Case No 378/2001
REPORTABLE
THE SUPREME COURT OF APPEAL
OF SOUTH AFRICA
In the matter between
TRANSNET LIMITED t/a PORTNET APPELLANT
and
THE OWNERS OF THE mv ‘STELLA
TINGAS’ FIRST RESPONDENT
THE mv ‘ATLANTICA’ SECOND RESPONDENT
CORAM : HEFER AP, SCOTT, FARLAM, CONRADIE, JJA et LEWIS AJA
HEARD : 4 NOVEMBER 2002
DELIVERED : 27 NOVEMBER 2002
___________________________________________________________
Admiralty - collision in Durban harbour - negligence of pilot - whether gross negligence - s6 of Act 105 of 1983 - non applicability of UK Pilotage Act of 1983
J U D G M E N T
______________________________________________________________
SCOTT JA/…
SCOTT JA:
[1] Shortly after midnight on 17 June 1997 the mv Atlantica, a bulk carrier, 224 metres in length and displacing some 65 000 metric tons, collided with the mv Stella Tingas in Durban harbour. At the time the latter vessel was alongside loading cargo at Island View berth 3. The Atlantica was in the process of entering the Island View Channel and was headed for berth 7 where she was to take on bunkers. Durban harbour is a compulsory pilotage port. The pilot navigating the Atlantica, Captain Peter Buffard, was an employee of the harbour authority, Transnet Limited, which is the present appellant. Both vessels were damaged in the collision. The owners of the Stella Tingas, the first respondent, (to whom I shall refer as the plaintiffs) instituted action in the Durban and Coast Local Division against the Atlantica, as first defendant (now the second respondent), and against Transnet, as second defendant, the action against the former being in rem and against the latter in personam.
[2] The claim against the Atlantica was founded on two grounds. The first was that the collision was caused by the negligence of the master and crew. The second was that it was caused by the negligence of the pilot for whose negligence the owners of the Atlantica were liable by reason of s 35 of the United Kingdom Pilotage Act of 1983 which, it was alleged, was applicable by virtue of the provisions of s 6(1) of the Admiralty Jurisdiction Regulation Act 105 of 1983 (‘the Act’).
[3] With regard to the claim against Transnet, the principal ground and the only one relied upon in this Court, was that the collision was caused by the gross negligence of the pilot. The reason for the allegation that the negligence was gross was an attempt to avoid the exemption of liability afforded to both Transnet and the pilot by subpara 10(7) of the First Schedule to the Legal Succession to the South African Transport Services Act 9 of 1989 (‘the Succession Act’). The subparagraph reads:
‘The Company and the pilot shall be exempt from liability for loss or damage caused by a negligent act or omission on the part of the pilot.’
The company referred to is Transnet, the company established in terms of s 2 of that Act. Another ground upon which the plaintiffs sought to hold Transnet liable was the negligence on the part of the master and crew of each of the two tugs in attendance at the time of the collision. However, this ground was formally abandoned during the course of the trial.
[4] In addition to denying liability, the Atlantica caused a Third Party notice to be issued citing Transnet and the pilot as third parties and claiming from them inter alia damages in respect of the damage caused to the Stella Tingas.
[5] The Court a quo (Booysen J) was asked to decide only the question of liability. The learned judge found that the pilot had been grossly negligent and that Transnet was accordingly liable to the plaintiffs for their damages. As to the claim against the Atlantica, he found that negligence on the part of the master had not been established and that the provisions of s 35 of the United Kingdom Pilotage Act of 1983 were not applicable in South Africa. The Atlantica was accordingly held not to be liable to the plaintiffs. No order was apparently sought in terms of the Third Party notice, nor was one granted. The judgment of the Court a quo has been reported: Owners of the mv Stella Tingas v mv Atlantica and Another (Transnet Ltd t/a Portnet and Another, Third Parties) 2002(1) SA 647 (D). Transnet now appeals against the order holding it liable to the plaintiffs. The plaintiffs, in turn, appeal against that part of the judgment in which the Atlantica was held not to be liable to the plaintiffs. This appeal is conditional on Transnet’s appeal succeeding. Both appeals are with the leave of the Court a quo.
