High Court of Australia Mason C.J., Brennan, Deane, Dawson, Toohey, Gaudron and McHugh JJ. Polyukhovich v The Commonwealth [1991] HCA 32 ORDER Answer the question reserved as follows:Is Section 9 of the War Crimes Act 1945 Cth, as amended, invalid in its application to the information laid by the second defendant against the plaintiff? Answer: No. No order as to costs. Cur. adv. vult. The following written judgments were delivered:— 1991, Aug. 14 Mason C.J. The plaintiff is an Australian citizen and a resident of South Australia. He brought an action in this Court seeking a declaration that the War Crimes Amendment Act 1988 Cth is invalid and, or in the alternative, a declaration that ss. 6(1), 6(3), 7, 9 and 11 of the War Crimes Act 1945 Cth as amended ("the Act") are invalid. The plaintiff's interest in seeking declaratory relief of this kind arises from the circumstance that the second defendant laid an information against the plaintiff alleging that, between 1 September 1942 and 31 May 1943, the plaintiff committed war crimes in the Ukraine. Each of the crimes is alleged to have been a "war crime" within the meaning of s. 9 of the Act, being a "serious crime" within the meaning of s. 6 of the Act. In each instance the crime was alleged to have been committed at a time when the Ukraine was under German occupation during the Second World War. It is common ground that at the time of the commission of the alleged offences there was no Australian legislation in force which purported to make it a criminal offence on the part of an Australian citizen or resident to do such acts in the Ukraine as the plaintiff is alleged to have done. In the course of the proceedings, at the request of the parties, I reserved for the consideration of the Full Court the question: "Is section 9 of the War Crimes Act 1945 as amended, invalid in its application to the information laid by the second defendant against the plaintiff?" The plaintiff submits that the question should be answered in the affirmative on two grounds. The first ground is that the section is beyond the legislative powers conferred upon the Parliament by s. 51(vi) and (xxix) of the Constitution with respect to defence and external affairs, these being the only two powers which, according to the defendants' case, could sustain the validity of the law. The second is that the section, because it attempts to enact that past conduct shall constitute a criminal offence, is an invalid attempt to usurp the judicial power of the Commonwealth, that power being vested by the Constitution in Ch. III courts. The War Crimes Act 1945 According to its long title, the Act is "[a]n Act to provide for the Trial and Punishment of War Criminals". The original preamble to the Act recited: Whereas it is expedient to make provision for the trial and punishment of violations of the laws and usages of war committed during any war in which His Majesty has been engaged since the second day of September, One thousand nine hundred and thirty-nine, against any persons who were at any time resident in Australia or against certain other persons. By s. 5(1) of the Act, as originally enacted, the Governor-General was authorized to "convene military courts for the trial of persons charged with the commission of war crimes". Section 7 conferred power on a military court so convened to "try persons charged with war crimes committed, at any place whatsoever, whether within or beyond Australia" and for that purpose "to sit at any place whatsoever, whether within or beyond Australia". "War crime" was defined by s. 3 to mean: "(a) a violation of the laws and usages of war; or (b) any war crime within the meaning of the instrument of appointment" of a certain Board of Inquiry appointed under the National Security (Inquiries) Regulations Cth committed in any place whatsoever, whether within or beyond Australia, during any war. By the War Crimes Amendment Act 1988 (No. 3 of 1989) the original Act was almost entirely repealed and replaced. The amended preamble recites: Whereas: (a) concern has arisen that a significant number of persons who committed serious war crimes in Europe during World War II may since have entered Australia and became Australian citizens or residents; (b) it is appropriate that persons accused of such war crimes be brought to trial in the ordinary criminal courts in Australia; and (c) it is also essential in the interests of justice that persons so accused be given a fair trial with all the safeguards for accused persons in trials in those courts, having particular regard to matters such as the gravity of the allegations and the lapse of time since the alleged crimes. Section 9 of the Act provides: (1) A person who: (a) on or after 1 September 1939 and on or before 8 May 1945; and (b) whether as an individual or as a member of an organisation; committed a war crime is guilty of an indictable offence against this Act. (2) Sections 5 and 7, and paragraph 86(1)(a), of the Crimes Act 1914 do not apply in relation to an offence against this Act. In order to ascertain whether an offence is a "war crime" within the meaning of the Act it is necessary to look to ss. 6, 7 and 8. Section 7 defines a "war crime" by reference to a "serious crime". What constitutes a "serious crime" is to be ascertained from s. 6. Section 6(1) provides: An act is a serious crime if it was done in a part of Australia and was, under the law then in force in that part one of a number of offences mentioned in the sub-section. One such offence is: "(a) murder"; another is: (k) an offence of: (i) attempting or conspiring to commit; (ii) aiding, abetting, counselling or procuring the commission of; or (iii) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the commission of; an offence referred to in [paragraph (a)]. Section 6(2), (3) and (6) provide: (2) In determining for the purposes of subsection (1) whether or not an act was, under the law in force at a particular time in a part of Australia, an offence of a particular kind, regard shall be had to any defence under that law that could have been established in a proceeding for the offence. (3) An act is a serious crime if: (a) it was done at a particular time outside Australia; and (b) the law in force at that time in some part of Australia was such that the act would, had it been done at that time in that part, be a serious crime by virtue of subsection (1). (6) For the purposes of [subsection (3)], the fact that the doing of an act was required or permitted by the law in force when and where the act was done shall be disregarded. Section 7 provides: (1) A serious crime is a war crime if it was committed: (a) in the course of hostilities in a war; (b) in the course of an occupation; (c) in pursuing a policy associated with the conduct of a war or with an occupation; or (d) on behalf of, or in the interests of, a power conducting a war or engaged in an occupation. (2) For the purposes of subsection (1), a serious crime was not committed: (a) in the course of hostilities in a war; or (b) in the course of an occupation; merely because the serious crime had with the hostilities or occupation a connection (whether in time, in time and place, or otherwise) that was only incidental or remote. (3) A serious crime is a war crime if it was: (a) committed: (i) in the course of political, racial or religious persecution; or (ii) with intent to destroy in whole or in part a national, ethnic, racial or religious group, as such; and (b) committed in the territory of a country when the country was involved in a war or when territory of the country was subject to an occupation. (4) Two or more serious crimes together constitute a war crime if: (a) they are of the same or a similar character; (b) they form, or are part of, a single transaction or event; and (c) each of them is also a war crime by virtue of either or both of subsections (1) and (3). Section 7 must be read in conjunction with the definition of war in s. 5. That section defines "war" to mean a war "in so far as it occurred in Europe in the period beginning on 1 September 1939 and ending on 8 May 1945". The section defines "occupation", inter alia, as meaning "(a) an occupation of territory arising out of a war". The effect of these definitions is to confine war crimes to conduct which took place outside Australia (see s. 7(1)(a) and (b)) except in so far as "serious crimes" under s. 7(1)(c) and (d) and s. 7(3) might conceivably be committed in Australia. However, the terms of the preamble and the provisions of s. 7(1)(a) and (b) make it clear that the primary and substantial concern of the Act is with war crimes committed outside Australia, in other words, with conduct on the part of persons outside Australia. Further, the primary and substantial concern of the Act is with war crimes committed in Europe during the Second World War. So much appears from the preamble and the definitions of "war" and "occupation". Only an Australian citizen or resident shall be charged with an offence under the Act (s. 11), but that means that the person charged must be an Australian citizen or resident only at the time that he or she is charged. It follows that the Act makes criminal acts done by a person who, at the time of the commission of those acts, had no relevant connexion with Australia. The information laid against the plaintiff rests on an application of the provisions of ss. 6(1), 6(3), 7, 9 and 11 of the Act. That is why he seeks a declaration that s. 9 is invalid, for the application of that section depends not only on its own validity but also on the validity of related provisions. Superior orders are not a defence in a proceeding for an offence under the Act: s. 16. However, s. 17(2) provides that, subject to s. 16, it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. Section 17(3) to (5) go on to provide: (3) To avoid doubt, the doing of the act by the defendant was permitted by the laws, customs and usages of war if it was reasonably justified by the exigencies and necessities of the conduct of war. (4) The defendant is not entitled to rely on a defence under subsection (2) unless there is evidence of the existence of the facts constituting the defence. (5) However, if there is such evidence, the onus of establishing, beyond a reasonable doubt, that those facts either do not exist or do not constitute the defence lies on the prosecution. The provisions of s. 68 of the Judiciary Act 1903 Cth apply in relation to an offence against the Act as if a reference in that section to a Territory did not include a reference to an external Territory: s. 13(1). By virtue of s. 68(2) and (5C) of the Judiciary Act, in relation to offences committed elsewhere than in a State or Territory, jurisdiction is conferred on the courts of a State or Territory notwithstanding any limits as to locality of the jurisdiction of that court under the law of that State or Territory. However, by s. 14(4) of the Act, on an application by a defendant, the magistrate or judge is bound to order that all proceedings for the offence charged be held in a State or Territory other than that in which it is being held, unless he or she is satisfied on the balance of probabilities that the defendant, when charged with the offence: (a) was a resident of the State or Territory referred to in subsection (1); or (b) was not a resident of that other State or Territory. The powers of a court to take action to prevent an abuse of process are expressly preserved: s. 13(4)(b). Moreover, s. 13(5) provides: Where, on the trial of a person for an offence against this Act, the person satisfies the judge, on the balance of probabilities, that: (a) the person is unable to obtain evidence that he or she would, but for the lapse of time or some other reason beyond his or her control, have been able to obtain; (b) the person's inability to obtain that evidence has substantially prejudiced, or will substantially prejudice, the preparation or conduct of his or her defence; and (c) the interests of justice require the making of an order under this subsection; the judge may make such order as he or she thinks appropriate for a stay of proceedings for the offence. The external affairs power (s. 51(xxix)) Discussion of the scope of the external affairs power has naturally concentrated upon its operation in the context of Australia's relationships with other countries and the implementation of Australia's treaty obligations. However, it is clear that the scope of the power is not confined to these matters and that it extends to matters external to Australia. I have previously expressed the view that the grant of legislative power with respect to external affairs should be construed with all the generality that the words admit and that, so construed, the power extends to matters and things, as well as relationships, outside Australia: New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [38] ; Robinson v. Western Australian Museum [39] ; Koowarta v. Bjelke-Petersen [40] . That view has prevailed in this Court since 1975: see the Seas and Submerged Lands Case, per Barwick C.J. [41] , per Jacobs J. [42] , per Murphy J. [43] ; Robinson , per Barwick C.J. [44] ; Viro v. The Queen [45] , per Murphy J.; Koowarta , per Stephen J. [46] ; The Commonwealth v. Tasmania (the Tasmanian Dam Case) [47] , per Gibbs C.J., per Murphy J. [48] . In Koowarta [40] , I said that the power extended to persons, as well as matters and things, outside Australia; in the Seas and Submerged Lands Case [49] , Jacobs J. had expressed the same view. Although not all of the judicial observations in the cases to which I have just referred mention explicitly the application of the power to persons, as distinct from matters, things or circumstances, external to Australia, there is no reason for drawing any distinction between the reach of the power in its application to persons on the one hand and matters, things and circumstances on the other hand. As Murphy J. stated in the Tasmanian Dam Case [50] , the power extends to conduct outside Australia. 1. (1975) 135 C.L.R. 337, at pp. 470-471. 2. (1977) 138 C.L.R. 283, at p. 335. 3. (1982) 153 C.L.R. 168, at p. 223. 4. (1975) 135 C.L.R., at p. 360. 5. (1975) 135 C.L.R., at p. 497. 6. (1975) 135 C.L.R., at pp. 503-504. 7. (1977) 138 C.L.R., at p. 294. 8. (1978) 141 C.L.R. 88, at p. 162. 9. (1982) 153 C.L.R., at p. 211. 10. (1983) 158 C.L.R. 1, at p. 97. 11. (1983) 158 C.L.R., at p. 170. 12. (1982) 153 C.L.R. 168, at p. 223. 13. (1975) 135 C.L.R., at p. 497. 14. (1983) 158 C.L.R., at p. 170. The existence of a connexion between the enacting State and the extraterritorial persons, things and events on which a State law operates has been held to be essential to the valid extraterritorial operation of that State law: Pearce v. Florenca [51] , per Gibbs J.; Union Steamship Co. of Australia Pty. Ltd. v. King [52] ; Port MacDonnell Professional Fishermen's Association Inc. v. South Australia [53] ; cf. Australia Act 1986 Cth, s. 2(1); Australia Act 1986 U.K., s. 2(1). The requirement for a relevant connexion between the circumstances on which the legislation operates and the State has been liberally applied and even a remote and general connexion suffices: Union Steamship [52] . According to traditional doctrine, the requirement for such a connexion or nexus stems from the circumstance that the legislative power of a State legislature is expressed to be for the peace, order and good government of the State. The opening words of the grant of legislative powers to the Parliament in s. 51 of the Constitution might have provided a foundation for a similar interpretation of the external affairs power, making the existence of a relevant connexion or nexus an essential prerequisite of the valid exercise of the power. However, I share the view expressed by Windeyer J. in Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [54] , where he said: So far as the Commonwealth is concerned, it is now for Parliament alone to judge whether a measure in respect of any topic on which it has power to legislate is in fact for the peace order and good government of the Commonwealth. This comment applies with particular force to an exercise of the external affairs power. 1. (1976) 135 C.L.R. 507, at pp. 518-519. 2. (1988) 166 C.L.R. 1, at p. 14. 3. (1989) 168 C.L.R. 340, at pp. 372-373. 4. (1988) 166 C.L.R. 1, at p. 14. 5. (1959) 103 C.L.R. 256, at p. 308. In the Seas and Submerged Lands Case [55] , Jacobs J. was of the same opinion, observing: There is no gap in the constitutional framework. Every power right and authority of the British Crown is vested in and exercisable by the Crown in Australia subject only to the Constitution. The State legislatures do not have that sovereignty which the British legislature and now the Australian legislature possess. A State can only legislate in respect of persons acts matters and things which have a relevant territorial connexion with the State, a connexion not too remote to entitle the law to the description of a law for the peace welfare and good government of the State. The words of s. 51 of the Constitution do not import any similar territorial limitation and there now is none in the case of the Australian legislature. The words "external affairs" can now be given an operation unaffected by any concept of territorial limitation. The result is that the Commonwealth, outside the boundaries of the States and subject to any particular constitutional injunctions, may make laws on all subject matters in exercise of its sovereignty. It follows that the legislative power of the Parliament with respect to matters external to Australia, using "matters" in a comprehensive sense, is not less in scope than the power of the Parliament of the United Kingdom with respect to such matters. 1. (1975) 135 C.L.R., at p. 498. The very recent cases on the external affairs power — Koowarta , the Tasmanian Dam Case, Richardson v. Forestry Commission [56] and Queensland v. The Commonwealth [57] — concerned the impact on the States of Commonwealth legislation implementing Australia's obligations under international conventions designed to protect human rights and the domestic environment. These cases vindicated the regulation by the Parliament of conduct within Australia pursuant to the external affairs power when that regulation is undertaken by way of implementation of an international convention. The cases serve to illustrate the proposition that legislation enacted pursuant to the power, though necessarily concerned with some aspect of externality, will have a domestic or internal operation. 1. (1988) 164 C.L.R. 261. 2. (1989) 167 C.L.R. 232. That, of course, is the situation here. The legislation makes conduct outside Australia unlawful, thereby visiting that conduct with legal consequences under Australian law. The conduct made unlawful constitutes a criminal offence triable and punishable in the ordinary criminal courts in this country. But, to the extent that s. 9 operates upon conduct which took place outside Australia and makes that conduct a criminal offence, the section is properly characterized as a law with respect to external affairs and is a valid exercise of power, subject to a consideration of the argument based on usurpation of judicial power. In conformity with what I have already said, I arrive at this conclusion on the footing that it is not necessary that the Court should be satisfied that Australia has an interest or concern in the subject-matter of the legislation in order that its validity be sustained. It is enough that Parliament's judgment is that Australia has an interest or concern. It is inconceivable that the Court could overrule Parliament's decision on that question. That Australia has such an interest or concern in the subject-matter of the legislation here, stemming from Australia's participation in the Second World War, goes virtually without saying. If, to the extent that the law operates upon conduct which took place within Australia, s. 9 cannot be supported as a valid exercise of legislative power, it is plainly severable so that, on any view, the section operates validly on conduct which took place outside Australia. In saying this, I do not mean to imply that s. 9 is invalid to the extent to which it may operate on conduct within Australia. That question does not arise for decision and there is no point in discussing it further. I should, however, indicate that I do not regard the circumstance that the law operates on the past conduct of persons who, at the time of the commission of that conduct, had no connexion with Australia as detracting in any way from the character of s. 9 as a law with respect to external affairs. The externality of the conduct which the law prescribes as the foundation of the criminal offence is enough without more to constitute it as a law with respect to external affairs. In this respect it makes no difference whether the law creates a criminal liability by reference to past or future conduct, so long as the conduct is external to Australia. My conclusion that the external affairs power supports the validity of s. 9 on the basis discussed above makes it unnecessary for me to examine the argument that the section may be supported on the ground that it is a law with respect to external affairs because it gives effect to an obligation arising under international law or because it implements a resolution of an international body. It is also unnecessary to deal with the alternative submission that the law is a valid exercise of the power because it facilitates the exercise of the universal jurisdiction under international law. Likewise, there is no need to consider the defence power. Usurpation of judicial power The plaintiff's argument on this aspect of the case is that the Act usurps the exercise of the judicial power of the Commonwealth in so far as it declares retrospectively certain past conduct to be a criminal offence and falls into the category of ex post facto laws. The submission is that one of the essential elements in the exercise of judicial power is the determination by a court of the issue whether the accused person infringed a rule of conduct prescribed in advance. According to the plaintiff's argument, the Act has determined that all persons who engaged in conduct of the kind declared unlawful are guilty of an offence, the only issue left for determination being whether the plaintiff is one of those persons. The judicial power of the Commonwealth is vested by s. 71 of the Constitution in Ch. III courts comprising this Court, federal courts created by Parliament and State courts exercising federal jurisdiction. Judicial power is an elusive concept; "it has never been found possible to frame a definition that is at once exclusive and exhaustive", to repeat the comment of Dixon C.J. and McTiernan J. in Reg. v. Davison [58] . According to the widely-accepted statement of Griffith C.J. in Huddart, Parker & Co. Pty. Ltd. v. Moorehead [59] , judicial power in s. 71 means "the power which every sovereign authority must of necessity have to decide controversies between its subjects, or between itself and its subjects, whether the rights relate to life, liberty or property". 1. (1954) 90 C.L.R. 353, at p. 366. 2. (1909) 8 C.L.R. 330, at p. 357. A more comprehensive statement of the content of judicial power is contained in the judgment of Kitto J. in Reg. v. Trade Practices Tribunal; Ex parte Tasmanian Breweries Pty. Ltd. [60] . His Honour observed at: Thus a judicial power involves, as a general rule, a decision settling for the future, as between defined persons or classes of persons, a question as to the existence of a right or obligation, so that an exercise of the power creates a new charter by reference to which that question is in future to be decided as between those persons or classes of persons. In other words, the process to be followed must generally be an inquiry concerning the law as it is and the facts as they are, followed by an application of the law as determined to the facts as determined; and the end to be reached must be an act which, so long as it stands, entitles and obliges the persons between whom it intervenes, to observance of the rights and obligations that the application of law to facts has shown to exist. It is right, I think, to conclude from the cases on the subject that a power which does not involve such a process and lead to such an end needs to possess some special compelling feature if its inclusion in the category of judicial power is to be justified. This statement contemplates as one element in the exercise of judicial power the application to the facts of a pre-existing or antecedent legal principle or standard, though it does not require that the rule or standard should have been ascertained or precisely defined before the determination is made in the exercise of judicial power. The need for an inquiry into what the law is presupposes that there may be uncertainty as to the nature, scope or content of the principle or standard to be applied. Indeed, it is widely recognized that courts, in exercising their judicial power, make and alter law in the sense of formulating new or altered legal principles. 1. (1970) 123 C.L.R. 361, at pp. 374-375. There is nothing in the statements which I have quoted to suggest that an exercise of judicial power necessarily involves the application to the facts of a legal principle or standard formulated in advance of the events to which it is sought to be applied. Indeed, there is powerful authority in this Court which supports the proposition that the application to the facts of a retrospective law which operates on past conduct so as to create rights and liabilities is an instance of the exercise of judicial power. In Nelungaloo Pty. Ltd. v. The Commonwealth [61] , the validity of the Wheat Industry Stabilization Act (No. 2) 1946 Cth was upheld, even though it validated an order for the acquisition of wheat, the validity of which was in issue in proceedings pending when the statute was enacted. The statute affected rights in issue in the litigation. At first instance, Williams J., in holding that the statute was a valid exercise of the defence power, remarked [62] : It was contended that this section infringes the judicial power because it does not amend the law prospectively but attempts to prescribe the construction to be placed upon an existing law by the court and the determination of the meaning of a statute is of the essence of the judicial power. The result of this contention, if sound, would be that the Commonwealth Parliament has no power to pass a declaratory statute which only has a retrospective operation. I cannot agree with this contention. On appeal, his Honour's decision upon the point was upheld [63] . Dixon J., with reference to the validating provision, said [64] : It is simply a retrospective validation of an administrative act and should be treated in the same way as if it said that the rights and duties of the growers and of the Commonwealth should be the same as they would be, if the order was valid. If such an enactment is a law with respect to the subject of defence, I can see no objection to its validity. Subsequently, in Reg. v. Humby; Ex parte Rooney [65] , I applied Nelungaloo and, in so doing, observed: Chapter III contains no prohibition, express or implied, that rights in issue in legal proceedings shall not be the subject of legislative declaration or action. 1. (1948) 75 C.L.R. 495. 2. (1948) 75 C.L.R., at p. 503. 3. (1948) 75 C.L.R., at pp. 579-580, 584. 4. (1948) 75 C.L.R., at p. 579. 5. (1973) 129 C.L.R. 231, at p. 250. More recently, in Australian Building Construction Employees' and Builders Labourers' Federation v. The Commonwealth [66] , this Court rejected the contention that the Builders Labourers' Federation (Cancellation of Registration) Act 1986 Cth was invalid because it was an exercise of judicial power or because it involved an interference with judicial power. The contention was that the impugned statute, which cancelled the registration of the plaintiff union as a registered organization, abrogated the function of this Court in pending proceedings concerning the cancellation of that registration. The Court drew a distinction between legislation affecting substantive rights in issue in litigation and legislative interference with the judicial process itself [67] . Liyanage v. The Queen [68] , where the statutes were directed to the trial of particular persons charged with particular offences on a particular occasion, was a case which fell into the second of the two categories. 1. (1986) 161 C.L.R. 88. 2. (1986) 161 C.L.R., at p. 96. 3. [1967] 1 A.C. 259. It is contended that the power of the Parliament to enact a retrospective or retroactive law dealing with substantive rights or liabilities does not extend to a law which makes past conduct a criminal offence. Such a law, it is said, stands in a very different position. It is suggested that support is to be found in Blackstone's Commentaries and in the decisions of the Supreme Court of the United States on Art. I, s. 9, cl. 3 and Art. I, s. 10, cl. 1 of the United States Constitution for the proposition that such a retrospective criminal law is beyond the power of the legislature on the ground that it is an interference with judicial power. The answer to this submission, as will appear, is that Blackstone does not assert that such a law is beyond the power of Parliament and that, to the extent to which the proposition is sustained by judicial decision, it rests upon the existence of a specific prohibition in the United States Constitution which has no counterpart in our Constitution. Blackstone in his Commentaries, 17th ed. (1830), vol. I, pp. 45-46, equated law to a "rule prescribed" and stated that, in the case of criminal conduct, such a rule must be prescribed as to future conduct. With reference to a law making past conduct a crime and inflicting punishment on the person who committed it, Blackstone said (at p. 46): Here it is impossible that the party could foresee that an action, innocent when it was done, should be afterwards converted to guilt by a subsequent law: he had therefore no cause to abstain from it; and all punishment for not abstaining must of consequence be cruel and unjust. All laws should be therefore made to commence in futuro, and be notified before their commencement; which is implied in the term "prescribed". But nowhere does Blackstone assert that it is beyond the power of Parliament to enact such a law, and still less that such a law would constitute an interference with the exercise of judicial power. He held strongly to the view that Parliament had power to enact that which was unreasonable and was vehemently opposed to the pretension that the courts had power to reject a statute on the ground that it was unreasonable "for that were to set the judicial power above that of the legislature, which would be subversive of all government": Commentaries (1830), vol. I, p. 91. Article I, s. 9, cl. 3 and Art. I, s. 10, cl. 1 of the United States Constitution prohibit any State as well as Congress from passing a bill of attainder or an ex post facto law. A bill of attainder is a legislative enactment which inflicts punishment without a judicial trial; initially a bill of attainder provided for punishment by death but in the context of the constitutional prohibition such a bill is now regarded as including what was formerly a bill of pains and penalties: Cummings v. Missouri [69] . An ex post facto law, of which a bill of attainder was, or might be, an instance, is a retrospective law which makes past conduct a criminal offence. An ex post facto law includes: 1st. Every law that makes an action done before the passing of the law, and which was innocent when done, criminal; and punishes such action. 2d. Every law that aggravates a crime, or makes it greater than it was, when committed: Calder v. Bull [70] , per Chase J. 1. (1798) 3 U.S. 386, at p. 390. The distinctive characteristic of a bill of attainder, marking it out from other ex post facto laws, is that it is a legislative enactment adjudging a specific person or specific persons guilty of an offence constituted by past conduct and imposing punishment in respect of that offence. Other ex post facto laws speak generally, leaving it to the courts to try and punish specific individuals. 1. (1867) 71 U.S. 277. 2. (1798) 3 U.S. 386, at p. 390. The constitutional prohibition against bills of attainder and ex post facto laws was not an expression of the antecedent common law of England. So much was acknowledged by Chase J. in Calder v. Bull [71] when he said: The prohibition against [State legislatures] making any ex post facto laws was introduced for greater caution, and very probably arose from the knowledge, that the parliament of Great Britain claimed and exercised a power to pass such laws, under the denomination of bills of attainder, or bills of pains and penalties; the first inflicting capital, and the other less, punishment. These acts were legislative judgments; and an exercise of judicial power. 1. (1798) 3 U.S., at p. 389. The absence of any similar prohibition in our Constitution against bills of attainder and ex post facto laws is fatal to the plaintiff's argument except in so far as the separation of powers effected by our Constitution, in particular the vesting of judicial power in Ch. III courts, imports a restraint on Parliament's power to enact such laws. In this respect the prohibition against bills of attainder has been seen "as an implementation of the separation of powers, a general safeguard against legislative exercise of the judicial function, or more simply — trial by legislature": United States v. Brown [72] . This doctrine applies to bills of attainder but not to the generality of other ex post facto laws. That is because it is of the essence of the prohibition of a bill of attainder "that it proscribes legislative punishment of specified persons — not of whichever persons might be judicially determined to fit within properly general proscriptions duly enacted in advance": Tribe, American Constitutional Law, 2nd ed. (1988), p. 643. The application of the doctrine depends upon the legislature adjudging the guilt of a specific individual or specific individuals or imposing punishment upon them. If, for some reason, an ex post facto law did not amount to a bill of attainder, yet adjudged persons guilty of a crime or imposed punishment upon them, it could amount to trial by legislature and a usurpation of judicial power. But if the law, though retrospective in operation, leaves it to the courts to determine whether the person charged has engaged in the conduct complained of and whether that conduct is an infringement of the rule prescribed, there is no interference with the exercise of judicial power. 1. (1965) 381 U.S. 437, at p. 442. That the application of the doctrine of separation of powers depends upon the notion of "trial by legislature" is demonstrated by United States v. Brown. The statute made it an offence for anyone who was, or had been within five years, a member of the Communist Party to serve as an officer or manager of a labour union. The Court held that the statute was invalid as a bill of attainder, observing [73] : The statute does not set forth a generally applicable rule decreeing that any person who commits certain acts or possesses certain characteristics (acts and characteristics which, in Congress' view, make them likely to initiate political strikes) shall not hold union office, and leave to courts and juries the job of deciding what persons have committed the specified acts or possess the specified characteristics. Instead, it designates the persons who possess the feared characteristics and therefore cannot hold union office without incurring criminal liability — members of the Communist Party. Later, the Court stated [74] that the command of the bill of attainder prohibition is "that a legislature can provide that persons possessing certain characteristics must abstain from certain activities, but must leave to other tribunals the task of deciding who possesses those characteristics " 1. (1965) 381 U.S., at p. 450. 2. (1965) 381 U.S., at p. 454. The view that a statute which contains no declaration of guilt and does not impose punishment for guilt is not a usurpation of judicial power is supported by the reasoning of the Privy Council in its decision in Kariapper v. Wijesinha [75] . The Privy Council upheld the validity of a statute enacted by the Parliament of Ceylon which imposed civil disabilities on persons to whom the statute applied, namely, persons, including the appellant, named in a schedule to the statute who were found guilty of bribery in a report by a commission of inquiry. The statute also provided for the vacation of the appellant's seat as a Member of Parliament. It was common ground that the Constitution of Ceylon insisted upon a separation of powers, at least to the extent that judicial power was vested exclusively in the courts to the exclusion of the legislature [76] . The Privy Council had so held in its earlier decision in Liyanage v. The Queen [77] . In Kariapper v. Wijesinha [78] , the appellant argued that the statute was an exercise of judicial power because it imposed punishment for guilt without trial by a competent court and was a bill of attainder, ex post facto legislation having an element of punishment being on the same footing as a bill of attainder. The Privy Council rejected the argument on the grounds that the statute contained no declaration of guilt and the disabilities which it imposed did not have the character of punishment for guilt but were to keep public life clean for the public good. 1. [1968] A.C. 717. 2. [1968] A.C., at p. 732. 3. [1967] 1 A.C., at pp. 287-289. 4. [1968] A.C., at p. 721. Sir Douglas Menzies, speaking for the Judicial Committee, referred with evident approval to the concurring opinion of Frankfurter J. in United States v. Lovett [79] . In that case, his Honour said [80] : All bills of attainder specify the offense for which the attainted person was deemed guilty and for which the punishment was imposed. Frankfurter J. went on to say [81] : No offense is specified and no declaration of guilt is made Not only does §304 lack the essential declaration of guilt. It likewise lacks the imposition of punishment in the sense appropriate for bills of attainder. Sir Douglas Menzies concluded [82] that, in conformity with the opinion of Frankfurter J., the Ceylon statute contained no declaration of guilt of bribery or of any other act. His Lordship continued [82] : It is the commission's finding that attracts the operation of the Act not any conduct of a person against whom the finding was made. Parliament did not make any finding of its own against the appellant or any other of the seven persons named in the schedule. The question of the guilt or innocence of the persons named in the schedule does not arise for the purpose of the Act and the Act has no bearing upon the determination of such a question should it ever arise in any circumstances. Accordingly, the statute did not interfere with the exercise of judicial power. 1. (1946) 328 U.S. 303. 2. (1946) 328 U.S., at pp. 322-323. 3. (1946) 328 U.S., at p. 323. 4. [1968] A.C., at p. 736. 5. [1968] A.C., at p. 736. All that I have said so far in refutation of this aspect of the plaintiff's case is confirmed by the decision of this Court in R. v. Kidman [83] . That case concerned the validity of the Crimes Act 1915 Cth so far as its provisions were retrospective. Section 2 of that Act added conspiracies to defraud the Commonwealth to the conspiracies which, by s. 86 of the Crimes Act 1914 Cth, were declared to be indictable offences. Section 3 of the 1915 Act provided that the Act was deemed to have been in force from the date of commencement of the 1914 Act. The accused were indicted for conspiracy to defraud the Commonwealth under the retrospective provisions of the 1915 Act. By a unanimous decision the validity of that Act was upheld. It was specifically contended that the Parliament had no power to enact retrospective criminal laws, reliance being placed on Calder v. Bull, the argument naturally being that the 1915 Act was an ex post facto law rather than a bill of attainder. 1. (1915) 20 C.L.R. 425. Griffith C.J. acknowledged [84] that an ex post facto law was forbidden by the United States Constitution but pointed out that no question of the validity of such a law could arise in the case of a legislature of plenary power. However, being of the opinion, mistakenly, that Parliament's power to enact a criminal law stemmed only from s. 51(xxxix), he concluded that this power did not extend to an ex post facto criminal law. He upheld the 1915 Act because in his opinion it did no more than re-enact the common law. 1. (1915) 20 C.L.R., at p. 432. Isaacs J. observed [85] : There is no prohibition in the Australian Constitution against passing ex post facto laws, as there is in the American Constitution The prohibition to the United States apparently assumes that Congress would otherwise have had the power. Therefore, in my opinion, no distinction can be validly drawn between ex post facto laws — regarding them as criminal only — and any other kind of retroactive laws. His Honour concluded by saying [86] : But the Parliament's powers are not confined to creating fear of punishment by threatening as to future acts, but extend to dealing with the conduct, which in its opinion deserves it, and so conveying the same warning and fear as a plenary Legislature within the ambit assigned to it. Higgins J. was of the same opinion [87] . His Honour specifically referred [88] to the fact that the Parliament of Great Britain had, by Acts of attainder and otherwise, made crimes of acts after they had been committed and held that, in the absence of a prohibition in the Constitution, the Commonwealth had like power to enact a retrospective criminal law. Likewise, Powers J. [89] considered that the Parliament had the same power to pass retrospective criminal laws with respect to the subject-matters committed to it by s. 51 of the Constitution as had the Imperial Parliament. 1. (1915) 20 C.L.R., at pp. 442-443. 2. (1915) 20 C.L.R., at p. 443. 3. (1915) 20 C.L.R., at pp. 451-454. 4. (1915) 20 C.L.R., at p. 451. 5. (1915) 20 C.L.R., at p. 462. In the light of what I have said earlier about the plaintiff's argument, the decision in Kidman was plainly correct. It has frequently been cited in subsequent decisions of the Court without any hint of disapproval: see R. v. Snow [90] ; Ex parte Walsh and Johnson; In re Yates [91] ; Millner v. Raith [92] ; Australian Communist Party v. The Commonwealth [93] ; University of Wollongong v. Metwally [94] . The only qualification relevant to the plaintiff's argument that needs to be made is that the separation of powers effected by our Constitution would invalidate a bill of attainder on the ground that it involves a usurpation of judicial power. To the extent that Higgins J. seems to suggest that such a bill, if enacted by the Parliament, would be valid, I am unable to agree. 1. (1917) 23 C.L.R. 256, at p. 265. 2. (1925) 37 C.L.R. 36, at pp. 86, 124-125. 3. (1942) 66 C.L.R. 1, at p. 9. 4. (1951) 83 C.L.R. 1, at p. 172. 5. (1984) 158 C.L.R. 447, at pp. 461, 484. True it is that the judgments in Kidman to which I have referred do not make any mention of the separation of powers. That is readily understandable. The challenge to the validity of the 1915 Act was, and could only be, that it was an ex post facto law, for it was not a bill of attainder. Before the present case it had never occurred to anyone to suggest that an ex post facto law of the kind under consideration here, not being a bill of attainder, could amount to a usurpation of judicial power because such an ex post facto law simply does not amount to a trial by legislature. It leaves for determination by the court the issues which would arise for determination under a prospective law. In the result, I would answer the question in the negative. Brennan J. The second defendant, with the claimed authority of the Commonwealth Director of Public Prosecutions, laid an information charging the plaintiff with the commission of a number of war crimes "within the meaning of section 9 of the War Crimes Act 1945 ". The War Crimes Act 1945 Cth was extensively amended by the War Crimes Amendment Act 1988 Cth which inserted a new preamble and new ss. 3 to 22 in the place of the preamble and ss. 3 to 14 of the original Act. The Act to which reference is made hereafter is the amended Act, except where the original Act (No. 48 of 1945) is expressly mentioned. The plaintiff, by his further amended statement of claim in an action brought against the Commonwealth and the second defendant, seeks declarations that the sections of the Act relevant to his prosecution are invalid and an injunction to restrain the defendants from taking any step in furtherance of the prosecution of the plaintiff for offences under the Act. The defendants assert that the Act is valid, relying on the external affairs power and the defence power conferred on the Parliament respectively by s. 51(xxix) and (vi) of the Constitution. The question reserved for consideration of the Full Court is as follows: "Is section 9 of the War Crimes Act 1945 as amended, invalid in its application to the information laid by the second defendant against the plaintiff?" That question is concisely drawn, but there is a more general question the answer to which may determine the answer to the question reserved and which is logically anterior to it. The more general question is this: is s. 9 as amended invalid? Or, even more generally, is the Act as amended invalid? The last of these questions should be addressed first. To appreciate the respective arguments for and against the validity of the Act, it is necessary to construe its terms and to ascertain its operation. 1. The meaning and operation of the Act Section 9 of the Act purports to create a new offence. It reads as follows: (1) A person who: (a) on or after 1 September 1939 and on or before 8 May 1945; and (b) whether as an individual or as a member of an organisation; committed a war crime is guilty of an indictable offence against this Act. (2) Sections 5 and 7, and paragraph 86(1)(a), of the Crimes Act 1914 do not apply in relation to an offence against this Act. The offence purportedly created by s. 9 is a statutory offence, the elements of which are to be found in other sections of the Act that give a statutory meaning to the term "a war crime". But it is clear from the terms of s. 9 itself that the acts which attract liability to conviction for the statutory offence are past acts. It is an offence that cannot now be committed, but it is an offence for which a person may be convicted by reason of past conduct. To be liable to conviction as for "an indictable offence against this Act" a person must have engaged in conduct of the prescribed kind — that is, conduct which answers the statutory definition of the term "a war crime" — during the period prescribed by par. (a) — that is, between 1 September 1939 and 8 May 1945 inclusive — and, when charged, be either: (a) an Australian citizen; or (b) a resident of Australia or of an external Territory: s. 11. The term "a war crime" is defined by reference to another term: "a serious crime". Section 6 contains the definition of "a serious crime" and conduct which answers that definition is "a war crime" if it was committed in circumstances which are prescribed by s. 7(1) or s. 7(3). Section 6 reads as follows: (1) An act is a serious crime if it was done in a part of Australia and was, under the law then in force in that part, an offence, being: (a) murder; (b) manslaughter; (c) causing grievous bodily harm; (d) wounding; (e) rape; (f) indecent assault; (g) abduction, or procuring, for immoral purposes; (h) an offence (in this paragraph called the "variant offence") that would be referred to in a preceding paragraph if that paragraph contained a reference to: (i) a particular intention or state of mind on the offender's part; or (ii) particular circumstances of aggravation; necessary to constitute the variant offence; (j) an offence whose elements are substantially the same as the elements of an offence referred to in any of paragraphs (a) to (h), inclusive; or (k) an offence of: (i) attempting or conspiring to commit; (ii) aiding, abetting, counselling or procuring the commission of; or (iii) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the commission of; an offence referred to in any of paragraphs (a) to (j), inclusive. (2) In determining for the purposes of subsection (1) whether or not an act was, under the law in force at a particular time in a part of Australia, an offence of a particular kind, regard shall be had to any defence under that law that could have been established in a proceeding for the offence. (3) An act is a serious crime if: (a) it was done at a particular time outside Australia; and (b) the law in force at that time in some part of Australia was such that the act would, had it been done at that time in that part, be a serious crime by virtue of subsection (1). (4) The deportation of a person to, or the internment of a person in, a death camp, a slave labour camp, or a place where persons are subjected to treatment similar to that undergone in a death camp or slave labour camp, is a serious crime. (5) Each of the following is a serious crime: (a) attempting or conspiring to deport or intern a person as mentioned in subsection (4); (b) aiding, abetting, counselling or procuring the deportation or internment of a person as so mentioned; (c) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the deportation or internment of a person as so mentioned. (6) For the purposes of subsections (3), (4) and (5), the fact that the doing of an act was required or permitted by the law in force when and where the act was done shall be disregarded. The acts which, by force of s. 6, are elements of the statutory offence may have been done during the prescribed period either in Australia (sub-s. (1)) or outside Australia (sub-s. (3)). If the relevant act was done inside Australia and if, under the law in force in that part of Australia at the time when the act was done, it amounted to any of the offences mentioned in sub-s. (1), the act answers the description of "a serious crime". If the act was done outside Australia but would have amounted to an offence of one of the kinds mentioned in sub-s. (1) had it been done at that time in some part of Australia by force of the law in that part, it is "a serious crime" for the purposes of sub-s. (3). The draftsman has not prescribed a particular Australian system of law to be the system to which reference must be made in determining, for the purposes of sub-s. (3), whether an act amounted to one of the offences mentioned. The consequence seems to be that, if an act, done within any part of Australia within the prescribed period, would have amounted to any of the offences mentioned under the law in force in that part of Australia at that time, the act is "a serious crime" for the purposes of s. 6(3). Presumably, sub-s. (2) is material in determining whether an act done outside Australia would have amounted to one of the offences mentioned in sub-s. (1) had it been done in some part of Australia, though it is artificial to apply a municipal system of law designed for the preservation of the King's peace to acts done by or on behalf of belligerents in war. An act which answers the description of "a serious crime" also answers the description of "a war crime" if it was done in circumstances defined by s. 7. That section reads as follows: (1) A serious crime is a war crime if it was committed: (a) in the course of hostilities in a war; (b) in the course of an occupation; (c) in pursuing a policy associated with the conduct of a war or with an occupation; or (d) on behalf of, or in the interests of, a power conducting a war or engaged in an occupation. (2) For the purposes of subsection (1), a serious crime was not committed: (a) in the course of hostilities in a war; or (b) in the course of an occupation; merely because the serious crime had with the hostilities or occupation a connection (whether in time, in time and place, or otherwise) that was only incidental or remote. (3) A serious crime is a war crime if it was: (a) committed: (i) in the course of political, racial or religious persecution; or (ii) with intent to destroy in whole or in part a national, ethnic, racial or religious group, as such; and (b) committed in the territory of a country when the country was involved in a war or when territory of the country was subject to an occupation. (4) Two or more serious crimes together constitute a war crime if: (a) they are of the same or a similar character; (b) they form, or are part of, a single transaction or event; and (c) each of them is also a war crime by virtue of either or both of subsections (1) and (3). The term "war" in this section is not left at large; it is defined by s. 5 to mean — (a) a war, whether declared or not; (b) any other armed conflict between countries; or (c) a civil war or similar armed conflict; (whether or not involving Australia or a country allied or associated with Australia) in so far as it occurred in Europe in the period beginning on 1 September 1939 and ending on 8 May 1945. Section 7 thus has effect, so far as it relates to "war", to any armed conflict occurring in Europe during the same period as that prescribed by s. 9(1)(a), whether the armed conflict involved Australia or not. The Commonwealth submitted that this section is to be read as importing by implication the requirement that the relevant conduct be a war crime or a crime against humanity by international law. This submission cannot be accepted. There is nothing in the framework of the Act which provides any foundation for implying an element which is not expressed. Although the suggested implied element would limit the exposure of persons to conviction, the Act expressly purports to deal with the requirements which, when added to the definition of "a serious crime", satisfy the element of "a war crime". As the preamble demonstrates, the Act was intended to provide for the bringing "to trial in the ordinary criminal courts in Australia" of "persons who committed serious war crimes in Europe during World War II", and ss. 6 and 7 were the legislative expression of what were perceived to be war crimes of serious degree. There is no legislative indication that the words of s. 7 should be read as importing into the definition of the term a war crime in s. 9 the definition of a war crime at international law. A war crime in international law is a violation of the laws and customs of war. Those laws and customs limit the belligerents' choice of means of injuring enemy forces, controlling enemy prisoners of war and treating enemy civilians, but are silent as to acts which do not contravene those limitations. Sections 6 and 7 do not import those limitations. Indeed, the sections assume that any killing, not justified or excused under the general law applicable (ordinarily to civilians) in Australia during the prescribed period, would amount to culpable homicide. To avoid so bizarre an operation of ss. 6, 7 and 9, the draftsman introduced s. 17(2) of the Act to confer immunity from conviction on persons whose acts, though falling within the category of "a war crime" as prescribed by ss. 6 and 7, were, at the time when they were done, "permitted by the laws, customs and usages of war" and were then "not under international law a crime against humanity". Section 17 provides: (1) This section has effect for the purposes of a proceeding for an offence against this Act. (2) Subject to section 16, it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. (3) To avoid doubt, the doing of the act by the defendant was permitted by the laws, customs and usages of war if it was reasonably justified by the exigencies and necessities of the conduct of war. (4) The defendant is not entitled to rely on a defence under subsection (2) unless there is evidence of the existence of the facts constituting the defence. (5) However, if there is such evidence, the onus of establishing, beyond a reasonable doubt, that those facts either do not exist or do not constitute the defence lies on the prosecution. The practical operation of sub-ss. (4) and (5) of this section may be open to debate, but its substantive effect is clearly to introduce a modification to the relevant system of Australian municipal law. Section 17 has been inserted in an attempt to provide, under municipal law, some justification or excuse for some acts of belligerents in war — a justification or excuse not usually provided by municipal criminal law: see Brierly, The Basis of Obligation in International Law (1958), pp. 297-298. In due course it will be necessary to consider whether the provisions of the Act correspond with the provisions of international law but, in considering whether the Act incorporates international law into its definition of "a war crime", we must refer briefly at this stage to some of the rules of international law. Thus, in construing s. 17, it is clear that the draftsman misunderstood the effect of the laws, customs and usages of war, for they give no permission for acts which would otherwise be forbidden; they simply forbid particular means of injuring enemy forces, controlling enemy prisoners of war or treating enemy civilians. Therefore, to give some effect to s. 17(2)(a), "permitted by" must be read as meaning "not in contravention of". However, by reading s. 17(2) in that way, any justification or excuse for "a serious crime" which was committed by a belligerent as a belligerent in time of war but which was not a violation of the law and customs of war must be found, if at all, in the relevant municipal law of a State or Territory. Yet the municipal law of the States and Territories contains no justification or excuse which might protect belligerents against liability to conviction for the deliberate killing of one who is not outside the protection of the municipal law. Alternatively, sub-s. 2(a) might be read as having the precise meaning which sub-s. (3) attributes to it, so that a "defence" is available when the act charged was reasonably justified by the exigencies and necessities of the conduct of war and was not under international law a crime against humanity. In due course, it will be necessary to consider more precisely the meaning and effect of s. 17(2) in order to ascertain whether it imports the elements of a war crime under international law. For the moment, it is sufficient to note that, once the reader is directed to the provisions of s. 17, it is impossible to read ss. 6 and 7 as though they impliedly imported international law to provide a definition of, or an element in the definition of, "a war crime" for the purposes of s. 9. The Act is drafted in language which not only bespeaks its turbulent legislative history but which rejects international law as the legal system by reference to which the elements of a war crime may be ascertained for the purposes of s. 9. (In this respect, as we shall see, the Act reverses the operation of the original Act which provided for the application of international law to the trial of war criminals.) The Act invokes in the first place the several systems of municipal law in force in Australia at the time when the relevant act was done: s. 6. Then, when a person is charged with the statutory offence, the reference in s. 6 to the law in force "in a part of Australia" is supplemented by s. 13(2) which reads as follows: Where a person is charged with an offence against this Act, then, for the purposes of: (a) determining whether a court of a State or internal Territory has jurisdiction in relation to the offence; (b) an exercise of jurisdiction by such a court in relation to the offence; (c) a proceeding connected with such an exercise of jurisdiction; and (d) an appeal arising out of, or out of a proceeding connected with, such an exercise of jurisdiction; this Act has effect, in relation to an act that is, or is alleged to be, the offence, as if: (e) a reference in subsection 6(3) or section 18 to a part of Australia were a reference to that State or Territory; and (f) without limiting subsection 6(2), all defences under the law in force in that State or Territory when the person is charged with the offence had been defences under the law in force in that State or Territory at the time of the act. The supplementation of the legal system applicable by force of s. 6 by the contemporary legal system of the place of trial is further modified by other provisions of the Act. Section 6(6) denies a defence of justification or excuse under the law of the place outside Australia where the relevant act was done. Not even the execution of judicial orders would, it seems, provide any justification or excuse for acts falling within s. 6(3), (4) or (5). Then s. 16 excludes the defence of superior orders. Section 16 reads: Subject to subsections 6(2) and 13(2), the fact that, in doing an act alleged to be an offence against this Act, a person acted under orders of his or her government or of a superior is not a defence in a proceeding for the offence, but may, if the person is convicted of the offence, be taken into account in determining the proper sentence. Section 20 instructs the reader that ss. 6(6) and 16 "are enacted to avoid doubt." Though s. 20 indicates that ss. 6(6) and 16 are not perceived by the Legislature to be statutory modifications of the applicable system of Australian municipal law, it is clear that s. 17 is intended to have that effect. From this review of the Act, it is clear that the elements of the statutory offence purportedly created by s. 9 are to be found solely in the provisions of the Act itself and, except for the limited purposes of s. 17, by reference to the municipal systems of law to which the Act refers. Indeed, s. 8(2) and (3) are consistent only with that view. Section 8 reads: (1) Subject to subsection 7(2), nothing in section 6 or 7 limits the generality of anything else in that section. (2) An act may be a serious crime by virtue of one or more of subsections 6(1), (3), (4) and (5), but not otherwise. (3) A serious crime may be a war crime by virtue of either or both of subsections 7(1) and (3), but not otherwise. (4) Two or more serious crimes may together constitute a war crime by virtue of subsection 7(4), but not otherwise. The Act thus exposes to punishment by a court of a State or Territory exercising jurisdiction under municipal law any person who (a) being an Australian citizen or resident when charged with the offence; (b) committed an act of the kind referred to in s. 6; (c) in Australia or outside Australia; (d) between 1 September 1939 and 8 May 1945 (inclusive); (e) in circumstances (defined in s. 7) connected with any armed conflict in Europe between the dates mentioned, whether Australia was involved in that conflict or not; (f) being an act that is not "permitted by" the laws, customs and usages of war or being under international law a crime against humanity. It is immaterial that, when the relevant act was done, the person who did it was not then an Australian citizen or resident, that no Australian citizen or resident nor any person under or entitled to the protection of Australian law was the victim or likely victim, that the armed conflict in the course of which the act was committed did not involve Australia or that the act was lawful according to the laws of the place where the act was done at the time when it was done. Whether or not any relevant Australian interest was involved when the relevant act was done, it is sufficient for the purposes of the Act that the person who did the act has become an Australian citizen or an Australian resident. The Act is truly retrospective in its operation: that is to say, it attaches penalties under Australian municipal law to the doing of an act to which that penalty was not attached when the act was done. Under this Act, the plaintiff is charged with offences allegedly committed in 1942 and 1943 in the Ukraine during Germany's occupation of that territory. It is alleged that he wilfully killed a number of people in pursuit of German policies — either a policy of persecuting the Jewish people or those opposed to German policies or a policy of annihilating suspected partisans or communists. The plaintiff was not then an Australian citizen or resident. If the allegations made in the charges be true, the plaintiff was guilty of heinous offences against the laws of the Ukraine and against the laws and customs of war but he is not charged with offences against the laws of the Ukraine nor, as we shall see, with offences against the laws and customs of war. He is charged with offences against the municipal law of Australia, created by s. 9 of the Act. The validity of the Act depends upon the legislative power of the Parliament to create the offence defined by the Act and to vest jurisdiction in Australian courts to try persons charged with that offence. As the Act is retrospective in its operation, I assume the defendants would seek constitutional support for the Act not only in sub-ss. (vi) and (xxix) of s. 51 but also in sub-s. (xxxix) of s. 51 (the sub-section on which reliance was placed in R. v. Kidman [95] to support the retrospective criminal law that was there upheld). It will be convenient to consider first the various ways in which the external affairs power is said to support the Act. 1. (1915) 20 C.L.R. 425. 