Some problems with the shared meaning rule as formulated in R. v. Daoust and 'The Law of Bilingual Interpretation'.This article offers a critical analysis of the account of the shared meaning rule adopted by the Supreme Court of Canada in R v Daoust and subsequently defended by the authors of The Law of Bilingual Interpretation, including retired Justice Michel Bastarache who wrote the unanimous judgment of the Court in Daoust. The author argues that, contrary to the position taken in Daoust, the presumption of shared meaning should not arise when one language version of a legislative text is clear but broad while the other is clear but narrow. She suggests that relying on the presumption in these circumstances focuses too narrowly on the norms of linguistic equality and the protection of minority language rights while inappropriately neglecting the norms of rule of law and linguistic security. The position taken in Daoust can only be justified by the textualist assumption that the literal meaning of a text -and by extension the shared literal meaning of the text--can be equated with legislative intent. This textualist assumption is inconsistent with the modern principle of interpretation that has been repeatedly endorsed by the Supreme Court of Canada. She endorses and justifies an alternative formulation of the shared meaning rule.
Cet article presente une analyse critique du principe du sens commun, adopte par la Cour supreme du Canada dans l'arret R c Daoust, que les auteurs de Le droit de l'interpretation bilingue ont par la suite defendu, notamment le juge a la retraite Michel Bastarache, qui a redige la decision unanime de la Cour dans Daoust. L'auteure soutient que, contrairement a la position adoptee dans Daoust, la presomption du sens commun ne devrait pas etre invoquee lorsque la version dans une langue d'un texte legislatif est claire mais large alors que la version dans l'autre langue est claire mais plus restreinte. Elle est d'avis que si l'on se fie a la presomption dans ces circonstances, on se concentre trop sur les normes regissant l'egalite linguistique et la protection des droits des minorites linguistiques au detriment du principe de la primaute du droit et des normes de la securite linguistique. La position degagee dans l'arret Daoust ne peut se justifier qu'en vertu du postulat de nature textuelle selon lequel la signification litterale d'un texte--et par extension le sens litteral commun du texte--peut equivaloir a l'intention du legislateur. Ce postulat va a l'encontre du principe moderne de l'interpretation que la Cour supreme du Canada a maintes fois enterine. L'auteure avalise et justifie une formulation de rechange a la regle du sens commun.
Table of Contents I. INTRODUCTION II. THE RELEVANT CONSTITUTIONAL NORMS A. Equal Authenticity and the Shared Meaning Rule B. The Steps to be Followed in Interpreting Bilingual Legislation III. WHEN DOES THE presumption arise? IV. THE GOVERNING PRINCIPLE OF STATUTORY INTERPRETATION V. JUSTIFYING THE SHARED MEANING RULE A. Norm Based Justification B. Textualist Justification VI. The Daoust Case A. The Strong Version Analysis B. The Weak Version Analysis VII. POST-DAOUST LAW VIII. CONCLUSION
In recent years the Supreme Court of Canada has had a lot to say about the interpretation of bilingual legislation, culminating in its judgment in R v Daoust, (1) written by Justice Bastarache on behalf of a unanimous panel. In 2008, Justice Bastarache along with others published a book entitled The Law of Bilingual Interpretation. (2) Almost half the book is devoted to surveying the evolution of the shared meaning rule as it applies to bilingual legislation in Canada and explaining the conception of the rule adopted in Daoust. The remainder of the book deals with the challenges of interpreting bijural legislation, bilingual case law, legislation drafted in Aboriginal languages and multilingual treaties. The Law of Bilingual Interpretation packs a great deal of information and analysis into relatively few pages.
In this paper, I have two goals. The first is to set out the understanding of the shared meaning rule adopted in Daoust and the book and to explain why I find it unsatisfactory. My first objection to the account of the rule in Daoust is that it effectively introduces a new strict construction rule to statutory interpretation without offering satisfactory justification for this innovation. My second objection is that it depends on a textualist approach to interpretation that is inconsistent with Driedger's modern principle, as explained and applied by the Supreme Court of Canada in its judgments of the past IS years. Of course, many of those judgments were written by Justice Bastarache himself and are not particularly textualist in their approach. But in the judgments addressing bilingual interpretation, the textualist bent is unmistakable. In these judgments, and in the book as well, it appears that respect for textual meaning is equated not only with legislative intent--the usual textualist assumption-but also with respect for linguistic equality and minority language rights. In my view, these equations do not stand up to a careful examination.
My second goal in this paper is to offer an account of the shared meaning rule that is grounded in a broader set of constitutional norms than those relied on by the authors of The Law of Bilingual Interpretation. No one challenges the fundamental assumption that if the two versions of bilingual legislation are authentic, rule of law requires that both express the same rule. It is unthinkable for there to be one law for English speakers and another for French speakers. And everyone agrees that interpreters must give effect to the intentions of the legislature, for this remains the official goal of statutory interpretation, grounded in the norm of legislative sovereignty. After that, however, views diverge. The account of the shared meaning rule offered in the book is grounded almost exclusively in the values of linguistic equality and the protection of minority language rights. The account offered in this paper emphasizes the rule of law, fairness and linguistic security in addition to linguistic equality and the protection of minorities.
Part II of the paper reviews the constitutional underpinnings of the equal authenticity rule and the problem of discrepancy between the versions that gives rise to the shared meaning rule. Linguistic equality was and remains an important ideal of confederation, as evidenced by section 133 of the Constitution Act, 1867, (3) section 23 of the Manitoba Act, 1870 (4) and sections 16 to 23 of the Canadian Charter of Rights and Freedoms. (5) Realization of linguistic equality in the enactment and publication of legislation is subject to many impediments, of which the most intractable is the inevitable discrepancy (or perceived discrepancy) between the two language versions. The shared meaning rule is an attempt to address this impediment. Part II sets out the steps an interpreter must go through in applying the shared meaning rule.
Part III then addresses two important questions in bilingual interpretation. When there is a discrepancy between the two language versions of a legislative text and it is possible to identify a shared meaning, should the presumption in favour of shared meaning be triggered in every case? If not, how should instances in which the presumption is triggered be distinguished from instances in which it is not? In this part, I contrast the book's expansive understanding of the presumption--which I call the "strong version"--with my own understanding--the "weak version."
Part IV steps back from the specific issue of bilingual interpretation and briefly sets out what in my view judges do--and should do--when they apply Driedger's modern principle to interpret legislation. My purpose here is to point out that the modern principle is inconsistent with textualism.
Part V compares the justifications underlying the weak and strong versions of the shared meaning rule, and points out the textualist underpinnings of the strong version. Like textualism, the strong version identifies textual meaning with intended meaning;. However, while it is plausible to suppose that the clear meaning, of a unilingual text expresses the rule that the legislature intended to enact, it is not plausible to suppose that the narrower version of a bilingual text expresses this intention. Furthermore, in the case of a clear unilingual text, it is plausible to suppose that the public will rely upon the clear meaning as an accurate expression of the law, so that adopting it protects the subject from surprise. However, in the case of bdingual legislation, when one text clearly expresses a narrow rule and the other text clearly expresses a broad one, adopting the shared meaning does not protect unilingual readers from surprise. If the narrow version prevails, those who relied on the broad version are necessarily surprised.