[6] Subsections 6(1) and (2) of the Act provide as follows:
‘6(1) Notwithstanding anything to the contrary in any law or the common law contained a court in the exercise of its admiralty jurisdiction shall -
(a) with regard to any matter in respect of which a court of admiralty of the Republic referred to in the Colonial Courts of Admiralty Act, 1890, of the United Kingdom, had jurisdiction immediately before the commencement of this Act, apply the law which the High Court of Justice of the United Kingdom in the exercise of its admiralty jurisdiction would have applied with regard to such a matter at such commencement, in so far as that law can be applied;
with regard to any other matter, apply the Roman-Dutch law applicable in the Republic.
(2) The provisions of subsection (1) shall not derogate from the provisions of any law of the Republic applicable to any of the matters contemplated in paragraph (a) or (b) of that subsection.’
The plaintiffs’ claims relate to ‘damage done by a ship’. Accordingly, and by virtue of s 7 of the English Admiralty Courts Act of 1861, a court of admiralty in South Africa would have had jurisdiction to entertain the claims immediately before the commencement of the Act on 1 November 1983. It follows that in terms of s 6(1)(a) the law to be applied is the law which the ‘High Court of Justice of the United Kingdom’ would have applied. (The reference to that Court is presumably intended to be a reference to the Supreme Court of England and Wales as constituted by the Supreme Court Act 1981. See Brady-Hamilton Stevedore Co and Others v mv Kalantiao 1987 (4) SA 250 (D) at 253 D.) It is clear that regard is to be had to the law as it existed on 1 November 1983. (See Transol Bunker BV v mv Andrico Unity and Others 1989 (4) SA 325 (A) at 334 H.) Subsection 6(1) of the Act is, however, subject to ss 6(2). Paragraph 10 of the First Schedule to the Succession Act contains detailed provisions relating to compulsory pilotage harbours and in particular any negligent act or omission on the part of a pilot. These provisions are clearly provisions within the meaning of ss 6(2) of the Act and the former must accordingly prevail over what for convenience may simply be referred to as the English admiralty law. Paragraph 10 is, of course, part of a South African statute and must be construed as such, despite the provisions of s 6(1) of the Act. It follows that for the purpose of determining what is ‘a negligent act or omission on the part of the pilot’ within the meaning of subpara 10(7) (quoted above) or whether an act or omission amounted to gross negligence so as not to enjoy the benefit of the exemption conferred by subpara 10(7), regard must be had to the South African law.
[7] I shall assume, without deciding, that the exemption would not apply if the pilot were found to have been grossly negligent. Gross negligence is not an exact concept capable of precise definition. Despite dicta which sometimes seem to suggest the contrary, what is now clear, following the decision of this Court in S v Van Zyl 1969 (1) SA 553 (A), is that it is not consciousness of risk-taking that distinguishes gross negligence from ordinary negligence. (See also Philotex (Pty) Ltd and Others v Snyman and Others 1998 (2) SA 138 (SCA) at 143 C - J.) This must be so. If consciously taking a risk is reasonable there will be no negligence at all. If a person foresees the risk of harm but acts, or fails to act, in the unreasonable belief that he or she will be able to avoid the danger or that for some other reason it will not eventuate, the conduct in question may amount to ordinary negligence or it may amount to gross negligence (or recklessness in the wide sense) depending on the circumstances. (Van Zyl’s case, supra, at 557 A - E.) If, of course, the risk of harm is foreseen and the person in question acts recklessly or indifferently as to whether it ensues or not, the conduct will amount to recklessness in the narrow sense, in other words, dolus eventualis; but it would then exceed the bounds of our modern-day understanding of gross negligence. On the other hand, even in the absence of conscious risk-taking, conduct may depart so radically from the standard of the reasonable person as to amount to gross negligence (Van Zyl’s case, supra, at 559 D - H.) It follows that whether there is conscious risk-taking or not, it is necessary in each case to determine whether the deviation from what is reasonable is so marked as to justify it being condemned as gross. The Roman notion of culpa lata included both extreme negligence and what today we would call recklessness in the narrow sense or dolus eventualis. (See Thomas Textbook of Roman Law at 250.) As to the former, with which we are presently concerned, Ulpian’s definition, D 50. 16. 213. 