2. The external affairs power The Constitution, by s. 51(xxix), confers on the Parliament plenary legislative power with respect to external affairs, but the support given by the external affairs power to a particular law depends upon the identification of some relationship, set of circumstances or field of activity which falls within the description of "external affairs" and with which the law has a connexion sufficient to satisfy the requirement imported by the words "with respect to" in s. 51: Koowarta v. Bjelke-Petersen [96] . 1. (1982) 153 C.L.R. 168, at p. 256. The Commonwealth submits that the Act is supported by s. 51(xxix) in a number of ways. First, the Act is said to be a law "with respect to things done outside Australia" and that a law "upon anything whatsoever done outside Australia" is supportable, "at least where there is a sufficient connection with Australia". Next, the Act is said to be a law enacted for purposes which attract the support of the power, the relevant purposes being identified as — (A) the discharge of an international obligation ; (B) the exercise of an international right ; (C) the meeting of an international concern; or (D) the implementation of resolutions or recommendations of the United Nations General Assembly ("UNGA") or any other international body. The first way in which the argument is put propounds the conduct in which individuals engage outside Australia as the relevant category of "external affairs" with which the Act is sufficiently connected. The next way in which the argument is put depends upon the international personality of Australia and propounds Australia's relationship with the community of nations as the relevant category of "external affairs". The two ways in which the argument is put call for separate consideration. (i) External acts Put at its broadest, the Commonwealth's argument attributes to the external affairs power a scope which would empower the Commonwealth to pass a law affecting the doing of anything by anybody outside Australia: acts which are external to Australia are said to be external affairs and a law relating to any external acts is said to be a law with respect to external affairs. In New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [97] Barwick C.J. said: The [external affairs] power extends, in my opinion, to any affair which in its nature is external to the continent of Australia and the island of Tasmania subject always to the Constitution as a whole. A wide view of the power was also taken by Mason, Jacobs and Murphy JJ. [98] . In Koowarta [99] , Mason J. said that " "external affairs" covers any matters or concerns external to Australia." In the contexts in which these observations were made, no question arose as to the connexion between Australia and the extraterritorial person, thing or event to which the particular Commonwealth law applied. Australia's connexion with its maritime boundaries in the one case and with its treaty obligations in the other was manifest. But it is argued that, although there must be some connexion between the extraterritorial operation of a State law and the extraterritorial persons, things or events on which a State law operates (Port MacDonnell Professional Fishermen's Association Inc. v. South Australia [1] ), no such connexion is required in the case of a Commonwealth law. If such a connexion were required, so the argument runs, there would be a lacuna in the plenitude of legislative powers which the Australian legislatures, Commonwealth and State, together possess: see per Jacobs J. in the Seas and Submerged Lands Case [2] . But the powers conferred by the Constitution are not to be expanded beyond their true scope merely to supply what is thought, from the public viewpoint, to be a desirable or convenient power. Limits on power are the measure of private immunity from legislative action by the State. The legislative powers of the Parliament are limited by the terms of the Constitution, and the connotation of the phrase "external affairs" must be ascertained from its context and purpose. Accepting fully that s. 51(xxix) is not to be narrowly construed (The Commonwealth v. Tasmania (the Tasmanian Dam Case) [3] ), nevertheless the power thereby conferred is limited. Although the phrase "for the peace, order, and good government of the Commonwealth" itself contains no territorial limitation (Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [4] ), it does not expand the connotation of the phrase "External affairs" in s. 51(xxix). Just as the power conferred by s. 51(xx) "remains a power to make laws with respect to immigration into and emigration from Australian territory", as Menzies J. pointed out in Reg. v. Foster [5] , so the power conferred by s. 51(xxix) remains a power to make laws with respect to Australia's external affairs. I do not understand the phrase "external affairs" to sweep into Commonwealth power every person who exists or every relationship, set of circumstances or field of activity which exists or occurs outside Australian territory. The "affairs" which are the subject matter of the power are, in my view, the external affairs of Australia; not affairs which have nothing to do with Australia. Although affairs which exist or occur outside Australia may be described as "external" in a geographical sense, I would not hold that the Constitution confers power to enact laws affecting affairs which, though geographically external, have nothing to do with Australia. There must be some nexus, not necessarily substantial, between Australia and the "external affairs" which a law purports to affect before the law is supported by s. 51(xxix). 1. (1975) 135 C.L.R. 337, at p. 360. 2. (1975) 135 C.L.R., at pp. 470-471, 497, 503. 3. (1982) 153 C.L.R., at p. 223. 4. (1989) 168 C.L.R. 340, at p. 372. 5. (1975) 135 C.L.R., at p. 497. 6. (1983) 158 C.L.R. 1, at pp. 220-221. 7. (1959) 103 C.L.R. 256, at pp. 300-301, 307-308. 8. (1959) 103 C.L.R., at p. 301. It may be that this view of the scope of the external affairs power corresponds with the legislative competence of Australia as recognized by international law. The internationally recognized competence of a domestic legislature is illustrated by " Le Louis " [6] , a case in which a question arose as to the limits of domestic legislative power under international law. In proceedings for the condemnation of a French ship for being employed in the slave trade, the seizors sought to rely upon a British statute which made the slave trade illegal. At that time, slavery was not illegal by the law of nations and it was held that "[t]he Legislature must be understood to have contemplated all that was within its power, and no more" [7] . Although the power of the Legislature to enact extraterritorial legislation was not denied, it was acknowledged that, by the law of nations, the plenitude of legislative power claimed by the United Kingdom Parliament would not be recognized internationally where there was no occasion for its exercise. The general terms of the statute were construed in accordance with the limits recognized by international law. However, the correspondence between the scope of the power conferred by s. 51(xxix) and that recognized by the law of nations is not a question which we have to decide in this case. The scope of the constitutional power is not determined by the law of nations, much less by international opinion as to Australia's connexion with a particular subject matter. We are here concerned solely with the limits of a power conferred by the Constitution. The recent cases relating to s. 51(xxix) show that the power thereby conferred enables the Commonwealth to legislate for the purpose of discharging the responsibilities and asserting to the full the interests of Australia as an independent member of the community of nations: the Seas and Submerged Lands Case; Koowarta ; the Tasmanian Dam Case; Gerhardy v. Brown [8] ; Richardson v. Forestry Commission [9] ;Mabo v. Queensland [10] . It is a plenary power exercisable as well in protection of Australia's international interests as in performance of its international obligations. But s. 51(xxix) does not arm the Commonwealth with power to enact laws governing affairs outside Australia with which Australia has no connexion. It is, of course, for the Parliament to determine in the first instance whether there is any connexion between Australia and a relationship, set of circumstances or field of activity which exists or occurs outside Australia and which a proposed law would purportedly affect, but, if the legislative judgment cannot reasonably be supported, the law will be held to be outside the power conferred by s. 51(xxix): see Richardson [11] ; Gerhardy v. Brown [12] . It leaves no great lacuna in the plenitude of Australian legislative power to deny the character of a law with respect to external affairs to a law which does no more than affect something or somebody unconnected with Australia occurring or existing outside Australia. To take an extreme example: would a law be properly characterized as a law with respect to external affairs if it imposed a criminal penalty on a person who, being a citizen and resident of France, had dropped litter in a Parisian street forty years ago? The limits of the power conferred by s. 51(xxix) are, in a real sense, a guarantee of the immunity from harassment by Australian law of persons who, having no connexion with Australia, engage in conduct elsewhere which does not affect Australian interests or concerns. 1. (1817) 2 Dods. 210 [165 E.R. 1464]. 2. (1817) 2 Dods., at p. 254 [165 E.R., at p. 1479]. 3. (1985) 159 C.L.R. 70. 4. (1988) 164 C.L.R. 261. 5. (1988) 166 C.L.R. 186. 6. (1988) 164 C.L.R., at p. 296. 7. (1985) 159 C.L.R., at p. 139. The requirement of some connexion between Australia and the relationship, set of circumstances or field of activity affected by a law is satisfied when the law purports to control extraterritorial conduct engaged in by Australian citizens or residents (whether natural or corporate) or by persons who are under or are entitled to the protection of Australian law. The extraterritorial conduct of such persons is an aspect of Australia's external affairs. Equally, the requirement of connexion is satisfied when the law, in protection of the interests of any such persons, purports to prohibit conduct engaged in by others outside Australia. The external protection of the interests of those under the protection of Australian law, wherever those interests might be located, is another aspect of Australia's external affairs. A law governing conduct of the kinds referred to in this paragraph answers the description of a law with respect to Australia's external affairs. However, the Act does not purport to affect the conduct of Australian citizens or residents outside Australia. It does not attach any penalty to conduct in which any such person might engage outside Australia. The conduct which is made a condition of present liability to conviction is past conduct of the kinds prescribed by the Act which occurred during the prescribed period, being conduct that was engaged in by any person. It is immaterial to liability whether the victims of any such conduct were under the protection of Australian law. If it be said that the Act bears the character of a law with respect to the "serious crimes" which fall within s. 7(1) or s. 7(3), the question is whether those crimes, at the time when they were committed, were then an aspect of Australia's external affairs. Nothing in the Act suggests that a serious crime falling within s. 7(1) or s. 7(3) must be an aspect of Australia's external affairs. The Act does not prescribe as an element of conduct answering the description of "a war crime" anything which connects Australia with that conduct at the time it was engaged in. The element which at first sight may seem to provide some contemporaneous nexus between "a war crime" in s. 9 and Australia is the definition of war in s. 5, for Australia was involved in the European war between 3 September 1939 and 8 May 1945. But, apart from the difference in the commencing date of the prescribed period, the definition of war in s. 5 rejects Australia's involvement in the Second World War as a necessary element in the definition, for the definition is satisfied by any armed conflict "(whether or not involving Australia or a country allied or associated with Australia)". If conduct of the prescribed kind which occurred during the prescribed period was not then an aspect of Australia's external affairs, a fortiori it could not have been an aspect of Australia's external affairs when the Act came into operation on 25 January 1989. Nearly forty-four years had passed between 8 May 1945 and the date when the Act came into operation. Unless Australia's external relations were affected by unpunished "serious crimes" committed more than forty-four years earlier (a question presently to be examined) I am unable to see how an act by an individual in the course of any European armed conflict occurring between 1 September 1939 and 8 May 1945 could be regarded, without more, as an aspect of Australia's external affairs in 1989. If "a serious crime" committed by an individual forty-four or more years earlier answering the statutory description of "a war crime" was not itself an aspect of Australia's contemporary external affairs before 8 May 1945, does it make any difference to the character of the Act that the only persons who are liable to conviction must now be Australian residents or citizens? The Act, as we have seen, is a retrospective law which attaches penal consequences to conduct engaged in before the law was made. If the Parliament has power to enact a law which prohibits Australian residents and citizens from engaging in particular conduct outside Australia in future, does it have power to enact a law which exposes to conviction Australian residents and citizens who have engaged in particular conduct in the past? Were the Act restricted in its application to persons who were Australian residents or citizens at the time when "a war crime" was committed, it would be strongly arguable, on the authority of Kidman [13] , that the Act is supportable under s. 51(xxxix) as a retrospective law with respect to the conduct of Australian residents and citizens overseas. In Kidman , a majority of this Court upheld the validity of a law deeming past conduct in defrauding the Commonwealth to be criminal and imposing penal consequences. By analogy, it could be argued that the extraterritorial commission of "a war crime" by Australian residents and citizens was, at that time, an aspect of Australia's external affairs and the retrospective imposition of a criminal penalty on them for engaging in that conduct is simply an exercise of the incidental power which empowers the Parliament to make retrospective laws to punish conduct that was at all material times a subject of legislative power. It is unnecessary in the present case to determine whether, if the Act had retrospectively created an offence which only persons who were Australian citizens or residents at the time could have committed, the Act might have been supported on the authority of Kidman . It is common ground that, at the times of the alleged occurrences of the acts charged against the plaintiff, the plaintiff had no connexion with Australia and was not then either a citizen or a resident of Australia. He subsequently became and is now an Australian citizen and a resident of the State of South Australia. Does his becoming a citizen and resident of Australia transform his earlier conduct into an aspect of Australia's external affairs? 1. (1915) 20 C.L.R. 425. In answering this question, I again put to one side the argument that Australia's international relations constitute the aspect of Australia's external affairs that enlivens the legislative power in s. 51(xxix) and supports the enactment of the Act. The argument which I now address is the argument that extraterritorial conduct, without more, has a sufficient connexion with Australia to enliven the external affairs power if the person who engaged in that conduct subsequently becomes an Australian citizen or resident. If it be right to say that the external affairs which may enliven legislative power under s. 51(xxix) are affairs with which Australia has some connexion, the scope of the power may alter from time to time as an Australian connexion with particular relationships, sets of circumstances or fields of activity comes into existence or disappears. Like the defence power (see, e.g., Hume v. Higgins [14] ), the scope of the external affairs power must be affected by the international conditions prevailing at the material time. It follows that, if there was no power to prohibit conduct outside Australia by a person who was not a citizen or resident at the time when the conduct was engaged in because that conduct was not an aspect of Australia's external affairs, the subsequent acquisition of Australian citizenship or residence cannot make that conduct an aspect of Australia's external affairs unless there be some international relationship which requires Australia to impose on a person who becomes a citizen or resident a penalty or liability for that conduct. There is no logical or legal reason why a power to forbid extraterritorial conduct engaged in by an Australian citizen or resident — which is an aspect of Australia's external affairs — should be expanded to a power to attach a penalty to persons for engaging in conduct which was not an aspect of Australia's external affairs unless, perhaps, the attaching of such a penalty (as distinct from the conduct) is itself required to satisfy Australia's international relations. 1. (1949) 78 C.L.R. 116, at pp. 133-136. If the law would have been beyond power had it been in force when the relevant conduct was engaged in, the power cannot now be enlarged by retrospective legislation. Legislative power cannot be exercised retrospectively to transform a relationship, set of circumstances or field of activity which was not an aspect of Australia's external affairs when it existed or occurred into an aspect of Australia's present external affairs. A law cannot create the facts which condition the power needed for its own support. A purported exercise of a legislative power cannot itself enliven the power to be exercised. Australian citizenship or residence in Australia is not a solvent of responsibility for crimes committed elsewhere; Australia's extradition obligations are discharged and its deportation powers are exercised in order to ensure that this country does not become an Alsatia for criminals from overseas. But the mere acquisition of Australian citizenship or residence in Australia does not transform earlier extraterritorial conduct that was not a matter of Australia's external affairs when it was engaged in into a matter of Australia's external affairs. Something more is needed to make transgressions of other laws in other places in other times an aspect of Australia's external affairs. It follows that, in my opinion, the Act cannot be supported as a law with respect to external affairs if no more than conduct answering the description of "a war crime" as that term is used in s. 9 is relied on as the relevant aspect of external affairs; nor can the Act be so supported if, in addition to that conduct, the subsequent acquisition of Australian citizenship and residence is relied on to constitute the relevant aspect of external affairs. The Commonwealth did not submit that the Act is supportable as a law with respect to immigration, citizenship or the influx of criminals. I pass now to a consideration of a more substantial argument, namely, that the external affairs power supports the Act because Australia's international relationships require or permit it to attach liability to criminal conviction to any person who has committed an act which amounts to "a war crime" as that term is used in s. 9. (ii) International relations (a) International obligation or concern and the right to try The Commonwealth submits that the Act discharges an international obligation or meets an international concern that persons alleged to be guilty of war crimes and crimes against humanity be sought out, brought to trial and, upon conviction, punished, irrespective of the place where the crime was committed or where the alleged offender is found and irrespective of the citizenship or residence of the alleged offender or the victim. The propounded matter of international obligation or international concern is not stated merely in terms of seeking out and punishing persons guilty of war crimes and crimes against humanity. A process of trial is concededly an integral part of the postulated obligation or concern. Therefore, the postulated international obligation or matter of international concern is that the guilt of an alleged criminal be ascertained by a court of competent jurisdiction administering a system of law which creates or recognizes the crime with which the alleged criminal is charged. The international instruments relating to the punishment of war criminals have uniformly contemplated trials according to law. In 1942, the Declaration of St. James placed among the main goals of war the punishment, "through the channels of organized justice", of war criminals. In 1943, the Moscow Declaration declared that those responsible for or taking part in atrocities, massacres and mass executions would be "sent back to the countries in which their abominable deeds were done in order that they may be judged and punished according to the laws of these liberated countries and of the free governments which will be created therein". That Declaration was echoed by a unanimous recommendation of the United Nations General Assembly in Resolution 3(I) of 13 February 1946 which was reaffirmed by Resolution 170(II) of 31 October 1947. The latter resolution "[r]easserts that trials of war criminals and traitors, like all other trials, should be governed by the principles of justice, law and evidence." Then, following a series of further resolutions, the General Assembly declared on 3 December 1973 (Resolution 3074(XXVIII)) that, inter alia — 5. Persons against whom there is evidence that they have committed war crimes and crimes against humanity shall be subject to trial and, if found guilty, to punishment, as a general rule in the countries in which they committed those crimes. In that connexion, States shall co-operate on questions of extraditing such persons. The problem of statutory limitations on the prosecution of war crimes had been the subject of a United Nations Convention in 1968 but most members of the Council of Europe found that Convention unacceptable because of its broad definition of "crimes against humanity". A European Convention, drawn more narrowly, was signed in 1974 calling on each Contracting State to lift time limitations on the prosecution of certain war crimes and genocide punishable "under its domestic law" committed after the particular State ratified or acceded to the Convention or prior to that time if the limitation period had not then expired. There can be no doubt but that, at least until 1974, there was a widespread desire expressed by the community of nations that the Axis perpetrators of war crimes during the Second World War should be apprehended, tried and, if found guilty, punished. Later evidence of this desire is meagre, though an agreement between the United States and the Union of Soviet Socialist Republics in 1989 relating to co-operation in investigating such war crimes suggests that some countries have continued the pursuit of war criminals. Although the instruments abovementioned gave jurisdictional priority to the courts of the States in whose territories the crimes were committed, there was also a general recognition that some crimes are international and are amenable to the jurisdiction either of international tribunals or of the courts of any State into whose hands the alleged criminal might fall. The Charter of the International Military Tribunal for the Prosecution and Punishment of the Major War Criminals (the Nuremberg Tribunal) which was attached to an Agreement dated 8 August 1945 between the major Allied Powers for the Prosecution and Punishment of the Major War Criminals of the European Axis contained a list of recognized war crimes in Art. 6(b). By its Resolution 95(I) of 11 December 1946, the General Assembly of the United Nations reaffirmed "the principles of international law recognized by the Charter of the Nürnberg Tribunal and the judgment of the Tribunal". Subsequent resolutions of the General Assembly or of the Economic and Social Council of the United Nations requested member States to ensure that "in accordance with international law and national laws, the criminals responsible for war crimes and crimes against humanity are traced, apprehended and equitably punished by the competent courts" (ECOSOC Resolution 1074D (XXXIX) of 28 July 1965). Moreover, each of the Geneva Conventions of 1949 prescribed a number of "grave breaches" which might be committed against persons or property protected by that Convention and contained a common provision in these terms: The High Contracting Parties undertake to enact any legislation necessary to provide effective penal sanctions for persons committing, or ordering to be committed, any of the grave breaches of the present Convention defined in the following Article. Each High Contracting Party shall be under the obligation to search for persons alleged to have committed, or to have ordered to be committed, such grave breaches, and shall bring such persons, regardless of their nationality, before its own courts. It may also, if it prefers, and in accordance with the provisions of its own legislation, hand such persons over for trial to another High Contracting Party concerned, provided such High Contracting Party has made out a prima facie case. (See Art. 49, Geneva Convention I; Art. 50, Geneva Convention II; Art. 129, Geneva Convention III; Art. 146, Geneva Convention IV.) These instruments establish that the trial and punishment of war crimes was, at least for many years after the Second World War, a matter of serious international concern and has become, under the Geneva Conventions, a matter of treaty obligation. In its Resolution 2712(XXV) of 15 December 1970, the General Assembly of the United Nations recognized that war crimes and crimes against humanity were still being committed, and requested States to take appropriate action to bring war criminals and persons guilty of crimes against humanity to justice. Yet Australia, in common with most other nations, failed for many years to respond to that request. The primary question on this branch of the case is whether the material relied on establishes that in 1989 there was either an obligation under customary international law or a matter of international concern that war criminals from the pre-1945 years be sought out and tried for their offences. As the sources of the postulated obligation and of the postulated concern are the same, there is no difference in content between the obligation and the concern. There are no relevant treaty obligations. The treaty obligations imposed by the Geneva Conventions of 1949 were not retrospective. The legislative obligations accepted by Australia under those Conventions were fulfilled by the Geneva Conventions Act 1957 Cth which substantially translated the Convention provisions into Australian municipal law. Although the material demonstrates that there was a widespread aspiration that the war criminals of the Axis powers should be brought to justice after the Second World War and although that aspiration was repeated in a series of resolutions in the UNGA and in the Economic and Social Council, the practice of States in the community of nations does not reveal a widespread exercise of jurisdiction to try alleged war criminals for extraterritorial war crimes. European States have exercised jurisdiction in respect of war crimes committed in their respective territories, but Israel and Canada are the only States which have asserted jurisdiction to try alleged war criminals in respect of extraterritorial war crimes. To determine whether, in these circumstances, there exists a customary law obligation to try alleged war criminals in respect of extraterritorial war crimes, it is necessary to refer to the sources of international law. Article 38(1) of the Statute of the International Court of Justice is generally regarded as a complete statement of the sources of international law: Brownlie, Principles of Public International Law, 4th ed. (1990), p. 3. It provides: The Court, whose function is to decide in accordance with international law such disputes as are submitted to it, shall apply: (a) international conventions, whether general or particular, establishing rules expressly recognized by the contesting States; (b) international custom, as evidence of a general practice accepted as law; (c) the general principles of law recognized by civilized nations; (d) subject to the provisions of Article 59, judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law. In the absence of international conventions, the custom required to evidence "a general practice accepted as law" must be "extensive and virtually uniform" (North Sea Continental Shelf Cases [15] ) and "followed on the basis of a claim of right and, in turn, submitted to as a matter of obligation" (MacGibbon, "Customary International Law and Acquiescence", British Year Book of International Law 115, vol. XXXIII (1957), at p. 117). In Nicaragua v. United States of America [16] , the International Court of Justice accepted that it is "sufficient that the conduct of States should, in general, be consistent with" a postulated rule of international law, but that was a view expressed in conjunction with an inquiry whether there was an opinio juris as to the binding character of the postulated rule (see pp. 99-101). An opinio juris supportive of a postulated rule of customary international law must explain and inform the practice of States in order to show that that practice is "accepted as law". The principle is conveniently stated by Dr Akehurst's summary of his article "Custom as a Source of International Law", British Year Book of International Law, vol. XLVII (1974-1975) 1, at p. 53: Opinio juris is necessary for the creation of customary rules; State practice, in order to create a customary rule, must be accompanied by (or consist of) statements that certain conduct is permitted, required or forbidden by international law (a claim that conduct is permitted can be inferred from the mere existence of such conduct, but claims that conduct is required or forbidden need to be stated expressly). It is not necessary that the State making such statements believes them to be true; what is necessary is that the statements are not challenged by other States. In the present case, there is no evidence of widespread State practice which suggests that States are under a legal obligation to seek out Axis war criminals and to bring them to trial. There is no opinio juris supportive of such a rule. 1. [1969] I.C.J.R. 1, at p. 43. 2. [1986] I.C.J.R. 14, at p. 98. Although there be no obligation in international law to which reference might be made in support of the validity of the Act, the external affairs power is not restricted to the support of laws passed to fulfil international treaty obligations (Richardson [17] ) or to fulfil other obligations under customary international law. It has been recognized that the power may be enlivened by circumstances which do not give rise to an obligation under international law: the Tasmanian Dam Case [18] . Those circumstances have been said to be sufficient if they reveal that the subject matter of the law is a matter of "international concern". The reason why matters of international concern enliven the power was explained by Stephen J. in Koowarta [19] : A subject-matter of international concern necessarily possesses the capacity to affect a country's relations with other nations and this quality is itself enough to make a subject-matter a part of a nation's "external affairs". 1. (1988) 164 C.L.R., at p. 289. 2. (1983) 158 C.L.R., at pp. 129-132, 171-172, 222, 258-259. 3. (1982) 153 C.L.R., at p. 217. It is clear that the term "international concern" possesses no very precise meaning, for it may cover a diverse multitude of topics. However, its meaning cannot carry the external affairs power beyond the fulfilment of the purpose for which it was conferred. The Constitution summoned into existence a new Commonwealth that was to take its place among the nations of the world and, in the fullness of time, to act and to be seen to be acting independently of the Imperial Government. The Commonwealth was equipped not only with executive power, so that its voice might be heard and its actions might be seen by the international community, but with legislative power, so that the Commonwealth might ensure that our municipal laws were conducive to the discharge of our international duties and the effective assertion of our international rights. As treaty obligations and other obligations in international law possess the capacity to affect Australia's relations with other countries, those obligations are a matter of international concern, but not every subject of international dialogue or even of widespread international aspiration has the capacity to affect Australia's relations. One purpose of the external affairs power is to furnish the Commonwealth with legislative authority to ensure that Australia acts in accordance with standards expected of and by the community of nations, even though those standards are not, or have not yet achieved the status of, obligations in international law. The observation of those standards may rightly be regarded as a matter of international concern. However, unless standards are broadly adhered to or are likely to be broadly adhered to in international practice and unless those standards are expressed in terms which clearly state the expectation of the community of nations, the subject of those standards cannot be described as a true matter of international concern. It may be that there are few occasions when the external affairs power is enlivened by the existence of a matter of international concern without a corresponding obligation in international law, but whether the enlivening factor be an obligation or a concern it is necessary to define it with some precision in order to ascertain the scope of the power. Understanding the notion of international concern in this way, the relationship between obligation and concern can be perceived — a relationship which requires elucidation in order to answer the question asked by Dawson J. in Richardson [20] : " if international concern is the touchstone, why is a treaty necessary at all? Why is international concern over a matter not sufficient of itself to bring it within the external affairs power?" It would be erroneous to attribute a scope to the external affairs power which depended on the broadest meaning which could be given to the imprecise phrase "international concern": that phrase is not a constitutional text and it is used to indicate that the power relates to matters affecting Australia's external relations even if those matters are not obligations under international law. 1. (1988) 164 C.L.R., at pp. 322-323. In the present case, the postulated matter of international obligation or concern is not simply the punishment of war criminals but the seeking out, bringing to trial and punishment of war criminals. In my view, the material relied on to establish that Australia is or was obliged to take steps now to bring to trial in Australia suspected war criminals from the Second World War fails to do so, but the material does establish that the apprehension and prosecution of war criminals from the Second World War before international tribunals or before courts of the country in which the crimes were committed were matters of international concern for many years after 1945. Whether those matters were still of international concern in 1989 may be doubted. There is insufficient material to show that the apprehension and trial of such war criminals before courts of countries other than those in which the crimes were committed were ever matters of international concern. However, I need not and I do not rest my judgment on this view for there is a further argument which depends simply on the existence of a universal jurisdiction to try international crimes. In the way in which this argument was first put by the Commonwealth, the Act was said to be a law adapted and appropriate to the exercise of a right which international law specially confers on each nation to try those charged with the commission of international crimes, especially war crimes. At first, the submission made by the Commonwealth identified the right to try as a right to try allegations of guilt of crimes defined by international law, albeit that definition was adopted by municipal law and applied as such. Although the terminology which equates universal jurisdiction with a right might be open to question, I would hold that a law which vested in an Australian court a jurisdiction recognized by international law as a universal jurisdiction is a law with respect to Australia's external affairs. Australia's international personality would be incomplete if it were unable to exercise a jurisdiction to try and to punish offenders against the law of nations whose crimes are such that their subjection to universal jurisdiction is conducive to international peace and order. As the material drawn from international agreements and UNGA resolutions acknowledges, international law recognizes a State to have universal jurisdiction to try suspected war criminals whether or not that State is under an obligation to do so and whether or not there is any international concern that the State should do so. Even if there were an international obligation or concern as postulated by the Commonwealth, its content must be this: that States should try war criminals in exercise of their universal jurisdiction to do so, for there is nothing to suggest that the postulated obligation or concern was to be satisfied or met by a trial in exercise of a jurisdiction which is not conferred by international law. A municipal law which provides for the exercise of a universal jurisdiction recognized by international law must prescribe an appropriate system of law by which to try an alleged criminal and an appropriate court to exercise jurisdiction in the case. The present argument thus raises for consideration two questions: 1. what system of law does the community of nations recognize as applicable to the creation and definition of war crimes and crimes against humanity? and 2. what courts are recognized to possess universal jurisdiction to apply the relevant system of law? International law recognizes certain international crimes in respect of which any country may exercise criminal jurisdiction regardless of the citizenship or residence of the alleged offender or of the place where the offence was committed: Halsbury's Laws of England, 4th ed., vol. 18, par. 1529; Professor Green, International Crimes and the Legal Process, International and Comparative Law Quarterly 567, vol. 29 (1980), at p. 568. Brownlie, op. cit., p. 305, states the principle in this way: It is now generally accepted that breaches of the laws of war, and especially of the Hague Convention of 1907 and the Geneva Convention of 1949, may be punished by any state which obtains custody of persons suspected of responsibility. This is often expressed as an acceptance of the principle of universality, but this is not strictly correct, since what is punished is the breach of international law; and the case is thus different from the punishment, under national law, of acts in respect of which international law gives a liberty to all states to punish, but does not itself declare criminal. Piracy is an older example of a crime under international law as war crimes now are. Indeed, one author has described war crimes as "international crimes par excellence " (Van den Wijngaert, War Crimes, Crimes Against Humanity, and Statutory Limitations in Bassiouni (ed.), International Criminal Law, vol. 3, (1987), p. 91). In In re List (Hostages Trial) [21] , the United States Military Tribunal sitting at Nuremberg said: An international crime is such an act universally recognized as criminal, which is considered a grave matter of international concern and for some valid reason cannot be left within the exclusive jurisdiction of the state that would have control over it under ordinary circumstances. The inherent nature of a war crime is ordinarily itself sufficient justification for jurisdiction to attach in the courts of the belligerent into whose hands the alleged criminal has fallen. 1. (1948) 15 Annual Digest 632, at p. 636. A war crime in international law consists in a violation of those laws and customs of war — "breaches of the laws of war" as Brownlie calls them — which oblige belligerents to abstain from prescribed anti-humanitarian acts in the waging of armed conflicts. The laws and customs of war are prohibitory in nature; they do not authorize the use of force. M.W. Mouton, the Netherlands Representative on the United Nations War Crimes Commission, writing on War Crimes and International Law in Grotius International Yearbook (1940-1946), said (at p. 59): The laws and customs of war never justify any killing, on the contrary they limit the right as to the choice of means of injuring the enemy (Hague regulations art. 22); forbid unnecessary cruelty and acts of barbarity against the enemy, limit the evils of war and unnecessary suffering and protect the civil population of an occupied country (see the preamble of the Hague regulations) and try to make lasting peaceful relations after the hostilities possible. Killing is not punishable in the course of combat. But killing is the form of execution between parties who have taken the "law" into their own hands for want of a court or a judge. The laws and customs of war are the procedural rules about this execution, laying down e.g. that torture is not allowed, that killing should be limited to armed combatants in action. The prohibitory nature of the laws and customs of war is illustrated by Art. 22 of the Hague Regulations of 1907 which provides: The right of belligerents to adopt means of injuring the enemy is not unlimited. War crimes, being violations of the laws and customs of war, thus consist in acts which transgress the limitations imposed by those laws and customs. Such transgressions are universally condemned and are internationally recognized as crimes which can be tried according to international law by the courts of any nation into whose hands the offender falls. The same national competence was recognized in relation to the offence of piracy, as the Privy Council observed in In re Piracy Jure Gentium [22] : With regard to crimes as defined by international law, that law has no means of trying or punishing them. The recognition of them as constituting crimes, and the trial and punishment of the criminals, are left to the municipal law of each country. But whereas according to international law the criminal jurisdiction of municipal law is ordinarily restricted to crimes committed on its terra firma or territorial waters or its own ships, and to crimes by its own nationals wherever committed, it is also recognized as extending to piracy committed on the high seas by any national on any ship, because a person guilty of such piracy has placed himself beyond the protection of any State. He is no longer a national, but "hostis humani generis" and as such he is justiciable by any State anywhere. Their Lordships' statement that recognition of crimes as defined by international law is "left to the municipal law of each country" should not be understood to mean that international law accepts whatever definition of an international crime the municipal law may contain. Rather, what is left to municipal law is the adoption of international law as the governing law of what is an international crime. So much appears from the judgment of Judge Moore (dissenting, but not on this point) in The S.S. "Lotus" [23] : in the case of what is known as piracy by law of nations, there has been conceded a universal jurisdiction, under which the person charged with the offence may be tried and punished by any nation into whose jurisdiction he may come. I say "piracy by law of nations", because the municipal laws of many States denominate and punish as "piracy" numerous acts which do not constitute piracy by law of nations, and which therefore are not of universal cognizance, so as to be punishable by all nations. Piracy by law of nations, in its jurisdictional aspects, is sui generis. Though statutes may provide for its punishment, it is an offence against the law of nations; and as the scene of the pirate's operations is the high seas, which it is not the right or duty of any nation to police, he is denied the protection of the flag which he may carry, and is treated as an outlaw, as the enemy of all mankind — hostis humani generis — whom any nation may in the interest of all capture and punish. The universal jurisdiction to try war criminals is a jurisdiction to try those alleged to have committed war crimes as defined by international law: see Simons, The Jurisdictional Bases of the International Military Tribunal at Nuremberg in Ginsburgs and Kudriavtsev (eds), The Nuremberg Trial and International Law (1990), pp. 48-49. But jurisdiction under municipal law to try a municipal law offence which is similar to but not identical with an international crime is not recognized as a jurisdiction conferred or recognized by the law of nations. Professor Green, op. cit., p. 571, comments: If a country introduces legislation describing some offence under its own criminal law as constituting, for example, piracy, and includes within that term offences which do not strictly fall within the international law definition, then that law can only be invoked to establish jurisdiction against nationals or residents of the country in question, a principle that was made crystal clear by Lord Stowell in his decision in the Le Louis [24] in 1817, and the situation is the same, even if the offence so described were to constitute an international crime under some other name or concept. Equally, if a country uses in its national criminal law a definition that only partly meets the conditions of international law, especially if the offence in question has been defined in a treaty, the courts of that country would only be entitled to try those whose actions fall within its own definition, although it might well be that the country concerned might have breached its international obligations by adopting so narrow a definition. 1. [1934] A.C. 586, at p. 589. 2. (1927) 2 W.C.R. 20, at p. 69. 3. (1817) 2 Dods. 210 [165 E.R. 1464]. However, when municipal law adopts the international law definition of a crime as the municipal law definition of the crime, the jurisdiction exercised in applying the municipal law is recognized as an appropriate means of exercising universal jurisdiction under international law. Brownlie, op. cit., p. 561, states the position thus: Since the latter half of the nineteenth century it has been generally recognized that there are acts or omissions for which international law imposes criminal responsibility on individuals and for which punishment may be imposed, either by properly empowered international tribunals or by national courts and military tribunals. These tribunals exercise an international jurisdiction by reason of the law applied and the constitution of the tribunal, or, in the case of national courts, by reason of the law applied and the nature of jurisdiction (the exercise of which is justified by international law). The jurisdiction of the courts of the United States to try cases of international crime was founded on the application by municipal courts of international law. Thus, in an early American case (United States v. Smith [25] noted in the report ofThe Magellan Pirates [26] ), the Supreme Court of the United States held that American common law "recognizes and punishes piracy as an offence, not against its own municipal code, but as an offence against the Law of Nations (which is part of the Common Law), as an offence against the universal law of society; a pirate being deemed an enemy of the human race." In that case and in Ex parte Quirin [27] , the Court spoke in terms which suggested that the courts of the United States applied international law directly as part of the municipal law of the United States. Thus in Ex parte Quirin [28] , the Court said: From the very beginning of its history this Court has recognized and applied the law of war as including that part of the law of nations which prescribes, for the conduct of war, the status, rights and duties of enemy nations as well as of enemy individuals. In In re Yamashita [29] the Court, speaking in more guarded language, observed that — We do not make the laws of war but we respect them so far as they do not conflict with the commands of Congress or the Constitution. (Cf. Art. I, s. 8, cl. 10 of the United States Constitution which confers on the Congress power "To define and punish Piracies and Felonies committed on the high Seas, and Offences against the Law of Nations".) 1. (1820) 5 Wheat. 153, at p. 161. 2. (1853) 1 Sp. Ecc. & Ad. 81, at pp. 90-91 [164 E.R. 47, at pp. 52-53]. 3. (1942) 317 U.S. 1. 4. (1942) 317 U.S., at pp. 27-28. 5. (1946) 327 U.S. 1, at p. 16. International law distinguishes between crimes as defined by it and crimes as defined by municipal law and it makes a corresponding distinction between jurisdiction to try crimes as defined by international law and jurisdiction to try crimes as defined by municipal law. Professor Lauterpacht (The Law of Nations and the Punishment of War Crimes, British Year Book of International Law, vol. XXI (1944) 58, at pp. 65-67) held the opinion that the law which must be applied in connexion with the prosecution and punishment of war criminals is primarily the law of nations and that, by applying the law of nations, municipal legislation providing for the trial of war criminals avoids the reproach of retroactivity and breach of the principle nullum crimen sine lege. He wrote: Once it is realized that the offenders are being prosecuted, in substance, for breaches of international law, then any doubts due to inadequacy of the municipal law of any given State determined to punish war crimes recede into the background. There is in this matter no question of any vindictive retroactivity arising out of the creation of crimes of which the accused could not possibly be cognizant. There is even no question of procedural retroactivity by subjecting him to a foreign jurisdiction in defiance of established law and principles. The law of Great Britain, of the United States, and of many other States, does not, as a rule, recognize the competence of national courts in respect of criminal acts committed by aliens abroad. But there would be no question of any retroactivity, contrary to justice and to established principles of law, if Great Britain were to alter her law so as to enable her tribunals, civil or military, whether functioning in Great Britain or abroad, to assume jurisdiction over German nationals who committed in Germany criminal offences against British prisoners of war or British civilians in circumstances not authorized by international law. It would thus appear that there is no novelty about the principle that a belligerent is entitled to punish such perpetrators of war crimes as fall into his hands; that that principle, far from being a mere assertion of power, grudgingly assented to by international law, on the part of the fortunate belligerent is, in turn, grounded in the fact of recognition by international law of the jurisdiction of States based on the territorial and cognate principles as well as in the fact that in punishing war criminals the belligerent applies and enforces, in essence, the rules of the law of nations which are binding upon the individual members of the armed forces of all belligerents; and that there is no question of any retroactive application of the law from any material point of view. In the Netherlands, war crimes committed by the occupying forces of Germany during the Second World War were, for a time after the War, prosecuted under municipal laws. This procedure, like the procedure contemplated by the Act, required some consideration of the relationship between international law and municipal law. A Special Court of Cassation refused to apply the municipal law of manslaughter to the trial of Ahlbrecht (1946-1947) — see Mouton, op. cit., p. 54 — an alleged war criminal who had been convicted and sentenced to death by the Special Court at Arnhem. The Special Court of Cassation was of the view that an army in occupation of foreign territory brings its own penal code, courts martial and penal procedure with it, but the members of the occupying forces may be brought to trial at the cessation of hostilities for violations of the laws and customs of war. The Court's judgment continued: however this conclusion in no way implies that as the Netherlands law now stands the Netherlands judge already possesses legal competence over enemy war criminals; for this, more is necessary, namely that the Netherlands judge can either directly apply an international agreement brought about to this effect with the co-operation of the Netherlands, or can base himself on a Netherlands law in which the legislator has set out in a concrete legal form his (the judge's) internationally recognized competence over enemy war criminals in the national sphere etc. The Court found no statement of international law in the municipal statutory provisions, as they then stood, observing that the view that the municipal statutory provisions applied to the case — proceeds from the mistaken train of thought that international law would give the Netherlands jurisdiction over all members of the enemy forces and officials who committed acts in this country which could be construed as coming under any act described and made punishable in the Netherlands Penal Code, unless they could put forward in their defence that a positive rule of international law nevertheless allowed them full liberty to commit those acts; however this version inverts the correct relationship of things and gives far too wide a scope to that jurisdiction which international law has granted to States over members of a hostile occupying force; on the contrary this jurisdiction is limited in principle to those among them who while in this country violated the laws or customs of war and who on those grounds alone can be tried by Netherlands courts; in other words, international law does not here function as a general ground of excuse for a foreign power whose actions in this country should in principle come fully under the provisions of the penal law, in force here, but on the contrary it does by way of exception give jurisdiction to that State where the hostile army operated with respect to the members of that army and its followers who violated there the laws or customs of war, and does so only for that purpose; it would be unreasonable to try foreign soldiers and officials according to Netherlands rules which were not written for them instead of trying them by those rules, written for them, which govern warfare and also taking into account the provision regarding superior orders which was accepted in article 8 of the Charter belonging to the London Agreement; indeed all those international documents quoted above take the line that alone the violation of the rules of war forms a basis for foreign jurisdiction: Mouton, op. cit., pp. 54-56 (emphasis added). Subsequently, the municipal law was amended by making punishable war crimes and crimes against humanity as defined by the Charter of the Nuremberg Tribunal (Baxter, The Municipal and International Law Basis of Jurisdiction over War Crimes, British Year Book of International Law 382, vol. XXVIII (1951), at p. 384). Whether or not it be right to say that contraventions of the laws and customs of war are the only crimes for which the members of the military forces of an enemy power in occupation of the territory of a State may be prosecuted in the courts of that State after the occupation ceases, it seems that municipal law can govern the acts done in the course and for the purposes of the occupation only if those acts fail to comply with the law of nations governing armed conflict and hostile occupation. This is the principle on which France and Norway acted in the trial of war criminals after the Second World War (Baxter, op. cit., at pp. 384-385) and there is authority to support that view in United States cases after the Civil War: Coleman v. Tennessee [30] ; Dow v. Johnson [31] . Whether contraventions of the laws and customs of war constitute the only crimes justiciable by the courts of the State whose territory has been occupied or whether such contraventions remove the immunity from municipal law to which occupying forces are entitled in respect of acts done in the course and for the purposes of hostile occupation, the prosecution must establish such a contravention in order to obtain a conviction. 1. (1878) 97 U.S. 509. 2. (1879) 100 U.S. 158. Historically, war crimes allegedly committed by the forces of a belligerent power were tried according to international law by military courts or commissions created for the purpose by the other belligerent: Mouton, op. cit., p. 39. In pursuance of the policy adopted by the Allied Powers in the Second World War to bring war criminals to justice, international military tribunals were constituted by the Allied Powers and military courts or commissions were appointed by particular nations for the purpose. The respective jurisdictions of these tribunals to try persons accused of war crimes were prescribed by the instruments which constituted them. Article 6 of the Nuremberg Charter prescribed the crimes within the jurisdiction of the Nuremberg Tribunal; similar provisions were inserted in the proclamation by the Supreme Commander for the Allied Forces in the Pacific War setting up the International Military Tribunal for the Far East: Mouton, op. cit., p. 53. Article 6(b) defined "war crimes" as follows: War crimes: namely, violations of the laws or customs of war. Such violations shall include, but not be limited to, murder, ill-treatment or deportation to slave labour or for any other purpose of civilian population of or in occupied territory, murder or ill-treatment of prisoners of war or persons on the seas, killing of hostages, plunder of public or private property, wanton destruction of cities, towns or villages, or devastation not justified by military necessity. The Nuremberg Tribunal observed, American Journal of International Law, vol. 41 (1947) 172, at p. 248: The Tribunal is of course bound by the Charter, in the definition which it gives both of War Crimes and Crimes against Humanity. With respect to War Crimes, however, as has already been pointed out, the crimes defined by Article 6, Section (b), of the Charter were already recognized as War Crimes under international law. They were covered by Articles 46, 50, 52, and 56 of the Hague Convention of 1907, and Articles 2, 3, 4, 46, and 51 of the Geneva Convention of 1929. That violation of these provisions constituted crimes for which the guilty individuals were punishable is too well settled to admit of argument. Apart from the coercive jurisdiction of international tribunals, enforcement of international law governing war crimes is a matter for the particular State into whose hands the alleged criminal falls. In In re Altstötter (The Justice Trial) [32] , the United States Military Tribunal at Nuremberg, dealing with the universality of punishment of war crimes, said: This universality and superiority of international law does not necessarily imply universality of its enforcement. As to the punishment of persons guilty of violating the laws and customs of war (war crimes in the narrow sense), it has always been recognized that tribunals may be established and punishment imposed by the State into whose hands the perpetrators fall. Those rules of international law were recognized as paramount, and jurisdiction to enforce them by the injured belligerent government, whether within the territorial boundaries of the State or in occupied territory, has been unquestioned. (Ex parte Quirin [33] ; In re Yamashita [34] .) 1. (1947) 14 Annual Digest 278, at pp. 282-283. 2. (1942) 317 U.S. 1. 3. (1946) 327 U.S. 1. In Australia, the original War Crimes Act (No. 48 of 1945) provided for the convening of military courts with "power to try persons charged with war crimes committed, at any place whatsoever, against any person who was at any time resident in Australia " (see s. 7), "war crimes" being so defined as to import the law of nations: see s. 3 and the instrument of appointment therein mentioned. Act No. 48 of 1945 did not exhaust the jurisdiction over war crimes which international law recognized Australia to possess, for the jurisdiction created by s. 7 was limited to war crimes against Australian residents. Although the jurisdiction vested in military courts by that Act was jurisdiction to administer international law, upon amendment of the War Crimes Act the jurisdiction vested in the competent courts of the several States and Territories (s. 13(1)) was jurisdiction to administer not international law but, as we have seen, a purely municipal law having a retrospective operation. There is no reason why jurisdiction to try persons alleged to be guilty of war crimes should be conferred only on military courts or commissions. I would respectfully agree with what Lord Simon L.C. said in the House of Lords on 7 October 1942 (quoted in War Crimes: Report of the War Crimes Inquiry (U.K.) — the Hetherington Report — Cm. 744, (1989), pars 6. 41, 6. 