Vetter v Lake Macquarie City Council - [2001] HCA 12

HIGH COURT OF AUSTRALIA
GLEESON CJ,
GUMMOW, KIRBY, HAYNE AND CALLINAN JJ
NATALIE JANE VETTER   APPELLANT
AND
LAKE MACQUARIE CITY COUNCIL  RESPONDENT
Vetter v Lake Macquarie City Council [2001] HCA 12
8 March 2001
S27/2000
ORDER
1.        Appeal allowed with costs.
2.Set aside orders 1, 2, 4 and 5 of the orders of the Court of Appeal of the Supreme Court of New South Wales made on 16 April 1999.
3.Order that the matter be remitted to the Compensation Court of New South Wales for retrial limited to the question of fault under s 10(1A) of the Workers Compensation Act 1987 (NSW) and that otherwise the appeal to the Court of Appeal be dismissed with costs.
4.Costs of the proceedings in the Compensation Court to be determined by the judge at retrial.

On appeal from the Supreme Court of New South Wales
Representation:
R C Kenzie QC with I D M Roberts for the appellant (instructed by Hunt & Hunt)
D F Jackson QC with A G Bell for the respondent (instructed by Palmieri Lawyers)
Notice: This copy of the Court's Reasons for Judgment is subject to formal revision prior to publication in the Commonwealth Law Reports.

CATCHWORDS
Vetter v Lake Macquarie City Council
Workers' compensation – Journey – Workers Compensation Act 1987 (NSW) – Compensation for worker injured on journey from work to place of abode – Worker travelling home after stopping for a meal with a relative – Whether deviation or interruption to the journey.
Workers' compensation – Practice and procedure – Compensation Court Act 1984 (NSW) – Appeal to Court of Appeal on a point of law only – Powers of the Court of Appeal.
Words and phrases – "journey" – "direct route" – "fault" – "place of abode".
Workers' Compensation Act 1926 (NSW), s 7.
Workers Compensation Act 1987 (NSW), s 10.
Compensation Court Act 1984 (NSW), s 32.

GLEESON CJ, GUMMOW AND CALLINAN JJ. This appeal requires the Court to construe sections of the Workers Compensation Act 1987 (NSW) ("the Act") and of the Compensation Court Act 1984 (NSW) relating to the ambit of a worker's journey from her place of work to her home and the nature of an appeal from the Compensation Court to the Court of Appeal.
Case History
On 1 June 1993 the appellant was working for the respondent as an accounts clerk. Her place of residence was 38.9 kilometres from her workplace, a journey by car, her usual means of travel, of about 45 minutes or so. The appellant had adopted a practice of calling upon her grandmother fortnightly after leaving work and before travelling to her place of residence. The practice was a regular one from which she departed very rarely, if at all. In order to call upon her grandmother the appellant was obliged to travel about 19 additional kilometres further than she would if she were to travel directly to her home.
On 1 June 1993 she undertook one of such visits to her grandmother at Belmont. After they ate a meal together she left her grandmother's residence and set out for home. It had been raining earlier in the day and it rained after she left her grandmother's house. At about 7:15 pm her car left the Pacific Highway and came into collision with an unattended truck, parked on the verge of the highway. The appellant suffered injuries including a severe head injury which caused brain damage and which has deprived her of any memory of the events leading up to, and the collision itself.
The appellant sought compensation for permanent injuries and consequential deprivations in the Compensation Court of New South Wales.
In order for the appellant to succeed in her claim she had to bring herself within s 10 of the Act which has since been amended but then provided as follows:
"Journey claims
(1)A personal injury received by a worker on any journey to which this section applies is, for the purposes of this Act, an injury arising out of or in the course of employment, and compensation is payable accordingly.
(1A)Subsection (1) does not apply if the personal injury was caused, partly or wholly, by the fault of the worker.
(1B)A personal injury received by a worker is to be taken to have been caused by the fault of the worker if the worker was at the time under the influence of alcohol or other drug (within the meaning of the Traffic Act 1909), unless the alcohol or other drug did not contribute in any way to the injury or was not consumed or taken voluntarily.
(1C) If the risk of injury on a daily or other periodic journey to which this section applies, compared with the risk of injury on the worker's normal journey, is materially increased for a reason connected with the worker's employment (including the distance travelled, the time of day or night, the method of travel or the route of the journey), subsection (1) is not excluded merely because the injury was caused by the fault of the worker.
(1D)Subsection (1) does not apply if the personal injury resulted from the medical or other condition of the worker and the journey did not cause or contribute to the injury.
(2) Subsection (1) does not apply if:
(a)the injury was received during or after any interruption of, or deviation from, any such journey; and
(b)the interruption or deviation was made for a reason unconnected with the worker's employment or the purpose of the journey,
unless, in the circumstances of the case, the risk of injury was not materially increased because of the interruption or deviation.
(3)The journeys to which this section applies are as follows:
(a)the daily or other periodic journeys between the worker's place of abode and place of employment;
(b)the daily or other periodic journeys between the worker's place of abode, or place of employment, and any educational institution which the worker is required by the terms of the worker's employment, or is expected by the worker's employer, to attend;
…
(4)For the purposes of this section, a journey from a worker's place of abode commences at, and a journey to a worker's place of abode ends at, the boundary of the land on which the place of abode is situated.
…
(5A)Nothing in this section prevents the payment of compensation for any personal injury which, apart from this section, is an injury within the meaning of this Act.
(6) In this section:
'fault' includes:
(a)negligence or other tort; and
(b) any failure to take reasonable care for the worker's own safety;
…
'place of abode' includes:
(a)the place where the worker has spent the night preceding a journey and from which the worker is journeying; and
(b)the place to which the worker is journeying with the intention of there spending the night following a journey."
The effect of the section is relevantly this. The worker will be entitled to compensation if he or she suffers injury on a journey between the workplace and home. A daily or other periodic journey between those places will be a journey to which the section applies. Subject to some matters referred to in sub-s (1C) that entitlement will be lost if the worker's injuries are caused wholly or partly by his or her negligence, other tort, or any failure to take reasonable care for his or her own safety. The sustaining of an injury during or after any interruption of, or deviation from a journey which would otherwise be a relevant journey will also disentitle a worker to compensation unless the risk of injury not be materially increased because of the interruption or deviation.
Accordingly the appellant could recover here, if, absent fault on her part, she was injured on a periodic journey between her workplace and home, or, if she was injured on such a journey after a deviation from, or an interruption to it, and, in the latter case, the risk of injury was not materially increased because of the deviation or interruption.
The appellant argued at the trial that she was entitled to succeed on the basis that she was injured after an interruption to a relevant periodic journey, or after or during a deviation from it, in circumstances in which there was no material increase in risk. The trial judge summarised the evidence with respect to the frequency of the appellant's visits to her grandmother's house, and held, in respect of the evening of the accident, that although she had visited her grandmother and stayed for a time there, the appellant was in the course of a single journey when she was injured: that she was on an "other periodic journey".
His Honour then turned to a consideration of events after the appellant's departure from her grandmother's residence on the basis that her stay there was an interruption to the journey home. He referred to the facts that, by this time darkness had fallen, it was raining, and the peak hour rush had abated. His Honour found that the rain was constant, and that, despite some evidence to the contrary, it was not heavy and the conditions were not particularly windy ones. He discussed the respondent's expert's evidence that the risk was greater while travelling at night in the rain, than in travelling in daylight in the rain. That expert had also relied on the known facts of the accident, as a basis for saying that in all probability the wet and dark conditions prevailing at the time, caused or contributed to it. The appellant's expert's evidence was that the reduction in the volume of traffic after peak hours, in effect, offset any increased risk of travel by car at night time in rain.
The trial judge preferred the evidence of the appellant's expert for a number of cogent reasons which are set out in the judgment. Both experts had produced statistical evidence of increases in risk by reason of rain and darkness. His Honour demonstrated in his reasons for judgment some fallacies in the assumptions and statistics relied upon by the respondent's expert. He accepted the evidence of the appellant's expert that any increase in risk in driving after dark in the rain, on the roads in question, in comparison with the risk of driving upon them in daylight, during rain, was very slight, and that the risk in either case, was, in real terms very low. His Honour accordingly held that there was no material increase in the risk of injury caused by the appellant's visit to her grandmother's house, on the basis that it was an interruption to her journey home.
His Honour also dealt with the other basis on which the appellant had argued her case, that the journey to the grandmother's house by the appellant was a deviation from her journey home and the time spent with her grandmother was an interruption to that journey. His Honour canvassed again much of the expert evidence and because he preferred the appellant's expert to the respondent's, held that there was no material increase in risk as a result of either the interruption or the deviation that he had identified. The appellant could recover therefore, subject to the question of fault, on either or both of the bases upon which she presented her claim.
The last matter therefore that his Honour had to decide was whether the appellant was disentitled to receive compensation because her injury was caused partly or wholly by her fault in terms of s 10(1A) and s 10(6) of the Act.
After reviewing the limited evidence that was available as to how the collision occurred, his Honour said that the possibilities suggested by the respondent amounted to no more than speculation: there was no evidence to show that a loss of control by the appellant of her vehicle resulted from negligence as opposed to inadvertence or error of judgment. His Honour accordingly held that the respondent had failed to discharge the onus that lay upon it of proving that the injuries were caused or contributed to by her own fault. His Honour therefore allowed the appellant's claim and made various orders and awards of compensation.
The respondent appealed to the Court of Appeal of New South Wales (Priestley, Handley and Powell JJA). The appeal was brought pursuant to s 32 of the Compensation Court Act which confers a right of appeal upon a party aggrieved in point of law[1]. It is unnecessary to decide in this case, whether once an error on a point of law is identified, the Court of Appeal is confined to that point only and has no power to decide any other matter[2] because all necessary factual questions were not addressed in the Compensation Court and the case will need to be remitted to the Compensation Court for complete determination.
[1]"Appeal to Court of Appeal from Judge on question of law
(1)If a party to any proceedings before the Court constituted by a Judge is aggrieved by an award of the Judge in point of law or on a question as to the admission or rejection of evidence, that party may appeal to the Court of Appeal.
(2)The Court of Appeal may, on the hearing of any appeal under this section, remit the matter to the Compensation Court for determination by the Compensation Court in accordance with any decision of the Court of Appeal and may make such other order in relation to the appeal as the Court of Appeal sees fit.
(3)A decision of the Court of Appeal on an appeal under this section is binding on the Compensation Court and on all the parties to the proceedings in respect of which the appeal was made.
