TALKING ABOUT THE SAME?
LEGAL SCHOLARSHIP & CULTURE FROM A COMPARATIVE PERSPECTIVE.*
I. Two kinds of legal scholarship
Scholarship in law typically assumes one of two forms. It either develops within the
realm of law, contributing to the legal materials and institutions’ capabilities to regulate
conduct; or it maintains a certain distance from that realm, defines it as a practice, and
attempts to provide an account so that it can be better understood or criticized. I will call
them internal and external legal scholarship, respectively.
Legal scholarship has traditionally been internalist for the most part: the label “doctrine,”
as internalism is known, clearly encompasses the dominant work of academic traditions
worldwide. The externalist approach represents a minor but growing tendency among
legal academics, being still very far from the central role it assumed on other disciplines,
like the social sciences. This description is roughly accurate with regard to most
academic legal cultures around the globe: Europe, Latin America, Asia, etc.1 There is,
however, one important exception worth considering: elite legal education and
scholarship in the United States of America, which focuses primarily on the externalist
approach and openly disregards doctrine.2 This paper will re-examine the foundations of
the internalism-externalism divide with recourse to a comparative analysis.
The study of constitutional law is probably one of the fields where the distinction can be
better appreciated. The standard doctrinal approach will typically dissect the
Constitution (the document) and will explore for any other source that can help assess
the meaning (history, normative theory, etc.), being especially attentive to authoritative
interpretations of the text by officials. The scholar will then arrange these materials in a
systematized presentation to make them more accessible and organized to the profit of
practitioners. On a second stage, he or she will develop the existing materials to
This paper represents work in progress. Please do not cite without the authors permission.
JSD Candidate, Yale Law School. firstname.lastname@example.org
For Europe, see Armin von Bogdandy, The Past and Present of Doctrinal Constructivism, 7 INT’L J.
CONST. L. 364 (2009).
See Robert C. Post, Constitutional Scholarship in the United States, 7 INT’L J. CONST. L. 416 (2009)
address questions that are not so simple to answer by glancing at the sources, where a
much deeper analysis is needed to find out what the law requires. Here, the scholar will
typically report the conclusions of other colleagues and confront them where he or she
thinks better answers can be provided. When carefully considered, the scholar’s job is
not substantially different from that of the judge or lawyer: very much like these
practitioners, our scholar will center his or her work on the legal materials.3
The focus of external constitutionalism is not the materials but the practice of
constitutional decision-making—the process by which the institutions and society at
large make political decisions of the constitutional kind.4 The externalist scholar will start
by delimiting the practice by way of pointing out the distinctive features it has, which will
have to be accounted for and justified to the light of shared values. Academic
communities typically share standard understandings of the practice, which defines
common problems and allows for scholars to address them in a fluid dialogue.5
However, it is distinctive of externalism that the practice’s boundaries are open to
discussion (unlike in doctrine, where the materials are identified a priori).6
The following chart stresses the main differences:
Internal Legal Scholarship External Legal Scholarship
Conceptualizes “the The universe of legal materials that are A social practice (the practice of law) by
law” as supposed to guide our conduct as which citizens abide by the law and
citizens and our decisions as officials officials coerce them into doing
something under the claim that it is the
law that requires it.
The problem that The technical fact that legal materials The practice is obscure, complex and
scholars address is cannot solve all the problems, because hard to understand. Additionally, it is not
they are chaotic, conflictive, or simply always clear how justified the existing
run out. practices are in relation to our societal
values and commitments.
The main The minimization of hard cases, by To reveal the several factors at play
contribution of means of: (i) a systematic presentation when deciding cases or legislating. This
scholars to the legal of the materials that allows for a better may enable practitioners to: (i) make
community identification of the rules and principles adjustments to the practice from inside
(lawyers, judges, and hence a better application of them to improve it; and (ii) think of, argue, and
legislators) is to cases; and (ii) the development of the decide hard cases.
law by discovering “implicit” solutions.
Academics They contribute to the rule of law by They contribute to deepen our
Von Bogdandy considers it “arguable” that doctrine can be considered a part of legal practice. Id. 367.
For an interesting example see Robert C. Post, Theorizing Disagreement: Re-conceiving the
Relationship between Law and Politics, 98 CAL. L. REV. 1319 (2010), 1347-50.
A classic example is judicial review in the United States and the “counter-majoritarian difficulty” as
Alexander Bickel exposed it. See THE LEAST DANGEROUS BRACH (Yale U.P.,1962).
Cf. Ronald M. Dworkin, LAW’S EMPIRE 91 (Belknap-Harvard, 1986).
consider their work developing a more refined legal understanding of how our social
valuable to society apparatus (easier to understand, practices are, and suggest adjustments
because clearer, more comprehensive) that can to better serve social goals.
better perform its conduct-guiding
The product of what A source of law. Although a source of A theoretic account (descriptive,
scholars do is secondary importance, scholars still normative, or interpretive).
purport to speak from the authority of
the law when they “develop” the
materials identifying solutions to cases
not explicitly covered.
On this paper, I will not explore the several issues that arise from this distinction taking it
as given. Instead, I argue that it appropriately captures the remarkable differences that
exist in legal scholarship when comparing the traditions mentioned above. For this
purpose, I will concentrate on one dimension in particular: the ideological one. By
exploring the foundations of modern American legal scholarship I intend to show that
what lies beneath this peculiar academic practice is precisely the internal-external
distinction (instead of some of the explanations typically offered). The argument is that
American Legal Realism shaped the current scholarship by means of externalizing it—
neither by providing a more accurate description of how judges decide, nor by
convincing us all that the law is indeterminate.
II. The Internalism-Externalism distinction at the base of the Formalist-Realist
It is generally said that American Legal Realism is responsible for the distinctively
externalist form of legal scholarship in the United States (US). Of the several features of
the American legal practice that have been attributed to the influence of realism, the
impact on the legal academy seems to be the one where the realist influences can be
clearly traced (after all, most—if not all—scholars today define themselves as “realists”).
American legal realists reacted against what they considered to be the dominant legal
This explains the standard argument internalists—as well as practitioners—use to draw the limit to their
claim of authority: de lege lata (within the limits of doctrine), and de lege ferenda (out of limits). This
argument is, of course, inexistent in an external-scholarship culture like the United States.
ideas of the time, later labeled as “formalism.”8 It has been extensively argued that the
legal cultures of Europe and Latin America, to name some that have been relatively
untouched by the powerful wave of legal realism that shocked the US, remain
characteristically “formalist.” There seems to be an important connection between this
formalist legal culture and the predominance of a doctrinal scholarship—
paradigmatically internalist—on these countries.9
The fact that an internalist or doctrinal approach to legal scholarship was common in the
US before realism, and conspicuously changed after it to conform to a typically
externalist perspective, shows that realism attacked the very foundations of doctrinal
scholarship. Indeed, legal realists explicitly declared that they wanted to reshape legal
scholarship in order to make it “more useful for lawyers.”
