Queensland Railways v Houston

[1994] QCA 529

IN THE COURT OF APPEAL	[1994] QCA 529
SUPREME COURT OF QUEENSLAND

Appeal No. 144 of 1994.

Brisbane

Before Pincus J.A. Davies J.A. Lee J.

[Qld Railways v. Houston]

BETWEEN:

DAVID ALLAN HOUSTON and SUZANNE
KIM HOUSTON (trading as DAVID HOUSTON

LIVESTOCK TRANSPORT)

(Plaintiffs)First Respondents

AND:

QUEENSLAND RAILWAYS

(Defendant) Appellant

AND:

WILLIAM JAMES ROBINSON

(Third Party)Second Respondent

REASONS FOR JUDGMENT - PINCUS J.A.

Judgment delivered 07/12/1994

I have read the reasons for judgment of Lee J, in which the facts of the case are explained in detail. The damage sued for was caused by the second respondent, Robinson, attempting to avoid the oncoming train; he braked hard and, fearing that he would not stop in time, swerved to the left, as a result of which his vehicle overturned.

The principal questions are whether the learned primary judge was entitled to find, first, that there was negligence on the part of the appellant in that its signalling system did not work, and secondly, that the train driver, Gilligan, was negligent in that he did not stop the train when the

ii

signal against the train was not green.

On the first point, there was no evidence of any particular fault in the signalling system; one simply has the fact that on the occasion in question it did not work at the time when the truck driver Robinson drove towards the crossing, although it operated well enough shortly afterwards. There was a temporary malfunction, the cause of which was not ascertained - or at least, not identified in the judge's reasons. His Honour's view was, in substance, that there was a very high standard of care demanded of the appellant in respect of the signalling system which should "be equated with a guarantee" that it would not fail. His Honour concluded that the failure of the lights to function was proof of negligence; it is unclear on what legal principle that conclusion rested. Counsel for the respondents contended that the very occurrence of the failure of the system was evidence of negligence: res ipsa loquitur.

The evidence was sufficient to exclude as a cause of the failure that the system was inherently deficient, and the respondents' consultant Mr Whisson made no criticism of the maintenance routine which, according to the evidence, was applied to the system. It therefore appeared that the temporary failure must have been due to either an error on the part of one of the appellant's employees in the course of maintenance work, or the temporary appearance of a fault

iii

in the system without negligence on anyone's part. There being evidence of three main possible causes, all of them appearing to be unlikely, counsel for the respondents suggested that we should hold one of them to be, of the three, the most likely. But that conclusion leads nowhere unless one is prepared to go further and say that the most likely of the three causes was the cause. The judge did not so find, and in my view the evidence does not warrant such a conclusion.

One is left with a malfunctioning of the system whose cause is unascertained -except that one should, in my view, exclude as causes both an inherent defect in the system and an inadequacy of the maintenance routine.

The argument for the appellant then becomes that even good systems the subject of adequate maintenance regimes sometimes malfunction, particularly if, as in the present case, they are fairly complex. It was not contended by the respondents' counsel that the system presently in issue was shown to be of a particular kind, such that a malfunction other than by the negligence of one of the appellant's employees is improbable; counsel put the contention on the broader footing that if a system designed to warn against serious dangers fails, then the failure is probably due to negligence.

I have found no authority to support that proposition,

iv

which appears to me to be too broad. One could concede that some systems designed to give warning might be shown to be so constructed that any failure in their operation would be reasonably presumed to be due to lack of care in maintenance; but in the absence of such proof it would be up to the person making the allegation positively to show a lack of care. This might perhaps be done without definitely identifying the cause of the failure, for example, by satisfying the Court that the failure was probably due to one of a number of identified causes, each being more consistent with negligence than no negligence. Here, there was no proof of that kind.

To return to the reasons given by the learned primary judge, his Honour did not say that the mere occurrence of a failure in the system was prima facie evidence of negligence on the part of the appellant. But if one treats the reasons as containing, as an implication, that step in reasoning, in my opinion it cannot possibly be supported. And unless that step is taken to be implicit, the reasons simply proceed from a statement that the standard of care required was very high to a conclusion of liability.

In my opinion the primary judge's finding of negligence on the part of the appellant in relation to the signalling system cannot stand.

I turn now to the second principal question, which is

v

whether the primary judge was right in finding that Gilligan failed to bring the train to a halt prior to the level crossing when "the driver's signal was other than green".