42): "I take it to be perfectly well-established International Law that the laws of war permit a belligerent commander to punish by means of his Military Courts any hostile offender against the laws and customs of war who may fall into his hands wherever be the place where the crime was committed National courts, in my view, are equally entitled to exercise whatever criminal jurisdiction would be conceded to them by International law The real question is not so much whether the domestic law of a particular nation has already conferred upon the particular national Courts concerned a particular jurisdiction. It may not have gone to the full length which International Law would recognise and permit. The important question is this: what is the ambit of the jurisdiction which might by International Law be conferred upon them, as for example, in the present case, by Parliament here actually legislating to enlarge, within permissible limits, the jurisdiction of our Courts to deal with crimes committed abroad?" (Official Report (Hansard): cols 578-579.) The Hetherington Report observes (par. 6.42): Legal opinion at the time seems to have been that jurisdiction over violations of the laws and customs of war existed, and that there was a need to legislate only to empower the domestic courts to utilise the jurisdiction which was already available under international law. It follows that it is no objection to the validity of the Act that it selects the ordinary courts of the States and Territories as the tribunals for the trial of persons charged with "a war crime". It is one thing to vest in a municipal court jurisdiction to administer the law of nations, albeit that that law is adopted by the municipal law. It is another thing to vest jurisdiction to administer municipal law that does not correspond with international law. The real objection to the validity of the Act is that the Act rejects international law as the governing law for the trial of persons allegedly guilty of war crimes and adopts a municipal law definition which operates retrospectively. That retrospectivity denies to the Act the capacity to satisfy an international obligation or to meet an international concern or to confer a universal jurisdiction recognized by international law. International law does not create an international crime retrospectively. The judgment of the International Military Tribunal at Nuremberg was concerned to repel the suggestion that its jurisdiction extended to the imposition of criminal penalties for acts that were not forbidden by international law when they were done. Subsequently, the question of retrospectivity arose in respect of Control Council Law No. 10 by which the Allied Control Council conferred jurisdiction upon Military Tribunals sitting in the respective occupation zones of Germany after the Second World War to try persons accused of crimes "recognized" by Law No. 10. An objection that the crimes so recognized (stated more broadly in the case of crimes against humanity than in the Nuremberg Charter) were retrospective was rejected by the United States Military Tribunal at Nuremberg (In re List (Hostages Trial)). The correspondence between the offences defined by Law No. 10 and crimes under international law may be doubted but the Military Tribunal asserted that there was correspondence in order to answer the argument that Law No. 10 was retrospective. The judgment contained the following passage [35] : (2) Control Council Law No. 10 and the Principle of " Nullum Crimen Sine Lege." — It is urged that Control Council Law No. 10 is an ex post facto act and retroactive in nature as to the crime charged in the indictment. The act was adopted on 20th December, 1945, a date subsequent to the dates of the acts charged to be crimes. It is a fundamental principle of criminal jurisprudence that one may not be charged with crime for the doing of an act which was not a crime at the time of its commission. We think it could be said with justification that Article 23 (h) of the Hague Regulations of 1907 operates as a bar to retroactive action in criminal matters. In any event, we are of the opinion that a victorious nation may not lawfully enact legislation defining a new crime and make it effective as to acts previously occurring which were not at the time unlawful. It therefore becomes the duty of a Tribunal trying a case charging a crime under the provisions of Control Council Law No. 10, to determine if the acts charged were crimes at the time of their commission and that Control Council Law No. 10 is in fact declaratory of then existing International Law. This very question was passed upon by the International Military Tribunal in the case of the United States v. Hermann Wilhelm Goering in its judgment entered on 1st October, 1946 [13 Annual Digest 203]. Similar provisions appearing in the Charter creating the International Military Tribunal and defining the crimes over which it had jurisdiction were held to be devoid of retroactive features in the following language: "The Charter is not an arbitrary exercise of power on the part of the victorious nations, but in [the] view of the Tribunal, as will be shown, it is the expression of International Law existing at the time of its creation; and to that extent is itself a contribution to International Law". The crimes defined in Control Council Law No. 10 which we have quoted herein, were crimes under pre-existing rules of International Law — some by conventional law and some by customary law. If the acts charged were in fact crimes under International Law when committed, they cannot be said to be ex post facto acts or retroactive pronouncements. It is true, of course, that customary International Law is not static. It must be elastic enough to meet the new conditions that natural progress brings to the world. It might be argued that this requires a certain amount of retroactive application of new rules and that, by conceding the existence of a customary International Law, one thereby concedes the legality of retroactive pronouncements. To a limited extent the argument is sound, but when it comes in conflict with a rule of fundamental right and justice, the latter must prevail. The rule that one may not be charged with crime for committing an act which was not a crime at the time of its commission is such a right. The fact that it might be found in a constitution or bill of rights does not detract from its status as a fundamental principle of justice. It cannot properly be changed by retroactive action to the prejudice of one charged with a violation of the laws of war. 1. (1948) 15 Annual Digest, at pp. 634-636. Whether or not all the offences recognized by Law No. 10 were at all material times offences against the law of nations, the principle against retrospectivity is rightly stated. Some texts have suggested that, in international law, there is no rule forbidding the enactment or adoption of criminal laws with retrospective force: see the discussion by the Supreme Court of Israel in Attorney-General (Israel) v. Eichmann [36] , the Report of the Canadian Commission of Inquiry on War Criminals (Deschênes Commission) (1986), pp. 137-138, and the writings therein mentioned. The better view now, if not in earlier times, is that the rule of customary international law is as expressed in Art. 15 of the International Covenant on Civil and Political Rights, (1966): 1. No one shall be held guilty of any criminal offence on account of any act or omission which did not constitute a criminal offence, under national or international law, at the time when it was committed 2. Nothing in this article shall prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of law recognized by the community of nations. Article 99 of the 1949 Geneva Convention (III) Relative to the Treatment of Prisoners of War contains a similar provision: No prisoner of war may be tried or sentenced for an act which is not forbidden by the law of the Detaining Power or by International Law, in force at the time the said act was committed. Oppenheim, International Law, 7th ed. (1952), vol. II, p. 390, comments: The reference to International Law, which did not appear in the Convention of 1929, is intended to remove any doubts as to the right of the detaining belligerent to try prisoners of war for war crimes. 1. (1962) 36 I.L.R. 277, at pp. 281-283. Thus international law not only refuses to countenance retrospective provisions in international criminal law; it condemns as offensive to human rights retrospective municipal criminal law imposing a punishment for crime unless the crime was a crime under international law at the time when the relevant act was done. It follows that there can be no international obligation to enact a municipal law to attach a penalty to past conduct unless that conduct, at the time when it was engaged in, was a crime under international law. International concerns must be qualified in like manner. For present purposes, the relevant time is between 1 September 1939 and 8 May 1945: s. 9(1)(a) of the Act. The jurisdiction of a national court to try a person for a war crime will be recognized as a universal jurisdiction in international law if the crime as defined by a post-war municipal law was an international crime at the time of its commission, but not otherwise. In ascertaining whether the jurisdiction of a national court to try alleged war criminals from the Second World War is a universal jurisdiction recognized by international law, the correct approach is that taken by the Hetherington Report in pars 6. 36 and 6. 37: 6.36. The principles nullem crimen sine lege and nulla poena sine lege are basic tenets of law, and are themselves included in the laws and constitutions of many countries. No state may enact legislation to deem an act a crime or render it punishable when it was not considered to be a criminal offence under the law at the time of its commission, and thus give its courts jurisdiction in respect of those acts. These principles are also enshrined in international conventions and declarations including the International Covenant on Civil and Political Rights 1966 (Article 15); the European Convention on Human Rights 1950 (Article 7); and the Universal Declaration of Human Rights 1948 (Article 11). 6.37. To apply these principles to war crimes as defined by this Inquiry's terms of reference it is necessary to consider whether the acts or omissions were crimes at the time of their commission, and, if so, whether there was jurisdiction over them at the time of their commission and whether the British criminal courts could now have jurisdiction over them. Therefore, the question is whether the statutory offence created by s. 9 of the Act corresponds with the international law definition of international crimes existing at the relevant time. If it does, the Act vests jurisdiction to try alleged war criminals for crimes which were crimes under the applicable (international) law when they were committed; its apparent retrospectivity in municipal law is no bar to the exercise of a universal jurisdiction recognized by international law and that is sufficient to enliven the external affairs power to support the Act which vests that jurisdiction. Even if there be no international obligation or concern calling for the exercise of the universal jurisdiction, a statutory vesting of the jurisdiction would be essential to its exercise by an Australian court and that would suffice to give the support of s. 51(xxix) to the law. But if the statutory offence created by s. 9 does not correspond with the international law definition of international crimes existing before 8 May 1945, the retrospective creation by Australian municipal law of the crime defined by the Act is offensive to international law. In that event, the Act cannot be seen to satisfy an international obligation, to meet an international concern or (subject to a further submission yet to be considered) to be appropriate and adapted to the vesting of a universal jurisdiction. This view is consistent with the view of Professor Baxter who, after the Supreme Court of Israel had affirmed Israel's jurisdiction to try and to condemn Eichmann for war crimes and other violations of international law committed before Israel came into existence, added a postscript to his article on "The Municipal and International Law Basis of Jurisdiction over War Crimes", op. cit. The postscript appears in Bassiouni and Nanda (eds), A Treatise on International Criminal Law (1973), vol. 2, p. 65. The author wrote (at p. 83): There could be no objection under international law to Israeli law's reaching out to "a person" of whatsoever nationality to the extent that the municipal law of that country merely incorporated in its law crimes under international law subject to universal jurisdiction. It is thus necessary to examine the consistency of the crimes defined by the law of Israel with those crimes recognized by international law. Baxter's conclusion was that Israeli municipal law defined war crimes restrictively and not more broadly than the international law definition, but he stopped short of affirming that Israel's definition of crimes other than war crimes conformed to international law. At this stage, the differing operation of s. 7(1) and s. 7(3) should be noted. Section 7(1) relates to "serious crimes" which, having been committed in connexion with a war or armed conflict, might have amounted to a war crime in international law; s. 7(3), on the other hand, relates to "serious crimes" committed in a country while it was at war or under hostile occupation, but it is immaterial whether the victims were of the same nationality as the offender or not. Serious crimes falling within s. 7(3) might have amounted to what has become known as crimes against humanity. Before considering crimes against humanity, we should inquire whether s. 9 of the Act, by the operation of ss. 6, 7(1) and 17, creates an offence which corresponds with a war crime as defined by the law of nations. "A war crime" in s. 9 and a war crime in the law of nations Section 6(3) of the Act requires a court, in determining whether an act committed outside Australia is "a serious crime" to "imagine another act committed [in a part of Australia] which is similar in all relevant respects" (per Lord Reid in Cox v. Army Council [37] ). Although Lord Reid went on to say that murder and theft "are the same all the world over", there are conceptual difficulties in translating a killing in the course and for the purposes of military activity in Europe (including hostile military occupation of a European State) into a part of Australia. Presumably, in the notional translation directed by s. 6(3)(b), the victim must be treated as having been within "the King's peace" (cf. Reg. v. Page [38] ) so that that element of culpable homicide can be established. But what is the position of an accused who, as a member of military forces, killed "in the course of hostilities in a war" or in any of the other circumstances referred to in s. 7(1)? If it be right to say that the members of an enemy force in occupation of a State are immune from the municipal law of the State provided they do not act in violation of the law and customs of war, an accused looking at ss. 6, 7(1) and 9 of the Act might have argued that his act would not have amounted to one of the offences under municipal law mentioned in s. 6(1) unless it were proved to be an act done in violation of the laws and customs of war. Had that been the effect of the Act, the law of nations would have governed his liability to conviction as for an offence under s. 9. Sections 6 and 7 would have operated as restrictions on liability to conviction for war crimes as defined in international law. But s. 17 precludes that construction. It is enacted as a critical part of a code which determines the liability to conviction of a person who is proved to have committed an act falling within ss. 6 and 7. One cannot postulate, for the purposes of s. 6, that the members of an enemy force are immune from punishment for the municipal law offences mentioned in s. 6(1) except where the act charged violated the laws and customs of war, for that would make s. 17(2) otiose. Section 6 must bring deliberate and wilful killing by a member of an enemy force into the category of "serious crime" unless the usual defences — self-defence, mistake, etc. — are open on the facts of the particular case, leaving s. 17(2) to take out of "a war crime" any killing which falls within its terms. The Act thus displaces the essential character of a war crime under the law of nations as a violation of the laws and customs of war and seeks to make the laws, customs and usages of war a source of excuse. Section 17(2) is misconceived, as Baxter (op. cit., p. 388) explains: If, in those states in which war crimes are tried under municipal law, the function of international law is to furnish a justification for acts of warfare which are thereby recognized to be lawful, the law of war loses its reasonableness. The international law of war is "prohibitive law" and its purpose is to place curbs upon the otherwise unrestrained violence of war. Belligerent acts in war are facts, not legal rights, and to set the law to justifying them, instead of keeping them within limits which comport with the dictates of humanity, leads to a law which places its emphasis on the rightness of war to the detriment of what is wrong in war. Any theory which relies on the law of nations as a defence for belligerent acts thus fails to accord with the true raison d'être of the law of war and with the many expressions of the intentions of those who have contributed to its development in recent years. Sub-section (2) does not stand alone as a defence. Sub-section (3) provides a dictionary for sub-s. (2) and calls for a factual evaluation of what was "reasonably justified by the exigencies and necessities of the conduct of war". It may be that the exigencies and necessities of the conduct of war could be construed to embrace any military objective which is not unlawful by the laws and customs of war (see Bindschedler-Robert, "Problems of the Law of Armed Conflicts" in Bassiouni and Nanda (eds), op. cit., vol. 1, p. 295, at pp. 304-306), a construction that would be repetitive of the concepts in sub-s. (2)(a). However that may be, the requirement that the act charged be "reasonably justified" leaves the way open for a jury to make an evaluation of the facts adverse to an accused although the legal restrictions on military action prescribed by the laws and customs of war have not been violated. By the combined operation of sub-ss. (2) and (3) of s. 17, the laws and customs of war are excluded as the legal criterion of criminal liability capable of application by direction of the trial judge and are replaced by the factual criterion of reasonable justification which a jury, probably inexperienced in the exigencies and necessities of the conduct of war, must apply as best it can. 1. [1963] A.C. 48, at p. 70. 2. [1954] 1 Q.B. 170. The disconformity between the statutory offence purportedly created by s. 9 of the Act and a war crime in international law can be illustrated by taking, as an example, the facts of the case in Osman Bin Haji Mohamed Ali v. Public Prosecutor [39] where Indonesian military personnel penetrated into enemy territory (Singapore) and, in doing sabotage, caused death. Had they been acting under legitimate orders and had they been wearing uniform, they may have been able to plead that their act was immune from municipal law as an act of State of a belligerent Power (Wright, War Criminals, American Journal of International Law, vol. 39 (1945) 257, at pp. 272-273) but, as they wore no uniform and concealed their identity in order to sabotage their objective, they were held to have forfeited the protection of a prisoner of war under the Geneva Convention (whether of 1929 or 1949) and were liable to conviction for murder under municipal law. As the laws and customs of war do not "permit" such an act, such an act would amount to "a war crime" under s. 9. Yet such an act is no contravention of the laws and customs of war, at least where the objective of the sabotage is legitimate. Oppenheim, International Law, 7th ed., vol. II, par. 255, deals with cases of this kind which are sometimes designated as "war treason". That designation includes cases of "[w]recking of military trains, destruction of the lines of communication or of telegraphs or telephones in the interest of the enemy, and destruction of any war material for the same purpose". The author writes, at pp. 574-576: Espionage and so-called war treason bear a twofold character. International Law gives a right to belligerents to use them. On the other hand, it gives a right to belligerents to consider them, when committed by enemy soldiers or enemy private individuals within their lines, as acts of illegitimate warfare, and consequently liable to punishment — though it seems improper to characterise such acts as war crimes. So-called "war treason" consists of all such acts committed within the lines of a belligerent as are harmful to him and are intended to favour the enemy. It may be committed, not only in occupied enemy country, or in the zone of military operations, but anywhere within the lines of a belligerent. Enemy soldiers — in contradistinction to private enemy individuals — may only be punished for such acts when they have committed them during their stay within a belligerent's lines under disguise. If, for instance, two soldiers in uniform are sent to the rear of the enemy to destroy a bridge, they may not, when caught, be punished for "war treason," because their act was one of legitimate warfare. But if they exchange their uniforms for plain clothes, and thereby appear to be members of the peaceful private population, they are liable to punishment. (Emphasis added.) By applying Australian municipal law, s. 9 would hold the out-of-uniform saboteur of a legitimate objective behind enemy lines in Europe guilty of a war crime consisting in the "serious crime" of murder if the sabotage was likely to cause death and did cause death — unless the accused can point to or adduce evidence to show that his act "was reasonably justified by the exigencies and necessities of the conduct of war". 1. [1969] 1 A.C. 430. Even if "permitted by" in s. 17(2)(a) were construed as meaning "not in violation of" the laws, customs and usages of war, s. 17 would not bring the offence created by s. 9 into correspondence with international law. Section 17 provides a defence to what has been proved under ss. 6 and 7. The issue of violation of the laws and customs of war would not arise unless the accused could point to or adduce "evidence of the existence of the facts constituting the defence" — and to do so in a trial more than forty years after the event. Instead of adopting the international law definition of a war crime, the Act has created an offence which holds members of an enemy force to be subject to municipal law in respect of acts done "in the course of hostilities in a war" and, instead of restricting criminal liability to acts done in violation of the laws and customs of war, holds an accused liable to conviction unless he can point to or adduce evidence which might lead a jury to acquit on the ground that the act "was reasonably justified by the exigencies and necessities of the conduct of war." Thus the offence which ss. 6, 7(1), 9 and 17 create is in disconformity with the offence of a war crime as defined by the law of nations. It is arguable that s. 16 (to which s. 17(2) is subject) creates a further disconformity between the municipal law and the international law of war crimes or, at least, the international law of war crimes as it existed during the period 1 September 1939 to 8 May 1945. Subject to s. 6(2), s. 16 excludes a defence of superior orders. It prescribes a rule which is the converse of Art. 443 of the Australian Manual of Military Law, as it stood at 1 September 1939. After setting out the more important violations of the laws of war, Art. 443 read: It is important, however, to note that members of the armed forces who commit such violations of the recognized rules of warfare as are ordered by their Government, or by their commander, are not war criminals and cannot therefore be punished by the enemy. He may punish the officials or commanders responsible for such orders if they fall into his hands, but otherwise he may only resort to the other means of obtaining redress which are dealt with in this chapter. This proposition, which had been copied from the British Manual of Military Law, 7th ed. (1929), was in substantial accord with the Rules of Land Warfare (1940) approved by the General Staff of the United States Army and then in force. Article 347 of those Rules read as follows: Individuals of the armed forces will not be punished for these offences in case they are committed under the orders or sanction of their government or commanders. The commanders ordering the commission of such acts, or under whose authority they are committed by their troops, may be punished by the belligerent into whose hands they may fall. The concurrent views of the Australian, British and United States military forces as at 1 September 1939 embraced the doctrine of respondeat superior advanced by Professor Oppenheim in his work International Law, 5th ed. (1935), vol. II, pp. 453-454: Violations of rules regarding warfare are war crimes only when committed without an order of the belligerent Government concerned. If members of the armed forces commit violations by order of their Government, they are not war criminals, and may not be punished by the enemy; the latter may, however, resort to reprisals. In case members of forces commit violations ordered by their commanders, the members may not be punished, for the commanders are alone responsible, and the latter may, therefore, be punished as war criminals on their capture by the enemy. The doctrine of respondeat superior had been applied in the case of Commander Karl Neumann (The "Dover Castle") American Journal of International Law, vol. 16 (1921) 704 by the Second Criminal Senate of the Imperial Court of Justice of Germany (the Reichsgericht), but that Court's view of the doctrine appears more clearly from its decision in the case of Lieutenants Dithmar and Boldt (The "Llandovery Castle") American Journal of International Law, vol. 16 (1921) 708. There it was held (at 722) that a defence of superior orders was not available "if [the] order is universally known to everybody, including also the accused, to be without any doubt whatever against the law". The correctness of an unqualified doctrine of respondeat superior was questioned by Lord Cave in an address to the Grotius Society in 1922: Grotius Transactions, vol. 8 (1922), p.xix. Lord Cave thought that the true scope of the doctrine "limits the impunity of the soldier to cases where the orders are not so manifestly illegal that he must or ought to have known they were unlawful": at p.xxiii. In 1942, Professor Lauterpacht rejected the unqualified defence of superior orders in a memorandum to a committee of the International Commission for Penal Reconstruction and Development, which was approved by the Commission and by the participating governments: see Dinstein, The Defence of Obedience to Superior Orders in International Law (1965), p. 104. Professor Lauterpacht said: "There ought to be no doubt that should courts entrusted with the trial of war crimes disregard altogether the plea of superior orders, they would be adopting a course which could not be regarded as defensible. On the other hand, while the fact of superior orders sets a limit to the punishment of acts which might otherwise constitute war crimes, it need not warp the effectiveness of the law in a manner which may rightly be regarded as a perversion of justice. It will not cover crimes committed by superior authorities and officers acting under their own responsibility and initiative; it will not protect criminal acts committed by subordinates for purposes of private gain and lust; it will not shield acts committed in pursuance of orders so glaringly offending against fundamental conceptions of law and humanity as to remove them from the orbit of any possible justification, including that of immediate danger to the person charged with the execution of the orders; it will not excuse crimes committed in obedience to unlawful orders in circumstances in which the person executing the crime was not acting under the immediate impact of fear of drastic consequences of summary martial justice following upon a refusal to act (the latter being crimes perpetrated by the vast army of officials in the occupied territories). If these limits of the doctrine of superior orders are taken into consideration, then its judicious application, far from defeating the ends of justice, may testify in a significant manner to the determination of the victorious belligerent to abide by the limitations of international law": British Year Book of International Law, vol. XXI (1944), pp. 73-74. Professor Lauterpacht edited the 6th edition of Oppenheim's International Law and rejected the earlier views of Professor Oppenheim: see vol. II, par 253. Following that edition, Art. 443 of the Australian Manual of Military Law was amended on 30 September 1944 to read: The fact that a rule of warfare has been violated in pursuance of an order of the belligerent Government or of an individual belligerent commander does not deprive the act in question of its character as a war crime; neither does it, in principle, confer upon the perpetrator immunity from punishment by the injured belligerent. Undoubtedly, a court confronted with the plea of superior orders adduced in justification of a war crime is bound to take into consideration the fact that obedience to military orders, not obviously unlawful, is the duty of every member of the armed forces and that the latter cannot, in conditions of war discipline, be expected to weight scrupulously the legal merits of the order received. The question, however, is governed by the major principle that members of the armed forces are bound to obey lawful orders only and that they cannot therefore escape liability if, in obedience to a command, they commit acts which both violate unchallenged rules of warfare and outrage the general sentiment of humanity. By 8 August 1945, when the Nuremberg Charter was agreed, the Oppenheim position had been abandoned. In its place, a doctrine of absolute liability was erected (see Dinstein, p. 117). Article 8 of the Charter provided that: The fact that the Defendant acted pursuant to order of his Government or of a superior shall not free him from responsibility, but may be considered in mitigation of punishment if the Tribunal determines that justice so requires. Article 6 of the Charter of the International Military Tribunal for the Far East (the Tokyo Tribunal), however, provided that a superior order should not "of itself" be a defence: see Whiteman, Digest of International Law, vol. 11, p. 972. Dinstein (op. cit., p. 157) comments that "the fact of obedience to orders was not prevented from contributing, in conjunction with other facts, to discharge from responsibility and was rejected only as a defence per se." If that was the state of international law during the prescribed period, there may be little disconformity between the provisions of the Act (ss. 16, 6(2) and 6(6)) and the relevant international law, despite the contrary provisions in Art. 443 of the Australian Manual of Military Law. The United States Military Tribunal at Nuremberg in the Hostages Trial adopted the formulation of the Nuremberg Charter, observing with reference to the Oppenheim doctrine (at p. 650): The fact that the British and American armies may have adopted it for the regulation of their own armies as a matter of policy, does not have the effect of enthroning it as a rule of International Law. However that may be, the significance in international law of obedience to superior orders remains a matter of "some incertitude" (Dinstein, p. 253) if not of "universal uncertainty" (Vogler, "The Defense of "Superior Orders" in International Criminal Law" in Bassiouni and Nanda (eds), op. cit., vol. 1, p. 619, at p. 634). Both of these jurists would reject the doctrine of absolute liability. Dinstein prefers a defence of lack of mens rea to which the fact of obedience to orders would be relevant. Vogler perceives international law to be moving to the position that the subordinate is criminally responsible "only if he had in reality recognized the criminal nature of the order, or if the fact of its criminality was obvious." In my respectful opinion, there is much to commend these views. Having regard to the uncertain state of international law during the prescribed period, one cannot be confident of the coincidence of the law prescribed by s. 16 and international law. However, there remains a disconformity between international law and the municipal law contained in ss. 6, 7(1), 9 and 17 of the Act. "A war crime" in s. 9 and a crime against humanity Section 7(3) brings within the statutory definition of "a war crime" conduct which, though heinous, is not necessarily in violation of the laws and customs of war. It is said that the crimes falling within s. 7(3) are crimes against humanity and that such crimes are, like war crimes, international. In the plaintiff's case, it is said that he was guilty of a particular species of crimes against humanity, namely, crimes of genocide. In the Nuremberg Charter, Art. 6(c) prescribed the crimes which were to be treated as crimes against humanity: Crimes against humanity: namely, murder, extermination, enslavement, deportation, and other inhumane acts committed against any civilian population, before or during the war, or persecutions on political, racial or religious grounds in execution of or in connection with any crime within the jurisdiction of the Tribunal, whether or not in violation of the domestic law of the country where perpetrated. (Emphasis added.) The words italicized limited the scope of the crimes falling within Art. 6(c) and, in effect, restricted them to crimes against humanity which were also war crimes or, pursuant to Art. 6(a), crimes against peace: see Schwelb, Crimes Against Humanity, British Year Book of International Law, vol. XXIII (1946), 178, at pp. 188, 194-195, 206. If war crimes and crimes against peace were, at all material times, crimes against the law of nations, it was unnecessary to consider whether crimes against humanity which were not connected with war crimes or with crimes against peace were crimes against the law of nations. Construing its Charter, the Nuremberg Tribunal said, at p. 249: With regard to Crimes against Humanity there is no doubt whatever that political opponents were murdered in Germany before the war, and that many of them were kept in concentration camps in circumstances of great horror and cruelty. The policy of terror was certainly carried out on a vast scale, and in many cases was organized and systematic. The policy of persecution, repression, and murder of civilians in Germany before the war of 1939, who were likely to be hostile to the Government, was most ruthlessly carried out. The persecution of Jews during the same period is established beyond all doubt. To constitute Crimes against Humanity, the acts relied on before the outbreak of war must have been in execution of, or in connexion with, any crime within the jurisdiction of the Tribunal. The Tribunal is of the opinion that revolting and horrible as many of these crimes were, it has not been satisfactorily proved that they were done in execution of, or in connexion with, any such crime. The Tribunal therefore cannot make a general declaration that the acts before 1939 were Crimes against Humanity within the meaning of the Charter, but from the beginning of the war in 1939 War Crimes were committed on a vast scale, which were also Crimes against Humanity; and insofar as the inhumane acts charged in the Indictment, and committed after the beginning of the war, did not constitute War Crimes, they were all committed in execution of, or in connexion with, the aggressive war, and therefore constituted Crimes against Humanity. In Reg. v. Finta [40] , the Ontario High Court rejected an argument that crimes against humanity were not recognized as international crimes before 1945. In doing so, Callaghan A.C.J.H.C. said [41] : I accept the reasoning of the Nuremberg Tribunal who indicated, when faced with the same sorts of arguments as above, that "by 1939 these rules laid down in the Convention were recognized by all civilized nations, and were regarded as being declaratory of the laws and customs of war". In other words, in light of the historical documents I have referred to and the words of the Nuremberg Tribunal, I am of the opinion that war crimes and crimes against humanity were, by 1939, offences at international law or criminal according to the general principles of law recognized by the community of nations. With great respect, the Nuremberg Tribunal used the words cited by his Lordship in reference only to "the laws and customs of war which are referred to in Article 6(b) of the Charter " (see American Journal of International Law, vol. 41, at pp. 248-249), not in reference to crimes against humanity. The words cited by his Lordship did not relate to crimes against humanity referred to in Art. 6(c) of the Charter; those crimes were dealt with by a subsequent passage in the judgment (at p. 249) which is the passage cited above. Schwelb points out (op. cit., p. 205) that, in the case of Streicher, the Tribunal found him guilty of crimes against humanity committed before 1 September 1939 in Germany against German nationals but the Tribunal pointed to a nexus between these activities and crimes committed on occupied Allied territory and against non-German nationals. Schwelb notes that this is "the most that can be said" and that — It cannot be said in the case of any of the defendants that he was convicted only of crimes committed in Germany against Germans before 1 September 1939. 1. (1989) 61 D.L.R. (4th) 85. 2. (1989) 61 D.L.R. (4th), at p. 101. Control Council Law No. 10, enacted by the military powers in occupation of Germany after the Second World War, provided for trial within the respective Zones of Occupation of persons charged with crimes "recognized" in Art. II. Among those crimes were "Crimes against Humanity" expressed more expansively than in the Nuremberg Charter. The breadth of that definition and the uncertainty of its content are manifest from the consideration given to it by the United States Military Tribunal in In re Altstötter (The Justice Trial) [42] : We hold that crimes against humanity as defined in C.C. Law 10 must be strictly construed to exclude isolated cases of atrocities or persecutions whether committed by private individuals or by a governmental authority. As we construe it, that section provides for the punishment of crimes committed against German nationals only where there is proof of conscious participation in systematic governmentally organised or approved procedures, amounting to atrocities and offences of that kind specified in the act and committed against populations or amounting to persecutions on political, racial, or religious grounds. Thus the statute is limited by the construction of the type of criminal activity which prior to 1939 was, and still is, a matter of international concern. Whether or not such atrocities constituted technical violations of laws and customs of war, they were acts of such scope and malevolence, and they so clearly imperilled the peace of the world, that they must be deemed to have become violations of international law. As the prime illustration of a crime against humanity under C.C. Law 10, which by reason of its magnitude and its international repercussions has been recognized as a violation of common international law, we cite "genocide". (Emphasis added.) The opinion of this Tribunal has not been accepted as an authoritative statement of customary international law. Law No. 10 and the tribunals which administered it were not international in the sense that the Nuremberg Charter and the International Military Tribunal were international. As Dr Schwelb observes (op. cit., p. 218): the difference between the Charter and Law No. 10 probably reflects the difference both in the constitutional nature of the two documents and in the standing of the tribunals called upon to administer the law. As we have attempted to show, the International Military Tribunal is, in addition to being an occupation court for Germany, also — to a certain extent — an international judicial organ administering international law, and therefore its jurisdiction in domestic matters of Germany is cautiously circumscribed. The Allied and German courts, applying Law No. 10, are local courts, administering primarily local (municipal) law, which, of course, includes provisions emanating from the occupation authorities. (Emphasis added.) 1. (1947) Annual Digest, at pp. 284-285. The crime of genocide, which the Tribunal described as the "prime illustration" of a crime against humanity attracting punishment, did not acquire the status of an international crime until after the Second World War: see Kunz, Editorial Comment: "The United Nations Convention on Genocide", American Journal of International Law, vol. 43 (1949) 738, at p. 742; Green, Canadian Law, War Crimes and Crimes Against Humanity, British Year Book of International Law, vol. LIX (1988) 217, at pp. 225-226; Shaw, Genocide and International Law in Dinstein (ed.), International Law at a Time of Perplexity (1989), p. 797. Genocide was not listed as a crime in Art. 6(b) of the Nuremberg Charter but, on 11 December 1946, the General Assembly of the United Nations by Resolution 96(I) affirmed that genocide is a crime under international law and initiated studies leading to the adoption of the Genocide Convention of 1948. There is some disagreement as to whether, by the time when the Genocide Convention was approved by the General Assembly, genocide had become a crime under customary international law. Shaw (op. cit., at p. 799) expresses the view that, after the General Assembly adopted Resolution 96(I) — Hand in hand with the speedy enshrinement of genocide as a crime under conventional law, has gone a process of parallel acceptance under customary international law. Kunz (at p. 742) and Green (at p. 225), on the other hand, take the view that the Convention created and defined the crime of genocide for the law of nations. The weight of opinion and international practice as evidenced by the drafting of Art. 6(c) of the Nuremberg Charter show that genocide was not a crime under international law until after the Second World War, and the history of the drafting of Art. 6(c) shows that crimes against humanity were not clearly established as crimes in international law independent of war crimes when the Nuremberg Charter was drafted: see Clark, Crimes Against Humanity in The Nuremberg Trial and International Law, pp. 177-199; but cf. Simons, at p. 49. Thus I come to the same conclusion as that expressed in the Hetherington Report whose summary I would adopt (par. 6.44) mutatis mutandis as applicable to Australia: To summarize, by 1939, before the offences which this Inquiry is required to investigate were allegedly committed, violations of the customs and uses of war, or war crimes as they were later called, were internationally recognized as crimes, both Britain and Germany being among the signatories of the Hague Conventions which confirmed them as such. The Nuremberg judgment also held that such acts were also recognized as crimes under customary international law, which bound even those nations which had not become party to the Conventions. Genocide was not so recognized until 1948 and we find the position of what were subsequently called crimes against humanity to be unclear. Under customary law belligerents had the right to try before military courts war criminals who fell into their hands, and also to provide for the surrender of others in the terms of the armistice or the peace treaty. Legal opinion then held that jurisdiction existed over such crimes and that a state had the right to legislate to incorporate that jurisdiction into its national domestic law. Therefore it can be argued that enactment of legislation in this country to allow the prosecution of "war crimes" in British courts would not be retrospective: it would merely empower British courts to utilise a jurisdiction already available to them under international law since before 1939, over crimes which had been internationally recognized as such since before 1939 by nations including both the United Kingdom and Germany. We are less certain that a similar stance can be adopted with regard to crimes against humanity. To legislate now for offences of genocide committed during the Second World War would, in our view, constitute retrospective legislation. (Emphasis added.) The crime created by s. 9 of the Act would thus expose to conviction persons whose act, if it fell only within sub-s. (3) of s. 7, would not (genocide) or might not (some other crime against humanity) have amounted to a crime under international law at the time when the act was done and did not then amount to a crime under municipal law. In the light of the objection of international law to retrospective criminal legislation unless the crime was, at the time of its commission, a crime under international law, there could be no international obligation which would have been satisfied nor any international concern which would have been met by creating retrospectively a crime one of the elements of which is furnished by s. 7(3). So far as s. 7(3) furnishes an element of the statutory offence created by s. 9, it is not consonant with pre-1945 international law. A jurisdiction to try a person for such an offence is not a universal jurisdiction to try a person for an international crime. The disconformity between the provisions of the Act defining a war crime falling within sub-s. (1) of s. 7 and international law equally denies that that provision, so far as it furnishes an element of the s. 9 offence, satisfies an international obligation or meets an international concern. The international obligation or concern, if any, can relate only to the seeking out and prosecution of persons alleged to be guilty of war crimes as recognized by international law. The prosecution of crimes which are the creation of a retrospective municipal law and which differ from international crimes cannot be the subject of a universal jurisdiction to try. As s. 8(3) provides that a "serious crime may be a war crime by virtue of either or both of subsections 7(1) and (3), but not otherwise", s. 9 cannot be supported as a provision which satisfies an international obligation or meets an international concern. Nor can it be supported as a law investing a universal jurisdiction to try a person for an international crime. If s. 9 be invalid, the Act must fall. To meet an argument that the statutory offence does not conform to the international law definition of international crimes (as international law stood during the prescribed period), the proposition that the Act invested a jurisdiction to try offences defined in international law was broadened in the course of the Commonwealth's submissions. It was submitted that there is a universal jurisdiction under international law to try allegations of guilt of crimes defined by municipal law if those definitions approximated the definitions of crimes in international law. The authority cited in support of the broadened submission was Brierly, op. cit., which must be set out in extenso. Asserting that jurisdiction over war crimes "in the strict sense" is "a special jurisdiction created by the international law of war" (p. 297), the author says (pp. 300-301): It is of course to be regretted that the laws of war do not define more precisely either the acts that it is permissible to treat as war crimes, or the procedure by which they ought to be dealt with. But lack of precision, and consequently scope for differing interpretations of the rule of law, are inevitable in any system of customary law, and it is notorious that they occur throughout the whole body of international law. Municipal courts are frequently called on to apply international law, yet they rarely have any absolutely authoritative guide to the right rule; they have to do their best, with the result that different national courts do not always arrive at uniform decisions on the same point of international law. Yet to allow this lack of uniform interpretation to persuade us that the law that a national court applies in such cases is simply its own municipal law is to come near to denying the existence of international law altogether; what the court does is to apply its own national view of what the rule of international law is, but that is a different matter, and in the present stage of international legal development it is unfortunately inevitable. It will be the same with national courts when they come to the trial of war crimes. They will not all arrive at exactly the same view as to the constituent elements of each particular offence; they will not all follow the same procedure of trial. But so long as the trial accords with natural justice, with "the general principles of law recognised by civilised nations" as principles proper to be applied on the trial of a criminal charge, they will not exceed the discretion that the laws of war have left to them. Nor is it at all likely that the absence of exact uniformity will be serious in practice. Even in the most favourable circumstances it will never be possible to bring to justice more than a small fraction of the crimes that Germans have committed in this war, and the possibility that in the cases that do come to trial any real difficulty, still less any real injustice, will be caused by the uncertainty of the laws of war is almost negligible. Murder, rape, robbery, no doubt it is desirable that crimes like these should be meticulously defined for the purposes of penal law, and in a developed municipal system that is done. But it is idle to expect such refinements in the laws of war. In practice courts will probably follow more or less closely the definitions and the procedures of their own municipal law, and in so doing they will be well within the latitude that the laws of war allow. But again that will not mean that they follow their own municipal law because that is the law which they are bound to apply; it will mean that in the absence of exact definition contained in the laws of war the municipal definition is likely to be the best available guide to the rule that natural justice requires them to apply. An analysis of the Act demonstrates that the last proposition is, if not inaccurate, at least subject to exceptions. The latitude which national courts must have in applying the law of nations cannot relieve them, when the question is in issue, from determining whether a municipal law does create or define an offence in respect of which international law recognizes a universal jurisdiction to try. The latitude available to national courts might allow them to accept that an absolute statutory exclusion of the defence of superior orders (ss. 6(6) and 16) is consistent with international law when the status of that defence is uncertain in international law, but the latitude cannot warrant the disregarding of divergences between the substantive offences defined in municipal law and the offences defined in international law. Nor are the two systems of substantive law brought into unison merely by according "natural justice" which has to do with the procedure of the court, not with the issues to be tried. Some violations of the laws and customs of war may be imprecisely defined, according to the particular law or custom that has been violated, but the Act is no more precise. The Act seems to stretch out to embrace international law in s. 17(2)(a) only to reject the laws, customs and usages of war as a legal touchstone of criminal responsibility in favour of a factual element in s. 17(3). To accept the Act as an approximate reflection of international law would not be to interpret international law but to abdicate the Court's duty of interpretation in favour of the provisions which commended themselves to the Legislature. Whatever right Australia may now possess under international law to bring alleged war criminals from the Second World War to trial before Australian courts, the Act does not exercise it. (b) Resolutions and recommendations of international bodies The resolutions and recommendations of the United Nations General Assembly and of the Economic and Social Council do not suggest that States should bring suspects to trial under a municipal law that is in disconformity with international law. Nor do they suggest that a municipal law which acts retrospectively is an appropriate means of bringing to justice a person whose act did not amount to a crime under international law at the time when the act was done. Indeed, in each of the Geneva Conventions of 1949, the obligation stated is to enact legislation to punish "persons committing, or ordering to be committed" any of the "grave breaches" defined therein: see, for example, Arts 146 and 147 of Geneva Convention IV. The obligation is to translate into municipal law the international law definition of those crimes to be effective in the case of future breaches. In the Genocide Convention of 1948, a similar obligation was imposed (Art. V) followed by a jurisdictional provision (Art. VI) which, as I read it, requires a person charged with a Convention offence to be tried ("shall be tried") either by "a competent tribunal of the State in the territory of which the act was committed, or by such international penal tribunal as may have jurisdiction with respect to those Contracting Parties which shall have accepted its jurisdiction." The Supreme Court of Israel read this provision as not excluding the jurisdiction of national courts but, with respect, I am unable to read it in that way (cf. Baxter's questioning of this jurisdictional decision in Bassiouni and Nanda (eds), op. cit., at p. 85). The material relied on gives no support for the notion that the jurisdiction of a national court to try a person suspected of committing a war crime or a crime against humanity should be exercised pursuant to a law which is in disconformity with international law. It follows that on none of the bases advanced is the Act to be characterized as a law with respect to Australia's external affairs. It thus derives no support from s. 51(xxix). The second power on which reliance is placed to support the Act is the defence power (s. 51(vi)). 3. The defence power The purpose of a law for which support is claimed under s. 51(vi) is "collected from the instrument in question, the facts to which it applies and the circumstances which called it forth": per Dixon J. in Stenhouse v. Coleman [43] . In determining whether a purpose of the Act is to advance the defence of the Commonwealth, it is legitimate to take into account the fact that, by punishing violations of the laws and customs of war, the conduct of future armed conflicts will be less likely to be attended by the barbarities those laws and customs proscribe. The deterrent effect of punishing such violations was noted by resolutions of the United Nations General Assembly passed between 1965 and 1973, but the deterrent effect of trials and punishments under the Act must now be doubtful. The passage of time since 8 May 1945, the occurrence of other armed conflicts in which Australia has been engaged during and after the Second World War (conflicts frequently attended by notorious violations of the laws and customs of war) and the selective operation of the Act upon acts done during armed conflicts in Europe between 1 September 1939 and 8 May 1945 not only diminish, if not destroy, the deterrent effect of the Act but throw doubt on the proposition that defence considerations are "the circumstances which called it forth". 1. (1944) 69 C.L.R. 457, at p. 471. Although the Act is capable of having a relevant deterrent effect and may, on that account, be said to be "appropriate and adapted" to serve defence purposes (as the original War Crimes Act was thought to do), the validity under s. 51(vi) of a law enacted in a time of peace depends upon whether the Parliament might have reasonably considered the means which the law embodies for achieving or procuring the relevant defence purpose to be appropriate and adapted to that end, a question of reasonable proportionality: see per Deane J. in the Tasmanian Dam Case [44] ; Richardson [45] . In times of war, laws abridging the freedoms which the law assures to the Australian people are supported in order to ensure the survival of those freedoms in times of peace. In times of peace, an abridging of those freedoms — in this case, freedom from a retrospective criminal law — cannot be supported unless the Court can perceive that the abridging of the freedom in question is proportionate to the defence interest to be served. What is necessary and appropriate for the defence of the Commonwealth in times of war is different from what is necessary or appropriate in times of peace: Richardson [46] , per Dawson J. "That is because the question of appropriateness and adaptation falls for determination by reference to the circumstances which engage the power": per Gaudron J. in Re Tracey; Ex parte Ryan [47] . The formation of the critical judgment as to whether the means adopted by a law are appropriate and adapted to serve defence purposes is entrusted to the Court: Stenhouse v. Coleman [48] . 1. (1983) 158 C.L.R., at p. 260. 2. (1988) 164 C.L.R., at pp. 291, 311-312, 336, 345-346. 3. (1988) 164 C.L.R., at p. 326. 4. (1989) 166 C.L.R. 518, at p. 597. 5. (1944) 69 C.L.R., at p. 470. The means which the Act adopts to secure future adherence to the laws and customs of war not only trample upon a principle which is of the highest importance in a free society, namely, that criminal laws should not operate retrospectively, but also select a specific group of persons from a time long past out of all those who have committed, or are suspected of having committed, war crimes in other armed conflicts. Respect for the laws and customs of war cannot be secured by a law having such an oppressive and discriminatory operation. I would hold the Act invalid. It follows that I would answer the question reserved for consideration thus: Section 9 of the War Crimes Act 1945 as amended is invalid in its application to the information laid by the second defendant against the plaintiff. Deane J. The question reserved for the consideration of the Court in this case is whether s. 9 of the War Crimes Act 1945 Cth ("the Act") is invalid in its application to the information laid against the plaintiff. That question raises problems of fundamental importance about the extent of the Commonwealth's legislative power with respect to "External affairs" and the implications, in so far as the trial and punishment of criminal offences are concerned, of the Constitution's exclusive vesting of the judicial power of the Commonwealth in the courts contemplated by Ch. III. It is convenient to turn at once to a consideration of the meaning and operation of the central provisions of the Act. Except to the extent necessary for discussion, I shall avoid repetition of either the background facts or the detailed text of the relevant statutory provisions. It suffices to note that the acts which are said to constitute offences under the Act were all allegedly done by the plaintiff in the Ukraine between 1942 and 1943. It is common ground that, at the time those acts were allegedly committed in the Ukraine, the plaintiff had no relevant connexion with this country and that, at that time, none of the alleged offences was a crime under any domestic law in force in this country. If any of the alleged offences is a crime under the domestic law of this country, it is only because it was made such a crime by the War Crimes Amendment Act 1988 Cth ("the 1988 Amendment Act ") which, while retaining the old title, substituted a completely new piece of legislation for the then provisions of the Act. Offences against the Act Section 7 of the Act identifies what constitutes a "war crime" for the purposes of the Act. It does this by building upon a "serious crime" as defined by s. 6. The effect of s. 7, when read in the context of the definitions of war and occupation in s. 5, is that any of the multitude of acts which might constitute a "serious crime" under s. 6(1), (4) and (5) (if done in a part of Australia) or s. 6(3),(4) and (5) (if done outside Australia) is a "war crime" for the purposes of the Act if it bears one or more of a number of specified relationships with the war or armed conflict which "occurred in Europe in the period" between 1 September 1939 and 8 May 1945 or with an occupation of territory arising out of that war or armed conflict (including the occupation of territory in Latvia, Lithuania or Estonia pursuant to the 1939 agreement between Germany and the U.S.S.R.). Section 9(1) of the Act provides that a person who committed a "war crime" during the period from 1 September 1939 to 8 May 1945 is guilty of an indictable offence against the Act. Section 10 prescribes the maximum penalties for an offence against the Act: imprisonment for life for an offence involving wilful killing; imprisonment for twenty-five years for any other offence. Section 11 restricts the persons who may be charged with an offence against this Act to "an Australian citizen" or "a resident of Australia or of an external Territory". Section 13 applies s. 68 of the Judiciary Act 1903 Cth to vest jurisdiction in relation to alleged offences against the Act in courts of the Australian States and internal Territories. The defence under s. 17 Section 17(2) constitutes a key to an understanding of the scope and operation of the Act. It provides that, subject to s. 16 (which excludes a defence of superior orders), "it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity". The past tense "was" in each limb of s. 17(2) makes plain that the temporal reference point for the operation of the sub-section is the time when the offence was allegedly committed. The law permits what it does not proscribe or penalize and the ordinary meaning of "permitted" when used with reference to a law or system of law is "not prohibited" or "not in contravention of". The word "permitted" is used in that sense in par. (a). It follows that, as a matter of ordinary language, the effect of the sub-section is that "it is a defence" if, at the time of its commission, "the act alleged to be the offence" was neither a contravention of the laws, customs and usages of war nor a crime against humanity under international law. The phrase "the laws, customs and usages of war" in par. (a) of s. 17(2) would seem to be the equivalent of the phrase "the laws and customs of war" which has long been used in international treaties and literature. The inclusion of the word "usages" presumably reflects a tendency in some places, including this country, to use that word instead of the more traditional "customs" (see, e.g., United Nations General Assembly Resolution 3(I), "Extradition and Punishment of War Criminals", 13 February 1946; Manual of Military Law 1941, Australian ed., Ch. XIV). The phrase "the laws, customs and usages of war" in par. (a) should be understood as referring to laws and customs which have reached the status of binding rules in the sense explained by Finch in The Nuremberg Trial and International Law, American Journal of International Law, vol. 41 (1947), at pp. 20-21: The laws and customs of war, including those of military occupation, are well established in international law. They are enacted in national legislation, codified in military manuals, incorporated in binding international conventions, and affirmed by the immemorial practice of states thus becoming a part of the common law of war. In their particular context as a "defence" to conduct which would otherwise be punishable as a "war crime" under the statutory definition contained in the main operative provisions of the Act, it appears to me to be clear enough that the words "was permitted by the laws, customs and usages of war" should be understood as meaning conduct which did not constitute a violation of the then rules of international law prohibiting certain types of conduct towards combatants, prisoners and civilians in or in relation to the waging of war. Such conduct is commonly described as a war crime under international law. That description should not, however, be allowed to conceal the fact that the rules of international law, as distinct from particular treaties, did not, at the time of the alleged offences in the present case (i.e. between September 1942 and May 1943), directly establish or make punishable the criminality of individuals for breaches of the law of war: see, generally, Stone, Legal Controls of International Conflict (1954), p. 357; United Nations War Crimes Commission, History of the United Nations War Crimes Commission and the Development of the Laws of War, (1948), pp. 262ff. International law had long recognized the right of a belligerent State to punish for war crimes not only members of its own forces and those who contravened the laws and customs of war upon its territory but also members of the armed forces of an opposing State who fell into its hands. The development of international law to the stage where it made individuals directly and universally punishable by the International Community or by a State or States acting on its behalf for conduct constituting a war crime was, however, something which occurred after the establishment of the United Nations War Crimes Commission in October 1943. It was closely associated with the development of the principle of international law that conduct of an individual could be a crime under international law notwithstanding that it was done in obedience to government or superior orders (see below). The phrase "crime against humanity" has, in the last half-century, also become a commonly used one in international treaties and the writing of publicists. There is little real difficulty about its meaning. It is a convenient general phrase for referring to heinous conduct in the course of a persecution of civilian groups of a kind which is now outlawed by international law but which may not involve a war crime in the strict sense by reason of lack of connexion with actual hostilities. Inevitably, there may be difficulties in determining whether a particular course of conduct (e.g. genocide) was made criminal and punishable as a "crime against humanity" under international law at any particular point of time. It is, however, unnecessary to pursue those difficulties for the purposes of the present judgment. As will be seen, the Act is not concerned with offences against international law in their character as such and does not confine the past acts which it makes punishable for the first time as offences against Commonwealth law to acts which were in contravention of international law. Sub-sections (3),(4) and (5) of s. 17 all bear upon the operation and effect of s. 17(2). They were introduced in the Bill for the 1988 Amendment Act by amendment made in the Senate. Sub-section (3) reads: To avoid doubt, the doing of the act by the defendant was permitted by the laws, customs and usages of war if it was reasonably justified by the exigencies and necessities of the conduct of war. That provision is ambiguous. It is open to be construed as an exhaustive definition of the scope of par. (a) of s. 17(2), that is to say, as confining what is "permitted by the laws, customs and usages of war" for the purposes of par. (a) to what can be reasonably justified by considerations of military necessity. The more obvious construction of sub-s. (3) is, however, not as an exhaustive definition of what falls within par. (a) but as no more than an express identification (to "avoid doubt") of but one of the possible sets of circumstances in which a defence under that paragraph is available. This more obvious construction is, in my view, the preferable one. So construed, the sub-section does not affect the scope of s. 17(2) otherwise than in relation to the special case of military necessity with which it deals. In so far as that special case is concerned, sub-s. (3) arguably extends the scope of sub-s. (2) in that it is arguable that it gives a more general application to a more lenient test than does international law. It would, however, presumably be a less generous test than the extended notion of necessity asserted under the German doctrine of Kriegsraison in the two World Wars: see Stone, op. cit., pp. 351-352. Section 17(4) provides that a defendant "is not entitled to rely on a defence" under s. 17(2) "unless there is evidence of the existence of the facts constituting the defence". Section 17(5) provides that, "if there is such evidence, the onus of establishing, beyond a reasonable doubt, that those facts either do not exist or do not constitute the defence lies on the prosecution". On first impression, sub-ss. (4) and (5) appear to have been framed on the assumption that the question whether the doing of any act falling within the very wide range of s. 7 was permitted by the laws, customs and usages of war and was not under international law a crime against humanity will always involve a defence allegation of additional fact by way of confession and avoidance. Such an assumption would be ill-founded in any case where, regardless of excuse or justification, "the act alleged to be the offence" was not, as a matter of international law, a contravention of the laws, customs and usages of war or a crime against humanity at the time when it was allegedly committed, e.g. some possible examples of the "serious crime" of "manslaughter" (see s. 6(1)(b)) "committed in the course of an occupation": see s. 7(1)(b). Be that as it may, s. 17(4) confirms what s. 17(2) itself clearly indicates, namely, that s. 17 does not add an element to what constitutes a "war crime" for the purposes of the Act. The elements of a "war crime" are to be found in the earlier sections to which reference has been made: in particular, s. 6 which defines a "serious crime"; s. 7 which builds upon a "serious crime" to define a "war crime"; and s. 9 which identifies a past period within which the commission of a "war crime" constitutes an indictable offence. Conduct which constitutes a "war crime" for the purposes of the Act is not punishable as such if a s. 17(2) defence is applicable. Where such a defence involves allegations of additional fact, there must be some evidence of the existence of those facts before the defence can be relied upon. Subject to that qualification, the onus of negativing the defence to the ordinary criminal standard of proof rests on the prosecution. It should be mentioned that it was submitted on behalf of the defendants that there should be implied in s. 7 of the Act a provision to the effect that it is an element of a "war crime" for the purposes of that section that the conduct in question constituted a war crime or crime against humanity under international law. There is, however, nothing in the Act which supports the importation of such a clause. To the contrary, the express provisions of s. 17 preclude the implication of any such general overriding provision in s. 7. The primary legislative purpose to be discerned in the Act The preamble to the Act, like all of the Act's operative provisions, was introduced by the 1988 Amendment Act. It expresses a legislative decision that it is "appropriate" that there should be "brought to trial in the ordinary criminal courts in Australia" any "persons who committed serious war crimes in Europe during World War II" and who " since entered Australia and became Australian citizens or residents" (emphasis added). While the Act in terms applies to "war crimes" committed in Australia, its requirement of one or other of the various specified relationships with the 1939-1945 war or armed conflict which "occurred in Europe" supports the conclusion that, as the preamble indicates, the ordinary operation of the Act was intended by the Parliament to be in relation to the trial and, in case of a conviction, the punishment of Australian citizens or residents accused of past conduct in Europe of the kind described by the Act. The present case falls within the area of that intended ordinary operation in that the plaintiff, an Australian citizen and resident, is charged with acts allegedly committed almost half a century ago in the Ukraine at a time when he had no disclosed connexion with Australia. The defendants' primary contention is that that ordinary operation of the Act (i.e. in relation to a case such as the present where the alleged offence is an act committed outside Australia by a person who was not at the time an Australian citizen or resident) can be justified by reference to the grant of legislative power contained in s. 51(xxix) of the Constitution. For reasons which will appear, it is unnecessary for me to consider a subsidiary submission of the defendants that the Act can also be justified as a law with respect to the naval and military defence of the Commonwealth and the States (Constitution, s. 51(vi)). Section 51(xxix): "External affairs" The first thing to be stressed about s. 51(xxix) of the Constitution for the purposes of the present case is that its reference to "External affairs" is unqualified. The paragraph does not refer to "Australia's external affairs". Nor does it limit the subject matter of the grant of power to external affairs which have some special connexion with Australia. The word "external" means "outside". As a matter of language, it carries no implication beyond that of location. The word "affairs" has a wide and indefinite meaning. It is appropriate to refer to relations, matters or things. Used without qualification or limitation, the phrase "external affairs" is appropriate, in a constitutional grant of legislative power, to encompass both relationships and things: relationships with or between foreign States and foreign or international organizations or other entities; matters and things which are territorially external to Australia regardless of whether they have some identified connexion with Australia or whether they be the subject matter of international treaties, dealings, rights or obligations. Such a construction of the phrase "External affairs" in s. 51(xxix) is supported by the settled principle of constitutional construction which requires that, subject to any express or implied general constitutional limitations and any overriding restrictions flowing from express or implied constitutional guarantees, the grants of legislative power contained in s. 51 be construed with all the generality which the words used admit and be given their full force and effect. The view that a law with respect to matters or things which are territorially outside Australia is a law with respect to "External affairs" for the purposes of s. 51(xxix) is supported by numerous statements in recent cases in this Court. Thus, Barwick C.J., in New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [49] , expressed the view that the legislative power with respect to external affairs "extends to any affair which in its nature is external to the continent of Australia and the island of Tasmania subject always to the Constitution as a whole". To the same effect were the comments of Mason J. [50] : The plaintiffs' argument proceeds on the footing that the power is no more than a power to make laws with respect to Australia's relationships with foreign countries. Why the power should be so confined is not readily apparent. The power is expressed in the widest terms; it relates to "affairs" which are external to Australia. "Affairs" include "matters" and "things" as well as "relationships" and a constitutional grant of plenary legislative power "should be construed with all the generality which the words used admit". Jacobs J. [51] was no less emphatic: In my opinion the Commonwealth has the power to make laws in respect of any person or place outside and any matter or thing done or to be done or prohibited to be done outside the boundaries of the Commonwealth. 