METHODS OF LEGAL REASONING
Law and Philosophy Library
VOLUME 78
Managing Editor
FRANCISCO J. LAPORTA, Department of Law,
Autonomous University of Madrid, Spain
ALEKSANDER PECZENIK, Department of Law, University of Lund, Sweden
FREDERICK SCHAUER, John F. Kennedy School of Government,
Harvard University, Cambridge, Mass., U.S.A.
Former Managing Editors
AULIS AARNIO, MICHAEL D. BAYLES†, CONRAD D. JOHNSON†,
ALAN MABE
Editorial Advisory Board
AULIS AARNIO, Research Institute for Social Sciences,
University of Tampere, Finland
ZENON BANKOWSKI, Centre for Law and Society, University of Edinburgh
PAOLO COMANDUCCI, University of Genua, Italy
ERNESTO GARZÓN VALDÉS, Institut für Politikwissenschaft,
Johannes Gutenberg Universität Mainz
JOHN KLEINIG, Department of Law, Police Science and Criminal
Justice Administration, John Jay College of Criminal Justice,
City University of New York
NEIL MacCORMICK, European Parliament, Brussels, Belgium
WOJCIECH SADURSKI, European University Institute,
Department of Law, Florence, Italy
ROBERT S. SUMMERS, School of Law, Cornell University
CARL WELLMAN, Department of Philosophy, Washington University
METHODS OF
LEGAL REASONING
JERZY STELMACH
.
BARTOSZ BROZEK
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN-10
ISBN-13
ISBN-10
ISBN-13
1-4020-4936-6 (HB)
978-1-4020-4936-1 (HB)
1-4020-4939-0 (e-book)
978-1-4020-4939-2 (e-book)
Published by Springer,
P.O. Box 17, 3300 AA Dordrecht, The Netherlands.
www.springer.com
Printed on acid-free paper
All Rights Reserved
© 2006 Springer
No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form
or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise,
without written permission from the Publisher, with the exception of any material supplied
specifically for the purpose of being entered and executed on a computer system, for exclusive
use by the purchaser of the work.
PREFACE
Anyone reflecting on the methodology of legal reasoning faces a difficult
task. The number of methodological theories in jurisprudence and the
vast literature on the subject are not the only problems that have to be
taken into account. Perhaps the most striking difficulty concerning the
methodology of legal argument is the heated debate between jurists, legal
theorists and philosophers of law that has been recurring since at least
nineteenth century.
Therefore a justification is needed for writing yet another book concerning the methods of legal reasoning; a book that aims to cover a lot
of what has already been proposed in legal theory. We believe that there
is such a justification. First, the perspective that we adopt in the present
book is unique, at least in some respects. We venture to look at the
methodology of legal reasoning “from the outside”, i.e. from a more general, philosophical perspective, while taking into account the “hard reality” of law. This perspective enables us to ask questions about the
justification for the methods of legal argument presented.
Second, we do not want to defend one, paradigmatic conception of
legal reasoning. On the contrary, we put forward the thesis that there is
a plurality of argumentative methods. The plurality, however, does not
lead to relativism in legal decision-making.
Third, we reject any hierarchy of the methods of legal reasoning, and
take the view that one can speak only of the precision and flexibility of
different methodologies.
Finally, we would like to show that the methodological conceptions of
jurisprudence constitute a coherent element of the humanistic methodology. The important aspect here is that the methodology of legal argument is much more precise and much better developed than the
“general” humanistic methodology. However, this does not mean that
there are no peculiarities of legal methodology, as we try to highlight in
the course of our presentation.
v
vi
FOREWORD
This is a substantial revision of our book that appeared in Polish
under the title “Metody Prawnicze” (Zakamycze Publishing House,
Kraków 2004). Chapters 1 and 3 have undergone the biggest revision and
a new chapter 6 has been added.
.
Jerzy Stelmach, Bartosz Brozek
Kraków – Kiel, September 2005
CONTENTS
1 Controversy over Legal Method in the Nineteenth
and Twentieth Centuries
1.1 Three Stances
1.1.1 The Rejection of Method
1.1.2 Methodological Heteronomy
1.1.3 Methodological Autonomy
1.2 Methods of Legal Reasoning
1.3 Logic – Analysis – Argumentation – Hermeneutics
2 Logic
2.1 Introduction
2.2 Classical Logic: Propositional Logic and First
Order Predicate Logic
2.2.1 Presentation of Calculi
2.2.2 Paradoxes of Material Implication
2.2.3 Examples
2.3 Deontic Logic
2.3.1 Possible World Semantics
2.3.2 Deontic Logic
2.3.3 Paradoxes of Deontic Logic
2.3.4 Examples
2.4 Logic of Action and Logic of Norms
2.4.1 Two Types of Obligation
2.4.2 Logic of Action
2.4.3 Jørgensen Dilemma
2.5 Defeasible Logic
2.5.1 The Concept of Defeasibility
2.5.2 Defeasible Logic
2.5.3 Objections Against Nonmonotonic Logic
2.5.4 Examples
2.5.5 Two Remarks
2.6 Summary
vii
1
1
1
2
6
9
12
17
17
20
20
23
24
29
29
30
33
35
40
40
41
46
49
49
51
53
54
63
63
viii
CONTENTS
3 Analysis
3.1 Introduction
3.1.1 The Notion of Analysis
3.1.2 History of the Concept
3.1.3 Analytic Philosophy
3.2 Linguistic Analysis
3.2.1 History and Basic Assumptions of Linguistic Analysis
3.2.2 Legal Conceptual Scheme
3.2.3 Speech Acts Theory
3.2.4 The Method and Its Limits
3.3 Economic Analysis of Law
3.3.1 Law and Economics
3.3.2 Idea of Economization
3.3.3 Limits of the Method
3.3.4 Conclusions
3.4 Summary
3.4.1 Features of Analysis
3.4.2 Analysis in Law
69
69
69
70
72
75
75
78
82
89
92
92
94
97
104
105
105
105
4 Argumentation
4.1 Introduction
4.1.1 Philosophies of Argumentation
4.1.2 Criteria of a Practical Discourse
4.2 Two Conceptions of a Legal Discourse
4.2.1 The Topical–Rhetorical Conception
of Legal Discourse
4.2.2 Procedural Conception of Legal Discourse
4.3 Legal Argumentation
4.3.1 Claim to Universality
4.3.2 Structure of Legal Discourse
4.3.3 Applications
111
111
112
123
130
5 Hermeneutics
5.1 Introduction
5.1.1 The Beginnings of Hermeneutics
5.1.2 What Do We Not Know About Hermeneutics?
5.2 Hermeneutics as Epistemology
5.2.1 Methodological Current in Philosophical
Hermeneutics
5.2.2 Legal Receptions
5.3 Hermeneutics as Ontology
167
167
167
173
175
132
139
145
146
147
163
176
183
187
CONTENTS
5.3.1 Ontology of Understanding
5.3.2 Legal Receptions
5.4 The Understanding of the Law
5.4.1 Claim to Universality
5.4.2 The Nature of Hermeneutic Cognition
5.4.3 Applications
ix
188
193
198
199
200
205
6 Methods of Legal Reasoning from a Postmodern Perspective
6.1 A Summary
6.2 Dilemmas of the Contemporary Philosophy of Law
6.3 The Epistemological Approach
6.4 Unfinished Projects
211
211
214
216
219
Index of Names
Subject Index
223
227
CHAPTER 1
CONTROVERSY OVER LEGAL METHOD IN THE
NINETEENTH AND TWENTIETH CENTURIES
1.1
THREE STANCES
A fundamental question that already preoccupied jurists in Roman times
concerned the existence of a “legal method”. A positive answer to that
question led, in turn, to controversy over how to characterize the method,
or methods, used in legal reasoning. Consequently, three substantially different perspectives on the methodology of legal argumentation developed
in nineteenth and twentieth century legal philosophy.
1.1.1
The Rejection of Method
The first of these stances, by far the rarest, although very important for
the purposes of the discussion in question, not only puts into doubt the
scientific character of jurisprudence, but questions the very existence of
any legal methods.
Kirchmann. This opinion was expressed, inter alia, by von Kirchmann,
who in 1847 delivered a lecture entitled Die Wertlosigkeit der Jurisprudenz als Wissenschaft. Attacking the historical school established by
Savigny and Puchta, Kirchmann not only criticized the condition of
jurisprudence but even put into doubt the usefulness of legal knowledge,
which he labeled valueless and parasitic.1
A similar, although less radical, stance was adapted later by
Hutcheson and the Critical Legal Studies movement.
Hutcheson. Hutcheson, a representative of the intuitionist version of
American realism, claimed that a judge who has to decide a concrete case
must have recourse to his own intuition and imagination, since every case
is only a stimulus to which the judge reacts in order to make a good (just)
decision. This reaction is an irrational, intuitive, or emotional hunch, a
tentative faculty of mind, a kind of imagination or intuition that is characteristic of good lawyers.2 Therefore, there is no objectively reconstructible method of legal reasoning.
1
2
CHAPTER 1
Critical Legal Studies. The representatives of Critical Legal Studies, in
turn, believe that the traditional methods of investigating and teaching
law are useless. For them, “law is politics”; consequently, there is no such
thing as the method of law. The space for the possibility of critical studies is created only when an anti-positivist, anti-legalistic and anti-formal
attitude towards the law is adopted. The consequent critique leads to the
“deconstruction” of traditional methodology. Some elements of this way
of thinking can be traced in contemporary postmodern philosophy. From
a general philosophical perspective it has to be observed that postmodernism “begs the question” while showing the bankruptcy of all methods,
for it does it with the use of some methods, like deconstruction. A more
sympathetic account of such schools as Critical Legal Studies illustrates
that postmodernism does not claim the bankruptcy of methodology, but
only shows precisely its limitations, relativity and pluralism.
1.1.2
Methodological Heteronomy
According to the second, more moderate stance, jurisprudence has some
features of a “real science” but only under the assumption that it uses
methods of other scientific disciplines, like mathematics, logic, physics,
biology, or – in some cases – linguistics, sociology or economics.
Jurisprudence enjoys, therefore, the status of a science but only at the
cost of losing its methodological identity and autonomy. This stance
cannot be analyzed exclusively within the context of Droysen and
Dilthey’s well known distinction between naturalistic and anti-naturalistic paradigms in science. Jurisprudence can take advantage not only of
the methods of social sciences and humanities, but also of methods
developed in logic and natural sciences. Such distinctions as “naturalistic – anti-naturalistic” are, in fact, pointless, and as far as legal theory is
concerned – false. In jurisprudence, “outside” methods are very rarely
straightforwardly applied. Rather, there has always been a kind of adaptation of methods to suit the specific needs of lawyers. We will return to
this point in Section 1.2 of the present chapter.
Analytic philosophy of law. The most rigorous attempts to incorporate
mathematical, logical and linguistic methods into jurisprudence have
been made by analytic philosophers. In analytic legal theory, as in analytic philosophy, one can distinguish between two “methodological
types” or “wings”: logical (horse-shoe analysis), and linguistic (soft-shoe
analysis), which is usually confined to the analysis of ordinary language.
The proponents of the “hard”, logico-mathematical methodology in
jurisprudence developed deontic logics, i.e., logics of such concepts as
CONTROVERSY OVER LEGAL METHOD
3
“forbidden”, “obligatory” and “allowed”. One should also acknowledge
their formal attempts to deal with concrete legal-theoretic problems, such
as the idea of a legal system, and the theory of legal rules and principles.
Among the “hard” analytic philosophers who contributed to jurisprudence, one should mention G.H. von Wright, O. Becker, J. Kalinowski,
A. Ross, J. Woleński, C. Alchourron and E. Bulygin.
The “soft” methods of analysis were applied by those philosophers of
law who followed the Oxford School of Ordinary Language and the “second” philosophy of L. Wittgenstein. These thinkers commonly prioritized ordinary language over artificial formal systems. They did not
attempt to reform given conceptual schemes. On the contrary, their aim
was to describe as precisely as possible how analyzed concepts function
in ordinary language. Among “soft” analytic legal philosophers the
name of H.L.A. Hart deserves special attention.
Legal realism. Legal realism is, in turn, an example of naturalistic
methodology in law. According to realists, jurisprudence can be labeled
“a science” only when it uses methods developed in natural sciences, or
at least in empirically oriented disciplines such as sociology or psychology. Legal theory should, moreover, be descriptive in character. This
kind of methodology was employed by the school of free law, American
realism, sociological jurisprudence, Scandinavian realism and
.
L. Petrazycki.
School of free law. The conception of naturalistic jurisprudence was
designed to oppose legal positivism, which was criticized, inter alia, by
representatives of the school of free law. In a work, Methode d’interpretation et sources en droit privé positif, published in 1899, Geny says that
continental positivism assumes wrongly that the only source of law is
statute law. He goes on to argue that in the process of legal interpretation one should take into account also three other sources of valid law:
customs, authority and free investigation of a judge. Independent courtdecisions are, ultimately, the results of the judge’s will, the needs of
society and balanced private interests. The main representative of the
school, Kantorowicz, advanced similar theories.3
.
.
Petrazycki. Petrazycki aimed to develop an adequate theory of law. He
argues that such a theory can be constructed when based on psychology
of emotions, which he developed himself. The psychological analysis
.
enables Petrazycki to identify a class of phenomena that is characteristic
both of moral and legal emotions.4
4
CHAPTER 1
American realism. American legal realism, like sociological jurisprudence, applied the methods of empirical sociology, and also of psychology and economics. Questioning positivist formalism, American realists
claimed to turn to practice and to investigate “real law,” especially the
behavior of judges. The instigator of this movement was Holmes. In his
“manifesto” paper The Path of the Law, Holmes argued that, in order to
explain what law is, it is necessary to adopt the perspective of the hypothetical “bad man”, who is not interested in the problem of justifying
legal decisions or the rationality of law; rather, he is concerned with predicting how the judge will act in given circumstances.5 Amongst the most
important representatives of American realism one should mention
Llewellyn, Frank and Moore.6
Sociological jurisprudence. From the many sources of American sociological jurisprudence Comte’s sociology, Bentham’s utilitarian philosophy, German jurisprudence of interests, and American legal realism
should be mentioned. The most important representative of this school
is Pound. According to him, law should realize and protect six social
interests: common security, social institutions (like family, religion and
political rights), sense of morality, social goods, economic, cultural and
political progress and protection of an individual’s life. The last of these
“social interests” Pound deems to be the most important.7 In order to
realize those goals a new sociological jurisprudence, Pound argues, must
be developed.8
Scandinavian realism. Scandinavian legal realism, like American realism, regards law as an empirical fact. However, American realists treated
law as a kind of behavior of a certain social group, consisting of people
professionally preoccupied with resolving conflicts; Scandinavian realists, on the other hand, searched for the “essence” of legal phenomena in
the psychological reactions of individuals. Hence such concepts as “law”
or “obligation” are regarded by Scandinavian realism as psychological
facts. The father of this school, Hägerström, deems valueless all ideas
and concepts that are not developed within the context of what is
real. Therefore, the world of norms and rules propounded by natural
law, or by legal positivism is unacceptable. Concepts such as “law”,
“obligation” and “validity” are purely metaphysical without a reference
to empirical facts. They make sense only if we associate them with concrete emotions or psychological reactions caused by the use of such
notions. Lundstedt and Ross are among other representatives of the
Scandinavian school.9
CONTROVERSY OVER LEGAL METHOD
5
Apart from analytic philosophy and legal realism, there are also other
traditions that utilized methods developed in non-legal disciplines.
Attempts have been made to adapt to legal aims methods found in systems theory, economics, argumentation and hermeneutic philosophies.
As a result, several legal-theoretic conceptions have developed, and
among them the systems theory of law, the economic analysis of law,
theories of legal argumentation and legal hermeneutics.
Systems theory. Systematic analyses were applied already in nineteenth
century sociology, as exemplified in Comte’s work. Describing the methods of sociology, the founder of positivism offered some examples from
biology, demonstrating analogies between living organisms and society.
This idea was further developed in the 1970s by two biologists, Maturana
and Varela, who formulated the theory of so-called autopoietic systems.
An autopoietic system controls the process of its creation (a nice example of such a system is a cell). The opposite are heteropoietic systems,
which are incapable of self-regulation, i.e., they must be controlled from
“the outside”. In the 1980s attempts were made to formulate a theory of
autopoietic systems in the social sciences, including legal theory. That
this happened was mainly due to Luhmann and Teubner, who claimed
that autopoiesis, i.e., the ability to self-regulate, is a characteristic of
some social systems in developed societies. A legal system, regarded as a
set of communicative acts, can be considered to be such a system. Its
function is not, as is usually claimed, to regulate social life and to solve
social conflicts, but to secure and promote the normative expectations of
a society.10
Although the ideas of Luhmann and Teubner were inspired heavily by
biological and sociological theories, it should be noted that Kelsen, who
developed the concept of a “pure theory of law” and advocated the
methodological autonomy of law, wrote in a somewhat similar spirit.
Economics of law. The economic school developed in the 1970s in
the USA. The school, being a conservative movement, rejected the
more “left” Critical Legal Studies. Its representatives made use of
the ideas of British utilitarianism, especially the work of J. Bentham and
J.S. Mill, and of the theses of American realism and sociological
jurisprudence. The main representative of Law & Economics, Posner,
tried to show that the processes of creating and interpreting law comply
with some economic rules. The law is, or at least should be, economically
effective, i.e., its aim is to minimize social costs and promote the increase
of social welfare. A rational decision is a decision that is economically
6
CHAPTER 1
justified, and, hence, leads to the maximization of the welfare of the
given society.11
The argumentation and hermeneutic theories of law are far harder to
interpret and analyze than those theories discussed above. We will analyze them in detail in Chapters 4 and 5. Here we limit ourselves to a few
general remarks.
Argumentation theory. Contemporary argumentation theories are
based on various philosophical traditions, from ancient logic, rhetoric,
dialectics, hermeneutics and eristic to contemporary conceptions of analytical ethics (Stevenson, Toulmin or Baier), constructivist theory of
practical advice (Lorenzen and Schwemmer) and the practical and theoretical discourse of Habermas. It is justifiable, then, to say that argumentation theories are based on methodological conceptions developed
in other scientific disciplines (philosophy, logic, linguistics).12
Legal hermeneutics. Similar things can be said of legal hermeneutics. The
nineteenth and twentieth century concepts of legal hermeneutics developed as a result of the absorption of various kinds of general hermeneutics: the “methodological” as created by Schleiermacher and Dilthey, and
the “phenomenological” as developed by Heidegger and Gadamer. Legal
philosophy also witnessed some attempts to develop analytic hermeneutics, based on Wittgenstein’s later philosophy. One should bear in mind,
however, that in Roman times jurists had already attempted to describe the
most important principles of legal method constituting a special legal
hermeneutics. In the seventeenth and eighteenth centuries several works
devoted exclusively to legal hermeneutics were produced; in this context
one should mention Eckhardi’s Hermeneutica iuris, recensuit perpetiusque
notis illustravit, Wittich’s Principia et subsidia hermeneuticae iuris, or
Sammet’s Hermeneutik des Rechts. The transition between old legal
hermeneutics and the contemporary version that adopts the ideas of general philosophical hermeneutics is marked by von Savigny’s theory of
interpretation, outlined in Juristische Methodenlehre.13
1.1.3
Methodological Autonomy
The third of the stances presented here assumes that jurisprudence
enjoys, at least to a certain degree, a methodological autonomy and
develops its own, “inner” criteria of what constitutes a science. In this
case we have to look at normative anti-naturalism, which ultimately
declares that legal science should work out its own methodology, different from that of logic, mathematics and natural sciences on the one
CONTROVERSY OVER LEGAL METHOD
7
hand, and that of other social and humanistic sciences on the other.
This thesis, concerning the methodological autonomy of jurisprudence,
is defended, usually, with ontological and pragmatic arguments. As a
consequence, jurisprudence determines itself the minimal – material or
procedural – conditions for the acceptance of its theses. Naturally, the
“border” between this stance and the two discussed earlier is not as sharp
as it may seem at first glance. Autonomy cannot mean methodological
isolation. Lawyers have always employed not only “their own” methods,
but also techniques developed in other disciplines. Thus, controversy
arises over the question of whether jurisprudence is methodologically
autonomous, and not over whether one can use, in addition to “legal”
methods, tools adapted from other sciences. The latter question is usually answered positively. According to the theory of methodological
autonomy, however, one can make use of such “adapted” methods only
if they support what is achieved by specifically “legal” arguments.
Roman jurisprudence. The thesis that law is methodologically autonomous was advocated by Roman jurists. Although Roman jurisprudence
did not develop any general philosophy of law, as Roman lawyers concentrated on concrete legal issues, nevertheless it addressed some problems of an abstract nature concerning legal ontology, axiology and
methodology.14
As regards methodology, several types of written works occurred.
Among them one should mention problem-oriented treaties (Quaestiones,
Disputationes, Epistulae and Digesta), commentaries, textbooks (mainly
on Institutiones), works concerning important legal concepts and rules
(Regulae, Definitiones, Sententiae, Opiniones, Differentiae), instructions
for civil servants (Libri de officio), monographs and collections of formulae and court decisions. As a by-product of their main “practical” activity, the Roman jurists constructed the fundamentals of general legal
disciplines: methodology and theory of interpretation. In particular, such
concepts as definitio, regula, interpretatio or rationes decidendi were elucidated and defined. Furthermore, several important directives of legal
interpretation were formulated.15 Even today, hundreds of Roman legal
topoi are used both in legal practice and for theoretical purposes
(especially in relation to theories of argumentation, see Chapter 4).
Historical school. The idea of the “full” methodological autonomy of
legal science was expressed clearly in two nineteenth century schools of
thought: the German historical school and legal positivism. The first
step towards postulating this autonomy was the “detachment” of legal
8
CHAPTER 1
philosophy from general philosophy. The step was taken by Kant in
“Metaphysical elements of legal theory” and by Hegel in “Principles of
the philosophy of law”. But the beginning of the heated debate about
legal method is marked by Savigny’s Juristische Methodenlehre.
Savigny rejects both the a priori deductive methodology offered by
natural law theorists and formal-dogmatic methods, which, until then,
had been used in legal science. Instead, he proposes empirical research of
law as a historical social fact. The law is a product of the national spirit,
originating from the natural inner forces of that spirit. It is not, therefore, a simple expression of the will of the legislator.
One has to remember, however, that although Savigny’s conception
contributed to the establishment of the methodological autonomy of
legal sciences, it was based on ideas from “outside the law”, especially
from Hegel’s philosophy and Schleiermacher’s hermeneutics.
Legal positivism. From this point of view, legal positivism seems to be
methodologically “purer”. However, despite appearances, the term “positivism” has been used in reference to different conceptions and schools.
One can maintain that, in order for a theory to count as positivistic, it
has to include at least some of the following seven theses: (1) legal norms
are created and not “discovered”, as the proponents of natural law maintain, (2) creation of law is an expression of the will of the sovereign,
(3) law consists exclusively of norms or rules, (4) lawyers should obey the
law without any exception, i.e., all legal decisions should be made on the
basis of legal norms (rules), (5) there exists no necessary connection
between law and morality or between the law as it is and as it should be,
(6) research of legal concepts has to be distinguished from historical,
sociological, or psychological research, (7) the legal system should
be considered a “closed logical system”, in which every decision can be
inferred from predetermined norms or rules using logical tools.
According to these criteria, Austin’s theory of law, the continental
Begriffsjurisprudenz, Kelsen’s normativism and Hart’s analytic legal
theory count as positivist. By contrast, Dworkin’s conception of law
challenges three of the seven theses mentioned above (3, 4 and 5). Therefore, Dworkin’s work cannot accurately be classified under the heading
“positivism”.
Within legal positivism, substantial effort has been devoted to analysis of fundamental legal concepts. This analysis serves both as a basis for
carrying out the process of legal interpretation and as the conceptual
framework for legal dogmatics and philosophy of law. Such analysis was
carried out by all the main representatives of legal positivism, including
CONTROVERSY OVER LEGAL METHOD
9
Austin, Hart and Kelsen. The most detailed formal-dogmatic analyses of
law were offered by proponents of the continental version of legal positivism that – not without a reason – is called Begriffsjurisprudenz
(jurisprudence of concepts). One should mention in this context the
names of: Jhering, Geber, Winscheid, Binding, Bergbohm, Merkl, Liszt,
Thon or Bierling.
Did legal positivism persuasively demonstrate that jurisprudence is a
special kind of normative science that has its own methods of argument?
It seems that the answer should be negative. The only exception is
Kelsen’s Reine Rechtslehre, for although radical, this conception is consequential enough to be successfully defended. The “methodological
purity” of other kinds of legal positivism is, however, questionable.
Positivist analysis very often mixed up elements from the sphere of what
is (Sein) and the sphere of what should be (Sollen). The methods it
employed often had a sociological or psychological pedigree (Austin,
continental positivism, Jhering’s jurisprudence of interests).
1.2
METHODS OF LEGAL REASONING
What has been said so far shows that it is hard to settle the discussion
concerning the methodology of legal reasoning. There are several different stances in this controversy: on the one hand, it is questioned whether
methods of legal argumentation have any autonomy, or even whether
such methods exist, whilst on the other hand, philosophers defend the
autonomous character of legal methodology. Another noteworthy
aspect of the debate is its terminological chaos, which makes reaching a
clear conclusion even more elusive. This terminological chaos has
already been displayed here, at the most general level. We wrote about
“the methods of legal reasoning”, “the methods of legal argument”, “the
method of jurisprudence”, and “the methods used by lawyers”. In order
to clarify some basic terminological issues, let us identify three different
categories of application of “the methods of legal reasoning”. First,
those methods can be applied in legal practice (by judges, prosecutors,
barristers) in the process of creating and interpreting law. Let us call this
application of the methods in question practical. Second, one can speak
of legal-dogmatic application of those methods, i.e., their application
within specialist analyses carried out in various areas of law. Finally, one
can point to the theoretical application thereof; this occurs in legal
theory and legal philosophy.
The philosophical traditions described in the previous section each
explored different aspects of the application of “the methods of legal
10
CHAPTER 1
reasoning”, only rarely confining their analysis exclusively to one
category. For instance, legal positivism developed “tools” both for legal
practice and legal dogmatics. Legal realism proposed the use of psychological and sociological methods both in legal theory and legal practice.
Analytic philosophy proposed methods to be used primarily in philosophical reflections on what the law is, but those methods could just as
easily be applied in legal dogmatics and legal practice. Hermeneutics, in
turn, attempted to reconstruct the basic structures of any act of understanding, i.e., legal-theoretic and legal-dogmatic as well as “practical”.
Finally, argumentation theories can easily be applied both in legal
practice and legal dogmatics.16
One can make similar remarks – concerning potential applicability in
two, or even three, different categories – about most conceptions of
method of legal reasoning presented above. The same may be said of the
structure of this book: although emphasis is placed on the methods that
can be used in legal practice, we present also “tools” that are applicable
to legal-dogmatic and legal-theoretic questions. In any event, the “borders” between “practical”, “legal-theoretic” and “legal-dogmatic” applications of a specific method are often far from sharp.
The next terminological problem we encounter is the lack of a commonly accepted definition of “a method”. We could have proposed one
of our own, saying, for example, that a method is a set of rules of proceeding, that determine what actions must be undertaken in order to
achieve a given aim. We have not done so, for any definition could easily
dismiss many legal-theoretic traditions as not offering a proper
“method”. Clearly, this would result in a severe restriction on what we
discuss.
Analysis of legal methodological theories enables the separation of
those theories into three groups. Each group offers a different perspective on whether a method of legal reasoning can be said to exist, and on
the autonomy of that method. According to the first group, lawyers reason applying no identifiable method. The second group claims that
lawyers do use certain methods, but those methods are adapted from
other disciplines: sociology, economics, linguistics or psychology. Within
this group, there are two subtypes: the first claims the “pure adaptation”,
i.e., the method as used in law does not differ from the method as used
in its original discipline. The second subtype, in turn, suggests that law’s
adopted methods are modified to take into account the special character of
law. Finally, the third group asserts both the existence and the autonomy
of specifically ‘legal’ methods of reasoning.
CONTROVERSY OVER LEGAL METHOD
11
It is difficult to settle the controversy between proponents of the three
stances. It seems, however, that there are strong reasons to consider
abandoning both the first conception, which questions the very existence
of methods of legal reasoning, and the third, which claims their full autonomy. We have already indicated that the first stance is troublesome, in our
discussion of Hutcheson’s ideas and the Critical Legal Studies movement.
Is not it true that Hutcheson’s hunch can be regarded simply as a kind of
intuitive method? Further, Critical Legal Studies scholars tend to deconstruct all legal methods, i.e., to show all the assumptions standing behind
the traditional methodology. Such deconstruction does not inevitably lead
to the conclusion that there are no methods of legal argument. Rather, it
is possible to reach a more moderate thesis: that there is no unique method
of legal reasoning. From this point of view, the Critical Legal Studies
movement is not an example of the first stance, but rather recognizes the
pluralism and relativism of the methods of legal reasoning.
Equally problematic are those theories which defend the autonomy of
the methods in question. The only significant theory that asserts the
claim to autonomy is Kelsen’s. It is, however, based on very strong ontological assumptions and as such cannot serve as a commonly acceptable
defense of the third stance.
These considerations allow us to formulate two conclusions. First,
there is no unique, universally acceptable methodology of legal reasoning,
as there is no “special” legal method. Both the heteronomy and the pluralism of such methods must be stressed. Heteronomy – because there
exists no specifically “legal” method. Pluralism – because there exists no
unique method of legal reasoning. Second, it is necessary to point out
that the “heteronomic” methods used by lawyers are in a way “specific”.
This “specificity” arises because any method used in law is subject to certain modifications and limitations, mainly because the creation and interpretation of law are regulated by certain procedures imposed by valid law
(think, for instance, of legal presumptions or the distribution of the burden
of proof). This conclusion does not concern, of course, the application of
the methods in question to legal-theoretic considerations.
Finally, it must be stressed that one cannot establish any hierarchy, or
system of application of the different methods of legal reasoning. The
order in which they are applied is determined by the individual case, the
difficulties it involves, the interpretive context, and – perhaps most
importantly – the methodological habits of the interpreter. Moreover, it
is easy to imagine that the same “interpretive activity” could be taken as
a manifestation of applying two different methods.
12
CHAPTER 1
1.3
LOGIC – ANALYSIS – ARGUMENTATION –
HERMENEUTICS
Below we will consider four methods used by legal practitioners and theoreticians: logic, analysis, argumentation and hermeneutics. Here, two
questions should be answered. First, do those four methods exhaust the
entire spectrum of methods used in legal thinking? And second: what are
the relationships between logic, analysis, argumentation and hermeneutics? Are they independent of each other, or do they overlap?
It is difficult to answer the first question. On the one hand, one can
name several “methods” that are not instances of one of the four men.
tioned. A good example is the psychological theory of Petrazycki. On
the other hand, however, the four listed methods (or better: groups of
methods) not only have a historically established position but also are
applicable to all the spheres we described above, i.e., in creating and
interpreting law, in legal-dogmatic analyses and in legal theory. Therefore
there is some justification backing our choice.
As for the second question: the “borders” between logic, analysis,
argumentation and hermeneutics are not sharp. In popular textbooks it
is usually held that there are two types of analysis: descriptive and reconstructive. Descriptive analysis aims to describe how ordinary language
functions. Reconstructive analysis, on the other hand, tries to reform
ordinary language with the use of logical tools. In contemporary philosophy (and legal theory) a strict differentiation between logical and
descriptive analysis is impossible. Elements of both types of analysis are
mixed together, as in the case of the “third way” – between the Scylla of
description and Harybdis of reconstruction – developed by J. Hintikka.17
He proposes to build formal “explicatory models” that would aim, not at
reforming ordinary language, but at precise explanation of some of its
fragments.
Logic is therefore a tool of analysis. One could ask why we have
decided to treat logic separately from analysis. There are several reasons.
First, unlike “analysis in general”, logic is rather a uniform method and
can hence be defined relatively easily. Second, logical methods can be
presented nicely from a historical perspective, which enables their consequent development to be tracked.
The relationship between logic and argumentation is more complicated, mainly because there are different theories of argumentation.
Those theories reconstruct the ways in which we use arguments.
Argumentation theories concentrate, then, on relationships between
arguments, on comparing them, and on bigger structures consisting of
CONTROVERSY OVER LEGAL METHOD
13
many arguments. They say much less about how concrete arguments are
built. As regards this issue, two stances are possible.
According to the first, represented, e.g., by R. Alexy,18 arguments
should be built in compliance with the rules of logic. From this perspective, logic and argumentation are complementary. The second stance –
exemplified in Ch. Perelman’s new rhetoric – says that the arguments used
in complex argumentation structures do not have to be logically correct.
This does not mean, however, that Perelman regards logic as useless.
Logical schemata may serve as a special kind of topoi. Moreover, many of
the classical legal topoi that play a crucial role in rhetorical argumentation, as for instance argumentum a fortiori or a contrario, can be regarded
as logically valid arguments. Nevertheless, logically invalid arguments can
also be rhetorically effective. Therefore, in Perelman’s conception, logic
does not occupy any special position and is only a possible source of
topoi. Argumentation and logic are not, on this account, complementary.
Informal analytic methods can also be reconciled with argumentation
theories. For instance, economic analysis can serve both to build arguments and to provide us with criteria for evaluating complex argumentation structures. It must be admitted, however, that the relationship
between analysis and argumentation is not inevitable. As already observed, the main aim of argumentation theories is to explain how different
arguments are to be compared with others and measured; the problem of
constructing arguments, fundamental from the point of view of logic
and analysis, is not that important for argumentation theories. It should
be added, however, that in contemporary legal theory some attention is
paid to the structural features of argumentation resulting in the development of logics, which take into account aspects of the process of
argumentation on the one hand, and create informal “logics of argumentation” on the other.
It seems, at least at first sight, that analysis has nothing to do with
hermeneutics. It turns out, however, that even those two methods are
linked in various ways. One example can be found in analytical hermeneutics. This is connected with the “later” philosophy of L. Wittgenstein.
Many analyses of “language games”, presented by Wittgenstein, resemble
the methods and results of hermeneutic philosophy. Among the “analytic
hermeneutic philosophers” one usually mentions: G.H. von Wright,
P. Winch and W.H. Dray. Also some legal-theoretic works have an analytical-hermeneutic character, like for instance Das Verstehen von
Rechtstexten by Hruschka, or some of Aarnio’s works.19
Analytical hermeneutics cannot be easily classified as one of the two
types of hermeneutics: methodological or phenomenological. It is much
14
CHAPTER 1
closer, of course, to the methodological type, which puts text and the
problem of its interpretation in the central place. However, phenomenological hermeneutics, which proposes an alternative-to-traditional
ontology, does not have to contradict analysis either. One of the main
representatives of hermeneutics, A. Kaufmann, paraphrased a well known
phrase of Kant’s, claiming that: “Analysis without hermeneutics is empty,
while hermeneutics without analysis is blind”.20
Some interaction, although not as clearly visible as those mentioned
above, can be traced “in-between” hermeneutics and logic and hermeneutics and argumentation. The existence of such interaction should not be
surprising, for all the four methods are accounts of the same phenomenon: human reasoning. On the other hand, the existence of some similarities and “common grounds” between logic, analysis, argumentation
and hermeneutics does not mean that we can speak of one theory.
Although we are considering four attempts to account for the same phenomenon, those attempts are drawn from diametrically different perspectives.
Finally, we must stress that the four essays presented below are independent of each other and are self-contained. Because of that, there are
some repetitions in the course of the book. We have allowed them for the
sake of the coherence of the presentation.
NOTES
1. See J.H. von Kirchmann, Die Wertlosigkeit der Jurisprudenz als Wissenschaft, Berlin
1847, p. 14 ff.
2. See J.C. Hutcheson, Judgment Intuitive, Chicago 1938, p. 21.
3. See J. Stelmach and R. Sarkowicz, Filozofia prawa XIX i XX wieku [Philosophy of
Law in 19th and 20th Centuries], 1st ed., Kraków 1998, p. 89 ff.
.
4. See L. Petrazycki, Teoria państwa i prawa [Theory of State and Law], v.I, Warszawa
1959–1960, pp. 72–73, 123.
5. See O.W. Holmes, “Path of the Law”, in idem, Jurisprudence, New York/London
1994.
6. See K. Llewellyn, Brumble Bush, New York 1969, p. 12 ff.
7. See R. Pound, Outlines of Lectures on Jurisprudence, Cambridge 1943, p. 104 ff.
8. See H. Lloyd, Introduction to Jurisprudence, New York/Washington 1972, p. 366.
9. See A. Ross, On Law and Justice, London 1958, p. 45 ff.
10. See G. Teubner, Recht als autopoietisches System, Frankfurt am Main 1989, p. 49 ff.
11. See R. Posner, Problems of Jurisprudence, Cambridge/London 1990, p. 360 ff.
12. Cf. J. Stelmach, Kodeks argumentacyjny dla prawników [Argumentation Code for
Lawyers], 1st ed., Kraków 2003, p. 31.
13. Cf. J. Stelmach, Die hermeneutische Auffassung der Rechtsphilosophie, Ebelsbach
1991, p. 19 ff.
CONTROVERSY OVER LEGAL METHOD
15
14. Cf. W. Litewski, Podstawowe wartości prawa rzymskiego [The Basic Values of Roman
Law], Kraków 2001, pp. 22, 51–52.
15. Cf. W. Litewski, Jurysprudencja rzymska [Roman Jurisprudence], Kraków 2000,
pp. 23, 119–133.
16. Cf. R. Alexy, A Theory of Legal Argumentation, translated by R. Adler and N. MacCormick, Clarendon, Oxford 1989.
17. Cf. J. Hintikka, “Epistemic Logic and the Methods of Philosophical Analysis”,
Australasian Journal of Philosophy 46, 1968, pp. 37–51.
18. Cf. R. Alexy, Theory . . ., op. cit.
19. Cf. J. Stelmach, Wsṕoclzesba filozofia interpretacji prawniczej [Contemporary
Philosophy of Legal Interpretation], Kraków 1995, pp. 71–72.
20. Quoted after K. Opalek, “G lówne kierunki niemieckiej teorii i filozofii prawa po II
wojnie światowej” [Main Currents in the German Theory and Philosophy of Law
After the Second World War], in idem, Studia z teorii i filozofii prawa [Studies in
Theory and Philosophy of Law], Kraków 1997, p. 41.
CHAPTER 2
LOGIC
2.1
INTRODUCTION
Logical studies have a very long and rich tradition that dates back to
antiquity. Despite this it is not easy to define logic. A consensus exists,
however, on the fact that logic is about reasoning: it helps us to evaluate
the validity of arguments. The question “which arguments are valid?” is
usually, however, answered in the following way: “the ones in which the
conclusion follows logically from the premises”. In this way we come
back to the question of the nature of logic, or – more precisely – of logical consequence.
A famous analysis of the notion of logical consequence was presented
by A. Tarski.1 Disregarding the details, one may summarize Tarski’s
findings in the following sentence:
A sentence A follows logically from the set of premises Γ if and only if in every case in
which the premises of Γ are true, A is also true.
The idea behind this analysis is that logic is a theory that describes the
“transmission of truth”. The “transmission” begins with the premises of
an argument, and ends with the conclusion. The aim of logic, therefore,
is to identify forms of argument that guarantee the transmission of
truth: if the premises of those arguments are true, their conclusions will
also be true.
In the previous sentence we used another notion that needs explanation. We said that the aim of logic is to indicate valid “forms of argument”. But what are these “forms of argument”? Let us look at the
following two arguments:
(1) If John is intelligent and hard working, he will succeed as
a lawyer.
(2) John is intelligent and hard working.
Therefore: (3) John will succeed as a lawyer.
(1) If the weather is good, John will go swimming.
(2) The weather is good.
Therefore: (3) John will go swimming.
17
18
CHAPTER 2
Let us substitute the sentence “John is intelligent and hard working”
with p, and “John will succeed as a lawyer” with q. Similarly, let p denote
“The weather is good” and q – “John will go swimming”. The first of the
arguments can now be presented as follows:
(1) If p, then q.
(2) p
(3) q
Naturally, using this substitution the second argument looks the same.
We shall say that both arguments have something in common: they have
the same form. It is relatively easy to speak of logical form in particular
cases; it is however much more difficult to define it in abstracto. One may
say that the logical form of an argument is determined by some key
terms in natural language. In our example these are “If . . ., then . . .”
and “Therefore”. The set of key terms also includes “and”, “or”,
“either . . ., or . . .”. It is clear, therefore, that the key terms shall have
special counterparts in the logical language. The counterparts are known
as logical connectives.2
The role of logic is to designate certain forms of argument as valid.
How can one know, however, that the designated forms are indeed valid?
The first of the possible answers is that acceptance of a given logic is
ultimately based on intuition: if the forms of argument indicated as
valid by a given logic are in accordance with our intuitive understanding of what is valid, the logic in question is adequate. The problem with
this solution is that, if intuition determines whether a given argument is
valid or not, we do not need logic at all. One may try to overcome this
difficulty by saying that the intuition in question is not just any “subjective intuition”, but an intuition that is shared by many reasonable
people. One may further observe that we are able to judge intuitively
only relatively simple arguments; more complicated cases have to be
analyzed using logical tools and we can trust logical theory in the
complex matters when it does not lead to counter-intuitive solutions in
the simple.