Gilligan said the light was green and there was no direct evidence to the contrary. The assistant driver McNiven said that he observed the lights change from red to green as the respondents' vehicle approached, and that supported Gilligan. In cross-examination, when asked whether he relied on his past experience to determine whether the red and green lights were working, he agreed that he had. I am not sure what either the question or the answer meant, but they were relied on as weakening McNiven's assertion that the train drove past a green light. That does not seem to matter for, as the appellant submitted, it seems clear that the judge was not prepared to reject McNiven's evidence. Indeed, his Honour impliedly accepted that McNiven saw the green light; the judge said: "His observation of a driver's signal change to green is consistent with the train's entry upon the shorter track circuit across and to either side of the crossing". It is conceded by the respondents that here the reasoning is in error; when the train entered upon the shorter part of the crossing, the red/green light was well behind the cabin;

further, his Honour overlooked that entry upon the shorter circuit to which his Honour referred did not operate the red/green light.

vi

The evidence showed that it is possible for the green light to come on in normal fashion as the train approached the crossing, without the warning signals against the road traffic operating. It was argued by counsel for the respondents that we should hold - as the judge did not - that the cause of the latter not operating was one which was inconsistent with the driver's green light operating normally; it seems to me that such a finding would be speculative.

One arrives, then, at the position that the driver said the green light was on, but the judge did not believe him, but his Honour did not disbelieve McNiven, who said it was on. There was no positive foundation for a finding that the green light failed to operate, and the view adopted by the trial judge that the driver was negligent in that respect is insupportable.

It follows that I agree with Lee J that the findings of negligence against the appellant must be set aside and the judgment reversed. I should add only that I would have had considerable difficulty, were it necessary to decide the matter, in assenting to the view that the truck driver was guilty of no negligence.

It was argued on behalf of the respondents that, if we
accepted the appellant's contentions, then there should be a
new trial. This Court will not ordinarily order a new trial

vii

where it upsets findings of liability against a defendant.
It is true that in the present case the outcome has been
affected by conceded errors made below and that provides
some foundation for an order for a new trial; in my
opinion, however, there should be no new trial because the
evidence led was insufficient to justify a finding of
negligence in respect of the appellant's signalling system
and there was no evidence to justify a finding that the
driver drove through a light which was "other than green".

I would allow the appeal with costs, set aside the judgment and orders of the primary judge and in lieu order that the plaintiffs' (first respondents') action be dismissed with costs. As to the third party proceedings, they should be dismissed, but with no order as to costs.

viii

IN THE COURT OF APPEAL

SUPREME COURT OF QUEENSLAND

Appeal No. 144 of 1994

Brisbane

Before Pincus J.A. Davies J.A. Lee J.

[Queensland Railways v. Houston & Ors.]

BETWEEN:

DAVID ALLAN HOUSTON and SUZANNE KIM HOUSTON
(TRADING AS DAVID HOUSTON LIVESTOCK TRANSPORT)

(Plaintiffs) First Respondent

- and -

QUEENSLAND RAILWAYS

(Defendant) Appellant

- and -

WILLIAM JAMES ROBINSON

(Third Party) Second Respondent

REASONS FOR JUDGMENT - DAVIES J.A.

Judgment delivered 7/12/94

I have read the reasons for judgment of each of Pincus J.A. and Lee J. I adopt the facts stated by Lee J. for the purpose of these reasons. Both of their Honours have concluded that the appeal should succeed. I agree. Moreover I am unable to detect any divergence in reasoning between them. I am therefore inclined to say that I agree with the reasoning by which each has reached the conclusions

ix

that the respondents failed to prove:

(1) that the failure of the flashing lights to operate was

caused by the negligence of the appellant; and

(2) that the driver of the locomotive proceeded through a railway signal that was other than green.

However in case others may see some difference in the reasoning by which each of their Honours reached the above conclusions I propose to state mine in summary form.

As to (1):
As Lee J. has pointed out, the expert opinion was that there
were three possible causes of failure of the flashing lights
and the respondents conceded that each was effectively
excluded as a probability by the evidence. The experts
having excluded other possible causes it was not open to the
court to infer one. Nor was it open to it to infer, in the
absence of supporting expert evidence, that, whatever the
cause was, it was more probably than not attributable to
negligent maintenance by the appellant. Both questions, the
cause of the failure and whether it would have been
prevented by the exercise of reasonable care by the
appellant, were highly technical ones upon which there was
no affirmative expert evidence; and without affirmative
expert evidence on both questions the respondents could not

prove their case on the balance of probabilities.

