BY DAVID TAN
In the Engineer’s case, Higgins J stated:
‘The question [of constitutional interpretation] is, what does the language mean; and when we find what the language means, in its ordinary and natural sense, it is our duty to obey that meaning’
Foreseeably, a large number of Australian lawyers and judges would agree with this: to identify constitutional laws we need to find out the meaning of provisions based solely on its language and not moral or policy considerations. This is what the legal philosopher Mark Greenberg calls the Linguistic Content Thesis.
Despite this seemingly innocuous assumption, I will use the potential fallout of disqualifications under Section 44 to show that there are times when the Linguistic Content Thesis is put under pressure. The main problem outlined here is that if parliamentarians are disqualified, the Linguistic Content Thesis entails that their ministerial decisions and votes on legislation should all be invalidated as well. I propose that if the reader wants to hold on to the Linguistic Content Thesis and uphold previous parliamentary actions, they will have to allow judges to change the law. I give reasons however for thinking that judicial change of the law is justified. As a clarification, I will not consider here the interpretive issues surrounding whether the parliamentarians should be disqualified in the first place but only the consequences of disqualification.
The Vardon and De Facto Officer Doctrine
The positions of several parliamentarians, some of whom are ministers, are in question due to s 44 of the Constitution. If they are disqualified, existing case law has interpreted their elections as being void. Consequently, two issues arise: first, whether their previous votes in parliament for legislation were valid and secondly, whether their ministerial decisions were valid.
On the first point, the High Court in Vardon v O’Loghlin decided that the votes of disqualified members would still be valid as it is necessary to do so. Why exactly it is necessary is not explained. I shall refer to this as the Vardon Doctrine. On the second point, Anne Twomey has commented that their decisions might still be upheld based on what is called the De Facto Officer Doctrine (‘the DF Doctrine’). The DF Doctrine states that if people have been relying on these decisions based on the (mistaken) belief that they had been made by actual public officials, then those decisions would be valid. I shall refer to the Vardon and DF Doctrines collectively as the Doctrines.
The justification for the Doctrines is that the public interest in the stability and predictability of the law would require them to hold. However, compared to the Vardon Doctrine, there is less support for the DF Doctrine in Australian case law. I am unaware of cases explicitly confirming the DF Doctrine for public actions under the constitution. Further Gummow and Hayne JJ are sceptical that the DF doctrine applies to officers under s 75(v) and Kirby J has expressed doubt more generally whether the DF Doctrine applies to constitutional invalidity.
Regardless of the level of acceptance of the Doctrines, those who subscribe to the Linguistic Content Thesis have problems with both as there is nothing in the explicit language of the Constitution that justifies them. On the matter of ministerial decisions, Section 64 says that Ministers can only hold the position if they are in parliament within three months of their ministerial appointment. Hence if they were never parliamentarians then they wouldn’t have been Ministers either. This would mean that they never had the power to make any decisions. The DF Doctrine’s proposition that citizens can still rely on those actions is not supported by any constitutional provision.
With regards to legislative voting, sections 23 and 40 of the Constitution state that a majority of both houses respectively is needed to answer questions arising in the respective houses. A reduction in the number of parliamentarians would affect the count. If for example, legislation was passed with one party having a majority of just one person, then the disqualification of one person who voted yes for the bill would mean that the bill should have failed. Following the language of ss 23 and 40 would be inconsistent with the Vardon Doctrine.
To reject the Doctrines would lead to the status of numerous legislative and Ministerial decisions put into doubt. All else being equal, a theory of interpretation that led to such an impractical and unstable outcome might be deemed unattractive compared to another theory that doesn’t. If the reader doesn’t think enough chaos will result, consider George Williams’ latest conjecture that an audit of parliament might give rise to the doubt of qualification of 20 or more members of parliament. For the purposes of this post I shall assume that rejecting the Doctrines is not an attractive option.
If the Linguistic Content Thesis is false, then we have a much easier time justifying the Doctrines. Judges would be able to use moral and policy considerations in interpreting provisions. Hence the stability of the legal system might be taken into account to justify the Doctrines. It is important to note that this is not the contention that moral and policy reasoning comes into play only when there is linguistic ambiguity or vagueness. Rather it is possible that while the language of a provision says one thing, there are moral and policy reasons to say that the law is something else. For arguments as to why the Linguistic Content Thesis doesn’t hold, see Mark Greenberg generally and Dale Smith for an Australian application.
Defending Linguistic Content
As someone who subscribes to the Linguistic Content Thesis (at least for the text of the Constitution, I have argued elsewhere that it doesn’t hold for case law), I want to provide some potential solutions to this issue to maintain the Linguistic Content Thesis without rejecting the Doctrines.
I shall consider two possible solutions here. The first is that the Doctrines are implied in the constitution. The second is to suggest that in some cases judges can change the law (the original law being what was originally communicated by the constitution). One method I won’t consider is to try and imply a rule of law requirement in the constitution and then state that such a requirement entails the Doctrines. I assume this method won’t work and defer to Lisa Burton Crawford who has written on this issue.
Solution 1: Implications
For reasons of space I will only consider two methods of constitutional implication discussed by Jeffrey Goldsworthy in a 2011 paper. I don’t think that these work. Remember that a key factor is that they must be explained in a linguistic manner.