An intuitive judgment is fortunately not the only way to demonstrate
the adequacy of a logical theory. Tarski’s analysis of logical consequence,
mentioned above, is of help here. It forms a basis for the so-called soundness and completeness theorems.3 Almost every logical system consists of
two parts: syntax and semantics. Syntax is a language-resembling structure, consisting of an alphabet, rules for constructing well formed formulas, rules of inference and axioms. Semantics, on the other hand, may
be regarded as a mathematical model of the world. If every sentence that
can be proved on the basis of the axioms and inference rules of a given
LOGIC
19
logic is true in all possible “world models” (and vice versa), the logic is
said to be sound (complete). Speaking metaphorically, and somewhat
loosely, the soundness and completeness theorems show that the language of a given logic “fits together” with the world, and therefore we
have grounds for believing that the logic adequately indicates valid forms
of argument.
This is not the end of the trouble, however. There exist various logics
with different syntaxes and semantics, identifying different forms of
argument as valid; moreover, the soundness and completeness theorems
hold for each of them. Thus the problem arises of choosing between
different logics that are sound and complete. The choice must be made
according to certain criteria. One can, for example, try to evaluate which
logic uses the semantics that best reflects the world. Intuition can also
help in picking out the correct logic.
At the end of this short introduction, two additional problems should
be addressed. Firstly, we have said that logical validity guarantees the
“transmission of truth” from the premises to the conclusion of an argument. This does not mean, however, that each argument, which is logically correct, has a true conclusion. For this to be true it is necessary for
the premises of the argument to be true. If the premises are not true,
there can be no “transmission of truth”. In this context internal and
external justifications are differentiated. We shall say that an argument is
internally justified if its conclusion follows logically from its premises.
An argument is externally justified, on the other hand, when it is: (a)
internally justified, and (b) its premises are true.
Secondly, one must mention the role that is usually ascribed to logic.
In the philosophy of science one distinguishes between the context of
discovery and the context of justification; the distinction is made in at
least two different ways. First, a scientific discovery could be divided into
two stages: the first ends with the formulation of a hypothesis of the analyzed phenomena (context of discovery), and the second serves to justify
(test, falsify) the hypothesis (context of justification). Second, in every
scientific discovery one may differentiate between two aspects: sociopsychological (context of discovery) and logical (context of justification).
The socio-psychological aspect consists of all factors that influence the
discovery – everything described by psychologists and sociologists may
be of interest here. The second aspect – the logical – enables one to look
at the scientific discovery as a purely rational undertaking, which meets
certain criteria for accepting and refuting scientific theories. Irrespective
of which version of the distinction is chosen, logic has a certain function
only in the context of justification. Additionally, it is easy to relate the
20
CHAPTER 2
distinction between both contexts to arguments other than scientific
ones. In this way the theory of the logical character of the context of justification is valid universally, i.e., it concerns all kinds of argument.4
Below we present a kind of “history” of the search for the logic of normative discourse (including legal discourse). We shall not, however, pay
much attention to chronology. We will rather try to show the basic ideas
behind different logics of normative discourse; we will also concentrate
on the criteria for comparing different logics, and the reasons for constructing new normative logics. We will discuss classical logic (propositional logic and first order predicate logic), deontic logic, the logic of
agency and, finally, defeasible logic.
2.2
CLASSICAL LOGIC: PROPOSITIONAL LOGIC
AND FIRST ORDER PREDICATE LOGIC
2.2.1
Presentation of Calculi
The history of contemporary logic began just over a hundred years ago
with the publication of the works of G. Boole, C.S. Peirce, and – first
and foremost – B. Russell and A.N. Whitehead. Begriffschrift (1902) by
Frege, and Principia Mathematica (1910–13) by Russell and Whitehead
constituted the turning point in the history of logic; both those works set
the stage for the incredible development of logic in the twentieth century.
The two basic logics elaborated by Frege, Russell and Whitehead are
classical propositional logic and first order predicate logic. The propositional calculus takes into account only those forms of argument, in
which elementary sentences are basic elements. The elementary sentences
can be, with just a few exceptions, identified with (grammatically) simple
(not compound) sentences. The compound sentences have also a complex logical structure. Because of that, the sentence connectives must
have logical counterparts. Those counterparts are called truth-functional
functors (or sentential connectives).
The alphabet of propositional logic consists of propositional variables
that are usually denoted by small letters p, q, r, etc. A propositional variable denotes an arbitrary elementary sentence. In the alphabet of propositional calculus one can also find symbols denoting the truth-functional
functors (sentential connectives): negation (¬), implication (→), conjunction (∧) and disjunction (∨).5 In order to provide a full syntactic characterization of propositional calculus it is necessary to recall the rules of
forming formulas, the rules of inference and axioms. According to the
rules of forming formulas, all propositional variables are well formed
LOGIC
21
formulas of propositional calculus. Additionally, if A and B are (arbitrary)
well formed formulas of propositional logic, ¬A, A ∧ B, A → B, and
A ∨ B are also well formed formulas of propositional calculus.
We will not present here the axioms of propositional logic, for the metalogical features of this logical system will not be analyzed. We will limit
ourselves to mentioning only one rule of inference that plays an important role in the considerations below. The rule is modus ponens, according to which an implication (A → B) and its antecedent (A) logically
imply its consequent (B). Schematically, this rule may be depicted in the
following way:
A→B
A
B
Besides the syntactical, a semantic characterization of propositional calculus is also needed. Interpretation in this calculus amounts to ascribing
the Boolean values of truth (1) or falsehood (0) to propositional variables. The sentential connectives are defined as follows:
¬A is true, if and only if A is false, otherwise it is false.
A ∧ B is true if and only if A is true and B is true, otherwise it is false.
A → B is false if and only if A is true and B is false, otherwise it is true.
A ∨ B is false if and only if A is false and B is false, otherwise it is true.
The above semantic characterization of sentential connectives is usually
presented in the form of truth-tables. The following is an example of a
table for implication:
→ 1
1 1
0 1
0
0
1
Elementary sentences are the most basic elements taken into account
when analyzing the validity of arguments using propositional logic. The
other type of logic developed by Frege and Russell – first order predicate
logic – enables one to take into account the inner structure of elementary
sentences. In the structure predicates (corresponding to the verb-part of
a sentence), the predicate’s arguments (corresponding to the subject) and
quantifiers (statements that the predicate refers to some or all objects in
the language’s domain) are distinguished.
The alphabet of first order predicate logic looks as follows. The predicates
are usually denoted by capital letters: P, Q, R, S. In order to increase the
22
CHAPTER 2
readability of the formalizations, however, we will denote predicates with
words, such as read, father_of, charged_with. Variables referring
to the predicate’s arguments will be denoted by small letters x, y, z.
Arguments will also sometimes be referred to with names written in italics,
e.g., john, car_of_peter, house_of_hanna, etc. The difference between names
and variables is clear: a given name (e.g., john) denotes a particular object
(e.g., a particular person), whilst a variable x is not ascribed to a particular
object, and only refers to an object determined in a general, abstract way.
In the alphabet of first order predicate logic there are also logical constants: truth-functional functors (the same as in propositional logic, i.e.,
¬, ∧, → and ∨) and two quantifiers: general ᭙, and existential ᭚.
In the above presented symbolism the sentence:
John is convicted of murder
could be written as follows:
convicted_of_murder(john)
whilst the sentence:
Some people are convicted of murder
is:
᭚ x(convicted_of_murder(x))
The rules for forming compound expressions of first order predicate
logic are as follows. First, atomic formulas, i.e., an n-ary predicate with
n individual variables or names (e.g., convicted_of_murder(john),
convicted_of_murder(x), father_of(john, bill), etc.) are well
formed formulas of first order predicate logic. Second, if A and B are
well formed formulas of first order predicate logic, ¬A, A ∧ B, A → B,
A ∨ B as well as ᭙ x A and ᭚ x A are also well formed formulas.
As in the case of propositional logic, we shall not present the axioms
of first order predicate logic. Furthermore, there exists a version of
modus ponens for first order predicate logic.
The semantic definitions (truth-tables) of functors in first order predicate logic are not different from definitions of the same functors in
propositional logic. Two additional logical constants – ᭙ and ᭚ – are
defined in the following way: the expression ᭙ x(predicate(x)) is true
if and only if all the objects belonging to the domain of the discourse
can truly be said to be predicate. The expression ᭚ x(predicate(x))
in turn, is true if and only if in the discursive domain there exists at least
one object that can truly be said to be predicate.
Interpretation in first order predicate logic is as follows. First, a set constituting the discourse’s domain is chosen (intuitively this set is comprised
LOGIC
23
of the objects existing in the world). To the 1-ary predicates there are
ascribed subsets of the discourse domain (containing only those objects of
which the given predicate can truly be predicated); the 2-ary predicates are
ascribed sets of ordered pairs of objects from the domain; the 3-ary predicates – sets of ordered triples, etc. In addition, every individual constant
(name) is ascribed a determined object from the domain of the discourse.
It is worth adding that both propositional logic and first order predicate logic are sound and complete.
2.2.2
Paradoxes of Material Implication
Before we attempt to show how legal reasoning is reconstructed with the
use of propositional logic and first order predicate logic, it is necessary
to mention the controversies surrounding material implication. As is well
known, different functors (sentential connectives) are designed to “correspond” to different connectives of natural language: disjunction to
“or”, conjunction to “and”, etc. It is usually held that the semantic characterization of ∧ and ∨ is in accordance with the use of “and” and “or”
in natural language. For example, the sentences in which “and” occurs
are held to be true only when both sentences connected by use of “and”
are true.
A number of controversies are connected, however, to implication.
This functor is said to correspond to conditionals of the form “if . . .,
then . . .”. The semantic characterization of implication, however, seems
not to meet – in certain circumstances at least – the criteria for using
conditionals. This is an extremely important problem as it is clear that
conditionals are the natural way of expressing legal norms.
Let us look once more at the truth-table for implication:
→ 1
1 1
0 1
0
0
1
The last row is troublesome. According to it, an implication is true in
each case in which its antecedent is false. This leads to the following sentences being established as true:
(1) If New York is the capital of the USA, then water boils at 30˚C.
(2) If 2 + 2 = 5, then Washington, D.C. is the capital of the USA.
The ascription of truth to the two above sentences may seem counterintuitive.6 However, as W.V.O. Quine observes rightly,7 the following
sentence may seem similarly counter-intuitive:
(3) If 2 + 2 = 4, then water boils at 100˚C.
24
CHAPTER 2
With (3) we still feel that “something is wrong”, although both the
antecedent and consequent of the implication are true. From this we may
conclude that sentences (1)–(3) are counter-intuitive because of what
they say (there is no connection between the meaning of the antecedent
and the consequent), and not because of the rules put forward in the
truth-table for implication.
Quine’s solution to the problems of material implication is not, however, fully satisfactory. The semantic characterization of implication
leads to acceptance of the following expressions as tautologies of the
classical propositional calculus (these are the paradoxes of the material
implication)8:
((p → q) ∧ (r → s)) → ((p → r) ∨ (q → s))
¬ (p → q) → p
According to the former, the following argument is logically valid:
If John is in Paris, then he is in France; and if John is in London, then he is in England.
Therefore, if John is in Paris, he is in London or if he is in France, he is in England.
The second of the tautologies leads us to accept the following chain of
reasoning:
It is not true, that if there exists God, the prayers of evil people will be heard. Therefore
there exists God.
It is difficult to apply Quine’s solution to the two presented arguments.
Should we say, then, that material implication does not correspond to
conditionals?
Both positive and negative answers to this question have been advocated.9 The positive answer leads usually to the development of new
functors that fit better the criteria for using conditionals in natural language; a case in point is the development of relevant logics.10 We will not,
however, discuss these formalisms in any detail. It is sufficient to note
that significant doubts exist over the relevance of material implication
for representing natural language conditionals. This is of great importance for us, because – as noted above – it is usually held that every legal
norm can be expressed in an “If . . ., then . . .” sentence.
2.2.3
Examples
We will start our reconstruction of legal reasoning by use of classical
logic with a simple example. According to Article 278§1 of the Polish
penal code (kk), “whoever takes somebody else’s property shall be imprisoned for term between 3 months and 5 years”. Let us imagine that the
LOGIC
25
accused, John, stole Adam’s bicycle. What is the argument that leads to
John’s conviction? It may look something like this:
(1) Whoever takes somebody else’s property shall be imprisoned for a term between 3 months and 5 years.
(2) John has taken somebody else’s property (for he has
stolen Adam’s bicycle).
Therefore: (3) John shall be imprisoned for a term between 3 months
and 5 years.
This argument may be reconstructed in classical propositional logic,
assuming that (1) is an implication:
(1) p → q
(2) p
(3) q
A valid form of argument has been applied here – modus ponens. What
are the variables substituted for? If (2) is substituted with p, then p must
stand for “John has taken somebody else’s property”; q is, of course,
“John shall be imprisoned for a term between 3 months and 5 years”. If
so, “p → q” should be read: “If John has taken somebody else’s property,
then John shall be imprisoned for a term between 3 months and 5 years”.
This sentence is not, however, equivalent to Article 278§1 kk! That is
clearly visible when one analyzes some other case, let us say a situation
in which Adam has stolen Ted’s car. Adam bears responsibility according to the same norm as John, namely Article 278§1 kk. The judge
applies then the same form of argument as in the previous case:
(1) Whoever takes somebody else’s property shall be imprisoned for a term between 3 months and 5 years.
(2) Adam has taken somebody else’s property (for he has
stolen Ted’s car).
Therefore: (3) Adam shall be imprisoned for a term between 3 months
and 5 years.
This time (2) cannot be substituted with p (for p stands already for “John
has taken somebody else’s property”); similarly (3) cannot be written
as q. Let us therefore choose other variables, r and s. Now (1) in our formalization is r → s. Our reconstructions show that, in classical propositional logic, it is difficult to formalize the first premise, which is a formal
counterpart of Article 278§1 kk, of either argument. In the case of John’s
crime we obtain the sentence: “If John has taken somebody else’s property, then John shall be imprisoned for a term between 3 months and
5 years”, whilst in Adam’s: “If Adam has taken somebody else’s property, then Adam shall be imprisoned for a term between 3 months and
5 years”.
26
CHAPTER 2
One can apply here the following trick to maintain that both arguments have the same form:
(1) If the accused takes somebody else’s property, he shall be
imprisoned for a term between 3 months and 5 years.
(2) The accused has taken somebody else’s property.
Therefore: (3) The accused shall be imprisoned for a term between
3 months and 5 years.
Substituting p with “The accused has taken somebody else’s thing”, and
q for “The accused shall be imprisoned for a term between 3 months and
5 years”, we obtain the same logical schema for the cases of both Adam
and John. This solution is, however, not acceptable. First, it is based on
the fact that the term “accused” refers to different persons in different
contexts. Second, whilst premise (1) of both arguments is – as required –
identical, premise (2) is likewise– and counter-intuitively – identical.
The indicated problems occur because of the fact that, in propositional logic, one cannot reconstruct the inner structure of the sentences
composing the analyzed arguments. Much more can be done in first
order predicate logic. Applying this calculus enables the following reconstruction of our examples:
(1) ᭙ x(takes(x) → imprisoned(x))
(2) takes(john)
(3) imprisoned(john)11
takes stands here for “takes somebody else’s property”, imprisoned
– “shall be imprisoned for a term between 3 months and 5 years”, and
john is a name for John. If we use adam as a name for Adam the judge’s
reasoning in Adam’s case may be presented as follows:
(1) ᭙ x(takes(x) → imprisoned(x))
(2) takes(adam)
(3) imprisoned(adam)
The presented formalization has the required features. In both arguments premise (1) is identical, but premise (2) is different; in other words,
the structure of the analyzed examples, as reconstructed with the use of
first order predicate logic, seems to resonate with our intuitions.
Arguments of the type presented above are traditionally called legal
syllogisms. The concept of a legal syllogism played a crucial role in legal
positivism. The continental positivists held that legal reasoning has (or
should have) the form of legal syllogism. Every such syllogism consists
of two premises and a conclusion. The first premise is a general and
abstract legal norm, as e.g., Article 278§1 kk from our example:
“Whoever takes somebody else’s property shall be imprisoned for a term
between 3 months and 5 years”. The second premise describes a state of
LOGIC
27
affairs, e.g., “John has taken somebody else’s property”. Finally, the conclusion is an individual and concrete legal norm; in the case of our example: “John shall be imprisoned for a term between 3 months and 5 years”.
Therefore, the logical reconstruction presented below:
(1) ᭙ x(takes(x) → imprisoned(x))
(2) takes(john)
(3) imprisoned(john)
is an instance of a legal syllogism.
Various arguments have been put forward against legal syllogism as a
correct reconstruction of legal reasoning. It has been maintained, for
example, that the syllogism is impossible, for the logic of norms is impossible. It has also been held that the essence of legal reasoning is the process
of valuation and not logical consequence, as positivists and other adherents of formal logic seem to suggest. These problems will be dealt with
below (see Section 2.4). Here, another issue must be addressed, namely the
thesis that legal syllogism is trivial because it can be applied only after all
the significant problems of legal reasoning have already been solved.
The core of this objection may easily be displayed using our earlier
example. In describing the state of affairs we said that John has stolen
Adam’s bicycle. Meanwhile, premise (2) says that John has taken somebody else’s property. These are certainly two different sentences. In one
of our informal reconstructions we dealt with this problem saying: “John
has taken somebody else’s property (for he has stolen Adam’s bicycle)”.
This formulation indicates that we are concerned here with an additional
stage in reasoning (please note our use of the word “for”), which has not
been accounted for in our logical reconstruction of legal syllogism. It is
relatively easy, however, to fix this problem, e.g., in the following way:
(1) ᭙ x(steals_bicycle(x) → takes(x))
(2) steals_bicycle(john)
(3) takes(john)
This argument follows from two premises, i.e., “Whoever steals Adam’s
bicycle takes somebody else’s property” and “John has stolen
Adam’s bicycle” to the conclusion that John has taken somebody else’s
property.
In legal theoretic literature, the conclusion of the kind of argument
that we have just presented is called the interpretational decision.12 Here,
Article 278§1 kk is interpreted; we say that the expression it contains (“to
take somebody else’s property”) refers, inter alia, to the act of stealing
Adam’s bicycle. The argument leading to the interpretational decision
can of course be adduced to our main syllogism:
(1) ᭙ x(takes(x) → imprisoned(x)) [general and abstract legal norm]
28
CHAPTER 2
(2) ᭙ x(steals_bicycle(x) → takes(x)) [premise of the interpretational decision]
(3) steals_bicycle(john) [a description of the state of affairs]
(4) takes(john) [from (2) and (3), modus ponens]
(5) imprisoned(john) [from (1) and (4), modus ponens]
When we look more closely at the above reconstruction, some other troublesome elements can be found. A case in point is, for example, the conclusion of our argument: “John shall be imprisoned for a period between
3 months and 5 years”. Judges never formulate their sentences in such a
manner, but stipulate exactly the period of imprisonment, saying, for
example, that John shall be imprisoned for 2 years. The argument of the
judge which leads to the determination of the duration of imprisonment
may also be reconstructed logically. Let us denote all the circumstances
that make a 2 year sentence a just one by circumstances;
two_years means “shall be imprisoned for 2 years”.The judge’s argument may be reconstructed as follows:
(1) ᭙ x((imprisoned(x) ∧ circumstances(x)) → two_years(x))
(2) imprisoned(john)
(3) circumstances(john)
(4) two_years(john)
The key premise (1) says that when somebody should be imprisoned for a
period between 3 months and 5 years (imprisoned(x)), and there are circumstances justifying a two year sentence (circumstances(x)), the person shall be imprisoned for two years (two_years (x)). Premise (2) is the
conclusion of our syllogism in its earlier version. Finally, premise (3) says
that the circumstances justifying the two year sentence do obtain in John’s
case. The conclusion of the analyzed argument, which Jerzy Wróblewski
calls consequences choice decision, says: John shall be imprisoned for two
years. We may now present our syllogism in a more complete form:
(1) ᭙ x(takes(x) → imprisoned(x)) [general and abstract legal norm]
(2) ᭙ x(steals_bicycle(x) → takes(x)) [premise of the interpretational decision]
(3) steals_bicycle(john) [description of the state of affairs]
(4) takes(john) [from (2) and (3), modus ponens]
(5) imprisoned(john) [from (1) and (4), modus ponens]
(6) ᭙ x((imprisoned(x) ∧ circumstances(x)) → two_years(x))
[premise of the consequences choice decision]
(7) circumstances(john) [further description of the state of affairs]
(8) two_years(john) [from (6), (5) and (7), modus ponens]
In this way we obtain a complex and logically valid schema. The
argumentative structure presented can be further elaborated upon. For
LOGIC
29
example, one can take into account the validity decision, i.e., the decision
leading to the establishment of the validity of Article 278§1 kk (that has
been reconstructed as (1)). Similarly, the arguments leading to the establishment of (3) and (7) (the evidential decisions) can be reconstructed
logically.
Legal syllogism, as it is usually presented, could be called trivial. It is
not, however, a complete logical reconstruction of legal reasoning. Every
decision made during a judge’s reasoning can be analyzed with the use of
logical tools, resulting in complicated argumentative structures.
2.3
2.3.1
DEONTIC LOGIC
Possible World Semantics
From our perspective, i.e., in the context of developing an adequate
“legal logic”, one cannot overlook modal logics. These are logics that,
besides traditional logical connectives, like negation or implication, offer
also modal functors – aletic (“it is possible that”, “it is necessary that”),
epistemic (“it is known that”, “it is believed that”) or deontic (“it is forbidden that”, “it is obligatory that”, “it is permitted that”). Aletic modal
logics were first described by the American logician C.I. Lewis. Their
development is, however, connected with the creation of possible world
semantics in the 1950s.13
The notion of a possible world is intuitively clear. In 2002 Brazil won
the World Cup. But one can easily imagine a situation in which they failed
to reach the final. They did not, and for that reason the world in which
Brazil lost is not an actual world; but it is a possible one. If you are sitting
in a chair right now, then in one minute you may still be sitting, but you
may also be standing. These are two possible worlds: in one of them you
are sitting and in the other – standing. But if you are in Kraków now, in
one minute you may be sitting or standing in Kraków, but not in New
York. There exists, of course, such a possible world, in which you may be
standing in New York in one minute’s time; but this world is not possible
relative to the actual world, in which you are in Kraków. Therefore, one
can differentiate between an “absolutely” possible world – i.e., every consistent state of affairs (the fact that you are in New York in a minute is not
logically inconsistent) – and a “relatively” possible world, i.e., a world that
is possible relative to the actual world (we will say that the possible world
is accessible from the actual world). Brazil winning the 1998 World
Cup is possible in the absolute sense. Today it is not, however, possible
“relatively”, for the 1998 World Cup was won by another team. At the
beginning of 1998, however, it was possible not only “absolutely”, but
also “relatively”.
30
CHAPTER 2
It is due to Saul Kripke14 that the above presented intuitions were
encapsulated in a very elegant mathematical form. Kripke showed how
one can build an adequate mathematical object that takes into account
the mentioned differentiations, and serves as a semantic model for
certain logical calculi. The object is an ordered quadruple:
〈 wa, W, R, ν 〉
wa stands here for the actual (our) world. W is a set of the possible
worlds. R is the accessibility relation. If the relation R holds between two
worlds (say wi and wj) – we will refer to this fact by (wiRwj) – it means
that the world wj is accessible from the world wi. Finally, v is the interpretation function that ascribes truth or falsehood to every sentence in
the given possible world (vw(p) = 1 or vw(p) = 0).15 Therefore, if we
substitute p for “Brazil won the World Cup in 2002”, then the interpretation function ascribes to p in the actual world wa, the value of truth
(vw (p) = 1), whilst in some other possible world, wi, in which Brazil failed
to win, vw (p) = 0.
The semantic structure thus constructed enables necessity and possibility to be defined. We will say that p is necessary in the actual world
■ p), if p is true in all the worlds accessible from wa; p is possible in
wa(■
wa(◊p), if p is true at least in one world accessible from wa. It is clear that
■ p is true if and only if ~◊~p is also true, i.e., if there does not exist such
a possible world accessible form wa in which ~p is true. The last remark
■ p ⇔ ~◊~p).
shows that necessity and possibility are mutually definable (■
a
i
2.3.2
Deontic Logic
S. Kripke developed possible world semantics with the aim of analyzing
the concepts of necessity and possibility. Rather quickly, however, it
turned out that Kripke’s mathematical tool could serve perfectly well the
analysis of other notions, such as “to know” and “to believe” or – important for us – “obligatory”, “forbidden” and “permitted”.
This does not mean, however, that the logic of obligations, i.e., deontic
logic, originated with the development of possible world semantics in the
mid-twentieth century. In “Nicomachean Ethics”, Aristotle had already
analyzed arguments like the following: “if ‘everything sweet should be
tasted’, and ‘a given thing is sweet’, i.e., it is one of the sweet things, then
a man who is capable should taste this given thing”.16 This is an example
of applying the practical syllogism that formalizes normative reasoning.
Another, much later work that includes some considerations of deontic
logic is Leibniz’s “Elementa iuris naturalis”.
LOGIC
31
In twentieth century philosophy analyses of normative sentences can
be found in the work of B. Bolzano, A. Höfler and E. Husserl. More
advanced conceptions of practical discourse were developed by E. Lapie
and E. Mally (logic of will), and by E. Menger (logic of habits). They
were created in 1902, 1926 and 1934 respectively. Only small parts of
those works were devoted to deontic logic. A different story can be told
about the works of W. Dubislav, J. Jørgensen, A. Hofstadter, J.C.C.
McKinsey, R.M. Hare and R. Rand, who each tried to formalize arguments that are normative par excellence.17
It is usually held, however, that the birth of deontic logic took place in
1951, the year of publication of G.H. von Wright’s paper “Deontic
Logic”. At the same time similar problems were investigated by
J. Kalinowski and O. Becker. The former published the results of his
research in 1953, the latter in 1952.
The first systems of deontic logic can be characterized as syntactic.
The situation changed with the above mentioned development of possible world semantics. Semantics enables one to define in a very intuitive
way the notion of obligation. Let O stand, as usually, for “it is obligatory
that . . .”. What does it mean if the sentence Op, “it is obligatory that p”,
is true in the actual world (wa)? One can imagine that enacting a norm of
behavior simply involves a legislator picking out a subset of the worlds
that are possible relative to wa. In the chosen worlds things stand as the
legislator wishes them to. Let us notice (see the figure) that the set DD of
deontically perfect worlds has to be a subset of the set M that contains
worlds possible relative to wa. This condition reflects the basic principle
of law, i.e., impossibilium nulla obligatio est. The legislator cannot make
obligatory (include into DD) what is not possible (i.e., what does not
belong to M).
With DD established by the legislator we can give the conditions of
truth for Op. Op is true in wa if and only if p is true in every world of DD.
As an illustration consider the sentence “It is obligatory that John does
not steal”. This sentence is true in wa, if the sentence “John does not
steal” is true in every world of DD. Were the latter sentence true only in
one of the worlds from DD, e.g., in w1, “It is obligatory that John does
not steal” would not be true; therefore, John would not be obliged to
refrain from stealing. The world w2 is a world that is accessible from
the actual world, but the legislator does not regard it as deontically
perfect. Finally, w3 is not accessible from wa, which amounts to saying
that it cannot become an actual world. Therefore, w3 cannot be included
in DD.
32
CHAPTER 2
w1
w3
w2
DD
wa
M
In deontic logic, in addition to the functor of obligation O, there are
also two other functors: the functor of prohibition F (forbidden) and the
functor of permission P (permitted). Those functors can be defined in a
natural way with the use of functor O. If p is forbidden, i.e., Fp, then it
is obligatory that ¬p:
Fp ⬅ O¬p
Furthermore, if p is permitted, Pp, then it is not true that it is obligatory
that ¬p:
Pp ⬅ ¬ O¬p
Let us address how these definitions work in practice. Suppose that a legislator forbids John to steal. Let p stand for “John steals”. According to
the definition Fp is equivalent to O¬p, i.e., John is under an obligation
not to steal. Therefore, in order to make the sentence “it is forbidden for
John to steal” true in wa, the sentence “John does not steal” has to be true
in every single world of DD. Let us now turn our attention to the sentence “It is permitted for John to steal” – Pp. According to the definition
this sentence is equivalent to ¬O ¬ p. And the latter sentence is true only
if the sentence O¬p is false. As we know, the falsity of O¬p requires that,
in at least one possible world of DD, the sentence ¬p is false – which
equates to p being true. In the described situation, the sentence “it is permitted for John to steal” is true, if there exists a world in DD in which
the sentence “John steals” is true. Let us observe that it is possible
that the sentence “John steals” is true in every world belonging to DD.
Then, both Pp (“It is permitted for John to steal”) and Op (“It is obligatory for John to steal”) are true.18
LOGIC
33
The semantic ideas presented above are realized in various systems of
deontic logic, as, for example, in standard deontic logic (SDL).19 It
should be noted that the soundness and completeness theorems have
been proved for SDL.
2.3.3
Paradoxes of Deontic Logic
There are numerous deontic logics that differ to greater or lesser
degrees. There are various reasons for the search for new logics of obligation, one of the most important being the paradoxes of deontic logic.
Amongst those paradoxes one can list the Ross paradox. This has to
do with the fact that, in SDL, the argument from
Op
to
O(p ∨ q)
is valid.
Let us substitute p with “send the letter” and q with “burn it”. Thus, a
paradoxical reasoning is constructed: “if it is obligatory to send the
letter, it is obligatory to send it or burn it”.
There is no consensus as to whether the Ross Paradox can be called a
real problem. Some adhere to the thesis that there is a problem in the reasoning, since the acceptance of the second norm seems counter-intuitive.
Others deny this, saying that the mere fact that one norm (O(p ∨ q)) follows logically from another norm (Op) does not mean that the latter
ceases to be binding. Therefore, if the norm “it is obligatory to send the
letter or to burn it” is fulfilled by burning the letter, the norm “it is obligatory to send the letter” is broken. In other words, the fulfillment of a
norm B that follows logically from a norm A, does not necessarily lead
to the fulfillment of norm A.20
It is not the Ross Paradox, however, nor problems of its kind that constitute the major puzzle of deontic logic. It is usually held that the most difficult
problems facing deontic logicians are contrary-to-duty (CTD) paradoxes.
Those paradoxes arise in connection with CTD-norms, i.e., norms that
require the breaking of another norm as a condition of their application.
Let us look more closely at a famous example known as the Chisholm
Paradox.21 This paradox is connected with the following four sentences:
(1) It is obligatory for a certain man to help his neighbors.
(2) It is obligatory that if he helps them, he tells them about it.
(3) If he does not help them, he should not tell them he helps them.
(4) The man does not help his neighbors.
It is easy to observe that the norm expressed in (3) is a CTD-norm,
for its antecedent (“if he does not help them”) describes the fact of
34
CHAPTER 2
breaking another norm (norm (1)). It is usually held that, intuitively,
sentences (1)–(4) are mutually consistent and logically independent of
each other (none of the sentences follows logically from others). Let us
try to formalize Chisholm’s example. If the sentence “A certain man
helps his neighbors” is substituted with p, and “He tells them about it”
with q, sentences (1)–(4) can be formalized in SDL in the following
manner:
(1) Op
(2) O(p → q)
(3) ¬p → O ¬ q
(4) ¬p
Unfortunately, this formalization is inconsistent. In SDL it follows from
sentences (1) and (2) that Oq, and from (3) and (4) – O ¬ q. The intuitively consistent set of sentences turns out in our formalization to be
inconsistent.
It is easy to observe that the above presented formalization is not the
only way of reconstructing sentences (1)–(4) in SDL. The issue here is
how the conditional duties should be formalized. Let us notice that legal
norms usually take the form of a conditional; traditionally it is maintained that every norm consists of an antecedent and a consequent. It is
held, moreover (as already noticed during the discussion of first order
predicate logic) that this very structure is captured by the material implication. In the language of deontic logic, however, there occurs a problem.
If p is the antecedent and q the consequent, the given norm can be
formalized in two ways – either as:
p → Oq
or
O(p → q)
In the analyzed example norm (3) was formalized in the former manner,
and norm (2) in the latter. One may consider it incorrect that the two conditional norms were treated differently (even though their natural language formulation encouraged this). In light of the above observation,
two solutions are possible: either represent norm (2) as p → Oq, or norm
(3) as O(¬p → ¬q). It turns out, however, that neither of those solutions
is acceptable. In the new formalizations some of the analyzed sentences
are logically dependent on others. In the case of the former, norm (2)
follows logically from sentence (4), and in the case of the latter, norm
(3) follows from (1). It can therefore be concluded that attempts to formalize Chisholm’s example lead to results that are intuitively unacceptable.
The CTD paradoxes, together with some other problems, serve as a
reason for creating deontic logics that differ to a greater or lesser degree
LOGIC
35
from SDL. We will not attempt to describe them here. We would only like
to indicate the mechanism that leads to developing new logical systems:
usually, when a new logical system is built, it is provided with an intuitively sound semantics and is tested on different sets of examples.
Sometimes such problems as CTD paradoxes occur. They highlight the
weaknesses of the developed systems and suggest that a search for other
solutions may be needed. It is vitally important, then, to consider when
and why something may seem paradoxical. Clear intuitions decide this
matter. Chisholm’s example is paradoxical, because we intuitively held the
sentences that constitute it to be consistent and logically independent.
However, it seems impossible to obtain a consistent and logically independent formalization of those sentences in SDL.
2.3.4
Examples
The examples presented in Section 2.2.3 may seem atypical as regards legal
reasoning. The way in which Article 278§1 kk is formulated does not make
it immediately obvious that we are dealing with a legal norm. In that provision such phrases as “ought to”, “it is forbidden” or “it is allowed” do
not occur. This is, however, the common way of formulating legal text:
legal provisions are usually expressed in an indicative mood. This does not
mean, however, that at the logical level they should be reconstructed without the use of deontic operators. If the legislator says that “whoever takes
somebody else’s property shall be imprisoned for a period between
3 months and 5 years”, it seems intuitive to reconstruct that statement in
the following way: “it ought to be the case that whoever takes somebody
else’s property shall be imprisoned for a period between 3 months and 5
years”. The following represents another way of introducing the deontic
operator into Article 278§1 kk: whoever takes somebody else’s property
ought to be imprisoned for a period between 3 months and 5 years. In the
first case Article 278§1 kk obtains the following symbolic form:
O (᭙ x(takes(x) → imprisoned(x)))
and in the second:
᭙ x(takes(x) → O(imprisoned(x)))
If we choose the second option, our basic legal syllogism in John’s case
will look as follows:
(1) ᭙ x(takes(x) → O(imprisoned(x)))
(2) takes(john)
(3) O(imprisoned(john))
In this formalization it is clear that both (1) and (3) are normative in
character.
36
CHAPTER 2
In the Article 278§1 kk example, one more thing may seem counterintuitive. Whilst reading a penal code one would expect to find norms
of behavior: statements of what we should and should not do. Article
278§1 kk, however, says nothing of this kind. It is often maintained that
penal codes contain only sanctioning and no sanctioned norms. And it is
the former that determine the obligations and rights of a citizen, whilst
the addressee of a sanctioning norm is a state authority. Such a norm
obliges the authority to act in a specific way where a sanctioned norm
has been broken. For instance, in the case of Article 278§1 kk, the action
of the authority consists in imprisoning the person who broke the sanctioned norm for a term of 3 months to 5 years.
In principle, in penal codes only sanctioning norms are expressed. It is
sometimes held that sanctioned norms are “outside the code”. This is, of
course, a metaphor. Sanctioned norms are not directly stated in penal
codes, but we can reconstruct them on the basis of directly stated sanctioning norms. If Article 278§1 kk says that “whoever takes somebody
else’s property shall be imprisoned for a term between 3 months and
5 years”, it is a basis for formulating the following sanctioned norm: one
should not in any circumstances take somebody else’s property.
It may be observed that a judge does not need the norm “one should not
in any circumstances take somebody else’s property” in order to give her
judgment. This is true, but one can easily imagine intuitively correct arguments in which sanctioned norms serve as premises or conclusions. For
instance: one should not in any circumstances take somebody else’s property, and because stealing a bicycle constitutes an instance of taking somebody else’s property, therefore one should not in any circumstances steal a
bicycle. In order to put forward such an argument one has to know that
there exists a norm stating that one should not in any circumstances take
somebody else’s property. This norm is not expressed directly in the penal
code. It has to be reconstructed and such reconstruction is not always a
trivial task. We shall not analyze this problem here in any detail. It is worth
observing, however, that there are no a priori reasons excluding the search
for logical schemata of such reconstruction.
One does not find similar problems in civil codes. Let us consider, as
an example, Article 415 kc of the Polish civil code: “whoever intentionally causes damage to someone has to redress it”. This provision is not
only addressed to “normal citizens”, but also includes a deontic operator “has to”. One can formalize Article 415 in the following way:
᭙ x(causes_damage(x) → O(redresses(x)))
Let us consider now whether there exist reasons to formalize legal reasoning with the use of deontic logic as opposed to first order predicate
LOGIC
37
logic. It can be maintained that, from a practical point of view, there is
no difference between the two. Let us look once again at Article 415 kc.
In our deontic-logic formalization it has the form:
᭙ x(causes_damage(x) → O(redresses(x)))
The following formalization that uses the predicate “has to redress the
damage” and sticks to first order predicate logic seems equally good:
᭙ x(causes_damage(x) → has_to_redress(x))
If a judge finds that John has intentionally caused damage to someone
(causes_damage(john)), she would conclude – in the first formalizationthat O(redresses(john)), and in the second: has_to_redress(john).
Both conclusions are the same: John is obliged to redress the damage.
Therefore, from a practical perspective, there is no difference here.
There are, however, at least two reasons for using deontic logic. Let us
try, first, to formalize both Article 415 kc and, e.g., Article 728§1 kc: “the
bank is obliged to inform the account holder about every change in the
account status”. We can formalize this in a deontic calculus as follows:
᭙ x((bank(x) ∧ change(x)) → O(inform(x)))
In first order predicate logic it becomes:
᭙ x((bank(x) ∧ change(x)) → obliged_to_inform(x))
Let us compare now both formalizations of Articles 415 kc and 728§1 kc.
In the former case we have:
and
᭙ x(causes_damage(x) → has_to_redress(x))
᭙ x((bank(x) ∧ change(x)) → obliged_to_inform(x))
whilst in the latter we have:
᭙ x(causes_damage(x) → O(redresses(x)))
and
᭙ x((bank(x) ∧ change(x)) → O(inform(x)))
Only in the latter case is it clearly visible that we are dealing with the
same notion of obligation (the deontic operator O). The formalization in
classical logic forces us to include the notion of obligation in the predicate letters. This may seem counter-intuitive.
The second reason why deontic logic is better for formalizing legal reasoning than classical logic is even more profound. There are situations in
which we infer one norm from another. Here is a simple example: if one
38
CHAPTER 2
should not kill then one should not kill on Sundays. Or: if it is obligatory
that a judge behaves responsibly and it is obligatory that a judge is honest, then it is obligatory that a judge behaves responsibly and is honest.
Finally: if it is obligatory that Adam does not steal Sven’s skies, then it
is forbidden for Adam to steal Sven’s skies.
Let us try to formalize those three arguments using different logical
systems. For the sake of simplicity we will confine ourselves to propositional logic: classical and deontic. Let us begin with classical logic.
Assume that p stands for “one should not kill” and q for “one should not
kill on Sundays”. The problem is that from just p, q does not follow. We
need some other paraphrase. The norm “one should not kill” can be formulated – as above – as a conditional norm, in which the conditions of
applications are tautological (T):
T→p
Because of the fact that T is true in any circumstances, T → p expresses
an unconditional obligation described by p. From this formula, it follows
by the rule of the strengthening of antecedent that:
(T ∧ q) → p
Naturally, q does not stand here for “one should not kill on Sundays”; it
can be expressed by something like “acts on Sundays”. The norm
“one should not kill on Sundays” is represented by the whole expression
(T ∧ q) → p.
Let us look, in turn, at the second argument. Let p stand for “it is obligatory that a judge behaves responsibly”, q for “it is obligatory that a judge
is honest” and r for “it is obligatory that a judge behaves responsibly and is
honest”. From p and q it does not follow that r. This time we can also try
another, perhaps slightly counter-intuitive, paraphrase. Let us assume
that we are dealing with conditional norms: “if someone is a judge, then
she should behave in a responsible way” and “if someone is a judge,
then she should be honest”. Let us formalize them in the following way:
p→q
p→r
From those two premises a sentence follows:
p → (q ∧ r)
which can be read: if someone is a judge then she should behave responsibly and should be honest. It is not exactly what we have been looking
LOGIC
39
for (in the consequent of the norm there are two sentences connected by
a conjunction, and not a single sentence in which there is the compound
predicate “behaves in a responsible way and is honest”). This problem,
however, results from using propositional logic instead of first order
predicate logic. It has nothing to do with the fact that we abstained from
using deontic logic here.
The last of the three arguments poses the biggest challenge for classical logic. If we substitute p for “it is obligatory that Adam does not
steal Sven’s skies” and q for “it is prohibited for Adam to steal Sven’s
skies”, then q, of course, does not follow from p. In the analyzed case,
however, no paraphrase can be found that would enable us to deal with
the problem.
In SDL there is no problem whatsoever! Let p stand for “Adam steals
Sven’s skies”. We can write now:
O¬p
Fp
This is a valid reasoning which leads – as we desire – from the sentence
O ¬ p (it is obligatory that Adam does not steal Sven’s skies) to the sentence Fp (it is forbidden for Adam to steal Sven’s skies). We apply here
simply the definition of the functor F.
SDL deals similarly elegantly with the two previous examples. The
second of them especially takes a simpler form than in the case of classical logic. Let us substitute p for “a judge behaves in a responsible way”,
and q for “a judge is honest”. Let us formalize now both norms that
serve as premises of our argument, i.e., “it is obligatory that a judge
behaves responsibly” and “it is obligatory that a judge is honest”:
Op
Oq
From those two sentences it follows in SDL that:
O(p ∧ q)
which reads: it is obligatory that a judge behaves responsibly and that a
judge is honest.
