BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

England and Wales Court of Appeal (Civil Division) Decisions


You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Gloster v Greater Manchester Police [2000] EWCA Civ 90 (24 March 2000)
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2000/90.html
Cite as: [2000] EWCA Civ 90

[New search] [Context] [View without highlighting] [Printable RTF version] [Help]


CASE NO: CCRTF 1999/0518/B2
IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE MANCHESTER COUNTY
COURT (HIS HONOUR JUDGE TETLOW)
ROYAL COURTS OF JUSTICE
STRND, LONDON WC2A 2LL
Friday 24 MARCH 2000

Before:
LORD JUSTICE PILL
AND
LADY JUSTICE HALE
__________________________
CHRISTOPHER JOHN GLOSTER

APPELLANTS
-V-
CHIEF CONSTABLE OF GREATER
MANCHESTER POLICE

RESPONDENTS
_________________________
(Transcript of the Handed Down Judgment of
Smith Bernal Reporting Limited, 180 Fleet Street
London EC4A 2HD
Tel No: 0171 421 4040, Fax No: 0171 831 8838
Official Shorthand Writers to the Court)
_________________________

Mr Nigel Poole (instructed by Russell, Jones & Walker) appeared for the Appellants
Mr Mark Turner QC (instructed by Berrymans Lace Mawer) appeared for the Respondents
_________________________
Judgment
As Approved by the Court
Crown Copyright ©


LORD JUSTICE PILL:

