Statutory Interpretation in A New Nutshell

STATUTORY INTERPRETATION IN
A NEW NUTSHELL
51
Ruth Sullivanl
Ottawa
This article attempts to update a Canadian classic - the realist account
ofstatutory interpretation published by John Willis in the Canadian Bar
Review in 1938. Willis' insights are compelling and they remain relevant
today. However, by focusing on the rhetoric of statutory interpretation, by
far its weakest point, Willis disregards the considerable work that goes on
when statutory interpretation is well done. This article draws attention to
that work.
Part 2 looks at the kinds of analyses relied on by good interpreters to
establish that elusive goal, the intention of the legislature . These include
textual, purposive, scheme, policy and consequential analysis . Part 2
examines the difference between easy and hard cases, then focuses on the
techniques used by interpreters to carry out the different kinds ofanalyses
and how these relate to the formal rules.
Part 3 looks at the range of arguments interpreters may construct based
on their preliminary analysis. Not every argument in statutory
interpretation is about the meaning ofwords. Interpreters also confront
drafter's mistakes, gaps in the legislative scheme, overlap and conflict,
and language that is over- or tinder-inclusive. The structure of these
different kinds ofarguments is set out and illustrated in Part 3.
Cet article essaie de mettre jour une thorie classique canadienne. En
1938, John Willis a publi dans cette revue un article prenant une
mthode raliste l'interprtation lgislative qui continue d'tre
pertinente aujourd'hui. Mais il se concentrait sur la rhtorique et laissait
de ct le travail considrable qui fait toujours une partie de
l'interprtation srieuse. Cet article essaie de mettre en relief ce travail
important.
La deuxime partie examine les types d'analyse utilises en cherchant un
objectif vasif., l'intention du lgislateur Ces types d'analyses
comprennent l'analyse textuelle, l'analyse fonde sur l'objet vis, le
rgime lgislatif, la position du lgislateur, et l'analyse des
I Ruth Sullivan, of the Faculty of Law, University of Ottawa, Ottawa, Ontario . In
preparing this article, I received many useful suggestions from Carol Diamond, Kate
Murray. Paul Salembier and Katie Wang . I also acknowledge the generous support of the
Social Science and Humanities Research Council of Canada.
52
LA REVIEW DU BARREAU CANADIEN
(Vol .82
circonstances. Cette partie examine aussi la diffrence entre des cas

simples et dociles et se concentre sur les techniques utilises par ceux
qui font l'interprtation pour effectuer les dierents types d'analyse, en
considerant les rgles formelles.
La troisime partie discute de la catgorie d'arguments qui peuvent tre
fonds sur l'analyse prliminaire. Les dbats ne concernent pas toujours
le sens des termes. Nous devons aussi confronter les erreurs du rdacteur
de loi, les lacunes dans le rgime legislatif le chevauchement et
l'incompatibilit entre les provisions, et les termes d'application trop
exhausive ou d'application restreinte. Une dmonstration de la structure
des diffrents types d'argumentation est incorpore dans cette partie.
Part 1 : Introduction . . ... ... ... ... ... . .. ... ... . .. . . . . .. . . . . .. . . .. . . . ... . ... .. ... . ... . .. ... .. . . . . . ..53
Part 11: Preliminary analysis . ... . .. ... ... . ... ... .. . ... . ... .. ... . ... . .. ... ... . ... ... .. . ... . ....56
(A) Textual analysis . . ... . ..... ... . ... ... ... ... . ... .. . .. . ... . .. ... ... . ... ... .. . ... . ....59
(B) Purposive analysis .. .. .. . .. . ... . .. .. . . .. . . .. . . . ... .... .. ... . ... . .. ... ... . ... ... .61
(C) Scheme analysis . . . . .. . . .. ._... ... .. . .... . .. .. . . .. . ... .. ... . ... . .. ... . ... . .. ... .61
(D) Note on extrinsic aids . .. ... . ... . .. ... . .... .. .... .... .. ... . ... . .. ... . .. . ... ... .62
(E) Policy analysis ... .. . ... . ... .. . . .. .. . . .. . . . . .. . . .. . .. .. .. ... ... . . .. . .. ... . .. . ... ....63
(F) Consequential analysis .. . ... ... . .. ... . ... . .. ... ... . ... ..... . .... ....... ... ... .63
Part III: Constructing Argument .. ... .. . ... . ... .. ... . ... . ..,.. ... . ... ... ... . ... ... .. . ... ... . .64
(A) Disputed meaning model .. . ... .. . .. . ... . .. ... ... . ... ... ... . ... ... .. . ... ... . .65
(1) Identifying sources ofdoubt .. . ... .. ... . ... . .. ... .... . .. ... ... ... . .. . .66
(2) Formulating an effective interpretation ... . ... . .. ... ... ... . .. ..67
(3) Establishing a linguistic basis .. . .. . ... . .. . .. .... ... . .. ... ... . .. . .. ..67
(4) Establishing legal argument infavour ofpreferred
interpretation . ., . .. .. . .. . ... . . .. . . . ... . . .. . . . ... . .. . .. .... ... . .. ... ... ... . ....69
(5) Conclusion ... .... ... .,. ... ... . ... .. . ... . ... .. . ... . ... .. . ... . .. . .. ... . ... .., ... .70
(B) Non-application of ordinary meaning... . ... .. .... . .. . .. ... ... . ... .... 71
(C) Avoiding overlap: paramountcy and exhaustive codes.. . ....73
(1) Paramountcy .. . ... . ... .. ... ._ . ... .... ... .. . . .. . .. . ... .. . . . .. . . .. . . .. . .., ... .73
(2) Exhaustive code . . .. . .. ... . ... . .. ... . .... ._... ... .. . ... . .. . .. . .. . . .. . .. ... .75
(D) Filling gaps . ... ... ... ... . ... .. .... ... . .. ... . ... . .. ... . ... . .. .... .... .. . ... ... ... . .. .77
(E) Supplementation ... . .. . ... .. . ... . ... .. . .. . ... . .. ... . . . . . .. .. . . ... . . . .. . .. . . . .. . .. .79
(F) Correcting (drafting) mistakes . ... . ... .. .... ... . .. ... . ... . .. ... ... ... . .. . .80
Part IV: Conclusion .. . . . . .. . .. .. . . ... . .. ... ... . ... ... .. . ... . ... .. . . .. . ... . . . .. . . . . . .. .. . .. . . .. . ....82
20031
Statutory Interpretations in a New Nutshell
53
Part 1: Introduction
In 1938 John Willis published an article called "Statutory Interpretation
in a Nutshell"? Willis' article is a significant contribution to the legal
realist literature of the twentieth century. In addition to being clever and
amusing, it offers a concise exposition ofthe main rules of interpretation,
with special focus on the so-called mischief rule, plain meaning rule, and
golden rule. Willis' thesis is that because these rules are open-ended and
inconsistent with one another, they are incapable of determining the
outcome in statutory interpretation disputes . Although judges purport to
apply them, in fact outcomes are determined by the politics and arbitrary
preferences of the presiding judge. "What use do courts make of the
`plain meaning rule'?" Willis asks. "The answer is that they use it, as they
use all other `rules of construction', as a device whereby to achieve some
desired reSUIL" 3
A current reader of Willis must be struck by how little has changed
since 1938. Judges continue to invoke the same rules, make the same
arguments and rely on the same rhetoric 4 as they did when Willis wrote.
And academics continue to deplore the contradictions and false premises
that disturbed Willis and other legal realists of the days
Yet some things have changed. Since 1938 there has been a
significant evolution in the way we think about rules and their role in
interpretation . For Willis and other legal realists, the inconsistency and
indeterminacy of the rules were unacceptable because these features
undermined their status as law. If the outcome of a dispute is not
determined by applying relevant rules to the facts of the case, then it is
not determined by law. If it is not determined by law, then it must be
determined by the personal preference of the individual judge. Those
appeared to be the only choices .
Most contemporary jurists have a broader conception of law and the
means by which legal disputes are resolved . They think of law as
including not only rules (which are binding) but also principles (which
2 (1939) 16 Can. Bar Rev . l .
3 Ibid. at 11 ; see also at 13 : "What use do the courts make of the `golden rule'today?
Again the answer is the same-they use it as a device to achieve a desired result . . ." and
at 15 : "What use do the courts make of the 'mischief rule'today? Once more, the answer
is that they use it to achieve a desired result . .."
4 In this article I use the term "rhetoric" in a technical rather than a pejorative sense.
Technically, rhetoric refers to the features of arguments that make them persuasive, the
means by which one person persuades another that a conclusion is appropriate and should
be accepted .
5 See, for example, R. Sullivan, "The Plain Meaning Rule and Other Ways to Cheat
at Statutory Interpretation" in Ejan Mackaay, ed ., Les certitudes du droit / Certainty and
the Law (Montreal: Les ditions Thmis, 2000) at 151 .
54
THE CANADIAN BAR REVIEW
[Vol.82
are not) as well as the values, assumptions and practices that contribute to
an evolving legal tradition. This post-realist conception permits
contemporary jurists to integrate statutory interpretation into law by
thinking of it as a principle-governed rather than a rule-governed activity.
As Dworkin might say, statutory interpretation is law because it is an
activity carried out within a practice-based legal principle . 6
This evolution in thinking about statutory interpretation is reflected in
the Supreme Court of Canada's adoption of Elmer Driedger's so-called
"modern principle" . In the first and second editions of The Construction
of Statutes,? Driedger explored the main rules discussed by Willis - the
plain meaning rule, the mischief rule and golden rule - in the first three
chapters of his book. He showed that each was dominant at one time in
history; and each reflects (albeit imperfectly) an important aspect of
interpretation, namely the reader's understanding of the text, the
intentions of the legislature, and harmony with established legal norms.
His thesis was that no single approach is adequale, that interpreters must
take into account and attempt to harmonize all aspects of interpretation.
In the fourth chapter he concluded:
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 8
Since its initial appearance in 1974, this passage has been continually
cited and relied on by Canadian courts . In 1998, in Rizzo v. Rizzo Shoes
Ltd., it was adopted by the Supreme Court of Canada as stating its
preferred approach . 9
While the modern principle is not immune from criticism, 10 it must
be admired for the way it accommodates the post-realist conception of
law as a practice within a tradition. Instead of reducing statutory
interpretation to a silly game of pick-up-rules - or worse, to a shell game
- it describes a purposeful, multifaceted activity . To resolve statutory
interpretation disputes, judges must analyze and integrate a variety of
6 R.M . Dworkin, Law's Empire (Cambridge : Belknap Press, 1986), especially
chapters 2, 7 and 9. Of course, Dworkin is also responsible for the rule-principle
distinction referred to in the text that has become a commonplace of legal discussion . See
R.M . Dworkin, Taking Rights Seriously (Cambridge : Duckworth, 1978), especially at 22
-27.
7 Elmer Driedger, The Construction ofStatutes, tst ed . (Toronto : Butterworths : 1974)
and 2nd ed . (Toronto : Butterworths, 1984).
8 Ibid. at 67 . As Driedger explains at 82 of the 1 st ed . and at 106 of the 2nd ed .,
"intention of Parliament" includes presumed as well as express and implied intent.
9 See Rizzo v. Rizzo Shoes Ltd, [1998] 1 S.C .R . 27.
10 SeeR. Sullivan, "Statutory Interpretation in the Supreme Court of Canada" (199899) 30 Ottawa L. Rev. 178 .
2003]
55
factors, including textual meaning, legislative purpose, acceptable

