[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 >> Commission for Racial Equality v Dutton [1988] EWCA Civ 17 (27 July 1988) URL: http://www.bailii.org/ew/cases/EWCA/Civ/1988/17.html Cite as: [1989] IRLR 8, [1989] QB 783, [1989] 1 All ER 306, [1989] 2 WLR 17, [1988] EWCA Civ 17

[New search] [Buy ICLR report: [1989] QB 783] [Buy ICLR report: [1989] 2 WLR 17] [Help]

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE WESTMINSTER COUNTY COURT
(His Honour Judge Harris)

		Royal Courts of Justice
		27th July 1988

1. LORD JUSTICE NICHOLLS: This case concerns gipsies, and whether they are a racial group within the Race Relations Act 1976 . The Mr. Patrick Dutton, has been the licensee of the Cat and Mutton public house at Broadway Market, London Fields, London E.8 for four years. Previously he was the licensee for a year at the Earl of Beaconsfield, in Southwark, and before that he was at the Lord Cecil, in Clapton, from 1978. At both those houses Mr. Dutton had unpleasant experiences with people who came from caravans which were parked illegally on nearby sites. They caused damage. They threatened Mr. Dutton and terrorised his wife. They behaved generally in such a way as to upset the Mr. Dutton's regular customers. So much so that, after such incidents, he put up a sign in the windows of the Earl of Beaconsfield and, subsequently, the Lord Cecil, which read "no travellers." By that he meant, as he said in evidence, a person who travels around in a caravan and parks on illegal sites and gives him "hassle." He wanted only to stop such people coming into his public house. Had the incidents continued he would have lost all his customers. After he put up the signs he had no more problems with such people.
2. One weekend, after Mr. Dutton had been at the Cat and Mutton for about 18 months, some 15 or so caravans parked opposite the public house on London Fields, illegally, about 150 yards away. On Sunday morning some of these "travellers" came into the Cat and Mutton. Mr. Dutton refused to serve one of them on the ground he was from the site. There was an incident. Mr. Dutton then put up handwritten signs in the windows of the door of the Cat and Mutton: "Sorry, no travellers." Since then he has had no more trouble with "travellers."
3. In June 1985 a local resident, who does not use the Cat and Mutton brought these signs to the attention of the Commission for Racial Equality. The commission took the view that the signs discriminated against gipsies. After correspondence this action was brought by the commission, in the exercise of its functions under section 63 . The commission seeks a declaration that by displaying the signs Mr. Dutton has contravened section 29 of the Race Relations Act 1976 and an injunction restraining him from continuing to display the signs. The action was heard by Judge J. P. Harris, Q.C. at the Westminster County Court. He sat with two assessors appointed from the list maintained by the Secretary of State under section 67(4) of persons who appear to him to have special knowledge and experience of problems connected with relations between persons of different racial groups. On 29 June 1987 the judge dismissed the action. The commission has appealed from that decision.

The statute

4. Section 29(1) provides:

"It is unlawful to publish or to cause to be published an advertisement which indicates, or might reasonably be understood as indicating, an intention by a person to do an act of discrimination, whether the doing of that act by him would be lawful or, by virtue of Part II or III, unlawful."

5. Discrimination for the purposes relevant in the present case is defined in section 1(1) . Two types of conduct are within the definition. Paragraph (a) defines what is generally known as "direct" discrimination, although not so called in the Act, as follows:

"(1) A person discriminates against another in any circumstances relevant for the purposes of any provision of this Act if –

(a) on racial grounds he treats that other less favourably than he treats or would treat other persons; ..."

6. Section 3(1) defines "racial grounds" as meaning any of the following grounds, namely colour, race, nationality or ethnic or national origins.
7. Mr. Dutton's notices are advertisements within the definition in section 78(1) . Further, they indicate an intention by him to treat "travellers" less favourably than he treats other persons, in circumstances relevant for the purposes of the Act. By the notices he is informing would-be customers that he will not serve any who are "travellers." They cannot use the Cat and Mutton. That is discrimination in the provision of goods, facilities or services: see section 20 . Thus arises the question on this part of the case: is that discrimination "on racial grounds?" On this, the first issue to be considered is the meaning of the expression "travellers" in the context in which the signs are being displayed.
8. The commission's case was that in these notices "travellers" is synonymous with gipsies. Before the judge there was material supporting the contention that, in recent years, the two expressions are sometimes used interchangeably. For example, a new edition of Chambers 20th Century Dictionary published in 1983, includes under the word "travel" the subheading "traveller:"

"travelling folk, people: the name by which itinerant people often call themselves, in preference to the derogatory names gipsies or tinkers."

9. Again, the Supplement to the Oxford English Dictionary (1986), added a further meaning to the word "traveller: ... Also, a gypsy." There was also evidence to the same effect, from two expert witnesses, Dr. Donald Kenrick and Dr. Thomas Acton.

The meaning of "gipsy"

10. Notwithstanding this material the judge rejected the view that the words are synonymous. I agree with him. But before proceeding further it is necessary for me to comment on the word "gipsy." One of the difficulties in the present case, in my view, is that the word "gipsy" has itself more than one meaning. The classic "dictionary" meaning can be found as the primary meaning given in the Oxford English Dictionary, (1933):

"A member of a wandering race (by themselves called Romany ), of Hindu origin, which first appeared in England about the beginning of the 16th century. and was then believed to have come from Egypt."