This is an appeal against a decision of His Honour Judge Tetlow given on 19 April 1999 at Manchester County Court. Mr Christopher John Gloster is a police officer in the Greater Manchester Force. While acting in the course of his duties, he was bitten by a police dog. His claim for damages was dismissed. The claim had been made under section 2 of the Animals Act 1971 ("the 1971 Act") and in negligence. Damages had been agreed, subject to liability, at £3,000.
The judge made detailed findings of fact which are not in dispute on the hearing of the appeal. On 18 February 1996, shortly after midnight, the claimant was driving a police car, with another police officer as passenger, in pursuit of a Ford Orion motor car which had allegedly been stolen. A van containing a police dog, Manpol Jack ("Jack") and his minder, Constable Owen, was also is pursuit.
Constable Owen stated over the radio that officers were to remain in their vehicles should the stolen car be abandoned. That was because he proposed to use his dog to effect an arrest. The claimant did not hear the message, probably because he was concentrating on his driving. The Ford Orion collided with another vehicles and the dog van and the police car stopped nearby. The driver of the Orion ran off.
Constable Owen took Jack, who is a German shepherd dog, otherwise known as Alsatian, out of the van and put him on a choker lead. He proposed to get the dog into a position where the dog could see the quarry. The claimant left his vehicle. He knew that when police dogs are about, you keep out of the way until the dog has gone on its way. The reason is that police dogs, once triggered to act and set off, cannot distinguish between friend and foe.
Constable Owen intended to release the dog with the object of detaining the fleeing driver, once the dog could see the quarry. Constable Owen shouted "This is the police with a dog. Stand still or I'll send the dog". When the quarry was sighted, the officer would have released the dog with the instruction "Hold him".
Unfortunately, when running with the dog, Constable Owen slipped and fell. Jack slipped his collar. In the meantime, the claimant had attempted to intercept the passenger in the car who was also running away. While he was going for the man, the dog went for him and bit him twice on the leg. As soon as he realised what was happening, Constable Owen shouted "Leave" and the dog obeyed immediately. Constable Owen's evidence was that police dogs, once triggered to act and sent off, cannot distinguish between friend and foe. Once the dog has seen a target, he sticks to that target.
Jack was a satisfactory police dog. He operated in accordance with its training and was not over aggressive. Jack conformed to the standards expected of trained police dogs. He was not aggressive when not working.
It was agreed between the parties that, but for its training, this breed of dog would not have acted in the way that Jack acted on this occasion. Police dogs will proceed to search for and, if possible, detain when given the appropriate command. The judge found that the attack on the claimant was a result of the dog's training and not a departure from it. It was conduct expected of the dog.
The judge rejected the claim that Constable Owen had been negligent either in slipping and thereby releasing the dog or in the instructions he subsequently gave. The judge found that it was "an accident pure and simple". It had occurred without negligence. There is no appeal against that finding.
The judge went on to consider the claim under section 2 of the 1971 Act. He held that the requirements of section 2(2)(b) were not satisfied. It was also held that "training is different from characteristics" and that "this unfortunate accident was due to the dog's training rather than natural inclination or characteristic". The judge stated that a trained animal did not seem to him to "fit comfortably" in the scheme of the Act and I respectfully agree. The judge referred to the statement of plaintiff's council in Cummings v Granger [1977] QB 397, 400, that "a trained or conditioned guard dog will never come within section 2(2)(b) of the Act of 1971".
Section 2 of the 1971 Act provides, insofar as is material:
"(1) Where any damage is caused by an animal which belongs to a dangerous species, any person who is a keeper of the animal is liable for the damage, except as otherwise provided by this Act.
(2) Where damage is caused by an animal which does not belong to a dangerous species, a keeper of the animal is liable for the damage, except as otherwise provided by this Act, if--
(a) the damage is of a kind which the animal, unless restrained, was likely to cause or which, if caused by the animal, was likely to be severe; and
(b) the likelihood of the damage or of its being severe was due to characteristics of the animal which are not normally found in animals of the same species or are not normally so found except at particular times or in particular circumstances; and
(c) those characteristics are known to that keeper ... ."
Dogs are not a dangerous species within the meaning of the section. By virtue of section 11 of the Act, "species" includes sub-species so that the comparison required in section 2(2)(b) is between Jack and other German shepherd dogs (Hunt v Wallis [1994] PIQR 128).
The difficulties presented by section 2(2) of the Act are well recognised. In Cummings Lord Denning MR stated, at p 404, that the section "is very cumbrously worded and will give rise to several difficulties in future". Ormrod LJ stated, at p 407, that the language of section 2(2)(b) was "remarkably opaque". In Curtis v Betts [1990] 1 WLR 459, Slade LJ considered the wording of the sub-section to be "somewhat tortuous" and Stuart-Smith LJ stated that he had not found the sub-section easy to construe. In Breeden v Lampard (unreported 21 March 1985) Lloyd LJ found the meaning of the concluding words of the sub-section "elusive" and Oliver LJ expressed "puzzlement as to what is the meaning of section".
The problem which arises in this case is the operation of section 2(2) when the characteristic relied on by the claimant is that Jack had been trained as a police dog, was engaged in searching for a suspect and was trained to be aggressive and bite in some circumstances. On behalf of the claimant, Mr Poole submits that the damage was due to Jack's training as a police dog and such training is a characteristic not present in the generality of German shepherd dogs. It was the characteristic of being trained which caused the dog to act atypically, as it was put. The relevant characteristic being the training, Mr Poole did not consider that he needed to rely on the "particular circumstances" addition to section 2(2)(b) but, in case he did, added the submission that Jack, having been trained as a police dog, was likely to cause damage in the particular circumstance which arose.
