South Africa: Constitutional CourtYou are here: SAFLII >> Databases >> South Africa: Constitutional Court >> 1997 >>  ZACC 11 | Noteup | LawCite
S v Lawrence , S v Negal ; S v Solberg (CCT38/96, CCT39/96, CCT40/96)  ZACC 11; 1997 (10) BCLR 1348; 1997 (4) SA 1176 (6 October 1997)
Download original files
Links to summary
Bookmark/share this page
CONSTITUTIONAL COURT OF SOUTH AFRICA
Cases CCT 38/96
REBECCA LAWRENCE Appellant in CCT 38/96
RODNEY GORDON NEGAL Appellant in CCT 39/96
MAGDALENA PETRONELLA SOLBERG Appellant in CCT 40/96
THE STATE First Respondent
THE MINISTER OF TRADE AND INDUSTRY Second Respondent
Heard on : 6 May 1997
Decided on : 6 October 1997
The constitutional issues on appeal
 The three appellants were each charged in a Magistrates’ Court and convicted of contraventions of the Liquor Act 27 of 1989 (“the Liquor Act”). The appellants, all employees of the Seven Eleven chain store, did not dispute the facts relied upon by the state at their trials. They were charged separately and the defence in each case was that the particular provisions of the Liquor Act under which that appellant was charged were inconsistent with the interim Constitution and were accordingly invalid.
 Each of the cases was concerned with a contravention of the terms of the grocer’s wine licence authorising the sale of wine at the stores at which the appellants were employed. In terms of the Liquor Act the holder of a grocer’s wine licence is prohibited from selling liquor other than table wine. There are also restrictions on the hours and days on which sales may be effected. The state’s case against Ms Lawrence was that she sold wine at a Seven Eleven store during a week day but after closing hours; the case against Ms Solberg was that she sold wine at a Seven Eleven store on a Sunday, which is a closed day for sales of wine by holders of grocers’ wine licences; and the case against Mr Negal was that he sold cider and beer at a Seven Eleven store despite the fact that the liquor licence of the store permitted only the sale of table wine.
 A Magistrates’ Court has no jurisdiction to declare the provisions of an Act of Parliament to be unconstitutional. At each of the trials the appellant concerned applied in terms of section 103(3) of the interim Constitution for a postponement of the trial to enable the constitutional issue to be referred to this Court for determination. On each occasion the application was refused and the trial proceeded. The trials followed the same pattern. The appellants admitted the material allegations made in the charge sheets, indicated that they would challenge the constitutionality of the provisions of the Liquor Act on which the charges were based, and led no evidence. The magistrates, as they were obliged to do in terms of section 103(2) of the interim Constitution, assumed the provisions of the Liquor Act to be valid and convicted the appellants.
 The appellants, who had been represented by the same counsel and attorneys at their trials, appealed to the Cape of Good Hope Provincial Division of the Supreme Court against their convictions. In each case the only ground of appeal was that the relevant provisions of the Liquor Act were inconsistent with the interim Constitution and accordingly invalid. The appeals were set down for hearing on the same day and were dealt with as one matter. The appellants did not ask the Court to refer the constitutional issues to this Court in terms of section 102(1) of the interim Constitution. Instead they conceded that the magistrates had correctly convicted them, that the only defence that could be offered was that the provisions were unconstitutional, and that the Provincial Division had no jurisdiction to set the convictions aside on such grounds. The appeals were accordingly dismissed and the appellants then noted an appeal in terms of rule 21(1) to this Court.
 The scheme of the Liquor Act is to control the sale of liquor through a licensing system. It is an offence under the Act to sell liquor without a licence or a special exemption, to fail to comply with a condition of a licence, and to sell liquor at a time or a place at which the sale of liquor is not permitted by the licence. The Act also contains a general prohibition against a liquor business being conducted on the same premises as any other trade or occupation, but exceptions are made in respect of businesses conducted in terms of a grocer’s wine licence or a sorghum beer licence.
 Sections 87 to 90 of the Liquor Act deal with conditions attaching to grocers’ wine licences. Section 87 provides that:
“The holder of a grocer’s wine licence . . . shall at all times carry on the business of a general dealer (which shall include dealing in groceries and foodstuffs), and may carry on or pursue any other business (excluding a business to which any other licence relates) or trade or occupation, on the licensed premises.”
Section 88(1) prohibits the sale under a grocer’s wine licence of any liquor other than table wine. Section 90(1) deals with the time when the table wine may be sold. The times are:
“(a) on any day, excluding a closed day and Saturday . . . between 08:00 and 20:00;
(b) on any Saturday, excluding a closed day . . . between 08:00 and 17:00.”
A closed day on which sales are not permitted under a grocer’s wine licence are Sundays, Good Friday and Christmas Day.
 The appellants contended that the prohibition imposed by section 90(1)(a) on the selling of wine “after hours” on week days and on closed days, and by section 88(1) on the sale of liquor other than wine, which made the sale of cider and beer unlawful, is inconsistent with the right to economic activity guaranteed by section 26 of the interim Constitution and that the prohibition against selling wine on Sundays was inconsistent with the right to freedom of religion, belief and opinion guaranteed by section 14.
 The Minister of Trade and Industry, the minister responsible for the administration of the Liquor Act, elected to intervene in the appeal in terms of section 102(10) of the interim Constitution and to present argument to the Court on behalf of the government. Subsequently, the South African Liquor Store Association was admitted as an amicus curiae and was given leave in terms of rule 9(9) to address oral argument to the Court at the hearing of the appeals which had been set down for hearing before this Court on the same day.
 As a result of the procedure followed in the Magistrates’ Court and in the appeals to the Cape of Good Hope Provincial Division the appeal record contained no evidence relevant to the constitutionality of the prohibitions challenged by the appellants. After the appeals had been noted to this Court an agreement was reached between the attorneys for the appellants and a representative of the Attorney General that the appellants would lodge affidavits from experts dealing with the issues on appeal, that the Attorney General would be entitled to lodge affidavits in answer to such contentions, and that the appellants would be entitled to lodge affidavits in reply.
 Claiming to act in pursuance of this agreement the appellants lodged affidavits from three experts in which it was said that the provisions of the Liquor Act limiting the hours and days of sale had no discernible impact on alcohol consumption, that there is no legitimate reason for distinguishing between the types of liquor which may be sold from particular premises, and in particular no legitimate reason for permitting grocers to sell wine, but not beer and cider. These affidavits referred to statistical information which was said to support these contentions and also contained averments that the legislation which permits grocers to sell wine, but not beer and cider or other liquor, resulted from political influence exercised by the “wine lobby” at the time the legislation was passed.
 The Attorney General, purporting to act in terms of rule 34 of the Constitutional Court Rules, lodged affidavits in which it was said that there is a relationship between the consumption of alcohol and violent crime, and that if restrictions on the times at which and the types of liquor which could be sold by supermarkets were removed there would be an increase in the consumption of liquor to the prejudice of the community. The Minister of Trade and Industry also lodged affidavits from experts disputing the averments made by the appellants’ experts. At the hearing of the matter counsel for the appellants tendered from the bar a number of extracts from publications which he said had been relied on by one of the appellants’ experts; there was no affidavit from the expert confirming this or explaining why the extracts had not been dealt with in his affidavit.
 In their written arguments counsel for the Attorney General and counsel for the Minister of Trade and Industry disputed the admissibility of the affidavits relied on by the appellants, contending that they went beyond what is permissible under rule 34. They also objected to the admission of the extracts from the publications which had been tendered from the bar by counsel for the appellants.
 In response to the averments made in the written arguments as to the admissibility of the affidavits, the appellants lodged a substantive application, in which it was contended that the affidavits were admissible in terms of rules 19(1)(a), (b)(ii) and (c)(i) of the Constitutional Court Rules, and rule 34, but asking in the event of it being held that the affidavits were not admissible, that they be admitted by this Court under its general power under rule 35 to condone non compliance with its rules. I deal later with the provisions of these rules and their application to the present case.
 The introduction of new evidence on appeal, even in a criminal case, is ordinarily permissible only in exceptional circumstances. Counsel for the appellants contended that this principle is not applicable to an appeal in which a constitutional question within the exclusive jurisdiction of the Constitutional Court has to be decided. In such circumstances, so he argued, an accused person does not have the opportunity to tender evidence relevant to the constitutional issue prior to the appeal. Any attempt to do so in the present case would have been met by an objection that the evidence was not relevant to any issue within the jurisdiction of the court dealing with the matter. He contended further that the Constitutional Court is the only court with the jurisdiction to receive such evidence, and that it ought therefore to construe its rules or regulate its process so as to permit such evidence to be tendered for the first time on appeal.
 The submission that the appellants could not have placed the relevant evidence on record before noting their appeals under rule 21 is not correct. There were at least two opportunities prior to the appeal to this Court when the evidence could have been placed on record. First, the appellants could have called the witnesses on whom they rely to give evidence at their trials indicating that the evidence was relevant to the determination of the constitutional issues that they wished to raise as a defence to the charges against them. If this had been refused the issue could have been raised as a ground of appeal. The appellants could also have tendered the evidence on which they now rely at the time of their appeals to the Cape of Good Hope Provincial Division. The constitutional issues were the only defences that the appellants had to the convictions and sentences imposed by the magistrates and were decisive for the appeals. The matters accordingly fell within the terms of section 102(1) of the interim Constitution. The appellants should have acted in terms of that section and asked for the matters to be referred to this Court for its decision, indicating that evidence was necessary for the purposes of resolving the constitutional issues. If the Cape Provincial Division considered the referral to be in the interests of justice, as it presumably would have done, it would have been obliged to refer the matters. It would also have been obliged to consider the implications of the appellants’ failure to place the necessary evidence on record at the time of their trials. If it took the view that evidence was admissible on appeal, it would have been obliged to receive the evidence and make findings thereon before referring the matter. If it took the view that evidence was not admissible on appeal, it would have referred the issues without receiving the evidence.
 The appellants did not avail themselves of either of these opportunities. The first time that they evinced any intention of adducing evidence on the constitutional issues was after an appeal in terms of rule 21 had been noted to this Court. In their application to have the new evidence admitted on appeal the appellants did not suggest that they could not have tendered the evidence prior to noting their appeals. They relied in the first instance on a contention that the evidence was admissible in terms of rules 19 and 34, and in the alternative, on the Court’s power in terms of section 173 of the 1996 Constitution to regulate its own process.
 Rule 19 deals with the procedure to be followed in appeals in which leave to appeal is required. Its provisions are made applicable to appeals noted under rule 21(1). They require the appellant to prepare and lodge the appeal record within a period of four months. Rule 19(1) deals with the contents of the record. It provides:
“(a) . . .
(b) The appeal record shall consist of –
(i) those portions of the judgment of the court a quo, and all relevant documentation lodged by the parties in the court a quo pertaining to the issues that are to be determined; and
(ii) only such evidence and exhibits or affidavits and annexures as may be relevant for the purpose of the appeal.
(c) (i) The parties shall endeavour to reach agreement on what should be included in the record and, in the absence of such agreement, the appellant shall apply to the President for directions to be given in regard to the compilation of the record.”
 The appellants contended that in terms of rule 19(1)(c)(i) the parties were entitled to agree that the trial record should be supplemented by the introduction of new evidence, that such an agreement was reached in the present case, and that in the circumstances the appellants were entitled to rely on the affidavits from the experts. Counsel for the Attorney General disputed this; he argued that the agreement should be construed as one determining the procedure to be followed in introducing affidavits admissible under rule 34, and that in the light of the material dispute of fact that existed on the affidavits, the requirements of rule 34 had not been met. I deal later with rule 34.
 I shall assume in favour of the appellants that their version of the agreement should be accepted. But even if this is so, the evidence would not be admissible in terms of rule 19. Rule 19 deals with the preparation of the appeal record, which according to the practice of our courts has always been understood to mean a record of the proceedings in the court against whose decision the appeal has been noted. Rule 19(1)(b) is directed to the exclusion from the record of evidence that may not be relevant to an appeal on constitutional issues only. It prescribes a procedure for circumscribing the record and not a means for introducing new evidence on appeal. That is apparent not only from the context, but also from the reference in rule 19(1)(b)(ii) to “evidence and exhibits” which can only be understood as referring to evidence and exhibits already on record.
 The interim Constitution requires that evidence relevant to the referral of issues to this Court be dealt with through procedures to be completed prior to the referral. This is emphasised in sections 102(1), 102(3), 102(15) and 103(4) of the interim Constitution. The same considerations that underlie these requirements apply to appeals and applications for direct access. The reason is obvious. All justices of the Court who are available to do so are required to deal with matters which are heard by this Court. It cannot be expected that eleven judges should sit to hear disputed evidence – a matter made clear by the judgment in Brink v Kitshoff NO which was reported prior to the noting of the appeal in the present matter.
 Counsel for the appellants suggested that the problem of resolving conflicts of fact could be addressed through section 7 of the Constitutional Court Complementary Act 13 of 1995 which empowers this Court to appoint commissions to receive evidence “necessary for the determination of any issue” in proceedings before it. The fact that this Court has the power to appoint a commission to receive evidence does not mean that litigants are entitled as of right to introduce new evidence on appeal. It is not necessary in the present case to express any opinion on the facts that have to be established to justify the appointment of a commission under section 7. No application was made for the appointment of such a commission and all that need be said is that the section provides no support for the construction of rule 19 on which the appellants rely. In effect, what the appellants contend is that the parties can agree to the introduction of disputed evidence for the first time on appeal, and require this Court to resolve the dispute, if necessary by hearing oral evidence, and that this can be done against the wishes of the Court and no matter how inconvenient such a procedure might be. Much clearer language than that used in rule 19 would be required to justify a conclusion that so unusual a procedure was contemplated by the rules of this Court.
 Rule 34 provides:
“(1) Any party to any proceedings before the Court, and an amicus curiae properly admitted by the Court in any proceedings shall be entitled, in documents lodged with the registrar in terms of these rules, to canvass factual material which is relevant to the determination of the issues before the Court and which do not specifically appear on the record: Provided that such facts –
(a) are common cause or otherwise incontrovertible; or
(b) are of an official, scientific, technical or statistical nature, capable of easy verification.
(2) All other parties shall be entitled, within the time allowed by these rules for responding to such document, to admit, deny, controvert or elaborate upon such facts to the extent necessary and appropriate for a proper decision by the Court.”
 Rule 34(1)(a) requires the facts relied upon to be “common cause” or “incontrovertible”. The rule has no application to disputed facts. Rule 34(1)(b) requires the facts to be of the character contemplated by the rule and to be capable of “easy verification”. Factual material in the affidavits which falls within these parameters is admissible under rule 34; but disputed facts which are not capable of easy verification are not.
Regulation of process
 Section 173 of the 1996 Constitution confers on this Court, the Supreme Court of Appeal and the High Courts an “inherent power to protect and regulate their own process, and to develop the common law, taking into account the interests of justice.” Counsel for the appellants contended that if the expert evidence on which they rely is not admissible under rule 19 or rule 34, this Court should exercise its powers under section 173 of the Constitution to admit it. The appellants do not, however, have to rely on section 173 which in any event seems not to be applicable to this case. This Court has power under its rules to admit new evidence on appeal. The question is whether that power should be exercised in the circumstances of the present case. For the reasons already given this Court should not, save in exceptional circumstances, permit disputes of fact or expert opinion to be raised for the first time on appeal. Such circumstances have not been established in the present case.
 The introduction of new factual material on appeal to this Court is regulated by rule 34. In so far as the evidence tendered by the appellants is not admissible under rule 34, no good reason exists to depart from the provisions of the rules. It follows that only those portions of the expert evidence that fall within the scope of rule 34 can be taken into account for the purposes of deciding these appeals. I will refer to the evidence relied upon by the appellants when I deal with each of the constitutional issues raised by the appeals.
The interpretation of section 26
 Section 26 of the interim Constitution, on which the appellants rely for their challenge to the provisions under which they were convicted, provides:
“26. Economic activity
(1) Every person shall have the right freely to engage in economic activity and to pursue a livelihood anywhere in the national territory.
(2) Subsection (1) shall not preclude measures designed to promote the protection or the improvement of the quality of life, economic growth, human development, social justice, basic conditions of employment, fair labour practices or equal opportunity for all, provided such measures are justifiable in an open and democratic society based on freedom and equality.”
 The appellants contend that section 26(1) should be interpreted expansively to encompass all forms of economic activity and all methods of pursuing a livelihood. The only exclusion that they would allow is in respect of those activities which are “innately criminal”. All other economic activities, so they contend, are protected by section 26(1) unless their curtailment can be justified under section 26(2).
 Relying on this construction of section 26 the appellants contended that subsection (1) constitutes the right and subsection (2) constitutes a special limitations clause which displaces section 33 as far as the limitation of “free economic activity” is concerned.
 The construction of section 26 advanced by the appellants does not give sufficient weight to the wording of the section or to the structure of chapter 3 of the interim Constitution. The words with which subsection (2) commences, “[s]ubsection (1) shall not preclude”, indicate that the two subsections must be read together to determine the content of the right. Reading section 26 in this way would also give full effect to the provisions of section 33 and to the structure of chapter 3 which makes provision for a general limitations clause. If the expansive interpretation of section 26 contended for by the appellants were to be adopted it would mean that the regulation of economic activity which cannot be justified according to the criteria specified in section 26 would be invalid no matter how reasonable or even necessary such regulation might be. There seems to me to be no justification for such a conclusion.
