The second formulation of the doctrine of Parliamentary Sovereignty is "Courts are bound to apply Acts of Parliament enacted in proper form, and cannot question their validity or constitutionality". This is a weaker formulation but it commands widespread support nowadays and it it usually treated of in the textbooks under the heading: "What is an Act of Parliament ?"

Basically, the argument turns upon a widely cited passage in Latham's The Law and the Commonwealth (1949), p 523: "When the purported sovereign is anyone but a single actual person, the designation of him must contain the statement of rules for the ascertainment of his will, and these rules since their observance is a condition of the validity of his legislation, are rules of law logically prior to him" A similar point is made by Salmond: "No statute can confer this power on Parliament for this would be to assume and act on the very power that is to be conferred".

Latham seeks to make the point clearer by comparing a single individual sovereign with a sovereign body. "It is not impossible", he writes, "to ascertain the will of an individual without the aid of rules" and he continues, "he may be presumed to mean what he says, and he cannot say more than one thing at a time". But even here, with all due respect to Latham, one might conclude that in the case of a single individual sovereign some "rule", however simple or tacit, must be operative, in order that mere domestic utterances be distinguished from legislative utterances. Otherwise every single utterance would be a legal enactment; rather like King Midas who turned whatever he touched into gold, so would Latham's King turn every utterance into law. King Henry II (1154 - ) asked "Who will free me from this turbulent priest" (Thomas a Becket) and his question was taken to be a sovereign command.

I any event, Latham's "presumptions" that such a single sovereign would mean what he says and could only say one thing at a time, look a bit like rudimentary "rules" for ascertaining the will of the sovereign. But we can let that point pass and continue with Latham's explanation, "... the extraction of a precise expression of will from a multiplicity of human beings is ... an artificial process and one which cannot be accomplished without arbitrary rules". Thus, he argues, at least some rudimentary manner and form rules must operate, however tacitly; otherwise we would have only "the simultaneous incoherent cry of a rabble, small or large, which cannot be law for it is unintelligible". [So what's new? And Must law be intelligible?]

At the time Latham wrote this was a revolutionary thought but it has long since passed into the orthodoxy of jurisprudential and constitutional thought. Thus, Professor Hart, in criticising Austin's thesis that there must be in every legal system a sovereign, one or number, concluded that in every developed legal system there must be a Rule of Recognition, that is a rule by reference to which we recognise what is - and what is not - law. Hans Kelsen has argued that in every legal system there must be a historically first material constitution, whether or not there is also a formal, written constitutions. Salmond, as we have already seen supposes some ultimate rule behind Parliament, granting sovereignty, because Parliament could not, logically, confer that power upon itself. Salmond's "ultimate rule", Hart's "rule of recognition", Kelsen's "historically first constitution" are all of a piece with Latham's logically prior rules.

If it is accepted that all legal systems logically must contain such a rule, then the question for our legal system is "What is the content of our rule ?" Heuston (p.8) comments on this as follows: "In the United Kingdom the rules on this matter are so simple and have been accepted without question for so long that it ios sometimes forgotten that they are rules of law". At common law they are simply two -

First, "Parliament is composed of three component parts, the Queen, the House of Lords, and the House of Commons, and each of those component parts must record its assent separately"

Secondly, "There must be some words of enactment, though probably they need not be in any particular form; otherwise there would be distinction between resolutions of either House and Acts of Parliament".

And so the question arises, if such rules were breached would what was produced be an Act of Parliament, at all ? Or could the courts refuse to apply what was, in reality, an improperly made rule ? Well the courts have done precisely that - Bowles v Bank of England [1913] 1 Ch 57 - A resolution of the Committee of the House of Commons for Ways and Means assenting to income tax at a certain rate for the ensuing financial year does not, either per se, or after adoption bu the House of Commons, authorise the Crown to levy on the subject the tax so assented to before the tax had actually been imposed by Act of Parliament. That is the ratio decidendi of the case; the actual decision was that the Bank of England could not, before the tax was imposed by statute, lawfully deduct any income tax from the dividends payable to a stockholder, without the stockholder's consent.