The first point taken by Mr Turner QC, on behalf of the respondent, is that the claimant fails to establish a causal link between the characteristic relied on and the likelihood of damage being severe. The judge had found as a fact, under section 2(2)(a), that the claimant had not established that the damage was of a kind which the animal, unless restrained, was likely to cause. The judge went on to find under the second limb of section 2(2)(a), that "if this dog bit, the damage would be severe, and therefore, sub-section (a) is satisfied". That likelihood was not due to the training of the animal, it is submitted, but to its big, sharp teeth. The second submission on behalf of the respondent is that the fact that Jack, like many German shepherd dogs, had been trained as a police dog did not imbue him with characteristics not normally found in German shepherd dogs. The dog did not cease to be normal as a result of its training as a police dog.
Mr Turner does not attempt to distinguish between the innate and acquired characteristics of a breed of dog or, as he put it, between nature and nurture, and does not seek to uphold the judge's decision on that ground. It will often not be possible to decide, he accepts, whether unusual behaviour displayed by a dog results from an innate or an acquired characteristic. The dog was not however acting uncharacteristically when acting in accordance with its training. It was trained to bite in limited and defined circumstances. Whether or not the particular circumstances in which the bites occurred are considered, the display of a characteristic not normally found in German shepherd dogs is not established.
I too find section 2(2)(b) puzzling. As Lloyd LJ pointed out in Breeden, the wording of section 2(2) is different from that of the equivalent clause in the draft Bill submitted by the Law Commission following its report entitled "Civil Liabilities for Animals", and laid before Parliament in 1967 (Law Com. 13). In the draft Bill, the sub-section provided:
"Where damage of any kind is caused by an animal which does not belong to a dangerous species, and--
(a) the animal has such characteristics that it is likely, unless restrained, to cause damage of that kind or that any damage of that kind that it may cause is likely to be severe; and
(b) those characteristics are known or treated as known to a person who is a keeper of the animal;
that person is liable for the damage, except as otherwise provided by this Act."
It is unnecessary to speculate on the reason for the change but it is clear that it was decided to include in the section the need to establish not only that characteristics likely to cause damage are present but also that the characteristics are not normally found in the relevant species or sub-species or not normally found except at particular times or in particular circumstances. It is the departure from the characteristics of a sub-species, which is by definition not dangerous, which can create the liability (provided the uncharacteristic behaviour is causative). The section does not provide for animals, which do not belong to a dangerous species, behaving normally.
Section 2(2) was explained in this way by Lloyd LJ in Breeden:
"The essential condition for liability now is that the characteristic which is known to the owner must be a characteristic which is abnormal for the species ... If liability is based on the possession of some abnormal characteristic known to the owner, then I cannot see any sense in imposing liability when the animal is behaving in a perfectly normal way for all animals of that species in those circumstances, even though it would not be normal for those animals to behave in that way in other circumstances, for example, a bitch with pups or a horse kicking out when approached too suddenly, or too closely, or from behind."
In the last part of that paragraph Lloyd LJ was expressing a view upon the "particular circumstances" addition to section 2(2)(b) which differs from the view expressed, for example, by Slade LJ in Curtis. The present case does not appear to me to turn upon those additional words, as to the effect of which judicial views differ, but upon the earlier words "characteristics" and "not normally found". I respectfully agree with Lloyd LJ that the section is not concerned with animals behaving in a perfectly normal way for animals of the species or sub-species.
The cases have been concerned with animals displaying characteristics not normally found in the species or sub-species. In Wallace v Newton [1982] 2 All ER 106, the court was concerned with a horse which had an unpredictable temperament not normally found in horses. In Hunt v Wallis, the border collie was alleged to have acted in an unruly manner uncharacteristic of that breed of dog. In Cummings v Granger [1977] QB 397 and Curtis v Betts [1990] 1 WLR 459 the so-called second limb in section 2(2)(b) was relied on. In neither case did the dog, an Alsatian in Cummings and a bull mastiff in Curtis, have characteristics not normally found in such dogs but they did display such characteristics in the "particular circumstances" present in those cases; an untrained dog roaming a yard which it regarded as its territory and a dog with a tendency to react fiercely when defending what he regarded as his own territory. In Smith v Ainger (transcript 16 May 1990), the characteristic relied on in the case of a "large dark grey long haired Alsatian cross" was its propensity to attack other dogs.
In Livesey v Chief Constable of Lancashire (transcript 10 November 1994), the claimant, a police officer, was attacked by a German shepherd dog trained for police work. The attack occurred when the claimant was off duty and exercising the dog on the fields adjacent to his house. He could offer no explanation for the dog's conduct and had never known a dog to behave in this way.
Smith J held that a "characteristic" should include traits or habits or inclinations which are inculcated by training as well as those which are innate. Smith J continued:
