You are on page 1of 26

Law and Philosophy

 Springer Science+Business Media Dordrecht 2014


DOI 10.1007/s10982-014-9206-9

FBIO P. SHECAIRA

GARDNER ON LEGAL REASONING

(Accepted 20 January 2014)


ABSTRACT. In Chapters 2, 3 and 7 of his new book, Law as a Leap of Faith, John
Gardner provides the elements of an account of legal reasoning. It is on the basis of
this account that Gardner defends or supports some of the most important theses
of his book, viz. theses pertaining to how law can be made, to the relation
between law and morality, and to the legitimacy of judicial law-making. A central
element of Gardners account is a distinction (suggested originally by Joseph Raz)
between two forms of legal reasoning, namely, reasoning about the law and reasoning according to law. In this paper I intend to describe and evaluate Gardners
account. Among the critical remarks that will appear in the paper is the claim that
Gardners concept of reasoning according to law is overly inclusive.
I. INTRODUCTION: GARDNER AND RAZ

John Gardner claims that his new book, Law as a Leap of Faith: Essays
on Law in General, does not contain a theory of law. The book is
supposed to collect quite a lot of thoughts about law in general,1
with only the modest ambition that those thoughts will turn out to
be consistent with one another. The books project is also described
as that of unbundling, i.e. separating out disparate thoughts that
have often been regarded, mistakenly, as part of some package
deal.2 The unbundling enterprise is undertaken without commitment to any firm philosophical agenda and hence without scruples
about searching for ideas across many supposedly rival positions and
traditions.3 Gardner sums up his approach to jurisprudence by
means of a familiar metaphor: I am such an intellectual squirrel
(way beyond a mere fox) that I dont care at all where or how I get
1
2
3

John Gardner, Law as Leap of Faith (Oxford: Oxford University Press, 2012), p. v.
Ibid.
Ibid., p. vi.

FBIO P. SHECAIRA

the little truths that I hoard away in the following pages. I may even
occasionally snatch them from under the hedgehogs very nose.4
Gardner deserves praise for his piecemeal, anti-dogmatic, and
non-adversarial ambitions. Of course, this is not to say that he
remains perfectly faithful to his avowed ambitions throughout the
book. In the preface, just a few pages after describing himself as an
intellectual squirrel, Gardner provides a summary of the principal
theses appearing in his book5:
(1) All law is made by people, but not all law is made intentionally, or even
knowingly, and in particular not all law is made by legislating.
(2) Some sound legal reasoning (usually by the higher courts) is capable of
making new law, often accidently.
(3) The content of all law is the content it was given by its makers and
changers (including those who make and change law by applying it),
never mind what content they ought to have given it.
(4) This includes the law that determines who counts as an ultimate maker
or changer of law (what Hart calls rules of recognition).
(5) There are therefore no moral criteria (necessary or otherwise) for
establishing what the law on any given matter says.
(6) Nevertheless, there are necessary (conceptually necessary) connections
between law and morality.
(7) A necessary connection: law by its nature holds itself out as morally
binding, even though that may be a mistake or a pretense.
(8) Another necessary connection: legal reasoning is moral reasoning with
one or more legal premises.
(9) Also, there is a moral ideal for law, the ideal of legality or the rule of
law, laws answerability to which is conceptually determined (i.e. is a
part of the very concept of law).
(10) And there is also, although more indirectly, a conceptual connection
between law and justice.
(11) Both the connection between law and legality, and that between law
and justice, confirm that law is a modal as opposed to a functional kind;
i.e. it is distinguished by how it does what it does, not by why.
(12) All of this, and more, is true of law in general.
4

Ibid., p. vi (footnote omitted).


Ibid., p. ix. What follows is nearly a quotation of Gardners text. I have omitted Gardners
parenthetical references to the chapters where the reader will find a defense or affirmation of each thesis
in the list.
5

GARDNER ON LEGAL REASONING

Recall that Gardner claims to have no theory of law in his book.


Either Gardner is using the word theory in an unusually narrow sense
or he is too modest to realize that his list of theses is more cohesive and
comprehensive than many sets of claims that other contemporary legal
philosophers would happily announce as genuine theories of law. Kevin
Toh has said not only that Gardner has a theory of law but that the
theory has a very recognizable provenance: [T]here is a theory of law
that provides a framework for Gardners thinking, and it is the theory
that Joseph Raz has constructed over the years.6 Toh sees this as a
problem. As someone who finds Razian views uncongenial, Toh
declares himself astonished at how often a substantial point made by
Gardner, of either philosophical or interpretive nature, is accompanied
by references to Razs works, and at how very seldom Gardner registers any disagreements with Raz.7
Astonished is a strong word. Gardner may not be quite the squirrel
he says he is in the preface to his book, but his Razian sympathies are
not surprising, and, for the most part, they do not detract from the
plausibility of his positions. Gardner has relied on Razs work consistently for many years. Nine of the eleven essays collected in Law as a
Leap of Faith were previously published, and they do not cite Raz more
often or more deferentially now than they did the first time they
appeared. Moreover, Gardner has a way of presenting Razs views
more clearly and forcefully than Raz himself, and then of using those
views fruitfully in the development of other ideas.
This is what happens, for instance, with Gardners treatment of
Razian claims about legal reasoning in particular, the suggestive distinction between reasoning about the law and reasoning according to
law.8 Gardner takes Razs distinction, re-describes it elegantly and
clearly, and then puts it to use in defense of three important theses. Two
of the three theses are identified as (2) and (8) in the list above; they
pertain respectively to how law can be made and to how the concept of
law relates to that of morality. The third thesis call it the legitimacy
thesis states that courts can make law in deciding cases without

6
Kevin Toh, Review of Law as Leap of Faith, Notre Dame Philosophical Reviews (2013), http://ndpr.
nd.edu/news/39497-law-as-a-leap-of-faith/.
7
Ibid.
8
Joseph Raz, Ethics in the Public Domain (revised edition, Oxford: Clarendon Press, 1995), pp. 326
340.