11. Hence the word "gipsy," also spelled as "gypsy." It is a corruption of the word Egyptian. We find this usage in Shakespeare, where Othello says to Desdemona (Act III, scene IV):

"That handkerchief

Did an Egyptian to my mother give.

She was a charmer, and could almost read

The thoughts of people."

12. Alongside this meaning, the word "gipsy" also has a more colloquial, looser meaning. This is expressed in the Longman Dictionary of Contemporary English, where two meanings are attributed to "gipsy." The first meaning is along the lines I have already quoted. The second is this:

"a person who habitually wanders or who has the habits of someone who does not stay for long in one place."

In short, a nomad.

13. I can anticipate here by noting that if the word "gipsy" is used in this second, colloquial sense it is not definitive of a racial group within the Act. To discriminate against such a group would not be on racial grounds, namely, on the ground of ethnic origins. As the judge observed, there are many people who travel around the country in caravans, vans, converted buses, trailers, lorries and motor vehicles, leading a peripatetic or nomadic way of life. They include didicois, mumpers, peace people, new age travellers, hippies, tinkers, hawkers, self-styled "anarchists," and others, as well as (Romany) gipsies. They may all be loosely referred to as "gipsies," but as a group they do not have the characteristics requisite of a racial group within the Act.
14. I give two further illustrations of this point. First, an extract from a report of the Greater London Conciliation Committee, set out in Appendix III to the Report of the Race Relations Board for 1967-68. This refers neatly to a difficulty arising in this field from the two meanings I have mentioned:

"There are the pubs who discriminate against gipsies. In tackling this problem the committee has been hampered by two ambiguities. There is, first, some doubt as to the status of gipsies under the Act. The committee feels that there is little or no justification for this doubt, but equally believe that it persists and that it does so largely because of the second ambiguity. This second ambiguity arises out of common parlance, for it seems that the word "gipsy" is used to designate wanderers generally as opposed to ethnic gipsies. The committee is, therefore, trying to prevent discrimination against gipsies in the one (proper) sense while being aware that it may not interfere with discrimination against gipsies in the other (vulgar) sense, and it is in a weak position in any argument with a publican about which way he uses the words."

Secondly, the decision of the Queen's Bench Divisional Court in Mills v. Cooper [1967] 2 Q.B. 459 . The court was there concerned with the meaning of the word "gipsy" in section 127 of the Highways Act 1959 . So far as material the section provides:

"If, without lawful authority or excuse ...

(c) a hawker or other itinerant trader or a gipsy pitches a booth, stall or stand, or encamps on a highway, he shall be guilty of an offence ..."

15. In Mills v. Cooper it was argued that the word "gipsy" should be "given its dictionary meaning, as being 'a member of the Romany race'." Lord Parker C.J. said (at p. 467):

"That a man is of the Romany race is, as it seems to me, something which is really too vague of ascertainment, and impossible to prove; moreover it is difficult to think that Parliament intended to subject a man to a penalty in the context of causing litter and obstruction on the highway merely by reason of his race. I think that in this context 'gipsy' means no more than a person leading a nomadic life with no, or no fixed, employment and with no fixed abode. In saying that, I am hoping that those words will not be considered as the words of a statute, but merely as conveying the general colloquial idea of a gipsy."

Likewise Diplock L.J.:

"I agree that the word 'gipsy' as used in section 127 of the Highways Act 1959 , cannot bear its dictionary meaning of 'a member of a wandering race (by themselves called Romany) of Hindu origin. ...' If it did it would mean that Parliament in 1959 had amended the corresponding section of the Highway Act 1935 (which referred to 'gipsy or other person'), so as to discriminate against persons by reason of their racial origin alone."

16. In the context provided by this difficulty, and the impossibility of ever being able to prove pure Romany origin, Lord Diplock preferred what he described (at p. 468) as the popular meaning of the word "gipsy:"

"a person without fixed abode who leads a nomadic life, dwelling in tents or other shelters, or in caravans or other vehicles."

17. The substance of that definition was then adopted by Parliament in the Caravan Sites Act 1968 . Section 6 of that Act imposed on local authorities a duty to exercise their powers, under the Caravan Sites and Control of Development Act 1960 , to provide caravan sites "so far as may be necessary to provide adequate accommodation for gipsies residing in or resorting to their area." Section 16 provides that:

"'gipsies' means persons of nomadic habit of life, whatever their race or origin, but does not include members of an organised group of travelling showmen, or of persons engaged in travelling circuses, travelling together as such."

18. I shall return at a later stage to the relevance of those statutory provisions. For the moment it is sufficient to note that there is ambiguity in the word "gipsy," and when considering reports and other material about gipsies it is essential therefore to identify what is the meaning with which the author is using the word "gipsy." In this judgment, save where I indicate otherwise, I shall henceforth use the word "gipsy" in the narrower sense, of the first of the two meanings mentioned above.

