The connotation/denotation distinction in constitutional interpretation *.
One more contribution to the long debate over constitutional interpretation calls for some justification. This Article tackles a specific problem of interpretation in Australian constitutional law, although it is a problem that will arise in regard to the interpretation of any document which must be read and applied over a long period of time. The Article seeks to bring the concepts of contemporary philosophy of language to bear upon the problem with the aim of delineating what we mean by "meaning." The problem chosen to explore the issues just described is the long recognised difficulty in applying terms in the Constitution to entities or activities that did not exist at the time the Constitution was drafted or passed into law. Difficulties of this sort for constitutional interpretation may be arising more often as the time when the document was written recedes and the volume of social change since its inception increases.
It will be argued that a solution to this problem of constitutional interpretation involves abandoning attempts to ascertain the meaning of the document as the sole means of using the Constitution. It will be further argued that a practice of constitutional interpretation that is restricted to ascertaining the meaning of the Constitution will be practicably unworkable. Finally, it will be argued that a practice of interpretation not based upon ascertaining meaning requires a justification quite different from one restricted solely to meaning. What is presently lacking is a justification of such practice. This Article will not seek to offer such a justification, but it will suggest some conditions that any such justification would need to satisfy.
A classic example of the interpretational problem described above, although one that proved relatively easy of solution, was posed by The King v. Brislan; ex parte Williams, (1) in which it was held that the words in section 51(v) of the Constitution (2) conferring power upon the Commonwealth to make laws with respect to "postal, telegraphic, telephonic and other like services" extended to radio broadcasting, even though such technology did not exist in the 1890s. The words "other like services" at the end of the provision, and the knowledge that already existed in the 1890s regarding the electro-magnetic spectrum made the finding rationally supportable. (3)
Other issues prove less tractable. Laws of the Commonwealth to implement treaties protecting the environment have been held to be laws with respect to external affairs, (4) a result that might have surprised at least some of the constitutional founders. By contrast, the High Court confirmed in Eastman v. The Queen (5) that the meaning of the word "appeal" in section 73 was to be construed in accordance with historical usage, which would not have carried the implication that fresh evidence could be admitted. (6)
Many more examples can be recited of instances in which social change has placed pressure upon the way in which one interprets provisions of the Constitution. Some of these have been before the High Court; others no doubt will be litigated in years to come. At Federation the word "marriage" undoubtedly meant a relationship between a man and a woman. The time fast approaches when the term may, at least amongst a substantial portion of the population, be used to describe certain relationships between people of the same sex. An issue will arise as to whether Commonwealth powers under section 51 (xxi) of the Constitution in regard to marriage, extend to making laws for such same-sex relationships.
Similar issues have arisen concerning challenges to the Commonwealth's power to legislate in regard to computing technology and rights in regard to genes and plant varieties. Scholars suggest that these rights are sui generis. (7) These arguments amount to the assertion that computing technology or genetic discoveries are not in truth copyright subjects or patentable inventions as those concepts were originally understood in section 51 (xviii) of the Constitution. (8)
Numerous further examples could be generated, in which entities or activities are said to be capable of regulation under Commonwealth laws, in turn said to be authorised by provisions of the Constitution, which were written at a time when the activities or entities did not exist.
One solution to the problems just described, and a solution frequently relied upon by judges of the High Court of Australia, has been to note the distinction between the connotation and denotation of words. Put simply, the connotation of a word is the sense or meaning of the word, while the denotation is the class of things identified or picked out by its meaning. Thus, although the connotation (or meaning, in the narrow sense) may remain unchanged, the denotation of the term is said to alter over time as new objects come into existence which are capable of being identified as members of the denoted class. (9)
The connotation/denotation distinction was developed within the field of semantics and logic, although it is little used in contemporary semantic theory. However, contemporary semantic theory does make much use of another distinction, similar in important respects, namely the distinction between the sense and the reference of words, a distinction first propounded by the German philosopher Gottlob Frege in 1892 in his paper Uber Sinn und Bedeutung (usually translated into English as On Sense and Reference). (10)
It is doubtful if Frege's distinction can do the work demanded by those who wish to maintain that the meaning of the Constitution may remain unchanged while the reference of its terms can vary. More recent criticisms of Frege's work make even more doubtful the possibility of a coherent account of meaning in which the terms of the Constitution could have a fixed meaning but a mobile reference. (11)
Contemporary semantic theory offers the judge or lawyer a difficult choice. If one wishes to maintain that the legal interpreter is solely concerned with meaning, then one may well be stuck with a single fixed meaning associated with the meaning of the words at the time they were uttered. This approach, however, should not offer much consolation to those so-called textual originalists such as Justice Scalia of the United States Supreme Court, since it will be argued that consistent with such an approach, the Constitution will have a narrow meaning in regard to substantial numbers of major legal controversies in contemporary times. It might even be said that the Constitution has no meaning in regard to these contemporary legal controversies. A purist would suggest that the Constitution means what it says and cannot therefore be accurately described as having no meaning. Nevertheless, the point is that these are controversies which are, by their nature, matters upon which the Constitution and its drafters have never spoken, and therefore, which have never been dealt with. In that sense, and in regard to these issues, the Constitution has no meaning.
The consequence, it will be argued, is that the process of constitutional interpretation must for practical reasons eschew a concern solely with the meaning of the document strictly understood. However, once the process of judicial decision making on constitutional cases is severed from the meaning of the Constitution, a new problem arises, namely, how or in what principled fashion can judges be constrained by the constitutional text. No theory presently on offer solves this problem.
The connotation/denotation concept was explained by Justice Windeyer in The Queen v. Commonwealth Conciliation & Arbitration Commission; ex parte Professional Engineers Association (12) in the following terms:
We must not, in interpreting the Constitution, restrict the denotation of its terms to the things they denoted in 1900. The denotation of words becomes enlarged as new things falling within their connotations come into existence or become known. But in the interpretation of the Constitution the connotation or connotations of its words should remain constant. We are not to give words a meaning different from any meaning they could have borne in 1900. Law is to be accommodated to changing facts. It is not to be changed as language changes. (13)
The use of this distinction in constitutional interpretation was also explained by Justice Dawson in Street v. Queensland Bar Association. (14) His Honour there said:
I speak of 1900, the time of Federation, because it is in accordance with the meaning given at that time that the limits of the phrase "trade and commerce" must be ascertained. The essential meaning of the Constitution must remain the same, although with the passage of time its words must be applied to situations which were not envisaged at federation. Expressed in the technical language of the logician, the words have a fixed connotation but their denotation may vary from time to time. That is to say, the attributes which the words signify will not vary, but as time passes new and different things may be seen to possess those attributes sufficiently to justify the application of the words to them. (15)
His Honour thereafter traced the technical use of the terms back to John Stuart Mill's A System of Logic and usefully collects the High Court decisions since 1908 that have relied upon the distinction.