This case is now authority only for the legal proposition that is a purported Act of Parliament lacks the proper form, i.e. the three assents, then it is no Act of Parliament. It is not authority for the legal proposition that taxes cannot be imposed prior to the annual Finance Act, because by virtue of the Provisional Collection of Taxes Act, 1968, resolutions of the House of Commons have the effect of law for not more than four months pending the passage of the annual Finance Act.

Further, and venerable authority is Princes Case (1606) 8 Coke Rep 48: "An Act of Parliament penned by the assent of the King, and of the Lords Spiritual and Temporal, and of the Commons, is a good Act. But an Act penned by the King with the assent of the Lords, or that of the King with the assent of the Commons is no Act of Parliament".

Again, there is a cluster of cases, arising furth of England, which suggest that courts can have jurisdiction to question the validity of an Act of Parliament on the grounds that the relevant rules concerning manner and form have not been observed. In R v Military Governor, N D U Interment Camp [1924] 1 IR 32, a Mrs O'Brien had been detained without trial by the army for six months. She applied for a writ of habeas corpus. This was resisted successfully at first instance on the (factual) grounds that a state of war or armed rebellion existed, and so the jurisdiction of the ordinary courts was ousted. This was reversed on appeal, the Court of Appeal not being satisfied that a state of war or armed rebellion existed. The Court of Appeal was therefor minded to order Mrs O'Brien's immediate release. The Attorney-General, however, sought to make a return to the Writ (i.e. argue against it) and so the Court proposed to make an order the next day. When the Attorney General returned the following day with Mrs O'Brien, he presented the Public Safety (Emergency Powers) Act, 1923 the Government had secured the passage of which through both Houses and obtained the Governor General's assent thereto, all on the previous day. This new Act authorised the Minister of Defence to order the detention of any person without trial. The Attorney General also produced just such an order in respect of Mrs O'Brien.

The Court of Appeal was hardly sympathetic but on strict Parliamentary Sovereignty arguments, there was little it could do - i.e. "all that a court can do with an act of Parliament is apply it". Mrs O'Brien's resourceful Counsel, however, offered he following argument: The Act, and with it the order, was invalid. Article 47 of the Irish Constitution enabled a referendum to be set in motion within seven days of the passage of a Bill unless both Houses had resolved that the Act was immediately necessary for the preservation of public order or peace. The government had acted in such haste that these declarations had not been made. Therefore the assent of the Governor-General was improper. The Court of Appeal gratefully accepted this argument and Mrs O'Brien was released. Notice that it was the form and manner of the enactment and not the content which the court used to declare the Act invalid. This case is authority for the proposition that a court is "not bound to assume that a measure has been properly converted from the Bill stage to the Act stage".

The second case is Harris v Minister of the Interior 1952 (2) S A 429 wherein the Appellate Division of the South African Supreme Court considered and annulled the Separate Representation of Voters Act, 1951 on the grounds that it contravened sections 35 and 152 of the South Africa Act, 1909. These are referred to as "the entrenched sections" and they provide that certain sections, including section 152 and any Acts made thereunder (the Act of 1951 admittedly being such an Act) can only be repealled or altered by a two-thirds majority of both Houses, sitting together. "Parliament" in South Africa could thus function unicamerally or bicamerally and the Sovereignty of such a Parliament appears therefore to be divided (cf formula 6) between Parliament as ordinarily constituted in two Houses and as specially constituted in one. Note that the decision is not a denial of "sovereignty" of the South African Parliament, but it is authority for the proposition that for some purposes the component elements of that parliament may combine in one way and for some purposes in another, and that (should the appropriate manner and form rules not be observed) the purported legislation is invalid.