 The criteria prescribed by section 26(2) and section 33 are different. Section 26(2) is directed in the first instance to the “design” of the measure. If it is “designed” to promote the protection or improvement of any of the matters referred to in the subsection, and is a measure justifiable in an open and democratic society based on freedom and equality, it does not infringe section 26. Section 33 calls for a proportionality test which does not form part of a section 26(2) analysis. If sections 26(1) and (2) are read together as defining the right effect can be given to both section 26(2) and section 33. There is accordingly no reason why section 26 should be construed as excluding the operation of section 33.
 Section 26 is concerned in the first instance with “who” may engage in economic activity and pursue a livelihood, and “where” this may be done. Section 26(1) lays down that “[e]very person” may do so “anywhere in the national territory.” In the light of our history of job reservation, influx control and monopolies it is understandable that there should be such a provision in the bill of rights.
 The meaning of section 26 is, however, by no means clear. There seem to be two possible approaches to its interpretation. The first focuses on the meaning of free participation in economic activity and in pursuing a livelihood. In a modern democratic society a right “freely” to engage in economic activity and to earn a livelihood does not imply a right to do so without any constraints whatsoever. As Van Dijkhorst J said in Directory Advertising Cost Cutters v Minister for Posts, Telecommunications and Broadcasting and Others:
“Section 26(1) goes no further than to enshrine the right freely to be active in the economic sphere wherever one wants – the economic sphere with all its inherent constraints.”
 Certain occupations call for particular qualifications prescribed by law and one of the constraints of the economic sphere is that persons who lack such qualifications may not engage in such occupations. For instance, nobody is entitled to practise as a doctor or as a lawyer unless he or she holds the prescribed qualifications, and the right to engage “freely” in economic activity should not be construed as conferring such a right on unqualified persons; nor should it be construed as entitling persons to ignore legislation regulating the manner in which particular activities have to be conducted, provided always that such regulations are not arbitrary. Arbitrariness is inconsistent with “values which underlie an open and democratic society based on freedom and equality”, and arbitrary restrictions would not pass constitutional scrutiny.
 On this approach to the interpretation of section 26 the right to engage in economic activity and to pursue a livelihood anywhere in the national territory would entail a right to do so freely with others. Implicit in this is that the participation should be in accordance with law. Thus nobody can claim that section 26 gives him or her the right to deal in stolen property or in harmful drugs or to break the law in any other way. Nor can anyone claim that the right entitles him or her to ignore laws having a rational basis which deal with town-planning, zoning, licensing, and other regulation of business, trades or professions. These are the constraints of the economic system applicable to all persons and those who wish to engage in it must do so subject to such constraints.
 If this is the correct approach to the interpretation of section 26, section 26(2) would then be construed as permitting legislation which curtails free participation as long as the legislation is designed to serve a purpose sanctioned by the section. Thus the right to “be active” in particular economic activities could be constrained by closed shop agreements or policies of affirmative action which are designed to meet the requirements of section 26(2), or by measures within the purview of section 26(2) which restrict the market, as for instance is presently the case in respect of commercial undertakings such as rail and air transport, telecommunications, and broadcasting.
 On this approach to the construction of section 26 the objections taken by the appellants would fail. The appellants would have to establish that they have been denied the right to engage “freely” in the selling of liquor. Liquor is a potentially harmful substance. It is part of the normal environment in which the liquor trade is conducted in South Africa, and other countries, for selling to be regulated by licences which control not only the right to sell liquor but also where, when and what liquor may be sold. For reasons that are given later in this judgment, there is a rational basis for such a regime which is not inappropriate to the regulation and control of the liquor trade in an open and democratic society, or so constraining or inhibiting as to justify the conclusion that their right to engage “freely” in such trade has been infringed.
 The alternative approach is to read sections 26(1) and (2) together as indicating that all constraints upon economic activity and the earning of a livelihood which fall outside the purview of section 26(2) will be in breach of section 26. This construction is less restrictive of “free economic activity” and for the purposes of this appeal I am prepared to assume in favour of the appellants that it is the correct approach.
The section 26 analysis on the basis of the assumption made
 The meaning of the words “designed to promote” in the context of section 26 are by no means clear. Counsel for the appellants contended that these words should be interpreted to mean: “constructed so as to achieve”.
 This is the meaning that Professor Mureinik suggested should be given to similar words used in section 8(3) of the interim Constitution. Section 8(3) deals with the issue of affirmative action in the context of the constitutional guarantee of equality. In the passage relied upon by counsel for the appellants Professor Mureinik said (at 47):
“The words ‘designed to achieve’ can mean (a) ‘intended to achieve’, in which case they mean much the same as ‘aimed at’; or (b) they can mean ‘constructed so as to achieve’. In this latter sense, the wording refers to measures which are not only intended to achieve something, but the design of which makes it objectively probable that they will in fact achieve that something; which are structured, in other words, to achieve that thing. In this sense ‘designed to’ embodies much of the ambivalence of ‘calculated to’, which, although it can mean ‘intended to’, the courts often interpret as ‘likely to’; indeed, it incorporates both those meanings.” (Footnote omitted)
Professor Mureinik was dealing with the interpretation of section 8(3). Counsel for the appellants adopted his reasoning and sought to apply it to the interpretation of section 26(2).
 According to the Oxford English Dictionary “design” means “[t]o purpose or intend (a thing) to be or do (something); to mean (a thing) to serve some purpose . . .”. In the context of section 26 of the interim Constitution this seems to me to be the meaning that should be adopted.
 This does not mean that there need be no connection between the “design” and the “end” sought to be achieved. The requirement that the measures be justifiable in an open and democratic society based on freedom and equality means that there must be a rational connection between means and ends. Otherwise the measure is arbitrary and arbitrariness is incompatible with such a society.
 In the passage relied upon by the appellants Professor Mureinik argued for a more stringent test of legislative purpose – that there be an “objective probability” that the purpose will be achieved. He was not, however, dealing there with “economic freedom”. To apply that test to economic regulation would require courts to sit in judgment on legislative policies on economic issues. Courts are ill equipped to do this and in a democratic society it is not their role to do so. In discussing legislative purpose Professor Hogg says:
“While a court must reach a definite conclusion on the adjudicative facts which are relevant to the disposition of litigation, the court need not be so definite in respect of legislative facts in constitutional cases. The most that the court can ask in respect of legislative facts is whether there is a rational basis for the legislative judgment that the facts exist.
The rational-basis test involves restraint on the part of the court in finding legislative facts. Restraint is often compelled by the nature of the issue: for example, an issue of economics which is disputed by professional economists can hardly be definitively resolved by a court staffed by lawyers. The most that can realistically be expected of a court is a finding that there is, or is not, a rational basis for a particular position on the disputed issue.
The more important reason for restraint, however, is related to the respective roles of court and legislature. A legislature acts not merely on the basis of findings of fact, but upon its judgment as to the public perceptions of a situation and its judgments as to the appropriate policy to meet the situation. These judgments are political, and they often do not coincide with the views of social scientists or other experts. It is not for the court to disturb political judgments, much less to substitute the opinions of experts. In a democracy it would be a serious distortion of the political process if appointed officials (the judges) could veto the policies of elected officials.”
 This accords with the approach of the United States Supreme Court to rational basis review. It has consistently held:
“This restriction upon the judicial function, in passing on the constitutionality of statutes, is not artificial or irrational. A state legislature, in the enactment of laws, has the widest possible latitude within the limits of the Constitution. In the nature of the case it cannot record a complete catalogue of the considerations which move its members to enact laws. In the absence of such a record courts cannot assume that its action is capricious, or that, with its informed acquaintance with local conditions to which the legislation is to be applied, it was not aware of facts which afford reasonable basis for its action. Only by faithful adherence to this guiding principle of judicial review of legislation is it possible to preserve to the legislative branch its rightful independence and its ability to function.”
 Section 26 should not be construed as empowering a court to set aside legislation expressing social or economic policy as infringing “economic freedom” simply because it may consider the legislation to be ineffective or is of the opinion that there are other and better ways of dealing with the problems. If section 26(1) is given the broad meaning for which the appellants contend, of encompassing all forms of economic activity and all methods of pursuing a livelihood, then, if regard is had to the role of the courts in a democratic society, section 26(2) should also be given a broad meaning. To maintain the proper balance between the roles of the legislature and the courts section 26(2) should be construed as requiring only that there be a rational connection between the legislation and the legislative purpose sanctioned by the section. I deal later with how, if it be disputed, the legislative purpose is to be established.
 The rational basis test fits the language of the section which, unlike section 33, sets as the criterion that the measures must be justifiable in an open and democratic society based on freedom and equality, but does not require in addition to this that the measure be reasonable. The proportionality analysis which is required to give effect to the criterion of “reasonableness” in section 33 forms no part of a section 26 analysis.
 In coming to the conclusion that section 26(2) calls for a rational basis review I have given consideration to the decision in Public Servants’ Association of South Africa and Another v Minister of Justice and Others. This case deals with the meaning of the words “designed to achieve” in section 8(3) of the interim Constitution and takes an approach that is different to the approach taken by me in this judgment. I specifically refrain from commenting in any way on the correctness of that decision or on the interpretation to be placed on section 8(3). That section raises difficult issues, the consideration of which must be deferred until the occasion arises for this Court to do so. It is not necessary to express any view as to whether the test for rational review under section 26 is in any way different to the test for rational review under section 8.
The burden of proof
 Counsel for the appellants contended that the state has the burden of establishing that the provisions of the Liquor Act on which the prosecutions were based were protected by section 26(2). In support of this contention they relied on the approach of the Indian Supreme Court to the interpretation of article 19 of the Indian Constitution and in particular to the way in which article 19(6) is applied by the Indian courts.
 Article 19(1) of the Indian Constitution provides:
“(1) All citizens shall have the right –
(a) to freedom of speech and expression;
(b) to assemble peaceably and without arms;
(c) to form associations or unions;
(d) to move freely throughout the territory of India;
(e) to reside and settle in any part of the territory of India;
(f) . . .
(g) to practice any profession, or to carry on any occupation, trade or business.”
There is no general limitations clause in the Indian Constitution but in terms of section 19(6):
“Nothing in sub-clause (g) of the said clause shall affect the operation of any existing law in so far as it imposes, or prevent the State from making any law imposing, in the interests of the general public, reasonable restrictions on the exercise of the right conferred by the said sub-clause, and, in particular, nothing in the said sub-clause shall affect the operation of any existing law in so far as it relates to, or prevent the State from making any law relating to, –
(i) the professional or technical qualifications necessary for practising any profession or carrying on any occupation, trade or business, or
(ii) the carrying on by the State, or by a corporation owned or controlled by the State, of any trade, business, industry or service, whether to the exclusion, complete or partial, of citizens or otherwise.”
 Article 19(6) has been construed as a limitations clause by the Indian Courts. The section requires laws which limit article 19(1) rights to be reasonable, and this element – that is, reasonableness – has to be established by the litigant who relies on the limitation. In this regard the approach of the Indian Courts bears a similarity to that adopted by South African courts to the application of section 33 of the interim Constitution. It is, however, an approach which is not necessarily applicable to section 26(2) of the interim Constitution which is not couched as a limitations clause.
 Indian jurisprudence is of little assistance to the appellants. The Supreme Court of India has held that liquor is a harmful substance and that laws which regulate and control the sale and production of liquor do not infringe article 19(1)(g) of the Indian Constitution and accordingly do not have to be justified in terms of article 19(6). The constitutional challenges raised by the appellants in the present matter would accordingly be dismissed by an Indian Court.
 In the view that I take of this matter it is not necessary to decide whether, in order to prove that their constitutional rights have been infringed, the appellants have to establish that the constraints imposed on them by the Liquor Act fall outside of what is sanctioned by section 26(2) or whether the Attorney General has to establish that they fall within the terms of section 26(2). I shall assume in favour of the appellants that in so far as there is any burden it falls upon the Attorney General.
 The purpose of particular legislative provisions has ordinarily to be established from their context which would include the language of the statute and its background. Where the purpose is one sanctioned by section 26(2) the question whether that purpose is justifiable in an open and democratic society based on freedom and equality is essentially a question of law; so too is the question whether there is a rational basis for the means used to achieve the legislative purpose. That is not to say that evidence will not be relevant to these enquiries; it may well be. The evidence, however, is more likely than not to consist of “legislative facts”. Professor Hogg in the article previously referred to says:
“The US literature draws a distinction between ‘adjudicative’ facts and ‘legislative’ facts, terminology originally coined by Professor Kenneth Culp Davis, the author of the major US treatise on administrative law. Adjudicative facts are facts about the immediate parties to the litigation – ‘who did what, where, when, how, and with what motive or intent’; legislative facts are facts of a more general character concerning the social or economic milieu which gave rise to the litigation.”
Legislative facts do not have to be proved as strictly as adjudicative facts and as Professor Hogg says:
“In constitutional cases in the United States there has been a substantial relaxation of the limits of judicial notice for findings of legislative facts.”
I will return to this issue later in my judgment. For the moment it is sufficient to say that the nature of the enquiry that has to be undertaken by a court for the purpose of deciding whether or not measures have been designed for a purpose sanctioned by section 26 and the type of evidence that might be relevant to such an enquiry, means that the question of “burden of proof” is likely to be less important than where adjudicative facts have to be established.
 The Attorney General contended that legislation controlling the manner in which the liquor trade is conducted is a measure “designed” to promote the protection or the improvement of the quality of life, economic development and human development and as such is sanctioned by section 26.
 The excessive consumption of liquor is universally regarded as a social evil. It is linked to crime, disturbance of the public order, impairment of road safety, damage to health, and has other deleterious social and economic consequences. These are legislative facts of which this Court can take judicial cognisance.
 The Liquor Act describes itself as an Act “[t]o provide for control over the sale of liquor; and for matters connected therewith.” The appellants did not dispute that the excessive consumption of alcohol is harmful and that some control over the sale of intoxicating liquor is needed. This much was common cause.
 The means employed by the Liquor Act to achieve this purpose is to prescribe a system of licensing under which liquor sales are controlled. This is prima facie sanctioned by section 26(2). Measures designed to curtail some of the harmful effects of trade in liquor are clearly measures designed to protect or improve the quality of life. This seems to have been accepted by the appellants who did not challenge the constitutionality of the licensing scheme of the Act, or of section 154(1)(a) which penalises the unlicensed selling of liquor. Their challenges were confined to the particular constraints imposed on them by the licensing system. It is necessary, therefore, to consider the particular constraints to which objection was taken in each of the appeals.
After hours sales – the Lawrence appeal
 Section 90(1) of the Liquor Act fixes the times at which business may be conducted under a grocer’s wine licence. It does so as follows:
“The holder of a grocer’s wine licence may, notwithstanding any law to the contrary –
(a) on any day, excluding a closed day and Saturday, sell or deliver his or her liquor between 08:00 and 20:00;
(b) on any Saturday, excluding a closed day, sell or deliver his or her liquor between 08:00 and 17:00.”
Section 159(a), which is the section of the Liquor Act which constitutes the offence that the appellant is alleged to have committed, provides that the holder of a licence who
“if it is not otherwise an offence in terms of this Act, refuses or fails to comply with a condition which is attached to the licence . . . shall be guilty of an offence.”
Section 163(1)(a) prescribes the penalties that may be imposed for a contravention of section 159(a) of the Act.
 It is not clear to me why the appellant was charged and convicted under these sections of the Act. Section 159(c) of the Act provides that any holder of a licence who
“keeps the licensed premises open for the sale, supply or consumption of liquor or sells or supplies any liquor at a time when the sale of liquor is not permitted by the licence . . . shall be guilty of an offence.”
and the penalties for a contravention of section 159(c) are fixed by section 163(1)(b) of the Act.
 Since it is an offence under section 159(c) to sell liquor at a time not permitted by the licence, section 159(a) would seem to have no application to the case against the appellant. The same applies to the charge of selling liquor on a closed day, on which the appellant Magdalena Petronella Solberg was convicted.
 That, however, is not an issue which could be raised before this Court, and for the purposes of this appeal it has to be assumed that the appellants were correctly convicted under section 159(a). As the case is a test case in which the appellants seek to challenge the constitutionality of the prohibition against selling wine after hours and on a “closed day” nothing turns on the fact that the conviction was entered under section 159(a) and not 159(c).
 The appellant’s objection is that the restrictions imposed by the Liquor Act on the hours during which the holder of a grocer’s wine licence may sell table wine during week days interferes with the freedom of such licence holders to trade lawfully and with the freedom of consumers to purchase wine at times most convenient to them.
 The appellant has to establish that the particular constraints to which she objects infringe her right under section 26.
 Freedom to engage in economic activity in an open and democratic society does not imply a totally unconstrained freedom. Economic activity is subject to regulation and a shop keeper cannot claim to have “an unconstrained right to transact business whenever one wishes”.