In Re Wakim; ex parte McNally, (16) Justice McHugh accepted the distinction, noting its similarity to the distinction drawn by Ronald Dworkin between "concepts," being the abstract notions expressed in the Constitution, and their application to present day "conceptions." His Honour said:
Indeed, many words and phrases of the Constitution are expressed at such a level of generality that the most sensible conclusion to be drawn from their use in a Constitution is that the makers of the Constitution intended that they should apply to whatever facts and circumstances succeeding generations thought they covered. (17)
Professor Leslie Zines has suggested that the distinction relied upon by the High Court is a now "outdated philosophical distinction," (18) bringing a riposte from Justice McHugh (19) and Justice Kirby. (20)
The distinction continues to do work for members of the High Court. In Re Patterson; ex parte Taylor, (21) those members of the Court who sought to deal with the constitutional power of the Commonwealth to make laws in regard to aliens pursuant to section 51(xix) of the Constitution had to grapple with the difficulty that in contemporary immigration law, an alien was thought to be anyone who was not a citizen of the Commonwealth. At the time the Constitution was drafted, citizenship was not an important legal concept, and anyone who was a British subject (even though born in Britain and without any other connection to Australia) would not have been considered an alien. The Court in Patterson overruled its previous decision in Nolan v. Minister of State for Immigration. (22) Members of the majority again concluded that while it was not possible for Parliament to alter the meaning of the word "alien" as it appeared in the Constitution, circumstances may bring it about that the class of people now denoted as aliens by that term could be quite different from those denoted as aliens in 1900. (23)
In Patterson, Justice McHugh further noted how in Sue v. Hill (24) the Court had held that the term "foreign power" in section 44(i) of the Constitution now includes the United Kingdom, although in 1901 and for long after, the United Kingdom was not a foreign power within the meaning of that term. (25)
JOHN STUART MILL'S SEMANTIC THEORY
In A System of Logic, (26) Mill said: Connotative names have hence been also called denominative, because the subject which they denote is denominated by, or receives a name from, the attribute which they connote. Snow, and other objects, receive the name "white", because they possess the attribute which is called "whiteness"; Peter, James and others receive the name "man" because they possess the attributes which are considered to constitute humanity. The attribute, or attributes, may therefore be said to denominate those objects, or to give them a common name. (27)
Mill acknowledged Archbishop Whately for having drawn attention to this distinction in his Elements of Logic, (28) a work by which Whately had revived the study of formal logic in England after some two centuries of neglect. (29) However, the term "to connote" was first used by Mill's father, James Mill, although in the sense conveyed by John Stuart Mill's term "denotation." (30)
By contrast with connotative names, the younger Mill referred to proper names, which denote only the individuals called by the name and do not indicate or apply any attributes as belonging to those individuals.
Finally, Mill recognised a category of connotative names given to individuals because the name signifies attributes, but only one individual possesses the attribute. Mill gives the example of "God" when used by a monotheist. On the other hand, a general name denotes that class containing the indefinite multitude of individuals so named.
Clearly, general names can refer to objects which do not exist at the time of an utterance. A general name will connote certain attributes which will then in turn denote all those subjects possessing the relevant attribute. (31) Thus the word "person" used today (or used in 1900) may denote the class of all beings belonging to the species homo sapiens, whether they existed in the past, present, or future, if that be the relevant meaning of the word in the utterance.
But it can be seen immediately that if the word "person" was used in the 1900s in the general sense to refer to all persons, past, present, and future, or to at least some category of persons which would include future persons, it denoted those persons who would come into existence in the twenty-first century, and it denoted those people existing in 1900 at the time the word was uttered. The denotation of the word "person" does not change with the coming into existence of each new object which may be described by the general name "person." Mill's concept simply does not permit the denotation of a general name to change while the connotation remains the same. Indeed how could it, since it is the connotation of a term which fixes its denotation?
How does a name or word come to connote some particular attribute or class of attributes by which the objects denoted may be ascertained? Mill's theory of connotation was not a complete attempt to explain how words have meaning, and treating meanings of general terms as a list of attributes might these days be thought an inadequate explanation of meaning and a replacement of one mystery by another.
Since at least the publication of Grice's seminal article Meaning, (32) philosophical analyses of the problem of meaning have recognised the importance of intention to an explanation of meaning. Grice's project, which remains controversial, was to demonstrate that the primary notion of meaning was explicable in terms of the intentions of an utterer to induce in the audience certain states of belief. Grice's explanation went on to show that the notion of meaning required that the utterer intended the belief to be induced by his or her communication, and the audience to recognise that it was the utterer's communication that was to cause them to have the belief. (33)
Grice's concept emphasised that meanings are essentially mental or psychological entities, but at the same time, language is a shared social practice, and dictionary or literal meaning (or what lawyers currently call textual meaning) is the mental state typically associated with, or shared by, speakers of a language, not the idiosyncratic mental state of some individual speaker.
Although proponents of truth-conditional semantics have sought to offer in opposition to Grice's view an analysis derived from the philosophy of logic, equating the meaning of a statement with the conditions that would render the statement true, at least some proponents of this view still recognise that for an utterance to be a meaningful utterance, it must be an intentional utterance. (34)
It is not necessary that every object referred to by a word be within the contemplation of the utterer as part of its meaning. Further, there will undoubtedly be inferences capable of being drawn from the meaning of terms, at least logical inferences, that will not have been within the contemplation of utterers. (35) However, the meaning of any particular text or utterance will still be based upon the shared intentions and practices of the community of language users from which the text or utterance arose. Applying these concepts back to Mill's theory, it is those shared intentions and practices that dictate the attributes connoted by a word.
Clearly it is quite consistent with Mill's terminology that a general name used in the Constitution in 1900 might denote something that was not then in existence; this will be because the meaning of the term as used in the Constitution in 1900 extended to that whole class of objects which includes the present object in issue. Thus, if the attributes connoted by the term "copyright" as used in the Constitution in section 51(xviii) in 1900 were of sufficient generality to encompass computer programs, then computer programs are denoted by that term. The non-existence of computing technology in 1900 is not a reason in itself to conclude that the term could not have denoted computer programs, any more than the fact that I did not exist in 1900 is a reason to find that denotation of the term "person" in the Constitution does not denote me.
