Lord Justice Clerk

Lord Brodie

Lord Philip

[2012] CSIH 91



delivered by LORD CARLOWAY,


in Reclaiming Motion



Pursuer and Reclaimer;



Defenders and Respondents:


Act:  I G Mitchell, QC, Wallace;  Drummond Miller LLP (for MacLachlan & Mackenzie)

Alt:  Milligan, QC; Paull & Williamsons LLP

4 December 2012

The pursuer’s Case on Record and at Proof
[1]        The pursuer seeks damages for injuries sustained by her on 13 August 2007. 

Quantum, on the basis of full liability, is agreed at £85,377 inclusive of interest to 22 November 2011.  The circumstances of the accident appear straightforward.  The pursuer parked her Vauxhall Signum car in a cul-de-sac, which had a 4.4 degree gradient.  She did not put it in gear.  It was accepted by both parties, although not established during the course of the proof, that, if the handbrake had not been applied to some extent when the car was parked, it would immediately have rolled backwards.  It did not do so.  For this reason, as the defenders explained at the hearing of the appeal, it was never contended by the defenders that the pursuer had not attempted to apply the handbrake, or that she had forgotten to do so.

[2]        The pursuer got out of the car and went round to the rear in order to open the hatchback and let her dog out.  As she was doing so, the car rolled backwards and the accident occurred.  After the proof, the Lord Ordinary found (Opinion para [3]) that “the pursuer applied the handbrake”.  He then narrated that the pursuer had testified “that it was her habitual practice to engage the handbrake by pulling it on quite aggressively and without depressing the button thereon” and had done so on the day of the accident.  Although the defenders did attempt to argue that the Lord Ordinary’s finding might be referable to a partial application of the handbrake, that is not how the court reads his Opinion.  Had the Lord Ordinary considered, on a balance of probability, that the pursuer had failed to apply the handbrake in the aggressive way which she described, the court is confident that the Lord Ordinary would have said so since that, in itself, would effectively have put an end to the case.  No issue would have arisen after proof, and certainly not upon appeal, as to whether any defect existed in the handbrake or otherwise; proof of which is the essential matter now in dispute.

[3]        Much turned at the proof upon the evidence concerning the handbrake and the effect of its application.  The pursuer led evidence from an expert consultant engineer, Alan Bathgate, who described the handbrake mechanism.  It is of conventional design and consists of a lever mounted on a central console between the front seats of the car at the front of a central armrest.  The lever moves through a vertical arc on a pivot.  This pivot, when turned, pushes a spring-loaded mechanism forwards, which in turn pulls a horizontal cable attached to the rear brakes.  Pulling that cable applies the brakes.

[4]        As is perhaps best illustrated by the above diagrams, when the lever is pulled, the mechanism is held in place by means of a pawl, the fork-like prongs of which slot over one of several teeth on a ratchet device.  Once in place, the pawl holds the ratchet, and hence the brake cable, in a particular tension.  The pawl can be raised free of the ratchet by pressing the release button at the front of the lever.  This pushes an internal rod backwards, causing the pawl to pivot upwards and free of the ratchet teeth.  If the button is not pressed, the two prongs of the pawl slot automatically, by means of spring loading, over one of the teeth of the ratchet. 

[5]        The brake can be applied by using one of two basic methods.  First, the release button can be pressed and the lever raised.  When the lever is in position, the button can be released, allowing the pawl to drop into position.  This method generates little sound.  The second method, preferred by some, is not to use the button at all when applying the brake, but simply to pull the lever up, dragging the pawl over the teeth of the ratchet until it stops in a suitable position.  This method creates the familiar handbrake ratchet sound.  It was this second method which was employed by the pursuer.  If the cabling was at the appropriate tension, there is no obvious reason why, on pulling up the lever using this second method, the rear brakes should not operate as they are designed to do.  Accordingly, the central issue is, assuming that the brakes did not operate in that way, how that eventuality came to pass.

[6]        The pursuer advanced a case based upon section 2 of the Consumer Protection Act 1987.  A common law negligence case had also been tabled, but it was not insisted upon in the appeal.  The case under the 1987 Act rests upon the pursuer demonstrating the existence of a “defect” in the car, and in particular its handbrake mechanism, which caused the accident.  It is of some importance, in that regard, to notice the precise terms of the case advanced by the pursuer on record, given the arguments which were ultimately presented at appeal.  In the fifth statement of fact, the pursuer avers that, at the time of the accident:

“The handbrake on the pursuer’s car failed to hold the car due to … defects in its design.  There was movement within the handbrake grip affecting the position of the pawl within the ratchet on the handbrake lever and that this could allow the pawl to slip on the ratchet and release the handbrake lever thus allowing the car to move towards the pursuer.  … The handbrake mechanism was situated between the driver’s seat and the passenger seat in a position that required the driver to reach back. When operating the handbrake forward pressure was placed on the handgrip tending to move it forward”.