1. (1975) 135 C.L.R. 337, at p. 360. 2. (1975) 135 C.L.R., at p. 470. 3. (1975) 135 C.L.R., at p. 497. In the Seas and Submerged Lands Case [52] , Murphy J. said that "external affairs" in s. 51(xxix) are not confined to subjects "of international concern" and are "not limited to the subject matters of treaties and conventions to which Australia is a party, and to the affairs of international bodies of which Australia is a member". His Honour concluded [53] that it mattered not whether the Seas and Submerged Lands Act 1973 Cth departed from the international conventions upon which the Commonwealth relied. It sufficed that the Act dealt directly with "aspects of external affairs". In the context of a challenge to the validity of legislation which dealt with (amongst other things) the continental shelf outside the limits of the territorial sea, it is arguable that Murphy J.'s comments should be read as reflecting an underlying view corresponding to that expressed by Barwick C.J., Mason J. and Jacobs J. in the above-quoted extracts from their judgments. If that be so, the view that the external affairs power extends to authorize the making of laws in respect of any place situate, and any matter or thing done or to be done, outside the boundaries of the Commonwealth was a basis of the decision in the Seas and Submerged Lands Case upholding the validity of the provisions relating to the continental shelf outside territorial Australia. That that view was subsequently held by Murphy J. is made clear by his judgment in Robinson v. Western Australian Museum [54] , where his Honour wrote that the Parliament "has plenary power under the external affairs power to legislate for shipping (whether intra-state, interstate or overseas) in and beyond the territorial sea and may make any provision it thinks fit for wrecks in or under the sea or on the coast", and added that "the practical limitation" upon the exercise of the power was international acceptance. In The Commonwealth v. Tasmania (the Tasmanian Dam Case) [55] , his Honour said: To be a law with respect to external affairs it is sufficient that it deals with circumstances or things outside Australia. (See also Viro v. The Queen [56] .) 1. (1975) 135 C.L.R., at p. 503. 2. (1975) 135 C.L.R., at pp. 503-504. 3. (1977) 138 C.L.R. 283, at p. 343. 4. (1983) 158 C.L.R. 1, at pp. 171-172. 5. (1978) 141 C.L.R. 88, at p. 162. One searches in vain in recent judgments in this Court in cases involving s. 51(xxix) for any denial of the view that the legislative power with respect to external affairs extends to any matter or thing done or situate outside Australia. The closest one comes to finding any express questioning of that view is in the judgment of Gibbs C.J. in Koowarta v. Bjelke-Petersen [57] , where his Honour (who had dissented in the Seas and Submerged Lands Case in relation to the territorial sea but not in relation to the continental shelf beyond territorial waters) saw the question as remaining an open one. Referring to the Seas and Submerged Lands Case, he said [58] : three members of the Court, Barwick C.J., Mason and Jacobs JJ., relied on the further ground that the power given by s. 51(xxix) was not limited to authorizing laws with respect to Australia's relationships with foreign countries, but extended to any matter or thing situated or done outside Australia. It is unnecessary to consider whether the words of par. (xxix) can have this dual operation, i.e. whether the phrase "external affairs" can be used to mean matters outside the Commonwealth as well as matters involving a relationship between Australia and other countries. In contrast, Mason J. in Koowarta [59] expressed the view that the question had been determined by the decision in the Seas and Submerged Lands Case which "decided that the power extends to matters and things, and I would say, persons, outside Australia". His Honour added [59] : That decision established, quite apart from the provisions of the Statute of Westminster, that the so-called principle denying an extra-territorial operation to the legislation of a colony had no application to laws enacted by the Commonwealth Parliament. There was, accordingly, no reason why the power should not apply to any matter or concern external to Australia. In Koowarta , Stephen J. (who, like Gibbs C.J., had dissented in the Seas and Submerged Lands Case in relation to the territorial sea) appears to me to have accepted that s. 51(xxix) conferred a general legislative power with respect to matters outside Australia. The word "External", his Honour wrote [60] , qualified "affairs" so as to restrict the meaning of the composite phrase "External affairs" in s. 51(xxix) "to such of the public business of the national government as relates to other nations or other things or circumstances outside Australia" (emphasis added). In the Tasmanian Dam Case [61] , Gibbs C.J. adopted as accurate that suggested "paraphrase" of the legislative power with respect to external affairs. 1. (1982) 153 C.L.R. 168. 2. (1982) 153 C.L.R., at p. 190. 3. (1982) 153 C.L.R., at p. 223. 4. (1982) 153 C.L.R., at p. 223. 5. (1982) 153 C.L.R., at p. 211. 6. (1983) 158 C.L.R., at p. 97. Whatever may have been the position before the emergence of Australia as a fully independent sovereign State, it should now be accepted that any law which can properly be characterized as a law with respect to any matter, thing or person occurring or situate outside Australia is a law with respect to "External affairs" for the purposes of s. 51(xxix). In referring to "a law with respect to any matter occurring outside Australia", I intend to include, among other things, what Jacobs J. described (see above) as "any matter or thing done or to be done or prohibited to be done outside the boundaries of the Commonwealth". As has been mentioned, that broad view of the scope of the power conforms with settled principles of constitutional construction. It is, as has been seen, arguably part of the ratio decidendi of the Seas and Submerged Lands Case. It is certainly strongly supported by statements in the judgments in that and other recent cases: see, in addition to the above references, Robinson v. Western Australian Museum [62] ; the Tasmanian Dam Case [63] . It is also supported by the consideration that Commonwealth laws with respect to matters, things or persons outside Australia are likely to operate in areas where there will commonly be no competing State interests with the result that, in the absence of Commonwealth legislative power, there would be a lacuna in the plenitude of combined legislative powers of the various Parliaments of the Australian federation. It has long been recognized in this Court that, subject to express and implied constitutional limitations and guarantees, no such lacuna exists in legislative authority in relation to internal matters: see, e.g., Colonial Sugar Refining Co. Ltd. v. Attorney-General (Cth) [64] ; Smith v. Oldham [65] ; Reg. v. Duncan; Ex parte Australian Iron and Steel Pty. Ltd. [66] . With the emergence of Australia as a fully sovereign and independent nation, there remains no acceptable basis for maintaining any such lacuna in the combined powers of the Parliaments of the federation to legislate for this country with respect to extraterritorial matters beyond that resulting from the limitations which the Constitution itself expressly or impliedly imposes. In that regard, it is important to bear in mind that the question whether s. 51(xxix) confers a general plenary legislative power with respect to matters, things and persons occurring or situate outside of Australia does not ordinarily involve what Stephen J. called the "conceptual duality" which applies in a case where the subject matter of a law with respect to external affairs is some matter, thing or person occurring or situate within Australia: see the Seas and Submerged Lands Case [67] . As Stephen J. commented [68] , "the absence of any State interests means that for that area the Commonwealth may act as if it were a unitary state, without need to draw any distinction, if distinction there be, between internal and external aspects of "sovereign rights" " (and see, generally, the discussion in the judgment of Jacobs J. in the Seas and Submerged Lands Case [68] , with which I respectfully agree). 1. (1977) 138 C.L.R., at pp. 294, 335. 2. (1983) 158 C.L.R., at p. 255. 3. (1912) 15 C.L.R. 182, at pp. 214-215. 4. (1912) 15 C.L.R. 355, at pp. 360-361, 365. 5. (1983) 158 C.L.R. 535, at pp. 590-591. 6. (1975) 135 C.L.R., at p. 458. 7. (1975) 135 C.L.R., at pp. 497-498. 8. (1975) 135 C.L.R., at pp. 497-498. To the extent that they provide for the trial and punishment of Australian citizens and residents for acts committed outside Australia and having a specified relationship with the 1939-1945 war or armed conflict which occurred in Europe, the provisions of the Act are clearly a law or laws with respect to matters or things which occurred or were done outside Australia. It follows from what has been said above that, at least to that extent, the Act is a law with respect to "External affairs" within the meaning of that phrase as used in s. 51(xxix). It is unnecessary for the purpose of deciding the present case to determine whether the provisions of the Act can also be characterized as a law with respect to "External affairs" to the extent that they provide for the trial and punishment of Australian citizens and residents in respect of acts committed within Australia which have one or other of the specified relationships with the 1939-1945 war or armed conflict which occurred in Europe. The offences involved in the present case are all alleged to have been committed by the plaintiff outside Australia and the operation of the Act in relation to alleged offences committed outside Australia could be severed from the provisions of s. 6(1) and (2) which have the effect of including acts committed within Australia in conduct punishable under the Act as a "war crime". Indeed, as I followed the argument, it was not submitted that the provisions of the Act dealing with overseas conduct are invalid on the ground that, even if they be valid to the extent of their own operation, they are inseverable from invalid provisions dealing with conduct within Australia. It should, however, be acknowledged that the question whether the Act can properly be characterized as a law with respect to external affairs for the purposes of s. 51(xxix) to the extent that its provisions are made applicable (by s. 6(1) and (2)) to acts committed within Australia is a more complicated one than the question whether it can be so characterized to the extent that its provisions apply to acts committed outside Australia. In its purported application to local conduct, the Act operates in an area in which it interacts with State laws. Indeed, s. 6(1) makes criminality under the law of the locus State an element of a "war crime". That interaction with State law raises difficult questions about whether the Act, if wholly within constitutional power, should be read down to avoid double jeopardy or double punishment in cases where there has been a trial or conviction under State law for that past criminal conduct and whether State laws are rendered invalid, under s. 109 of the Constitution, to the extent that they would provide for future punishment of a "serious crime" that is liable to be punished, under the Act, as a "war crime". In a context where I am unpersuaded that there is any obligation at all upon Australia under customary international law or under any treaty to enact legislation providing for the further punishment of an individual's pre-1945 conduct within Australia (i.e. punishment additional to the penalty already provided under "the law then in force" in the relevant "part of Australia": see s. 6(1) and (2)), the answer to the question whether the provisions of the Act are, to the extent that they deal with conduct within Australia, properly to be characterized as a law with respect to external affairs for the purposes of s. 51(xxix) ultimately turns upon whether, being a law with respect to acts committed within Australia, they can, in all the circumstances, also be properly characterized as a law with respect to the 1939-1945 war or armed conflict in Europe or whether, if they cannot properly be so characterized, their operation is capable of being reasonably considered to be "appropriate and adapted" to deal with some matter of international concern or to achieve an identified purpose or object which is itself a legitimate subject of external affairs in the sense explained in the Tasmanian Dam Case [69] ; see, also, Richardson v. Forestry Commission [70] . 1. (1983) 158 C.L.R., at pp. 259-261. 2. (1988) 164 C.L.R. 261, at pp. 291, 309-312, 324, 336, 344-346. There remains to be considered the question whether, notwithstanding that the Act (at least to the extent that the provisions apply to acts committed outside Australia) is properly to be characterized as a law with respect to external affairs, it is nonetheless beyond the competence of the Parliament of the Commonwealth by reason of some express or implied limitation upon power to be found in the Constitution. Two possible sources of such a restriction upon the Parliament's legislative power with respect to "External affairs" were raised in the course of argument of the present case. The first suggested source of such a restriction is the phrase "for the peace, order, and good government of the Commonwealth" in the introductory words of s. 51. The second is Ch. III of the Constitution. "for the peace, order, and good government of the Commonwealth" Each of the legislative powers conferred by s. 51 is expressly stated, by the introductory words of the section, to be a "power to make laws for the peace, order, and good government of the Commonwealth" with respect to the specified subject matter. It was argued that the words "for the peace, order, and good government of the Commonwealth" introduce a requirement of some identified connexion with Australia and that no such connexion exists in relation to acts committed between 1939 and 1945 in Europe by a person who was not, at the relevant time, a citizen or resident of this country. There is a short answer to that argument. The words "for the peace, order, and good government of the Commonwealth" in s. 51 do not impose an objective qualification upon the subject matters of legislative power. They refer to the intended beneficial operation of laws made with respect to those subject matters and simply express the fact that in "a general and remote sense the purpose and design of every law is to promote the welfare of the community" for which the law is made: see Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [71] ; Union Steamship Co. of Australia Pty. Ltd. v. King [72] . It is for the Parliament, not for the Court, to decide whether a law will be or is "for the peace, order, and good government of the Commonwealth". Where the Parliament decides, by enactment, that a law is "for the peace, order, and good government of the Commonwealth", that, of itself, provides sufficient connexion between this country and the subject matter of that law for the purposes of s. 51: see, generally, King [73] . 1. (1959) 103 C.L.R. 256, at p. 308. 2. (1988) 166 C.L.R. 1, at pp. 12-13. 3. (1988) 166 C.L.R., at pp. 9-13. Moreover, even if, contrary to my own view and to authority, the words "for the peace, order, and good government of the Commonwealth" introduced a requirement that the particular law have some special connexion with this country over and beyond the fact that it is made for Australia by the Australian Parliament, that requirement would be satisfied by the Act. The operation of a law which provides for the trial and, in the event of conviction, punishment of Australian citizens or residents in Australian courts for past conduct in connexion with a war in which this country was a belligerent obviously has a close and special connexion with Australia regardless of whether the alleged offence was committed overseas or whether, at the time of its commission, the alleged offender had any connexion with this country. Chapter III of the Constitution The Constitution is structured upon the doctrine of the separation of judicial from legislative and executive powers. Chapter III gives effect to that doctrine in so far as the vesting and exercise of judicial power are concerned. Its provisions constitute "an exhaustive statement of the manner in which the judicial power of the Commonwealth is or may be vested": see, e.g., Reg. v. Kirby; Ex parte Boilermakers' Society of Australia [74] . The specific grants of legislative powers contained in s. 51 are expressly made subject to the Constitution and it is settled law that those grants of legislative power are subject to the provisions of Ch. III identifying the permissible repositories, and controlling the manner of exercise, of Commonwealth judicial power. 1. (1956) 94 C.L.R. 254, at p. 270. The main objective of the sometimes inconvenient separation of judicial from executive and legislative powers had long been recognized at the time of the federation. It is to ensure that "the life, liberty, and property of the subject [is not] in the hands of arbitrary judges, whose decisions [are] then regulated only by their own opinions, and not by any fundamental principles of law" (Blackstone, Commentaries, 17th ed. (1830), vol. I, p. 269; and see, to like effect, Story, Commentaries on the Constitution of the United States (1833), §1568). That objective will, of course, be achieved only by the Constitution's requirement that judicial power be vested exclusively in the courts which it designates if the judicial power so vested is exercised by those courts in accordance with the essential attributes of the curial process (cf. Re Tracey; Ex parte Ryan [75] ). Indeed, to construe Ch. III of the Constitution as being concerned only with labels and as requiring no more than that the repository of judicial power be called a court would be to convert it into a mockery, rather than a reflection, of the doctrine of separation of powers. Common sense and the provisions of Ch. III, based as they are on the assumption of traditional judicial procedures, remedies and methodology (see below), compel the conclusion that, in insisting that the judicial power of the Commonwealth be vested only in the courts designated by Ch. III, the Constitution's intent and meaning were that that judicial power would be exercised by those courts acting as courts with all that that notion essentially requires. Accordingly, the Parliament cannot, consistently with Ch. III of the Constitution, usurp the judicial power of the Commonwealth by itself purporting to exercise judicial power in the form of legislation. Nor can it infringe the vesting of that judicial power in the judicature by requiring that it be exercised in a manner which is inconsistent with the essential requirements of a court or with the nature of judicial power. It would, for example, be beyond legislative competence to vest jurisdiction to deal with a particular class of matter in a Ch. III court and to provide that, in the exercise of that jurisdiction, the judge or judges constituting the court should disregard both the law and the essential function of a court of law and do whatever they considered to be desirable in the public interest. 1. (1989) 166 C.L.R. 518, at p. 580. The ordinary object of the exercise of judicial power is the ascertainment of rights and liabilities or of guilt or innocence under the law. The point was made by the United States Supreme Court in Prentis v. Atlantic Coast Line Co. [76] , in a passage which has been quoted with approval on a number of occasions in this Court (see, e.g., Rola Co. (Australia) Pty. Ltd. v. The Commonwealth [77] ; Reg. v. Davison [78] ): A judicial inquiry investigates, declares and enforces liabilities as they stand on present or past facts and under laws supposed already to exist. That is its purpose and end. Legislation on the other hand looks to the future and changes existing conditions by making a new rule to be applied thereafter Prima facie, the relevant substantive law for determining rights and liabilities is the law which operated at the time of the circumstances from which those rights and liabilities are alleged to arise. Thus, it is a rule of construction that it is to be presumed that it was not the legislative intent that a statutory provision which affects rights or liabilities should operate retrospectively. Nonetheless, the focus of civil litigation is upon the determination of rights and liabilities under the law as it exists at the time of the proceedings. Civil legislation which operates retrospectively in the sense that it extinguishes or alters pre-existing rights or liabilities or deems rights and liabilities which it creates to have existed at an earlier time may, depending on the circumstances, be susceptible of legitimate criticism as unfair or unjustified. Such legislation will not, however, contravene the doctrine of separation of powers merely because it retrospectively creates, extinguishes or alters civil rights and liabilities or because it requires the courts to recognize and enforce, in subsequent civil litigation, the retrospective operation of its provisions (cf., e.g., Reg. v. Kirby; Ex parte Boilermakers' Society of Australia [79] ). Putting to one side cases of impermissible interference with the proper discharge of judicial power by the courts, it will do so only if it is properly to be seen as involving a purported legislative exercise of the judicial function. Except in quite extreme cases (e.g. a statute providing that there be a verdict for the plaintiff in the amount of $500,000 in pending defamation proceedings in a court), the boundary between what is permissible as falling within the limits of legislative power and what is forbidden as a usurpation of judicial power is likely to be blurred in civil matters. The reason is that both the legislature and the judicature may, within the limits of their respective functions under the doctrine of separation of powers, each settle questions of rights and liabilities under the civil law. The position is different, however, in the case of a law which operates to make criminal an act which was not a crime when done. I turn to explain why that is so. For convenience of discussion, I shall refer to such a law as an "ex post facto criminal law". 1. (1908) 211 U.S. 210, at p. 226. 2. (1944) 69 C.L.R. 185, at p. 211. 3. (1954) 90 C.L.R. 353, at p. 370. 4. (1956) 94 C.L.R., at p. 281. There are some functions which, by reason of their nature or because of historical considerations, have become established as incontrovertibly and exclusively judicial in their character. One — and the most important — of such functions is the adjudgment of guilt of a person accused of a criminal offence. As such, that function is a matter "appertaining exclusively to [judicial] power" (per Griffith C.J., Waterside Workers' Federation of Australia v. J.W. Alexander Ltd. [80] ) which "could not be excluded from the judicial power": see, e.g., Reg. v. Davison [81] . It follows that, subject only to limited qualifications (see below), the Parliament is incapable of vesting jurisdiction to try a person accused of an offence against a law of the Commonwealth otherwise than in one of the "courts" contemplated by Ch. III or of directing that part of such jurisdiction be exercised in a manner inconsistent with the requirements of that Chapter. More important for present purposes, the Parliament is incapable of substituting a legislative enactment of criminal guilt of an offence against a law of the Commonwealth for a trial by a Ch. III court. To appreciate the implications of that fundamental constitutional truth, it is necessary to understand the central function of a criminal trial under our system of government. 1. (1918) 25 C.L.R. 434, at p. 444. 2. (1954) 90 C.L.R., at pp. 368-369, 383. The basic tenet of our penal jurisprudence is that every citizen is "ruled by the law, and by the law alone". The citizen "may with us be punished for a breach of law, but he can be punished for nothing else" (Dicey, Introduction to the Study of the Law of the Constitution, 10th ed. (1959), p. 202). Thus, more than two hundred years ago, Blackstone taught (see Commentaries (1830), vol. I, pp. 45-46) that it is of the nature of law that it be "a rule prescribed " and that, in the criminal area, an enactment which proscribes otherwise lawful conduct as criminal will not be such a rule unless it applies only to future conduct. Consequently, it would be contrary to the nature of the legislative function under our system of government, if: after an action (indifferent in itself) is committed, the legislator then for the first time declares it to have been a crime, and inflicts a punishment upon the person who has committed it. Here it is impossible that the party could foresee that an action, innocent when it was done, should be afterwards converted to guilt by a subsequent law: he had therefore no cause to abstain from it; and all punishment for not abstaining must of consequence be cruel and unjust. All laws should be therefore made to commence in futuro, and be notified before their commencement; which is implied in the term " prescribed ". Put differently, it is basic to our penal jurisprudence that a person who has disobeyed no relevant law is not guilty of a crime. Of its nature, a crime "is an act committed, or omitted, in violation of a public law, either forbidding or commanding it" (Blackstone, Commentaries (1830), vol. IV, p. 5). It necessarily involves a contravention of a prohibition contained in an existing applicable valid law. As Lord Atkin wrote, for the Judicial Committee of the Privy Council, in Proprietary Articles Trade Association v. Attorney-General (Canada) [82] : Criminal law connotes only the quality of such acts or omissions as are prohibited under appropriate penal provisions by authority of the State. The criminal quality of an act cannot be discerned by intuition; nor can it be discovered by reference to any standard but one: Is the act prohibited with penal consequences? Morality and criminality are far from coextensive; nor is the sphere of criminality necessarily part of a more extensive field covered by morality — unless the moral code necessarily disapproves all acts prohibited by the State, in which case the argument moves in a circle. It appears to their Lordships to be of little value to seek to confine crimes to a category of acts which by their very nature belong to the domain of "criminal jurisprudence"; for the domain of criminal jurisprudence can only be ascertained by examining what acts at any particular period are declared by the State to be crimes, and the only common nature they will be found to possess is that they are prohibited by the State Accordingly, the whole focus of a criminal trial is the ascertainment of whether it is established that the accused in fact committed a past act which constituted a criminal contravention of the requirements of a valid law which was applicable to the act at the time the act was done. It is the determination of that question which lies at the heart of the exclusively judicial function of the adjudgment of criminal guilt. 1. [1931] A.C. 310, at p. 324. That basic tenet of our penal jurisprudence has at times been blurred in the area of so-called "judge-made" criminal law (see, generally, Spencer, Nulla Poena sine Lege in English Criminal Law, The Cambridge-Tilburg Law Lectures 1980; Kelsen, General Theory of Law and State (1945), p. 146). In so far as legislation is concerned, history records some departures from it particularly in times of civil disturbance. At least since the time of Bentham and Mill, however, ex post facto criminal legislation has been generally seen in common law countries as inconsistent with fundamental principle under our system of government. The point was well made by a very strong Court of Exchequer Chamber (Kelly C.B., Martin, Channell, Pigott and Cleasby BB., Willes and Brett JJ.) in a judgment delivered by Willes J. in Phillips v. Eyre [83] . Having recognized that some retrospective legislation was "beneficial and just", their Lordships wrote [84] : The retrospective Attainder Acts of earlier times, when the principles of law were not so well understood or so closely regarded as in the present day, and which are now looked upon as barbarous and loosely spoken of as ex post facto laws, were of a substantially different character. They did not confirm irregular acts, but voided and punished what had been lawful when done. Mr. Justice Blackstone [1 Bla. Com. 46] describes laws ex post facto of this objectionable class as those by which "after an action indifferent in itself is committed, the legislature [Blackstone wrote "legislator"] then for the first time declares it to have been a crime, and inflicts a punishment upon the person who has committed it" (Emphasis added). Obviously, their Lordships were not suggesting — any more than was Blackstone — that ex post facto criminal legislation was beyond the legislative competence of the Imperial Parliament whose powers have never been confined by an entrenched doctrine of the separation of judicial from legislative and executive powers. Nonetheless, their Lordships' comments — like those of Blackstone — are directly relevant to the determination of what lies beyond the limits of the legislative function under a constitution which, like ours, entrenches the doctrine and subjects legislative power to it. In that regard, it is important to note that their Lordships identified the central vice of a Bill of Attainder not as lying in its specific naming of an individual but as lying in its ex post facto operation as a legislative decree that an act which was not criminal when done was "voided and punished" as a crime. A statute which decreed that "any person" who had supported the unsuccessful party in some past period of civil disturbance was, notwithstanding that he had contravened no then existing law, guilty of treason and subject to a death penalty would not be a Bill of Attainder in the strict sense in that a trial would be necessary to determine whether a particular accused had in fact supported the unsuccessful party and the actual sentencing would be by a court. It would, nonetheless, fall squarely within the category of laws which their Lordships condemned as inconsistent with a proper understanding of "the principles of law". So also does any statute which, like s. 9(1) of the Act, declares that a person is guilty of a crime against the law of the Commonwealth if he has committed a past act which did not, when committed, contravene any then existing and applicable law of the Commonwealth and was therefore not such a crime. 1. (1870) L.R. 6 Q.B. 1. 2. (1870) L.R. 6 Q.B., at p. 25. The perception that ex post facto criminal legislation lies outside the proper limits of the legislative function is not confined to countries whose legal traditions can be traced to the British system of government. It is shared by all the nations of the European Economic Community: see Case 63/83 Reg. v. Kirk [85] . It is reinforced by the provisions of international conventions concerned with the recognition and protection of fundamental human rights: see, e.g., Universal Declaration of Human Rights (1948), Art. 11(2); European Convention for the Protection of Human Rights and Fundamental Freedoms (1950), Art. 7; American Convention on Human Rights (1969), Art. 9. In some international conventions and national declarations of fundamental rights, a disavowal of retrospective criminal legislation is made subject to a qualification in respect of an act which was, at the time it was committed, a crime against international law. Such a qualification is not in point in the present case where the conduct made criminal by the Act is not made punishable as, and need not necessarily have been, a crime against international law at the time it occurred. 1. [1984] E.C.R. 2689, at p. 2718. A statutory provision, such as s. 9 of the Act, that a "person who" in the past "committed" a specified act "is guilty of" a punishable crime prescribes no rule of conduct. It prohibits nothing. It trespasses upon the exclusively judicial field of determining whether past conduct was a crime, that is to say, whether it was in fact an act or omission which the law "prohibited with penal consequences". Within that field, it negates the ordinary curial process by enacting, and requiring a finding of, criminal guilt regardless of whether there was in fact any contravention of any relevant law. If the specified act was not prohibited by such a law when done, such a statutory provision is a retroactive legislative declaration of past criminal guilt when in fact there was none. If it nominates, either individually or by reference to an identifiable group, the person or persons who have committed the specified act, it constitutes a Bill of Attainder or a Bill of Pains and Penalties, depending upon the punishment. In such a case, the statutory provision constitutes a legislative declaration of guilt without any trial at all. Plainly, it involves a usurpation of judicial power. As Murphy J. wrote in Victoria v. Australian Building Construction Employees' and Builders Labourers' Federation [86] : For centuries the finding that a person has broken the criminal law has been regarded as within the judicial sphere, and outside the sphere of the Parliament and the executive. Bills of attainder by which Parliament entered the sphere of criminal justice are inconsistent with this basic constitutional scheme. (Emphasis added.) 1. (1982) 152 C.L.R. 25, at p. 107. The position is less obvious where such a statutory provision does not nominate a particular person or group of persons but identifies the persons whom it makes punishable for past "crime" by reference only to their having committed some past act which was not criminal when done. In such a case, there will be a need for a trial to determine whether a particular accused falls within the class of those whose past conduct is retroactively made criminal. Nonetheless, such a statutory provision declaring past conduct to have been a criminal offence constitutes a usurpation of judicial power in that, once it is established that the accused has committed the past act, the question whether that act constituted a criminal contravention of the law is made simply irrelevant. To that extent, curial determination of criminal guilt is ousted by legislative decree. The point can be illustrated by dividing the legislation in such a case into its essential components. One component of such legislation is the requirement that there be a "trial" in the courts, in which judicial process must be observed, to determine whether it is established beyond reasonable doubt that a particular person knowingly engaged in the designated conduct. The second component is the enactment that, if it be established that the particular person did in fact engage in that past conduct which was not criminal when done, he is guilty of a punishable crime. That second component of the legislation invades the heart of the exclusively judicial function of determining criminal guilt, that is to say, of determining whether past conduct constituted a criminal contravention of the law. It pre-empts and negates what would otherwise be an inevitable judicial determination that, since the act of the particular person did not constitute a criminal contravention of any Commonwealth law which was applicable at the time when it was done, that person committed no crime under our law. In the place of that inevitable judicial determination, it imposes a legislative enactment of past guilt which it requires the courts, in violation of the basic tenet of our criminal jurisprudence and the doctrine of separation of judicial from legislative and executive powers, to apply and enforce. It is simply not to the point that the first component of the legislation camouflages the usurpation of judicial power involved in the second by requiring a display of the full panoply of judicial process for the purpose of determining whether it is established beyond reasonable doubt that the accused person knowingly did a specified act which was not criminal when done. It follows from what has been written above that Ch. III's exclusive vesting of the judicial power of the Commonwealth in courts acting as such does not, at least prima facie, permit the conviction of a person by a court under a statute which declares that a person is guilty of a punishable crime if he has done a past act which was not criminal when done. There are two closely related reasons why the process leading up to such a conviction would prima facie contravene the doctrine of separation of powers embodied in Ch. III. It is convenient to restate them in summary form. First, the legislature's interference in that process would go beyond the limits of the legislative function under a constitution structured upon the separation of judicial and legislative powers in that it would involve a usurpation and partial exercise of what lies at the heart of the exclusively judicial function in criminal matters, namely, the determination of whether the accused person has in fact done an act which constituted a criminal contravention of the then applicable law. Second, a court's participation in that process would also be inconsistent with the doctrine of the separation of powers in that it would represent an abdication of the judicial function of determining in a criminal trial whether past conduct had contravened the law in favour of the legislature's decree that a past non-criminal act is to be punished as a crime. It becomes necessary to consider whether anything within the Constitution itself or considerations of context or authority require that the provisions of Ch. III be given some more limited effect than that which the doctrine of separation of judicial from legislative and executive powers would prima facie support. I shall consider that question under three headings: "Constitutional provisions and structure", "The United States precedent" and "The effect of decided cases". Constitutional provisions and structure It can be said at once that there is nothing in the Constitution which militates against construing Ch. III as precluding the enactment (by the Parliament) or the application and enforcement (by the courts) of an ex post facto criminal law. To the contrary, some of the specific provisions and the general structure of the Constitution support such a construction. The provisions of Ch. III are based on an assumption of traditional judicial procedures, remedies and methodology. They confer "jurisdiction", that is, the curial power of declaration (dictio) of the law (jus): "the power and authority of a court to hear and determine a judicial proceeding" (In re Estate of De Camillis [87] ). The jurisdiction so conferred is upon "courts" rather than upon the judge or judges who constitute a particular court. Most important, it is with respect to "matters". In a context such as Ch. III, the word "matters" denotes controversies of a kind appropriate to come before a court of justice: see, e.g., per Griffith C.J., South Australia v. Victoria (" the Boundaries Case ") [88] . A matter "must be such that it can be determined upon principles of law" ( [88] ; and see, to the same effect [89] , per O'Connor J.). The "expression" (i.e. matter in Ch. III) "includes and is confined to claims resting upon an alleged violation of some positive law " ( [90] , per Isaacs J.). In their joint judgment in In re Judiciary and Navigation Acts [91] , Knox C.J., Gavan Duffy, Powers, Rich and Starke JJ. cited the judgments in the Boundaries Case as authority for the proposition, that "a matter under the judicature provisions of the Constitution must involve some right or privilege or protection given by law, or the prevention, redress or punishment of some act inhibited by law " (emphasis added). 1. (1971) 322 N.Y.S. 2d 551, at p. 556. 2. (1911) 12 C.L.R. 667, at p. 675. 3. (1911) 12 C.L.R. 667, at p. 675. 4. (1911) 12 C.L.R., at p. 708. 5. (1911) 12 C.L.R., at p. 715. 6. (1921) 29 C.L.R. 257, at p. 266. Section 80 of the Constitution provides clear confirmation of Ch. III's assumption of traditional curial standards, procedures and methodology in so far as the criminal law is concerned. It assumes that the determination of criminal guilt will be by "trial" and provides that, when the "trial" is "on indictment", it "shall be by jury". Traditionally, the ultimate question in a criminal trial by jury has been whether the accused did something which was, at the time he did it, contrary to the then law. In that regard, it is quite contrary to all traditional notions of a criminal trial by jury to require the jury to convict a person of a crime by reason of an act which was not criminal when done. In requiring that the trial by jury "shall be held in the State where the offence was committed " (emphasis added), s. 80 assumes, what would in any event be implicit in the requirement of trial by jury, that the question of guilt or innocence will be determined by reference to criminality or otherwise as at the time of the alleged offence. Moreover, the structure of the federation and the paramountcy of the Commonwealth's legislative powers with respect to designated subject matters could give rise to extraordinary injustice, difficulty and uncertainty if it lay within the Commonwealth's legislative competence to enact an ex post facto criminal law. If the proscribed act under such a Commonwealth law had been mandatory under an operative and applicable State law at the time when the act was done, the accused would be made retroactively guilty of a crime for doing what the law actually required him to do at the time he did it. If the Commonwealth's ex post facto criminal law covered the field and purported to exclude the operation of any State law, the question would arise whether the exclusion and consequent pro tanto invalidity of the State law could be made fully retroactive (cf. University of Wollongong v. Metwally [92] ). If it could, State criminal law would assume a provisional character in any area within the reach of Commonwealth legislative power. If, in such a case, the accused had been convicted and punished for an offence against the inconsistent State law, he would have been accused and punished under what retroactively had been made invalid. If the Commonwealth law did not apply retroactively to invalidate the State law but had full cumulative operation, double jeopardy and double punishment would be the result. If it be objected that these problems are hypothetical and far-fetched, one need do not more than refer to the provisions of ss. 6(1), 7(1) and 9(1) of the Act. Under those provisions, a person who had been convicted of the offence of "wounding" in 1942 in New South Wales would be retroactively made guilty of a "war crime" under the Act if his offence was committed in any of the circumstances referred to in s. 7(1)(c) and (d). In such a case, the elements of the offence of "wounding" under New South Wales law would be elements of the "war crime" under the Act. Unless one can read down the provisions of the Act so that they do not apply to the case where a person has been convicted of the underlying "serious crime" under s. 6(1), those provisions would inevitably give rise, in such a case, to serious difficulties and uncertainties about, among other things, the past validity of the State law and the effect of the past trial and conviction in respect of the State offence. 1. (1984) 158 C.L.R. 447, at pp. 478-479. The United States precedent The doctrine of the separation of powers which is incorporated in the Constitution differs from that embodied in the United States Constitution in so far as the relationship between the legislative and executive arms of government is concerned. Chapter III's separation of judicial from executive and legislative power was framed on the United States model. Even in that area, however, there is a relevant and important difference between the two in that the United States Constitution contains express prohibitions of any "Bill of Attainder or ex post facto Law" (Art. I, §9, cl. 3 (Federal) and Art. I, §10, cl. 1 (State): "the Bill of Attainder Clause") which does not appear in our Constitution. It can be argued that the failure of the Constitution to follow the United States precedent in that regard provides grounds for rejecting a construction of Ch. III which would preclude the enactment by the Parliament of an ex post facto criminal law. Such an argument is, of course, equally applicable to a Bill of Attainder in the strict sense. A general answer to it is that, in circumstances where the framers of our Constitution took a quite different approach to the need to incorporate express and detailed statements of underlying individual rights and guarantees, it is impermissible to treat the absence of one or other of the express statements of individual rights or guarantees contained in the United States Constitution as, of itself, constituting a reason for rejecting the possibility that the provisions of our Constitution may give rise to the implication of a corresponding individual right or guarantee (cf. Street v. Queensland Bar Association [93] ). There is, however, a more particular answer which turns the argument on its head. 1. (1989) 168 C.L.R. 461 pp. 521-522. At the time when the Constitution was adopted by the Australian people, it had long been recognized in the United States that the Bill of Attainder Clause did no more than make express what was, in any event, implicit in the doctrine of the separation of judicial from legislative and executive powers. Thus, in Calder v. Bull [94] , Chase J., in the course of what has subsequently been accepted as the classic exposition of the effect of that clause, wrote: The prohibition against their making any ex post facto laws was introduced for greater caution, and very probably arose from the knowledge, that the parliament of Great Britain claimed and exercised a power to pass such laws, under the denomination of bills of attainder, or bills of pains and penalties; the first inflicting capital, and the other less, punishment. These acts were legislative judgments; and an exercise of judicial power. (Emphasis added.) In a context where the express prohibition of "ex post facto laws" was seen as merely declaratory of an aspect of the doctrine of separation of powers, it was a natural step to confine the scope of that phrase to its traditional meaning of ex post facto laws dealing with criminal offences. That step was taken in Calder v. Bull where it was laid down that the reference to ex post facto laws in Art. I should be construed in a technical sense and confined to criminal laws which "create or aggravate the crime; or increase the punishment, or change the rules of evidence, for the purpose of conviction": see per Chase J. [95] ; Watson v. Mercer [96] . As I have indicated, I am, in this judgment, using the phrase "ex post facto criminal law" in the even narrower sense explained by Blackstone, that is to say, as referring only to a law which retroactively makes criminal an act which was not criminal when done. 1. (1798) 3 U.S. 386, at p. 389. 2. (1798) 3 U.S., at p. 391. 3. (1834) 33 U.S. 88, at pp. 109-110. Chase J.'s view that the Bill of Attainder Clause, construed as confined to criminal laws, merely states what is implicit in the doctrine of separation of powers remains the accepted view in the United States. Thus, one finds the United States Supreme Court in United States v. Brown [97] , acknowledging that: The best available evidence, the writings of the architects of our constitutional system, indicates that the Bill of Attainder Clause was intended not as a narrow, technical (and therefore soon to be outmoded) prohibition, but rather as an implementation of the separation of powers, a general safeguard against legislative exercise of the judicial function, or more simply — trial by legislature. (Emphasis added.) (And see also, e.g., Ogden v. Blackledge [98] ; Weaver v. Graham [99] ; Tribe, American Constitutional Law, 2nd ed. (1988), pp. 656ff.; Lehmann, The Bill of Attainder Doctrine: A Survey of the Decisional Law, Hastings Constitutional Law Quarterly, vol. 5 (1978), pp. 777-780; The Bounds of Legislative Specification: A Suggested Approach to the Bill of Attainder Clause, Yale Law Journal, vol. 72 (1962), pp. 330ff.; but cf. Berger, Bills of Attainder: A Study of Amendment by the Court, Cornell Law Review, vol. 63 (1978), pp. 379ff.) 1. (1965) 381 U.S. 437, at p. 442. 2. (1804) 2 Cranch (6 U.S.) 272, at p. 277. 3. (1981) 450 U.S. 24, at p. 29. It follows that the fact that the provisions of the Constitution incorporating the doctrine of the separation of judicial from legislative and executive powers were obviously influenced by corresponding provisions of the United States Constitution provides no reason for denying that it is implicit in those provisions that the Parliament lacks competence to enact an ex post facto criminal law. To the contrary, the recognition, long before 1900, that a prohibition of ex post facto criminal laws was implicit in the doctrine of separation of judicial power which those United States provisions embodied supports the conclusion that Ch. III precludes the enactment of such a law. In that regard, it is relevant to note that at least one of the principal framers of our Constitution saw its adoption of the doctrine of the separation of judicial from legislative and executive powers as involving significant limitation on the powers of the Parliament corresponding to those which had been recognized by the United States decisions as flowing from that doctrine. In his Studies in Australian Constitutional Law, (1901), Mr. Justice Inglis Clark defined a "law" of the Commonwealth as "a rule of conduct prescribed by the Parliament in regard to any matter in respect of which the Parliament is authorised by the Constitution to make laws" (emphasis added). In the course of the following paragraph — opposite the marginal note "The distribution of governmental powers implies a limitation on the power of Parliament" — he wrote (pp. 39-41, emphasis again added): The Constitution does not prohibit the Parliament of the Commonwealth from making retroactive laws; and none of the Constitutions of the States imposes any such prohibition upon the Parliament of the State. But by limiting the governmental authority of the Federal Parliament to the exercise of legislative functions the Constitution of the Commonwealth has indirectly but effectually prohibited to the Parliament any legislation similar in character to some of the legislation which has been enacted in some, if not all, of the colonies which have become States of the Commonwealth. The legislation to which this statement refers has been usually enacted to remedy defects in previous legislation which have been discovered in the course of litigation and have defeated the expectations of the promoters of the previous law. Such amending laws have been repeatedly declared by the American courts to be invalid because they were encroachments upon the exclusive province of the Judiciary under a constitution which conferred separately upon different departments of the government the legislative, the executive and the judicial powers exercisable under it [See People v. Board of Supervisors, [1] ; Governor v. Porter, [2] ; Mayor &c. v. Horn, [3] ]. The underlying principle of the decisions of the American courts upon this subject was concisely stated by Thompson, J., in the case of Dash v. Van Kleeck [4] in which he said — " To declare what the law is or has been is a judicial power; to declare what the law shall be is legislative. " Clearly enough, the first sentence in the above passage must be read, in context, as referring to a direct constitutional prohibition of retroactive laws. Equally clearly, Inglis Clark's apparent acceptance of the "underlying principle" referred to in the last sentence as applicable to our Constitution would preclude the enactment of an ex post facto criminal law. 1. (1857) 16 N.Y., 424. 2. (1844) 5 Humph., 165. 3. (1866) 26 Md., 294. 4. (1811) 7 Johns., 477. The effect of decided cases The only case in this Court in which reasoned consideration has been given to the question of the legislative competence of the Parliament of the Commonwealth to enact an ex post facto criminal law is R. v. Kidman [5] . In that case, the Court upheld the validity of s. 3 of the Crimes Act 1915 Cth which had been assented to on 7 May 1915. That Act (by s. 2) added conspiracy to defraud the Commonwealth to the conspiracies which had been declared by s. 86 of the Crimes Act 1914 Cth to be indictable offences. Section 3 provided that the amendment was to be deemed to have been in force from the date of the commencement of the 1914 Act (29 October 1914). The Court held, unanimously, that it had been within the legislative competence of the Parliament to make the legislation retrospective. 1. (1915) 20 C.L.R. 425. The first thing to be noted about Kidman for the purposes of the present case is that all members of the Court saw the question of the validity of the retrospective provisions as depending "entirely on the meaning of sub-sec. xxxix [i.e. the "incidental" power] of sec. 51 of the Constitution " (per Isaacs J. [6] ; and see also per Griffith C.J. [7] , per Higgins J. [8] , per Gavan Duffy and Rich JJ. [9] , per Powers J. [10] , who also relied upon the "defence power" conferred by s. 51(vi)). Notwithstanding that reference was made to Calder v. Bull, it apparently was not suggested in argument in Kidman that there was any relevant limitation of the plenary legislative powers conferred upon the Parliament by s. 51 to be derived from other provisions of the Constitution (see per Higgins J. [11] ). The result was that none of the judgments adverted to the question whether a retroactive criminal law was consistent with s. 71's exclusive vesting of judicial power in Ch. III courts. With the benefit of hindsight, that may seem to be surprising. It must, however, be remembered that Kidman was decided some three years before the Court was first called upon to "make a critical examination" of the effect of s. 71's vesting of judicial power in Ch. III courts: see Alexander [12] . 1. (1915) 20 C.L.R., at p. 440. 2. (1915) 20 C.L.R., at pp. 434-435. 3. (1915) 20 C.L.R., at pp. 450-451. 4. (1915) 20 C.L.R., at p. 455. 5. (1915) 20 C.L.R., at pp. 460-461. 6. (1915) 20 C.L.R., at p. 451. 7. (1918) 25 C.L.R., at p. 441. Kidman is clear authority for the proposition that the fact that a criminal law operates retrospectively does not necessarily preclude it from being properly characterized as a law for the peace, order and good government of the Commonwealth with respect to one or other of the heads of legislative power set out in s. 51. In that regard, the case provides support for the conclusion reached earlier in this judgment that the operative provisions of the Act are properly to be characterized as a law with respect to "External affairs" notwithstanding that they relate only to past acts. Thus, Gavan Duffy and Rich JJ. said [13] : in any case Parliament may think and rightly think that punishment of offences committed before the passing of the Act would be likely to deter persons from obstructing the execution of powers in the future. If we take the collection of Customs duties by way of a concrete example of the execution of a power, might not Parliament reasonably think that the most effective means to prevent the perpetration of frauds in the course of such collection, and so protect and facilitate the execution of the power, would be not merely to provide for the punishment of those who offended in the future but to actually punish offenders in the past? If so, Parliament would be at liberty to adopt those means. There are statements to like effect in the judgments of Isaacs J. [14] , Higgins J. [15] and Powers J. [16] . By analogy, the Parliament might "reasonably think" that the punishment of present Australian citizens and residents for past acts of barbarism in other countries would act as a deterrent to other Australian citizens and residents from engaging in such conduct in the future. That of course says nothing on the question whether such a law is consistent with Ch. III of the Constitution. 1. (1915) 20 C.L.R., at pp. 456-457. 2. (1915) 20 C.L.R., at p. 443. 3. (1915) 20 C.L.R., at p. 450. 