The third case in this series is Attorney-General for New South Wales v Trethowan [1932] AC 526. The legislature in New South Wales had in 1929 enacted that the Upper House should not be abolished except by a Bill which before being presented to the Governor for the Royal Assent should be approved by the electors in a referendum. This requirement for a referendum could not itself be repealed except by the same process. In 1931 a New South Wales government secured the passage of two Bills , one purporting to abolish the requirement of a referendum, the other purporting to abolish the upper House. It was admitted that these Bills were to be presented for the Royal Assent without any prior referendum. A declaration of invalidity and an injunction was sought and granted. This was upheld,, first, by the High Court of Australia, and, secondly, by the Privy Council but only on the grounds that section 5 of the Colonial Laws Validity Act, 1865 laid down express requirements as to manner and form.

Well these are interesting and instructive cases, but do they tell us anything about Parliamentary Sovereignty in the United Kingdom ? there are some formidable arguments to overcome, including the following seven:-

(1) All three case deal with provisions under a written constitution or similar instrument. There is nothing (the Act of Union apart) analogous in the United Kingdom.

(2) If these cases are thought of as concerned with delegated or subordinate legislation, and a plausible case certainly can be made out, then they are simply irrelevant to a consideration of Parliamentary Sovereignty, since even in the United Kingdom, subordinate legislation is subject to judical review [Delete ?e.g. Belfast Corporation v O D Cars [1960] AC 490 , an instance of striking down legislation as ultra vires a constitutional statute (namely, section 5 of the Government of Ireland Act, 1920) although the outcome was reversed by the House of Lords].

(3) The Irish case in particular would, in a United Kingdom context, have to face the formidable authority of Bradlaugh v Gossatt [1884] 12 QBD 271 which decided that the internal proceedings of Parliament are not to be impeached in the Courts, an authority all the more compelling in light of the traditional view of Parliament as a High Court whose proceedings for that reason, too, are unimpeachable in then lower courts.

(4) The Australian case raises the question whether the courts could or should issue an injunction to prevent an allegedly defective Bill being presented for the Royal Assent. It is thought unlikely that the court would of their own volition intervene in the internal working relations of the component elements of Parliament. [my note says new case ? 17/1/97 what was that ?]

(5) The view of the role of the courts indicated by these three cases seems to run counter to the so-called "enrolled Act" rule as expounded by Lord Campbell in Edinburgh and Dalkeith Ry Co v Wauchope 1842 - "All that a court of justice can do is to look at the Parliamentary roll. If it appears from that that a Bill has passed both houses and received the Royal Assent, no court of justice can inquire into the mode in which it was introduced into Parliament". On the other hand, Lord Campbell was dealing not with the question whether the three component elements of Parliament have properly combined, but with the presence of fraud in the pedigree of a private Act. Again, presumable, the corollary of Lord Campbell's dictum is that if it does not appear that a Bill has passed both Houses, etc then it is not an Act, and it is surely open to the Courts to hold invalid a purported Act that bears its own dearth-would on its face.

(6) It is widely thought that the consequences of such decisions as Bradlaugh, Edinburgh and Dalkeith Ry Co, Pickin, etc is that all and any of the ills in Parliamentary procedure are cured by the Royal Assent - that the Royal Assent rectifies any prior defective procedure. Pushed to its logical conclusion this would mean that manner and form does not operate as a condition precedent to the validity of an Act of Parliament. Here Bribery Commission v Ranasingh [1965] AC 172 (PC) is highly significant; at 172-3, "a legislature has no power to ignore the conditions of law making that are imposed upon it ... this restriction exists independently of the question whether the legislature is sovereign". On that view it follows that a Bill that has received the Royal Assent may be held ultra vires and invalid and where a Speaker's certificate is a necessary part of the legislative process, the absence of such certification may prove fatal to the validity of the legislation.

(7) The argument that the South African case is authority for a common law rule about manner and form as well as an interpretation of section 5 of the Colonial Laws validity Act, 1865 is hardly supported by the Privy Council decision itself which is almost wholly based upon section 5.

Well, where does all this leave the second formulation ?

(a) We know that in theory and in practice that there must be at least tacit manner and form rules in all legal systems;

(b) We can note that at least some of the arguments against manner and form review in the United Kingdom depend upon assuming Parliamentary Sovereignty and that, of course, is precisely what is in question. [Circularity/ question-begging]

(c) We can observe that in all the legal systems considered: UK, Ireland, South Africa, Australia, the exercise of sovereignty appears to be divided among several elements, e.g. the consent of the electorate by way of a referendum, is required validly to enact some legislation.