consequences, and presumptions of intent . The attention paid to these
factors and the amount of emphasis each receives depend on the
circumstances of the case - the type of legislation, the subject matter and
audience, how precise the language is, the lapse of time since enactment,
and the like. On this approach, judges have considerable discretion, but
this discretion is structured and constrained by a principle-based practice
of decision- making .
So what role do the so-called rules of statutory interpretation play in
the current approach? Do those classic objects of realist scorn continue to
serve any purpose in statutory interpretation? Or have rules like ejusdent
generis and the plain meaning rule become obsolete?
In my view, although we no longer imagine that interpretation

disputes are resolved by applying rules to texts, we continue to rely on the
old rules and to develop new ones because they help to structure
interpretation, and they aid communication . The rules operate as a
checklist of relevant considerations . They suggest different lines of
inquiry and ensure that no possibility has been overlooked . They are
relied on by counsel in developing arguments and by judges to justify
outcomes in interpretation disputes . Although they are sometimes used to
disguise choice, or avoid explaining the real basis of choice, they need not
be used in that way. The rules are a medium within which interpreters
work, the bricks used to build a structure. The medium constrains, but
does not control." The controlling factor is the interpreter and the work
that he or she does in producing an opinion, factum or judgement.
In speaking of interpreters I include both judges and lawyers.
Obviously there is a difference in the way these two groups approach
statutory interpretation, particularly in the context of a dispute. Lawyers
are obliged to seek a resolution that favours their client; judges are
obliged to adopt a resolution that accords with the law. To do their jobs
well, however, both groups must engage in the type of analysis and
argument construction described below.
In Part II of this paper I describe the type of analysis that interpreters

should carry out before turning to the business of constructing arguments.
The natural impulse of most interpreters is to minimize this work, or skip
it altogether. It is difficult to resist the siren call of those tried and true
arguments - the meaning is plain, the mischief must be overcome, the
legislature cannot have intended such an absurd result. However, these
arguments are unpersuasive if they are not tied to a sound textual
I t The notion of law as a medium within which judges work to construct decisions
the way builders make structures out of bricks is suggested and developed by D. Kennedy
in "The Phenomenology of Judging" in A. Hutchinson & P. Monahan, eds., The Rule of
Law: Ideal or Ideology (Toronto : Carswell, 1987).
56
LA REVIEW DU BARREAUCANADIEN
[Vol .$Z
analysis, to cogent evidence of legislative intent and to justification of the

norms relied on in calling something absurd . Preliminary analysis enables
the interpreter to:
" identify the sources of doubt, conflict or confusion that have arisen in attempting to
apply the legislation to particular facts
" work out the legislative purpose and scheme
" appreciate which contextual factors are relevant to interpretation and how they
support one outcome or another
In Part III ofthis paper I set out the varying models of argument that
can be used in statutory interpretation and suggest how argument is
structured under each model . My purpose in Part III is to illustrate how
disputes about the meaning of legislation are argued and to emphasize
that there is more to statutory interpretation than such disputes .
Sometimes the meaning is clear, but there is a gap in the legislative
scheme, and the question is whether the court can do anything about it.
Sometimes there is overlap between a clear provision and the common
law, and the issue is whether both apply. Many disputes are about the
circumstances in which a court should update a statute or decline to apply
a legislative rule even though its meaning appears to be clear. In short,
determining the meaning of words in a legislative text is an important task
ofinterpreters, and a necessary first task, but it is only part of the work of
interpretation.
Part II: Preliminary analysis
When an interpreter sits down to deal with an issue in statutory
interpretation, he or she begins with a set of facts. The job is to determine
whether any legislative provision applies to those facts, and if so, to what
effect . To make this determination the interpreter has to answer the
following questions:
" What is the meaning of the legislative text?
" What was the legislature's intent? That is, when the text was enacted, what rule did
the legislature intend to adopt? What purposes did it hope to achieve? What specific
intentions (if any) did it have regarding these facts?
" What would the consequences be if a proposed interpretation were adopted? Would
these consequences violate important, legally recognized norms?
The interpreter asks these questions because, as Driedger's modern

principle tells us, these are the determinants of outcome in statutory
interpretation.
If all goes well, the meaning of the text (as understood by the
interpreter) coincides with the legislature's intention (as reconstructed by
the interpreter) and produces results that (in his or her view) accord with
2003]
57
important, legally recognized norms . This convergence of factors makes

for an easy case . All an interpreter need do is persuade others that this
analysis ofmeaning, intention and norms is sound.
Suppose, for example, that the provision to be interpreted is a

prohibition against bringing vehicles into a park.lz John (looking for a
friend) drives his truck into a large green space that has no roadways,
rutting the grass and crushing flower beds until he stalls in the sand under
the children's swings . Assuming the green space is a park, this is an easy
case. Although the word "vehicle" has fuzzy edges and can sometimes be
hard to apply, cars are the very prototype of vehicles . A car is certain to
be within the definition of "vehicle" in any ordinary language dictionary.
Furthermore, driving is a form of "bringing" when it comes to cars . As for
legislative intent, there are many reasons why a legislature might want to
keep cars out of parks : the safety of the occupants; protection of the
grounds from damage; creation of an exhaust-and noise-free
environment. However the purpose of the legislation is defined, John's
incursion into the park is clearly within the mischief. Finally, keeping
John's car out of the park would not violate any important legal norms.
In harder cases, one or more of the factors relied on in interpretation