Part VI reviews the facts and reasoning in the Daoust case and suggests that the case was rightly decided not because the court gave effect to the shared meaning, but because of rule of law concerns. When dealing with penal legislation, strict construction--a presumption in favour of the narrow version is appropriate. It protects the subject from unfair surprise. However, in the case of benefitconferring or quasi-constitutional legislation, to automatically prefer the narrow version serves no one's interest.
Part VII looks at bilingual interpretation in the aftermath of Daoust. While several judgments of the Supreme Court of Canada have approved and applied Daoust, the majority judgment in Canada (Citizenship and Immigration) v Khosa (6) suggests that not all members of the Court are committed to the strong version of the presumption. Furthermore, in recent writing, Pierre-Andre Cote himself has rejected the strong version. It is likely that Daoust will not prove to be the final word on the law governing the interpretation of bdingual legislation.
H. THE RELEVANT CONSTITUTIONAL NORMS
One of the strengths of The Law of Bilingual Interpretation is its recognition of the important connection between the rules applied by courts to resolve interpretation disputes and the political values and ideological commitments underlying those rules. The following passage is from the introduction to The Law of Bilingual Interpretation:
It is obvious that the law governing languages embodies an ideological and political position. In Canada, language rights are designed to favour national unity and respect for minorities. The necessity of developing principles of interpretation respectful of the two official languages and two legal systems is therefore a constitutional imperative. It has to do with the assurance of minimum conditions permitting the official language minorities to participate fully in public affairs, a component of the constitutional principle of protection of minorities. The rules of interpretation applicable to bilingual legislation are thus an important part of the legal framework created to implement the constitutional principles just mentioned and realize the moral objectives of the nation. They constitute some of the tools permitting minorities to consolidate their identity and affirm their dignity as citizens of Canada. (7)
This paragraph is a welcome recognition that legal rules, including interpretation rules, are grounded in ideological and political positions. Once an ideological and political position is sufficiently recognized by law, particularly constitutional law, it becomes a legal norm and, as I will argue below, plays a legitimate role in the creation, enactment and interpretation of legislation. The most accepted rule of statutory interpretation, that courts should give effect to the intention of Parliament, is grounded in the norm of legislative sovereignty. This norm is obviously important but it is never the only norm applicable in an interpretation dispute. One of the themes I hope to develop in this paper is that the scope and weight to be attributed to a rule of interpretation in a given case is not only a function of the importance of the norm in which it is grounded, but also of its relevance to the dispute to be resolved by the court.
A. Equal Authenticity and the Shared Meaning Rule
The Law of Bilingual Interpretation identifies two principles of interpretation that are grounded in respect for the two official languages recognized by Canada's Constitution. The first is the equal authenticity rule; the second is the presumption in favour of shared meaning, which is conventionally referred to as the shared meaning rule. (8)
The equal authenticity rule provides that both versions of bilingual legislation are equally official and authoritative expressions of the law enacted by the legislature. (9) This rule is a necessary condition of linguistic equality and the protection of minority language rights. Publishing the law in the minority language does not achieve equality if that language version has a lesser status than the majority language version and cannot be relied on as an official statement of the law. Linguistic equality requires that both versions be authentic.
By declaring both language versions of legislation to be authentic, the legislature holds both versions out as an accurate statements of the law on which subjects may safely rely. In so far as both versions live up to this promise of reliability, Frenchand English-speaking subjects have secure access to the law in their own language, in keeping with the rule of law. Both groups know in advance what the legal consequences of their conduct will be so that no one is unfairly taken by surprise. Unilingual speakers of French or English can live in their own linguistic community and enjoy the full benefits of citizenship without having to master the other official language. In making law and applying law, and to a lesser extent in offering governmental services, the state is obliged to be bilingual so that individual Canadians can successfully function unilingually. (10)
In principle, equal authenticity is a complete solution to the problem of accommodating French- and English-speaking subjects within a single political entity. It respects and protects both language communities, and it provides the members of both groups with direct and secure access to the law.
The fly in the ointment is that equal authenticity can fulfill its promise only if both language versions actually do state the same law--and this is a condition that can never be fully realized. In practice, apart from the inherent divergences between any two languages, mistakes in drafting or translation inevitably occur. Given that discrepancies between versions are unavoidable, the best the law can do is to minimize their impairment of equal authenticity and its underlying norms.
When it comes to respect for linguistic equality, minimal impairment is easily achieved. So long as the method of resolving discrepancy does not involve an automatic preference for one language over the other and both versions have an equal chance of being identified as the version that correctly states the law, the norms of linguistic equality and the protection of minority language rights are respected. When it comes to respect for the rule of law and linguistic security, however, the challenge is much greater. To the extent a subject reasonably relies on a version that turns out to be an inaccurate statement of the law, the rule of law and, in many cases, fairness will be undermined. To the extent a subject is forced to read both versions to ensure an accurate understanding of the law, linguistic security is undermined. What is called for is a response to discrepancy that, as much as possible, permits subjects to rely on their own language version (in accordance with the norm of linguistic security) without being taken by surprise (in accordance with the norms of the rule of law and fairness).
The primary tool developed by Canadian courts to respond to discrepancies between the two versions of bilingual legislation is the shared meaning rule. As formulated in The Law of Bilingual Interpretation, the shared meaning rule states that courts should, as far as possible, ascertain legislative intent by determining the shared meaning of the two versions. (11) This formulation is based on two distinct assumptions. First, in so far as a shared meaning can be identified, that meaning should be presumed to be the meaning intended by the legislature. Second, interpreters should strive to identify a shared meaning in the two versions of bilingual legislation. In my view, these assumptions overstate both the importance and the relevance of shared meaning. I would suggest that shared meaning should be presumed only in so far as doing so, as opposed to not doing so, can be shown to advance the relevant constitutional norms.
B. The Steps to be Followed in Interpreting Bdingual Legislation
Some years ago, in The Interpretation of Legislation in Canada, (12) Pierre-Andre Cote described the sequence of steps that courts follow in applying the shared meaning rule. His account was approved and adopted by the Supreme Court of Canada in the Daoust case. (13) It was adopted and further elaborated in The Law of Bilingual Interpretation. (14) Basically, it consists of three steps. First, the interpreter must read both versions of the provision to be applied (15) and determine whether there is a discrepancy between the versions. Second, if there is a discrepancy, the interpreter must determine whether there is nonetheless a shared meaning between the two language versions. Third, the interpreter must consider the other rules of statutory interpretation, including in particular, other evidence of legislative intent. In cases where there is a shared meaning, that meaning is presumed to express the intention of the legislature, and the point of looking at other rules of interpretation is to determine whether the presumption in favour of the shared meaning has been rebutted. In cases where there is no shared meaning, the presumption plays no role, and the correct outcome must be determined exclusively through reliance on the other rules of statutory interpretation.