One can reasonably question whether such arguments, having norms
as premises and conclusions, are important. It seems that the answer
should be positive. First, such arguments are actually carried out, therefore a complete theory of legal reasoning should account for them.
Second, the notion of the set of logical consequences of a given set of
40
CHAPTER 2
norms is used to define the set of valid legal norms. According to the
systemic conception of the validity of law, valid norms are the norms
explicitly enacted by a legislator, plus what follows logically from them.
The analyzed examples allow us to say that formalizations in deontic
logic are better suited to our intuitions than classical formalizations.
Similarly, possible world semantics, as used in deontic logic, seems intuitive. Moreover, SDL is an extension of classical logic. This means that,
as far as the arguments that do not include the operators of obligation,
permission or prohibition are concerned, deontic logic is equivalent to
classical logic. It does not necessarily follow from all this that deontic
logic is not problematic. The most important problems are connected
with paradoxes. But the role of paradoxes is positive: they indicate what
is wrong and encourage the search for new, better deontic systems.
2.4
LOGIC OF ACTION AND LOGIC OF NORMS
2.4.1
Two Types of Obligation
Amongst objections against deontic logics, in addition to the problem of
paradoxes, there are several problematic questions of a more general,
philosophical nature. We will try to look more closely at two such objections: first, the thesis that deontic logic formalizes the notion of oughtto-be, and does not take into account the notion of ought-to-do; second,
the thesis that deontic logic is not a logic of norms because we cannot
say that norms are either true or false. The former problem will serve as
a pretext for discussion of the logic of action. The latter, in turn, will
allow us to comment on the Jørgensen Dilemma.
Philosophers sometimes distinguish between two concepts of obligation: the first stating what ought to be the case and the second stating
what ought to be done.22 The importance of this distinction is questioned
by those who claim that the latter can be reduced to the former. They
insist23 that the sentence “person α ought to do p” is equivalent to the
sentence “it ought to be the case that person α does p”. If one approves
of this reduction then deontic logic, as described in the previous sections,
which is a logic of the ought-to-be operator, is adequate. There are, however, strong objections against reducing ought-to-do to ought-to-be.24 Let
us present one of them. P. Geach suggested analyzing the following sentence25: “Fred ought to dance with Ginger”. According to the reductionist conception, that sentence is equivalent to this: “it ought to be the case
that Fred dances with Ginger”. The sentence “Fred dances with Ginger”
is, however, equivalent to “Ginger dances with Fred” (for the relations
of dancing are symmetrical). Instead of saying “it ought to be the case
LOGIC
41
that Fred dances with Ginger” we could also say: “it ought to be the case
that Ginger dances with Fred”. But now, reversing the direction of the
first transformation, we can write that “Ginger ought to dance with
Fred”. This seems counter-intuitive. From the sentence that Fred ought
to dance with Ginger it does not necessarily follow that Ginger ought to
dance with Fred.
This problem may also be illustrated from a legal perspective. Let us
imagine two persons, John and Adam, concluding an agreement according
to which, when a certain condition is fulfilled, John will be under an obligation to sell Adam his car, but Adam will have the right to choose whether
he wants to conclude the final agreement and buy the car or not. If a similar analysis to that carried out in the case of Fred and Ginger is carried out
here, from the sentence “John ought to conclude a sale agreement with
Adam”, and from the fact that concluding an agreement is a symmetrical
relations, it would follow that Adam ought to conclude a sale agreement
with John, which in the described circumstances is paradoxical.
Both presented examples show that the idea of reducing “α ought to
do p” to “it ought to be the case that α does p” can lead to counter-intuitive results. One may wonder, however, whether the problems are really
connected with the failure to distinguish between ought-to-be and oughtto-do. The Fred and Ginger example indicates one more feature of SDL:
that the concept of obligation involved is impersonal. In our example we
started with the sentence “Fred ought to dance with Ginger” and substituted it with “it ought to be the case that Fred dances with Ginger”. It is
suggested, sometimes, that distinguishing the “obligation from the point
of view of Fred” from the “obligation from the point of view of Ginger”
suffices to solve the puzzle in question. The sentence “it ought to be the
case that Fred dances with Ginger” and the equivalent sentence “it ought
to be the case that Ginger dances with Fred” express obligations from the
point of view of Fred. In consequence, from the sentence “(From the point
of view of Fred) it ought to be the case that Ginger dances with Fred”
one cannot derive that Ginger ought to dance with Fred. In this way we
concede that obligations differ (for they always relate to a specific person), but we are not forced to say that ought-to-do cannot be reduced to
ought-to-be. This observation will be confirmed when we look in greater
detail at the deontic logic of action.
2.4.2
Logic of Action
We would like to analyze now some formalizations of ought-to-do. It is
interesting that the first attempts at constructing deontic logic aimed to
42
CHAPTER 2
capture the ought-to-do. G.H. von Wright in 1951 proposed a system
equipped with symbols A, B, C,. . . to which deontic operators, like O
(“it is obligatory that. . .”) were added. A, B, C,. . . stood for “general
actions”, e.g., theft, sale, etc. Contemporarily, various strategies of constructing deontic logics of action are employed.26 We will look more
closely at two of them.
The following intuition is behind the first of the strategies. Human
actions bring about changes in the world. For instance, if the action is
building a bridge, the change in the world consists in the appearance of
a bridge. In describing the state of the world prior to the action, the sentence “there is a bridge here” is false, whereas after the action it is true.
Human actions lead us, therefore, from one state of the world to another.
Or, in other words, they constitute a move from one possible world to
another. This simple analysis prompts the following idea: on the semantic level actions are represented by pairs of possible worlds – the first
world of the pair is the situation in which the action is undertaken, and
the second is the situation in which the action ends. Observe that an
expression denoting an action is not defined by a single pair – “world
before action – world after action” – but by all such possible pairs.
Therefore, the action “John builds a bridge” is, on the semantic level, a
set of all pairs of possible worlds, of which the first is a world in which
there is no bridge and the second is a world in which there is a bridge.
Each pair can be labeled an execution of the formalized action.27
Now, it suffices to apply a procedure similar to that of SDL in order
to define which actions are obligatory, which are forbidden and which
are allowed. Recall that in the logic described in the previous section the
act of creating a norm consisted in identifying a set of possible worlds
which we called deontically perfect. In the deontic logic of action
“moves” between the worlds (ordered pairs of worlds) are divided into
legal (the Leg set) and illegal (the Illeg set). We shall say that an action
A is forbidden in world w if the set of all executions of A in w is included
in Illeg. Similarly, an action A is allowed in world w if at least one execution of A in w belongs to Leg.28 In order to define obligation let us
assume that OmA means nonexecution of A, i.e., it is an execution of
any action which is not A. We shall say that the action A is obligatory in
w if all executions of OmA in w belong to Illeg.
One can query whether a logic thus constructed is better – and in what
respects – than “normal” deontic logics. The first reason to claim this is
the philosophical motivation that stands behind the proposed system –
the distinction between ought-to-be and ought-to-do. Another advantage
of this system over “normal” logics is connected to the fact that, in the
LOGIC
43
logic of action, certain kinds of obligation can be expressed that cannot
be reconstructed in SDL. The latter concerns only “ideal” situations but
cannot deal with “sub-ideal” ones, i.e., obligations which must be fulfilled in situations in which other obligations have already been violated.
In order to make such a reconstruction in our logic of action executions
of certain actions that lead from one sub-ideal world to another subideal must be included in the set Leg. Another desirable feature of the
present system is that it can easily be “personalized”, i.e., obligations can
easily be ascribed here to specific persons.
Our second example of a formal system that tries to capture the
“ought-to-do” is a deontic logic developed with the use of STIT logic,
created in the 1980s by N. Belnap.29 STIT is a logic that includes the
operator “See To It That”. This operator is defined in a very rich semantic structure, constructed with the use of a technique called branching.
The basic semantic ideas of STIT are extremely simple and intuitive. Two
fundamental concepts of STIT are that of a moment and a history.
Moments are ordered (they form transitive and nonreflexive relation).
Two moments can belong to the same, or two different, histories. This is
depicted in the figure below as a tree, which, from the bottom-up represents the direction of the flow of time.
h4
h3
h2
h1
m4
h5
h6
h7
m3
m5
m2
m1
m6
h8
44
CHAPTER 2
Let moment m1 be the actual moment. As we can see, “below” m1 the tree
does not branch. Intuitively, this corresponds to the thesis that the past
is fixed (fully determined). However, “above” m1 our tree has several
branches that, taken together, represent the undetermined future. Every
“maximal” branch of the tree represents a certain history. For instance,
the branch that goes through m1, m2, m3 and onward, constitutes the history h1 (or h2), and the branch that goes through m1, m2 and m4 constitutes the history h3 (or h4). It is useful to denote by Hm the set of all
histories “going through” the moment m (therefore, for instance, Hm1 =
{h1,h2,h3,h4, h5, h6,h7,h8}, and Hm4 = {h3,h4}). From the perspective of the
logical systems we have already presented, every moment is a certain possible world (possible state of affairs); hence, in a propositional logic a
moment is represented by the valuation function that ascribes to all the
atomic expressions of a language the values of truth or falsehood.
In such semantics the concept of action is encoded by defining the
functor “see to it that”, which we write formally:
[α stit A]
this is read: person ⬀ sees to it that A. In order to give a semantic characteristic to [α stit A] we must introduce the concept of choice. From the
intuitive point of view this is simple. In every moment m person α can
choose from various actions and her choice determines the future, i.e.,
determines which history will be realized. It is not the case, however, that
the choices of a given person determine the future univocally. The below
figure illustrates this. In moment m person α has three possible choices:
w1, w2 and w3. Choices w1 and w3 determine one, concrete history (in the
case of the former – h1, and of the latter – h5). The choice of w2 does not
determine the future univocally, reducing only the set of possible histories to h2, h3 and h4.
h2
h3
h1
Ought(m)
h4
h5
A
-A
A
A
-A
w1 w2 w3
m
An important feature of STIT is that, in it, we ascribe to different people the choices they can make at moment m. If person α has at m three
LOGIC
45
possible choices, it does not mean that person β also has three choices;
moreover, the choices of β can “divide” the set of possible histories in a
different way than the choices of α can.30 Let us use Choiceαm to denote
the set of all choices of α at m. Let Choiceαm (hi) stand for the choice that
includes history hi. Therefore, e.g., Choiceαm (h1) = w1.
We can now define the functor [. . . stit . . .]. The expression [α stit A]
(“Person α sees to it that A”) is true at moment m and history h, if and
only if A is true for every history h’, which belongs to Choiceαm (h). Look
at the example depicted in the figure above. [α stit A] is true at moment
m for h1 (since h1 is the only history belonging to w1 and A is true in h1);
in the case of the remaining histories [α stit A] is false. For h5 this is obvious. The falsity of [α stit A] at m for h3 and h4 results from the fact that
in one of the histories (h2) belonging to the same choice as h3 and h4, A
is false. It seems that this definition of the truth of [α stit A] is sound.
Person α sees to it that A, if her choice results in A being true.
In order to take the last step – introduce to our semantic structure the
functor of obligation O – we need, as in the case of the deontic logics
discussed above, a norming function. This time “norming” means picking out for every moment m a set Ought(m) of those histories that are
desirable from the point of view of a legislator. We will say that at a
moment m and for a history h it ought to be that X (we write: OX), if and
only if X is true at moment m and a history hi for every history hi belonging to Ought(m). For instance, in the situation depicted in the figure, the
sentence OA is true at m and h1, for A is true at m for the only history
belonging to Ought(m), i.e., for h1.
From our point of view the most interesting question is when the
expression O[α stit A] (“it ought to be the case that person α sees to it
that A”) is true. According to the above definition O[α stit A] is true at
moment m and history h, if [α stit A] is true at m for every history belonging to Ought(m). In our example, only h1 belongs to Ought(m), and we
have already determined that [α stit A] is true at m for h1. Therefore, O[α
stit A] is true at m for all the histories h1–h5. If, however, the set Ought(m)
included, apart from h1, also h3, the expression O[α stit A] would be false
at m for all the histories, since at m/h3 [α stit A] is false. What is interesting, is that in the same situation OA would be true (for A is true both in
h1 and in h3). This shows that, in the semantics under consideration, the
expression “it ought to be the case that A” is not equivalent to “it ought
to be the case that person α sees to it that A”.
The system described above will not be developed any further.31 We
would like, however, to point out several facts. First, the expression O[α
stit A] is not yet the ought-to-do obligation. J.F. Horty, the author of the
46
CHAPTER 2
presented conception, indicates that even the operator O, as now
defined, suffices to solve the “Fred and Ginger puzzle”. Observe that the
expression “it ought to be the case that person α sees to it that A” cannot
be reduced to the expression “it ought to be the case that A”. The difference between OA and O[α stit A] is based on the fact that, in the
latter case the obligation is in a way personalized. This supports our
diagnosis that the paradoxical character of the Fred and Ginger
case stems from the fact that obligations in traditional deontic logics are
not “personalized”; a contrario, it is not a result of reducing ought-to-do
to ought-to-be. Horty shows, however,32 that there exist situations
which cannot be adequately accounted for with the use of the functor
O[. . .stit. . .]. In order to describe those situations he proposes – in the
framework provided by STIT – another deontic operator which encodes
ought-to-do.33
Secondly, the semantics of STIT enables Horty to build a system that
takes into account the obligations of many persons. Such logic has
important consequences for the problem of the “group ought”. One can
identify at least a few such obligations. Let us look at two examples. A
group of pupils has two obligations: to clean the blackboard before the
lesson begins, and to keep quiet during the lesson. The first of the obligations will be fulfilled if any of the pupils cleans the blackboard.
Fulfillment of the second requires that all the pupils stay quiet. The distinction between those two kinds of obligation is possible only when the
language of our logic can not only express the fact that there are different obligations on different persons, but can also account for the complicated relations between those obligations.
The logics of action constitute, as we have seen, an interesting alternative for traditional deontic logics. They may be characterized, first and
foremost, as using rich and intuitively sound semantic structures. The
search for such structures, which are able to model more and more complex situations, is one of the most important directions of contemporary
logical research.34
2.4.3
Jørgensen Dilemma
Up to now the focus has been on when sentences which take the form “it
ought to be the case that p”, can be labeled true. But can such sentences
be true or false at all? It seems that one can ascribe truth or falsehood to
descriptive sentences that inform us about facts. Questions, orders and
norms, on the other hand, do not seem to fall into categories possessing
truth values. One may maintain that this is not an important problem;
but the fact is that contemporary logic – or at least the commonly
LOGIC
47
accepted part of it – concerns expressions that are true or false. If we
reflect on these two observations, i.e., that:
(1) Only true or false sentences can serve as premises or conclusions in
logically valid arguments.
and
(2) Norms cannot be ascribed truth values.
then we should conclude that:
(3) Norms cannot serve as premises or conclusions in logically valid
arguments.
Our conclusion (3) puts into doubt the possibility of developing any logic
of legal reasoning. However, we do put forward many legal arguments
every day, and they seem intuitively correct. Therefore, we can note:
(4) Intuitively correct normative arguments do exist.
Theses (1) – (4) constitute a dilemma that was first described by
Jørgen Jørgensen in a paper published in 1938.35 It should be added that,
in the original formulation, the Jørgensen Dilemma concerned imperatives, not norms.
It is not difficult to argue that the Jørgensen Dilemma poses a fundamental challenge for any formal reconstruction of legal, or, more
generally, normative, discourse. As we have already observed, the acceptance of thesis (3) of the Dilemma leads directly to questioning the very
possibility of a logic of norms. For that reason, it is necessary to devote
some space to analyzing possible solutions to the Dilemma.
Thesis (3) of the Jørgensen Dilemma is a logical consequence of theses
(1) and (2). Therefore, in order to resolve the Dilemma one can question
thesis (1), (2) or (4). We would like to start with the latter possibility,
observing that thesis (4) can be understood in two ways. First, the expression “normative arguments” can mean “arguments using norms as
premises and conclusions”. On the other hand, however, “normative
arguments” can consist of other expressions than norms. We have therefore two versions of thesis (4): (a) when “normative arguments” means
“arguments using norms”; or (b) when we maintain that intuitively correct normative arguments are built of some other expressions. The distinction between (4a) and (4b) is useless if we cannot identify those
“other expressions”. Philosophers and logicians have named and defined
such a category of expressions called deontic sentences. A deontic sentence is an expression stating the existence of an obligation relative to a
certain deontic system. The following is an example of a deontic sentence: “According to valid law, John ought not to steal”. It is usually held
that such sentences, which describe only what is obligatory, prohibited or
permitted relative to a certain normative system can be ascribed truth
48
CHAPTER 2
values. If, however, deontic sentences are true or false, then there can be
logical relationships between them.36
Therefore, if we choose to understand thesis (4) of the Dilemma in the
(4b) sense, then the Dilemma can be resolved, assuming that the “normative arguments” in question are arguments made up of deontic sentences.
If we accept (4a), however, the problem remains. Of course, the distinction between norms and deontic sentences is problematic. We will not go
into the details of the debate on this issue. However, we must address a
terminological problem. Sometimes deontic logic is distinguished from
the logic of norms. The former is thought to concern deontic sentences,
i.e., expressions that are true or false. The latter concerns norms, i.e.,
expressions which are neither true nor false. This conception – that norms
cannot be ascribed the value of truth – is called noncognitivism. From the
perspective of noncognitivism it is essential to distinguish carefully
between deontic logic and the logic of norms, whilst in the rival theory –
cognitivism – this distinction is not required.
As already observed, the Jørgensen Dilemma can also be resolved by
questioning thesis (1) or thesis (2). According to thesis (1) logical relations obtain only between sentences that are true or false. It must be conceded that not all logicians and philosophers agree with this stance. This
is proven by some logical systems that are not based on truth values. The
proponents of such logics have to deal with very serious problems,
mainly related to the fact that basic and commonly accepted metalogical
concepts, such as satisfaction or soundness, are based on the notion of
truth. The adaptation of those concepts to a logic that makes no use
of notions of truth and falsehood is not a trivial task. Therefore, the
solution to the Jørgensen Dilemma that consists in abandoning thesis
(1) remains highly problematic.
The last of the proposed solutions to the Dilemma is to abandon thesis (2). This thesis is questioned by cognitivists, i.e., those who ascribe
truth values to norms. There is some agreement that the cognitivist
theory of Aquinas is one of the most coherent. According to him, the
norms that we should follow in our lives are only an inferior copy of eternal divine law. Because of their pedigree, those norms can be ascribed
truth (and norms incompatible with them can be designated as false).
A closer analysis of Thomistic philosophy reveals, however, that one can,
with a sufficient degree of certainty, express only one – the most general –
norm: bonum est faciendum, good should be done. To deduce more precise rules of behavior from this general norm is a question of individual
decision, rather than of well established logical operations.37 It is therefore difficult to “transfer” truth from the general norm (which is true
LOGIC
49
because of its pedigree) to the specific rules we use every day. Other versions of cognitivism are also troublesome. Usually, the notions of truth
they apply are problematic. Therefore, this solution to the Jørgensen
Dilemma is not commonly accepted either.
In recent years the discussion surrounding the Jørgensen Dilemma has
become less and less intense, although there emerge, from time to time –
new attempts to deal with it.38 Despite the fact that there is still no
commonly accepted solution to the Dilemma, the research on logical
reconstruction of legal reasoning goes on, and each year new deontic
logics or logics of norms are developed. This may well stem from the fact
that the Jørgensen Dilemma continues to be a challenge for logicians and
thus forces them to search for new formalisms and ideas. Most of the
constructed logics of normative discourse are based on notions of truth
and falsehood. This is also true of the logical systems presented above.
This does not mean that we claim the impossibility of a logic of norms
in which expressions cannot be ascribed the value of truth or falsehood.
Our choice was motivated only by the importance the presented systems
enjoy in contemporary legal theory. Whether we can treat them as proper
logics of norms, or “only” as deontic logics depends on the accepted
solution to the Jørgensen Dilemma, which, in turn, is based on some
ontological choices.
2.5
2.5.1
DEFEASIBLE LOGIC
The Concept of Defeasibility
We would like to turn now to a discussion of defeasible logic. Research
on this type of logical system began in the 1970s. The concept of defeasibility, however, was introduced much earlier. It appeared in H.L.A.
Hart’s paper “The Ascription of Responsibility and Rights”, published
in 1948. Hart writes:
When the student has learnt that in English law there are positive conditions required for
the existence of a valid contract, (. . .) his understanding of the legal concept of a contract is still incomplete (. . .). For these conditions, although necessary, are not always
sufficient and he has still to learn what can defeat a claim that there is a valid contract,
even though all these conditions are satisfied. The student has still to learn what can follow on the word “unless”, which should accompany the statement of these conditions.
This characteristic of legal concepts is one for which no word exists in ordinary English.
The words “conditional” and “negative” have the wrong implications, but the law has a
word which with some hesitation I borrow and extend: this is the word “defeasible”, used
of a legal interest in property which is subject to termination or defeat in a number of
different contingencies but remains intact if no such contingencies mature. In this sense,
then, contract is a defeasible concept.39
50
CHAPTER 2
Hart’s idea amounts to a declaration that certain legal concepts, like that
of a binding contract, lack definite conditions of application. Unforeseen
circumstances can always occur, causing us to withdraw the claim that we
are dealing with a valid contract, although the usual conditions for its
validity are fulfilled.40 An important point must be stressed here. We call
certain contracts “valid” because they were concluded in accordance
with certain legal rules. Therefore, the ascription of a legal concept hangs
together with the fulfillment of a legal norm. If we say – after Hart –
that legal concepts are defeasible, then so are legal norms. A legal norm
is defeasible if there are situations in which the conditions of that norm’s
application obtain, but the norm is not applied.
Defeasibility thus defined leads to some logical problems. If we reconstruct a legal norm, as we did above, with the use of material implication:
h→d
(h stands for the norm’s antecedent, and d for the consequent), we will
not be able to say that the norm is defeasible. This is because in the case
of defeasible norms, it is possible that h → d is valid, h obtains, but we
cannot deduce d. In classical logic (including deontic logic based on classical calculi) this cannot be the case, since if we have h → d together with
h, d follows on the basis of modus ponens.
It is clear from the above that acceptance of the thesis that legal rules
are defeasible forces us to look for an alternative logic of legal discourse.
Such logic has been developed, not in the field of legal theory, but within
research on artificial intelligence. It turns out that the problem of defeasibility is important not only for legal or normative reasoning, but also in
theoretical discourse. Logicians developing artificial languages for computer systems encountered the following problem.41 It happens that a man
(or a computer system) has to reason with uncertainty as to whether all
relevant information has been collected. For instance, when we know that
Tweety is a bird, it is reasonable to say that Tweety flies. If, however, we
had additional information that Tweety is a penguin or a baby bird, then
we would have to withdraw from saying that Tweety flies (for we know
from elsewhere that if a bird is a penguin or a baby bird it does not fly).
It follows from this example that such conditionals as “if x is a bird, then
x flies” are simply false, but nevertheless, we sometimes use them in our
reasoning. Naturally, such conditionals cannot be formalized as a material implication. A new, nonclassical implication needs to be sought.
Defeasible logic constitutes such a nonclassical system. It is an example of nonmonotonic logic.42 It is instructive to expand here on the
meaning of “nonmonotonic”. Classical logic is monotonic. This means
LOGIC
51
that if a sentence p follows from a set of premises A, then p follows also
from a set B, which is a superset of A. Every logic which lacks this feature is nonmonotonic. It is easy to show that our Tweety example
requires a nonmonotonic logic. In the example we first infer from two
premises – “if x is a bird, then x flies” and “Tweety is a bird” – that
“Tweety flies”. Later, we add the information that “Tweety is a penguin”
(and we know that penguins do not fly). From this extended set of
premises the conclusion that Tweety flies no longer follows.
2.5.2
Defeasible Logic
There are many defeasible logics.43 In this section we would like to present one of them,44 concentrating on its main ideas and omitting technical details.
Our defeasible logic (in short: DL) operates on two levels. On the first
level arguments are built from a given set of premises; on the second level
the arguments are compared in order to decide which of them prevails.
The conclusion of which argument is “best” becomes the conclusion of
the given set of premises.
The language of DL is the language of first order predicate logic,
extended by the addition of a new functor, the defeasible implication, for
which we will use the symbolfi. For defeasible implication there exists a
defeasible modus ponens, analogous to that of the material implication:
A⇒B
A
B
The difference between material and defeasible implications is visible
only on the second level of DL.
The language of DL serves the building of arguments. In our Tweety
example we have two situations. In the first, three sentences belong to
our set of premises: “if x is a bird then x flies”, “Tweety is a bird” and “if
x is a penguin then x does not fly”. The first of the premises can be formalized in the following way:
bird(x) fi flies(x)
The second premise is, of course:
bird(tweety)
And the third:
penguin(x) fi ¬(flies(x)) 45
52
CHAPTER 2
This set of premises enables us to construct only one argument. With the
help of defeasible modus ponens we obtain:
bird(x) fi flies(x)
bird(tweety)
flies(tweety)
The addition of a fourth premise:
penguin(tweety)
enables us to build the following argument:
penguin(x) fi ¬(flies(x)
penguin(tweety)
¬flies(tweety)
Having those two arguments we can move to the second level of DL, in
which the arguments are compared in order to decide which is better,
and in consequence which of the sentences – flies(tweety) or
¬flies(tweety) – should be regarded as the conclusion of our set of
four premises.
In the second level of DL two concepts play a crucial role: attack and
defeat. We shall say that an argument A attacks an argument B if the
conclusions of both arguments are logically inconsistent.46 In our example that is the case since flies(tweety) and ¬flies(tweety) are contradictory. If two arguments compete with one another, one must know
how to decide which argument prevails, i.e., which defeats the other.
Various ways of comparing attacking arguments have been developed.47
The easiest and most flexible is the following. One checks what the defeasible implications that served to build the attacking arguments are. It is
assumed that those implications are ordered. In a comparison an argument wins when it is built with the use of a defeasible implication that is
higher in the order. In our example the first argument is based on the
implication bird(x) fi flies(x), whilst the second is based on penguin(x) fi ¬flies(x). It is reasonable to assume that the second implication is higher in the ordering, since it represents a stronger tie – there
are exceptions to the rule that if something is a bird then it flies, but the
second rule that penguins do not fly, is exceptionless. If penguin(x) fi
¬flies(x) is higher in the ordering than bird(x) fi flies(x), then the
second argument defeats the first.
LOGIC
53
The conclusion of which argument prevails in a comparison of all
competing arguments built from the given set of premises, is the logical
conclusion of this set. In the first situation our set of premises contained
only three sentences (bird(x) fi flies(x), penguin(x) fi ¬flies(x),
bird(tweety)), which enabled us to build only one argument. The conclusion of this argument, flies(tweety), is the logical conclusion in the
first situation. In the second situation another sentence is added to our
premises: penguin(tweety). This made it possible to construct the second argument. Both arguments attack one another and the second argument wins. Therefore, its conclusion, ¬flies(tweety), and not the
conclusion of the first argument, follows logically in the second situation. It is clear from this that DL is nonmonotonic. In the first situation
flies(tweety) was the logical conclusion, but in the second, in which
the set of premises is extended, flies(tweety) no longer follows.
2.5.3
Objections Against Nonmonotonic Logic
At the beginning of this chapter we attempted to define what logic is. The
definition we proposed poses a serious challenge for defeasible (or, more
generally, nonmonotonic) logics. It prompts doubts as to whether these
systems are logics at all.
As already observed, the key insight regarding the nature of logic was
formulated by Tarski in his definition of logical consequence. The definition may be somewhat boldly presented as follows:
A sentence A follows logically from the set of premises Γ if and only if in every case in
which the premises of Γ are true, A is also true.
A short reflection enables us to say that Tarski’s analysis shows our intuitive notion of logical consequence to be monotonic (even if we extend
the set of premises Γ, it still will be a case in which all the sentences of Γ
are true; therefore if after the extension of Γ, A ceases to follow from it,
as is the case with nonmonotonic logics, such a notion of logical consequence is incompatible with Tarski’s analysis). Moreover, the concept of
truth also seems to be “monotonic” (it is difficult to assume that the
addition of a new premise can make false a conclusion thus far considered true). In such a situation, the idea of logic as a set of rules for the
“transmission of truth” must be abandoned. Instead some theoreticians
are inclined to speak of the “transmission of justification”. The role of
nonmonotonic logic would be to determine which forms of reasoning
lead from justified premises to a justified conclusion.
Abandoning Tarski’s analysis also results in abandoning the intuitively
appealing soundness theorems. This is problematic as regards the
54
CHAPTER 2
question of whether one can “trust” nonmonotonic logic. On the other
hand, the concepts of the second level of DL seem sound and the analyses carried out using DL demonstrate the flexibility and usefulness of
this formal system.
Let us look more closely now at some examples of formalizations in
DL. This will enable us to formulate several arguments in favor of nonmonotonic logic. We will identify two important features of DL formalizations: modularity and structural resemblance between legal texts and
their formal counterparts. Further, we will show how DL deals with
some hard cases. A comparison of defeasible and classical techniques
will highlight some additional problems of the formal reconstruction of
legal reasoning.
2.5.4
Examples
Some peculiar logical problems are connected with the structure of legal
texts. Let us look at the following example. Let Article 1 say that the full
capacity to perform legal acts is granted once a person is 18 years old;
Article 2, in turn, constitutes an exception to Article 1, stating that
persons declared mentally ill by a court do not have the capacity to
perform legal acts. An attempt to formalize those two provisions in classical monotonic logic leads to the following results:
A1: ᭙x((18_years(x) ∧ ¬mentally_ill(x)) →
capacity(x))
A2: ᭙x(mentally_ill(x) → ¬capacity(x))
A distinctive feature of this formalization is that the formula representing Article 1 includes the predicate “mentally_ill”, and therefore it
takes into account the exception stated in Article 2. Our formalization
mixes up, then, information from two different provisions. Such circumstances do not occur when a nonmonotonic system is used. In DL
Articles 1 and 2 take the following form:
A1: 18_years(x) fi capacity(x)
A2: mentally_ill(x) fi ¬capacity(x)48
In DL the information contained in Articles 1 and 2 is not “mixed up”.
Therefore, the defeasible formalization resembles structurally legal texts.
The presented formalization, apart from being structurally similar to
legal texts, displays modularity. Imagine introducing Article 3, stating
another exception to Article 1, for instance that married men do not have
LOGIC
55
the capacity to perform legal acts. In the classical formalization this
causes a revision of the formula representing Article 1:
A1: ᭙x((18_years(x) ∧ ¬mentally_ill(x) ∧
¬married(x)) → capacity(x))
A formula representing the new provision is also needed:
A3: ᭙x(married(x) → ¬capacity(x))
In DL the introduction of Article 3 is much easier. It suffices to add:
A3: married(x) fi ¬capacity(x)
The formalization in DL displays modularity because adding a new provision does not lead to the revision of the formulas formulated earlier.
Modularity and structural resemblance in nonmonotonic systems,
such as DL, can be fully appreciated when we imagine that, together with
Article 3, the legislator enacts also Article 4, which states an exception to
Article 3 saying that those married men whose last names begin with C
have the capacity to perform legal acts.
Let us recall that the formalization of Articles 1–3 in classical first
order predicate logic looks like this:
A1: ᭙x((18_years(x) ∧ ¬mentally_ill(x) ∧
¬married(x)) → capacity(x))
A2: ᭙x(mentally_ill(x) → ¬capacity(x))
A3: ᭙x(married(x) → ¬capacity(x))
Assume that name_C stands for the predicate from Article 4. Then, this
article can be formulated as follows:
A4: ᭙x((married(x) ∧ name_C(x)) → capacity(x))
However, we have to change also the formalization of Article 3:
A3: ᭙x((married(x) ∧ ¬name_C(x)) → ¬capacity(x))
In DL, in which we had:
A1: 18_years(x) fi capacity(x)
A2: mentally_ill(x) fi ¬capacity(x)
A3: married(x) fi ¬capacity(x)
56
CHAPTER 2
we only need to add:
A4: (married(x) ∧ name_C (x)) fi capacity(x)49
Modularity and structural resemblance may seem weak arguments in
favor of nonmonotonic systems. There are some facts, however, which
testify to the contrary. The nonmonotonic formalizations lead not only
to simpler results as regards “quality”, but also as regards “quantity”.
For instance, an attempt to formalize the provisions of the Polish penal
code concerning killing in classical logic results in more than 100 formulas. A similar formalization in DL requires only 33 formulas.50
In order to illustrate this, and to formulate one more argument in favor
of nonmonotonic logics, let us try to formalize Article 148§1 of the
Polish penal code (kk). The provision says that whoever kills a man shall
be imprisoned for at least 8 years. This can be formalized in classical
logic the following way:
148§1 kk: ᭙x(kills(x) → punishment(x))
This is not a complete formalization, however. It does not take into
account, for instance, the exception stated in Article 148§2 kk, which
qualifies some types of killing. If we cover them with the predicate qualified, our formalization must be changed in the following way:
148§1 kk: ᭙x((kills(x) ∧ ¬qualified(x)) →
punishment(x))
One must add to this a formula representing Article 148§2. Exceptions to
Article 148§1 kk can be found also in the remaining part of Article 148
and in Articles 149–151 kk. It should also not be forgotten that in the
general part of the penal code there are provisions concerning guilt and
self-defense that also constitute exceptions to Article 148 kk. As a result,
a formalization of Article 148 kk in classical logic – due to its lack of
modularity and structural resemblance – requires that at least ten, if not
more, exceptions be taken into account. This causes the following problem: if this formalization is accepted as the basis of a judge’s decision,
the judge would be required to check whether any of the exceptions to
Article 148 kk have occurred in making a decision; so, the judge would
have to question whether the killing in question is an act of euthanasia,
killing with particular cruelty, etc. In actual cases such justifications do
not exist. The judge tackles directly only those questions, which are
obviously relevant. It seems that such a process of the application of law
can be successfully modeled using nonmonotonic systems, for they offer
formalizations that are modular and display structural resemblance.
LOGIC
57
Even more important theoretical and logical problems are connected to
hard cases. The most widely popularized such case seems to be Riggs vs.
Palmer, described in Taking Rights Seriously by R. Dworkin.51 These are
the facts: Elmer Palmer murdered his grandfather, Francis Palmer.
According to the applicable law of succession, Elmer was to inherit part of
Francis’ property. The law in question did not contain any provision that
would deprive Elmer of his right to inherit because of what he had done.
The New York Court of Appeals decided, however, that Elmer had no right
to the inheritance, because “no man should profit from his own wrong”.
Dworkin interprets the court’s decision in the following way: in a legal
system there are two types of legal norms – rules and principles. Legal
rules, such as the norm that gave Elmer the right to inherit are applied in
an “all-or-nothing” fashion: they are either fulfilled or not, tertium non
datur. Legal principles, on the other hand, have the “dimension of
weight”, i.e., they may be taken into account to greater or smaller
degrees. Moreover, principles can, in particular cases, “produce” exceptions to legal rules. In Riggs vs. Palmer we are dealing with such a case.
The legal principle “No man shall profit from his own wrong” ‘produces’
an exception to the rule that gives Elmer his right to inherit.
Let us try to look at the situation from a logical point of view.
We have the following predicates: dies, grandfather and inherits.
Observe that both grandfather and inherits are two-place predicates for we will not say “Francis is a grandfather”, but “Francis is
Elmer’s grandfather”; similarly, we will say “Elmer inherits from
Francis”, and not “Elmer inherits”. In classical first order predicate logic
the rule of the law of succession, which determines that if someone dies
and has a grandson, the grandson has a right to the inheritance, can be
formalized in the following way:
R: (᭙x) (᭙y) ((dies(x) ∧ grandfather(x,y)) →
inherits(y,x))
The principle, in turn, which says that “No man shall profit from his own
wrong” can be written:
P: (᭙x) (wrong(x) → ¬profit(x))
Since Francis died (dies(francis)), and he was Elmer’s grandfather
(grandfather(francis,elmer)), then on the basis of modus ponens we
can conclude that Elmer benefits from Francis’ inheritance (inherits(elmer, francis)):
(᭙x) (᭙y) ((dies(x) ∧ grandfather(x, y)) → inherits
(y, x))
58
CHAPTER 2
dies(francis)
grandfather(francis, elmer)
inherits(elmer, francis)
On the other hand, Elmer did wrong (killing Francis) and therefore,
according to the principle we formulated, he shall not profit from his act
(¬profit(elmer)):
(᭙x) (wrong(x) → ¬profit(x))
wrong(elmer)
¬profit(elmer)
If we assume what seems obvious – that the fact of inheriting is an
instance of profit (it can be formalized as: (᭙x) (᭙y) (inherits(yx) →
profit(y))), then our formalization of the rule R and the principle P
produces a contradiction. Using the rule we obtain inherits(elmer,
francis), and hence, on the basis of the just formulated relationship,
profit(elmer); applying the principle, on the other hand, leads us to the
conclusion ¬profit(elmer).
Our analyses suggest a way out of this problem: in the formalization
of the rule R we must include the exception “produced” by the principle
P. R thus becomes:
(᭙x) (᭙y) ((dies(x) ∧ grandfather(x,y) ∧ ¬wrong(x)) →
inherits(y,x))
Now, the argument leading to the conclusion that Elmer benefits from
Francis’ inheritance (inherits(elmer, francis)) is blocked:
(᭙x) (᭙y) ((dies(x) ∧ grandfather(x,y) ∧ ¬wrong(x)) →
inherits(y,x))
dies(francis)
grandfather(francis, elmer)
wrong(elmer)
Now we cannot apply modus ponens to R for wrong(elmer) obtains, and
not ¬wrong(elmer).
Is the presented solution satisfactory? It is easy to observe that R does
not resemble structurally the rule it stands for. This formalization is not
modular either. One can easily imagine that the norm saying that a
LOGIC
59
grandson benefits from the inheritance of his late grandfather, could
“lose” against some other principle. This other exception would also have
to be included in the formulation of R. If there is interaction between
rules and principles, the lack of modularity has, however, catastrophic
consequences. Principles can “produce” exceptions to rules in particular
cases and the number of those exceptions is theoretically unforeseeable
and potentially infinite. Therefore, one can never construct “the final”
formalization of any legal rule, for there is always a possibility that in a
certain case a principle will “produce” an additional exception.
Those problems are omitted when one shifts to nonmonotonic logic.
In DL, R becomes:
R: (dies(x) ∧ grandfather(x, y)) fi inherits(y, x)
and the principle:
P: wrong(x) fi ¬profit(x)
We have to add, as in the case of the classical formalization, that:
(᭙ x) (᭙ y) (inherits(y, x) → profit(x))
Modularity in the nonmonotonic formalization makes it possible to deal
with the potentially endless list of exceptions to R “produced” by different principles very easily. Those exceptions do not have to be included in
the formulation of R.
Up to now, we have not looked at how legal norms are applied in
DL. This process is highly characteristic and may even be deemed
problematic.
Let us recall, first, the classical, monotonic formalization of Articles
1, 2 and 3 introduced above (for the sake of simplicity we omit Article 4):
A1:
A1: ᭙x((18_years(x) ∧ ¬mentally_ill(x) ∧
¬married(x)) → capacity(x))
A2: ᭙x(mentally_ill(x) → ¬capacity(x))
A3: ᭙x(married(x) → ¬capacity(x))
Imagine two situations. In the first John is more than 18 years old, is not
mentally ill and is not married. On the basis of Article 1 we conclude that
John has the capacity to perform legal acts (capacity(john)):
᭙x((18_years(x) ∧ ¬mentally_ill(x) ∧ ¬married(x))
→ capacity(x))
60
CHAPTER 2
18_years(john)
¬mentally_ill(john)
¬married(john)
capacity(john)
We reached this conclusion by applying to the formalization of Article 1
and to the known facts the simple scheme of modus ponens. The
same scheme can be applied in the second situation, in which John, in
addition to being over 18 years old, is married. This time we conclude on
the basis of Article 3 that John does not have the capacity to perform
legal acts:
᭙x(married(x) → ¬capacity(x))
married(john)
¬capacity(john)
Determination of the logical consequences of legal norms in both
situations is more complicated in the case of the nonmonotonic
formalization. In DL the first situation looks as follows. We have three
legal norms:
A1: 18_years(x) fi capacity(x)
A 2: mentally_ill(x) fi ¬capacity(x)
A3: married(x) fi ¬capacity(x)
and the following facts obtain:
18_years(john)
¬mentally_ill(john)
¬married(john)
From those premises only one argument can be built
18_years(x) fi capacity(x)
18_years(john)
capacity(john)
Since we have only one argument, its conclusion – capacity(john) – is
the logical consequence in the first situation.
LOGIC
61
In the second situation, besides the formulas representing our three
norms, we have also:
18_years(john)
married(john)
We can now construct two arguments leading to contradictory conclusions:
(A)
18_years(x) fi capacity(x)
18_years(john)
capacity(john)
and
(B)
married(x) fi ¬capacity(x)
married(john)
¬capacity(john)
In order to determine the logical consequence in the second situation we
must compare arguments (A) and (B), or, more precisely, “weigh” two
legal norms occurring in the arguments: Article 1 and Article 3. As the
second provision constitutes an exception to the first, it can be placed
“higher” in the ordering, and hence argument (B) prevails over argument
(A). Therefore, it is the conclusion of argument (B) – ¬capacity(john) –
that is the required logical conclusion in the second situation.