proves unhelpful. For example, the meaning of the legislative text is
obscure. Or the legislature's intentions are unknowable, contradictory, or
too vague to offer much insight in the circumstances. Or the relevant
norms point in different directions . In such cases, the interpreter
appropriately gives less weight to the weak or indeterminate factors and
more to the factors (if any) that seem clear and compelling .
Suppose, for example, that instead of driving a truck into the park,
John rode a bicycle. It is not self-evident that a bicycle is (or is not) a
vehicle. It depends on the context. If a landlord asks a tenant whether he
needs a parking space for his vehicle, chances are "vehicle" does not
include bicycle. But if an employer asks a prospective employee whether
he has a vehicle to get him to work, "vehicle" could well include bicycle .
Similarly, it is not self-evident that John's incursion into the park is within
the contemplated mischief. While bicycle riding can threaten the security
ofpedestrians (or wandering children or pets), bicycles do not create foul
smells or loud noises . Depending on the circumstances, they may or may
not create a significant threat to the landscape. To resolve a case such as
this, the interpreter must try to establish the legislature's intention and
then work out the implications of that intention for bicycles in parks.
In the hardest cases, two or more factors are strongly suggestive but
they suggest different outcomes . In such cases, the apparent meaning of
lz This example is one of H .L .A . Hart's enduring contributions to legal philosophy.
See H .L .A . Hart, The Concept ofLaw (New York : Oxford University Press, 1961) at 8995 .
sa
wol.a2
the text differs from the meaning that was probably intended by the
legislature . Or applying the text as written leads to anomalous
consequences that were probably unforeseen by the legislature . In such
cases the interpreter is forced to choose an outcome, ideally on a
principled basis, which he or she can both articulate and support .
Suppose, for example, that John drove an ambulance into the park, in
response to an emergency call, to save the life of a seriously injured child .
An ambulance is clearly a vehicle . Moreover, it has the potential to do
damage to grass and flower beds ; it produces exhaust and more than its
share of disruptive noise; and it is a threat to the safety of pets and
pedestrians who may be using the park. Nonetheless, most people would
agree that the ambulance should have legal access to the park. It would
be absurd to exclude it, The legislature must have taken it for granted that
rational interpreters would read down the prohibition to exclude obvious
exceptions such as these.
The point of these examples, like the point of Driedger's modern
principle, is that in every case, interpreters must explore all aspects of
interpretation . Before deciding which factors to rely on and how they will
construct their arguments, interpreters must know whether the case is
easy or hard, and if hard, why and how it is hard. They must identify
sources of doubt or conflict and the relevant indicators of legislative
meaning and intent; and they must assess the weight of the competing
factors. Preliminary analysis of this sort is necessary to ensure sound
judgement and persuasive argument .
There is no standard way of classifying the different types ofanalyses
that must be carried out at the preliminary stage . In the list below I
describe them under the headings textual analysis, purposive analysis,
scheme analysis, poliv analysis, and consequential analysis . I also
include a short comment on the use of extrinsic aids. Others might prefer
fewer, or more, or different categories.13 The important thing is not how
this preliminary work is categorized, but how it is carried out. The goal is
to be as comprehensive and as explicit as possible .
Being explicit is particularly important . Interpretation normally
operates at a sub-conscious level. Upon reading the words of a text, a
reader may intuitively reach certain conclusions - the text is clear (or
unclear), it expresses a particular meaning or purpose, it leads to absurd
13 My categories are not mutually exclusive . For example, scheme analysis overlaps
with purposive analysis on the one hand and textual analysis on the other; policy analysis
overlaps with purposive and consequential analysis. The descriptions of the different
categories of analysis and of the rules relied on by interpreters in carrying out these
analyses is found in R. Sullivan, Statutory Interpretation (Concord, OnU irwin Law,
1996) and R. Sullivan, Sullivan and Driedger on the Construction of Statutes, 4th ed.,
2002.
2003]
Statutory Interpretations in
a New Nutshell
59
(or reasonable) results. But legal interpreters cannot afford to rely on

intuition alone . To be persuasive, they have to expose the basis of their
intuitive judgements - the chain of inferences that leads to their
conclusions about meaning or purpose or consequence : and the
assumptions upon which those inferences are based.
(A) Textual analysis
In textual analysis, the interpreter focuses on the language of the
legislative text and the conventions of legislative drafting . The task is to
show how a particular meaning emerges from reading the words to be
interpreted in their immediate context and the more extended literary
context consisting of the Act as a whole and any other related legislation .
Reference may also be made to the legislation of other jurisdictions, to
treaties or agreements, or to model legislation .
The first and most fundamental rule of legislative drafting is ordinary
language use. Drafters choose their words and arrange them in sentences
using the same lexicon and grammar, and relying on the same knowledge,
as the public that is governed by the legislation. In legislation like the
Criminal Code, this is the general public, the "average person" on the
street . 14 In legislation aimed at a specialized audience-manufacturers of
aircraft, for example - this is a limited segment of the public with
specialized knowledge and a technical vocabulary. In either case, the idea
is to use language that conforms to the standards of usage of competent
speakers within the audience addressed .
This fundamental idea underlies a number of interpretative rules,
including the ordinary meaning rule, the technical meaning rule and the
plausible meaning rule:
" Ordinary meaning rule: the meaning that spontaneously comes to the mind of a
competent reader upon reading the legislative text is presumed to be the meaning
intended by Parliament . This meaning governs unless the evidence suggests some
other meaning was intended .
" Technical meaning rule: when legislation deals with a specialized subject and uses
language that people governed by the legislation would understand in a specialized
way, that specialized understanding is preferred over ordinary usage.
14 In fact, of course, the average person or the general public is a convenient fiction .
The audience to which legislation is addressed, particularly legislation like the Criminal
Code, is diverse in terms of culture, gender, race, and the like. These differences
undermine traditional assumptions about "common" meaning and "common" sense.
60
[Vol.82
" Plausible meaning rule : if the ordinary meaning is rejected to give effect to the actual
or presumed intentions of the legislature, the meaning adopted must be one thewords
are capable of bearing.
To determine the ordinary meaning of bilingual legislation, reference

must be made to both language versions and the same meaning must be
assigned to each. Both the English and French versions of legislation are
authentic and equally authoritative at the federal level and in New
Brunswick, Quebec, Ontario and Manitoba. If the two versions share a
particular meaning, that shared meaning is presumed to be intended .
However, the presumption is rebutted if other evidence shows that some
other meaning - one not shared by the two versions - better reflects the
legislature's intent .
While drafters observe the conventions of ordinary language use,
they also rely on conventions that reflect the special status and function
of legislation. For example, they write in a formal and impersonal style
and try to avoid figurative or decorative language . They choose language
with unusual care and use it consistently, without stylistic variation. They
say things in a straightforward fashion, using as few words- as possible.
These conventions form the basis for the following presumptions relied
on in analyzing legislative texts:
" Straightforward expression : the legislature chooses the clearest, simplest and most
direct way of stating its meaning ("had the legislature intended [the rejected
interpretation], it would have used other words")
" Uniform expression: the legislature uses the same words and techniques to express
the same meaning and different words and techniques to express different meanings
: there are no superfluous
" No tautology ("the legislature does not legislate in vain")
words in legislation; every feature of the text has an identifiable role in the legislative
scheme
" Internal coherence: all the provisions of a legislative text fit together logically and
work together coherently to achieve the purposes of the legislature
These presumptions are the basis of a number of standard

interpretation rules, including:
" Implied exclusion (expressio unius est exclusio alterius): if something is not
mentioned in circumstances where one would expect it to be mentioned, it is
excluded by implication
" Associated words (noscitur a sociis) : the meaning of a word is affected by the other
words to which it is linked in a sentence
20031
61
" Limited class (ejusdem generis) : when general language follows a series of more
specific terms, the class of things referred to by the general language may be read
down to refer to a narrower class of things to which the specific terms all belong .
(B) Purposive analysis
In purposive analysis, the interpreter identifies the objects that the

legislature wanted to achieve in enacting its legislation - the practical
impacts it hoped to produce, the policies, attitudes and behaviour it
wished to promote, or the evils or "mischief' it wished to cure.
Sometimes the purpose of legislation is narrowly focussed and specific ;
this is often the case in amending legislation, for example . Normally,
legislation seeks to promote a range of purposes, some competing and
some complementary .
The challenge in purposive analysis is threefold : (1) to put words to
each purpose - that is, to articulate each one in an express formulation
that is rhetorically effective ; (2) when there are two or more purposes, to
work out the relationship among the purposes and, if necessary, to rank or
balance them; and (3) to provide evidence that the legislature intended to
pursue the purposes identified and to rank or balance them as suggested .
Legislative purposes are often described in a preamble or purpose
statement or discussed in extrinsic materials like a government
commission report . Such material provides direct evidence of the
legislature's intentions . Alternatively, purpose may be inferred by reading
the legislative text in its external context, taking into account the
circumstances existing at the time of enactment and the conditions in
which the legislation was expected to operate . Purpose can also be
inferred quite compellingly from amendments introduced to the
legislation over the years .
When inferring the purpose oflegislation, interpreters rely heavily on
so-called "common sense", the store of knowledge and values that are
presumed to be shared by everyone . Interpreters can infer that a
prohibition against vehicles in parks is designed to protect safety, create
a pollution- free environment, and prevent the destruction of grass and
flowers because they know what "everybody knows" - the sort of thing
that parks are used for, what vehicles typically do, and the need for safe
havens for children's play and adult leisure activity.
(C) Scheme analysis
In scheme analysis, the interpreter explores how the legislation is
intended to operate so as to bring about desired goals . The idea is to
determine how each provision of the legislation was meant to relate to the
others and work with them in a purposeful and coherent fashion .
62
(Vol .82
Whereas in textual analysis the legislation is approached as a literary

genre, with a specialized set of conventions for conveying meaning, in
scheme analysis it is approached as a means to an end. The fundamental
premises are that :
" the legislature has devised a coherent and effective implementation plan
" each provision has a role to play in implementing that plan
" a provision's role in the scheme controls its meaning and scope
Scheme analysis often turns out to be the most compelling indicator