There were further potential defects mentioned, but these fell by the wayside in the course of the first instance proceedings.  In the eighth statement, the pursuer set out her case under the 1987 Act, as follows:

“The handbrake was of defective design and manufacture.  As a result the safety of the pursuer’s car was not such as persons generally are entitled to expect”.

[7]        It is necessary to set out the evidence of Mr Bathgate in some detail, since much of the pursuer’s criticisms of the Lord Ordinary, on appeal, were directed towards his interpretation of Mr Bathgate’s evidence.  Mr Bathgate initially examined the car on 12 October 2007.  He was unable to identify any obvious fault in the handbrake mechanism.  On the contrary, he was able to apply the brake successfully simply by pulling the pawl over the fourth tooth on the ratchet.  It could be pulled up a further three teeth from that position. 

[8]        Mr Bathgate carried out a more detailed inspection by arranging for the removal of the handbrake several months later in January/February 2008.  Having done so, he ascertained that the plastic grip on the outside of the lever was not retained in position and could travel forwards and backwards along its length.  This was either because a “tang”, which was designed to retain the lever onto the central consol, had become worn and disengaged, or because the plastic grip of the lever itself, again as a result of wear and tear, had become loose.  It was Mr Bathgate’s position that this looseness in the grip had been present when he had first examined the car, even although he had made no note of the problem at that time.

[9]        Mr Bathgate was able to demonstrate to his own satisfaction that, on rare occasions, the semi-detached handbrake grip could slip and jam the release button fully forward, thus preventing the brake from operating properly.  As a result of this, the pawl could rest higher up on the ratchet’s teeth, rather than fully engage with them.  In terms of his report dated 26 September 2011 [no 6/15 of process] he wrote:

“There is then a possibility that a sudden movement of the vehicle, such as closing a door, could cause the pawl to slip on the edge of the teeth, releasing the handbrake. It would normally be expected in such a situation that the pawl would then catch on another tooth on the ratchet as it released, but this is dependent on the travel of the release button rod not being obstructed by the position of the hand grip”.

Having been referred to this passage in the witness box, Mr Bathgate continued [transcript p 74]:

“I have spent literally hours moving the handbrake back and forward, and I think on two or maybe three occasions only did I manage to get the handbrake into such a position that it was holding the release button in sufficiently far to cause it not to contact the ratchet at all… I can manage to pull it on and have the handgrip in such a position when it was on that it was possible by applying reasonable pressure to it to bump the handbrake off again and…the pawl did not catch on the ratchet…

…What I did manage to do is find a position with the handbrake grip where it was neither fully forward nor fully back but it did hold the…pawl slightly raised from the ratchet and it was holding, but just by bumping it I was able to release the entire handbrake, so, in other words, there seemed to be a position whereby with the handbrake on and appearing secured, the position of the hand grip was such that it held the pawl on the edge of the tooth, not fully engaged but not fully released from it, and it would appear that the button was partially depressed and locked in that position by the forward movement of the hand grip”.

Mr Bathgate was then asked, and answered, a leading question as follows (p 77):

“… you said at the beginning of this explanation that…in many attempts to see what might have happened, you were able to do that on two or three occasions?

Yes, that’s so. I had spent many hours [trying hard] to replicate it again and it’s almost impossible. It seems that possibly when it’s being raised, if the release button is pushed in slightly and the hand grip slips forward at the same time, it is locking and holding the release button in that position, which is just enough to allow the…pawl to sit on the edge of the ratchet and, if that position is maintained, it would be possible to bump it and release the handbrake”.

He repeated his position, about only being able to replicate this problem on 2 or 3 occasions despite “many, many occasions”, in answer to specific questions from the Lord Ordinary (p 81).

[10]      Mr Bathgate explained that, if the problem which he postulated had occurred, the driver would have thought that she had applied the handbrake properly.  But, again in an exchange involving a leading question (p 105), he stated that he could not say whether what he thought might have happened, did happen even though he could not think of any other way that could cause the car with the handbrake applied to roll backwards.  Various competing causes, enumerated by the defenders’ expert, Dr Jan Graham, were put to Mr Bathgate, but he excluded them all (other than handbrake defect), on the assumption that the pursuer had correctly applied the brake. In that context, he concluded his examination-in-chief as follows (p 146):

“…the only … explanation is … to do with the position of the hand grip holding the release button in a position that the pawl is not engaging into a ratchet and is effectively sitting almost tooth-on-tooth and then, when it does release when [the pursuer] is out of the car, because the pawl is held in its raised position, it’s not catching on subsequent teeth but is fully releasing the handbrake”.