FBIO P. SHECAIRA

thereby usurping legislative powers or violating the ideal of the rule of


law, so long as they make law by reasoning according to law.9
My goal in this paper is to evaluate and suggest some ways of
improving Gardners account of legal reasoning. It is a plausible
account that goes a long way towards establishing the three foregoing theses. Yet in several respects it needs to be corrected or at
least made more precise. Section 2 begins to explain Gardners account. Section 3 considers how the account is used in support of
thesis (2), thesis (8), and the legitimacy thesis. Since considering
Gardners arguments for these theses is an important part of
understanding his account of legal reasoning, the aims of Section 3
are more explanatory than critical, although Gardners case for thesis
(8), in particular, is the target of some criticism. Section 4 continues
to asses Gardners account and identifies a set of problems pertaining
to the concepts that make up the definitions of reasoning about the
law and reasoning according to law (namely, premise, redundancy, and
decisiveness). Section 5, the final section, takes a step back and discusses a jurisprudential assumption motivating Gardners account.
His commitment to legal positivism is not criticized, but his inattention to the phenomenon of theoretical disagreement is put in
question.
II. GARDNERS ACCOUNT OF LEGAL REASONING

Let us begin with some preliminary remarks about Gardners approach to the study of legal reasoning. First, Gardners account of
legal reasoning is really an account of legal argument. Like many
other legal philosophers, Gardner is interested in discussing how
legal actors, especially judges, publicly give reasons in support of
their decisions. There is little discussion in Gardners book of legal
actors private reasoning processes, the characteristic elements of
which may or may not be fully disclosed in public legal argument.10
Second, Gardners approach to legal reasoning is primarily logical
(as opposed to dialectical or rhetorical). He does not focus on the
9
The legitimacy thesis is not mentioned in Gardners preface in the same list as (2) and (8), but it
plays an important part in Chapter 2, where Gardner exposes various myths that have surrounded legal
positivism.
10
The clearest indication that Gardner equates legal reasoning with legal argument is to be found at
the end of the book, in the index of subjects, where there is an entry for legal reasoning (or legal
argument). Gardner (2012), p. 309.

GARDNER ON LEGAL REASONING

interaction between legal arguers or on the standards that govern


their (often protracted) exchanges of reasons, objections, counterobjections, etc. The focus is rather on the internal structure of
arguments understood as discrete sets of claims composed of premises and conclusions. Another way to put the point would be to
emphasize that Gardners topic is legal argument, not legal argumentation (i.e. debate, discussion, dialogue).
Third, Gardners account of legal reasoning is essentially nonevaluative. His major aim is to explain what counts as legal argument,
not to inquire into what it takes for legal argument to be good or bad.
Thesis (2) above makes reference to sound legal reasoning, but one
gets the impression from Gardners essays that genuine or true could
be substituted for sound without distortion. It is probably the case
that Gardner also regards his examples of genuine legal arguments as
being sound in some further epistemic or logical sense. But he does not
present careful arguments to that effect; and the soundness of which
he speaks, even if interpreted to mean that the relevant arguments are
(inter alia) logically meritorious, should not be taken to imply deductive validity. For, as we will see, Gardners discussion of legal argument
focuses on cases of argument from analogy and argument weighing
reasons for and against a conclusion. Arguments of these types are
often non-deductive and thus cannot be evaluated by reference to the
standard notion of soundness.11
Moreover, although Gardner commits to the legitimacy thesis (as I
baptized it earlier), he does not present it as being entailed by his
account of the defining features of legal reasoning. In his discussion of
11
To be sure, the precise status of these arguments is a matter of debate; but not many people will
assert today with confidence that analogical arguments are deductive (for my own position, that
analogical arguments are partly deductive compounds, see Fbio Shecaira, Analogical Arguments in
Ethics and Law: A Defence of a Deductivist Analysis, Informal Logic 33(3) (2013)); and those who wish
to assert the deductive nature of pro and con arguments have to grapple with Carl Wellmans powerful
case to the effect that they are conductive, i.e. neither deductive nor inductive. Carl Wellman,
Challenge and Response: Justification in Ethics (Carbondale: Southern Illinois University Press, 1971).
A terminological note is in order. A deductive argument is understood here as an argument formulated
by someone who (rightly or wrongly) believes in its deductive validity (see Mark Vorobej, Defining
Deduction, Informal Logic 14(2) (1992)). To call an argument deductive is not to give it praise, for a
deductive argument can fail to be deductively valid. In the same way, an inductive argument may be
inductively weak. Calling an argument deductive or inductive (or conductive or abductive or whatever)
does not settle the question of its logical merits; although it does settle the question of the standards by
reference to which the argument should be appraised. It is futile, if not wrong, to dismiss as deductively
invalid an argument formulated by someone whose avowed aim is to establish the conclusion as
probable given the evidence. In saying that analogical and pro and con arguments are (often) nondeductive I am not suggesting that they are (often) bad but only that they (often) are not to be judged
according to deductive standards.

FBIO P. SHECAIRA

legitimacy, Gardner argues that legal reasonings capacity to produce


new legal norms does not render it indistinguishable from the sort of
unrestrained, all-things-considered sort of reasoning that is characteristic of (non-subordinate) legislators. No claim is made to the effect that
the exercise of legal reasoning is legitimate just in virtue of its defining
characteristics. Gardners intention is to insulate judicial law-making
from attacks upon its legitimacy that assume that judges can only make
law by engaging in all-things-considered deliberation.
To sum up, Gardners discussion of legal reasoning in Law as a
Leap of Faith is chiefly a non-evaluative study of the logical structure
of legal arguments (i.e. premise-and-conclusion sets). One central
feature of Gardners account is a distinction between two forms of
legal reasoning. First there is reasoning about what legal norms
already apply to [a] case12 reasoning about the law and then there
is reasoning that has already-valid legal norms among its major or
operative premises, but combines them non-redundantly in the same
argument with moral or other merit-based premises13 reasoning
according to law.
The contrast between already-valid [i.e. existing] legal norms
and merit-based norms is consistent with Gardners endorsement of
a positivist account of the validity of legal norms.14 According to
positivism,15 a norm can only be said to be (already) legally valid if it
can be identified independently of merit-based reasoning. That
includes: (i) being able to locate the relevant source-materials (e.g.
statutes, precedents, customs) independently of merit-based reasoning (i.e. by relying on the customary practices of officials in the
relevant jurisdiction); and also (ii) being able to derive a norm from
those source-materials by means of some merit-free mode of interpretation (e.g. by looking to existing conventions of interpretation,
or to some persons or constituencys actual understanding, etc.16).
The following is a simple example of legal argument used by
Gardner call it Example 1:
12

Gardner (2012), p. 39.