"no travellers"

19. I can now state my reasons for agreeing with the judge's conclusion on the "direct" discrimination issue. Like most English words, the meaning of the word "traveller" depends on the context in which it is being used. It has one meaning when seen on a railway station. For some time now the refreshment service provided at railway stations and on trains has been styled "Travellers Fare." The word has a different meaning when in its context it is directed at travelling salesmen. In my view, in the windows of the Cat and Mutton "no travellers" will be understood by those to whom it is directed, namely, potential customers, as meaning persons who are currently leading a nomadic way of life, living in tents or caravans or other vehicles. Thus the notices embrace gipsies who are living in that way. But the class of persons excluded from the Cat and Mutton is not confined to gipsies. The prohibited class includes all those of a nomadic way of life mentioned above. As the judge said, they all come under the umbrella expression "travellers," as this accurately describes their way of life.
20. It is estimated that nowadays between one-half and two-thirds of gipsies in this country have wholly or largely abandoned a nomadic way of life, in favour of living in houses. I do not think that the notices could reasonably be understood as applying to them, that is, to gipsies who are currently living in houses. Gipsies may prefer to be described as "travellers" as they believe this is a less derogatory expression. But, in the context of a notice displayed in the windows of a public house near a common on which nomads encamp from time to time, I do not think "no travellers" can reasonably be understood as other than "no nomads." It would not embrace house-dwellers, of any race or origin.
21. For this reason I cannot accept that Mr. Dutton's notices indicate, or might reasonably be understood as indicating, an intention by him to do an act of discrimination within section 1(1)(a) . Excluded from the Cat and Mutton are all "travellers," whether or not they are gipsies. All "travellers," all nomads, are treated equally, whatever their race. They are not being discriminated against on racial grounds.

Indirect discrimination: a racial group

22. That suffices to dispose of the claim based on section 1(1)(a), but that is not the end of the action. I must now turn to consider section 1(1)(b), which is in the following terms:

"(1) A person discriminates against another in any circumstances relevant for the purposes of any provision of this Act if –

….

(b) he applies to that other a requirement or condition which he applies or would apply equally to persons not of the same racial group as that other but –

(i) which is such that the proportion of persons of the same racial group as that other who can comply with it is considerably smaller than the proportion of persons not of that racial group who can comply with it; and

(ii) which he cannot show to be justifiable irrespective of the colour, race, nationality or ethnic or national origins of the person to whom it is applied; and

(iii) which is to the detriment of that other because he cannot comply with it."

23. On this the first question which arises is whether gipsies are a racial group. If they are not, paragraph (b) cannot apply to Mr. Dutton's notices. He cannot apply to a gipsy who wishes to have a drink at the Cat and Mutton a condition, in this case, of not being a traveller, which he applies equally to persons "not of the same racial group" unless gipsies are a racial group within the Act. Indeed, if gipsies are not a racial group, a notice saying "No gipsies" would be lawful.
24. The definition of "racial group" in section 3(1) includes a group of persons defined by reference to "ethnic ... origins." This definition was considered by the House of Lords in Mandla v. Dowell Lee [1983] 2 AC 548 There the context was whether Sikhs constituted a group defined by reference to ethnic origins. Lord Fraser of Tullybelton observed (at p. 563) that the word "ethnic" in the Act should be construed relatively widely, in a broad cultural/historic sense. He approved the following passage from the judgment of Richardson J. sitting in the New Zealand Court of Appeal in King-Ansell v. Police [1979] 2 N.Z.L.R. 531 , 543:

"a group is identifiable in terms of its ethnic origins if it is a segment of the population distinguished from others by a sufficient combination of shared customs, beliefs, traditions and characteristics derived from a common or presumed common past, even if not drawn from what in biological terms is a common racial stock. It is that combination which gives them an historically determined social identity in their own eyes and in the eyes of those outside the group. They have a distinct social identity based not simply on group cohesion and solidarity but also on their belief as to their historical antecedents."

25. Lord Fraser (at p.562) summarised his opinion on the construction of the Act in his own words:

"For a group to constitute an ethnic group in the sense of the Act of 1976, it must, in my opinion, regard itself, and be regarded by others, as a distinct community by virtue of certain characteristics. Some of these characteristics are essential; others are not essential but one or more of them will commonly be found and will help to distinguish the group from the surrounding community. The conditions which appear to me to be essential are these:

(1) a long shared history, of which the group is conscious as distinguishing it from other groups, and the memory of which it keeps alive;

(2) a cultural tradition of its own, including family and social customs and manners, often but not necessarily associated with religious observance. In addition to those two essential characteristics the following characteristics are, in my opinion, relevant;

(3) either a common geographical origin, or descent from a small number of common ancestors;

(4) a common language, not necessarily peculiar to the group;

(5) a common literature peculiar to the group;

(6) a common religion different from that of neighbouring groups or from the general community surrounding it;

(7) being a minority or being an oppressed or a dominant group within a larger community, for example a conquered people (say, the inhabitants of England shortly after the Norman conquest) and their conquerors might both be ethnic groups.

A group defined by reference to enough of these characteristics would be capable of including converts, for example, persons who marry into the group, and of excluding apostates. Provided a person who joins the group feels himself or herself to be a member of it, and is accepted by other members, then he is, for the purposes of the Act, a member. That appears to be consistent with the words at the end of section 3(1): 'references to a person's racial group refer to any racial group into which he falls.' In my opinion, it is possible for a person to fall into a particular racial group either by birth or by adherence, and it makes no difference, so far as the Act of 1976 is concerned, by which route he finds his way into the group."

