It is requisite that [a] resolution be notified to the people who are to obey it. … [I]t is incumbent on the promulgators to do it in the most public and perspicuous manner; not like Caligula, who (according to Dio Cassius) wrote his laws in a very small character, and hung them up upon high pillars, the more effectually to ensnare the people.

Sir William Blackstone, Commentaries on the Laws of England

In a recent post on this blog, Julian R Murphy examined early signs of Edelman J’s approach to the ‘original public meaning’ of the Constitution. While that post focussed mainly upon the first element of that formula, the aim of this post is to explore an issue raised in Re Canavan [2017] HCA 45 by Edelman J in argument and reserved for future consideration in the judgment concerning the second element – that the meaning of laws be public.

The Colonial Office and its secret memoranda

In June 1897, Joseph Chamberlain’s Colonial Office examined the most recent iteration of the draft Constitution, prepared at Adelaide. Anxious not to be seen to be interfering in the Federation process, the Colonial Office considered that its views on the Adelaide Bill would be best communicated through a conduit. New South Wales Premier George Reid agreed to perform that role.

In July 1897, Chamberlain wrote to Reid enclosing three memoranda expounding the Colonial Office’s views. He told Reid they were ‘for your private and independent consideration’. As Professor Williams shows in his invaluable 2005 work, The Australian Constitution: A Documentary History (at 711–13), Reid made some rather clumsy though not unsuccessful attempts to promote the Colonial Office’s positions during debates. While it is unclear precisely with whom Reid shared the memoranda (although importantly, we do know that he gave them to Edmund Barton as head of the Drafting Committee), their existence was first made public in 1900 when extracts were published in London.

Memorandum C was a marked-up version of the Draft Bill containing what Chamberlain called ‘friendly suggestions’ from his draftsmen. Next to the provision that became s 44(i) — the effect of which at that point was that disqualification would result only from a positive act acknowledging foreign allegiance — the author asked, ‘suppose he returns to his allegiance?’ This reflected a concern that a person who had previously taken an oath of allegiance to a foreign power would be disqualified on the wording of the draft clause regardless of any naturalisation process in Australia.

In short, the Attorney-General’s contention in Re Canavan was that the change in wording to s 44(i) between the Sydney session and the Melbourne session (on its face removing the requirement for a positive act) should be understood, in the light of the history, as a response by Barton’s Drafting Committee to the relatively benign query of the Colonial Office described above, not meant to drastically alter the meaning of the earlier draft.

Upon the Solicitor-General raising the memoranda on the opening morning in Re Canavan, the following exchange ensued:

EDELMAN J: What date did the Colonial Office memoranda become public?

MR DONAGHUE: Your Honour, I am not certain of the answer to that question. They became widely publicly available in the Williams book provided, quite recently in 2005. The article I mentioned was written in 1926, so clearly the author of that article had access to the material and was familiar with them, so at least for that long. But the memoranda were not, it appears, widely distributed amongst the convention delegates. They were given to the drafting committee and, according to that article by de Garis, most of the changes were unobtrusively incorporated into the Constitution, a very large majority of them.

These are memoranda that, since they have become more widely available with the publication of the Williams text, were referred to by Chief Justice French in a number of judgments. His Honour referred to them in Fortescue Metals Group Ltd v Commonwealth and in Williams v Commonwealth as part of the extrinsic material informing the construction of the Constitution.

EDELMAN J: But if they are to have some decisive effect or some effect which might tip the balance in terms of construction, does that then mean that the construction of the Constitution, the meaning of the Constitution, changes according to when a particular document becomes available to the public?

MR DONAGHUE: We submit not, your Honour. We are using this material only to help identify the purpose or the target of the particular provisions that were being drafted in the same way that different drafts of the Constitution and the Convention Debates have been used. The availability – or the ready availability – of those documents to the public has varied historically and, indeed, as your Honour knows, for much of the history of this Court, no reference was permitted to be made to the Convention Debates but we submit that the endorsement in Cole v Whitfield of resort to that material did not change the meaning of the Constitution. It changed the material to which the Court was prepared to look in identifying the true meaning of the Constitution.

EDELMAN J: Yes. But, is there any other written instrument that has ever construed by reference to material that is not reasonably available at the time the instrument is promulgated to those persons to whom it is directed – contracts, trusts, wills, statutes?

The only parties to raise any specific objection to the memoranda were the Greens Senators, who argued that no reliance could be placed on ‘memoranda of uncertain authorship and narrow distribution (notably not being distributed to the Convention Delegates)’.