The first method is that of “practical necessity” which Goldsworthy suggests is used in Lange. A particular proposition of law will arise, by implication, if it is practically necessary to give effect to what is explicitly stated in the constitution. The problem with practical necessity is that it is hard to find a linguistic basis for it. People can communicate impractical things. A madman can tell someone “Bring me a million dollars!” and we wouldn’t think that he also implied “but not if you don’t have that much money”. Now the framers of the Constitution were (presumably) not madmen but Goldsworthy rightly is sceptical about the practical necessity test as framers can make mistakes or overlook matters. Suppose my billionaire friend says he will give me a million dollars. One night I see a figure whom I confidently think is my friend and jokingly call out “Bring me my million dollars!” It turns out that this figure is actually someone else who doesn’t have that kind of money. It would be odd for someone to say that I implied “but not if you don’t have that much money” since I genuinely thought it was my friend who was a billionaire. In the current case, the framers when debating s 44 were concerned about the wording of s 44 and who it would capture – see here, here and here. They did not however discuss what would happen once they were disqualified. As I see it, this is exactly a case where there has been an oversight or mistake.
A second method is using presuppositions via what Goldsworthy calls background assumptions. He gives this example: if I went to a café and ordered a burger, it would be assumed that the burger I have asked for would be an edible one and not one encased in unbreakable plastic. This would be the presupposition even if I, as the speaker, did not consciously think “I want a burger that can be eaten”. To see if this approach is workable, we need to consider in more detail what the “background” actually is. As I understand Goldsworthy, background assumptions are something like the set of commonly held beliefs in a given society. In the example above, a commonly held belief both by the speaker and audience is that burgers are edible. Unlike practical necessity, some theories of language do involve concepts similar to background assumptions. Nonetheless, the background doesn’t help in s 44; there isn’t enough communal beliefs to think that parliamentarians who get disqualified won’t have their decisions reversed. Goldsworthy might try to limit this to communal beliefs of lawyers, but remember that on his linguistic method it is fixed to communal beliefs held (at least) by the speaker. I don’t think it was obvious even among lawyers in 1901 that the Doctrines were clearly true (or that anything like them even existed).
Solution 2: Changing the Law
I want to suggest that in cases where laws lead to unintended consequences, which obviously occurred due to an oversight or mistake of the lawmakers, there can be a legal duty to change the law. The law is still what the original linguistic meaning of the text is (so the Linguistic Content Thesis is still true), but judges can replace the law with their own decisions. Goldsworthy in his 2011 paper does suggest that something like this can be done (at p 20) but is not entirely clear how this is justified. I want to present an argument that judges can change the law not because of moral or policy reasoning, but by tracing it back to the will of the framers.
To do this I want to talk about what economists call preference changes. A preference is just what choice I would prefer between two or more options. For example, between coke and tea I prefer drinking coke. However, preferences can change depending on what we believe about our actions and their consequences. I might prefer coke today but if tomorrow I found out that coke causes cancer I would drink tea instead. We can distinguish between the preference of someone with the actual information they have and their preferences if we gave them new information. Call the former their original preferences and the latter their updated preferences. Suppose that Daniel is someone who is very health conscious and would never ingest anything unhealthy no matter how good it tastes. He also doesn’t know that coke has bad effects on your health. Based on his original preferences, perhaps he treats tea and coke equally. However, if we update his beliefs by telling him that coke is really bad for your health, we would get an updated preference that he would rather drink tea than coffee. It is important to point out that we can predict that Daniel would decide to stop drinking coke simply because we knew that he values health. We don’t need Daniel to actually form the updated preferences in his mind.
To show why this is relevant consider this scene in the movie Interstellar: the main character tells his son to keep driving the truck they are both in, and not to stop for anything, in order to catch a drone. The son sees a cliff coming up but keeps driving while the father is distracted. The father notices it in time, stops the son and is puzzled as to why the son didn’t stop. The son’s answer was that the father asked him to keep driving no matter what. In such a case we might say that the father had an original preference for the son keep to driving to catch the drone. However, when we update the father’s information to knowing that they might die his updated preference is to stop driving. The son should have easily predicted this. This explains why I don’t think anyone would blame the son if he had disobeyed the instruction to keep driving. Analogously, in the case of s 44, the original preference of the framers is for judges to follow s 44. However, if we had given them the piece of information that this would lead to much governmental chaos, it seems to me quite likely that their updated preference would be to avoid instability. As stated above, the framers don’t actually need to form this updated preference and instead it can be predicted. Similar to the son about to drive over a cliff, judges can ignore s 44 because they have predicted what the framer’s updated preferences.
Some might think that this approach leads to too many constitutional provisions being in danger of judicial change. I cannot fully defend against this here but allow me to make two quick points. Firstly, this is only supposed to apply in cases of mistakes and oversights. If there is evidence that the framers thought about the issue and didn’t choose to include a certain provision in the Constitution, then it is not for the judge to second guess the framers’ judgment. Secondly, in most cases it will be unclear what the updated preferences would be. Suppose the framers knew about modern economies, what would they think about s 92? I’m not sure anyone really knows. Changing the law is only allowed in very transparent cases (e.g. s 44) where there is very strong evidence that they would update their preferences.
 See the Background in John Searle, Making the Social World (Oxford University Press, 2010) 31-2; the common ground in Robert Stalnaker, ‘Common Ground’ (2002) 25 Linguistics and Philosophy 701.
 In the literature they are actually called unconditional and conditional preferences, but I have used this terminology for simplicity.
David Tan is a Lecturer at Deakin University. David thanks Patrick Emerton, Jeffrey Goldsworthy and Jayani Nadarajalingam for providing useful comments on this post.
Suggested citation: David Tan ‘Section 44, Interpretation and Changing the Law’ on AUSPUBLAW (20 September 2017) <https://auspublaw.org/2017/09/section-44-interpretation-and-changing-the-law/>