[11]      In cross-examination, Mr Bathgate accepted (p 167) once again that he had only managed to replicate the circumstances of his theory 2 or 3 times over a period of 5 hours, or at least a whole afternoon.  He went on (p 168) to describe that what he was replicating was a “tooth-on-tooth” situation, whereby, if the pawl then slipped, it would come fully off the ratchet.  The cross-examination developed with the defenders suggesting to Mr Bathgate that, so far as the mechanism of failure was concerned, there was not much difference between him and Dr Graham.  The following exchange occurred (p 173):

“…essentially it’s the tooth-on-tooth that’s the problem.  How you get there might be slightly different?

Yes, but for the handbrake to fail completely, it would have to be tooth-on-tooth, and in that respect I do agree with Dr Graham”.

He did not agree, however, that this could not be achieved by dragging, as distinct from lifting, the pawl over the ratchet teeth.  

[12]      When Dr Graham’s report of being unable to replicate Mr Bathgate’s experiment was put to him, Mr Bathgate again explained (p 202) that he had, albeit on only two or three occasions, managed to re-create the exact circumstances which could explain the accident.  These he described as the pawl and ratchet “sitting tooth-on-tooth”.  He had not been able to replicate the experiment after his initial achievements.  He was asked (p 233) to agree with Dr Graham’s view that it is possible to set the handrake mechanism “tooth-on-tooth” by depressing the release button, pulling up the lever and carefully lowering the pawl onto the ratchet.  He agreed with this and that it required “considerable trial and error”.

[13]      At the very end of cross-examination, there was an important exchange (at p 233 et seq), which begins with further parts of Dr Graham’s report being put to Mr Bathgate, as follows:

“’The likelihood of achieving an inadvertent tooth-on-tooth setting in a ‘real’ handbrake application with the release button depressed is extremely low, requiring as it does that the handbrake lever is fortuitously moved to exactly the right position to allow the pawl to be released onto the correct, neutrally-stable position on a ratchet tooth, without upsetting the delicate balance’, … you would agree with that?

I would agree, yes.

‘If the release button is not depressed, the action of pulling the handbrake lever sets the pawl of the handbrake mechanism ‘bouncing’ over the teeth of the ratchet to create the characteristic buzz…?’


‘The possibility that a ‘real’ handbrake application can be made without the release button being depressed and yet the pawl comes to rest in exactly the right position to achieve a tooth-on-tooth setting is, in my view, sufficiently remote as to be implausible’?

I wouldn’t agree with that”.

The Lord Ordinary intervened to ascertain that it was the “implausibility” that Mr Bathgate disagreed with, since he thought that there was a “higher degree of possibility than implausibility” even if he could not quantify it.  Mr Bathgate concluded (p 235):

“You could probably apply the handbrake hundreds of times and never achieve it and once just get it on that crucial point. …it would require more luck than anything else to achieve that point, but, nonetheless, it is possible to do it”.

[14]      Finally, in re-examination, it was explained (perhaps for the first time, p 239) that Mr Bathgate was of the view that, because of the awkward positioning of the handbrake, the pursuer would have been pulling it forward as well as up, thus causing it to become loose and free from the tang.

[15]      The defenders elected not to lead any evidence, but submitted that the pursuer had failed to demonstrate the existence of a defect, or that it had caused the accident.  The defenders appear to have accepted the existence of the problem with the handbrake grip, but submitted that that was produced by wear and tear over the years since the car’s manufacture in 2004.  The pursuer could not establish that the defect had been present when the car had left the factory and that, according to the defenders, was fatal to her case. 

The Lord Ordinary’s Findings
[16]      The Lord Ordinary accepted (para [12]) the evidence of Mr Bathgate that he had been able to demonstrate on rare occasions that it was possible to operate the handbrake lever in such a way as to achieve a result whereby the pawl rested on top of the ratchet without engaging a tooth.  Indeed, he gave Mr Bathgate a general note of credit as a witness who spoke in a straightforward manner designed to assist the court.  Nevertheless, the Lord Ordinary stated that he was not:

“however persuaded, on the facts, that Mr Bathgate proffered an explanation satisfactory to explain the cause of this accident”.

He elaborated his reasons for this.  First, he was conscious that the failure of the lever had been demonstrated only in artificial laboratory conditions, once the handbrake mechanism had been removed from the car, rather than taking place in the car itself.  Secondly, he noted that, although he regarded this as “quite proper”, in carrying out his experiments, Mr Bathgate had set about attempting to achieve the result of having a pawl ending up resting on top of the ratchet teeth in a manner which did not hinder the movement of the ratchet.  This was “far removed” from the situation in which a person repeatedly applied the handbrake without any intention of achieving a particular result.