But it is now plain that the connotation/denotation distinction offers little assistance in answering the question, whether in 1900 the term "copyright" could be applied to, amongst numerous other subjects, computer programs. However, when we now put the question in those terms, it obviously seems far less likely that in fact it did. To use Mill's term, the "attributes" connoted by the word "copyright" in 1900 will have been determined by the shared understandings and intentions of competent users of the word at that time. The list of attributes that any competent speaker of English in 1900 would have used to enumerate the characteristics of a work in regard to which copyright might be claimed would not have extended to computing technology. Still more telling arguments could be developed, in regard to plant variety rights or genetic discoveries, for example, as subjects of Commonwealth legislative power.
Returning to our central problem, if we attribute one particular meaning to the Constitution in 1900 and we wish to continue applying that meaning (indeed we may doubt whether there can be any other meaning), how do we apply the terms of the Constitution to facts and circumstances not foreseen or imagined at the time that the document was written? There may well be terms of the sort referred to by Justice McHugh in Wakim (36) couched at such a level of generality that they can be said to have meant or denoted objects of a type never envisaged in 1900. Trade and commerce, even on the meaning possessed by those terms in 1900, plainly denoted contracts to buy and sell computers, television sets, and other such goods never dreamt of at the turn of the century. On the other hand, where the term denotes a bundle of attributes (to use Mill's language) which had a commonly accepted meaning connoting those attributes in 1900, it is a far more difficult question to determine whether the term could also mean a different, albeit a similar or related bundle of attributes, denoting objects of a sort that did not exist in 1900.
FREGE AND THE CONCEPTS OF SENSE AND REFERENCE
Gottlob Frege relied upon a distinction very similar to Mill's connotation/denotation distinction in the development of his semantic theories. The importance of Frege's work to modern logic and semantics leaves no room to suggest his distinctions are philosophically outdated. His work provides the best chance of making good, in some revised form, the connotation/denotation concept.
One of Frege's principal concerns was to determine why identity statements in regard to proper names were necessarily true. Using Frege's famous example, it is true that the morning star is the evening star (both are the planet Venus). If, however, one imagines the truth of the identity statement to be a consequence of the object referred to by each phrase being the same, a paradox arises. We cannot substitute one phrase for the other in all possible applications and still preserve the truth of the resultant proposition. Thus, while the sentence "the evening star is the morning star" is true, it is not necessarily true that "John believes that the evening star is the morning star" (he may mistakenly believe them to be different planets). (37)
For Frege, that one could not substitute the phrases "morning star" and "evening star" in all contexts while preserving truth was explained by words having both sense as well as reference. Thus evening star and morning star have the same reference; they both designate the planet Venus. On the other hand, they have a different sense. Frege's explanation of the concept of sense is illusive, something like mode of presentation of reference, or way of thinking of something.
Frege's theory was largely elaborated in regard to problems associated with the semantics of proper names. Much of Frege's important work on this subject was published posthumously, and it has been a source of philosophical controversy as to whether the concepts of sense and reference can be applied to general concept words as opposed to proper names, although some contemporary philosophers have argued that it is capable of such application. (38)
Frege's theory provides no avenue for justifying a theory of interpretation which would permit the meaning of terms to remain fixed while the range of objects to which they refer varies. One cannot treat Frege's concepts of sense and reference as referring to meaning on the one hand, and the actual objects denoted by a meaningful expression on the other. Rather, Frege considered sense and reference were both aspects of the meaning of an expression. To each sense there will be a reference; however, the very point of the distinction is that although a single reference may be associated with a number of senses (the Reverend Dobson, Lewis Carroll, the author of Alice in Wonderland are three separate senses all associated with the one reference). It will offer no assistance in solving the problems of constitutional interpretation associated with new and unforeseen activities. Frege's theory will not permit a single meaning to have a changing or evolving reference.
MEANING AND SCRIPT
Another way of seeking to resolve the problems caused by social and technological change also needs to be examined and rejected. In recognising that meanings are tied up with the shared intentions and practices of a community of language users at some particular time, it follows that the meaning of a text does not change even if a script for that text may be read and given a meaning other than that given to it by the community by which it was written. Language tokens such as writing or the sounds made by human vocal chords during speech are ways of representing meaning but are not themselves meanings. Thus, the quite different scripts "dog" and "hund" are respectively the English and German signs that mean "dog."
The process by which common words of English acquire new meanings is familiar (39). However, it would be misleading to suggest that meanings have changed; rather, the signs or symbols of the English language have altered in the meanings they now represent. Words such as "gay" or "cool" may now be used to represent the meanings of homosexual or fashionable, but these were meanings always understood by English speakers; what in truth has happened is that we can now represent these meanings with different symbols. To put it this way is not to deny that the symbols with which meanings are represented may bring their own resonance to words, so that "gay" and "homosexual" may not enjoy perfect synonymity.
In the case of a text written or drafted over 100 years ago, such as the Constitution of Australia, it would be surprising if there had not been at least some drift so that the symbols that constitute the text no longer represent the same meanings in modern English that they represented in Australian English of the late nineteenth century. This drift is unlikely to have occurred in any coherent and principled fashion. One should not expect, if one reads the text of the Australian Constitution as a text of modern Australian English, that it would read as a coherent and consistent body of constitutional rules. The drift or slippage between meanings, and the signs in English used to represent them does not bring about some automatic updating of the Constitution; rather, it is part of the noise or static which impedes our properly understanding the meanings to be conveyed by the text. (40)
The problem posed by the drift which causes the symbols of the language to represent different meanings with the passage of time is more chronic than may be imagined. Once it is recognised that words do not have an evolving or mobile denotation or reference, permitting them to gather up new activities different in kind from those originally described, then it must be understood that those new activities were not subsumed by the meaning of the word in the original text, but rather, the language symbol (or word) which represents the old meaning has now come to represent a meaning encompassing the old activity and the new activity. This is no more or less than the language symbol now representing a new and expanded meaning. The process by which the symbols of our language gather and shed meaning, and whether new meanings are added to existing symbols or acquire new symbols of their own, is no doubt a complex social process, but not one related to a consistent reinterpretation of the Constitution that would track some group of coherent and rational legal principles.
It was pointed out at the beginning of this Article that many of the terms used in a statute or a constitution will refer to or denote matters or objects that did not exist at the time the text was created. The individual facts could not have been in the contemplation of the drafters. It was pointed out that in many instances this will not be a problem. Statutes are couched in general language and usually apply to all people or things coming into existence, including those born or created long after the Act was passed.