Ibid.
14
Ibid., pp. 1923.
15
Or, at any rate, to Gardners interpretation of it.
16
Ibid., p. 47. Gardners contrasting examples of merit-based modes of interpretation include
looking to what would make the norm morally defensible, or more fit for its intended purpose, etc.
Ibid.
13

GARDNER ON LEGAL REASONING

Tortfeasors are liable to pay full reparative damages to those whom they tortiously
injure;
Jones tortiously injured Smith to the tune of $50;
therefore, Jones is liable to pay Smith $50 in reparative damages.17

Example 1 is a case of reasoning about the law, since its single


operative premise expresses an already-valid legal norm.18 There is
nothing in the argument as formulated to indicate that the norm at
issue is legally valid; the example simply assumes that the norm is to
be found in some conventional source-material (e.g. a judicial decision) by appeal to some non-evaluative interpretative procedure (e.g.
by reference to the explicit ratio of the judicial decision). To bring
out this assumption, one would have to add at least one more line to
the argument, thus transforming it into a composite argument made
up of two linked inferences. For instance, Example 1A:
A judicial ruling by a high court was issued on the explicit ground that tortfeasors
are liable to pay full reparative damages to those whom they tortiously injure;
therefore, tortfeasors are liable to pay full reparative damages to those whom they
tortiously injure;
Jones tortiously injured Smith to the tune of $50;
therefore, Jones is liable to pay Smith $50 in reparative damages.19
17

Ibid., p. 186.
It is with some hesitation that I classify Example 1 as a case of reasoning about the law. Recall that
Gardner explains reasoning about the law as reasoning about what legal norms apply to a case, i.e. as if
its purpose were only to establish a conclusion that the abstract content of the law is such and such at a
particular time. But Example 1 seems to go further: it indicates how a particular legal dispute ought to
be settled, given that the law is such and such. When Raz distinguishes reasoning about the law and
reasoning according to law sometimes he seems to have a distinction between theoretical and practical
argument in mind (e.g. Legal reasoning is reasoning either about what the law is or about how legal
disputes should be settled according to law. Raz (1995), p. 327). I do not think, however, that Gardner
follows Raz in this respect. Gardner uses the about/according terminology in Chapter 2. He does not
use it again in Chapter 7, where Example 1 appears, but in the latter chapter he contrasts Example 1
with instances of legal argument where cases are adjudicated not only on the basis of existing legal
norms but also with appeal to merit-based norms. In other words, in Chapter 7 Gardner contrasts
Example 1 with arguments that fit the definition of reasoning according to law provided in Chapter 2.
All it takes to conclude that, for Gardner, Example 1 is a case of reasoning about the law is the
assumption that Chapters 2 and 7 are continuous in their reflections about the different forms that legal
reasoning can take. In any event, if I misclassify Example 1, the main points of this paper will not be
significantly affected. The paper focuses on the idea that legal reasoning, in one of its possible forms,
combines legal and non-legal premises in the way described by Gardner. My claims about that
description would need but slight restatement were Gardner to object to my presentation of his
account.
19
The first inference is not valid as formulated. For it to become valid, a further (normative)
premise would have to be added to the effect that the explicit grounds for the rulings of the high court
are authoritative or otherwise binding.
18

FBIO P. SHECAIRA

For the sake of simplicity, Gardner leaves out of his examples the
sub-arguments that would be necessary to establish the (sourcebased, merit-independent) validity of the legal norms functioning as
operative premises.20 The next example is one of reasoning according
to law (i.e. reasoning that combines valid legal norms non-redundantly with moral or otherwise non-legal norms). This type of reasoning is more complex than reasoning about the law, not because
the source-based validity of the legal norms is expressly argued for
but because of the way in which the norms that are (ex hypothesi)
legally valid interact with other norms that are not legally valid.
Consider Example 2:
Tortfeasors are liable to pay full reparative damages to those whom they tortiously
injure [a legal norm];
Jones tortiously injured Smith to the tune of $100;
but the tort was also the breach of a contract between Jones and Smith;
the contract provided for maximum reparative damages of $50 for any breach;
contracts and the limits on damages they set are legally binding as between parties
to the contract [a legal norm];
and it is unjust to let someone avoid a legally binding contractual limit on damages
by instead suing the other contracting party in tort [a moral norm];
therefore, Jones is liable to pay Smith only $50 in reparative damages.21

Example 3 below is a further example of reasoning according to


law that appears in Gardners book. Again, a legal norm occurs as a
non-redundant yet non-decisive operative premise alongside a moral
premise:
The Civil Rights Act of 1964 gives everyone the legal right not to be discriminated
against in respect of employment on the ground of his or her sex (a legal norm);
denying a woman a job on the ground of her pregnancy is morally on a par with
discriminating against her on the ground of her sex, even though there is no exact
male comparator to a pregnant woman that would allow the denial to count as
sex-discriminatory in the technical sense (a moral claim);
20
In Section 5 I will touch on the question of whether Gardners simplification may have distortion
as its price.
21
Gardner (2012), p. 187.

GARDNER ON LEGAL REASONING

therefore, women have a legal right not to be denied a job on the ground of their
pregnancies (a new legal norm);
now, this woman P has been denied a job by D on the ground of her pregnancy;
therefore, D owes P a job.22

Like Example 1A, Example 3 is a composite argument with two


inferences. But in contrast to Example 1A, its composite nature has
nothing to do with the fact that an explicit argument is given in
support of the validity of the legal norm figuring as a premise. The
first inference making up Example 3 establishes a new legal norm by
appeal to a moral norm and an existing legal norm. The second
inference applies the new legal norm to the facts of the case. The first
premise of the argument should give us pause. It is supposed to
express a norm, and yet it is formulated in a way that makes it look
like a factual claim describing the content of a source-material
(namely, the Civil Rights Act of 1964). This is a minor blunder which
likely comes of Gardners wish to indicate the source-based provenance of the legal norm against discrimination on the ground of sex
(in contrast to the merit-based provenance of the new legal norm
pertaining to discrimination on the ground of pregnancy). The best
way to indicate the provenance of the former would be to add
another premise to the argument, which would then become a
composite of three inferences. Consider Example 3A:
The Civil Rights Act of 1964 states that everyone has the legal right not to be
discriminated against in respect of employment on the ground of his or her sex;
therefore, everyone has the legal right not to be discriminated against in respect of
employment on the ground of his or her sex (a legal norm);
denying a woman a job on the ground of her pregnancy is morally on a par with
discriminating against her on the ground of her sex, even though there is no exact
male comparator to a pregnant woman that would allow the denial to count as
sex-discriminatory in the technical sense (a moral claim);

22
This is not an exact quotation of Gardners text. I have made slight changes to Gardners
formulation at pp. 3940 in order to render the form of Example 3 analogous to that of Examples 1, 1A,
and 2.