Al of this contributes to the wide arrayed impression that academic cultures that engage
in internalist legal scholarship hold distinctively formalist (or pre-realist) conceptions
about the law, while those that—like the US—have a consolidated practice of
externalism have been highly influenced by legal realism. However, while the decisive
influence of legal realism on the distinctively externalist American scholarship seems
more less clear, the connection between formalism and internalism that purportedly
dominates other cultures has mostly been conjectured rather than established. The
present paper attempts to analyze the way in which legal realism shaped the American
legal scholarship to convert it to externalism. Although several related claims are
normally associated with legal realism, I argue that the influence of this school of
thought is better understood in terms of reshaping the academic conception of law,
which ultimately influenced the conception of practitioners.
The formalism-realism dispute is surely complex to define.10 American legal realists
See William W. Fisher, Morton J. Horwitz, and Thomas A. Reed (eds.), AMERICAN LEGAL REALISM
(Oxford U.P., 1993), xi-xv (where “formalism” is more precisely labeled “classical legal thought”).
This is the argument of Patrick Atiyah & Robert Summers, FORM AND SUBSTANCE IN ANGLO-AMERICAN
LAW (Oxford U.P., 1987), Ch. 14.
Just a note on method. To reconstruct the dispute, one can (a) provide an historical account of the
ideas and their influence, as a chapter in the history of ideas. Such an attempt would try to explain who
the realists were, which ideas they had, and what they were reacting to. It should provide first hand
evidence and is vulnerable to attack with evidence pointing on a different direction. On the other hand,
one can (b) develop a substantial account of the dispute that tries to present the issues that were
originally discussed (or some of them) in a way that probably disregards most of the actual discussions
and their context, and reframes (some or most of) the basic concerns so that they directly address
were not a movement, and their ideas deferred from one another in very important
ways. Conversely, the formalists have never been considered a solid and well-defined
camp but rather a grouping of general ideas and beliefs about the law that were
allegedly common before the realist attack; furthermore, it seems that the grouping was
mostly done by the realists themselves to better serve the targeting.11 Despite these
complexities, the formalism-realism dispute has been at the centre of the American
legal universe for decades and has influenced dramatically the way in which Americans
learn, think about, practice, and teach the law.
There are probably two ways in which the dispute can be reconstructed, although
arguably most accounts have not been careful enough to separate both kinds of
claims.12 According to the standard story on the formalism-realism divide, American
legal realists of the 1920´s and 30´s reacted to an account of the practice of
adjudication—purportedly dominant among academics and practitioners during the
formalist age—that they considered to be wrong and misleading. This account, realists
exposed, claimed that American appellate judges decided the legal questions on cases
before them solely on the basis of what the legal materials required them to. The
realists challenged an understanding of the American judicial practice that reconstructed
awkwardly what appellate judges did (and do), and therefore served as a foundation for
legal studies that were of no use to the practicing lawyer. Put this way, the core claim of
legal realism is fundamentally a descriptive one and is supposed to confront an equally
descriptive formalist counterpart on its lack of empirical fit to explain what American
judges were (and are) doing.
There is yet another popular reconstruction of the dispute—one that attempts to stress
more fundamental differences between formalism and realism, on the basis of the
problems that we presently find interesting and illuminating. Historic fidelity is, of course, not a primary
concern of such an attempt—one can be selective in choosing topics and arguments that better fit our
contemporary interest. My argument should be read in this second sense (b).
Brian Z. Tamanaha has very recently argued that there are many holes in the story of formalism and
realism. See BEYOND THE FORMALIST-REALIST DIVIDE (Princeton, U.P., 2010).
For a famous account mixing the two see Brian Leiter, Positivism, Formalism, Realism, 99 COLUM. L.
REV. 1138 (1999), 1145-6 (“Given these definitions, we may characterize formalism as the descriptive
theory of adjudication according to which (1) the law is rationally determinate, and (2) judging is
mechanical. It follows, moreover, from (1), that (3) legal reasoning is autonomous, since the class of legal
reasons suffices to justify a unique outcome; no recourse to non-legal reasons is demanded or
required.”). (My emphasis.)
conceptual debate about legal indeterminacy. Do the legal materials and the legitimate
forms of legal reasoning determine the solution to cases? Or does the law (understood
in that sense) suffer from indeterminacy? According to this second interpretation,
American legal realism is responsible for making the point that the legal materials are
indeterminate, therefore being inapt to constrain the judge to decide cases in one
sense. Note that the issue here is not how the judges work with the legal materials at
their disposal, but rather what they can and cannot do given the very nature of the legal
materials. The first claim is descriptive, empirical, and hence dependent on the legal
culture and practice. The second claim, however, is conceptual: it refers to the way in
which the legal materials and valid forms of legal reasoning do influence the business of
judging, with independence of the particular legal system and cultures.13 Formalism, in
this latter view, has to be presented as the rival theory that advances the claim of legal
In this paper I argue that both accounts of formalism are either false, or at least
incompatible with the formalist foundation of internalism, assuming there is one. The
presentations of formalism suggested by each reconstruction of the formalism-realism
divide could serve as caricature models to help emphasizing the virtues of realism, but
they are certainly very problematic when they try to explain which ideas can actually be
found in the world supporting the practices we engage in. By all means, “formalism” has
long been an obscure label.15 I address each account on a separate section.
a. The descriptive project of realism, and formalism
American legal realists’ primary concern was to understand the way in which appellate
judges decide cases, in order to explain the process of judicial decision-making and, if
possible, be able to predict it. This represents, in their opinion, the only research
agenda that will prove useful for practicing lawyers. The project is a descriptive one:
empirical observation reveals that judicial outcomes cannot be explained on the basis of
the legal materials and legal reasoning, therefore, the goal is to identify the factors that
See Scott Shapiro, LEGALITY (forthcoming: Harvard U.P., 2011), 241-2.
See, for example, Brian Leiter, NATURALIZING JURISPRUDENCE (Oxford U.P., 2007), 92.
Herbert L.A. Hart, THE CONCEPT OF LAW (Oxford U.P., 1994), 297 (“It is not always clear precisely what
vice is referred to in these terms”).
determine the judges’ decisions in the cases before them.16
The formalist counterpart to such a conception of realism has normally been conceived
as a theory of adjudication in the following terms:
The contrasting view here—usually dibbed “Formalism”—is committed to the descriptive
claim that judges respond primarily (indeed, perhaps exclusively) to the rational
demands of the applicable rules of law and models of legal reasoning. We can gloss the
Formalist’s descriptive claim as saying: judges are (primarily) rule-responsive; while the
Realist claims that judges are (primarily) fact-responsive. […] The real dispute between
the Formalist and the Realist then concerns whether the reasons that determine judicial
decision are primarily legal reasons or non-legal reasons.