4. (1915) 20 C.L.R., at pp. 460-461. Even though the possible effect of Ch. III was neither raised nor considered in Kidman , Griffith C.J. was troubled by the notion that the Parliament possessed legislative competence to enact an ex post facto criminal law, that is, a law which retroactively makes criminal an act which was indifferent when done. His Honour saw the making and enforcement of such a law as standing quite outside the ordinary enactment and administration of criminal laws with respect to, or for the purposes of, the subject matters of legislative power contained in s. 51 (17). He expressed the view that, if it were necessary to assign an ex post facto criminal law to some definite category, "the true category would be "Control over the liberty of the subject" or "Reward and punishment of citizens who have deserved well or ill of the State" " and concluded [17] that a law which operates merely as an ex post facto criminal law was not within the legislative power conferred by s. 51(xxxix). The Chief Justice's difficulty with the notion that the Parliament could, as a matter of characterization under s. 51, have legislative competence to enact an ex post facto criminal law was not shared by the other members of the Court. The discussion of that difficulty in his judgment tends, however, to underline the importance of the fact that no consideration at all was given in Kidman to the possible relevance of Ch. III of the Constitution. 1. (1915) 20 C.L.R., at pp. 432-434. The basis of Griffith C.J.'s conclusion in Kidman that the retrospective provision of the 1915 Act was valid was that the law was not really an ex post facto criminal law at all. In his Honour's view, conduct which constituted conspiracy to defraud the Commonwealth had already been a crime at common law before the enactment of the 1915 Act. On that view, all that the 1915 Act did was to enact "in the form of a Statute the unwritten law of the Commonwealth" [18] . Griffith C.J.'s view that the legislation in Kidman was not an ex post facto criminal law in the narrow sense in which I have been using that phrase, was apparently shared by Isaacs J. [19] . The judgments of a majority of the Court proceeded, however, on the assumption that the 1915 Act was a truly ex post facto criminal law. That being so, Kidman is properly to be seen as providing technical authority for a denial of the proposition that Ch. III of the Constitution precludes the enactment by the Parliament of such a law. However, in circumstances where the effect of Ch. III and the question of the validity of that proposition were neither raised in argument nor considered in any of the judgments and where the case was decided before the implications of the adoption by Ch. III (and by the Constitution as a whole) of the doctrine of the separation of judicial from administrative and legislative power had been critically examined in this Court, the authority of Kidman in that regard is weak indeed. This Court is the creation and the servant of the Constitution. It would be quite wrong for it to treat an implicit decision of its own in an earlier case on a point which was neither raised in argument nor considered in the judgments as of itself providing a justification for refusing to acknowledge the implications of Ch. III's fundamental guarantee of judicial process. 1. (1915) 20 C.L.R., at pp. 436-437. 2. (1915) 20 C.L.R., at p. 442. Kidman has been mentioned in a number of subsequent cases in the Court. In Moss v. Donohoe [20] , Griffith C.J. referred, without discussion, to the dissenting opinion he had expressed in Kidman about the invalidity of a truly retroactive criminal law. The case was referred to, again without discussion, in R. v. Snow [21] , and Millner v. Raith [22] as authority for the proposition that the Parliament has authority to enact an ex post facto criminal law. Otherwise, subsequent references to the case in the Court have been merely passing and obiter or they have been in civil proceedings or deportation cases without reference to ex post facto criminal legislation. In none of those subsequent cases has there been any advertence to the possible relevance of Ch. III of the Constitution. In these circumstances, the subsequent cases in the Court add little to the weight of Kidman as authority on the effect of Ch. III. 1. (1915) 20 C.L.R. 615, at pp. 620-621. 2. (1917) 23 C.L.R. 256, at p. 265. 3. (1942) 66 C.L.R. 1, at p. 9. Kidman and the cases referred to in the preceding paragraph aside, there is no real authority on the question whether Ch. III precludes the enactment by the Parliament of an ex post facto criminal law. Cases concerning the constitutional law of the United Kingdom, Canada, New Zealand and other countries without comparable constitutional provisions vesting national judicial power can offer but limited assistance in this particular area. Such cases do, however, disclose a general perception of the inconsistency of ex post facto criminal legislation with accepted minimum standards of fairness in the administration of criminal justice see, e.g., Waddington v. Miah [23] ; Federal Republic of Germany v. Rauca [24] ; Reg. v. Kirk [25] . In view of the influence of the United States precedent upon the framing of our Constitution's vesting of judicial power exclusively in designated courts, the most helpful persuasive decisions from other jurisdictions are the cases in the United States Supreme Court, particularly Calder v. Bull, decided before the adoption of the Australian Constitution. Specific reference has already been made to them (see above). They support the proposition that ex post facto criminal legislation is inconsistent with the exclusive vesting of judicial power in courts under a constitution structured upon the separation of judicial from legislative and executive powers. Specific reference should also be made to the decision of the Judicial Committee of the Privy Council (Lord MacDermott, Lord Morris of Borth-y-Gest, Lord Guest, Lord Pearce and Lord Pearson) on appeal from the Supreme Court of Ceylon in Liyanage v. The Queen [26] . 1. [1974] 1 W.L.R. 683, at p. 694; [1974] 2 All E.R. 377, at p. 379. 2. (1982) 38 O.R. (2d) 705, at p. 717. 3. [1984] E.C.R., at p. 2718. 4. [1967] 1 A.C. 259. The impugned legislation in Liyanage had been purportedly enacted by the Parliament of Ceylon to deal with a past abortive coup d'état. It retrospectively widened the existing statutory offence of conspiring to wage war against the Queen. In that regard, however, it would seem not to have been an ex post facto criminal law in the sense explained above since the conduct which was retrospectively included in the offence would clearly have constituted some other criminal offence at the time when it occurred. The substantive retrospective operation of the legislation was plainly directed to the trial and punishment of the persons involved in the coup. Its "pith and substance" was summarized as follows [27] : "It legalised their imprisonment while they were awaiting trial. It made admissible their statements inadmissibly obtained during that period. It altered the fundamental law of evidence so as to facilitate their conviction. And finally it altered ex post facto the punishment to be imposed on them." The last sentence is a reference to the fact that the challenged legislation retrospectively prescribed a minimum penalty (ten years' imprisonment) and decreed compulsory forfeiture of property in the case of a conviction of conspiracy "to wage war against the Queen". 1. [1967] 1 A.C., at p. 290. As their Lordships recognized [28] , the Constitution of Ceylon differed from that of Australia and the United States of America in that it did not expressly vest national judicial power in designated courts. However, their Lordships considered [29] that other provisions of the Constitution of Ceylon manifested a similar intention that "judicial power shall be vested only in the judicature". In circumstances where the challenged legislation applied to acts which were, in any event, criminal when done, their Lordships' consideration was confined to the question whether legislation which retrospectively increased the penalty for past criminal conduct and which operated retrospectively to alter the course of particular proceedings was consistent with the exclusive vesting of judicial power in the judicature. In their Lordships' view, the answer to that question depended upon whether the effect of the legislation was that judicial power had been "usurped or infringed by the executive or the legislature". They went on to explain [30] : Section 29 (1) of the Constitution says: "Subject to the provisions of this Order Parliament shall have power to make laws for the peace order and good government of the Island." These words have habitually been construed in their fullest scope. Section 29 (4) provides that Parliament may amend the Constitution on a two-thirds majority with a certificate of the Speaker. Their Lordships however cannot read the words of section 29 (1) as entitling Parliament to pass legislation which usurps the judicial power of the judicature — e.g., by passing an Act of attainder against some person or instructing a judge to bring in a verdict of guilty against someone who is being tried — if in law such usurpation would otherwise be contrary to the Constitution. There was speculation during the argument as to what the position would be if Parliament sought to procure such a result by first amending the Constitution by a two-thirds majority. But such a situation does not arise here. In so far as any Act passed without recourse to section 29 (4) of the Constitution purports to usurp or infringe the judicial power it is ultra vires. 1. [1967] 1 A.C., at p. 286. 2. [1967] 1 A.C., at p. 287. 3. [1967] 1 A.C., at p. 289. The importance of the Judicial Committee's decision in Liyanage for present purposes lies in its clear recognition that the exclusive vesting of judicial power in the judicature under the written Constitution of a British Dominion has important and far-reaching implications in so far as the integrity of the judicial process and the exercise of judicial power are concerned. In that regard, the decision that the impugned legislation was invalid for the reason that it represented a legislative usurpation and infringement of judicial power underlined the difference between the situation under such a written Constitution defining the scope of legislative power and the situation in the United Kingdom where the Constitution is unwritten and legislative power is unconfined [31] . Their Lordships stated [32] that they were "not prepared to hold that every enactment in this field which can be described as ad hominem and ex post facto must inevitably usurp or infringe the judicial power". In the case before them, however, they had "no doubt" that the challenged legislation constituted "such interference" [33] . The decisive invalidating features of the legislation were the fact that it was directed against those involved in the abortive coup and the fact that it retrospectively prescribed a minimum term of imprisonment and automatic forfeiture of property. Their Lordships said [34] : The true nature and purpose of these enactments are revealed by their conjoint impact on the specific proceedings in respect of which they were designed, and they take their colour, in particular, from the alterations they purported to make as to their ultimate objective, the punishment of those convicted. These alterations constituted a grave and deliberate incursion into the judicial sphere. Quite bluntly, their aim was to ensure that the judges in dealing with these particular persons on these particular charges were deprived of their normal discretion as respects appropriate sentences. They were compelled to sentence each offender on conviction to not less than ten years' imprisonment, and compelled to order confiscation of his possessions, even though his part in the conspiracy might have been trivial. The position would obviously have been a fortiori if the conduct which the legislation retrospectively subjected to the minimum punishment of ten years' imprisonment had not even been an offence under the law of Ceylon at the time when it occurred. 1. [1967] 1 A.C., at p. 288. 2. [1967] 1 A.C., at p. 289. 3. [1967] 1 A.C., at p. 290. 4. [1967] 1 A.C., at pp. 290-291. It becomes necessary to weigh the considerations favouring acceptance of the general proposition that the effect of s. 71's exclusive vesting of the judicial power of the Commonwealth in Ch. III courts precludes the enactment by the Parliament of an ex post facto criminal law against any considerations militating against acceptance of that general proposition. On the one hand, considerations of fundamental principle and the implications of the Constitution's doctrine of the separation of judicial from legislative and executive powers dictate acceptance of the proposition. In that, they are reinforced by the general structure of the Constitution, by a number of specific constitutional provisions dealing with other matters and by judgments in United States cases in relation to the effect of the exclusive vesting of the judicial power of the United States in designated courts under the doctrine of separation of powers embodied in that country's Constitution. On the other hand, the bare decision in Kidman [35] , unsupported by relevant express statement or reasoning, stands in opposition supported only by the fact that it was uncritically and unquestioningly accepted in some later cases in which, as in Kidman , no reference at all was made to Ch. III. In these circumstances, the validity of the general proposition must be accepted and, to the extent that it denies it, Kidman must be overruled. 1. (1915) 20 C.L.R. 425. Some exclusions from the scope of Ch. III The cases establish that s. 71's vesting of the judicial power of the Commonwealth exclusively in Ch. III courts is subject to at least two exceptions in the area of the trial and punishment of unlawful conduct. The first exception relates to the power of each of the Houses of Parliament to punish for contempt or breach of privilege: see Reg. v. Richards; Ex parte Fitzpatrick and Browne [36] . The second relates to the powers of military tribunals to enforce military discipline by dealing, at least in some circumstances, with offences against military law committed by those who fall under military authority: see R. v. Bevan; Ex parte Elias and Gordon [37] . The first of those exceptions flows from the provisions of s. 49 of the Constitution which, unlike s. 51, were not expressly made subject to Ch. III: see Re Tracey; Ex parte Ryan [38] . The second is a largely pragmatic exception based on historical and practical considerations [39] . 1. (1955) 92 C.L.R. 157. 2. (1942) 66 C.L.R. 452. 3. (1989) 166 C.L.R., at p. 581. 4. (1989) 166 C.L.R., at pp. 581-583. The Court has not had occasion in the past to examine the extent to which Ch. III of the Constitution is applicable in relation to the trial and punishment of persons accused of crimes against international law. In so far as Australia's participation in the establishment and functioning of an international tribunal for the trial and punishment of such alleged crimes is concerned, the provisions of Ch. III would be inapplicable for the reason that the judicial power of the Commonwealth would not be involved. Australia's participation would be as a member State of the International Community and the judicial power involved would be the judicial power of that Community. The position is not so clear in a case where a local tribunal is purportedly vested with jurisdiction in relation to an alleged crime against international law. It may be arguable that, in such a case, the judicial power of the Commonwealth is not involved for so long as the alleged crime against international law is made punishable as such in the local court: see, e.g., Brierly, The Basis of Obligation in International Law (1958), p. 304. Alternatively, at least where violations of the laws and customs of war are alone involved, analogy with the disciplinary powers of military tribunals and largely pragmatic considerations might combine to dictate recognition of a special jurisdiction standing outside Ch. III. Those are, however, questions which it is unnecessary to answer for the purposes of the present case. The purported conferral of jurisdiction in the present case is upon Ch. III courts. Conduct which the Act makes punishable as a "war crime" is not punishable under the Act as an offence against international law. It is punishable as a distinct and independently defined indictable offence under the municipal law of the Commonwealth. Nor is that conduct confined to violations of international law. The conduct which the Act retroactively makes punishable as a crime under municipal law need not have been a crime against international law at the time it was committed. The elements of the indictable offence of a "war crime" are identified by ss. 6, 7, 9 and 10 of the Act without reference to international law. International law becomes relevant only if a defence is raised, under s. 17(2), that the conduct which constitutes a "war crime" under ss. 6 and 7 was "permitted by the laws, customs and usages of war" and "was not under international law a crime against humanity". That defence can be relied on only if "there is evidence of the existence of the facts" which constitute it (s. 17(4)). More important, even where there is such evidence, the validity of a defence under s. 17(2) is not to be determined by reference only to international law. Section 17(2) is expressly made subject to s. 16 which, subject to presently irrelevant qualifications, excludes (except as a plea of mitigation) the defence of government or superior orders. At the time of the alleged offences in the present case (i.e. prior to June 1943), it is clear beyond real argument that the stage had not been reached where customary international law unqualifiedly regarded a subordinate (as distinct from his State) as being individually guilty of a breach of the law and customs of war in circumstances where he had merely acted in obedience to superior orders which he was bound to obey under his national law: see, e.g., Wheaton's Elements of International Law, 6th ed. (1929), vol. II, pp. 1159-1160; Oppenheim, International Law, 5th ed. (1935), vol. II, pp. 453-454; the United States Rules of Land Warfare, (1940), Art. 347; British Manual of Military Law, 7th ed. (1929), Art. 443; Lauterpacht, The Law of Nations and the Punishment of War Crimes, The British Year Book of International Law 58, vol. XXI (1944), at pp. 69-74; Pal, Crimes in International Relations (1955), pp. 382-384; Röling, The Law of War and the National Jurisdiction since 1945, Recueil des Cours (1960-II), vol. 100, pp. 372-377; History of the War Crimes Commission, pp. 274-288; Stone, op. cit., p. 362; Detter De Lupis, The Law of War (1987), pp. 357-358; and, as regards Australia, the Manual of Military Law 1941 which recognized the defence (Ch. XIV, par. 443) but which was amended on 30 September 1944 to deny it. The importance of the express exclusion of a defence of superior orders is not so much that such a defence was available at international law at least in some circumstances at the time of an alleged offence under the Act. It is that the legislative exclusion of the defence — like the legislative indifference to the question whether, at the time of an alleged offence, customary international law directly established or made punishable the criminality of individuals for breaches of the law of war (see above) — underlines the fact that what the Act is concerned to make punishable is a new crime against the law of the Commonwealth which the Act itself retrospectively creates. If the elements of an indictable offence as defined by ss. 6, 7, 9 and 10 are present and the qualified defence under s. 17(2) is not raised or, if raised, does not prevail, it matters not for the purposes of the Act whether the accused was or was not personally guilty of a crime under international law. In these circumstances, the jurisdiction purportedly conferred by the Act is in relation to crimes under Commonwealth law. The consideration that the alleged conduct was not criminal under international law when it occurred may, but will not necessarily, constitute a defence. When conduct is punishable under the Act, it is punishable, in the ordinary municipal courts, solely by reason of the fact that it is an indictable offence under the municipal law of this country. The trial and punishment within Australia of an Australian citizen or resident accused of such an indictable offence under municipal law is part of the ordinary judicial power of the Commonwealth. That being so, the Act is necessarily beyond the competence of the Parliament to the extent that it is properly to be seen as an ex post facto criminal law in the sense that I have explained. Is the Act an ex post facto criminal law? As a general rule, the question whether an enactment of a particular legislature declaring that a past act is punishable as a crime is an ex post facto criminal law will fall to be determined by reference to the system of law which is subject to the legislative powers of that legislature. The fact that the particular act was a crime under some other system of law at the time when it was done will ordinarily be irrelevant. In a federation such as Australia where the national law encompasses the enactments of a number of different legislatures with authority in designated areas, the question whether a law of a particular legislature is an ex post facto criminal law will ordinarily fall to be determined by reference to the particular area of the national law in relation to which that particular legislature has authority. Thus, it will ordinarily be irrelevant that an act retrospectively made criminal under an enactment of the Parliament of the Commonwealth was already a crime under an enactment of one or more of the State Parliaments. It is, however, possible to envisage circumstances in which the interaction between Commonwealth and State laws could give rise to difficulty in determining whether a law of the Parliament which retrospectively declared a past act to be punishable as a crime was properly to be seen as an ex post facto criminal law. If, for example, the Commonwealth Parliament were, pursuant to a reference (under s. 51(xxxvii) of the Constitution) or a request (under s. 51(xxxviii)) by the Parliaments of all the States, to enact a national code in relation to a certain type of crime which retrospectively replaced earlier Commonwealth and State laws, it would be arguable that the code was not an ex post facto criminal law to the extent that it retrospectively made criminal under Commonwealth law an act which had been, at the time when it was done, a corresponding crime punishable by a corresponding or greater penalty under the law of the State where it was done. It is, however, unnecessary to pursue such questions for the purposes of the present case. It is not suggested that the acts committed outside Australia to which the Act applies were crimes under any law of this country — Commonwealth, State or Territory — at the time when they were done. In so far as the Act applies to past conduct within Australia, the fact that elements of that conduct (i.e. the s. 6(1) elements) were punishable as different, and in many cases much less heinous, crimes under the law of the State or Territory in which the conduct occurred is clearly insufficient to prevent the Act's retroactive declaration that the past conduct (i.e. the s. 6(1) elements aggravated by the s. 7 elements) is punishable under Commonwealth law as "war crimes" from being an ex post facto criminal law in the relevant sense. Indeed, no submission to the contrary was advanced on behalf of those who argued in support of the validity of the Act. Nor is it necessary to consider whether a law of the Parliament which confers retrospective jurisdiction upon a municipal court to punish a past crime under international law (as such) is an ex post facto criminal law in the relevant sense. As has been seen, the Act does not confine the conduct which it retrospectively makes punishable to acts which were, when they occurred, crimes under international law. Nor does it make that conduct punishable as a crime under international law. What the Act does is to declare past conduct, which may or may not have been a crime under international law, to have been a crime under Commonwealth law which is punishable in its capacity as such. In those regards, the Act differs from the War Crimes Act Cth as enacted in 1945 and from legislation enacted in some other countries for the punishment of past acts as war crimes: see, e.g., the Canadian Criminal Code, s. 7(3.71)-(3.76); the Netherlands Penal Code, Art. 27a; and cf. the War Crimes Act 1991 U.K., s. 1. It also makes inapplicable statements in other cases or in learned writings to the effect that an offence is being punished in its capacity as a contravention of international law (see, e.g., Baxter, The Municipal and International Law Basis of Jurisdiction over War Crimes, The British Year Book of International Law, vol. XXVIII (1951) 382, at pp. 382-387) and that, for that reason, the retrospective conferral of jurisdiction is not a retroactive or ex post facto criminal law: see, e.g., Reg. v. Finta [40] ; and, as to the recent United Kingdom legislation, Steiner, "Prosecuting War Criminals in England and in France", [1991] Criminal Law Review 180, at p. 186. 1. (1989) 61 D.L.R. (4th) 85, at pp. 94-95. It follows that the operative provisions of the Act constitute an ex post facto criminal law and are inconsistent with s. 71's exclusive vesting of the judicial power of the Commonwealth in Ch. III courts. Accordingly, s. 9 of the Act is invalid in its application to the information laid by the second defendant against the plaintiff and the question reserved for consideration should be answered in the affirmative. There is one further matter which I would briefly mention. It is that the enormity of the acts of barbarism perpetrated before and in the course of the 1939-1945 war in Europe should not be allowed to disguise the nature of the critical question addressed in this judgment. The charges against the plaintiff relate to acts allegedly done in the Ukraine. It is common ground that those acts, when and if done, did not contravene any applicable law of the Commonwealth. Yet the charges against the plaintiff are not of breaches of international law or of breaches of the system of law which applied in the Ukraine during the wartime occupation of that country. They are of crimes against the law of the Commonwealth. The only arguable basis of those charges is a statutory provision (the Act, s. 9(1)) which purported to declare in 1988 that a "person is guilty of an indictable offence" against the law of the Commonwealth if, between 1939 and 1945, he had done an act which was not, when done, an offence against the law of the Commonwealth at all. The critical question upon the answer to which this judgment turns is ultimately one of abstract constitutional law. It is whether the Commonwealth Parliament possesses power to legislate that a "person is guilty" of a crime against Commonwealth law if, in the past, he has done some specified thing which was not, when done, such a crime. That question must, in my view, be answered in the negative for the reason that a law which declares that a person "is guilty" of a crime against a law of the Commonwealth if he has done an act which did not, when done, in fact contravene any such law is inconsistent with Ch. III of the Constitution. Both in substance and in form, the central operation of the Act is as such a legislative declaration of criminal guilt. It prohibits nothing, prescribes no rule of conduct and is incapable of being contravened since, by its terms, it is inapplicable to acts committed after its enactment. As I have endeavoured to explain, it is not to the point that the Act identifies a "person" whom it declares to be "guilty" of past crimes against the law of the Commonwealth not by name but, in the case of the plaintiff, by reference to whether, within a long past period and in another country, he did an alleged act which was not such a crime when done and which has never, if done where it was allegedly done, been prohibited by any applicable law of the Commonwealth, including the Act. Nor is it to the point that the operation of the Act to declare that such a person "is guilty" of such a past crime is obscured by the requirement of a trial to determine whether a particular accused is in fact such a person. What is to the point for the purposes of the present case is the combined effect of two propositions which are basic to the criminal jurisprudence of this country. The first of those propositions is almost a truism. It is that criminal guilt, under our system of law, means being guilty of a contravention of the requirements of a then existing and applicable penal law: a crime is, as Blackstone wrote (see above), "an act committed, or omitted, in violation of a public law, either forbidding or commanding it". That proposition lies at the heart of Lord Atkin's comments in the Proprietary Articles Case (see above) when he wrote (emphasis added) that criminal law "connotes only the quality of such acts or omissions as are prohibited under appropriate penal provisions" and that the "criminal quality of an act" cannot "be discovered by reference to any standard but one: Is the act prohibited with penal consequences?" The second of those two propositions is that the function of determining whether a person is in fact guilty of a crime against a law of the Commonwealth is a function which appertains exclusively to, and which cannot be excluded from, the judicial power which our Constitution vests solely in the courts which it designates. That being so, it is beyond the competence of the Parliament to declare, as s. 9(1) of the Act purports to do, that a "person is guilty" of a crime against a law of the Commonwealth by reason of having committed a past act which did not, when done, contravene any applicable Commonwealth law and was therefore not in fact such a crime. Dawson J. Whilst I have never been able to accept the view that the external affairs power (Constitution, s. 51(xxix)) extends to matters of a purely domestic character merely because they form the subject of an international treaty or may otherwise be shown to be of international concern, I have never doubted that the power extends to places, persons, matters or things physically external to Australia. The word "affairs" is imprecise, but is wide enough to cover places, persons, matters or things. The word "external" is precise and is unqualified. If a place, person, matter or thing lies outside the geographical limits of the country, then it is external to it and falls within the meaning of the phrase "external affairs". Whether the external affairs power was always construed in this way is something which need not trouble me. Section 3 of the Statute of Westminster 1931 Imp. declares and enacts that the Parliament of a Dominion has full power to make laws having extraterritorial operation. That section, amongst others, was adopted by the Statute of Westminster Adoption Act 1942 Cth, the adoption being given effect from 3 September 1939. As Menzies J. pointed out in Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [41] : [Section 3] is not an amending section; it is an amplifying section directed to negativing a conception which in the course of time had perhaps come to be regarded as a misconception as to the nature of a power to legislate granted by the Imperial Parliament to a Dominion Parliament without any affirmation that the power granted extended to the making of laws with extra-territorial operation. The conception or misconception, which was never anything but obscure, was, in any event, largely disposed of in 1932 by the Privy Council, without any reliance upon the Statute of Westminster, in Croft v. Dunphy [42] when it said of the Canadian Parliament: Once it is found that a particular topic of legislation is among those upon which the Dominion Parliament may competently legislate as being for the peace, order and good government of Canada or as being one of the specific subjects enumerated in s. 91 of the British North America Act, their Lordships see no reason to restrict the permitted scope of such legislation by any other consideration than is applicable to the legislation of a fully Sovereign State. So far as the Australian States were concerned the power contained in each of their Constitutions to legislate for the peace, order and good government of the State (or the equivalent power) continued to be construed as requiring that State legislation with an extraterritorial operation be connected, not too remotely, with the enacting State. That test, however, was never narrowly applied and its operation must now be governed by s. 2(1) of the Australia Act 1986 Cth and s. 2(1) of the Australia Act 1986 U.K., each of which declares and enacts that each State has full power to make laws for the peace, order and good government of the State with extraterritorial operation: see Union Steamship Co. of Australia Pty. Ltd. v. King [43] . 1. (1959) 103 C.L.R. 256, at p. 300. 2. [1933] A.C. 156, at p. 163. 3. (1988) 166 C.L.R. 1, at p. 14. Whatever the position may have been or may be with the States, since the Statute of Westminster extended to Australia there is no doubt at all that the Constitution bestows upon the Commonwealth Parliament full and complete power to legislate extraterritorially with respect to those matters enumerated in s. 51 which, of their nature, can be the subject of extraterritorial legislation. This was the view of Windeyer J., with which I respectfully agree, in Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [44] : If the Commonwealth Parliament were to legislate gratuitously in respect of foreign persons in foreign territory, in one of the ways fancifully suggested in argument, an Australian court could not hold the legislation was invalid — provided, always, that it was in respect of one of the matters in s. 51. Vis-à-vis the States, the competence of the Commonwealth Parliament remains limited and the Statute of Westminster does not affect this. But in respect of the matters set out in s. 51 the Parliament is now in reality fully sovereign, except perhaps in a theoretical unrealistic sense satisfying to convinced Austinians who see the Statute of Westminster as a repealable enactment of the Imperial Parliament. Whatever limitations international comity may impose are the consequences of considerations of political propriety and of the limitations of political power, not of legal capacity. 1. (1959) 103 C.L.R., at pp. 306-307. While the old conception or misconception persisted, it may have suggested a restrictive interpretation of the external affairs power in its application to circumstances external to Australia. This was recognized by Jacobs J. when he said in New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [45] : It is true that the operation of the [external affairs] power may have been limited in 1900 by the concept that Australia, lacking sovereignty, could legislate only for its territory; but that limitation, if it existed, did not alter the meaning of the words. That limitation, while it was thought to exist, may have required legislation extending to matters external to Australia to exhibit, even under the external affairs power, some relationship between those matters and Australia in order to provide the necessary territorial connexion. But now that the extraterritorial legislative power of the Commonwealth has been put beyond doubt, there is no justification for reading down the words "external affairs" so as to require any such relationship. Externality alone is sufficient. 1. (1975) 135 C.L.R. 337, at p. 497. That means that, subject to any express or implied constitutional prohibitions or limitations, the external affairs power is capable by itself of providing the authority for Commonwealth legislation extending to circumstances which are geographically external to Australia, without reference to the other legislative powers conferred by s. 51 of the Constitution. That must be the case, for it is beyond question that s. 51(xxix) stands as an independent legislative power. A contrary suggestion was shortly disposed of by Latham C.J. in R. v. Burgess; Ex parte Henry [46] : It has been argued that sec. 51(xxix.) should be construed as giving power to make laws only with respect to some external aspect of the other subjects mentioned in sec. 51. Prima facie it would be as reasonable to argue that any other single power conferred by sec. 51 is limited by reference to all the other powers conferred by that section — which is really an unintelligible proposition. There is no reason whatever why placitum xxix. should not be given its natural and proper meaning, whatever that may be, as an independent express legislative power. It is true that in 1936 Latham C.J. viewed s. 51(xxix) as being incapable of providing general extraterritorial legislative power because of covering cl. 5 of the Constitution. Covering cl. 5 relevantly provides that the laws of the Commonwealth shall be in force on all British ships whose first port of clearance and whose port of destination are in the Commonwealth. But if in so providing covering cl. 5 reflected a limited conception of Commonwealth extraterritorial legislative power, it has been overtaken by s. 3 of the Statute of Westminster. In any event, covering cl. 5 appears to have been intended to extend Commonwealth domestic laws to the ships designated, upon the customary basis that they were to be treated as part of Australia, rather than to authorize a limited exercise of extraterritorial legislative power: see Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd [47] . 1. (1936) 55 C.L.R. 608, at p. 639. 2. (1959) 103 C.L.R., at p. 309. I must confess that I have never entirely understood why the words "peace, order, and good government" should have been thought to impose a territorial limitation upon State legislative power, when they otherwise have no practical limiting effect. Indeed, Windeyer J. in Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [48] , thought that the origin of the limitation was otherwise. Nevertheless, it must now be accepted that the limits upon the extraterritorial operation of a State law — now subject to the Australia Act — inhere in the grant of power to the legislature: Port MacDonnell Professional Fishermen's Association Inc. v. South Australia [49] . Be that as it may, in the absence of any territorial limitation upon the Commonwealth's legislative powers, it is clear that the words "peace, order, and good government of the Commonwealth" in s. 51 of the Constitution impose no practical limits upon those powers: see Reg. v. Burah [50] ; Hodge v. The Queen [51] ; Powell v. Apollo Candle Co. [52] ; Riel v. The Queen [53] . These cases were examined by this Court recently in Union Steamship Co. of Australia Pty. Ltd. v. King, a case which raised questions concerning the extent of the power of the Parliament of New South Wales to legislate for the peace, welfare and good government of the State. This Court said [54] : These decisions and statements of high authority demonstrate that, within the limits of the grant, a power to make laws for the peace, order and good government of a territory is as ample and plenary as the power possessed by the Imperial Parliament itself. That is, the words "for the peace, order and good government" are not words of limitation. They did not confer on the courts of a colony, just as they do not confer on the courts of a State, jurisdiction to strike down legislation on the ground that, in the opinion of a court, the legislation does not promote or secure the peace, order and good government of the colony. Just as the courts of the United Kingdom cannot invalidate laws made by the Parliament of the United Kingdom on the ground that they do not secure the welfare and the public interest, so the exercise of its legislative power by the Parliament of New South Wales is not susceptible to judicial review on that score. Whether the exercise of that legislative power is subject to some restraints by reference to rights deeply rooted in our democratic system of government and the common law (see Drivers v. Road Carriers [55] ; Fraser v. State Services Commission [56] ; Taylor v. New Zealand Poultry Board [57] ), a view which Lord Reid firmly rejected in Pickin v. British Railways Board [58] , is another question which we need not explore. The effect of the authorities is that, save possibly for quite extraordinary circumstances, it is for the Commonwealth Parliament alone to judge whether legislation which otherwise falls within power is for the peace, order and good government of the Commonwealth: Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [59] ; Pearce v. Florenca [60] . 1. (1959) 103 C.L.R., at p. 308. 2. (1989) 168 C.L.R. 340, at p. 370. 3. (1878) 3 App. Cas. 889. 4. (1883) 9 App. Cas. 117. 5. (1885) 10 App. Cas. 282. 6. (1885) 10 App. Cas. 675. 7. (1988) 166 C.L.R., at p. 10. 8. [1982] 1 N.Z.L.R. 374, at p. 390. 9. [1984] 1 N.Z.L.R. 116, at p. 121. 10. [1984] 1 N.Z.L.R. 394, at p. 398. 11. [1974] A.C. 765, at p. 782. 12. (1959) 103 C.L.R., at p. 308. 13. (1976) 135 C.L.R. 507, at pp. 515-516. The conclusion that the reach of the external affairs power extends to all places, persons, matters or things geographically external to Australia is, I think, supported by recent authority. In the Seas and Submerged Lands Case [61] , Barwick C.J. expressed the view that the "power extends to any affair which in its nature is external to the continent of Australia and the island of Tasmania subject always to the Constitution as a whole". A similar view was expressed by Mason J. [62] , and by Jacobs J. [63] . It would seem that Murphy J. placed at least as wide a construction upon the reach of the external affairs power outside Australia as Barwick C.J., Mason and Jacobs JJ. Subsequently, in The Commonwealth v. Tasmania (the Tasmanian Dam Case) [64] , Murphy J. expressed the view that the external affairs power authorizes the Parliament to make laws which govern conduct outside Australia: see also Viro v. The Queen [65] . 1. (1975) 135 C.L.R., at p. 360. 2. (1975) 135 C.L.R., at p. 470. 3. (1975) 135 C.L.R., at p. 497. 4. (1983) 158 C.L.R. 1, at p. 170. 5. (1978) 141 C.L.R. 88, at p. 162. In Koowarta v. Bjelke-Petersen [66] , Gibbs C.J. said, referring to the Seas and Submerged Lands Case: three members of the Court, Barwick C.J., Mason and Jacobs JJ., relied on the ground that the power given by s. 51(xxix) was not limited to authorizing laws with respect to Australia's relationships with foreign countries, but extended to any matter or thing situated or done outside Australia [67] . It is unnecessary to consider whether the words of par. (xxix) can have this dual operation, i.e. whether the phrase "external affairs" can be used to mean matters outside the Commonwealth as well as matters involving a relationship between Australia and other countries. In the same case [68] , Mason J. reiterated his view, which he said had been accepted in the Seas and Submerged Lands Case, that "the power extends to matters and things, and I would say, persons, outside Australia". In Robinson v. Western Australian Museum [69] , Barwick C.J. repeated his opinion that "the Commonwealth may take as the subject matter of its law some fact or circumstance which is actually outside the territorial limits of the Commonwealth". Mason J. also repeated his view [70] . And in the Tasmanian Dam Case [71] , Gibbs J. adopted as accurate the paraphrase of par. (xxix) suggested by Stephen J. in Koowarta v. Bjelke-Petersen [72] , namely " such of the public business of the national government as relates to other nations or other things or circumstances outside Australia". 1. (1982) 153 C.L.R. 168, at p. 190. 2. (1975) 135 C.L.R., at pp. 360, 470-471, 497. 3. (1982) 153 C.L.R., at p. 223. 4. (1977) 138 C.L.R. 283, at p. 294. 5. (1977) 138 C.L.R., at p. 335. 6. (1983) 158 C.L.R., at p. 97. 7. (1982) 153 C.L.R., at p. 211. In perceiving that the Constitution requires the exclusion of domestic matters from the ambit of the external affairs power, I have elsewhere pointed to the division of legislative power between the Commonwealth and the States and have observed that, if international concern over entirely domestic matters were sufficient to bring those matters within the external affairs power, par. (xxix) would have the potential to obliterate the division which s. 51 was intended to effect. To construe par. (xxxix) in that way would be to disregard entirely its constitutional setting. But if, as I think to be the case, it is necessary to have regard to the scheme of the Constitution in construing the external affairs power, the result is different with regard to circumstances external to Australia. For although the sovereignty of the Australian nation is divided internally between the Commonwealth and the States, there is no division with respect to matters which lie outside Australia. There the sovereignty of the nation is the sovereignty of the Commonwealth which may act as if it were a unitary state without regard to the "conceptual duality" within Australia to which Stephen J. referred to in the Seas and Submerged Lands Case [73] . There is no corresponding capacity on the part of the States, either singly or together. Indeed, any limitation upon the power of the Commonwealth to legislate with respect to matters outside the country would leave a gap in the totality of legislative power which the Constitution bestows upon the Commonwealth and the States. An interpretation of the Constitution which denies the completeness of Australian legislative power is unacceptable in terms of constitutional theory and practice. Apart from express or implied constitutional prohibitions or limitations, it is not to be contemplated that there are laws which no Parliament has the power to pass: Attorney-General (Ontario) v. Attorney-General (Canada) [74] ; Colonial Sugar Refining Co. Ltd. v. Attorney-General (Cth) [75] ; Smith v. Oldham [76] ; Reg. v. Duncan; Ex parte Australian Iron and Steel Pty. Ltd. [77] . 1. (1975) 135 C.L.R., at p. 458. 2. [1912] A.C. 571, at pp. 583-584. 3. (1912) 15 C.L.R. 182, at pp. 214-215. 4. (1912) 15 C.L.R. 355, at pp. 360-361, 365. 5. (1983) 158 C.L.R. 535, at pp. 590-591. The War Crimes Act 1945 Cth as amended by the War Crimes Amendment Act 1988 Cth ("the Act") extends to serious crimes committed in the course of hostilities in a war or in the course of an occupation of territory arising out of a war. That appears from s. 7. "War" is restricted by s. 5 to a war in the period beginning on 1 September 1939 and ending on 8 May 1945 in so far as it occurred in Europe. Under s. 7 war crimes also extend to serious crimes committed in pursuing a policy associated with the conduct of war or with an occupation or on behalf of, or in the interests of, a power conducting a war or engaged in an occupation. The Act does, therefore, contemplate war crimes committed within Australia. Difficult questions could arise in relation to those provisions which deal with war crimes committed within Australia, particularly as war crimes are defined so as to include acts which are criminal acts under the laws of the States, but those questions do not arise in this case. The war crimes alleged against the plaintiff, who is now an Australian citizen and a resident of South Australia, were said to have been committed in the Ukraine between 1 September 1942 and 31 May 1943. It was not argued that the invalidity of any provisions relating to the commission of war crimes within Australia would, by reason of inseverability or otherwise, affect the validity of the provisions relating to war crimes committed outside Australia if those provisions are otherwise valid. The question reserved for the consideration of the Court is whether s. 9 of the Act is invalid in its application to the information laid against the plaintiff. That information relates entirely to events which took place in Europe almost fifty years ago. Section 9 of the Act relevantly provides: (1) A person who: (a) on or after 1 September 1939 and on or before 8 May 1945; and (b) whether as an individual or as a member of an organisation; committed a war crime is guilty of an indictable offence against this Act. Section 11 provides that a person shall not be charged with an offence against the Act unless he is an Australian citizen or a resident of Australia or of an external Territory, although s. 5 defines "person" as meaning a natural person whether or not the person is or has ever been an Australian citizen or a resident of Australia. A serious crime is, under s. 7, a war crime if it was committed in the circumstances to which I have already referred. Section 6 provides: (1) An act is a serious crime if it was done in a part of Australia and was, under the law then in force in that part, an offence, being: (a) murder; (b) manslaughter; (c) causing grievous bodily harm; (d) wounding; (e) rape; (f) indecent assault; (g) abduction, or procuring, for immoral purposes; (h) an offence (in this paragraph called the "variant offence") that would be referred to in a preceding paragraph if that paragraph contained a reference to: (i) a particular intention or state of mind on the offender's part; or (ii) particular circumstances of aggravation; necessary to constitute the variant offence; (j) an offence whose elements are substantially the same as the elements of an offence referred to in any of paragraphs (a) to (h), inclusive; or (k) an offence of: (i) attempting or conspiring to commit; (ii) aiding, abetting, counselling or procuring the commission of; or (iii) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the commission of; an offence referred to in any of paragraphs (a) to (j), inclusive. (2) In determining for the purposes of subsection (1) whether or not an act was, under the law in force at a particular time in a part of Australia, an offence of a particular kind, regard shall be had to any defence under that law that could have been established in a proceeding for the offence. (3) An act is a serious crime if: (a) it was done at a particular time outside Australia; and (b) the law in force at that time in some part of Australia was such that the act would, had it been done at that time in that part, be a serious crime by virtue of subsection (1). (4) The deportation of a person to, or the internment of a person in, a death camp, a slave labour camp, or a place where persons are subjected to treatment similar to that undergone in a death camp or slave labour camp, is a serious crime. (5) Each of the following is a serious crime: (a) attempting or conspiring to deport or intern a person as mentioned in subsection (4); (b) aiding, abetting, counselling or procuring the deportation or internment of a person as so mentioned; (c) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the deportation or internment of a person as so mentioned. (6) For the purposes of subsections (3), (4) and (5), the fact that the doing of an act was required or permitted by the law in force when and where the act was done shall be disregarded. In addition to setting out in sub-s. (1) the circumstances in which a war crime may be committed, s. 7 in sub-ss. (2) and (3) provides: (2) For the purposes of subsection (1), a serious crime was not committed: (a) in the course of hostilities in a war; or (b) in the course of an occupation; merely because the serious crime had with the hostilities or occupation a connection (whether in time, in time and place, or otherwise) that was only incidental or remote. (3) A serious crime is a war crime if it was: (a) committed: (i) in the course of political, racial or religious persecution; or (ii) with intent to destroy in whole or in part a national, ethnic, racial or religious group, as such; and (b) committed in the territory of a country when the country was involved in a war or when territory of the country was subject to an occupation. Under s. 16 it is no defence for a person charged under the Act that he or she acted upon superior orders, although that may be taken into account in determining the proper sentence, but under s. 17(2) it is a defence that the act alleged against a person charged with an offence was permitted by the laws, customs and usages of war and was not under international law a crime against humanity. It was submitted that s. 17 operates so as to confine a war crime under ss. 6 and 7 to a crime which was not permitted by the laws, customs and usages of war or which was under international law a crime against humanity. However, it is clear that s. 17 does not operate in this way. It provides a defence only. Indeed, under s. 17(4) a defendant is not entitled to rely on a defence under s. 17(2) unless there is evidence of the existence of the facts constituting the defence. If there is evidence, under s. 17(5) the onus of displacing the defence lies upon the prosecution in accordance with the criminal standard of proof. The provisions of the Act which I have set out show that there are not inconsiderable problems of interpretation arising from the way in which it was framed, but it is unnecessary for me to tackle those problems for it is quite apparent that, in its application to the plaintiff, the Act deals entirely with places, persons, matters or things which were physically external to Australia. For the reasons which I have given, I consider that the external affairs power extends the legislative reach of the Commonwealth Parliament to those places, persons, matters or things because of their externality and nothing further is required to bring them within the description of "external affairs". Nor do I think that the externality of the circumstances with which the Act relevantly deals is affected by its retrospectivity. Those circumstances, as exemplified by the allegations against the plaintiff, remain physically external to Australia, notwithstanding that they occurred in the past. For that reason they fall within the ambit of the external affairs power. The submission that the retrospectivity of the Act otherwise places it beyond the competence of the Commonwealth Parliament must, in my view, fail. The Act is truly retrospective — that is to say, retroactive — in its application to past events. Actions in the past which were, as a matter of domestic law, not criminal, are made criminal. The Act is, therefore, an ex post facto law. Ex post facto laws may be either civil or criminal, but the description is frequently used to refer only to criminal laws, perhaps because the creation of crimes ex post facto is, for good reason, generally considered a great deal more objectionable than retrospective civil legislation. Blackstone in his Commentaries, 17th ed. (1830), vol. I, pp. 45-46, after referring to Caligula's method of prescribing laws by writing them in very small characters and hanging them up on high pillars in order to ensnare the people, said: There is still a more unreasonable method than this, which is called making of laws ex post facto: when after an action (indifferent in itself) is committed, the legislator then for the first time declares it to have been a crime, and inflicts a punishment upon the person who has committed it. Here it is impossible that the party could foresee that an action, innocent when it was done, should be afterwards converted to guilt by a subsequent law: he had therefore no cause to abstain from it; and all punishment for not abstaining must of consequence be cruel and unjust. All laws should be therefore made to commence in futuro, and be notified before their commencement; which is implied in the term " prescribed ". But Blackstone was not denying the capacity of Parliament to pass ex post facto laws, however undesirable they may be: see Commentaries , 16th ed. (1825), vol. I, p. 90. The resistance of the law to retrospectivity in legislation is to be found in the rule that, save where the legislature makes its intention clear, a statute ought not be given a retrospective operation where to do so would be to attach new legal consequences to facts or events which occurred before its commencement: Fisher v. Hebburn Ltd. [78] , per Fullagar J.; see also Maxwell v. Murphy [79] ; Geraldton Building Co. Pty. Ltd. v. May [80] ; Rodway v. The Queen [81] . However, the injustice which might be inflicted by construing an enactment so as to give it a retrospective operation may vary according to its subject matter. Indeed, justice may lay almost wholly upon the side of giving remedial legislation a retrospective operation where that is possible: see George Hudson Ltd. v. Australian Timber Workers' Union [82] . With legislation of that character, if the ordinary rule be couched in terms of a presumption against retrospectivity, it must, at best, be a weak presumption: see Doro v. Victorian Railways Commissioners [83] . With a criminal law, where the injustice of giving it an ex post facto operation will ordinarily be readily apparent, the presumption must be at its strongest. 1. (1960) 105 C.L.R. 188, at p. 194. 2. (1957) 96 C.L.R. 261, at p. 267. 3. (1977) 136 C.L.R. 379. 4. (1990) 169 C.L.R. 515, at p. 518. 5. (1923) 32 C.L.R. 413, at p. 434. 6. [1960] V.R. 84, at pp. 85-86. However, the ex post facto creation of war crimes may be seen to be justifiable in a way that is not possible with other ex post facto criminal laws, particularly where the conduct proscribed would have been criminal conduct had it occurred within Australia. The wrongful nature of the conduct ought to have been apparent to those who engaged in it even if, because of the circumstances in which the conduct took place, there was no offence against domestic law. And, of course, if the conduct amounted to genocide or a crime against humanity, that comment would be the stronger. This justification for a different approach with respect to war crimes is reflected in the International Covenant on Civil and Political Rights to which Australia became a signatory on 18 December 1972. Article 15(1) of that Covenant forbids the ex post facto creation of criminal offences, but Art. 15(2) provides: "Nothing in this article shall prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of law recognized by the community of nations." Because of the view which I take of the external affairs power, I have no need to enter upon the question whether before 1945 genocide or crimes against humanity constituted offences under customary international law; it is sufficient to observe that, even if they did not, the wrongful nature of the conduct would nevertheless have been plainly evident. War crimes of the kind created by the Act simply could not, in any civilized community, have been described as innocent or blameless conduct merely because of the absence of proscription by law. In any event, the intention of the legislature that the Act should have a retrospective operation could not have been more clearly expressed. There is no room for the application of the rule or presumption against giving the legislation a retrospective construction. There is ample authority for the proposition that the Commonwealth Parliament may in the exercise of its legislative powers create retrospective laws, including criminal laws with an ex post facto operation. I have earlier referred to the authorities which establish that the power of the Parliament to make laws for the peace, order and good government of the Commonwealth is, in constitutional terms, a sovereign legislative power with respect to the matters enumerated in s. 51: see also Ibralebbe v. The Queen [84] . And sovereignty necessarily involves the power to legislate retrospectively. Whatever the objections which might be raised to ex post facto laws — and as the passage cited from Blackstone shows, they are considerable — there can be no doubt about the capacity of Parliament to pass them. 1. [1964] A.C. 900, at p. 923. The power of the Parliament to create criminal offences is incidental to the power to legislate with respect to the specific matters enumerated in pars (i) to (xxxviii) of s. 51. And s. 51(xxxix) gives to the Parliament express power to legislate with respect to matters incidental to the execution of any power vested by the Constitution in Parliament. In R. v. Kidman [85] , Griffith C.J. thought that, for a matter to be incidental to the execution of a power, it must relate to its present execution. An ex post facto law could not, in his view, be incidental to the execution of a power because it operated upon a past event and not with reference to its present execution. This view (which, with respect, is difficult to understand since past events may have a direct bearing upon a present exercise of legislative power as this case demonstrates) was rejected by the rest of the Court. They were of the view that it is incidental to the power of Parliament to make laws with respect to the matters confided to it, to express its condemnation of past actions falling within those matters by attaching criminal penalties to them. As Higgins J. said [86] : We have not been referred to any words in the Constitution which point to any limitation of the plenary powers of the Federal Parliament so long as the Parliament keeps within the ambit of the subjects of legislation specifically assigned to it. The British Parliament, admittedly, has power to make its laws retroactive; and I know of no instance in which a Legislature created by the British Parliament has been held to have overstepped its powers by making legislation retroactive. There are plenty of passages that can be cited showing the inexpediency, and the injustice, in most cases, of legislating for the past, of interfering with vested rights, and of making acts unlawful which were lawful when done; but these passages do not raise any doubt as to the power of the Legislature to pass retroactive legislation, if it see fit. The maxim runs: Nova constitutio futuris formam imponere debet, non proeteritis. The word used is "debet," not "potest." 1. (1915) 20 C.L.R. 425, at p. 434. 2. (1915) 20 C.L.R., at p. 451. In R. v. Kidman, legislation creating the indictable offence of conspiring to defraud the Commonwealth was held to be valid, notwithstanding that the legislation was given an ex post facto operation by its application to conspiracies committed before the legislation came into force. The correctness of the decision has never, so far as I am aware, been subsequently doubted: see R. v. Snow [87] ; Ex parte Walsh and Johnson; In re Yates [88] ; Millner v. Raith [89] ; Nelungaloo Pty. Ltd. v. The Commonwealth [90] ; Australian Communist Party v. The Commonwealth [91] ; University of Wollongong v. Metwally [92] . 1. (1917) 23 C.L.R. 256, at p. 265. 2. (1925) 37 C.L.R. 36, at pp. 81, 127. 3. (1942) 66 C.L.R. 1. 4. (1948) 75 C.L.R. 495, at pp. 503-504. 5. (1951) 83 C.L.R. 1, at p. 172. 6. (1984) 158 C.L.R. 447, at pp. 456, 461, 480, 484. In the United States both Congress and the State legislatures are prohibited by the Constitution from passing any bill of attainder or ex post facto law: Art. I, ss. 9 and 10. The case law in the United States has grown up around bills of attainder rather than ex post facto laws generally. This would seem to be due in part to the restrictive interpretation given to ex post facto laws within the meaning of the prohibition, whereby they are confined to criminal laws: Calder v. Bull [93] . Moreover, bills of attainder have been held to extend beyond anything which could be so described historically. They have been held to embrace not only what would be described, as a matter of history, as bills of pains and penalties but also as extending to any kind of punishment legislatively inflicted. A bill of attainder in England imposed the penalty of death, forfeiture of land and possessions and "corruption of blood" whereby the heirs of the person attainted were prevented from inheriting his property. A bill of pains and penalties inflicted lesser punishment, involving forfeiture of property and, on occasions, corporal punishment less than death. The wide interpretation of the bill of attainder clause in the United States led the Supreme Court to observe in United States v. Brown [94] : In 1810, Chief Justice Marshall, speaking for the Court in Fletcher v. Peck [95] , stated that "[a] bill of attainder may affect the life of an individual, or may confiscate his property, or may do both." This means, of course, that what were known at common law as bills of pains and penalties are outlawed by the Bill of Attainder Clause. The Court's pronouncement therefore served notice that the Bill of Attainder Clause was not to be given a narrow historical reading (which would exclude bills of pains and penalties), but was instead to be read in light of the evil the Framers had sought to bar: legislative punishment, of any form or severity, of specifically designated persons or groups. 1. (1798) 3 U.S. 386, at p. 390. 2. (1965) 381 U.S. 437, at p. 447. 