limits to this cause of action. However, having said this, I respectfully agree that the passage cited in that judgment from Damhouder's. Practycke in Criminele Saken 142 Capitel no 3 and Johannes Voet Commentarius ad Pandectas 9.1.6 are in point and can validly be relied upon as authority for the recognition of the

8/...

8.

exception in issue.

In the absence of further authorities, it would seem that the court a quo relied mainly on two other considerations. (i) What was said - or not said - in O'Callaghan's case (supra) and in South African Railways and Harbours v Edwards 1930 AD 3; and (ii) the so-called "risk principle".

At page 433H - 434H the learned judge said:

"The authorities, many of which are collected and analysed in the decisions in O'Callaghan NO v Chaplin 1927 AD 310 and South African Railways and Harbours v Edwards 1930 AD 3, indicate that the liability of the owner of a domesticated animal which causes damage acting from some inner excitement or vice is based on mere ownership, and not on fault on the part of the owner. It is a case of strict liability. In O'Callaghan's case supra at 329 Innes CJ summarised the position thus:

'By our law, therefore, the owner of a dog that attacks a person who was lawfully at the place where he was injured, and who neither provoked the attack nor by his negligence contributed to his own injury, is liable, as

9/...

9.

owner, to make good the resulting damage.'

He adds (at 329 - 30):

'The conclusion is satisfactory for two reasons especially. In the first place it provides a remedy in cases where otherwise persons injured would be remediless. Instances must occur where a dog, a bull, or other domesticated animal inflicts damage under circumstances which make it impossible to bring home negligence to the owner. Yet of two such persons it is right that the owner, and not the innocent sufferer, should bear the loss. And in the second place the adoption of culpa as the sole basis of liability would inevitably lead us towards the scienter test - a doctrine definitely ruled out by Graham v Viljoen (Buch 1878 p 126) and which it is common cause is not the test which our law applies in cases of this kind. Because it would be difficult to hold a man negligent who had no knowledge of the vicious propensities of his dog, unless it were held that the mere attack implied negligence - res ipsa loquitur - which might in its practical application approximate to the ownership basis, but which would not equally be defensible in law'.

It is significant that while the Chief Justice specially mentions that if the injured person either provoked the attack or by his own

10/...

10.

negligence contributed to his own injury he has no claim, it is not said that the negligence of another may affect the owner's liability.

In the South African Railways and Harbours case the Chief Justice ventured a further summary of 'the relevant principles of our law laid down in the more authoritative cases'. Of those the following should be quoted (at 10 of the report):' [Five of the ten principles summarised by de Villiers CJ are then cited by Eloff DJP and his judgment continues.]

It is again to be noted that while the role of a third party is specifically mentioned as something which could affect the claim of the injured party, the Court said that it is only when that third party provokes the animal that the plaintiff's claim might be defeated. Negligence of a third party is not mentioned as a defence."

Thus the absence of any reference to a situation falling within the purview of the exception in issue is regarded by the court a. quo as some indication that this exception ought not to be approved. But in the first-quoted extract from O'Callaghan's case, I do not understand Innes CJ to have pronounced exhaustively

11/...

11.

on the acto de pauperie. And as regards the other

decision - as de Villiers CJ himself prefaces - his

summary of principles is based upon what has been

decided "in the more authoritative cases" none of

which, as far as I am aware, adjudicated upon the

exception in issue. In both these cases the courts

were concerned with the facts before them, facts which

do not correspond with those in the present case. For

this reason the validity of the exception in issue did

not arise and any general pronouncement in these

judgments cannot be taken to detract from its

acceptance. In the circumstances I do not consider that

these earlier decisions of this court lend weight to

the argument that the exception in issue ought not to

be acknowledged.

At page 438 C - D of the judgment of the court a quo it is stated that:

12/...

12.

"The policy of the law in imposing strict liability on an owner of a domestic animal is that he creates a source of danger and should be answerable even if the neglect of another contributes to the event causing damage. Only conduct of a third party which occasions the animal to behave in the manner complained of will relieve him of liability." (438C - D)

This principle is likewise relied upon in a note by

Prof C G van der Merwe on the decision of the court a

quo, which decision the writer endorses: see

[1992] 109 South African Law Journal 398 particularly

at 401 and 402. It is true that the creation of risk

is a reason put forward as justification for strict

liability featuring in our law in certain

circumstances. (On risk liability generally cf Minister

of Law and Order v Nqobo 1992(4) SA 822(A) at 832E -

834A). But the underlying reason for such a principle

is not in itself the determinant of its content or

13/...

13. ambit. As pointed out by Schreiner JA, though in reference to vicarious liability, in Carter & Co (Pty) Ltd v McDonald 1955(1) SA 202(A) at 211H:

"the fact that a main reason for the existence of the principle of the master's liability may be that he has created a risk for his own ends does not mean that wherever by his words or actions he has created or increased the risk to other persons he is liable. It is often useful to examine the reason which probably gave rise to the rule, in order to discover the rule's limits, but the reason, even if certainly established, is not the same as the rule."

It must also be borne in mind that liability without fault runs counter to fundamental legal precept, though in certain instances considerations of social policy no doubt justify its existence. Where the owner of an animal has taken care to entrust it to another as its custodian, the former has ex hypothesi no means of exercising control over it. Competing

14/...

14.
interests are plainly at stake. Should the owner in
such a case be held liable in the absence of any fault

on his part or should the injured person be restricted to an action against the negligent custodian? Dictates of fairness and justice, to my mind, favour the owner and warrant the recognition of the exception in issue.

For these reasons, and on the authorities quoted, I agree that the appeal should succeed. It is hardly necessary to add that fault on the part of the owner in the selection of a custodian (culpa in eligendo) could give rise to an Aquilian action against the owner at the instance of the injured party and that in certain circumstances the former, if the custodian was his employee, would be vicariously liable.

M E KUMLEBEN JUDGE OF APPEAL

NIENABER JA - Concurs


▲ To the top