[17]      Thirdly (“second circumstance”), the Lord Ordinary took cognisance of Mr Bathgate’s evidence that he had only been able to achieve the result on 2 or 3 occasions.  The Lord Ordinary had regard to the view of Dr Graham, as expressed in his report (supra), that the likelihood of “this” happening in real life was “sufficiently remote as to be implausible”.  The Lord Ordinary did not consider that Mr Bathgate had “seriously” sought to rebut that contention.

[18]      In adopting the approach recommended by Lord Brandon of Oakbrook in The Popi M [1985] 1 WLR 948, the Lord Ordinary came to the conclusion that the pursuer had failed, on a balance of probability, to prove causation, in the sense of demonstrating that the accident had been caused by the defect alleged. 

[19]      The Lord Ordinary also determined that the defect in the handbrake grip, which may have existed at the time of the accident, had not been proved by the pursuer to have been present when the car had left the defenders’ factory.  He interpreted section 3 of the 1987 Act in a manner which did not provide a consumer with a “life-long warranty of goods” on the part of the manufacturer.  On this basis, he considered that the pursuer’s case under the 1987 Act failed.

[20]      Had he been in favour of the pursuer, the Lord Ordinary would have found her to have been contributory negligent to a “small” extent in failing to leave the car in gear, although he did not consider that such a finding was appropriate in respect of the defenders’ other contention that the pursuer’s failure to run round and re-enter the car as it rolled towards her also amounted to fault on her part.  He did not attempt to quantify the level of contributory negligence.

[21]      The first problem facing the court is to determine precisely what grounds of appeal are being maintained by the pursuer.  It has before it three written documents: (1) the grounds of appeal; (2) a “note of argument”; and (3) written “submissions”.  It then has the content of the oral presentation at the appeal hearing.  Neither the written documents, nor the oral submission, follow the same structure and it is difficult to extract from them the essence of the pursuer’s appeal.  What is attempted here is a distillation of the content of the written material as supplemented by, what appeared to be insisted upon, in oral argument. 

[22]      The terms of the Consumer Protection Act 1987 required to be construed in accordance with the Product Liability Directive 85/374/EEC.  Both imposed liability on the producer for damage caused by a “defect in a product” without the need to prove any fault.  “Defect” had a special statutory meaning, different from its normal meaning in the English language.  This was that “the safety of the product is not such as persons generally are entitled to expect” (1987 Act section 3(1)).  It was for the courts to determine what it was that persons were entitled to expect (A v National Blood Authority [2001] 3 All ER 289, Burton J at para 31 vii).  What the pursuer required to show was that there was something in the product’s safety, which the public were entitled to expect, but which was not present.  It was expectation at the time of manufacture which was important, but the defect did not require to be visible at that time.  There required to be an inherent safety problem in the manufacture of the product, but it was not necessary for the pursuer to demonstrate the precise mechanism whereby the defect thereby identified caused the damage.  It was sufficient for the pursuer to demonstrate that the damage must have resulted from some defect (Ide v ATB Sales Ltd [2008] PIQR P13 Thomas LJ at paras 7 and 19).  It was accepted that it was not enough merely to show that the product had failed and caused damage (Tesco v Pollard [2006] EWCA Civ 393, Laws LJ at para 17); but, where the only plausible explanation for the failure was a defect in the product, the pursuer must succeed (Clerk and Lindsell, Torts (19th ed) para 11-55; Richardson v LRC Products Limited [2000] Lloyd’s Rep Med  280; Foster v Biosil [2001] 59 BMLR 178, Booth QC at 180; Abouzaid v Mothercare (UK) Ltd 2000 WL 1918530, Pill LJ at para 22). 

[23]      It was common ground that, if there was a failure in the product, it was in the handbrake.  There had been two theories advanced by the competing experts, although only one gave evidence.  The defenders’ explanation on implausibility related to the pawl and ratchet being “point to point”.  However, the pursuer’s expert had not said that there was a need to achieve a precise “point to point” match between pawl and ratchet.  On the contrary, he was saying that it was possible for the pawl to slip if the pawl, although not “point to point”, had rested higher up than it should have on the ratchet tooth.  The Lord Ordinary had conflated the two theories of the experts and misunderstood Mr Bathgate’s evidence when he concluded that Mr Bathgate had accepted the rarity of his own theory.

[24]      If, as in this case, the car had left the factory gates in a state whereby the slipping of the handbrake grip would, in due course, happen, that constituted a lack of safety in the product and hence a defect in terms of the legislation.  Once the pursuer was able to present in evidence a mechanism for the failure of the handbrake, and there was no other competing theory supported by the evidence nor any basis for the court to reject the evidence of the mechanism or failure, she had demonstrated that the car was unsafe and therefore ought to succeed.