3. (1810) 10 U.S. 87, at p. 138; 6 Cranch 87, at p. 138. Historically, bills (or, more correctly, acts) of attainder constituted a particular form of law, generally of an ex post facto character, whereby punishment was inflicted upon a designated person or group of persons adjudged by the legislature to have been guilty of crimes, usually of a capital nature, such as treason or murder. The particular objection to bills of attainder was not so much that they may have had an ex post facto operation, but that they substituted the judgment of the legislature for that of a court. In England, the practice evolved of giving the person with whom a bill dealt some sort of a hearing, but the result was still secured by legislation and not by judicial action: see Australian Communist Party v. The Commonwealth [96] . In the United States, where, as I have said, an expanded version of bills of attainder was adopted, it was this aspect which was seen as the vice, not only because it was oppressive, but also because it was thought (at least by 1965) to offend against the separation of powers doctrine. In United States v. Brown [97] the Supreme Court said: While history thus provides some guidelines, the wide variation in form, purpose and effect of ante-Constitution bills of attainder indicates that the proper scope of the Bill of Attainder Clause, and its relevance to contemporary problems, must ultimately be sought by attempting to discern the reasons for its inclusion in the Constitution, and the evils it was designed to eliminate. The best available evidence, the writings of the architects of our constitutional system, indicates that the Bill of Attainder Clause was intended not as a narrow, technical (and therefore soon to be outmoded) prohibition, but rather as an implementation of the separation of powers, a general safeguard against legislative exercise of the judicial function, or more simply — trial by legislature. 1. (1951) 83 C.L.R., at p. 172. 2. (1965) 381 U.S., at p. 442. The perception that the bill of attainder clause in the United States Constitution was intended to implement the separation of powers came somewhat late in the history of that clause. It is not without its difficulties, the chief of which is, as was pointed out by White J. in dissent in United States v. Brown [98] , in quoting from Dreyer v. Illinois [99] , that State legislatures, as well as Congress, are prohibited from passing bills of attainder and "[w]hether the legislative, executive and judicial powers of a State shall be kept altogether distinct and separate is for the determination of the State". Nevertheless, it is perfectly clear that it is only the prohibition against bills of attainder and not the prohibition against ex post facto laws generally which might be traced to the doctrine of the separation of powers. That is because it is not the ex post facto nature of bills of attainder which offends against the doctrine, but the substitution of the judgment of the legislature for that of the courts. That is not a necessary feature of an ex post facto law. Where an ex post facto law penalizes a past activity by means of a generally applicable rule rather than by specifying the persons to be subjected to the penalty, it is not a bill of attainder whatever other objections might be raised to its ex post facto operation. In that situation a court is still left to determine whether an individual is guilty of having engaged in the prohibited activity, albeit an activity which took place before the law created the offence, and the legislature has done no more than lay down a rule of general application which is part of its true function. Even a law which penalizes persons who possess specified characteristics may not be a bill of attainder, provided the characteristics are specified in sufficiently general terms. Legislation will amount to a bill of attainder only where it is apparent that the legislature intended the conviction of specific persons for conduct engaged in in the past. The law may do that by penalizing specific persons by name or by means of specific characteristics which, in the circumstances, identify particular persons. A court in applying such a law is in effect confined in its inquiry to the issue of whether or not an accused is one of the persons identified by the law. If he is, his guilt follows. The proper judicial inquiry as to whether an accused has been guilty of prohibited conduct has thus been usurped by the legislature. Alternatively, a bill of attainder may designate the persons it seeks to penalize by means of some characteristic (such as membership of an organization) that is independent of and not equivalent to the criminal activity which it is the purpose of the law to prohibit or prevent. In United States v. Brown, the Supreme Court struck down as amounting to a bill of attainder an Act which made it a crime for a member of the communist party to serve as an officer or (except in clerical or custodial positions) as an employee of a labour organization. The Court said [1] : The statute does not set forth a generally applicable rule decreeing that any person who commits certain acts or possesses certain characteristics (acts and characteristics which, in Congress' view, make them likely to initiate political strikes) shall not hold union office, and leave to courts and juries the job of deciding what persons have committed the specified acts or possess the specified characteristics. Instead, it designates in no uncertain terms the persons who possess the feared characteristics and therefore cannot hold union office without incurring criminal liability — members of the Communist Party. The Court went on to observe [2] : The designation of Communists as those persons likely to cause political strikes is not the substitution of a semantically equivalent phrase In a number of decisions, this Court has pointed out the fallacy of the suggestion that membership in the Communist Party, or any other political organization, can be regarded as an alternative, but equivalent, expression for a list of undesirable characteristics. 1. (1965) 381 U.S., at p. 473. 2. (1902) 187 U.S. 71, at p. 84. 3. (1965) 381 U.S., at p. 450. 4. (1965) 381 U.S., at p. 455. The Commonwealth Constitution contains no provision which corresponds to the bill of attainder clause in the United States Constitution. Nor does it adopt in its entirety the United States theory of the separation of powers. That is readily apparent from the fact that responsible government is a central feature of our constitutional scheme. But in Reg. v. Kirby; Ex parte Boilermakers' Society of Australia [3] , this Court pointed out that the federal structure required the provision of a judicature with paramount powers to determine the demarcation of governmental functions between independent governments. That provision is to be found in Ch. III of the Constitution, which establishes that no resort can be had to judicial power except in conformity with that Chapter. For that reason, it is beyond the competence of the Parliament to invest with any part of Commonwealth judicial power any body or person except a court created or invested with jurisdiction pursuant to Ch. III. To that extent there is clearly a separation of the judicial power of the Commonwealth from its legislative and executive functions. 1. (1956) 94 C.L.R. 254. But even accepting for present purposes the expanded notion of a bill of attainder which has prevailed in the United States and accepting that a measure which would fall within that notion would offend against the separation of Commonwealth judicial power, it cannot be said that the War Crimes Act amounts to a bill of attainder. True it is that it operates ex post facto, but that does not, as I have endeavoured to explain, convert it into a bill of attainder nor does it import that quality of a bill of attainder which has been said to constitute an intrusion upon the exercise of judicial power, namely, the substitution of legislative judgment for the judgment of the courts. Those activities which are said by the Act to constitute war crimes are defined in general terms without any attempt to designate any person or group of persons as having engaged in any of those activities. Nor does the Act pick out the persons it seeks to penalize by means of a characteristic which is independent of the prohibited criminal activity; it may have done so had it specified that any person who had been a member of a particular enemy organization was thereby to be held guilty of a war crime, but that is not the approach adopted. The Act confines the period during which, and the field of war in relation to which, those activities will have constituted war crimes, but that in no way usurps the judicial function of determining whether a particular person charged with an offence engaged in conduct which the Act describes as a war crime. The fact that the period specified by the Act occurred before the Act in its amended form came into operation is what gives it an ex post facto operation. But otherwise its provisions are in such a form as to be capable of operating prospectively and, were they to do so, the function of a court applying those provisions would be the same, namely, ascertaining whether a person alleged to have engaged in conduct which the Act proscribes had in fact done so. The Act does not, merely because it penalizes conduct which occurred before it came into operation, supplant the court in the exercise of its ordinary, traditional function. That function is, in the application of the criminal law, the ascertainment in accordance with the law of the guilt or innocence of a person charged with an offence under the law. In designating conduct — whether in the future or in the past — as criminal, a law does not intrude upon the judicial function. It is when the legislature itself, expressly or impliedly, determines the guilt or innocence of an individual that there is an interference with the process of the court. The Act requires no finding of guilt on the part of a person charged with an offence under its provisions, save where that person is found to have contravened the rules of conduct which it lays down. That finding requires the determination of a court after a trial. The fact that the Act lays down rules of conduct in relation to events which occurred before it came into effect does not invest it with the attributes of a bill of attainder, however widely such an instrument is defined. Of course, the real question is not whether the Act amounts to a bill of attainder, but whether it exhibits that characteristic of a bill of attainder which is said to represent a legislative intrusion upon judicial power. In Liyanage v. The Queen [4] there was no bill of attainder (although bills of attainder were referred to by analogy), but the Privy Council held that legislation passed by the Parliament of Ceylon constituted the usurpation of the judicial function in contravention of the separation of judicial power which is embodied in the Constitution of Ceylon. The legislation followed an abortive coup d'état in January 1962, in respect of which a number of persons identified in a White Paper were charged with a variety of conspiracies. The accused were all (save for one) in custody at the time the two Acts constituting the legislation in question were passed. The Acts were attacked in argument in a manner which the Privy Council said fairly described their effect [5] : The first Act was wholly bad in that it was a special direction to the judiciary as to the trial of particular prisoners who were identifiable (in view of the White Paper) and charged with particular offences on a particular occasion. The pith and substance of both Acts was a legislative plan ex post facto to secure the conviction and enhance the punishment of those particular individuals. It legalised their imprisonment while they were awaiting trial. It made admissible their statements inadmissibly obtained during that period. It altered the fundamental law of evidence so as to facilitate their conviction. And finally it altered ex post facto the punishment to be imposed on them. The legislation was held to be invalid, not because of its ex post facto operation (although that was a necessary feature of it), but because, far from laying down any general rule of conduct, it was designed to secure the conviction of identifiable individuals. It was, therefore, a legislative attempt to exercise judicial power which, under the Constitution of Ceylon, was confided to the courts. Indeed, the Privy Council said of the Acts [6] , that one might fairly apply the words of Chase J. in Calder v. Bull [7] that they were "legislative judgments; and an exercise of judicial power". Earlier [8] , the observation was made that: It goes without saying that the legislature may legislate, for the generality of its subjects, by the creation of crimes and penalties or by enacting rules relating to evidence. But the Acts of 1962 had no such general intention. They were clearly aimed at particular known individuals who had been named in a White Paper and were in prison awaiting their fate. The fact that the learned judges declined to convict some of the prisoners is not to the point. That the alterations in the law were not intended for the generality of the citizens or designed as any improvement of the general law is shown by the fact that the effect of those alterations was to be limited to the participants in the January coup and that, after these had been dealt with by the judges, the law should revert to its normal state. 1. [1967] 1 A.C. 259. 2. [1967] 1 A.C., at p. 290. 3. [1967] 1 A.C., at p. 291. 4. (1798) 3 U.S., at p. 389. 5. [1967] 1 A.C., at p. 289. The legislation in this case was, by way of contrast, enacted by the legislature "for the generality of its subjects". It is not aimed at "particular known individuals" but at any Australian citizen or resident who, within the specified period, engaged in conduct amounting to the commission of a war crime as defined by the Act. It does not, therefore, represent a legislative usurpation of judicial power. For these reasons, I would answer the question asked in the negative. Toohey J. The relevant facts and statutory provisions appear in the judgments of other members of the Court. I shall avoid undue repetition. The plaintiff, who is an Australian citizen and a resident of South Australia, is charged with having committed "war crimes" contrary to s. 9 of the War Crimes Act 1945 Cth. All references in this judgment to "the Act" are to the War Crimes Act 1945 Cth as amended by the War Crimes Amendment Act 1988 Cth, unless otherwise stated. It is unnecessary to set out the information in detail. It contains thirteen paragraphs. Some allege that the plaintiff murdered one or more persons in the Ukraine, the persons being identified by name or description. Others allege that the plaintiff was knowingly concerned in or party to the murder of about 850 Jews from the village of Serniki in the Ukraine. All offences are said to have taken place between 1942 and 1943, during and in the course of the German occupation of the Ukraine in World War II. They are said to have been committed "in pursuing a policy of Germany associated with the conduct of the said war and occupation or on behalf of or in the interests of that power being a policy of persecution of the Jewish people of Europe on political, racial or religious grounds" or "with intent to destroy the Jewish people" in the places concerned and, in five cases, in pursuing a policy of annihilating those suspected to be partisans or communists. The question reserved for the consideration of the Court is whether s. 9 of the Act is invalid in its application to the information laid against the plaintiff. Nevertheless, the arguments went to the validity of the Act generally. The issues canvassed may be summed up in this way: 1. Is the Act valid as an exercise of the power of the Parliament to make laws for the peace, order and good government of the Commonwealth with respect to external affairs (s. 51(xxix) of the Constitution)? 2. Is the Act valid as an exercise of the power of the Parliament to make laws for the peace, order and good government of the Commonwealth with respect to the naval and military defence of the Commonwealth (s. 51(vi) of the Constitution)? 3. If the Act is otherwise valid, does it involve an improper use of the judicial power of the Commonwealth? If it does, what consequences follow for the charges against the plaintiff? But, in the end, the question to be answered by the Court is the question reserved for its consideration. External affairs — externality The Commonwealth argued that the external affairs power supports the validity of the Act by reference to what might be termed geographical externality and, in the alternative, on the basis that the Act implements an obligation or a concern to be found in international law or a resolution of an international body or that it is an exercise of a universal jurisdiction in international law. (The Director of Public Prosecutions supported the Commonwealth in its arguments. Nevertheless, it is convenient to speak only of "the Commonwealth" except where some additional argument was advanced by the Director.) I shall deal first with the argument based on geographical externality. It is true that there are passages in earlier judgments of members of the Court suggesting that a law of the Commonwealth may operate with respect to persons or events by reason only of their externality. In New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [9] , Barwick C.J. spoke of "any affair which in its nature is external to the continent of Australia and the island of Tasmania subject always to the Constitution as a whole". Mason J. [10] said that the power conferred by s. 51(xxix) "extends to matters or things geographically situated outside Australia". Jacobs J. [11] , spoke of the power "to make laws with respect to places outside, or matters or things done outside the boundaries of the Commonwealth". Murphy J. did not speak in these terms. But in the light of what his Honour said in Viro v. The Queen [12] , and in The Commonwealth v. Tasmania (the Tasmanian Dam Case) [13] , he must be taken to have indorsed this general view of s. 51(xxix). 1. (1975) 135 C.L.R. 337, at p. 360. 2. (1975) 135 C.L.R., at p. 471. 3. (1975) 135 C.L.R., at p. 497. 4. (1978) 141 C.L.R. 88, at p. 162. 5. (1983) 158 C.L.R. 1, at pp. 171-172. However, these statements must be taken in context and, when so taken, they do not establish a proposition that there need be no connexion whatsoever between Australia and the subject matter of a law which the external affairs power is said to support. In the Seas and Submerged Lands Case the statements were made in answer to an argument that the external affairs power was limited to a power to make laws with respect to Australia's relationships with foreign countries: see, for instance, Mason J. [14] . The legislation under attack in the Seas and Submerged Lands Case did not directly involve another country but it went to the sovereignty of the Commonwealth over the territorial sea, that is, to the very boundaries of Australia. In Koowarta v. Bjelke-Petersen [15] Mason J., described "external affairs" as covering "any matters or concerns external to Australia". But the context in which the statement was made cannot be ignored; it was the implementation of a treaty obligation. 1. (1975) 135 C.L.R., at pp. 470-471. 2. (1982) 153 C.L.R. 168, at p. 223. Certainly the word "external" means situated outside and of itself requires no connexion with that which it is outside. But in sub-s. (xxix) "External" is an adjective qualifying "affairs" and it is necessary to have regard to the composite expression. Dictionaries commonly define "affair" by reference to "concern". An external affair is a matter which is external in the sense that it lies outside Australia but it is a matter which is of concern to Australia. This does not mean that it must be a matter touching Australia's relations with another country though that will ordinarily be the case. But it does mean that it is a matter in which Australia has an interest. Whether a matter so qualifies is, in this respect, for the Parliament to determine. But the power assumes the existence of a national interest in some person, thing or matter that enables one to say that the subject of legislation concerns Australia. I am content to accept the explanation given by Stephen J. in Koowarta [16] , where his Honour says: The word "External" must in this constitutional context qualify "affairs" so as to restrict its meaning to such of the public business of the national government as relates to other nations or other things or circumstances outside Australia. It is legislation for the peace, order and good government of the Commonwealth with respect to such a subject-matter that the words of par. (xxix) appear to envisage. 1. (1982) 153 C.L.R., at p. 211. Quick and Garran, Annotated Constitution of the Australian Commonwealth (1901), pp. 631-632, give sub-s. (xxix) a restricted operation. But they do so in the context of the relationship between Australia and the Imperial Parliament at the time of federation. It was through the Imperial Parliament that Australia's external affairs were conducted. With the emergence of Australia as a nation, that restrictive view could no longer stand. However, the issue is one of constitutionality, that is, the power of the Parliament to legislate with respect to a particular matter. A matter does not qualify as an external affair simply because it exists outside Australia. It must be a matter which the Parliament recognises as touching or concerning Australia in some way. Indeed it might be thought more than passing strange that the Constitution solemnly conferred power on the Parliament to legislate with respect to a matter in which it had no interest. Externality — War Crimes Act The Commonwealth urged that the Act deals with matters geographically external to Australia, hence that it is justified by reference to the external affairs power. The Solicitor-General for the Commonwealth put the matter this way: The Commonwealth is not relying upon the external affairs power to intrude in any way into a field which would otherwise be within State power. The Act is confined to dealing with something done, or the consequence of something done, overseas where the national Parliament thinks that that legislation is desirable. But the Act is not confined in the way suggested by the Solicitor-General. The Commonwealth argued that s. 6(1), which refers to conduct "done in a part of Australia", is definitional only; that it provides "the foundation for the structure of a municipal offence", going no further than is warranted by international law; that s. 6(3), which picks up the "serious crimes" listed in s. 6(1) where the conduct occurred outside Australia, effectively comprises the range of acts which can constitute a war crime; and that the expression "serious crime" in s. 6(1) has no operation other than as a step towards an understanding of what is meant by a "war crime" for the purposes of ss. 7 and 9. And, it was said, s. 7(1) must be read in light of the definition of war in s. 5. But none of this assists the Commonwealth's broad argument on externality. It is true that the definition in s. 5 confines "war" to a "war (whether or not involving Australia or a country allied or associated with Australia) in so far as it occurred in Europe in the period beginning on 1 September 1939 and ending on 8 May 1945" (emphasis added). But a serious crime, which is a "war crime" if committed, inter alia, in pursuing a policy associated with the conduct of such a war or with an occupation (s. 7(1)(c)) or on behalf of, or in the interests of, a power conducting such a war or engaged in an occupation (s. 7(1)(d)), may be committed anywhere in the world including Australia. To say that is not to overlook s. 7(2) which provides that a serious crime was not committed in the course of hostilities in a war (see s. 7(1)(a)) or in the course of an occupation (see s. 7(1)(b)) merely because the connexion between the serious crime on the one hand and the hostilities or occupation on the other was only incidental or remote. The connexion may be very real indeed, wherever the serious crime was committed. A person may not be charged with an offence against the Act unless he or she is an Australian citizen or a resident of Australia or of an external Territory: s. 11 of the Act. But that is a status required at the date of charging. The person charged need not have been an Australian citizen or resident at the time of the alleged offence. Indeed, it is clear that the Act is aimed at situations in which neither an Australian offender nor victim is likely to have been involved: see, for instance, the preamble. Furthermore, the definition of war in s. 5 does not require Australia's involvement. Unquestionably, the Act encompasses conduct which may have occurred within or without Australia. That is borne out not only by the sections to which I have already referred, but in addition by s. 18. That section deals with alternative verdicts and speaks of conduct done in Australia and elsewhere. Nevertheless, there is no difficulty in concluding that, in the context of a war in which Australia was directly involved, in which many Australian service personnel and civilians were killed, wounded, imprisoned or ill-treated and which had such significant social, economic and political consequences for this country, an Act purporting to render those who are Australian citizens or residents liable for conduct associated with that war legislates with respect to a matter which is of concern to Australia and to which the public business of the national government relates. It is true, of course, that the definition of war in s. 5 does not require Australia's involvement. But it would be to turn a blind eye to history to see no connexion between the dates and area identified in s. 5 and World War II, or to conclude that "war" as defined could relate to a conflict in which Australia had no interest at the time, even if not directly involved. For these reasons the law is one with respect to "External affairs" within s. 51(xxix). The fact that the Act creates a liability for past conduct does not of itself remove the legislation from the capacity to deal with an external affair though that matter has other implications to which it will be necessary to refer later in these reasons. Whether the Act can be said to be a law with respect to external affairs, in so far as it encompasses conduct which occurred within Australia, is another matter, with which it is unnecessary to deal. If, in that respect, the Act is beyond power, the relevant provisions are severable. Their severance would not affect the operation of the balance of the Act. External affairs — international law In its terms, s. 9 does not involve an inquiry into international law and, on the view I have taken that the Act is a law with respect to external affairs, it might be thought unnecessary to enter upon a discussion of international law. However, acknowledging the breadth of the argument before the Court, I propose to consider some aspects of that law in arriving at a conclusion as to the validity of the Act. The general scope of the Act has been explained. Reference to international law in terms appears only in s. 17(2) which, subject to the exclusion in s. 16 of the defence of superior orders, makes it a defence to a charge under the Act, if the doing of the act: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. The Commonwealth supported the validity of the Act as an exercise of the external affairs power, not only by reference to geographical externality, but also by contending that the Act gives effect to an obligation or concern in international law or that it implements a resolution of an international body. The Commonwealth further argued that the Act facilitates the exercise of a right existing in international law, namely, the universal jurisdiction. External affairs — international obligation The obligation upon which the Commonwealth relied to support s. 9 as a valid exercise of the external affairs power was expressed as an obligation "to search out, bring to trial and, if found guilty, to punish war criminals". In satisfying that obligation, it was said, Australia may itself prosecute and punish or it may extradite the offender to the country where the offence is alleged to have been committed. The plaintiff's response was that no such obligation exists in international law. He said further that, if there is any comparable obligation, it does not go beyond a duty to extradite the offender to the place of the offence. To establish the existence of the wider obligation for which the Commonwealth contended, it is necessary to point to general practice by States and opinio juris. In the Commonwealth's submission, each limb has a bearing on the other in the sense that if State practice is widespread it is less crucial to demonstrate opinio juris. And, it was contended, the obverse is true. I have read what Brennan J. has written in regard to the existence of an international obligation to prosecute and punish as opposed to extraditing. I agree with his Honour's analysis and with his conclusion: "there is no evidence of widespread State practice which suggests that States are under a legal obligation to seek out Axis war criminals and to bring them to trial. There is no opinio juris supportive of such a rule." I have nothing to add to Brennan J.'s analysis and conclusion in this regard. It follows that the Act cannot be supported as an exercise of the external affairs power on the ground that it gives effect to an international obligation. External affairs — international concern It is convenient to approach this aspect of the case against the background of what was said by Stephen J. in Koowarta [17] : A subject-matter of international concern necessarily possesses the capacity to affect a country's relations with other nations and this quality is itself enough to make a subject-matter a part of a nation's "external affairs". It was appropriate for Stephen J., in the context of his judgment, to speak of "a country's relations with other nations". But I do not understand his Honour to exclude any other consideration as constituting a subject matter of international concern. In the light of what he said in Koowarta [18] (the passage quoted earlier in these reasons), his Honour would accept a matter touching the public business of Australia in relation to an event outside Australia as capable of giving rise to a concern that would answer the description — an external affair: see also Mason J. in Koowarta [19] . 1. (1982) 153 C.L.R., at p. 217. 2. (1982) 153 C.L.R., at p. 211. 3. (1982) 153 C.L.R., at p. 234. Much of the learning that is relevant to the question whether there is an international obligation on countries to prosecute and punish war criminals whose offences were committed elsewhere is naturally relevant to the question whether it is a matter of international concern to do so. Again, I agree with Brennan J.'s conclusion: "There is insufficient material to show that the apprehension and trial of such [i.e. World War II] war criminals before courts of countries other than those in which the crimes were committed were ever matters of international concern." But this aspect of the case tends to be subsumed in what was referred to in argument as the universal jurisdiction. This jurisdiction, it was said by the Commonwealth, supported the Act as a law with respect to external affairs. External affairs — universal jurisdiction The Commonwealth contended that, in the event that the Court found no relevant international obligation or concern to exist, the Act is nevertheless a valid exercise of the external affairs power because Australia has jurisdiction in international law to prosecute war crimes and crimes against humanity which occurred outside Australia against non-nationals. The focus of this analysis shifts from inquiry into a substantive obligation or concern, requiring or justifying action on the part of the Australian Government, to the concept of crimes existing in international law and principles of jurisdiction which provide Australia with authority to prosecute those crimes. The term "jurisdiction" has different meanings in international and municipal law. In international law it is used in various ways but it may be taken to refer to "a state's general legal competence and is an aspect of state sovereignty": Triggs, "Australia's War Crimes Trials: A Moral Necessity or Legal Minefield?", Melbourne University Law Review, vol. 16 (1987) 382 (hereafter "Triggs"), at p. 387. Relevantly, it "refers to a state's legitimate assertion of authority to affect legal interests": Randall, "Universal Jurisdiction Under International Law", Texas Law Review, vol. 66 (1988) 785 (hereafter "Randall"), at p. 786. The term has legislative, adjudicatory and enforcement dimensions: Randall, p. 786; Triggs, p. 387; Wagner, "U.S. Prosecution of Past and Future War Criminals and Criminals Against Humanity: Proposals for Reform Based on the Canadian and Australian Experience", Virginia Journal of International Law, vol. 29 (1989) 887 (hereafter "Wagner"), at p. 899. We are here concerned with Australia's authority to make criminal laws applicable to certain persons, events or things with the aim of dealing with an international law crime. We are concerned, therefore, not only with Australia's legislative power in constitutional law, but also with Australia's enforcement and adjudicatory authority in international law because the Commonwealth relies on that authority to support its legislative power. The subjects of international law are primarily, though not exclusively, states: Brownlie, Principles of Public International Law, 4th ed. (1990) (hereafter Brownlie), ch. III and see ch. XXIV. Individuals are recognized by international law in so far as they are protected by, or, more importantly here, are subject to, international law. There is no exhaustive list of bases upon which a state may exert authority over an individual in international law nor is there precise agreement between commentators as to categorization. But a common and convenient analysis is that five principles emerge by which the legitimacy of an asserted jurisdiction in criminal matters may be assessed: Kobrick, "The Ex Post Facto Prohibition and the Exercise of Universal Jurisdiction over International Crimes", Columbia Law Review, vol. 87 (1987) 1515 (hereafter "Kobrick"), at p. 1519; Randall, pp. 787-788; Wagner, pp. 899-900. Cf. Brownlie, pp. 300-307; Triggs, pp. 387-389. They are: 1. the territoriality principle, which applies when an offence occurs within the territory of the prosecuting state; 2. the nationality principle, which applies when the offender is a national of the prosecuting state; 3. the protective principle, which is excited where an extraterritorial act threatens the integrity of the prosecuting state; 4. the passive personality principle, which applies where the victim of the offence is a national of the prosecuting state; and 5. the universality principle. The last of these principles permits jurisdiction to be exercised over a limited category of offences on the basis that the offender is in the custody of the prosecuting state. The jurisdiction is based on the notion that certain acts are so universally condemned that, regardless of the situs of the offence and the nationality of the offender or the victim, each state has jurisdiction to deal with perpetrators of those acts. Since the Act focuses primarily (and, in practice, possibly entirely) on acts committed by non-nationals against non-nationals outside Australia, its likely basis for jurisdiction over war criminals from World War II is the principle of universal jurisdiction: see Wagner, p. 901. There appears to be no consensus that the "nationality of offender" basis for jurisdiction will include the situation where an offender later becomes a national. Cf. Triggs, p. 393, where it is suggested that "territorial jurisdiction over Australian citizens and residents" may be applicable. Before examining material which is relevant in deciding whether war crimes and crimes against humanity in international law are subject to universal jurisdiction, it is useful to look at the doctrine itself because views differ as to its nature. The principle of universality is, at times, used to refer to the authority of states to exercise jurisdiction over certain conduct, regardless of whether it constitutes a crime under customary international law. On this view of the principle it is the universality of the condemnation of, for example, the common crime of murder which allows every state to exert authority over an offender in the absence of other jurisdictional links: Brownlie, pp. 304-305; Triggs, p. 389. In other words, a state may assume jurisdiction over an alleged murder offence where it would otherwise fall within the jurisdiction of another state under its own municipal law. No question of an international law offence arises. However, the principle is most often formulated so that it applies only to crimes which are already constituted as such under international law. In this respect, the principle rests on the existence of an offence in international law; the municipal law under which an individual is prosecuted must be in conformity with that international law. On this view, authority to prosecute the relevant conduct extends to every state under its own laws, even in the absence of one or more of the other jurisdictional links such as territoriality or nationality. But it is the existence of the crime in international law, and not simply the universality of condemnation in states' own municipal laws (though this may be evidence supporting the existence of the crime), which justifies the exception to the requirements of the other jurisdictional bases: Randall, pp. 795-798; American Law Institute, Restatement of the Law, Third: The Foreign Relations Law of the United States (hereafter Restatement), §404; Williams and Castel, Canadian Criminal Law: International and Transnational Aspects (1981) (hereafter Williams and Castel), p. 137. In the context of war crimes and crimes against humanity, it is this formulation of the universality principle, relying as it does on the existence of an offence in international law, which is relevant: Brownlie, p. 305; Williams and Castel, ch. 5. And it is this formulation on which the Commonwealth relies. The Commonwealth's use of the concept of a "right" existing in international law by reason of the universality principle is misleading, especially if it is (as it was in argument) associated with a right vested in Australia by treaties and other international agreements. We are concerned here with authority to proceed legally; in that sense Australia may have a right but it is not in the nature of a substantive right created by treaty or, by analogy, contract. The Commonwealth's concern was to emphasize the potency of the principle for the purposes of relying on it for constitutional validity. Thus a distinction was drawn between a "special and limited right" and mere permission. Again terminology may be misleading because universality of jurisdiction is in fact a permissive doctrine. But the proposition that universal jurisdiction is positively conferred by international law and is not merely the absence of prohibition is well founded. Specifically conferred authority to exercise that jurisdiction is a sufficient foundation on which to base a law of the Parliament with respect to external affairs because the universality of the condemnation necessarily touches and concerns Australia. If jurisdiction conferred by principles of international law is a component of sovereignty, then, in the absence of constitutional prohibition or conflict between the scope of federal and State powers, that jurisdiction is a necessary aspect of the Commonwealth's capacity to function effectively in the international community. Therefore the exercise of that jurisdiction where it exists — or rather the perceived commission of an international crime subject to that jurisdiction — is, or may give rise to, an external affair for the purposes of the legislative power of the Commonwealth. That Australia has a choice whether or not to exercise the jurisdiction does not alter that characterization as an external affair. Universal jurisdiction: war crimes and crimes against humanity Whether the rationale for the universality principle lies in the proposition that those committing certain offences lose their national character and are therefore subject to any state's jurisdiction, or whether it lies in the fundamental nature of the crime — its particular gravity and heinousness (see Randall, pp. 792-795; In re List (Hostages Trial) [20] ; Attorney-General (Israel) v. Eichmann [21] (Supreme Court); [22] , (District Court)), there appears to be general agreement that war crimes and crimes against humanity are now within the category subject to universal jurisdiction: see Brownlie, pp. 305, 562; Kobrick, pp. 1522-1523, 1529; Randall, p. 800; Wagner, p. 905 (with respect to war crimes). 1. (1948) 15 Annual Digest 632, at p. 636. 2. (1962) 36 I.L.R. 277, at pp. 282-283. 3. (1962) 36 I.L.R. 18, at p. 50. In numerous cases of prosecution of war criminals after World War II, for both violations of the international laws of war and crimes against humanity, reliance was placed, inter alia, on the universality principle. For example, in In re Eisentrager (Shanghai) [23] , the United States Military Commission rejected the argument of the defendants that, because they were German citizens residing in China, they were subject only to Chinese law and jurisdiction. The Commission said [24] : A war crime is not a crime against the law or criminal code of any individual nation, but a crime against the ius gentium. The laws and usages of war are of universal application, and do not depend for their existence upon national laws and frontiers. Arguments to the effect that only a sovereign of the locus criminis has jurisdiction and that only the lex loci can be applied, are therefore without any foundation. See also The Hadamar Trial (In re Klein) (Wiesbaden) [25] ; In re Tesch (Zyklon B Case) [26] ; In re List; Attorney-General (Israel) v. Eichmann; Demjanjuk v. Petrovsky [27] . 1. (1947) 14 Law Reports of Trials of War Criminals 8. (hereafter "L. Rep. Trials War Crims"). 2. (1947) 14 L. Rep. Trials War Crims, at p. 15. 3. (1945) 1 L. Rep. Trials War Crims 46. 4. (1946) 13 Annual Digest 250. 5. (1985) 776 F. 2d 571, at p. 582. In In re List the United States Military Tribunal ("U.S.M.T.") said [28] : An international crime is such an act universally recognized as criminal, which is considered a grave matter of international concern and for some valid reason cannot be left within the exclusive jurisdiction of the state that would have control over it under ordinary circumstances. 1. (1948) 15 Annual Digest, at p. 636. And the District Court of Jerusalem, in Attorney-General (Israel) v. Eichmann, based its jurisdiction "on a dual foundation: the universal character of the crimes in question and their specific character as intended to exterminate the Jewish people": [29] . The Court further explained [30] : The State of Israel's "right to punish" the accused derives, in our view, from two cumulative sources: a universal source (pertaining to the whole of mankind), which vests the right to prosecute and punish crimes of this order in every State within the family of nations; and a specific or national source, which gives the victim nation the right to try any who assault its existence. Both the District Court and Supreme Court judgments described the precedent of universal jurisdiction over piracy, drew an analogy between piratical acts and Nazi atrocities, and found support for the universality principle in the earlier war crimes cases [31] . 1. (1962) 36 I.L.R., at p. 26. 2. (1962) 36 I.L.R., at p. 50. 3. (1962) 36 I.L.R., at p. 26; (1962) 36 I.L.R., at pp. 290-295. However, to say that war crimes and crimes against humanity were, sometime after World War II, subject to universal jurisdiction does not answer the question whether the conduct of which the plaintiff is accused was a war crime or a crime against humanity before the end of the War nor whether, if they existed, those crimes could then be prosecuted by any state. It may be said that, if a crime is found to have existed at some point in the past but was not the subject of universal jurisdiction, the subsequent expansion of jurisdiction is a procedural matter only and that a state with no other jurisdictional link can prosecute legitimately after the status of universal jurisdiction has been achieved. A better approach in this instance, however, is to examine the relationship between the concepts of "international crime" and "universal jurisdiction". The question whether a crime is constituted as such in international law is, conceptually, distinct from the question whether that crime is the subject of universal jurisdiction: Kobrick, pp. 1522, 1528. A crime created by treaty will not be the subject of universal jurisdiction merely by reason of its conventional existence: see Restatement, §404. It is less clear, however, that crimes having their source in custom can be said not to be the subject of universal jurisdiction unless limitations on the right to prosecute are contained in the definition of the crime itself. Certainly, the two questions — whether a crime exists and the scope of jurisdiction to prosecute — are inextricably linked. An international crime is constituted, precisely, where conduct is identified which offends all humanity, not only those in a particular locality; the nature of the conduct creates the need for international accountability. Where conduct, because of its magnitude, affects the moral interests of humanity and thus assumes the status of a crime in international law, the principle of universality must, almost inevitably, prevail: see Zoller, Territorial Effect of the Norm on Responsibility in Ginsburgs and Kudriavtsev (eds), The Nuremberg Trial and International Law (1990), p. 106. This is particularly true of crimes against humanity since they comprise, by definition, conduct abhorrent to all the world. Therefore, while the question whether war crimes and crimes against humanity were subject to universal jurisdiction during World War II remains theoretically distinct, the question whether the crimes existed as such at that time is basic. If such conduct amounted, then, to customary international crimes, their very nature leads to the conclusion that they were the subject of universal jurisdiction. It is convenient to look, first, at the concepts of "war crime" and "crime against humanity" to determine whether they existed in international law between 1942 and 1943 (the period in respect of which the plaintiff is charged) and, if so, to examine the Act in some detail to determine whether its provisions accurately reflect those concepts and are an effective exercise of universal jurisdiction. International crimes — war crimes and crimes against humanity The term "war crime" in s. 9 of the Act looks to two distinct, though overlapping, concepts in international law: "war crimes" and "crimes against humanity". War crimes in international law are contraventions of the laws and customs of war recorded in such documents as the Hague Conventions of 1907 and in military manuals. "Crimes against humanity" in international law is a generic term which refers to crimes of persecution, that is, persecution on political, racial or religious grounds, and to crimes of extermination. It is important to note that the difference between war crimes and crimes against humanity lies in the context in which they are committed. Traditionally, the laws and customs of war governed only conduct between belligerents or between a belligerent and the inhabitants of an occupied country. This is a reflection of the fundamental doctrines of sovereignty and non-intervention between states. Crimes against humanity, on the other hand, are not so confined. They may be carried out by a national against another national of the same country, and in peacetime. Conduct may therefore constitute both a war crime and a crime against humanity. The Commonwealth submitted that s. 7 of the Act embraces both kinds of crime at international law. The submission continued in this way. War crimes, including crimes against humanity which also amount to a war crime, are reflected in s. 7(1) and (2). And if the terms of sub-s. (3) are satisfied, the conduct may be prosecuted under that sub-section also. Those crimes against humanity which would not also have amounted to a war crime are particularly reflected in s. 7(3). War crimes There is no doubt that war crimes were crimes in international law during World War II. The fourth Hague Convention of 1907, the international Convention concerning the Laws and Customs of War on Land ("the Hague Convention"), was ratified by Germany and Russia as well as the major Allied powers. The first Article of the Convention required the contracting states to issue instructions to their land forces in conformity with the Regulations respecting the Laws and Customs of War on Land annexed to the Convention ("the Hague Rules"): see Manual of Military Law, 7th ed. (1929) (Great Britain); Manual of Military Law 1941, Australian ed. (hereafter " Australian Military Manual "); the "Kriegsbrauch in Landkriege", instructions issued to the German armed forces following the Hague Conventions. The matters dealt with by the Hague Rules included: the status of belligerents, the humane treatment of prisoners of war, and Arts 42-56 dealt with rules of conduct of a hostile state in occupied territory. Article 46 provided that, where a territory is occupied, "[f]amily honour and rights, individual life, and private property, as well as religious convictions and worship, must be respected." Article 50 read: No collective penalty, pecuniary or otherwise, shall be inflicted upon the population on account of the acts of individuals for which it cannot be regarded as collectively responsible. The Hague Convention and its annexed Rules provided an undisputed reference in peace negotiations between Germany and the Allies after World War I. With respect to war crimes, the International Military Tribunal ("I.M.T.") exercised jurisdiction over major war criminals after World War II on the basis of Art. 6(b) of its Charter ("the Nuremberg Charter"). That Article defines war crimes as: violations of the laws or customs of war. Such violations shall include, but not be limited to, murder, ill-treatment or deportation to slave labour or for any other purpose of civilian population of or in occupied territory, murder or ill-treatment of prisoners of war or persons on the seas, killing of hostages, plunder of public or private property, wanton destruction of cities, towns or villages, or devastation not justified by military necessity. (Emphasis added.) Article 6(b) does not extend beyond the treatment of civilian populations in occupied territory: see In re Altstötter (The Justice Trial) [32] . The I.M.T. claimed that the law contained in Art. 6(b) represented existing international law: "With respect to War Crimes, the crimes defined by Article 6, Section (b), of the Charter were already recognized as War Crimes under international law. They were covered by Articles 46, 50, 52, and 56 of the Hague Convention of 1907, and Articles 2, 3, 4, 46, and 51 of the Geneva Convention of 1929": judgment of the International Military Tribunal (Nuremberg), reproduced in American Journal of International Law vol. 41 (1947) 172, at p. 248. 1. (1947) 14 Annual Digest 278, at p. 282. It was argued before the I.M.T. that this law did not apply generally because Art. 2 of the Hague Convention expressly stated that its provisions bind only contracting parties and do not apply if all parties to an international conflict are not parties to the Convention. Several countries involved in World War II were not parties to the Hague Convention. However, in its judgment the I.M.T. said, at pp. 248-249: The rules of land warfare expressed in the Convention undoubtedly represented an advance over existing international law at the time of their adoption. but by 1939 these rules laid down in the Convention were recognized by all civilized nations, and were regarded as being declaratory of the laws and customs of war which are referred to in Article 6(b) of the Charter. The United Nations General Assembly subsequently adopted, in December 1946, by a unanimous vote, Resolution 95(I) which affirmed the principles of international law "recognized by" the Nuremberg Charter and the I.M.T.'s judgment. So, by 1939 the Hague Convention had been in existence for thirty-two years. By 1941, forty-one states had signed the Convention; twenty-five had deposited ratifications, including Germany (with the reservation of Art. 44 of the Rules) though not Australia: Australian Military Manual, p. 340. The Convention provided a reference after the major world conflict which occurred during that time and its provisions have been widely reflected and disseminated in various states' military manuals. In light of the precision of these rules and the length of time they had been in existence, together with states' reliance on them, there is sufficient evidence that a contravention of these conventional laws of war amounted to an offence in customary international law at the commencement of World War II. The relevant questions to be asked with respect to a war crime in this narrow sense, then, concern the scope of that crime and whether the Act, on its proper construction, properly implements its prosecution. Those questions will be considered later. Crimes against humanity As already noted, "crime against humanity" is a generic term in international law encompassing different kinds of maltreatment of civilian populations, including those of the same nationality as the perpetrator. The Commonwealth submitted that there are, in international law, three classes of crimes against humanity: crimes of persecution, which are reflected in s. 7(3)(a)(i); crimes of extermination, reflected in s. 7(3)(a)(ii); and other serious crimes against members of any civilian population. The Commonwealth said that it was "probably" the case that crimes of persecution and other serious crimes against civilian populations must be committed in the execution of or in connexion with war or occupation to be a crime at international law. In the case of crimes of extermination, on the other hand, it was said that no such connexion is — and presumably the Commonwealth meant, also, was — required. Although the submission referred to conduct in execution of or in connexion with "war or occupation", the thrust of the argument seemed to be in conformity with the limitation imposed on crimes against humanity by Art. 6(c) of the Nuremberg Charter, to be discussed later. That is to say, crimes against humanity, apart from crimes of extermination, must have been committed in execution of or in connexion with war crimes or the crime of waging aggressive war. There is little doubt that crimes against humanity, in each of these classes, now exist in international law either as treaty law or, probably, as a matter of customary international law: cf. Meron, "The Geneva Conventions as Customary Law", American Journal of International Law, vol. 81 (1987) 348. But the question is whether crimes against humanity were crimes in international law before 1945. There was no international agreement creating a crime against humanity. If the crime existed, it was a matter of customary law. A customary law comprises two elements: (i) general practice by states; and (ii) opinio juris, in other words, expressed opinion that such a crime exists. Material sources produced before 1945 are evidence of both of these elements; those produced after 1945 are evidence of opinio juris only, as they are statements of opinion as to the state of international law in the past. A survey of the material is useful. Although the Hague Convention and the Hague Rules did not themselves deal with the conduct of belligerents towards their own citizens, there is some suggestion in the Convention that its provisions (and therefore those of military manuals produced in consequence) were not intended to cover the field of legal protection accorded to civilian populations. The preamble to the Convention includes the so-called Martens clause: Until a more complete code of the laws of war can be drawn up, the High Contracting Parties deem it expedient to declare that, in cases not covered by the rules adopted by them, the inhabitants and the belligerents remain under the protection and governance of the principles of the law of nations, derived from the usages established among civilized peoples, from the laws of humanity, and from the dictates of the public conscience. (Emphasis added.) As can be seen, there is an acceptance that binding humanitarian norms existed apart from the rules dealt with by the Convention itself. After World War I the Commission on the Responsibility of the Authors of the War and on Enforcement of Penalties presented a report to the Preliminary Peace Conference at Versailles in March 1919 (reproduced in American Journal of International Law, vol. 14 (1920) 95). The Commission made findings as to the progress of the War and made recommendations for prosecutions and for the establishment of an international tribunal. Several times in its report the Commission used the term "laws of humanity" or "dictates of humanity" or similar phrases. For example, one conclusion drawn by the Commission, at p. 117, was: All persons belonging to enemy countries who have been guilty of offences against the laws and customs of war or the laws of humanity, are liable to criminal prosecution. In relation to the tribunal to be set up to try the crimes the Commission said, at p. 118: Two classes of culpable acts present themselves: (a) Acts which provoked the world war and accompanied its inception. (b) Violations of the laws and customs of war and the laws of humanity. In In re Altstötter, as reported in (1947) 6 L. Rep. Trials War Crims 1, at p. 46, the U.S.M.T. said: Since the World War of 1914-1918, there has developed in many quarters evidence of an international interest and concern in relation to what was previously regarded as belonging exclusively to the domestic affairs of the individual State; and with that interest there has been an increasing readiness to seek and find a connexion between domestic abuses and the maintenance of the general peace. Reference was also made to instances in which states had intervened to prevent abuse by another state of its own subjects, including French intervention to check religious atrocities in Lebanon in 1861 and national protests directed towards Roumania and Russia with respect to aggression against Jews and towards Turkey on behalf of persecuted Christian minorities. Oppenheim, International Law, 3rd ed. (1920), vol. I, expressed doubt that there was, then, "really a rule of the Law of Nations" which permitted intervention on humanitarian grounds. However, he said, at p. 229: Many jurists maintain that intervention is admissible, or even has a basis of right, when exercised in the interest of humanity for the purpose of stopping religious persecution and endless cruelties in time of peace and war. That the Powers have in the past exercised intervention on these grounds, there is no doubt. Thus Great Britain, France, and Russia intervened in 1827 in the struggle between revolutionary Greece and Turkey. And Bluntschel, Das Moderne Volkerrecht der Civilisierten Staaten, 3rd ed. (1878), (quoted in In re Altstötter [33] , said, at p. 270: States are allowed to interfere in the name of international law if "humanity rights" are violated to the detriment of any single race. 1. (1947) 2 L. Rep. Trials War Crims, at p. 47. Next, the conduct which forms the substance of a "war crime" for the purposes of the Act, namely, murder, manslaughter, wounding, kidnapping and various sexual offences under Australian municipal law, attracted criminal sanctions before 1945 in most, if not all, of the states which were parties to World War II. This does not constitute that conduct an international crime but it is evidence of state practice concerning conduct between nationals of the same country. In determining whether a rule of justice may be declared an international law, it is relevant that each individual state condemns the conduct the subject of the rule: cf. In re List [34] . 1. (1948) 15 Annual Digest, at p. 633. At the end of World War II, crimes against humanity were dealt with in the following way. Article 6(c) of the Nuremberg Charter of 1945 defined "crimes against humanity" as: murder, extermination, enslavement, deportation, and other inhumane acts committed against any civilian population, before or during the war, or persecutions on political, racial or religious grounds in execution of or in connexion with any crime within the jurisdiction of the Tribunal, whether or not in violation of the domestic law of the country where perpetrated. (Emphasis added.) The paragraph contains two important limitations. First, a crime against humanity must comprise conduct directed at a civilian population. Isolated acts against individuals, unconnected with a larger design to persecute or exterminate a population, are not within the definition of the crime, whether committed by an individual or by a state authority: see, for instance, In re Altstötter (1947) 14 Annual Digest, at p. 284. The second limitation is that crimes against humanity must have been committed "in execution of or in connexion with any crime within the jurisdiction of the Tribunal", that is, war crimes (Art. 6(b)) or crimes against peace or waging aggressive war: Art. 6(a). The second limitation applies both to acts of persecution and to acts of extermination. As to the grammatical change made to Art. 6(c), which makes the intention of the contracting parties to this effect unequivocal, see Schwelb, "Crimes Against Humanity", The British Year Book of International Law, vol. XXIII (1946) 178 (hereafter "Schwelb"), pp. 193-195. Only crimes against humanity committed during the period of the war were held to be capable of founding a conviction because only those acts could be seen to satisfy the limitation. In its judgment, the I.M.T. concluded, at p. 249: To constitute Crimes against Humanity, the acts relied on before the outbreak of war must have been in execution of, or in connexion with, any crime within the jurisdiction of the Tribunal. The Tribunal is of the opinion that revolting and horrible as many of these crimes were, it has not been satisfactorily proved that they were done in execution of, or in connexion with, any such crime. The Tribunal therefore cannot make a general declaration that the acts before 1939 were Crimes against Humanity within the meaning of the Charter, but from the beginning of the war in 1939 War Crimes were committed on a vast scale, which were also Crimes against Humanity; and insofar as the inhumane acts charged in the Indictment, and committed after the beginning of the war, did not constitute War Crimes, they were all committed in execution of, or in connexion with, the aggressive war, and therefore constituted Crimes against Humanity. The I.M.T. judgment does not throw much light on what "in execution of, or in connexion with" means. The Tribunal found that conduct amounting to a crime against humanity either was a war crime or was done in execution of or in connexion with a crime against peace. And, if done during the War, that connexion seems to have been assumed. Be that as it may, Art. 6(c) of the Charter evidences a conceptually distinct crime where the conduct constituting the crime is, in a practical sense, associated with other conduct which amounts to a war crime or a crime against peace. On 20 December 1945 the Control Council for Germany, comprising representatives of Britain, the United States, France and U.S.S.R., enacted a law for the punishment of persons guilty of, inter alia, "crimes against humanity". This law, generally known as Control Council Law No. 10 ("C.C. Law No. 10"), was passed to effect the prosecution of war criminals other than the major actors dealt with by the I.M.T. Article II 1(c) of C.C. Law No. 10 defined crimes against humanity in substantially the same terms as did the Nuremberg Charter but, significantly, did not contain the words "in execution of or in connexion with any crime within the jurisdiction of the Tribunal". The U.S.M.T., whose jurisdiction emanated from C.C. Law No. 10, said, in In re Altstötter [35] : The [Nuremberg] Charter, the I.M.T. Judgment, and C.C. Law 10 constitute authoritative recognition of principles of individual penal responsibility in international affairs which had been developing for many years. Surely C.C. Law 10, which was enacted by the authorized representatives of the four greatest powers on earth, is entitled to judicial respect when it states, "Each of the following acts is recognized as a crime". Although the U.S.M.T. drew a general conclusion concerning C.C. Law No. 10, the context of the statement is a discussion about Art. 6(b) of the Nuremberg Charter and its equivalent, Art. II 1(b) of C.C. Law No. 10, not about crimes against humanity. However, later [36] , the Tribunal said: Whether the crime against humanity is the product of statute or of common international law, or, as we believe, of both, we find no injustice to persons tried for such crimes. They are chargeable with knowledge that such acts were wrong and were punishable when committed. 1. (1947) 14 Annual Digest, at p. 282. 