We should be aware of three features that formalism presents when construed as a
“contrasting view” to realism in such terms. (i) First, it needs to be understood as a
theory with strong descriptive ambitions. (ii) Then, it focuses on adjudication as the
social practice it tries to explain, instead of being concerned with legal events more
broadly considered (i.e. instances where our conduct voluntarily abides by, or is forced
to comply with the law other than when the members of the judiciary adjudicate a
dispute). (iii) Furthermore, in order to maintain a valid argument with realism, it needs to
direct its descriptive ambitions to one particular portion of the adjudicative practice: the
appellate review (of questions of law), instead of being more comprehensive on its
explanation of adjudication. I will bracket the second and third points for now and focus
only on the first one.
It has been pointed out that theories of adjudication typically make both descriptive and
normative claims; that is to say, authors purport to explain how and why judges decide
the way they do and, at the same time, intend to prescribe what they understand is the
right way to decide cases. At first glance, such a combination of descriptive and
normative assertions constitutes a rarity among theoretical elaborations, like those on
epistemology or morality, which only advance claims of one sort.18 There seem to be
several reasons for this rarity. Brian Leiter, for instance, has pointed out to the fact that
Following such a line, most of the realist literature is devoted to the identification of the several non-
legal elements that influence judicial decision-making. For instance, Jerome Frank famously claimed: “the
personality of the judge is the pivotal factor in law administration.” See his LAW AND THE MODERN MIND
(Brentano’s, 1930), 111.
Brian Leiter, NATURALIZING JURISPRUDENCE (2007), 23-4 (citations omitted).
Brian Leiter, Heidegger and the Theory of Adjudication, 106 YALE L. J. 253 (1996), 255-8.
among theorists “[t]here is a presumption that current adjudicative practice is roughly
right,” which would explain why “a good description (of a roughly right normative
practice) necessarily constrains prescription.”19 According to this view, then, the
propositional composition of the discipline is shaped by a conjuncture: the fact that the
academic community shares a particular attitude toward the practice being studied.
Absent that, we would be left with competing descriptions of the practice without any
Moreover, the conjuncture seems not only to be responsible for the propositional
composition of the theories, but also for the dominant content of their normative claims.
As Leiter himself explains, “the normative ambition of a theory of adjudication is that
judges ought to do «more explicitly and more consciously» what it is the theory claims
(as a descriptive matter) they largely do already.”20 Here, again, this appears to be the
result of the development of a descriptive account by authors who happen to consider
the current practice a good one, rather than the development of an ambitious normative
theory of judicial decision-making. I will call it the “normative appendix” [n] of a primarily
descriptive account [D] of adjudication.
But a different situation could be imagined—one where scholars are not so pleased with
the current practice and their main objective is to produce a normative account of how
judges ought to perform. The question is, then: Do the authors of such (primarily)
normative theories have reasons to formulate descriptive claims too? If they do not,
their contributions would be limited to the prescriptive plane, which would contradict the
assertion that theories of adjudication typically combine descriptive and normative
components. But we need not go that far. I could mention at least two strong reasons
for normative theories of adjudication to incorporate a descriptive chapter. On one hand,
sound normative proposals need to be feasible, and it is their burden to prove so.
Drawing a connection between the author’s normative proposals and a description of
some portion of the current practice that she finds praiseworthy will do the trick of
proving the ideal attainable (after all, to a certain extent judges would be doing it
Id. 257. Leiter calls this “The Presumption.”
Id. 258. Leiter calls this “the standard relation” between a theory’s normative and descriptive claims.
already).21 On the other hand, feasible normative proposals about adjudication intend to
be influential: Authors need to convince judges to modify their practices, and hence
address their work to these and other members of the legal community as well. Here,
reference to actual practice is a powerful argumentative and rhetoric tool to get judges
to adapt their decision-making to some concrete examples they are already familiar
with.22 The less distance there is between the ideal practice and the actual practice, the
more successful the normative proposal can be; which provides authors with an
incentive to present their ideal models as close as possible to some portions of the
existing practice.23 I will refer to the “descriptive appendix” [d] to normative theories of
adjudication [N] to account for the inclusion of descriptive-content propositions within
primarily prescriptive theories.
In sum, theories of adjudication do combine descriptive and prescriptive claims, albeit
for different reasons and with different intentions, in a much richer landscape than that
suggested by Leiter.24 The function of a descriptive appendix [d] within an ambitious
normative theory is very different from that of the propositions constituting the core of a
primarily descriptive one [D]. Regarding the first case, the common feature to both
instances referred above—and to any other that may exemplify the use of a descriptive
appendix—is that the normative core of the theory is not dependent on any description.
Claims about how judges decide cases are introduced either to prove a crucial
conceptual point (which could be done otherwise25), in the first case; or for strategic
reasons, in the second. Of course, every descriptive claim is subject to evaluation on its
True, it seems the theory could not connect the prescribed practice to a description that purports to be
comprehensive of the whole current practice, unless it does so to criticize some portions of that practice
as deviations of the proposed standard. Certainly, feasibility could be covered without description, but
pointing out to a practice already in existence is an easier way of making the same point.
This is basically the strategy of Einer Elhauge in STATUTORY DEFAULT RULES: HOW TO INTERPRET
UNCLEAR LEGISLATION (Harvard U.P., 2008).
This is a very common strategy, exampes of which can be found in John H. Ely, DEMOCRACY AND
DISTRUST: A THEORY OF JUDICIAL REVIEW (Harvard, 1980), and Carlos S. Nino, THE CONSTITUTION OF
DELIBERATIVE DEMOCRACY (Yale U.P., 1996) (both arguing that what they are asking courts to do is, to a
certain extent, what judges are already doing if observed carefully enough).
Leiter’s univocal understanding of theories of adjudication sits at the very base of his reconstruction of
the formalist-realist dispute. See his NATURALIZING JURISPRUDENCE (2007), 23-4 n.47, and Legal
Formalism and Legal Realism: What is the Issue? 16 LEGAL THEORY 111 (2010), 111 n.1.
By showing that the practice of adjudication, if amended in the prescribed way, remains possible. A
descriptive appendix does not intend to provide accurate and complete descriptions of the practice, it only
seeks to establish a baseline for normative implications.
accuracy and those that compose some descriptive appendix are not immune to such
scrutiny. But attempting to rebut this kind of theory by proving a descriptive appendix
inaccurate will do little justice to the normative core of the proposal. Conversely, claims
composing the core of a descriptive theory are supposed to provide a comprehensive
account of the practice (or most of it, at least) and will be defective if they fail to do so.
Legal realism, in my view, is a clear case of a primarily descriptive theory with a
normative appendix. For one, if the Realists wanted to study adjudication in a way that it
is “useful for lawyers,” it is not only obsolete but also close to self-defeating to offer
normative advice for judges (which, if followed, could render the description
inaccurate!). However, more explicitly, the main exponents of realism either saw no
point in telling judges how they ought to decide (Frank), or just thought that judges
should keep on doing what they already would do anyway (Llewellyn).26
If Realism and Formalism are to be compared on their respective descriptive merits,
therefore, it is not enough to think of them as theories-of-adjudication simpliciter. Once
this is understood, the first case to be discarded is (a) the construction of Formalism as
a primarily normative theory with a descriptive appendix [N; d], as explained, because of
the lack of equivalence of its descriptive aims with those of realism [d≠D]. Indeed, for
such a comparison to work Formalism has to be construed either as (b) a primarily
descriptive theory of adjudication—one that, like Realism, intends to provide a complete
and accurate explanation of how (appellate) judges decide cases—or, at least, as (c) a
theory of adjudication with strong descriptive ambitions over appellate adjudication. The
difference between both possibilities is the ambition of their normative project: for the
former, a normative appendix suffices [D+n]; for the latter, both the descriptive and
normative dimensions are somehow equivalent in weight [D+N]. Let us explore the
plausibility of each of these alternative constructions of formalism.