FBIO P. SHECAIRA

therefore, women have a legal right not to be denied a job on the ground of their
pregnancies (a new legal norm);
now, this woman P has been denied a job by D on the ground of her pregnancy;
therefore, D owes P a job.23

III. THE USES OF GARDNERS ACCOUNT

What I have called Gardners account of legal reasoning is composed of three elements: a distinction between two forms of legal
reasoning namely, reasoning about the law and reasoning according
to law , a definition of each form of legal reasoning,24 and a small
set of examples of legal arguments. Example 1 is a case of reasoning
about the law; Examples 2 and 3, cases of reasoning according to
law.25 We will come back to these examples. For now, let us consider how Gardner puts his account to use in support of the three
theses mentioned earlier in the paper. Consideration of the uses of
Gardners account will serve to clarify the account and also to set the
stage for a critique of it in Section 4.
Thesis (2) states that some sound legal reasoning (usually by the
higher courts) is capable of making new law, often accidently. Gardner
recognizes that courts routinely make new legal norms of a non-general
type when they issue binding rulings, such as the ruling that Jones is
liable to pay Smith $50 in reparative damages. This is a legal norm
because
[the] making of the ruling has legal consequences: it changes the application of
other legal norms. Not until the ruling has been made in his favour, for
example, can Smith lawfully enlist petty officials who will auction Joness property,
or attach Joness earnings.26

23
The first inference is not valid as formulated, viz. in the absence of a further normative premise
stating, say, that the terms of the Civil Rights Act are binding.
24
The word definition may seem strong. Indeed, it is unlikely that Gardner has in mind a definitive
list of necessary and sufficient criteria for the application of the concepts reasoning about and reasoning
according to law. I do believe, however, given the work they are supposed to do in his book, that
Gardners explanations of these concepts are meant as fairly precise statements of their meaning.
25
To be clear, Examples 1A and 3A were introduced here as variations on Gardners examples; they
are not found in Gardners book.
26
Gardner (2012), p. 186.

GARDNER ON LEGAL REASONING

But thesis (2) is not concerned with this relatively trivial type of
judicial norm creation; it refers to the more contentious judicial ability
to create new general legal norms in the process of deciding concrete
cases. Example 3 includes such a legal norm: Women have a legal right
not to be denied a job on the ground of their pregnancies. And,
implicitly, so does Example 2: Tortfeasors are liable to pay full reparative damages to those whom they tortiously injure, except where the
tort is also a breach of contract and awarding full reparative damages
for the tort would allow the recipient to circumvent a legally binding
contractual limit on damages for the breach of contract.27 To be sure,
the mere fact that a court refers to or implicitly relies on such general
norms does not turn them into valid law:
In virtue of (and subject to) the judges legal powers to decide cases on this subject,
these new norms become legally valid in the process, at least for the purposes of
the present case. If the judge sits in a sufficiently elevated court, then, depending
on the workings of the local stare decisis doctrine, the new norms may also
become legally valid for the purposes of future cases, subject always to future
judicial powers of overruling and distinguishing.28

Gardners argument for thesis (2) is unobjectionable. If Examples


2 and 3 are cases of sound (i.e. genuine) legal argument, then,
keeping in mind the proviso that the extent of a judges law-making
powers depends (inter alia) on the shape of the local stare decisis
doctrine, it can be said that some sound legal reasoning is capable of
making new law. We will later discuss whether Gardners definition
of reasoning according to law might be so broad as to cover other
argument patterns, beyond those instantiated by Examples 2 and 3,
that are not plausibly deemed sound types of legal reasoning. For
now, notice that the status of Examples 2 and 3 would hardly raise
controversy. Example 3 looks like a case of legal argument from
analogy. Gardner characterizes it as a classic example of legal reasoning.29 Other authors have gone further, treating argument from
analogy as the basic pattern of legal reasoning30 and as an identi-

27
28
29
30

p. 1.

Ibid., p. 188.
Ibid., pp. 3940.
Ibid., p. 40 (my emphasis).
Levi, Edward, An Introduction to Legal Reasoning (Chicago, IL: University of Chicago Press, 1949),

FBIO P. SHECAIRA

fying characteristic not only of legal reasoning itself but also of legal
education.31 Indeed, Gardner seems to be safe in his assumption that
Example 3 is a case of sound legal reasoning.
I am inclined to say the same about Example 2. It is a case of a conflict
of legal norms that (ex hypothesi) is not resolvable by appeal to any precise
criterion set out in a further legal norm (e.g. lex specialis derogat lex
generalis). That being the case, the appeal to a moral norm seems a legally
legitimate way of resolving the conflict: It is a legal argument only
because the question of how to apply the two [legal] norms and in
particular which of them to depart from makes the moral norm
argumentatively relevant.32 Reliance on merit-based considerations is
unavoidable even when the conflict of norms is resolved by means of
more structured deliberative procedures (short of a precise criterion
comparable to, e.g., lex specialis). Think, for instance, of proportionality
tests employed by constitutional courts in order to resolve conflicts
between fundamental rights or between rights and policies.33
In sum, Examples 2 and 3 are instances of sound legal argument
patterns (respectively: arguments dealing with a conflict between
norms and arguments from analogy) where new general legal norms
result from the combination of source-based legal norms with meritbased norms. So, again, Gardner is right to say that some sound legal
reasoning is capable of creating new law. Now, the case for thesis (8)
is more dubious. Consider again how Gardner formulates that thesis:
Another necessary connection [between law and morality]: legal
reasoning is moral reasoning with one or more legal premises.34
Shouldnt Gardner have said (as he did in formulating thesis (2)) that
some legal reasoning viz. reasoning according to law, but not

31
Weinreb, Lloyd, Legal Reason: The Use of Analogy in Legal Argument (New York: Cambridge
University Press, 2005), p. vii.
32
Gardner (2012), p. 188.
33
For an explanation of the limited power of constraint that proportionality tests exercise upon the
deliberation of legal actors, see Frederick Schauer, Balancing, Subsumption, and the Constraining Role
of Legal Text, in Matthias Klatt (ed.), Institutionalized Reason: The Jurisprudence of Robert Alexy (Oxford:
Oxford University Press, 2012). For a more technical discussion about the logical structure of conductive arguments (focusing on examples of proportionality reasoning in the Supreme Court of Canada), see Derek Allen, Conductive Arguments and the Toulmin Model: A Case Study, in J. Anthony
Blair and Ralph Johnson (eds.), Conductive Argument: An Overlooked Type of Defeasible Reasoning (London:
College Publications, 2011).
34
Gardner (2012), p. ix.

GARDNER ON LEGAL REASONING

reasoning about the law35 is moral reasoning with one or more


legal premises?36
Why does Gardner suddenly elide the distinction between the
different forms of legal reasoning in his formulation of thesis (8)?
Why does he claim that legal reasoning simpliciter which readers
would naturally take to include reasoning about the law combines
moral and legal premises? There are at least two alternative explanations for Gardners formulation of thesis (8). One explanation is
that Gardner makes a mistake that might come of his wish to put
forward a strong claim about the necessary connections between law
and morality. After all, it is more striking to say that legal reasoning
(simpliciter) is moral reasoning than to say that legal reasoning in one
of its possible forms is moral reasoning. The origin of this putative
mistake could be traced to one specific passage, where Gardner
begins to slip from using legal reasoning in a generic sense to using
it as a synonym of reasoning according to law:
For judges admittedly have a professional obligation to reach their decisions by
legal reasoning. And even in a case which cannot be decided by applying only
existing legal norms it is possible to use legal reasoning to arrive at a new norm
that enables (or constitutes) a decision in the case, and this norm is validated as a
new legal norm in the process.
Obviously, legal reasoning, in this sense, is not simply reasoning about what legal
norms already apply to the case.37

From this point on, in Chapter 2, Gardner drops the qualification


(in this sense) and appears to use legal reasoning and reasoning
according to law interchangeably. Eventually he speaks of judges
doing things by legal reasoning, or according to law.38 The generic
term stops being used generically and acquires what originally was
one of its two special senses. It could be, thus, that thesis (8) was
formulated once Gardner had already fused the two terms in his
mind.