2, is helpful : ‘culpa lata is extreme negligence, that is not to realise what everyone realises’ (culpa lata est nimia neglegentia, id est non intellegere quod omnes intellegunt). Commenting on this definition, Lee in The Elements of Roman Law 4 ed at 288 describes gross negligence as being ‘a degree of negligence which indicates a complete obtuseness of mind and conduct’. Buckland in A Textbook of Roman Law 3 ed at 556 suggests that what is contemplated is a ‘failure to show any reasonable care’. Dicta in modern judgments, although sometimes more appropriate in respect of dolus eventualis, similarly reflect the extreme nature of the negligence required to constitute gross negligence. Some examples are: ‘no consideration whatever to the consequences of his acts’ (Central South African Railways v Adlington & Co 1906 TS 964 at 973); ‘a total disregard of duty’ (Rosenthal v Marks 1944 TPD 172 at 180); ‘nalatigheid van ’n baie ernstige aard’ or ‘’n besondere hoë graad van nalatigheid’ (S v Smith en Andere 1973 (3) SA 217 (T) at 219 A - B); ‘ordinary negligence of an aggravated form which falls short of wilfulness’ (Bickle v Joint Ministers of Law and Order 1980 (2) SA 764 (R) at 770 C); ‘an entire failure to give consideration to the consequences of one’s actions’ (S v Dhlamini 1988 (2) SA 302 (A) at 308D). It follows, I think, that to qualify as gross negligence the conduct in question, although falling short of dolus eventualis, must involve a departure from the standard of the reasonable person to such an extent that it may properly be categorized as extreme; it must demonstrate, where there is found to be conscious risk-taking, a complete obtuseness of mind or, where there is no conscious risk-taking, a total failure to take care. If something less were required, the distinction between ordinary and gross negligence would lose its validity.
[8] Against this background, I turn to the facts. It was common cause that ships entering Durban harbour bound for a berth in Island View Channel follow more or less the same route. After proceeding through the harbour entrance channel and upon entering the area known as Basin B they steer several degrees to starboard to keep clear of the coal berths along the
Bluff on the port side, then alter back to port before steering again to starboard in order to line up with the leading lights and proceed up Island View Channel. The leading lights are placed beyond the channel and indicate its centre line. Because there are berths on the port side of the channel it is usual for ships to proceed up the channel slightly to the starboard of the centre line. The distance between the entrance to Basin B and the entrance to Island View Channel was not given in evidence, but from other distances given it would appear to be in the region of just under a sea mile.
[9] By the time the trial commenced the pilot, captain Buffard, had died. Nonetheless, from his statement made shortly after the incident, his evidence at a subsequent inquiry, the Atlantica’s logs and a transcript of the conversation between the pilot and the masters of the two tugs in attendance, it is possible to obtain a fair idea as to both the route taken by the Atlantica and the events in the critical period immediately preceding the collision. I pause to mention that these documents, if for no other reason, were admissible in terms of s 6(3) of the Act. (Cf Cargo Laden and Lately Laden on Board the mv Thalassini Avgi v mv Dimitris 1989 (3) SA 820 (A) at 842 B - D.)
[10] It appears that whether as a result of a misunderstanding between the pilot and the helmsman or otherwise, the Atlantica on entering Basin B proceeded further to starboard than is usual and had to alter back to port a correspondingly further distance. Just what the extent of the deviation was is not clear. Ultimately its consequence was that the angle at which the vessel approached the mouth of Island View Channel was different from the usual and, in turn, required a harder turn to starboard in order to proceed up the channel. According to the pilot (as appears from the transcript of his evidence at a subsequent fact-finding commission), by the time he had steadied the ship following the turn back to port, she was abeam of coal berth 2 on the Bluff and heading in the direction of Island View berth 2. A reference to a chart of the harbour suggests that the deviation had been corrected and the ship set on course for the channel when still some distance from its mouth. The pilot said he maintained this course until he was ready to turn to starboard into Island View Channel, by which time the starboard bow was approaching buoy I1. This was then positioned off the point and adjacent to the starboard side of the channel. The Stella Tingas, as I have said, was moored at Island View berth 3 on the port side of the channel. According to the duty officer who was standing on the starboard side of that vessel, he saw the Atlantica rounding buoy I1 and approaching at an angle of about 25 degrees.