(4)The following appeals under this section may be made only by leave of the Court of Appeal:
(a)an appeal from an interlocutory decision,
(b)an appeal from a decision as to costs only,
(c) an appeal from a final decision or award, other than an appeal that involves (directly or indirectly) a claim for, or a question relating to, an amount of $20,000 or more."
[2]cf Krew v Commissioner of Taxation(Cth) (1971) 45 ALJR 324 at 325-326.
Handley JA (with whom Powell JA agreed on the point on which the respondent succeeded[3]) having posed for himself the question whether the appellant had made one or two journeys[4] said that the trial judge was in error in holding that when the appellant was injured she was engaged in a single journey, or nothing more than a roundabout journey to her home[5]. The factors which his Honour thought decisive were that the direction of travel to the appellant's grandmother's house was in an opposite direction to the appellant's home, and the length of her stay at her grandmother's house[6]. Handley JA said that the repetition of the pattern of travel did not support a finding of a single journey any more than would a finding that once a fortnight a worker went to watch a sporting fixture, or to see a film, or to play basketball[7]. It was his Honour's opinion that the journey on which the appellant was embarked when she was injured, was not, as the Act required, a single journey between her workplace and home, and for that reason her claim should fail, even though he was of the opinion that the appellant's journey to her grandmother's residence was a periodic one[8].
[3](1999) 18 NSWCCR 34 at 50 [55]-[56].
[4](1999) 18 NSWCCR 34 at 37-38 [7]-[8].
[5](1999) 18 NSWCCR 34 at 40 [17].
[6](1999) 18 NSWCCR 34 at 40-41 [17]-[18].
[7](1999) 18 NSWCCR 34 at 41 [18].
[8](1999) 18 NSWCCR 34 at 41 [20].
His Honour did however consider the question of the materiality of any increase in risk, if he were to assume that the appellant was injured during or after a relevant deviation from, or interruption in the journey, a question which, his Honour said, was within broad legal limits "a question of fact"[9]. Handley JA said that the only question of law was whether an increase in risk which appears significant in percentage terms, must be a material increase even if it is not significant in absolute terms. He concluded that that question had to be answered in favour of the appellant[10].
[9](1999) 18 NSWCCR 34 at 43 [28].
[10](1999) 18 NSWCCR 34 at 43 [28].
Handley JA then discussed the issue of fault. He said that the evidence in the present case did not exclude any cause of the accident which would have established negligence on the part of the appellant, that the appellant's loss of control of her vehicle on the curve, its movement off the roadway, and collision with the truck could constitute evidence of fault[11]. In consequence, Handley JA was of the opinion that the trial judge fell into legal error in holding that the evidence raised possibilities as to how the accident might have happened but did not provide any foundation for a finding of fault on the part of the appellant[12]. There was a further error, his Honour said, on the part of the primary judge in placing an onus on the employer to establish the precise cause of the accident. Although the trial judge's error on the question of fault would of itself only lead to an order for a new trial, the conclusion of Handley JA with respect to the nature of the appellant's journey, against the appellant, required the entry of judgment for the respondent.
[11](1999) 18 NSWCCR 34 at 44 [34].
[12](1999) 18 NSWCCR 34 at 46 [37].
Priestley JA, who also regarded the relevant question as being whether there had been one or more than one journey[13], differed from Handley and Powell JJA in his characterisation of the appellant's travel. His Honour identified the question of law which was raised in the Court of Appeal as being whether the facts found by the trial court could support the legal description given to them by the trial court of one journey[14]. It was open, Priestley JA held, for an objective bystander, using ordinary language, to describe the appellant's travel on the day of the appellant's accident as "going to see her grandmother on her way home". Accordingly, in choosing to prefer a description of that kind to another description, such as, "making a trip to her grandmother's and then making a second trip to home from there", the trial judge did not fall into error[15].
[13](1999) 18 NSWCCR 34 at 48 [44].
[14](1999) 18 NSWCCR 34 at 48 [44].
[15](1999) 18 NSWCCR 34 at 48 [45].
Although Priestley JA agreed with the other members of the Court on the other matters that they decided, his different conclusion as to the nature of the appellant's travel would lead to an order for a new trial limited to the questions of fault.
In the event the appeal to the Court of Appeal was allowed and judgment was entered for the respondent. The orders that the Court of Appeal made were as follows:
"1. The Appeal be allowed.
2. The Respondent to pay Appellant's costs of these proceedings.
3.The award in favour of the applicant in the Compensation Court proceedings be set aside.
4.In lieu thereof there be an award for the respondent in respect of those proceedings.
5.The Respondent in these proceedings to have a certificate under the Suitors Fund Act."
The Appeal to this Court
The appellant appealed to this Court on the following grounds:
"The Court of Appeal:-
(a)erred in concluding that the question of whether or not the appellant was injured in the course of a daily or other periodic journey between her place of employment and her place of abode was a question of law;
(b)erred in concluding that the appellant was injured in the course of a journey which was not a daily or other periodic journey between her place of employment and her place of abode;
(c)erred in concluding that it was not open as a matter of law to find that the injury to the appellant occurred during a single journey from her place of employment to her place of abode".
The respondent has filed a notice of contention that the Court of Appeal ought to have found that the appellant's injuries were caused wholly or partly by the fault of the appellant within the meaning of s 10(1A) of the Act as it read at the date that the appellant was injured.

The appellant's first submission is that the majority in the Court of Appeal erred in determining that the only reasonable conclusion was that the appellant suffered her injuries on a journey between her grandmother's residence and her own residence, rather than while she was on a periodic journey between her place of employment and her residence, as was held by the trial judge. The appellant further submits that the finding on this issue was a finding on a question of fact only and not on a point of law which is the only matter that may be the subject of an appeal to the Court of Appeal pursuant to s 32 of the Compensation Court Act.
Whether facts as found answer a statutory description or satisfy statutory criteria will very frequently be exclusively a question of law. To put the matter another way, indeed, as it was put by Priestley JA in his judgment, whether the facts found by the trial court can support the legal description given to them by the trial court is a question of law[16]. However, not all questions involving mixed questions of law and fact are, or need to be susceptible of one correct answer only. Not infrequently, informed and experienced lawyers will apply different descriptions to a factual situation. That is why the test whether legal criteria have been met has been expressed in language of the kind used by Jordan CJ in The Australian Gas Light Co v Valuer-General[17]:
"[I]f the facts inferred … from the evidence … are necessarily within the description of a word or phrase in a statute or necessarily outside that description, a contrary decision is wrong in law".
[16](1999) 18 NSWCCR 34 at 48 [44]-[45].
[17](1940) 40 SR (NSW) 126 at 138.
In his speech in Edwards (Inspector of Taxes) v Bairstow[18] Lord Radcliffe identified an error of law as arising if "the true and only reasonable conclusion contradicts the determination". Mason J (with whom Gibbs, Stephen, Murphy and Aickin JJ agreed) discussed the matter comprehensively and stated the law on this topic in this country as follows in Hope v Bathurst City Council[19]:
"Many authorities can be found to sustain the proposition that the question whether facts fully found fall within the provisions of a statutory enactment properly construed is a question of law. One example is the judgment of Fullagar J in Hayes v Federal Commissioner of Taxation[20], where his Honour quoted the comment of Lord Parker of Waddington in Farmer v Cotton's Trustees[21], which was adopted by Latham CJ in Commissioner of Taxation v Miller[22], that where all the material facts are fully found, and the only question is whether the facts are such as to bring the case within the provisions properly construed of some statutory enactment, the question is one of law only. Fullagar J then said[23]:
'… this seems to me to be the only reasonable view. The distinction between the two classes of question is, I think, greatly simplified, if we bear in mind the distinction, so clearly drawn by Wigmore, between the factum probandum (the ultimate fact in issue) and facta probantia (the facts adduced to prove or disprove that ultimate fact). The "facts" referred to by Lord Parker … are the facta probantia. Where the factum probandum involves a term used in a statute, the question whether the accepted facta probantia establish that factum probandum will generally – so far as I can see, always – be a question of law.'
However, special considerations apply when we are confronted with a statute which on examination is found to use words according to their common understanding and the question is whether the facts as found fall within these words. Brutus v Cozens[24] was just such a case. The only question raised was whether the appellant's behaviour was 'insulting'. As it was not unreasonable to hold that his behaviour was insulting, the question was one of fact."
[18][1956] AC 14 at 36.
[19](1980) 144 CLR 1 at 7.
[20](1956) 96 CLR 47 at 51.
[21][1915] AC 922 at 932.
[22](1946) 73 CLR 93 at 97.
[23](1956) 96 CLR 47 at 51.
[24][1973] AC 854.
Earlier in Williams v Bill Williams Pty Ltd[25], Mason JA had observed[26]:
"[I]t may happen that the tribunal at first instance is confronted with the task of applying the statutory expression to primary facts in such circumstances that it is reasonably possible to arrive at different conclusions, the question being largely one of degree upon which different minds may take different views. Here, again, it is not possible to conclude that the decision appealed from is erroneous in point of law.
The principle has been enunciated that, if different conclusions are reasonably possible, the determination of which is the correct conclusion is a question of fact[27]."
[25][1971] 1 NSWLR 547.
[26][1971] 1 NSWLR 547 at 557.
[27]NSW Associated Blue-Metal Quarries Ltd v Federal Commissioner of Taxation (1956) 94 CLR 509 at 512; Australian Iron & Steel Pty Ltd v Luna (1969) 44 ALJR 52; Hall v Yellow Cabs of Australia Ltd (1970) 92 WN (NSW) 426.
In Hope v Bathurst City Council[28], Mason J pointed out that when it is necessary to engage in a process of construction of the meaning of a word (or phrase) in a statute a question of law will be involved, but that the question may be a mixed one of fact and law. His Honour's reasons make it clear that a question exclusively of law arises, as the respondent sought to argue was the position in this case, if, on the facts found only one conclusion is open[29].
[28](1980) 144 CLR 1 at 8.
[29](1980) 144 CLR 1 at 9. See also Collector of Customs v Agfa-Gevaert Ltd (1996) 186 CLR 389.
Priestley JA was of the opinion that this was not an appeal in which there arose a question exclusively of law, and that it was possible to characterise the appellant's visit to her grandmother as "going to see her grandmother on her way home" just as easily as "making a trip to her grandmother's and then making a second trip to home from there"[30]. We agree with his Honour, that it is possible to characterise the appellant's travel in the former way but not that the matter is to be decided by asking whether there were one or two journeys.
[30](1999) 18 NSWCCR 34 at 48 [45].