As to (2):

x

The rejection of the locomotive driver's evidence that the light was green did not prove that it was other than green;

and there was no evidence to the contrary. Nor was it said
that a green light in the locomotive driver's favour could
not occur when the flashing lights were not working.
Moreover the assistant driver's evidence, also that the
light was green, does not appear to have been rejected by
the trial judge although his Honour's reason for thinking
that the assistant driver's recollection may have been
correct was based on a misapprehension of technical
evidence.

In the light of the concession, rightly made by the respondent, that the learned trial judge was wrong in imposing on the appellant a duty of care equated with a guarantee that the flashing lights would never fail, these conclusions are sufficient to dispose of the appeal in the appellant's favour. I agree, for reasons given by each of Pincus J.A. and Lee J., that there should be no new trial.

I wish only to add some observations about the so-called doctrine of res ipsa loquitur because so much time in this appeal was taken up with its operation and application. In cases where there is compulsory insurance or the defendant is a government instrumentality and where, as here, knowledge of what (if anything) went wrong with a system under the defendant's control is more likely to be within the exclusive knowledge of the defendant, there is a

xi

temptation to invoke the doctrine. But it is important to
appreciate how narrow is its operation.

Inferring negligence from proven facts is, more often than not, a question of commonsense. But where the question whether a person is negligent depends on the cause of failure of a complex piece of machinery, that question is likely to involve highly technical questions upon which even those with expertise may differ. And a court must be careful to ensure that, if an inference of negligence does depend on such technical questions, there is reliable evidence on those questions from which that inference can be drawn. And it must be remembered that the so-called doctrine of res ipsa loquitur is no more than an assertion that, as a matter of commonsense, negligence may be inferred from proven facts. It is unlikely, as a matter of commonsense, that an inference of negligence may be safely drawn as a probability from the fact alone of the failure of a complex piece of machinery.

I agree with the orders proposed by Lee J.

xii

IN THE COURT OF APPEAL

SUPREME COURT OF QUEENSLAND

Appeal No. 144 of 1993

Brisbane
[Queensland Railways v. Houston & Anor]
BETWEEN:

DAVID ALLEN HOUSTON and SUZANNE KIM HOUSTON trading as DAVID HOUSTON LIVESTOCK TRANSPORT

(Plaintiffs) First
Respondents

AND:

QUEENSLAND RAILWAYS

(Defendant) Appellant

AND:

WILLIAM JAMES ROBINSON

(Third Party) Second
Respondent

Pincus JA Davies JA Lee J

Judgment delivered 07/12/1994

Each member of the Court delivering separate reasons.
Davies JA and Lee J agreeing as to the orders to be made;
Pincus JA dissenting in part.

APPEAL ALLOWED. SET ASIDE THE JUDGMENT AND ORDERS MADE BELOW AND IN LIEU ORDER THAT JUDGMENT BE ENTERED FOR THE APPELLANT AGAINST THE FIRST RESPONDENTS WITH COSTS BOTH HERE AND BELOW TO BE TAXED; THAT THE THIRD PARTY PROCEEDINGS BE DISMISSED; AND THAT THE APPELLANT PAY THE SECOND RESPONDENT'S COSTS BOTH HERE AND BELOW TO BE TAXED

CATCHWORDS: NEGLIGENCE - proof of - failure of railway crossing lights maintained by appellant - respondents unable to point to likely cause of failure - unable to fault maintenance system employed by appellant - res ipsa loquitur - whether mere fact of accident bespoke negligence on part of appellant - whether occurrence one which according to

xiii

common experience is normally associated with negligence - application of principle to complex piece of machinery

APPEAL AND NEW TRIAL - interference with findings of fact - whether any basis for finding - whether new trial ought to be ordered in light of failure to establish prima facie case.

Counsel:	N M Cooke QC with him P J Flanagan, for the appellant
A J H Morris QC with him J Miles, for the respondents.
Solicitors:	Crown Solicitor for the appellant Cranston McEachern and Co for the respondents

Hearing date: 4 November 1994
IN THE COURT OF APPEAL

SUPREME COURT OF QUEENSLAND

Appeal No. 144 of 1994

Brisbane

Before Pincus JA

Davies JA

Lee J

[Queensland Railways v. Houston & Anor]
BETWEEN:

DAVID ALLEN HOUSTON and SUZANNE KIM HOUSTON trading as DAVID HOUSTON LIVESTOCK TRANSPORT

(Plaintiffs) First
Respondents

AND:

QUEENSLAND RAILWAYS

(Defendant) Appellant

AND:

WILLIAM JAMES ROBINSON

(Third Party) Second
Respondent

JUDGMENT - W C LEE J

Judgment delivered 07/12/1994

This is an appeal from a judgment of the District Court awarding the first respondents damages for negligence against the appellant, in the agreed sum of $62,238.87. The action was in respect of a motor vehicle accident occurring on the outskirts of Goondiwindi on 15 August, 1991.