Ultimately, these questions did not need to be resolved. At [33] of its unanimous judgment, the Court said:

Whether or not it is appropriate to have regard to the confidential Colonial Office memorandum, the extent of the redrafting of the predecessors of both ss 44(i) and 45(i) which occurred in the period between the Sydney session and the Melbourne session is such that it cannot adequately be explained as doing no more than responding to that query.

The question is thus left to be resolved on another day.

It should go without saying that it is not my intention to attribute a settled position to Justice Edelman. Judges ask questions for any number of reasons; those questions seldom invite close scrutiny. But where they raise unsettled issues of importance (and given that the question was reserved for future consideration in the Court’s judgment), there is no reason not to use what appears on its face as an embryonic suggestion as a starting point for further inquiry.

Confidential documents and original public meaning in Australia

The conceptual difficulty of relying upon documents that were not available at Federation to determine the meaning of the Constitution is evident from Edelman J’s statement at [76] of Graham v Minister [2017] HCA 33:

[T]he essential meaning [of the Constitution’s text] is not necessarily literal and it proceeds by reference to the way that the essential meaning would be understood by a reasonable, legally informed person at the time of utterance, which in this case is Federation. That process … is avowedly historical.

If one adopts that approach, the question raised by the memoranda is whether the hypothetical ‘reasonable, legally informed person at the time of utterance’ should have information that we know only a privileged few possessed in 1900 imputed to him.

It seems intuitively unsatisfactory that an inquiry into the meaning of the Constitution could yield one answer today and a contradictory answer the following day, nothing having changed in between other than the release of a secret document. A concern of this kind was made explicit by Edelman J in Graham (at [83], quoting from Heydon J’s dissent in Pape v Commissioner of Taxation (2009) 238 CLR 1), where he expressed an aversion to the notion that ‘the Constitution can have “two contradictory meanings at different times, each of which is correct at one time but not another”.’

In his 1997 defence of ‘moderate originalism’ in interpreting the Australian Constitution, Professor Goldsworthy said (at 20) that that approach:

holds that the meaning of the Constitution depends on evidence of the founders’ intentions which in 1900 was readily available to their intended audience, but not on other evidence of their intentions. … The meaning of a law, like that of any other utterance, is something public, not hidden. It is partly determined by what its intended audience knows, or can be reasonably expected to know, about its author’s intentions, but not about concealed intentions.

At first blush, that passage reads as an emphatic rejection of recourse to the memoranda. Considered in context, however, Goldsworthy’s approach arguably does not demand exclusion of the memoranda for all purposes. What it points to, rather, is the risk that in considering a confidential document it will be used to give effect to the subjective intentions of its author.

An orthodox solution

Wong v The Commonwealth (2009) 236 CLR 573 concerned the meaning of s 51(xxiiiA) of the Constitution, which was added after a 1946 referendum, providing a head of power with respect to: ‘the provision of maternity allowances, widows’ pensions, child endowment, unemployment, pharmaceutical, sickness and hospital benefits, medical and dental services (but not so as to authorize any form of civil conscription), …’. Two doctors who had been deprived of Medicare benefits by a Commonwealth legislative scheme that denied such benefits to practitioners who engaged in ‘inappropriate practice’ argued that the deprivation amounted to ‘civil conscription’ and was therefore invalid. Thus, the case substantially turned upon the legal meaning of the term ‘civil conscription’.

The legislative history showed that the parenthetical words had been added by the Leader of the Opposition, Mr R G Menzies. The Attorney-General and Minister for External Affairs, Dr H V Evatt, had sought written advice on Menzies’s amendment from the Solicitor-General, Sir George Knowles. That advice assuaged Evatt’s concern that Menzies’s insertion would prevent the Commonwealth from passing legislation to prevent medical practitioners from refusing to treat patients entitled to Commonwealth benefits.

Justice Kirby, in his staunchly and explicitly anti-originalist judgment, did not engage with Knowles’s advice; but the other members of the Court did. Justices Hayne, Crennan, and Kiefel said (at [183]):

The advice is of use only as one of a number of different sources which disclose the issue (or issues) to which the amendment proposed by Mr Menzies was directed, and which offer some indication of then current usages of the language that was ultimately incorporated in s 51(xxiiiA).