(a) Formalism as a primarily normative theory of adjudication [d+N], although ruled-out
by the interpretation of the Formalist-Realist dispute here discussed, is certainly a
plausible reconstruction. It represents the most commonly-held formalist position
nowadays: it does not matter that the judicial practice is or is not formalist, as it is clear
See Brian Leiter, American Legal Realism, in Martin P. Goding and William A. Edmundson (eds.), THE
BLACKWELL GUIDE TO THE PHILOSOPHY OF LAW AND LEGAL THEORY (2005) 50, 58.
that to a certain extent judges do not decide formalistically—the aim is that the practice
should be made more “formalistic” (for whatever that means).27 Under this version,
descriptive-like claims that may be included have no ambition to adequately (and much
less completely) describe the adjudicative practice, but are directed to make the
conceptual point that the normative version of adjudication advocated is possible.28
They constitute a descriptive appendix. It should be noted how the very point of
departure of normative formalism is that the practice is not formalistic enough: although
no anti-formalist (e.g., realist) description is typically provided, it constitutes an
assumption that the practice does not comport with the formalist standards that are
prescribed—although it perfectly could.
(b) One option to allow for a comparison with realism is to think of formalism as a
primarily descriptive theory of judicial decision-making, according to which judges do
decide following formalistic patterns [D] and it is desirable that they continue to do so
[n]. The construction, nevertheless, seems implausible because it does not correspond
to any of the formalist views in existence, as it openly disregards one of their major
chapters—namely, their normative project, which is reduced to a mere appendix.29 From
“vulgar” formalism to its most “sophisticated” elaborations, proponents of these ideas
are concerned with the way in which at least some of the practice of adjudication is
conducted, and intend to exert influence on judges to make them adapt their practices
This version has been called “Sophisticated Formalism” (Brian Leiter, Legal Formalism and Legal
Realism: What is the Issue? 16 LEGAL THEORY, 130-1).
The importance of conceptual claims for normative positions regarding adjudication is stressed for
other familiar arguments against formalism in Martin Stone, Formalism, in Coleman and Shapiro (eds.),
THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW, 166 (2002), 174-80.
Brian Tamanaha has argued that this view was not endorsed in the United States before realism. See
BEYOND THE FORMALIST-REALIST DIVIDE, 28-40.
Not even vulgar formalism in America endorses a descriptive formalist picture of judicial decision-
making with a normative appendix (claiming that judges just need to continue being formalists as they
supposedly already are). It is exactly the other way round. In the context of both confirmation hearings
and conservative rhetoric about the judicial function, the general descriptive picture is that some judges
are formalists (follow the law) and others are “activists” (don’t follow the law). Because vulgar formalists
believe that the law is mostly or absolutely determinate (conceptual formalism), they understand that
“activism” in the judicial function is solely caused by judicial will, hence pathological and should be
eradicated (normative formalism). So the concern is precisely that some judges do not perform as
formalists, although they could, and should be required to do so. The strong dimension of American
vulgar formalism is clearly the normative one. Accordingly, when Elena Kagan said during her
confirmation hearings that it was “law all the way down” in appellate review, conservatives were not
(c) The other possibility for comparing is to consider formalism as both an ambitious
descriptive and normative theory. To explore it, we need to understand the way in which
these theories seek equilibrium between their descriptive and normative projects.31
Typically, authors strive to provide a descriptive account of some major portions of the
adjudicative practice, which they praise identifying the values it embodies, and finally
signal as the right way to engage in judicial decision-making. Any instance of the judicial
practice that falls outside of this characterization both remains without a thorough
description, and is normatively disapproved. In this way, descriptive and normative
projects overlap so that there generally seems to exist direct connection between the
core of the practice that is described and the one that is judged normatively right. At the
same time, the rest of the practice—the one that does not comport with those
standard—is both left unaccounted for and considered defective. This approach has
become dominant with the proliferation of interpretive accounts of the adjudicative
practice, which flourished pursuing both descriptive and normative endeavors.32
According to this observation, theories with strong normative ambitions, even when they
undertake strong descriptive goals (not merely an appendix), coordinate both projects in
a way that do not attempt to produce a description of those portions of the practice that
they consider wrong and suggest eradicating. One good example comes from American
Legal Realists themselves. While some prominent realists—like Frank and Llewellyn—
either rejected or reduced to a minimalist appendix their normative projects, some
others did think that the practice of judicial decision-making had to be adjusted in order
to accommodate major normative suggestions. The second group, then, is susceptible
of being read in interpretive key—like that described under (c). Indeed, Oliver W.
Holmes and Felix Cohen understood that realist findings about the real reasons judges
pleased to find in her someone who shared their descriptive intuitions about the practice of adjudication,
they were happy to find that a SC-Justice-to-be understood that the law, conceptually, is enough to
decide cases and should be followed when doing so.
Ronald Dworkin uses the term “equilibrium” borrowing it from Rawls. See LAW ’S EMPIRE 90, 424 n.17.
The most famous account of an interpretive methodology about legal practices is the one provided by
Ronald Dworkin in LAW ’S EMPIRE, Ch. 2 and 3. Dworkin’s theory of adjudication is in itself a great example
of how and interpretive equilibrium actually works: to the extent that Dworkin cannot be thought of as
describing the actual practice of adjudication is that he purports to prescribe it. See John Mackie, The
Third Theory of Law, in Marshall Cohen (ed.), RONALD DWORKIN AND CONTEMPORARY JURISPRUDENCE
(1983) 161, 163.
had for deciding cases notwithstanding the justifications provided in judicial
pronouncements ought to be projected into strong normative claims—namely, that
judicial discursive practices should be abandoned in favor of a more explicit exposition
of the real motives behind the decisions.33 It takes little effort to see that this constitutes
no normative appendix, but an ambitious attempt to change a central aspect of the
adjudicative practice. How, then, does this normative claim about judicial discourse
relates with the realist description of adjudication? Interestingly enough, the feature of
the practice that these authors proposed to abandon is precisely one that they poorly
account for: American realism certainly overlooked judicial reasoning as a salient
attribute of adjudication and declined to provide an explanation for it other than in terms
of a pretense.34 Of course, this has been widely denounced as a major weakness of its
In sum, all of the tentative reconstructions of formalism for a comparison with realism
seem problematic. The one commented under (a), however plausible, does not serve
the comparison because it lacks descriptive aspirations of significance. Conversely, an
understanding of formalism as a primarily descriptive theory such as (b) would certainly
solve this issue, but at the expense of overlooking what appears to be the most
conspicuous section of modern formalism—its normative ambitions. Finally, an
Oliver W. Holmes, The Path of the Law, 10 HARV. L. REV. 457 (1897). (“The fallacy to which I refer is
the notion that the only force at work in the development of the law is logic[; …] the notion that a given
system, ours, for instance, can be worked out like mathematics from some general axioms of conduct.