[25]      The defenders’ arguments at the appeal hearing followed broadly the approach taken by them in their written submissions.  In terms of the 1987 Act, as interpreted in light of the Product Liability Directive 85/374/EEC, in order to succeed the pursuer did not require to prove fault on the part of the defenders, but she did need to demonstrate that: (1) the product was defective and (2) the defect caused the injury complained of.  The word “defect” ought to be construed, in the first place, in terms of the Shorter Oxford English Dictionary definition as involving something “wanting or falling short” or a “lack or absence of something essential to completeness”, or a “deficiency”.  The explanation as to its meaning in the statute resulted in a narrower construction, since the defect had to be something which affected the safety of the product. 

[26]      A mere failure of a product was not sufficient to achieve success (Clark and Lindsell (supra) para 11-55; Richardson v LRC Products Limited (supra); Foster v Biosil (supra) Booth QC at 181).  If all that had been required was proof of failure of the product, then there would be no prospect of a manufacturer ever being able to invoke the defence, that the defect did not exist at the time the product was initially supplied (section 4(1)(d)).  Almost by definition, failure cannot be the same as defect, since any failure occurs post supply.  The test of safety, in relation to what persons are generally entitled to expect, should not be overly strict and the 1987 Act should not be used by the courts as an equivalent of a contractual warranty on the part of the manufacturer (Tesco v Pollard (supra) Laws LJ at para 15). 

[27]      The pursuer’s primary position was that there had been a particular defect in the handbrake.  She had a subsidiary argument that, even in the absence of a proven defect, the mere fact that the handbrake did not work on this occasion was sufficient to enable her to succeed.  If that argument were correct, it would mean that the pursuer could succeed if she demonstrated that on a particular occasion the product had not worked as expected, even although it had worked perfectly well before and after the incident and no defect had been found on subsequent inspection.  There was no precedent, under the 1987 Act, which indicated that this was the law.

[28]      A fundamental problem underlying the pursuer’s case was that the precise nature of the defect had never been identified.  There were a number of defects referred to on record but, ultimately at the proof, the one complained of was the slipping of the handbrake grip.  It had only been at the stage of submissions that the pursuer had attempted to argue that there was no need to prove a particular defect and that it was sufficient to show that the brake did not work on this occasion.  That approach is not permissible.  The pursuer’s case had to be tested on the premise that she had to prove that there was a defect in the grip, which caused a “tooth on tooth” application, which in turn led to the accident. 

[29]      Mr Bathgate had given extensive evidence on the timing of the creation of the alleged defect.  He considered that the grip had broken after manufacture and supply, possibly as a result of wear and tear.  It could not even be said that the grip had been broken at the time of the accident, given that Mr Bathgate had not made any reference to it at the time of his initial report.  Although the Lord Ordinary may have misdirected himself as to the onus of proving the existence of the defect at the time of supply, it was implicit from his reasoning that he considered that it was probable that the defect arose post supply.  That was the only reasonable finding to make. 

[30]      The issue of causation was one purely of fact.  The court should be slow to interfere with findings of fact made by a judge who has seen and heard all the evidence in the case and, in particular, should not do so unless it could be demonstrated that the Lord Ordinary was “plainly wrong” (Thomas v Thomas 1947 SC (HL) 45, Lord Thankerton at 54 and Lord Macmillan at 59, see also Piglowska v Piglowski [1999] 1 WLR 1360, Thomson v Kvaerner Govan 2004 SC (HL) 1 and Barber v Somerset County Council [2004] 1 WLR 1089, Lord Rodger at para 16 and Lord Walker at paras 67 and 70). 

[31]      The Lord Ordinary had been entitled to reject the evidence of Mr Bathgate regardless of what the contradictory evidence might have been.  The law on this matter was clear from The Popi M (supra) where Lord Brandon of Oakbrook undermined the application of Sherlock Holmes’s logic that when the impossible was eliminated “whatever remains, however improbable, must be the truth”.  The mere fact that a particular mode of failure was, in the opinion of an expert, “possible” did not mean that the Lord Ordinary was bound to accept it as a cause of the accident.  Mr Bathgate had been able to replicate the failure only when careful and delicate positioning of the handbrake had taken place.  Yet, the evidence of the pursuer was that on the day of the accident she had wrenched the handbrake lever up, making the possibility of achieving a “tooth-on-tooth” application too remote to be plausible.  Although it was not necessary for the defenders to offer an alternative explanation, there were at least two other possibilities: (1) that the pursuer simply failed to apply the handbrake properly; and (2) that the pursuer achieved a “tooth on tooth” application regardless of any defect in the handbrake.  In all these circumstances, it could not be said that the Lord Ordinary had not been entitled to come to the view that the explanation put forward by the pursuer was too unlikely to be the probable cause of the accident.  If the court was against the defenders, then the finding of contributory negligence should be one of 20%.