of legislative intent and it often provides a sound basis for preferring or
rejecting a particular interpretation . The only drawback to this form of
analysis is that it requires considerable effort . The interpreter has to read
the entire Act, and perhaps other Acts as well . He or she has to work out
the legislature's purposes and its envisaged implementation plan, then
work out how the provision to be interpreted contributes to that plan, and
finally work out how to communicate that analysis to the audience . This
last step is particularly difficult when the scheme in question is complex
or subtle or presupposes specialized knowledge.
(D) Note on extrinsic aids
Textual, purposive and scheme analysis are often assisted by socalled "extrinsic aids", including the following:
" Legislative sources: consists of conventions, treaties or model legislation that
legislation is designed to implement or on which it has been based in whole or in part .
" Legislative history: consists of statements formally brought to the attention of the
legislature during thelegislative process, including Ministerial statements, committee
reports, recorded debate and tabled background material.
" Legislative evolution: consists of the successive amendments and reenactments a
provision has undergone from initial enactment to time of application; subsequent
evolution is excluded .
" Expert opinion : consists of precedent, administrative opinion, and scholarly legal
publications, as well as expert testimony.
The rules governing the admissibility and use of this material are
complex and in a state of flux . Generally speaking, legislative sources and
legislative evolution are always admissible, as is precedent. Legislative
history and other forms of expert opinion are subject to specialized rules.
20031
63
(E) Policy analysis
In policy analysis, the interpreter considers the values, principles and

concerns the legislature is presumed to respect when it enacts legislation.
Whereas purposive analysis focuses on actual intent-the goals that were
self-consciously pursued by proponents of the legislation - policy
analysis focuses on values and concerns that are presumed to inform
every legislative initiative. In purposive analysis, the interpreter must
tender evidence that establishes legislative intent as a matter of fact. In
policy analysis, the interpreter relies on law, or the evolving legal
tradition, to establish presumed intent.
In carrying out a policy analysis, interpreters rely on the following
presumptions, each of which reflects particular values or norms:
" strict construction of penal legislation
" strict construction of legislation that interferes with individual rights
" strict construction of exceptions to the general law
" liberal construction of human rights codes
" liberal construction of remedial legislation
" liberal construction of social welfare legislation
" liberal construction of legislation relating to Aboriginal peoples
" presumed compliance with constitutional law
" presumed compliance with the rule of law
" presumed compliance with international law
" presumed continuation of common law
" presumed non-interference with common law rights
presumption against the extra-territorial application of legislation
" presumption against the retroactive application of legislation
" presumption against interference with vested rights (both common law and statutory)
" presumption against applying legislation to the Crown and its agents
The list is not closed. In recent years, Canadian courts have added the
following policy-based presumptions :
" the legislature intends to combat the feminization of poverty
" the legislature intends to protect the environment from pollution
(F) Consequential analysis
In consequential analysis, the interpreter predicts and evaluates the

consequences of adopting a particular interpretation. Consequences that
arejudged to be good are presumed to have been intended . Consequences
that are judged to be absurd or otherwise unacceptable are presumed not
to have been intended .
64
[Vol.8Z
An interpretation may be considered absurd if it has one or more of

the following effects:
" it creates irrational distinctions (treating like things differently or different things the
same way)
" it defeats the purpose of the legislation
" it leads to incoherence, contradiction or anomaly
" it undermines the efficient administration ofjustice
" it violates important norms of justice or fairness
The clearer and more precise a text seems to be, the greater the
absurdity required to depart from its ordinary meaning. The greater the
absurdity that flows from a particular interpretation, the more justified an
interpreter is in rejecting it.
Having carried out these different types of analysis, the interpreter
must then consider how the results can be used to support a preferred
position (in the case of litigants) or to explain andjustify the correct legal
result (in the case of judges). This involves choosing an appropriate
model of argument and developing persuasive supporting arguments.
Part III: Constructing Argument
This part surveys the various models of argument that can be used to
structure disputes about the proper interpretation of a legislative text . It is
based on the following framework :
If the problem is. . .,
the model of argument is . . .
inaccurate or garbled text
corrigible mistake
ambiguous, vague or incomplete

text
disputed meaning
over-inclusive text
non-application
under-inclusive text
incorrigible gap in legislative

scheme, or supplementation with
common law rule or remedy
overlapping provisions
(no conflict) : exhaustive code

(conflict) : paramountcy
2003]
65
In a corrigible mistake argument, the interpreter claims that the

provision in question contains a drafting mistake, which must be
corrected before determining whether the provision applies to the facts,
He or she must establish what the legislature clearly intended and what
the text would have said had it been properly drafted .
In a disputed meaning argument, the interpreter claims that,
properly interpreted, the provision in question has the following meaning :
[here the interpreter sets out an appropriate paraphrase or definition or
test.] He or she must establish that this is the ordinary meaning, or the
intended technical meaning, or at least a plausible meaning . If this
interpretation is adopted, the provision obviously applies (or does not
apply) to the facts .
In a non-application argument, the interpreter identifies a reason

not to apply a provision to the facts even though, given its ordinary
meaning, it would otherwise apply . For example, the provision may be
read down to promote the purpose or to avoid absurdity or to comply with
presumed intent .
In a incorrigible gap argument, the interpreter claims that the
legislation as drafted cannot apply to the facts even though, given its
purpose, it should apply. Whether this omission is deliberate or
inadvertent, the court has no jurisdiction to fill a gap in a legislative
scheme or otherwise enlarge the scope of legislation .
In a supplementation argument, the interpreter concedes that the
legislation as drafted does not apply, but claims that the common law
applies so as to supplement the under-inclusive legislation .
In the absence of conflict, if two or more provisions apply to the same
facts, each is to be applied as written . In an exhaustive code argument,
the interpreter concedes that the overlapping provisions are not in conflict
but claims that one of them was meant to apply exhaustively, to the
exclusion of the others .
In a paramouncy argument, the interpreter claims that the
overlapping provisions are in conflict and one is paramount over the
others .
(A) Disputed meaning model
Whether legislation applies to particular facts, and how it applies,
depend fundamentally on what the legislation "says" : the content of the
legal rule as declared by the words of the text . A majority of disputes in
statutory interpretation are understood as disputes about the meaning of
particular words in the legislative text .
To mount an effective argument using the disputed meaning model, a
number of steps should be followed and a number of claims should be
66
wol.a2
made . To illustrate this approach, thejudgement of the Supreme Court of

Canada in R. v. Hasselwander 15 will be used. The dispute arose in
Hasselwander when the defendant's Mini-Uzi sub-machine gun was
seized under s. 102 of the Criminal Code, 16 which authorized the seizure
and forfeiture of so-called prohibited weapons. As defined in s. 84(1) of
the Code, "prohibited weapon" included what is colloquially known as
automatic weapons, meaning:
any firearm . . . that is capable of toute arme feu . . . pouvant tirer
fixing bullets in rapid succession rapidement plusieurs balles
during one pressure of the pendant la dure d'une pression
trigger
sur la dtente
In the condition in which it is manufactured and sold, a Mini-Uzi submachine gun cannot fire bullets in rapid succession during a single
squeeze of the trigger. However, once a certain part is removed, it
becomes a fully automatic weapon . The defendant claimed the seizure
was unlawful because the Mini-Uzi was not an automatic weapon as
defined in the Code .
(1) Identifying sources ofdoubt
The interpreter's first step is to identify which words in the relevant
provision are possible candidates for dispute. These are words or
expressions whose meaning is doubtful in relation to the facts of the case .
In any attempt to apply legislation to facts, some things will be perfectly
obvious or at least reasonably clear, but there may be one or two words
whose meaning or scope or import in relation to the facts is doubtful .
Because ofthis doubt, it is unclear whether the legislation applies or how
it applies. These are the words in dispute on which argument will focus .
To identify potential sources of doubt, the interpreter tests how

clearly each word in the relevant provision applies (or does not apply) to
the facts of the case. In Hasselwander, no one-not even the defendant
could plausibly deny that the Mini-Uzi was a "firearm" as defined in the
Code . 17 Similarly no one-not even the Crown-could insist that the gun,
at the moment of seizure, was one that "fired bullets in rapid succession
during one pressure of the trigger." However, it was established at trial
that converting the Mini-Uzi to a fully automatic weapon was easy and
inexpensive, an adjustment that any gun-owner could make . In light of
that fact, it was not so obvious whether the appellant's Mini-Uzi was
15 R. v. Hasselwander, [199312 S .C .R . 398.
16 R.S .C . 1995, c. C-46 .
2003]
67
"capable of firing bullets in the relevant way. On this point the appellant
and the Crown could each make plausible, but conflicting claims . These,
then, were the words of doubtful application.
(2) Formulating an effective interpretation
The second step is to develop an interpretation of the doubtful
language that effectively clarifies whether the legislation applies and
leads to the preferred result .ls Such an interpretation usually takes the
form of a paraphrase or definition, or in some cases a set of criteria that
operates as a legal test. The formulation should be as precise and concise
as possible and should be set out as the initial claim in the argument : "In
this provision, properly understood, the words in dispute mean . . ."
Coming up with an effective definition, paraphrase or set of criteria
is perhaps the most challenging and important task in interpretation. It is
here that the skill and imagination of counsel can make a real difference.
In the Hasselwander case, the defendant claimed that the phrase "is
capable of means "has a current or existing capacity to do a thing" . Since
at the moment of seizure the Mini-Uzi had no existing capacity to fire
bullets in rapid succession during one pressure of the trigger, it was not a
prohibited weapon within the meaning of the Code. The defendant wisely
included the word "is" in identifying the language in dispute to
emphasize existing as opposed to future capability.
Not surprisingly, the Crown focussed more narrowly on the words