 The out of hours sale for which the appellant was convicted took place on Monday 22 January 1996. In terms of section 90(1)(a) of the Liquor Act the appellant was permitted on that day to effect sales from 8 in the morning until 8 at night. It is a requirement which applies not only to sales under grocers’ wine licences, but also to sales under liquor store licences. These hours of trading allow more than sufficient time for the appellant or any “ licensed grocer” to engage “freely” in the business of selling wine.
 The scheme of the legislation is to effect controls through licences. The licences control who may sell liquor, what liquor may be sold, and when and where sales may take place. It is not necessary to deal with the different types of licences that may be given or the conditions attaching to them. Restriction on the hours of selling apply to all licences though different selling times are fixed for different types of licences. The distinctions drawn between the different types of licences are rationally related to the differences in the nature of the businesses and no point was made of this in argument. The basis of the appellant’s argument was that restrictions on the hours of sale do not reduce alcohol related problems, and that an increase in the hours of sale would not lead to an increase in alcohol consumption or alcohol related problems. The restrictions on hours of selling were therefore irrational.
 The appellant’s argument presumes that a restriction on the consumption of liquor is the only legitimate basis for restricting trading hours. This is not necessarily the case, and the Attorney General in fact contended that the restriction on trading hours was also conducive to economic growth and to human development. The regulation of the hours of trade allows small traders to close their shops at the end of a long day without fearing that they will lose trade to larger undertakings that are able to employ persons to work at night; it also frees employees from the pressure of being required to work overtime or at night. These considerations are relevant to economic growth, human development and fair labour practices. The case was, however, argued on the basis that the control of the supply of liquor was the main purpose of the legislation and it is on this basis that I will deal with the matter.
 The appellants sought support for their contentions on this issue from studies undertaken in other countries which are referred to in the affidavit of one of the expert witnesses. The correctness of the proposition that an increase in selling hours does not lead to an increase in consumption or an increase in alcohol related problems is disputed in the affidavit of the expert, Marcelle Christian, on whom the Minister of Trade and Industry relies. In her affidavit she says:
“To this day accumulated research has continued to show that:
– Alcohol-related problems are highly correlated with per capita consumption. This appears to hold over time and in different counties [sic].
– Decreases in per capita consumption produce reductions in alcohol-related problems regardless of whether this was from control measures or not.”
And she goes on to say that:
“a large body of recent research studies (30+) has found that relaxation of controls over availability such as expansion of the type and number of outlets; extended days and hours of operation and lower age restrictions, tends to be positively related to levels of alcohol consumption and indices of alcohol-related problems, not only among moderate drinkers, but also amongst so called heavy drinkers.” (Emphasis omitted)
 Counsel for the appellants sought to persuade us that the views of the appellants’ experts should be accepted rather than those of the Minister’s expert. The conflict is not capable of being resolved in this way. The expert evidence was not placed before the Court in a proper form and the attempt to cure the defect by tendering unverified extracts from publications on which the expert is said to have relied is unacceptable. The proposition relied upon by the appellants is, moreover, not “common cause or otherwise incontrovertible” nor does it depend on “official, scientific, technical or statistical” material that is capable “of easy verification”. In any event the conflict is not decisive of the case. The question to be decided is not whether the policy underlying the Liquor Act is an effective policy; it is whether there is a rational basis for such policy related to the purpose of the legislation.
 What is clear from the affidavit of Mr Makan – one of the appellants’ own experts – is that the control of the availability of alcohol is a recognised means of combatting the adverse effects of alcohol consumption. Mr Makan acknowledges this and says that:
“One of the strongest advocates of the Control of Availability theory is the World Health Organisation (‘WHO’). In 1993 the WHO suggested controls on the availability of alcohol . . . as one of [the] cornerstone[s] of its European Alcohol Action Plan.”
Mr Makan disputes the efficacy of the control of availability theory and says that few governments rely on it. The means to be adopted to control the liquor trade is, however, a matter for the government and not the courts to decide.
 I am satisfied that even if the burden of proof is on the Attorney General to establish the rational basis, and not on the appellants to negative it, it has been established that there is a rational basis for measures restricting the hours of sale as part of a legislative scheme designed to curtail the consumption of liquor. In the circumstances the restrictions do not in my view constitute a breach of section 26 of the interim Constitution.
Beer and cider – the Negal appeal
 The provisions of sections 87 to 90 of the Liquor Act dealing with grocers’ wine licences were introduced in 1962 after the report of the Malan Commission which had been appointed to consider possible changes to the Liquor Act. The Malan Commission recommended that grocers be licensed to sell beer and wine, and the appellant contended that the only reason for the failure to implement this recommendation was the influence of the “wine lobby”. The appellant contended further that there is no reason to distinguish between wine, beer and cider and that it is irrational to permit grocers to sell wine but not beer and cider.
 It is not clear to me why it would be irrational for the legislature to facilitate the sale of wine as part of a policy directed to encouraging the growth and development of the South African wine industry – a purpose prima facie sanctioned by section 26(2). It is, however, not necessary to consider this or the Attorney General’s argument that there is a rational basis for the prohibition against the sale by grocers of liquor other than wine that is related to the legislative policy of curtailing the consumption of liquor.
 The prohibition against the sale of beer and cider in a grocery store does not arise from sections 87 to 90 of the Liquor Act. It arises firstly from the general prohibition against the selling of liquor without a licence, and secondly from section 40(1) of the Liquor Act which provides:
“Subject to sections 87 and 99 and without derogating from section 41,[] the holder of a licence . . . shall not conduct his or her business under the licence on premises on which any other business (including a business to which any other licence relates) or any trade or occupation is carried on or pursued.”
 The appellant does not challenge the constitutionality of the licensing system imposed by the Act or the provisions of section 40 which restrict the goods that may be sold on licensed premises. For the purposes of this appeal, therefore, the constitutionality of these provisions must be assumed.
 Section 87 is one of the two provisions specifically exempted from the provisions of section 40. It provides:
“The holder of a grocer’s wine licence . . . shall at all times carry on the business of a general dealer (which shall include dealing in groceries and foodstuffs), and may carry on or pursue any other business (excluding a business to which any other [liquor] licence relates) or trade or occupation, on the licensed premises.”
The exception constitutes an extension and not a curtailment of the right to participate in the liquor trade. It allows grocers the special concession of being able to sell wine without being subject to the restrictions of section 40. Having taken advantage of that extension the appellant was bound by the terms on which it was granted.
 In his notice of appeal the appellant confined the constitutional challenge to the provisions of section 88, 159(a) and 163 of the Liquor Act. Section 88(1) of the Act provides that :
“The holder of a grocer’s wine licence shall not sell liquor other than table wine.”
Section 88(2) defines “table wine”. Section 159(a) makes it an offence to refuse or fail to comply with a condition attached to a licence and section 163 prescribes the penalties that can be imposed for a breach of section 159(a).
 At the hearing of the appeal counsel for the appellant did not ask for section 88 to be struck down. Striking down section 88 would not make it legal for wine and cider to be sold by grocers; that would still be prohibited by section 154(a) of the Act. Indeed, the effect of striking down section 88 might well be to invalidate the provisions of the Act dealing with grocers’ wine licences, for without section 88 there would be no provision dealing with what might be sold under such licences. If this were to be the result it would put an end to wine selling in Seven Eleven stores and make it unlawful for any liquor to be sold there.
 In an attempt to avoid these difficulties counsel for the appellant asked the Court to order that “to the extent that” a grocer’s wine licence prohibits the sale of beer and cider it is inconsistent with the Constitution and invalid. This, however, would not solve the appellant’s problem. The prohibition against the sale of liquor is contained in section 154(a) which prohibits the sale of liquor otherwise than under a licence or an exemption. To be able to sell beer and cider the Seven Eleven stores must procure a licence authorising them to do so. And the only licence in the Liquor Act as presently framed that permits this to be done, other than a grocer’s wine licence, is a liquor store licence. But that precludes the holder of the licence from conducting a grocery business.
 The fallacy in the appellant’s argument is that it treats section 88 as the obstacle to grocers selling beer and cider whereas in substance the section deals with the scope of the exception to the prohibition against selling any liquor from a grocery store. If the appellant wishes to challenge the constitutionality of prohibiting grocers from selling beer, cider or any other liquor the challenge should be directed against section 40 and not against the exception to the prohibition made by sections 87 and 88.
 Instead of doing this, the appellant has approached the Court for an order that the scope of the exception made by sections 87 and 88 be enlarged. In effect what the appellant has asked this Court to do is amend the Liquor Act so as to make provision for a “grocer’s wine, beer and cider licence”, as an exception to the prohibition imposed by section 40 of the Act. A court can strike down legislation that is unconstitutional and can sever or read down provisions of legislation that are inconsistent with the Constitution because they are overbroad. It may have to fashion orders to give effect to the rights protected by the Constitution, but what it cannot do is legislate.
 The exception established by section 87 and the related provisions dealing with grocers’ wine licences do not constitute an infringement of grocers’ rights to engage “freely” in the liquor trade. On the contrary they constitute an enlargement of their rights under the Liquor Act. If the appellant wishes to challenge the Liquor Act as such or the licensing system which restricts the types of business which can be conducted in association with the sale of liquor, the challenge should be directed against those provisions, and not against the exempting provisions of sections 87 to 90. I express no opinion as to whether a challenge to section 40 would have any substance; that was not an issue in the present appeals and does not call for any comment in this judgment.
Sunday trading – the Solberg appeal
 The appellant was convicted of contravening section 159(a) read with sections 2, 90(1)(a) and 163(1)(a) of the Liquor Act. Section 2 defines closed day as meaning Sunday, Good Friday and Christmas Day.
 The appellant contends that the prohibition infringes her right under section 14 of the interim Constitution to freedom of religion as well as her right under section 26 to free economic activity.
 It will be convenient to deal first with the arguments addressed to the issue of religious freedom. The terms of section 14 are as follows:
“14. Religion, belief and opinion
(1) Every person shall have the right to freedom of conscience, religion, thought, belief and opinion, which shall include academic freedom in institutions of higher learning.
(2) Without derogating from the generality of subsection (1), religious observances may be conducted at state or state-aided institutions under rules established by an appropriate authority for that purpose, provided that such religious observances are conducted on an equitable basis and attendance at them is free and voluntary.
(3) Nothing in this Chapter shall preclude legislation recognising –
(a) a system of personal and family law adhered to by persons professing a particular religion; and
(b) the validity of marriages concluded under a system of religious law subject to specified procedures.”
 The appellant contended that the purpose of prohibiting wine selling by grocers on “closed day[s]” was “to induce submission to a sectarian Christian conception of the proper observance of the Christian sabbath and Christian holidays or, perhaps, to compel the observance of the Christian sabbath and Christian holidays.” This, so the argument went, “coerced individuals to affirm or acquiesce in a specific practice solely for a sectarian Christian purpose”, and was inconsistent with the freedom of religion of those persons who do not hold such beliefs and do not wish to adhere to them.
 In support of this contention it was argued that the history of the legislation showed that closed days were introduced into the Liquor Act for a religious purpose. Sunday, Good Friday and Christmas Day, which are the only days presently covered by the definition of “closed day” in section 2 of the Act, are all of particular significance to the Christian religion. When the interim Constitution came into force on 27 April 1994 the definition in section 2 included another day of significance to the Christian religion, Ascension Day, as well as the Day of the Covenant, which was a day of religious significance to a section of the community. The reference to these two days was deleted from the definition in 1995. Counsel for the appellant argued that the selection of these particular days as closed days showed that the legislation had a religious purpose, and that this was sufficient in itself to constitute an infringement of section 14.
 In R v Big M Drug Mart Ltd, a decision of the Supreme Court of Canada on which the appellant relied in support of this argument, Dickson CJC said:
“If the acknowledged purpose of the Lord’s Day Act, namely, the compulsion of sabbatical observance, offends freedom of religion, it is then unnecessary to consider the actual impact of Sunday closing upon religious freedom. Even if such effects were found inoffensive, as the Attorney-General of Alberta urges, this could not save legislation whose purpose has been found to violate the Charter’s guarantees. In any event, I would find it difficult to conceive of legislation with an unconstitutional purpose, where the effects would not also be unconstitutional.”
 The Big M Drug Mart case concerned the provisions of the Canadian Lord’s Day Act. Its name proclaimed its purpose as did its provisions. It appears from the judgment in that case that the Act prohibited any work or commercial activity on the “Lord’s Day” – Sunday – as well as any games or performances where an admission fee was charged, any transportation for pleasure where a fee was charged, any advertisement of anything prohibited by the Act, the shooting of firearms and the sale or distribution of foreign newspapers. Certain exemptions were made in respect of “work of necessity or mercy” and the railways were allowed to operate. There was also a provision that persons could be exempted from the provisions of some of the prohibited activities by “provincial legislation or municipal charter”. The Canadian courts had previously held that the object of the Act was to compel the observance of the Christian sabbath. This led Dickson CJC to say:
“A finding that the Lord’s Day Act has a secular purpose is, on the authorities, simply not possible. Its religious purpose, in compelling sabbatical observance, has been long-established and consistently maintained by the courts of this country.”
 I would have no difficulty in holding that a law which compels sabbatical observance of the Christian sabbath offends against the religious freedom of those who do not hold such beliefs:
“If I am a Jew or a Sabbatarian or a Muslim, the practice of my religion at least implies my right to work on a Sunday if I wish. . . . any law purely religious in purpose, which denies me that right, must surely infringe my religious freedom.”
This does not mean that the selection of a Sunday for purposes
which are not purely religious and which do not constrain the practice
religions would be unlawful simply because Sunday is the Christian
 The Canadian Lord’s Day Act had a purely religious purpose and was designed to compel adherence to the Christian sabbath. The provisions of the Liquor Act relating to grocers’ wine licences are, however, materially different in their scope and effect from the Lord’s Day Act and I have difficulty in seeing how they can be said to compel sabbatical observance or to promote any particular religion. The Liquor Act permits the selling of liquor on closed days under certain licences but prohibits selling under other licences. Thus liquor may be sold under “on-consumption licences” on Sundays to lodgers and their guests at licensed hotels, to persons eating at licensed restaurants, to persons eating at licensed wine houses, and to club members and their guests at licensed clubs. The purpose of the closed day provisions seems to be to curtail the selling of liquor to the general public on such days. Thus hotel liquor licences do not permit liquor to be sold on closed days to persons other than lodgers and their guests, unless such persons take meals at the hotel, and restaurant licences and wine house licences do not permit the sale of liquor on closed days to persons who do not take meals at the licensed premises, though this is permissible on other days. Consistently with this policy liquor may not be sold on closed days under a theatre licence, a sportsground licence, or under off-consumption licences such as liquor store licences, wine farmers’ licences, sorghum beer licences and grocers’ wine licences. Closed day sales may also not be made under licences which authorise sales to liquor traders such as wholesale liquor licences, and brewers’ licences, or without the permission of the chairperson of the liquor board, under a producer’s licence. Special off-consumption licences can, however, be granted by the liquor board if “exceptional circumstances warrant the granting of the licence”, and these licences are apparently not subject to any conditions as to the times of doing business other than those imposed in terms of section 32(2) of the Act. This section does not require that sales on closed days be prohibited and it was not disputed that special licences have been granted authorising sales on such days.
 In the present case we are concerned with the narrow question whether section 90 of the Liquor Act, to the extent that it prohibits the selling of wine under a grocer’s wine licence on Sundays, infringes religious freedom by doing so. We are not required to consider whether other provisions of the Act which prohibit Sunday selling under different licences infringe the Constitution, nor are we required to consider the prohibition against sales of wine by grocers on Christmas Day or Good Friday. I will confine myself to this narrow question.
 In the Big M Drug Mart case Dickson CJC said:
“The essence of the concept of freedom of religion is the right to entertain such religious beliefs as a person chooses, the right to declare religious beliefs openly and without fear of hindrance or reprisal, and the right to manifest religious belief by worship and practice or by teaching and dissemination.”
I cannot offer a better definition than this of the main attributes of freedom of religion. But as Dickson CJC went on to say freedom of religion means more than this. In particular he stressed that freedom implies an absence of coercion or constraint and that freedom of religion may be impaired by measures that force people to act or refrain from acting in a manner contrary to their religious beliefs. This is what the Lord’s Day Act did; it compelled believers and non-believers to observe the Christian sabbath.
 I am not unmindful of the fact that constraints on the exercise of freedom of religion can be imposed in subtle ways and that the choice of Christian holy days for particular legislative purposes may be perceived to elevate Christian beliefs above others; and that as a result adherents of other religions may be made to feel that the state accords less value to their beliefs than it does to Christianity.
 Section 90 does not, however, prohibit grocery stores from doing business on Sundays and in fact the Seven Eleven stores are kept open for business on such days; nor does it force people to act or refrain from acting in a manner contrary to their religious beliefs.
 In South Africa, Sundays have acquired a secular as well as a religious character. This had happened before the interim Constitution came into force and before the 1995 amendments were made to the Liquor Act. Patterns established over the years by legislation have resulted in Sundays being the most common day of the week on which people do not work. Weekends consisting of Saturdays and Sundays are times at which most South Africans take a rest from work. Many take a whole weekend, some take Saturday afternoons and Sundays and some (but probably a minority) take only Sundays or Sunday afternoons. Shops are open in many parts of the country on Sunday mornings and sports stadia and places of entertainment are also open on Sundays. There are, however, only very few who work on Sunday afternoons.