26. In the present case the judge expressed his conclusion on the conditions enunciated by Lord Fraser in this way:

"It may well be, as I have said, that there is a small number of travelling people who can claim either by looks or characteristics to be true gipsies but these people have been absorbed into a larger group. Some have abandoned the nomadic way of life and some are indistinguishable from any ordinary member of the public. The larger group of travellers or gipsies forming a part of a larger group cannot in my judgment on the evidence before the court satisfy those two essential conditions and can satisfy barely any of the other five conditions. Although there may be a Romany language, some may be able to trace their ancestry back to people who came to England many hundreds of years ago, the language does not seem to be in general use. There is no common religion, they have no literature. Although it was urged on the court that there should be some relevance in the fact that they have what was described as oral literature passing on myths and other old stories I do not think that was what Lord Fraser was referring to."

He decided that gipsies were not a group defined by reference to ethnic origins.

27. I come here to a further difficulty about the present case. The evidence on this part of the case consisted principally of evidence called by the commission: the two experts I have mentioned, and a Mr. Peter Mercer, who is a gipsy. No expert evidence was led by Mr. Dutton. But although there was no contrary evidence called by Mr. Dutton, the judge was not impressed by either of the commission's expert witnesses. He approached their evidence with much caution and doubt. Mr. Sedley criticised the judge's comments in this regard, but on this the judge's advantage, of having seen and heard the witnesses, is obviously of paramount importance. We are not in a position to conclude that the judge erred in his assessment of the reliability of these witnesses.
28. Nevertheless, taking the judge's assessment of the witnesses fully into account, and with all respect to the judge, I am unable to agree with his conclusion on what have been called the Mandla conditions when applied, not to the larger, amorphous group of "travellers" or "gipsies," colloquially so-called, but to "gipsies" in the primary, narrower sense of that word. On the evidence it is clear that such gipsies are a minority, with a long-shared history and a common geographical origin. They are a people who originated in northern India. They migrated thence to Europe through Persia in medieval times. They have certain, albeit limited, customs of their own, regarding cooking and the manner of washing. They have a distinctive, traditional style of dressing, with heavy jewellery worn by the women, although this dress is not worn all the time. They also furnish their caravans in a distinctive manner. They have a language or dialect, known as "pogadi chib," spoken by English gipsies (Romany chals) and Welsh gipsies (Kale) which consists of up to one-fifth of Romany words in place of English words. They do not have a common religion, nor a peculiar, common literature of their own, but they have a repertoire of folktales and music passed on from one generation to the next. No doubt, after all the centuries which have passed since the first gipsies left the Punjab, gipsies are no longer derived from what, in biological terms, is a common racial stock, but that of itself does not prevent them from being a racial group as widely defined in the Act.
29. I come now to the part of the case which has caused me most difficulty. Gipsies prefer to be called "travellers" as they think that term is less derogatory. This might suggest a wish to lose their separate, distinctive identity so far as the general public is concerned. Half or more of them now live in houses, like most other people. Have gipsies now lost their separate, group identity, so that they are no longer a community recognisable by ethnic origins within the meaning of the Act? The judge held that they had. This is a finding of fact.
30. Nevertheless, with respect to the judge, I do not think that there was before him any evidence justifying his conclusion that gipsies have been absorbed into a larger group, if by that he meant that substantially all gipsies have been so absorbed. The fact that some have been so absorbed and are indistinguishable from any ordinary member of the public, is not sufficient in itself to establish loss of what Richardson J., (in the King-Ansell case, at p.543) referred to as "an historically determined social identity in [the group's] own eyes and in the eyes of those outside the group." There was some evidence to the contrary from Mr. Mercer, on whose testimony the judge expressed no adverse comment. He gave evidence that "we know who are members of our community" and that "we know we are different." In my view the evidence was sufficient to establish that, despite their long presence in England, gipsies have not merged wholly in the population, as have the Saxons and the Danes, and altogether lost their separate identity. They, or many of them, have retained a separateness, a self-awareness, of still being gipsies.
31. I feel less constrained than otherwise I would to depart from the judge's conclusions on this point because of the importance attached by him to the meaning borne by the word "gipsy" in the Highways Act 1959 and the Caravan Sites Act 1968 . He said this:

"Although the Highways Act 1959 and the Caravan Sites Act 1968 are statutory examples of the use of the word "gipsy" the meaning given to the word in those Acts does have great weight in my mind. If you find a word defined in a definition section of one Act of Parliament and defined by the Divisional Court on another use of the same word in another statute it would be difficult to say: well when you are looking at the Race Relations Act 1976 you must have a wholly and totally different meaning attached to it. I consider, agreeing as I do with the Divisional Court in Mills v. Cooper, that it would be impossible to discover if any person or any body of persons were members of the Romany race or true gipsies. It is not difficult to discover whether they are leading a nomadic life, whether they are travelling from place to place with no fixed abode and no fixed employment. But having ascertained these matters one might justifiably come to the conclusion that they being travellers were not clearly gipsies. As I say I do not think one can be a gipsy or a non-gipsy in one statute and not in another."