2. (1947) 14 Annual Digest, at p. 285. In In re List the defendants were charged with crimes which all came within the scope of war crimes in international law. But the U.S.M.T. discussed C.C. Law No. 10 generally, saying [37] : The crimes defined in Control Council Law No. 10 were crimes under pre-existing rules of International Law — some by conventional law and some by customary law. 1. (1948) 15 Annual Digest, at p. 634. There were, therefore, significantly different claims by the U.S.M.T., operating under the authority of C.C. Law No. 10, and by the I.M.T., operating under the authority of the Nuremberg Charter, as to the state of international law before 1945. The former tribunal claimed that a crime against humanity was an independent crime under customary international law; the latter tribunal required its connexion with a war crime or a crime against peace. The differences may, as Dr Egon Schwelb points out, reflect the difference in the legal nature of the two instruments and in the status of the two tribunals created to exercise jurisdiction: Schwelb, pp. 218-219. The I.M.T. was, in addition to being an occupation court for Germany, also, to some extent, an international judicial body administering international law. Being an international judicial organ, the I.M.T.'s jurisdiction in domestic matters of Germany was circumscribed. The zonal tribunals applying C.C. Law No. 10, on the other hand, were arguably in the nature of local courts administering primarily municipal law and therefore not limited by the settled boundaries of international law: cf. In re Altstötter [38] , where the U.S.M.T. itself concluded that it exercised international jurisdiction. 1. (1947) 14 Annual Digest, at pp. 278-279, 287. One more kind of evidence relating to crimes against humanity should be considered: enabling legislation of countries which do not have territorial jurisdiction in international law with respect to war criminals from World War II, and the resulting prosecutions of those persons in municipal courts some time after 1945. These sources are evidence of opinio juris of states though their law-making capacity, as practice, is not relevant when considering the state of international law in the past. In 1950, the Israeli Parliament enacted legislation for the prosecution of, inter alia, crimes against humanity, defined substantially in accordance with C.C. Law No. 10: the Nazi and Nazi Collaborators (Punishment) Law 1950 Israel. Attorney-General (Israel) v. Eichmann was prosecuted under this legislation. The Court concluded [39] , that the crimes for which the appellant was convicted, including crimes against humanity, "must be regarded as having been prohibited by the law of nations since "time immemorial" ". 1. (1962) 36 I.L.R., at p. 283. In 1987 the Canadian Parliament amended the Canadian Criminal Code to provide for the prosecution of war crimes and crimes against humanity. The conduct in the definition of crimes against humanity is similar in scope to that in Art. 6(c) of the Nuremberg Charter; but no connexion with other crimes is required. The Canadian definition, however, does require conduct amounting to a crime against humanity to constitute a "contravention of customary international law or conventional international law" or to be "criminal according to the general principles of law recognized by the community of nations": s. 7(3. 76). This latter formulation corresponds to the terms of the relevant Canadian constitutional provision. Reg. v. Finta [40] , in the Ontario High Court of Justice, held the Canadian legislation to be constitutionally valid, in that it was not retroactive and because crimes against humanity, as defined in the legislation, existed in international law before 1945. This conclusion was based on a survey of conventions and agreements and other relevant material [41] . However, the Court's opinion as to the limits of a crime against humanity is not clear. In part, the decision was based on the judgment of the I.M.T. Callaghan A.C.J.H.C. said [42] , that he accepted the reasoning of the I.M.T.: who indicated that "by 1939 these rules laid down in the [Hague] Convention were recognized by all civilized nations, and were regarded as being declaratory of the laws and customs of war". But, as Brennan J. points out in his judgment in the present case, the words quoted referred only to war crimes as defined in Art. 6(b) and not to crimes against humanity defined in par. (c). Also, as noted earlier, "crimes against humanity" are defined very broadly in the Canadian legislation. Callaghan A.C.J.H.C. concluded [42] : I am of the opinion that war crimes and crimes against humanity were, by 1939, offences at international law or criminal according to the general principles of law recognized by the community of nations. (Emphasis added.) The distinction between "general principles" of law and international law, although corresponding to the distinction made in the relevant Canadian constitutional provision (and in equivalent international law, for example, in Art. 15 of the International Covenant on Civil and Political Rights), was not spelt out. 1. (1989) 61 D.L.R. (4th) 85. 2. (1989) 61 D.L.R. (4th), at pp. 97-103. 3. (1989) 61 D.L.R. (4th), at p 101. 4. (1989) 61 D.L.R. (4th), at p 101. The United Kingdom War Crimes Act 1991 provides for prosecutions for murder, manslaughter or culpable homicide committed between 1 September 1939 and 5 June 1945 in a place which was then part of Germany or under German occupation and which "constituted a violation of the laws and customs of war": s. 1(1). Thus no question directly arises as to crimes against humanity. The conclusions contained in War Crimes: Report of the War Crimes Inquiry, Cm. 744 (1989) ("the Hetherington Report") informed the preparation of the United Kingdom legislation. The report concluded, at par. 5.43: In 1939 there was no internationally accepted definition of crimes against humanity, as there was of violations of the laws and customs of war. The Nuremberg definition of 1945 appears partly to be based on the principle that some crimes are so patently against the laws of all civilized nations as to be regarded as crimes in international law, prosecutable by any nation. [However] while the moral justification for trying crimes against humanity at Nuremberg is understandable, the legal justification is less clear. So, there is support in Israel's legislation for the existence at the relevant time of crimes against humanity defined according to C.C. Law No. 10. Also, Canadian legislation, defining crimes against humanity broadly, has been held to be an accurate reflection of the law at the time. But the United Kingdom legislation, by omission, carries the implication that crimes against humanity were not formulated sufficiently before 1945 to be binding rules of law. Crimes against humanity — conclusion It is impossible, perhaps, to say definitively what were the limits of crimes at international law between 1939 and 1945. This is not merely because of the state of historical record, but because of the nature of international law. The sources of international law and their relative status are not, and were not then, finally fixed. Documents such as those emanating from the United Nations and states' legislation are strong authority, but there is no heirarchy of judicial and legislative organs creating a system of binding precedent as in municipal law. For example, practice contrary to express intention does not necessarily attract legal sanction; and its status — the status of contravening practice — is unsettled also. Since no permanent international court of criminal justice exists to determine authoritatively the scope of international criminal law or to enforce sanctions for its breach, agreements and other documents evidencing international crimes do not function in the same way as statutes in municipal law. This is the case especially where crimes develop from customary practice of nations, but even where treaties exist between states. The absence of consistent enforcement and sanction means that documents evidencing international criminal laws cannot be scrutinized with the same intensity for the exact limits of the provisions they contain. It is not only unrealistic but incorrect to take an excessively technical approach. In In re Piracy Jure Gentium [43] Lord Sankey said: Speaking generally, in embarking upon international law, their Lordships are to a great extent in the realm of opinion, and in estimating the value of opinion it is permissible not only to seek a consensus of views, but to select what appear to be the better views upon the question. 1. [1934] A.C. 586, at pp. 588-589. There is a certain unease and evident moral, and legal, tension surrounding the question of crimes against humanity. This shows itself in various ways, as in the sometimes peremptory dealing with the question by the tribunals. See, for example, the statement of the U.S.M.T. in In re Altstötter [44] (already quoted), after a discussion of war crimes in the narrow sense: Surely C.C. Law 10, which was enacted by the authorized representatives of the four greatest powers on earth, is entitled to judicial respect when it states, "Each of the following acts is recognized as a crime". And also the readiness of the I.M.T. to make the connexion between the conduct in question and other crimes within its jurisdiction. The tension is further illustrated in the divergence of views represented in the current war crimes legislation in Israel, Canada and the United Kingdom. 1. (1947) 14 Annual Digest, at p. 282. Upon analysis, the moral tension is seen to be between a desire to ensure that fundamental justice is not avoided by an overly technical scrutiny and a fundamental objection to individuals being called to account by victors in a war according to laws which did not exist at the time; a fear, also, of justice being undermined. When analysed legally, and from the perspective of the time, the tension is seen to be between two fundamental notions: on the one hand, the doctrine of sovereignty with its concomitant principle of non-intervention between nations; and, on the other, fundamental principles of human rights, including the right of a people to be protected by the world community if abused by a sovereign power. Certainly, with the development of principles of human rights and the joint responsibility for their protection since, and largely as a result of, World War II, the limitations of a strictly defined doctrine of sovereignty, and exclusive rights with respect to the welfare of a group of people, have become increasingly evident: see, for example, Brilmayer, Justifying International Acts (1989), ch. 5 and ch. 7. However, humanitarian norms cannot be said to have been absent before 1945. Fundamental values and the laws of war themselves, being rules limiting the means of aggression not rules permitting violence, arise from a desire to preserve humanity and humaneness in relations between all people. They are themselves, in this sense, humanitarian norms. With this analysis in mind, I have reached the following conclusions. There is, on a survey of relevant material, evidence of the existence before 1939 of a consciousness of acts which offend fundamental human rights; these may be called crimes against humanity. This is to be found in diplomatic instances and legal commentary in the nineteenth century; in the report to the Preliminary Peace Conference of 1919; in the Martens clause in the Hague Convention (by implication); and in the consistency of sanction of similar crimes in municipal laws of individual states. Crimes which extended, conceptually, beyond war crimes were contemplated. But before 1939 there was no real indication of the boundaries of these crimes. Reference is made to the "laws of humanity" or the "dictates of conscience" but the scope of the offence does not emerge. Two statements of the scope of crimes against humanity appeared in the Nuremberg Charter and C.C. Law No. 10 in 1945, containing an important difference between them. Given that the I.M.T. was most clearly exercising international jurisdiction and that no precise definition of the crime had emerged prior to that time, the narrower view of the crime contained in the Nuremberg Charter must be preferred. It follows that, at the relevant time, conduct which amounted to persecution on the relevant grounds, or extermination of a civilian population, including a civilian population of the same nationality as the offender, constituted a crime in international law only if it was proved that the conduct was itself a war crime or was done in execution of or in connexion with a war crime. The conduct in respect of which the plaintiff is charged is the murder of a number of persons, either Jews or those suspected of being partisans or communists. The paragraphs of the information relating to those suspected of being partisans or communists allege expressly that the conduct was in pursuance of a policy of annihilation "contrary to the laws of war", in other words that the conduct was a war crime. The paragraphs relating to Jewish people allege that the murders were committed during and in the course of the German occupation of the Ukraine, and either in pursuing Germany's policy of persecution of Jewish people or with intent to destroy Jewish people. In so far as the information alleges a murder in pursuance of Germany's policy of persecution, the conduct is done in execution of or in connexion with a war crime in international law, the war crime being Germany's planned persecution in occupied Europe of people by reason of their race, or their political or religious beliefs. In so far as the information alleges a murder with intent to destroy the Jewish people, the conduct is alleged to have been committed "during and in the course of" the German occupation. "In the course of", which reflects the terminology of s. 7(1) of the Act, implies, in this context, more than a temporal connexion between the murder and the occupation. If it meant otherwise, it would add nothing to "during" the occupation. The allegation is sufficient, therefore, to amount itself to a war crime, as well as a crime against humanity. If, for any reason, the allegation of murder with intent to destroy the Jewish people is insufficient, in context, to amount to a war crime, the information would fail to describe conduct which amounted to a crime against humanity as defined in international law at the relevant time and would fail to be within the universal jurisdiction to prosecute that crime. On its face, however, the conduct alleged against the plaintiff constituted a war crime or a crime against humanity at the relevant time. The Act and crimes in international law The next matter for consideration is whether the definition of war crime in s. 7 of the Act represents a valid exercise of Commonwealth legislative power as an appropriate exercise of universal jurisdiction conferred in international law with respect to war crimes, in the narrow sense, or crimes against humanity. "Appropriate" in this context means, of course, "effective"; it does not imply a power in this Court to review the desirability of the exercise of the jurisdiction. The question for the Court here is: does the Act permit a person to be convicted of an offence under s. 9 in circumstances where the conduct alleged against the person would not have amounted to a war crime or a crime against humanity under international law? Section 17 of the Act plays a decisive role in this determination. Sub-section (2) of that section is the central provision. It reads: Subject to section 16, it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. It follows from the earlier discussion as to crimes against humanity that par. (b) of s. 17(2) can have only an operation consistent with the restricted understanding of that crime in international law. As to par. (a), since, as Baxter says, the international law of war is " "prohibitive law" Belligerent acts in war are facts, not legal rights" ("The Municipal and International Law Basis of Jurisdiction over War Crimes", The British Year Book of International Law vol. XXVIII (1951) 382 (hereafter "Baxter"), at p. 388), it is not accurate to speak of acts which are "permitted" by the laws, customs and usages of war. However, although inaccurate, the meaning of s. 17(2)(a) is not unclear. Questions of onus of proof aside, there is no difficulty in construing the positive language of the paragraph as invoking the prohibitive provisions of the laws of war. The words "was permitted by the laws, customs and usages of war" must be taken to refer to conduct which "did not contravene the laws, customs and usages of war". The Commonwealth submitted that s. 17 is properly to be considered integral to the formation of the offence contained in s. 9, with the result that the offence created by that section is confined by the limits of the relevant crimes at international law. This conclusion was said to result from two possible constructions of the relevant sections. The first is that the provisions of s. 17 have the effect of creating an implied element in the offence created by the Act. In other words, it was submitted, words should be read into both s. 7(1) and s. 7(3), after "A serious crime is a war crime", to the effect: "which is a war crime or a crime against humanity at international law". This would involve the prosecution proving, in every case, beyond reasonable doubt, the facts which constitute the elements of the international crime alleged to have been committed. The analogy, on this construction, is the Crown's onus in a prosecution for murder to prove an intention on the part of the accused to kill or do grievous bodily harm. If an "element of an offence" is to be understood in this strict sense, the suggested construction of ss. 7 and 17 is not supported by the words and structure of the Act. Section 9 contains the elements of the offence created by the Act. A person is guilty of an indictable offence if he or she has (1) committed a "war crime", (2) between 1 September 1939 and May 1945. With no express reference to war crimes or crimes against humanity in international law in s. 9, or in ss. 6 or 7, and without more compelling reason than a general evocation of international law in, for example, the terms of s. 7(3)(a)(i) and (ii), there is no justification for implying those crimes as an element (in the strict sense referred to) of the offence in s. 9. This does not mean, however, that the offence created by the Act may not be limited by the scope of crimes at international law referred to in s. 17. As an alternative construction, the Commonwealth submitted that s. 17 and the provisions of the Act as a whole have this effect because s. 17 makes contravention of an international criminal law a "prerequisite of guilt", rather than an element of the offence. There are two ways, it was said, in which the defence in s. 17(2), may be so raised. The accused may, in effect, demur to the indictment by claiming that the allegations in the charge do not, as a matter of law, constitute an offence against the laws, customs and usages of war or a crime against humanity. This would be a matter for the judge to decide and, if satisfied of the accused's submission, the prosecution would not proceed. If no facts needed to be determined, the jury would not be involved. An analogy here is a question as to the jurisdiction of a court. The prosecution, or a plaintiff in a civil case, need not prove that the claim made falls within the jurisdiction of a court; but if the issue is raised it must be resolved in favour of the prosecution, or plaintiff, before the proceedings may continue. Mason C.J. and Dawson J. observed in Thompson v. The Queen [45] : Proof of jurisdiction is a prerequisite of guilt but otherwise it is not an element in proof of the commission of the offence except in those cases in which the offence is so defined that commission of it in a place or locality is made an element of the offence charged. 1. (1989) 169 C.L.R. 1, at p. 12. The second way in which it was said that an accused may raise the "defence" and claim that liability under the Act does not attach is to introduce additional facts which take the conduct alleged out of the category of a crime in international law. The facts alleged by the prosecution may, if proved, constitute a war crime in international law and for the purposes of the Act. For example, the facts alleged may amount to the manslaughter of civilians in occupied territory. But additional facts may raise a sufficient doubt as to the military necessity of the conduct. Section 17(3) expressly provides that conduct arising from military necessity is permitted by the laws, customs and usages of war. If this is the way liability is denied — by the assertion of additional facts — s. 17(4) requires that there be evidence of the existence of those facts. The language of s. 17(4) does suggest that this second way is the only way to deny liability under s. 17(2) but this literal reading cannot be taken to preclude the right to demur. Sub-section (4) is concerned to preclude reliance on sub-s. (2) where the defendant relies on additional facts but there is no evidence of them, merely assertion. As to that situation, see s. 13(5). In so far as a denial of liability does require evidence of additional facts, a defendant has an evidentiary onus in that regard. The word "defence" is used inaptly in s. 17 because of the onus cast upon the prosecution if there is evidence of the existence of the facts constituting the defence: see s. 17(5). However, there is nothing unusual in a defendant having an evidentiary onus nor, for that matter, in an inapt use of the word "defence" in a statute. An analogy for this view of the operation of s. 17(2) is self-defence in the criminal law. The absence of circumstances which may amount to self-defence is not, in the strict sense, an element of murder. An evidentiary onus lies on an accused to adduce facts which are capable of raising a doubt as to whether the conduct in relation to which he or she is charged was done by way of self-defence. But, despite the name, self-defence is not a defence in the sense that an accused must prove the facts on which he or she relies. I have dealt in some detail with this view of s. 17, involving the proposition that the section has the effect of limiting the operation of ss. 7 and 9 without creating an implied element in s. 7, because it seems to me the correct construction. Prosecutions under s. 9 must be in accordance with the whole Act. Apart from the particular provisions of s. 17, there are indications in the Act — in specific provisions and general structure — which point strongly to the offence created in s. 9 being predicated on the international law crimes of war crimes and crimes against humanity as formulated at the relevant time. The indications are as follows: Specific provisions (i) The terms of s. 7(3) correspond to phrases and concepts in both Art. 6(c) of the Nuremberg Charter and Art. II of the 1948 Convention on the Prevention and Punishment of the Crime of Genocide (the "Genocide Convention"). Article 6(c) speaks of "persecutions on political, racial or religious grounds"; s. 7(3)(a)(i) speaks of a serious crime committed "in the course of political, racial or religious persecution". The difference between sub-pars (i) and (ii) of s. 7(3)(a) also corresponds to the distinction between acts of persecution and acts of extermination, central to Art. 6(c). With respect to acts of extermination, the reference in s. 7(3)(a)(ii) to serious crimes committed "with intent to destroy in whole or in part a national, ethnic, racial or religious group" corresponds precisely to the language of Art. II of the Genocide Convention. (ii) Section 7(1) provides that a serious crime may be a "war crime" in either of two contexts: during hostilities or in the course of an occupation. These are the two contexts dealt with by the laws and customs of war and which distinguish violations of those laws and customs from crimes against humanity per se. And nothing in s. 7(1) requires there to be conduct, for the purposes of s. 9, which is inconsistent with conduct which amounts to a war crime in the narrow sense. The plaintiff relied on s. 7(2), which makes a merely incidental or remote connexion, whether in time, in time and place, or otherwise, insufficient to constitute conduct "in the course of" hostilities or an occupation, in order to support the proposition that a mere temporal or spatial connexion would, in some circumstances, be sufficient to found a prosecution so long as it was not "incidental or remote". This, on its face, would exceed international law. The Commonwealth, on the other hand, argued that s. 7(1), read in the light of sub-s. (2), requires a strong, substantial connexion and that mere temporal or spatial connexion is insufficient. This latter construction is the better view. The use of brackets, instead of commas, in s. 7(2) has the effect of emphasising the main text of the sub-section, which simply makes incidental or remote connexion insufficient. The subject matter of the provision is the degree, not the kind of connexion. Although not conclusive, this construction lends support to the proposition that s. 7(1) remains within the limits of international law. Such a construction is supported by the ordinary meaning of the phrase "in the course of" which carries with it the idea of a relationship between the substance of the two things connected. Conduct need not be calculated by sober military minds to achieve a military end; neither need war crimes in international law. The sexual assault of women inhabitants by occupying soldiers during an occupation would be "in the course of" the occupation and would be a war crime in international law but need not be (though it may be) calculated to further the ultimate military object of occupation. (iii) The language of ss. 16, 6(6) and, by implication, s. 20 of the Act also refers to concepts familiar in international law. See the discussion of ss. 16 and 20 later in these reasons. Structure (iv) The structure of s. 7 itself, divided into two substantive parts — sub-ss. (1) and (3), reflects the two categories of crimes under discussion and contemplated in virtually all literature in this area of international law: war crimes and crimes against humanity. (v) The structure of the "definition" of a "war crime" for the purposes of the Act indicates that the Act should properly be construed to stay within the limits of international law referred to in s. 17. A "serious crime" in s. 6 is defined by reference to municipal law; s. 7 defines the circumstances within which those notional municipal offences can be prosecuted under the Act. Section 7 itself would be unnecessary if no invocation of international law were intended. The dual structure of the s. 9 offence, therefore, supports the proposition that two systems of law are contemplated in the application of the Act. (vi) Section 9 is the first section in Pt III of the Act, entitled "War Crimes". Many of the sections in that Part which follow s. 9 have the effect of limiting the scope of the prosecutions possible under s. 9. Some limit the prosecutions procedurally only: see, for example, s. 14. Some sections, on the other hand, have a substantive operation, creating, in effect, a condition precedent for prosecution. For instance, s. 11 limits prosecutions to those against Australian citizens or residents. This is not an element of the offence but, if the question is raised, the prosecution must be able to establish the citizenship or residency of the accused. Thus it is consistent with the scheme of the Act that s. 9 be read in accordance with the limitations contained in Pt III of the Act. The Act and international law Subject to a qualification to be discussed shortly, the Act evidences a sufficient intention that it does not seek to trespass beyond the boundaries of international law as they have been discussed in this judgment. Although a "war crime" is defined in terms which do not of themselves attract notions of international law, the scope of s. 17 and the Act as a whole warrant a conclusion that the Act is in accord with what international law understands as war crimes and crimes against humanity at the relevant time. The qualification in question exists by reason of s. 16 which reads: Subject to subsections 6(2) and 13(2), the fact that, in doing an act alleged to be an offence against this Act, a person acted under orders of his or her government or of a superior is not a defence in a proceeding for the offence, but may, if the person is convicted of the offence, be taken into account in determining the proper sentence. The language of s. 16 is very close to that of Art. 8 of the Nuremberg Charter agreed to on 8 August 1945 and, no doubt, was taken from that instrument. Section 20 of the Act reads: "Subsection 6(6) and section 16 are enacted to avoid doubt." The "doubt" was not clearly identified in the submissions of counsel; indeed virtually nothing was said of s. 16 in the course of argument. If s. 16 reflects international law at the times specified in the information, the correspondence between the Act and that law may, broadly speaking, be said to be complete. But what if it does not? Referring to the various trials by Allied courts after World War II, Professor Starke has commented: It appears clearly established also by the above-mentioned post-war trials (see, for example, the judgment of the Nuremberg Court) that orders by superiors, or obedience to national laws or regulations, do not constitute a defence, but may be urged in mitigation of punishment: Introduction to International Law, 9th ed. (1984), p. 529. Of course the question in the present case is the state of international law between 1942 and 1943 when the offences alleged against the plaintiff are said to have been committed. The view that the doctrine of superior orders was available to service personnel charged with war crimes was asserted by Professor Oppenheim in his International Law as late as the fifth edition, published in 1935: see vol. II, pp. 453-454. That view was reflected in the Australian Military Manual which, at the outbreak of World War II, said, in ch. XIV, Art. 443: It is important, however, to note that members of the armed forces who commit such violations of the recognized rules of warfare as are ordered by their Government, or by their commander, are not war criminals and cannot therefore be punished by the enemy. But when Professor Lauterpacht edited the sixth edition of Oppenheim in 1940, he rejected the earlier view and expressed his opinion in terms corresponding to s. 16 of the Act: see vol. II, pp. 453-454. The Australian Military Manual was amended to give effect to the changed opinion on 30 September 1944. In In re List [46] , the U.S.M.T. said of the defence of superior orders: The defence relies heavily upon the writings of Prof. L. Oppenheim to sustain their position. It is true that he advocated this principle throughout his writings. As a coauthor of the British Manual of Military Law, he incorporated the principle there. It seems also to have found its way into the United States Rules of Land Warfare (1940). We think Professor Oppenheim espoused a decidedly minority view. The fact that the British and American armies may have adopted it for the regulation of their own armies as a matter of policy, does not have the effect of enthroning it as a rule of International Law. 1. (1948) 15 Annual Digest, at p. 650. It is significant that as early as 1921, in the case of The Llandovery Castle [47] , a German court held the defence of superior orders to be unavailable if the order were "universally known to be against the law". 1. (1921) 2 Annual Digest 436. Now it is true that the author of a major study — Dinstein, The Defence of Obedience to Superior Orders in International Law, (1965) (hereafter Dinstein) — ends, at p. 253, with these words: "I cannot help but conclude my study of the subject of obedience to superior orders in international law with some incertitude." But, as noted, there is a body of authority and opinion that between 1942 and 1943 the defence of superior orders was not available in the case of a prosecution for a war crime. The point does not seem to have been debated in relation to crimes against humanity as opposed to war crimes in the narrower sense. But, given that crimes against humanity need not be defined by reference to military relationships, this may not be surprising. Section 16 of the Act says no more than that "the fact that a person acted under orders is not a defence". It says nothing of duress, other defences made available by ss. 6(2) and 13(2)(f) of the Act, or the implications of an order for the mens rea of the defendant: in this respect, see the discussion in Dinstein, at pp. 251-252. The section is in truth very limited in its operation. To the extent that there may be "doubt" (s. 20) as to the availability of the defence of superior orders in the "war" to which the Act applies, I do not think that the doubt is sufficient to destroy the general correspondence between the Act and international law. It follows then that the Act is supportable as an exercise of the universal jurisdiction and, in that regard, as an exercise of the external affairs power. Return to externality I return now to where this judgment began, namely, that the Act is a law with respect to external affairs because it relates to conduct that took place during a war which touched and concerned Australia, whether or not Australia was directly involved in the particular conflict from which the conduct arose. On this approach, the Act is viewed simply as municipal law which penalizes certain conduct answering the statutory definition of a "war crime" between 1 September 1939 and 8 May 1945. And, subject to what follows concerning Ch. III of the Constitution, the Act is a valid exercise of the external affairs power. Defence power The Commonwealth contended that the Act is supportable as an exercise of the defence power in s. 51(vi). But, for the reasons given by Brennan J., to which I do not wish to add, the Act is not so supportable. It is therefore unnecessary to say more about this aspect of the case. Constitutional limitations — Chapter III Sub-section (xxix) of s. 51 of the Constitution, as with the other sub-sections of that section, is expressed to be "subject to this Constitution". It is necessary therefore that the Act does not offend any part of the Constitution, relevantly, that it does not offend Ch. III — The Judicature. The provisions of Ch. III of the Constitution function to achieve the independence of the judiciary for two related ends. First, they ensure the institutional separation of the site of judicial power from those of executive and legislative powers so that the courts may operate as a check, through review, on the other arms of government. Secondly, the independence of the judiciary is protected so as to ensure that cases are decided free from domination by other branches of government and in accordance with judicial process: see Harris v. Caladine [48] . 1. (1991) 172 C.L.R. 84, at pp. 133-135. The limits of judicial power are difficult to ascertain precisely; the line between an essentially legislative or executive function and a judicial function is not always obvious. However, it is clear that the determination of criminal liability is, historically and by its nature, the exercise of a purely judicial power: Waterside Workers' Federation of Australia v. J.W. Alexander Ltd. [49] . 1. (1918) 25 C.L.R. 434, at pp. 443-444. The essence of judicial power was stated by Kitto J. in The Queen v. Trade Practices Tribunal; Ex parte Tasmanian Breweries Pty. Ltd. [50] , when his Honour said: Thus a judicial power involves, as a general rule, a decision settling for the future, as between defined persons or classes of persons, a question as to the existence of a right or obligation, so that an exercise of the power creates a new charter by reference to which that question is in future to be decided as between those persons or classes of persons. In other words, the process to be followed must generally be an inquiry concerning the law as it is and the facts as they are, followed by an application of the law as determined to the facts as determined; and the end to be reached must be an act which, so long as it stands, entitles and obliges the persons between whom it intervenes, to observance of the rights and obligations that the application of law to facts has shown to exist. Legislation enacted by the Parliament pursuant to s. 51 of the Constitution, which purported to require a court to which Ch. III applies to act otherwise than in accordance with these principles, would offend Ch. III. 1. (1970) 123 C.L.R. 361, at p. 374. Bills of attainder Bills of attainder (which impose the death penalty) and bills of pains and penalties (which impose a lesser penalty) may be defined as legislative acts imposing punishment on a specified person or persons or a class of persons without the safeguards of a judicial trial: see Lehmann, The Bill of Attainder Doctrine: A Survey of the Decisional Law, Hastings Constitutional Law Quarterly, vol. 5 (1978) 767, at pp. 790-791. Legislative acts of this character contravene Ch. III of the Constitution because they amount to an exercise of judicial power by the legislature. In such a case, membership of a group would be a legislative assessment as to the certainty, or at least likelihood to the criminal standard of proof, of an accused doing certain acts or having certain intentions. Those acts or intentions would not themselves be open to scrutiny by the court. The vice lies in the intrusion of the legislature into the judicial sphere: Murphy J. in Victoria v. Australian Building Construction Employees' and Builders Labourers' Federation [51] . 1. (1982) 152 C.L.R. 25, at p. 107. The Act, however, does not answer the description of a bill of attainder. It specifies no particular persons and it requires a trial. As to the kind of bill of attainder which confers guilt according to membership of a group, the Act confines the class of persons in relation to which it operates in the following ways. It deals only with conduct occurring between 1939 and 1945; war crimes outside of that time are not within its operation. More importantly perhaps, the Act's operation is confined to the European field of war, indicating that, at least in practice, it is the agents of European Axis countries who are the focus of the Act. But this falls short of the Act impliedly operating with respect to a specific class of persons. Guilt of a war crime for the purposes of s. 9 is to be determined, in substance, not just form, by an assessment of conduct and intent on the part of the accused. Proof of conduct and of the necessary state of mind which constitutes murder, manslaughter, wounding or the various sexual offences is too particular in its nature to amount, in these circumstances, to a disguised description of group membership. Even though it may be that members of a group are likely to be the ones prosecuted, no assumption is made that members of that group, by virtue of that membership, have, or are likely to have, acted in one or more of these ways. Retroactivity The Commonwealth submitted that there is no consitutional prohibition against the Parliament enacting retroactive legislation. It is of some importance to see how the submission was couched. Mr Rose, of counsel for the Commonwealth, said in relation to the external affairs power: of course the power is, like all the other powers in s 51, subject to Ch III, but, once the law is characterized as one with respect to external affairs, then like any of the legislation enacted under the other powers, Ch III does not enable a court to say, for example, that this is an unjust law. Whether a court may declare a statute to be invalid because it is unjust is a question that goes to the very heart of the relationship between the courts and Parliament: see, among the writings on the subject, Walker, Dicey's Dubious Dogma of Parliamentary Sovereignty: A Recent Fray with Freedom of Religion, Australian Law Journal vol. 59 (1985) 276. But that question does not arise here. The question is rather whether a law, which requires a court to which Ch. III applies to act contrary to accepted notions of judicial power, contravenes Ch. III and is therefore invalid to that extent. The question is not answered by the characterization of the law with respect to the sub-sections of s. 51. If the law is not in truth a law with respect to a particular head of power, it is not a valid law. It is only if the law meets that test that Ch. III need arise for consideration. When it does so arise, it operates to invalidate a law that does not comply with what it requires. I do not accept the submission of the Commonwealth in the absolute terms in which it was proffered. In legislation, judicial decisions and statements of principles, both of municipal and international law, there has emerged a general abhorrence of retroactive criminal law. The notion that there should be no crime or punishment, except in accordance with law, was recognized as early as 1651, when Hobbes wrote: "No law, made after a fact done, can make it a crime For before the law, there is no transgression of the law": Leviathan , (1651), Chs. 27-28, quoted in Glanville Williams, Criminal Law: The General Part, 2nd ed. (1961) (hereafter Williams), p. 580. The principle was formulated in Art. 8 of the French Declaration of the Rights of Man of 1789, it re-appeared in the French Constitution of 1791 and remains in the French Code Pénal. It headed the German Penal Code of 1871 and was guaranteed by the Weimar Constitution. Article I, s. 9, cl. 3 of the United States Constitution expressly prohibits ex post facto laws: see generally Williams, pp. 575-581; Hall, Nulla Poena Sine Lege, Yale Law Journal, vol. 47 (1937) 165. In international law the principle of non-retroactivity is enshrined in Art. 15(1) of the International Covenant on Civil and Political Rights (1966), which reads, inter alia: No one shall be held guilty of any criminal offence on account of any act or omission which did not constitute a criminal offence, under national or international law, at the time when it was committed. (See also Art. 7 of the European Convention for the Protection of Human Rights and Fundamental Freedoms, (1950); Art. 11(2) of the Universal Declaration of Human Rights, (1948); Art. 9 of the American Convention on Human Rights, (1969); Art. 7 of the African Charter on Human and Peoples' Rights, (1981).) There has been considerable debate surrounding Art. 15, tending to focus on a concern that it should not function to undermine the validity of agreements and legislation and resulting prosecutions of war criminals after World War II: see Bossuyt, Guide to the Travaux Préparatoires of the International Covenant on Civil and Political Rights (1987), p. 330; also the comments of The European Commission of Human Rights in X v. Belgium, Yearbook of the European Convention on Human Rights, vol. 4 (1961) 324. It is generally agreed, however, that non-retroactivity is a fundamental principle in international law: Reshetov, The Temporal Operation of Norms on Criminal Responsibility in Ginsburgs and Kudriavtsev (eds), The Nuremberg Trial and International Law (1990), 111, at pp. 111-113. All these general objections to retroactively applied criminal liability have their source in a fundamental notion of justice and fairness. They refer to the desire to ensure that individuals are reasonably free to maintain control of their lives by choosing to avoid conduct which will attract criminal sanction; a choice made impossible if conduct is assessed by rules made in the future. Thus, in Calder v. Bull [52] Chase J. said: no man should be compelled to do what the laws do not require; nor to refrain from acts which the laws permit. Laws should function to give reasonable warning of their operation and permit individuals to rely on that scope and meaning until expressly altered. Another nineteenth-century rationale for the principle was expressed in terms of specific deterrence: The reason why these laws are so universally condemned is, that they overlook the great object of all criminal law, which is, to hold up the fear and certainty of punishment as a counteracting motive, to the minds of persons tempted to crime, to prevent them from committing it. But a punishment prescribed after an act is done, cannot, of course, present any such motive: Jacquins v. Commonwealth [53] . 1. (1798) 3 U.S. 386, at p. 388. 2. (1852) 63 Mass. 279, at p. 281. Prohibition against retroactive laws protects a particular accused against potentially capricious state action. But the principle also represents a protection of a public interest. This is so, first, in the sense that every individual is, by the principle, assured that no future retribution by society can occur except by reference to rules presently known; and secondly, it serves to promote a just society by encouraging a climate of security and humanity. In so far as the principle of non-retroactivity protects an individual accused, it is arguably a mutable principle, the right to protection dependent, to some extent, on circumstances. Where, for example, the alleged moral transgression is extremely grave, where evidence of that transgression is particularly cogent or where the moral transgression is closely analogous to, but does not for some technical reason amount to, legal transgression, there is a strong argument that the public interest in seeing the transgressors called to account outweighs the need of society to protect an individual from prosecution on the basis that a law did not exist at the time of the conduct. But it is not only the issue of protection of an individual accused at the point of prosecution which is raised in the enactment of a retroactive criminal law. It is both aspects of the principle — individual and public interests — which require fundamental protection. In Calder v. Bull, Chase J. delineated the scope of the provision in the United States Constitution which prohibits ex post facto laws. He concluded [54] , that it proscribed four types of laws. They were: (i) laws that make an action done before the passing of the law, and which was innocent when done, criminal; and punish such action; (ii) laws that aggravate a crime, or make it greater than it was, when committed; (iii) laws that change the punishment, and inflict a greater punishment, than the law annexed to the crime, when committed; and (iv) laws that alter the legal rules of evidence, and receive less, or different, testimony than the law required at the time of the commission of the offence, in order to convict the offender. 1. (1798) 3 U.S., at p. 390. It is not the case that a law (even a criminal law) that operates retroactively thereby offends Ch. III of the Constitution. It is only if a law purports to operate in such a way as to require a court to act contrary to accepted notions of judicial power that a contravention of Ch. III may be involved. It is conceivable that a law, which purports to make criminal conduct which attracted no criminal sanction at the time it was done, may offend Ch. III, especially if the law excludes the ordinary indicia of judicial process. Such a law may strike at the heart of judicial power: see Liyanage v. The Queen [55] . But it is unnecessary to pursue this topic further because, as will appear, I do not consider that the Act, in its application to the information laid against the plaintiff, is retroactive in any offensive way. Likewise, I find it unnecessary to consider whether the decision in R. v. Kidman [56] is consistent with the operation of Ch. III as described in this judgment, though it is apparent that I do not share dicta which may be thought to suggest that an ex post facto law can never offend Ch. III: see, for instance, Isaacs J. [57] ; Higgins J. [58] ; Powers J. [59] . The Court was not invited to overrule Kidman and, in the absence of full argument on the point, it is preferable to say no more on this aspect. 1. [1967] 1 A.C. 259, at pp. 289-290. 2. (1915) 20 C.L.R. 425. 3. (1915) 20 C.L.R., at pp. 442-443. 4. (1915) 20 C.L.R., at p. 451. 5. (1915) 20 C.L.R., at p. 462. Retroactivity: The Act The Act is not offensively retroactive in relation to the information laid against the plaintiff. The information, except for par. 2 and the relevant part of par. 12, alleges that the plaintiff murdered one or more persons. Paragraphs 2 and 12 contain allegations that the plaintiff was "knowingly concerned in or party to the murders" of about 850 Jewish people in the village of Serniki. Section 9 of the Act refers only to conduct done in the past. Those who "committed a war crime" are guilty of an offence. The offence under s. 9 (strictly, unnamed, since "war crime" simply identifies certain conduct which amounts to the s. 9 offence) did not exist in Australian law at the time of the conduct, so the section is, in some sense at least, retrospective. But the Act causes no detriment to an accused where the offence charged rests on the notional crime of murder referred to in s. 6(1) and, by implication, s. 6(3). Under the laws of the Australian States, the intentional killing of a person in the absence of authorization, justification or excuse was at the relevant time an offence punishable by, at least, a maximum penalty no less than that imposed by s. 10 of the Act for an offence involving the wilful killing of a person: imprisonment for life. And there was no suggestion in argument that the law prevailing in the Ukraine at the relevant time was any more beneficial to an accused. Further, the universality of the condemnation of murder in municipal laws generally is reflected in the existence in international law of war crimes and crimes against humanity, both of which (so far as it is relevant here) are predicated on conduct which all Australian States and Territories would have identified as murder. Where the conduct alleged is murder, the circumstances described in s. 7, which must be satisfied to establish the offence in s. 9, do not alter the nature of the conduct which is punishable so as to take it outside the scope of the municipal crime of murder. They describe the circumstances in which a crime of murder will be a "war crime". Conduct constituting such an offence under the Act was conduct which attracted the sanction of criminal laws generally, not just the censure of moral codes. In those circumstances, it cannot be said that an individual is caused detriment to which he or she would not have been subject at the time of the conduct, or that he or she had "no cause to abstain" from that conduct. In its application to the information in pars 2 and 12, alleging that the plaintiff was "knowingly concerned in" the murder of several hundred Jewish people, the scope of the Act is less clear. The phrase reflects that in s. 6(1)(k)(iii) of the Act which is part of the sub-section increasing the range of parties to a crime who may be caught by the scope of s. 6. Those who attempt, conspire to commit, aid, abet, counsel or procure, or who are knowingly concerned in one of the crimes mentioned in pars (a) to (j) of the sub-section are also subject to its provisions. To be "knowingly concerned" is not generally an element of criminal conduct or a component of being a party to an offence in State criminal law involving offences to the person. The breadth of this phrase is made even larger by the language preceding it. Those who are "in any way, directly or indirectly, knowingly concerned in" a relevant offence may be prosecuted. Nevertheless, par. (k)(iii) is part of s. 6(1) and is therefore qualified by the words "and was, under the law then in force , an offence". The plaintiff did not argue that the words "knowingly concerned" should be struck out and I am not persuaded that their inclusion in s. 6 affects the validity of s. 9. Section 6 of the Act invokes State law offences whereas s. 9 creates a Commonwealth offence. The notional offences on which the "war crime" is based did not exist in Commonwealth law. But it does not follow from this that the Act is retroactive in any relevant sense. If it is insufficient that the crime of murder existed as a crime within Australian law as a whole, then, at least in the case of murder, it is sufficient that the principles on which the doctrine of non-retroactivity is based are not contravened. Questions may perhaps arise with respect to other criminal offences as to whether there was cause to abstain, but not in the case of murder. Likewise, questions may arise in relation to par. (j) of s. 6(1) which speaks of "an offence whose elements are substantially the same as the elements of an offence referred to" in the preceding paragraphs. But it too is qualified by the opening words of sub-s. (1). It is unnecessary to explore these matters. In its application to the information against the plaintiff the Act is not retroactive in a way offensive to Ch. III of the Constitution. Summary of conclusions There is an element of risk in attempting to summarise the contents of any judgment. But, having regard to the length of this judgment and the range of issues it canvasses, there is some justification for making the attempt. What follows is a summary of the judgment; it need hardly be said that the summary cannot be divorced from the context in which it appears. 1. The power of the Parliament to make laws with respect to "External affairs" (s. 51(xxix) of the Constitution) includes a power to make laws with respect to matters external to Australia which touch or concern Australia in some way. 2. The Act is a law with respect to a matter external to Australia, touching or concerning the national interest of Australia, in so far as it relates to conduct occurring outside Australia arising from "war" as defined. 3. There is insufficient evidence of any international obligation to seek out war criminals and bring them to trial to support the Act as an exercise of the external affairs power. 4. Likewise, there is insufficient evidence of any international concern that war criminals be tried in countries other than those in which their crimes were committed to support the Act as an exercise of the external affairs power. 5. The power of the Parliament to make laws with respect to "External affairs" includes a power to make laws with respect to international crimes which are subject to the universal jurisdiction. 6. The Act is a law with respect to external affairs in so far as it is an exercise of the universal jurisdiction to prosecute war crimes and crimes against humanity as formulated in international law at the relevant time. 7. The Act cannot be supported by reference to the defence power in s. 51(vi) of the Constitution. 8. The validity of the Act may be tested against the requirements of Ch. III of the Constitution, that is, the Act must not call for an exercise, by a court to which the Chapter applies, of what is not truly judicial power. 9. In its application to the information against the plaintiff, the Act does not offend Ch. III of the Constitution. It follows then that I answer the question reserved for the consideration of the Court as follows: Section 9 of the Act is not invalid in its application to the information laid against the plaintiff. Gaudron J. The relevant facts, statutory provisions and legislative history are set out in the judgments of Brennan J. and of Deane J. They reveal, amongst other things, that the War Crimes Act 1945 Cth ("the Act") was amended with effect from 25 January 1989 to create an indictable offence, called a "war crime", to be prosecuted in the courts of the States and internal Territories of Australia. See ss. 9 and 13 of the Act. The acts which constitute the offence created by s. 9 of the Act are confined to acts which occurred between 1 September 1939 and 8 May 1945 and which were connected in one or other of the ways specified in s. 7 of the Act with the Second World War in Europe or with an occupation arising out of that war: see the definitions of occupation and war in s. 5 of the Act. See also s. 7 and s. 9(1)(a). As a matter of construction, the offence may be constituted by acts committed in Australia or acts committed outside Australia. See ss. 6 and 7. Also as a matter of construction, the acts may have been committed by persons who were then citizens or residents of Australia or by persons then having no connexion with this country: see, again, ss. 6 and 7. However, s. 11 of the Act provides that only an Australian citizen or resident may be charged with a "war crime". It seems from the preamble to the Act, inserted by the 1988 amendments, that the expectation of the Parliament may have been that the Act would, in the main, operate with respect to "persons who committed serious war crimes in Europe during World War II [and who] since have entered Australia and [become] Australian citizens or residents". An information has been laid against the plaintiff charging him with a number of "war crimes". The crimes were allegedly committed on various dates between 1 September 1942 and 31 May 1943 in the Ukraine during or in the course of German occupation of that territory, in pursuit of German policy associated with that occupation or with the war, or on behalf of or in the interests of Germany, or, in some cases, with intent to destroy a racial group: see s. 7(1)(b), (c) and (d) and s. 7(3)(a)(ii) and (b). At no time during the period 1 September 1942 to 31 May 1943 was the plaintiff a citizen or a resident of Australia. He is now an Australian citizen, resident in South Australia. The question asked in the present case pursuant to s. 18 of the Judiciary Act 1903 Cth is this: "Is Section 9 of the War Crimes Act 1945 as amended, invalid in its application to the information laid against the plaintiff?" The assumption implicit in that question is that the operation of s. 9 with respect to the acts alleged in the information may be severed from the wider operation effected by ss. 6 and 7 of the Act. The arguments of the plaintiff and of the defendants were based on the further assumption that severability may be effected on the basis that the acts alleged occurred wholly outside Australia. For reasons which will later appear, it is unnecessary to decide the issue of severability, but, in my view, the assumptions to which I have referred are correct: see Bank of N.S.W. v. The Commonwealth [60] , per Dixon J.; R. v. Commonwealth Court of Conciliation and Arbitration; Ex parte Whybrow & Co. [61] , per Isaacs J. 1. (1948) 76 C.L.R. 1, at pp. 369-371. 2. (1910) 11 C.L.R. 1, at p. 54. The question referred pursuant to s. 18 of the Judiciary Act may be approached from various perspectives. By reason of the assumptions implicit in the question and in the arguments, I shall first consider whether, to the extent and by reason of its application to events which occurred wholly outside Australia, s. 9 is a law with respect to external affairs. In Koowarta v. Bjelke-Petersen [62] , Mason J. said that New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case") [63] "decided that the [external affairs] power extends to matters and things outside Australia". And Stephen J. [64] was of the same view. However, I do not think that the decision in the Seas and Submerged Lands Case was to that effect. 1. (1982) 153 C.L.R. 168, at p. 223. 2. (1975) 135 C.L.R. 337. 3. (1982) 153 C.L.R., at p. 211. In the Seas and Submerged Lands Case the sufficiency of the externality of the matter or thing to which the law in question applied was expressly accepted by Mason J. [65] and by Jacobs J. [66] . It may be that, as was accepted by Gibbs C.J. in Koowarta [67] , Barwick C.J. expressed his acceptance of its sufficiency when, having indicated that various matters did not limit the power, his Honour stated [68] that "[t]he power extends to any affair which in its nature is external to the continent of Australia and the island of Tasmania". However, that depends on what his Honour meant by the word "affair". Nothing else in the judgments in that case tends in favour of the view that externality is, of itself, sufficient to attract the power. Although that view was later expressly adopted by Murphy J. in The Commonwealth v. Tasmania (the Tasmanian Dam Case) [69] , his Honour's decision in the Seas and Submerged Lands Case was based on the view [70] that "[t]he Constitution, particularly s. 51(xxix), is intended to enable Australia to carry out its functions as an international person, fulfilling its international obligations and acting effectively as a member of the community of nations". 1. (1975) 135 C.L.R., at p. 471. 2. (1975) 135 C.L.R., at p. 497. 3. (1982) 153 C.L.R., at p. 190. 4. (1975) 135 C.L.R., at p. 360. 5. (1983) 158 C.L.R. 1, at pp. 171-172. 6. (1975) 135 C.L.R., at p. 503. Had the Seas and Submerged Lands Case decided that the external affairs power extends to matters or things outside Australia it would dictate the conclusion that, in its application to the matters alleged in the information — being, as already indicated, matters which occurred wholly outside Australia, s. 9 of the Act is a law with respect to external affairs. Equally, were there a decision to the opposite effect, that would dictate the opposite result. But there is no decision one way or the other. Thus, it is necessary to ascertain whether the power with respect to external affairs extends to acts, matters or things which are outside Australia by reference to the meaning of the words conferring the power. The legislative power conferred by s. 51(xxix) of the Constitution is a power "to make laws for the peace, order, and good government of the Commonwealth with respect to [e]xternal affairs". The extent of that power depends on the meaning of the words "external affairs", ascertained in accordance with the ordinary rule of constitutional interpretation that they "be construed with all the generality which the words admit": Reg. v. Public Vehicles Licensing Appeal Tribunal (Tas.); Ex parte Australian National Airways Pty. Ltd. [71] . In particular, their meaning is not cut down by the words "for the peace, order, and good government of the Commonwealth". Those words neither identify nor limit the subject matter of legislative power, and they do not now (and it is very much to be doubted that they ever did) import a limitation upon the ability to legislate extraterritorially. Moreover, they indicate that, so long as the proposed measure is one with respect to some subject matter entrusted to the Parliament, it is for the Parliament to decide the nature and extent of the measure, if any, to be enacted on that topic: see Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [72] . See also Pearce v. Florenca [73] , and Union Steamship Co. of Australia Pty. Ltd. v. King [74] . 1. (1964) 113 C.L.R. 207, at p. 225. 2. (1959) 103 C.L.R. 256, at pp. 306-308. 3. (1976) 135 C.L.R. 507, at pp. 515-516. 4. (1988) 166 C.L.R. 1, at pp. 12-13. The word "affairs" in the expression "external affairs" is a word of considerable generality. In its ordinary usage, "affair" may import a connexion or relationship, but not necessarily of any precise or formal kind. Rather, "affair" signifies a relationship of a vague and general kind involving some interest or concern transcending mere curiosity. If that notion of connexion or relationship is imported into the phrase "[e]xternal affairs" in s. 51(xxix), the phrase extends to acts, matters or things external to Australia which attract the interest or concern of the Australian body politic. And that interest or concern is necessarily established if an external act, matter or thing has been selected by the Parliament of the Commonwealth as the act, matter or thing to which its legislation should apply. Thus, when the validity of legislation is in question, it will necessarily be a law with respect to external affairs if it applies to "matters or things geographically situated outside Australia": the Seas and Submerged Lands Case [75] , per Mason J. 1. (1975) 135 C.L.R., at p. 471. The view that a law enacted by the Parliament of the Commonwealth is necessarily a law with respect to external affairs to the extent that it operates upon acts, matters or things external to Australia, although not dictated by the decision in the Seas and Submerged Lands Case, is one which received considerable support in that case. It also receives support from statements in the judgments in Robinson v. Western Australian Museum [76] , per Barwick C.J., per Mason J. [77] , and per Murphy J. [78] , Viro v. The Queen [79] , per Murphy J., Koowarta [80] , per Stephen J., and per Mason J. [81] and the Tasmanian Dam Case [82] , per Murphy J. However, and more fundamentally, it is a view which proceeds from the ordinary meaning of the words "[e]xternal affairs" in s. 51(xxix). Accordingly, in my view, s. 9 of the Act is properly to be characterized as a law with respect to external affairs to the extent and by reason that it operates upon acts, matters or things outside Australia. More particularly, it is a law on that topic to the extent of its claimed application to the information laid against the plaintiff, that information being based on acts which are said to have taken place in the Ukraine, a place geographically external to Australia. 