35
Recall that I am using the phrase reasoning about the law to refer to practical arguments that
apply existing legal norms to the facts of a case (as opposed to arguments that simply state the abstract
content of the law at a given moment). See footnote 18 above.
36
In fact, Gardner seems to admit to this where he says that [legal] arguments often need moral
premises. Gardner (2012), p. 190 (my emphasis).
37
Ibid., p. 39.
38
Ibid., p. 41 (my emphasis).

FBIO P. SHECAIRA

A different explanation is that Gardner makes no simple mistake


no half-conscious terminological blunder in his formulation of
thesis (8), but is deliberately pointing to the fact that all legal reasoning (including reasoning about the law) is subordinate to moral
reasoning. If legal actors engage in legal reasoning it is because they
conscious as they must be about the moral significance of that
engagement regard themselves as having moral reason to do so.
Even if this is not how legal actors truly think (i.e. even if sometimes
they engage in legal argument for undisclosed prudential reasons or
for no determinate reason at all) Gardner would insist that they
necessarily speak as if they had moral reasons for arguing legally. For
legal actors necessarily put forward moral claims when arguing for
the existence of legal obligations, rights, permissions, powers, liabilities, and so on.39 Gardner appears to be arguing along these lines
when he says that moral premises enter into legal reasoning not in
virtue of any source-based authorization to that effect but rather
because legal reasoning is a morally significant practice that is bound
by moral principles:
Should we suppose that moral norms of justice are relevant to a legal argument
like that in [Example 2] only because of some (undisclosed) legal norm according
to which moral norms, or moral norms of justice, are legally binding or at any rate
admissible in legal argument? Of course not. This turns the world upside down.
The main puzzle about law, as a practice, is not the problem of how legal practitioners, including judges, come to be legally permitted or required to apply moral
norms. It is the problem of how legal practitioners come to be morally permitted or
required to apply legal norms. Legal practitioners should act morally in their work
for the same reason that doctors and soldiers should: because their work affects
peoples lives in morally significant ways. There is no further problem of why they
should act morally. Whereas there is a further problem a moral problem of
why they should defer to legal norms when they do so.40

A similar argument was once offered by Raz:


Given that legal decisions have a significant impact on peoples lives they must
at least appear to be morally acceptable. This means that legal reasoning can be
autonomous only on the basis of the two-stage argument, i.e. only if there is a
body of considerations which can be applied without using moral considerations,
which is such that its application by the courts appears to be morally acceptable.

39
40

Ibid., p. 132.
Ibid., p. 189.

GARDNER ON LEGAL REASONING

The doctrine of authority provides such sanction for the application of sourcebased considerations.41

Raz asserts the priority of moral reasoning over legal reasoning and
holds that a legal argument can only be described as autonomous (i.e.
as relying exclusively on source-based considerations) insofar as it is
regarded as composing the second stage in a two-part argument, the
first stage of which appeals to moral considerations that sanction the
use of legal reasoning. On this view, even an apparently simple argument like Example 1 would amount to the second inference in a
composite argument beginning with a moral sub-argument (in italics):
It is morally permissible for courts to adjudicate cases by appeal to source-based norms;
among the source-based norms of the legal system there is one to the effect that tortfeasors
are liable to pay full reparative damages to those whom they tortiously injure;
therefore, tortfeasors are liable to pay full reparative damages to those whom they
tortiously injure;
Jones tortiously injured Smith to the tune of $50;
therefore, Jones is liable to pay Smith $50 in reparative damages.

If we assume that Gardner wants to endorse Razs two-stage


conception of legal argument, then we can make sense of the claim
that legal reasoning is moral reasoning with one or more legal
premises. For what is the argument above but a complex moral
argument with one legal premise (to wit, the first premise of the
second inference)? It is still not clear to me, however, whether the
second explanation is better than the first explanation of Gardners
case for thesis (8). The first explanation suggests that thesis (8) should
be understood as a poorly worded claim to the effect that reasoning
according to law (not legal reasoning in general) is moral reasoning
with one or more legal premises. The second explanation assigns no
similar mistake to Gardner but, on the other hand, portrays the case
for thesis (8) as relying on a contentious aspect of the Razian theory
that supplies the framework for Gardners book. Behind the view
that all legal reasoning is moral reasoning lies, it seems, the idea that
the normative claims made by legal actors are necessarily moral
claims. If we accept that judges statements of source-based legal
41

Raz (1995), p. 334.

FBIO P. SHECAIRA

norms (e.g. tortfeasors are liable to pay) are really statements of


moral norms (tortfeasors are morally liable to pay), then we have
reason to believe that judges must assume the existence of some
moral authorization to apply the norms issuing from the relevant
sources. The question, of course, is whether we should indeed accept
that legal statements are moral statements.42
So, the trouble with the second explanation is that it renders the
case for thesis (8) dependent not only on Gardners plausible account
of legal reasoning but also on other, less persuasive, aspects of his
legal theory. Not knowing what is more charitable to suggest that
thesis (8) is poorly worded or instead that it rests on a contentious
argument I think it is best to leave the question open and allow
Gardner to explain exactly what he means by thesis (8).
Gardners plausible case for the legitimacy thesis was mentioned
in Section 2. He does not claim that his account of legal reasoning
has any direct moral implications; instead he claims that the account
serves to refute one of the assumptions underlying a well-known
argument against the legitimacy of judicial law-making. Positivists
believe that law is often indeterminate with respect to particular
cases. This belief is grounded on the view that all law is positive law
and that positive law, dependent as it is on the actions and intentions
of human agents with limited foreknowledge, cannot fail to have
some gaps. Judges are not normally empowered to refuse to adjudicate a case on the ground that the law does not provide a unique
solution to it. When law is indeterminate, judges must adjudicate by
having recourse to extra-legal considerations. But, it is thought,
judges also have a constant duty to apply the law. When they do
otherwise, they act as legislators, creating new legal norms on the
basis of non-legal considerations. Not only is this to usurp legislative
power, it also involves a violation of the ideal of the rule of law. For
judicial legislation is typically retroactive: it occurs in the context of a
legal dispute involving parties who did not act with the benefit of
knowing in advance the law that would be applicable to them.
42

This is precisely one of the Razian commitments that Kevin Toh complains about in his review
of Gardners book (see text accompanying footnote 6 above). I agree that Gardners case for this
position is not entirely convincing. But this is not to accept the suggestion (which comes with the
phrase Razian commitment) that Gardner accepts Razs view uncritically or as a result of some sort of
bias. In fact, Gardner argues carefully for the view at issue between pages 132 and 139 of his book. I am
left unconvinced but not unimpressed by his efforts. For a detailed analysis and critique of Gardners
argument, see the paper by Lus Duarte dAlmeida and James Edwards in this symposium.