 These rest days are recorded in labour agreements, in business practices, in contracts of service and in provincial legislation. Amongst those who observe Sundays as rest days, are many who do so because it has become the most convenient day for such purpose, and not because of any wish to observe the Christian sabbath. The secular nature of Sundays is evidenced by the ways in which many people spend their Sundays, engaging in sport and recreation rather than in worship.
 No evidence was placed before this Court as to how, if at all, the provisions of section 90 of the Liquor Act interfere with the appellant’s freedom of religion or the freedom of religion of any other person, or serve any other religious purpose. It is difficult to discern any coercion or constraint imposed by section 90 of the Liquor Act on the religious beliefs of holders of grocers’ wine licences or any other person, or any religious purpose served by such prohibition. The section does not compel licencees or any other persons, directly or indirectly, to observe the Christian sabbath. It does not in any way constrain their right to entertain such religious beliefs as they might choose, or to declare their religious beliefs openly, or to manifest their religious beliefs. It does not compel them to open or close their businesses on a Sunday.
 There is also no evidence as to whether, if there has been an interference with freedom of religion, the legislation would or would not be justifiable as a reasonable limitation of such right. This is an issue which has troubled the Canadian courts since the decision in the Big M Drug Mart case and may have to be considered by this Court at some future time. That is not necessary in the present case; nor is it necessary to decide whether this Court should follow the Canadian approach or the less exacting approach of the United States Supreme Court to legislation requiring “Sunday closing”. In view of the decision to which I have come it would be inappropriate to comment on such matters in the present case and I refrain from doing so.
 In the judgments of Sachs J and O’Regan J reference is made to decisions of the United States Supreme Court dealing with the First Amendment to the United States Constitution. The First Amendment provides:
“Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof . . . .”
It is clear from the United States decisions that although there is an area in which the “establishment” clause and the “free exercise clause” overlap, the two clauses have different concerns. In developing our own jurisprudence under section 14 of the interim constitution and section 15 of the 1996 Constitution we should be careful not to blur this distinction.
 The primary purpose of the “establishment clause” in the US Constitution is to prevent the advancement or inhibition of religion by the state. The primary purpose of the “free exercise” clause is to permit adherents of different faiths to pursue their religious beliefs without being impeded from doing so by state coercion.. Our Constitution deals with issues of religion differently to the US Constitution. It does so under the equality provisions of section 8, the freedom of religion, belief and opinion provisions of section 14, and the education provisions of section 32.
 The only provision relied on by the appellant in the present case is section 14. Section 14 does not include an “establishment clause” and in my view we ought not to read into its provisions principles pertaining to the advancement or inhibition of religion by the state. To do so would have far reaching implications beyond the apparent scope and purpose of section 14. If such obligations on the part of the state are to be read into section 14 does this mean that Christmas Day and Easter Friday can no longer be public holidays, that “Family Day” is suspect because it falls on Easter Monday, that the SABC as public broadcaster cannot broadcast church services (as it does regularly on Sunday mornings, though it does not regularly broadcast Muslim services on Fridays or Jewish services on Saturdays or Hindu services on any particular day of the week), that its daily religious programmes must be cancelled, and that state subsidies to denominational schools are prohibited? These examples can be multiplied by reference to the extremely complex United States law which has developed around the “establishment clause”.
 I should add that I can see nothing in the text of section 14(1) or in the historical background to a constitution which made no provision for an establishment clause, which would require such a principle to be read into its provisions. The Constitution deals with unequal treatment and discrimination under section 8. Unequal treatment of religions may well give rise to issues under section 8(2), but that section was not relied upon by the appellant in the present case. To read “equitable considerations” relating to state action into section 14(1) would give rise to any number of problems not only in relation to freedom of religion but also in relation to freedom of conscience, thought, belief and opinion, which would go far beyond the difficulties raised by the “establishment clause” of the US Constitution.
 Section 14(2) does not in my view provide justification for giving an extended meaning to section 14(1). Compulsory attendance at school prayers would infringe freedom of religion. In the context of a school community and the pervasive peer pressure that is often present in such communities, voluntary school prayer could also amount to the coercion of pupils to participate in the prayers of the favoured religion. To guard against this, and at the same time to permit school prayers, section 14(2) makes clear that there should be no such coercion. It is in this context that it requires the regulation of school prayers to be carried out on an equitable basis. I doubt whether this means that a school must make provision for prayers for as many denominations as there may be within the pupil body; rather it seems to me to require education authorities to allow schools to offer the prayers that may be most appropriate for a particular school, to have that decision taken in an equitable manner applicable to all schools, and to oblige them to do so in a way which does not give rise to indirect coercion of the “non-believers”. But whatever section 14(2) may mean, and we have heard no argument on this, it cannot in my view be elevated to a constitutional principle incorporating by implication a requirement into section 14(1) that the state abstain from action that might advance or inhibit religion.
 There may be circumstances in which endorsement of a religion or a religious belief by the state would contravene the “freedom of religion” provisions of section 14. This would be the case if such endorsement has the effect of coercing persons to observe the practices of a particular religion, or of placing constraints on them in relation to the observance of their own different religion. The coercion may be direct or indirect, but it must be established to give rise to an infringement of the freedom of religion. It is for the person who alleges that section 14 has been infringed to show that there has been such coercion or constraint. In my view this has not been established in the present case.
 Whatever connection there may be between the Christian religion and the restriction against grocers selling wine on Sundays at a time when their shops are open for other business, it is in my view too tenuous for the restriction to be characterised as an infringement of religious freedom. In the circumstances I hold that the appellant has failed to establish that section 90 of the Liquor Act is inconsistent with section 14 of the interim Constitution.
 The alleged infringement of section 26 can be dealt with briefly. The appeal on this ground is no different in substance from the Lawrence appeal. Sunday is the day of the week on which most South Africans do not work. A restriction on the sale of liquor on Sundays is, therefore, likely to be more effective in curtailing the consumption of liquor than a restriction on the sale of liquor on any other day of the week. For the reasons given in the Lawrence appeal I am satisfied that the appellant in the present matter has failed to establish that the prohibition of the sale of wine by grocers on
Sundays infringes section 26 of the interim Constitution.
 It follows that the appellants have failed to establish that the legislation under which they were charged and convicted was inconsistent with the interim Constitution.
 The following order is made: the appeals in the matters of Rebecca Lawrence v The State, Rodney Gordon Negal v The State, and Magdalena Petronella Solberg v The State are dismissed.
Langa DP, Ackermann J, and Kriegler J concur in the judgment of Chaskalson P.
 I have had the opportunity of reading the judgments prepared in this matter by Chaskalson P and Sachs J. I agree, for the reasons given by Chaskalson P, that section
88(1) and section 90(1) of the Liquor Act 27 of 1989 (“the Liquor Act”) are not in breach of section 26 of the Constitution of the Republic of South Africa, 1993 Act 200 of 1993 (“the interim Constitution”), and therefore I agree that the appeals of Ms Lawrence and Mr Negal should fail. I cannot, with respect, agree with the conclusions of Chaskalson P or Sachs J concerning the challenge arising from section 14 of the interim Constitution brought by Ms Solberg (“the appellant”) for the reasons I give below.
 The appellant was charged with and convicted of selling liquor on a Sunday in breach of the provisions of the Liquor Act. It is against that conviction which she now appeals. Her legal representative argues that the provisions in terms of which she was convicted are unconstitutional and therefore invalid and that her conviction should be set aside. There are two constitutional arguments raised: first, that the provisions of section 90(1) read with the definition of closed day in section 2 of the Liquor Act are in breach of section 26 of the interim Constitution. As I agree with the reasoning of Chaskalson P on that score, I shall say no more about it. The second constitutional attack raises the question of the right to freedom of conscience and religion.
 Section 90(1) of the Liquor Act prohibits the selling of liquor by wine licensees on “closed day[s]”. “Closed day[s]” are defined in section 2 of the Liquor Act as Sunday, Good Friday and Christmas Day. The appellant in this case argues that the purpose of this prohibition is to “induce submission to a sectarian Christian conception of the proper observance of the Christian sabbath and Christian holidays” and that the prohibition is therefore in breach of section 14 of the interim Constitution.
 This is the first case in which we have had to consider section 14 which provides:
“(1) Every person shall have the right to freedom of conscience, religion, thought, belief and opinion, which shall include academic freedom in institutions of higher learning.
(2) Without derogating from the generality of subsection (1), religious observances may be conducted at state or state-aided institutions under rules established by an appropriate authority for that purpose, provided that such religious observances are conducted on an equitable basis and attendance at them is free and voluntary.
(3) Nothing in this Chapter shall preclude legislation recognising -
(a) a system of personal and family law adhered to by persons professing a particular religion; and
(b) the validity of marriages concluded under a system of religious law subject to specified procedures.”
 There is no evidence on the record to establish the appellant’s religious beliefs. There can be no doubt however that she has a direct interest in the constitutionality of the provisions under scrutiny. If they are held to be unconstitutional on the grounds she has raised, then her conviction may be set aside. If they are found not to be inconsistent with the interim Constitution on those grounds, then the conviction will stand. This interest is clearly sufficient to found her challenge to the provisions.
 Chaskalson P has found that in prohibiting the sale of liquor on Sundays, Good Fridays and Christmas Day, section 90(1) of the Liquor Act does not constitute a breach of section 14. He acknowledges that constraints upon freedom of religion can be imposed in subtle ways (at paragraph 93) but finds that in this case there is no constraint upon people’s “right to entertain such religious beliefs as they might choose, or to declare their religious beliefs openly, or to manifest their religious beliefs.” Nor, he finds, is anyone compelled to open or close a shop on a Sunday (at paragraph 97). He concludes that any constraint imposed by the provisions is too “tenuous” to be characterised as an infringement of religious freedom (at para 105).
 Sachs J, in his concurring judgment, finds that while there has been a breach of section 14, that limitation is justified in terms of section 33. He concludes that the provision contains a sectarian message which constitutes a breach of section 14 (at para 163). He concludes however that that breach is justified.
 I shall commence by considering the purpose and meaning of section 14 in our Constitution. Unlike the Constitution of the United States, our Constitution contains no establishment clause prohibiting the “establishment” of a religion by the state. Nevertheless, the interim Constitution contains a range of provisions protecting religious freedom. In section 8, the interim Constitution prohibits “unfair discrimination” on grounds of religion. In section 32(c), every person is given the right
“to establish, where practicable, educational institutions based on a common culture, language or religion, provided that there shall be no discrimination on the ground of race.”
And, of course, section 14 protects the freedom of religion. It is not possible to read this array of constitutional protections without realising that our Constitution recognises that adherence to religion is an important and valued aspect of the lives of many South Africans and that the Constitution seeks to protect, in several ways, the rights of South Africans to freedom of religion.
 The provisions of section 14 themselves are instructive as to the manner in which the right should be developed in our law. Section 14(1) protects the right to freedom of religion and conscience. Section 14(2) then provides that religious observances may be conducted at state or state-aided institutions provided that they are conducted on an equitable basis and attendance at them is free and voluntary. And section 14(3) permits legislation recognising systems of personal and family law shared by members of a religion.
 It is clear from these provisions, and particularly sections 14(2) and (3), that the strict approach of the United States Supreme Court to the provisions of the First Amendment of the Constitution of the United States of America in relation to the separation between state and religious bodies has been avoided. That approach was perhaps most memorably stated by Black J in Everson v Board of Education of the Township of Ewing:
“The ‘establishment of religion’ clause of the First Amendment means at least this: Neither a state nor the Federal Government can set up a church. Neither can pass laws which aid one religion, aid all religions, or prefer one religion over another. Neither can force nor influence a person to go to or to remain away from church against his will or force him to profess a belief or disbelief in any religion. . . . In the words of Jefferson, the clause against establishment of religion by law was intended to erect ‘a wall of separation between church and State’.”
This approach to the First Amendment has led to a jurisprudence which generally prohibits any state endorsement or funding of religion. The US Supreme Court has, for example, held unconstitutional a New York practice recommending the reading of a non-denominational prayer in state schools; has struck down a Pennsylvania statute requiring that schools commence the day with a reading without comment of ten verses of the Bible; and held unconstitutional a practice whereby public schools permitted the use of their facilities and released their students for religious education during school hours.
 The provisions of section 14(2) of the interim Constitution make it clear that religious observances at public institutions will not give rise to constitutional complaint if the observances meet three requirements: the observances must be established under rules made by an appropriate authority; they must be equitable; and attendance at them must be free and voluntary. It seems appropriate to imply from this provision and from the absence of an express establishment clause that a strict separation between religious institutions and the state is not required by our Constitution.
 On the other hand, it also seems plain from the provisions of section 14(2) that state endorsement of religious practices is subject to certain qualifications. First, it should not be coercive. The requirement of free and voluntary attendance at religious ceremonies is an explicit recognition of the deep personal commitment that participation in religious ceremonies reflects and a recognition that the freedom of religion requires that the state may never require such attendance to be compulsory. It protects the rights to conscience both of non-believers and of people whose religious beliefs differ from those which are being observed. Direct coercion, of course, is only half the problem. As Black J stated in Engel v Vitale:
“When the power, prestige and financial support of government is placed behind a particular religious belief, the indirect coercive pressure upon religious minorities to conform to the prevailing officially approved religion is plain.”
 The stipulation of voluntariness is not the only precondition established by section 14(2). The subsection requires that even where attendance is voluntary, the observance of such practices must still be equitable. In my view, this additional requirement of fairness or equity reflects an important component of the conception of freedom of religion contained in our Constitution. Our society possesses a rich and diverse range of religions. Although the state is permitted to allow religious observances, it is not permitted to act inequitably.
 In determining what is meant by inequity in this context, it must be remembered that the question of voluntary participation is a consideration separately identified in section 14(2). The requirement of equity must therefore be something in addition to the requirement of voluntariness. It seems to me that, at the least, the requirement of equity demands the state act even-handedly in relation to different religions. As Brennan J held in Larson v Valente:
“This constitutional prohibition of denominational preferences is inextricably connected with the continuing vitality of the Free Exercise Clause . . . Madison’s vision – freedom for all religion being guaranteed by free competition between religions – naturally assumed that every denomination would be equally at liberty to exercise and propagate its beliefs. But such equality would be impossible in an atmosphere of official denominational preference. Free exercise thus can be guaranteed only when legislators – and voters – are required to accord to their own religions the very same treatment given to small, new, or unpopular denominations.”
It is important to emphasise that Brennan J is concerned here not with the establishment clause of the US Constitution but with the free exercise clause which provides that Congress shall make no law prohibiting the free exercise of religion. Requiring that the government act even-handedly does not demand a commitment to a scrupulous secularism, or a commitment to complete neutrality. Indeed, at times giving full protection to freedom of religion will require specific provisions to protect the adherents of particular religions as has been recognised in both Canada and the United States of America. The requirement of even-handedness too may produce different results depending upon the context which is under scrutiny. For example, in the context of religious observances at local schools, the requirement of equity may dictate that the religious observances held should reflect, if possible, the religious beliefs of that particular community or group. But for religious observances at national level, however, the effect of the requirement is to demand that such observances should not favour one religion to the exclusion of others.
 The requirement of equity in the conception of freedom of religion as expressed in the interim Constitution is a rejection of our history, in which Christianity was given favoured status by government in many areas of life regardless of the wide range of religions observed in our society. Sachs J in his judgment in this case has provided a valuable account of the ways in which Christian principles were endorsed by legislation and its practices often imposed upon all South Africans regardless of their beliefs (see paragraphs 148 – 152). The explicit endorsement of one religion over others would not be permitted in our new constitutional order. It would not be permitted, first, because it would result in the indirect coercion that Black J adverted to in Engel v Vitale. And secondly because such public endorsement of one religion over another is in itself a threat to the free exercise of religion, particularly in a society in which there is a wide diversity of religions. Accordingly, it is not sufficient for us to be satisfied in a particular case that there is no direct coercion of religious belief. We will also have to be satisfied that there has been no inequitable or unfair preference of one religion over others.
 I return now to the facts of this case. Under section 90 of the Liquor Act, stores holding grocers’ wine licences are not required to close on Sundays, Good Friday and Christmas Day, but are merely prohibited from selling wine, the alcoholic beverage which their grocer’s wine licence permits them to sell.
 In my view, it is not possible to read the inclusion of Sundays in the definition of “closed day” in the abstract. The inclusion of Sundays is accompanied, and ineluctably coloured, by the inclusion of Good Friday and Christmas Day. Good Friday and Christmas Day are, without doubt, important days in the Christian calendar. In addition, many Christian denominations consider, as a central tenet of their religion, that Sundays should be observed as a day of rest and religious observance. It is true that both Good Friday and Christmas Day are days which have been declared as public holidays. However they are only two of twelve statutorily recognised public holidays. And they are the two days of the twelve which have a direct foundation in the practice and observance of Christianity. It seems an unavoidable conclusion, that these two days together with Sundays were selected to comprise the definition of closed day because of their religious significance for Christians. If the purpose had been to provide for days of rest, the days selected would have been all days recognised as public holidays. The inevitable effect of choosing these days was to give a legislative endorsement to Christianity, but not to other religions.