It turns out, then, that DL, which displays structural resemblance and
modularity, leads to relatively complicated application of legal norms
(determining the logical consequences in the given case). Classical formalizations are simpler in this regard. This advantage of classical calculi
diminishes, however, as soon as cases more difficult than the application
of Articles 1–3 are at stake. For instance, let us look at Riggs vs. Palmer.
In the classical formalization, after the exception resulting from the principle “No man shall profit from his own wrong” has been introduced, we
have the following, complex formula:
R: (᭙x) (᭙y) ((dies(x) ∧ grandfather(x, y) ∧
¬wrong(x)) → inherits(y, x))
We will not apply this norm in Riggs vs. Palmer, since one of the conjuncts
is not fulfilled, i.e., ¬wrong(x) does not obtain (for Elmer did wrong).
62
CHAPTER 2
In DL we have:
R: (dies(x) ∧ grandfather(x, y)) fi inherits(yx)
P: wrong(x) fi ¬profit(x)
In the analyzed case the following facts obtain:
dies(francis)
grandfather(francis, elmer)
wrong(elmer)
this enables us to construct two arguments:
(A)
(dies(x) ∧ grandfather(x, y)) fi inherits(yx)
dies(francis)
grandfather(francis, elmer)
inherits(elmer, francis)
and
(B)
wrong(x) fi ¬profit(x)
wrong(elmer)
¬profit(elmer)
and since benefiting from inheritance is profitable ((᭙x) (᭙y) (inherits(y, x) → profit(x))), the conclusions of both arguments contradict
one another. Comparing arguments (A) and (B) we “weigh” the norms
(dies(x) ∧ grandfather(x, y)) fi inherits(yx) and wrong(x) fi
¬profit(x). The New York Court of Appeals gave priority to the latter
norm and concluded that in Riggs vs. Palmer the logical conclusion of
argument (B) prevails.
Let us modify the case slightly and imagine that Elmer killed his
grandfather but did it unintentionally. He most certainly did wrong and,
according to the principle employed by the court, he should not benefit
from his act. The application of rule R as formalized in classical logic
leads to the conclusion that, in the modified circumstances, Elmer does
not benefit from Francis’ inheritance. It could be argued, however, that
the modified case is different from the original and that it is unjust to
deprive Elmer of his rights. Such reasoning can easily be represented in
LOGIC
63
DL. Here, in the process of “weighing” the norms of arguments (A) and
(B), priority would be given to the first of the norms. It is clear, then, that
the flexibility of the “complicated” application of norms in DL may have
profound practical consequences.
2.5.5
Two Remarks
At the end of our presentation of DL we would like to add two remarks.
First, the substitution of the idea of “transmission of truth” with the
idea of “transmission of justification” enables one to regard DL as a
logic that captures some pragmatic aspects of legal reasoning, and to
look for a pragmatic notion of logical consequence. Second, nonmonotonic systems may serve as a basis for questioning the thesis that the role
of logic is confined to the context of justification. The complicated procedure of applying norms in DL can be seen as an attempt to capture the
formal aspects of the context of discovery.
2.6
SUMMARY
Our analyses of the logical reconstruction of legal reasoning, although
not all-embracing, may serve as a basis for some conclusions regarding
the nature and limits of applying logical methods. First and foremost,
they show that there is no common agreement over what the logic of
legal discourse looks like. It should be added that the formalisms that we
presented are not complementary. These are, in most cases, formal systems that are incompatible. For instance, the proponents of defeasible
logics put forward arguments against classical logic, whilst the constructors of the deontic logic of action oppose the way obligation is defined
in SDL.
Secondly, every attempt to develop a logic of legal discourse faces two
kinds of problem. On the one hand, there are issues of general, philosophical nature; the Jørgensen Dilemma, considerations of various kinds
of obligation, and objections against labeling nonmonotonic systems
“logics” are cases in point. On the other hand, there are more specific
problems, such as the various paradoxes of deontic logic. What is important, however, is that these problems do not result in the abandonment
of attempts to construct a logic of legal discourse; on the contrary, they
only encourage new research in the field.
We would like to stress one more thing: the important role intuition
plays in constructing normative logics. It is intuition that stands behind
the feeling of “a paradox” in certain situations. This is not to say that
intuition decides everything, but its role should not be underestimated.
64
CHAPTER 2
There is one more characteristic feature of contemporary research on
normative logic: the way in which new systems develop to overcome
recurring paradoxes leads through more and more complex semantics.
This “semantic strategy” has recently been extended by the addition of a
pragmatic ingredient. We must stress that this feature of contemporary
normative logic – intuition plus semantics, with a bit of pragmatics – can
be found in almost any logical research carried out nowadays.
It is necessary yet to ask what the conclusions of our analyses of the
limits of applying logical methods should be. It should be observed that
the contemporary logic of legal discourse aims to “conquer” more than
classical logic did. It is appropriate to recall the question of whether the
role of logic should be confined to the context of justification, or
attempts to analyze logically such hard cases as Riggs vs. Palmer. This
shows that there is no such thing as issues that cannot be analyzed from
a logical point of view. Even hard cases have a logical dimension.
Naturally, it is not the case that logic establishes algorithms for solving
every legal case imaginable. However, with the expansion of logical
methods, it is impossible to identify any strict limits on the application of
formal tools. The only indication of such limits may be the fact that the
role of logic remains to point out when, on the basis of given premises,
we can accept some conclusion. However, this is also the aim – at least
prima facie – of analysis and argumentation and, one could even argue,
also of hermeneutics.
In concluding, we would like to mention those logics, which have not
been discussed above: the logic of induction and probability logic. Our
omission of those logics does not mean that they are unimportant for
modeling logically legal reasoning. They can serve well the reconstruction of some arguments regarding evidence. We have decided, however,
not to present them because there is nothing “peculiar” about their
application in legal discourse. In other words, these formalisms are not
connected with practical discourse in any special way.
Apart from those mentioned above, it is possible to find other kinds
of “logic” in literature: informal, discursive, dialectical, etc. We have
deliberately inserted quotation marks around “logic”, for the theories
in question have nothing to do with how logic is understood in this
chapter. We do not want to say that we regard those conceptions as useless. Their introduction would ruin, however, the coherence of our
presentation. Furthermore, they are based on ideas that resemble those
on which theories of legal argumentation, discussed in Chapter 4, are
based.
LOGIC
65
NOTES
1. Cf. A. Tarski, “O poje˛ciu wynikania logicznego” [On the Concept of Logical
Consequence], Przegla˛d Filozoficzny, vol. 39, 1936, pp. 58–68.
2. The logical form of ordinary language expressions is not usually “visible at first
sight”. Thus, in order to judge the logical validity of arguments carried out in ordinary language, they are usually “translated” (paraphrased) into the chosen logical
equation. As we will see, such paraphrasing is rarely universal or unproblematic.
3. Cf. J. Etchemedy, The Concept of Logical Consequence, Harvard University Press,
1990, p. 5 ff.
4. It seems obvious that one can reconstruct logically legal arguments (e.g., judicial reasoning) only from the point of view of justification; what “really happens” in the
judge’s head must be disregarded, whilst what is intersubjectively controllable is
taken into account.
5. These are not, of course, all the possible functors. In a two-valued logic there are
16 possible functors.
6. One can demonstrate this with the following example: if we assumed that norms have
the form of material implication, then all norms that had a false (or contradictory)
hypothesis would be true (valid).
7. W.V.O. Quine, Methods of Logic, 4th edition, Cambridge, Massachussets., 1982, p. 45 ff.
8. Cf. G. Priest, An Introduction to Non-classical Logic, Cambridge, 2001, p. 13.
9. Cf. K. Ajdukiewicz, “Okres warunkowy a implikacja materialna” [Conditionals and
Material Implications], Studia Logica, IV, 1956.
10. Cf. G. Restall, Introduction to Substructural Logics, London–New York, 2000.
11. In this argument we apply, of course, modus ponens, but for the sake of simplicity we
omit the step of universal instantiation, as we do also in the examples below.
12. The terminology used is due to J. Wróblewski. Cf. his Sa˛dowe stosowanie prawa
[Judicial Application of Law], 2nd edition, Warszawa, 1988.
13. A survey of modal logics can be found in: G.E. Hughes, M.J. Cresswell, A New
Introduction to Modal Logic, London–New York, 1996.
14. And Stig Kanger and Jaakko Hintikka.
15. That is how it looks in the propositional calculus. In modal predicate logic semantics
is, of course, more complex, but the main ideas are the same.
16. Aristotle, Etyka Nikomachejska [Nicomachean Ethics], 1147a, in Aristotle, Dziela
Wszystkie [Collected Works], vol. V, Wydawnictwo Naukowe PWN, Warszawa,
2000, p. 216.
17. Cf. J. Kalinowski, Logika norm [Logic of Norms], Daimonion, Lublin 1993,
pp. 48–63.
18. This shows that Pp is the so-called weak permission, which means that ¬p is not
obligatory, but does not guarantee that p is not obligatory, as is the case with strong
permissions.
19. Cf. R. Hilpinen, “Deontic Logic”, in L. Goble (ed.), The Blackwell Guide to
Philosophical Logic, Malden–Oxford, 2001, pp. 159–182.
20. Cf. J. Wolenski, Logiczne problemy wykladni prawa [Logical Problems of Legal
Interpretation], Zeszyty Naukowe UJ, Warszawa–Kraków, 1972.
21. See J. Carmo, A.J.I. Jones, “Deontic Logic and Contrary-to-Duties”, in D. Gabbay
(ed.), Handbook of Philosophical Logic, 2nd edition, vol. IV, Dordrecht, 2001,
pp. 287–366.
66
CHAPTER 2
22. Cf. G.H. von Wright, “Ought to be – Ought-to-do”, in E.G. Valdes, W. Krawietz,
G.H. von Wright and R. Zimmerling (ed.), Normative Systems in Legal and Moral
Theory – Festschrift for Carlos E. Alchourrón and Eugenio Bulygin, Berlin, 1997,
pp. 427–438 and J.W. Forrester, Being Good and Being Logical – Philosophical
Groundwork for a New Deontic Logic, New York, 1996.
23. For instance: Meinong, Hartmann and Chisholm, cf. J.F. Horty, Agency and Deontic
Logic, Oxford, 2001.
24. Cf. ibidem.
25. It is not Geach’s original example but its revised version proposed by J.F. Horty and
N. Belnap in “The deliberative stit: a study of action, omission, ability, and obligation”, Journal of Philosophical Logic, 24, 1995, pp. 583–644.
26. Cf. K. Segerberg, “Getting started: beginnings in the logic of action”, Studia Logica,
51, 1992.
27. From the mathematical point of view an action in the universe of possible worlds W
is therefore a two-argument relationship, i.e., a set of ordered pairs <u, w>, such that
u, w ∈ W.
28. Here, we are dealing, once again, with the weak permission.
29. Cf. J.F. Horty, op. cit.
30. Therefore, we have the following structure: <Tree, <, Agent, Choice>, where Tree is
a set of moments, < is the relation that orders the moments, Agent is the set of agents
and Choice is a function ascribing to every agent α at the moment m a subset of the
set Hm of all the histories “going through” m.
31. For the details, see J.F. Horty, op. cit.
32. Ibid., p. 55 ff.
33. Ibid., p. 59 ff.
34. We would not like to suggest that logics of action are developed only in order to solve
the “Fred and Ginger problem”. There are also other problems in which those logics
are developed. One can point out, for instance, the definition of other-than-standard
deontic operators, the analysis of mutual relations between those operators or the
problem of expressing the conflict between obligations, etc.
35. Cf. J. Jørgensen, “Imperatives and Logic”, Erkenntnis 7, 1938, pp. 288–296.
36. See for instance J. Woleński, Z zagadnień analitycznej filozofii prawa [Issues in the
Analytical Philosophy of Law], Zeszyty Naukowe UJ, Prace Prawnicze, Warszawa–
Kraków, 1980.
37. J. Kalinowski in Le problème de la vérité en morale et en droit, Lyon 1967 argues to
contrary.
38. Cf. the discussion in Ratio Juris, caused by R. Walter’s paper “Jørgensen’s Dilemma
and How to Face It”, Ratio Juris 9, pp. 168–171.
39. H.L.A. Hart, “Ascription of responsibility and rights”, in A. Flew (ed.), Logic and
Language, Blackwell, 1951, p. 152.
40. Such a thesis seems more justified in relation to common law systems than continental systems. However, as shown below, the idea of defeasibility can be useful for
analyzing certain aspects of legal reasoning as carried out within the continental
tradition.
41. On other problems that caused the development of nonmonotonic systems see J.F.
Horty, “Nonmonotonic Logic”, in L. Goble (ed.), Blackwell Guide to Philosophical
Logic, Malden–Oxford, 2001, pp. 336–361.
LOGIC
67
42. Beside the nonmonotonic logics, the so-called formal theories of belief revision have
also been used to solve the “Tweety problem”.
43. Cf. H. Prakken, G. Vreeswijk, “Logics for Defeasible Argumentation”, in D. Gabbay
and F. Guenthner (eds.), Handbook of Philosophical Logic, Kluwer Academic
Publishers, Dordrecht, 2002, 2nd edition, vol. 4, pp. 219–318.
44. The presentation will be based on the idea of a defeasible logic as developed by
H. Prakken in Logical Tools for Modelling Legal Argument, Dordrecht, 1997.
45. This premise could also be reconstructed with the use of the material implication →,
which naturally is still at our disposal. The resulting formalization would be easier,
as every argument that is based on → prevails over an argument based on fi (see
below).
46. As this presentation is elementary, we apply here a simplified definition of attack.
47. See H. Prakken, op. cit.
48. In order for this formalization to “work” we need additional information that
Article 2 is “higher” in the ordering than Article 1, with the result that, in the case of
a conflict between both articles, it is the argument based on Article 2 that prevails.
49. Here, one has yet to update the ordering between the defeasible implications.
50. One can presumably formalize those provisions in classical logic with the use of a
smaller number of formulas. However, even this alleged formalization would be
much more complex than the one in DL.
51. Cf. R. Dworkin, Taking Rights Seriously, Harvard University Press, 1977.
CHAPTER 3
ANALYSIS
3.1
INTRODUCTION
It is difficult to define analysis. The main reason for this is that there are
many procedures labeled “analysis”. Here, of course, we are concerned
only with a few of them; for instance, chemical analysis will not be considered. We shall be preoccupied only with those analytical tools which
can be used in reasoning, interpretation or argumentation.
In what follows three basic meanings of the notion of analysis will be
identified and a short history presented of each. Then two methods of
analysis, which play an important role in the philosophy of law will be
described. These methods are linguistic analysis and economic analysis.
3.1.1
The Notion of Analysis
Historically speaking, a plethora of definitions of analysis have been
offered. It is convenient to start with three paradigmatic examples. The
Ancient Greek mathematician, Pappus, says:
For in analysis we suppose that which is sought to be already done, and we inquire from
what it results, and again what is the antecedent of the latter, until we on our backward
way light upon something already known and being first in order. And we call such a
method analysis, as being a solution backwards (anapalin lysin).1
Descartes, in turn, in the 13th of his work “Regulae ad directionem
ingenii” remarks:
If we are to understand a problem perfectly, we must free it from any superfluous conception, reduce it to the simplest terms, and by a process of enumeration split it up into
the smallest possible parts.2
Several hundred years later B. Russell put forward yet another definition
of analysis:
We start [the analysis] from a body of common knowledge, which constitutes our data.
On examination, the data are found to be complex, rather vague, and largely interdependent logically. By analysis we reduce them to propositions which are as nearly as possible simple and precise, and we arrange them in deductive chains, in which a certain
number of initial propositions form a logical guarantee for all the rest.3
69
70
CHAPTER 3
The quoted definitions seem to describe distinct, albeit not entirely
different, procedures. In the first definition analysis is a search for logical reasons (let us label it analysis1). In the second it is “decomposition”
(analysis2). Finally, in the third definition, analysis is a translation into a
certain language, which meets specific conditions, like simplicity or precision (analysis3). Of these three conceptions4 – analysis as a search for
logical reasons, analysis as decomposition and analysis as translation – it
is analysis2 that seems best to correspond with our intuitions. In the
dictionaries of various European languages “analysis” is usually understood as a method that leads from complex concepts to more simple
ones. Analysis – thus defined – is usually contrasted with synthesis,
which is a process of building complex entities from simple elements.
Let us note that analysis2 differs from the two other conceptions as it
has as its subject concepts – which are to be decomposed into simpler
concepts. Analysis1 and analysis3, on the other hand, concern sentences or
propositions. Only propositions can have logical reasons and only sentences can be translated into other sentences (sentences can also be
decomposed but then they are treated as syntactic phenomena and not
as bearers of meaning). It has to be noted that this difference is not as
important as it may seem at first sight. One of the members of the Vienna
Circle and a great representative of analytic philosophy, R. Carnap,
defined logical analysis in the following way: “The logical analysis of a
particular expression consists in the setting-up of a linguistic system and
the placing of that expression in this system”.5 The expression in question
can, of course, refer to a specific concept. Therefore, there may exist very
close links between analysis2 and analysis3.
3.1.2
History of the Concept
Ancient thinkers applied different methods of analysis. It is of special
interest to observe the procedures of searching for proper definitions.
This method is usually tied with the name of Socrates and is illustrated
well in many of Plato’s dialogues. The Socratic method is an example of
analysis2. Analysis1, in turn, can be found in those fragments of dialogues where Plato puts forward a hypothesis and seeks reasons for its
acceptance. Analysis1 is also carried out in Aristotle’s “Analytics”, the
work that founded logical research. It is interesting to note that, in antiquity, only analysis1 was theoretically described (especially in the already
quoted work of Pappus).
In the middle ages, especially after the “renaissance” of the twelfth century, analytical methods were widely used, as witnessed in the achievements
ANALYSIS
71
of the scholasticism. One can easily agree with J.M. Bochenski,6 that
analysis of a comparable level of precision recurred not earlier than the
twentieth century. The middle ages, by contrast, did not offer any new,
fully developed methods of analysis (with the exception of some theories
of John Buridan).
The works of Descartes constituted the next important step in the
methodological discussion of analysis. In “Discourse on method” one
finds a passage, which distinguishes between four principles of reasoning
and describes one of them as follows:
Divide each of the difficulties I examined into as many parts as possible and as may be
required in order to resolve them better.7
There are few doubts what Descartes has in mind here: analysis as
decomposition (analysis2). This conception of analysis was popularized
in the famous “Logic of Port Royal”, a book written in Cartesian spirit
by Arnauld and Nicole. The work, published in 1662, was used as the
basic textbook on logic as late as the nineteenth century, influencing the
common acceptance of analysis2. The methods of analysis2 can be found
in the works of the greatest philosophers of modern times: Leibniz, the
British empiricists, and Kant.
The third of the meanings given to analysis – analysis as translation –
started to enjoy the status of an important methodological concept
after the so-called linguistic turn in philosophy, which occurred with the
publication of the works of Frege, Moore, Russell and Whitehead. It is
convenient here to devote a few words to the reasons why the “linguistic turn” caused a change in how the notion of analysis was understood.
Speaking somewhat boldly, philosophers before the linguistic turn, and
at least since Descartes, were interested in what constitutes thinking or
reasoning. “The inner discourse” or as the author of the “The discourse
on method” would put it, “the chain of ideas in the human mind”, constituted the subject of the philosophers’ interest. This interest was
supplemented with a peculiar understanding of reasoning, which was
based on a metaphor of seeing: for Descartes “to think” means “to see
ideas with the mind’s eyes”. It is not surprising then that, in such a
framework, what interests philosophers are ideas or concepts and the
method of analysis leads to decomposing the complex images into
simpler ones. After the linguistic turn the situation changes: sentences
now occupy the central place that had been reserved for concepts (even
though concepts play an important role in some kinds of analytic
philosophy).8
72
CHAPTER 3
This is not to say, however, that the idea of analysis as decomposition
is completely unknown to philosophers of the linguistic turn. B. Russell
wrote:
Analysis may be defined as the discovery of the constituents and the manner of combination of a given complex. The complex is to be one with which we are acquainted; the
analysis is complete when we become acquainted with all the constituents and with their
manner of combination, and know that there are no more constituents and that that is
their manner of combination (. . .).9
This passage suggests that it is not always possible to differentiate
between the three kinds of analysis outlined above. One can even say that
there are not three kinds of analysis, but three aspects of the same
process and that the different aspects play more or less important roles
depending on the aim of the analysis, the philosophical assumptions
accepted, etc.10 However, such a solution – simple as it may be – makes a
precise definition of analysis impossible.
To sum up: analysis cannot be easily defined. That said, one can identify three models of analysis and any given method is likely to resemble
more or less one of those models. For example, mathematical analysis is
first and foremost connected with the search for logical reasons. Anyway,
the three presented models can serve as a basis for characterizing any
applied (or postulated) methods of analysis.
3.1.3
Analytic Philosophy
In the twentieth century there emerged a philosophical movement that is
usually referred to as “analytic philosophy”. It would be inappropriate to
characterize analytic philosophy as a school of thought that uses analysis as its method; there are numerous other philosophical traditions making use of analysis. The problem is that because of the diversity of
schools labeled “analytic” it is difficult to establish the necessary or sufficient conditions for referring to someone as an “analytic philosopher”.
J.M. Bocheński maintains that there are four keywords pertinent to
analytic philosophy: analysis, language, logic and objectivism.11
Bocheński understands analysis as opposed to philosophical synthesis;
the task of a philosopher is not to construct all-embracing systems but
to solve concrete problems. The keyword “language” refers to accepting
language as the basic medium of philosophizing. “Logic”, in turn,
underlines analytic philosophers’ trust in formal tools. Finally, “objectivism” is opposed to any kind of subjectivism.
The above mentioned principles are realized to varying degrees in
different schools of analytic philosophy. For instance, members and
ANALYSIS
73
followers of the Vienna Circle put special emphasis on applying logical
methods. The Oxford School of Ordinary Language, by contrast, dismissed logic as useless and concentrated on painstaking analysis of how
words are actually used in ordinary language. Bocheński’s proposal is
not, therefore, a definition of analytic philosophy but only indicates the
principles important to analytic philosophers.
Analytic philosophers are usually divided into two “wings” (camps).
The first wing is occupied by the reconstructionists, whose aim is to
reform ordinary language using logical tools. The second wing – the
descriptionists – aims only to describe how ordinary language functions
and uses methods of an informal character. The methods of reconstruction are often referred to as “hard”, whilst descriptive techniques are
called “soft”. Therefore, we have “hard” (formal) analytic philosophers
on the one hand, and “soft” (anti-formal) on the other.
Today the above mentioned division is questionable. But in the middle
of the twentieth century there were good reasons to justify it. The “formal approach” was represented then by members and followers of the
Vienna Circle (who, after the Second World War stayed in the United
States or in the United Kingdom). On the other hand, the “anti-formal”
methods were applied by Wittgenstein in Cambridge and by the Oxford
School of Ordinary Language. The differences between the two
approaches were clearly visible then, as was the animosity between the
representatives of both wings of analytic philosophy.
G.E. Moore is considered the “founding father” of analytic philosophy.
His Principia Ethica (1903) marks symbolically the beginning of a philosophy that – being in opposition to neo-Hegelianism – concentrated on
conceptual analysis based on common sense. Moore can easily be classed
as a representative of the anti-formal wing of analytic philosophy. The
same may be said of the “second” philosophy of L. Wittgenstein or
the Oxford School of Ordinary Language.
Around the time Moore “founded” anti-formal analytic philosophy,
the “formal” wing began to emerge. German logician G. Frege is usually
regarded as the first representative of this way of philosophizing. In
addition to Frege’s work, as a milestone in the history of logic – and of
“hard” analysis – one should also mention Russell’s and Whitehead’s
Principia Mathematica, published in 1910–1913. In the spirit of Principia
L. Wittgenstein wrote his first major work, Tractatus LogicoPhilosophicus. The idea of applying logical methods in philosophy was
radicalized by the Vienna Circle, founded by M. Schlick in 1920s. Among
members of the Circle one should mention O. Neurath, R. Carnap and
F. Waissmann. Less radical, but still logically oriented, was another
74
CHAPTER 3
distinguished philosophical school: the Lvov-Warsaw School, with
its founder K. Twardowski and famous members: J. Lukasiewicz,
S. Leśniewski, K. Ajdukiewicz, A. Tarski, T. Kotarbiński and others.
More contemporary methods of analysis, represented mainly by
American and British philosophers, cannot easily be classified as formal
or informal, although in some cases such a classification is to a certain
extent justified. For instance, W.V.O. Quine can be labeled a representative of “hard” analysis; J. Hintikka likewise. G. Evans, on the other hand,
used rather “soft” methods. However, some first rate philosophers, such
as H. Putnam, M. Dummett and D. Davidson, do not easily qualify as
representatives of either of the two groups.
Of the many approaches employed by analytic philosophers we examine below only linguistic analysis, which is characteristic of the Oxford
School of Ordinary Language. There are several reasons for this. First,
the methodological richness of analytic philosophy makes it impossible
to present everything. Second, linguistic analysis is a method that is relatively well defined. Third, it has played an important role in legal theory
and the philosophy of law. Finally, it is a model example of anti-formal
analysis.
In Chapter 2 we also wrote about hard analysis; the chapter is devoted
to presenting different attempts at the logical reconstruction of legal reasoning. The logical systems presented there are tools of “hard analysis”.
This formal analysis consists in translating (paraphrasing) natural language expressions into certain logical systems and in drawing conclusions
from the obtained formalizations. Although in Chapter 2 we did not pay
proper attention to the actual process of paraphrasing, in all those places
where legal arguments are logically reconstructed a paraphrase has been
employed. The development of logical methods presented in Chapter 2 is
also an illustration of another kind of analysis. The logical systems we
described constitute the analysis2-definition of a legal norm, and some of
them – of the concepts of obligation, permission, etc.
The decision to devote separate chapters to logic and analysis is not
designed to exaggerate the differences between “soft” and “hard” analysis. What we had in mind was the coherence of the presentation. An historical approach to the logical reconstruction of legal reasoning enables
one to show the development of formal methods and identify the key
logical problems of legal reasoning. From this perspective mixing up
“hard” with “soft” methods would make less clear the picture of logical
research in legal theory.
We have to add one terminological remark. Below we will describe linguistic analysis which – in its purest form – was applied by the members of
ANALYSIS
75
the Oxford School of Ordinary Language. As already indicated, it is a
model example of “soft”, anti-formal analysis. Instead of “linguistic analysis” one can also say: “linguistic philosophy”. Linguistic analysis (philosophy) has to be distinguished from the philosophy of language. Linguistic
philosophy is identified by a certain method, whilst philosophy of language
is identified by a certain subject. Philosophy of language can therefore be
developed with the use of linguistic analysis but also of formal methods,
phenomenology, etc. Linguistic philosophy, in turn, is not confined to developing a theory of language; it may concern any philosophical problem.
3.2
LINGUISTIC ANALYSIS
We would like to look now at the basic assumptions of linguistic analysis, in particular the special role it attaches to ordinary language. Then
several important legal-theoretic conceptions that have grown out of
linguistic analysis will be presented.
3.2.1
History and Basic Assumptions of Linguistic Analysis
Philosophical analyses which are labeled “soft” can be found in the writings of one of the “founding fathers” of analytic philosophy, G.E. Moore.
In such works as Principia Ethica (1903), Moore developed a common
sense philosophy, which made extensive use of analysis of the ordinary
meaning of words. Similar methods were applied by other philosophers
belonging to the Cambridge School of Analysis such as J. Wisdom,
M. Black and S. Stebbing.
In reference to linguistic analysis one cannot fail to mention the “second” philosophy of L. Wittgenstein. Wittgenstein claimed to have solved
all the major problems of philosophy in Tractatus logico-philosophicus
and, after publishing it in the early 1920s, decided not to follow any “academic career”. He returned to Cambridge and to philosophical thinking
at the beginning of the 1930s. He developed then a new conception,
which – in opposition to the one employed in Tractatus – abandoned
logic in favor of analyzing ordinary language in its diverse forms.
Linguistic analysis in its classic form was developed in Oxford by J.L.
Austin and his followers. Austin, the author of How to Do Things with
Words (1962), put ordinary language at the center of philosophical attention, maintaining that analysis of the subtle ways in which ordinary language functions is the proper way to approach philosophical problems.
Among other representatives of the Oxford School one should mention
P.F. Strawson and G. Ryle. One should also add to this list J. Searle,
although he did not regard himself as a linguistic philosopher.
76
CHAPTER 3
Strawson devoted much attention to linguistic analysis as a method
and showed that it can be applied to all traditional philosophical disciplines, including metaphysics (Individuals, 1959). Ryle was a highly influential figure in British philosophy from the 1930s. His main work is The
Concept of Mind (1949), in which Cartesian dualism is attacked. Searle,
before turning to the problems of the philosophy of the mind, developed
further in Speech Acts (1969) the Austinian conception of illocutions,
which we describe below.
Among the legal philosophers making use of linguistic analysis one
should mention H.L.A. Hart, and some elements of this kind of analysis can also be traced in the works of R. Dworkin or J. Raz. It is not
improper to mention K. Opalek, who in his monograph “Z teorii dyrektyw i norm” (1974) applied both “hard” and “soft” methods to analyze
such concepts as “directive”, “norm” or “imperative”.
It is difficult to state precisely what linguistic analysis consists of. One
could say that it is the analysis of concepts that are used in ordinary language; this, however, is neither precise nor adequate. We will begin our
attempt to describe linguistic analysis by looking at what Strawson has
to say on the matter.
The most characteristic feature of linguistic analysis is the special role
of ordinary language. Strawson says that concepts which are basic from
the philosophical point of view – if there is such a thing – are to be
looked for in non-technical, ordinary language and not in specialized languages.12 Those specialized languages, which include the language of sciences, are secondary relative to ordinary language, for getting acquainted
with the theoretical concepts of various sciences assumes earlier familiarity with the pre-theoretic notions of everyday practice. Strawson says that
ordinary language is based on the so-called conceptual scheme, a general
structure of our every day and scientific thinking13; the aim of philosophizing is to discover and analyze this structure. This is, naturally, a very
strong assumption, although some reasons backing it can be formulated.
For instance, one can observe with Austin that
our common stock of words embodies all the distinctions men have found worth
drawing, and the connections they have found worth marking, in the lifetimes of many
generations: these surely are likely to be more numerous, more sound, since they have
stood up to the long test of the survival of the fittest, and more subtle, at least in all
ordinary and reasonably practical matters, than any that you or I are likely to think up
in our arm-chairs of an afternoon – the most favored alternative method.14
What is striking in this passage is the phrase “the survival of the fittest”,
which allows us to call Austin’s argument that ordinary language has
a special status – the “evolutionary argument”: if ordinary language
ANALYSIS
77
results from a very long evolution, one can conclude that it “fits” the
world well. In other words: if it is an efficient tool, it surely encodes some
knowledge about the world.
The evolutionary argument – as persuasive as it is – is not sufficient
for Austin to claim that ordinary language is the “last word” in philosophical thinking. As the author of How to do Things with Words
observes himself:
[The ordinary language] embodies, indeed, something better than the metaphysics of the
Stone Age, namely, as was said, the inherited experience and acumen of many generations of men. (. . .) [But] certainly, ordinary language is not the last word: in principle it
can everywhere be supplemented and improved upon and superseded. Only remember,
it is the first word.15
What Austin says does not only weaken the conclusions of the evolutionary argument. In this quoted passage one can find much more: a
certain idea of linguistic analysis, which is an alternative to that of
Strawson. According to Strawson, the analysis of the conceptual scheme,
which is hidden beneath the ordinary language is the final aim of
philosophy. Austin, on the other hand, suggests analyzing the way ordinary language functions; but the analysis in question is designed only
as a point of departure for philosophy and not as its final aim. Only
through an analysis of language can one prepare the ground for serious
philosophical undertakings. From what has been said we can label
Strawson’s idea of analysis the maximalist approach, and Austin’s the
minimalist.
Below we will see examples of both minimalist and maximalist approaches. Hart’s analyses presented in Section 3.2.2 are maximalist rather than
minimalist. Speech acts theory, presented in Section 3.2.3, is an example of
the minimalist program.
The most important objection to the program of linguistic analysis
can be formulated as follows: philosophy which confines itself to describing how concepts function in ordinary language is a caricature of what it
ought to be, namely the serious pursuit of important problems. Antiformal philosophers reply to this with the following metaphor, which
may be labeled “grammatical”. Most of us are able to use our language
without making many serious mistakes. This does not mean, however,
that we are aware of all the rules of grammar. Like a linguist, who discovers the rules of grammar, a philosopher analyzes rules that govern our
experience, but of which we are not (and do not have to be) aware. Such
analysis cannot be termed a mere “caricature”.
Another problem is connected to a specific feature of linguistic analysis that is very strongly stressed by Strawson. The aim of linguistic
78
CHAPTER 3
analysis is not to discover “elementary factors”. It is rather, speaking
metaphorically, an attempt to draw a map of the conceptual scheme. The
author of Individuals appeals to us to abandon the idea that analysis has
to lead towards simplicity. He suggests, instead, imagining a model of
a net, a system of interconnected units such that every unit can be
understood properly from the philosophical perspective only when its
interconnections with others are captured.16
It is plain to see that this philosophical method does not fit any of the
three models of analysis we identified above. It has nothing to do with
searching for logical reasons, or translating from a complex language into
a simpler one. Moreover, Strawson expressly states that it does not assist
in achieving greater simplicity. We believe, however, that one should not
take for granted what Strawson says. Is not it true that the Strawsonian
conceptual scheme is something basic? It is the task of a philosopher to
discover the scheme, but also to relate philosophical problems to it. For
instance: Strawson addresses the problem of Cartesian dualism – a serious philosophical issue by any standards – by determining that, in the
conceptual scheme, there exists a concept of a person which differs in
some important respects from the concept of a thing. The problem of
dualism can now be analyzed against the background of this reconstructed
fragment of the conceptual scheme of ordinary language. Therefore, the
aim of analysis according to Strawson is not only to discover (describe)
the conceptual scheme, but also to deal with philosophical problems by
relating them to that scheme (this will be illustrated below, especially in
Section 3.2.2). One can say therefore that linguistic analysis in its maximalist form is an example of analysis3. We will come back to this controversial issue later.
The proponents of linguistic analysis are aware of the many objections
raised against their method, and they have formulated various responses.
In order to assess more accurately the benefits and limits of applying linguistic analysis to law, it is necessary to look more closely at how the
method works in practice.
3.2.2
Legal Conceptual Scheme
Let us look more closely at how H.L.A. Hart used the methods of linguistic analysis in order to elucidate “the concept of law”. Hart’s undertaking can be seen from the maximalist perspective as an attempt to
show how the concept of law functions in our conceptual scheme.
Consequently, one will not find in Hart’s analysis a definition of law that
characterizes it by reducing it to more elementary elements. Moreover, in
The Concept of Law Hart adapts an anti-definitional approach. This
ANALYSIS
79
approach is connected to the diagnosis of ordinary concepts of language
that was popular in J.L. Austin’s school.
Such concepts are necessarily vague. Accordingly, every concept has a
core of determinate meaning and a penumbra of indeterminacy; in other
words, there are objects that are certainly covered by the concept’s denotation, but there are also many borderline cases – objects which cannot
easily be placed inside or outside the concept’s denotation. In the case of
such concepts, defining – understood traditionally, as determining the
set of necessary and sufficient conditions for predicating the given concept of the given object – is pointless for it leads to the elimination of
borderline cases, and therefore creates a false picture of how the concept
functions in ordinary language.
This anti-definitional approach does not, however, mean that one cannot say anything about ordinary language concepts. On the contrary: the
principle of the primacy of ordinary language enables the subtle analysis of various concepts and interrelations between concepts. As a result,
however, we never arrive at a simple definition. As Hart puts it himself
in his opus magnum: “(. . .) this book is offered as an elucidation of the
concept of law rather than a definition of ‘law’”.17
Hart commences his discussion by criticizing the definition of law proposed by a nineteenth century English philosopher, J. Austin, the author
of The Province of Jurisprudence Determined18. Austin’s definition is
usually reduced to the following slogan: “law is an order backed by a
threat of sanction”. In order to illustrate how Hart attacks Austin’s definition by means of linguistic analysis, let us quote a longer passage from
The Concept of Law. Considering in which situations the word “imperative” is appropriate, the English philosopher writes:
[They] can be illustrated by the case of a gunman who says to the bank clerk: ‘Hand over
the money or I will shoot!’ Its distinctive feature which leads us to speak of the gunman
ordering not merely asking still less pleading with the clerk to hand over the money, is that,
to secure compliance with his expressed wishes, the speaker threatens to do something
which a normal man would regard harmful or unpleasant, and renders keeping the
money a substantially less eligible course of conduct for the clerk. If the gunman succeeds, we would describe him as having coerced the clerk, and the clerk as in that sense
being in the gunman’s power. Many nice linguistic questions may arise over such cases:
we might properly say that the gunman ordered the clerk to hand over the money and the
clerk obeyed, but it would be somewhat misleading to say that the gunman gave an order
to the clerk to hand it over, since this rather military-sounding phrase suggests some right
or authority to give orders not present in our case. It would, however, be quite natural to
say that the gunman gave an order to his henchman to guard the door.19
This passage typifies the way in which “anti-formal” philosophers argue.
Hart presents a simple counter-example to challenge Austin’s definition.
80
CHAPTER 3
The sentence uttered by the gunman – “Hand over the money or I will
shoot” – can be called an order backed by a threat of sanction, but it is
not “law”. In this way Hart shows that Austin’s definition over-simplifies
how the concept of law functions in ordinary language.
By developing many similar distinctions, Hart continues his analysis
of Austin’s definition. He shows that even if we changed the definition,
saying that the orders in question must be general and commonly
obeyed, we would still be far from determining how the concept of law
operates. This can easily be observed when norms which confer powers,
or determine proper procedures, e.g., for the writing of a will, are taken
into account. These are legal norms, but they are not general and commonly obeyed orders backed by the threat of sanction.
This analysis of how the word “law” is used leads Hart to claim that law
consists of a union between primary and secondary rules. Primary rules
are rules that confer powers or state obligations. Secondary rules are
about the primary rules: they determine how to introduce, change and
interpret the primary rules. Among the secondary rules there is also the
rule of recognition, which enables us to say which legal rules are valid.
This hypothesis of the concept of law is then analyzed by Hart in various contexts; the problems of the sovereign, justice, morality, and international law are addressed with the goal of elucidating the concept of law
in mind. The last of these issues is of special interest for, reflecting on the
question whether international law can be labeled “law”, Hart applies a
method sometimes called “argumentation from paradigms”.20 The
method involves determining whether a given phenomenon can be called
“X” by showing either that the phenomenon fits well a paradigm, a
“model” X, or that it differs from it and in what respects. Applying this
strategy, Hart asks whether international law is an exemplary incidence of
the union of primary and secondary rules. His answer is, for various reasons, negative. Having analyzed the problems of international law, such as
the lack of sanctions and lack of a uniform rule of recognition, Hart concludes that there are some analogies between international law and state
law, which is a paradigm of the union of primary and secondary legal
rules. International law is, therefore, a phenomenon which displays some,
but not all, of the features of a “model” law.21 Let us note that Hart’s
method does not demand a conclusive answer to the question of whether
international law is law. It is sufficient to say that it is a borderline case.
This, of course, stems from the fact that Hart does not aim to give a precise definition of law, but only to demonstrate how the concept of law
functions in ordinary language. This approach is different from the one
proposed by Austin in The Province of Jurisprudence Determined. Having
ANALYSIS
81
defined the “positive law”, Austin does not hesitate to claim that international law is not positive law for it is not enacted by a sovereign.22
The above example can lead to the false conclusion that linguistic
analysis cannot help to deal with “authentic” problems and can only
serve to criticize other approaches. Leaving aside the fact that – as
Wittgenstein would put it – many authentic problems are just pseudoproblems resulting from our lack of understanding of how language
works, linguistic philosophy is filled with attempts to solve old philosophical problems. Such attempts can also be traced in Hart’s work. For
instance, the Oxford philosopher considers the concept of obligation,
criticizing the so-called predictive theory thereof. In short: being obliged
means that there is a sufficiently high probability of sanction if the obligation is not fulfilled. Hart puts forward two arguments against this sociological conception. First, he observes that the validity of a rule should
not be confused with its applicability, as it is only applied because it is
considered valid.
Second, he says in the “linguistic” spirit:
If it were true that the statement that a person had an obligation meant that he was likely
to suffer in the event of disobedience, it would be a contradiction to say that he had an
obligation, e.g., to report for military service but that, owing to the fact that he had
escaped jurisdiction, or had successfully bribed the police or the court, there was not the
slightest chance of his being caught or made to suffer. In fact, there is no contradiction
is saying this, and such statements are often made and understood.23
It is characteristic that, in addition to criticizing the predictive interpretation of obligations, Hart proposes his own solution to the problem of
the relationship between the concept of obligation and statements concerning the efficiency of legal system. He claims that those statements
can be regarded as presuppositions of statements, which refer to the
validity of legal norms. In order to grasp this conception it is necessary
to say a few words about the concept of presuppositions.24
The phenomenon called “presupposition” was mentioned initially by
Frege but it was given the status of a full-blown philosophical problem in
Strawson’s paper “On Referring”.25 Strawson’s idea was a reply to
Russell’s solution of a certain logical puzzle. The puzzle was connected
with the sentence: “The present king of France is bald”. Because of the
fact that France is a republic, and there exists no king of France, the
quoted sentence should be deemed false. For the same reason, however,
its negation: “It is not the case that the present king of France is bald”
also seems false. And here we encounter a real problem, for accepting that
both sentences are false we break the basic law of logic, i.e., the principle
of the excluded middle.