GARDNER ON LEGAL REASONING

Gardner acknowledges that, given the indeterminacy of law,


judges must sometimes appeal to non-legal considerations. In doing
so, they may make new law. But Gardner insists that this judicial
exercise in law-making is not to be confused with legislation. Legislation is not constrained by positive law.43 The power to legislate is
the power to change structural features of positive law, often
abruptly. Judges who make law, on the other hand, can do so by
reasoning according to law, in a much more piecemeal and controlled manner. The moral considerations used by judges do not
operate freely but are combined with legal considerations in arguments that resemble Examples 2 and 3. So long as judges make law
by these means, they are not legislating not in any strong sense of
the word:
What is really morally important under the heading of the separation of powers is
not the separation of law-making powers from law-applying powers, but rather the
separation of legislative powers of law-making (i.e. powers to make legally
unprecedented laws) from judicial powers of lawmaking (i.e. powers to develop the
law gradually using existing legal resources). Similarly, the only morally credible
rule-of-law ban on retroactive legislation is just that; namely a ban on retroactive
legislation, not a ban on the retroactive change of legal norms, even when that
change is made in accordance with law.44

IV. PROBLEMS IN GARDNERS ACCOUNT

One problem in Gardners account pertains to his use of the word


premise. Legal reasoning is distinguished by Gardner from other
types of reasoning (e.g. moral, political) by the presence of alreadyvalid legal norms among its operative premises. Legal reasoning
itself can take two forms, the difference between them relating to the
decisiveness of the legal norms employed as premises. In reasoning
about the law, the legal norms are decisive (i.e. together with the
facts of the case, they establish a definite ruling). In reasoning
according to law, the legal norms are non-redundant yet non-decisive (i.e. to establish a ruling, they must be combined with other,
43
Gardners point presumably applies to non-subordinate legislators whose powers are not closely
circumscribed by source-based duties concerning the goals their laws should promote. For discussion,
see Raz (1995), pp. 241244, on the directed powers of subordinate legislators.
44
Gardner (2012), p. 41.

FBIO P. SHECAIRA

merit-based norms). Recall Example 2 (now with parenthetical letters, for ease of reference):
(A) Tortfeasors are liable to pay full reparative damages to those whom
they tortiously injure;
(B) Jones tortiously injured Smith to the tune of $100;
(C) but the tort was also the breach of a contract between Jones and Smith;
(D) the contract provided for maximum reparative damages of $50 for any
breach;
(E) contracts and the limits on damages they set are legally binding as
between parties to the contract;
(F) and it is unjust to let someone avoid a legally binding contractual limit
on damages by instead suing the other contracting party in tort;
(G) therefore, Jones is liable to pay Smith only $50 in reparative damages.

Gardner says that the legal standing of this argument is due to its
having two already-valid legal norms as operative premises: namely,
claims (A) and (E).45 But notice that there is a conflict between the
norms expressed in (A) and (E).46 The moral norm in (F) serves to
resolve that conflict by tipping the balance in favor of the conclusion
supported by the norm in (E). The norm in (A) is overridden in the
process and therefore plays no role in the justification of the conclusion, (G). Yet Gardner deems (A) a premise. This is unusual, since
premises are most often defined as claims put forward in support of
(or as evidence for) a conclusion.47 Is this an innocuous terminological quirk in Gardners book? Not really. Gardners use of the
word premise invites a problem that can be perceived if we recall
that legal norms figure in legal reasoning not only as premises but
also non-redundantly.
Since a premise, for Gardner, need not be a claim put forward in
support of a conclusion, then a non-redundant premise presumably
is any consideration that somehow affects (i.e. is relevant to) the
truth or tenability of the conclusion, positively or negatively. But if
45

Ibid., p. 188.
To be precise, the conflict is not between the norms themselves but between the consequences
that they entail in conjunction with other premises of the argument.
47
It is unusual but not unprecedented: Wellman understands a premise as any consideration
which counts or is thought to count for or against the conclusion. Wellman (1971), p. 90 (my emphasis).
For discussion of Wellmans unusual definition, see Rongdong Jin, The Structure of Pro and Con
Arguments: A Survey of the Theories in Blair and Johnson (2011), pp. 1113. Jin refers to Irving Copis
widely known textbook on logic, where premises are defined as propositions affirmed (or assumed) as
providing support or reasons for accepting the conclusion. Ibid., p. 11.
46

GARDNER ON LEGAL REASONING

that is the case, then the following argument would have a nonredundant legal norm as an operative premise and would consequently fit Gardners definition of legal reasoning (more specifically,
of reasoning according to law). Consider Example 4:
Homicide is to be punished by death (an already-valid legal norm);
Jones committed homicide;
but the death penalty is morally repugnant (a legally unprecedented moral norm);
therefore, Jones is not to be punished by death.

The source-based legal norm in the first line of the argument


figures as a non-redundant premise (in Gardners sense), since it
counts against the conclusion that Jones should be exempted from the
death penalty. The fact that it is overridden by a moral norm does
nothing to change its standing as a non-redundant premise. Yet it is
hard to accept the implication that Example 4 is a case of reasoning
according to law. Gardners definition is overly inclusive as it stands.
It could be suggested that, to correct the problem, Gardner might
simply turn to the traditional use of the word premise. He could
then insist that legal reasoning is exemplified by arguments that
include non-redundant legal norms that count (or are thought to count)
in favor of the ruling, even though they may fail to be decisive and
thus require the support of some additional moral norm. Examples 2
and 3 would still fit the amended definition, whereas Example 4
(fortunately) would be excluded. However, this adjustment takes
care of only part of the problem of over-inclusion. There are other
dubious cases of legal argument that would still fit Gardners definition, even if the word premise were taken in its traditional sense.
Consider Example 5:
The liberty of the individual is to be protected against state interference (an
already-valid legal norm);
the liberty of the individual encompasses the right to purchase or sell labor;
therefore, the liberty of the individual to purchase or sell labor is to be protected
against state interference;
therefore, the liberty of bakery employees to agree to work for more than 60 h a
week or 10 h a day is to be protected against state interference.