Courts in the United Kingdom are not likely to disqualify what purports to be an Act of Parliament but where the purported Act of Parliament "carries the death-wound in itself", as it was put in the Prince's Case, the courts might contrive to refuse to apply it. And, since parliamentary privilege is part of the common law - Ashby v White (1703) 14 St Tr 695 , it is settled that it cannot affect rights outside the House; i.e. internal procedure is one thing but the constitutional rights and liberties of the citizen is another. It follows that there may be circumstances in which the courts would enquire beyond the face of purported legislation.

Three examples will make this clearer:

1. Regency Acts 1937-53

2. Parliament Acts 1911 and 1949

3. Northern Ireland Constitution Act, 1973.

1. The Regency Acts, 1937-53 provide for a series of untoward events concerning the monarch, by providing for a regent to carry out all the royal functions including assenting to bills in the event of infancy, incapacity, or temporary absence abroad . The regent is however limited in that he may not assent to a Bill for changing the order of the succession to the Crown, or to a Bill repealing the provisions of the Act of Union, 1707 securing Presbyterian Church Government in Scotland. Arguably the courts would be able to hold a purported Act altering the order of the succession to the Crown invalid as to manner and for, even if the regent had assented to it - and similarly re the Act of Union 1707 ?

2. Better known are the Parliament Acts of 1911 and 1949. These provided that under specified circumstances a Bill may become an Act without the Lord's assent. But these Acts do not apply to an Act to extend the life of Parliament . So, could a court declare such an Act invalid on the basis of manner and form if it had been passed without the assent of the Lords ? Again, this does not sem wholly absurd. One problem here is that the jurisdiction of the ordinary courts might well be ousted by the Speaker's certification that the requirement of the Parliament Acts have been complied with. Section 3 of the 1911 Act declares such a certificate to be "conclusive for all purposes2. But in such circumstances, were it the case that the House of Lords had not assented to the Bill extending the life of parliament the courts might be prepared to hold that the legislation was beyond the powers of the delegated legislature (Commons and Queen) created by the Parliament Acts - one recollects the dictum in the Prince's Case about an Act penned only by the Monarch and the Commons. In addition, the courts might be willing to hold the "ouster clause" inapplicable in such circumstances and inept to exclude judicial review. Certainly the courts, including the House of Lords, have treated similar ouster clauses most unsympathetically and robustly - Anisminic v Foreign Compensation Commission [1969] 2 AC 147 is one justly celebrated example. [Yardley bet story half-bottle and on Anisminic see Denning, The Discipline of the Law, Part II Ch 2. I also have a reference to Denning, the Misuse of Power but it is not clear to me what that is about.] And, indeed, the question arises whether the House of Lords can be lawfully abolished without its consent; certainly there are arguments suggesting that this is not possible - see Mirfield. [Future, i.e. after 1996, I should write up the potted summary of his argument and my assessment.]

(3) The Northern Ireland Constitution Act, 1973, section 1 provides that Northern Ireland shall not cease ro be part of the United Kingdom without the consent of the majority of the people of Northern Ireland voting in a border-poll. Clearly the statute is of some perhaps considerable political significance, but what is its legally binding content ? If we take its content seriously, then we must conclude that for the purposes of ceding Northern Ireland, four consents rather than the more usual three consents are legally necessary. If that is so, then manner and form review of legislation purporting to cede Northern Ireland without a border poll appears open to the courts, though problems of locus standi (i.e. who has an interest in suing) and remedies (injunction before presentation for Royal Assent ?) would still remain. It is arguable on the basis of these examples that Parliament has redefined itself for specific purposes and that the manner and form rules - the logically prior conditions of valid legislation - differ according to the type of legislation proposed:-

Thus only one consent is necessary to pass the budget resolution into law for a period of four months (Provisional Collection of Taxes Act, 1968).