[11] On the face of it, there would appear to be nothing untoward about the manoeuvre the pilot sought to execute. Indeed, there was evidence to the effect that ships coming from other parts of the harbour would not infrequently approach the channel at a similar angle before steering to starboard and proceeding up the channel to take on bunkers.
[12] In the event, the Atlantica failed to respond to the helm and sheered to port. The duty officer on the Stella Tingas saw that a collision was imminent and fled. The Atlantica struck the Stella Tingas a glancing blow before proceeding up the channel.
[13] The experts were agreed that the sheer to port was caused by phenomena referred to respectively as ‘squat’ and ‘bank effect’, both of which were related to speed. The former can occur in shallow water and particularly in a narrow channel. I interpose that what is shallow or narrow is, of course, related to the ship’s size and draft. A ship displaces its own weight in water. Simply put, as it moves forward it leaves a void behind which has to be filled. In a confined area the flow of water to fill the void is retarded by the bulk of the ship and this, in turn, as it was said, plays havoc with the steerage. The ‘bank effect’ in the instant case would have been caused by water displaced by the ship, as she entered the channel and rounded the starboard point of the so-called island, pushing up against the bank, which presumably shelves steeply, and then returning to create pressure on the bow forcing it to port. As it appears that the sheer to port commenced as the Atlantica was in the process of entering Island View Channel, it would seem likely that the cause of the sheer, at least initially, was the ‘bank effect’, rather than the ‘squat’.
[14] Much of the evidence by the experts, mainly former pilots, related to the question of the Atlantica’s speed as she proceeded through the entrance channel of the harbour and across Basin B. It is necessary to mention at this stage that the only witness at the trial who was on board the Atlantica at the time was the master. Admittedly he was not in control of the navigation of the ship and may not have paid as much attention to what was happening as he would have had the pilot not been in charge. Nonetheless, he was familiar with Durban harbour; he said that the speed seemed reasonable to him and that he had entered the harbour before at the same speed. The master impressed the Court a quo as a good witness.
[15] Prior to the Atlantica entering the harbour the fuel supply was changed from heavy to diesel oil. The latter would have enabled the ship to achieve a speed of approximately 9 knots. According to the ‘engine room movement book’ (‘the engine book’) and the ‘deck bell book’ (‘the bell book’), as the Atlantica approached the harbour entrance she proceeded at full speed ahead for some 8 minutes. The speed of the engine was then reduced to half speed ahead and then almost immediately thereafter to slow ahead. This was some 10 minutes before the vessel, according to the ship’s log, passed the breakwater light. A change in the engine speed would not have had an immediate effect on the speed of the ship through the water, particularly having regard to her size and displacement. Much of the debate in the evidence related to the extent to which the ship would have lost headway while proceeding through the harbour entrance and across Basin B. The engine book and the bell book reflect several changes from slow ahead to half ahead, then back to slow ahead. The pilot’s report confirms that these were made not so much to increase or reduce the speed of the ship through the water but rather to maintain control of the steering. Indeed, after the initial period of full ahead the ship would have been steadily losing headway. The engine changes in relation to when the ship entered the port and when the collision occurred are, in any event, far from clear. The reason is that not only were the clocks on the bridge and in the engine room not synchronised but some of the entries were presumably not made immediately so that the time difference between the two logs is not constant. In addition, entries in the engine book are not always recorded in the bell book.