1. (1977) 138 C.L.R. 283, at p. 294. 2. (1977) 138 C.L.R., at p. 335. 3. (1977) 138 C.L.R., at p. 343. 4. (1978) 141 C.L.R. 88, at p. 162. 5. (1982) 153 C.L.R., at p. 211. 6. (1982) 153 C.L.R., at p. 223. 7. (1983) 158 C.L.R., at pp. 171-172. The conclusion that s. 9 of the Act is a law with respect to external affairs to the extent and by reason that it operates upon acts, matters or things outside Australia makes it unnecessary to consider whether some other feature also renders it, either to some greater or lesser extent, a law on that topic. Nor is it necessary to consider whether, as argued on behalf of the defendants, it is a law with respect to defence. However, as s. 9 operates only with respect to events which occurred during the period of and in connexion with the Second World War in Europe or an occupation arising out of that war and as that conflict came to an end almost forty-five years before s. 9 came into effect, it is very difficult to see any connexion at all between it and "the defence of the Commonwealth [or] of the several States" (s. 51(vi) of the Constitution). The defence power has been described as a "purposive power", meaning that it will support a law which is reasonably capable of being seen as appropriate and adapted to the purpose of defence: see Stenhouse v. Coleman [83] . Even so, save for those considerations embodied in the phrase "pour encourager les autres", there is no basis on which s. 9 can be said to be in the slightest degree relevant to defence. And those considerations, of doubtful validity in circumstances which are proximate in time, become even more questionable as events recede into history. But cf. R. v. Kidman [84] per Powers J. In my view, those considerations will not stamp a law applying to events which occurred during a war which came to an end some forty-five years earlier as a law with respect to defence. 1. (1944) 69 C.L.R. 457, at p. 471. 2. (1915) 20 C.L.R. 425, at p. 460. The legislative powers conferred by s. 51 are conferred "subject to this Constitution". Thus, laws passed on a subject entrusted to the Parliament under s. 51 are invalid if they offend a constitutional prohibition. On behalf of the plaintiff it was argued that s. 9, in its application to the information laid against him, offends a prohibition deriving from Ch. III of the Constitution or, more particularly, from s. 71 which provides that "[t]he judicial power of the Commonwealth shall be vested in the High Court of Australia, and in such other federal courts as the Parliament creates, and in such other courts as it invests with federal jurisdiction." The latter courts are identified in s. 77(ii) as "the courts of the States". It is not in doubt that Ch. III is the source of important prohibitions which, amongst other things, operate to guarantee the independence of the federal judiciary: see, e.g., Waterside Workers' Federation of Australia v. J.W. Alexander Ltd. [85] ; Reg. v. Kirby; Ex parte Boilermakers' Society of Australia ("the Boilermakers' Case") [86] ; and Attorney-General (Cth) v. The Queen [87] . The prohibition which was asserted on behalf of the plaintiff was not identified in precise terms. Rather, Mr. Charles, senior counsel for the plaintiff, pointed to the features of the offence created by s. 9 and the features of the proceedings contemplated by the Act and argued that one or more of those features offend Ch. III. 1. (1918) 25 C.L.R. 434. 2. (1956) 94 C.L.R. 254. 3. (1957) 95 C.L.R. 529. The offence created by s. 9 of the Act is constituted by a "serious crime" (a concept which is elaborated in s. 6) committed during the period specified in s. 9(1)(a) and in one or other of the circumstances specified in s. 7. As earlier indicated and in broad terms, s. 7 and s. 9(1)(a) require the offence to have been committed during the period of and to have been connected with the Second World War in Europe or with an occupation arising out of that war. The information in this case is based on s. 6(3) of the Act which provides:— An act is a serious crime if: (a) it was done at a particular time outside Australia; and (b) the law in force at that time in some part of Australia was such that the act would, had it been done at that time in that part, be a serious crime by virtue of subsection (1). The information relies on "serious crimes" (murder, and being knowingly involved in murder) as specified in s. 6(1)(a) and (k)(iii) of the Act. Sub-section (1) of s. 6 operates by reference to several legal categories and provides as follows: An act is a serious crime if it was done in a part of Australia and was, under the law then in force in that part, an offence, being: (a) murder; (b) manslaughter; (c) causing grievous bodily harm; (d) wounding; (e) rape; (f) indecent assault; (g) abduction, or procuring, for immoral purposes; (h) an offence (in this paragraph called the "variant offence") that would be referred to in a preceding paragraph if that paragraph contained a reference to: (i) a particular intention or state of mind on the offender's part; or (ii) particular circumstances of aggravation; necessary to constitute the variant offence; (j) an offence whose elements are substantially the same as the elements of an offence referred to in any of paragraphs (a) to (h), inclusive; or (k) an offence of: (i) attempting or conspiring to commit; (ii) aiding, abetting, counselling or procuring the commission of; or (iii) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the commission of; an offence referred to in any of paragraphs (a) to (j), inclusive. Mention should also be made of s. 6(2) which provides: In determining for the purposes of subsection (1) whether or not an act was, under the law in force at a particular time in a part of Australia, an offence of a particular kind, regard shall be had to any defence under that law that could have been established in a proceeding for the offence. Assuming it was committed during the period and in the circumstances required by s. 9(1)(a) and s. 7 of the Act, the conduct which, by virtue of s. 6(1) and (3), constitutes a "war crime" is any conduct which is capable of being assigned to a legal category designated in s. 6(1) by the hypothetical application of a law of the Commonwealth or of some State or Territory as that law stood at the time of the conduct involved. Section 17(2) allows a defence to a charge of war crime as follows: Subject to section 16, it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. That defence is further elaborated by s. 17(3), (4) and (5). And s. 16 provides that, subject to any defence available pursuant to s. 6(2) or s. 13(2), it is not a defence that the person concerned "acted under orders of his or her government or of a superior". A number of matters should be noted by reference to s. 17. First, it is sufficient to state, without going to the terms of s. 17(3), (4) and (5), that I agree with Deane J., for the reasons that his Honour gives, that s. 17(2)(a) must be construed to mean that the acts in question did not infringe the rules of international law prohibiting certain conduct in the course of and in connexion with the waging of war. Secondly, and subject to s. 16(2) and any wider area of defence that may be permitted by s. 17(3), I also agree with Deane J. that s. 17(2) permits a defence, as a matter of law, that the conduct in question was neither a war crime nor a crime against humanity under the rules of international law as those rules stood at the time of the commission of the acts alleged, as well as a defence that, as a matter of fact, that conduct did not offend those rules. It is with respect to a defence of the latter kind that s. 17(4) operates so that the defence cannot be relied upon unless there is evidence of facts which raise the defence. Thirdly, I agree with Brennan J. and with Deane J. that it is not an element of the offence created by s. 9 that the conduct should amount to a war crime or a crime against humanity under the rules of international law. There is no need for an implication to that effect, and the presence of s. 17(2) makes that implication impossible: see John v. Federal Commissioner of Taxation [88] . 1. (1989) 166 C.L.R. 417, at pp. 434-435. Fourthly, the availability of a defence under s. 17(2) does not effect the consequence that only that conduct which, at the time of its commission, involved an infringement of the rules of international law creating war crimes or crimes against humanity is made punishable by the Act. I agree with Deane J., for the reasons that his Honour gives, that, at least for a large part of the period of 1 September 1939 to 8 May 1945, international law did not exclude the defence of superior orders which is excluded by s. 16 of the Act. There is thus a substantial and significant disconformity between the area in which the Act operates and the area occupied by the rules of international law by reference to which s. 17(2) is formulated. And, as Deane J. points out, that disconformity extends to the period covered by the information. Moreover, as Brennan J. points out, to the extent that a defendant might wish to rely on a factual defence as permitted by s. 17(2), he or she must be able to point to or call evidence before the defence can be raised. The first and most unusual feature of the offence created by s. 9 of the Act is that it is confined to past conduct and, it is common ground, to conduct which, at the time of its commission, was not subject to any law of this country. In particular, it was not then governed by the criminal law of this country and, thus, could not then form the basis of a criminal prosecution in this country. Another unusual feature of the offence created by s. 9 is that, at least to the extent involved in the information laid against the plaintiff, it is constituted by acts which are described in terms of particular legal categories rather than in terms assigning particular acts or omissions to some legal category. The proceedings contemplated by the Act for an offence against s. 9 also have an unusual feature in that, to the extent that they are based on s. 6(3), the assignment of conduct to a legal category involves the application of a body of law contrived for the proceedings. The first step in that process is dictated by s. 6(1) which, as earlier indicated, requires that the law should have been in force when the act was done. The next step is directed by s. 13(2) which relevantly provides that: Where a person is charged with an offence against this Act, then, for the purposes of: (b) an exercise of jurisdiction by a court [of a State or internal Territory] in relation to the offence; this Act has effect, in relation to an act that is, or is alleged to be, the offence, as if: (e) a reference in subsection 6(3) or section 18 to a part of Australia were a reference to that State or Territory; and (f) without limiting subsection 6(2), all defences under the law in force in that State or Territory when the person is charged with the offence had been defences under the law in force in that State or Territory at the time of the act. Section 18 provides for alternative verdicts. It need not be further considered. In a prosecution based on s. 6(3), the combined effect of s. 6(1) and s. 13(2) is to contrive a body of law to determine, not whether the offence of "war crime" was committed, but whether the acts constitute a "serious crime". That body of law is comprised of the law of the State or Territory in whose court the prosecution is brought as that law stood at the time of the acts alleged, but notionally amended to incorporate defences available when the charge was laid. Because it must be contrived for each prosecution, that body of law is necessarily variable. There is some rational basis for variation according to the time of the acts alleged. However, the contrived body of law may also vary according to the time when the charges are laid and according to the State or Territory in which the proceedings are brought. The initial selection of the State or Territory in which proceedings are to be brought is for the prosecutor. However, s. 14 of the Act allows for an application to be made for an order transferring the proceedings to another State or Territory. It is provided by s. 14(3) that such an application must be made "as soon as reasonably practicable" after a charge is laid or "at such later time as the magistrate or judge allows". By s. 14(4), the application must be granted unless the person charged is a resident of the State or Territory in which the proceedings were brought or is not a resident of the State or Territory to which the transfer is sought. Although s. 14 makes residence the dominant factor in determining the State or Territory in which the matter is to be prosecuted, it stops short of an absolute requirement that a prosecution be brought in the State or Territory in which the defendant resides. Apart from the convenience of the judge and, possibly, the lawyers involved, there is no obvious reason for the selection of the law of the State or Territory in whose court the proceedings are brought as the reference point for contriving a body of law to determine whether the acts charged constitute a "serious crime". And, at least where the conduct involved occurred outside Australia and, hence, outside that State or Territory, no reason is provided because, at least ordinarily, it will be where the defendant resides. The State or Territory in which he or she resides may be the result of chance or of the merest circumstance, and one might expect that its laws will have been framed without regard to the events in wartime Europe. To the extent that a body of law is contrived by reference to the law of some State or Territory other than that in which the defendant resides, that will often be the consequence of the initiative of the prosecutor, the inertia of the defendant, the discretion of the judge or a combination of one or more of those matters. Thus, to a significant extent, the body of law contrived to determine whether the acts involved in any particular case constitute a "serious crime" is not only variable, but results from circumstances which are either remote from or irrelevant to conduct which, as here and as contemplated by s. 6(3), occurred outside Australia and is made the subject of s. 9 of the Act. One other feature of the proceedings contemplated by the Act should be mentioned. Section 13(4)(b) relevantly provides that "[n]othing in Part II or subsection 9(1) shall be taken to limit the powers of a court to take action to prevent an abuse of process." However, s. 13 further provides: (5) Where, on the trial of a person for an offence against this Act, the person satisfies the judge, on the balance of probabilities, that: (a) the person is unable to obtain evidence that he or she would, but for the lapse of time or some other reason beyond his or her control, have been able to obtain; (b) the person's inability to obtain that evidence has substantially prejudiced, or will substantially prejudice, the preparation or conduct of his or her defence; and (c) the interests of justice require the making of an order under this subsection; the judge may make such order as he or she thinks appropriate for a stay of proceedings for the offence. (6) Nothing in subsections (4) and (5) limits the generality of anything else in those subsections. It was argued that s. 13(5), not being within Pt II of the Act, must be taken as an exhaustive statement of the circumstances which will justify a stay of proceedings under the Act and, hence, a limitation upon the inherent power of a Court to control its own process. The inherent power of a court to stay proceedings is a power which is closely confined: see Jago v. District Court (N.S.W.) [89] . However, the power is an essential attribute of a superior court and exists for the purpose of ensuring that proceedings serve the ends of justice and are not themselves productive of or an instrument of injustice: see Cocker v. Tempest [90] . In my view, an intention to interfere with such an important and essential power must be revealed by unmistakable language before a provision of a statute will be construed as having that effect: see Cameron v. Cole [91] , per Rich J. And, in that event, a question might arise, at least in circumstances which would call for the exercise of that power, whether its curtailment or abrogation transformed the power purportedly vested in the court into something other than judicial power and, thus, brought the provision into conflict with Ch. III. Notwithstanding that s. 13(4), (5) and (6) are curious both in form and in substance, they do not, in my view, reveal an unmistakable intention to curtail or abrogate the inherent power of a court to stay its proceedings. 1. (1989) 168 C.L.R. 23. 2. (1841) 7 M. & W. 502, at pp. 503-504 [151 E.R. 864, at p. 865]. 3. (1944) 68 C.L.R. 571, at p. 589. It was said in the Boilermakers' Case [92] that s. 71, being a complete and exhaustive statement with respect to the judicial power of the Commonwealth, requires that only judicial power and powers ancillary or incidental thereto be conferred on a court as named or indicated in s. 71. The correctness of the principal conclusion of the Boilermakers' Case was doubted by Barwick C.J. and by Mason J. in Reg. v. Joske; Ex parte Australian Building Construction Employees & Builders' Labourers' Federation [93] . However, it is not in doubt that s. 71 imposes limits as to the powers which the Parliament may confer on a court: see Hilton v. Wells [94] . See also Reg. v. Spicer; Ex parte Australian Builders' Labourers' Federation [95] ; Reg. v. Commonwealth Industrial Court; Ex parte Amalgamated Engineering Union, Australian Section [96] ; and Mikasa (N.S.W.) Pty. Ltd. v. Festival Stores [97] . 1. (1956) 94 C.L.R., at pp. 270-271, 289. 2. (1974) 130 C.L.R. 87, at pp. 90, 102. 3. (1985) 157 C.L.R. 57, at p. 68 and at pp. 81-82. 4. (1957) 100 C.L.R. 277, at pp. 290-291. 5. (1960) 103 C.L.R. 368. 6. (1972) 127 C.L.R. 617. An essential feature of judicial power is that it be exercised in accordance with the judicial process. I attempted to identify the features of that process in Harris v. Caladine [98] , and in Re Nolan; Ex parte Young [99] . To adopt the words of Kitto J. inReg. v. Trade Practices Tribunal; Ex parte Tasmanian Breweries Pty. Ltd. [1] , the essential features of that process include the determination of legal rights, obligations or consequences by the ascertainment of the facts as they are and as they bear on the matter for determination, and the identification of the applicable law, followed by an application of that law to those facts. Those features may be more or less obvious, depending on the issues involved and the nature of the law to be applied. At one extreme, the law to be applied may require the assigning of future rights and obligations attaching to or in consequence of a legal relationship in the exercise of a judicial discretion in which no particular matter is decisive. At the other extreme, the law may assign distinct legal consequences by reason that a person has committed a particular proscribed act. Criminal laws are laws of the latter kind. But whatever the issues and whatever the nature of the law being applied, the power vested in a court can be exercised only on the basis of the discovered facts and by application of the law which determines the legal consequences attaching to those facts. 1. (1991) 172 C.L.R. 84, at pp. 150-152. 2. (1991) 172 C.L.R. 460. 3. (1970) 123 C.L.R. 361, at 374. A power to be exercised by the application of law to facts invented by Parliament or invented according to some statutory formula or prescription would not be a power to be exercised in accordance with the judicial process and would not be judicial power. That is not to say that statutory fictions may not be employed in the course of and for the purpose of formulating the legal rights, obligations or consequences attaching to a relationship or to conduct. And, of course, they may be applied by the courts when those rights, obligations or consequences are in issue. However, the relationship or the conduct which is the basis of those rights, obligations or consequences must be real and not fictitious. A law assigning legal consequences on the basis of fictitious or invented facts may sometimes, on that account, be characterized as other than a law on a subject matter within legislative power: see Actors and Announcers Equity Association v. Fontana Films Pty. Ltd. [2] . Quite apart from that consideration, a law conferring power on a court to determine legal consequences on the basis that a person is who he is not or on the basis that he did what he did not would be invalid for offending Ch. III. It would be invalid because the power in question would involve a travesty of the judicial process and would, thus, be a power which, by virtue of s. 71, could not validly be conferred on a court. 1. (1982) 150 C.L.R. 169. Equally, it would be a travesty of the judicial process if, in proceedings to determine whether a person had committed an act proscribed by and punishable by law, the law proscribing and providing for punishment of that act were a law invented to fit the facts after they had become known. In that situation, the proceedings would not be directed to ascertaining guilt or innocence (which is the function of criminal proceedings and the exclusive function of the courts), but to ascertaining whether the Parliament had perfected its intention of declaring the act in question an act against the criminal law. That is what is involved if a criminal law is allowed to take effect from some time prior to its enactment. Of course, the position is different if the law re-enacts an earlier law which applied when the acts were committed. At least that is so to the extent that that earlier law has not been brought to bear on conduct falling or alleged to fall within it. In this regard, it is sufficient to state that, in my view, a law would not be a law reenacting an earlier law if it purported to apply cumulatively upon it. And the position is different again in the case of a law which acts retrospectively upon civil rights, obligations or liabilities. The function of a court in civil proceedings is the determination of present rights, obligations or liabilities. In that context, a retrospective civil law is very much like a statutory fiction in that it is a convenient way of formulating laws which, by their application to the facts in issue, determine the nature and extent of those present rights, obligations or liabilities. It was argued on behalf of the defendants that it was decided in Kidman [3] that it is within the power of the Parliament to pass a criminal law taking effect from some time prior to its enactment. Kidman concerned a provision creating the statutory offence of conspiracy to defraud the Commonwealth. 1. (1915) 20 C.L.R. 425. It is now obvious, if it was not when Kidman was decided, that conspiracy to defraud the Commonwealth constituted on offence at common law. Thus, the provision considered in Kidman merely reenacted a law which applied at the earlier time. It was that feature which, in the judgment of Griffith C.J. [4] , served to signal that the provision was within legislative power. And Isaacs J. [5] , expressed a similar consideration, namely, that the provision was not "ex post facto converting a lawful act into an unlawful act". However, the other Justices in that case proceeded on the basis or assumed without deciding (see Higgins J. [6] ) that the provision created a new offence constituted, so far as it operated during a period prior to its enactment, by acts which were not then proscribed by law. 1. (1915) 20 C.L.R., at p. 436. 2. (1915) 20 C.L.R., at p. 442. 3. (1915) 20 C.L.R., at p. 448. As Deane J. points out in this case, the significance of Ch. III was neither raised nor considered in Kidman . Accordingly, in my view, Kidman is authority only for what it actually decided, namely, that it is within the legislative power of the Parliament to create a statutory offence taking effect prior to its enactment if it merely gives statutory form to an earlier common law offence. However, if Kidman is authority for some wider proposition, then in my view, for the reasons given by Deane J., it should be reexamined. In particular, it does not rest upon a principle that has been carefully worked out in a succession of cases and, it is not a decision which, outside the present case, appears to have been acted upon: see The Commonwealth v. Hospital Contribution Fund [7] . 1. (1982) 150 C.L.R. 49, at pp. 56-57. The true nature of what is commonly called an "ex post facto law" or a "retroactive law" is revealed in the judgment of Powers J. in Kidman [8] where his Honour described a law of that kind as "a law by which, after an act has been committed which was not punishable at the time it was committed, the person who committed it is declared to have been guilty of a crime and to be held liable to punishment". As is there made clear, it is the statute or the Act of the Parliament, and not the determination of a court, by which a person is declared to have been guilty. That is the usurpation (see Liyanage v. The Queen [9] ) of power which is exclusively judicial. See Huddart, Parker & Co. Pty. Ltd. v. Moorehead [10] , per Griffith C.J., and per Isaacs J. [11] ; Alexander [12] , per Griffith C.J., and per Isaacs and Rich JJ. [13] ; Trade Practices Tribunal [14] , per Kitto J.; and Re Tracey; Ex parte Ryan [15] , per Deane J. 1. (1915) 20 C.L.R., at p. 457. 2. [1967] 1 A.C. 259, at p. 289. 3. (1909) 8 C.L.R. 330, at p. 357. 4. (1909) 8 C.L.R., at p. 383. 5. (1918) 25 C.L.R., at p. 442. 6. (1918) 25 C.L.R., at pp. 463, 465. 7. (1970) 123 C.L.R., at p. 374. 8. (1989) 166 C.L.R. 518, at p. 580. The usurpation of judicial power by a law which declares a person guilty of an offence produces the consequence that the application of that law by a court would involve it in an exercise repugnant to the judicial process. It is repugnant to the judicial process because the determination of guilt or innocence is foreclosed by the law. The only issue is whether the person concerned was a person declared guilty by the law. And all that involves is the determination, as a matter of fact, whether some person is the person, or answers the description (whatever form it takes) of the persons, declared guilty by the Act. It does not involve, and indeed negates, that which is the essence of judicial power in a criminal proceeding, namely, the determination of guilt or innocence by the application of the law to the facts as found. Accordingly, such a law is invalid as infringing s. 71 because it involves the exercise by Parliament of a power which can be exercised only by the courts named or indicated in s. 71 and because its application by a court would involve it in exercising a power repugnant to the judicial process. If Kidman holds otherwise, it should no longer be followed. It remains to be considered whether, in its application to the information laid against the plaintiff, s. 9 is invalid by reason that it offends Ch. III. I have already indicated that, in my view, a law which merely re-enacts a law which applied to acts at the time of their commission, provided that that earlier law has not already been brought to bear on those acts, would not come into conflict with Ch. III. And I see no reason why, assuming it to be a law with respect to a subject matter entrusted to the Parliament of the Commonwealth, the Parliament might not re-enact the common law, the law of a State or of a Territory or reproduce the law of some foreign country or a rule of international law. However, validity would depend on the faithful re-enactment or reproduction of that law and, should an issue arise, it would be for the Commonwealth to establish — by evidence if necessary — that its law faithfully re-enacted or reproduced the earlier law: see Australian Communist Party v. The Commonwealth [16] , per Dixon J., per Williams J. [17] , per Webb J. [18] , and per Fullagar J. [19] . 1. (1951) 83 C.L.R. 1, at pp. 200-202. 2. (1951) 83 C.L.R., at pp. 222-223. 3. (1951) 83 C.L.R., at p. 244. 4. (1951) 83 C.L.R., at pp. 261-262. It was not suggested that the Act is a re-enactment or a reproduction of an earlier law, save to the extent that the defendants pointed to the rules of international law with respect to war crimes and crimes against humanity. As already indicated, the Act is not confined to the area of operation of those rules as they stood from time to time in the period 1 September 1939 to 8 May 1945 and by reference to which the defence allowed by s. 17(2) is formulated. Accordingly, it is not a faithful reproduction of those rules. Moreover, despite the complexity of ss. 6(3) and 13 and the way in which they are formulated, the proceedings contemplated by the Act do not involve the application of law to facts as found to determine their legal consequence. Rather, the complexity and the contrived nature of those provisions indicate that what is involved is the application of a formula which, in essence, directs an inquiry whether the person charged did acts outside Australia which, if done within Australia, would have amounted to an offence in Australia if the law had then been the same as that contrived for the proceedings. That task is not altered by reason that the Act allows a "defence" which, as already indicated, differs from the rules of international law by reference to which it is formulated. The "defence" is merely part of the formula providing a description of the persons whom the Act declares to be guilty. And, thus, the task involved in the proceedings contemplated by the Act is no more than that of ascertaining, as a matter of fact, whether a person charged under the Act fits that description. Were that task to be divorced from the consequence which ss. 9 and 10 seek to attach, it might properly be conferred on a Royal Commission of Inquiry or some similar body possessed of investigative powers. However, whether divorced from or attached to that consequence, it cannot be conferred on a court named or indicated in s. 71 of the Constitution for the simple reason that the process involved is one that denies an essential feature of the judicial process, namely, the application of law to facts to determine their legal consequence. It makes no difference to the validity of s. 9, in its application to the information laid against the plaintiff, that jurisdiction is also conferred on the courts of the internal Territories. Although the courts of the Territories are in a somewhat different position from the courts named and indicated in s. 71 (see Capital T.V. and Appliances Pty. Ltd. v. Falconer [20] and Hilton v. Wells [21] ), s. 9 is, in any event, invalid by reason that it involves the usurpation of a power which can be exercised only by a court in accordance with the judicial process. 1. (1971) 125 C.L.R. 591. 2. (1985) 157 C.L.R., at pp. 67-68. The question referred pursuant to s. 18 of the Judiciary Act should be answered "Yes". McHugh J. The question reserved for consideration of the Full Court in this case is whether s. 9 of the War Crimes Act 1945 Cth as amended ("the Act") is invalid in its application to the information laid by the second defendant against the plaintiff. In my opinion that question should be answered in the negative. The facts and contentions of the parties are set out in other judgments. Except for the purpose of explaining my reasons, I shall not repeat them. Upon its proper construction, the Act applies to conduct which constitutes a war crime (within the meaning of the Act) whether that conduct occurred inside or outside Australia. The terms of s. 6(1) of the Act when read with s. 7(1)(c) or (d) or s. 7(3) and the terms of s. 6(3) when read with s. 7 make it clear that s. 9(1) applies to a person who committed an act constituting a war crime whether that act occurred inside or outside Australia. The most relevant provisions of the Act are as follows: 6.(1) An act is a serious crime if it was done in a part of Australia and was, under the law then in force in that part, an offence, being: (a) murder; (b) manslaughter; (c) causing grievous bodily harm; (d) wounding; (e) rape; (f) indecent assault; (g) abduction, or procuring, for immoral purposes; (h) an offence (in this paragraph called the "variant offence") that would be referred to in a preceding paragraph if that paragraph contained a reference to: (i) a particular intention or state of mind on the offender's part; or (ii) particular circumstances of aggravation; necessary to constitute the variant offence; (j) an offence whose elements are substantially the same as the elements of an offence referred to in any of paragraphs (a) to (h), inclusive; or (k) an offence of: (i) attempting or conspiring to commit; (ii) aiding, abetting, counselling or procuring the commission of; or (iii) being, by act or omission, in any way, directly or indirectly, knowingly concerned in, or party to, the commission of; an offence referred to in any of paragraphs (a) to (j), inclusive. (2) In determining for the purposes of subsection (1) whether or not an act was, under the law in force at a particular time in a part of Australia, an offence of a particular kind, regard shall be had to any defence under that law that could have been established in a proceeding for the offence. (3) An act is a serious crime if: (a) it was done at a particular time outside Australia; and (b) the law in force at that time in some part of Australia was such that the act would, had it been done at that time in that part, be a serious crime by virtue of subsection (1). Section 7 provides: (1) A serious crime is a war crime if it was committed: (a) in the course of hostilities in a war; (b) in the course of an occupation; (c) in pursuing a policy associated with the conduct of a war or with an occupation; or (d) on behalf of, or in the interests of, a power conducting a war or engaged in an occupation. (2) For the purposes of subsection (1), a serious crime was not committed: (a) in the course of hostilities in a war; or (b) in the course of an occupation; merely because the serious crime had with the hostilities or occupation a connection (whether in time, in time and place, or otherwise) that was only incidental or remote. (3) A serious crime is a war crime if it was: (a) committed: (i) in the course of political, racial or religious persecution; or (ii) with intent to destroy in whole or in part a national, ethnic, racial or religious group, as such; and (b) committed in the territory of a country when the country was involved in a war or when territory of the country was subject to an occupation. (4) Two or more serious crimes together constitute a war crime if: (a) they are of the same or a similar character; (b) they form, or are part of, a single transaction or event; and (c) each of them is also a war crime by virtue of either or both of subsections (1) and (3). Section 9(1) provides: A person who: (a) on or after 1 September 1939 and on or before 8 May 1945; and (b) whether as an individual or as a member of an organisation; committed a war crime is guilty of an indictable offence against this Act. Section 16 provides: Subject to subsections 6(2) and 13(2), the fact that, in doing an act alleged to be an offence against this Act, a person acted under orders of his or her government or of a superior is not a defence in a proceeding for the offence, but may, if the person is convicted of the offence, be taken into account in determining the proper sentence. Section 17 provides: (1) This section has effect for the purposes of a proceeding for an offence against this Act. (2) Subject to section 16, it is a defence if the doing by the defendant of the act alleged to be the offence: (a) was permitted by the laws, customs and usages of war; and (b) was not under international law a crime against humanity. (3) To avoid doubt, the doing of the act by the defendant was permitted by the laws, customs and usages of war if it was reasonably justified by the exigencies and necessities of the conduct of war. (4) The defendant is not entitled to rely on a defence under subsection (2) unless there is evidence of the existence of the facts constituting the defence. (5) However, if there is such evidence, the onus of establishing, beyond a reasonable doubt, that those facts either do not exist or do not constitute the defence lies on the prosecution. The Commonwealth contended that s. 6(1) was definitional and that the Act operated only in respect of conduct which occurred in Europe during the period 1 September 1939 to 8 May 1945. It is true that the first paragraph of the preamble, the definitions of "war" and "occupation", and the terms of ss. 16 and 17 of the Act suggest that the Act was aimed only at those persons who committed war crimes in Europe during that period. But the terms of s. 6(1) are too emphatic to be read down by these considerations. The words "[a]n act if it was done in a part of Australia" in that sub-section cannot be read as if they meant "an act done outside Australia". Moreover, the contrast between s. 6(1), which makes an act of a certain kind a serious crime "if it was done in a part of Australia", and s. 6(3), which makes an act of a certain kind a serious crime "if it was done at a particular time outside Australia", is explicable only on the ground that the Parliament of the Commonwealth intended that an act committed in Australia can constitute a war crime for the purposes of the Act. By s. 11, however, only a person who is at the present time an Australian citizen or resident of Australia or an external Territory can be charged with an offence against the Act. The Commonwealth also submitted that s. 7 contained an implication that the conduct in question had to constitute a war crime or a crime against humanity under international law before it amounted to a war crime for the purposes of s. 7. But nothing in the Act gives any support for that proposition. Indeed, the inference to be drawn from s. 17 is to the contrary. That section makes it a defence "if the doing by the defendant of the act alleged to be the offence was permitted by the laws, customs and usages of war; and was not under international law a crime against humanity". The word "permitted" in this context is erroneous. As Baxter has pointed out, in "The Municipal and International Law Basis of Jurisdiction over War Crimes", The British Year Book of International Law, vol. XXVIII (1951) 382, at p. 388: The international law of war is "prohibitive law" and its purpose is to place curbs upon the otherwise unrestrained violence of war. Belligerent acts in war are facts, not legal rights To make sense of the section, therefore, the words "permitted by" should be read as meaning "not in breach of". When the section is so read, it provides an answer to a charge under s. 9(1) where the act of the defendant was not in breach of the laws, customs and usages of war and was not under international law a crime against humanity. Nevertheless, s. 17(4) provides that the "defence" is not open "unless there is evidence of the existence of the facts constituting the defence". Unless there is such evidence, the prosecution is not required to prove that those facts do not exist or do not constitute such a "defence". It follows that a conviction under s. 9(1) may ensue without proof that the act of the defendant was in breach of the laws, customs and usages of war (that is to say, was a "war crime" for the purposes of international law) or was under international law a crime against humanity. That being so, it is not a necessary element of a charge under s. 9(1) that the act of the defendant was in breach of the laws, customs and usages of war or was under international law a crime against humanity. Neither s. 7 nor any other provision of the Act can be read as containing any implication to the contrary. The constitutional validity of the Act, therefore, depends upon two questions: (a) Did the Parliament have power in 1988 to punish an Australian resident or citizen for conduct falling within ss. 6(3) and 7 of the Act when that conduct occurred outside Australia more than forty years previously? (b) Did the Parliament have power in 1988 to punish an Australian resident or citizen for conduct falling within ss. 6(1) and 7 of the Act when that conduct occurred in Australia more than forty years previously? The external affairs power In my opinion, in so far as the Act penalizes conduct constituting a war crime which occurred outside Australia, the Act is validly enacted pursuant to the external affairs power. According to a substantial body of dicta, but no express decision, in this Court, s. 51(xxix) of the Constitution authorizes a law for the peace, order and good government of the Commonwealth with respect to any matter, thing or conduct occurring outside Australia: New South Wales v. The Commonwealth ("the Seas and Submerged Lands Case ") [22] ; Robinson v. Western Australian Museum [23] ;Viro v. The Queen [24] ; Koowarta v. Bjelke-Petersen [25] . In the Seas and Submerged Lands Case [26] , Jacobs J. said that, under the external affairs power, "the Commonwealth has the power to make laws in respect of any person or place outside and any matter or thing done or to be done or prohibited to be done outside the boundaries of the Commonwealth". 1. (1975) 135 C.L.R. 337, at pp. 360, 471, 497, and cf. pp. 503-504. 2. (1977) 138 C.L.R. 283, at p. 294, and cf. pp. 335, 343. 3. (1978) 141 C.L.R. 88, at p. 162. 4. (1982) 153 C.L.R. 168, at pp. 222-223. 5. (1975) 135 C.L.R., at p. 497. I can see no reason for giving the power conferred by s. 51(xxix) any narrower interpretation than that expounded by Jacobs J. Subject to any express or implied limitations in the Constitution, a grant of power conferred by s. 51 is to "be construed with all the generality which the words used admit": Reg. v. Public Vehicles Licensing Appeal Tribunal (Tas.); Ex parte Australian National Airways Pty. Ltd. [27] . The ordinary meaning of the word "external" is "outside". The meaning of the word "affairs" varies considerably according to its context. But it is a word which is capable of a very wide meaning. The Macquarie Dictionary, 2nd rev. ed. (1987), includes among the meanings of "affair": 1. anything done or to be done; that which requires action or effort; business; concern: 2. (pl.) matters of interest or concern; particular doings or interests: 3. an event or a performance; a particular action, operation, or proceeding: 4. thing; matter (applied to anything made or existing, with a descriptive or qualifying term): 5. a private or personal concern; a special function, business or duty. 1. (1964) 113 C.L.R. 207, at p. 225. Moreover, since the adoption of s. 3 of the Statute of Westminster Act 1931 Imp. by the Statute of Westminster Adoption Act 1942 Cth, no question can arise as to the extraterritorial operation of Commonwealth legislation. But, independently of the effect of that legislation, once a subject matter is found to be within the terms of s. 51, the Commonwealth has, and always has had, power to give extraterritorial operation to a law on that subject matter: see Croft v. Dunphy [28] . The contrary view was based on a misconception: Reg. v. Foster; Ex parte Eastern and Australian Steamship Co. Ltd. [29] . In so far as the nature of any subject matter specified in s. 51 is capable of being given an extraterritorial operation, therefore, the Commonwealth has the constitutional power to give it that operation. If the Parliament of the Commonwealth in exercising its powers under s. 51 "were to legislate gratuitously in respect of foreign persons in foreign territory, an Australian court could not hold the legislation invalid": Reg. v. Foster [30] . 1. [1933] A.C. 156, at p. 163. 2. (1959) 103 C.L.R. 256, at p. 300. 3. (1959) 103 C.L.R., at p. 306. Furthermore, the words "peace, order, and good government of the Commonwealth" do not require that a law with respect to "external affairs" should be a law whose subject matter has some recognizable connexion with Australia. Those words do not impose any limitation on the power conferred on the Parliament of the Commonwealth by s. 51 of the Constitution. They do not require legislation to be held invalid because, in the opinion of the Court, the legislation does not promote or is not in fact for the peace, order and good government of the Commonwealth: Union Steamship Co. of Australia Pty. Ltd. v. King [31] . The words "peace, order, and good government" are a recognition of the fact that, for the purposes of constitutional theory, "the purpose and design of every law is to promote the welfare of the community": Harrison Moore, Constitution of the Commonwealth of Australia, 2nd ed. (1910), pp. 274-275, cited in Reg. v. Foster [32] , per Windeyer J. 1. (1988) 166 C.L.R. 1, at p. 10. 2. (1959) 103 C.L.R., at p. 308. Accordingly, the term "external affairs" should be interpreted to include any matter, thing, event or relationship existing or arising or which might exist or arise outside Australia. Section 51(xxix) is not confined, therefore, to the making of laws authorizing arrangements with other nations or implementing arrangements properly entered with other nations: see the Seas and Submerged Lands Case [33] . Nor is it confined to affairs which concern Australia's relations with other countries or affect Australia's standing in the community of nations or which have some recognizable connexion with Australia. A law which punishes an Australian resident or citizen in respect of conduct occurring outside Australia is a law for the peace, order and good government of the Commonwealth with respect to "external affairs". Thus, in so far as the Act operates in respect of "war crimes" committed by Australian residents or citizens outside Australia, it would have been validly enacted under the external affairs power if it had been enacted on 1 September 1939. On that hypothesis, the Act would have operated prospectively to punish Australian citizens and residents in respect of conduct occurring outside Australia with respect to a war occurring in Europe. The critical question, however, is whether the retrospective operation of the Act means that it is not a law with respect to external affairs even though the Act punishes conduct which has occurred outside Australia. 1. (1975) 135 C.L.R., at p. 360. In R. v. Kidman [34] , this Court unanimously held that the Parliament of the Commonwealth had the power to deem a law, which made conspiracy to defraud the Commonwealth an offence, to have been in force from a date prior to the enactment of that law. A majority of the Court were of the opinion that the retrospective operation of the law was authorized by s. 51(xxxix) of the Constitution, which provides that the Parliament has power to make laws with respect to: Matters incidental to the execution of any power vested by this Constitution in the Parliament or in either House thereof, or in the Government of the Commonwealth, or in the Federal Judicature, or in any department or officer of the Commonwealth. In Kidman [35] Higgins J. said: It is clear that pl. xxxix. of s. 51 was not meant to limit, it was meant to increase, the powers of Parliament to make laws; and there is not one word, from first to last, to indicate an intention to withhold from the Federal Parliament the same absolute discretion as the British Parliament itself has, with regard to past events as well as present and future — provided that the Federal Parliament confine itself to the specified subjects and matters incidental to the execution of the legislative executive and judicial powers. 1. (1915) 20 C.L.R. 425. 2. (1915) 20 C.L.R., at p. 453. The present Act punishes conduct which occurred outside Australia more than forty years before the enactment of the Act. For the reasons which I have given, however, s. 51(xxix) of the Constitution vests in the Parliament the power to make a law with respect to such conduct. Upon the assumption that Kidman was correctly decided — a question with which I have to deal in determining whether the Act offends the provisions of Ch. III of the Constitution — the retrospective operation of the Act was authorized by s. 51(xxxix) of the Constitution since that operation was a matter "incidental to the execution of [a] power vested by this Constitution in the Parliament". Accordingly, in my opinion, in so far as the Act punishes conduct occurring outside Australia at a time prior to the commencement of the Act, the Act is authorised by s. 51(xxix) and (xxxix) of the Constitution. In my opinion, the Act is also valid in so far as it punishes an Australian resident or citizen for an act "done in a part of Australia" which is a "serious crime" within the meaning of s. 6(1) of the Act. Before a person can be punished under s. 9(1) for such an act, the "serious crime" must be a "war crime" within the meaning of s. 7 of the Act: see s. 8(2), (3) and (4) and s. 9(1). The terms of s. 7 have already been set out. Section 5 of the Act defines "occupation" to mean: (a) an occupation of territory arising out of a war; or (b) without limiting the generality of paragraph (a), an occupation of territory in Latvia, Lithuania or Estonia as a direct or indirect result of: (i) the agreement of 23 August 1939 between Germany and the Union of Soviet Socialist Republics; or (ii) any protocol to that agreement. That section also defines "war" to mean: (a) a war, whether declared or not; (b) any other armed conflict between countries; or (c) a civil war or similar armed conflict; (whether or not involving Australia or a country allied or associated with Australia) in so far as it occurred in Europe in the period beginning on 1 September 1939 and ending on 8 May 1945. Because of the definitions of "war" and "occupation", it is not possible for a person who has committed an act which is a serious crime within the meaning of s. 6(1) to have committed a war crime within the meaning of s. 7(1)(a) or (b). But a person who has committed an act which is a serious crime within the meaning of s. 6(1) will have committed a "war crime" within the meaning of s. 7(1)(c) or (d) if his or her act was done in pursuing a policy associated with the conduct of a war or occupation (as defined) or on behalf of or in the interests of a power conducting such a war or engaged in such an occupation. Moreover, since Australia was involved "in a war", it is theoretically possible for a person who committed an act in Australia to have done so in circumstances which make it a "war crime" by reason of s. 7(3). It would be surprising if there is any person who has committed a war crime within the meaning of the Act by reason of an act done in Australia. However, the constitutional validity of s. 9(1), in so far as it depends upon the combined operation of ss. 6(1) and 7, has to be determined by reference to the terms of those provisions and not by reference to the likelihood of their operation in respect of acts committed in Australia. The question, therefore, is whether a law which punishes an act committed in Australia in circumstances which make it a "war crime" by reason of s. 7 is a valid enactment under the external affairs power. In my opinion, the external affairs power extends to conduct engaged in in Australia for the purpose of carrying out some object external to Australia. Thus, a law which prohibits persons from doing any act or thing in Australia with intent to overthrow by force or violence a foreign government is a law with respect to external affairs: cf. s. 24AA of the Crimes Act 1914 Cth. Similarly, the external affairs power would support a law which prohibits persons from preparing in Australia for an incursion into a foreign State for the purpose of engaging in hostile activities: cf. s. 7 of the Crimes (Foreign Incursions and Recruitment) Act 1978 Cth. Although the acts which ss. 6(1), 7 and 9(1) of the Act penalize were acts occurring within Australia, they were acts which were committed in the course of or for the purpose of events occurring outside Australia. Consequently, ss. 6(1), 7 and 9(1), which punish such acts, were validly enacted under the external affairs power. For the reasons I have already given, the retrospective operation of these sub-sections does not take them outside the scope of the external affairs power. Chapter III Counsel for the plaintiff contended that, by reason of Ch. III of the Constitution, the Parliament of the Commonwealth cannot "pass legislation requiring Federal Courts to act in a manner which is inconsistent with basic requirements of justice". In support of his contention that the Act infringed Ch. III, he relied on its retroactive operation, the burden of proof of an accused under s. 17(4), the nature of the proof required from an accused under s. 17(2) and the effect of s. 13(5) and (6) of the Act on the common law doctrine of abuse of process. Only the retroactive nature of the Act, however, raises any arguable question that the Act is invalid because of the provisions of Ch. III of the Constitution. Unless Kidman is overruled — and counsel for the plaintiff did not suggest that it should be — the decision in that case seems to me to be a complete answer to the suggestion that the retrospective operation of a law of the Parliament constitutes an infringement of Ch. III of the Constitution. Kidman upheld the constitutional validity of a statute which retrospectively made it a criminal offence to conspire to defraud the Commonwealth. It is a direct authority on the point in issue in this case. It is true that the legislation in Kidman made it an offence to do what was already a common law offence. But, so far as the present case is concerned, that distinction is one without a difference. Like the present case, Kidman was concerned with the retrospective operation of a law of the Parliament which created a criminal offence. The correctness of the decision in Kidman has not hitherto been doubted in any judgment of this Court. To the contrary, it has frequently been cited as an authority: see, e.g., R. v. Snow [36] ; Ex parte Walsh and Johnson; In re Yates [37] ; Australian Communist Party v. The Commonwealth [38] ; University of Wollongong v. Metwally [39] . In Ex parte Walsh and Johnson [40] , Isaacs J. said: Whatever the Parliament enacts with respect to any of the named subject matters for any period or point of time subsequent to the establishment of the Commonwealth is, subject to any prohibition or qualification in the Constitution itself or in any controlling Imperial law, binding throughout the Commonwealth and on certain British ships beyond the Commonwealth (ss. 51 and 52 and covering s.V). Such laws may be made prospectively or retrospectively, the Commonwealth Parliament having in this respect the power of the Imperial Parliament. In Nelungaloo Pty. Ltd. v. The Commonwealth [41] Williams J. pointed out: It is trite law that the powers conferred upon the Commonwealth Parliament by s. 51 of the Constitution are plenary powers of legislation as large and of the same nature as those of the Imperial Parliament itself. No one doubts the power of the Parliament of the United Kingdom to enact retrospective criminal laws. English law has rejected Hobbes' statement that: "No law, made after a fact done, can make it a crime For before the law, there is no transgression of the law." (Leviathan (1651), Chs 27-28, cited by Williams in Criminal Law: The General Part, 2nd ed. (1961), p. 580.) Even Blackstone, who was strongly critical of ex post facto laws, did not deny the power of Parliament to pass such laws: see Commentaries on the Laws of England (1765), p. 46. Throughout the centuries, the Imperial Parliament has passed numerous laws, civil and criminal, having a retrospective operation. 1. (1917) 23 C.L.R. 256, at p. 265. 2. (1925) 37 C.L.R. 36, at pp. 86, 124-125. 3. (1951) 83 C.L.R. 1, at p. 172. 4. (1984) 158 C.L.R. 447, at p. 484. 5. (1925) 37 C.L.R., at p. 81. 6. (1948) 75 C.L.R. 495, at pp. 503-504. Moreover, numerous Commonwealth statutes, most of them civil statutes, have been enacted on the assumption that the Parliament of the Commonwealth has power to pass laws having a retrospective operation. Since Kidman , the validity of their retrospective operation has not been challenged. And I can see no distinction between the retrospective operation of a civil enactment and a criminal enactment: see Kidman [42] , per Isaacs J. 1. (1915) 20 C.L.R., at pp. 442-443. Notwithstanding this course of authority and practice, however, two members of the Court are of the opinion that the Parliament of the Commonwealth has no power to pass criminal laws which have a retrospective operation. They hold that such laws are an invalid exercise of judicial power. They point out that the effect of Ch. III was not the issue in Kidman . In those circumstances, it seems proper to re-examine the question of the validity of retrospective criminal laws notwithstanding the course of authority in this Court, the accepted constitutional doctrine and the practice of successive Parliaments during the last seventy-five years in enacting such laws. In my opinion, the enactment of laws having a retrospective operation does not infringe the constitutional guarantee that the judicial power of the Commonwealth can be exercised only by courts established and judges appointed in accordance with Ch. III of the Constitution, and by such other courts as are invested with federal jurisdiction. The Constitution of the United States of America expressly prohibits the passing of Bills of Attainder and ex post facto laws: Art. I, s. 9, cl. 3 and Art. I, s. 10, cl. 1. At common law, Bills of Attainder were special Acts under which the legislature inflicted capital punishments upon persons alleged to be guilty of treason and other felonies "without any conviction in the ordinary course of judicial proceedings": Story, Commentaries on the Constitution of the United States, 5th ed. (1891), vol. II, p. 216. The sentence of death also meant that the person concerned was "attainted" which had the consequence, inter alia, that his or her goods and chattels and land and tenements held in fee or in tail for term of life were forfeited to the Crown: Lord Coke's First Institute (1818), vol. 3, pp. 567-568. Other special Acts inflicting lesser punishment than death were technically known as Bills of Pains and Penalties. In the sixteenth and seventeenth centuries Bills of Attainder and Bills of Pains and Penalties were passed on numerous occasions by the English Parliament, particularly "in times of rebellion, or of gross subserviency to the Crown, or of violent political excitements": Story, p. 217. During the American Revolution, a number of such Bills were passed by the thirteen colonies: see United States v. Brown [43] . 1. (1965) 381 U.S. 437, at p. 442. The Supreme Court of the United States has given the term "Bill of Attainder" in the U.S. Constitution a wide meaning. It extends to all "legislative acts, no matter what their form, that apply either to named individuals or to easily ascertainable members of a group in such a way as to inflict punishment on them without a judicial trial": United States v. Lovett [44] . On the other hand, the Supreme Court has construed the term "ex post facto Law" in the U.S. Constitution narrowly so that it applies only to criminal and penal laws. In Calder v. Bull [45] Chase J. said: Every ex post facto law must necessarily be retrospective; but every retrospective law is not an ex post facto law: the former only are prohibited. Every law that takes away or impairs rights vested, agreeable to existing laws, is retrospective, and is generally unjust, and may be oppressive; and it is a good general rule, that a law should have no retrospect: but there are cases in which laws may justly, and for the benefit of the community, and also of individuals, relate to a time antecedent to their commencement; as statutes of oblivion or of pardon. They are certainly retrospective, and literally both concerning and after the facts committed. But I do not consider any law ex post facto, within the prohibition, that mollifies the rigor of the criminal law: but only those that create or aggravate the crime; or increase the punishment, or change the rules of evidence, for the purpose of conviction. In Fletcher v. Peck [46] Marshall C.J. said: An ex post facto law is one which renders an act punishable in a manner in which it was not punishable when it was committed. Such a law may inflict penalties on the person, or may inflict pecuniary penalties which swell the public treasury. In Johannessen v. United States [47] the Supreme Court said: It is, however, settled that this prohibition is confined to laws respecting criminal punishments, and has no relation to retrospective legislation of any other description. 1. (1946) 328 U.S. 303, at p. 315. 2. (1798) 3 U.S. 386, at p. 391. 3. (1810) 10 U.S. 87, at p. 138. 4. (1912) 225 U.S. 227, at p. 242. The framers of our Constitution were much influenced by the model of the U.S. Constitution. They "felt the full fascination of its plan": Dixon, Jesting Pilate (1965), p. 113. Yet, although Chs I, II and III reflect Arts I, II and III of the United States model, our Constitution does not prohibit Bills of Attainder or ex post facto laws. The omission must have been deliberate. It is a powerful indication that the Parliament was intended to have the power to enact ex post facto laws. Furthermore, I have not seen anything in the historical materials which would indicate that the framers of the Commonwealth Constitution believed or assumed that giving a criminal statute a retrospective operation was an exercise of, or an interference with the exercise of, judicial power. Inglis Clark later wrote that "any exposition of the purport of the language of an existing law, or any declaration of the existence of any rights or liabilities as the result of its enactment, is not an exercise of legislative power it is an attempted encroachment on the province of the Judiciary and is therefore invalid" (my emphasis): Studies in Australian Constitutional Law (1901), p. 39. But he accepted that "[t]he Constitution does not prohibit the Parliament of the Commonwealth from making retroactive laws": ibid., pp. 39-40. A law which creates a criminal offence but operates retrospectively is not the same as a Bill of Attainder or a Bill of Pains and Penalties. Such Bills are an interference with the exercise of judicial power. Bills of Attainder and Bills of Pains and Penalties constitute a legislative punishment "of specifically designated persons or groups": United States v. Brown [48] . Such Bills are "legislative judgments; and an exercise of judicial power": Calder v. Bull [49] . I think that the enactment of a Bill of Attainder or a Bill of Pains and Penalties would infringe the provisions of Ch. III of the Constitution. But the fact that the Constitution impliedly forbids the enactment of Bills of Attainder and Bills of Pains and Penalties does not mean that the Constitution prohibits the making of criminal laws having a retrospective operation. Retrospectivity is not itself sufficient to offend Ch. III of the Constitution. I cannot accept the argument that the determination of guilt or innocence is foreclosed by a criminal law which has a retrospective operation. Under such a law, it is still the jury, and not the legislature, which determines what the facts of the case are and which applies the law, as determined by the judge, to those facts for the purpose of determining whether the accused is guilty or innocent of the charge against him or her. Such a law is not an exercise of, or an interference with the exercise of, judicial power. 1. (1965) 381 U.S., at p. 447. 2. (1798) 3 U.S., at p. 389. Kidman was correctly decided. The Act in question in this case is not a Bill of Attainder or a Bill of Pains and Penalties. It differs from an ordinary criminal statute only in the fact that it operates retrospectively and not prospectively. It does not select a specifically designated person or group and impose a punishment on that person or group. It does not make any determination of fact. It does not adjudge any person or group to be guilty of any offence. There is not a scintilla of difference between the roles of the judge and jury in a trial under this Act and the roles of the judge and jury in a trial under a hypothetical law, in substantially identical terms to this Act, passed on 1 September 1939 and operating prospectively. The only difference between the present Act and that hypothetical law would be that the present Act makes it a legislative offence to do what was not a legislative offence at the time when it was done. That is to say, the difference is that the present Act retrospectively, and not prospectively, imposes penal sanctions on proscribed conduct. The imposition of penal sanctions on proscribed conduct, however, is an exercise of legislative, not judicial, power. Accordingly, the present Act does not interfere in any way with the judicial process or with the judicial power of the Commonwealth. Conclusion The Act is a valid enactment under s. 51(xxix) and (xxxix) of the Constitution. It does not infringe the provisions of Ch. III. I would answer the question asked in the negative.