82
CHAPTER 3
Russell solved the problem by saying that the logical structure of the
two sentences is significantly different from what their grammatical form
suggests. The sentence “The present king of France is bald” is logically a
conjunction of three sentences: “There exists x, and that x is king of
France” and “x is bald” and “there is only one x”. This sentence is false
because the first of the conjuncts is false (“There exists such x, that x is
king of France”). Analogously one can deal with the sentence “It is not
the case that the present king of France is bald”.
Strawson criticized Russell’s solution mainly because of the artificial
distinction between grammatical and logical structures of a sentence.26
He suggested instead that the concept of presupposition be used. We say
that a sentence p is a presupposition of a sentence q (or, in other words,
that q presupposes p), if the possibility of ascribing any logical value to
q and ¬q depends on p being true. We are able to say that q
(or ¬q) is true or false if and only if the presupposition of q – the sentence p – is true. The analyzed sentences “The present king of France is
bald” and “It is not the case that the present king of France is bald” can
be established as truth or false under the condition that their presupposition – the sentence “There exists the present king of France” – is true.
If the presupposition is false – as is the case in our example – the analyzed sentences cannot be designated as either true or false. Usually it is
maintained that sentences which cannot be given the value of truth or
falsehood are senseless.27
Let us return to the problem of the relationship between statements
concerning obligations and statements concerning the efficiency of law.
Hart suggests that the latter are presuppositions of the former. He does
not use here, presumably, the notion of presupposition in its exact
meaning. Rather, he intends to say that a full meaning of statements concerning obligations can be grasped only under the assumption that statements concerning law’s efficiency are true.28
We would like to add one observation here. It seems that the entire
method of linguistic analysis, especially the maximalist approach, is a
search for “loosely” understood presuppositions. In trying to discover
the elements of the conceptual scheme one asks what concepts are presupposed by the given linguistic phenomena.
3.2.3
Speech Acts Theory
One of the most important achievements of linguistic philosophy is the
analysis of various aspects of language. It explores not only how certain
concepts, like “law”, function in language, but also what these “ways of
functioning” are. An important thread in this research is the analysis
ANALYSIS
83
of practical discourse (both ethical and legal) and the demonstration of
the role rules play within it.29 The concept of a rule is one of the central
subjects of Wittgenstein’s Philosophical Investigations. Other authors analyzing this subject are, inter alia, A. Ross,30 K. Opalek31 or R. Dworkin.32
Another important aspect of linguistic philosophy is the theory of
speech acts. Let us look more closely at this. The conception of speech
acts was developed initially by J.L. Austin and later developed further by
J. Searle and others.
Austin developed first the theory of performatives. He analyzed a class
of linguistic expressions that had been disregarded by earlier philosophers. Some favorite Austinian examples of expressions belonging to the
class are, for instance: “I name this ship ‘the Queen Elizabeth’” (uttered
in circumstances when a bottle of Champagne is smashed against the
helm of a ship); “I give and bequeath this watch to my brother” (in a
will); “I do” (as an answer to the question “Do you take this woman as
your wife?”). Austin observes that such expressions play a different role
than the sentences which we use to describe the world: they are uttered
with the aim of changing the world. The utterance (in specific circumstances) of the first expression causes a specific ship to be called “Queen
Elizabeth” from that moment on; of the second – that the will as regards
a specific watch has been expressed; and of the third (assuming the same
answer from the woman) – that a marriage has been concluded.
Those and similar expressions Austin labels performatives and characterizes them in the following way. They:
(A) do not ‘describe’ or ‘report’ or constitute anything at all, are not ‘true or false’; and
(B) the uttering of the sentence is, or is a part of, the doing of an action, which again
would not normally be described as, or as ‘just’, saying something.33
Performative expressions cannot be ascribed truth or falsehood, they can
be however described as successful (fortunate, happy) or unsuccessful
(unfortunate, unhappy). For instance, if the will is not written in the
legally correct way, then the sentence “I give and bequeath this watch to
my brother” will be unsuccessful – in other words, it will be an unfortunate performative. Similarly, the “I do” of the prospective husband
uttered in the presence of a person who has no power to conduct a marriage ceremony would be an unhappy performative.
What is characteristic of the given examples is that there exists a certain conventional procedure which determines the necessary conditions
of a successful action (and therefore of the happiness of a performative).
In the case of the will and the marriage, the procedure is regulated by
legal norms; in the case of the naming of a ship, the rules of custom set
84
CHAPTER 3
the procedure. Those performatives which do not follow the procedure
are unfortunate.
However, as Austin observes, not every performative is tied to a conventional procedure. If one says: “I promise to come tomorrow”, the
very uttering of those words – independent of any external circumstances – results in the making of a promise, and hence changes the
world; the expression “I promise to come tomorrow” is therefore a performative expression, for it fulfills the Austinian specification of such
expressions. The only thing that can “go wrong” in this situation, and
result in the performative being unfortunate, is my intention: if I uttered
the words “I promise to come tomorrow” with no intention of fulfilling
the promise, one would be in a position to say that it is an unfortunate
performative.
Austin pays a lot of attention to situations in which performatives turn
out to be unfortunate (so-called infelicities). These considerations are of
special interest for us. Austin developed a typology of infelicities. It has
to be stressed, first, that we are addressing a typology and not a classification (logical division). This shows that Austin does not want to simplify the described phenomena and intends to present mechanisms
governing the use of language in the full richness of the complicated and
sometimes vague connections between them. It may seem surprising and
unnecessary to dedicate so much effort to a typology of infelicities. In
fact, however, it is sufficient to observe that a description of “what can
go wrong” is a modo negativo analysis of the phenomena that obey the
rules, i.e., the typology of infelicities helps us to draw the boundaries of
the set of happy performatives.34
The way Austin develops his typology requires our special attention,
for the method characteristic of linguistic philosophy manifests itself
clearly in this process. We have already observed this in referring to
Austin’s desire to pay the richness of ordinary language its due. Another
typical analytic maneuver Austin applies in his work is to present a long
list of examples in favor of, or against, the proposed hypothesis. The use
of simple linguistic intuitions is visible even in the way Austin labels different kinds of infelicities. Those cases in which a performative is
unhappy because a certain procedure has not been followed, Austin calls
misfires; the situations, in which “something went wrong” because of the
attitude of the speaker, are termed abuses. This distinction is commented
upon as follows:
When the utterance is a misfire, the procedure which we purport to invoke is disallowed
or is botched: and our act (marrying etc.) is void or without effect etc. We speak of our
act as a purported act, or perhaps an attempt – or we use such an expression as ‘went
ANALYSIS
85
through a form of marriage’ by contrast with ‘married’. On the other hand, in the (. . .)
cases [of abuses] we speak of our infelicitous act as ‘professed’ or ‘hollow’ rather than
‘purported’ or ‘empty’, and as not implemented, or not consummated, rather then as
void or without effect. But let me hasten to add that these distinctions are not hard and
fast (. . .).35
What strike us in the quoted passage are not only the numerous linguistic examples, but also Austin’s abstention from introducing exceptionless
distinctions.
Among the misfires Austin distinguishes errors regarding the invocation of a procedure and the execution of a procedure. The first class
includes situations in which there exists no proper convention and situations in which an existing procedure is wrongly applied. In the second
class there are flaws (the procedure is partially wrongly executed) and
hitches (the procedure has not been completed). Austin distinguishes also
between different kinds of abuse, identifying insincerities (where I give a
promise but do not intend to fulfill it) and infractions or breaches (where
I promise something but do not fulfill it later). Naturally, for each category Austin presents a long list of examples. For instance, in considering
insincerities he distinguishes between “not having the requisite feelings”
(e.g., “I congratulate you” said when I do not feel at all pleased), not having the requisite thoughts (e.g., “I find you not guilty” when I do believe
you are guilty), or “not having the requisite intentions” (e.g., “I promise”
when I do not intend to do what I promise). Austin comments in detail on
the distinctions, pointing out various features of insincerities and
observes: “the distinctions are so loose that the cases are not necessarily
easily distinguishable: and anyway, of course, the cases can be combined
and usually are combined. For example, if I say “I congratulate you!”,
must there really be a feeling, or rather a thought, that you have done or
deserved well?36
As in the case of other above mentioned distinctions, Austin points
out the limitations of the introduced typologies, underlining their vagueness, etc. This arises, once again, from the philosopher’s conviction that
his aim is not to offer a simplified sketch of language, but to record all
the nuances thereof. There are numerous other examples of vague or
uncertain conceptions in Austin’s works. The most striking, probably,
can be found in the very structure of How to Do Things with Words. The
work is constructed in a very peculiar way: the first part of it introduces
the concept of a performative and distinguishes between performatives
and constatives (expressions describing reality); the second part, in turn,
questions the usefulness of the notion of a performative and introduces
a new theory of the expressions with which we “do something in the
86
CHAPTER 3
world”. It is interesting to look closer at the reasons (and the method)
that led Austin to abandon the concept of a performative: those reasons
are as important as those that prompted the development of Austin’s
initial theory.
Austin observes, first, some similarities in the behavior of performatives and constatives. Let us inspect the following sentence: “John’s children are bald”, uttered in a situation when John has no children. The
presupposition of the sentence “John’s children are bald” – i.e., the sentence “John has children” – is false. According to the theory of presuppositions we discussed earlier, a sentence, whose presupposition is false,
can be ascribed neither truth nor falsehood; we say such a sentence is
senseless. According to Austin the sentence “John’s children are bald”,
even if John has no children, cannot be called senseless; it should rather
be described as “void”.
This observation allows us to see the similarity between the described
situation and, for instance, a situation in which a bridegroom’s “I do” is
uttered but some procedural conditions have not been fulfilled, i.e., a situation of a misfire. In this context Austin suggests that the sentence
“John’s children are bald”, uttered in a situation where John has no children, should also count as a misfire. The same can be said of the sentence
“A cat is on a mat”, uttered even though I do not believe that a cat is on
a mat. This situation is similar to the one in which I say “I promise”,
although I do not intend to keep the promise. Both can be labeled
“abuses”.
Austin summarizes these remarks in the following way: “[It] suggests
that at least in some ways there is danger of our initial distinction
between constative and performative utterances breaking down”.37 This
conclusion leads us to the essential question: is it possible to develop a
precise criterion for distinguishing between performatives and constatives? Austin considers first a grammatical criterion: performatives are
expressions that use verbs in first person singular present indicative
active (e.g., “I promise”, “I give”, etc.). This criterion will not do, however. In performatives one can use verbs in the third person (e.g., “You
are hereby authorized to pay . . .”, “Passengers are warned . . .”), in plural (“We promise . . .”), in various tenses (“You did it” instead of “I find
you guilty”), in passive (“You are obliged. . .”) and in different moods (“I
order you. . .”).
Consequently, instead of a purely grammatical criterion, Austin proposes a criterion that can be labeled “a paraphrase criterion”: a performative is every expression that is “reducible, or expandable, or analyzable
into a form or reproducible in a form, with a verb in the first person
ANALYSIS
87
singular present indicative active”.38 But in such a case we are entitled to
say that the sentence “I read a book” is a performative, although it seems
not to be. The above definition should therefore be augmented with the
condition that performatives are expressions that include (or their paraphrase includes) some special verbs; “to promise” is such a verb, but “to
read” is not.
Austin observes that those special verbs display a certain asymmetry
as regards their use in first person singular present indicative active and
their use in different persons and tenses. For instance, the utterance of
“I promise” is certainly an extra-linguistic action but “You promised” or
“He promised” are only descriptions of such actions. Similar things
can be said of such verbs as “to bet”, etc. The asymmetry cannot be
observed, however, in the case of such verbs as “to read” or “to walk”.
There exist verbs, however, that show the futility of Austin’s attempt.
For instance, the verb “to state” exhibits the asymmetry in question (“I
state . . .” as opposed to “You state, that . . .”) but to call this a performative verb is problematic (similarly with the verb “to classify”). On the
other hand one can insult someone with verbs but the paraphrase
“I insult you” seems inadequate.
It turns out, then, that all the attempts to formulate a criterion differentiating performatives from constatives fail. Neither the grammatical
criterion nor the paraphrase criterion with the addition of the class of
“asymmetrical” verbs can do the work demanded. This result causes
Austin to rethink the problem. It is helpful to pause here for a moment
and ask what the reasons are that caused Austin to abandon his initial
theory, and what role the assumptions underlying his method played in
his decision. We should ask why Austin even cares that the distinctions
he introduced are not exceptionless. Is not it true that vagueness is something natural for ordinary language, and that the fact that some divisions
are not exclusive and precise does not mean they are wrong or useless?
The only answer that is in compliance with the method of linguistic
analysis, is to say that the distinction between performatives and constatives presents wrongly the function of ordinary language. The next natural question: how do we know that this theory is wrong, has the following
answer: the analyzed examples show a more adequate and elegant conception. The conception in question is the theory of speech acts.
A speech act is a basic unit of linguistic communication. The examples
that motivated the abandonment of the distinction between performatives and constatives show that every speech act can be viewed from different perspectives; thus one can say that every speech act has at least two
“dimensions”: performative and constative. Austin proposes that speech
88
CHAPTER 3
acts be looked at from three different angles: one can treat them as
expressions that have certain meaning, as utterances that have certain
conventional force and as expressions that cause a certain result which is
not determined conventionally.
Pursuing this argument, Austin distinguishes between three aspects of
a speech act: the so-called locutionary act, illocutionary act and perlocutionary act. A speech act as a locutionary act is the act of saying
something; at the same time, as an illocutionary act, it is performing
some other thing (e.g., asking, ordering, sentencing, apologizing); finally,
as a perlocutionary act it is bringing about some consequence in the
actions, thoughts or feelings of the listeners or the speaker. Let us revert
to an Austinian example. A speech act in which someone says to me
“Shoot her” is:
– a locution: he said to me “Shoot her” meaning by “shoot”, shoot,
and referring by “her”, to her;
– an illocution: he urged (or advised, or ordered) me to shoot her;
– a perlocution: he got me to shoot her.
Austin’s analysis does not stop at this distinction. The philosopher
considers the adequacy of his conception, paying attention to the relations between locutionary, illocutionary and perlocutionary acts, and the
criteria for distinguishing between them. We will not follow these analyses in detail, although they constitute a good example of the application
of the method of linguistic analysis. We would like to mention, however,
that the Austinian theory is not “the last word” of linguistic philosophy
as regards speech acts. There were numerous philosophers that tried to
develop it further. Of special interest is J. Searle’s work, Speech Acts.39
Searle criticizes the differentiation between locutionary and illocutionary
acts, pointing out that it is not unusual that meaning constitutes illocutionary force (as is the case with “I promise to do this”). In such cases it
is impossible to distinguish between locutionary and illocutionary acts.40
Searle set out to replace Austin’s theory with one of his own. Searle
treats speech acts as units of linguistic communication that consist of an
utterance act, a propositional act, an illocutionary act and a perlocutionary act. The utterance act is the act of uttering certain words. The propositional act is the act of expressing a certain proposition. Both
illocutionary and perlocutionary acts, as described by Searle, can be
broadly compared to their Austinian counterparts. We will not analyze
in detail Searle’s conception. The conception is another illustration of
how linguistic analysis operates and of the problems it encounters. The
richness of ordinary language makes it easy to find counter-examples
to any hypothesis. This does not mean, however, that such fallible
ANALYSIS
89
hypotheses and theories are useless; it is certain that they say something
important about language and about how we use it.
Another illustration of this fact is the so-called typology of speech acts.
In the last chapter of How to Do Things with Words Austin identifies five
general classes of speech acts labeling them with “the following, more-orless rebarbative names”41: verdictives, exerctives, commissives, behabitives,
expositives. Verdictives are typified by the giving of a verdict; a list of
verbs characteristic of those acts includes: “acquit”, “convict”, “find (as
a matter of fact)”, “calculate”, “measure”, “analyze”. Exerctives are the
exercising of powers, rights or influence (“nominate”, “dismiss”, “order”,
“vote for”, etc.). Commisives are typified by promising or otherwise
undertaking (“promise”, “contract”, “declare intention”, etc.).
Behabitives, in turn, concern attitudes and social behavior (“welcome”,
“bid farewell”, “toast”, “condole”, etc.). Finally, expositives – difficult to
define, as Austin admits himself – locate our utterances in a structure
of argumentation or discussion (“I reply . . .”, “I give as an example . . .”,
“I assume that . . .”).
It is not surprising that a closer analysis of the relationships between
the mentioned categories reveals that this is not an exhaustive classification but only a typology. It is equally unsurprising that Austin’s typology
was subjected to severe criticism, notably by Searle, who proposed his
own taxonomy of speech acts.42 To date, there have been developed several such typologies, sometimes employing very complicated conceptual
structures. Research concerning speech acts is carried out today not only
by philosophers, but also by linguists and jurists. Of special interest are
pragma-dialectical analyses, i.e., analyses of argumentation structures
consisting of numerous speech acts.43
At the end of our analysis of the application of linguistic methods to
the conception of speech acts we should observe that this theory has special import for legal theory and the philosophy of law. One does not need
to go to great lengths to show that the notion of illocution (or performative) is necessary for an adequate account of legal language.
3.2.4
The method and Its Limits
Methods. From what has been said so far, we can formulate several
remarks about the methods of linguistic analysis. First of all one can say
that those methods are based on a general directive that may be
expressed in the following way:
Put forward hypotheses concerning the problem that interests you, and test them on
examples motivated by intuitions regarding the use of ordinary language.
90
CHAPTER 3
This directive is formulated in a very general way, which is inevitable if
one wishes to state a directive that embraces both maximalist and minimalist approaches to linguistic analysis. The general directive can be
applied in particular cases using different “procedures”. In the course of
our presentation we have identified and described some of them. One
example is developing typologies, a method often applied by proponents
of the minimalist approach. The fact that they construct typologies
rather than logical divisions is rooted in features of ordinary language,
such as vagueness. Other procedures of linguistic analysis include argumentation from paradigms and – especially interesting – the method of
presuppositions. It has to be stressed that the list of methods covered
here does not exhaust the list of methods serving the realization of the
general directive of linguistic analysis. The examples are highly characteristic, but they do not comprise the whole list.
It is appropriate to consider which of the models of analysis is represented by linguistic methods. This question can be troublesome, especially in the case of Austinian “minimalism”. We have analyzed Austin’s
conceptions of language, but these do not aim to solve a particular philosophical problem, they try only to capture some features of the “basic
medium” of philosophizing. Austin’s method is, however, also capable of
dealing – at least initially or on a preparatory basis – with extra-linguistic
problems. For instance, in “A Plea for Excuses” (1956) the philosopher
considers the problem of freedom. Strawson’s maximalist method, on
the other hand, aims to discover the conceptual scheme. This “discovering” is, however, connected with reducing or relating analyzed problems
to the scheme. Therefore, one can call it, as well as Austin’s approach,
analysis3.
In linguistic analysis there are also methods of reasoning that may
count as instances of analysis2. A good example is found in Austinian
typologies. Finally, some methods used by “soft” analytic philosophers
can be labeled analysis1. Recall the method of presuppositions that we
described above. It displays a striking resemblance with analysis1. The
difference is that, whereas in analysis1 one seeks logical reasons for the
analyzed sentences, the method of presuppositions consists in looking
for sentences presupposing the given ones. In both cases the relationship
between what is given and what is searched for has a logical character.
Our considerations inevitably lead to the conclusion that a precise definition of linguistic analysis is impossible. Moreover, one should not
really speak of linguistic analysis as one, unique method. Rather, what we
have here are different strategies for dealing with philosophical problems
that each aim at realizing the general directive we formulated: put forward
ANALYSIS
91
hypotheses concerning the problem that interests you and test them on
examples motivated by intuitions regarding the use of ordinary language.
The limits. Our discussion of linguistic analysis was centered on the
legal-theoretic problems. The discussion has shown that this method can
serve legal philosophers well. It is not surprising then that linguistic
analysis has been used to investigate numerous problems, ranging from
legal language, discourse and the ontology of law, to more specific problems, like the validity of norms. It is not difficult, either, to envisage applying this kind of analysis in the field of legal dogmatics and legal practice.
An interesting relationship between the methods of linguistic analysis
and the interpretation of law was observed by R. Sarkowicz, who developed a three-layer conception of legal interpretation.44 Among the three
layers of interpretation, he identifies the level of presuppositions, on
which the interpreter reconstructs the worldview, the system of values,
and the conception of social institutions as assumed by the legislator.
These pieces of information are presuppositions (in a loose sense of
the word) of legal text. Therefore, in Sarkowicz’s conception, one of the
characteristic methods of linguistic analysis becomes an important component of the process of legal interpretation.
Summary. In summarizing our presentation we would like to underline
three issues.
First, the method of linguistic analysis is based (especially in the
maximalist version) on a very strong defense of the distinguished role of
ordinary language. From this assumption important consequences
follow in the results of the analyses. In particular, features of ordinary
language, such as vagueness and imprecision, are “inherited” by the
results of linguistic analysis.
Second, it would be difficult today to find any theorists who apply
linguistic analysis in its pure form. Despite this, many fruitful analyses
do take advantage of linguistic methods.
Third, some of the conceptions of “soft” analytic philosophers have
profound importance. Historically speaking, one cannot overlook the
fact that linguistic analysis highlighted a sphere of language that was not
properly analyzed in a philosophical world dominated by theories of the
Vienna Circle. This sphere includes practical discourse and the pragmatic
dimension of language. From a more contemporary perspective, “soft
analysis” has given rise to achievements, such as speech acts theory,
which are regarded as highly important contributions to legal theory and
the philosophy of law.
92
CHAPTER 3
3.3
ECONOMIC ANALYSIS OF LAW
3.3.1
Law and Economics
Economic analysis of law – i.e., applying economic tools to the analysis
of what the law should be – is not a new idea. Some elements of this
approach can be traced in the writings of Machiavelli, Montesquieu and
the representatives of the German historical school. Economic analysis
of law par excellence originated, however, in the Anglo-Saxon world.
This is hardly surprising, given that the Anglo-Saxons built the foundations of economics itself, as witnessed by the Scottish Enlightenment and
the philosophies of Bentham and Mill.
Already in 1897, one of the greatest representatives of American realism in legal theory, O.W. Holmes, wrote:
For the rational study of the law the black-letter man may be the man of the present, but
the man of the future is the man of statistics and the master of economics.45
Holmes’ prophecy had already become reality by the beginning of the
twentieth century, when American, German and Scandinavian jurists
attempted to apply economic methods to the analysis of law.46 Economics
was not, however, the only perspective they adopted. Economic tools were
only part of the theoretical arsenal used by legal realism.
Out of this tradition the Chicago School of Law & Economics originated. It is usually held that the school was established in the 1970s.
However, by the end of the 1950s, research on the borderline between law
and economics was undertaken. Amongst the seminal articles of those
times one should mention R. Coase’s “The Problem of Social Cost”
(1960), A. Alchian’s “Some Economics of Property Right” (1965) and
G. Calabresi’s “Some Thoughts on Risk-Distribution and the Law of
Torts” (1961). In 1972 R. Posner published his Economic Analysis of Law,
defending the thesis that the traditional institutions of common law are
economically effective. In this way the Chicago School was fully established in the mid-1970s. Posner is regarded as its main representative. The
idea of applying economic tools to law proved useful enough to prompt
research in various academic centers in the United States and around the
world. Today it is difficult to count the many schools carrying out their
research under the heading of the economic analysis of law.47 Both the
variety and number of enthusiasts of the economic analysis of law make
it impossible to talk about a unified research program. It seems however
that all the representatives of Law & Economics share a basic thesis,
which says that the law is (or should be) economically effective (i.e., it
leads or should lead to economically effective allocation of goods).
ANALYSIS
93
This basic thesis is surrounded by some troublesome issues. An explanation must be given, for instance, of what exactly constitutes “an economically efficient allocation of goods”. Economists and lawyers have
offered at least two explanations. First, one can speak of Paretoefficiency. The allocation of goods is Pareto-efficient in the given society
if it is not possible to better the situation of any member of the society
without worsening the situation of others. Secondly, one can apply
Kaldor–Hicks efficiency: the allocation of goods is efficient if it is
impossible to better the sum of the welfare of the society, even with the
assumption that the situation of some of the members of the society can
be worsened. It is not our aim to analyze the above mentioned definitions of efficiency. We need only stress that the concept of economic efficiency is not univocal, and that controversy over the understanding of
efficiency plays an important role in debates concerning the foundations
of the economic analysis of law.
The problem of defining economic efficiency should be placed within
a more general framework. The basic thesis of Law and Economics
demands that efficiency is taken as the main indicator of what the law is
(should be). This suggestion seems to contradict traditional conceptions
of what a legal system is: according to the traditional account, the legal
system should secure justice, not efficiency. For the proponents of Law
and Economics, however, there exists no tension between justice and efficiency. We can even say that efficiency is an economic explication of justice. One does not have to add that this assumption is controversial. We
leave the issue here, only highlighting its existence.48
Another premise of the economic analysis of law is based on the fact
that economic models employ the concept of a person who acts rationally (homo oeconomicus). This assumption is often questioned by psychologists who maintain that much human behavior is far from being
“rational”. This particular objection can be challenged by saying that the
criticized assumption has proved useful – at least to a certain degree – in
the history of economics. One should add here that, in contemporary
economics, models are built in which the assumption in question is in
various ways “loosened”.49
Let us observe further that the basic thesis of the economic analysis of
law can be read in two ways: descriptively or normatively. In the first case
we will say that the law is economically efficient. In the second – that it
should be. The pioneers of Law and Economics attempted to justify the
descriptive thesis. For instance, one of the aims of Posner’s Economic
Analysis of Law is to show that the precedents of American courts in
tort law lead to economically efficient results. As time has passed, more
94
CHAPTER 3
emphasis has been placed on the normative dimension of the basic
thesis.50 Below we will read the basic thesis normatively for it enables us
to show that the tools of economics can serve not only the theory of law,
but also legal practice.
3.3.2
Idea of Economization
Let us look at the following example. Let us assume we have a simple
legal system consisting of just one norm:
(A) Whoever causes damage to someone else must redress it.
The obligation expressed in (A) is not very precisely determined. Almost
all the concepts used in its formulation give rise to interpretational problems. How should one define “damage”? What does it mean that a person A caused damage to a person B? Should it be the case that only those
actions of A that directly result in a damage count as torts? What does it
mean to “redress” the damage? These are the questions to be answered
in the process of interpreting of (A). What are the criteria of interpretation to be employed?
Let us look at the following situation. John has caused damage to
Adam. Therefore, on the basis of (A), he has to redress it. But in certain
circumstances we may ask: should John redress the entire damage? Such
a question is sound in a situation in which Adam’s behavior contributed
to the damage. What are the criteria for answering it?
Let us try to build a simple economic model, which can serve as a basis
for the required answer. What indicators need to be taken into account?
It is clear that we need to know the value of the damage. Let us call it S.
Let us assume further that Adam could have undertaken an action that
resulted in no damage. Such an action naturally has its own cost. Let KP
stand for the cost of Adam’s preventive actions. The comparison of S
and KP does not tell us much. Obviously, if the cost of the preventive
action KP is higher than the damage S, then undertaking the preventive
action would be irrational.
Let us consider now whether we should expect the plaintiff (in our
case: Adam) to undertake preventive action when KP < S. The answer to
this question will be positive only if the damage occurs in every case in
which the plaintiff refrains from acting. If, however, the plaintiff’s refrain
does not in every case result in damage, the inequality KP < S is useless.
Let us assume then that by P we will understand the probability of damage where the plaintiff does not undertake the preventive action. Let us
assume, further, that undertaking the preventive action eliminates the
possibility of the damage. In such circumstances, every time the unit cost
ANALYSIS
95
of prevention is lower than the unit cost of the damage (KP < P · S), it is
rational to undertake preventive action. Let us illustrate this using our
example. Assume that the damage S, John has caused to Adam is 100.
Let the probability of the damage P equal 10%, and the cost KP – 4. It is
easy to observe that KP < P · S, for 4 < 10. It would be rational then for
Adam to undertake preventive action. If he refrained from doing this,
the court could say that his behavior contributed to the damage. John
will not, therefore, be responsible for the entire damage.51
Let us look more closely at how the court’s reasoning runs here. John
has caused damage to Adam; the damage amounts to 100 and the question the court has to answer reads: should John redress the entire damage, or only a part of it? The court will choose the former possibility only
if Adam’s behavior did not contribute to the damage. When can one say
that this is the case? Our simple economic model suggests that it is the
case when the unit cost of prevention is lower than the value of the damage multiplied by the probability of its occurrence.
The model presented above is, of course, quite simplistic as it rests on
many idealistic assumptions. We have assumed, for instance, that the
preventive actions reduce to 0 the probability of the occurrence of the
damage. What if, however, the preventive actions reduce the probability
only by half ? Let us inspect the table below (from now on by P we
understand the probability of the occurrence of the damage in the given
circumstances):
Situation 1
Situation 2
KP
P (%)
S
Expected damage
Total social costs
0
4
10
5
100
100
10
5
10
9
Under these changed circumstances, in Situation 1, in which the plaintiff does not undertake the preventive actions, the social cost is 10. In
Situation 2, in which preventive actions are undertaken reducing the
probability of the occurrence of the damage to 5%, the social cost
equals 9 (that includes the expected damage and the cost of the preventive action). Since the social costs in Situation 2 are lower, this is the
desired situation; the plaintiff should therefore undertake the preventive
actions. If he refrained from doing so and damage occurred, the court
will hold him responsible for contributing to the damage. If P1 stands for
the probability in Situation 1, and P2 for the probability in Situation 2,
the rule forming the basis of the court’s decision can be formulated in the
following way: if (P2 · S) + KP < P1 · S, then the plaintiff contributed to
the damage.
96
CHAPTER 3
Another assumption made in the model constructed above was risk-neutrality. A risk-neutral person values the certainty of receiving 5,000 USD
equally with a 50% chance of receiving 10,000 USD. A risk-averse person,
by contrast, values the certainty of receiving a certain amount of money
more highly than a 50% chance of receiving double that sum. Finally,
a risk-lover prefers a 50% chance of receiving 10,000 USD to the certainty
of receiving 5,000 USD. If we replaced our assumption of risk-neutrality
with the assumption that the person in question is risk-averse, the difference in the total social cost between Situation 1 and Situation 2 would
increase: in Situation 2 the risk of the occurrence of damage is lower
than in Situation 1, and this constitutes an added value for the riskaverse person (note, however, that there may be no good reason to abandon the assumption of risk-neutrality in constructing a model for the
purpose of answering a general interpretational question).
The following further assumptions are also implicit in the model given
above: (a) that the parties’ level (frequency and duration) of activity does
not affect the determination of the cost of the preventive actions; (b) that
the court is capable of determining precisely the costs and results of both
parties’ actions; (c) that all the costs are measurable and can be expressed
in monetary terms.52 It is vital to question whether acceptance of these
assumptions is sound. In other words: how do the assumptions influence
the results of the analysis, and is their influence significant enough that
the assumptions should be analyzed themselves? With regard to riskneutrality and the level of activity, it is possible to extend the mathematical structure of our model to include those additional factors.
Assumptions (b) and (c), however, highlight an essential problem concerning the economic analysis of law.
One of the greatest difficulties with the economic analysis of a legal
case is the “economization” of the case, i.e., quantification, estimation of
the costs and values of states of affairs and actions. If John damages
Adam’s car, which is worth 10,000 USD, then the value of the damage
can easily be estimated. Imagine, however, that John’s car crashes into
Adam’s bicycle. If the only thing damaged is the bicycle, the values can
be straightforwardly calculated. But if Adam breaks his leg in the accident, the situation becomes more complicated. And in such a case how
can one verify whether or not Adam contributed to the occurrence of the
damage? How should the cost of the preventive actions be calculated?
The problem of estimating the amount of the damage is not only troublesome for the economic school; this is a typical problem in legal practice. In civil codes one usually finds some rules that help to determine the
values in question. In other systems the rules are developed by the courts.
ANALYSIS
97
Therefore, the indicated problem does not exclusively affect the economic analysis of law.
From the above discussion a picture can be constructed in which application of the economic method is divided into three stages. In the first
stage a case is “economized”, i.e., the relevant aspects of the case are
quantified and expressed in monetary terms. At the second stage, a suitable economic (mathematical) model is developed to assist in answering
the questions that arise in the case. Finally, at the third stage, conclusions
are drawn from that economic model. Problematic decisions are made in
all the three stages. In the first stage actions and states of affairs to which
it is difficult to ascribe a definite economic value are nonetheless “estimated”. In the second stage one can always question whether the constructed model takes into account all the important factors, or whether
it is based on over-simplistic assumptions. The third stage, leading “from
economics back to the law”, can also be troublesome, although the principle that governs it – to promote solutions minimizing the total social
cost – seems clear.
The three-stage account of the application of the economic method
is, of course, a simplification. For instance, the first stage cannot be
completely detached from the second; the possible economic models
determine what can be “economized” in a given case. Similarly, the conclusions we draw from the model do not have to end the analysis. If they
are unacceptable then they may lead to a revision of the model or to a
new “economization” of the case. Even so, the three-stage scheme
described above does give a satisfactory account of the key elements of
the economic method.
3.3.3
Limits of the Method
The limits of the application of economic analysis to law will now be considered. In order to do this, several examples from various areas of law
(especially private law and criminal law) will be addressed more closely,
and some typical situations faced by lawyers will be analyzed (interpretation of law, creation of law, determination of sanction to be applied, etc.).
It is not surprising that economic analysis has been used most widely
in private law, in which economic efficiency may be considered the highest aim. In the example presented in the previous section, the construction of an economic model proved useful in the interpretation of a
generally stated legal norm constituting tort liability:
Whoever causes damage to someone must redress it.
Imagine now a legislator who is about to introduce such a norm, but
they want to make more precise the concept of tort liability. Let us
98
CHAPTER 3
assume that the legislator has to choose between two different ways of
determining liability. The first bases liability on the condition that the
actor could have undertaken steps to minimize the probability of the
occurrence of damage. If the action that caused the damage was undertaken with due care the agent is not liable; otherwise he/she is held
responsible for the damage (this type of liability is called negligence).
The second way of determining liability involves declaring responsible
anyone who has caused damage, irrespective of whether or not he acted
cautiously (this is called strict liability). From the economic perspective,
the first conception assesses liability according to whether the agent
incurred the relevant cost of the preventive actions (=acted with due
care) or not. In the second conception the agent is held liable irrespective
of any costs he incurred to undertake the preventive actions.
Let us consider which of the conceptions should be applied by the legislator. From the point of view of the economic analysis of law, the question is which of the models of liability leads to the most economically
efficient solution, i.e., which minimizes the total social cost. Let us look
at the two tables:
Preventive actions
No
Yes
Preventive actions
No
Yes
KS
P (%)
S
Expected damage
Total social cost
0
4
20
5
100
100
20
5
20
9
KS
P (%)
S
Expected damage
Total social cost
0
4
20
18
100
100
20
18
20
22
KS stands here for the costs of the preventive actions. In the case of
the first table the preventive actions that cost 4, reduce the probability
of the occurrence of the damage from 20% to 5%. In this way the social
cost without the preventive actions equals 20, with them it equals 9. This
means that undertaking of the preventive actions is reasonable, for it
leads to better results in terms of social cost. The second table illustrates
a situation in which the preventive actions reduce the probability of the
occurrence of the damage only by 2%, from 20% to 18%. In this case the
total social cost without prevention is 20, whilst, with it, it is 22. This
time refraining from the preventive actions is the optimal decision.
Consider now what would be reasonable behavior for an agent in both
cases depending on the conception of liability accepted. In Situation 1,
accepting the conception of negligence leads to the following results: (a) if
ANALYSIS
99
the agent does not act with due care (cost 4), he will have to redress the
entire damage (20), whilst (b) if he acts with due care (cost 4), he will not
be held liable, therefore his costs are limited to the cost of the preventive
actions (4). It is clear, then, that a rational actor would rather pay 4, in
order to avoid paying 20. If, in Situation 1, the strict liability conception
is adopted, the agent who does not undertake the preventive actions
would pay 20, otherwise – 9 (costs of prevention + the costs if damage).
Therefore, in Situation 1, the agent would choose the economically
efficient route of undertaking the preventive actions no matter which
conception of liability is followed. Let us observe, however, that in the
case of strict liability his costs are significantly higher than in the case of
negligence (9 to 4). This means that the frequency of his activity will be
much lower with strict liability than with negligence (for higher costs will
make the activity less efficient).
In Situation 2, following the conception of negligence leads to the
agent paying 4, if preventive actions have been undertaken, and paying 20
if they have not. Rational behavior requires undertaking those preventive
actions, which in Situation 2 lead to a solution that is economically inefficient (social cost 22). The conception of strict liability, on the other
hand, results in the agent who undertakes the preventive actions paying
22, or 20 otherwise. In this case there will be no prevention, which is an
efficient solution. Let us observe that in cases like Situation 2, acceptance
of the negligence approach leads to inefficiency and enables the agent to
increase the frequency of his activities (low cost that equals 4). Strict liability, by contrast, is an efficient solution in Situation 2, and decreases the
frequency of the activities in question (high costs equaling 20).
The following conclusions can be drawn from the analysis of the problem above: should negligence or strict liability be chosen? It is easy to
observe that in such cases as Situation 1, undertaking the preventive
actions produces desirable results: the probability of the occurrence of
damage decreases from 20% to 5%. The efficiency of preventive actions
in Situation 2 is much worse (from 20% to 18%). What does this say
about the kind of activity described by the tables? The activities in
Situation 2 have to be dangerous, for the probability of the occurrence of
damage is in their case very high, and the preventive actions do not help
considerably here. Demolishing old buildings is a good example of such
activity. From the point of view of a society, the level of frequency of
this activity should be as low as possible. Thus strict liability is recommendable here. The activities of Situation 1, on the other hand, are less
dangerous, for it is relatively easy to decrease the possibility of damage
in their case. Building a highway may serve as an example. This activity
100
CHAPTER 3
is often socially acceptable and the law should encourage it. Acceptance
of the conception of negligence, rather than strict liability, acts as
encouragement, for its costs are lower than those of strict liability.
A short survey of legal regulations in different countries shows that
legal systems comply with this analysis. In the United States, Germany,
France or Poland, tort liability is usually based on negligence or similar
ideas. Only dangerous activities are treated differently, using strict liability or something similar.
The examples discussed above may give the impression that application of the economic analysis of law is restricted only to those parts of
private law that concern torts. It is true that that area of law was explored
in early works on Law and Economics. Today, however, economic analysis is applied to all kinds of problems in private law. A lot of work is
devoted to the concept of property, and there are also important contributions concerning contract law. An interesting shift of perspectives may
be observed here. Consider interpreting the provisions of an agreement.
In the provisions of law, the main purpose of interpretation – from the
economic point of view – is to reduce social cost. Agreements, however,
should be interpreted in a way that minimizes the costs of the parties.
Proponents of the economic analysis of law also analyze other areas of
private law, such as insurance law, legal procedure, etc.53
More problematic is the application of the economic method to criminal law. An exception is the determination of a sanction, and designing
the system of sanctions, which does seem to fit with economic analysis.
J. Bentham wrote in 1788: “the profit of the crime is the force which
urges man to delinquency: the pain of the punishment is the force
employed to restrain him from it. If the first of these forces be the
greater, the crime will be committed; if the second, the crime will not be
committed”.54 Bentham’s observation can be presented as a simple economic formula. Let Z stand for the income expected from the crime, K
for the loss connected with the punishment, and P for the probability of
punishment. Someone who acts rationally will commit a crime only if
the expected income will be higher than the cost of punishment multiplied by the probability of punishment:
Crime will be committed if Z > P · K.
As in the previous cases, this model is significantly simplified. It can
naturally be extended in various ways, for instance by taking into
account different attitudes towards risk (one can suppose that there are
a relatively high number of risk lovers amongst criminals). Especially
problematic is the estimation of the value of Z, P and K. For example,
the expected income, or cost, of many types of crime should include such
ANALYSIS
101
factors as “psychological value”. This is an important problem, but the
analysis here is ex definitione simplified. It is only necessary for present
purposes to outline what the economic method is.
Suppose for instance that we would like to determine whether it is
rational – within a given system of sanctions and law enforcement – to
enact norms that apply more severe sanctions and to increase the expenditure on law enforcement. In order to answer this question let us return,
for a while, to the perspective of a potential criminal. According to the
model above, the crime will be committed if Z > P · K. Clearly the probability of punishment P depends on the state of law enforcement. The
value of K, on the other hand, is connected to how the system of sanctions is designed. One has to add that this system has its economic
dimension. If most crimes are punished with imprisonment, then the
expenses for penitentiary system are high.
On the one hand, therefore, we have the social costs resulting from
crimes, and on the other, the costs of prevention, which are equal to the
expenses of law enforcement. A system of criminal law is efficient if the
costs of prevention are lower than the costs of the crimes that were
avoided because of the prevention. In other words, investing in law
enforcement is efficient up to the point when the marginal cost of prevention equals the marginal cost of the crime avoided thanks to the additional prevention.55 These simple dependencies are helpful in resolving
the question of whether to increase expenditure on law enforcement
and/or whether to introduce more severe sanctions: the move will be
rational as long as the system of criminal law remains efficient.
This solution is not, of course, the result of a painstaking economic
calculation; it follows rather from general considerations, which are nevertheless based on a simple economic model. More precise analyses can
be carried out, for instance, regarding what kind of sanction is the most
efficient for specific types of crimes. One can consider, e.g., whether it is
better to apply monetary or non-monetary sanctions.56 The sanction
should be determined in a way that deters the potential criminal from
committing the crime. Therefore, with the given probability of punishment P, K has to be calculated in such a way that Z < P · K holds. If this
aim can be achieved with monetary sanctions then these should be
applied. The reason for this is simple: from the economic point of view
monetary sanctions are less expensive than the non-monetary, because
they enable us to avoid expenses for penitentiary system.