"capable of'. It argued that this expression refers to a potential as well as
an existing capacity to do a thing. It put forward a test of potentiality,
consisting of criteria to be met for a semi-automatic firearm to be
considered "capable of fully automatic discharge: the conversion to
automatic discharge must be quick, easy and inexpensive for the average
gun-owner. Relying on a test of this sort is often helpful when the
doubtful language is vague or otherwise confers discretion on those who
apply the legislation. Even though the criteria adopted here are neither
precise nor objective, they offer a measure of guidance and they create a
sense of impersonal evaluation. Thus, from a rhetorical perspective, the
formulation is effective .
(3) Establishing a linguistic basis
The third step is to work out a sound linguistic basis for the preferred
interpretation. Because of the presumption in favour of ordinary meaning,
17 "Firearm" is defined in s . 84(l) of the Code as a barrelled weapon
from which a
bullet can be discharged and which is capable of causing serious bodily harm .
18 For counsel, the preferred result is the one that favours the client; for a court, it is
the result that thejudge believes to be legally correct
68
(Vol.82
interpreters should if possible claim that their preferred interpretation

corresponds to the ordinary meaning of the text. If that claim cannot be
made, they may try to establish a technical or legal meaning. At the least,
they must try to show that their interpretation is a meaning the text can
plausibly bear.
It is surprisingly difficult to come up with evidence to support a claim
about the ordinary or plausible meaning of words . Dictionaries are a
readily available source and interpreters rightly rely on them. However,
relying on dictionaries fosters a tendency to focus on single words as
opposed to phrases, which may not always be appropriate, and to equate
ordinary meaning with dictionary definition, which is never appropriate .
Dictionaries can indicate the range ofplausible usage. But the meaning of
words in a text is the meaning of its words in that particular context. For
this reason interpreters often start with a dictionary but go on to refine the
definition in light of relevant contextual factors . This approach was used
by Cory J. in the Hasselwander case:
The word "capable" as it is defined in the Oxford English Dictionary (2nd ed . 1989)
includes an aspect of potential capability for conversion . . . .
. . . [T]he French definition as well clearly conveys the idea that the word "pouvant"
includes a potential which has yet to be realized . . .
Yet, that potential aspect must be given some reasonable restriction . It is the proper role
of the court to define the meaning of "capable" as it is used in the definition of
"prohibited weapon" in s . 84(l) . In my view, it should mean capable of conversion to
an automatic weapon in a relatively short period of time with relative case . 19
Apart from dictionaries, interpreters can appeal directly to the linguistic

intuitions of their audience by citing (or making up) analogies or related
contexts in which the disputed words clearly bear the preferred meaning.
In Hasselwander, for example, counsel for the defendant might have put
the following scenario to the court. Suppose that a person looking to buy
Mr. Hasselwander's gun on the day it was seized said, "Well, the MiniUzi looks to be in good condition, but is it capable of firing
automatically?" To avoid misrepresentation, Mr. Hasselwander would
have to answer no to this question. It is hard to see how the same gun can
be simultaneously both capable and not capable of firing automatically.
Authoritative assertions of ordinary or plausible meaning can
sometimes be found in case law. Finally, interpreters may try to introduce
the expert testimony of a professional linguist. Normally, the ordinary
and legal meaning of words is judicially noticed and the courts receive
expert opinion evidence only to establish the technical meaning ofwords .
19
Supra note 15 at 415-16 . Emphasis added .
2003]
69
In recent years, however, professional linguists have developed new

techniques to establish contextual word meaning and their expertise could
be of assistance to the courts? 0 Of course, the admissibility of such
evidence would have to be established .
(4) Establishing legal argument infavour ofpreferred interpretation
The final step in mounting a disputed meaning argument is to show
that the preferred interpretation fits the context and overall scheme of the
Act, promotes relevant purposes, avoids unacceptable consequences and
is supported by extrinsic evidence of legislative intent. This is where the
interpreter draws on the various analyses carried out at the preliminary
stage.
In flasselwander, counsel for the defendant relied heavily on textual
analysis to support his interpretation . He pointed out, for example, that in
other Criminal Code provisions dealing with illegal weapons, when
Parliament wanted to address the potential for alternation or adaptation,
it used different and more pointed language - such as "adapted" and
"can be made capable of' . Given the convention of consistency in
legislative drafting, he argued that the court should infer that Parliament
did not intend to deal with the possibility for adaptation when it used the
contrasting expression "is capable of' .
The defendant also relied on scheme analysis . He pointed out that
both before and after the 1977 amendment, the Criminal Code
distinguished two categories of firearms : restricted and prohibited . One
purpose of the 1977 amendment was to transfer automatic weapons from
the restricted to the prohibited category, subject to this consideration :
semi-automatic weapons registered as restricted weapons when the new
legislation came to force were to be grandfathered. Under the former
legislation, automatic weapons were defined as "any firearm that is
designed, altered or intended to fire bullets in rapid succession during one
pressure of the trigger" . This description clearly excluded semi-automatic
firearms such as Mr. Hasselwander's, which had only a potential capacity
to fire automatically. Under the new regime to which Mr. Hasselwander
was subject, restricted weapons were redefined as follows :
Restricted : any firearm that is designed, altered or intended to fire bullets in rapid
succession during one pressure of the trigger and that on January 1, 1978, was
registered as a restricted weapon ....
Prohibited : any firearm that is capable of firing bullets in rapid succession during one
pressure of the trigger
20 For an excellent example, see C. Cunningham et al ., "Plain Meaning and Hard
Cases", (1994) 103 Yale L.J . 1561 .
70
lvol.8z
Under this scheme, counsel argued, weapons that were designed, altered
or intended to fire automatically were merely restricted before 1978 but
after that date the same weapons became prohibited unless they were
registered as restricted on January 1, 1978. In order for this scheme to
work, the class of weapons described as "capable of firing automatically
must be coextensive with the class described as "designed, altered, or
intended" to fire automatically. Otherwise, those who owned convertible
semi-automatic weapons at the start of 1978 would have had to give up
their weapons whereas those with filly automatic weapons could have
kept them - an obvious absurdity. To avoid this absurdity, the class of
firearm that is capable of firing automatically must be limited to firearms
actually altered and not just capable of being altered to fire automatically.
Finally, the defendant's counsel relied on both purposive and policy
analysis . He pointed out that the goal of protecting the public was
adequately served by including semi-automatic weapons in the category
of restricted weapons. He also pointed out that penal legislation
traditionally receives a restrictive interpretation. If there is a choice
between a narrow and a broad interpretation of disputed language, the
narrow interpretation should prevail.
Counsel for the Crown relied primarily on purposive and
consequential analysis, and its arguments were adopted by a majority of
the Court. Cory J. wrote:
What then should "capable" mean as it is used in the s. 84(1) definition of prohibited
weapon? It should not be restricted to the narrow meaning of immediately capable.
Such a definition would mean that the simple removal of a part, which could be
replaced in seconds, would take the weapon outside the definition . This surely could
not have been the intention of Parliament. If it were, the danger from automatic
weapons would continue to exist just as strongly as it did before the prohibition was
enacted .zl
Parliament's purpose was to protect the public from weapons whose only
purpose, Cory J. suggested, is to wound or main people . Given the
pressing importance of this purpose, and the minimal social utility of
permitting ordinary citizens to possess such weapons, achieving the
purpose was assigned considerable weight - enough to defeat the textand-scheme-based arguments of the defendant.
(5) Conclusion
if the interpreter has done her job of identifying the disputed
language, interpreting it, and supporting her interpretation on both
linguistic and legal grounds, the result will be obvious. Once the court in
21 Supra note 15 at 415 .
20031
71
Hasselwander decided that the words "is capable of meant "have a