FBIO P. SHECAIRA

Example 5 is inspired by the famous case of Lochner v. New York.48


Like the majority argument in Lochner, Example 5 would probably
raise some eyebrows. Lochner is often regarded as involving a disingenuous attempt to present a decision resting on a highly contentious moral-political argument as if it flowed inexorably from the
meaning of abstract terms like liberty. Notice that, on Gardners
definition, Example 5 is no less a case of legal reasoning than
Examples 1, 2 and 3. In Example 5 there is an operative premise
expressing a valid legal norm (in the first line of the argument) which
is non-redundant yet non-decisive. The problem is that the legal
premise is highly indeterminate. The argument relies on a moral
premise (in the second line) whose use is hardly circumscribed by the
meaning of the deeply contested phrase liberty of the individual.
Although there is no indication of this in the argument as formulated, there is ample space for reasonable debate about the meaning
and implications of the phrase.
Now, Example 5 does not work against Gardners account in
exactly the same way as Example 4. It is straightforwardly counterintuitive to classify the latter as an example of reasoning according
to law, since it involves a departure from a legal norm on exclusively
moral grounds. The moral consideration in Example 5, on the other
hand, complements the relevant legal norm, which is not defeated or
overridden in the process. We should, however, be able to perceive
the challenge posed by Example 5 once we recall the uses of
Gardners account, in particular the case for the legitimacy thesis.
For what is the difference between Example 5 and the sort of reasoning we expect from senior legislators who are constrained by
nothing beyond the broad terms of a terse constitutional text? The
plausibility of Gardners case for the legitimacy thesis rests on the
idea that judicial law-making powers are not illegitimate insofar as
they consist in powers to develop the law gradually using existing
legal resources.49 But there is nothing gradual about the reasoning
in Lochner or in Example 5. Perhaps Gardners account is not
straightforwardly over-inclusive in respect of Example 5; but the
account still fails to support the legitimacy thesis insofar as it includes

48
49

198 U.S. 45 (1905).


Gardner (2012), p. 41.

GARDNER ON LEGAL REASONING

arguments where law is not made in piecemeal and controlled


fashion.
It is not easy to find a way of helping Gardner to avoid the
problem of over-inclusion. Part of the difficulty stems from the fact
that the moral considerations to which a legal actor resorts can be
more or less closely circumscribed by existing legal standards. There
is a categorical difference (i.e. a difference in kind) between reasoning about the law and reasoning according to law. But within the
latter category there are many different degrees of control that the
existing legal norms can exercise over complementary merit-based
considerations. As Timothy Endicott has put it, a new norm or
standard that is repugnant to the existing posited law would certainly be new in a stronger sense than a standard that is narrowly
controlled by [though not yet a definite part of] the existing law.50
As a consequence, there are intermediate degrees between clear
exercises in legislation and clear exercises in judicial law-making
hedged by legal reasoning. The problem is to know where to we
draw the line. How narrowly must law-making be controlled for it to
be properly exercised by judges?
In the absence of a plausible answer to that question, it might pay
to consider a distinction that Gardner fails to discuss. It is the distinction between legally authorized reasoning and legally unauthorized
reasoning. This distinction is not to be confused with the distinction
between legal reasoning and non-legal reasoning. Legal analogies are
classic examples of legal reasoning. Their legal character does not
entail, however, that a judge is authorized to use them in every
context. There are countries, for instance, that restrict the use of
analogies in criminal law cases, especially where the analogies may
be used to establish a ruling that does not favor the defendant. That
is a case of legal reasoning that does not enjoy legal authorization.
On the other hand, there are cases of legally authorized non-legal
reasoning. For instance, the Swiss Civil Code instructed judges, in
the absence of other aids to statutory interpretation, to enforce the
rule they would adopt if they were legislators themselves.51

50

Timothy Endicott, Adjudication and the Law, Oxford Journal of Legal Studies 27(2) (2007), p. 319.
John Henry Merryman and Rogelio Prez-Perdomo, The Civil Law Tradition: An Introduction to the
Legal Systems of Europe and Latin America (3rd edition, Palo Alto, CA: Stanford University Press, 2007),
p. 46.
51

FBIO P. SHECAIRA

In fulfilling their duties, judges should not only pay heed to the
legal standards more or less directly pertinent to the facts of a case.
They should also review any higher-level standards that regulate the
application of the immediately relevant standards. For higher-level
standards may proscribe what otherwise would be regarded a sound
form of legal reasoning, and they may also render lawful a form of
argument that judges would normally not be allowed to engage in.
The latter possibility is more important for our purposes. There can
be no talk of usurpation of legislative powers by a judge who creates
legally unprecedented norms with the authorization of legislated
positive law.52 It would also be wrong to accuse such a judge of
disregarding the rule of law if she acts under the protection of a
higher-order source-based legal rule. The rule of law does not discriminate in principle among the levels at which the various sourcebased rules of a legal system are located.
In other words, the legitimacy of judicial law-making does not
depend only on the degree to which judicial reasoning is constrained
by existing legal norms. Indeed, prior to asking whether the law is
able to constrain legal reasoning, a judge should ask if the law itself
prescribes that legal reasoning be constrained. Let us go back to
Example 5 (my version of the majority argument in Lochner). By
hypothesis, there is no express legal authorization in that case for
judges to engage in non-legal reasoning (i.e. there is nothing like the
provision of the Swiss Civil Code mentioned earlier). But there is
also a shortage of minimally precise legal standards by which a
dutiful judge might abide. If there is anything morally or politically
problematic here it is the shortage of precise legal standards itself,
not the law-making that is exercised by judges as a consequence of
the shortage. The shortage may have a number of causes, only one
of which is legislative incompetence or indifference. Another salient
cause, especially where constitutional documents are at issue, is a
definite (albeit often undisclosed) legislative wish to avoid settling
52
The word legislated is important here. Some may question the legitimacy of a courts ability to
legislate (i.e. to forge a new general legal norm on the basis of all-things-considered deliberation) if that
ability is provided for in case law. A judge does not usurp legislative power if she was awarded power to
legislate by the legislators themselves. But it may be argued that a judge exceeds the limits of her
powers if she legislates on the basis of a prior judicial authorization which lacks legislative sanction. For
instance, there is a stronger argument to be made for the legitimacy of judicial review in those countries
where the institution is legislatively authorized than in those countries, such as the US, where its origins
lie in case law.