It is easy to identify some factors that prevent monetary sanctions
from serving their purpose (i.e., they fail to deter). One such factor is
the limited property of the potential criminal. If a sufficiently deterring
102
CHAPTER 3
monetary sanction was significantly higher than the value of the property then non-monetary sanctions would have to be applied (including
imprisonment). Similarly, the lower the probability of punishment is, the
more severe the monetary sanctions have to be (in order that P · K > Z);
in such a case it may turn out that, even for a potential criminal who has
some property, the level of K will be so high that it will not serve its purpose as a deterrent. The application of non-monetary sanctions can also
prove necessary when the expected income from a crime (Z) will be so
high that the rule Z < P · K would make K extremely high.57 Therefore,
the monetary sanctions do not serve their function, either in the case of a
person who has limited property, or in the case of a rich man who is
planning to steal 1 million USD. The soundness of this analysis is reflected
in the fact that, in penal codes, most “basic” crimes carry non-monetary
sanctions.58
This, and other economic analyses of sanctions, can provoke two
kinds of objection. First, one can question the assumption that people
who commit crimes act rationally. Second, the “economization” of punishment makes us forget about a basic dimension of criminal law, namely
the notion of a just punishment.
It is indeed the case that the simple model of behavior for the potential
criminal is based on an assumption that the acting person is rational and
that he/she will calculate potential income and loss. As has already been
stated, the assumption of rationality is the basis for most of the economic
models. The problem is, however, that unlike private law, in criminal law
this assumption seems highly counter-intuitive. This is because, traditionally, criminologists and sociologists speak of crimes in psychological and
sociological terms, underlining the a-typical qualities of criminals. This,
however, does not constitute a decisive argument against the application
of economic analysis to criminal law. It can – at the very least – serve as
an alternative to traditional criminology. Observe, for instance, that a lot
of crime is committed in anticipation of profit. Moreover, the rationality
assumption is also questioned in relation to the classic economic models.
Despite this, the assumption proves useful, because what is analyzed is
not the behavior of a specific person; we apply economic tools to model
the behavior of certain markets (of wheat or of crimes!) in which their
members “statistically” act in a rational way.
The second of the mentioned doubts, which concerns the substitution
of justice with efficiency, has already been analyzed above. Let us repeat
here that economic efficiency can be treated as an explication of justice.
If this controversial thesis is accepted then one should not speak of a
“substitution” of justice with efficiency. A sanction which is optimal in
ANALYSIS
103
terms of economic efficiency is also just. Moreover, the acceptance of the
thesis that efficiency is an explication of justice facilitates not only the
analysis of how sanctions should be applied, but also of the most basic
concepts concerning criminal responsibility. In literature, different analyses concerning intent, error, self-defense, etc. can be found.59 On the
other hand, the dismissal of the thesis that efficiency is an explanation of
justice, does not necessarily lead to questioning the usefulness of the economic analysis of criminal law. On this second reading, the economic
analysis may be regarded as an alternative approach to the problems of
criminal policy.
Apart from private and criminal law, other areas of law can also be
analyzed with the use of the economic method – for instance legal procedures, or constitutional law.60 The latter is connected to a field of
research which makes extensive use of economic analysis, e.g., political
science and the theory of social choice. The problems analyzed in this
context aim not only to answer questions of what the law should be, but
also enable one to look at the role of legal systems from a more general
perspective. A good point of departure for such analyses is the notorious
Coase Theorem. Coase, the Nobel Prize winner for economics in 1991,
published at the beginning of the 1960s a famous article, “The Problem
of Social Cost”.61 In the article a theorem is formulated that can be
reconstructed in the following way: in a world where the transaction
costs equal zero, the allocation of goods is efficient irrespective of the
initial distribution of property rights.62 The notion of “transaction
costs” used above causes heated debates in economic and legal-economic
literature. Usually by “transaction costs” one means either the cost of
establishing and maintaining property rights, or the cost of transferring
the property rights.63 Coase Theorem says in effect that if there are no
such costs then, no matter how the legal system is built, (irrespective of
the initial distribution of property rights), an efficient allocation of
goods will be achieved. In other words, assuming that the transaction
costs equal zero, the form of law has no importance; what counts is
whether there is any law or not.
One should not of course conclude that – because of what the Coase
Theorem says – the law is useless, or that we can enact anything in the
belief that the market will “take care of itself ”. Nevertheless, some interesting conclusions for legal theory and philosophy follow from the theorem as it points out an important relationship between the form of law
and transaction costs. In reality those costs never equal zero, and hence
the way the legal system is built matters. If, following proponents of
Law and Economics, it is assumed that law should promote economic
104
CHAPTER 3
efficiency then, on the basis of the Coase Theorem, at least two directives
for creating (or interpreting) law can be formulated. The first says that –
if possible – legal norms should minimize transaction costs. According
to the second, if high transaction costs cannot be eliminated, then the
law should aim at an efficient allocation of goods not counting on the
“invisible hand of the market”.
Not all the assumptions behind, and consequences following from the
Coase Theorem will be considered here. It is only necessary to show that
economic analysis can be suitably applied to the most general legaltheoretic issues. Moreover, economic analysis here does not consist in the
construction of a mathematical model (although the Coase Theorem has
a very precise mathematical form).
A good summary of the examples presented above is given in the
following words of G. Becker: “Indeed, I have come to the position that
the economic approach is a comprehensive one that is applicable to all
human behavior, be it behavior involving money prices or imputed
shadow prices, repeated or infrequent decisions, large or minor decisions,
emotional or mechanical ends”.64 The method of economic analysis can
be applied not only in private law but in any area of law, including most
general problems of the philosophy of law, such as the justification of
the existence of law.
3.3.4
Conclusions
The analysis in the previous section demonstrates that there are various
branches of the economic analysis of law: they use various tools, take
advantage of mathematical modeling more or less directly, and are based
on different assumptions. They are all similar, however, in that they aim
to express a legally relevant case in the language of economics and to
attempt to draw conclusions in which economically efficient solutions
are promoted.
Admittedly, the assumptions standing behind the economic analysis
of law are objectionable. The “substitution” of justice with efficiency,
“calculating” of values that seem unquantifiable, the acceptance of the
counter-factual model of homo oeconomicus – all this is problematic.
One can argue, however, that the weakness of the economic method is
also its strength. In applying this method, it is not necessary to have
recourse to intuitions, or other vague categories. Moreover, although it is
true that economic models (especially those presented above) simplify
significantly the modeled reality, there is no reason why they could not be
extended, taking into account all these important elements highlighted
by our case (illustrated in Section 3.3.2). Still, the presented method,
ANALYSIS
105
based on the principles of economics, displays consistency and is also of
some consequence. If the economic method is applied only in the sphere
of private law, then two kinds of utility, or justice (one for private law,
and another for other areas of law) must be justified. From this point of
view it seems only reasonable to apply the economic method in any area
of law, and to any legal issue whatsoever.
3.4
3.4.1
SUMMARY
Features of Analysis
In this chapter two special methods of analysis have been discussed:
linguistic and economic. Economic analysis can relatively easily be
identified as analysis3, i.e., analysis as translation. A proponent of Law
and Economics attempts to “translate” the case that interests her into the
language of economics and interprets the obtained result. It is difficult
to classify linguistic analysis in a similar way. It displays the features of
analysis3, but in some contexts also of analysis1 and analysis2.
It is appropriate to point out now the most important feature of the
presented methods and, more generally, of any kind of analysis. In analysis1 one seeks logical reasons for the analyzed sentence; the reasons have
to be self-evident or accepted earlier on some basis. Analysis2 leads to
decomposition of a given entity into more basic elements. Finally, analysis3 aims to translate the “interpreted case” into a language, which is simpler, clearer, “more basic”. This key feature of any analytic method can
be, somewhat broadly, expressed in the following way: any analysis leads
to reducing (expressing) the analyzed case (example) to a certain chosen
conceptual scheme (the conceptual scheme thesis). In the case of economic
analysis, the scheme in question is the conceptual scheme of contemporary economics. Linguistic analysis, in turn, reduces analysanda to the
conceptual scheme of ordinary language. The conceptual scheme thesis is,
as has been observed in describing the two methods, both the weakness
and the strength of analysis. It is a weakness because it is easy to object
to the selection of a “chosen” conceptual scheme as an arbitrary decision.
It is a strength because the choice of such a scheme makes analysis a well
determined method, the assumptions of which can be easily identified
and, the results of which can be estimated similarly easily.
3.4.2
Analysis in Law
It is difficult to assess the possible applications of analysis in law.
From what has been said so far it follows that, in legal reasoning, at least
analysis2 and analysis3 are of certain value. Analysis1, i.e., the search for
106
CHAPTER 3
logical reasons, cannot be straightforwardly applied, although one cannot exclude it (the similarity between analysis1 and the method of presuppositions has already been remarked upon). The examples discussed
above, in which analytic methods are applied to law show that lawyers
find it hard to accept a single conceptual scheme as “the chosen” one. It
is relatively easy to find legal applications of various domain-related
analytic methods; at the same time it is maintained, or at least assumed,
that there exists a pluralism of conceptual schemes. Lawyers construct
their arguments taking advantage of economics, common sense, ordinary language, ethics, etc.
A good illustration of this is found in the various “theories” of legal
dogmatics and legal practice. Precepts of civil law (e.g., concerning property), of criminal law (the structure of crime), and of constitutional law
(construction of the proportionality principle) can all be treated as
instances of analysis (analysis2 or analysis3). They are analyses, however,
that are confined to a specific domain and make use of various “conceptual schemes” (usually referring to the vague category of common
sense); in other words they do not form part of any wider project which
analyzes the entire legal system within the framework of a unique, chosen conceptual scheme.
This suggests that what lawyers do cannot be called analysis (it is not
possible to have an analysis without a chosen conceptual scheme).
Naturally, this is only a descriptive diagnosis: a statement of how things
are, rather than a statement of what lawyers should do. On a normative
level one can support the application of “full blooded” methods of
analysis. However, the idea of using only some analytic tools in legal discourse, which would be easier to accept by the lawyers than a “full
blooded analysis”, leads to some theoretical problems. One can maintain, of course, that various analytic methods should be used (locally) for
constructing arguments. But in that case, a new theory is needed, one
that is capable of comparing arguments built with the use of analytic
methods based on different conceptual schemes. Argumentation theories
may provide an answer to this theoretical challenge.
NOTES
1. Quoted after J. Hintikka, U. Remes, The Method of Analysis, Dordrecht, D. Reidel,
1974, p. 8.
2. R. Descartes, Regulae ad directionem ingenii, Rodopi Bv Editions, 1998.
3. B. Russell, Our Knowledge of the External World, Routledge, London 1993,
p. 214.
ANALYSIS
107
4. Similar types of analysis are identified by M. Beaney, in “Analysis”, Stanford
Encyclopedia of Philosophy, plato.stanford.edu. Beaney tags them: regressive analysis, decompositional (resolutive) analysis and transformative (interpretive) analysis.
5. R. Carnap, “Die Methode der logischen Analyse”, in Actes du huitième Congrès international de philosophie, à Prague 2–7 Septembre 1934, Prague: Orbis, 1936, p. 143.
6. See for instance J.M. Bocheński, “Subtelnoś ć” [Subtlety], in idem, Logika i filozofia.
Wybór pism [Logic and Philosophy. Selected Writings], PWN, Warszawa 1993,
pp. 133–149.
7. Descartes, Discours de la méthode, French and European Pubns, 1965.
8. Cf. I. Hacking, Why Does Language Matter to Philosophy, Cambridge University
Press, 1975.
9. B. Russell, Theory of Knowledge, George Allenand Unwin, London, 1984, p. 119.
10. That is how M. Beaney puts it, see his “Analysis”, op. cit.
11. J.M. Bochenski, “O filozofii analitycznej” [On the Analytic Philocophy], in idem,
Logic . . ., op. cit., p. 38 ff.
12. P.F. Strawson, Analysis and Metaphysics, Oxford, 1992, p. 2 ff.
13. See the entry “Metaphysics” in: The Concise Encyclopedia of Western Philosophy and
Philosophers, Routledge, 1992.
14. J.L. Austin, “A plea for excuses”, Proceedings of the Aristotelian Society, 1956–7, p. 125.
15. Ibidem, p. 126.
16. See Analysis . . ., op. cit., p. 15 ff.
17. H.L.A. Hart, The Concept of Law, 2nd edition, Oxford University Press 1994, p. 213.
18. J. Austin, The Province of Jurisprudence Determined, Hackett Publishing,
Indianapolis, 1998.
19. H.L.A. Hart, The Concept . . ., op. cit., p. 19.
20. Cf. J. Woleński, “Wste˛p. Harta Poje˛cie prawa” [Introdution: Hart’s The Concept of
Law], in H.L.A. Hart, Poje˛cie prawa, J. Woleński (transl.), PWN, Warszawa 1998,
p. XIX. See also J. Woleński, Issues . . ., op. cit.
21. H.L.A. Hart, The Concept . . ., op. cit., chapter X passim.
22. The Province . . ., op. cit., p. 201.
23. The Concept . . ., op. cit., p. 84.
24. Cf. J. Woleński, “Introduction . . .”, op. cit.
25. P.F. Strawson, “On Referring”, Mind 59, 1950, pp. 320–344.
26. Cf. A. Grobler, “Presupozycje” [Presuppositions], in R. Wójcicki, Ajdukiewicz. Teoria
znaczenia [Ajdukiewicz. Theory of Meaning], Prószyński i S-ka, 1999, pp. 96–105.
27. A semantics for presuppositions was developed by B. van Frassen, “Presuppositions,
supervaluations, and self-reference”, Journal of Philosophy 65, pp. 136–152.
28. J. Wolenski, “Introduction. . .”, op. cit., p. XX.
29. Cf. R. Alexy, Theory . . ., op. cit.
30. A. Ross, Directives and Norms, London, 1968.
31. T. Opalek, Z teorii dyrektyw i norm [Theory of Directives and Norms], Warszawa,
1974.
32. R. Dworkin, Taking Rights . . ., op. cit.
33. J. Austin, How to do Things with Words, 2nd edition, Oxford, 1976, p. 5.
34. A. Stroll claims that the analysis of what “does not count as normal” in order to
establish “what is the normal case” is one of the most characteristic features of
Austin’s philosophy”. See A. Stroll, Twentieth Century Analytic Philosophy,
Columbia University Press, New York, 2000, pp. 170–171.
108
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
CHAPTER 3
How to do Things . . ., op. cit. p. 16.
Ibidem, p. 41.
Ibidem, p. 54.
Ibidem, pp. 607–608.
J. Searle, Speech Acts, Cambridge University Press, Cambridge (Mass.) 1977. We
have to mention that Searle stresses that his book is a work in the philosophy of language and is not an example of linguistic philosophy. However, although Searle’s
argumentation contradicts sometimes Austin’s methodology, Searle’s debt towards
Austin is obvious. See also A. Grabowski, Judicial Argumentation and Pragmatics,
Ksie˛garnia Akademicka, Kraków 1999, 61 ff.
See for instance J. Searle, “Austin on Locutionary and Illocutionary Acts”, The
Philosophical Review, vol. LXXVII, no. 4, 1968, pp. 405–424; A. Grabowski, Judicial
. . ., op. cit., p. 77 ff.
How to do Things . . ., op. cit., p. 151.
Cf. A. Grabowski, Judicial . . ., op. cit., chapter III.
Cf. ibdem, chapter V.
Cf. R. Sarkowicz, Poziomowa interpretacja tekstu prawnego [Three Level Conception
of Legal Intepretation], Wydawnictwo UJ, Kraków, 1995.
O.W. Holmes, “The Path of Law”, Harvard Law Review 10, 1897, p. 469.
Cf. H. Pearson, Origins of Law and Economics – The Economists’ New Science of
Law. 1830–1930, Cambridge, Cambridge University Press, 1997.
Cf. E. Mackaay, “Schools: General”, in Encyclopedia of Law and Economics, http://
encyclo.findlaw.com.
Cf. L. Kaplow, S. Shavell, “Fairness vs. Welfare”, Harvard Law Review, February
2001, pp. 967–1380; R. Posner, Economic Analysis of Law, Aspen Publishers, 6th edition, 2002.
Cf. Th. S. Ulen, “Rational Choice Theory in Law and Economics”, in Encyclopedia
of Law and Economics, op. cit.
Cf. J.D. Hanson, M.R. Hart, “Law and Economics”, in D. Patterson (ed.), A
Companion to Philosophy of Law and Legal Theory, Blackwell, Malden – Oxford
2000, pp. 311–331.
Such a rule was actually formulated by Justice L. Hand in US vs. Carroll Towing Co.
case; cf. J.D. Hanson, M.R. Hart, “Law and . . .”, op. cit.
Cf. J.D. Hanson, M.R. Hart, “Law and . . .”, op. cit.
Cf. S. Shavell, Foundations of the Economic Analysis of Law, Belknap, 2004.
Quoted after E. Eide, “Economics of Criminal Behavior”, in Encyclopedia of Law
and Economics, op. cit., p. 346.
The marginal cost is the additional cost incurred for production of an addition unit of
the given good or of conducting the given service. In our example the marginal cost of
prevention is the additional cost for law enforcement, while the marginal cost of crime
is the cost of crimes that are avoided thanks to the investments in law enforcement.
Cf. S. Shavell, Foundations . . ., op. cit.
.
Cf. ibid. See also K. Paw lusiewicz, B. Brozek, “Prawo karne w świetle ekonomicznej
analizy prawa (Uwagi krytyczne)” [Criminal Law in Light of the Economic Analysis
of Law. Critical Remarks], Państwo i Prawo, 12, 2002.
Cf. S. Shavell, Foundations . . ., op. cit., chapter 24.
.
Cf. S. Shavell, Foundations . . ., op. cit.; K. Paw usiewicz,
B. Brozek, “Criminal Law . . .”,
l
op. cit.
ANALYSIS
109
60. See S. Shavell, Foundations . . ., op. cit.
61. R.H. Coase, “The Problem of Social Cost”, Journal of Law and Economics, 3, 1960,
pp. 1–44.
62. See S.G. Medema, R.O. Zerbe, “The Coase Theorem”, in Encyclopedia of Law and
Economics, op. cit., passim.
63. See D.W. Allen, “Transaction Costs”, in Encyclopedia of Law and Economics, op. cit.,
passim.
64. G. Becker, The Economic Approach to Human Behavior, Chicago, The University of
Chicago Press, 1976. p. 81.
CHAPTER 4
ARGUMENTATION
4.1
INTRODUCTION
As a philosophy of interpretation, argumentation provides the humanities, including legal theory, with methods which appeal to logic and
analysis (both discussed in the preceding chapters), as well as to
hermeneutics (to be discussed in the final chapter of this book); it therefore has a wide application. Given that this philosophy occupies a position amidst formal logic and “hard” analysis on the one hand, and “soft”
hermeneutics on the other, it is rightly described as “the third way” in the
methodology of the humanities. By means of argumentation one can
justify interpretative theses of normative character. This kind of justification is usually based upon the criteria of fairness, equity, validity, reliability or efficiency, rather than on the criterion of truth (these criteria
can be regarded as counterparts to the criterion of truth which are to
be applied in the normative sphere). Proposed definitions of the first
three of these criteria usually appeal to the notion of rationality; as
regards the criterion of efficiency, its definitions are based on empirical
(psychological) considerations rather than on the notion of rationality.
Even though argumentation rejects the criterion of truth, replacing it
with the above mentioned counterparts, it nevertheless aspires to be
consistent, both with logic and with every kind of rational analysis.
More specifically, argumentation aims both to create its own analysis –
informal logic – and to draw widely from various other methods of
analysis (ranging from linguistic to economic). However, the uniqueness
of argumentation, which enables one to distinguish it from more formal
methods, lies in its openness to other philosophies of interpretation,
hermeneutics in particular. It is difficult to refute the fact that the basic
rules determining the criteria for accepting practical discourse are justified by an appeal to intuition. The only controversial issue is what kind
of intuition is invoked when these rules are formulated – namely,
whether this intuition is purely rational (analytical), phenomenological
(assumed by the proponents of hermeneutics), or, rather, psychological.
The role of psychological intuition is particularly important in theories
111
112
CHAPTER 4
which hold that the aim of a discourse is to persuade an opponent, to
win a dispute at any price – in other words, in those theories which assert
that argumentation is to be evaluated solely in terms of its efficiency.
Furthermore, argumentation is a specifically “legal” method. A history of the most important theories of argumentation may serve as a justification of this thesis: they either arose from legal theory or were
invented with a view to being applied primarily in the field of law. Two
views are typical in this context. The first one (formulated by Perelman)
holds that legal argumentation (a judge’s reasoning) is paradigmatic for
all other types of practical reasoning. The second one (whose author is
Alexy), in turn, holds that practical legal discourse is a special example
of general argumentative discourse (it is the so-called Sonderfallthese).
Whether these views are convincing or not is of course another issue, to
which we will return in Section 4.2. It should be noted here, though, that
these views express an unequivocal conviction that a practical discourse
occupies a particular position within the framework of a general discourse and for that reason deserves separate and special treatment.
4.1.1
Philosophies of Argumentation
Contrary to popular opinion, controversy about good reasons in practical discourse began, not in twentieth-century metaethics and legal theory, but in the ancient philosophical schools. The proponents of
contemporary theories of argumentation continue to pursue the
methodological investigations begun by specialists in hermeneutics,
dialectics, topics, and – at least to some extent – even sophistry and eristic. The fact that philosophies of argumentation draw on such different
traditions and sources has given rise to an endless discussion about their
essence and the scope of their application. This discussion is usually pursued at the meta-theoretic level; in consequence, the problem of practical applications of particular conceptions is almost entirely left out of
the account. Theories of argumentation which appeal to different, not
always compatible, sources, provoke the justifiable objection of eclecticism. Some of these theories contain elements of two different
approaches – transcendental (objective) and psychological (subjective).
Moreover, some philosophers want to attribute the characteristics of formal logic to the informal logic of argumentation, and, in consequence, to
assess a practical, argumentative discourse by means of the criterion of
truth. Furthermore, one has often given to particular – key – concepts
of the philosophy of argumentation different, constantly redefined, meanings with the sole view of immediately justifying one’s thesis. Terminological confusion and the multiplication of philosophical problems have
ARGUMENTATION
113
resulted in decreasing interest in the theories of argumentation at the end
of the twentieth century. The “existence” of these theories seems to be
confirmed only by seminars, conferences and a huge number of publications. The theories, themselves, however, have not exerted any serious
influence on legal practice; research of the theories of argumentation has
remained unknown to most representatives of legal practice, and useless
to those few who have become acquainted with it. For all this, it must be
conceded that in many hard cases, in which interpretation based on logic
and analysis “is not sufficient”, and in which one does not wish to apply
relativistic (to a greater or lesser degree) hermeneutics, the only method
that guarantees certainty and objectivity is argumentation. It is, in our
view, one of the most important methodological alternatives for humanities as a whole and (above all) for jurisprudence. Accordingly, we shall
attempt to present this philosophy of interpretation, describing in Section
4.3 a variant of it which, in our opinion, can be universally applied.
Let us begin, however, with a historical digression. As was said earlier,
the problem of argumentation (together with questions concerning the
understanding, interpretation and justification of interpretative decisions)
was taken up by representatives of the ancient philosophical schools.
Hermeneutics. Problems connected with argumentation were discussed
very early on in the oldest hermeneutical theories, especially in biblical,
philological and legal hermeneutics. These theories were to provide universally valid rules concerning the interpretation and understanding of all
kinds of texts (religious, literary, philosophical and legal). These universal rules were to be used in the very process of interpretation as well as in
the context of that interpretation’s application and justification. Thus
hermeneutics, at least in the early stages of its development, was strictly
connected with the philosophies of argumentation, especially with logic,
dialectics, rhetoric and topic. Even though in the nineteenth century,
hermeneutics became fully independent of the above philosophies (there
arose general humanistic hermeneutics), its ties with them remained relatively close. In support of this claim, it is appropriate to recall the views
of Schleiermacher, Dilthey, Misch, Lipps, Betti, Gadamer, or Ricoeur
(regarding nineteenth- and twentieth-century philosophy of hermeneutics), and, for instance, Reinach and Kaufmann (regarding the philosophy
of law). Hermeneutics will be dealt within Chapter 5.
Logic. There is also a close connection between the philosophies of
argumentation and logic. Since antiquity, authors of various theories of
argumentation have appealed to logic. Logic was used either directly – as
114
CHAPTER 4
a method of argumentation – or indirectly – as one of the methodological
foundations of a given philosophy of discourse (logic, for instance, made
more precise the criteria for the acceptability of an argumentative discourse). Of course, the philosophies of argumentation appealed to various formal and more informal (soft) conceptions of logic. The resultant
dilemma can be generalized in the following way: (1) Theories of argumentation are supposed to attain maximum precision, or, to put it differently, to make possible the logical justification of theses advanced in
an argumentative discourse. This explains why these theories frequently
appeal to formal conceptions of logic, including classical logic, as well as
different varieties of deontic and modal logic. (2) However, the possibility of formal logic’s application in practical discourse turned out to be
very limited. Thus, advocates of philosophies of argumentation have at
their disposal at best less formal varieties of logic (frequently constructed
only for the purposes of these philosophies).
In consequence, every philosophy of argumentation developed its own
logic, and those logics varied greatly in terms of their value and scope of
application. Another difficulty was that in antiquity the term “logic” was
associated with other terms, which – like logic – were also important for
argumentation (including analytics – the most easily understandable –
as well as dialectics, rhetoric and topic). Besides, in those times, in at least
some contexts, the terms “logic” and “dialectics” were used as synonyms;
the point to be emphasized, though, is that in antiquity the term “logic”
was already associated with such activities as thinking, reflecting and
calculating, whereas the term “dialectics” was associated directly with
discourse, i.e. with dialogue. A more strict definition of both terms can
be found in Aristotle’s works: he asserted that logic and analytics enable
the derivation of true (apodictic) conclusions, whereas dialectics can
serve only as a tool for deriving conclusions which can be regarded as
right (i.e. probable). The reason why Aristotle treated dialectics, rhetoric
and topic as different in nature from logic, was that they constituted
a means of persuading opponents in a discussion, rather than of establishing the truth. Differences in the function of these two groups of theory underly Aristotle’s distinction between logical conclusions – based
on the criterion of truth – and dialectic conclusions – based on the criterion of rightness. This distinction is of special, if not fundamental,
importance for contemporary theories of argumentation. We pay special
attention to relations between various notions in the philosophies of
argumentation. This is for the simple reason that the differences between
these philosophies are frequently derived from the differences in the way
these relations are understood. The point to be emphasized is that giving
ARGUMENTATION
115
arbitrary – torn from context – meanings to these notions is likely to give
rise to rather fruitless academic discussions, which – fortunately – have
little effect on those philosophies of argumentation that are of real
importance.
Dialectics. Even though dialectics aspires to be a rather strict (in the
logical sense) method, it can be most aptly characterized as a “soft”
method of argumentation which aims to yield conclusions that are probably right. This is consistent with Aristotle’s view, according to which
dialectics is a method of practical philosophy appealing to the criterion
of good, rather than a method making use of the criterion of truth (the
latter criterion is used within theoretical philosophy). However, there
exist some arguments which testify to the existence of a close relationship between logic and dialectics. More specifically, argument a fortiori
and argumentum per reductio ad absurdum – frequently used in practical
discourse – are simultaneously logical and dialectical in nature.
According to Kalinowski, argument a fortiori in the form a maiori ad
minus (if one is allowed more, then one is allowed less) can be regarded
as a theorem of formal logic, provided that everything which is less
important is contained in what is considered as more important.1
Analogously, it seems that argumentum per reductio ad absurdum can be
“transferred” from formal logic to the area of normative reasoning.
Let us, however, return to history. For Socrates, dialectics was a
method of conducting a discussion (a philosophical controversy)
embracing two separate parts: negative – elenctic, and positive – maieutic.
The elenctic method (i.e. a method of refutation) involves pressing the
false thesis of a disputant to its absurd consequences. This method was
designed to purge a disputant’s mind of false views. An important component of this method was irony, which provisionally assumes a disputant’s false thesis to be true and forces the disputant – by means of
skilled argumentation – to formulate a true thesis inconsistent with the
one she initially defended (thus, this is in fact argumentum per reductio ad
absurdum). “Socratic irony” in fact rests on “the knowledge of one’s
ignorance” – a particular capacity to recognize falsity, which should be
acquired by every participant in a discussion. The maieutic method, i.e.
“the obstetrician’s method”, in turn, consists of “eliciting the truth”. The
role of a person leading a discussion is essentially to ask questions and
thereby “help the truth come to the world”. According to Socrates, what
constitutes the starting-point of a dispute is the establishment, by asking
the simplest questions, of commonly known and empirically verified
facts; more complex facts were to be established by means of analogy at
116
CHAPTER 4
a later stage of the dispute. This kind of inductive generalization led to
defining “common places” (Aristotle was to speak in this context about
topics), i.e. universally valid theses.
According to Diogenes Laertios, the term “dialectics” was first used
by Plato, who understood it to mean the activity of permanently applying reason, as well as acquiring practice in this activity. Plato uses dialectics both in order to investigate the world of ideas and to explain
phenomena. In Phedo, he comes to the conclusion that the only truly scientific way of analyzing phenomena consists of an appeal to dialectics,
rather than to teleological or causal considerations. Plato gives the term
“dialectics” a very broad meaning: in his view, it is a method of pursuing
a discourse (a conversation – a philosophical dialogue), a method of
establishing relations between different theses, and philosophy tout court.
It is philosophy tout court, because it enables one to grasp the world of
ideas and phenomena in a non-empirical way. Given that dialectics is a
science of ideas, i.e. about true autonomous beings, it can rightly be
called metaphysics, i.e. philosophy tout court. Accordingly, since dialectics explores the relations between general notions and theorems containing these notions and ideas, it is a deductive method, which gives rise
to and underlies logic in a strict sense.
Dialectics was also a central object of concern for Aristotle. He treated
his logic and analytics as preparation for dialectics. The object of logic is
merely the form of statements, whereas the object of dialectics is the
content – the substance – of statements. An analysis of form (as something which is general) should precede an analysis of content (as something
which is particular). According to Aristotle, the main goal of dialectics is
ultimately discussion, though he also asserts that dialectics may serve as a
tool for finding out the truth (in one of his works, he states that theses
should be treated logically – through the prism of the criterion of truth,
and dialectically – through the prism of the approval or the opinion of
others, i.e. according to appearance). As was mentioned above, Aristotle
distinguished logic (and analytics) – theory intended to provide true (apodictic) conclusions – and dialectics – theory intended to yield conclusions
which pass as right, or find themselves in circulation where they pass as
right. Thus, ultimately, the main function of dialectics is not to decide
whether a given thesis is true or false, but, rather, to enable an audience to
be persuaded that a given thesis is right; it is therefore a method of practical rather than theoretical philosophy. (A similar interpretation of
Aristotle’s dialectics was put forward by Schopenhauer in his Eristic.)
The philosophies of interpretation formulated in the nineteenth and
twentieth centuries often made use of dialectic methods. Philosophy of
ARGUMENTATION
117
interpretation in the nineteenth century was developed, above all, by
Schopenhauer (mentioned above), who was the author of a very influential conception of eristic dialectics. As for the twentieth century, it is
worth mentioning (besides hermeneutics, which often drew on the
method of dialectics – vide Reinach, Gadamer, Kaufmann) the views of
Perelman, who built his philosophy of argumentation on “classical” –
dialectical – grounds.
Rhetoric. Rhetoric may also confirm the interaction and close relationships between different argumentative methods and techniques. This discipline, which was originally purely philological, found its application in
various philosophies of interpretation relatively early on, thus becoming
(beside dialectics and topics) the third of the most important components
of almost every theory of argumentation. This term always referred to the
skill (art) of good, honest and reliable persuasion in speech and in writing. To be numbered amongst the most eminent representatives of ancient
rhetoric are – from Greek philosophy – Gorgis, Isokrates, Aristotle,
Demetrius from Faleron, Dionysus from Halocarnas, Hermogenes from
Tarsus as well as – from Roman philosophy – Cicero and Quintilian. Over
the course of time, the term “rhetoric” began to be used, not only in reference to oratory art, but also to the theory of prose, manuals of public
speaking, a sort of pedagogical system, and of course, a certain type of
philosophy of argumentation. The division of rhetorical disciplines corresponds to the structure of preparing and making a speech. The first
stage (invention) involves collecting arguments that are relevant to
a given case; the second one (composition) consists in constructing a
speech which must be adapted to the given case; the third one (elocution)
involves determining the requirements of style, as well as the rules and
conditions of correct and precise expression – through the medium of
language – of one’s thoughts; the fourth stage (mnemonic) concerns
memorizing the text of a speech; and the final stage (actio) encompasses
all the techniques of delivering a speech (specifying how to impress the
audience by one’s voice, gestures, posture, mimic). The goal of a speech –
which was an efficient persuasion – was to be attained by simultaneously
affecting the reason, will and emotions of the audience. The following
principles were ranked amongst the basic prescriptions of the art of rhetoric (i.e. the art of properly shaping “a persuasive message”): the principle of limitability (a speech should be an inherently coherent “living
organism”); the principle of adequacy – of “a rhetorical tact” (a speech
should take account of all relevant circumstances); the principle of functionality (a speaker should make use of rhetorical means of persuasion
118
CHAPTER 4
in a reflective and purposeful way). Finally, the kind of speech to be
delivered determined the “repertoire” of rhetorical techniques to be used
(for instance, the following kinds of speeches were distinguished: encouraging, discouraging, accusing, defending; and, in addition, laudatory and
circumstantial speeches, such as, welcome and farewell speeches).2
For Aristotle, rhetoric is the art of finding adequate means of persuasion
in every situation. One may even say that rhetoric and dialectics are in
fact the same discipline, since both share the same goal – to convince by
means of speech – and both appeal to the same criterion of evaluation –
efficiency. It should be emphasized, though, that rhetoric emphasizes
this criterion more strongly than dialectics. Dialectics also appeals to the
criterion of rightness, and even to the criterion of truth (since it retains
close relations with logic and analytics). Developing Aristotle’s definition of rhetoric, Perelman states the goal of rhetoric to be the analysis of
discursive techniques designed to elicit or strengthen support for theses
submitted to an audience for acceptance. This means that rhetoric is a
way of convincing by means of discourse, rather than by means of the
truth. For that reason, according to Perelman, rhetoric sensu largo also
embraces dialectics and topics. Rhetoric cannot be identified with formal
logic, since it is not possible to prove the truthfulness of premises which
are used in the process of discourse. In a practical discourse, the degree
to which particular theses are accepted may vary, since a controversy
concerns values, or, put more precisely, the rightness of theses as
opposed to their truth. Besides, one of the main goals of a practical discourse is to convince someone to accept one’s reasons; thus, convincing
always entails convincing someone (one or many persons) – and is therefore directed at an audience.3 The rhetoric of Aristotle, Cicero and
Quintilian became a starting-point for one of the most influential contemporary philosophies of argumentation – namely, “the new rhetoric”
of Perelman (to be discussed in detail in the next section of this chapter).
Topics. Philosophy (or rather, the problems) of argumentation based
on topics is connected with rhetoric (arguably most closely), as well as
with logic and dialectics. The word “topos” meant in Greek and Latin “a
place” from which a speaker or a writer derives “an inventive material”.
A topos may be located either in an indefinite place – a thought – or in a
definite place – a sign, a symbol, a gesture, a word, a text. According to
Aristotle (who – incidentally, does not provide a definition of topos),
topoi were “elements” or “premises” out of which a dialectician could
construct his syllogisms, and a rhetorician, his enthymemes. The structure of a dialectic syllogism differs from that of a logical syllogism.
ARGUMENTATION
119
The former variety of syllogism is intended to show the rightness of
a thesis (it has therefore a distinctly normative tinge). An enthymeme, in
turn, is a rhetorical (shortened) syllogism i.e. a syllogism in which one
of the premises (an obvious one) is implicit rather than stated. It consists
therefore of two parts – a supposition and a conclusion – and discounts
a major premise owing to its being obvious (to give an example, it follows
from the minor premise “Socrates is a man” that “Socrates is mortal”;
the implicit – obvious and thereby unmentioned – major premise is “all
men are mortal”). A rhetorical syllogism may be either logical (when
both of its implicit and stated premises are descriptive sentences), or
normative (when one or both of its premises – including the implicit
one – assumes the form of a directive or a value judgment).4
Aristotle distinguished common and special topoi. He understood
common topoi – loci communes – to be “places” (in thought or memory)
referring to general (universal) issues, and constituting a starting-point,
as well as a basis, for all practical discourses; by special topoi – loci specifici or loci propriae causae – in turn, he understood “places” inherent in
a concrete case or located in a given branch of knowledge (topoi of the
latter type were frequently use in legal discourse). To summarize, topics
is a method for establishing relations between notions that are crucial
in a given discourse (these notions appear in the thesis of a discussed
problem and, later, in hypotheses that are supposed to explain and justify
that thesis).
In Topics, Aristotle examines four kinds of relations: definitional relations, relations concerning genus, essential characteristics (proprium),
and accidental characteristics (accidens). He concludes that problems
emerging in every discussion fall under at least one of these relations. In
the next eight books of Topics, Aristotle offers an analysis of mutual
relationships between notions generated by combinations of the four
types of relation. In consequence, he formulates 382 rules (he calls them
topoi) that capture general interrelations between particular categories of
notions.5 Therefore, topoi are not purely material (since they do not refer
to a concrete object or notion) – they always concern entire categories of
notions and may therefore function in a discourse as “common places”.
Put differently, topoi constitute arguments which are simultaneously universal and “non-specific”, because they do not belong to a concrete discipline (this is true with reference to loci communes, though of course not
to loci specifici), and are not a type of purely scientific (logical) argumentation. The universality and “non-specificity” of dialectic and
rhetorical syllogisms lies in the fact that they make possible – unlike classical formal logic – simultaneous argumentation in favor of two different
120
CHAPTER 4
aspects of a disputed issue. Topics – an ancillary discipline of the type
ars inveniendi – may help other philosophies of argumentation to find
good reasons (arguments, premises), thus facilitating success in an interpretative controversy.
An interest in classical topic and rhetoric arose anew in the twentieth
century, owing especially to schools of legal philosophy, which will be
explored in more detail in Section 4.2.
Eristic and sophistry. Eristic and sophistry have the same philosophical
origins (since eristic techniques were widely used and taught by sophists)
and the same goal, which is to win a dispute at any price – per fas et nefas
(i.e. by permissible and forbidden means) – without paying much heed to
the plausibility of advanced reasons. By relating these disciplines to
dialectics, Schopenhauer makes their scope even broader. Hence the reason he speaks, in his treaty on argumentation, about “dialectic eristic”.
This classification, introduced by Aristotle, takes into account not only
logical and dialectic inferences, but also eristic and sophistic ones. In the
case of eristic inferences, the form is correct, but the statements themselves are not true – they only appear to be true. As for sophistic inferences, their form is fallacious – it only creates an appearance of truth.
However, it is difficult to agree with Schopenhauer’s claim that the goal of
dialectic eristic is to prove the rightness of advanced theses. This was the
goal of dialectics (rather than dialectic eristic), which suggests that the
justification of advanced theses’ rightness was another – besides proving
truth – objective of discourse. Justifying the rightness of an advanced
thesis was the objective of rhetoric and topics, but certainly not of eristic
and sophistry. Eristic and sophistic discourse had only one goal – to win
a dispute; and each was to be evaluated only according to the criterion of
efficiency. Rightness always makes reference to some morally acceptable
good, whereas efficiency makes reference only to purely instrumental values. It is not merely by chance that neither Plato nor Aristotle held eristic
in high esteem. According to Aristotle, eristic is a dishonest way of conducting a verbal struggle in a discussion, and the relationship between an
adherent of eristic and a dialectician resembles the relationship between
a person who draws false diagrams and a geometer.6
Eristic and dialectic use an extensive repertoire of methods and techniques. The most important and frequently used “eristic catches” are: (1)
the use of eristic expansion, i.e. acting in a way that introduces chaos into
an argument to confuse an opponent, (2) introducing side plots, which
have little bearing on the discussed issue, with a view to diverting an
opponent’s attention from crucial theses, which, if developed, might be
ARGUMENTATION
121
“dangerous” for the other party, (3) appealing to the true or apparent
acceptance of a thesis by the audience (by pointing out that an opponent’s views are inconsistent with the views of the audience, irrespective
of whether such inconsistency exists), (4) fighting an opponent with his
own weapon, i.e. turning his arguments to one’s own advantage (retorsio
argumenti), (5) making use of sui generis dialectics of “thesis – antithesis”,
where an antithesis that is defended by an opponent is formulated in
such an unconvincing and absurd manner that the opponent is led to
question and reject it and thereby to assume a thesis which he initially
rejected, (6) “fabricating consequences”, meaning deriving, via fallacious
inference, theses from an opponent’s statements that those statements
did not actually contain, (7) concealing the end one really pursues in an
argument; this can be achieved, for instance, by “eristic expansion” and
making a discussion drag on, (8) responding to an opponent’s sophistic
argument with another sophistic argument, (9) requiring an opponent to
justify his self-evident theses, (10) ironically admitting one’s incompetence in order to suggest that theses advanced by an opponent are simply preposterous.7
Most (it cannot be said for sure whether all) of the eristic and sophistic “methods” stand in contradiction with the rules of rational discourse,
which serve to provide a right (and according to some philosophers even
true) solution to the matter in hand. Rightness undoubtedly constitutes
the basic criterion for evaluating a practical discourse, yet another
important criterion is efficiency. Provided that an argument fulfils necessary requirements of “minimum morality”, there is nothing to prevent it
appealing to methods which ensure maximum efficiency, as well as – if it
is at all possible – to eristic techniques (so long as these techniques do
not violate the principles of rational discourse). Consider, for instance,
retorsio argumenti, which amounts to turning an opponent’s argument to
one’s own advantage (to give an example of this argument’s application:
when an opponent says “he should be indulged, since he is still a child”,
we may reply “since he is still a child, he should be punished so that these
bad habits do not become rooted in him”). In a practical (normative)
discourse in which two or more right solutions are possible, one is justified in using this kind of argument, all the more so as it does not necessarily infringe upon other rules of the rightness of a discourse. This
observation is especially important, given that a generally negative
opinion about eristic argumentation is expressed here.8
Schopenhauer, author of the treaty Eristic, or Art of Leading a Dispute
which was devoted to eristic dialectics, initiated renewed discussion of
eristic in the contemporary philosophies of argumentation.