III. WHEN DOES THE PRESUMPTION ARISE?
The position adopted in Daoust and in The Law of Bilingual Interpretation is that when a discrepancy occurs between the two versions of bilingual legislation, there are three possibdities. First, there is a shared meaning because one version lends itself to two or more possible meanings while the other clearly expresses only one of those meanings. This is labelled "ambiguity shared meaning." (16) Second, there is a shared meaning when the meaning expressed by one version is narrower than the meaning expressed by the other and the narrow meaning is contained in the broader meaning. This is labelled "breadth shared meaning." (17) Both types of shared meaning give rise to a presumption in favour of shared meaning, although the former often attracts more weight. Third, in the absence of either form of shared meaning, the two versions are said to be in "absolute conflict" and no presumption arises. (18)
Here we reach the key difference between the formulation of the shared meaning rule adopted in the Daoust case and the book, and the formulation that is proposed here. In my view, not every instance of shared meaning should give rise to the presumption that the shared meaning was intended. The presumption should arise only in cases where one version is ambiguous and the other is clear (ambiguity shared meaning). In all other cases, the presumption should not arise and the interpreter should go directly to the other rules of statutory interpretation to resolve the discrepancy. Even if there is an overlap between the two versions so that the range of reference of one version is encompassed within the range of reference of the other, this type of shared meaning (breadth shared meaning) does not give rise to a presumption. From a legal point of view, breadth shared meaning is an instance of absolute conflict, so there is no presumption to be overcome.
On the surface, my objection to the formulation of the shared meaning rule adopted in Daoust and the book may seem pedantic. Regardless of whether shared meaning gives rise to the presumption, the full panoply of rules must be applied in every case. In the end, if it appears that the legislature intended a meaning other than the shared one, that other meaning will prevail. This point is repeatedly made in the book in response to criticism that Daoust embraces a textualist approach. (19) Strictly speaking, this is a fair response. The Law of Bilingual Interpretation does not say that interpretation ends once a shared meaning is found. On the contrary, it emphasizes that interpretation must continue until all legitimate evidence of legislative intent is fairly canvassed.
Although the differences between the two formulations may seem more formal than substantive, in my view they are highly significant, not only for the outcome in particular cases, but also for the judicial approach to interpretation generally. Presumptions matter; otherwise there would be no point in creating them. When the existence of a shared meaning gives rise to a presumption in favour of the shared meaning, it transforms the interpretive question from "what is the best or most plausible meaning having regard to the purpose and context?" into "is there other evidence of legislative intent that is sufficient to rebut the presumption?" In cases where the other evidence of legislative intent is weak or inconclusive, the presumption is likely to be determinative. More fundamentally, the competing formulations reflect different conceptions of legislative intent, different justifications for the shared meaning rule and different assumptions about the judicial role in interpretation.
IV. THE GOVERNING PRINCIPLE OF STATUTORY INTERPRETATION
In Re Rizzo & Rizzo Shoes Ltd., (20) the Supreme Court of Canada established Driedger's modern principle as the fundamental rule governing statutory interpretation in Canada. The following is the key passage, written by Justice Iacobucci:
Although much has been written about the interpretation of legislation ... Elmer Driedger in Construction of Statutes (2nd ed. 1983) best encapsulates the approach upon which I prefer to rely. He recognizes that statutory interpretation cannot be founded on the wording of the legislation alone... Today there is only one principle or approach, namely, the words of an Act are to be read in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament. (21)
An important aspect of Justice Iacobucci's pronouncement is his preference for the modern principle, as set out in the second rather than the third edition of the Construction of Statutes. (22) There are differences between Driedger's formulation and the formulation in subsequent editions. First, the subsequent editions are clear about the legitimacy of relying on norms such as the rule of law and fairness in interpreting legislation. Second, they acknowledge the possibility of conflict between textual meaning, legislative intent and judge-made norms and resolve it by assigning to the court the duty to attach appropriate weight to these competing factors in light of the circumstances of each case. With this approach, the modern principle does not assign a pre-determined weight to textual meaning or to legislative purpose or to particular presumptions of legislative intent; the weight attaching to these factors is determined in light of the circumstances of each case.
Although there are differences between the second and subsequent editions of Driedger, every edition is clear in its repudiation of textualism (also known as the "plain meaning rule"). In his judgment in Rizzo, Justice Iacobucci points out that, according to the modern principle, interpretation cannot be founded on wording alone. (23) In other words, legislative intent cannot be equated with the textual meaning of the provision to be interpreted--even if that meaning seems clear. The meaning of a text is merely evidence--albeit important evidence--of the rule that the legislature intended to enact. Textual meaning must be complemented by other evidence, most notably by legislative purpose and scheme, but also by the traditional and evolving presumptions of legislative intent. (24)
Driedger says that the words of a provision to be interpreted must be read in their "entire context," (25) a concept that he defined expansively. In the first and second editions of Construction of Statutes, (26) chapters one to three review the common law approaches to interpretation from the 16th to the 20th century culminating in the modern principle as set out in chapter four. Most of the remainder of the book is then devoted to elucidating what Driedger meant by context. (27) This part of the Construction of Statutes is generally ignored by the Supreme Court of Canada. This practice has allowed the Court to rely, at least some of the time, on a narrow conception of context, which has led to a degree of backsliding into textualism--one of the evils the modern principle was meant to overcome. Here too, Justice Iacobucci has led the way. In Bell ExpiessVu Limited Partnership v Rex, (28) he wrote:
[W]hen a statute comes into play during judicial proceedings, the courts (absent any challenge
on constitutional grounds) are charged with interpreting and applying it in accordance with
the sovereign intent of the legislator. In this regard, although it is sometimes suggested that
"it is appropriate for courts to prefer interpretations that tend to promote those [Charter]
principles and values over interpretations that do not", it must be stressed that, to the
extent this Court has recognized a "Charter values" interpretive principle, such principle
can only receive application in circumstances of genuine ambiguity, i.e., where a statutory
provision is subject to differing, but equally plausible, interpretations. (29)
This passage reveals two important assumptions about the understanding of the modern principle promoted by Justice Iacobucci and other members of the Court. First, the modern principle is a tool to discover legislative intent, which is the goal of statutory interpretation. Second, legal context has a limited role to play in that endeavour. As explained by Justice Iacobucci, context consists of the text as a whole and the purpose for which the text was enacted, but does not necessarily include the other contexts explored by Driedger in his first and second editions and carried forward in subsequent editions. (30) The Court's pronouncement in Bell ExpressVu implies that the fundamental values of the Canadian legal system, many of which are enshrined in our entrenched constitutional law, are not front and centre when the executive branch develops a legislative proposal, nor when the legislative branch enacts it, and they should not be front and centre when the judicial branch interprets it. In my view, this assumption is faulty.
I have no quarrel with a passage like the one in Bell ExpressVu to the extent it emphasizes that presumed intent does not automatically trump other considerations. My quarrel is with the attempt to preclude consideration of these norms unless and until the legislation--considered apart from these norms--is found to be ambiguous. I also quarrel with the idea that the weight of the considerations affecting interpretation should be determined in the abstract, outside the context of the particular problem to be dealt with. In my view, the weight of textual meaning depends on the clarity and precision of the text as it applies to the facts in question, just as the weight of a presumption depends on the importance of the underlying constitutional norm, the consequences of its violation in the circumstances of the case and its relation to other relevant norms.