The reference in the Constitution to trade and commerce is properly and sensibly read as including contracts to buy and sell computers, spaceships, television sets, and microwave ovens, none of which existed and most of which were beyond the comprehension of the drafters of the Constitution. Nevertheless, there still needs to be a link with the beliefs and intentions of speakers of English in 1900 if these matters are to be resolved semantically rather than by reference to some interpretative rule which no longer depends upon meaning.
LEGAL INTERPRETATION AND SEMANTIC THEORY
Ambiguity, uncertainty, or meaninglessness may be resolved in normal dialogue, whether oral or written, by reformulations and restatements until the recipient is satisfied that he or she has understood what is being conveyed. What distinguishes problems of interpretation in law from so many other uses of language is that the very purpose for which the document or text was created prevents restatement or recapitulation. Whether a legal problem concerns a contract, deed, or other private instrument, or a statute, constitution, or international instrument, the document exists because people have bound themselves to have their legal rights and entitlements determined in the future by reference to an authoritative document. The document is appealed to just because one party does not wish the other to now seek to restate or reformulate what was originally communicated. The document is relied upon as the bulwark against one or another party resiling from a promise, agreement, or compact or seeking to now restate a version of some earlier agreement in a fashion calculated to give advantage.
These functions of legal texts also generate the often described difficulties where drafters must use general terms, yet seek to deal with whole categories of future occurrence, and where it is impossible to foresee all types of future occurrence or comprehend all of the implications that may be inferred from the language used in the document.
Despite these difficulties, it might still be thought that the only way an authoritative document can be used to resolve a legal dispute is to search for the relevant meaning of the document applicable to the dispute and apply that meaning. Many lawyers and legal philosophers treat the search for the meaning of an authoritative legal text as the sole means by which it can operate as a dispute solver. To the extent that there appear to be different modes of interpretation, it is sought to subsume these under the concept of meaning or to search within the competing theories of meaning for that one which represents the appropriate way of decoding the text. (41)
It is of course theoretically possible for texts to operate in many other ways. To be fanciful for a moment, but to help drive the point home, let me suggest that a text might operate along with an authoritative tribunal where the tribunal is obliged to read the text and to take it into account in resolving the dispute, but would not be expected to use it only by seeking to apply sentences of it in accordance with their meaning directly to the circumstances of the dispute. Such use of a document would be analogous to that which might be made by a religious adherent for whom a sacred text may be applied by being read as a source of inspiration but without the suggestion that it can be applied to the solution of some religious or moral dilemma by subsuming the problem or dispute under the terms of the text.
This use of texts where they are applied, not by seeking to read them as rules or instructions applicable directly to the circumstances of the dispute, I shall refer to as a non-semantic use of the text. This may perhaps be an unfortunate label since any use of the text will depend upon it having a meaning. What, however, is emphasised, is that the use will not be a direct application of the meaning of the text.
One can imagine ways in which an authoritative legal text is used to resolve disputes without applying its meaning in the conventional sense, for example, by using a recognised decision procedure whenever it lacks a meaning or has ambiguous meanings. Thus, if we apply the decision rule that in the case of any ambiguity or uncertainty in interpretation of a document we shall adopt that interpretation which will produce the best consequences, all things considered (an act-utilitarian decision rule), (42) then although we could describe the interpretation of the document arrived at in this fashion as a meaning of the document, our interpretation has not been arrived at by any semantic rule. We have not resolved the ambiguity by finding the meaning most probably intended by its author, or that meaning which would most probably be adopted by a skilled speaker of the document's language. From a semantic perspective there may be something arbitrary about such an interpretation; however, its virtue lies in its utility maximising consequences, not in its consistency with the rules of language or grammar.
Once rules of construction are looked on in this light, it can be seen that many rules of legal construction may be concerned with elucidating meaning, but not necessarily all. Some decision rules such as ejusdem generis might be thought to represent presumptions about meaning to be applied in cases of uncertainty or ambiguity. (43) On the other hand, a presumption of the sort that common law rights or liberties will not be revoked or interfered with without express words may be considered a presumption as regards meaning (the legislature must have intended not to interfere with such rights or it would have used express words), or it may be treated as a substantive rule protective of rights and liberties. (44) This latter justification of the rule would be supported on the basis that it is a protection of rights, which ought only to be abrogated by clear words.
For many jurists, the legislative task in construing a statute is giving effect to the legislative will as expressed through the words by which the legislature has formally spoken. From this perspective, the process of statutory construction is always and only an exercise in ascertaining the meaning which the legislature has expressed. This will be the case whether one concerns oneself with the actual intentions of those who promoted the bill that underlies the statute in question, or whether one considers that the legislative intention is to be ascertained objectively, or in other words, whether one views meaning as determined by the intentions of speakers or by the rules governing ordinary usage. (45)
Justice Scalia, a key proponent of textual originalism, recognises that, consistently with his view that the meaning of the United States Constitution is determined by the practices of language use governing English at the time the Constitution was written, there is therefore no room for rules of interpretation in ascertaining the meaning of the text:
Can we really just decree that we will interpret the laws that Congress passes to mean less or more than what they fairly say? I doubt it. (46)
The central importance of ascertaining the meaning of the text, and scepticism or disapproval of principles of construction that may involve departure from meaning, no doubt reflect the theory of legislation that views it as an expression of the will of the legislature, and views the legislature as having authority to exercise its will by virtue of the democratic sources of its political power. (47) Even in the case of a constitution, for which no people alive have voted, its authority might still be seen to derive from its adoption through a constitutional plebiscite and from the fact that the electors at large, when properly constituted, retain the power to amend or appeal it and have not seen fit to do so. (48)
Any attempt to consistently apply the principle that the text of a constitution or statute shall be determined by looking to the actual intentions of the drafters or legislators, however, has proven impossible to sustain in light of difficulties in knowing which individual's intentions are relevant, and because in regard to many legal problems, no individual in the lawmaking process held an actual intention on the issue that has arisen. Recourse to the hypothetical intention, or the intention the individual lawmakers would have had, had they considered the issue, proves more problematic than a straightforward appeal to the ordinary canons of language use. (49)
It is often because of the anomalies and difficulties of seeking to give effect to the actual intentions of the drafters or legislators that appeal is made to the principles of textualism. (50) In Gricean terms, textual meaning is often referred to as dictionary meaning or literal meaning, in contrast to speakers' meaning. Jurists who embrace textualism as the appropriate way to ascertain the meaning of a legal document need not abandon semantic theories that stress intentionality as an essential ingredient of meaning. As argued above, Grice saw dictionary or literal meaning as capable of being explicated in terms of speakers' meaning.