The first respondents were at all material times the owners of a Western Star prime mover, commonly described as a road train. On the day in question, that vehicle was being driven by their employee, the second respondent, in a southerly direction along the Leichardt Highway from Moonee, where it had just picked up a load of cattle, to Goondiwindi, approximately 90kms away. The prime mover was towing two stock trailers, each capable of holding two levels of stock. The first or lead trailer was fully loaded, however, only the bottom level of the second or dog trailer was occupied. Somewhere along the journey the prime mover was joined by three other vehicles of similar description. Those vehicles were unassociated with the first respondents' but were, as amongst themselves, travelling in convoy. About half way to Goondiwindi, the lead truck in the convoy overtook the first respondents' vehicle. The four vehicles then remained in that formation for the rest of the trip.

About 2.5kms north of Goondiwindi the highway is intersected by a railway level crossing maintained by the appellant. That part of the railway line in the immediate vicinity of the crossing is notionally divided into three segments. The first, track "A", consists of a section of the line commencing approximately 320 metres from the intersection and ending 15 metres from the commencement of the bitumen. The second, track "B", comprises that part of the line immediately following track "A", extending 5 metres over the width of the highway and ending a further 15 metres on the other side. The third section of the line, track "C", is irrelevant for the purposes of this appeal.

Installed at the crossing is a standard signalling system which, when operational, is triggered by train entering onto track "A". At that stage a circuit is completed which triggers two automatic signalling devices.

The first consists of two sets of flashing lights located on the highway, one on either side of the crossing, which warn users of the highway of the approaching danger. The second is a railway signal, located soon after the commencement of track "A", which, when green, indicates to the driver of the locomotive that it is safe to proceed over the crossing.

Soon after the train enters onto track "A", a further marker indicates the point at which the locomotive driver is to manually blow the train's whistle. The learned trial Judge found, based on the observations of the second respondent and the drivers of the other three prime movers, that at the relevant time the flashing lights were not functioning.

That finding is not challenged before us.

The experts called on both sides were able to suggest three likely causes of the signal failure: rusty rails which would have prevented the circuit from energising; a stuck relay which would have prevented the flashing lights alone from operating; and a loose wire which would have resulted in a total loss of power to the circuit. Of the three, the first was discounted on the basis that the relatively heavy traffic travelling over the line would have prevented corrosion. The second was considered unlikely as a stuck relay would be expected to have resulted in a permanent loss of power to the circuit. The signals were manually operated by one of the appellant's employees approximately 15 minutes after the accident and were observed to be functioning properly. A further two hours after the accident they were also observed to be working when another train passed over the line. Inspections shortly before and after the accident failed to reveal the presence of a loose wire and this too was considered an unlikely cause. A lack of arcing in the circuit also assisted to exclude this conclusion. In that state of the evidence, it was conceded by the respondents in this Court, rightly or wrongly, that each of these causes was "effectively excluded" in the present case.

His Honour also found that the locomotive driver proceeded through a railway signal that was other than green. It will be necessary to return to that finding later as it is the subject of attack by the appellant.

The approach to the crossing on the northern side of the town is marked by a straight, flat section of highway, about 2 to 3kms in length. On the day in question, the weather was fine and visibility excellent. The second respondent was familiar with the crossing, having driven on the same stretch of highway on past occasions. About 200 to 300 metres prior to the crossing, a standard traffic sign marks a decrease in the speed limit from 100 to 80kph. A further sign located 100 to 120 metres from the crossing marks its further reduction to 60kph.

For most of the journey in question the second respondent was travelling at 85kph. On approaching the 80kph sign just mentioned he lifted his foot from the accelerator. That movement activated an automatic braking device within the vehicle which slowed it somewhat. Before reaching the 60kph sign the second respondent, over his UHF radio, heard the driver of the lead prime mover, who by that time had apparently traversed the crossing, warn that the flashing lights were not working. The message was repeated 2 or 3 times. At about the same time as the second respondent approached the 60kph sign he observed a train approaching from his right. According to his evidence, he was then doing 60 to 65kph. On appreciating the danger, he braked heavily and, after realising that he was unable to stop in time to avoid a collision with the train, swerved his vehicle off to the left hand side of the highway. The vehicle skidded for some time, went off the side of the road, hit a gate post and rolled over. The truck and the lead trailer both tipped over on to their right hand sides and ended up in a position perpendicular to the highway.