32. In my view those two statutes do not materially assist in the present case, and the judge misdirected himself on this point. The present case is quite different from Mills v. Cooper. In the present case the issue is not which of two or more meanings of the word "gipsy" is to be preferred in the context of a particular statute or document. The question is whether there is an identifiable group of persons, traditionally called "gipsies," who are defined by reference to ethnic origins. That is essentially a question of fact, to be determined on the evidence, applying the approach set out in Mandla case. On that question the definition of "gipsy" used in the Caravan Sites Act 1968 , and the meaning of the word "gipsy" in the Highways Act 1959 as interpreted in Mills v. Cooper, are of little assistance, if any. Furthermore the difficulty, mentioned in Mills v. Cooper (at pp. 467, 468) of determining today whether a person is of "the Romany race" or is of "pure Romany descent" or "Romany origin," seems to have led the judge into thinking that that difficulty constituted an obstacle to the commission's success in the present case. But that is not so. The material provision in the Act of 1976 is concerned with ethnic origins, and "ethnic" is not used in that Act in a strictly biological or racial sense. That was decided in Mandla.
33. In my view, accepting the judge's doubts about the evidence of Dr. Kenrick and Dr. Acton, the evidence was still sufficient to establish that gipsies are an identifiable group of persons defined by reference to ethnic origins within the meaning of the Act.

Indirect discrimination: adverse impact

34. Having concluded that gipsies are a racial group, each of the subparagraphs (i) - (iii) in section 1(1)(b) must be satisfied before the conduct complained of amounts to discrimination within the meaning of the Act. I shall consider the three sub-paragraphs one by one, starting with sub-paragraph (i).
35. Clearly the proportion of gipsies who will satisfy the "no travellers" condition is considerably smaller than the proportion of non-gipsies. Of the estimated gipsy population in the United Kingdom of some 80,000, between one-half and two-thirds now live in houses. But this still means that a far higher proportion of gipsies are leading a nomadic way of life than the rest of the population in general or, more narrowly, than the rest of the population who might wish to resort to the Cat and Mutton.
36. Mr. Samuels submitted that the word "can" in the expression "can comply" in sub-paragraph (i) means "can comply without giving up the distinctive customs and cultural rules of gipsies." He submitted that gipsies can cease to be nomadic, and become house-dwellers, and comply with the "no travellers" condition, without giving up their customs and culture and that, therefore, sub-paragraph (i) is not satisfied in this case. I do not accept this. Lord Fraser's words in the Mandla case (at p.566) which Mr. Samuels embraced in this submission, were used in the context of a "no turban" condition being applied in relation to a Sikh. Lord Fraser was rejecting the submission that "can" meant "can physically." But that does not assist the solution of the present case. Indeed, gipsies can and do cease to be nomadic, but that will be of little use to a particular nomadic gipsy when he chances upon the Cat and Mutton and wishes to go in for a drink. At that stage he is, in practice, unable to comply. In the present case the problem is a different one: at what moment of time does ability to comply fall to be judged? Is it when the condition is invoked (in this case, when the gipsy is outside the public house wishing to enter)? or is it at some earlier date (which would give the gipsy sufficient opportunity to acquire housing accommodation for himself before turning up at the Cat and Mutton)?
37. A similar question was considered by the Employment Appeal Tribunal in Clarke v. Eley (IMI) Kynoch Ltd. [1983] ICR 165 , with regard to section 1(1) of the Sex Discrimination Act 1975 , the wording of which does not differ materially from section 1(1)(b) of the Act of 1976. Browne-Wilkinson J. delivered the judgment of the tribunal, to the effect that the relevant point of time at which the ability or inability to comply has to be shown is the date at which the requirement or condition has to be fulfilled. I find his reasoning, set out at pp. 171-172, compelling, and I agree with his conclusion.
38. In my view, therefore, sub-paragraph (i) is satisfied in the present case.

Indirect discrimination: detriment

39. Sub-paragraph (iii) requires the applied condition to be to the relevant person's detriment because he cannot comply with it. Rightly, it was not disputed that sub-paragraph (iii) is satisfied in the present case, by the hypothetical nomad gipsy being excluded from the Cat and Mutton: I say hypothetical, because there was no evidence that (there were any gipsies amongst the travellers on the nearby sites).

Indirect discrimination: justification

40. I have left sub-paragraph (ii) to the end, for this reason. On the admitted or proved facts it is possible for this court to decide whether sub-paragraphs (i) and (iii) are satisfied, even though the judge himself did not decide these points. Sub-paragraph (ii) is different. On the facts before us it would not be satisfactory for this court to attempt to decide this point, which the judge expressly left open.
41. In these circumstances for my part I would remit the action to the county court for the judge to determine whether section 1(1)(b)(ii) is satisfied in the present case and, if it is, for him to make such order as he considers appropriate. I would allow this appeal accordingly.
42. LORD JUSTICE TAYLOR: I agree. The commission's case under section 1(1)(a) of the Race Relations Act 1976 must fail for the reasons given by my Lord. The word "traveller" is not synonymous with the word "gypsy."
43. The case under section 1(1)(b) turns essentially on whether gipsies are a racial group within the meaning of the Act. It is only on that issue that I wish to add some observations.
44. The learned judge considered four different approaches to the issue. First he said he would consider the evidence of the two expert witnesses called on behalf of the commission and their gipsy witness Mr. Mercer. In fact, he made no reference to Mr. Mercer in dealing with this first approach. Clearly, the two experts made an unfavourable impression on the learned judge who described their views as wholly obsessive, biased and totally preconceived. In particular, he found Dr. Acton to be a very bad witness in that he could not be contained within the ordinary question and answer routine. The learned judge summarised his conclusions on this approach (i.e. via the commission's witnesses) by saying merely that he approached the experts "with much caution and doubt" and did not consider "that their evidence overrides or displaces the views" he later set out. But, however difficult or partisan the experts may have been, it was surely an excessive reaction to reject their evidence altogether. The historical account they gave of gypsies, their origins and customs was not contradicted by any other evidence. Furthermore, Mr. Mercer gave first hand evidence confirming much of what they said and the learned judge made no finding adverse to his qualities as a witness.
45. The second approach was headed by the learned judge "Statutes and statements in Government Reports on the legal interpretation of the word 'Gipsy'."
46. It is clear that the word gypsy bears at least two broad meanings. Historically it referred to:

"a member of a wandering race (by themselves called "Roman"), of Hindu origin, which first appeared in England about the beginning of the 16th c. and was then believed to have come from Egypt:" see The Oxford English Dictionary .