For a Court asked to apply the terms of a contract, justice and fairness dictate that the meaning of the terms be ascertained by reference to dictionary or literal meaning in the Gricean sense, it being fairer to hold a party to what the other side, to a reasonable observer of the bargain, would be thought to have agreed to, rather than to poorly expressed subjective intentions. However, even this principle may be displaced if the contracting party was aware that the other did not understand the words in their usual sense and sought to take unconscionable advantage of the mistake.
I shall proceed on the assumption that some version of textualism will be the standard theory in most instances of legal interpretation. Few appellate courts in the common law world presently espouse speakers' or utterers' intention theories of interpretation in either contractual, statutory, or constitutional contexts. A distinction between the two perspectives is important in those instances where textualism leads one to put aside an actual intention held by individual lawmakers. In dealing with problems caused by new social and legal problems not foreseen at the time a constitution or statute was drafted, the difference between textualism and the original intention theory of interpretation becomes less relevant. Those constitutional framers responsible for drafting the Australian Constitution probably did not have in mind, when drafting the external-affairs power, the possibility of international treaties concerned with environmental protection, or discrimination on the grounds of race and gender. Individual framers probably did not have an actual intention that marriage might ever be thought one day to include same-sex relationships. They certainly did not have an intention that computing technology would be included or excluded from the copyright power. However, none of these problems are resolved by moving from the level of speakers' or utterers' meaning to textualism.
If one substitutes for the intentions of those who drafted the Constitution the meanings that would be attributed to the words of the Constitution by competent English speakers in Australia in 1900, one is no further advanced in resolving those interpretative difficulties created by historical and cultural change since the time the document was drafted. Properly understood, textualism as a theory of meaning will simply cause us to substitute for the intentions and beliefs of the actual constitutional drafters, the intentions and beliefs about words of competent speakers of English in late nineteenth century Australia. No doubt the pool of beliefs that might be held about the meaning of words amongst competent speakers of English in 1900 was more extensive than the actual beliefs of Sir Samuel Griffith (51) or Inglis Clark, (52) but nobody in 1900 had any beliefs about computing technology, and no one used the word "copyright" in connection with computer programs.
One could seek to escape the dilemma just described by seeking to attribute a hypothetical intention to a competent English speaker in 1900. Thus, one might ask whether a normal English speaker in 1900 would, if presented with an appropriate package of information about the nature of computers, be willing to treat computer programs as the sort of thing in regard to which one could make a copyright claim. However, one of the attractions of textualism is that it seems to avoid the difficulties and paradoxes that arise whenever one has to resort to a hypothetical speaker's intention. Just how does one take a hypothetical nineteenth-century speaker of English and present him with an appropriate package of information? What knowledge would such a hypothetical speaker have to have attributed to him to provide us with an answer? Does one assume that the hypothetical speaker of 1900 otherwise brings to bear upon the problem all the other values and beliefs of an Australian in 1900? In which case, the answer one derives will be anachronistic and probably incoherent; any intelligible view about control of computing technology requires some knowledge about the function of computers in contemporary life and their economic and technical operation. On the other hand, if one attributes to one's 1900 speaker of English the knowledge and beliefs of a twenty-first century Australian, then firstly, it makes a mockery of any claim that one is applying the canons of language use of Australians who lived over one hundred years ago, and secondly, the exercise is so filled with artificiality and hypothesis that it is clear in any event that the answer one arrives at will be dictated by the inferences the Court chooses to draw and will not be constrained by any historical linguistic facts.
Ronald Dworkin has used the phrase "semantic originalism" to describe the views of jurists such as Justice Scalia, who maintain that texts such as the United States Constitution should be read and interpreted in accordance with the meaning that they had at the time that they were written and that the meaning must be ascertained by applying the canons of language use of American English from that time. Dworkin argues for a more flexible mode of interpretation. He acknowledges that this approach would involve departure from speakers' meaning, but in labelling his view "expectation originalism," he maintains that his view is faithful to what the constitutional drafters would have expected to occur as succeeding generations sought to apply the Constitution to new problems. (53)
Dworkin claims to be an originalist, but he separates his views from textualists, or those he calls semantic originalists, by arguing that those abstract terms within the Constitution (particularly those rights amendments in the United States Constitution) are terms that need to be interpreted afresh in each historical epoch to deal with historical changes. Dworkin does not use a connotation/denotation distinction to show how these fresh interpretations may be unpacked from the abstract or general terms of the Constitution. In arguing that the general and abstract terms of the Constitution are statements of moral principle which may be interpreted to cover circumstances unforeseen by the constitutional framers, Dworkin's view is a non-semantic approach in the sense described above.
The approach of Dworkin, in common with those Australian jurists who apply a textualist approach to interpreting the Constitution, is to have the best of both worlds. Purporting on the one hand to be originalists for whom the interpretation of the Constitution is said to be grounded in the text created by the framers, on the other hand they seek to avoid some of the implications of rigorous semantic originalism. (54) In Australia this has been done by use of the connotation/denotation distinction, which may be thought to give room for manoeuvre without abandoning appeal to original meaning.
Theoretically, there is nothing to prevent semantic originalism being adopted as the sole canon for interpreting the Constitution. Such an approach can be properly justified by an appeal to semantic theory, and it represents a consistent and intelligible view concerning the interpretation of the constitutional text. However it is doubtful if such an approach has ever been, or could be, rigorously adhered to. Once it is acknowledged that the meanings conveyed by the original constitutional text are bound up with the shared intentions of the community of language users that gave rise to the document, it must be recognised that there is no scope for a changing or evolving denotation or reference that could leap ahead of those shared intentions. The meaning of the Constitution is therefore governed by the knowledge, beliefs, and social values of the community of language users at the time it was written. Its meaning will consequently be narrow, and its terms will not encompass new social practices and technologies that were unforeseen at the time it was written.
Apart from the difficulty semantic originalism poses once it is recognised that there is no way one can continue to adapt the Constitution to new and unforeseen circumstances without altering or expanding the meaning of its terms, there is a further problem, that of knowing or ascertaining what was the original meaning of the terms.
Because of the problem of linguistic drift, although many words (55) in modern Australian English have meanings which overlap with the meaning those words connoted in 1900, it is also likely that many of these terms (for example, trade and commerce, external affairs) have had their meanings expanded or displaced. To now simply apply those terms as terms of modern English would be to abandon semantic originalism. To be a consistent semantic originalist would require complex and sophisticated linguistic analysis to determine and trace the displacement and change of meaning in these terms.