The appellant's real destination was her own residence. She always intended to spend the night at her own residence. As Mahoney JA said in Minchinton v Homfray[31]:
"The term 'journey', used in the relevant sense, has an indeterminate meaning or meanings. In the Shorter Oxford English Dictionary, the relevant meanings are:
'a day's travel; the distance travelled in a day or a specified number of days …; a spell of going or travelling, viewed as a distinct whole; an excursion or expedition to some distance; a round of travel ...'.
Having regard to the statutory reference to 'interruption or duration', the relevant meaning for present purposes is, I think: 'a spell of going or travelling, viewed as a distinct whole'."
There is no obligation upon a worker to take the shortest and most direct route from the worker's place of work to the worker's abode so long as the journey can be said to be a journey between the worker's place of abode and place of employment. And there is no reason why a worker might not, within the statutory meaning of a journey, choose a route, albeit an indirect and longer one, which may enable the worker to achieve a purpose in addition to the purpose of reaching the worker's residence in order to spend the interval between ceasing and recommencing work, again provided that the journey still has a character of a journey between his or her place of work and place of abode, and there is no material increase in risk during or after any deviation or interruption. That is what the Act requires. Any question whether that requirement has been satisfied is not to be answered by posing and answering a different question altogether and of the kind posed by the Court of Appeal, was the appellant engaged in one or more journeys[32].
[31](1994) 10 NSWCCR 778 at 785.
[32](1999) 18 NSWCCR 34 at 37-41 [6]-[19] per Handley JA, 48-49 [44]-[50] per Priestley JA.
The decision of this Court in Young v Commissioner for Railways[33] upon earlier legislation, the Workers' Compensation Act 1926 (NSW), was relied upon by the respondent. It was an appeal in a case which had fallen to be decided on its facts and holds in terms of principle no more than that "the journey must commence as a journey between the place of abode and the place of employment"[34]. It was, unlike this case, one in which there was no evidence upon which the Court could act to find that the journey there was of such a kind. Dixon CJ posed the issue for decision as follows[35]:
"Was the Workers' Compensation Commission justified in saying, on the facts as I have stated them, that when he left the works he was then commencing the journey home? In my opinion, they were not. All that was known was that he was proceeding upon some business that he had to transact, that he was attending to some business. Where, or what it was, in which direction it led him, is just simply unknown. And it seems to me impossible to say that when he left the works or factory he commenced a journey which was a journey to his home."
[33][1961] ALR 258.
[34][1961] ALR 258 at 260 per Dixon CJ.
[35][1961] ALR 258 at 260.
A journey may sometimes aptly be described in more than one way. This is perhaps such a case, but because the trial judge's description or characterisation was one reasonably available description, the majority in the Court of Appeal erred in insisting on the different description that they did.
It follows in our opinion that the primary judge made no error of law in deciding that the appellant was undertaking a journey, properly to be described as a periodic journey between her place of employment and place of abode within the meaning of s 10(3)(a) of the Act. There was no issue as to the periodicity of the journey and the appellant has concurrent findings in her favour on this matter in both the trial court and the Court of Appeal. However, that she was injured after an interruption to, or after or during a deviation from such a journey, cannot be disputed. As to this the appellant has concurrent findings of fact in her favour that in the circumstances of the case the risk of injury was not materially increased because of the interruption or deviation, and, subject only to questions of fault, she would be entitled to succeed.
It remains to deal with the respondent's notice of contention that the appellant's injuries were caused or contributed to by fault on her part. Priestley and Handley JJA in the Court of Appeal were of the opinion that the trial judge was in error in point of law in the way in which he dealt with this aspect of the case. But the respondent wants this Court to go further than Priestley and Handley JJA were prepared to go and to make an affirmative finding of fault now.
We do not think that the Court of Appeal or this Court should, or indeed is able to, make such an affirmative finding in this case. Before saying why we think this to be so we would refer to Davis v Bunn[36] in which Dixon J said of a motor vehicle accident:
"[U]nless and until the cause of the vehicle's change of direction was explained, I think mere proof that it suddenly swerved from one side of the road to the other and hit the plaintiff's stationary car would constitute sufficient evidence".
Dixon J accepted that such an occurrence could be consistent with one or more causes unrelated to the driver's negligence, such as fainting by the driver or mechanical failure, but that such unavoidable events are sufficiently unusual to raise a probability of an absence of due care by the driver[37]:
"In the absence of all explanation, the probability would be high enough to justify an inference [against the driver] … a presumption of fact would arise and its strength would be a matter for the jury to estimate, in whose province it would be to draw or refuse to draw the inference."
[36](1936) 56 CLR 246 at 260.
[37](1936) 56 CLR 246 at 260. See also Government Insurance Office of NSW v Fredrichberg (1968) 118 CLR 403 at 414 per Barwick CJ.
Reference should also be made to what Barwick CJ said in Government Insurance Office of NSW v Fredrichberg[38]:
"[The jury] must yet be satisfied in their minds that more probably than not the defendant was in fact negligent and that his negligence, even though they cannot identify the particular negligent act or omission, caused the plaintiff's injuries".
[38](1968) 118 CLR 403 at 415.
A trial judge in a case of this kind is, with respect to questions of fact in much the same position as a jury. The trial judge here was in error as Priestley and Handley JJA held him to be. The observations that were made in the two jury cases which we have quoted apply to the exercise in which the trial judge was engaged in this case, but in respect of which he erroneously said that the possibility of negligence on the part of the appellant would be speculative only. The respondent accepts that the onus is upon it to establish fault. It may be that on a full consideration of all of the relevant facts, notwithstanding the apparent absence so far of a very convincing case to meet a case of fault capable of being inferred against her, a court might still take the view that the respondent's onus has not been discharged. As long ago as 1774 Lord Mansfield said[39] that all evidence is to be weighed according to the proof which it is in the power of one side to have produced and the power of the other to have contradicted. The evidence in this case will fall to be considered against the background of the appellant's impaired memory.
[39]Blatch v Archer (1774) 1 Cowp 63 at 65 [98 ER 969 at 970]. See also Weissensteiner v The Queen (1993) 178 CLR 217 at 226-227 per Mason CJ, Deane and Dawson JJ.
Neither the appellant nor the respondent has had the benefit of a trial at which the correct questions on the issue of fault were addressed and answered. All facts enabling the correct questions to be answered have not themselves therefore necessarily all been decided. The answers to those questions are bound to depend, among other things, upon impressions formed of witnesses, and preferences for one asserted set of facts over a contrary assertion, and upon the capacity of each side to adduce evidence. For these reasons there should be a new trial limited to the question of fault as proposed by Priestley JA in the Court of Appeal.
It is unnecessary, in view of the conclusion that we have reached, to express any opinion about the correctness or otherwise of the limited construction that the Court of Appeal of New South Wales (Beazley and Giles JJA, Davies A-JA) put upon s 32 of the Compensation Court Act in North Broken Hill Ltd v Tumes[40], which was referred to in argument, or of the possible application of s 75A of the Supreme Court Act 1970 (NSW) to appeals from the Compensation Court.
[40](1999) 18 NSWCCR 412.
We would allow the appeal to this Court with costs. We would set aside orders 1, 2, 4 and 5 of the orders of the Court of Appeal made on 16 April 1999, order that the appeal to the Court of Appeal be dismissed with costs and that the matter be remitted to the Compensation Court of New South Wales for a retrial limited to the question of fault under s 10(1A) of the Workers Compensation Act 1987 (NSW). The costs of further proceedings in the Compensation Court will be for the judge at the retrial.
KIRBY J. This appeal[41] concerns the meaning and application of the Workers Compensation Act 1987 (NSW) ("the Compensation Act") and the Compensation Court Act 1984 (NSW) ("the Court Act"). In issue are the entitlements to compensation of a worker, injured on a journey, and the powers of the Court of Appeal deciding an appeal from an award entered by the primary judge in favour of the worker in the Compensation Court of New South Wales ("the Compensation Court").
[41]From a judgment entered by the New South Wales Court of Appeal: Lake Macquarie City Council v Vetter (1999) 18 NSWCCR 34 ("Vetter") per Handley JA (Powell JA concurring), Priestley JA dissenting.
The facts
The appellant, Ms Natalie Vetter ("the worker"), was employed by the respondent, Lake Macquarie City Council ("the employer"). Her place of employment was at Boolaroo, near Speers Point, in New South Wales. Her place of abode was at Nelsons Plains, which is north of Newcastle, where she lived with her parents.
It was the worker's practice to drive her car to and from work. Her usual route home after work took her northwards, a distance of 38.9 km. It ordinarily required a journey of about 45 minutes. However, as the primary judge found, it was also the worker's practice, each fortnight, to visit her grandmother and to have dinner with her at her home which was in Belmont. This the worker did on 1 June 1993, arriving at her grandmother's home at about 5.00 pm. She had dinner with her grandmother and, at about 7.00 pm, departed for her place of abode. The complete journey, taking in the worker's visit to the grandmother's home, covered a total distance of 58.6 kms. It thus added 19.7 kms to the journey, with a corresponding increase in the time of travel. It necessitated first travelling south-east to Belmont, instead of north on the direct journey to Nelsons Plains. In effect, it involved an almost semicircular diversion through the outlying districts of Newcastle for a private purpose of no concern to the employer.
At about 7.15 pm, while proceeding north in the direction of her home, the worker's vehicle left the road and collided with a parked truck. The worker, who was alone in her vehicle, was severely injured. Because she suffered brain damage, she was unable to recollect anything of the collision or why it had occurred. There were no witnesses. All that was known was that the worker's car had crashed into the back of the stationary vehicle. There were no marks on the roadway to explain what had happened. There was no evidence of the speed of the worker's car at the time of the collision. Nor was there evidence of any conduct of another vehicle that had played a part in the accident.
The worker made a claim under the Compensation Act. She did not suggest that her injuries were received whilst engaged on a journey required for her employment[42]. Her claim was for benefits provided for workers injured in defined circumstances on specified journeys[43] between identified places to and from work[44]. Her case, put simply, was that she had been injured on a journey home from work. The employer declined to pay compensation. It justified its refusal on the grounds that: (1) the worker was not injured on a "journey", as defined; (2) if she was, such "journey" was not a "daily or other periodic [journey]"[45] as provided by the Compensation Act; and (3) in any case, there had been a substantial "interruption of, [and] deviation from" any compensable journey, which was made for a reason unconnected with the worker's employment or the purpose of the compensable journey, which had materially increased the risk of injury[46]. If all of these defences failed, the employer finally argued that (4) the injuries suffered by the worker were caused "partly or wholly" by fault on her part[47].