Justice Heydon (at [277]) relied upon the advice as evidence of the ‘contemporary understanding of the words “civil conscription” in the context of medical services’. A footnote to that paragraph comes close to addressing the issue at the heart of this post:

A ‘moderate originalist’ has thoughtfully argued that only evidence of ‘the founders’ intentions which … was readily available to their intended audience’ may be examined – which would exclude private communications like that of Sir George Knowles and his colleagues: see Goldsworthy, ‘Originalism in Constitutional Interpretation’ (1997) 25 Federal Law Review 1 at 20. However, if the correct approach is to search, not for the actual intention of the framers, but for what their words meant at the time they were used – and it is this which Cole v Whitfield seems to favour – the objection is not open. Even quite secret contemporary material could cast light on contemporary meaning. (emphasis added)

These passages support the view that regard may be had to the Colonial Office memoranda, provided that the reliance placed upon them accords with the ordinary limitations upon the use of extrinsic materials. Those limitations are found in Cole v Whitfield (1988) 165 CLR 360 (at 385):

Reference to the history of s 92 may be made, not for the purpose of substituting for the meaning of the words used the scope and effect — if such could be established — which the founding fathers subjectively intended the section to have, but for the purpose of identifying the contemporary meaning of language used, the subject to which that language was directed and the nature and objectives of the movement towards federation from which the compact of the Constitution finally emerged.

Evidently the authors of the above passages in Wong were conscious of the Cole v Whitfield limitations (the joint judgment echoing them and Heydon J citing them) and exercised discipline in observing them.

If follows that the High Court would likely take the view that the extent to which the memoranda were in the public domain at Federation is less important than the way in which they are to be relied upon in the case at hand. In other words, secrecy is probably not a determinative ground for exclusion in interpreting the Constitution, provided that the material is deployed in accordance with the limitations laid down in Cole v Whitfield.

The United States Experience

The United States Constitution is the most prominent document answering Edelman J’s appeal for ‘any other written instrument that has ever [been] construed by reference to material that is not reasonably available at the time the instrument is promulgated’.

The Constitutional Convention of 1787 met in private, and on 29 May 1787 resolved ‘[t]hat nothing spoken in the House be printed, or otherwise published, or communicated without leave.’ The Secretary, William Jackson, kept an official journal of proceedings and various deputies kept notes of varying quality and detail.[1]

James Madison recorded that after debate on whether the official journal should be either destroyed or turned over to the President, the latter option was chosen. On 17 September 1787, Jackson told George Washington that he would wait upon him that evening to deliver the journal (but only ‘after burning all the loose scraps of paper which belong to the Convention’!).

Washington retained the journal until March 1796, when he deposited it with the Department of State. The journal remained under seal with the Department until Congress resolved in 1818 to publish it under the President’s direction, which was achieved in 1819.

Pausing here, this history shows that both Marbury v Madison (1803) and McCulloch v Maryland (1819) were decided before any record of the Constitutional Convention was published. Indeed, a full 12 years after Marbury, Thomas Jefferson asked John Quincy Adams whether he knew:

that there exists in manuscript the ablest work of this kind ever yet executed, of the debates of the constitutional convention of Philadelphia in 1788? The whole of every thing said and done there was taken down by Mr. Madison, with a labor and exactness beyond comprehension.

Madison decided that his notes should be published posthumously, delaying release of his authoritative account until 1840. Other deputies’ notes continued to be published sporadically during the late-nineteenth and early-twentieth centuries.

Once Madison’s notes became public, courts began relying upon them almost immediately. In Carpenter v Pennsylvania (1854), the Supreme Court said that Madison’s notes showed that the Framers never intended for the Ex Post Facto Clause to apply to civil laws. Since that decision the Court has developed a rich tradition of engaging with the material.[2]

Insofar as any attempt has been made in the United States to deal with the problem raised by Edelman J, the solution reached by the strictest American proponents of fidelity to original public meaning substantially accords with the approach in Wong. While on the one hand Robert Bork was comfortable engaging with the once-confidential debates of 1787 to illuminate the subject to which the Framers’ language was directed (Plaquemines Port v Federal Maritime Commission 838 F.2d 536, 546), on the other hand he was dismissive of private documents going to subjective motives (The Tempting of America at 144):

If someone found a letter from George Washington to Martha telling her that what he meant by the power to lay taxes was not what other people meant, that would not change our reading of the Constitution in the slightest. Nor would the subjective intentions of all the members of a ratifying convention alter anything.