 It is true that the recognition of these days is, in all likelihood, a relic from a former era when almost all trading activities and public sport and entertainment were prohibited on Sundays and religious holidays for express religious purposes. For all that it may be an anachronism, it is important to note that the definition of “closed day” was amended in 1995 by the deletion of Ascension Day and the Day of the Vow (16 December), but Christmas Day, Good Friday and Sundays were retained in the definition. Whatever the historical provenance of the provision, its current purpose and effect are to give special recognition to the holy days of one religion and not others. That contravenes the requirement in section 14 that the legislature should act even-handedly in relation to religion and not prefer one to the exclusion of others.
 I cannot accept that the legislature’s purpose in enacting the definition of closed day is a secular one. Even if it were that would not necessarily be the end of the matter. In my view, the question in each case will not be the question of purpose alone, but the question of whether the overall purpose and effect of the provision constitutes a breach of freedom of religion.
 I also cannot agree with Chaskalson P when he concludes that because the provisions do not constrain individuals’ “right to entertain such religious beliefs as they might choose, or to declare their religious beliefs openly, or to manifest their religious beliefs”, there is no infringement of section 14 (at para 97). In my view, the requirements of the Constitution require more of the legislature than that it refrain from coercion. It requires in addition that the legislature refrain from favouring one religion over others. Fairness and even-handedness in relation to diverse religions is a necessary component of freedom of religion.
 In sum, it is my view that the focus of section 14 is both purpose and effect. In interpreting section 14, we must recognise first, the value accorded to religious beliefs and the diversity of such beliefs in our society by our Constitution; and secondly, the fact that religious beliefs are a matter of personal faith and commitment which should not be the subject of coercion, whether direct or indirect, by the state. In this case, the legislation results in a breach of section 14 of the interim Constitution in that it results in the favouring of one religion over others. The appellant did not argue that the provision was in breach of section 8(2), the right not to be discriminated against unfairly on the grounds of religion. It is not necessary, in view of my conclusion, to consider whether section 90 would constitute a breach of that constitutional provision as well.
 The question that remains for consideration is whether that breach is justified in terms of section 33 of the Constitution. It is now well established in our jurisprudence that to pass that hurdle, we must be persuaded that the legitimate purpose and effect of the infringing legislation outweigh the extent of the infringement caused. In this case, we are concerned with an infringement of section 14(1) of the interim Constitution. Where there is an infringement of that right, the infringement must be shown not only to be reasonable and justifiable in an open and democratic society based on freedom and equality, but also necessary in such a society. In S v Makwanyane, the following was said of the different standards of justification:
“What is clear is that s 33 introduces different levels of scrutiny for laws which cause an infringement of rights. The requirement of reasonableness and justifiability . . . clearly envisages a less stringent constitutional standard than does the requirement of necessity. In both cases, the enquiry concerns proportionality: to measure the purpose, effects and importance of the infringing legislation against the infringement caused. In addition, it will need to be shown that the ends sought by the legislation cannot be achieved sufficiently and realistically by other means which would be less destructive of entrenched rights. Where the constitutional standard is necessity, the considerations are similar, but the standard is more stringent.”
Unfortunately, in this case, little evidence was placed before us by government of the precise purpose and effect of the provisions. It is not the first time that this Court has been left with little assistance in this regard. Inevitably, the absence of such evidence is an obstacle to the exercise we have to conduct in terms of section 33. It makes it far less likely that we will conclude that the infringement is justified.
 I accept that if the purpose of section 90 were to provide a uniform day of rest for all South Africans, that would be an important legislative purpose for the purposes of section 33. Indeed, it may well be that the purpose and effect of the definition of “closed day” in relation to provisions in the Liquor Act other than section 90 is to achieve a day of rest. That is because other provisions require the closure of liquor outlets on closed days. However, section 90 does not require that grocery stores close on “closed day[s]”, it merely prohibits the sale of alcoholic beverages on such days.
 It is not clear to me precisely what the purpose of the challenged provision is, but I am willing to accept that at least one of its purposes is to restrict the availability of liquor on closed days in order to restrict consumption. Such a purpose or effect is sufficient to ensure that there is no breach of section 26 of the interim Constitution, but it is far less persuasive in relation to the breach of section 14. This is so because even if one of its purposes is the restriction of supply to restrict consumption, it is hard to conclude that this is the primary purpose of the definition of closed day in section 90(1). First, because the Liquor Act does not prohibit the sale of all liquor on closed days, only certain types of sale. In addition, it does not prohibit sale on non-religious public holidays, such as the Day of Goodwill (26 December), New Year’s Day or Family Day (the day after Easter Sunday), when the roads are particularly full and the restriction of consumption would appear to be particularly desirable. To the extent, therefore, that this is a purpose of the legislation I cannot consider it to weigh heavily for the purposes of proportionality in the context of a breach of section 14. Nor am I satisfied that this purpose of the legislation is effectively achieved. To the extent that the Liquor Act permits the consumption of liquor in a variety of circumstances on closed days, it is not clear at all how effective it is in achieving a restriction of consumption by prohibiting sales from grocery stores and liquor stores. On the other hand, in identifying as closed days, days of Christian significance, the legislature displays an endorsement of Christianity in conflict with the Constitution. It is true that the scope of the infringement of section 14 is not severe or egregious, but in my view, the purpose and the effect of the legislation is not sufficient to meet the test of justification required by section 33.
 In terms of section 98(5) of the interim Constitution, when this Court finds that a provision in a law is unconstitutional it must declare it to be invalid to the extent of its unconstitutionality. In this case, I have found that the prohibition on trading upon closed days contained in section 90(1) of the Liquor Act is unconstitutional. Section 90(1) provides as follows:
“The holder of a grocer’s wine licence may, notwithstanding any law to the contrary –
(a) on any day, excluding a closed day and Saturday, sell or deliver his or her liquor between 08:00 and 20:00;
(b) on any Saturday, excluding a closed day, sell or deliver his or her liquor between 08:00 and 17:00.”
In my view, the constitutionality of section 90(1) could be cured by the excision of the words “closed day and” from section 90(1)(a) and the phrase “excluding a closed day” in section 90(1)(b).
 For these reasons, I cannot concur in the order of Chaskalson P concerning the appeal of Ms Solberg.
Goldstone J and Madala J concur in the judgment of O’Regan J.
 The relevant facts of the case are set out in the judgment of Chaskalson P at paragraphs 1 to 4. My judgment deals only with the challenge based on the conviction for selling wine on a Sunday.
 On the face of it, the prohibition of the sale of liquor by grocery stores on Sundays has nothing to do with freedom of religion. It applies equally to all sellers and purchasers of liquor, operating independently of their religious persuasion, and making no distinction between believers and non-believers. It in no way interferes with the practice of religion as internationally understood. I am unaware of any form of worship, observance, practice or teaching that actually requires the sale of liquor on Sunday or that is directly inhibited by its not being sold on that day.
 There are two ways, however, in which the determination of Sundays as closed days for the sale of liquor by grocery stores might involve violations of section 14 of the interim Constitution. The first relates to the impact that the choice by the state of the Christian sabbath as a closed day might have on non-Christian liquor sellers who, because their religion obliges them to cease trading on a different sabbath on top of the statutory limit on trading on the closed day, are placed at a competitive disadvantage. In the trenchant words of Stewart J of the US Supreme Court (dissenting):
“Pennsylvania has passed a law which compels an Orthodox Jew to choose between his religious faith and his economic survival. That is a cruel choice. It is a choice which I think no State can constitutionally demand. For me this is not something that can be swept under the rug and forgotten in the interest of enforced Sunday togetherness. I think the impact of this law upon these appellants grossly violates their constitutional right to the free exercise of their religion.”
 The second way in which section 14 might be involved is through the negative radiating symbolic effect that state endorsement of the Christian sabbath might have. The manner in which this might happen is well brought out in a passage from a judgment by O’Connor J of the US Supreme Court:
“The Establishment Clause[] prohibits government from making adherence to a religion relevant in any way to a person’s standing in the political community. Government can run afoul of that prohibition in two principal ways. One is excessive entanglement with religious institutions, which may interfere with the independence of the institutions, give the institutions access to government or governmental powers not fully shared by nonadherents of the religion, and foster the creation of political constituencies defined along religious lines. . . . The second and more direct infringement is government endorsement or disapproval of religion. Endorsement sends a message to nonadherents that they are outsiders, not full members of the political community, and an accompanying message to adherents that they are insiders, favored members of the political community. Disapproval sends the opposite message.”
The question in the present case is whether the prohibition of the sale of liquor by grocery stores on Sundays amounts to such an endorsement, thereby sending out a message that is inclusionary for some and exclusionary for others in violation of section 14. It should be emphasised that the constitutional enquiry in the present matter in no way relates to the rights of Christians to observe Sunday as a day of special religious significance. There can indeed be no doubt that such rights are firmly protected by section 14. What is in issue is the authority of the state to impose a particular religious view on the whole of society.
 In responding to the question in the context of the facts of the present case, I have encountered two major difficulties. The first relates to the scope that should be granted to the operation of the maxim that the law does not concern itself with trifles (de minimis non curat lex) in the area of belief and conscience. The second concerns whether or not it is appropriate or even possible to apply objective criteria when determining the significance law has in respect of something as subjective and personal as religious belief. The problem that faces a court in a multi-faith country is to decide whose viewpoint or frame of reference should be adopted when such an evaluation is made.
 To complicate the matter further, the challenge based on section 14 came not from believers whose faith was being threatened, but from grocers whose profits were being limited. The applicants were, of course, quite entitled to raise the issue of the constitutionality, in terms of section 14, of a law which placed restrictions on their commercial activities. Yet, the result was an air of artificiality in relation to this aspect of the case, and a lack of evidence, from the side both of the applicants and of the state, on the question of the purpose and impact of closed days. If ever there was a case which required close contextual rather than purely abstract analysis, it was this one, and if ever a cupboard was bare of concrete contextual information it was the one in the present matter.
The text and context of the interim Constitution
 Our solutions to all these problems and difficulties will, of course, be found not in the complex and often contradictory North American jurisprudence on the subject but in the text and context of our own Constitution. In Prinsloo v Van der Linde and Another this Court cautioned against simplistic transplantation into our jurisprudence of formulae, modes of classification and legal doctrine developed in other countries where the constitutional texts and socio-historical situations were different from ours. At the same time we stated that in developing doctrine we had to take account both of our specific situation and of problems which we shared with all humanity. Furthermore, section 35(1) required us when interpreting the bill of rights to promote the values of an open and democratic society based on freedom and equality. We emphasized that we should be astute not to lay down sweeping interpretations at this stage but should allow doctrine to develop slowly and, hopefully, surely, on a case by case basis with special emphasis on the actual context in which each problem arose. Although our observations in that case were specially directed towards the interpretation of section 8 equality rights in our country, they can, in my view, be applied with profit to the interpretation of section 14 as well. It is with these considerations in mind that my analysis proceeds. If I draw on statements by certain United States Supreme Court justices, I do so not because I treat their decisions as precedents to be applied in our courts, but because their dicta articulate in an elegant and helpful manner problems which face any modern court dealing with what has loosely been called church/state relations. Thus, though drawn from another legal culture they express values and dilemmas in a way which I find most helpful in elucidating the meaning of our own constitutional text.
 The principal provision in our Constitution we have to consider is section 14. Headed “Religion, belief and opinion”, its first subsection reads as follows:
“14(1) Every person shall have the right to freedom of conscience, religion, thought, belief and opinion, which shall include academic freedom in institutions of higher learning.”
This is the central provision as far as our enquiry is concerned, but it by no means exhausts the text with regard to questions of religion, belief and opinion. In the first place, the drafters of the interim Constitution emphasized the importance of section 14 rights by providing that they could not be derogated from during a state of emergency (section 34(5)(c)), and furthermore, that they should be amongst the relatively small number of rights which could only be limited (in terms of section 33(1)(b)(aa)) on the condition that such limitation was not only reasonable but also “necessary”. Secondly, section 8 complements section 14 by identifying discrimination on the grounds of religion, conscience and belief as presumptively constituting unfair discrimination.
Thirdly, section 17 guarantees that everyone shall have the right to freedom of association, which clearly includes the right of religious bodies to function freely as part of civil society.
 Fourthly, section 14(2) provides that:
“Without derogating from the generality of subsection (1), religious observances may be conducted at state or state-aided institutions under rules established by an appropriate authority for that purpose, provided that such religious observances are conducted on an equitable basis and attendance at them is free and voluntary.”
This subsection has been referred to as “a prime example of a provision attesting to the negotiators’ unwillingness to erect walls of separation between church and state”, allowing for the conduct of religious observances at state or state-aided institutions, which, subject to certain conditions to be strictly adhered to, would include educational institutions, prisons, and state hospitals.
Fifthly, section 14(3) opens the way to the possible recognition of religiously based family law by providing:
“(3) Nothing in this Chapter shall preclude legislation recognising –
(a) a system of personal and family law adhered to by persons professing a particular religion; and
(b) the validity of marriages concluded under a system of religious law subject to specified procedures.”
 Sixthly, the language provisions also testify to the importance which the interim Constitution attributes to religion as part of national life and culture. Section 3(10)(c) provides that:
“The Pan South African Language Board shall be responsible for promoting respect for and the development of German, Greek, Gujerati, Hindi, Portuguese, Tamil, Telegu, Urdu and other languages used by communities in South Africa, as well as Arabic, Hebrew and Sanskrit and other languages used for religious purposes.” (My emphasis)
Seventhly, section 32(c) provides that every person shall have the right:
“to establish, where practicable, educational institutions based on a common culture, language or religion, provided that there shall be no discrimination on the ground of race.”
And, eighthly, persons taking official oaths are offered the choice either of swearing the oath and adding the words “So help me God”, or else of making a solemn affirmation without reference to God.
 There are other provisions which, although not directly concerned with religion and belief, have an important bearing on how section 14 should be interpreted. It is noteworthy, for example, that section 15(1) provides inter alia that every person shall have the right to freedom of expression and freedom of artistic creativity and scientific research, while section 15(2) requires that all media financed by or under the control of the state shall be regulated in a manner which ensures impartiality and the expression of a diversity of opinion. Freedom of opinion and freedom of expression go hand in hand, and this section testifies to a strong constitutional concern for openness and diversity.
 The same theme is adverted to in the limitations clause and the interpretation clause, both of which establish the notions of an open and democratic society as a primary point of reference for evaluating the bill of rights. The concept of an open society must indeed be regarded as one of the central features of the bill of rights and a key element in the interpretation of section 14. Such a society is a pluralistic one in which there is no official orthodoxy or faith. In the ringing words of the US Supreme Court:
“If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion or force citizens to confess by word or act their faith therein.”
 Further evidence of the importance attributed by our Constitution to the respect for diversity is contained in the postscript, where the emphasis on reconciliation so as to overcome the strife and division of the past, underlines the importance of tolerance and mutual accommodation as one of the underpinnings of our new constitutional order. Openness coupled with diversity presupposes that persons may on their own, or in community with others, express the right to be different in belief or behaviour, without sacrificing any of the entitlements of the right to be the same in terms of common citizenship.
 To my mind, read in the context of all of the above provisions and of the Constitution as a whole, section 14 was intended at least to uphold the following principles and values: South Africa is an open and democratic society with a non-sectarian state that guarantees freedom of worship; is respectful of and accommodatory towards, rather than hostile to or walled-off from, religion; acknowledges the multi-faith and multi-belief nature of the country; does not favour one religious creed or doctrinal truth above another; accepts the intensely personal nature of individual conscience and affirms the intrinsically voluntary and non-coerced character of belief; respects the rights of non-believers; and does not impose orthodoxies of thought or require conformity of conduct in terms of any particular world-view. The Constitution, then, is very much about the acknowledgement by the state of different belief systems and their accommodation within a non-hierarchical framework of equality and non-discrimination. It follows that the state does not take sides on questions of religion. It does not impose belief, grant privileges to or impose disadvantages on adherents of any particular belief, require conformity in matters simply of belief, involve itself in purely religious controversies, or marginalise people who have different beliefs.
State bias in the pre-constitutional period
 In the pre-constitutional era there were a number of statutory provisions with a religious foundation that in no way purported to maintain neutrality in relation to “different confessional alignments”. According to Professor J D van der Vyver, writing in 1986, “[i]n cases where the legislature . . . expressed a particular religious preference it . . . clearly sided with Christianity.” He points out that the Publications Act 42 of 1974 “seemingly subject[ed] the entire censorship system to the dictates of Christian morality.” Furthermore, primary and secondary education in public schools for white children was based on the principle of Christian national education, while education in black schools had to have a Christian character. A further indication of Christian bias in the law was that the crime of blasphemy “applied to the slandering of the God confessed by Christianity only”. His survey goes on to point out that “[t]he Christian bias of certain branches of statutory law [was] also evidenced by a series of Sunday observance laws covering a wide range of regulative and prescriptive measures.” In broad outline, the legislation fell into two main categories, namely, commercial and labour law, and public entertainment and recreation. The former included restraints on retail trade on Sundays and a number of detailed Sabbatarian prescriptions relating to bills of exchange, pegging of claims to mineral rights, and conducting the business of butchers and fishmongers. The latter cluster of penal statutes declared it an offence on Sundays to show films or permit public entertainment at a place where an admission fee was charged, while in Natal horse-racing on this day was unlawful even if gratuitous, as was dancing at a place of amusement or recreation in the Orange Free State.