[42]cf Bull v Schweppes (Australia) Pty Ltd [1960] WCR (NSW) 67.
[43]Compensation Act, s 10(1), with cross-reference to s 10(3).
[44]Relevantly between the "worker's place of abode and place of employment": Compensation Act, s 10(3)(a).
[45]Compensation Act, s 10(3)(a).
[46]Compensation Act, s 10(2)(a).
[47]Compensation Act, s 10(1A).
In the Compensation Court, the primary judge found that all of the foregoing defences failed. He entered the award in favour of the worker. There was no relevant dispute about the facts in the Court of Appeal or in this Court. The defence concerning a material increase in the risk of injury by reason of the interruption or deviation was not re‑agitated. The issues tendered for this Court's consideration stemmed from a challenge to the view adopted by the Court of Appeal concerning the journey provisions of the Compensation Act and about the particular disqualification for "fault of the worker"[48]. In the course of arguing the appeal, questions arose as to the meaning of the Court Act in relation to the jurisdiction and powers of the Court of Appeal in disposing of such an appeal.
[48]Compensation Act, s 10(1A).
The applicable legislation

The applicable provisions of s 10 of the Compensation Act which deal with journey injuries are stated elsewhere[49]. Provisions such as those in s 10(3), entitling workers to compensation for personal injuries received on "daily or other periodic journeys between the worker's place of abode and place of employment", date back to the Workers' Compensation Act 1926 (NSW) ("the 1926 Act")[50].
[49]Reasons of Gleeson CJ, Gummow and Callinan JJ at [5].
[50]1926 Act, s 7(1)(b).
The 1926 Act, as originally enacted, provided that an injury was compensable if it was received on a journey of defined regularity between defined places. Under that Act, the worker was disqualified from recovery if the injury was "received during any substantial interruption of, or deviation from the journey" for reasons "unconnected with the worker's employment"[51]. Such provisions had a clear social purpose. They imposed on an employer compensation liability in respect of the journeys of workers to and from work, but only as defined.
[51]1926 Act, s 7(1).
In the years since their first enactment, the journey provisions have changed many times. The original sub-section was eliminated by amending legislation in 1929[52]. However, an entitlement was restored in 1942. Various preconditions were then added, including one which required that the injury be received "without [the worker's] own default or wilful act"[53]. Although the termini of the journeys were successively expanded[54], the main contours of the provisions were substantially settled by the 1950s. Workers were entitled to compensation for injuries on journeys of specified kinds. They lost that entitlement where the injury was received during or after a substantial interruption or deviation; but only if such interruption or deviation materially increased the risk of injury. They also lost the entitlement if it could be shown that there was a relevant default or wilful act on the worker's part. Provisions of this kind passed into the present Compensation Act when it replaced the 1926 Act in 1987.
[52]Workers' Compensation (Amendment) Act 1929 (NSW), s 3. See Boulter, Workers' Compensation Practice in New South Wales (1966) at 107-108.
[53]Workers' Compensation Act and Workmen's Compensation (Broken Hill) Act (Amendment) Act 1942 (NSW), s 3(a). The provisions were originally restricted to apply for a period lasting six months after the termination of the Second World War: s 3(a). This limitation was repealed by the Workers' Compensation (Amendment) Act 1947 (NSW), s 2.
[54]See eg Workers' Compensation Act and Workmen's Compensation (Broken Hill) Act (Amendment) Act 1942 (NSW), s 3(a); Workers' Compensation (Amendment) Act 1951 (NSW), s 2(b).
Obviously, the journey provisions represented an expansion of the liability of employers to pay compensation for injuries occurring outside their control and immediate or direct economic interests. This fact made such provisions controversial to the minds of some earlier judges[55]. The provisions also had to operate in social circumstances which changed significantly after they were first enacted. The relevant changes included altered use of public and private transport, increased use of private motor vehicles to transport workers to and from work and different patterns of pre- and post-work activity.
[55]Landers v Dawson (1964) 110 CLR 644 at 652 per Windeyer J; cf Hobsons Pty Ltd v Thorne [1954] WCR (NSW) 59 at 59-60 per Street CJ; Mills, Workers Compensation (New South Wales) (1969) at 59.
In 1989, the disqualification for fault under the Compensation Act was enlarged by the insertion of s 10(1A)[56]. This sub-section disentitled the worker, who had received personal injury on an otherwise compensable journey, if that worker's injury "was caused, partly or wholly, by the fault of the worker". It eliminated notions of wilfulness and seriousness. It was to be given effect even if considered "Draconian"[57]. After the worker in the present case received her injuries, s 10(1A) of the Compensation Act was amended once again to substitute a more limited disqualification, which provides that s 10(1) does not apply where "the personal injury is attributable to the serious and wilful misconduct of the worker"[58]. This is something that could not apparently be proved against the present worker.
[56]Workers Compensation (Amendment) Act 1989 (NSW), Sched 1, item 1.
[57]WorkCover Authority (NSW) v Billpat Holdings Pty Ltd (1995) 11 NSWCCR 565 at 595. See Aardvark Security Services Pty Ltd v Ruszkowski unreported, Court of Appeal of New South Wales, 19 March 1993 at 6-7 of my own reasons.
[58]WorkCover Legislation Amendment Act 1996 (NSW), Sched 1, item 1.3.
The history of the journey provision in the Compensation Act therefore reflects changing attitudes and varying legislative objectives. It was agreed that the present appeal fell to be decided by the application of the provisions of s 10 as that section stood at the time the worker received her injuries[59].
[59]WorkCover Legislation Amendment Act 1996 (NSW), Sched 1, item 1.3, [5].
The decision of the Court of Appeal
In the Court of Appeal the judges divided. The leading judgment was given by Handley JA. His Honour rejected two of the arguments advanced by the employer for reasons with which Priestley JA concurred. The first of these (on an assumption contrary, so far as Handley JA was concerned, to his main conclusion about the proper characterisation of the "journey") concerned the primary judge's decision that the worker had received her injuries in the course of an "other periodic journey" within the meaning of the Compensation Act[60]. The second concerned the primary judge's conclusion that any such increase in the risk of injury, as had occurred on the journey, was slight and not a material increase of the disqualifying kind contemplated by the Act[61]. On each of these points, Handley JA affirmed that there was evidence to support the findings at first instance. There was therefore no error of law that would authorise intervention by the Court of Appeal.
[60]Vetter (1999) 18 NSWCCR 34 at 41-42 per Handley JA, 49 per Priestley JA.
[61]Vetter (1999) 18 NSWCCR 34 at 43 per Handley JA, 49 per Priestley JA.
On the final point argued before the Court of Appeal, Handley JA concluded that the primary judge had erred in law[62]. In this respect too, Priestley JA agreed[63]. The error concerned the application by the primary judge of an incorrect legal test for determining what the employer had to show in order to establish the "fault" for which s 10(1A) of the Compensation Act provided. In effect, Priestley and Handley JJA concluded that the primary judge had erred in law in finding that there was no evidence that the injury had been caused, partly or wholly, by the fault of the worker. In their Honours' conclusion, there was some evidence. It was not necessary in law for the employer to identify the particular negligent act or omission relied on[64]. The primary judge had erred in approaching the question of fault on the assumption that it was.
[62]Vetter (1999) 18 NSWCCR 34 at 47.
[63]Vetter (1999) 18 NSWCCR 34 at 49-50.
[64]Vetter (1999) 18 NSWCCR 34 at 46 per Handley JA referring to Government Insurance Office of NSW v Fredrichberg (1968) 118 CLR 403 at 415 per Barwick CJ.
The third member of the Court of Appeal, Powell JA, refrained from expressing an opinion on any of the foregoing conclusions[65]. However, Powell JA concurred with Handley JA in his determination of the primary issue argued for the employer in the Court of Appeal. This was described as the resolution of the question whether, in the circumstances of this case, "there were two journeys or one"[66].
[65]Vetter (1999) 18 NSWCCR 34 at 50.
[66]Vetter (1999) 18 NSWCCR 34 at 37 per Handley JA. The primary judge had asked the same question in his reasons: Vetter v Lake Macquarie City Council unreported, Compensation Court of New South Wales, 29 May 1998 at 8-9 per Judge Duck.
On this question, Handley JA decided that there were two journeys: one from the place of employment to the grandmother's home and the other from the grandmother's home to the place of abode. The worker was injured on the second journey which was not one between a specified origin and destination as provided by the Compensation Act. In finding that the worker was not entitled to compensation, Handley JA said[67]:
"Not without doubt, I have concluded that the true and only reasonable conclusion contradicts the finding of the trial Judge on this question and that it must be held that the respondent was injured on a journey between her grandmother's home and her place of abode which was outside the protection of s 10."
Because Powell JA agreed in this part of Handley JA's reasons[68], his agreement produced the Court of Appeal's decision upholding the employer's appeal. The award in the worker's favour was set aside. Her claim for compensation was dismissed.
[67]Vetter (1999) 18 NSWCCR 34 at 40-41.
[68]Vetter (1999) 18 NSWCCR 34 at 50.
In his minority opinion, Priestley JA also approached the primary question as depending on whether there was "[o]ne journey or two"[69]. Like Handley JA[70], and presumably in search of the same objective standard, Priestley JA decided the answer to that question by reference, not to his own classification of the facts but to how others would reasonably categorise the journey[71]. In his reasons, Priestley JA explained[72]:
"[T]he facts of the present case left it open to the trial Judge to categorise the respondent's journey in the way that he did. It seems to me that an objective bystander, using ordinary language, could describe the respondent's after work travel on the day of the accident as 'going to see her grandmother on her way home' just as easily as 'making a trip to her grandmother's and then making a second trip to home from there'. If I am right in this view, it seems to me to show that the case does not fall into the category of those cases where there is only one way of truly and reasonably categorising the description to be derived from the facts."
[69]Vetter (1999) 18 NSWCCR 34 at 48.
[70]Vetter (1999) 18 NSWCCR 34 at 38.
[71]This was also the approach of the primary judge, quoted by Handley JA in Vetter (1999) 18 NSWCCR 34 at 38.
[72]Vetter (1999) 18 NSWCCR 34 at 48 (original emphasis).