[32]      There is little doubt that the Lord Ordinary has posed, for this court, a significant conundrum.  If a handbrake is properly applied, as has been held to have occurred in this case, how does a car come to roll backwards on a modest incline?  The obvious possibility, as a matter of common sense, is that the handbrake must have failed as a result of some form of defect, giving that term its normal English meaning.  Yet the Lord Ordinary has found that such a defect has not been proved; thus leaving the cause of the accident as, presumably, an unexplained and perhaps inexplicable mystery.  Accordingly, this court requires to analyse the Lord Ordinary’s reasoning for his conclusion with some considerable care and to determine whether he, as the primary assessor of fact, has misdirected himself or reached a conclusion which is not justified by the evidence. In relation to the latter, whilst affording the Lord Ordinary due deference because of his advantage in having seen and heard the witness, this court will be entitled to intervene if the Lord Ordinary’s decision on the facts: (i) contains material inconsistencies or inaccuracies; (ii) fails to appreciate properly the weight or bearing of facts admitted or proved; or (iii) is otherwise plainly wrong (Thomas v Thomas 1947 SC (HL) 45, Lord Thankerton at 54 and Lord Macmillan at 59).

[33]      The principal issue is whether, having rejected the idea that the pursuer had failed to apply the handbrake (or failed to do so adequately), the Lord Ordinary was nevertheless entitled to reject also the only other explanation tendered to him in evidence; that the handbrake mechanism was defective because of the possibility of the pawl being jammed free of the ratchet as a result of the loose grip.  This, not for the first time, requires the court to consider whether it is appropriate to use the logic, attributed by Sir Arthur Conan Doyle to his fictional detective Sherlock Holmes in, for example, the Sign of Four, that once the impossible has been eliminated, “whatever remains, however improbable, must be the truth”.  This logic is sound as stated, but it is only applicable if all possibilities are known and hence capable of elimination.  In real situations, closed systems of possible cause do not always exist.

[34]      The difficulty about a court of law applying the detective’s analysis was thoroughly examined in England in The Popi M [1985] 1 WLR 948.  The judge at first instance (Bingham J, [1983] 2 Lloyd’s Rep 235) had been left with two explanations for the sinking of a cargo vessel in calm seas.  The first was some unexplained action of wear and tear and the second was collision with a submarine.  The judge regarded (at 248) the submarine hypothesis as containing an “inherent improbability” yet found it established on a balance of probability since he had rejected wear and tear and it had been the only alternative postulated in the evidence.  The Court of Appeal ([1984] 2 Lloyd’s Rep 555) considered that the judge was “plainly right” (Donaldson MR at 558).  However, the House of Lords disagreed.  Lord Brandon of Oakbrook, who delivered the only substantive speech, began (at 951) by stressing that, where the burden of proving the cause of an event is on a pursuer, although it is open to defenders to advance an alternative cause, there is no obligation on them to do so or to prove it.  Following the obiter dictum of Scrutton LJ in La Compania Martiartu v Royal Exchange Assurance Corporation [1923] 1 KB 650 (at 657) (which Bingham J had also cited), he continued:

“…it is always open to a court, even after the kind of prolonged inquiry with a mass of expert evidence… to conclude…  that the proximate cause of the ship’s loss, even on a balance of probabilities, remains in doubt, with the consequence that the [pursuer has] failed to discharge the burden of proof…”.

He regarded it inappropriate to use Holmes’ logic for three reasons.  It is worth quoting these:

“The first reason is…that the judge is not bound always to make a finding one way or the other with regard to the facts averred by the parties.  He has open to him the third alternative (sic) of saying that the party on whom the burden of proof lies in relation to any averment made by him has failed to discharge that burden. No judge likes to decide cases on the burden of proof… There are cases, however, in which, owing to the unsatisfactory state of the evidence or otherwise, deciding on the burden of proof is the only just course for him to take.

The second reason is that the dictum can only apply when all relevant facts are known, so that all possible explanations, except a single extremely improbable one, can properly be eliminated…

The third reason is that the legal concept of proof of a case on a balance of probabilities must be applied with common sense.  …If a judge concludes… that the occurrence of an event is extremely improbable, a finding… that it is nevertheless more likely to have occurred than not, does not accord with common sense…”.

Lord Brandon identified Bingham J’s error as being that he thought that he required to choose between the (only) two theories presented to him, even if he regarded one as “extremely improbable” and the other as “virtually impossible”.