[16] Based on the engine speed of full ahead for eight minutes prior to the Atlantica entering the entrance channel, Captain Dominy, who testified on behalf of the plaintiffs, estimated that the ship proceeded through the entrance channel at a speed of about 9 knots. Captain McGregor, who testified for Transnet, thought that the speed was more likely to have been in the region of 7 knots. Both estimates are little more than guesses. Captain Martin, another of the plaintiffs’ witnesses, noted that according to the ship’s log the vessel passed the breakwater light at 00h21 and collided with the Stella Tingas at 00h35, ie a difference of 14 minutes. Although the actual times differed, the port record similarly showed a difference of 14 minutes between the time the ship ‘crossed into port’ and the collision. Captain Martin assumed the reference to the breakwater light to be the light on the south breakwater and measured the distance between that light and where the collision occurred to be 1,69 sea miles. This gave him an average speed of 7,24 knots. If the measurement was taken from the light on the north breakwater he arrived at an average of 6,3 knots. No evidence was led as to the point at which the port authority regarded the ship as having crossed into port. Captain Dominy did a similar exercise, but measured the distance between the south light and the point where the collision occurred at 1,67 sea miles, which gave him an average of 6,7 knots. However, counsel for Transnet pointed out, correctly I think, that because the times, 00h21 and 00h35 were recorded to the nearest minute, there could have been a difference of up to one minute between the two points in time; in other words, it could have taken the ship closer to 15 minutes to cover the distance in question. In view of the short time involved, he argued, this could have made a considerable difference to the calculation of the average speed. In this regard, it is not without significance that the master rejected the suggestion that the ship covered the distance in question at an average speed of as much as 7,24 knots. Finally, with regard to the actual speed of the Atlantica, it was recorded in the log that her speed ‘as per GPS [Global Positioning System] during the collision’ was 3,4 knots. No attempt was made to explore when the system was activated or over what distance the speed was measured. Captain Martin conceded that he was unable to dispute that when the collision occurred the speed of the Atlantica was 3,4 knots.
[17] The evidence of what would have been an appropriate speed in the circumstances is also somewhat contradictory. Captain Martin expressed the view that a reasonable speed, if the ship had followed the usual route, would have been about 5 knots. The Atlantica’s expert, Mr Fiddler, on the other hand, regarded it not unreasonable for the ship to have proceeded through the entrance channel at 7,5 knots. Both Captain Martin and Captain Dominy testified that ships normally proceed up Island View Channel at a speed of approximately 3 to 4 knots, but that, of course, was the speed registered on the GPS when the collision occurred at a point close to the mouth of the channel.
[18] On the basis of the aforegoing, I am unpersuaded that the evidence establishes that the Atlantica proceeded through the harbour entrance and across Basin B at a speed which would have been excessive had the ship followed the usual route. I am also unpersuaded that the angle at which the Atlantica approached Island View Channel or the turn to starboard which the pilot proposed to execute was shown to have been so untoward as to give rise to an inference of negligence on his part. On the other hand, the very fact that the ship was unable, whether as a result of the ‘squat’ or ‘bank effect’ or both, to execute the turn to starboard safely is indicative of a speed which was excessive for the manoeuvre contemplated. For this the pilot must take the blame; he was accordingly guilty of negligence.
[19] The plaintiffs’ experts expressed the view that once the pilot became aware of the deviation, he should have stopped the ship by putting the engine astern and then with the assistance of the tugs manoeuvred the ship to port so as to enable her to enter Island View Channel in the usual way. This, of course, was a view expressed with hindsight and strikes me as perhaps requiring a standard of excellence which is not reasonably required. But it is nonetheless clear that when approaching the channel at the angle he did, the pilot by whatever means should have ensured that the ship was proceeding at a slower speed and if necessary maintained steering control by using what was described as ‘short kicks on the engine’.
[20] The question remains whether the pilot’s negligence amounted to gross negligence. In finding that it did, Booysen J said the following:
‘He knew that he was about to enter a narrow and shallow channel where the danger of a sheer caused by the suction effect or the banking effect, or both, was always present if he went too fast. He knew that in the event of a sheer he would have very little time and space to avoid a collision. He knew that the consequences of a collision in the harbour involving a vessel of the size of the “Atlantica” would be considerable. It does seem that he had somewhat of a dare-devil attitude.