47. More recently it has come to mean "a nomad." The latter meaning has been adopted in certain statutes. Thus, as the judge noted, the word gipsy in section 127 of the Highways Act 1959 was held to have "the colloquial meaning of a person leading a nomadic life with no fixed employment and with no fixed abode:" (Mills v. Cooper [1967] 2 Q.B. 459 , 466). Again, section 16 of the Caravan Sites Act 1968 defines gypsies as meaning "Persons of nomadic habit of life, whatever their race or origin, but does not include members of an organised group of travelling showmen, or of persons engaged in travelling circuses, travelling together as such." Those statutes, however, have nothing whatsoever to do with race relations or discrimination. They are concerned with highways and the provision and regulation of caravan sites. The statutory adoption of the second broad meaning of gypsy in those contexts cannot be taken to consign a racial group called gypsies to oblivion if it still exists in fact. I therefore agree that the learned judge misdirected himself in relying on the statutory meaning of "gypsy" in contexts quite different from that of the present case. At page 9 of his judgment he said:

"Although the Highways Act 1959 and the Caravan Sites Act 1968 are statutory examples of the use of the word "gipsy" the meaning given to the word in those Acts does have great weight in my mind. If you find a word defined in a definition section of one Act of Parliament and defined by the Divisional Court on another use of the same word in another statute it would be difficult to say: well when you are looking at the Race Relations Act 1976 you must have a wholly and totally different meaning attached to it. I consider, agreeing as I do with the Divisional Court in Mills v. Cooper, that it would be impossible to discover if any person or any body of persons were members of the Romany race or true gipsies. It is not difficult to discover whether they are leading a nomadic life, whether they are travelling from place to place with no fixed abode and no fixed employment. But having ascertained these matters one might justifiably come to the conclusion that they being travellers were not clearly gipsies. As I say I do not think one can be a gipsy or a non-gipsy in one statute and not in another."

That approach assumes "gypsy" must have the same meaning in all contexts and fails to identify the two different meanings mentioned above. In fact, the word gypsy does not occur in section 1 of the Race Relations Act 1976 . The phrase which has to be construed is a "Racial Group" as defined in section 3(1) and as interpreted in Mandla v. Dowell Lee [1983] 2 AC 548 , 560.

48. Confusion of the two meanings of gipsy continued when the judge came to refer to certain reports which had been put before him. He quoted from the report of the Greater London Conciliation Committee (the passage referred to in Lord Justice Nicholl's judgment). He went on to cite a report "Accommodation for Gipsies," (1976), by Mr. John Cripps for the Department of the Environment and the Welsh Office. However, they were not dealing with the same subject matter. The first report was concerned with the type of problem presented by the present case. The second was clearly made in the context of provision of caravan sites and defined the word gipsy in that context as, in effect, a person of "nomadic habit of life."
49. The third approach was to examine dictionary definitions. Here, the judge expressed the view that one could pick and choose the meaning one wished to find. He cited six definitions ranging from "A member of a dark haired race which may be of Indian origin etc. " through the broader meaning of "A person who habitually wanders" to the merely abusive "Cunning rogue." Having set out those definitions the learned judge said "Accordingly in my judgment the plaintiffs cannot really derive any assistance from dictionary definitions. People obtaining the meaning from the dictionary could not think that a gipsy was a member of a racial group or had basic ethnic origins." Here, I do not follow the learned judge's reasoning. The fact that dictionaries give more than one meaning for the word gipsy does not prevent the word from having, at any rate in some contexts, the meaning given in four out of six of the definitions.
50. Finally the learned judge considered the approach laid down in Mandla v. Dowell Lee [1983] 2 AC 548 It is important first to emphasise that ethnic origin is not now limited to or to be equated with strict racial or biological origins. Lord Fraser said, at pp. 561:

"My Lords, I recognise that 'ethnic' conveys a flavour of race but it cannot, in my opinion, have been used in the Act of 1976 in a strictly racial or biological sense. For one thing, it would be absurd to suppose that Parliament can have intended that membership of a particular racial group should depend upon scientific proof that a person possessed the relevant distinctive biological characteristics (assuming that such characteristics exist). The practical difficulties of such proof would be prohibitive, and it is clear that Parliament must have used the word in some more popular sense.

….

In my opinion, the word 'ethnic' still retains a racial flavour but it is used nowadays in an extended sense to include other characteristics which may be commonly thought of as being associated with common racial origin."