[…] The language of judicial decision is mainly the language of logic. […] But [… b]ehind the logical form
lies a judgment as to the relative worth and importance of competing legislative grounds, often an
inarticulate and unconscious judgment, it is true, and yet the very root and nerve of the whole proceeding.
You can give any conclusion a logical form. […] I think that the judges themselves have failed adequately
to recognize their duty of weighing considerations of social advantage. The duty is inevitable, and the
result of the often proclaimed judicial aversion to deal with such considerations is simply to leave the very
ground and foundation of judgments inarticulate, and often unconscious.” At 465-7.) (My emphasis.)
Edward Levi, AN INTRODUCTION TO LEGAL REASONING (Chicago, 1949), 1.
It was Ronald Dworkin who most prominently exposed that this description fails to capture important
features of our adjudicative practices (or, in Dworkinean terms, lacks of “fit”). See LAW ’S EMPIRE, 154-61.
(“[An interpretive reading of legal realism] can be rescued as a good explanation for […] adjudication only
by procrustean machinery that seems widely inappropriate. It can be rescued only if we do not take
judicial opinions at face value at all; we must treat all the judges who worry about problematical statues
and precedents as practicing some unmotivated form of deception. They must be seen as inventing new
rules for the future in accordance with their convictions about what is best for society as a whole, freed
from any supposed rights flowing from consistency, but presenting these for unknown reasons in the false
uniform of rules dug out of the past.” At 159-60.) A similar point was made earlier by H.L.A. Hart when
criticizing the realist’s views on adjudication in THE CONCEPT OF LAW, 138-41.
interpretive reading like the one advanced under (c) could arguably satisfy both
requirements. It would also imply, however, that formalism is distinctively odd as a
theory, as it would implement a coordination of projects that appears somehow unique
to the genre: theoreticians of formalism, according to this understanding, take it very
seriously to describe and make sense of some portions of the practice that end up
rejecting as undesirable.36 For formalists have a strong idea of how judges should
decide cases before them, namely, “following-the-law”-whether-they-like-it-or-not, and
as part of the interpretive tradition it does not seem clear why their theory ought to
provide an accurate description of the many ways in which judges actually depart from
that guideline—they should follow the law, period.37
There is, however, one sense in which it is not spurious to articulate such a challenge.
Surely, a formalist interpretation of adjudication must identify the salient features of that
practice and explain them in light of the values it serves. It can leave some features
unaccounted for, and typically will, but needs to make sense of a significant part of the
practice—its most central features—in order for it to remain appealing.38 It is altogether
Realists, as we know, will prove how mistaken the formalist description is and how it fails to make
sense of the practice of adjudication (at an appellate level).
Let’s think of another example. John Hart Ely was among the most important critics of the Supreme
Court’s decision in Roe v. Wade, 93 S. Ct. 905 (1973). See his The Wages of Crying Wolf: A Comment
on Roe v. Wade, 82 YALE L. J. 920 (1973), where Roe is criticized “because it is not constitutional law and
gives almost no sense of an obligation to try to be” (Id. 947), as it “lacks even colorable support in the
constitutional text, history, or any other appropriate source of constitutional doctrine” (Id. 943). According
to Ely, because the Court “is under an obligation to trace its premises to the charter from which it derives
its authority,” it “has no business imposing [a principle]” when “it lacks connection with any value the
Constitution marks as special” (Id. 949). Nevertheless, on occasion to provide his comprehensive views
on the practice of judicial review in DEMOCRACY AND DISTRUST, despite the fact that Ely explicitly
acknowledged the importance of Roe as a salient note on the practice he was interpreting (Id. 2-3, 248
n.52), he still didn’t account for the case other than to exemplify the illegitimate use of “substantive due
process” (Id. 15, 21) and to anticipate it wouldn’t last, following the fate of Dred Scott and Lochner (Id.
228-9 n.91). After studying Ely’s views, however, we are not left with the feeling that he owes us an
account of Roe only because his theory purports to interpret the practice of judicial review: to Ely, Roe
doesn’t fit the best account of the practice and therefore is just wrong; end of the story.
Dworkin identifies, separately, two kinds of limits that constrain the interpreter when providing an
account of a practice: On a first level, there seems to be a quantitative limit concerning the amount of
features that can be left out (“[T]here must be an interpretive stage at which the interpreter settles on
some general justification for the main elements of the practice identified [in advance]. This will consist of
an argument why a practice of that general shape is worth pursuing, if it is. The justification need not fit
every aspect or feature of the standing practice, but it must fit enough for the interpreter to be able to see
himself as interpreting that practice, not inventing a new one.” LAW ’S EMPIRE, 66). On a second level,
some features are so central to the practice that is very hard for an interpretation that overlooks them to
be successful (“Every community has paradigms of law, propositions that in practice cannot be
a different story to criticize an interpretation for not being able to account for some
salient features of the practice than to do it because of the author’s rejection of some
portions of the same practice as unjustified. The interpretive dispute germinates on the
first plane, not on the second one.39 If there is a valid challenge to the interpretive
version of formalism by its realist counterpart, it can’t be that formalism doesn’t explain
how adjudication works on those cases where judges can’t be pictured as following the
law, because formalism openly regards that section of the practice as wrong and worthy
of abandonment. Hence, it can only be over the fact that formalism, like any successful
interpretation, should account for that phenomenon, since it constitutes a central feature
of the adjudicative practice.40
Put this way the dispute revolves around which features of the practice are central to it
and therefore cannot be left outside of any appealing interpretation. It is a dispute, thus,
about the shape of the practice itself—about the elements that constitute it. Jerome
Frank, one of the main realists, protested against scholarship of formalist inspiration—
here labeled a “liberal understanding or science of law”—in precisely these terms:
It is fair to ask how there can be any such understanding, liberal or otherwise, which
omits consideration of any part of the process by which judges arrive at their decisions.
[…] With few exceptions, our discussions of law posit an “ideal,” super-human,
passionless, judge. In an occasional aside we admit that a judge may be affected by
“weakness” when he allows his feelings to enter into his reasoning. But the manner of
referring to these “weaknesses” indicates a belief that they are exceptional and
challenged without suggesting either corruption or ignorance,” like “den[ying] that the traffic code [is] part
of the law” Id. at 88).