Justice Scalia’s rationalisation closely reflected that of Heydon J in Wong. That approach was to treat the Convention records as historical evidence of the understanding of legally educated men of the meaning of the words they had enacted. In the colourful words of the late Justice (Scalia Speaks at 183): ‘It is not that “the Constitution must mean this because Alexander Hamilton thought it meant this, and he wrote it”; but rather that “the Constitution must mean this because Alexander Hamilton, who for Pete’s sake must have understood the thing, thought it meant this.’

Notwithstanding the well-known influence of many aspects of the United States Constitution upon the Framers of the Australian Constitution, it has been clear since at least the Engineers’ Case (1920) 28 CLR 129 that where some fundamental feature distinguishes a constitutional issue arising in the United States from a seemingly similar issue in Australia, the American position will be ignored. Hence, in considering the respective rights of the Commonwealth and States, the majority in that case said (at 146) that the common sovereignty of all parts of the British Empire, together with the principle of responsible government ‘radically distinguish [the Australian Constitution] from the American Constitution.’ The issue considered in this post, which to some extent inheres in all written constitutions, is not affected by any such radically distinguishing feature. The values at stake — including that laws should be promulgated in the public domain and be clear and stable in meaning, and that the original meaning and purpose of the language enacted is at least relevant to its interpretation — are, broadly speaking and allowing for differences in emphasis, shared by the two legal traditions. While the United States practice would not be adopted uncritically, it is at least worthy of consideration.

Application to Re Canavan

Considering the Attorney-General’s argument in Re Canavan in the light of the above discussion, it seems that reliance upon the memoranda was permissible. The relevant criteria from Cole v Whitfield were that they be used as evidence of: (1) what the Framers understood the words to mean; or (2) the objectives pursued by the Framers to the extent that such objectives informed their linguistic choices. In Wong, the members of the joint judgment appeared to place Knowles’s advice into both categories, while Heydon J restricted his use to the first. The use of the memoranda in Re Canavan fell within the second. In essence, as is clear from the Solicitor-General’s answer to Edelman J’s second question in the above transcript excerpt, the argument was that the memoranda bespeak a narrow objective – that of responding to a very specific concern of the Colonial Office.

The only remaining objection appears to be that reliance upon once confidential material would mean that our understanding of the Constitution would not remain fixed as it was at Federation. There are of course numerous instances where orthodoxy demands such a result. The range of new interpretive possibilities arising from examination of the debates permitted by the Court in Cole v Whitfield is but one example. Moreover, as McHugh J observed in Eastman v The Queen (2000) 203 CLR 1 (at [154]):

[E]ven when we see meaning in a constitutional provision which our predecessors did not see, the search is always for the objective intention of the makers of the Constitution. A commitment to discerning the intention of the makers of the Constitution … does not equate with a Constitution suspended in time. … This Court has not accepted that the makers’ actual intentions are decisive, and I see no reason why we should regard the understandings of the immediate audience as decisive. (emphasis in original)

Ultimately, the Court in Re Canavan held that the extent of the amendment did not accord with the narrow objective suggested in the memoranda. That conclusion confirms that it is the words enacted in the Constitution that govern, and the search for legal meaning is not an exercise is divining the Framers’ subjective intentions.


Once it is acknowledged that a confidential document could never be used to give effect to subjective intention at the expense of objective meaning, and that the use of private documents must always be subject to the orthodox limitations upon the use of extrinsic material in interpreting the Constitution, the grounds for objection to reliance upon documents that were not publicly available at Federation seemingly fall away. To the extent that a document merely evidences a hidden intention (like Bork’s hypothetical letter from Washington to his wife), it ought to be excluded. But where the material reliably informs us of the contemporary understanding of language and the stimuli to which the Framers’ language was a response (like the Colonial Office’s memoranda), it need not be ignored.


[1] For a more detailed account of the below history, Professor Max Farrand’s 1911 three-volume documentary history of the Convention presents the material chronologically and includes an illuminating introduction. Farrand’s Records are available in full on the Library of Congress’ website.

[2] While I am unable to engage in a full account of this tradition in this post, a summary appears in Gregory E Maggs, ‘A Concise Guide to the Records of the Constitutional Convention of 1787 as a source of the Original Meaning of the US Constitution’ (2012) 80 George Washington Law Review 1707.


Henry Cooper is a tipstaff to a Judge of the New South Wales Court of Appeal.

Suggested citation:  Henry Cooper, ‘”For Your Private Consideration”: Secret Documents and the Public Meaning of the Constitution’  on AUSPUBLAW  (12 February 2018) <>