51. Lord Fraser then set out at page 562 what he described as the two essential conditions followed by five other relevant characteristics. The passage has already been cited in full by Nicholls L.J. The learned judge's (Lord Parker C.J) application of the Mandla tests is set out at page 20 of this judgment, he said:

"It may well be, as I have said, that there is a small number of travelling people who can claim either by looks or characteristics to be true gipsies but these people have been absorbed into a larger group. Some have abandoned the nomadic way of life and some are indistinguishable from any ordinary member of the public. The larger group of travellers or gipsies forming a part of a larger group cannot, in my judgment, on the evidence before the court satisfy those two essential conditions and can satisfy barely any of the other 5 conditions."

52. With respect to the judge, the fact that some gypsies, even a substantial proportion, have abandoned the nomadic way of life or have become assimilated in the general public is not decisive of the issue. There are no doubt other religious, racial or ethnic groups whose numbers diminish due to intermarriage, lack of adherence to the group or lapsed observance. But if there remains a discernible minority which does adhere it may still be a racial group within Lord Fraser's criteria.
53. On the evidence, and perhaps that of Mr. Mercer in particular, there is still a discernible group of gypsies with "a long shared history of which the group is conscious as distinguishing it from other groups, and the memory of which it keeps alive." There may well be individuals on the borderline between membership and assimilation whom it might be difficult to classify, but that does not deny the existence of the group. Likewise, the fact that some of those within the group prefer to call themselves travellers rather than gypsies is not indicative of whether a discrete racial group has ceased to exist.
54. As to Lord Fraser's second essential characteristic, I agree that the evidence summarised by Nicholls L.J. does show gypsies have a cultural tradition of their own including family and social customs and manners.
55. Accordingly, I conclude that the four approaches rightly identified by the learned judge, ought to have led him to a different conclusion from the one he reached.
56. I too would remit the action to the county court for the judge to determine whether section 1(1)(b)(ii) is satisfied and to make the appropriate order. I would allow this appeal.
57. LORD JUSTICE STOCKER: I have had the benefit of reading in draft the judgments of my Lords. I agree with their conclusions that the matter should be remitted for determination under section 1(1)(b)(ii), though for my part I have entertained considerable doubt on one aspect of this matter in the light of some of the findings of fact made by the judge.
58. I agree that in order to succeed under section 1(1)(a) - direct discrimination - the commission had to prove that the words "Traveller" and "Gypsy" were synonymous. For the reasons given by Nicholls L.J., they plainly are not and the claim under this subsection must fail.
59. So far as indirect discrimination is concerned the judge dealt with the question whether or not gypsies were of "ethnic origin" and thus a racial group in general terms in his consideration of direct discrimination and applied his findings that they were not to indirect discrimination and dismissed the claim under 1(1)(b) in a single paragraph. There is no reason why he should not have adopted this approach but it seems to me that it may have concealed the problem that "Gypsy" is used in two different senses. On the one hand it is used to embrace the category which may be described as "true gypsies" - i.e. one who is, or believes he is, of Romany descent, or by long established adherence is a member of that class. On the other, it embraces all those of nomadic habit and disposition.
60. I agree that the judge's reasoning whereby he concluded that gypsies, in the strict sense, are not an ethnic group was in many respects flawed. I say nothing of his rejection of the evidence of the expert witnesses - he saw and heard them and was entitled to regard their evidence with doubt and caution - though their evidence on the historical migration, settlement, and customs of gypsies was not the subject of any evidence to the contrary. I also agree with my Lords that dictionary definitions can support a conclusion either way on the essential question of the ethnic origins of gypsies where the word is capable of the two distinct meanings referred to, but this fact does not itself resolve the problem raised on this appeal. I further agree that contrary to the judge's finding no assistance is to be derived from the meaning of the word "gypsy" for the purpose of section 127 of the Highways Act 1959 , or the Caravan Sites Act 1968 , since both statutes would be unworkable in practice if "gypsy" for the purposes of those Acts were to be defined in the strict sense. The definition accorded to the word for the purpose of the Highways Act 1959 in Mills v. Cooper [1967] 2 Q.B. 459 and the definition in the Caravan Sites Act 1968 , do not assist at all when the issue under the Race Relations Act 1976 is whether or not the word "gypsy" for the purpose of that Act imports the conception of "ethnic origin." I refer hereunder to the dicta of Lord Parker C.J., in that case in a different context. Before considering the fourth basis of the judge's reasoning, his application of the decision of the House of Lords in Mandla v. Dowell Lee & Ors. [1983] 2 AC 548 - I think it convenient to refer to four passages in his judgment which are findings of fact made by him. He said:

"There is, in my judgment, no easily identifiable group of gypsies as there were Sikhs. The evidence is that persons who had hitherto regarded themselves as true gypsies or Romanies no longer wish to be known as gypsies because they think that is pejorative and they wish to adhere to a larger amorphous group known apparently as 'Travellers'."

61. At page 14 the judge expressed his findings in these terms:

"Gypsies may, as I have said, be a part of a group of travelling people, they may well be accurately called 'Travellers,' but they themselves do not, in my judgment, form any clearly identifiable group."

At page 15 the judge said:

"It may well be, as I have said, that there is a small number of travelling people who can claim either by looks or characteristics to be true gypsies but these people have been absorbed into a larger group. Some have abandoned the nomadic way of life and some are indistinguishable from any ordinary member of the public. The larger group of travellers or gypsies forming a part of a larger group cannot, in my judgment, on the evidence before the court satisfy those two essential conditions and can satisfy barely any of the other five conditions."