potential that can be met in a relatively short period of time with relative
ease", it was obvious that the appellant's semi-automatic was within the
definition of "prohibited weapon".
(B) Non-application ofordinary meaning
A provision is over-inclusive when, taking its ordinary meaning, it
applies to facts to which, for one reason or another, it should not apply.
The courts respond to over-inclusive provisions by reading them down,
declaring implied qualifications or exceptions or otherwise declining to
apply the provision to particular facts. Much of the law of statutory
interpretation is about identifying legitimate reasons for invoking one of
these non-application techniques .
Some non-application arguments are grounded in purposive analysis .
The interpreter argues that the legislature intends its legislation to apply
only in so far as it promotes the purposes of the legislation. Applications
outside that purpose are impliedly excluded, especially if the legislation
limits freedom or interferes with private property. Other non-application
arguments are grounded in consequential analysis . In Re Vabalis? 2 for
example, a married woman applied to change her name from Vabalis to
Vabals under Ontario's Change of Name Act. Section 4(1) of the Act was
in the following terms:
A married person applying fora change of surname shall also apply for a change of the
surnames of his or her spouse and all unmarried minor children of the husband or of
the marriage .
Since Ms. Vabalis had not adopted her husband's name when she married,
applying this provision to her would have required her husband, whose
surname was different, to change his name to Vabals . Here is how the
court responded :
In our opinion the double application requirement contained in s. 4(1) should be
construed as being applicable only where the married person applying for the change
of surname uses th same surname as the spouse. We are all agreed that the literal
interpretation of s. 4(1) as requiring a change of name of the applicant's spouse in the
present situation would lead to an obvious absurdity. A statute enacted by the
Legislature of this province should not be so interpreted.23
Many non-application arguments are grounded in common law

presumptions of legislative intent . Several of these presumptions
explicitly address the application of legislation - for example, the
22 (1983), 2 D.L.R. (4th) 382 (Ont . C.A .),
23 Ibid. ai 383-84 .
72
(Vol.82
presumptions against retroactive or extra-territorial application and

against interference with vested rights. Others address the relation of
legislative provisions to other sources of law - for example, the Charter
or international law. The remaining presumptions address common law
values that the legislature is presumed to respect, most notably private
property and the rule of law. If applying a provision to facts would violate
one of these presumptions, the interpreter may argue that, given the
absence of evidence to the contrary, the legislature did not intend the
provision to apply to those facts.
In Re Goodman and Manitoba Injuries Compensation Board,24 for
example, the appellant claimed an entitlement to compensation on the
ground that he had been injured while trying to protect his wife from a
criminal assault. Under s . 6(I) of Manitoba's Criminal Injuries
Compensation Act, 1970, if injury resulted to a person while he "was
endeavouring to. . . prevent lawfully the commission of a criminal
offence", the person was entitled to compensation .
The Manitoba Court of Appeal conceded that the husband's attempt
to thwart the attack on his wife was an endeavour to lawfully prevent the
commission of a criminal offence, but it nonetheless concluded that the
provision did not apply because this endeavour had occurred while the
couple was vacationing in Nevada. The Court said :
It would be unthinkable that the legislature of Manitoba intended to give a right of
compensation to all citizens of Russia, China, Jamaica, Uganda, etc ., who are injured
in those countries as a result of endeavouring to preserve the peace . If the subsections
are to have any reasonable construction, it is necessary to imply some qualification :
either they are limited by the residence of the applicant or they are limited by the place
of occumence .z5
Because non-residents injured in Manitoba were expressly entitled to

apply for compensation under the legislation, the court concluded that the
qualification must be place of occurrence . Therefore, residents of
Manitoba injured outside Manitoba were not entitled to compensation .
Notice that the issue in the Vabalis and Goodman cases was not the
meaning ofthe words used in the legislation. On their face, the provisions
in question in these cases clearly applied. However, the appellant in
Vabalis and the Commission in Goodman were able to provide
convincing legal reasons for non-application .
Two other grounds for non-application are (1) conflict between two
overlapping provisions, such that it is impossible for both to apply, and
(2) a provision constitutes or is part of an exhaustive code, such that all
24
(1980), 120 D .L.R. (3rd) 235 (Man . C.A .) .

at 238 .
25 Ibid.
2003]
73
other provisions are impliedly excluded . These grounds are examined

below.
(C) Avoiding overlap: paramountcy and exhaustive codes
In Canada, persons wishing to determine their rights and obligations
must look to federal and provincial statute law, to regulations, by-laws
and other forms of delegated rule-making, and to private law rules
governing such matters as contract, civil liability and restitution. Even
though these rules occupy different positions in the legal hierarchy,
assuming they are validly made, all are presumed to apply in accordance
with their terms . There is nothing impermissible or surprising in attaching
more than one legal consequence to a given set of facts . Subjects are
obliged to comply with every applicable obligation and prohibition, and
may equally claim every benefit to which they are entitled under the law .
The presumption that overlapping rules apply regardless of overlap
can be rebutted in two ways . First, if there is a conflict between two
applicable rules, such that it is impossible to comply with both, then
obviously one must prevail over the other. It would be an intolerable
violation of the rule of law if citizens were subject to contradictory rules .
Second, even in the absence of conflict, if a rule-maker intends its rule to
apply to the exclusion of other rules of equal or lesser status in the
hierarchy, the courts must give effect to this intention . In such cases, the
court concludes that the rule in question (or the regime of rules of which
it is a part) was intended to be an exhaustive "code" .
(1) Paramountcy
An interpreter who wishes to use the paramountcy model of argument
must establish two things : first, that a conflict exists between two
applicable provisions or rules, and second, that one of them is paramount
and therefore applies to the exclusion of the other.
Establishing that there is conflict between legal rules is relatively
straightforward : two rules are in conflict if and only if it is impossible to
comply with both? 1) Establishing which rule is paramount in cases of
conflict is rather more challenging . Over the years the courts have
established a number of paramountcy rules . Some are based on hierarchy :
" entrenched constitutional law prevails over legislation

" federal legislation prevails over provincial legislation
" legislation prevails over common law
26 Friends of the Oldman River u Canada (Minister of Transport), [1992] 1 S .C .R .

3 at 39- 39 .
74
[Uo1.82
These paramountcy rules have the force of genuine rules: they are binding
on the courts .
Other rules establish paramountcy among legislative provisions :
"
"
"
"
primary legislation prevails over delegated legislation

quasi-constitutional legislation prevails over ordinary legislation
specific provisions prevail over general ones (specialibus generalia non derogant)
subsequently enacted provisions prevail over previously enacted ones
Unlike the rules based on hierarchy, these paramountcy rules are based on
presumed legislative intent. They are therefore liable to be displaced by
evidence of actual intent revealed through textual, purposive, scheme,
policy and consequential analysis .
A good example of the paramountcy model of argument is found in
R. v Greenwood.27 In that case an accused was charged with sexual
assault and sexual interference involving a thirteen-year-old child. At the
preliminary inquiry, the only evidence adduced was the unsworn
testimony of the child. The issue was whether this evidence was
admissible . Section 16(3) ofthe Canada Evidence Act provided :
A person referred to in subsection (1) [i .e . a person under 14 years of age] who does
not understand the nature of an oath or a solemn affirmation but is able to communicate
the evidence may testify on promising to tell the truth.
After questioning the child, the provincial courtjudge concluded that the
child did not understand the nature of an oath or a solemn affirmation, but
did understand the importance of telling the truth in the circumstances.
He therefore ruled that the child's evidence was admissible, given the
ordinary meaning of s. 16(3).
Counsel for the accused did not challenge this interpretation of s.
16(3), but argued that the section did not apply because it was superceded
by s. 540(l)(a) ofthe Criminal Code. That provision required evidence at
a preliminary inquiry to be given under oath :
Where an accused is before ajustice holding a preliminary inquiry, the justice shall
(a) take the evidence under oath....
It was not difficult to establish a conflict between the two applicable

rules. A witness cannot both be required to give evidence under oath and
be permitted to give unsworn testimony. The challenge was to establish
which of the two rules should be regarded as paramount in the
circumstances of the case. Since both rules were established by
27 (1992), 7 0.2 (3rd) 1 (C .A.).
2003]
75
enactments of the federal Parliament, they occupied the same place in the
legal hierarchy.
Counsel for both parties tried to rely on the specialibus rule: when
two provisions conflict, the more specific one impliedly creates an
exception to the more general one. Counsel for the accused focussed
narrowly on the conflicting provisions themselves. He argued that s.
16(3) of the Canada Evidence Act should be considered general because
that provision applies to proceedings of all kinds whereas s. 540 of the
Criminal Code should be considered specific because it is limited to
preliminary hearings. Conversely, counsel for the Crown focussed
broadly on the two statutes as a whole. He argued that the Canada
Evidence Act should be considered the specific legislation because it
deals with the specialized subject of evidence in contrast to the Code,
which deals with crime in general . He could also have argued, focussing
on the two conflicting provisions, that s. 16(3) of the Canada Evidence
Act is specific because it deals with the special case of minors whereas
the Criminal Code provision is general because it applies to witnesses in
general .
In the end, the court concluded that the Canada Evidence Act
provision should prevail . However, the basis for this conclusion was not
the specialibus rule but rather the court's purposive and consequential
analysis . Griffiths J.A . wrote:
If s. 540(1)(a) of the Code were literally interpreted28 to exclude the application of s.
16(3) of the Canada Evidence Act, then this would lead to a highly undesirable and
anomalous result, which surely could not reflect the intention of Parliament . To deny
the applicability of s. 16(3) to the preliminary inquiry is to defeat its manifest purpose,
most especially in cases of sexual assault and cognate crimes against children 29
After pointing out other unacceptable consequences that would flow from
assigning paramountcy to the Code, Griffiths J.A. concluded that s. 16(3)
of the Canada Evidence Act was meant to operate as an exception to the
general rule set out s. 540(1)(a) of the Code.
(2) Exhaustive code
An interpreter may wish to rely on an exhaustive code model of
argument in either of the following circumstances :
" provisions from different statutes apply without conflict to particular facts, or
" legislative provisions and the common law apply without conflict to particular facts.
28 Griffiths J. says "literally interpreted" but he means "applied". The problem here
is not meaning, but conflict among applicable rules.
29 Supra note 27 at 8.
76
LA REVIEW DU 13ARREAU CANADIEN
[Vol.82
Although in these circumstances the overlapping provisions and rules are