Some might consider the difficulties just described as being exaggerated, that the level of social change and meaning displacement is capable of being handled by interpreters and that the only proper way of updating the Constitution remains the method of amendment, in accordance with section 128 (if one is concerned with the Australian Constitution). Some might criticise the view on broad political grounds, but it could not be criticised for misunderstanding the nature of language or meaning. Provided interpreters avowing semantic originalism are content to bear the burdens that such a theory imposes, it should be acknowledged that they are striving to apply the meaning of the Constitution.
NON-SEMANTIC THEORIES OF INTERPRETATION
Not everyone will be prepared to accept the limitations imposed by semantic originalism. In the United States, Ronald Dworkin has been perhaps the best known critic of the view, but there have been many other jurists who have suggested that the Constitution may have an evolving interpretation. This Article will not seek to review the many theories that have been propounded in justification of a "living constitution," or changing or evolving interpretation. (56) Rather, this Article only seeks to make the point in regard to these theories, that they are non-semantic theories in the sense described above. They do not apply the meaning of the Constitution to the solution of legal disputes. If they do not apply the meaning of the Constitution to legal disputes, then if they are to be justified interpretations of the Constitution, they require an appropriate theory justifying their non-semantic modes of interpretation.
Hopefully, enough has been already said regarding the determinative role, in regard to the meaning of words, of the shared intentions of the community of language users, to now permit it to be said without further argument that the meaning of a text cannot simply shift with the passage of history in order to encompass new and unforeseen circumstances. Dworkin's theory of expectation originalism disclaims an attempt to apply the original meaning of the Constitution, but says that the drafters must themselves have envisaged that a process of reinterpretation to deal with new and unforeseen circumstances was inevitable, and provided later interpreters keep faith with the central and underlying principles of the Constitution, its revised meanings will be proper interpretations. (57)
Dworkin criticises the argument that the meaning of the Constitution can be confined by the intentions of some presumed set of framers or drafters. Those framers may have believed that the provisions they drafted about equality and due process had legal implications in concrete cases different from those you or I would now draw from a contextual application of their words, but Dworkin argues that it does not follow they meant to say anything different from what you or I would have meant to say if we had used the same words. This argument is ultimately unworkable. While Dworkin does not wish to abandon the notions that language is intentional and that the framers' intentions in some fashion must count, they now count for so little that I can come to a different conclusion from them despite apparently applying the same words. This will not be an application of the meaning of the words. This may be merely a result of linguistic drift, in which case the words in the Constitution have now acquired a new meaning. Alternatively, it may be that through a process of interpretation, a new meaning has been attributed to the words of the Constitution.
Dworkin believes that his theory is strong enough to demonstrate the existence of an objectively correct answer at any particular time as to what the Constitution means, albeit it may mean different things at different times. (58) Stanley Fish has argued persuasively that Dworkin's theory cannot establish that the constraints on interpretation are in the text; rather, interpreters are constrained by the practices and understandings shared by the community within which they function. Dworkinian principles or moral readings of the Constitution only operate as an external constraint upon the possible interpretations that may be read into the document to the extent that one maintains there is an objectively correct public morality. (9)
It has already been argued that one of the principal purposes of a legal text is to constrain citizens and governments and create a framework for permissible conduct. However, it is not just any constraint, or constraint for its own sake, that the Constitution seeks to lay down, but specific constraints defined by the meanings contained within the Constitution. Any attempt to sever the Constitution's effect in later times from the meaning possessed by the text at the time it was written also severs the purpose and function served by the Constitution from that which it had at the time of Federation.
The significance of this might best be understood by contemplating the effects of the adoption in Australia of a bill of rights with similar force to a constitutional provision. Such a bill could and would only be adopted at the present time because a substantial portion of the community believes that certain specific constraints on governmental action and certain legal effects so far as the rights of citizens are concerned should be guaranteed or underwritten by an authoritative text. However, no sooner will such a text be agreed upon and adopted, than the problems of linguistic drift and unforeseen social change will start to afflict it. Unless its interpreters all espouse semantic originalism, the bill of rights will inevitably come to guarantee or constrain conduct in ways never intended or foreseen by those who promoted it. This ought not to be seen as necessarily an argument against adopting such a bill of rights--proponents might consider this an acceptable price, although given the uncertainties, a price difficult to measure.
If one is not to be driven back to semantic originalism, then legal theory needs to propound a basis that can justify an authoritative tribunal, such as the High Court of Australia, regularly replacing the meaning attributed to the Constitution with a fresh meaning. This theory would need to have two important characteristics. Firstly, it would need to show how at least ideally, there is a preferred solution. A theory so weak that it merely invited judges to interpret the document as best they thought would suit the contingencies of the day would be tantamount to simply conferring upon High Court judges a general discretion to resolve governmental and political disputes in accordance with the dictates of their political and moral values. A constitution open to reinterpretation in a fashion which is essentially unconstrained can no longer serve the purpose of a compact between the arms of government and the citizens of the State.
No doubt there are possible decision rules for reinterpreting the Constitution that would produce determinant results. On the other hand, these rules may be deficient for other reasons. A decision rule which requires the Constitution to be read as if it were a text of modern English would produce as determinate a result as semantic originalism. On the other hand, it would not produce a consistent and rational reinterpretation of the text.
Proponents of living constitutions might desire a mode of interpretation that will permit new constitutional rules to be fashioned reflecting the solutions a wise and rational legislature would adopt, but it is a tall order to expect that a theory of interpretation can perform that function.
The second matter that any adequate theory of interpretation should provide is a political justification of the function that will be performed by the authoritative tribunal vested with the power to construe the Constitution. Doctrines emphasising the role of courts as declarers of law will be inadequate for this purpose. An emerging justification for such a role may be based simply on its inevitability, perhaps coupled with maxims of restraint and a ukase to decide issues narrowly, and only if deciding them is unavoidable, thus seeking to minimise the interpretative task and protect courts against the accusation that they have seized a role that more properly belongs with the legislature or with the electors gathered in a constitutional plebiscite. However, it is not obvious why this task, once undertaken, should adopt a conservative default rule rather than any other.
Neither constitutional interpreters nor legal theorists have devised a theory sufficiently strong to meet the two principal requirements just described. Constitutional interpretation thus presently dwells in an uncomfortable zone. Few jurists are prepared to fully embrace the rigors of semantic originalism, while they lack a comprehensive justification for alternative practices. Regrettably, current legal theory does not appear to have a ready solution to these dilemmas. One possible answer is the pessimistic one, that no solution to these dilemmas exists. This could be because we expect of constitutions things they cannot deliver. If the Constitution is to operate as a constraint that is determinant and predictable, not open to infinite future variation with new interpretation, then one will be driven towards semantic originalism with the difficulties already highlighted. The removal of those difficulties may not be reconcilable with the purpose of a constitution, or with the political theories which seek to justify it and the governmental and judicial institutions created under it.