122
CHAPTER 4
Contemporary theories of argumentation. Contemporary theories of
argumentation were most often formulated in response to growing conflict between the most important philosophies of argumentation. The
“positivistic-analytic” and “phenomenological-hermeneutic” paradigms
took their final forms. Besides, mainly owing to psychoanalysis, psychological interpretation maintained an important position in the humanities.
Controversy over the choice of a method of humanistic interpretation
became even more vigorous in connection with discussion – pursued
mainly in the areas of metaethics and legal philosophy – concerning the
epistemological (semantic-logical) characteristics of normative statements.
In consequence, the main goal of theories of argumentation in the latter
half of the twentieth century was to find an epistemological equilibrium –
a sui generis “third way” between competing philosophies of interpretation. The philosophy of argumentation was supposed to substitute formal
logic, together with analysis (which are inapplicable or applicable on a limited scale in practical – normative – discourse) and the “softer” method of
phenomenological hermeneutics, relying upon intuition. The result of this
process was the rise of a wide variety of philosophies of argumentation.
Some of these philosophies were still formulated on the basis of analytical philosophy – in connection with controversy over “good reasons”
in ethics. An important contribution to construction of the analytical
theory of practical discourse was made by Wittgenstein, Ayer, Stevenson,
Austin, Hare, Toulmin and Baier, among others.9 Thus, for instance,
Hare bases his theory of argumentation on analysis of the language of
morals and, in consequence, introduces a distinction between the
descriptive and prescriptive (evaluative) meaning of ethical predicates.
The two main rules – the rule of universality and the rule requiring ethical statements to be prescriptive – upon which every moral argumentation rests confirm this distinction. And even though the rules of moral
discourse are different from the rules of argumentation in exact sciences,
they appeal to the same criteria of rationality.
An investigation devoted to the use of constructivist method (logic)
for the needs of practical advice was conducted by Lorenzen and
Schwemmer. According to Schwemmer, “constructivist ethics” (rational
moral discourse) relies upon two fundamental principles: the principle of
reason (Vernunftsprinzip) – called also the principle of advice
(Beratungsprinzip), and the moral principle (Moralprinzip).
As for Habermas, he found a basis for justifying the rational conception of discourse in the consensual theory of truth. This theory enabled
him to distinguish an action and a discourse (a dialogue). He treats discourse as a process of rational communication which takes place in a
ARGUMENTATION
123
communicative community and, in addition, in an ideal situation of
speech, which enables the participants of a discourse to arrive at “right
conclusions”, i.e. at an agreement (a consensus). Whilst theoretical discourse is to be evaluated according to the criterion of truth (since in this
kind of discourse empirically verifiable and logically decidable theses are
being formulated), a practical discourse is to be evaluated through the
prism of rightness (since in this kind of discourse, attempts are made to
justify the rightness of normative statements). However, the point to be
stressed is that what constitutes the final criterion both of truth and rightness is consensus legitimated by the “force of the better argument” and
reached in the process of rational language communication. This process
can be said to fulfill the demands of rationality only if it observes some
formal rules such as, for instance, the rule of the equality of participants
in a discourse (the members of a communicative community), the rule of
freedom of argumentation, the rule of veracity, and the rule canceling
privileges, and the compulsion of any participant in a discourse.10
In a relatively short time, the discussion concerning rational discourse
had permeated the field of legal philosophy. This process was mainly due
to Viehweg, the author of Topik und Jurisprudenz (published in 1954) as
well Perelman and Olbrechts-Tyteca, the authors of La nouvelle rhétorique. Traité de l’argumentation (published in 1958). The latter was the
first of a whole series of works devoted to the problems of legal argumentation and written by the representatives of the Brussels school
(especially by its leader, Perelman). One work that made a considerable
contribution to the maintenance and revival of the diminishing interest
in problems of legal argumentation was Alexy’s Theorie der juristischen
Argumentation. Die Theorie des rationalen Diskurses als Theorie der
jurstischen Begründung. Classical philosophies of argumentation, such as
ancient topics and sophistry, provided a point of departure both for
Viehweg and for Perelman. As for Alexy, he draws on different sources,
in particular: the philosophy of Kant (the Kantian conception of practical reason), analytical philosophy and Habermas’ discourse theory. As
a result, his procedural theory of legal discourse has little in common
with the ancient philosophies of interpretation. Both conceptions of
discourse will be discussed at greater length in the next section.
4.1.2
Criteria of a Practical Discourse
A fundamental problem of every philosophy of argumentation concerns
the choice of which criteria are to be used to “measure” a discourse, and
thereby decide whether it should be accepted or rejected. It is suggested
that this problem is not only the most important, but also the most
124
CHAPTER 4
controversial one. First, a given criterion is at issue, next – the scope of
its application: the problem here is whether a given criterion should be
used only in the process of internal justification – interne Rechtfertigung,
only in the process of external justification – externe Rechtfertigung, or
in both (internal justification questions whether a proposed solution follows logically from premises assumed in a discourse, whereas external
justification aims to assess the rightness of the very premises).11
Let us, however, return to the problem of the criteria of a practical discourse. The earliest theories of argumentation were intended to provide
an answer to the question of when a given discourse should be accepted
or rejected. This question was answered in at least three ways. According
to some philosophers, the only acceptable and final criterion of an argumentative discourse is truth: if one can demonstrate that a proposed
solution to a controversy is true, then acceptance of this solution is not
only possible, but rather demanded from all the participants of a discourse. Other philosophers opted for a “softer” criterion – namely, that
of rightness (a winning solution is one that can be demonstrated to be
the most right – just – one). Having given up the claim to logical certainty (truthfulness), these philosophers wanted to relate the criterion of
rightness to rationality (what is right in the ethical sense, must be rational
as well; and conversely, a rational solution should be accepted as right by
all participants in a discussion). Still other philosophers held that the
only justifiable and valid criterion for evaluating a practical discourse is
efficiency. A total separation of the criterion of efficiency from the criteria of truth and rightness led to a radically instrumental and relativistic
understanding of argumentative discourse, which found its expression,
for instance, in sophistic and eristic conceptions.
Truth. It was within those philosophies of argumentation which made
allowance for the methods of logic and analysis that truth was regarded
as a basic criterion for evaluating solutions advanced during a discourse.
Aristotle had already pointed out that logic and analysis should serve as
tools for deriving true (apodictic) conclusions. The situation of dialectics
is more complex: it is to serve either truth alone (Plato), or both truth
and rightness (Aristotle) – in the latter case, dialectics is a theory
intended to yield conclusions which pass as valid or find themselves in
circulation where they pass as valid (probabilia).
It should be stressed that the relationship between logic and dialectic
has always been a close one (an analysis of a certain type of argument
formulated within normative logics shows this clearly, as was shown
in the above discussion of dialectics). Truth returns as a criterion for
ARGUMENTATION
125
evaluating argumentative resolutions in contemporary conceptions. The
conviction, held by some representatives of the philosophies of argumentation, that this criterion can be applied may be a result either of the
adoption of a cognitive view of the meaning of normative predicates
and statements (directives and norms), or of some “moderately” noncognitive views. If we assume that directives (norms) are statements
which can be assigned logical values, then, of course, we can also assume
that argumentation, invoking these kinds of statement, can be assessed
by means of purely logical categories. In some cases, however, even if we
deny that directives (norms) have cognitive sense, we are not thereby
forced to forgo entirely the possibility of constructing certain types of
logic (formal or informal) which might be used as a tool for constructing
a criterion – based on truth – for evaluating a practical discourse.
Without getting involved in philosophical controversies which, in our
view, cannot be resolved, we wish to point out that in the process of
argumentation people very often appeal to theses and arguments that are
simply true or false. These theses and arguments, however, are being formulated within a discourse that can be called theoretical. Theses and
arguments of normative nature are formulated within the other kind of
discourse – namely, practical. As usual, this kind of discourse appeals to
other criteria: rightness and efficiency. The necessity of making a distinction between these two kinds of discourse was understood by
Aristotle, who distinguished theoretical and practical philosophy. Let us
recall that theoretical philosophy is based on the criterion of truth,
whilst practical philosophy is based on the criterion of good – rightness.
Also Kant distinguished, and even placed in opposition, two cognitive
powers of the transcendental subject – namely, theoretical (scientific)
reason, which appeals to the criterion of truth, and practical (normgiving) reason, which appeals to the criterion of formally understood
rightness. A distinction between the two kinds of discourse can also be
found in Habermas’ works. In his view, a theoretical discourse is measured by truth, whereas a practical discourse is measured by rightness, yet,
ultimately, both discourses appeal to the same criterion, which is consensus reached in a rational way.12
In our opinion, the process of legal cognition (broadly understood as
interpretation) encompasses two interlinked discourses: theoretical and
practical. By mixing theses formulated within each type of discourse,
participants in a discussion often become embroiled in heated, though at
the same time fruitless, controversy. Multiple argumentative conclusions
are fully discursive, cognitive and logically verifiable. We have at our disposal a whole repertoire of scientific means – both strictly formal (such
126
CHAPTER 4
as, for instance, logic or “hard” analysis) and empirical. As is well
known, there should be no controversy over logically provable and
empirically verifiable conclusions. Thus, controversy may arise only as a
result of incompetence or opportunism, becoming in consequence an
expansive – eristic – sort of argumentation that aims to yield success at
any price – even at the expense of truth. Confusion in argumentation will
become aggravated if cognitive theses are given a normative sense (that
is to say, if we discuss the rightness of statements which we have already
recognized as unquestionably true), or if normative theses (that is to say,
theses that are to be evaluated by means of other criteria than the criterion of truth) are given a fully cognitive sense. For that reason, one
should thoroughly examine a thesis formulated in the process of a discourse before one decides whether it can be evaluated in terms of its
truthfulness and falsity or, rather in terms of some other criteria – such
as rightness, justice, validity, reliability, or efficiency. In our view, only the
latter kind of theses can be the object of a “proper argumentation”, i.e.
a practical discourse.
At this point, we would like to turn our attention to some terminological issues. In discussing the concept of argumentation from a historical
perspective, we have associated it with such terms as hermeneutics
(though in a rather limited sense), logic, dialectics, rhetoric, and, finally
sophistry and eristic. All these terms identified various philosophies,
methods and techniques directly connected with an argumentative
process (activity). At the same time, the concept of discourse has been
used to analyze its two varieties – theoretical and practical. One could
justifiably introduce further distinctions – for instance, between general
and special discourse (the latter type of discourse embraces legal discourse, among others). The concept of discourse, however, may give rise
to many doubts on account of its ambiguity (which is increased by the fact
that it is “fashionable” in the sense that it is invoked by many areas of
knowledge to explore essentially different processes). It seems difficult –
if not impossible – to give a sound definition of this term: it would be
difficult to build an analytic definition (i.e. reflecting the received –
historical – sense) of the concept of discourse, since that would require
that a whole range of intuitions concerning the meaning of this concept
be taken into account. Accounting for all these intuitions would almost
certainly make the concept fuzzy. More specifically, in the process of creating such a definition, it would be necessary to allow for the fact that the
concept of discourse refers to cognitive process, communication, logical
argumentation, discussion, making a speech, convincing through the
medium of speech etc. Thus, the concept was used to designate both
ARGUMENTATION
127
general processes of cognition and communication, as well as special
activities connected with ways of leading a discussion.
On the one hand, given the lack of a plausible criterion for choosing
between these meanings, it must be conceded that the choice of only one
of them would have to be qualified as arbitrary and therefore unjustifiable; on the other hand, a definition that accounted for many different
meanings would be useless for more specialized kinds of analysis and, in
addition, inconsistent with the principle of the economy of argumentation. A synthetic (stipulative) definition of the concept of discourse, in
turn, would be arbitrary, for the simple reason that it would have to be
based on only one intuitive view (that of the author) of which of the
concept’s many plausible meanings is “the most plausible”.
Rightness. It follows from the preceding considerations that it is rightness that should constitute the measure, standard, or criterion for evaluating a practical discourse. If it is really the case that a practical discourse
(argumentation) cannot be “measured” by means of the criterion of
truth, it is necessary to adopt some other measure or criterion, such as
that of rightness, for example. Of course, the concept of rightness is by no
means clear, and, accordingly, in trying to provide its definition, we
encounter the same problems we have run into when examining the concept of discourse. Thus, first, we have at our disposal many other concepts
which in some cases can be used as its synonyms, especially the concepts
of rationality (i.e. “trans-logical” rationality which is precisely rightness),
fairness, validity, reliability, and even efficiency if it can be legitimated in
a rational way). Second, rightness is always associated with certain moral
values, which determine some “ethical minimum” which must be complied with for every practical discourse to be possible. Third, the concept
of rightness may be interpreted materially or formally, i.e. procedurally
(the latter interpretation seems more adequate if rightness is to be used
as a criterion for evaluating practical discourse). Fourth, notwithstanding all the reservations outlined, rightness may be connected with truth,
because both discourses – practical and theoretical – are interlinked. In
consequence, as was emphasized by Habermas, truth in a sense legitimates rightness and rightness, in a sense, legitimates truth. Fifth, rightness may also be conceived of as efficiency. It may plausibly be argued
that only what is right may work efficiently – bring about real effects (we
mean here rational efficiency rather than “efficiency at any price”, which
is characteristic, for instance, of eristic). It is particularly evident in
the case of economic arguments, which are frequently put forward in
practical discourse: they imply that for a solution to be right it must be
128
CHAPTER 4
economically efficient, i.e. that one should not sacrifice efficient solutions
for the sake of solutions which realize some abstract ideal of fairness or
rightness, yet are impossible to realize. Sixth, rightness can be replaced
with other (most often formally – procedurally understood) criteria, for
instance those of reliability or validity, on condition that these criteria
have a moral dimension. Seventh, what is of special significance for the
understanding of the criterion of rightness are its connections with the
concept of rationality, especially when this concept is given more formal
and procedural interpretation; accordingly, those things are right which
are simultaneously – from the standpoint of formal rules of the accepted
procedure – rational.
It seems that evaluation of the majority of theses formulated in the
course of argumentative discourse is not possible unless the criterion of
rightness is applied. We are, so to speak, doomed to apply this criterion,
given the impossibility of directly applying strictly logical criteria, which
would justify – by demonstrating their truthfulness – solutions adopted
in discourse (these solutions can simply be directives or norms, or take
the form of special kinds of normative statement). In addition, we also
have at our disposal the instrumental criterion of efficiency; however, the
use of it beyond an ethical context (without making due allowance for
the requirements of rightness) may lead to multiple abuses in argumentation and, in consequence, to flagrantly unfair resolutions in a dispute.
The thesis that other criteria – such as, for instance, fairness, rationality,
validity or reliability – enable the construction of a more precise criterion
for evaluating practical discourse is not very plausible: A “new word” may
be gained, but the same definitional problems that plague the discussion of
rightness remain.
Efficiency. Undoubtedly, efficiency can be one valid criterion for evaluating legal discourse, since what is ultimately at stake is success in an
argumentative dispute. The problem boils down to the question: should
success be achieved at any cost? Difficulty in offering a precise definition
of this criterion stems from the fact that it can be understood in at least
three different ways.
First, it can be interpreted in a purely formal and rational way.
Efficiency in a practical discourse would be attained by applying a previously accepted procedure. In consequence, the result would be both fair
(because consistent with the requirements of a procedure accepted by all
members of a discourse) and rational (for the same reason). A procedure
determines both the formally understood ethics and logic of a discourse.
Efficiency, understood in this way, appeals to the concept of instrumental
ARGUMENTATION
129
rationality (the result of a practical discourse is legitimated by a procedure applied in a given case of argumentation). This is precisely how the
problems of discourse and rationality (as a criterion for evaluating the
results of discourse) were conceived in theories of communication
(meaning the theories of Habermas and Apel), and in theories of system
(above all Luhmann’s theory13). The above interpretation of the concept of
efficiency does not rule out the possibility of associating the concept
with the criterion of rightness (as formally, or procedurally, understood),
which we analyzed earlier.
Second, the criterion of efficiency may be referred to empirical reality
rather than the social world presupposed by our normative considerations. Accordingly, a practical discourse can be described as efficient if it
functions in the real – empirical – world. It was due to utilitarianism that
this account of efficiency was introduced into philosophies of argumentation and then developed by the representatives of pragmatism and
American legal realism. These three views – utilitarianism, pragmatism
and American legal realism – suggest that actual law (law in action) is
the law that produces real effects in a given social sphere. Thus understood, the criterion of efficiency certainly narrows the criterion of rightness, though it does not necessarily contradict it. Not every right (fair)
solution is efficient, but every efficient solution is right (fair), since it
realizes some values which are fundamental from the point of view of
utilitarian, pragmatic or realistic ethics.
It becomes easier to examine this interpretative tradition when one
analyzes legal cases. It may be the case that law (a legal verdict) which is
fair (rational in a metaphysical sense) requires the impossible, or that law
(a legal verdict) which is fair (both in a metaphysical and instrumental
sense) turns out to be totally inefficient (for instance, from the economic
point of view). What is to be done in such a situation? According to the
advocates of the presented view, one should appeal to the criterion of
efficiency. Efficiency understood in this way does not have to be (and, in
fact, was not in the philosophies discussed) an ethically neutral category.
These philosophies only assumed a different hierarchy of values, according to which the most fundamental values were those which enabled the
construction of an empirical criterion for assessing the results of a practical discourse. For the representatives of utilitarianism this value was
pleasure (happiness), for the adherents of pragmatism – utility, and, for
instance, for the proponents of economic analysis of law – social wealth.
Conflict between the criteria of rightness and efficiency might arise only
if the latter criterion were entirely separated from an ethical context (i.e.
values understood in a material or formal way) and used in a purely
130
CHAPTER 4
instrumental – operational – way. Of course, in such a case priority
should be given to the criterion of rightness.
Third, rightness may be referred only to a positive result, that is to say,
to the assumption that an argumentative dispute has to be won at any
price (per fas et nefas) – by using permissible and forbidden methods – in
accordance with the principle “the end justifies the means”. Success in an
argumentative dispute was to be achieved by different sophistic and eristic techniques and methods, as well as by psychological artifices proposed within some contemporary philosophies of argumentation. Thus
understood, efficiency has no ties with ethics – a practical discourse is
not to be evaluated through the prism of fundamental values. Interesting
in this context, is Perelman’s distinction between two kinds of activity
undertaken in the course of argumentation; these activities – convincing
and persuading – are correlated with two types of audience, respectively,
universal and particular. In those philosophies of argumentation which
appeal to morality, it has always been emphasized that the function of a
discourse is convincing, which is – in the last instance – directed at some
“ideal audience”. Convincing, relies both upon rightness and rationality,
upon, as Perelman put it, the objective “validity of an argument”: it is
the main – and what’s more, “ethically active” – goal of a practical discourse. Persuasion is a little different as it can only be evaluated using the
criterion of efficiency, understood in a purely instrumental and subjective sense. Persuasion is directed at a particular audience, at concrete
people upon whom one attempts to enforce some solution, irrespective
of whether this solution satisfies even the minimum requirements of
rightness and rationality. Hence, in the case of a particular audience,
argumentative theses are regarded as justified even when accepted only
by part of the auditorium. Thus, a situation may emerge where, in spite
of errors (or even abuses) committed in a practical discourse, one may
achieve the final result, that is to persuade part of the audience to accept
the proposed solution, thereby succeeding in the argumentative dispute,
in defiance of the requirements of rightness and rationality. This is the
reason why it is argued here that a narrowly – instrumentally and subjectively – understood criterion of efficiency cannot be used automatically as the measure of a practical discourse.
4.2
TWO CONCEPTIONS OF A LEGAL DISCOURSE
It is possible to engage endlessly in controversies surrounding practical
discourse’s reception of the philosophy of argumentation, multiply divisions and classifications. However, the discussion here will be confined to
ARGUMENTATION
131
two, important and characteristic, conceptions of practical discourse: (a)
procedural; (b) topical–rhetorical. Although these conceptions appeal to
somewhat different philosophies of argumentation, they are nevertheless
complementary to each other. In order to speak reasonably about legal
discourse, it is necessary to explore not only the problem of the reception
of philosophies of argumentation directed at legal discourse, but also
(above all) the problem of the autonomy and specificity of this discourse,
i.e. the problem of the relationship between general discourse and legal
discourse.
According to the first view of this relationship, there exists only one –
universal – general discourse. It is within the framework of this discourse
that universally valid and rules are formulated to be applied later in all
kinds of argumentation. Thus, there exists only one philosophy of argumentation, and its different applications (this kind of situation exists in
hermeneutics, especially in its phenomenologically oriented variety, discussed in the following chapter). On the above understanding of philosophy of argumentation, the distinction between general discourse and
particular discourses essentially loses its significance, retaining only a
didactic meaning.
The second view describes these relations from the opposite perspective. The “argumentative reality” is the reality of concrete (particular) discourses – the only discourses that actually exist. Advocates of this view
often assumed that a certain type of discourse, say legal or ethical, is of
paradigmatic significance for all other practical discourses. According to
Perelman, the paradigmatic discourse is the legal one. He asserts that the
reasoning of a judge is exemplary, not only for other types of legal reasoning, but also for other – particular – practical discourses. Yet whether
Perelman’s view is correct is a matter of some dispute. In our opinion, the
reasoning of a judge should be regarded as an “exception” rather than an
“exemplary case”, i.e. generalizable, “universal pattern”. It is difficult not
to notice the specificity of the reasoning of a judge even with reference to
legal discourse: the judge is a fully autonomous arbiter, rather than a participant in a discourse who possesses the same – equal – rights as other
participants. Participants in other particular discourses hardly ever make
use of legal argumentation (and – a fortiori – of the type of the reasoning
applied by a judge), one reason being that they lack the required legitimation (the specificity of legal discourse, which lies, in particular, in its
close connections with valid law, make this kind of argumentation inaccessible to representatives of other humanistic disciplines).
Finally, according to the third view, legal discourse is a special case of
general practical discourse – this view is the so-called Sonderfallthese,
132
CHAPTER 4
formulated most distinctly by Alexy. He distinguishes three interpretative
theses concerning the fact that legal discourse is a special case of general
practical discourse: Sekundaritätsthese, Additionsthese, Integrationsthese.14
We shall discuss them briefly. Sekundaritätsthese, i.e. a thesis about “the
façade character of legal discourse”, asserts that, in cases which cannot
be decided exclusively on the grounds of the rules of valid law, it is
general practical discourse that constitutes the real basis for making a
decision; the role of legal discourse is to provide only “secondary legitimation” and thereby conceal “behind the façade of valid law” the real
reasons for the decision. Additionsthese, i.e. a thesis about “the complementary character of general practical discourse”, implies that legal
argumentation is sufficient only up to the point where specific legal arguments are exhausted and, in consequence, it becomes necessary to supplement them with arguments from general practical discourse.
Integrationsthese, a thesis about “the integrality of discourse”, says that
specific legal arguments should be used in conjunction – at every stage of
a given discourse – with arguments from general discourse. Sonderfallthese, the thesis presented here, also gives rise to many doubts: note, for,
instance, that the first two theses are descriptive, whereas the third one is
arguably normative. True, it is hard to call into dispute the view that legal
discourse is specific, yet it is not clear in relation to what it is specific (to
a general discourse, or rather to other, particular practical discourses?).
Furthermore, considering the fact that a general discourse can hardly be
applied independently of some particular discourse, it may plausibly
be argued that a general discourse does not exist as such – it exists only in
concrete applications, that is to say, as some practical discourse. Only
some general rules (common places – loci communes) have, so to speak,
independent existence: they constitute a recurring element of every
particular practical discourse and thereby fulfill the role of peculiar
“argumentative axioms”.
The above reservations notwithstanding, it seems that the third view
is – on certain conditions – acceptable; we wish to stress, though, that
should some additional presuppositions be accepted, the first view may
also be defended. Contrary to appearances, these two views can be easily
reconciled with each other.
4.2.1
The Topical–Rhetorical Conception of Legal Discourse
In the second half of the twentieth century many attempts were made to
make use of ancient philosophies of argumentation for the purposes of
the law. Intensification of the controversy over choosing a method
of jurisprudence, the growing criticism – especially after 1945 – of legal
ARGUMENTATION
133
positivism, as well as disappointment with the methodologies proposed
by analytical philosophy and hermeneutics, naturally contributed to
increased interest in theories of argumentation. These theories were to
enable construction of a philosophy of interpretation which would constitute a methodological alternative for legal theory – the sui generis
“third way” between analytical philosophy and hermeneutics. It is presumably not by chance that the first conceptions of legal argumentation
(those of Perelman and Viehweg) drew from the reliable traditions of
ancient topics and rhetoric. It should be noted, though, that over time, the
connections between legal argumentation and this tradition became
looser (it is clearly visible both in later works of Perelman as well as in the
German current Methodenlehre) and, accordingly, theories of legal argumentation started to be regarded as “specifically ‘legal’ methodologies”.
Perelman. In 1958 La nouvelle rhetorique: Traité de l’argumentation by
Perelman and Olbrechts-Tyteca was published. The authors treated this
work as the renewal (reception) of the tradition of ancient rhetoric, especially as represented by Aristotle, Cicero and Quintilian. In his later
works, Perelman frequently returned to issues connected with classical
topics and rhetoric, yet he introduced multiple changes to solutions
advanced in antiquity, especially by Aristotle. Both the continuity and
change can be easily tracked, for instance, in his later work, Logique
juridique. Nouvelle rhétorique, published in 1979.
Perelman considers topics to be an essential element of every possible
theory of argumentation. In particular, he scrutinizes the Aristotelian
problem of relations between common and special places. “Common
places” (loci communes) – viewpoints or values to be taken into consideration during every discussion – enable a speaker to formulate theses,
rules or maxims to be used in the process of a given discourse (usually in
its initial phase). “Common places” stand in the same relation to “unspecialized reflections” as “special places” (loci specifici) stand in relation to
special disciplines. To give an example, general principles of the law are
only loci specifici of the law (as a special discipline), whereas the most
general theses (like those analyzed by Aristotle in his Topics) are a point
of departure for “an unspecialized reflection” and fulfill in every discourse a role analogous to the role fulfilled by axioms in a formal system.
Having chosen “common places”, a speaker must see to it that they
become “present” in the consciousness of her interlocutors, or of the
audience. This goal is to be realized by the various techniques (above all
rhetorical ones) of a discourse. An especially important role in bringing
“common places” to the consciousness of the audience is played by
134
CHAPTER 4
rhetorical figures which include reinforcing (an oratory development of
a theme), repetition, using apparently indirect speech, visualizing (an
event is described suggestively so that it may be “seen” by the audience)
and – finally – reversing tenses (this technique frequently leads to violation of the grammatical rules concerning the sequence of tenses, yet it
enables a speaker to increase the effect of her argumentation).15
Perelman tries to describe the main types of legal topic, making a separate catalogue of arguments and legal principles.16 Both types of topic
fulfill the same role in legal discourse – that is, they enable the conducting
and completion of an argument, and the interpretation of valid law.
Topics do not possess a strictly logical structure, since, as Perelman points
out, they do not refer to the form, but to the object of reasoning; in other
words, they help establish principles on the basis of valid law. Perelman
(following Tarell) numbers the following arguments amongst such quasilogical arguments (i.e. arguments in a narrower sense): a contrario, a
simile (or per analogiam), a fortiori, a completudine, a cohaerentia, per
reductio ad absurdum, teleological, ab exemplo, systematic, naturalistic,
psychological, historical and economic. He also analyzes a catalogue
(drawn up by Struck) of 64 principles of the law (legal topics).17
Especially interesting in the context of Perelman’s consideration of
topics, is the question of relations (which we have already discussed)
between “common places” and “special places”. Specifically, it should be
noted that an apparently distinct line between both types of topic (general and special) is no longer easy to draw in the case of legal discourse.
This is due to two reasons. First, legal discourse depends on a whole
range of external circumstances connected, inter alia, with the complexity
level of an interpreted case, tradition and even the psychological situation of an interpreter. These circumstances will influence the choice of
certain principles (which in point of fact make it possible to enter upon
argumentation). It may turn out that the same topics will be regarded in
some situations as general topics – unspecialized – loci communes, and
in other situations – as special (detailed, specialized) topics – loci specifici.
It is the context and time of use that will decide whether the same
principles are used either as loci communes or as loci specifici.
Second, multiple legal topics (arguments and principles) are of universal nature. They can be used also in other practical discourses, where
they fulfill the role of unspecialized general principles (this observation
applies to such arguments as, for instance, a simili, a contrario, or a fortiori) or principles: pacta sunt servanda (contracts ought to be observed),
clara sunt interpretanda (what is clear requires no interpretation) or nemo
iudex idoneus in propria causa (no one can be a good judge in her own
ARGUMENTATION
135
case). Thus, one can hardly agree with the view that these topics are of a
specifically legal nature and thereby constitute “special places” (loci
specifici) only in legal discourse. According to Perelman, rhetoric is a
technique of legal discourse aimed to bring about the “effect of presence” i.e. to make topics “present”, both in the consciousness of interlocutors and the audience. However, the role Perelman attributes to
rhetoric goes much further: it is on the basis of rhetoric that he builds his
whole conception of argumentation. This “new rhetoric” decidedly
exceeds the scope of rhetoric as understood in antiquity. It is by means
of the “new rhetoric” that Perelman builds a procedurally oriented theory of legal argumentation (to be discussed in Section 4.2.2).
According to Perelman, the main task of rhetoric is analysis of those
discursive techniques designed to elicit or strengthen support for theses
submitted for discussion. As mentioned earlier, Aristotle defined rhetoric as the art of finding usable means of convincing in every situation.
Now, Perelman accepts this definition, but he supplements it by adding
four detailed theses of the nature of rhetoric: (1) The goal of rhetoric is
to convince by means of a discourse. In the process of a discourse, one
is allowed to make use of the techniques of rhetoric sensu largo (which,
in addition, embraces topics, dialectics and all the other techniques
applied during disputes and discussions), (2) Rhetoric makes no use of
formal logic. In the process of a practical discourse, one does not
attempt to locate truth (since it is not possible to prove the truthfulness
or falsity of normative statements), but, rather, to convince the audience,
(3) Truth – as an objective category – can be described as “impersonal”,
whereas convincing can be described as “personal” (since the act of
convincing is always directed at a person or persons, i.e. to some audience),
(4) The category of convincing can be graded. The degree to which a thesis is accepted may vary where a dispute concerns values other than truth.
What is more, the degree to which a thesis is accepted may depend on the
type of audience to which the argumentation is directed (the audience may
be either universal or particular; it should be noted, though, that it is necessary to persuade, rather than convince, the latter type of audience).18
As has been illustrated, rhetoric and topic – two formerly distinguishable philosophies of argumentation – were tied by Perelman so that they
now form a coherent whole – a new theory of legal argumentation,
which, as will be shown, can also be interpreted procedurally.
Viehweg and the current Methodenlehre. Topik und Jurisprudenz – a book
by Viehweg published in 1954 – gave rise to discussion about the possibility of applying topic in legal interpretation (this discussion can be
136
CHAPTER 4
situated within the framework of the “fundamental discussion” pursued
in the German science of law, where it is called Methodenlehre). As a
matter of strict fact, it should be noted that in 1928 Salomon – a legal
philosopher – spoke about topic as a method of jurisprudence; it remains
true, however, that it was not until the book by Viehweg had been published that the main discussion of this issue developed. The ancient
philosophies of argumentation, especially the topic of Aristotle and
Cicero, provided a starting-point for Viehweg (just as for Perelman).19
Viehweg tries to show that legal reasoning, the work of lawyers and
legal method each have the nature of a topic. Accordingly, his approach
is decidedly anti-positivist. Let us recall that legal positivism – emphasizing the methodological autonomy of jurisprudence – stresses the systematic aspect of legal thinking (reasoning), whereas topic, proposed by
Viehweg, exemplifies the anti-systematic, focused on specific problems,
approach to legal method. Legal positivism (its two classical versions –
Gesetzposistivismus and Begriffsjurisprudenz as well as Kelsen’s normativism) presupposed that legal thinking (reasoning) – corresponding to
the structure of the system of valid law – possesses a logical structure.
This presupposition found its expression in the conception of a legal syllogism (regarded as the basic type of legal reasoning commonly applied
in the process of legal interpretation), and, later, in the Kelsenian conception of a static system, i.e. one in which hierarchical connections
between norms in a given system are unquestionably of a legal (inferential) nature (within such a system, one can directly derive the content of
a lower norm from the content of a higher norm).
The method of topic is thoroughly different (the very word “method”
is perhaps not quite appropriate in this context – Viehweg himself preferred describing topic as “an argumentative technique”). Topic is simply
a certain technique of thinking focused on particular problems – the
technique traces its origins back to rhetoric and its goal is to solve concrete dogmatic problems. It is by no chance that topic, in Viehweg’s view,
was always an immanent part of civil law (this was perfectly realized by
Roman lawyers, who, as is well known, concentrated on solving genuine
legal problems, trying to avoid superfluous formalisms). In being simultaneously anti-positivistic, anti-formalistic, anti-theoretical and antisystematic, topic becomes in many respects close to the hermeneutic
approach: topic appears to imply that all “activities” are to serve interpretation and, accordingly, that it is in the process of solving particular
problems that “concrete law” comes into being. The approach based on
topic remains in exact opposition to all those views (above all legal
positivism) which are, so to speak, “of paradigmatic nature”, i.e. which
ARGUMENTATION
137
assume some methodological axioms, true and universally valid types of
reasoning, and appeal to the notion of an a priori accepted and uniquely
possible system.
Of course, the topic approach to the method of legal reasoning (if it is
a method!) may engender many doubts. In applying topics (both general
and special) in legal discourse, one also introduces elements of systematic thinking and uses a sort of argumentative logic. Moreover, in legal
discourse, these topics (arguments and principles) are continually
applied, since they are considered to be the universally valid standards of
legal thinking. One may, therefore, justifiably assert that legal topic is
also associated with systematic thinking, because reasoning based on it
relies upon the rules of some sort of – informal – logic.
Numerous philosophers, as well as the theorists of law representing
the strand of Methodenlehre, took part in a discussion over the possibility of using topic as a method of jurisprudence. There was no lack of
critical opinions of Viehweg’s conception. The main objections raised
were that the representatives of topic did not advance any plausible arguments against the systematic and formal (axiomatic–deductive) method
of the analysis of law, and that the alternative method they put forth
(which is in fact a mixture of questions unconnected with one another)
implies that an accidentally invoked topos, rather than valid law, will
determine a concrete legal solution. Esser – an otherwise moderate advocate of the method of topic – points out that it does not make much
sense to radically oppose systematic reasoning and reasoning based on
topic, given that topics themselves (being in fact general rules or principles) legitimate thinking in terms of a system (as already noted in the
preceding paragraphs). The place of topic is, in Esser’s view, in the area
of jurisdiction. In particular, the interpretative activity of a judge is a
paradigmatic example of thinking focused on particular problems
(which is a characteristic of topic). All in all, according to Esser (as well
as to Viehweg), the function of topic boils down to finding reasonable
arguments, which help an interpreter to properly describe facts relevant
to a given case, as well as to make the final decision (i.e. establish a legal
norm appropriate for the case).
According to Kriele the essence of topic as applied in the science of
law can be summarized by three theses: (1) legal arguments are not
deductive, (2) these arguments are essentially legitimated by the opinion
of the majority, (3) they must be analyzed in each particular case and
may not neglect any single view or opinion. Even though Engisch and
Larenz questioned multiple particular assumptions of Viehweg’s conception, they nevertheless share his view that it is not possible to apply
138
CHAPTER 4
the axiomatic–deductive method in legal reasoning. As for Zippelius, he
asserted that legal topic plays a particularly important role in the process
of interpretation concerning axiological gaps in the law (so-called
Wertungslücken). More specifically, topic enables one to apply general
clauses to a concrete case. Of primary importance for the rise of topical–
rhetorical theories of argumentation were the works of representatives of
the Brussels school (especially Perelman, whose conception of rhetoric
and topic we have already discussed). What salvaged and intensified
falling interest in the issues of legal topic was the work of Struck
Topische Jurisprudenz, Argument und Gemeinplatz in der juristischen
Arbeit (published in 1971).
At this point it would not be inappropriate to query whether the discussion devoted to the topical–rhetorical conception of legal discourse
amounts to “much ado about nothing”. It is submitted that the answer to
this question should definitely be negative for the following reasons. In
our view, topic and rhetoric open up the possibility of constructing realizable (applicable) philosophies of argumentation. The applicability of
these philosophies flows, among other things, from the fact that “the elements of topics” are incessantly present in legal thinking (legal argumentation), which as a rule is focused on particular problems (particular
cases). It is hard to overestimate the significance of a second reason: the
problem of topic concerns an absolutely fundamental issue, namely
whether the solutions to legal cases must be based solely on legal rules or
norms. Positivists (for instance Hart and Kelsen) gave different, though
always positive, answers to this question. According to Hart, legal decisions must be based on legal rules, yet judges are free to choose or formulate these rules (when, for instance, they establish a precedent). Kelsen,
by contrast, asserted that judges have no such liberty: a norm is either
derived (deduced) from an immediately higher one belonging to a system
of valid law (in the case of a static system), or passed in accordance with
a competence contained in a norm belonging to a system of valid law (in
the case of a dynamic system). However, according to Dworkin – who, as
is well known, rejects the positivistic approach – legal decisions may be
taken on the basis of both legal rules and standards (the latter embraces
principles and policies). In point of fact, the new word “standard” designates an old idea – namely, some legal topic. The first type of standard –
principle – refers to basic moral values (justice, honesty etc.), whilst the
second type – policies – refers to economic, political or social values.
It is worth pointing out one more – rather paradoxical – fact: the conception of law and legal discourse based on topic was initially realized in
systems of precedent law, not in systems of continental law (the latter
system is a direct descendant of the tradition of Roman jurisprudence,
ARGUMENTATION
139
which gave priority to thinking focused on concrete problems, and,
accordingly, to the methods of interpretation based on topic).
4.2.2
Procedural Conception of Legal Discourse
The procedural approach enables one to look at the issue of legal discourse
from a different – more abstract and formal – perspective. By no means do
we wish to assert that only two approaches – namely, topical–rhetorical
and procedural – are acceptable. Yet it remains the case that these two
approaches played an essential role in shaping the contemporary philosophy of argumentation.
In the second half of the twentieth century, theories of argumentation
formulated in the field of jurisprudence became paradigms for those
examined in the fields of philosophy (ethics), sociology, political sciences
and economics. It was also at this time that procedural and formal variants of theories of argumentation began to be formulated more frequently: given that these variants are essentially different from ancient
conceptions of argumentation, they cannot be treated simply as acceptance of those conceptions. As a rule, these theories are strongly normative: they are not disturbed by the limitations built into real legal
discourse, which depends on the complexity of the case being discussed
and its entire interpretative context, tradition, and all the characteristics
of those who directly participate in it (meaning limitations of a
phenomenological nature – connected with the consciousness level of
participants in a discussion – and limitations of a psychological nature,
meaning the psychological experience of participants in a discussion). The
main ambition of the authors of procedural theories is to create an ideal
argumentative model, or, more precisely, to describe formal conditions
which must be satisfied by each acceptable, i.e. right and rational, practical discourse. These theories, then, aim to provide a “universally valid”,
theoretical model to function as a measure of rightness and rationality
for every possible practical discourse. Thus, the specified goals of procedural theories constitute both their strength – since if one accepts certain
idealistic presuppositions, the goals will hardly be questionable – and
their weakness – since, among other things, the goals are usually formulated on the meta-theoretical level, which makes it impossible to apply
these theories directly in legal discourse. Besides, procedural theories
contain not only purely normative theses, but also some, arguably
unwanted, descriptive ones.
Perelman. As announced earlier, we once again return to Perelman’s
conception. Ancient topic and rhetoric constituted a starting-point for
his philosophy of argumentation, and its final result is a procedurally
140
CHAPTER 4
oriented theory of argumentation, called by him “the new rhetoric”. At
the root of “the new rhetoric” it is possible to find Aristotle’s philosophies of argumentation (especially topic and rhetoric), which Perelman
considerably supplements and modifies, as well as Kantian ideas of the
transcendental subject and the categorical imperative. As mentioned earlier, the goal of “new rhetoric” is to convince the audience by means of
a discourse in decision making situations, i.e. situations in which there
exist different practical (normative) alternatives. Thus, the standard for
evaluating of an argument is its persuasive force; the criterion of effectiveness is, in Perelman’s view, of secondary importance. The main goal
of every argumentation is to gain or strengthen the support of the universal audience. In order to achieve this goal, a speaker must adapt her
speech to the demands of the universal audience.20 Thus, the universal
audience becomes a key notion in understanding any theory of practical,
legal discourse. It is precisely due to this notion that this theory acquired
procedural and formal characteristics. This is because the notion of the
universal audience is interpreted by Perelman in an abstract, ideal and
formal manner. The fact that an argumentation, i.e. certain reasons
advanced in the process of a discourse, has been recognized by the universal audience implies that the argumentation is right, valid, rational
and objective. Furthermore, the acceptance of argumentative decisions
by the universal audience is likely to guarantee the efficiency of the
relevant practical discourse.