Dworkin is also careful to separate both dimensions. On one hand, he situates standard interpretive
disputes around the notion of “legal paradigms” that refer to some central features of the practice, the
exclusion of which would make an interpretation “deeply suspect” (“These paradigms give shape and
profit to interpretive debates about the law. They make possible a standard form of argument: seeking to
embarrass an interpretation by confronting it with a paradigm it cannot explain.” LAW ’S EMPIRE, 91-2). If a
theory fails to account for a salient feature of the practice, it has a problem of “fit.” On the other hand,
although interpretations can controversially differ about their normative projections toward the practice
(“No conception need justify every feature of the political practices it offers to interpret: like any
interpretation, it can condemn some of its data as a mistake, as inconsistent with the justification it offers
for the rest, and perhaps propose that this mistake be abandoned” Id. at 99), this dimension seems only
open to internal criticism within each conception.
Let us follow with the Ely example that I started on note 37. Of course, to conclude that the Justices
didn’t decide Roe like they should have, according to Ely’s evaluation, tells us almost nothing about why
they decided in the way they did. However, we may think it important that an interpretation of the practice
of judicial review accounts for Roe as an important episode in itself, or conspicuous example of a line of
cases that we understand too relevant to be discarded as “wrong.” The most popular criticism of Ely’s
theory hasn’t been to confront his views that Roe is a mistake, but rather that his interpretation misses an
important portion of the practice by leaving Roe (and other fundamental-rights cases) unattended.
pathological. Now even if the humanness of judges were pathological, it would deserve
explicit attention as part of “a liberal understanding or science of law.”
Notwithstanding, a quick review of the dispute between formalism and realism (or their
respective versions of judicial decision-making, to be precise) suggests that the
disagreement is too profound for it to be merely about an unaccounted feature, even an
important one: realists blame formalism for having the picture of adjudication completely
wrong—they do not seek to amend or complement a formalist interpretation, since to
their minds there is nothing on it that is worth preserving.42 Indeed, what formalists
consider the margins, the pathology, is actually the core for realism. As Frank goes on
But calm observation discloses that such “frailties” are normal, not diseased; recurrent,
not exceptional. And a study of law which shoves the consideration of the normal and
usual into a footnote and labels it “unusual and morbid” cannot lead to anything like an
adequate understanding of the subject.
The way in which the dispute can be fairly reconstructed suggests, then, that the rough
data formalism and realism are trying to interpret is different. Their understanding of
what constitutes the practice that is to be interpreted overlaps to a certain extent but not
completely, far from it. Both interpretive perspectives name their object-practice in the
same way (“the law” or “adjudication”); this and the partial overlapping contribute to the
perception that we are before an interpretive dispute, which sits at the base of the
understanding of formalism we have been considering so far. But such conception
seems to completely underestimate the disagreement between both positions.
To sum up, so far I have explored one of the implications of situating the formalist-realist
dispute on the descriptive plane, namely, (i) to assume a descriptive reading of
formalism. For this purpose, I have identified three possible versions of formalism as a
theory of adjudication: (a) primarily normative with a descriptive appendix; (b) primarily
descriptive with a normative appendix; and (c) interpretive. Only the two latter versions,
(b) and (c), will allow for a comparison with realism along descriptive lines;
Jerome Frank, LAW AND THE MODERN MIND (1931), 143-4.
For an example of a less radical critique to formalism see Herbert Hart’s famous argument about the
“open texture of the law,” in THE CONCEPT OF LAW , 124-36. His analysis, however, acknowledged that
formalism had a point in describing how the law may work for what we call “easy cases.” Shapiro
analyzes Harts response as a commitment to “moderate anti-formalism.” See LEGALITY, 227-31, 248.
Frank, Id. at 144.
nevertheless, only one of these two, (c), appears as a plausible reconstruction taking
into account the strong normative claims that formalists advance. (The first one, (a), is
also perfectly plausible, although irrelevant to this discussion). I have further showed,
however, how unfitting it is to think of the formalist-realist controversy in terms of an
interpretive dispute. The review was also useful because it gave us a sense of what
may be at play on the formalism-realism debate.
Let us briefly explore the other two implications of this reconstruction. For a comparison
with realism to work, formalism has to be understood as (ii) a descriptive theory of
adjudication, instead of a theory that purports to explain how the law works more
broadly (on the courts and outside them). Furthermore, (iii) formalists have to be
thought of as attempting to account for what goes on not when all judges decide cases,
but particularly when appellate judges do.
Even if we imagine a descriptive version of formalism, what would it be describing?
Surely, the formalist picture of the legal materials determining conduct appears to fit
some portions of the legal practice much better than others. Two issues stand out here.
First, the law seems to influence our conduct more directly on everyday life, in a way
similar to what formalism suggests: we pay taxes, stop at red lights, comply with the
legal requirements to write a contract, etc., merely because the norms say we should.44
Secondly, lower judges deal with a majority of disputes where the law is not an issue:
facts require proving, compensation needs to be determined; but the law governing the
matter is rather clear. A formalist theory may do just fine explaining how the law works
on everyday life or how judges decide easy cases.45 Moreover, if we are to consider
formalism as an interpretation of a practice, unless it is an absolutely inapt interpretation
(which seems to be the realist’s evaluation), it should account for at least some salient
features. Why can’t realists recognize that formalists describe these portions reasonably
well? The short answer is that realism doesn’t care about the law outside the courts or
even litigation concerning easy cases.
Hart’s criticism of “rule-scepticism” (i.e. legal realism), for instance, points out that the law is much more
than the court’s decisions and their prediction. See THE CONCEPT OF LAW , 137-8.
Coincidentally, Dworkin concedes that what he calls “strict conventionalism” (basically, legal positivism)
can explain how the practice works with easy cases and cases that do not come to court. See LAW ’S
Realists are only interested in appellate review.46 As Llewellyn famously claimed, the
realist description of how judges decide cases can be appreciated “in any case doubtful
enough to make litigation respectable.”47 Max Radin also pointed out that judicial
“[d]ecisions will consequently be called for chiefly in what may be called marginal cases,
in which prognosis is difficult and uncertain. It is this fact that makes the entire body of
legal judgments seem less stable than it really is.”48 Radin is clearly aware of the
distortion that such a selection of cases may produce with regard to the picture of
adjudication one develops. But it is Llewellyn himself who provides us with a detailed
account of the issues at stake when we write, investigate, and discuss about a practice
that is labeled “the law” or “adjudication,” but only addresses one peculiar sample of
what we purport to explain or justify:
There is one more matter to take up […]; it is the most important matter in all this odd-lot
jumble of importances. […] That matter is the almost bias of all present and past
discussion about law. We talk about legislatures and of courts of last resort. We talk of
almost nothing else. […] You find me trying to analyze the work of “courts” and
“judges”—criticizing here, moved there to admiration. What judges? Judges of appellate
courts. You find me dealing with “the law” and what it is and does. I say its center is the
action of officials, all law officials—and no sooner say it than I slip off my own platform to
land for lecture after lecture in discussion purely of these courts of high review: what
[…] Law is, to the community, what law does. What picture of the doing can you find in
all this study of appellate courts alone?