* Based upon an article previously published by the author under the title JS Mill, Gottlob Frege, and the High Court: The Connotation/Denotation Distinction in Constitutional Interpretation, 23 Australian Bar Rev. 296 (2003).
(1.) (1935) 54 C.L.R. 262 (Austl.).
(2.) Australia is governed by a Federal system in which there are six States and a central government referred to in the Constitution as the Commonwealth. The Federal system was formed after campaigns in the late 19th Century to federate the original British colonies and provide them with independence. The campaign culminated in the adoption of the Australian Constitution in 1901 following a plebiscite held in each of the colonies. The Constitution is a written document of some 128 sections. Although in setting up a Federal structure it shares some similarities with the American Constitution the system of government provided for under the Constitution is a Westminster-style government.
(3.) Justice Dixon nevertheless expressed doubt that broadcasting was a like service merely because of its use of similar mechanical or electromagnetic technology and thought services like telegraphy or telephony were like if they were means for transmitting messages from one person to another rather than broadcasting at large. Ironically, applying that meaning to the words, radio and television broadcasting would not be like services, but email presumably would. See The King v. Brislan; ex parte Williams (1935) 54 C.L.R. 262, 289-90 (Austl.) (Dixon, J., dissenting).
(4.) Commonwealth v. Tasmania (the Tasmanian Dam Case) (1983) 158 C.L.R. 1 (Austl.).
(5.) (2000) 203 C.L.R. 1 (Austl.).
(6.) Id. at 49 (opinion of McHugh, J.); but see id. at 85 (Kirby, J., dissenting) ("Even on the assumption that the content of the appellate jurisdiction of this Court is to be decided by reference to what 'appeals' meant in 1900 (which I would not accept) there are many indications that it is factually incorrect to suggest, as a universal rule, that the notion of 'appeals' excluded absolutely the reception of new evidence.").
(7.) Jill McKeogh & Andrew Stewart, Intellectual Property in Australia [paragraph] 9.8 (2d ed., Butterworths 1997).
(8.) In Grain Pool of Western Australia v. Commonwealth (2000) 202 C.L.R. 479, the High Court upheld the constitutional validity of the Plant Variety Rights Act 1987 (Commonwealth) and the Plant Breeders Rights Act 1994 (Commonwealth). It follows from what is argued below that this should be treated as a non-semantic interpretation of the Constitution. The same point is made in regard to the Court's conclusion in Nintendo Co. Ltd. v. Sentronics Systems Proprietary Ltd. (1994) 181 C.L.R. 134, regarding the Circuit Layouts Act 1909 (Commonwealth).
(9.) See the entry "Connotation and Denotation" by Gim Del Villar in the Oxford Companion to the High Court of Australia 135 (Tony Blackshield et al., eds., Oxford U. Press 2001).
(10.) Gottlob Frege, On Sense and Reference, in Translations from the Philosophical Writings of Gottlob Frege 56 (Peter Geach & Max Black, eds., Blackwell 1977).
(11.) An example of such recent criticism is found in Michael Beaney, Frege, Making Sense 257ff. (Duckworth 1996). For further references to commentary on Frege's notions of sense and reference, see note 33, infra.
(12.) (1959) 107 C.L.R. 208 (Austl.).
(13.) Id. at 267.
(14.) (1989) 168 C.L.R. 461 (Austl.).
(15.) Id. at 537 (opinion of Dawson, J.).
(16.) (1999) 198 C.L.R. 511 (Austl.).
(17.) Id. at 552.
(18.) Leslie Zines, The High Court and the Constitution 17 (4th ed., Butterworths 1997).
(19.) See Re Wakim, 198 C.L.R. at 552:
Philosophers are now said to regard the distinction between connotation and denotation as outdated.... But whether criticism of the distinction is or is not valid should not be seen as decisive. What is decisive is that, with perhaps only two exceptions, the Court has never hesitated to apply particular words and phrases to facts and circumstances that were or may have been outside the contemplation of the makers of the Constitution.... [O]nce we have identified the concepts, express and implied, that the makers of our Constitution intended to apply, we can give effect to the present day conceptions of those concepts.
(20.) See Eastman, 203 C.L.R. at 80 (speaking of the interpretation of constitutional language according to present-day meaning, Justice Kirby contested the "distinction between the connotation and denotation of verbal meaning").
(21.) (2001) 207 C.L.R. 391 (Austl.).
(22.) (1988) 165 C.L.R. 178 (Austl.).
(23.) See Patterson, 207 C.L.R. at 410 [paragraph] 44 (opinion of Gaudron, J.) & 426 [paragraph] 108 (opinion of McHugh, J.).
(24.) (1999) 199 C.L.R. 462 (Austl.).
(25.) Patterson, 207 C.L.R. at 427 [paragraph] 110.
(26.) John Stuart Mill, A System of Logic, Ratiocinative and Inductive (Longmans, Green & Co. 1898) (first published 1843).
(27.) Id. at 20.
(28.) Richard Whately, Elements of Logic (J. Mawman 1826).
(29.) John Passmore, A Hundred Years of Philosophy 21 (Penguin Books 1968).
(30.) Mill, supra n. 26, at 25.
(31.) Id. ch. 2, [section] 7.
(32.) H.P. Grice, Meaning, 66 Phil. Rev. 377 (1957).
(33.) H.P. Grice's theory is further developed in Utterers' Meaning and Intention, 78 Phil. Rev. 147 (1969) and Meaning Revisited, in Mutual Knowledge (N.V. Smith, ed., Academic Press 1982).
(34.) Mark de Bretton Platts, Ways of Meaning 92 (2d ed., MIT Press 1997). Truth-conditional semantics was first devised by Donald Davidson. The critical essays are collected in Donald Davidson, Inquiries into Truth and Interpretation (Oxford U. Press 1984). The current preference for semantic theory to rely upon the psychological states of speakers is not only the preference of the author, see the entry by Tim Crane, Meaning, in The Oxford Companion to Philosophy 542 (Ted Honderich, ed., Oxford U. Press 1995).
(35.) Further there is the point made by Professor Goldsworthy that general terms by their nature are such that the objects which fall under them at one time may not fall under them at another time. Hence, the phrase "legal tender" in 1965 included pounds, shillings and pence but not in 1967. Plainly, however this is still governed by the connotation of the term, which in turn is governed in the fashion described above. See Jeffrey Goldsworthy, Originalism in Constitutional Interpretation, 25 Fed. L. Rev. 1, 42 (1997) (publication of the Australian National University).