 It should be noted that in almost all of the above cases the restrictions extended not only to Sundays, but to what Professor Van der Vyver referred to as all public holidays “with a religious base”, namely, Good Friday, Ascension Day, the Day of the Vow and Christmas Day. Ascension Day is no longer a public holiday, and the Day of the Vow is now the Day of Reconciliation, so that the only holidays with a religious base that survive are Good Friday and Christmas Day.
 Not only did the state require observance of certain aspects of the Christian religion, it also refused to recognise the validity of marriages that did not conform to the Christian prototype. The identification of Christianity with what a judge called “civilized peoples” emphasized the role of the Christian religion as a specific source of values for the interpretation and development of the law. The hurt caused by the non-recognition of Hindu and Muslim marriages by the courts has been well documented. Comparing the old situation to the new, Farlam J recently indicated his agreement with the proposition that:
“. . . it is quite inimical to all the values of the new South Africa for one group to impose its values on another and that the Courts should only brand a contract as offensive to public policy if it is offensive to those values which are shared by the community at large, by all right-thinking people in the community and not only by one section of it. It is clear, in my view, that in the Ismail case the views (or presumed views) of only one group in our plural society were taken into account.”
The contract to which he referred related to property arrangements pursuant to a Muslim marriage. In a later case, Mahomed DP speaking for this Court had this to say on the lack of recognition accorded to Muslim marriages:
“Unions which have been solemnised in terms of the tenets of the Islamic faith, for example, are not recognised in our law because such a system permits polygamy in marriage. It matters not that the actual union is in fact monogamous. As long as the religion permits polygamy, the union is ‘potentially polygamous’ and for that reason, said to be against public policy. The result must therefore be that the father of a child born pursuant to such a religious union would not have the same rights as the mother in adoption proceedings pursuant to s 18 of the Act. The child would not have the status of ‘legitimacy’ and the consent of the father to the adoption would therefore not be necessary, notwithstanding the fact that such a union, for example under Islamic law, might have required a very public ceremony, special formalities and onerous obligations for both parents in terms of the relevant rules of Islamic law applicable.” (Footnotes omitted)
 The marginalisation of communities of Hindu and Muslim persuasion flowed from and reinforced a tendency for the norms of “Christian civilisation” to be regarded as points of departure, and for Hindu and Muslim norms to be relegated to the space of the deviant “Other”. Any echo today of the superior status in public law once enjoyed by Christianity must therefore be understood as a reminder of the subordinate position to which followers of other faiths were formerly subjected. Indeed, the concern expressed by O’Connor J about the message sent by state endorsement of religion to non-adherents to the effect that they are outsiders and not full members of the political community, has special resonance in South Africa. Religious marginalisation in the past coincided strongly in our country with racial discrimination, social exclusion and political disempowerment. Similar although not identical observations may be made about anti-semitism, which targeted members of the Jewish community for disadvantageous treatment in the public as well as the private sphere. Thus, any endorsement by the state today of Christianity as a privileged religion not only disturbs the general principle of impartiality in relation to matters of belief and opinion, but also serves to activate memories of painful past discrimination and disadvantage based on religious affiliation.
 Professor Laurence H Tribe points out further, correctly in my view, that any actual or perceived alliance between government and religion can undermine free political discourse. “The more political leaders wrap themselves in the mantle of religion,” he writes “the more readily those who oppose them may be accused of opposing God. That, in turn, may polarize citizens and leaders around a religious axis, creating the sort of divisiveness that the first amendment was partly intended to minimize.” Finally, we should remember that the movement for freedom of belief has preceded every other in the history of the struggle for human rights and fundamental freedoms, while conversely, religious persecution, sectarian strife, and ideological totalitarianism have undermined democracy and respect for fundamental rights in many parts of the world. State enforcement of a particular belief or ideology can in an extreme case do more than marginalise citizens and block free debate, it can threaten the whole system of constitutional democracy. The present case, of course, comes nowhere near raising these distressing spectres, but it does highlight how sensitive these matters are, and how potentially deep the implications of apparently harmless provisions may be. Painful history in our country and abroad reminds us that the values underlying section 14 can never be taken for granted, and must always be jealously nurtured. At the same time the wide range of matters covered also indicates how broad and varied the spectrum of potential violations is, starting with the most minor infraction which barely impinges upon the values protected by the section, and extending to the most egregious invasion which threatens the whole constitutional order. The appropriate location of the challenged law in this spectrum will accordingly have a significant influence on the way in which it will to be evaluated.
Competitive disadvantage because of Religion
 As I have said, although the section 14 issue of principle is real, the way it came to us was artificial. The objective was to abolish a commercial restraint, not to secure a religious freedom. Thus, the matter before us arises out of a prosecution of an employee of a grocery chain store whose actual complaint was that she was compelled by the state not to sell liquor on a Sunday. She did not allege that she was obliged by her religion not to sell liquor on a day other than Sunday as well, and, as a result of her belief, subjected by the state to an invidious choice between following her religion or pursuing her trade. Nevertheless, it was not a precondition for her bringing of the case that she establish that her own rights of religion, belief or opinion were trespassed upon. It was sufficient for her to complain that her rights were infringed as a result of her being prosecuted in terms of a statutory provision which, objectively speaking, was invalid because it violated section 14.
 As a result, however, of what appears to have been the tacking on of a complaint of Sabbatarian disadvantage to a more general charge of unconstitutional commercial regulation, no evidence was placed before us to indicate whether in practice there actually was such competitive disadvantage. Furthermore, we have no clear factual foundation for deciding whether, if it did exist, it was substantial or trivial. General knowledge does not provide much help. Because the tenets of Islam, as I understand them, prohibit the use or distribution of liquor at any time on any day, Muslim shopkeepers could hardly be expected to complain. Orthodox Jews might feel that by closing their stores on Saturday and not selling liquor on Sunday as well, they were at some commercial disadvantage because Sunday was chosen as the closed day and not Saturday. Whether or not Sabbatarian exemptions should be provided for in the case of complete cessation of trading by grocers on Sundays is not the issue before us. One could hardly require the state to do a survey of Sabbatarian practice and then choose as a closed day the one that caused the least inconvenience to believers. It could be argued that exemptions would involve the state in invidious determinations of religious persuasion and subject the affected persons to the embarrassment of being forced to make special applications which would emphasise rather than diminish their non-majority status. Yet in the present case we are not required to enter this difficult terrain. We are dealing only with one item on the sale racks that is locked away while the customers push their trolleys through massed ranks of foodstuffs inviting purchase, so that any competitive disadvantage suffered, say, by a Jewish storekeeper, who because of religious observance, closed his or her shop on a Saturday, would indeed be trivial.
 The mere fact that the closed day coincides with a day that has its origins in Christian practice cannot automatically mean that it continues to serve the sectarian purpose of compelling observance of that day as a Christian day of rest. Even if Sunday continues to have special religious significance for many South Africans, it has also become secularised as a common pause day for believers of all persuasions and believers of none. Its special position has been recognised in labour law. Its origins in Christianity have no more intrinsically sectarian consequences than does the use of terms AD and BC to establish dates in our calendar. Accordingly, I find it difficult to accept that state-imposed temperance on a common pause day is in itself enough to implicate section 14 simply on the grounds that that day of rest originated from and continues to coincide with the Christian sabbath.
 It is not always easy to distinguish between observances and practices that are purely sectarian, those that are completely secular and those that combine elements of both. In this respect, I would associate myself with the broad sweep of the remarks of Brennan J in the context of the case which arose out of the public display by a municipality of a nativity scene at Christmas-time.
“Intuition tells us that some official ‘acknowledgement’ is inevitable in a religious society if government is not to adopt a stilted indifference to the religious life of the people . . . It is equally true, however, that if government is to remain scrupulously neutral in matters of religious conscience, as our Constitution requires, then it must avoid those overly broad acknowledgments of religious practices that may imply governmental favoritism toward one set of religious beliefs.
. . . .
[A]t least three principles – tracing the narrow channels which government acknowledgments must follow to satisfy the Establishment Clause – may be identified. First, although the government may not be compelled to do so by the Free Exercise Clause, it may, consistently with the Establishment Clause, act to accommodate to some extent the opportunities of individuals to practise their religion. . . . [T]hat principle would justify government’s decision to declare December 25th a public holiday.
. . . .
Second . . . while a particular governmental practice may have derived from religious motivations and retain certain religious connotations, it is nonetheless permissible for the government to pursue the practice when it is continued today solely for secular reasons. . . . [T]he mere fact that a governmental practice coincides to some extent with certain religious beliefs does not render it unconstitutional. Thanksgiving Day, in my view, fits easily within this principle, for despite its religious antecedents, the current practice of celebrating Thanksgiving is unquestionably secular and patriotic.
. . . .
Finally, we have noted that government cannot be completely prohibited from recognizing in its public actions the religious beliefs and practices of the American people as an aspect of our national history and culture. . . . While I remain uncertain about these questions, I would suggest that such practices as the designation of ‘In God we Trust’ as our national motto, or the references to God contained in the Pledge of Allegiance to the flag can best be understood . . . as a form of ‘ceremonial deism,’[] protected from Establishment Clause scrutiny chiefly because they have lost through rote repetition any significant religious content.” (Footnotes omitted)
 I am not persuaded, therefore, that the selection of Sunday as a closed day either imposes unacceptable commercial disadvantage on non-Christians in a constitutionally meaningful sense, or that it results in state-imposed observance of the Christian sabbath in any significant way.
 The main problem, as I see it, lies elsewhere, and arises from other considerations. The crucial aspects are, firstly, that all of the closed days, namely, Sundays, Good Friday and Christmas Day, are what Professor van der Vyver in pre-constitutional times called “religiously-based holidays”, and secondly, that the prohibition relates simply to the sale of liquor. Putting those two factors together, there appears to be no getting away from the inference that although part of the objective might have been purely secular, the means used, namely the selection of religiously-based days as closed days, was intended to acknowledge and comply with the sentiments of those Christians who regarded these days as days requiring special observance. The identification of these days suggests that the manifest object was not simply to serve as a means of economic regulation or as a way of achieving a measure of temperance at selected times. The sectarian message might not be powerful, but it is inescapable. Had Sundays and all public holidays been included, the situation would have been different, and the choice of days could have been considered neutral; had only Sundays been referred to, the signal would have been mixed. The fact is that Sundays have been coupled with Good Friday and Christmas Day.
 My view then is that the identification of Sundays, Good Friday and Christmas Day as closed days for purposes of selling liquor, does involve an endorsement by the state of the Christian religion in a manner that is problematic in terms of section 14. The functional impact of the law may be marginal, and its symbolic effect muted, yet the communication it makes cannot be disregarded. Even if there is clear scope for the application of the de minimis rule to the question of some ancillary economic costs resulting from being true to one’s faith, it should be used with extreme caution when it comes to deciding such sensitive and not easily measurable questions as freedom of conscience, religion and belief. One of the functions of the Constitution is precisely to protect the fundamental rights of non-majoritarian groups, who might well be tiny in number and hold beliefs considered bizarre by the ordinary faithful. In constitutional terms, the quality of a belief cannot be dependent on the number of its adherents nor on how widespread or reduced the acceptance of its ideas might be, nor, in principle, should it matter how slight the intrusion by the state is. The objective of section 14 is to keep the state away from favouring or disfavouring any particular world-view, so that even if politicians as politicians need not be neutral on these questions, legislators as legislative drafters must.
 The strength of the O’Connor J’s approach, namely its all-encompassing character which lifts it out of formulaic reasoning and combines the relationship between purpose and effect, also appears to be its weakness. It indicates the broad question to be asked, but not the specific criteria to be used for the answer. More especially, it does little to establish from whose standpoint the message by the state should be considered. What comes through as an innocuous part of daily living to one person who happens to inhabit a particular intellectual and spiritual universe, might be communicated as oppressive and exclusionary to another who lives in a different realm of belief. What may be so trifling in the eyes of members of the majority or dominant section of the population as to be invisible, may assume quite large proportions and be eminently real, hurtful and oppressive to those upon whom it impacts. This will especially be the case when what is apparently harmless is experienced by members of the affected group as symptomatic of a wide and pervasive pattern of marginalisation and disadvantage.
 In testing whether in the present case the state endorsed a particular set of beliefs in a manner which violated section 14, I shall attempt to apply the sensibilities and perspectives neither of what has been called the “reasonable Christian”, nor, for example, of the reasonable Jew, Muslim, or Hindu, nor of the reasonable atheist, but of the reasonable South African (of any faith or of none) who is neither hyper-sensitive nor overly insensitive to the belief in question, but highly attuned to the requirements of the Constitution. In my opinion, such a reasonable South African is a person of common sense immersed in the cultural realities of our country and aware of the amplitude and nuanced nature of our Constitution. He or she neither attempts relentlessly to purge public life of even the faintest association with religion for fear of otherwise descending the slippery slope to theocracy, nor, at the other extreme, regards the religiously-based practices of the past to be as natural and non-sectarian as the air one breathes simply because of their widespread acceptance.
 I accordingly endorse the compactly expressed approach of Farlam J when he holds that:
“. . . it is quite inimical to all the values of the new South Africa for one group to impose its values on another and that the Courts should only brand a contract as offensive to public policy if it is offensive to those values which are shared by the community at large, by all right-thinking people in the community and not only by one section of it.” (My emphasis)
Such right-thinking persons, in my view, would have little difficulty in accepting that whatever may be their deep and continuing special significance for Christians, the survival of Sunday, Good Friday and Christmas Day as secularised public holidays integrated into the programmes of rest, travel and, in the case of Christmas Day, festivity, of all South Africans, no longer represents state endorsement of religion. At the same time, however, they would have equally few doubts that the choice of these days as closed days for the purposes of the sale of liquor, and not to establish a common day of rest, does indicate a maintenance of the pre-constitutional sectarian bias referred to by Professor Van der Vyver. Accordingly, adopting the approach of the reasonable non-sectarian South African conscious of the values of openness and tolerance enshrined in the Constitution, aware of the importance in this area of not regarding a majoritarian view as a national one and sensitive to the need to show special regard for the sensibilities of those who may feel excluded or offended by the measure, I come to the following conclusion: the inescapable message sent out by the particular choice of these closed days is that despite the enactment of section 14, the state still shows special solicitude to Christian opinion or, to put it more accurately, to the views of certain Christians, and thereby infringes section 14.
Severance and the limitations clause
 As I have stated above, it is the conjunction of Good Friday and Christmas Day with Sunday that manifests a state endorsement of Christianity as a religion requiring special observance and meriting more respect than other religions. Implicit in this is the assumption that Christians occupy central positions in the political kingdom, while non-Christians live on the periphery. If these two public holidays were severed from the rest of the definition of closed days, then, in my view, the identification of Sunday as a closed day would not implicate section 14. Its legitimate secular purpose would then be shorn of its illegitimate sectarian one. In the present case, however, severance was not urged upon us, and we have not heard argument about it. Furthermore, the prosecution related to selling liquor on a Sunday and not to selling it on Christmas Day and Good Friday. Counsel brought these two days into the picture basically to serve as colourative proof that Sunday was chosen not as a mutual day of rest but because of its Christian associations. For the reasons which follow, however, I do not find it necessary to decide whether severance would be either appropriate or competent in the present case.
 In my view, the application of section 33 resolves the matter. I have stated that the de minimis rule may have little application in deciding whether or not there has been an infringement of section 14 by virtue of state endorsement of a particular religious creed. I nevertheless believe that the degree of infringement is highly relevant in the balancing process involved in the second phase of the enquiry, namely, the test of proportionality required when considering whether the intrusion against the right qualifies in terms of section 33 as a reasonable, justifiable and necessary one. This Court has on a number of occasions emphasized that the test established by section 33 relies on proportionality, a process of weighing up the individual’s right which the state wishes to limit against the objective which the state seeks to achieve by such limitation. As Langa J put it in S v Williams and Others:
“This evaluation must necessarily take place against the backdrop of the values of South African society as articulated in the Constitution and in other legislation, in the decisions of our Courts and, generally, against our own experiences as a people”.
 The requirement that limitations on section 14 rights must be not only reasonable and justifiable, but also necessary, clearly identifies section 14 as one of the core provisions of the bill of rights requiring special solicitude by this Court. It reduces the margin of appreciation which a test of reasonableness on its own would allow the legislature, and places special emphasis on the selection of options which are clearly not unduly burdensome, overbroad or excessive, considering all the reasonable alternatives. In S v Makwanyane and Another Chaskalson P emphasised that there was no absolute standard which could be laid down for determining reasonableness and necessity; principles could be established, but the application of those principles to particular circumstances could only be done on a case by case basis. The importance of looking at the actual context in which an alleged infringement is to be evaluated was further underlined in President of the Republic of South Africa and Another v Hugo where Goldstone J referred to the need for “. . . a careful and thorough understanding of the impact of the discriminatory action upon the particular people concerned. . . . A classification which was unfair in one context may not necessarily be unfair in a different context.”