presumed to apply, this presumption can be rebutted by establishing that
the legislature intended one of its provisions (or series of provisions) to
deal exhaustively with the facts in question, to the exclusion of
everything else.
To establish that intention, interpreters rely on the types of analysis
described in Part 2 : textual, purposive, scheme, policy, and consequential .
They focus in particular on the following considerations :
" whether the provision is part of a scheme that is detailed and comprehensive and
appears to be complete
" whether the provision provides an adequate solution or remedy to the facts of the case
" whether resorting to other provisions or the common law would merely duplicate the
effect of the provision
" whether resorting to other provisions or the common law would interfere with the
scheme of the Act or tend to defeat its purpose
Because rules made by the legislature prevail overjudge-made rules, it is

relatively easy to establish that legislation is meant to exclude the
common law . To the extent legislation changes the common law, it
operates like an amendment : in effect it "repeals" inconsistent common
law. However, when legislation codifies, or partially codifies, the
common law, the outcome can be difficult to predict, for the court must
determine whether the legislation was intended to definitely displace the
common law.
It is harder to establish that the provisions of one statute are meant to
exclude the provisions of other statutes issuing from the same legislature .
However, if there is evidence that this is what the legislature intended, the
courts will rule accordingly. In Re British Columbia Teachers'Federation
and Attorney General for British Columbia, 30 for example, British
Columbia's Court of Appeal found that the province's School Act was a
complete code which excluded the application of other legislation .
The issue in the case was whether the Treasury Board of British
Columbia had authority to issue a directive that reduced appropriations to
any local school board that reduced the number of teachers it employed
below a certain number and gave the remaining teachers a raise. On the
basis of s. 24(1) of the Financial Administration Act, it appeared that a
directive along these lines was authorized:
24.(1) The Treasury Board may
(a)
by directive, control or limit expenditure under any appropriation ;

(b) by directive, . . . set conditions for any kind of expenditure under an appropriation .
30 (1985) 23 D .L .R. (4m) 161 (B .C .C .A .) .
2003]
77
However, the Court of Appeal ruled otherwise. Anderson J.A . for the
majority offered the following explanation:
The above quoted provisions [from the Financial Administration Act], although framed
in wide and general language, do not, in my opinion give the power to issue directives
in respect of matters specifically dealt with by the School Act.
... the School Act provided a comprehensive code dealing with local self-government
in the public education field.... [T]he very matters dealt with by the directive were
specifically reserved to school boards subject to the collective bargaining and
arbitration process outlined in the School Act31
Anderson J.A. emphasized several features of the School Act suggesting

that it should be treated as an exhaustive code. The first was its
constitutive character - it established an elaborate institutional
framework and conferred powers on various offices or agencies . In
addition, it dealt with the school board matters in a comprehensive
fashion. Every aspect of running a school system was covered by the Act:
there were no gaps. The Act set out general principles governing the
operation oflocal schools as well as detailed provisions for implementing
those principles . Finally, the Act dealt with an important area of
governmental activity that historically had been dealt with at the local
level. On the basis of these features, the court concluded that the School
Act was meant to be exhaustive and that s. 24 of the Financial
Administration Act could therefore not be used to issue directives to
school boards .
(D) Filling gaps
Gaps arise when the purpose of legislation is clear, but the means
devised to achieve the purpose are inadequate because a provision is
under- inclusive or there is a lacuna in the legislative scheme. In contrast
to mistakes, which are always inadvertent, gaps may or may not be
inadvertent; they may in fact reflect the legislature's intent . Thus, unlike
mistakes, which are always blamed on the drafter and which can be fixed
by courts, gaps in legislation are attributed to the legislature itself and
courts have no jurisdiction to fill them .
The key consideration in distinguishing incorrigible gaps from
corrigible mistakes is the difference between under-inclusion on the one
hand, and the other types of deficiencies that may occur in drafting
legislation on the other. In order to fill a gap, a court must enlarge the
scope of a provision or add a new provision to the legislative scheme - a
31
Ibid. at 173 and 175-76 .
78
[Vol.82
new right, obligation or power, or an additional curtailment of freedom.

Only the legislature has the constitutional authority to subject citizens to
new prohibitions or obligations, create new entitlements or confer new
powers on agents of the state.
In Pacific Press Ltd. v. Cain,32 for example, the issue was whether a
coroner had the power to find a witness at an inquest in contempt if the
witness refused to answer a question. Section 38(3) of the Coroner's Act
provided :
A coroner has the same powers to compel the attendance of witnesses and to punish a
witness for disobeying a summons to appear, refusing to be sworn, or refusing without
lawful excuse to give evidence as are conferred on a justice by the Offence Act.
However, the powers conferred on a justice by the Offence Act were

limited to punishment for failure to attend or remain in attendance and did
not extend to punishment for failure to answer a question .
Counsel pointed out that the legislature clearly intended coroners to
have the power to compel witnesses to answer questions, and this power
would be incomplete and useless without the power to punish for refusal .
Although the court conceded that this might be so, it was nonetheless
unwilling to fill the gap:
Although it seems that the legislature might have intended that the coroner have some
sort of power to compel a witness to give evidence, the [statute as drafted does not
confer] the intended power. That gap clearly might be regarded as a drafting error, but
[there is nothing] which justifies the court filling the gap rather than waiting for a
legislative remedy. 33
Even though the legislative intent was fairly clear, and even though the
failure to achieve that intent was probably due in fact to a mistake by the
drafter, the court could not correct it because to do so would effectively
confer a new power on the coroner - a power to send journalists to jail .
Not surprisingly, the court held that it lacked jurisdiction to do this .
Notice how paramountcy apd exhaustive code arguments differ from
the disputed meaning model of argument. Whereas the disputed meaning
model addresses doubt about the meaning of a provision in relation to
particular facts, paramountcy and exhaustive code arguments are
concerned with the relationship among legal rules .
32 (I997), 147 D .LK (4th) 339 (B .C .S .C.) .

33 Ibid. a t 357 .
2003]
(E)
79
Supplementation
Although courts cannot fill gaps in legislation, in appropriate cases

they can supplement an under-inclusive provision or scheme by resorting
to the common law. Supplementation is clearly inappropriate if the
provision or scheme is meant to be exhaustive . Conversely,
supplementation is clearly appropriate if the subject matter of the
legislation is within the inherent jurisdiction of the court. Legislation
dealing with the welfare of children, for example, is often supplemented
by invoking the parens patriae jurisdiction . Similarly, procedural
legislation may be supplemented when a court invokes its inherent
jurisdiction to control its own process.
Even outside the areas of inherent jurisdiction, the courts on occasion
appeal either to established common law rules or to their jurisdiction to
add to the common law (within appropriate limits) as justification for
supplementing a defective legislation scheme . A rather striking example
occurred in R. v. Jobidon,34 which concerned the scope of the offence of
assault as defined in s. 265(1) o the Criminal Code:
265(1) A person commits an assault when
(a) without the consent of another person, he applies force intentionally to
that other person, directly or indirectly; . . .
In interpreting this language, Canadian courts traditionally incorporated

various common law limitations on consent, including the rule that a
person did not consent to the application of force by voluntarily engaging
in a fight or brawl . In 1983, the common law limitations on consent were
partially codified in the following provision, added to s. 265 :
(3) For the purpose of this section, no consent is obtained where the complainant
submits or does not resist by reason of

(a) the application of force to the complainant or to a person other than
complainant;
(b) threats or fear of the application of force to the complainant or to a person other
than the complainant;
(c) fraud; or
(d) the exercise of authority.
The appellant argued that by including only some common law

limitations on consent and leaving out the others, Parliament clearly
indicated its intention to enlarge the scope of the offence. The voluntary
fight limitation was impliedly excluded .
This was a plausible argument. Supposing Parliament intended to

34 (199112 S.C.R. 714.
80
[Vol.82
retain the full panoply of common law limitations, it clearly failed to

implement that intention when it enacted s. 265(3) . By mentioning some
but not all of the limitations, it created a gap in the intended scheme .
However, the court did not purport to fill this gap by reading in an
additional limitation. Rather it applied existing common law to
supplement the legislative scheme . Of course, the end result was the
same. Gonthier J. wrote:
While at first glance the appellant's argument may seem cogent, it is ultimately
unpersuasive . Parliament did not set foot into new territory when listing the four
vitiating factors in s. 265(3) . On the contrary it will be seen that, for the most part, that
list merely concretized, and made more explicit, basic limits on the legal effectiveness
of consent, which had for centuries formed part of the criminal law in England and in
Canada. Their expression in the Code did not reflect an intent to remove the existing
body of common law, which already described those limitations and their respective
scope. 35
In effect, the Court here relied on supplementation to fill a gap.