The dog trailer remained upright and parallel to the highway, just off the side of the bitumen. The overturned vehicle was some 48 metres from the level crossing. The second respondent escaped unhurt.

The technical evidence relating to the stopping distance of a vehicle of a similar weight to the first respondents' carrying a similar load was that, travelling at the speed alleged, it could have stopped safely in under 27m. The vehicle's brakes were said to be in good condition. The second respondent also said that he did not hear the train's whistle although it seems to have been accepted that it was blown; the other drivers heard it.

These matters were relied on by the appellant as evidence of contributory negligence and it will be necessary to refer to this aspect of this case later.

The first respondents' claim was said to consist of damage to the prime mover and trailers and the loss of and damage to stock which it was carrying. Although many particulars of the negligence alleged were pleaded, in the end the case before the learned District Court Judge apparently turned on the two findings which I have outlined: that the flashing lights were not operating at the relevant time and that the locomotive driver proceeded through a railway signal that was other than green.

Standard of Care

It was conceded at the hearing of the appeal that the trial Judge acted under an error of law. That error was expressed in his Honour's conclusion that the standard required of the appellant should be "equated with a guarantee" that "the flashing lights never failed". The relevant standard, of course, was not absolute but rather was to be determined by ascertaining what a reasonable person would do by way of response to the foreseeably created risk. Chief among the factors to be considered in setting that standard were the magnitude of the risk, both in terms of the frequency of its occurrence and the extent of its consequences, and the expense and inconvenience of any alleviating action: Wyong Shire Council v. Shirt (1979- 80) 146 C.L.R. 40, 47.

On that basis this was obviously a case in which the dangers inherent in the appellant's operations demanded of it an extremely high degree of care, but no matter how highly that standard may be expressed it does not assist by speaking of the appellant's obligations in absolute terms.

It was required to do no more than was reasonable in the circumstances.

Nor does it assist to describe the appellant's obligations as non-delegable. Even assuming that the facts of the present case display the requisite degree of control on behalf of the appellant and the corresponding dependence on the part of the respondents necessary to give rise to a non-delegable duty on the part of the appellant to ensure that reasonable care was taken in respect of the carrying on of its operations, that again is not an absolute obligation.

A duty to ensure that reasonable care is taken does not amount to a guarantee that an accident will not occur. The appellant is not an insurer of risks and it advances the argument no further to say that the appellant, as an occupier carrying on a dangerous activity, was under a higher duty than persons in less analogous positions: Burnie Port Authority v. General Jones Pty Ltd (1992-94) 68 A.L.J.R. 331, 345-7; Kondis v. State Transport Authority (1984) 154 C.L.R. 672, 686. The expression "non-delegable duty" is but another way of expressing the proposition that the particular duty is and continues to be a personal one.

It is one which remains on the defendant throughout with the consequence that if reasonable steps are not in fact taken to avoid or minimise the risk of injury, it will be no sufficient answer for the defendant to say that he engaged a competent or an apparently competent person to take them.

If reasonable care is not in fact taken, the defendant is liable. It is a guarantee, not that accidents will not occur, but that reasonable steps will be taken to avoid or minimise the risk of their occurrence.

Before passing from this point I cannot help but note
that His Honour's error seems to have been precipitated by
the embarrassing nature of the first respondents' pleading.
The first respondents themselves allege as part of the
particulars of the appellant's negligence, inter alia, a
failure to ensure that the signalling system was operative
at all times and, in effect, in constant perfect condition.
Indeed if the first respondents' pleading were taken as
accurately representing the current state of the law, the
strict liability which the High Court in Burnie subsumed
into the general law of negligence would be revived. The
appellant could not even escape liability by proving that
the failure of the railway signals occurred by reason of an
act beyond its control or by reason of an act which it took
reasonable care to prevent. Such a conclusion is contrary
to even the most basic understanding of the law of
negligence.

Having conceded that they could not support the judgment on the primary basis on which it was entered, the respondents sought to support it on other grounds, namely:

1.   That the case was one for the application of the principle of res ipsa loquitur and that this Court, being in no worse position than the trial Judge, should draw the inference which the application of that principle is said to give rise to; and

2.    That his Honour was correct in finding that the driver of the locomotive negligently proceeded through a railway signal that was other than green, an act in respect of which the appellant is said to be vicariously liable.