V. JUSTIFYING THE SHARED MEANING RULE
In The Law of Bilingual Interpretation, Justice Bastarache and his co-authors argue that historically the courts have relied on two different justifications for resolving discrepancies on the basis of the shared meaning rule. The first is that, when one version is clear and the other ambiguous, the clear version helps to clarify the ambiguous one. They comment,
In this view of the rule, the clear version plays a role equivalent to that played in a unilingual interpretation where two different provisions in the same piece of legislation refer to or use the same term, and one provision's use of the term can help to clarify the other provision's unclear use of it. (31)
In their view, this justification is unsatisfactory.
A. Norm Based Justification
It would be difficult to disagree that this justification for the shared meaning rule is unsatisfactory. It ignores the essential point, which is that both versions of the Act are equally authentic expressions of legislative intent. In proposing an alternative formulation, however, I do not rely on it. My justification for the weak version is not that clear language clarifies ambiguous language, but rather that adopting the shared meaning in such a case minimizes unfair surprise and loss of linguistic security, whereas adopting the meaning that is not shared would undermine those norms. Adopting the shared meaning in this context is also consistent with the presumption of competent drafting. Given the attention lavished on the preparation of legislative texts, they are presumed to be accurate statements of the law on which subjects can safely rely. By preferring the interpretation that fits both language versions, it is unnecessary to treat either as a mistake.
To illustrate, I will use the discrepancy considered by the Supreme Court of Canada in K. v Tupper. (32) The English version of the Criminal Code, as it read at the relevant time, made it an offence to possess "any instrument for house-breaking" while the French version made it an offence to possess "un instrument pouvant servir aux effractions de maisons." (33) The English version can be understood to mean: (1) an instrument that is capable of being used for house-breaking or (2) an instrument that is designed for house-breaking. The French version clearly refers to an instrument that is capable of being used for house-breaking If meaning (1) (the shared meaning) is adopted, the rule that Parliament is presumed to intend will in fact be expressed in both language versions and neither language community will be taken by surprise. Because the French text accurately expresses the presumably intended rule, French-speaking subjects can safely rely on the text without having to resort to the other language version. For them, there is no threat to either the rule of law or linguistic security. English speakers are not much worse off. Since the ambiguity is apparent on the face of the English text, the adoption of meaning (1) will not take readers by surprise. However, to identify (1) as the correct expression of the rule, English speaking readers will have to read the French version, thereby experiencing some loss of linguistic security. Nonetheless, on balance, by adopting the shared meaning, readers of both versions do relatively well. By contrast, if meaning (2) were adopted (the meaning that is not shared), the position of English readers would remain the same, but the position of French readers would be completely undermined. On the face of the French version, there is nothing to indicate that it is not reliable and that recourse to the other language version is needed. If meaning (2) were adopted, French readers would be taken by surprise and their linguistic security would be destroyed. As this example shows, in cases of ambiguity shared meaning, there are good reasons to presume the shared meaning, and to discourage adopting the meaning that is found in one version alone.
The analysis plays out very differently in cases of breadth shared meaning. In such cases, each version clearly states a different rule; that is, each version lends itself to one interpretation only and there is nothing on the face of either to indicate a need to resort to the other language version for clarification. Unlike presuming the clear version, presuming the narrow version does not enhance the rule of law or minimize impairment of linguistic security. No matter which language version prevads, the intended rule is going to be accurately expressed in one version only and members of the other language community are going to be surprised and suffer a loss of linguistic security. Furthermore, unlike a discrepancy between a clear and an ambiguous version, which can be resolved through interpretation alone, a discrepancy between two clear versions requires the court to identify and correct a drafting mistake. In a contest between a clear broad and a clear narrow version, there is no reason to presume that the narrow version is the one intended by the legislature.
Consider the discrepancy resolved by the Federal Court of Appeal in the Deltonic Trading case. (34) The English version of the Customs Tariff at the relevant time imposed a tariff on "lobster," while the French version referred to "homard." In English, the word "lobster" is used to describe a variety of crustaceans, including some (such as rock lobster) without claws. In French, the word "homard" refers exclusively to crustaceans with claws. For purposes of discussion, let us assume that the English word "lobster" is not ambiguous and that it clearly refers only to any crustacean named "lobster," with or without claws. That is the broad meaning. Let us also assume that the French word "homard" is clear and refers to crustaceans with a certain type of claws. That is the narrow meaning.
Because the narrow meaning is contained in both versions, adopting it--like adopting the clear version--ensures that the same meaning is found in both. However, here the similarity ends. In the case of ambiguity shared meaning, readers of the ambiguous version have no expectation of being able to rely on their text without consulting the other language version. If the clear version prevails, they will not be taken by surprise. In the case of breadth shared meaning, because the broad rule is clearly expressed, readers of that version have no reason to doubt the accuracy of their text, and if the narrow rule prevails, they will be taken by surprise. English-speaking Canadian lobster fishers, for example, would fairly assume that clawless lobsters imported into Canada are subject to a tariff and might arrange their affairs so as to benefit from this competitive advantage. The fact that the narrower rule is encompassed within the clearly expressed broad rule does nothing to protect them from disappointment should the narrow version prevail.
Further, in the case of ambiguity shared meaning, adopting the meaning that is not shared would completely undermine readers of the clear version, whereas adopting the shared meaning avoids this effect. In the case of breadth shared meaning, however, it does not matter which version is adopted. In either case, the readers of one of the versions will be taken by surprise and their linguistic security will be undermined. In the case of crustacean fishers, if the narrow meaning were adopted, fishers who invested in clawless lobster fishing, in reliance on the broad English version, would be surprised and might suffer economic loss as a result of their reliance. Conversely, if the broad meaning were adopted, fishers who avoided clawless crustaceans in order to concentrate on the protected clawed ones would be equally surprised and might equally complain that they had lost an opportunity to exploit a competitive advantage.
B. Textualist Justification
In the analysis set out above, I attach little significance to the fact that the narrow version is a subset of the broad version, and therefore its meaning is "contained" in both versions. Equal authenticity requires the text of each version to be a reliable expression of the single rule that the legislature intended to enact. If the two texts express clear but different rules, in my view, neither is one is reliable, and some basis other than textual meaning must be found to determine which version is to be preferred. I see no justification for supposing that because the narrow rule is subsumed in the broader one, it is the rule intended by the legislature.
In this respect, my justification for the shared meaning rule is very different from the justification offered in The Law of Bilingual Interpretation, which is grounded in textualist assumptions. A textualist assumes that the textual meaning of a properly drafted legislative text, one that is clear, is its "actual" meaning and is appropriately equated with the rule that the legislature intended to enact. By giving effect to clear textual meaning, a court automatically gives effect to the intentions of the legislature. And since that meaning, is also apparent to, and relied upon by the public, giving effect to it also respects the rule of law. For these reasons, textualists conclude that further interpretation is neither necessary nor appropriate when the meaning of a legislative text is clear. It is legitimate to rely on purposive or contextual analysis only if the textual meaning proves ambiguous.