(36.) See Wakim, 198 C.L.R. at 552 [paragraph] 44.
(37.) See Frege, supra n. 10.
(38.) Key developments from Frege's work were Bertrand Russell's theory that proper names are definite descriptions, explained in his essay On Denoting, in Logic and Knowledge, Essays of Bertrand Russell 39 (Robert Charles Marsh, ed., George Allen & Unwin Ltd. 1956), On the other hand, Saul Kripke rejected the Fregean approach to proper names in his causal theory of reference. See Saul A. Kripke, Naming and Necessity (Blackwell Publishers 1972). Michael Beaney, see supra n. 11, notes the attempts to apply Frege's theory to concept words and functional expressions and further notes that posthumously published papers of Frege suggested he likewise intended it to apply, However, Frege's work has remained underdeveloped in this application.
(39.) In this sentence, the reference to "words" is made to them as signs or symbols, not the concepts or meanings represented by those signs or symbols
(40.) Surprisingly, it has been suggested in some cases that statutes should be interpreted in accordance with the current meaning that would be attributed to their terms even if that would not be the meaning they had when they were enacted. This method even has a label, the "always speaking approach." See D.C. Pearce & R.S. Geddes, Statutory Interpretation in Australia 93ff (5th ed., Butterworths 2001), discussed infra n. 43. Although Pearce and Geddes suggest this approach shows "obvious good sense," it is difficult to see why the new or altered meaning picked up by the terms used in a statute as a result of linguistic drift should be given effect. It is hard to imagine a justification of the always speaking approach that would satisfy the principles discussed in the last section of this Article.
(41.) See e.g. Andrei Marmor, Interpretation and Legal Theory, 28ff (Oxford U. Press 1992), (suggesting that interpretation is a search for meaning, although acknowledging the different senses of meaning).
(42.) See H.L.A. Hart, The Concept of Law 135 (2d ed., Clarendon Press 1994), and Neil MacCormick, Legal Reasoning and Legal Theory 129ff (Clarendon Press 1978), who both suggest the use of such a rule in limited circumstances.
(43.) Pearce & Geddes, supra n. 40, at 103.
(44.) Id. at 131ff.
(45.) See Larry Alexander, All or Nothing at All? The Intentions of Authorities and the Authority of Intentions, in Law and Interpretation 357, 360ff (Andrei Marmor, ed., Clarendon Press 1995), for the view that legal texts are attempts by authorities to communicate their determinations of what ought to be done, and the compendious list of those jurists holding this view, id. at 361 n. 11; see also Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 37ff (Princeton U. Press 1997), for the view, expressed by one of its leading exponents, that the meaning of a statute or constitution is the original meaning of the document ascertained objectively, namely, by the usages of competent speakers of the language at the time it was written.
(46.) Scalia, supra n. 45, at 29.
(47.) Alexander, supra n. 45, at 361.
(48.) G.J. Lindell; Why is Australia's Constitution Binding?--The Reasons in 1900 and Now, and the Effect of Independence, 16 Fed. L. Rev. 29 (1986).
(49.) See Ronald Dworkin, Law's Empire 325 (Fontana 1986); Michael Stokes, Constitutional Commitments Not Original Intentions: Interpretation in the Freedom of Speech Cases, 16 Sydney L. Rev. 250 (1994); David Lyons, Original Intent and Legal Interpretation, 24 Australian J. Leg. Phil. 1 (1999).
(50.) I shall use textualism as a label for the interpretative principle that a text should be read in accordance with its meaning ascertained from the linguistic customs and practices of the speakers of the language in which the text was written at the time in which it was written. I understand this to be what Scalia means by textualism. For a recent statement of a textualist view of interpretation from Australia, see Wilson v. Anderson (2002) 190 A.L.R. 313, 316 (Austl.) (opinion of Gleeson, C.J.).
(51.) Sir Samuel Griffith (1845-1920) was one of the principal drafters of the Australian Constitution and a campaigner for Federation. He became Chief Justice of Queensland in 1893, and in 1903, was appointed the first Chief Justice of the High Court of Australia.
(52.) Inglis Clark (1848-1907) was a constitutional lawyer and campaigner for Federation. He prepared an early complete draft for the Constitution for consideration at the 1891 Constitutional Convention. Clark was strongly influenced by the United States model, having corresponded with Oliver Wendell Holmes, Jr. His Studies in Australian Constitutional Law, first published in 1901, has been an influential text.
(53.) The distinction is made by Dworkin in his comment in reply to Scalia published in Scalia's book. See Scalia, supra n. 45, at 119. Dworkin makes a similar point in several other places. See e.g. Dworkin, Law's Empire, supra n. 49 (chs. 9, 10); Ronald Dworkin, Life's Dominion 118ff (Harper 1993); Ronald Dworkin, Freedom's Law (Oxford U. Press 1996) (particularly the Introduction).
(54.) One of the commonest criticisms of Justice Scalia's semantic originalism is that he fails to follow it faithfully in his own judicial work. See R Dworkin's comments, in Scalia, A Matter of Interpretation, supra n. 45, at 120; see also Jeffrey Rosen, Originalist Sin, in New Republic 26 (May 5, 1997).
(55.) The term "words" is again now used to indicate the signs or symbols used by speakers to represent meanings.
(56.) For a contemporary Australian example, see Hon. Justice Michael Kirby, Constitutional Interpretation and Original Intent: A Form of Ancestor Worship? 24 Melbourne U. L. Rev. 1, 11 (2000).
(57.) Dworkin, Freedom's Law, supra n. 53, at 75-76,
(58.) Dworkin, Law's Empire, supra n. 49, at 364ff.
(59.) Stanley Fish, Doing What Comes Naturally (Clarendon Press 1989) (particularly ch. 4, at 99, and chs. 5, 16); see also Stanley Fish, There Is No Such Thing As Free Speech and It's A Good Thing, Too (Oxford U. Press 1994) (especially chs. 11, 13). John Rawls, with his concept of public reason expounded, and Cass Sunstein, with his notion of untheorised disagreement, highlight the intractability of many moral disagreements and the importance of judicial decisions being justified without recourse to the global moral views of the adjudicator. See John Rawls, Political Liberalism (Columbia U. Press 1996); Cass Sunstein, Legal Reasoning and Political Conflict (Oxford U. Press 1996).
Christopher Birch **
** Christopher Birch SC BA LLB PhD is a barrister practising at the Sydney Bar and lecturer in legal philosophy of the University of Sydney.