Two consents suffice under the provisions of the Parliament Acts, 1911 and 1949 insofar as they restrict the power of the Lords in order ultimately that the will of the Commons should prevail. In effect, with the exception of a Bill to extend the life of Parliament the Lords enjoy only a delaying power - a power which increases in political significance as the life of a Parliament draws to a close. Bills can be presented for the |Royal Assent after having been approved only by the Commons in two situations:-

(i) if the Lords fail within one month to pass a Bill which, having passed the Commons, is sent up at least one month before the end of the session and is endorsed by the Speaker as a "money bill";

(ii) if the Lords refuse in two consecutive sessions whether of the same Parliament or not, to pass any Bill (other than an "money bill" or a Bill to prolong the life of Parliament) which has been passed by the Commons in these two sessions, provided one year has elapsed, between the Bill's second reading in the first session and its third reading in the second session. Normally the Lords do not press their objection to legislation to the extent necessary to trigger the provisions of the Parliament Acts but that is the political reality, not the legal rule.

Three consents represents the usual or "normal" situation and is widely treated as normative and is the one favoured by traditionalists such as Austin, Dicey, and Wade.

Four consents might exceptionally be required as under the Statute of Westminster, 1931, s 4 or the Northern Ireland Constitution Act, 1973,s 1.

This variety of the requisite consents, all vouched for by Acts of Parliament, suggest that the "manner and form" rules of the United Kingdom constitution are somewhat more subtle than is generally allowed in the trite saw that what the Queen in Parliament assembled enacts is law and it allows of the possibility of courts refusing to regard as procedurally valid purported Acts contrary to such rules.

And this "limit" upon Parliament must be accepted-

(1) on the logical argument already developed;

(2) on pain of refusing to regard the plain words of Acts of Parliament as having any meaning.

However such "limitations" upon Parliament may not be very significant and few indeed are the situations under which courts might be called on to disqualify a purported Act by virtue of a "manner and form" test.

Further, proponents of classical Parliamentary Sovereignty have two counter -arguments:

(1) All "deviant" consent cases are merely instances of delegated legislation. Parliament, understood classically as the Monarch, Lords, and Commons remains the supreme legislative authority but it can and sometimes does authorise law making by other bodies such as the Commons alone, for the purposes of tax collection under the provisional Collection of Taxes Act, 1968 or, under the Parliament Acts by the consent of only the Queen and the Commons. Since judicial review is competent in the case of subordinate legislation any manner and form review in deviant consent cases is nothing out of the ordinary and is not even the minimal encroachment on Parliamentary Sovereignty that I have suggested it to be. However, it seems a strained construction, if not a desperate rear-guard action, to interpret deviant consent cases in this way, given the pre-eminence of the Commons over the Lords. It is even more strained to regard EU law as delegated legislation, although some staunch theorists have argued that case. This approach does, however salvage the classic doctrine from it "modern" critics, e.g Latham, Jennings, and myself.

(2) even if Parliament is bound by these rules for the time being, it is always open to Parliament to change any such rules, thereby reasserting its sovereignty.

I will consider some of these matters more fully in my seventh lecture but here I draw attention to argument developed by Hart, especially at pp 145-148 of his Concept of Law. The first point worthy of extraction from Hart's pages is that manner and form may be a greater constraint on Parliamentary Sovereignty even than I have argued. He puts a hypothetical example to show how the self-redefinition of Parliament might achieve "... very much the same results as those which the accepted doctrine, that Parliament cannot bind its successors, seems to put beyond its power" (147). A statute which, after fixing a minimum wage for engineers, went on to provide that no future Bill concerning engineers' pay should have effect as law unless confirmed by resolution of the Engineers Union , and that this provision itself could not be repealed without the consent of the Engineers' Union, might secure all that, in practice, could be done by a statute that purported to fix engineers' wages "for ever" and then crudely prohibited its repeal altogether. Hart, continues "Yet an argument which lawyers would recognise as having some force, can be made to show that although the latter would be ineffective under the present rule of continuing parliamentary sovereignty, the former would not" (147). If Hart is correct about this, then "manner and form" arrangements, used imaginatively, threaten to undermine the classical conception of Parliamentary Sovereignty.