[35]      Lord Brandon’s reasoning was explored in Ide v ATB Sales [2008] PIQR P13 in which Thomas LJ referred to the circumstances of The Popi M (supra) as being very unusual and stated that:

“4        …In the vast majority of cases where the judge has before him the issue of causation of a particular event, the parties will put before the judges two or more competing explanations as to how the event occurred, which though they may be uncommon, are not improbable.  In such cases, it is… a permissible and logical train of reasoning for a judge, having eliminated all of the causes of the loss but one, to ask himself whether, on the balance of probabilities, that one cause was the cause of the event.  What is impermissible is for a judge to conclude in the case of a series of improbable causes that the least improbable or least unlikely is nonetheless the cause of the event; such cases are those where there may be very real uncertainty about the relevant factual background… or the evidence might be highly unsatisfactory.  In that type of case the process of elimination can result in arriving at the least improbable cause and not the probable cause.

6.         As a matter of common sense it will usually be safe for a judge to conclude, where there are two competing theories before him neither of which is improbable, that having rejected one it is logical to accept the other as being the cause on the balance of probabilities.  …if there were only three possible causes of an event, then it was permissible for a judge to approach the matter by analysing each of those causes.  If he ranked those causes in terms of probability and concluded that one was more probable than the others, then, provided those were the only three possible causes, he was entitled to conclude that the one he considered the most probable, was the probable cause of the event provided it was not improbable”.

[36]      This is all somewhat convoluted but no doubt the general sense is clear.  The dictum proposes that an improbable cause cannot, on a balance of probability, be the cause!  However, Holmes’ thinking should not be discounted entirely.  What might, at first sight, be improbable or implausible may become increasingly more probable as and when alternative causes are eliminated.   In a case such as the present, involving a very simple road traffic accident, it can only be regarded as at least unfortunate that the case required to be decided on onus of proof.   This is an echo of Lord Brandon’s comment that judges do not like to decide cases purely on onus of proof.  As Lord Reid stressed, in McWilliams v Sir William Arrol & Co 1962 SC (HL) 70 (at 83):

“In the end when all the evidence has been brought out, it rarely matters where the onus originally lay: the question is which way the balance of probability has come to rest”.

The question is whether the Lord Ordinary has erred in deciding the case purely on the basis of a failure to overcome the onus of proving that the car rolled backwards because of a defect, in the face of evidence that a defect may possibly have existed.

[37]      In order to succeed in a claim under section 2 of the Consumer Protection Act 1987, a pursuer must prove that he suffered damage “caused… by a defect”.  This phraseology ought not to pose any complex questions of understanding.  It involves proof of the presence of a defect and of damage caused by that defect.  “Defect” should, in the first place, be given its ordinary English meaning.  It involves the product lacking something.  Given the terms of section 3(1) relative to there being a defect in the product if its safety is not such as persons are generally entitled to expect, the defect must be a lack relative to the safety of the product (see Abouzaid v Mothercare (UK) 2000 WL 1918530, Pill LJ at para 22).  Whether such a lack of safety exists is determined according to what the court considers the public would expect. The court does not understand there to be any dispute with the proposition that a car would have a defect if its handbrake mechanism was such that it did not, upon proper application, engage at least the rear brakes of the car.  The public are entitled to expect that handbrakes do perform this function and do not, suddenly and without warning, fail.  If this was the position at the time of supply to the first customer, there would be a defect which sounded in damages if loss occurred as a result.  Equally, if the mechanism was such that such a failure could occur, without any problem being noticed in advance, in the future, the same result would follow.  If, therefore, it were possible, in terms of Mr Bathgate’s evidence, that the pursuer could have used the handbrake properly, yet the rear brakes were not engaged, leaving aside any issue of wear and tear, a defect in the handbrake and the car would exist.

[38]      The court does not consider that it is of any moment that the product is described as the car or the handbrake, or both.  However, it is clear from the pursuer’s pleadings that what the pursuer set out to prove was the existence of “defects” in the “design and manufacture” of the handbrake.  The particular one founded upon after proof was “movement in the handbrake grip affecting the position of the pawl”.  If the pursuer proved that, she would succeed.  If she did not, she would be bound to fail as not having proved her case on record which, as a matter of fair notice, the defenders had come to meet and which remained the only case against them when the pursuer closed her proof. 

[39]      There will be cases in which the failure of the product to act in a particular way will lead inevitably to the inference that a defect of some sort must have existed (Clerk and Lindsell, Torts (19th ed) para 11-55; Ide v ATB Sales (supra)).  The court accepts also that, in such cases, a pursuer may not have to prove the precise mechanism of how the defect led to the failure.  However, a car which has had its handbrake applied at one point may roll backwards at a later point for a variety, albeit a limited variety, of reasons which do not involve a defect in the handbrake (or the car) itself.  One obvious example is that, after applying the handbrake, the brake was accidentally released somehow.  It may even be possible to consider some “inexplicable failures” (Richardson v LRC Products [2000] Lloyd’s Rep Med 280, Ian Kennedy J at 285), although there was no evidence of such causes in this case.  However, the short point remains that this pursuer set out to prove that it was a defect in the car’s handbrake which prompted the movement.  The Lord Ordinary was not bound to hold that all possible explanations, other than defect in the handbrake or even the car, had been eliminated at the proof.  In these circumstances, he was entitled to the view that the movement of the car itself did not, without more, provide sufficient evidence of defect.  Rather the pursuer’s success depended upon the acceptance of Mr Bathgate’s theory of handbrake failure.