GARDNER ON LEGAL REASONING

controversial moral or political issues that are best resolved by


judges in the context of particular cases.
V. BACK TO GARDNERS ASSUMPTIONS

The upshot of the previous section is that Gardners account of legal


reasoning, though generally plausible, has some significant limitations.
The account provides definitions of different forms of legal argument
in terms of valid legal norms operating as non-redundant but not
necessarily decisive premises which need to be adjusted in order to
avoid a problem of over-inclusion. One way of mitigating the problem
is for Gardner to alter his understanding of a premise, i.e. to make it
include only considerations that count or are thought to count in favor
of a conclusion. But even if Gardner did alter his notion of a premise,
we would still be able to find examples of arguments with nonredundant legal norms figuring as premises that should not be regarded
as examples of genuine legal arguments at least not for the purpose of
establishing the legitimacy thesis. In the previous section I suggested a
distinction (between legally authorized and legally unauthorized reasoning) that may serve to supplement Gardners account and to some
extent help him deal with the issue of legitimacy.
The problems in Gardners account over-inclusion and need of
supplementation are limitations that pertain to the definitions of
reasoning about the law and reasoning according to law. Recall,
however, that these definitions do not exhaust the components of
Gardners account. As we discover problems affecting the definitions,
Gardners examples emerge as the mainstay of his account. Examples
2 and 3, in particular, continue to serve well to pump our intuitions
toward the view that reasoning may be assigned legal standing even
if it includes legally unprecedented moral norms as premises. The
problem is to associate these examples with general definitions that
are not affected by the problem of over-inclusion. I will not attempt
to provide my own definitions here, but I will suggest a precaution
that Gardner may want to keep in mind if he eventually agrees that
his definitions should be revised.
To put it simply, Gardner may want to expand his set of examples. Gardners current sample is not only small; it seems to have
been selected on the basis of contentious jurisprudential assumptions. Gardners examples of legal reasoning invariably involve an

FBIO P. SHECAIRA

appeal to already-valid legal norms. The examples vary in respect of


how the legal norms interact among themselves or with merit-based
norms, but they all involve existing legal norms. It could be argued
that Gardners focus on arguments that rely on existing legal norms
is explained by his commitment to legal positivism. Positivists
assume that valid law can be identified by appeal to non-evaluative,
source-based reasoning; and if that is the case, then there is nothing
of particular interest to be said about reasoning regarding what the
content of the law is at a given moment except that it relies on the
fact that certain actions took place, that they were undertaken with
certain intentions, that the rest of the law is thus and so, etc.53 For
positivists, in other words, arguments concerning the content of the
law are non-evaluative and thus have nothing to teach us about how
legal norms may interact with moral norms in legal reasoning.
But the point I want to make in this final section has little to do
with positivism. Even if Gardner is right to hold that the validity of a
legal norm is not dependent on its merits, he is not thereby free to
ignore the sorts of arguments that legal actors often formulate in
(what they conceive as) attempts to discern the content of the law.
Gardner may want to insist that judges are presenting their own
reasoning inaccurately when they, while relying on moral considerations, claim to do no more than search for existing law. The point
is that legal actors, mistaken or not, often have discussions about
what the law is that involve appeal to merit-based considerations;
and the arguments that they formulate in the context of such discussions do not clearly fall into the patterns of argument that
Gardner explicitly considers with Examples 2 and 3. Could Gardner
simply dismiss those arguments on the assumption that they do not
count as genuine cases of legal reasoning?
The types of argument that Gardner leaves out of his account
have been taken to be at the center of legal practice by legal theorists
of different schools of thought. Karl Llewellyn, for instance,
famously held that statutes can be read according to different canons
of interpretation, thereby yielding conflicting solutions to legal
cases.54 In the same vein, it could be said that a statutory provision
will have different effects for a legal case depending on whether it is
53

Joseph Raz, Between Authority and Interpretation (Oxford: Oxford University Press, 2009), p. 378.
Karl Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons about
How Statutes are to Be Construed, Vanderbilt Law Review 3 (1950).
54

GARDNER ON LEGAL REASONING

interpreted by an intentionalist, a textualist, or a purposivist judge.


And the point could be generalized to apply to other source-materials: a precedent, for instance, will have different effects depending
on whether its binding element its ratio is associated with the
material facts of the case, the explicit rationale for the ruling, the
moral principle justifying a line of related cases, and so on.
Disagreement about proper interpretative methodology is also
emphasized by Ronald Dworkin in his famous discussion of theoretical disagreement55 in law, a broad category which also includes
disagreement about what source-materials are relevant to a case.56
The latter sort of disagreement goes even deeper than the type of
disagreement discussed by Llewellyn; for prior to determining how
one is to go about interpreting any given bit of source-material, one
must determine whether that bit of source-material is relevant to the
case (e.g. is the case regulated by statute or precedent?).
It could of course be argued that Llewellyn, Dworkin and others
have exaggerated the centrality to or the pervasiveness in legal
practice of different forms of theoretical disagreement.57 But its
presence in law, if mostly at the appellate level, is not to be ignored.
Common lawyers might be especially concerned with the fact that
Gardner puts no argument in standard form that could represent a
judges attempt to discern a doctrine running through a line of
related cases. That is a classic form of legal argument (is it not?), but
it involves an attempt (prior to applying any legal norm) to establish
that certain legal norms are to be applied as valid.
There is no guarantee that Gardner will arrive at an improved
definition by expanding his set of examples. But if Gardner did expand his sample we would have less reason to worry that his perception of legal practice may be skewed by contentious theoretical
commitments. It is hazardous to pursue an explanation of legal
reasoning while ignoring precisely those types of argument that have
generated most jurisprudential discussion.
55

Ronald Dworkin, Laws Empire (Cambridge, MA: Harvard University Press, 1986), Chapter 1.
For discussion of the several types of disagreement that count as theoretical, see Dale Smith,
Theoretical Disagreement and the Semantic Sting, Oxford Journal of Legal Studies 30 (2010), pp. 641642.
57
For an argument to that effect, see Brian Leiter, Explaining Theoretical Disagreement, The
University of Chicago Law Review 76 (2009) (claiming that theoretical disagreements are relatively
marginal phenomena within the scope of a general theory of law, emerging primarily at the pinnacle of
the pyramid of legal questions that arise. Ibid., p. 1249). For another argument along the same line, see
Fbio Shecaira, Dealing with Judicial Rhetoric: A Defence of Hartian Positivism, Australian Journal of
Legal Philosophy 37 (2012).
56

FBIO P. SHECAIRA

ACKNOWLEDGMENTS

For helpful comments on earlier versions of this paper I would like to thank
Lus Duarte dAlmeida, Ben Hamby, Lucas Miotto, Stefan Sciaraffa, and a
referee for Law and Philosophy.
Faculty of Law
Federal University of Rio de Janeiro,
Rio de Janeiro, Brazil
E-mail: fabioperin@direito.ufrj.br

You might also like