In The Law of Bilingual Interpretation, this reasoning is transposed to the interpretation of bilingual texts by equating shared meaning in a bilingual text with clear meaning in a unilingual one. Thus, the authors write: "a finding of a shared meaning creates a presumption in favour of the shared meaning as the actual meaning of the provision--it is, after all, a tool to determine legislative intent ..." (35) They also adopt the two step approach of textualism, once again with shared meaning taking the place of "clear" or "plain" meaning:
[T]he search for legislative intent has, in the past, been seen as an integrated part of the task of interpreting the two versions, rather than as a separate step to be performed after the application of the Shared Meaning Rule. However, since Daoust the law is clear that the search for shared meaning and the appeal to external rules of statutory interpretation are best seen as distinct tasks. (36) If [the] two versions taken together express a clear legislative rule, then appeal to other sources of information about legislative intent is neither necessary nor advisable. As the Supreme Court noted in Bell ExpressVu: Other principles of interpretation--such as the strict construction of penal statues and the "Charter values" presumption--only receive application where there is ambiguity as to the meaning of a provision. (37) When the Shared Meaning Rule results in a clear understanding of the shared meaning of the bilingual provision, the situation is quite similar to that of a clearly stated unilingual provision, and in such a case appeal to these "other principles" does not appear to be necessary. Of course this is not to say that a shared meaning disclosed by application of the rule should overcome strong evidence that the intention of the legislator was different; at the same time, some evidence as to legislative intent--such as ancient presumptions--may be seen to be somewhat less persuasive than an interpretation that arises out of the words of the legislation. (38)
This passage reveals what is at best ambivalence toward the rejection of textualism implicit in Driedger's modern principle. In Bell ExpressVu, the Court reintroduces a two-step approach to interpretation: first determine whether the legislation is clear, and second resort to certain presumptions of legislative intent only if the legislation is ambiguous. (39) However, the Court there and in subsequent decisions (40) emphasizes that the determination of whether the legislation is clear or ambiguous is based not only on textual meaning, but also on purposive and contextual analysis, (41) including at least some aspects of the legal context and the external context. The only things clearly excluded in Bell ExpresVu are the presumptions based on constitutional norms (of which shared meaning is itself an example). In The Law of Bilingual Interpretation, however, context and purpose, along with the other presumptions are labelled "external rules." (42)
VI. The Daoust Case
The purpose of this part is to compare the Supreme Court of Canada's analysis in Daoust, based on the strong version of shared meaning, to an analysis based on the weak version. In my view, an analysis based on the weak version offers a clearer and more compelling justification of the outcome in the case.
The issue in Daoust was whether the operators of a pawn shop, by receiving property as a pledge for a loan knowing that the property was stolen, violated subsection 462.31(1) of the Criminal Code, which reads as follows in English and French respectively: (43)
462.31 (1) Every one commits an offence who uses, transfers the possession of, sends or delivers to any person or place, transports, transmits, alters, disposes of or otherwise deals with, in any manner and by any means, any property or any proceeds of any property with intent to conceal or convert that property or those proceeds, knowing or believing that all or a part of that property or those proceeds was obtained ... as a result of ... an enterprise crime offence....
462.31 (1) Est coupable d'une infraction quiconque--de quelque facon que ce soit--? utilise, enleve, envoie, livre a une personne ou a un endroit, transporte, modifie ou aliene des biens ou leurs produits--ou en transfere la possession--dans l'intention de les cacher ou de les convertir sachant ou croyant qu'ils ont ete obtenus ou proviennent, en totalite ou en partie ... de la perpetration ... d'une infaction de criminalite organise....
To follow the steps for interpreting bilingual legislation, the first question to ask is whether there is a discrepancy between the two versions. And clearly there is: whereas the English version makes any form of dealing with the property or proceeds an offence, (44) the French version makes only the enumerated forms of dealing an offence. Under both formulations of the shared meaning rule, this is a discrepancy.
The second question to ask is whether there is a shared meaning. If the weak analysis is adopted, there is no shared meaning because both versions of the legislation state clear rules and they are different. The English version criminalizes not only the enumerated forms of dealing mentioned in the French version, but also other forms of dealing, which would include the facts of the case. The French version is equally clear in including only the enumerated forms of dealing, which would exclude the facts of the case. To resolve this discrepancy, resort must be had to other indicators of legislative intent.
However, if the strong version is adopted, there is a shared meaning. Although both versions state a clear rule, the narrow meaning--the list of enumerated forms of dealing--is found in both versions. Since there is a shared meaning, other indicators must be sufficiently strong to overcome the presumption in favour of the shared meaning.
A. The Strong Version Analysis
What is striking about Daoust is that the Court finds strong evidence that Parliament in fact intended to enact the rule expressed by the broader English version. It concludes that the narrow French version was the result of a drafting mistake. Under step three of the shared meaning ride, this is supposed to rebut the presumption. Nonetheless, the Court concludes that the shared meaning prevails:
We can conclude from the legislative history of the enactments pertaining to the laundering of proceeds of crime that Parliament's true intent was to criminalize all acts ("or otherwise deal with") in relation to the proceeds of crime where the intent is to conceal or convert them. This intent is explicit in the English version of s. 462.31. Nevertheless, the legislative intent revealed by the history must be one that could reasonably be supported by the text of the statute. Such is not the case here. Parliament did not achieve what it intended when it drafted s. 462.31. For this reason, the French version, the one with the narrower meaning, must be favoured. Here, we are concerned with discovering not only the intent that Parliament was pursuing, but also the intent it expressed ... The two versions of s. 462.31 are divergent because of an error or an omission on the part of Parliament, but that does not give this Court the authority to amend a clearly drafted enactment: Gaysek v. The Queen,  S.C.R. 888, at p. 895; Ville de Montreal v. ILGWU Center Inc.,  S.C.R. 59, at p. 66. If this Court did have such authority, an accused could not know the limits of his or her liability. It therefore follows that the text to be analysed here is the one that allows us to establish the common meaning, that is, the more restrictive of the two versions. Since the two versions arc identical, with the exception of an addition in the English version, the French version must prevail for the purposes of this analysis. (45)
The Court finds that the legislative intent revealed by the legislative evolution of the relevant provisions must be rejected because that intent is expressed by the broad but not by the narrow version of the legislative text. The narrow meaning must prevail, not because it is intended, but because "here, we are concerned with discovering not only the intent that Parliament was pursuing, but also the intent it expressed." (46) In other words, despite repeated insistence that the shared meaning must be tested against other indicators of legislative intent, in practice the narrow meaning must always prevail because it is the only meaning expressed in both versions of the text. What matters is not what the legislature intended, but what it actually said.
This part of the reasoning in Daoust is inconsistent with the many cases in which Canadian courts have indeed rejected the narrow version of a bilingual text in favour of the broader one. (47) Further, it incorrectly denies the courts' jurisdiction to correct drafting errors. A drafting error occurs whenever a legislative text does not express the rule that the legislature intended to enact. Such an error occurs every time the two versions of bilingual legislation clearly express different rules, one broad and one narrow. Only one version is correct. And once the correct version is identified by the court, the mistake in the other version is corrected, in effect. This occurs as much when the narrow meaning prevails as it does when the broad meaning is preferred. In either case, one of the versions is notionally redrafted to bring it in line with the correct version.