The pilot knew that he was going too fast; he knew that he was not aligned with the leading lights; he knew that he was about to enter a narrow, shallow channel with a huge ship; he knew that in the event of a sheer there was likely to be a collision, yet in true dare-devil spirit he tried to perform a manoeuvre which had little chance of success. I find pilot Buffard’s actions [not] to have been reckless but at least grossly negligent in the circumstances.’
[21] This finding appears to attribute to the pilot a conscious taking of a risk in circumstances which would amount to dolus eventualis. The judge says, for example, the pilot ‘in true dare-devil spirit’ attempted ‘to perform a manoeuvre which had little chance of success’. In my view, there is no justification for such a finding on the evidence. The passage quoted contains a number of misdirections. The statement that the pilot knew that he was going too fast is presumably based on the fact that in his evidence before the fact-finding commission he appears to have understated his speed. But, by then, of course, he had the advantage of hindsight. The actual speed of the ship as she approached the mouth of the channel is unknown. All we do know is that it was excessive to the extent that it contributed to the ‘squat’ and ‘bank effect’ which, in turn, caused the sheer to port. Indeed, there is nothing to suggest that in the absence of the sheer the speed was such that the ship would not have completed the turn to starboard and proceeded up the channel without mishap. I can see no justification for the conclusion that the pilot knew in advance that he was going too fast. The master certainly did not get the impression that the speed was excessive. Nor, as I have said, was there anything to suggest that the pilot in a dare-devil spirit was attempting to perform a manoeuvre which had little chance of success.
[22] Much was made in evidence of the phenomena of ‘squat’ and the ‘bank effect’. But it was not a common occurrence and would not necessarily have been at the forefront of the pilot’s mind. No doubt he ought to have foreseen this as a possibility. But it is clear that he did not. This much appears from the transcript of the conversation between the pilot and the masters of the tugs. When the Atlantica failed to respond to the helmsman’s turn to starboard, the pilot’s first reaction was to assume that the aft tug was ‘up against’ the port quarter. This, it was explained, would have caused the tug to serve as a rudder and to force the ship to port. The pilot immediately ordered that tug ‘to take the weight off the ship’s back’. It was only when the tug master responded that he was not touching the ship that the pilot would have realised that there was something else causing the sheer. It was at this stage that he announced that things were ‘going wrong’. In the event, the tugs were unable to arrest the sheer and the forward tug, which was attempting to ‘push in’ the port bow, had to abandon the attempt and move to the stern in order to avoid being crushed.
[23] The trial judge’s observation that the pilot knew that he was not aligned with the leading lights also requires comment. It is true that because of the deviation the ship did not follow the route usually followed by vessels coming into port to berth in Island View Channel. But there was nothing untoward, as such, about the angle at which the Atlantica approached the mouth of the channel. As previously mentioned, there was evidence that ships moving from elsewhere in the harbour would approach the channel at a similar angle.
[24] A further point raised in argument was that the pilot erred or demonstrated a lack of caution by electing to bring the Atlantica in at night, having regard to the size of the ship, the fact that she was 22 years old with a single propeller and was not fitted with bow-thrusters to facilitate lateral movement. The short answer to this is that at all times there were two tugs in attendance and there is nothing in the evidence to suggest that darkness played a role or that the collision would not have occurred had the Atlantica been brought into harbour in daylight.
[25] There can be no doubt that the evidence establishes that the pilot was negligent. In my view, however, the plaintiffs failed to discharge the burden of showing that the pilot was negligent to such a degree that his conduct constituted gross negligence.
[26] In passing I should mention that the pilot was also criticised for not adopting one or more other measures to combat the sheer once it had begun. These included putting the engine astern and dropping the port anchor. Counsel for the plaintiffs conceded, correctly in my view, that the evidence did not establish that any of these measures would have had the desired effect of preventing the collision. In any event, the failure of the pilot to adopt one measure rather than another to combat the sheer would not in the circumstances have amounted to gross negligence.
[27] It follows that Transnet’s appeal against the judgment of the Court a quo in favour of the plaintiffs must be upheld.