Priestley JA accordingly decided that, on this point, there was no error of law[73]. The Court of Appeal, on this point, was not entitled to substitute conclusions on the facts for those reached by the primary judge. His Honour applied what he took to be the approach of the Court of Appeal in an earlier case of roundabout journeying in The Old Spaghetti Factory v Oughtred[74]. In that decision, Samuels JA had cautioned against attempts to "erect a point of law when none in truth exists"[75]. In light of his conclusion of an error in the approach taken to the "fault" issue, Priestley JA favoured allowing the appeal and ordering a retrial in the Compensation Court. However, the order of the Court was that proposed by Handley and Powell JJA rejecting the worker's claim. It is from the judgment that followed that order that, by special leave, the appeal now comes to this Court.
[73]Vetter (1999) 18 NSWCCR 34 at 49.
[74][1975] WCR (NSW) 231.
[75][1975] WCR (NSW) 231 at 234.
The issues
As argued before this Court, the following issues require decision:
1.Whether the Court of Appeal, in the approach taken by Handley and Powell JJA, misconceived its own jurisdiction and effectively entered into an impermissible evaluation and redetermination of the facts.
2.Whether, as a matter of law, in the facts found, it was open to the primary judge to conclude that the worker, when injured, was on a "journey" within the meaning of the Compensation Act.
3.Whether, as a matter of law, in the facts found, it was open to the primary judge to find that there was "no evidence" to show that the worker's loss of control of her vehicle, which led to her injuries, resulted from negligence; that is was not caused, partly or wholly, by fault on her part.
4.Whether, if the matter were returned to the Compensation Court, that Court should be subject to orders limiting the retrial to the facts and issues litigated in the original hearing in that Court.
The worker eventually accepted that the best result that she could hope for (conceding the correctness of the opinion of Priestley and Handley JJA on the "fault" issue) was the order which Priestley JA had favoured, namely retrial in the Compensation Court. She contested the entitlement of the Court of Appeal, or this Court, to dismiss her claim on the basis of that error. She disputed the employer's submission that any orders should be made governing the actual conduct of the retrial in the Compensation Court.
The employer, by notice of contention, argued that, as a matter of law, the Court of Appeal was bound to conclude that the injuries sustained by the worker were within s 10(1A) of the Compensation Act and hence that no compensation was payable. In oral argument, counsel for the employer submitted that this Court, standing in the shoes of the Court of Appeal, should make such orders as the Court of Appeal ought to have made. In the event that it lost on all issues, the employer urged that it would have been open to the Court of Appeal, by order, to limit the evidence on which any retrial was ordered and that this Court should so provide.
The jurisdiction and powers of the Court of Appeal
The answer to the first issue depends on the jurisdiction and powers of the Court of Appeal in hearing an appeal from the Compensation Court. These are stated in the provisions of s 32 of the Court Act. These are set out in other reasons[76].
[76]Reasons of Gleeson CJ, Gummow and Callinan JJ, n 1.
The unusual terms in which s 32 of the Court Act is expressed were noticed during argument. In terms, s 32(1) is addressed to the precondition for the right of appeal of a party to proceedings before the Compensation Court. Aggrieved parties may appeal "in point of law or on a question as to the admission or rejection of evidence"[77]. The sub-section does not expressly state the powers of the Court of Appeal once an appeal is validly before that Court. Those powers are left to an inference, derived from the grounds which an appellant must make out to engage the jurisdiction of that Court.
[77]Court Act, s 32(1).
If a "point of law" can be shown to attract that Court's jurisdiction, is it then open to that Court, having jurisdiction, to reconsider the facts found by the Compensation Court and reach its own conclusions about such facts? In my view, that question cannot be avoided. An answer to it is required so that this Court can measure the approach adopted by the Court of Appeal against the criticisms of what that Court did, raised by the grounds of appeal, and so as to frame the relief that is appropriate to such conclusion.
Appeal provisions of the kind appearing in s 32 are not unknown in Australian legislation. The language of ss 32(1) and (2), strictly construed, would appear to give some support to a construction that, once jurisdiction is attracted by establishing a grievance "in point of law", broad powers are conferred on the Court of Appeal to make consequential orders. Certainly, the section does not expressly limit the powers of the Court of Appeal. Support for treating the Court's powers expansively would derive from the nature, function and general powers of the Court of Appeal itself and the regular functions of that Court, as part of the Supreme Court of a State, exercising its own jurisdiction and powers, to review fact-finding by trial courts[78]. Generally speaking, in default of restrictive legislation, where by statute jurisdiction is conferred, and functions are imposed, on a superior court, a broad view is taken of the powers which that court then enjoys[79].
[78]cf Supreme Court Act 1970 (NSW), s 75A.
[79]cf Ward v Williams (1955) 92 CLR 496 at 505; Jackson v Sterling Industries Ltd (1987) 162 CLR 612 at 639; Cardile v LED Builders Pty Ltd (1999) 198 CLR 380 at 393-394 [26] per Gaudron, McHugh, Gummow and Callinan JJ, 423-424 [110] of my own reasons.
Although the point concerning the jurisdiction and powers of the Court of Appeal was not raised in a ground of appeal to this Court, the importance of the point, in disposing of the appeal, was eventually recognised by both parties when questioned during argument. Because the worker eventually accepted, and the record reflected, that an error of law had been shown in the original disposition of the proceedings in the Compensation Court, the precondition to the jurisdiction of the Court of Appeal, provided by s 32(1) of the Court Act, was certainly established in this case. The employer was aggrieved in point of law by the award in favour of the worker. The employer was therefore entitled to appeal. Its appeal necessarily engaged the jurisdiction of the Court of Appeal. It enlivened that Court's powers.
Although, on one view, it might have been in the employer's interests to contend that the Court of Appeal was therefore entitled to review, for itself, the fact-finding of the Compensation Court on other issues (at least those such as whether the journey undertaken by the worker was, on the facts, a "journey" within the Compensation Act), that argument was declined by the employer. The employer accepted the conventional view. It was the view apparently adopted in the Court of Appeal in these proceedings. According to that view, not only is the jurisdiction of the Court of Appeal engaged only by a grievance in point of law (or in relation to the admission or rejection of evidence), its powers then arising are also controlled by such preconditions. Notwithstanding this concession for the employer, because the issue is squarely presented, because it controls the powers of the Court of Appeal, and because it is relevant to the disposition of the present appeal, it is in my view necessary for this Court to resolve the matter, happy as I would be to pass it by.
The resolution lies, I think, in the history of s 32 of the Court Act. Originally, under the 1926 Act, appeal to the Supreme Court of New South Wales from the predecessor of the Compensation Court (the Workers' Compensation Commission) was by way of stated case. By s 37(4) of the 1926 Act, it was provided that "[w]hen any question of law arises in any proceeding before the commission", the Commission was empowered "of its own motion [to] state a case for the decision of the Supreme Court thereon"[80]. For a very long time, the Workers' Compensation Commission had control over the identification of the point of law submitted to the Supreme Court for decision. It did so by the stated case which engaged the jurisdiction of the Supreme Court[81]. It was not until 1960[82] that a right of direct appeal was provided in terms which obviously represent the source for s 32(1) of the Court Act. In the context of the 1926 Act, and against more than 30 years of history, it was assumed that the "point of law" required to attract the jurisdiction to appeal also limited the subject matter of that jurisdiction and the powers of the Supreme Court once such jurisdiction was engaged.
[80]See Boulter, Workers' Compensation Practice in New South Wales (1966) at 301.
[81]cf Roberts v Jones [1928] WCR (NSW) 194: as a result of which s 37(4) was amended by the Workers' Compensation (Amendment) Act 1929 (NSW), s 8.
[82]Workers' Compensation (Further Amendment) Act 1960 (NSW), s 3(1)(f)(iii).

Following the establishment of the Compensation Court in 1984, the appeal provisions have followed a meandering course. At one stage, provision was introduced to permit appeals concerning points other than points of law and the admission or rejection of evidence. Such appeals lay as of right so long as the compensation involved was more than a specified sum[83]. Later still, however, the Court Act was again amended[84], presumably to stem the tide of appeals taken as of right involving no more than questions of fact.
[83]Compensation Court (Amendment) Act 1989 (NSW), Sched 1, item 12 amending s 32 of the Court Act.
[84]WorkCover Legislation Amendment Act 1995 (NSW), Sched 3, [13]; Courts Legislation Further Amendment Act 1997 (NSW), Sched 1, item 1.3, [3].
In the context of this legislative history, and despite the ungainly language, there can be little doubt that the purpose of Parliament was to limit both the jurisdiction and powers of the Court of Appeal to the determination of appeals on a point of law (or in relation to the admission or rejection of evidence). The alternative construction would be capricious, involving the need for jurisdiction to establish, relevantly, an error in point of law but thereafter allowing, and probably requiring, the Court of Appeal to exercise its powers to decide purely factual disputes. I say that this would be capricious because factual disputes of great consequence would not alone engage the jurisdiction of the Court of Appeal unless, accidentally, some point of law was shown. But on such a view of the sub-section, a trivial factual dispute could do so, so long as the precondition expressed in s 32(1) was otherwise established[85].
[85]The terms of s 32(4) of the Court Act, introduced by Sched 1, item 1.3, [3] of the Courts Legislation Further Amendment Act 1997 (NSW), are also relevant and support this view.
It follows that the better view of s 32(1) of the Court Act is that which has long been accepted. The jurisdiction of the Court of Appeal is engaged by, and its powers are then limited to, correcting decisions on a "point of law or on a question as to the admission or rejection of evidence". Only this construction gives the sub-section a workable meaning, taking into account its language and apparent purpose, and the history of the sub-section and its predecessors.
This view being established, it follows that the threshold objection of the worker to what was done by the majority in the Court of Appeal requires decision. Given that the jurisdiction and powers of that Court were limited to the correction, relevantly, of an error of law, the worker submitted that the majority of the Court of Appeal itself fell into error of law by an overexpansive view of what constituted an error of law in fact‑finding on the part of the primary judge.
The jurisdiction to correct only errors of law
As this Court observed in Collector of Customs v Agfa-Gevaert Ltd[86], the distinction between questions of fact and questions of law is important in many fields of law[87]. Despite the attempts of many "distinguished judges and jurists to formulate tests for finding the line between the two questions, no satisfactory test of universal application has yet been formulated"[88]. Some of the attempts of the past to establish rules were described in Collector of Customs v Agfa-Gevaert Ltd as importing "artificial, if not illusory"[89] distinctions, as well as rules of limited help in particular circumstances[90].
[86](1996) 186 CLR 389 at 394.