 The reason why context is so important in constitutional matters is well explained by Wilson J in Edmonton Journal v Alberta AG:
“. . . a particular right or freedom may have a different value depending on the context. It may be, for example, that freedom of expression has greater value in a political context than it does in the context of disclosure of the details of a matrimonial dispute. The contextual approach attempts to bring into sharp relief the aspect of the right or freedom which is truly at stake in the case as well as the relevant aspects of any values in competition with it. It seems to be more sensitive to the reality of the dilemma posed by the particular facts and therefore more conducive to finding a fair and just compromise between the two competing values . . . .”
In deciding what is reasonable and necessary in the present case we should accordingly look to the actual dilemma triggered by its particular facts, and not deal with it in a formulaic way simply because section 14 has been infringed.
 Although the section 14 right is in general a weighty one, not each and every breach of the right carries the same weight. A trivial breach of a specially protected right might be easier to justify in terms of section 33 than a grievous infringement of an “ordinary” right. The intensity or severity of the breach must accordingly be a highly relevant factor in any proportionality exercise; the more grievous the invasion of the right, the more compelling must be its justification. Conversely, the lighter the transgression, the less stringent the requirements of justification. Thus, I have no doubt that any state action which interfered directly with or compelled a particular form of religious observance would rarely pass the tests of reasonableness and necessity, if at all, and then only if the most compelling justificatory circumstances were established. Indeed, there is a core to the individual conscience so intrinsic to the dignity of the human personality that it is difficult to imagine any factors whatsoever that could justify its being penetrated by the state. At the other extreme, there are transgressions of section 14 so marginal in themselves, or so slight in relation to manifestly legitimate objectives with which they are inextricably interlinked, that the burden of persuasion on the facts imposed by section 33 would be far easier to discharge.
 With these considerations in mind I turn to the concrete act of balancing that we are required to do in the present case. I start with an attempt to characterise the severity of the invasion of the right. We may assume, even though it has not been proved, that there is some but not very significant competitive disadvantage suffered by persons whose sabbath happens to be on days other than Sunday, Good Friday and Christmas Day. Such persons, as well as Christians who feel that their religious observances are matters of private confession that have nothing to do with the state, are obliged by the state not to sell liquor on these days because of deference to a particular form of Christian sensibility. Yet, the activities involved are so limited that I cannot regard any economic disadvantage flowing from belief as being substantial at all, and I doubt whether they even enter the scales of proportionality.
 Of greater significance is the signal given to the public at large that the state regards the Christian religion as worthy of special respect above other religions. In the words of Professor Tribe, government’s gratuitous use of a religious means is likely to convey a message of exclusion to all those who do not adhere to the favoured religion; when such people learn that government has gone out of its way to adopt the religion’s tools, they may believe that government must have adopted its tenets as well, and quite reasonably feel, in O’Connor J’s words, as if they are not full members of the political community.
 This is where the problem of weighing things of a completely different order arises. The difficulty which we must overcome is how to assess the intrinsically intangible, but very real (even if rather reduced) symbolic effect of religious favouritism, as against the very palpable and quite terrible consequences of alcohol abuse which the state wishes to diminish. I have already indicated the factors which led me to the conclusion that the selection of days chosen by the state amounted to endorsement of religion in a manner that breached section 14. I will now consider in its context how grave that breach is, or to put it another way, how powerful or weak the exclusionary message is which the state is sending by such endorsement.
 To begin with, the negative symbolic effect of such state favouring of Christianity must be seen in its legislative and historical context. The closed days are a small part of a statute designed to control the sale of liquor and not, as in the Big M case, a central aspect of a statute primarily intended to compel religious observance. Secondly, the challenged provisions do not impose beliefs or interfere in a direct way with observance. They relate to a situation of favouritism coupled with indifference, rather than one of orthodoxy combined with persecution, and represent a relatively insignificant relic of a vanishing era, rather than a pungent symbol of continuing hegemony. Even the objectionable pre-constitutional period referred to in my historical survey was not remotely as egregious in respect of religion as it was, for example, in relation to race. The result is that section 14 must be construed in a context of a legacy of institutionalised religious favouritism that is far less pronounced and pervasive than, say, the systematic racism and sexism which must influence the meaning we must give to the equality provisions in section 8. Then, just as the economic effects on non-Christians are at most of a marginal character, since grocers are not obliged to close their stores on certain days, but merely required not to sell liquor on them, so the symbolic effect of compulsory observance is correspondingly reduced.
 In the fourth place, the signal was further muted by the fact that prohibition of the sale of liquor on closed days was far from complete, since the legislation in question made provision for sales on a Sunday by persons in possession of hotel, wine house or club house licences, or persons specially granted licences to sell liquor on Sundays. The more modulated the prohibition, the less manifestly Sabbatarian the message. Fifthly, as has already been pointed out, the days in question have become highly secularised. Shops, cinemas, sportsgrounds and jazz clubs do thriving business on Sundays. Significant though Good Friday is to a great many practising Christians, the Easter weekend has become for the South African population as a whole a time for enjoying late summer holidays. To the regret of many Christians, Christmas Day has become a holiday when, as it has been said, more homage is paid to Mammon than to God. The result is that choosing these days as closed days now sends out a far less powerful signal than it once would have done. Finally, there is relatively little sectarian significance to the prohibition of the sale of liquor as such.
 The overall consequence is a law that, while indeed offending against section 14, does so in an indirect and marginal way, imposing relatively little obligatory observance, in respect of a matter of slight sectarian import, in relation to days that have become highly secularised. The message of inclusion coupled with exclusion is accordingly a notably subdued and insubstantial one.
 Balanced against these relatively minor infractions of section 14 are strong factors which operate to justify the singling out of Sunday, Good Friday and Christmas Day as days when the state makes special legislative attempts to encourage temperance. O’Regan J correctly points out in her lucid analysis that no evidence in support of section 33 justification was led. Yet, there are facts of common knowledge to which we cannot blind ourselves. Sunday comes at the end of the weekend, consisting of two pause days, the first of which, with Friday night, is a time of relatively heavy drinking. Good Friday represents the first day of the Easter weekend which is a period when there is exceptionally high traffic on the roads and when drunken driving constitutes a specially serious menace. Christmas Day is the first of two public holidays widely dedicated to festivity. In my view, the state’s interest on behalf of society in encouraging temperance on these particular days is a powerful and legitimate one.
 Although contested evidence was placed before us to the effect that regulation of the sale of liquor in general has failed to reduce the damage caused by alcohol abuse, I am unaware of anything on record, or any information of common knowledge, which suggests that once such regulation is regarded as legitimate governmental activity, and once the purpose of reducing alcohol intake on these particularly high-risk days is treated as an appropriate and compelling governmental objective, then such objectives could reasonably have been achieved by less intrusive means. In this respect it is highly relevant that it is only the purchase of liquor in transportable form that is made difficult, not the purchase of liquor as such.
 My conclusion, then, is as follows: on the one hand, the scope and intensity of the invasion of section 14 rights is relatively slight. On the other hand, the dangers of excessive drinking, particularly on weekends, at the beginning of the Easter weekend and at Christmas-time, are grave. Pay packets are reduced, domestic violence is intensified and exceptionally high slaughter on the roads resulting from drunken driving becomes a matter of national concern. There are accordingly strong reasons for adopting suitably focused measures which are designed to and hopefully will restrict the consumption of alcohol on these particular days and not on others. I accordingly feel that in the particular circumstances of this case the legislative restrictions in question are both reasonable and necessary.
 The result is that I agree with O’Regan J that the provisions relating to closed days involve a breach of section 14. Since, however, I am of the opinion that such infringement is sanctioned by section 33, I concur with Chaskalson P in his conclusion that the provisions in question are not unconstitutional.
 In doing so, I must state that although in general I support the spirit of realism and common sense which I find in his judgment, I cannot agree with the observation in para 104 that coercion, whether direct or indirect, must be established before section 14 can be said to have been breached. The state as the state is animated by the values expressed or implied in the Constitution, and by those alone. By endorsing a particular faith as a direct and sectarian source of values for legislation binding on the whole nation, it exceeds the competence granted to it by the Constitution. Even if there is no compulsory requirement to observe or not to observe a particular religious practice, the effect is to divide the nation into insiders who belong, and outsiders who are tolerated. This is impermissible in the multi-faith, heterodox society contemplated by our Constitution.
 I agree with Chaskalson P’s judgment in relation to the other issues raised, and accordingly concur in the order he proposes.
Mokgoro J concurs in the judgment of Sachs J.
For the Appellants: P Hodes SC and AM Breitenbach instructed by Bill
Tolken Hendrikse Inc
For the First Respondent: B Morrison SC and J Slabbert instructed by the
Attorney General, Cape Town
For the Second Respondent: JL van der Merwe SC and SK Hassim instructed by
the State Attorney, Pretoria
For the Amicus Curiae (SALSA): L Wessels instructed by Du Plessis, de Heus and van Wyk
Section 88(1) of the Liquor Act.
“If in any proceedings before a court referred to in subsection (1), the presiding officer is of the opinion that it is in the interest of justice to do so, he or she may postpone the proceedings to enable the party who has alleged that a relevant law or provision is invalid, to apply to a provincial or local division of the Supreme Court for relief in terms of subsection (4).”
Section 103(2) stipulates:
“If in any proceedings before a court referred to in subsection (1), it is alleged that any law or provision of such law is invalid on the ground of its inconsistency with a provision of this Constitution and the court does not have the competency to enquire into the validity of such a law or provision, the court shall, subject to the other provisions of this section, decide the matter on the assumption that the law or provision is valid.”
That section provides:
“If, in any matter before a provincial or local division of the Supreme Court, there is an issue which may be decisive for the case, and which falls within the exclusive jurisdiction of the Constitutional Court in terms of section 98(2) and (3), the provincial or local division concerned shall, if it considers it to be in the interest of justice to do so, refer such matter to the Constitutional Court for its decision: Provided that, if it is necessary for evidence to be heard for the purposes of deciding such issue, the provincial or local division concerned shall hear such evidence and make a finding thereon, before referring the matter to the Constitutional Court.”
Rule 21(1) of the Constitutional Court Rules states that if:
“(a) an accused has appealed unsuccessfully to a provincial or local division of the Supreme Court against a conviction or sentence imposed on him or her in the magistrate’s court; and
(b) the grounds of appeal include a constitutional issue within the exclusive jurisdiction of the Court; and
(c) the accused wishes to appeal against the conviction or sentence solely on the grounds of such constitutional issue,
the accused shall be entitled to appeal on such grounds to the Court against the conviction or sentence.”
Section 40 of the Liquor Act. In terms of section 41(1)(c) a liquor business would include the sale of “mineral waters, other drinks (other than liquor as defined in section 2(1)), tobacco, cigars, cigarettes, matches, [and] cooler bags . . .”. The list of products may be extended by the chairperson of the provincial liquor board.
Section 87. The terms of section 87 are set out in paragraph 6 below.
Section 99. Any business, other than the sale of liquor, may be conducted on premises licensed for the sale of sorghum beer.
“If the validity of a law is in dispute in any matter, and a relevant government is not a party to the proceedings, it shall be entitled to intervene as a party before the court in question, or shall be entitled to submit written argument to the said court.”
The Attorney General prosecutes in the name of the state under a constitutional authority to do so (section 108(1) of the interim Constitution; section 179(2) of the Constitution of the Republic of South Africa, 1996 Act 108 of 1996 (“the 1996 Constitution”)), and represented the state, not the “government” in the prosecution of the appellants and in opposing their appeals.
Paras 17 – 21 and 22 – 23.
See above n 6.
Section 100(3) of the interim Constitution.
Item 17 of schedule 6 to the 1996 Constitution provides that “All proceedings which were pending before a court when the new Constitution took effect, must be disposed of as if the new Constitution had not been enacted, unless the interests of justice require otherwise.” The appellants rely on this provision for their arguments based on section 26 of the interim Constitution which they see as being more favourable to their case than the comparable provision of the 1996 Constitution, section 22.
See rule 33 which incorporates by reference the provisions of section 22 of the Supreme Court Act 59 of 1959.
It makes provision for the admission of “legislative facts” and other material that may be common cause.
Moulded Components and Rotomoulding South Africa (Pty) Ltd v Coucourakis and Another 1979 (2) SA 457 (W) at 462H–463B; Krygkor Pensioenfonds v Smith  ZASCA 47; 1993 (3) SA 459 (A) at 469E–I.
Id at para 104.
Section 35(1) requires the provisions of the bill of rights to be interpreted so as to promote such values.
These are merely illustrations of the application of section 26 and should not be construed as a definition of what is permitted. The restrictions could be imposed on a regional basis which would have a bearing on the provision that economic activity can be conducted “anywhere”.
Paragraphs 54 – 56 and 65 – 70.
“This section shall not preclude measures designed to achieve the adequate protection and advancement of persons or groups or categories of persons disadvantaged by unfair discrimination, in order to enable their full and equal enjoyment of all rights and freedoms.”
English Dictionary vol IV 2 ed (Clarendon Press, Oxford 1989) meaning 10.
Professor Mureinik does not indicate the source of the meaning
so as to achieve” which he explains as meaning
“objectively probable that they will in fact achieve”. None of
meanings given to “design” in the Oxford Dictionary has this
connotation. The only reference to the design of laws
in the Oxford Dictionary
is a quotation from Burke given in respect of the meaning accepted in this
judgment: “Ask of politicians
the end for which laws were originally
designed; and they will answer, that the laws were designed as a protection for
the poor and
“Proof of Facts in Constitutional Cases” (1976) 26 University of Toronto law Journal 386 at 396–7.
Carmichael, Attorney General of Alabama v Southern Coal & Coke Co  USSC 108; 301 US 495, 510 (1937). See also FCC v Beach Communications, Inc  USSC 64; 508 US 307, 315 (1993), and Lehnhausen, Director, Department of Local Government Affairs of Illinois v Lake Shore Auto Parts Co  USSC 64; 410 US 356, 364–5 (1973).
S v Makwanyane above n 28 at paras 103–4.
See paragraphs 26 – 30 above.
There is another possible construction, namely that the party that relies on the law has to show that it is designed to meet a purpose specified in section 26(2), and if this is done, that the party who contends that a particular provision of the law does not serve that purpose, to establish that there is no rational connection between the purpose and the provision. Compare Basu Constitutional Law of India 6 ed (1991) at 39.
In the United States the courts do not require evidence on this issue. “[A] legislative choice is not subject to courtroom factfinding and may be based on rational speculation unsupported by evidence or empirical data.” (FCC v Beach Communications, Inc above n 38 at 315). See also Cabinet for the Territory of South West Africa v Chikane and Another 1989 (1) SA 349 (A) at 368A–369D and 382H–383F. As the Beach Communications case shows (at 314–5) the United States Supreme Court requires those challenging the constitutionality of legislation under a rational basis review “to negative every conceivable basis which might support it” (quoting from Lehnhausen v Lake Shore Auto Parts Co above n 38 at 364).
University of Toronto Law Journal above n 37 at 395.
Paragraphs 67 – 70.
In fact this was acknowledged by one of their experts – Mr Makan – in the affidavits lodged by the appellants and was “common cause” within the meaning of rule 34(1)(a). It was presumably for this reason that the challenge was to the statutory conditions attaching to grocers’ wine licences and not to the licensing system as such.
Section 85(1) of the Act.
In the United States the burden of negativing a rational connection between the legislation and a legitimate government purpose is on the person challenging the validity of the legislation. “[T]hose attacking the rationality of the legislative classification have the burden ‘to negative every conceivable basis which might support it,’” FCC v Beach Communications Inc above n 38 at 315. Whilst a section 26 analysis may call for a different approach a rational basis review is one in which the legislature will be allowed considerable lee-way.
Section 41 identifies the products other than liquor that may be sold on various types of licensed premises. It does not deal with products that may be sold on premises in respect of which a grocer’s wine licence has been issued. That is dealt with in section 87.
Section 99 deals with the sale of sorghum beer and is not relevant to this appeal.
Section 2(c) of Act 57 of 1995.
Id at 95.
Id at 93.
Id at 99. My emphasis.
Section 54(1)(a) of the Act.
Section 57 of the Act.
Section 61 of the Act.
Section 65 of the Act.
Section 54(1)(b) of the Act.
Sections 57 and 61 of the Act.
Section 63(1) of the Act.
Section 72(1) of the Act.
Section 85 of the Act.
Section 93 of the Act.
Section 100 of the Act.
Section 77 of the Act.
Section 81 of the Act.
Section 103 of the Act. A producer’s licence authorises the sale of wine and other alcoholic beverages by the “manufacturers” of such liquor.
Section 20(b)(viii) read with section 22(2)(b) of the Act.
If the restriction in its application to Christmas Day or Good Friday should be seen as having a purely religious purpose which infringes the freedom of religion of non-Christians, I can see no reason why the reference to those days cannot be severed from the definition leaving only Sundays as closed days. That issue does not, however, arise in the present case.
Above n 58 at 97.
See above n 57.