[…] We can, too, call attention, now and again, to the limitations of the picture thus set
up. But we cannot keep the picture from distortion, as you go more and more deeply into
a single part of it, neglecting all the others. We cannot take you into the other parts. We
have neither the wit nor the knowledge—neither we, nor any man.
Leiter insistently reminds this note on his several accounts of realism. See his American Legal Realism
(2005), 52-3, and 64; and NATURALIZING JURISPRUDENCE (2007), 20, 23, 79, and 115.
Karl Llewellyn, Some Realism about Realism—Responding to Dean Pound, 44 HARV. L. REV. 1222
Max Radin, In Defense of an Unsystematic Science of Law, 51 YALE L. J. 1269 (1942), 1271.
Karl Llewellyn, THE BRAMBLE BUSH (1960), 101-3 (Emphasis on the original). Llewellyn also explains
why he particularly regrets not being able to consider how the law works outside the courts and in
litigation before lower courts (“If, as I claim, what appellate judges do is vastly more important than what
appellate judges say, that can be only because importance to other people, to the layman, to the poor
devils to whom they do it, appears to be the primary measure of importance. [… However] [h]ere in th[e]
moving mountain of the cases unappealed, is the impact of the officials on society—even within the realm
of litigation. Beyond, there is the massive impact of the administrative machine. By my own showing, on
my own premises, these are what count. I pass them by. Out of my own mouth, damned.” Id. at 101-2).
(Emphasis on the original).
In conclusion, a descriptive version of formalism is evidently hard to defend unless we
interpret it as explaining how the law works outside the courts or how judges decide
easy cases. From this point of view, realists seem to have dragged formalism to explain
a particular portion of the practice that falls outside their primary area of interest; and
formalism failed. The collapse, of course, is not complete: some descriptive formalism
could still work as a theory of how the law influences our conduct daily, or as a theory of
adjudication for easy cases. But this is somehow left out of the picture. All we seem to
care now is for the practice as the realists defined it: the interesting cases, the hard
ones, the ones that get to appellate review.
b. Legal indeterminacy and conceptual formalism
There is yet another familiar reconstruction of the formalism-realism disputes. American
Realists could be thought of as advancing the claim of legal indeterminacy, particularly
for cases before appellate review.50
According to this plausible version, the realists pointed out that legal materials and legal
reasoning are not determinative of one legal solution for most cases reaching appellate
courts. That is to say, even extremely competent and honest judges, who master the
law and take the required time to study the case and identify and research the
materials, will find that most of the times those materials do not require them to make a
determinate decision but instead are compatible with several remarkably different
This accounts for saying that there is something about the law (i.e., legal materials plus
reasoning) that impedes judges to work “mechanically” in reaching decisions to many
relevant cases, namely, the difficult ones. Because difficult cases constitute the critical
mass in the docket of every appellate court, any serious study of adjudication at this
level must depart from the premise that legal materials simply cannot—not merely “do
See Brian Leiter, American Legal Realism, 51-2; and Legal Indeterminacy, 1 LEGAL THEORY 481 (1995).
See also Martin Stone, Formalism, in Jules Coleman and Scott Shapiro (eds.), THE OXFORD HANDBOOK OF
JURISPRUDENCE AND PHILOSOPHY OF LAW , 166 (2002).
See, for example, Llewellyn, Some Realism about Realism (“in any case doubtful enough to make
litigation respectable the available authoritative premises — i.e., premises legitimate and impeccable
under the traditional legal techniques — are at least two, and that the two are mutually contradictory as
applied to the case in hand.” at 1239); and THE BRAMBLE BUSH, 74-6.
not”—control the outcomes but in a very marginal portion of cases. At first sight, such a
point of departure would certainly undermine any scholarly agenda that aims to
contribute to legal determinacy by expanding the influence of the legal materials, like
that of doctrinal internalism—after all, by making new factual situations more
determinate scholars at best reduce the chances of judicial disputes based on those
facts to get to appellate review, and hence do not change the composition of the
Let us take a closer look to this picture of adjudication. There are surely many reasons
why cases get to court. Among them, we may roughly distinguish two major groups:
reasons concerning the complexity of the case itself and other circumstantial reasons
(like the fact that the parties may hate each other so badly as to foreclose the possibility
of a settlement; strategic decisions concerning time; litigation-prone lawyers; economic
incentives; etc.). If we set aside the second group to concentrate in the nature of the
cases at hand, the standard story tells us that a dispute may be worth judicial
intervention because of one or a combination of the following two reasons: controversial
determination of facts, and vagueness, ambiguity, or other source of indeterminacy of
the legal materials governing the issue.52 Typically, in a legal system where trial courts
settle authoritatively the facts and appellate courts have jurisdiction over matters of law
alone, it is to be expected that only the cases that present some difficulty regarding the
law have a chance to reach appellate review. To put it more clearly, new factual
situations challenging the existing legal materials come up recurrently in real life and
although this is a feature that would increase the likelihood that a case is brought to
court, it is certainly neither sufficient nor even indicative of it. However, because of
institutional arrangements regarding the courts and their jurisdiction, there are reasons
to believe that among the cases that for whatever reason are litigated, an appellate
court will review in particular those where there is a difficulty with respect to the law
governing the facts.
Furthermore, if the legal materials can be expanded so as to rule more clearly new
Cases might be “hard” because of several reasons concerning the law, some of which are not captured
by the concepts of vagueness and ambiguity of the legal rules. Sometimes, there are cases where it is
simple to identify what the law requires and yet judges may have doubts about how they ought to decide.
See Stone, Formalism, 179.
factual situations (by means of legislation, precedent, or even doctrinal scholarship),
such a development would be innocuous to the picture just presented—appellate courts
would simply loose any interest on these cases on the basis that they are no longer
“difficult.” Appellate courts operate at the vanguard of the law: their docket evolves
through addition of new factual situations that constitute a challenge to the present legal
materials (hard cases), and detraction of cases already subsumable under them (easy
cases). This scheme, however, does not purport to account for the whole composition of
the appellate docket: easy cases may as well get appellate review, for instance, on the
grounds of mistakes made by trial judges in the application or interpretation of the law—
after all, it is in the spirit of judicial review to allow higher courts to control for mistakes
that could have been made earlier in the process. Notwithstanding that, the picture does
provide a reasonable explanation for why the appellate docket may show an important
concentration of difficult cases—cases that would serve as strong examples for the
claim of legal indeterminacy.53
Now, if legal indeterminacy in such terms is legal realism’s main point, then what
version of formalism is entailed by this reconstruction? Not a very plausible one, to be
sure. Assuming formalists understand the law to be mostly determinate in everyday life
and even that most cases that get to courts can be decided employing some sort of
mechanical reasoning, none of these means that they would consider the law to be
absolutely gapless. The position that the law, on a minimalist definition of the class of
legal reasons,54 is completely determinate is absolutely implausible as a part of any
jurisprudential account: one hard case is enough to dismiss it.55
Such a conception of formalism is even less plausible when thought of as a foundation
for doctrinal internalism, for it is an assumption of doctrinal work that the legal materials
See Leiter, Legal Indeterminacy, 487-8 (“All that is needed for an interesting and worrisome claim
about indeterminacy is the truth of a certain thesis about the Local Indeterminacy of reasons—to wit, that
the law is indeterminate in, for example, the important cases or, more generally, in cases that are actually
It is important to clarify this to keep Ronald Dworkin’s conception out of the picture, since he does not
endorse a minimalist or austere definition of the class. See Brian Leiter, Legal Indeterminacy, 1 LEGAL
THEORY 481 (1995), 485.