A little later, on the same page, the judge said:

"I am wholly satisfied that the group, whether you call them gypsies or travellers, are not a group forming a racial group referred to by reference to their ethnic origins as provided by section 3(1) of the Act of 1976."

62. If the reference in the earlier passage to the larger group is a reference to the wide meaning of the word "gypsy" and the small number to the word in its strict sense then this conclusion may beg the question rather than answering it. Finally the judge adopted as a finding of fact in the instant case the dicta of Lord Parker C.J. in Mills v. Cooper [1967] 2 Q.B. 459 , 467:

"That a man is of the Romany race is, as it seems to me, something which is really too vague of ascertainment, and impossible to prove."

63. My hesitation arises from the conclusion to be drawn from these findings, if justified on the facts, in the context of Mandla, read in its full context. It seems to me clear from the speeches of Lord Fraser of Tullybelton and Lord Templeman that the fact alone that a group may comply with all or most of the relevant criteria does not itself establish that such a group is of ethnic origin. Examples of such groups which might comply with the criteria but which would not be of ethnic origin were cited by the respondent at p. 555D of the report. Indeed, such groups might themselves be of multi-racial composition where no question of racial discrimination on the grounds of ethnic origin could possibly arise though many of the criteria could apply to them. No doubt there are many other examples. At page 561, Lord Fraser of Tullybelton said:

"... I recognise that 'ethnic' conveys a flavour of race but it cannot, in my opinion, have been used in the Act of 1976 in a strictly racial or biological sense."

64. and at page 562D he says:

"….the word 'ethnic' still retains a racial flavour but it is used nowadays in an extended sense to include other characteristics which may be commonly thought of as being associated with common racial origin.

For a group to constitute an ethnic group in the sense of the Act of 1976, it must, in my opinion, regard itself, and be regarded by others, as a distinct community by virtue of certain characteristics."

65. It is in this context that he sets out the criteria which in his opinion were essential or helpful in "distinguishing the group from the surrounding community." Lord Templeman at page 569E aid:

"In my opinion, for the purposes of the Race Relations Act a group of persons defined by reference to ethnic origins must possess some of the characteristics of a race, namely group descent, a group of geographical origin and a group history. The evidence shows that the Sikhs satisfy these tests. They are more than a religious sect, they are almost a race and almost a nation."

66. It is in the context of these comments that the House considered the question whether or not the Sikh community complied with the relevant criteria. It seems to me relevant to observe that the main issue before the House was not so much whether the Sikh community did or did not comply with the criteria but what was the correct test to apply in deciding the question of ethnic origin? The question whether or not the Sikh community complied with the criteria seems to me to have been one which was almost self evident once the appropriate criteria was established and for my part I doubt very much whether an ordinary member of the public would have had any doubts about this. Most people would regard Sikhs as a "race" even if they falsely believed that their race was "biologically derived." Many, if not all, of the general public would know that there had been two Sikh wars and would know that for generations Regiments of Sikhs formed a part of the Indian Army and were often a symbol, through their presence on guard at British Embassies and establishments, of British Imperial power based on the Indian Army and the British Army in India. They would know that they fought in two world wars as distinctive units. They would know of their distinctive dress and probably some of their customs regarding hair and the wearing of turbans. They would know that the Sikhs had a distinct religion or would at least have heard of the Golden Temple of Amritsar. The question whether or not Sikhs were of ethnic origin within the criteria was, in my view, a simple and obvious one and would have been regarded as such by the general public once the appropriate criteria for the phrase "ethnic origin" was established. A Sikh would certainly have so regarded himself and his fellow Sikhs. The same does not necessarily apply to gypsies and if the judge's findings of fact were justified by the evidence I would, for my part, be inclined to agree that even if individual gypsies fall within many of the Mandla criteria they were not an ethnic group because on the judge's finding such a group was not in any true sense identifiable as a group even by the gypsies themselves or by others, and no sufficient racial flavour existed. If the judge's findings were justified by the evidence the fact that the conclusion was reached by a process of flawed reasoning would not necessarily be fatal to the decision.
67. Was the finding justified on the evidence? Accepting that the judge was entitled, having heard them, to form an unfavourable view of the experts and to regard their evidence with caution it is not easy to understand how he can have wholly rejected their historical discourse nor their evidence with regard to the customs and traditions and traditional way of life peculiar to gypsies since no evidence to controvert this was tendered. The evidence of Mr. Mercer who described himself as "a gypsy by birth" and whose people "were gypsies back in 1888" was to the effect that he could identify "our own people." However, the evidence of the continued separate identity of gypsies as people "who regarded themselves and who were regarded by others as a distinct community" was scant and it is for this reason I have hesitated whether or not it could be said that the ethnic identity of gypsies in the strict sense was established. The validity of the judge's finding above recited "that there may well be a small number of travelling people who can claim either by looks or characteristics to be true gypsies, but these people have been absorbed into a larger group" must depend on whether or not there is sufficient evidence that such absorption has occurred. There was, at least, some evidence that it has not and for these reasons I do not feel I can properly dissent from the conclusions of my Lords. I reach this conclusion with some regret. I doubt whether if the claim for breach of the Act is finally established, benefit rather than detriment will result to either side.
68. For the reasons I have given I agree with the conclusion of my Lords and would allow this appeal. I agree with the directions proposed.

Appeal allowed with costs. Judge's order below set aside. Costs reserved to the judge pending the outcome of the issue.