[28] It accordingly becomes necessary to consider the plaintiffs’ conditional appeal against the finding of the Court a quo that the Atlantica was not liable to the plaintiffs for the latter’s damages. The principal ground upon which it was contended that the Atlantica was so liable, shortly stated, is the following. The maritime claim in question relates to damage done by a ship. In terms of s 6(1) of the Act, English admiralty law is therefore applicable. Section 35 of the United Kingdom Pilotage Act of 1983 (which became law prior to 1 November 1983) imposes liability on the owner of a vessel under compulsory pilotage for damage caused by the vessel or by the fault of the navigation of the vessel in the same manner as the owner would be liable if the pilotage were not compulsory. Accordingly, so it was argued, the Atlantica is liable for the damage caused by the negligence of the pilot.
[29] Subsections 6(1) and 6(2) of the Act are quoted in para 6 above. I shall return to them later but it is first necessary to say something about the English admiralty law relating to compulsory pilotage. At common law a shipowner is liable for the negligence of a pilot voluntarily engaged just as it would be liable for the negligence of the master. Where, on the other hand, the pilotage is compulsory, the shipowner would not be liable for the negligence of the pilot. The reason for the distinction was explained by Dr Lushington in The Maria (1839) 1 W Rob 95 at 99 [166 ER 508 at 510]. Simply put, in the case of compulsory pilotage the master is compelled to take the pilot on board and the owners are not liable for the acts of a person over whom they have no control and whom they are compelled to employ. A voluntary pilot, by contrast, is employed in the discretion of the master and is considered a servant of the owners. The immunity of owners for damage caused by the fault of compulsory pilots was furthermore reinforced by various statutes in the 19th century, including the Merchant Shipping Act of 1894. This immunity was abolished with effect from 1 January 1918 by the Pilotage Act of 1913 following the International Collisions Convention signed at Brussels in 1910. (For details of the Convention and its implementation, see Owners of the Steamship Towerfield v Workington Harbour and Dock Board [1949] P 10 at 22 - 23, 46 - 47. See generally McGuffie Marsden on the Law of Collisions at Sea 10 ed 246 - 248.) Section 15(1) of the 1913 Act was repeated in identical terms in s 35 of the Pilotage Act of 1983 which repealed the earlier Act. Section 35 reads:
‘Notwithstanding anything in any public or local Act, the owner or master of a vessel navigating under circumstances in which pilotage is compulsory shall be answerable for any loss or damage caused by the vessel or by any fault of the navigation of the vessel in the same manner as he would if pilotage were not compulsory.’
The effect of the section is to render the shipowner liable for loss or damage caused by the fault of a compulsory pilot in the same way as the shipowner would be liable at common law for loss or damage caused by a voluntary pilot. For the purpose of the present case it is important to emphasize that the basis of such liability is that a voluntary pilot (and now by statute a compulsory pilot) is regarded as the servant of the shipowner. This is, and has been for many years, the basis of the shipowner’s liability. It was expressly confirmed to be so by the House of Lords in Esso Petroleum Co Ltd v Hall Russell & Co Ltd [1989] AC 643 (HL) at 683 C - 685 H, [1989] 1 All ER 37 (HL) at 58 f - 60 g.
[30] As previously indicated, the provisions of para 10 of the First Schedule to the Succession Act are provisions within the meaning of s 6(2) of the Act and must prevail over the English admiralty law. For convenience I quote para 10 in full.
‘(1) The harbours of the Company are compulsory pilotage harbours with the result that every ship entering, leaving or moving in such a harbour shall be navigated by a pilot who is an employee of the Company, with the exception of ships that are exempt by statute or regulation.
It shall be the pilot’s function to navigate a ship in the harbour, to direct its movements and to determine and control the movements of the tugs assisting the ship under pilotage.
(3) The pilot shall determine the number of tugs required for pilotage in consultation with the Port Captain, whose decision shall be final.
A master shall at all times remain in command of his ship and neither he nor any person under his command may, while the ship is under pilotage, in any way interfere with the navigation or movement of the ship or prevent the pilot from carrying out his duties except in the case of an emergency where the master may intervene to preserve the safety of his ship, cargo or crew and take whatever action he deems necessary to avert the danger.