Res Ipsa Loquitur
As with all matters susceptible of proof, evidence of
negligence may be given either directly or circumstantially.
It is given directly when, in a particular case, eye
witnesses testify as to the facts which are said to
constitute the elements of the cause of action; it is given
circumstantially when, from the facts proved, the tribunal
of fact is asked infer those matters against the defendant.

In both cases the legal or ultimate onus always remains on

the party alleging the negligence, i.e. the plaintiff.
Again in both cases, however, the practical or evidentiary
onus may shift according to the state of the evidence at any
particular stage in the proceedings: Purkess v. Crittenden
(1965) 114 C.L.R. 164. So it is that if a plaintiff
establishes a prima facie case against a defendant, the
defendant is placed in the very real position where he must
choose whether or not to introduce evidence in order to
explain or contradict that of the plaintiff. If he decides
not to, he leaves the plaintiff's case to stand or fall by
its own merits and runs the obvious risk that the tribunal
of fact will make the finding which the plaintiff asks of
it. If he does then the question simply becomes one
whether, on the whole of the evidence, the tribunal of fact
is satisfied on the balance of probabilities that the
plaintiff has made out his case. It is a question of law
whether a plaintiff has raised a prima facie case, that is
whether he has adduced sufficient evidence on which a
tribunal of fact, acting reasonably, could find in his
favour. Once that point is reached, however, it is then for
the tribunal to decide as a matter of fact whether, applying
its common sense, it is prepared to accept that evidence as
establishing the plaintiff's case as more probable than not.

These are all uncontroversial propositions. They are fundamental to the way in which our rules of evidence work and are of almost universal application in our Courts. In particular, they do not differ according to how the plaintiff chooses to establish his prima facie case.

In many cases, that stage of the proceedings is reached by the plaintiff adducing credible evidence as to the cause of the accident and the circumstances surrounding its occurrence. In others, it is reached, not by pointing to any specific causative fault on the part of the defendant, but rather by demonstrating some general system or practice which could have been adopted to reduce the risk of its occurrence. But there is no fixed rule regarding how a plaintiff goes about establishing a prima facie case. Each case by necessity depends on its own facts. Sometimes even the mere fact of the occurrence of the accident, without any further reference to its probable cause or surrounding circumstances, will be sufficient to discharge the plaintiff's onus in that regard. Or to put it another way some accidents, by their very nature, do not occur without negligence on the part of the person against whom it is alleged. In those cases, the mere fact of the accident provides the basis on which the tribunal of fact is asked to make the necessary inference. Cases of that kind are conveniently reflected in the Latin phrase res ipsa loquitur meaning literally "the thing speaks for itself". The thing, i.e. the accident, speaks for itself because it is an event which, according to common experience, would not ordinarily have occurred without some want of due care on behalf of the defendant. It is therefore an occurrence which, as a matter of reasonable inference, the defendant caused by his negligence.

For so simple a concept, however, it is surprising how
often it is misunderstood. The tendency has been to apply
it to cases in which the logic behind it has no relevance.
There is of course no harm in the phrase provided it is
properly understood but to some the expression of the
principle in such a convenient form transforms it into some
rigid legal canon. But the words themselves contain no
magic, nor do they express any peculiar or exceptional
doctrine of law. As the High Court has reinforced time and
time again, they do no more than neatly embody the common
sense rule of evidence that facts may be proved
circumstantially: Fitzpatrick v. Walter E. Cooper Ltd (1935)
54 C.L.R. 200; Davis v. Bunn (1936) 56 C.L.R. 246; Mummery
v. Irvings Pty Ltd (1956) 96 C.L.R. 99; Anchor Products Ltd
v. Hedges (1966) 115 C.L.R. 493; The Nominal Defendant v.
Haslbauer (1967) 117 C.L.R. 448; Piening v. Wanless (1967-
68) 117 C.L.R. 498; The Government Insurance Office of New

South Wales v. Fredrichberg (1968) 118 C.L.R. 403.

As with all circumstantial cases however, the inference
which is sought to be drawn from the established facts must
be one which arises as a matter of reason, not as a matter
of conjecture, speculation or hypothesis: Mummery, 117.
Obviously, not every case can be so described; not every
accident bespeaks negligence either on the part of the
defendant or even at all: Anchor Products, 497; Mummery,
113. If it did, the defendant would be placed in the
position of having to positively establish the propriety of
his conduct by demonstrating the exercise of due care on his
part. An indiscriminate application of the principle
therefore would produce the absurd result that a plaintiff
who merely establishes his injury casts on onus on the
defendant to explain its cause or justify his conduct. The
legal onus of proof would be reversed, a result inconsistent
with fundamental law. So it is only in a limited category
of cases that the principle has application. Consequently
it is necessary to identify those factors which act as a
general indicator to the circumstances in which one may
reasonably draw an inference of negligence in the absence of
an explanation as to the cause of the accident.