[87]Occasionally an attempt is made in legislation to identify what constitutes an "error of law" for the purpose of the relevant legislation. Thus, in s 476(1)(e) of the Migration Act 1958 (Cth), error of law is defined as "being an error involving an incorrect interpretation of the applicable law or an incorrect application of the law to the facts as found by the person who made the decision".
[88]Collector of Customs v Agfa-Gevaert Ltd (1996) 186 CLR 389 at 394.
[89](1996) 186 CLR 389 at 396.
[90](1996) 186 CLR 389 at 396.
This appeal is not the occasion for a lengthy re-exploration of these questions. Specifically, it is not a suitable vehicle in which to explore some of the debates that have been agitated in past decisions concerning whether "perverse" findings of fact can ever of themselves amount to an "error of law"[91]. There are analogies in the law which would support such an approach. These include the well-known ultimate category of the erroneous exercise of a discretionary power. An appellate court, even where it cannot point to specific error, is authorised to intervene in such a case if the outcome challenged is so "unreasonable or plainly unjust, the appellate court may infer that in some way there has been a failure properly to exercise the discretion which the law reposes in the court of first instance"[92]. However, no one suggested that the findings of fact made by the primary judge in this case were of that perverse variety. The most that was claimed was that, upon the findings made, the proper application of the Compensation Act required the conclusion that the journey on which the worker was embarked at the time she received her injuries fell outside the "journeys" for which that Act provided.
[91]Azzopardi v Tasman UEB Industries Ltd (1985) 4 NSWLR 139 at 155-157 per Glass JA (Samuels JA concurring at 157); cf at 146-151 of my own reasons; Ambulance Service of New South Wales v Daniel (2000) 19 NSWCCR 697 at 709‑711.
[92]House v The King (1936) 55 CLR 499 at 505.
In the Court of Appeal, Handley JA took his criterion for classifying the "journey" in question to be whether the description of it as a single journey contradicted the "true and only reasonable conclusion" open on the facts[93]. This was one variant of the test propounded by Lord Radcliffe in Edwards (Inspector of Taxes) v Bairstow[94]. By that test, Handley JA concluded that the "true and only reasonable" classification of the journey on which the present worker was injured was that it was a journey between her grandmother's home and her place of abode[95]. It therefore fell outside the Compensation Act.
[93]Vetter (1999) 18 NSWCCR 34 at 40.
[94][1956] AC 14 at 36. At one stage in his Lordship's speech, the text reads "the true and only reasonable conclusion". However, 10 lines later it is stated as the "only one true and reasonable conclusion".
[95]Vetter (1999) 18 NSWCCR 34 at 40-41.
Although others regard that test as useful[96], respectfully, I regard it as presenting a somewhat sterile criterion by which to decide whether the Court of Appeal has stepped outside its own powers. What appears to one judicial observer to be the "true and only reasonable conclusion" may strike another as open to argument. Another may regard it as manifestly incorrect. "True and reasonable" is a description of a conclusion not of the criterion by which to arrive at that conclusion. Obviously, s 32 of the Court Act, in limiting the Court of Appeal's jurisdiction and powers to correcting errors of law, imposes on the Court of Appeal significant constraints. Both Priestley JA and Handley JA recognised this. Obviously too, those constraints do not warrant the appellate court's simply substituting its own view of the facts and calling the substitution a correction of an erroneous statutory construction. The primacy accorded to the Compensation Court's factual decisions, by limiting appeals from its awards, relevantly, to the establishment of a grievance in point of law, must be respected so long as what is involved is a finding of fact. The giving of meaning to straightforward statutory language, according to the facts as found, is such a finding.
[96]Reasons of Gleeson CJ, Gummow and Callinan JJ at [25].
With respect, two elements in the reasoning of Handley JA may suggest error in the approach which his Honour took. The first lies in the apparent acceptance (expressly or inferentially shared by the other participating judges) that the relevant question was whether there were "two journeys or one". There is no doubt that, in many decisions in the past, judges have posed and answered this question[97]. However, strictly speaking, it is an unnecessary one. It may even sometimes tend to mislead. The issue, in point of law, is not whether there were two, three or more journeys. It is whether the "journey" propounded by the worker is one which, within the findings of fact, falls within or outside the statutory definition. By contemplating that the choice had to be made between classifying the worker's travel as one continuous journey or two journeys, the judges in the Court of Appeal erred.
[97]Mills, Workers Compensation (New South Wales) (1969) at 264-266. See eg Walker v Imperial Chemical Industries of Australia and New Zealand Ltd [1947] WCR (NSW) 47; Miller v The Commissioner for Government Transport [1955] WCR (NSW) 211; Price v Robert Corbett Pty Ltd [1958] WCR (NSW) 104; Werkspoor v Scowen [1960] WCR (NSW) 109.
It may have been equally erroneous to ask how the worker herself would have described the journey if requested to do so[98]. Her description would doubtless have depended upon the questioner and the purpose for which the question was posed. Under the applicable legislation, the duty of classification falls upon the primary judge. That judge's classification must stand unless an error "in point of law" authorises the Court of Appeal to disturb it. Usually, a view of the facts taken by the primary decision-maker will not amount to an error of law. It will only do so if there is no evidence to support the conclusion, if the conclusion itself or the reasoning offered to support it betray a mistaken view of the applicable law, or if no reasonable decision‑maker could have come to that view of the facts[99].
[98]Vetter (1999) 18 NSWCCR 34 at 38 per Handley JA.
[99]Edwards (Inspector of Taxes) v Bairstow [1956] AC 14 at 33-34 per Lord Radcliffe.
The second error in Handley JA's reasoning may appear in his Honour's ultimate conclusion. As stated, his Honour reached his conclusion "[n]ot without doubt"[100]. This may have been no more than a courteous phrase meant to soften the blow of reversal of the primary judge's award. However, if Handley JA was in doubt, that was tantamount to acknowledging that a decision-maker, charged with the duty to find the facts and to apply to them a statute expressed in ordinary words of the English language, could not unreasonably have come to a different conclusion. If that were so, the finding would not be such an error as to attract the powers of the Court of Appeal to correct errors of law on the part of the Compensation Court. If there was any reasonable doubt about the matter, it is difficult to sustain a conclusion that the evidence was "necessarily within the description of a word or phrase in a statute or necessarily outside that description"[101].
[100]Vetter (1999) 18 NSWCCR 34 at 40.
[101]The Australian Gas Light Co v The Valuer-General (1940) 40 SR (NSW) 126 at 138 per Jordan CJ (emphasis added).
I prefer not to decide this appeal on the first issue concerning the jurisdiction and powers of the Court of Appeal. In this Court, the decision should not turn on any passing infelicity in the reasons of the Court of Appeal or a conclusion that, on the primary question, that Court mistakenly saw an error of law where none existed. After all, the reference to whether there were "two journeys or one" might arguably be only another way of saying that, although the worker was obviously on a journey home from work in one sense of that phrase, because there had been an earlier separate journey from work, the later journey was to be classified as a second non-compensable journey rather than one falling within the statutory definition. I will assume that it was within the jurisdiction and powers of the Court of Appeal to regard the question presented by the appeal as tendering a point of law which asks "whether the facts are such as to bring the case within the provisions properly construed of some statutory enactment"[102]. But was a suggested error of law of that kind established?
[102]McPhee v S Bennett Ltd [1934] WCR (NSW) 372 at 373 per Jordan CJ; (1935) 52 WN (NSW) 8 at 9; cf Farmer v Cotton's Trustees [1915] AC 922 at 932.
The requirements for a "journey"
In a number of decisions, this Court has emphasised that claims for compensation in respect of injuries upon the journeys defined in the applicable Act may only succeed if the journey in question is properly classifiable as one between a specified origin and a specified destination[103]. This point was made most clearly in Young v Commissioner for Railways[104].
[103]See eg Landers v Dawson (1964) 110 CLR 644 at 651-652.
[104][1961] ALR 258 ("Young").
Young was a case where a worker, with leave from his employer to do so, left work on the day of his death at 2.35 pm because he had private business to attend to. Normally, he left work at 4.05 pm and arrived home at 6.10 pm. There was no evidence of what the worker did between 2.35 pm and 5.05 pm. At the latter time he met another worker in a hotel and left with him at about 5.40 pm. Just before 6.00 pm, on the road leading to his home, the car, driven by the worker, was involved in an accident and he was killed. This Court held that the claim failed. Dixon CJ said[105]:
"It is clear that the road upon which the accident occurred led to the deceased's home, and no one need doubt that he was then on his way home when the accident occurred. But that is not enough to satisfy the provision of the statute, which says that the journey must have been between the worker's place of abode and place of employment. Of course, as I have already shown, the proviso allows of interruptions, deviations and other breaks in the journey, but the journey must commence as a journey between the place of abode and the place of employment.
…
It is, of course, plainly true that ultimately on that day he intended to reach home. That intention was no doubt present with him when he left to go to the works in the morning; it was probably an intention which he entertained day after day. But it does not follow that he was on the journey between the factory, the place of employment, and his abode. He obviously had some other destination which doubtless was intermediate, but was a destination at which he was going to transact business. The time is long. It is not a short time. ... [It is] not right or in accordance with any probability of fact to treat that interval of time as a mere interruption of a journey which he commenced to go home. He commenced the journey to transact whatever business it was."
[105]Young [1961] ALR 258 at 260.
The decision in Young is distinguishable on the facts from the present case. Here, the route of the journey was known. It conformed to the worker's "periodic" pattern. There was no gap in the evidence. Nor was there any suggestion that the worker had commenced the journey with some unidentified person or business in mind. The origin of the journey was undoubtedly the "place of employment". The ultimate destination of the journey was the worker's "place of abode". It is true that there was both an interruption of, and deviation from, the direct journey to the worker's home. But there was no interruption of, or deviation from, the journey home which she took each Friday fortnight to interpose a call on her grandmother. And, as found and affirmed, any such interruption or deviation caused no material increase in the risk of injury. Accordingly, the relevant question became whether, because of the interruption and deviation by way of the grandmother's home, it must be said, objectively, that the worker's "journey" was not one falling within the Compensation Act.
The employer, in support of its construction, pointed to the fact that the journey provisions represent an additional benefit under the Compensation Act, conferred in terms limited to the provisions of s 10. A fiction was involved, as demonstrated by the way in which s 10(1) of that Act deemed a journey injury to be one "arising out of or in the course of employment". Being a fiction, it was argued, entitlements should be confined to the kind of case to which s 10 applied, namely journeys to and from home with a fair degree of directness and without undue interruption. This did not include journeys to or from a place of private concern which involved the worker "in the role of a civilian" and not as an "employee"[106].