Where a master intervenes, he shall immediately inform the pilot thereof and, after having restored the situation, he shall permit the pilot to proceed with the execution of his duties.
The master shall ensure that the officers and crew are at their posts, that a proper look-out is kept and that the pilot is given every assistance in the execution of his duties.
The Company and the pilot shall be exempt from liability for loss or damage caused by a negligent act or omission on the part of the pilot.
For the purpose of this item, “pilot” shall mean any person duly licensed by the Company to act as a pilot at a particular harbour.’
It will immediately be observed that the pilot is expressly stated to be an employee of the Company, ie Transnet, (subpara 1). In terms of subpara 2 the pilot is to navigate a ship in the harbour. Subparagraph 4 prohibits the master, who is the shipowner’s agent, from in any way interfering with the navigation or movement of the ship or preventing the pilot from carrying out his duties while the ship is under pilotage except in a case of emergency. These provisions are wholly inconsistent with the position in England where the pilot, whether voluntary or compulsory, is pro hac vice the shipowner’s servant. Expressed differently, to hold the shipowner liable for the negligence of a compulsory pilot would be contrary to the provisions of para 10. Indeed, if the shipowner were vicariously liable, subpara 10(7), to the extent that it exempts the Company, would be unnecessary.
[31] It follows that the effect of s 6(2) of the Act, read with para 10 of the First Schedule to the Succession Act, is to preclude the application of s 35 of the 1983 Pilotage Act in South Africa. The plaintiffs’ first ground of appeal must therefore fail.
[32] The further ground on which it was contended that the Atlantica was liable to the plaintiffs was that the master was negligent for failing to take steps to prevent the collision. In terms of subpara 10(4) of the Schedule quoted above, the master was in effect prohibited from interfering in the navigation or movement of the ship until such time as there was an emergency. The evidence of the master was that he had full confidence in the pilot and believed that with the assistance of the tugs the ship would be able to execute the turn to starboard and proceed up Island View Channel; it was only when the forward tug abandoned the attempt to push the bows around to the starboard that he realised that there was going to be a collision. It was common cause that by then a collision was inevitable. The master denied having heard the pilot say to the tug masters that things were ‘going wrong’. But even if he had, I do not think he can be held to have been negligent for failing to intervene. The words in question were uttered no more than some three minutes before the collision. The pilot was then heavily engaged in issuing commands to the tugs in order to retrieve the situation. No doubt there were various options available in the attempt to avoid a collision, but in my view the master was entitled to assume that the pilot knew what he was doing. As a master of a ship he would have appreciated the dangers of wrongly interfering with the conduct of the pilot. Nearly a century ago Lord Alverstone CJ in The Tactician [1907] P 244 (CA) at 250 said the following:
‘The cardinal principle to be borne in mind in these pilotage cases, that raise difficult questions of law, and very often difficult questions of fact, is that the pilot is in sole charge of the ship, and that all directions as to speed, course, stopping and reversing, and everything of that kind, are for the pilot; and I entirely agree, if I may say so, with great
respect, with the opinions of the very learned judges, from Dr Lushington downwards, to which attention has been called, as to the danger of a divided command, and the danger of interference with the conduct of the pilot; and that if anything of that kind amounts to an interference or a divided command serious risk is run of the ship losing the benefit of the compulsory pilotage.’
I readily endorse these views. In my view the collision was not caused by any negligence on the part of the master. On this ground, too, the appeal must therefore fail
[33] In the result the following order is made.
(a) The appellant’s appeal against the judgment of the Court a quo in the first respondent’s favour is upheld with costs.
(b) The first respondent’s appeal against the dismissal by the Court a quo of the first respondent’s claim against the second respondent is dismissed with costs.
(c) The order of the Court a quo is altered so as to read as follows:
(i) The plaintiffs’ claim against the first defendant is dismissed with costs, such costs not to include the costs incurred in consequence of the first defendant’s Third Party Notice.
The plaintiffs’ claim against the second defendant is dismissed with costs.
D G SCOTT
JUDGE OF APPEAL
CONCUR:
HEFER AP
FARLAM JA
CONRADIE JA
LEWIS AJA