In a case where the plaintiff seeks to rely solely on the occurrence of the accident and adduces no further evidence in support of his case, the tribunal of fact is required to draw on its common sense and general experience to make the inference which the plaintiffs seeks of it. To avoid the conjectural process of which I have spoken, the accident must be one which "in the normal course of things" or which "in the ordinary affairs of mankind", would not ordinarily have occurred without negligence on the part of the defendant: Mummery, 116; Haslbauer, 452; Piening, 507-8; Fredrichberg, 413. Of course that need not be the sole inference which the facts may support. Others may be suggested, some of which are consistent with the exercise of due care on the defendant's behalf. In order to establish its prima facie case and base the finding sought, it is sufficient if the fact of the occurrence could more probably than not point to a conclusion consistent with the defendant's negligence. In other words it is sufficient if the other suggested causes are improbable enough having regard to the nature of the accident that the tribunal of fact could, on balance, reasonably discard them. That is the first vital condition to the operation of the principle.

The second point which can be made is that if the res is to speak for itself then it must be left for itself to speak. The application of the principle assumes an absence of explanation i.e. that the cause of the accident has not been identified. That does not mean that a plaintiff cannot couch his case in the alternative. It simply means that if a credible explanation for the accident is offered, either on the plaintiff's or the defendant's case, which is accepted by the tribunal of fact, then the question ceases to be one as to whether negligence can be inferred from the mere fact of the accident and becomes one whether on the whole of the evidence that cause was due to the negligent failure of the defendant to avoid or prevent it: Mummery, 115; Anchor Products, 496-8, 506; Haslbauer, 453.

Bearing in mind those matters, the short question for the Court in the present case is whether the mere fact of the occurrence of this accident bespoke negligence on behalf of the appellant. Clearly, in my opinion, it did not. To sheet home to the defendant any responsibility for the accident one must attach the failure of the flashing lights to the defendant's negligent conduct. Quite apart from any question whether that event is to be properly characterised as the res or the explanation for it, it is hardly an event which, as a matter of reasonable inference, can be said to point to negligence on behalf of the appellant. It is not within the common experience of mankind to say whether the failure of such lights is normally associated with a want of due care on behalf of the person charged with their proper maintenance. No expert was prepared to go that far and I do not see how a lay person could. We simply know, on the plaintiff's case, that a piece of highly technical equipment failed: cf. Piening, 508. To go the further step and say that that event pointed towards negligence on the part of the appellant requires a leap in reasoning which I am not prepared to legitimise. With the increasing complexity of modern life that which is said to fall within the Court's or the community's common experience will inevitably lose track with new developments and the application of the principle will be correspondingly restricted. It follows that in my opinion the judgment below cannot be supported on this basis.

I should add that even if I am wrong in this conclusion, I would not be disposed to resolve the matter now. To say that the principle is properly invoked is not to say that the plaintiff automatically succeeds. The further step required is for the tribunal of fact to decide whether it is prepared to drawn the inference of negligence: Anchor Products, 500; Haslbauer; 453. In taking that step, the tribunal would be required to assess the strength of the inference which is said to arise from the occurrence and decide whether it is in any way weakened or qualified by the evidence adduced by the appellant as to the possible causes of the signal failure and the steps taken by it in response to the foreseeable risk of that event. Quite apart from the question of whether matters relating to witnesses' credit might be involved, I am persuaded that it is not legitimate for the respondents to come to this Court and seek it to make such an involved finding merely on the basis that it was not made below.

Vicarious liability

The second basis on which the respondents seek to uphold the judgment is, in effect, his Honour's finding that the appellant was vicariously liable for its employee, the locomotive driver, having negligently proceeded through a railway signal that was other than green. Importantly, no finding was made in respect of the state of the signal at the time i.e. whether it was red or whether it was simply inoperative. In making his finding, his Honour rejected the contrary evidence of the locomotive driver on this point on the basis that his, the driver's, observation of the green railway signal was "but [a] reconstruction unsupported by his recollection of having seen the state of the signal".