The second important point to be extracted from Hart's pages is conceptual: he draws a distinction between two versions of the doctrine which he calls "continuing parliamentary sovereignty" and "self-embracing parliamentary sovereignty". The first conception requires that parliament is eternally "free at every moment of its existence as a continuing body, no only from legal limitations imposed ab extra but also from its own prior legislation" (145). Hart himself believes that this conception of parliamentary sovereignty is "now established" (145) and that this is "clear" (146) and that no parliament can prohibit its successors from repealing its legislation. The conception is part of the Rule of Recognition. Hart also believes that there is no logical reason why this should be so and that a different doctrine might well have been established. That doctrine might even better deserve the name "sovereignty" and it is the doctrine "that Parliament should not be incapable of limiting irrevocably the legislative competence of its successors but, on the contrary it should have this self-limiting power" (145). Hart believes that the choice between these two doctrines - and a choice there must be because the two doctrines are inconsistent - is an "empirical" question and, as noted, he also believes that the "continuing" doctrine prevails. This I question. It seems to me to be a matter of what counts as evidence if, indeed, it is an empirical question at all. So long as one looks at the doctrines espoused by the great gurus of 20th century English constitutional theory, especially Dicey and Wade, especially if one projects back to their intellectual grounding in the fairly crude Austinian jurisprudence (Bentham's contribution being recognised only later in the history of jurisprudential and constitutional ideas) which was accepted and espoused by many lesser figures, then it certainly appears that the continuing conception has most support. But if one examines the actual practice of parliaments, and the decisions of courts, there is a great deal more to be said for the self-embracing conception than Hart, himself, appeared to acknowledge. I will address these issues more fully when I consider the fifth formulation but here I simply want to identify these two competing views and indicate that the self-embracing conception elevates the manner and form critique, as adumbrated in Hart's hypothetical, to the status of a clear refutation of the classical conception. Certainly the self-embracing conception fits better with ideas such as "Freedom once given cannot be revoked". Just as it is half a century or more too late to treat the effect of the Parliament Acts as merely establishing a delegated legislature, so it is to late to deny the widely shared perception that even before, let alone after the Canada Act, 1982, the United Kingdom Parliament simply could not reclaim exclusive and supreme legislative authority over the territory and people of Canada. (see Dickinson)

The trouble is, of course that one conception - self-embracing - denies too much and the other - continuing - asserts too much. Nor is it possible to sythesise these contradictory conceptions into a coherent doctrine. Marshall sees very clearly that "both ... [doctrines] assume an odd character" (p45), if we conceive of sovereignty as legal omnipotence. "...continuing omnipotence can only be guaranteed by making it subject to a paradoxical restriction, limiting its ability to affect its own competence ... self-embracing competence may embrace itself in such a way as to bring its omnipotence to an end" (p45). It seems from this that neither doctrine is internally coherent and it also seems that we cannot logically adopt both position simultaneously even though it is this ambiguity as the core of the doctrine which contributes to its longevity and popularity - it can be made to mean very different things as the circumstances require.

Since both Hart and Marshall, turn to theological doctrines of God's omnipotence to illustrate the nature of continuing (read non-continuous) and of self-embracing (read self-defeating) sovereignty, I need not be more papal than either, but turn myself to an illusrtrative question which I hope throws light on the debate and exposes the central illogicalities of both the classical and the modern doctrines. My question about God's purported omnipotence is this: "Can God create a stone so heavy that he, himself cannot lift it ?" If "Yes", then God is not Omnipotent. If "No" then God is not omnipotent. Therefore God is not omnipotent. Moving swiftly from theology to constitutional theory, the parallel question is "Can Parliament make law that Parliament cannot change ?". If "Yes" then Parliament is not sovereign. If "No" then Parliament is not sovereign. Therefore Parliament is not sovereign. Q E D.

[The Marshall stuff is less easy to "thin" but what is interesting there is an argument that both conceptions are flawed].