B. The Weak Version Analysis
There can be little doubt that Daoust was rightly decided. However, an analysis of the facts in Daoust based on the weak version of shared meaning offers a better justification of the outcome. Under the weak version, the shared meaning between the two versions of subsection 462.31(1) of the Code (48) does not give rise to a presumption that Parliament intended the narrow version and resort must be had to the other rules of statutory interpretation. The legislative evolution examined by the Court strongly suggests that in fact Parliament intended to enact a broad anti-laundering rule. However, it is clearly unacceptable that a person who might legitimately have relied on the narrow rule expressed in the French version could be convicted under the broad rule expressed in the English version. Since the state holds out both versions as authentic, it cannot penalize a person for relying on the narrower one--not without violating; important constitutional norms. However, the determinative norms here are not linguistic equality or the protection of minority language rights, but rather fairness and the rule of law.
When the English and French versions of bilingual legislation express clear but different rules, the state has made a mistake and all subjects are at risk of being taken by surprise. In deciding which rule to adopt, courts should strive as much as possible to minimize the surprise. Further, in a public law context, courts might appropriately presume that the state should bear the burden of the mistake. When interpreting penal legislation or legislation that interferes with the rights of the subject, that would entail adopting a strict construction (in effect, favouring, the narrow version). Conversely, when interpreting quasi-constitutional legislation or legislation that confers benefits, it would entail adopting a liberal construction (in effect, favouring the broad version). I do not mean to suggest that these presumptions should be determinative. In every case, the weight attaching to the various interpretive factors must depend on the circumstances of the case. In Daoust, despite persuasive evidence that the legislature intended the broad rule, the court declined to give effect to that intention because the consequences of adopting the intended rule were unacceptable. In the circumstances of that case, as occasionally happens, constitutional norms trumped legislative intent.
VII. POST-DAOUST LAW
Since it was decided, Daoust has been widely recognized as the leading authority on the interpretation of bilingual legislation, and it has been relied on in a number of judgments to justify a presumption in favour of breadth shared meaning. (49) It is clear that several members of the Supreme Court of Canada are committed to this approach. (50) In these subsequent cases, however, the courts consistently acknowledge the real possibility of rebutting the presumption. Furthermore, Daoust has been recast to some extent as a case in which the narrow meaning prevailed because of the principle that penal legislation is narrowly construed. (51)
Although Daoust is widely cited and its teachings are now reinforced by The Law of Bilingual Interpretation, there are some dissenting voices. First, there is Pierre-Andre Cote himself. Although his account of the shared meaning rule as set out in the third edition of his text (52) is relied on by Justice Bastarache in Daoust, Cote has since rejected the notion that breadth shared meaning should give rise to a presumption. In a subsequently published paper, he wrote:
Il y a lieu de faire observer que si la prevalence de la version claire sur la version ambigue se justifie rationnellement, puisque l'on doit presumer que la meilleure expression de la volonte legislative est celle qui est exempte d'ambiguite, il en va autrement de la prevalence de la version restreinte: il n'y a,a notre avis, aucun motif rationnel de preferer le sens le plus restreint, car rien ne permet d'affirmer qu'il represente mieux l'intention legislative que le sens large. (53) [It is worth noting that the primacy of the clear version over the ambiguous version is rationally justified, for it must be presumed that the better expression of the legislature's will is the one that is free of ambiguity; however, the same cannot be said of the primacy of the narrow version over the broader one: in my opinion, there is no rational basis to automatically prefer the narrow meaning, for there is no justification for saying that it is a better expression of the legislature's will than the broader one.] (54)
A second source of potential resistance can be found in the majority judgment of the Supreme Court of Canada in Khosa. (55) The issue in that case was whether section 18.1 of the Federal Court Act (56) precluded reliance on the common law standards for judicial review established by the Court in Dunsmuir. (57) The majority found that the section did not exclude the common law because it dealt with the grounds for review rather than with review standards. Subsections 18.1(3) and (4) are in the following terms:
Powers of Federal Court
(3) On an application for judicial review, the Federal Court may
(a) order a federal board, commission or other tribunal to do any act or thing it has unlawfully failed or refused to do or has unreasonably delayed in doing; or
(b) declare invalid or unlawful, or quash, set aside or set aside and refer back for determination in accordance with such directions as it considers to be appropriate, prohibit or restrain, a decision, order, act or proceeding of a federal board, commission or other tribunal.
Pouvoirs de la Cour federale
(3) Sur presentation d'une demande de controle judiciaire, la Cour federale peut :
a) ordonner a l'office federal en cause d'accomplir tout acte qu'il a illegalement omis ou refuse d'accomplir ou dont il a retarde l'execution de maniere deraisonnable;
b) declarer nul ou illegal, ou annuler, ou infirmer et renvoyer pour jugement conformement aux instructions qu'elle estime appropriees, ou prohiber ou encore restreindre toute decision, ordonnance, procedure ou tout autre acte de l'office federal.
Grounds of review
(4) The Federal Court may grant relief under subsection (3) if it is satisfied that the federal board, commission or other tribunal
(a) acted without jurisdiction, acted beyond its jurisdiction or refused to exercise its jurisdiction;
(b) failed to observe a principle of natural justice, procedural fairness or other procedure that it was required by law to observe;
(c) erred in law in making a decision or an order, whether or not the error appears on the face of the record;
(d) based its decision or order on an erroneous finding of fact that it made in a perverse or capricious manner or without regard for the material before it;
(e) acted, or failed to act, by reason of fraud or perjured evidence; or
(f) acted in any other way that was contrary to law.
(4) Les mesures prevues au paragraphe (3) sont prises si la Cour federale est convaincue que l'office federal, selon le cas :
a) a agi sans competence, outrepasse celleci ou refuse de l'exercer;
b) n'a pas observe un principe de justice naturelle ou d'equite procedurale ou toute autre procedure qu'il etait legalement tenu de respecter;
c) a rendu une decision ou une ordonnance entachee d'une erreur de droit, que celleci soit manifeste ou non au vu du dossier;
d) a rendu une decision ou une ordonnance fondee sur une conclusion de fait erronee, tiree de facon abusive ou arbitraire ou sans tenir compte des elements dont il dispose;
e) a agi ou omis d'agir en raison d'une fraude ou de faux temoignages;
f) a agi de toute autre facon contraire a la loi.