No doubt in isolation that is a finding which it would be difficult to challenge. But even if one were to accept that the driver's evidence is to be rejected, it is difficult to see any positive basis for the finding which his Honour made. Three things can be noted about this. The first is that as a general proposition the rejection of the defence's witnesses does not provide positive support for the opposite case, namely that the signal was either red or not functioning at the time. Even if his Honour was correctly of the opinion that the driver had reconstructed this aspect of his evidence, that does not of itself support a finding that the signal was other than green. It merely leaves the evidence in a neutral state. Secondly, in the absence of a specific finding as to the cause of the failure of the crossing lights, it seems impossible for his Honour to have concluded on any sound basis that the railway signal was not green. Indeed, one of the three explanations advanced for the signal failure, namely the stuck relay theory, would have produced the results that a green signal would have been apparent to the driver of the locomotive even though the crossing lights would have been inoperative.

The last point which can be made on this topic is that

the finding was made in the context of the trial Judge's
misapprehension that the driver's evidence as to the colour
of the signal was relevantly unsupported. But that is not
the case. The assistant locomotive driver also gave
evidence that the railway signal was green at the time. The
trial Judge sought to explain this observation away on the
basis that it was made at a stage when the locomotive had
triggered the circuit by passing onto track "B" but this, it
was conceded, was erroneous. The track "B" circuit did not
trigger the railway signal and in any event by that stage
the locomotive would have well passed it. In response to
this admitted error it was said that his Honour's finding
that the locomotive passed through a railway signal that was
other than green involved an implicit rejection of the
assistant driver's evidence. But it is difficult to see how
this is so. On the contrary, by not explicitly rejecting
his evidence but rather by seeking to explain it, it could
equally be said that the trial Judge accepted the accuracy
of his observation but in view of his misapprehension about
its timing, discounted it without giving any real
consideration to what the position might have otherwise have
been. He simply did not turn his mind to the question of
whether the driver's evidence was supported either by the
testimony of his assistant or by the expert evidence as to
the likely cause of the system's failure. So viewed it
cannot in my opinion be said either that his Honour rejected
the assistant driver's evidence or that he would inevitably
have made the same finding if cognisant of the nature of the
misapprehension he was acting under. In any event, again a
rejection of the assistant driver's evidence does not
support the contrary proposition and in the absence of any
indication that the trial Judge has turned his mind to the
cause of the signal failure, no basis for the finding
remains. This is a case where clearly the trial Judge has
failed to use his advantage in making the finding which he
did: Abalos v. Australian Postal Commission (1990) 171
C.L.R. 167; Devries v. Australian National Railways
Commission (1992-93) 177 C.L.R. 472. Accordingly, the
second basis on which it was sought to support the judgment
also fails and it follows that the judgment below cannot
stand.

New Trial

It was urged on behalf of the respondents that should this Court come to the conclusion that the judgment below be set aside a new trial should be ordered. But a new trial is not a matter of right, it will only be ordered if this Court thinks it appropriate: O. 70 r. 12 Rules of the Supreme Court. In particular it is not a procedure designed to allow a party the opportunity to re-litigate a defective case.

The inability of the first respondents to point towards any probable cause of the failure of the flashing lights and, more importantly, their complete failure to launch any attack on the maintenance system employed by the appellant, a system described by one of the respondents' own experts as "very efficient", persuades me that the appellant should have judgment entered in its favour. In short it is the complete failure of the respondents to point towards any step which the appellant could reasonably have taken to avoid or minimise the risk of such an occurrence which is fatal to their submission.

I should add that in light of these conclusions it is unnecessary to go further and consider the issue of whether, had the judgment been supportable or a new trial ordered, any finding of contributory negligence ought to have been made. I indicate, however, that my preliminary view is that particularly in light of the technical evidence as to the safe braking distance of the first respondents' vehicle had it been travelling at the speed alleged, it would be difficult to escape the conclusion that the second respondent was either driving at an excessive speed or had failed to keep any or any proper lookout.

In the result I would allow the appeal, set aside the judgment and orders made below and in lieu order that judgment be entered for the appellant against the first respondents with costs both here and below to be taxed. It was the appellant's decision to join the second respondent as third party, when it was admitted in the first respondents' own pleading that the second respondent was their employee. Any finding of contributory negligence against the second respondent would therefore have attached to the first respondents and gone in reduction of any award of damages in their favour. It is difficult to see what could have been gained by issuing the third party notice.

In any event in view of the judgment in favour of the appellant there is no liability which the second respondent could contribute towards or provide an indemnity in respect of. Accordingly, the third party proceedings are dismissed.

The appellant is to pay the second respondent's costs both

here and below to be taxed.