[106]Tallar v K D Welding Co Pty Ltd [1955] WCR (NSW) 214 at 216 per Rainbow CCJ quoted in Vetter (1999) 18 NSWCCR 34 at 38 per Handley JA.
In support of this approach to characterising the kinds of "journey" for which the Compensation Act provides, the employer pointed to the permitted places of origin and destination of compensable journeys under the Compensation Act. The origins allowed included, where the worker was required or expected by the terms of the worker's employment to reside temporarily, a camp or like place[107]. The destination extended to the "place of employment", educational institutions, medical and hospital venues attended for compensation or like purposes, a place of work pick-up and where payment was made of wages or other moneys connected with the employment[108]. This list indicated, according to the employer, a high degree of specificity in a "journey" falling within s 10. From the list, as a matter of statutory construction, could be derived a conclusion that journeys for purely private purposes fell outside the statutory cover. So, it was argued, the journey to the grandmother's home was outside s 10. It was so, not because of a disqualifying interruption or deviation, but because the journey was properly characterised, when the worker set out, as a purely private affair with a destination different from those covered by the Compensation Act. The employer submitted that this conclusion was necessary when regard was had to the considerations which Handley JA had listed: (1) it involved travel in a direction opposite to the place of abode; (2) it increased the distance and time of travel significantly; and (3) it was for a private purpose having nothing to do with the employment[109].
[107]Compensation Act, s 10(3)(e).
[108]Compensation Act, s 10(3).
[109]Vetter (1999) 18 NSWCCR 34 at 40-41.
This approach to s 10 of the Compensation Act is mistaken. That Act is intended to apply to the employment journeys of workers in a great variety of employment and domestic situations. It provides a valuable benefit to such workers. This benefit should not be narrowly construed nor confined to journeys in which the employer has some direct or notional interest. From the first provision of this benefit in the 1926 Act, the realities of interruptions of, and deviations from, direct journeys to and from work were accepted as inevitable and not necessarily disqualifying. The fact that s 10 of the Compensation Act contemplates such interruptions and deviations contradicts a narrow classification of a "journey" as one to a private, non-statutory, destination, simply because it involves a departure from the direct journey between a permitted work origin and the place of abode. So long as the journey in question can fairly be characterised as a "journey", or part of a "journey", within s 10, questions of interruption and deviation must be left to the judge to determine in accordance with s 10(2). It is inconsistent with the terms of that sub-section for interruptions and deviations, as such, to deprive the travel involved of the character of a "journey" within the Compensation Act.
Of course, a point will be reached where an interruption is greatly protracted, a deviation is very substantial or the purposes are wholly private or unknown[110]. In such circumstances, the "journey" will never have had, or will during its course lose, the character of a "journey" for the purposes of s 10 of the Compensation Act. In appropriate cases, the trier of fact will be entitled to reach a conclusion that the worker, even if setting out on a compensable journey, abandoned that journey and began the pursuit of a purely private purpose[111]. Such decisions involve factual findings that will rarely present points of law. This is so because they are commonly dependent on an initial impression or a judgment concerning the intention of the worker derived from the evidence or from an absence of evidence[112].
[110]cf Young [1961] ALR 258.
[111]Price v Robert Corbett Pty Ltd [1958] WCR (NSW) 104; cf Miller v The Commissioner for Government Transport [1955] WCR (NSW) 211.
[112]Young [1961] ALR 258.

The extent to which s 10 of the Compensation Act and its predecessors have been interpreted to provide cover to a worker on a journey, although that journey involved significant interruptions and deviations, can be seen in the earlier decision in the Court of Appeal in The Old Spaghetti Factory v Oughtred[113]. Handley JA correctly saw that decision as apparently inconsistent with the conclusion to which he had come[114]. In that case, an interruption of almost two hours, after midnight, which took the worker in a direction opposite to his home and involved his staying for most of that time with a co-employee at a place of entertainment before delivering her home, was held by the primary judge to be a "journey" of the statutory kind. Unanimously, the Court of Appeal found that this involved no error of law. Priestley JA correctly regarded the decision as relevantly indistinguishable[115]. The issues presented concerning the journey in that case and in the present case were equally issues of fact and degree. They tendered, as such, no point of law.
[113][1975] WCR (NSW) 231.
[114]Vetter (1999) 18 NSWCCR 34 at 40.
[115]Vetter (1999) 18 NSWCCR 34 at 48.
In this Court, it matters not that the Court of Appeal departed from its own earlier precedent. But the sharp contrast between the decisions in somewhat similar facts, presents a choice between competing approaches and conclusions. In my opinion, the approach in The Old Spaghetti Factory v Oughtred was correct. It was open to the primary judge in this case, as in that case, to conclude, for the purpose of the applicable statutory provisions, that, notwithstanding the deviations and interruptions, the propounded "journey" remained one between the worker's place of employment and place of abode. Once that conclusion was reached, there was no basis upon which the Court of Appeal could find error, and particularly an error "in point of law". What was involved was no more than a classification of the facts by reference to legislation expressed in ordinary, non-technical language. On this point, I therefore agree with the conclusion of Priestley JA[116]. Without more, this opinion would support disposition of this appeal in accordance with the orders which Priestley JA favoured.
[116]Vetter (1999) 18 NSWCCR 34 at 49.
The issue of "fault" of the worker
The employer, by its notice of contention, argued the third issue. In the undisputed facts, was the Court of Appeal bound in law to find that the worker was disqualified from entitlements under s 10 of the Compensation Act because her personal injuries were "caused, partly or wholly, by ... fault"[117] on her part? If so, should the Court of Appeal have disposed of the appeal to it, consistent with a finding of "fault", by ordering that the award in favour of the worker be set aside on that equally determinative, but different, ground? Within the Court of Appeal, doubts have been expressed that that Court is authorised to make its own findings of fact in such circumstances, as distinct from requiring that such facts be found by the tribunal of fact, here the Compensation Court[118].
[117]Compensation Act, s 10(1A).
[118]North Broken Hill Ltd v Tumes (1999) 18 NSWCCR 412 at 421 per Beazley JA.
Since s 10(1A) was introduced in the Compensation Act, the Court of Appeal has, on a number of occasions, considered the approach proper to the determination of "fault of the worker" as that sub‑section has provided. Clearly enough, by departing (temporarily as it has proved) from expressing the disqualification in terms of "serious" and "wilful" misconduct on the part of a worker, s 10(1A) evinced a purpose to limit the entitlements of workers on a "journey" otherwise coming within the Compensation Act in which "fault" of the worker could be proved. It was common ground that the onus of proving such "fault" rested on the employer. In Aardvark Security Services Pty Ltd v Ruszkowski[119], by reference to the parliamentary debates, attention was called to the apparent purpose of the legislation that introduced the sub‑section to deprive workers of compensation benefits even where the "fault" proved was of the "minutest" degree[120].
[119]Unreported, Court of Appeal of New South Wales, 19 March 1993 at 6-7 of my own reasons.
[120]See also WorkCover Authority (NSW) v Billpat Holdings Pty Ltd (1995) 11 NSWCCR 565 at 595.
Nevertheless, as the authorities in the Court of Appeal have developed in and since that case, a number of points have been stated about s 10(1A) which are relevant to the present case. The notion of "fault" expressed by the sub‑section is to be read with regard to the purpose for which the disqualification is provided. That purpose means that the "fault" mentioned is not exactly the same as "fault" for the purpose of recovering damages in a claim framed in negligence brought against the worker, for example, by a passenger in the worker's vehicle[121]. The kind of "fault" to which the disqualification in s 10(1A) refers has been held to involve more than "some temporary inadvertence to danger, some lapse of attention, some taking of a risk or other departure from the highest degree of circumspection, [which may be] excusable in the circumstances because [they are] not incompatible with the conduct of a prudent and reasonable man"[122]. Moreover, to enliven the disqualification, the employer must show affirmatively that "fault" of the requisite kind played an effective role in causing the worker's personal injuries[123].
[121]WorkCover Authority (NSW) v Billpat Holdings Pty Ltd (1995) 11 NSWCCR 565 at 604 per Priestley JA.
[122]cf Sungravure Pty Ltd v Meani (1964) 110 CLR 24 at 37 as followed in WorkCover Authority (NSW) v Billpat Holdings Pty Ltd (1995) 11 NSWCCR 565 at 605 per Priestley JA; Paul Perry Horse Training Pty Ltd v Harker (1996) 12 NSWCCR 689 at 695-696.
[123]WorkCover Authority (NSW) v Billpat Holdings Pty Ltd (1995) 11 NSWCCR 565 at 610 per Clarke JA.
Because the sub‑section has been amended, and now has significance only for residual cases such as the present, I see no good purpose in re-examining the foregoing interpretations, adopted in the Court of Appeal. Whilst courts must give effect to the objects of Parliament as stated in the language used in the applicable legislation, it is not unreasonable, where the consequence of one view of legislative provisions may be wholly disproportionate and apparently unjust, to approach their meaning in a way as favourable as the words permit, to avoid such extreme and arbitrary outcomes. The way adopted by the majority in the Court of Appeal has that advantage. In the circumstances, whatever doubts I once expressed[124], I am prepared to accept that approach.
[124]Aardvark Security Services Pty Ltd v Ruszkowski unreported, Court of Appeal of New South Wales, 19 March 1993 at 7-9 of my own reasons.
Can it, therefore, be said, as a matter of law, that the employer in this case established "fault" of such a kind, on the part of the worker, so as to attract the disqualification from entitlements provided in s 10(1A)? Put another way, can it be said that a conclusion to that effect was the only one available to the primary judge and the Court of Appeal on the evidence? The employer relied on the language of the sub‑section, its purpose to cut back entitlements in respect of journey claims and the clear provision that even a partial contribution of fault on the part of the worker was enough to disqualify the worker from benefits under s 10. In the facts of this case, the employer submitted that the necessary, indeed only, inference to be drawn from the evidence was that there was "fault", at least to some degree, in the worker's losing control of her vehicle at the time and in the weather conditions that she did and running off the road, so as to come into collision with a parked vehicle.
In addition to relying upon the fact that motor vehicles, under proper control of drivers such as the worker, do not ordinarily leave a road without "fault" to some degree on the part of the driver, the employer invoked the evidence set out in a traffic collision report which was received in evidence in the Compensation Court. The police officer who made this report had taken apparent regard of the features of the road immediately before the place where the worker's vehicle left the bitumen. The fact that there was a "left sweeping bend on a downhill grade" suggested to that officer that the worker appeared "to have lost control whilst travelling north [and] negotiating" this bend. Another police officer called before the Compensation Court repeated this view. An expert called by the employer testified that there was nothing inherently unsafe in the section of road and that a mechanical defect in the worker's car was unlikely. He therefore also attributed the incident to the manner in which the worker had operated and controlled her vehicle. The worker's expert did not give evidence as to his conclusions about the cause of the accident. His reports did not address that issue. At trial, he was asked no questions concerning the subject.