Speaking for the majority, Justice Binnie pointed out what he took to be a discrepancy between the French and English versions of subsection 18.1(4): "The English version of s. 18.1(4) is permissive; the court is clearly given discretion. In the French version, the words 'sont prises' translate literally as 'are taken' which do not, on the face of it, confer a discretion." (58) As indicated by the federal Interpretation Act, (59) the present indicative in the French version is presumed to impose a duty. (60) Here is how Justice Binnie deals with the apparent discrepancy in the Khosa case:
[T]he linguistic difference must be reconciled as judges cannot be seen to be applying s. 18.1(4) differently across the country depending on which language version of s. 18.1(4) they happen to be reading. In R. v. Daoust, ... the Court cited with approval the following approach: Unless otherwise provided, differences between two official versions of the same enactment are reconciled by educing the meaning common to both. Should this prove to be impossible, or if the common meaning seems incompatible with the intention of the legislature as indicated by the ordinary rules of interpretation, the meaning arrived at by the ordinary rules should be retained. (61) ... Linguistic analysis of the text is the servant, not the master, in the task of ascertaining Parliamentary intention: see Slaight Communications Inc. v. Davidson ... (62) A blinkered focus on the textual variations might lead to an interpretation at odds with the modern rule because, standing alone, linguistic considerations ought not to elevate an argument about text above the relevant context, purpose and objectives of the legislative scheme ... Here the English version cannot be read so as to compel the court to grant relief: the word "may" is unquestionably permissive. In Bastarache et al., (63) it is said that "the clearer version provides the common meaning" (p. 67), but it cannot be said that the French text here is ambiguous. Accordingly, the linguistic issue must be placed in the framework of the modern rules of statutory interpretation that give effect not only to the text but to context and purpose. There is nothing in the context or purpose of the enactment to suggest a Parliamentary intent to eliminate the long-standing existence of a discretion in judicial review remedies. As mentioned earlier, the principal legislative objective was simply to capture the judicial review of federal decision makers for the Federal Court. (64)
Two features of this analysis are striking. First, in applying Daoust or the shared meaning rule as formulated in Daoust, Justice Binnie ignores breadth shared meaning. Having found that neither version of subsection 18.1(4) is ambiguous, he turns immediately to the "external" rules of interpretation, in particular purposive and contextual analysis. Second, he expressly repudiates a textualist approach to the interpretation of bilingual legislation: "linguistic considerations ought not to elevate an argument about text above the relevant context, purpose and objectives of the legislative scheme ...." (65)
Obviously, Khosa does not overturn Daoust, or significantly weaken its authority, but it suggests that not all members of the Court are committed to the strong version of shared meaning adopted in Daoust, and that for some members of the Court its textualist assumptions may prove problematic. My hope is that despite the claim in The Law of Bilingual Interpretation that Daoust definitively states that law, some aspects of it remain open for debate.
The Law of Bilingual Interpretation is an important contribution to Canadian legal scholarship. It raises and addresses issues that have received little attention in the past, such as the interpretation of bilingual judgments and multilingual treaties. This article has focussed on what I have called the strong version of the shared meaning rule and the authors' efforts to justify it on the basis of certain constitutional norms. In the early pages of the book, their emphasis is on linguistic equality and the protection of minority language rights. Later, the focus shifts to respect for legislative intent. I have tried to show that these norms are not particularly well served by the strong formulation of the rule.
Linguistic equality is achieved, to the extent it can be in bilingual legislation, by insisting on the equal authenticity of both language versions and by resolving discrepancies in a way that does not automatically favour one language over the other. The strong rule meets these criteria, but no more so than the weak rule does. Neither formulation ensures an equal distribution of "correct answers" between the two languages. Both look to something other than language to decide which version is correct. In fact, both look to the intention of the legislature to settle the matter. Under the strong rule, legislative intent is presumptively identified with the meaning found in both versions, which is the narrower one, and that presumed meaning is tested against the overall context and purpose. This approach, I have argued, is a partial return to the textualism rejected by the Supreme Court of Canada in the Rizzo case. Under Driedger's modern principle, the intention of the legislature is determined after, not before, context and purpose are taken into account. More importantly, a presumption in favour of the narrower version of a legislative rule is a form of strict construction that lacks a rational basis. There is no reason to presume that a legislature would have intended, or would prefer, a narrow rather than a broad rule.
As a proponent of the weak formulation of the shared meaning rule, I think we do not need a presumption to deal with cases where both language versions are clear, but one is broader than the other. Supposing we do, however, I would favour a presumption based on the idea underlying the contra proferentum rule in contract law: if a legislature creates and publishes a text with a mistake that gives rise to an interpretation dispute, it is fair to make the legislature bear the cost of its bad drafting. In a penal law context, this would favour the narrower version of the text, but in the context of benefit-conferring legislation, the broader version would prevail. (66) Like all presumptions, the weight attaching to this one should depend on the circumstances of each case. With or without the benefit of presumptions, good interpretation is almost always a balancing act.
(1) 2004 SCC 6,  1 SCR 217, 235 DLR (4th) 216 [Daoust cited to SCR].
(2) The Honourable Mr. Justice Michel Bastarache et al, The Law of Bilingual Interpretation (Markham, Ont: LexisNexis Canada, 2008).
(3) Constitution Act, 1867 (UK), 30 & 31 Vict, c 3, s 133, reprinted in RSC 1985, App II, No 5.
(4) Manitoba Act, 1870, s 23, 33 Vict, c 3, reprinted in RSC 1985, App II, No 8.
(5) Canadian Charter of Rights and Freedoms, ss 16-23, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act, 1982 (UK), 1982, c 11.
(6) 2009 SCC 12,  1 SCR 339, 304 DLR (4th) 1.
(7) Bastarache, supra note 2 at 10.
(8) Nearly all the so-called "rules" of statutory interpretation operate as presumptions.
(9) See Bastarache, supra note 2 at 23-32 for discussion of the equal authenticity rule. The book rightly points out that publishing law in two languages does not logically entail that both are authentic. In Canada, however, the authenticity of both versions has long been established and is now confirmed in various constitutional and quasi-constitutional texts.
(10) The book points out that linguistic security is the primary norm underlying section 133 of the Constitution Act, 1867 (supra note 3). See Bastarache, supra note 2 at 16-17. However, the significance of linguistic security is later played down and the symbolic importance of bilingual legislation is emphasized: bilingualism tells speakers of both languages that they "are valued enough to have [their] government's rules and decisions conveyed to them in their own language." See ibid at 30-31.
(11) Bastarache, supra note 2 at 15.
(12) Pierre-Andre Cote, The Interpretation of Legislation in Canada, 3d ed (Scarborough, Ont: Carswell, 2000) at 326-28. Cote's description of the steps first appeared in the first edition of his text Interpretation des lois (Montreal: Les Editions Yvon Biais, 1982) at 274-75. This formulation was carried forward without change to the third edition, which is the one quoted in Daoust, infra.
(13) Daoust, supra note 1 at para 27.
(14) Bastarache, supra note 2 at 42-48.
(15) See ibid at 53, where Mr. Justice Bastarache says that "the process of comparison between the two versions should take place at the level of the provision as a whole, rather than at the level of individual words which might differ." It is also to take place at the level of the provision, rather than at the level of the Act as a whole.
(16) Ibid at 64.
(17) Ibid at 74.
(18) Ibid, at 46-48.
(19) See example, ibid at 34, 82, 83-84, 88.
(20)  1 SCR 27, 154 DLR(4th) 193 [Rizzo cited to SCR].
(21) Rizzo, supra note 20 at para 21.
(22) Please see Elmer A Driedger, Construction of Statutes, 2d ed (Toronto, Ont: Butterworths, 1983) and Ruth Sullivan, Driedger on the Construction of Statutes, 3d ed (Markham, Ont: Buttersworth, 1994).
(23) Supra note 21.