(F) Correcting (drafting) mistakes
Although in principle legislation should not contain mistakes, in
practice mistakes are commonplace. To rely on the corrigible mistake
model of argument, an interpreter must establish one of the following
claims :
(1) The wording of the text is contradictory or incoherent and it is impossible to make
sense of it using the standard techniques of interpretation . The court must redraft
the provision so as to give it some sensible application.
(2) The wording of the text states or suggests one thing, but it is clear that the
legislature intended something else. This discrepancy between meaning and
intention is due to a drafting mistake . The court has jurisdiction to correct this
mistake by notionally redrafting the text so as to bring it in line with the clear
intention of the legislature.
An example of a court assigning meaning to an incoherent text is found

in the judgement of the New Brunswick Court of Appeal in Association
ofParentsfor Fairness in Education, Grand Falls District, 50 Branch u
Minority Language School Board No.50 et aL 36. New Brunswick's
Schools Act provided that a school board organized on the basis of one
official language could not provide classes in the other official language
to "persons of the other official language / des personnel parlant fautre
35 Ibid. at 739.
36 (l987), 40 D.L.R. (4ch) 704 (N .B .C.A .)
2003]
81
langue officielle". Angers J.A. found it impossible to make sense of this

language, in part because the English and French versions established
different bases for the categories : "the French version of the Act identifies
people on the basis of their phonetic characteristics, whereas the English
version does so by their being members of an ethnic group ."37 More
importantly, neither basis was workable because of the large bilingual,
bicultural population in New Brunswick . This group of people would not
fall into either class . Here is how Angers J.A. dealt with the problem:
There are rare circumstances where the court may remove or substitute words so that
the meaning of an Act is intelligible and to ensure that it can be put to practical use . . .
I believe that such circumstances exist in the case before us .
Thus, if we remove . . . the words "for persons of the other official language", words
which are of no practical use, the section reads as follows :
3 .3 . . . a school board (organized on the basis of one official language) shall not provide
classes . . . in which the language of instruction is (the] other official language . . . 8
An example of a court bringing the text in line with what the legislature
obviously intended is found in the judgement of the Ontario Court of
Appeal in R. v. Creaghan 39 The defendant was charged with mischief
under the Criminal Code and relied on what was then s. 386(2) as a
defence:
No person shall be convicted of an offence under sections 387 to 402 where he proves
that he acted with legal justification or excuse and colour of right .
Although the defendant could not establish that he acted with legal
justification or excuse, he did establish to the satisfaction of the court that
he acted with "colour of right" . In the view of the court, despite the
wording of s. 386(2), it was not necessary for the defendant to establish
that he acted not only with colour of right but also with legal justification
or excuse :
We are all of the view that the word "and" which precedes the words "with colour of
right" in s. 386(2) should be read as "or" . Manifestly, it would not be sensible to
require the accused to prove not only that he acted with legal justification or excuse
and with colour of right . If the accused acted with legal justification or excuse he is not
criminally liable and that is the end of the matter and there is no need to resort to colour
of right 40
37 ibid. at 712 .
38 1bid at 713-14 .
39 (I982), 31 C .R . (3rd) 277 (Ont. C .A.) .

40 Ibid at 282 .
82
[Vol .82
Notice that the claim being made by the interpreter who relies on this
model of argument is not that courts are entitled to amend legislation or
disregard the will of the legislature. The claim is that, due to incompetent
drafting, the provision as written fails to express the true intent of the
legislature, and if the court is to give effect to that intention, it must repair
the drafting mistake . In order to establish the true intention of the
legislature, the interpreter relies on the various forms of analyses
canvassed in Part 2 and on any relevant extrinsic aids .
Part IV: Conclusion
In this article I have tried to show that statutory interpretation disputes are
resolved, not by arbitrary preference, but by work. During the preliminary
analysis stage, the interpreter analyzes the legislation to locate sources of
doubt and conflict . He or she then identifies reasons to resolve the doubt
or conflict one way as opposed to another. During the argument
construction stage, the interpreter decides which model of argument is
best and formulates the claims associated with that model . He or she then
relies on the preliminary analyses of text, purpose, scheme, policy and
consequences in support of the claims.
t have also tried to show that there is more to statutory interpretation
than argument about the meaning of text and whether it is plain or
ambiguous . Determining whether legislation applies, and how it applies,
depends on all admissible evidence of legislative intent, both actual and
presumed. While some interpretation problems can be dealt with in one
way only, many lend themselves to more than one model of argument. In
such cases, the choice of model is strategic: the interpreter adopts the
model that best supports his or her preferred outcome . This aspect of
interpretation is nicely illustrated by the judgement ofthe Supreme Court
of Canada in R. v Mclntosh,41 which considered the self-defence
provisions of the Criminal Code . The majority used the disputed meaning
model of argument and purported to decide the case on the basis that the
ordinary meaning was plain. The dissent relied on the corrigible mistake
model of argument.
It is obvious that interpreters enjoy significant discretion when they
argue and resolve interpretation disputes . However, they are obliged to
justify what they do by showing how their preferred outcome relates to
the meaning of the text, the intentions of the legislation and evolving
legal norms. These considerations determine the outcome in easy cases,
where they all point in the same direction . In hard cases, they limit the
range of acceptable outcomes. These considerations also underlie the
rules of statutory interpretation, which, properly used, do not disguise
choice but expose it and justify it in a persuasive way.
41 l19951 1 S .C.R . 686 .

Statutory Interpretation in A New Nutshell

Uploaded by

Copyright:

Available Formats

Statutory Interpretation in A New Nutshell

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Statutory Interpretation in A New Nutshell

Uploaded by

Copyright:

Available Formats

STATUTORY INTERPRETATION IN

LA REVIEW DU BARREAU CANADIEN

circonstances. Cette partie examine aussi la diffrence entre des cas

Statutory Interpretations in a New Nutshell

THE CANADIAN BAR REVIEW

Statutory Interpretations in a New Nutshell

factors, including textual meaning, legislative purpose, acceptable

In my view, although we no longer imagine that interpretation

In Part II of this paper I describe the type of analysis that interpreters

analysis, to cogent evidence of legislative intent and to justification of the

The interpreter asks these questions because, as Driedger's modern

Statutory Interpretations in a New Nutshell

important, legally recognized norms . This convergence of factors makes

Suppose, for example, that the provision to be interpreted is a

In harder cases, one or more of the factors relied on in interpretation

THE CANADIAN BAR REVIEW

(or reasonable) results. But legal interpreters cannot afford to rely on

LA REVIEW DU BARREAU CANADIEN

To determine the ordinary meaning of bilingual legislation, reference

These presumptions are the basis of a number of standard

Statutory Interpretations in a New Nutshell

In purposive analysis, the interpreter identifies the objects that the

THE CANADIAN BAR REVIEW

Whereas in textual analysis the legislation is approached as a literary

Scheme analysis often turns out to be the most compelling indicator

Statutory Interpretations in a New Nutshell

(E) Policy analysis

In policy analysis, the interpreter considers the values, principles and

In consequential analysis, the interpreter predicts and evaluates the

LA REVIEW DU BARREAU CANADIEN

An interpretation may be considered absurd if it has one or more of

If the problem is. . .,

the model of argument is . . .

inaccurate or garbled text

ambiguous, vague or incomplete

incorrigible gap in legislative

(no conflict) : exhaustive code

Statutory Interpretations in a New Nutshell

In a corrigible mistake argument, the interpreter claims that the

In a non-application argument, the interpreter identifies a reason

THE CANADIAN BAR REVIEW

made . To illustrate this approach, thejudgement of the Supreme Court of

To identify potential sources of doubt, the interpreter tests how

Statutory Interpretations in a New Nutshell

Not surprisingly, the Crown focussed more narrowly on the words

LA REVIEW DU BARREAU CANADIEN

interpreters should if possible claim that their preferred interpretation

Apart from dictionaries, interpreters can appeal directly to the linguistic

Supra note 15 at 415-16 . Emphasis added .

Statutory Interpretations in a New Nutshell

In recent years, however, professional linguists have developed new

THE CANADIAN BAR REVIEW

Statutory Interpretations in a New Nutshell

Hasselwander decided that the words "is capable of meant "have a

Many non-application arguments are grounded in common law

LA REVIEW DU BARREAU CANADIEN

presumptions against retroactive or extra-territorial application and

Because non-residents injured in Manitoba were expressly entitled to