[40]      A court is seldom bound to accept evidence presented to it, even if it is unchallenged or uncontradicted.  In the absence of agreed fact apparent on the face of the record or a joint minute, it may reject the testimony of a particular witness, or parts of it, as, in this case, not sufficiently reliable.  Where the evidence is of a crucial nature, the judge is bound to give a reason for its rejection and, when it is unchallenged or uncontradicted, the reason ought to be correspondingly clear.  The appellate court will not, however, reverse a decision on the reliability of a particular piece of testimony where the judge has given cogent reasons for its rejection.

[41]      The court rejects the submission that the Lord Ordinary misunderstood the nature of Mr Bathgate’s evidence or the challenge to it.  As set out above, although Mr Bathgate began by describing a situation in which the pawl was resting high up in the grooves of the ratchet teeth, his references to being able to replicate the circumstances of the accident only 2 or 3 times related to the creation of a situation whereby either the pawl sat “on the edge of the ratchet” or, as he said at the end of his examination-in-chief, “not engaging into a ratchet” but “sitting almost tooth-on tooth”.  This was emphasised in cross-examination.  It is clear that Mr Bathgate’s 2 or 3 times related to being able, one way or another, to create a situation in which pulling the lever up by whatever method left the pawl effectively free of the ratchet in the event of movement.

[42]      The court considers that the Lord Ordinary has given cogent reasons for rejecting Mr Bathgate’s evidence that this “tooth-on-tooth” contact could occur in reality.  First, no such defect was discoverable on normal application of the handbrake in the car. Secondly, it was only created when deliberately attempted.  The experiment was designed to achieve a particular result and not to discover empirically whether there was a fault in the system.  Thirdly, the recreation was only achieved on 2 or 3 occasions over a period of at least an afternoon.  It was not achieved again and could not be demonstrated at the proof.  Hearsay of Dr Graham’s view that it was “implausible” that the “tooth-on-tooth” positioning could occur in real life was admitted in the course of the pursuer’s proof and the Lord Ordinary cannot be faulted for referring to it, even if Dr Graham was not called upon to testify.  The description of the likelihood as “remote” was regarded by the Lord Ordinary as one which was not seriously rebutted by Mr Bathgate.  That was not an unreasonable conclusion, given Mr Bathgate’s responses in cross-examination (supra).  Mr Bathgate certainly classified the possibility of his explanation actually occurring as greater, in possibility terms, than implausibility but the Lord Ordinary was entitled, in the absence of any attempt at greater quantification, to regard the differences between Mr Bathgate and Dr Graham in this area as not particularly significant. 

[43]      Ultimately, the Lord Ordinary was left with what he regarded as an implausible explanation which, in terms of Lord Brandon’s cautionary words, he could therefore hardly hold to have been the cause of the car’s movement.  For the reasons given above, the court is unable to fault the Lord Ordinary’s reasoning on this critical issue and, on that basis, this reclaiming motion must fail.

[44]      Were the court to have reached the view that the Lord Ordinary had erred in a material respect, it would then have required to reconsider the printed version of the evidence in order to determine for itself where the balance of probability had come to rest.  Suffice it to say that, upon such a reconsideration, the court may have struggled to come to any conclusion other than that this handbrake had simply not, for whatever reason, been properly applied.  In that regard it may have been forced to conclude that the pursuer’s testimony was not to be relied on in this crucial area of the proof. 

[45]      The court would have concluded that the defenders had made out their case that, even if there had been a defect in the car at the time of the accident, it could not be said to have existed three years earlier, at the time of its first supply after manufacture in 2004.  There was no material that would have justified the conclusion that there was anything wrong with the manufacture of the handbrake.  If the handgrip had come free of the “tang” or lever by the date of the accident, that was something which was proved to have occurred as a result of wear and tear of the type which the public would anticipate might occur.  It was not something which could be regarded as a flaw or deficiency inherent in the design or manufacture of the handbrake mechanism or the car itself.

[46]      Finally, had the court been in favour of the pursuer, it would have assessed contributory negligence in respect of the failure to leave the car in gear at 10%.  In that regard, it stresses that quantifying contributory negligence is a function to be carried out by the court of first instance irrespective of the finding on liability (Hogan v Highland Regional Council 1995 SC 1, LJ-C (Ross) at 2).