(24) See, for example, Pierre-Andre Cote, "Le mot 'chien' n'aboie pas: reflexions sur la materialite de la loi" in Gerard Cohen-Jonathan et al, Melanges Paul Amselek (Bruxelles: Bruylant, 2005) at 279.
(25) Elmer A Driedger, Construction of Statutes, 2d ed (Toronto, Ont: Butterworths, 1983) at 87.
(26) Please see Elmer A Driedger, Construction of Statutes, 1st ed (Toronto, Ont: Butterworths, 1974) and Elmer A Driedger, Construction of Statutes, 2d ed (Toronto, Ont: Butterworths, 1983).
(27) See supra note 25 at 107-108. Driedger writes: "There is Internal Context and External Context. Internal Context is everything contained within the four corners of the Act....External Context is the setting of the Act, and will here be considered under the headings Social, Legal, Language and Intellectual Contexts."
(28) 2002 SCC 42,  2 SCR 559, 212 DLR (4th) 1 [Bell ExpressVu cited to SCR].
(29) Bell ExpressVu, supra note 28 at para 62 [emphasis in original].
(30) See supra note 25 at 158. In the second edition, Driedger writes: "The general body of the law--statutes and judicial decisions--are included in external context. Law outside the statute under consideration may not be relevant; but it may always be looked at and if it is relevant it may have a bearing on the construction of the statute."
(31) Bastarache, supra note 2 at 64.
(32)  SCR 589, 63 DLR (2d) 289 [Tapper cited to SCR].
(33) Please see Criminal Code, RSC.
(34) Deltonic Trading Corp. v Deputy M.N.R., Customs and Excise (1990),  113 NR 7 (FCA), 21 ACWS (3d) 823 [Deltonic Trading Corp. cited to NR].
(35) Bastarache, supra note 2 at 90.
(36) Ibid at 84-85.
(37) Bell ExpressVu, supra note 28 at 28.
(38) Bastarache, supra note 2 at 85.
(39) Bell ExpressVu, supra note 28.
(40) See, for example, Bristol-Myers Squibb Co. y Canada (AG)  1 SCR 533, 253 DLR (4th) 1; Pharmascience Inc v Binet,  2 SCR 513, 273 DLR (4th) 192.
(41) See Bell ExpressVu, supra note 28 at para 30. Justice Iacobucci writes:"[I]t is necessary, in every case, for the court charged with interpreting a provision to undertake the contextual and purposive approach set out by Driedger, and thereafter to determine if 'the words are ambiguous enough to induce two people to spend good money in backing two opposing views as to their meaning'" [emphasis original).
(42) See Bastarache, supra note 2 at 85, where all "factors other than the Shared Meaning Rule" are considered under the heading "External Rules Apply in Case of No Shared Meaning."
(43) Criminal Code, RSC 1985, c C-46, s 462.31(1).
(44) This claim is subject to an interpretation based on the associated words or limited class rule.
(45) Daoust, supra note 1 at para 44-47.
(46) Daoust, supra note 1 at para 44.
(47) See, for example, R v Compagnie Immobiliere BCN Ltee,  1 SCR 865; R v Jean B,  I SCR 80; Beothuk Data Systems Ltd, Seawatch Div v Dean,  1 FC 433 (FCA).
(48) Supra note 43.
(49) See, for example, the dissenting judgment of Justice Deschamps in Caisse populaire Desjardins de l'Est de Drummond v Canada, 2009 SCC 29, 2009 2 SCR 94 at para 84, 309 DLR (4th) 323 [Caisse populaire cited to SCR]; R v S.A.C., 2008 SCC 47,  2 SCR 675 at para 13, 295 DLR (4th) 407 [S.A.C. cited to SCR]; the dissenting judgment of Justice Charron in Montreal (City of) v Quebec (Commission desdroits de la personne et des droits de la jeunesse), 2008 SCC 48,  2 SCR 698 at paras 51-55, 295 DLR (4th) 577 [Commission des doits de la personne cited to SCR]; Medovarski v Canada (Minister of Citizenship and Immigration); Esteban v Canada (Minister of Citizenship and Immigration), 2005 SCC 5 1,  2 SCR 539 at paras 23-26, 258 DLR (4th) 193 [Medovarski cited to SCR]; D.K. c R, 2009 QCCA 987,  RJQ 1261 at para 37, 68 CR (6th) 120, 252 CCC (3d) 332 [D.K.avec renvois aux RJQ]; FederationJrancotenoise c Canada (Procureur general), 2008 NWTCA 6, 2009 WWR 259 at paras 162-64, 176 CRR (2d) 116 [Federation franco-tenoise cited to WWR]; Maritimes and Northeast Pipeline Limited Partnership v Elliott, 2004 FC 553,  3 FCR 612 at para 18, 238 DLR (4th) 358 [Elliott cited to FCR]; Greater Toronto Airports Authority v International Lease Finance Corp (2004), 69 OR (3d) 1 at paras 82-84, 235 DLR (4th) 618 [Greater Toronto Airports Authority cited to OR].
(50) See the judgments of Deschamps J. in Caisse populaire, ibid; S.A.C, ibid and Charron J. in Commission des doits de la personne, ibid.
(51) See S.A.C., ibid at 16.
(52) See Cote, supra note 12.
(53) Pierre-Andre Cote, "L'interpretation des textes legislatifs bilingues au Canada" in Rodolfo Sacco, dir, L'interpretation des textes juridiques rediges dans plus d'une langue (Torino: L'Harmattan, 2002) 12 [emphasis added].
(54) My translation.
(55) Khosa, supra note 6. The dissenting judgment did not address the issue.
(56) RSC 1985, c F-7.
(57) Dunsmuir v New Brunswick, 2008 SCC 9,  1 SCR 190, 291 DLR (4th) 577 [Dunjmir cited to SCR].
(58) Khosa, supra note 6 at para 39. In my view, on a proper interpretation of subsection 18.1 (4), there is no discrepancy between the two language versions. Both declare the grounds on which the powers set out in subsection 18.1(3) may be exercised. Both should be interpreted as declarations of those grounds: the grounds ... are / les motifs ... sont.
(59) RSC 1985, c 1-21.
(60) Loi d'interpretation, LRC 1985, c 1-21, art 11.
(61) Daoust, supra note 1 at para 26.
(62)  I SCR 1038 at 1071-72, 59 DLR (4th) 416 (Lamer J. dissenting in part, but not on this point).
(63) Bastarache, supra note 2.
(64) Khosa, supra note 6 at paras 39-40.
(65) Khosa, supra note 6 at para 39.
(66) This proposal is obviously limited to public law contexts. In legislation dealing with private legal relationships, the adversarial relation arises between private citizens rather than between private citizens and the state.
RUTH SULLIVAN *
* Ruth Sullivan is a Professor of Law at the University of Ottawa, retiring in 2011 after 27 years in the Common Law Faculty. Nine of those years were spent in the Legislative Services Branch of the Department of Justice of Canada, on exchange, sabbatical or secondment. She owes much to colleagues and friends at both places for inspiration, encouragement and laughs. Although the author has learned a great deal as a result of her association with the Department of Justice, the views expressed in this article do not necessarily express or reflect the Department's views.