Of course, here too, “[t]he ease of this maneuver ought to raise questions about the interest of the
«premise» being refuted.” Brian Leiter, Legal Indeterminacy, 486, n. 19 (a contrariu sensu, pondering the
significance of the argument against the thesis that the law is indeterminate for all cases which consists of
showing the existence of at least one easy case).
run out and therefore need to be developed by the scholar, i.e. they need to be
expanded to cover cases not explicitly regulated. Moreover, as noted, American legal
realists were not completely skeptic about the determinacy of the legal materials or,
more to the point, about the possibility that more carefully designed rules could prevent
cases from getting to appellate courts. Karl Llewellyn, for example, seems to have
drafted Article 2 of the Uniform Commercial Code with this in mind, which apparently
contributed to eliminate appellate litigation regarding some of its sections.56 Similarly, to
at least some of the realists, their empirical observations of legal rules failing to be
determinative was just an invitation to reformulate those rules “in more fact-specific
ways,” hence contributing to their determinative power.57 This, of course, would not
eliminate indeterminacy (on appellate review), but it would represent a worthy
contribution from legal scholars to identify technical problems of indeterminacy and help
to correct them by perfecting the rules. If this is the case, then the agenda of doctrinal
internalism and the realist construction of legal indeterminacy seem more than
compatible. And if one can be a good realist and still be committed to (allegedly
formalist-based) doctrinal work, the reconstruction of the formalism-realism dispute in
terms of different positions regarding legal determinacy does not seem to make a lot of
sense: formalists could easily agree to the picture of adjudication that I presented
above, which would not only be compatible with an agenda of doctrinal scholarship, but
would probably encourage one too.
Finally, there is one more point to make that will help appreciate the impact of the realist
argument on indeterminacy. H.L.A. Hart famously criticized realism because of its
extreme depiction of legal indeterminacy. He and the realists differed on the causes of
indeterminacy and yet, both agreed that to a certain extent the legal materials were far
from determinative. Hart’s point is then that the realists exaggerated this feature of the
See James White, The influence of American Legal Realism on Article 2 of the Uniform Commercial
Code, in Krawietz, MacCormick & von Wright, PRESCRIPTIVE FORMALITY AND NORMATIVE RATIONALITY IN
MODERN LEGAL SYSTEMS 401 (Duncker, 1994), 412.
Leiter, in NATURALIZING JURISPRUDENCE, discusses Oliphant’s approach in this sense (“Oliphant’s
critique was that the ‘legal rules,’ as articulated by courts and scholars, had become too general and
abstract, ignoring the particular factual contexts in which the original disputes arose. […] [He] argued that
a meaningful doctrine of stare decisis could be restored by making legal rules more fact-specific.” at 30).
Herman Oliphant, A Return to Stare Decisis (1928), in Fisher, Horwitz, and Reed (eds.), AMERICAN LEGAL
law, even when considering only the realm of adjudication.58 On an interesting article,
Brian Leiter has tried to identify the disagreement between the two, concluding that:
While both acknowledge indeterminacy in law, and while both acknowledge, accordingly,
that rules do not determine decisions in some range of cases, they clearly disagree over
the range of cases about which these claims hold true. Theirs, in short, is a
disagreement as to a degree, but it is a real disagreement nonetheless. While Hart
would locate indeterminacy, and thus the causal irrelevance, of rules “at the margin,”
Realist skepticism encompasses the “core” of appellate litigation.59
So, in the end, to Leiter the disagreement can be thought of in terms of a “half-full/half-
empty” picture. It is interesting to note, however, that while the realists are concerned
with “appellate litigation,” Hart never refers to it specifically and merely analyzes “the
function of rules in judicial decision.”60 Thus, it is perfectly reasonable to think that,
contrary to what Leiter suggests, the disagreement between the two cannot be
understood as a matter of degree—it may not even be a legitimate disagreement at all.
As we have seen, difficult cases tend to concentrate at the appellate level, which may
explain the realist perception that indeterminacy affects the core of the practice instead
of the margins.
III. Conclusion: the portion of the law that is relevant
As I have shown, both standard explanations of the formalist-realist divide have failed to
provide a reasonable account of formalism. On one hand, it seemed that formalism and
realism couldn’t be competing theories at the descriptive level because (i) an
interpretive reconstruction doesn’t account for the level of disagreement among them,
and (ii) even if both had descriptive ambitions it is impossible to think of them as being
designed to account for the same segment of the legal practice. On the other hand,
once extreme theories of legal determinacy are discarded, the picture of appellate
adjudication mostly dealing with cases of indeterminacy seems accepted by both
formalists and realists. Moreover, if their dispute is reconstructed as one of degree, our
See Hart, THE CONCEPT OF LAW, 124-54.
Leiter, NATURALIZING JURISPRUDENCE, 78 (Original emphasis).
Hart, THE CONCEPT OF LAW, 138.
doubts about there being a meaningful disagreement start coming back.
These failures, however, have pointed to what may be a more successful explanation.
Formalists and realists appear to be describing or interpreting different segments of the
legal practice: while formalists seem to be focusing on adjudication in general (or even
legal events more broadly), realists only care about appellate courts.
Legal realism has clearly externalized American legal scholarship. The realists have
redefined the study of law as the study of the practice of law (in particular, the practice
of judicial decision-making). Once this move is made, the shape and boundaries of the
practice are open for discussion and it is perfectly legitimate to drop portions that are
not interesting to explain—trial courts and extra-judicial legal events can be easily
forgotten. The legal picture that doctrinal scholarship knew is dramatically altered: legal
indeterminacy and judicial discretion are exacerbated to the point where they can’t
continue to be overlooked—from the margins to the core. Doctrine is useless; not
because there is nothing it can do (judges and lawyers could still find it helpful), but
because it can offer no answer to our new problems. Externalism is here to stay.
When scholars in the United States talk about constitutional law, their distinctively
externalist understanding of the subject shapes their discourse in a way that is not so
easy to grasp for scholars of other parts of the world. Conversely, European or Latin
American discussions about constitutionalism are developed on top of an internalist
basis that to a certain extent can make it difficult for an American scholar to fully
understand them.61 Sometimes, in spite of using virtually the same words, it is not silly
to think they might not be talking about the same.
For an example of this problems, see the exchange of Von Bogdandy (The Past and Promise of
Doctrinal Constructivism) and Post (Constitutional Scholarship in the United States), 7 INT’L J. CONST. L.