"If a dog is trained to behave in particular way, it seems reasonable to infer that he is likely to behave in a different way from a dog which has not been trained in that way. I consider that Sabre did have characteristics which were different from those of the ordinary run of German shepherd dogs because he had been trained to behave in certain ways. However his training had only taught Sabre to bite either when instructed to do so or if a suspect failed to stop when he had been chased and caught up with. There is no evidence he had any tendency to bite outside those circumstances. Nor was there any evidence before me that his training would make him more likely than other German shepherd dogs to bite a person outside the circumstances in which he had been trained to bite ... I am not prepared to hold that Sabre had a general propensity to bite ... I hold that Sabre had two characteristics not normally found in German shepherd dogs: first an inculcated propensity to attack and bite after chasing a person who was running away and second an inculcated propensity to hold onto his victims arm."
Having considered Curtis, Smith J concluded:
"The damage suffered by the plaintiff was not due to Sabre's inculcated characteristic of attacking and biting humans in circumstances akin to criminal work. The attack occurred in circumstances wholly unlike criminal work. ... For the plaintiff to show that the damage was due to a characteristic of the dog, it would have to be shown that the dog had a tendency to make a sudden and unprovoked violent attack on a person. There is no evidence that Sabre had such a tendency. Thus by that approach I conclude that sub-paragraph (b) is not made out ... [The] conduct was not consistent with any characteristic he had previously displayed and was not in any way related to the characteristic which had been inculcated in him by his training."
The judge held that in those circumstances the claim failed. It appears however that Smith J might have found in the plaintiff's favour had the attack occurred in circumstances "akin to criminal work".
The question is whether the "inculcated propensity" to bite is to be treated as the characteristic for the purposes of the section. That may be said to be a characteristic not normally found in German shepherd dogs. If, on the other hand, it is the ability of the dog to respond to specific training and instruction which is the relevant characteristic, it cannot be said to be a characteristic not normally found in German shepherd dogs. Indeed, it is the hallmark of the breed.
On balance, I consider the second approach to be that intended by the statute. The particular form the training takes, which included instruction to conduct a limited form of attack in limited circumstances, is a manifestation of the characteristic rather than the characteristic itself. The acts themselves are not characteristics. Moreover, the answer to the question should not depend on the proportion of German shepherd dogs which is trained; it may be only a minority. It is the ability to respond to training which is the relevant characteristic.
It is a characteristic of German shepherd dogs that they are capable of being trained to perform certain acts under instructions. That is why they are used as police dogs and for similar purposes. The training includes instructions to conduct a limited form of attack in limited circumstances. The characteristic which is relied on as not normally found in German shepherd dogs, training to act on instructions, is a characteristic of the sub-species. Jack acted as he was trained to act and in a way characteristic of the sub-species. Section 2(2) was not intended to cover German shepherd dogs acting in accordance with their character. It was most unfortunate that, due to a combination of circumstances, the claimant was bitten. However on that basis the dog's behaviour cannot be said to be due to a characteristic not normally found in German shepherd dogs, which is the undoubted basis for liability in section 2(2)(b).
Moreover, on the judge's finding of fact, the claimant also fails on causation. The "likelihood of the damage being severe" was not due to the characteristic relied on. I have referred to the findings of fact and understand the reluctance of the judge to make a finding that the bites were damage of a kind which the dog was "likely to cause" within the meaning of the first limb of section 2(2)(a). On the evidence, the bites were a most unlikely event.
However, it is also most unlikely that when providing the alternatives in section 2(2)(a) the draftsman (or the Law Commission) had in mind a situation such as the present. It would in my view be a curious result if the claimant were to fail on this ground alone, following findings of fact properly made and assuming the meaning of characteristic more favourable to the claimant. The claimant would have established a relevant characteristic and an attack due to the existence of that characteristic but still failed on causation. I consider it unlikely that the section was intended to involve a definition of characteristic which could produce that result in circumstances such as the present.
The fact that some breeds of dog can be trained to bite may, if my view of the section is correct, leave members of the public without a remedy under the Act in some circumstances. The training of police dogs has a valuable social purpose in the prevention and detection of crime. Plainly a risk of injury to members of the public does arise from the use of dogs. Skilled training and supervision are required. The absence of a remedy under section 2(2) does not exclude a remedy in negligence and a possible remedy in assault and battery. The negligence claim failed in this case but in my view a high standard of care must be required of those who use dogs such as German shepherds for security purposes. If the training, in contrast to the present case, is for other than a legitimate purpose, claims in assault and battery may arise as well as one in negligence.
I would dismiss the appeal.
LORD JUSTICE HALE:
I have found this a most anxious and difficult case and am grateful to both counsel for the quality of their arguments. Two questions arise. First, did Jack have 'characteristics which are not normally found in animals of the same species or are not normally so found except at particular times or in particular circumstances'? Mr Turner wisely conceded that such characteristics might be innate or learnt. There are at least two good reasons for this. First, it would often be impossible to know whether an animal's unusual characteristics were the result of his character or of his early experiences. Secondly, some animals do not fall within the definition of a 'dangerous species' in section 6(2) because they are 'commonly domesticated in the British Isles' but require early training in order to become so domesticated and thus not to have 'characteristics which are not normally found in animals of the same species'. Horses are an obvious example: I venture to think that an unbroken and unschooled horse would fall within section 2(2)(b). The same may apply to some breeds of dog.