The Lord’s Day Act dealt with in the Big M case above n 58 is an example of such legislation.
Engel v Vitale  USSC 116; 370 US 421 (1962). The prayer read as follows: “Almighty God, we acknowledge our dependence upon Thee, and we beg Thy blessings upon us, our parents, our teachers and our Country.”
School District of Abington Township, Pennsylvania v Schempp; Murray v Curlett  USSC 162; 374 US 203 (1963). Provision was made in the legislation for children to be excused from the reading upon parental request.
Above n 2 at 431.
See, for example, Brennan J’s concurrence in McDaniel v Paty  USSC 61; 435 US 618 (1978) where he held that “government [may] take religion into account . . . to exempt, when possible, from generally applicable governmental regulation individuals whose religious beliefs and practices would otherwise thereby be infringed, or to create without state involvement an atmosphere in which voluntary religious exercise may flourish.” (At 639, footnote omitted.) See also O’Connor J’s concurring opinion in Wallace v Jaffree; Smith v Jaffree 472 US 38, 83 (1985). And the decision of the Canadian Supreme Court in Edwards Books and Art Ltd v The Queen (1986) 28 CRR 1.
“Everyone shall have the right to freedom of thought, conscience and religion. This right shall include freedom to have or to adopt a religion or belief of his choice, and freedom, either individually or in community with others and in public or private, to manifest his religion or belief in worship, observance, practice and teaching.”
It appears that in the United States during the Prohibition era, the sale of communion wine was not prohibited. See Stone et al Constitutional Law 3 ed (Little, Brown and Company, Boston 1996) at 1608.
“Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof . . . .”
Mayor of Pawtucket v Donnelly  USSC 47; 465 US 668, 687–8 (1984).
In this respect I would like to endorse the views on “Sundays” of Dickson CJC in R v Big M Drug Mart Ltd (1985) 13 CRR 64 at 108:
“I would like to stress that nothing in these reasons should be read as suggesting any opposition to Sunday being spent as a religious day; quite the contrary. It is recognised that for a great number of Canadians, Sunday is the day when their souls rest in God, when the spiritual takes priority over the material, a day which, to them, gives security and meaning because it is linked to Creation and the Creator. It is a day which brings a balanced perspective to life, an opportunity for man to be in communion with man and with God. In my view, however, as I read the Charter, it mandates that the legislative preservation of a Sunday day of rest should be secular, the diversity of belief and non-belief, the diverse socio-cultural backgrounds of Canadians make it constitutionally incompetent for the federal Parliament to provide legislative preference for any one religion at the expense of those of another religious persuasion”.
The danger to constitutional analysis is that if you ask an artificial question you get an artificial answer.
See Redlich et al Constitutional Law 3 ed (Matthew Bender, New York 1996) at 1527.
Id at paras 18–9.
“In interpreting the provisions of this Chapter a court of law shall promote the values which underlie an open and democratic society based on freedom and equality and shall, where applicable, have regard to public international law applicable to the protection of the rights entrenched in this Chapter, and may have regard to comparable foreign case law.”
“The rights entrenched in this Chapter may be limited by law of general application, provided that such limitation –
(a) shall be permissible only to the extent that it is –
(i) reasonable; and
(ii) justifiable in an open and democratic society based on freedom and equality; and
(b) shall not negate the essential content of the right in question, and provided further that any limitation to –
(aa) a right entrenched in section 10, 11, 12, 14(1), 21, 25 or 30(1)(d) or (e) or (2); or
(bb) a right entrenched in section 15, 16, 17, 18, 23 or 24, in so far as such right relates to free and fair political activity,
shall, in addition to being reasonable as required in paragraph (a)(i), also be necessary.”
This is underlined by Constitutional Principle XII in schedule 4 which required the Constitutional Assembly, when drafting the new Constitution to ensure that:
“Collective rights of self-determination in forming, joining and maintaining organs of civil society, including linguistic, cultural and religious associations, shall, on the basis of non-discrimination and free association, be recognised and protected.” (My emphasis)
Du Plessis and Corder Understanding South Africa’s Transitional Bill of Rights (Juta & Co Ltd, Kenwyn 1994) at 157.
See schedule 3 to the Constitution.
West Virginia State
Board of Education v Barnette  USSC 130; 319 US 624, 642 (1943) per Jackson J giving
the opinion of the
“This Constitution provides a historic bridge between the past of a deeply divided society characterised by strife, conflict, untold suffering and injustice, and a future founded on the recognition of human rights, democracy and peaceful co-existence and development opportunities for all South Africans, irrespective of colour, race, class, belief or sex.
The pursuit of national unity, the well-being of all South African citizens and peace require reconciliation between the people of South Africa and the reconstruction of society.
The adoption of this Constitution lays the secure foundation for the people of South Africa to transcend the divisions and strife of the past, which generated gross violations of human rights, the transgression of humanitarian principles in violent conflicts and a legacy of hatred, fear, guilt and revenge.”
judgment of Dickson CJC in Big M above n 7 at 105 where he states that
what unites various principles and values “is the notion of the centrality
conscience and the inappropriateness of governmental intervention
to compel or to constrain its manifestation.”
Van der Vyver “Religion” in Joubert The Law of South Africa vol 23 (Butterworths, Durban 1986) at 197.
“[South Africans] have been subjected to a system of censorship which was intended to impose the Calvinist morality of a small ruling establishment on the entire population.”
Also quoted in the judgment of Mokgoro J in Case and Another v Minister of Safety and Security and Others; Curtis v Minister of Safety and Security and Others  ZACC 7; 1996 (3) SA 617 (CC); 1996 (5) BCLR 609 (CC) at para 11. See also Dlamini “Culture, Education, and Religion” in Van Wyk Rights and Constitutionalism 573 at 597:
“Although there has been no religious intolerance of Christianity in South Africa, the same cannot be said of other religions. Moreover, even when it comes to Christianity, the government in the past did not allow the free expression of Christian convictions. . . . Those who opposed [the] policies and practices [of apartheid] in the name of Christianity were ruthlessly suppressed.”
In both instances the appropriate legislation provided that the religious convictions of the parents and the pupils should be respected in regard to religious instruction and religious ceremonies. Van der Vyver above n 24 at 197.
Id at 199 and 198.
“Dancing” was defined in section 1 of the Control of Dancing Ordinance 12 of 1957 (OFS) as “to dance with a partner to the accompaniment of music”. See also Van der Vyver above n 24 at 199.
Above n 24 at 198.
See, generally, Sinclair The Law of Marriage vol 1 (Juta & Co Ltd, Kenwyn 1996) at 164–5.
Executors v The Master (Natal) 1917 AD 302 at 307 (per Innes CJ). See also
Ismail v Ismail 1983 (1) SA 1006 (A) at 1026; Kerr “Back to the
Problems of a Hundred or More Years Ago: Public Policy Concerning Contracts
Relating to Marriages
that are Potentially or Actually Polygamous” (1984)
101 South African Law Journal
In S Narayan (ed) The Selected Works of Mahatma Gandhi: Satyagraha in South Africa vol 3 (Navajivan Publishing House, India 1928) at 377–8, M K Gandhi refers to “the terrible judgment” in the Cape Supreme Court setting aside the practice of forty years, which
“. . . thus nullified in South Africa at a stroke of the pen all marriages celebrated according to the Hindu, Musalman and Zoroastrian rites. The many married Indian women in South Africa in terms of this judgement ceased to rank as the wives of their husbands and were degraded to the rank of concubines, while their progeny were deprived of their right to inherit the parents’ property. This was an insufferable situation for women no less than men, and the Indians in South Africa were deeply agitated”.
The shock to Indian women was so great that for the first time they joined in the Satyagraha campaign. Gandhi continued (at 388):
“It was an absolute pure sacrifice that was offered by these sisters, who were innocent of legal technicalities, and many of whom had no idea of country, their patriotism being based only upon faith. Some of them were illiterate and could not read the papers. But they knew that a mortal blow was being aimed at the Indians’ honour, and their going to jail was a cry of agony and prayer offered from the bottom of their heart, and was in fact the purest of all sacrifices.”
The “terrible judgment” is reported as Esop v Union Government (Minister of the Interior) 1913 CPD 133. Counsel for the government argued that “Mariam is in law the concubine and not the wife of the applicant.” (At 134). And Searle J said that “The courts of this country have always set their faces against recognition of these so-called Mahommedan marriages as legal unions”. (At 135).
See also Cachalia “Citizenship, Muslim family law and a future South African constitution: a preliminary enquiry” (1993) 56 THRHR 392 at 398–9.
“. . . the [Islamic] customs and the contract in question are contrary to public policy and are, consequently, unenforceable.”
For the purposes of the present matter it is not necessary to express any views on whether polygyny as such would be contrary to public policy as developed in the light of the values of the Constitution. The present case does not require us to examine the extent to which section 14 gives constitutional protection to practices of a cultural or economic character which have a basis in or are associated with religious belief.
Milton Shain The Roots of Anti-Semitism in South Africa (WUP, Johannesburg 1994) at 148 refers to the way in which anti-semitism moved from the private or ideational sphere into the public realm.
“Few of the major human religions have not at one time or another suffered persecution, or themselves persecuted – through the authority of a State in which they have become established – the members of other religions, or heretics within their own fold. For a substantial proportion of the worst atrocities perpetrated in recorded history, the ostensible justification has been the alleged need for the dominance or maintenance of one belief system rather than another. This is not the place to recite a catalogue of religious persecutions over the ages, let alone to describe the iniquities perpetrated either by, or against, any particular religious group. Suffice it to recall that the movement for ‘freedom of belief’ precedes every other in the history of the struggle for human rights and fundamental freedoms”.
The International Law of Human Rights (Clarendon Press, Oxford 1983) at 324.
This issue has divided
judges in a number of North American cases. See, for example, Braunfeld
above n 4, Edwards Books and Art Ltd v The Queen (1986) 28 CRR 1, and
Jones v The Queen (1986) 25 CRR
See Nowak and Rotunda Constitutional Law 5 ed (West Publishing Co, St Paul 1995) at 1315:
“To hold otherwise would be to require the state to pursue its goal of establishing a uniform day of rest by choosing a day when the least number of people might use the time to attend religious services.”
See, for example, section 10(1)(a) of the Basic Conditions of Employment Act 3 of 1983.
The French Revolutionary Calendar, described as “a splendid instrument for the protection of a national, republican religion against international Christianity”, lasted only from 1794 until 1804 when the Gregorian Calendar with its Saints’ Days and Church holidays was restored. See Gottschalk The Era of the French Revolution (1715–1815) (Houghton Mifflin Company, Cambridge 1957) at 259 and 341. The United Nations, containing member countries which officially recognise religions such as Buddhism, Islam and Judaism as constituting elements of their constitutional character, uses the internationally accepted calendar, though certain individual members do not. We could hardly be expected in South Africa to strike out the case number of the matter before us because the year 1997 is dated from the year given by Christians as that when Christ was born.
If Christian origin alone were enough to establish unconstitutionality, then pictures of religious inspiration would have to be removed from public galleries, and the masses, cantatas and requiems originally composed for performance in a Cathedral could no longer be played in the City Hall; the province of Kwa-Zulu Natal would have to shed the second part of its name, and Port St Johns would also have to secularise itself; more fundamentally, it cannot be said that obedience by the criminal law to the Biblical Commandment “Thou shalt not kill” involves an invasion of section 14.
Lynch above n 6 at 714–6. In reading this passage it should be noted that the US Establishment Clause is manifestly less accommodating to religion than is section 14 of the Constitution. Moreover, the difficulties of establishing bright lines in the area of religion and conscience are borne out by the fact that Brennan J and O’Connor J applied similar approaches in this case, but came to different conclusions, the former writing in dissent and favouring unconstitutionality of the display of the creche, the latter finding it not to be unconstitutional.
See Certification of the Constitution of the Western Cape, 1997 (CC) Case No CCT 6/97, 2 September 1997, not yet reported, at para 28.
Nor, for the same reason,
does it involve unfair discrimination on grounds of religion, conscience and
belief in terms of section
8(2). The issue of unfair discrimination on those
grounds was not fully argued before
See Tribe above n 39 at 1262:
“A more plausible dichotomy may be that between those government measures (1) that put an individual to a choice between adherence to religious duties and enjoyment of government benefits (or . . . avoidance of a government burden like criminal prosecution), and (2) those government measures that are not triggered by the religious choice in question but burden religious activity only in a manner ancillary to an undeniably secular choice.” (My emphasis)
“The injury caused by the government’s action . . . is that the State, in a school setting, in effect required participation in a religious exercise. . . . [T]he embarrassment and the intrusion of the religious exercise cannot be refuted by arguing that these prayers, and similar ones to be said in the future, are of a de minimis character.”
“. . . I do not see how we can separate an issue as to what is believed from considerations as to what is believable. . . . Religious symbolism is even used by some with the same mental reservations one has in teaching of Santa Claus or Uncle Sam or Easter bunnies or dispassionate judges.”
“The legislature may be able to divorce the secular Sunday from the religious Sunday of history, but the Orthodox Jew, the Seventh Day Adventist, and the atheist cannot.”
Tribe above n 39 at 1205 suggests that O’Connor J in the Lynch case asked the right question, but gave the wrong answer because she approached the matter from the point of view of “the reasonable Christian”.
See Wallace, Governor of Alabama v Jaffree; Smith v Jaffree 472 US 38, 76 (1985) where O’Connor J held in effect that something should be deemed sufficiently secular in aim if a fully informed, independent observer would judge it to have a predominantly secular purpose at the time it is challenged. See also Tribe above n 39 at 1205.
Above n 35 SALR at 707G–H; BCLR at 90F–G.
See section 33(1)(b)(aa) above n 15.
In Coetzee v Government of the Republic of South Africa; Matiso and Others v Commanding Officer, Port Elizabeth Prison, and Others  ZACC 7; 1995 (4) SA 631 (CC); 1995 (10) BCLR 1382 (CC) at paras 55–60, I dealt extensively with the way the term “necessary” had been interpreted when used in international human rights instruments and came to the conclusion that the term “necessary” was not made the subject of rigid definition, but rather regarded as implying a series of interrelated elements in which central place was given to the proportionality of the means used to achieve a pressing and legitimate public purpose.
See S v Bhulwana; S v
Gwadiso  ZACC 11; 1996 (1) SA 388 (CC); 1995 (12) BCLR 1579 (CC) at para 18 where
O’Regan J says: “The more substantial the inroad into fundamental
rights, the more persuasive the
grounds of justification must
See Hogg Constitutional Law of Canada 3 ed (Carswell, Scarborough 1992) at 861 where he says:
“The severity of the contravention [of a Charter right] would not be irrelevant, of course, because it would be harder to establish that a severe contravention was reasonable and demonstrably justified.”
Compare Edwards Books above n 41 at 38 (competitive pressure on retailers to abandon the observance of a Saturday sabbath could not be characterised as insubstantial or trivial) and Jones above n 41 at 85 (legislative or administrative action whose effect on religion is trivial or insubstantial is not a breach of freedom of religion).
Tribe above n 39 at 1224. See also Lynch above n 6.
Above n 7 at 97–8.
See Prinsloo above n 10 at para 20.
See paragraph 90 of this judgment.
In this sense, the message
per se has far less religious tonality than does the nativity scene over which
the US Supreme Court divided
in Lynch above n
It would indeed be odd if this Court could not take judicial notice of matters massively testified to in criminal and family cases which regularly come before the courts. Hoffmann J in Stoke-on-Trent City Council v B & Q Plc; Norwich City Council v B & Q Plc (1991) ChD 48 at 65 (which concerned balancing out respect for store-closing on the “English Sunday” against the principle of unrestricted trade everywhere in the European Union) deals with judicial notice in the following terms:
“[I]f the court is satisfied on the basis of judicial notice that the requirements of proportionality have been met, there is no need for the prosecution to adduce oral or documentary evidence. Judicial notice is not confined to questions which everyone would be able to answer of his own knowledge. It includes matters of a public nature such as history, social customs and public opinion which may have to be culled from works of reference. As the late Professor Sir Rupert Cross said in Cross on Evidence 5th ed. (1979), p. 160, judicial notice is important for two reasons:
‘In the first place, it expedites the hearing of many cases. Much time would be wasted if every fact which was not admitted had to be the subject of evidence which would, in many cases, be costly and difficult to obtain. Secondly, the doctrine tends to produce uniformity of decision on matters of fact where a diversity of findings might sometimes be distinctly embarrassing’.”
Similar points are made by Hogg above n 63 at 858–9 where he says that it would be unfortunate if a law was struck down because of a deficiency in the evidence which could be supplied by a common sense finding. After dealing with the problems of costs involved in parading experts, he offers the opinion that:
“it would be desirable for Charter review to become less dependent on evidence, even if the courts have to strain somewhat to make ‘obvious’ or ‘self-evident’ findings.”
Rejecting the argument that the absence of an establishment clause in the Canadian Charter meant that a Canadian court should not adopt principles relating to religious freedom developed in the United States of America, Dickson CJC said that: “[t]he theological content of the legislation remains as a subtle and constant reminder to religious minorities within the country of their differences with, and alienation from, the dominant religious culture.” Big M above n 7 at 98 and 99–101.