In these circumstances, the employer submitted that the necessary, and indeed only, inference available on the facts was that the worker's injuries were caused, at least partly, by her own fault thereby putting her, as a matter of law, outside the Compensation Act. The employer emphasised that it was unnecessary to establish recklessness to attract the disqualification in the sub‑section[125]; that the trier of fact was entitled to draw reasonable inferences from established facts; that the maxim res ipsa loquitur remained part of Australian law and, in the absence of explicit proof of the cause of the mishap, that the maxim could be invoked in a case such as this[126].
[125]cf Medida Pty Ltd v Tobin (1995) 12 NSWCCR 576 at 582-583; Paul Perry Horse Training Pty Ltd v Harker (1996) 12 NSWCCR 689 at 696.
[126]cf Schellenberg v Tunnel Holdings Pty Ltd (2000) 200 CLR 121 at 140-141 [46]‑[47], 168 [124], 176 [153].
It is important to remember the basis upon which Priestley and Handley JJA concluded that the primary judge had erred in his treatment of the issue of "fault". Their Honours did not hold that only one finding was open on this issue in the facts which the judge had found. Their decision was that the primary judge had misdirected himself in law by concluding that there was "no evidence" to support a factual finding of fault. Their Honours held that, as a matter of law, there was some such evidence. It therefore remained for the Compensation Court to conclude whether such evidence was sufficient to establish the existence of fault. Only if there was "no evidence" upon which the contrary finding could be made would it be open to the Court of Appeal, in an appeal limited to a point of law, to substitute a finding to that effect and, on that basis, a decision in favour of the employer.
In my view, on this point, Priestley and Handley JJA were correct. The employer's submission on the notice of contention should be rejected. It is not possible to say that the evidence reaches the point beyond reasonable argument as the employer asserted. It is true that the worker herself failed to give evidence of an explanation consistent with an absence of fault on her own part, as would ordinarily have been expected. But in this case, such absence of evidence was fully explained. By reason of her injuries, the worker had no recollection of the accident or how it had occurred. Accordingly, her lack of evidence, far from assisting the drawing of an inference of fault, is neutral in that respect. It leaves the onus squarely on the employer to prove a causatively relevant "fault" and, in the absence of other evidence in the worker's case, effectively, to do so from its own evidence.
Whilst it is correct to say that motor vehicles do not ordinarily leave a road in circumstances such as occurred in this case, in the conditions of rain that prevailed at the time of the incident, the kind of collision that occurred could have been a result of momentary inattention judged to fall short of the "fault", as that word has been defined, that would disqualify the worker from recovery for an otherwise compensable injury. Whilst it is true that an inference of negligence may be drawn from facts of the kind disclosed in the evidence here and, indeed, that there is some evidence to support that inference, it is not inevitable that the trier of fact will draw that inference. The primary judge is not bound to do so. Within the constraints of rational decision-making, and subject to any appellate scrutiny of the reasons given, it is left to the trier of fact to determine whether or not the inference should be drawn[127].
[127]Government Insurance Office of NSW v Fredrichberg (1968) 118 CLR 403 at 414; cf Schellenberg v Tunnel Holdings Pty Ltd (2000) 200 CLR 121 at 132-134 [22]‑[24], 148 [72], 161-162 [105].
It follows, in accordance with the opinion of Priestley and Handley JJA (which the worker did not contest in this Court), that the primary judge has never decided whether or not to draw, or to reject, the inference available from the evidence relevant to the issue of fault. In the orders favoured by Priestley JA, the matter would be returned to the Compensation Court for retrial so that the tribunal of fact may perform its functions according to law and without the misdirection identified by the Court of Appeal.
Remitting the matter to the Compensation Court also upholds the respective functions of that Court and of the Court of Appeal. The latter, being confined, relevantly, to correction of errors in point of law, is not ordinarily authorised to remake findings of fact that have miscarried at trial. The only circumstance in which, as a matter of law, it would be entitled to do so, would be if the facts disclosed by the evidence yielded, in law, but one consequence. The whole point of Handley JA's analysis of the mistake made in this respect at trial (in which Priestley JA agreed) was that the fact-finder had not completed his task according to law. The completion of that task is what should now occur.
The scope of the retrial
The employer, however, argued that the order for a retrial, favoured by Priestley JA, was not appropriate if it meant that, before the Compensation Court, the worker would have a completely free hand to re-present her case calling new evidence, different from that called at the first trial. It was complained that this would amount to giving the worker a second opportunity to establish her entitlements as, it was claimed, she had failed to do at the first trial. Suggestions were made in argument that, in some unidentified way, the worker might be estopped from presenting her case in a different fashion. Because, by the Court Act, the Court of Appeal has large powers to make "such other order in relation to the appeal as the Court of Appeal sees fit"[128], a question arises as to whether the worker should, by order now made by this Court in disposing of the appeal to it, be limited to the record of the first trial, the only question to be reconsidered being the inference to be drawn about "fault" from the evidence as it stood at the end of the first trial.
[128]Court Act, s 32(2).
There are at least two answers to that question. First, it is not raised by any ground of appeal or by the notice of contention, as such. It does not therefore arise on the record tendered for determination by this Court. Secondly, and in any case, in an appeal limited to a point of law, where such an error has been demonstrated, it is normal for the appellate court to identify the error, to set aside the order affected by the error and to require a retrial in the court of trial, freed of the error which has been found, at least of the issues that miscarried. This was the order which Priestley JA favoured in this case[129]. It is not normal for the appellate court to interfere in the conduct of the retrial by the court of trial. Occasionally, where departures from the rules of natural justice or like error are shown, the appellate court may order, or suggest, that the trial court be differently constituted for the retrial. This is not such a case. The usual convention should be observed, as Priestley JA proposed. The employer sought, ultimately, to reserve any rights that it might have in the conduct of a retrial. What those rights (if any) may be, will remain for the future.
[129]Vetter (1999) 18 NSWCCR 34 at 50.
The costs certificate
A final issue arose concerning the certificate which the Court of Appeal granted to the worker when she lost the appeal before that Court. It is not competent, within present authority, for this Court to grant a certificate under the Suitors' Fund Act 1951 (NSW)[130]. The discretion to grant such a certificate belongs, in this case, to the Court of Appeal. The certificate which that Court earlier granted was provided on a footing different from that which would now apply. In short, it was based on the joint opinions of Handley and Powell JJA concerning the "journey" issue. By the decision of this Court affirming the approaches favoured by Priestley and Handley JJA, on the "fault" issue, it would now rest on Priestley JA's proposed orders, not given effect in the Court of Appeal because of Handley JA's opinion (with which Powell JA agreed) on the primary issue regarding the journey.
[130]Commissioner of Stamp Duties (NSW) v Owens [No 2] (1953) 88 CLR 168; Gurnett v The Macquarie Stevedoring Co Pty Ltd [No 2] (1956) 95 CLR 106.
Returning the matter unnecessarily to the Court of Appeal, solely to obtain a costs certificate (assuming such to be available in those circumstances[131]) where a certificate has already been granted, would involve needless delay, expense and possible injustice. It is sufficient in the circumstances to leave the order for the certificate to stand. The reasons may differ but the order itself does not require correction.
[131]cf "Suitors' Fund Act", (1956) 30 ALJ 117 at 118; Selby Shoes (Australia) Pty Ltd v Erickson (1955) 73 WN (NSW) 116.
It follows that the proceedings will have to be returned to the Compensation Court. But the only basis upon which the matter should be returned for retrial is to permit that Court to determine the issue of whether, within s 10(1A) of the Compensation Act, as it stood at the relevant time, it is proved that the personal injury suffered by the worker was caused partly or wholly by fault on her part.
Orders
I agree in the orders proposed by Gleeson CJ, Gummow and Callinan JJ.
HAYNE J. I agree that the appeal should be allowed and orders made in the form proposed by Gleeson CJ, Gummow and Callinan JJ. I agree generally with their Honours' reasons but wish to add something about whether the principal question which it was sought to agitate in the Court of Appeal was a point of law.
In both the Court of Appeal and in this Court, there was no dispute that "journey" is used throughout s 10 of the Workers Compensation Act 1987 (NSW) ("the Act") in its ordinary meaning. There was no live question between the parties about the construction of the relevant provisions of the Act. The only question was whether it could be said that the appellant was on a journey between her place of abode and place of employment when she was injured. That is, adopting what was said by Mahoney JA in Minchinton v Homfray[132], could it be said that the appellant was on a spell of going or travelling, viewed as a distinct whole, between those two places when she was injured? In the circumstances of this case, that was a question of factual classification about which minds might differ. It was not a question which did not admit of the answer which was given by the primary judge. It was not a question which, if answered in the sense in which the primary judge answered it, inferentially revealed a misunderstanding of the proper construction of the section. As Mason J pointed out in Hope v Bathurst City Council[133], when a statute uses words according to their common understanding and it is not unreasonable to conclude that the facts as found fall within those words, the conclusion that they do, is a conclusion on a question of fact[134]. The contention which the respondent made, that the conclusion that the facts fell within the statutory words was not reasonably open, is a contention on a point of law[135]. Once that question is resolved against the respondent, as I agree it should be, the remaining question of factual classification is a question of fact.
[132](1994) 10 NSWCCR 778 at 785.
[133](1980) 144 CLR 1 at 7.
[134]NSW Associated Blue-Metal Quarries Ltd v Federal Commissioner of Taxation (1956) 94 CLR 509 at 511‑512 per Kitto J.
[135]Australian Slate Quarries Ltd v Federal Commissioner of Taxation (1923) 33 CLR 416 at 419 per Isaacs and Rich JJ; Federal Commissioner of Taxation v Broken Hill South Ltd (1941) 65 CLR 150 at 155 per Starke J; NSW Associated Blue‑Metal Quarries (1956) 94 CLR 509 at 512 per Kitto J.

In answering that question of fact, it is unhelpful to speak of one journey or two. Those are shorthand descriptions of the answer given. The question to be answered is that posed by the statute – whether the worker was on a daily or other periodic journey between his or her place of abode and place of employment when injured. It is not whether there was one journey or two.