Hence I have no difficulty with the idea that the fact of training, or the particular susceptibility to training of some breeds of dog, is not such as to take an animal into the category of one with 'characteristics not normally found' . . . . But what about an animal which is trained but in a way which is unlike that of most animals of his species? The experts in this case were agreed that (a) but for its training this breed of dog would not act in the way that this dog acted on the occasion in question, and (b) this dog and other trained police dogs (of whatever breed) will proceed to search for and, if possible, detain when given the appropriate command. Clearly this is not something which other dogs are trained to do. It makes them different from other dogs, including other German shepherd dogs.
In the case of police dogs that propensity is put to a socially useful purpose, the apprehension of persons reasonably suspected of having committed arrestable offences. But it is not difficult to imagine cases in which such a propensity could be put to most anti-social purposes. Of course, if such a trained dog is deliberately set upon a quarry without lawful justification then it will be an assault (or in some circumstances the deliberate infliction of harm by indirect means under the principle in Wilkinson v Downton [1897] 2 QB 57). Of course, if such a dog escapes through carelessness then there will be liability in negligence. There may even be circumstances in which it is negligent to have such a dog at all. But what about the retired or off duty police dog (as in the unreported first instance case of Livesey v Chief Constable of Lancashire)? The whole point of section 2(2) of the 1971 Act, which replaced the old scienter action, was to impose a stricter liability upon those who knew that the animal they had possessed characteristics which other such animals did not have.
However, the second question relates to the consequences of having those characteristics. The keeper of an unusually docile or well trained animal has no such stricter liability. Section 2(2)(a) requires that the damage caused must be 'of a kind which the animal, unless restrained, was likely to cause or which, if caused by the animal, was likely to be severe'. Section 2(2)(b) then requires that 'the likelihood of the damage or of its being severe was due' to the characteristics in question.
This clearly raises two alternatives. Mr Turner argues that the second, the likelihood of the damage in this case 'being severe', was due to the size of the animal and the size of his teeth and not to the unusual characteristics inculcated by his police training. The police training might have made it more likely that Jack would bite (see below) but not that any bite he did inflict would be severe. That was due to the characteristics normal to his breed. I cannot find any fault with that logic. In Curtis v Betts [1990] 1 WLR 459, both Stuart Smith LJ and Slade LJ had difficulty with the words 'the likelihood of' in section 2(2)(b), which they thought 'inept'. Stuart Smith LJ would have excised them and concentrated simply upon the need for a causal link between the characteristic in question and the damage done. But these comments were clearly obiter and do not require this court to reject Mr Turner's argument on this point.
But what about the first alternative? Was the 'likelihood of the damage' due to the characteristics in question? Mr Turner has an easy answer to this. The judge expressly found that this was not satisfied 'as the dog was not likely to cause damage unrestrained, save where he was triggered by the police handler's action and command'. Mr Turner argues that this was a finding of fact which is not challenged in this appeal. That might depend upon whether the judge asked himself the right question. Was it simply whether 'the likelihood of the damage' was due to the characteristics in question; or was it whether 'the likelihood of damage of a kind which the animal, unless restrained, was likely to cause' was due to those characteristics? On a straightforward reading of the subsection, the words 'the damage' relate back to where they are first used, which is at the outset: 'Where damage is caused by an animal which does not belong to a dangerous species, a keeper of the animal is liable for the damage . . . if' (a), (b) and (c) are satisfied. On this reading, the likelihood of the damage in fact caused must be due to the characteristics in question. However, it would be very odd if section 2(2)(b) related back to section 2(2)(a) in one respect - the likelihood of the damage being severe - but not in the other - the damage being of a kind which the animal unless restrained was likely to cause.
Hence in my view the judge did ask himself the right question. He was also entitled to make the finding he made. The damage here was not only a bite: it was the second bite which the dog is trained to administer if the quarry does not obey the instruction to stand still. It was highly unlikely that this dog would cause the damage which he did cause except in the very unusual combination of circumstances in this case: where he had slipped his leash in circumstances where he thought that he had been set on the quarry when in fact his handler had not meant to release him until he had seen the quarry. The judge must have been right to say that the fact of being unrestrained would not have made the animal likely to cause this damage: but he would also have been right had he held that on the particular facts of this case the likelihood of this particular damage was not due to the fact of this dog's police training.
Hence, while I may express some surprise at the finding that there was no negligence in this case and considerable sympathy for the claimant's plight, I cannot disagree with the judge's finding on causation in the very unusual circumstances of this case. But in my view it would be most unwise to draw any general conclusions as to the potential liability of those who, for whatever reason, choose to keep dogs who are trained to attack human beings.
It follows that I agree in the result but because the case does not fall within section 2(2)(a) of the 1971 Act rather than because it does not fall within section 2(2)(b).
Order: Appeal dismissed with costs.
(order does not form part of the approved judgment)


BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2000/90.html