Doubts may arise here as to whether criteria of efficiency may be
established by the universal – purely formal – audience. Aristotle
asserted that the task of rhetoric is to convince any unspecialized, but
real, audience. In addition to the notion of the universal audience,
Perelman introduces the notion of the particular audience. We think that
the problem of efficiency should be associated with the latter type of
audience: an argumentation can be called efficient if it is accepted by at
least part of a particular audience. Moreover, even mutually exclusive
argumentative theses can be described as efficient, provided that a
speaker succeeds in persuading part of the particular audience to accept
his argumentation. As mentioned earlier, Perelman distinguished the
notion of convincing (convaincre) from the notion of persuading (persuader): the former is connected with an arguments’ validity and, accordingly, with the concept of the universal audience, whereas the latter is
connected with an arguments’ efficiency and, accordingly, with the particular audience. The particular audience exists in a specific place and
time, for some concrete discourse, whereas the universal audience is
the criterion for assessing every possible kind of particular practical
ARGUMENTATION
141
discourse. Perelman emphasized that a speaker must use rational arguments, adjusting them to the categorical imperative as understood by
Kant, and that her postulates and reasoning must be valid for the whole
human community. The universal audience is made up of all wellinformed and reasonable people – the whole community of potential
participants in a discourse – or at least some ideal representation of such
a community.21
Does the above definition of the universal audience provide sufficient
reason for treating “the new rhetoric” as a theory of procedural type? It
seems that even Perelman would have difficulty in answering this question, the reason being that his theory of argumentation constitutes a specific combination of rules and theses which can be interpreted materially
(these rules and theses can be found as early as in the ancient philosophies of argumentation, especially topic and rhetoric), and purely procedural and formal rules and theses (introduced into considerations about
practical discourse – practical reason – by Kant). This specific combination of material and formal elements in Perelman’s theory is most clearly
visible in the relationship between two types of audience: universal and
particular.
Alexy. Free from these material intercalations is Alexy’s procedural
theory of practical, legal discourse. His theory of argumentation appeals
to at least three philosophical traditions. These are the Kantian conception of practical reason, analytical philosophy and the theory of
Habermas (especially his consensual theory of truth). Alexy presented
his theory in the work Theorie der juristischen Argumentation. Die
Theorie des rationalen Diskurses als Theorie der juristischen Begründung
(published in 1978).
Alexy makes a clear distinction between theoretical and practical discourse. He does not deal with theoretical (scientific) discourse, which
appeals to the formally understood criterion of truth. His focus is on
practical discourse, the goal of which is to provide justification for normative statements. Thus, the starting-point of Alexy’s considerations is
different from that of Habermas, who also distinguishes the two types of
discourse, although he does not juxtapose them so decidedly as Alexy
does. This is due to the fact that Habermas regards consensus as a criterion to be applied in evaluating those theses formulated during a
theoretical discourse and those formulated during a practical discourse.
Alexy’s dualist account of both kinds of discourse enables him, by
contrast, to justify the thesis of the cognitive specificity of a practical
discourse. What is more, not only does a practical discourse appeal to
142
CHAPTER 4
rules and principles other than those, which “govern” a theoretical
discourse, but legal discourse itself is a special example of a general,
practical discourse (Sonderfallthese).
The procedure of a practical discourse is determined by formal rules.
Proceeding in accordance with these rules enables argumentative decisions to be made, which fulfill the requirements of rationality and rightness. The indefeasibility of the rules of a practical discourse stems from
the fact that they are confirmed by many types of scientific and philosophical justification: technical, empirical, analytical, transcendental or
universal-pragmatic.22 Thus, to reject these justifications would be to
reject elementary scientific intuitions (empirical and analytical) and
philosophical ones (appealing to practical reason or common sense). The
manner in which these rules are justified, as well as their purely formal
character, make them universal; in consequence, these rules play the role
of sui generis axioms in a practical discourse.
Alexy speaks of several types of formal rule “governing” a practical
discourse. These include rules that are basic, appealing to reason,
argumentative, justificatory, and, finally, the so-called rules of “passage”.
Basic rules are formulated on the basis of the most fundamental intuitions
concerning the process of language communication; they embrace, in particular, the following rules: (1) no speaker may contradict herself, (2) every
speaker may defend only what she herself believes, (3) every speaker who
uses a certain predicate to designate a given object should use this predicate with reference to every other similar – in respect of relevant features –
object, (4) different speakers may not assign different meanings to the
same expression. The following, according to Alexy, is an example of a
rule that appeals to reason: upon the demand of some other participant
in a discourse, every speaker must justify her thesis and may not invoke
any circumstances that would legitimate her refusal to provide such justification. As for argumentative rules, one may point to the rule, which
forbids the justification of a situation in which one participant in a discussion treats another participant differently (better or worse) from others.
Alexy associates justificatory rules with the principle of generalizability
(Verallgemeinerbarkeitsprinzip) and he examines its three variants
(proposed by Hare, Habermas and Baier). Rules belonging to the last
group – i.e. the rules of “passage” – grant every participant in a practical
discourse the right to appeal to the arguments of empirical, analytical or
theoretical type. That is to say, they enable the participant to “pass”, at
any point in the discourse, from practical to theoretical discourse.23
According to Alexy, what is ultimately at stake in a legal discourse is
the process of justifying (Rechtfertigung) adjudications (verdicts), i.e.
ARGUMENTATION
143
normative statements of a special kind. Two aspects of this process may
come into play – namely, internal justification (interne Rechtfertigung)
and external justification (externe Rechtfertigung). The former is
intended to demonstrate that an adjudication (verdict) follows logically
from premises assumed for its justification, while the latter is intended to
demonstrate the rightness of those premises. The goal of practical, legal
discourse is surely to demonstrate the rightness of normative premises
assumed for the needs of justification. Alexy presents six groups of rules
and forms of external justification: (1) rules and forms of interpretation,
(2) dogmatic argumentation, (3) law-making adjudications (precedents),
(4) practical argumentation, (5) empirical argumentation, and (6) special
forms of legal argument (for instance, such argumentative forms as a
simili, a contrario, a fortiori, or ad absurdum).24
The relationships between a general, practical discourse and a practical, legal discourse are captured in the form of three theses, which constitute the development and specification of the above discussed
Sonderfallthese. Let us recall these theses: Sekundaritätsthese, that is a
theory of “the superficial character of a legal discourse”, Additionsthese,
a thesis about “the complementary character of a general, practical
discourse” and Integrationsthese, a thesis about “the integrality of a discourse”. The first thesis asserts that, in cases which cannot be decided
exclusively on the grounds of the rules of valid law, it is a general, practical discourse that constitutes the real basis for making a decision – the
role of a legal discourse is to provide only “a secondary legitimation”;
the second one implies that legal argumentation is sufficient only up to
the point when specific legal arguments become exhausted and, in consequence, need to be supplemented with arguments from a general, practical discourse; the third thesis, in turn, says that specific legal arguments
should, at every stage of a given discourse, be used in tandem with arguments from a general discourse. Movement from one discourse to
another is, in light of the above mentioned theses, not only possible, but
also, in many cases of argumentation, simply necessary.
A practical, legal discourse possesses a fully formal structure, made
up of the rules of a general, practical discourse as well as special legal
rules. Proceeding in accordance with the procedure determined by these
rules enables an unquestionably acceptable – from the viewpoint of the
criteria of rationality and rightness – result to be accomplished. The
idea of procedurally understood rationality is ultimately expressed in
the following six principles: (1) consistency, (2) teleological rationality,
(3) verifiability, (4) coherence, (5) generalizability, and (6) veracity and
openness.25
144
CHAPTER 4
Of course, the procedural account of practical, legal discourse may
(just as the topical–rhetorical account) give rise to doubts. Problemthinking – preferred within the topical–rhetorical approach – has been
replaced with systematic thinking. Systematic thinking, though, can
plausibly be defended only if one accepts a number of idealistic (normative) presuppositions, which, as a rule, are of meta-theoretical nature.
This is because the procedural theory of legal discourse only constructs
presuppositions and conditions for the “proper” theory of legal argumentation. The possibility of applying such a complex theory in practice
is in fact very limited. Besides, not all theses and rules are self-evident in
an analytical or common-sense way, and the catalogue can in principle
be supplemented and modified arbitrarily. In consequence, the procedural theory of a discourse cannot be definitely “closed”, and each
example is only one of many possible versions.
What distinguishes the topical–rhetorical account of a discourse from
the procedural one is something that may be termed “an argumentative
perspective”. Topical thinking takes a concrete problem (a particular
legal issue) as its starting-point; only then does it adapt to that problem
a given type of argumentation, drawing from topics that can be applied
to that problem. Thus, problem-thinking is inductive in its nature: we
begin with a concrete issue (normative fact), which we subsequently
interpret by appealing to “common places” of legal thinking, i.e. to specific topics. Procedural theories of a practical legal discourse propose
the reverse way of proceeding. In an attempt to prove the rightness of the
premises assumed to justify argumentative decisions, we appeal in the
first place to the general rules of a practical discourse and the rules of a
legal discourse. These rules enable us to ascertain whether a proposed
solution to a concrete case fulfils the criterion of “an ethical minimum”,
i.e. the criterion of formally understood rationality or rightness. It is not
until this test is passed that an argumentative dispute may be finished.
Thus we are dealing with a kind of systematic proceeding, which can be
described as deductive, since the argumentation begins “from what is
general” (i.e. from rules (axioms) of a (practical and legal) discourse) to
reach “what is particular” (i.e. a solution to a concrete legal problem).
Ultimately, however, both topical and procedural conceptions provide
an answer to the question of whether legal decisions should be made
solely on the basis of legal norms (rules). Even though procedural theories favor systematic thinking in the law, they nevertheless admit the possibility of appeal to extra-legal elements, such as the general rules of a
practical discourse, in a legal discourse.
ARGUMENTATION
4.3
145
LEGAL ARGUMENTATION
The time has now come to attempt to summarize the above discussion of
legal argumentation. At the beginning of this chapter it was suggested
that theories of a legal discourse in jurisprudence “want” to occupy the
place between formal logic and “hard analysis” on the one hand, and
hermeneutics on the other. This opinion can be confirmed in light of
what has already been established. The procedural theories of legal argumentation make use – to a relatively large extent – both of logic and of
all types of linguistic analysis. Topical–rhetorical theories of a legal discourse, which reject systematic methods in the analysis of law, make use
(just as hermeneutical philosophy does), in the first place, of informal
logic and “soft analysis”. Theories of legal argumentation depart from
the standards of thinking about law, which were fixed by legal positivism, the law of nature and legal realism. Within these theories, according to the spirit of Kantian philosophy, ontological problems have been
eliminated, and the center of gravity has fallen on methodological issues.
These are theories of purely interpretative nature, and, accordingly, they
assert that a concrete (real) law emerges only as a result of the process of
argumentation. Their orientation is anti-positivistic, because they
assume that decisions made in a legal discourse may be based not only
on legal norms (rules), but also on the general rules of a practical
discourse, ethical standards, arguments and legal topics.
In our view, it is not possible to show correctly and rationally that any
one conception of a legal discourse is better than another. We deem controversies over this issue to be entirely academic. The complexity of the
structure of a practical, legal discourse, the openness of theories of argumentation to all other philosophies of interpretation, and, accordingly,
the constant possibility of their modification and improvement, and the
impossibility of applying most of these theories in argumentative practice (the more theoretically sophisticated a theory of argumentation, the
less practically useful it becomes) are all reasons which support the assertion that it is necessary to use both (discussed above) approaches. That
is, the topical–rhetorical approach as well as the procedural approach
should be combined with a view to producing an adequate theory of a
practical, legal discourse, i.e. a theory that can be realized – applied in
interpretative and argumentative practice. The topical–rhetorical conception – if deprived of the theoretical (systematic) perspective provided
by the procedural approach – will be too narrow (lame). The procedural
theory, in turn will be too wide (leaping) if detached from topics that can
be interpreted materially.26
146
CHAPTER 4
Opposing the problem approach and the systematic approach, and
persistently trying to prove that the former or the latter characterizes
legal thinking makes little sense, for the simple reason that we appeal to
both approaches when we make interpretative decisions, or when we
conduct a more widely understood legal discourse. Moreover, these types
of thinking are interlinked: problem (topic)-thinking turns into systematic thinking and vice versa – this is precisely what the sui generis dialectic of a legal discourse consists of. Joining the two perspectives –
topical–rhetorical and procedural – enables the avoidance of one more
objection, which can almost always be raised, namely, that theses or rules
of a purely formal, procedural and normative nature are often confused
with theses or rules possessing a material, descriptive and empirical character. The conception of a practical discourse proposed below possesses
features characteristic of both normative theory (in the part of the conception which concerns formal rules of a legal discourse) and descriptive
theory (in the part which concerns legal topics that can be interpreted
materially).
4.3.1
Claim to Universality
Legal argumentation as a method is intended to be of universal application. This is by no means surprising, given that similar claims are set up
in the other philosophies of interpretation, i.e. logic, analysis and
hermeneutics, though each asserts its claim slightly differently. Since we
regard the distinction (already made in the earlier part of this book)
between practical and theoretical discourse as plausible, we can examine
this thesis, concluding that it is only with reference to its specific field of
application that legal argumentation is universal. Naturally, this field is
the field of practical cognition. Thus, according to this thesis, a proper
argumentation can be led only within a practical discourse (and what is
more, this argumentation is to be limited only to “hard cases” – see
Section 4.3.2, rule 5). As for a theoretical, legal discourse, however, it is
open to all the other scientific methods (logic, analysis), arguably with
the exception of argumentation (since one does not discuss facts). Thus,
argumentation may be regarded as a specifically ‘legal’ method, because
it enables us to operate in the world of practical (normative) reasoning,
where it is no longer possible to decide issues according to the criterion
of truth and falsity. Just as logic and the methods of analysis are
irreplaceable in a theoretical discourse, argumentation is irreplaceable in
a practical discourse. It follows from this fact that the claim to universality contained within theories of argumentation can reasonably be discussed only in relation to a practical discourse. From this perspective,
ARGUMENTATION
147
it is possible to affirm that only phenomenologically oriented hermeneutics
may justifiably assert a full claim to universality (i.e. with reference to
both kinds of discourse). This is because “the problem of understanding”
has neither ontological nor methodological limits.
The universality of argumentation, however, can also be defended
along different lines. For example, by attempting to demonstrate the universal validity of general rules of practical discourse, a universally valid
procedure and ethics for all possible communications between people
can be established. Accordingly, the rules thus interpreted are fully transcendental in nature and, for that reason, the scope of their application
cannot be limited to a practical discourse, still less to a legal discourse.
4.3.2
Structure of Legal Discourse
Now the time has come to deal in more detail with legal discourse. As the
foregoing considerations imply, both views of a practical legal discourse
ought to be discussed: procedural (formal) and topical–rhetorical (material). General rules establish a universally valid procedure for every possible practical discourse (including legal discourse). These rules enable
criteria determining the viability of a practical discourse or, more precisely, concrete argumentative decisions, to be formulated. The criteria
are rationality and rightness. To put it more simply, a practical discourse
is rational and right if, and only if, it is conducted in accordance with
general rules (to be discussed below). The rules are elementary, which is
why accepting them is not a matter of the good will of participants in a
discourse, but a sort of ethical imperative. In excluding these rules from
a practical discourse, every possibility of achieving a morally acceptable
communication is excluded. The indefeasibility (universal validity) of
these rules follows from the fact that they are purely formal (alluding to
Kant’s philosophy, we might say that these rules are sui generis categorical imperatives of a practical discourse, determining not the content but
the form of concrete argumentative resolutions), and from the fact that
they can be justified (confirmed) in many different ways (e.g., by appealing to common sense, the criterion of self-evidence, or to the Kantian
conception of practical reason, which implies the existence of universally
valid ethical intuitions confirmed by the universally valid moral law –
the formally understood categorical imperative). In relation to legal topics, and in consequence, to final argumentative resolutions, general rules
play the role of sanctioning rules. Without these rules, an argumentation,
based solely on material premises, could always be undermined. Bearing
in mind the fact that the functions of a legal discourse are not only ethical but also instrumental, it is necessary to reasonably limit a number of
148
CHAPTER 4
general rules commonly regarded as basic. The rationale for doing so is
that, by leading to feasible argumentative resolutions, a legal discourse
should above all put an end to an interpretative controversy. And criteria of rationality and rightness that are too rigorous might lead to the
detachment of a practical discourse from a concrete argumentative case.
It is the stage of application (the topical–rhetorical stage), which guarantees that a legal discourse will be realized. At this stage, one becomes
entangled in problem-thinking about a legal issue (being the object of a
discourse), as one begins to appeal to material legal topics. A selection of
topics will be different in each legal discourse (only general rules will
remain the same); more specifically, it will turn on the difficulty level and
argumentative context of a given case, as well as on the traditions and
the habits of participants in the legal discourse engendered by this case.
In a legal discourse, depending on the place where they appear, topics
(arguments and legal principles) may play either the role of loci communes or loci specifici. Some arguments and principles may be interpreted in a legal discourse as “common places” (loci communes), since
their application is universal both in the law (in relation to all kinds of
legal reflection) and in other practical discourses (this concerns the
majority of arguments and at least some principles, such as, for instance,
pacta sunt servanda, ignorantia iuris nocet, audiatur et altera pars, which
can be appealed to also in other discourses – for instance, ethical or
political). There are also topics which, in a legal discourse, play the role
of “special places” (loci specifici). These are above all specialized (i.e.
referring to a concrete kind of legal reflection) legal principles (such as,
for instance, lex specialis derogat legi generali, nullum crimen, nulla poena
sine lege poenali anteriori, whose application beyond a legal discourse
can hardly be imagined: in non-legal discourses, it is rather the converse
of the former principle that is accepted; and, as for the latter, as a rule, it
refers only to criminal law).
This issue is highlighted, since, according to Aristotle and Perelman,
legal topics (general rules of the law) are only “special places” (loci specifici) of the law. As mentioned above, this view flows from their conviction that “common places” (loci communes) are always connected with a
general – unspecialized – type of reflection. Of course, from the standpoint of a general, practical discourse (if such discourse may exist at all
autonomously), legal topics may be treated only as “special places” (loci
specifici), because they concern a specialized kind of reflection connected
with the law. We also want to point out that, although from the standpoint of the topical–rhetorical approach, topics form the material content of a legal discourse, many arguments and principles can be interpreted
ARGUMENTATION
149
in a formal way. Such an assumption essentially undermines the opposition (accepted by the proponents of the topical account of legal argumentation) between two types of thinking in a practical discourse –
namely, problem thinking and systematic thinking: topics, if interpreted
materially, may be an element of the former; if interpreted formally – of
the latter.
In our view, an analysis of the structure of a practical discourse should
embrace general rules of a practical discourse, the rules of passage (joining a general discourse with a legal one) and legal topics (arguments and
legal principles). General rules and rules of passage determine the formal
and procedural character of a legal discourse, while legal topics inform
its material content. These rules and topics will be discussed below in the
order proposed above.
General rules. Here, a catalogue is constructed of rules, which appeal to
the criterion of self-evidence. That is to say, we are take into account only
rules, which can hardly (or which simply cannot) be questioned from the
viewpoint of common sense or from the viewpoint of elementary ethical
standards. In this sense, they are universally valid and indefeasible. A formal (in the sense which Kant gave to the categorical imperative) character of these rules enables defining the criteria of rationality and rightness
to be used while assessing decisions made in the course of a practical
discourse. This is because an argumentation satisfies the “minimum”
requirements of rationality or rightness (i.e. it can be accepted, or regarded as valid) if and only if it is made in accordance with these rules. General
rules establish both the procedure and formally understood ethics of a
practical discourse. Thus, ultimately, a list of these rules should embrace
only those, which are uncontroversial, universally accepted and unequivocal (as far as the way they are formulated is concerned). There is a
danger that every attempt at building a catalogue of these rules may
provoke a difficult controversy as to whether successively proposed rules
are universally valid and should always be complied with in a practical
discourse:
1. One should engage in a practical discourse only if one is convinced that
the discourse can justifiably be called right. More specifically, one
should be above all convinced that the methods used in the discourse
are right; the conviction of the rightness of the discourse’s objective is
of somewhat lesser importance. If an argumentative discourse has
been conducted in accordance with general rules, and, consequently,
with procedure accepted through these rules, then the final result –
being rational and right – also has to be acceptable in an axiological
150
CHAPTER 4
sense. Thus, one needs to be convinced of the rightness of the rules of
argumentation. Otherwise there would be a danger of a participant
molding the discourse to a result she has earlier accepted as right, and
then trying to achieve it at any price – per fas et nefas. Were that to be
the case, then the only effectively applicable criterion in a legal discourse would be efficiency (understood eristically). In a practical,
argumentative discourse, which appeals to the transcendentally
understood criteria of rationality and rightness, the principle that
“the end justifies the means” cannot be accepted.
2. A practical discourse ought to be conducted in such a way that the principle of veracity may be respected. This rule also seems uncontroversial. In a discourse, one may not lie, tell untruths, or omit to tell the
truth by remaining silent. It is also forbidden to invoke any circumstance that might justify lying (e.g., some sort of external pressure or
coercion). What is more, even lying “in a good cause” is not allowed.
Even though the view that the principle of veracity ought to be realized in a practical discourse without exception is defended, it is
important at the same time to be mindful of the possibility of limiting this principle in a legal discourse. Examples of limiting the
principle of veracity in a legal discourse are provided especially by
legal rules concerning criminal action, which formally sanction the
right to “passing over truth in silence”.27
3. A practical discourse ought to be conducted in such a way that the principles of freedom and equality are respected. A practical discourse
ought to proceed in accordance with the formally understood principles of freedom and equality. This means that, in the course of a discourse, we must proceed in accordance with at least eight special rules
that follow: (1) an argumentative discourse ought to be accessible to
everyone possessing sufficient knowledge of the discourse’s object
and, in some cases (e.g., in the case of a legal discourse), a justifiable
interest in it, (2) each participant in a practical discourse ought to
possess the same privileges and be subject to the same limitations,
(3) additional privileges and limitations can be introduced only if agreed
by all the participants in a practical discourse (these privileges and
limitations ought to affect each of them to the same degree), (4) each
participant in a legal discourse ought to have the same opportunities
to participate in it, especially with respect to opening the discourse,
submitting theses and presenting views, giving answers, suggesting
that the discourse should be suspended or finished, (5) no participant
in a legal discourse may be subject – in connection with the discourse –
to pressure or to any limitations, unless these limitations affect each
ARGUMENTATION
151
participant to the same degree, (6) each participant in a practical discourse ought to justify the thesis she submits in the discourse, or to
answer a question, if another participant requires that she do so, (7) if
a person wishes to treat one participant in a discourse differently from
the others, she is obliged to justify her doing so, (8) if a decision (or
the establishment of a fact) made in the course of a practical discourse satisfies only some of those participating in the discourse, the
remaining participants ought to agree to that decision (or that establishment of fact). The content of rules 4–8 reflects the principles
formulated by Habermas and Alexy.28
4. A practical discourse ought to take the basic principles of language
communication into account. This rule also is a fundamental one, since
a practical discourse ought to satisfy the condition of inter-subjective
communicability, expressed by the following seven special rules: (1) a
practical discourse ought to be transparent, (2) a practical discourse
ought to be conducted by means of the most simple language possible, (3) in a practical discourse, a concept can be given a different
meaning than that accepted in ordinary language only if this meaning
is accepted by all participants in the discourse, (4) each participant in
a practical discourse who uses some predicate to define an object
ought to use the same predicate with reference to every object having
similar – essential – properties, (5) each participant in a discourse
should give the same meaning to a given expression29, (6) each participant in a practical discourse ought to use the methods of language
analysis, possibly extensively, (7) the course and the results of a practical discourse ought to be generalizable. The rule of generalizability
played a crucial role in the theories of argumentation (Perelman,
Schwemmer, Habermas and Alexy wrote extensively about it). This
rule also contains the essence of the condition of inter-subjective
communicability of a practical discourse: if argumentation could not
be generalized, then a practical discourse could not satisfy the conditions of openness and transparency.
5. A practical discourse ought to be conducted only in hard cases. This rule
seriously limits the scope of application of a practical discourse – it
implies that it is necessary to narrow the “claim to universality” of a
practical discourse, discussed above. At the same time, though, it is
important to remember that, in principle, the whole philosophy of
humanistic interpretation has been “invented” for hard cases, i.e.
cases which cannot be interpreted and adjudicated by means of standardized (algorithmic) methods. Another question arises concerning
the role of a practical discourse: is there one, and only one, rational
152
CHAPTER 4
and right decision in a given hard case, or are there many such decisions?
Let us recall that, according to Hart, a judge who has to adjudicate in
a case should appeal only to legal rules. Only when there are no applicable rules (which is typical for hard cases), she must have recourse to
extralegal standards. This requirement, however, does not exclude the
possibility that a judge will pass an entirely new – precedential –
judgment in a given case: she is limited by the rules of valid law, yet
she has unlimited freedom in making her final decision.
Dworkin approaches this problem in a different way. He admits
that a judge may appeal not only to rules but also to legal standards
(principles, i.e. legal topics and policies), yet at the same time he
asserts that there exists only one right decision (right answer) in a
hard case, and that the judge-Hercules should find it. We are convinced, though, that a hard case may have more than one – rational
and right – solution (which, by the way, may be the reason why it
qualifies as a hard case). In order to settle a hard case, one must
appeal to the rules (norms) of valid law, general rules of a practical
discourse, as well as legal topics. A legal discourse not only allows us
to discuss possible varying decisions in a hard case, but also provides
us with the criteria for choosing one decision from amongst those proposed (the criteria of choice are formulated in the process of the
external justification). A rule which confines a practical discourse
only to hard cases has a decidedly anti-eristic and anti-sophistic character. It is possible to derive from this rule a ban on undertaking a
practical discourse without good reasons for doing so, e.g., only for
the needs of “games of negotiation”, which give rise to expansive,
though superfluous, argumentation. When fully engaging in a practical discourse, it is necessary to be convinced of its rightness (and, consequently, of its necessity), and of the fact that the case being
interpreted is a hard one.
6. A practical discourse ought to take account of established facts. This
rule expresses the conviction that at each stage of a practical discourse
we should make use of facts already established in a theoretical discourse, as well as of those which can be established in the future. Alexy
spoke in this sense about the rules of transition (Übergansregeln) from
a practical discourse to a fully cognitive discourse, appealing to empirical theses, purely theoretical theses and theses based on language
analysis.30
7. A practical discourse ought to move directly towards its end. This rule
establishes at least two important principles: “the economy of argumentation” and “the directness of a discourse”. This rule, just like
ARGUMENTATION
153
rule 5, has an anti-eristic and, in some cases, also anti-rhetorical character. It prohibits the use in a legal discourse of methods (eristic and
rhetoric) which might unjustifiably prolong an argumentative controversy. To give some examples, it is forbidden to begin a discourse with
exceedingly lengthy introductions, the purpose of which is only “to
tire the opponent”, to introduce into a discourse many superfluous
digressions and questions, to make partial summaries which make it
difficult to follow the main plot, to create apparent complications, to
call persistently into dispute the opponent’s theses which are obviously correct, or to engender chaos and confusion. At least five
special rules can be related to the rule of the directness of a legal discourse: (1) each participant in a practical discourse should confine
herself to submitting only such theses, rules and arguments which she
is convinced will contribute directly to the settlement an interpretative
case (being the object of the discourse), (2) no participant in a practical discourse may submit a thesis which contradicts already accepted
theses, unless she sufficiently justifies it or convinces all other participants in the discourse that the thesis is to be accepted, (3) no participant in a discourse should contradict herself, (4) each participant in a
discourse who submits a thesis not connected directly with the discourse’s object should present her reasons for having done so, (5) each
participant in a practical discourse who attacks a thesis which is not
connected directly with the discourse’s object should present her
reasons for having done so.
8. A practical discourse ought to allow for generally accepted standards,
practices and customs. The principle of inertia (Prinzip der Trägheit)
provides that decisions which have already been accepted in a legal
discourse are not to be changed or rejected without sufficient grounds.
This principle implies that in a practical discourse generally accepted
argumentative standards (topics), practices and customs should be
taken into account as often as possible. According to Perelman, this
principle is fundamental for our spiritual and social life. Since this
principle treats the process of interpretation as a historical sequence
of events, which ultimately shape our argumentative pre-understanding
(Vorverständnis or Vorurteil), it can be said that it leaves a practical
discourse open to tradition. Yet this principle has also been criticized
on various grounds. Some saw in it a manifestation of argumentative
conservatism – of a static understanding of a practical discourse.
What may count as an argument against treating it as a universally
valid principle of legal discourse is the fact that, according to rule 5,
an argumentative discourse should be pursued only in hard cases,
154
CHAPTER 4
which often change standards, practices and customs previously
established and confirmed by argumentative tradition. Yet, undoubtedly, the principle in question strengthens the principle of the economy of argumentation, because if a standard (a rule etc.) satisfying
the criteria of rationality and rightness has been accepted and
applied, then it should not be changed without sufficient reason. By
giving this principle only a formal meaning (since we do not propose
what the material content of standards, practices and customs
invoked in a discourse should be), we are able, to include it in our
catalogue of general rules of a practical discourse.
Rules of passage. These rules concern “passage” of “transition” from
a general, practical discourse to a particular, legal discourse as well as
from formal assumptions (deductive, systematic thinking) to material
assumptions (problem, inductive thinking, i.e. thinking connected with
legal topics and the valid law). The rules of passage enable one to
understand the essence of the relationship between a general, practical
discourse and a legal discourse, and, in consequence, to make more precise the thesis about “the specificity of a legal discourse” (our version
of Sonderfallthese). This is so because a legal discourse is, on the one
hand, a development of a general discourse – one of its special cases.
On the other hand, though, a general discourse exists only in practical
applications. If detached from particular discourses, it becomes a collection of general rules and principles determining some “ideal procedure”; the problem, though, is that it is unclear where this procedure
should be applied. Only by combining this discourse with concrete
material topics and material law do we achieve a complete whole. On
the other hand, it is important to remember that a legal discourse is a
specific discourse, mainly because it has to be pursued in connection
with valid law, which may limit the scope of application of at least some
general rules. Both issues find expression in the following three rules of
passage:
1. A legal discourse ought to accommodate the general rules of a practical
discourse.
2. A legal discourse ought to be pursued in direct connection with valid law.
This rule remains closely related to four other rules: (1) a legal discourse’s participants must not defend themselves by claiming their
ignorance of the rules of valid law, (2) a legal discourse’s participants
should make extensive use of dogmatic arguments, yet at the same
time, (3) they must not invoke rules of valid law which are not directly
connected with the case (which is the discourse’s object of concern),
ARGUMENTATION
155
(4) in a legal discourse the use of a concept which is given a different
meaning than that commonly accepted is allowed only if there exists
a legal definition of this concept, or if the newly proposed definition
has been accepted by all participants in the discourse.
3. The scope of application of general rules in a legal discourse can be limited only if explicitly demanded by the rules of valid law. Limitations in
question may concern the following situations: (1) different positions
of parties in legal proceedings, which may lead to limiting rule 3,
which spells out the principle of the freedom and equality of every
participant in a discourse (what needs to be emphasized in this context is the special position of a judge, who is independent and plays
the role of an arbiter in a controversy), (2) an accused person’s right
of refusal to respond to questions, or to offer explanations, and the
right of refusal to offer explanations vested in those sharing close
relationships with the accused – these rights may cause the limitation
of rules 2 (the principle of veracity), and 6 (the principle of taking
account of established facts), (3) a barrister’s duty to undertake only
legal actions that support an accused person, and an accused person’s
right not to provide evidence against herself; these principles may lead
to the limitation of rules 2 and 6, (4) the duty of a barrister and a legal
adviser to keep secret all information gathered in connection with a
case – this duty may limit rules 2 and 5, (5) hearing in private, which
may limit rule 4, point 1 (the rule of transparency), (6) the necessity
of using legal definitions in a legal discourse, which may limit rule 4,
point 3 (the principle of ordinary language).
Legal topics. A space between general rules of a practical discourse and
valid law is filled by legal topics. It is thanks to topics that we can connect general – formal – rules with a concrete – material – case. At this
stage of argumentation, problem thinking is substituted for systematic
thinking, appealing to general rules and principles. Thus, an argumentative process begins from the general, and then progresses towards the
particular, concrete issue in dispute; this is a procedural stage of a legal
discourse. Later, by discussing and deciding the controversy at issue, the
direction of thinking shifts and there is a move from the particular to the
general. At this topical–rhetorical stage of a legal discourse, we carry out
the inductive task of generalizing the decision that has already been
made (since according to rule 7, point 4, the course and the result of each
practical discourse can be generalized). Only by combining these two
argumentative perspectives can we achieve a coherent method of legal
interpretation, and simultaneously succeed in avoiding the one-sidedness
156
CHAPTER 4
implied by the acceptance of only procedural, or only topical–rhetorical,
conceptions of a legal discourse.
The term legal topics, is intended to cover both arguments and legal
principles. Let us recall that topics can be treated in a legal discourse
either as common places (loci communes), if they concern universal
issues, or as special places (loci specifici), if they concern specialized legal
issues (issues that are specifically legal and – usually – connected with a
concrete field of the law). The fact that in a legal discourse one combines
particular topics with a concrete legal case allows one to interpret those
topics materially. However, we have already pointed out that it is also
possible to confer a formal meaning on at least some topics. Finally, a
legal topic is each argument which is generally known, as well as being
accepted and justified by a valid legal tradition (in the European legal
culture the tradition in question is Roman law). Legal topics often provide strong – barely defeasible – arguments in a legal discourse. They
never guarantee, however, absolute certainty because a case may always
emerge in which a given argument or legal principle cannot be applied.
What may count as the only exception to this regularity, are legal principles explicitly expressed in the rules of valid law. Then they are simply
norms valid within a given system of law.
Arguments. In principle, these are certain rules of legal logic (informally understood) whose level of reliability (strength) and potential
scope of application in a legal discourse are highly differentiated. It
should also be noted that all of the arguments mentioned below (the catalogue is, of course, not exhaustive) can be used, not only in a legal discourse, but also in other kinds of practical discourse. Here, the 16
arguments that we consider to be the most important are considered.31
1. Argument a simili, or argument from analogy (similarity), is one of
the most frequently used arguments in a legal discourse. The conviction underlying this type of argument is that we have the right to
apply the same interpretation in similar (analogous, comparable)
normative situations. Argument a simili may be used in a legal discourse in different situations: with direct reference to a concrete legal
norm, a whole statute (analogia legis), a legal order, i.e. a system of
internal law (analogia iuris), a precedent, a custom, or, finally, any
other legal rule or principle. This argument may be used only if a similarity is shown between the case being decided and another – earlier
decided – case.
2. Argument a contrario is the reverse of argument a simili: the latter
demands reasoning “from similarity”, whereas the former demands
ARGUMENTATION
157
reasoning “from difference”. The borderline between these two
arguments is by no means distinct, however. These arguments are
very often used in interpretative (argumentative) situations. To give
an example, a hypothetical legal norm dealing with the succession of
the sons of a deceased may be interpreted, by means of argument
a simili, as applying also to the daughters of a deceased, or by means
of argument a contrario, as applying solely to his sons, since only
they are explicitly mentioned in the norm. Which argument is used
in a practical discourse will ultimately depend on whose interests are
being represented in a dispute.
3. Argument a fortiori literally means “argument from the stronger
(scope)”. This argument appears in two forms: argument a minori ad
maius, where a narrow inference is given a wider scope, and argument a maiori ad minus, where a wide inference is given a narrower
scope. Argument a minori ad maius is constructed on the basis of a
negative rule (a prohibition): if less is forbidden, then more is also
forbidden (e.g., if it is forbidden to injure a person, then it is forbidden to kill a person). The argument a maiori ad minus, in turn, is
based upon a positive rule (permission): if more is allowed, then less
is allowed (e.g., if one is allowed to kill a person in self-defense, then
one is allowed to injure a person in self-defense). According to
Kalinowski, the argument a maiori ad minus can be treated as a theorem of formal logic, provided that everything less important is contained in the more important (if all X may do A and each B is A,
then all X may do B). Yet the logical character of even this argument
was called into question. Take Perelman’s example, which undermines Kalinowski’s interpretation. In accordance with the argument
a maiori a minus, a person entitled to buy three bottles of alcohol in
a liquor shop, may also purchase one bottle of alcohol. Yet a statute
in force in Belgium in 1919 prohibited sales of less than 2l of alcohol. Thus, one could purchase one bottle with a capacity of 2l, but
not three bottles each with a capacity of half a liter. Nowadays, analogous situations exist in various kinds of wholesale trade.
4. Argument ab exemplo, i.e. “from example”, often appears in a legal
discourse. What can be regarded as exemplary are some other kinds
of a practical discourse, particular theses, rules or principles formulated in the course of this – exemplary – discourse as well as precedent decisions concerning particular cases. Argumentation ab
exemplo is numbered among the less formal methods of analytical
philosophy. In a legal discourse argument a simili and a contrario are
often used interchangeably. In point of fact, argumentation “from
158
5.
6.
7.
8.
9.
CHAPTER 4
example” and “from exemplary cases” are to a large extent based
upon the assumption of similarity (analogy) between “an exemplar”
and an object (an argumentative case) to which this exemplar is
extended.
Argument per reductio ad absurdum is used above all in formal logic,
where its structure is fairly simple. Suppose we want to prove that
statement A is true. We can do this indirectly, assuming that the contradictory statement, i.e. non-A is true. In proving that statement
non-A is false, we prove by means of the law of double negation that
statement A is true. In a practical (legal) discourse, though, we do not
prove truthfulness, but rationality and rightness. Yet we must proceed
in a similar way. We may prove that thesis A is rational or right by
proving that the opposite thesis, i.e. non-A is absurd. Thus, it is
possible indirectly to confirm the rationality and rightness of initial
thesis A. Use is made of a certain type of reasoning per reductio ad
absurdum, because we show that thesis non-A is nonsensical (e.g., that
a rational and fair legislator passed a nonsensical or unjust rule).
Argument a rerum natura, i.e. “from the nature of things”, assumes
that in a legal, as well as in any other, discourse, one should not formulate theses and make decisions which cannot be – ontologically –
realized. This argument, built upon material grounds, cannot in
principle be questioned in a legal discourse. It allows us to reject
norms or legal rules which prescribe impossible actions. Fuller considered this argument to be one of eight conditions to be fulfilled for
the law to come into being.
Argument a loco communi, i.e. “from common places”, appeals to
general topics connected with unspecialized fields of a practical discourse (including a legal discourse), to fundamental values, general
rules of a practical discourse and at least some legal arguments and
principles.
Argument a loco specifici, i.e. “from special places” appeals to special
topics connected with the specialized fields of a legal discourse, to
specifically legal arguments and principles, and to precedents from
similar argumentative cases.
Argument a cohaerentia, i.e. “from coherence” is based upon the
assumption that a legal discourse should be free from contradictions
as far as possible. This argument underlies the requirement that
each thesis which stands in conflict with theses already accepted by
participants in a legal discourse should be removed. Let us recall that
rule 7 (especially points 2 and 3) also concerns these issues.
ARGUMENTATION
159
10. Argument a completudine, i.e. “from completeness”, in turn, is based
upon the assumption that a legal discourse should be as complete as
possible. This argument underlies the requirement that each thesis
which might cause “an argumentative gap”, or constitute a threat to
coherence should be removed from a legal discourse. Rule 7 (especially points 4 and 5), mentioned in the context of argument
a cohaerentia, is concerned, at least indirectly, with this issue.
11. Systematic argument presumes that a legal discourse constitutes a
certain closed, ordered, coherent and consistent whole, i.e. a system.
This argument is strictly connected both with argument a cohaerentia and argument a completudine. It entitles us to require that each
thesis which is not contained in a legal discourse, i.e. which is neither
a rule of the system nor a consequence of a rule (argument or principle) be removed from the discourse. Of course, adherents of the
purely topical approach will question this argument, because it prioritizes, against their convictions, a systematic way of thinking and
arguing.
12. Teleological argument is a kind of reasoning focused upon the objective of a legal discourse. This argument enables one to emphasize
and ultimately confirm a presupposed objective. According to rules
1 and 7 of a legal discourse, argumentation should be aimed at
directly deciding an issue and should be conducted in the conviction
of its (i.e. argumentation’s) rightness. Thus, one may question all
argumentative theses that do not serve the realization of an objective
which was assumed as formally right and accepted by all.
13. Psychological argument may help in explaining the motives of participants in a legal discourse. Its scope of application, though, is
rather narrow. Besides, it is not by chance that this kind of argument
was often used in various sophistic and eristic conceptions.
14. Sociological argument concerns the behavior of participants in a
legal discourse. This argument appeals to purely material (empirical)
premises. For that reason theses formulated by reference to this argument are not easy to question, though it must be conceded that the
application of these theses in the practical part of a legal discourse
is limited.
15. Historical argument is based on the assumption that every concrete
discourse occurs in a given time and is preceded by something, i.e. has
its own history. From rule 8, spelling out “the principle of inertia”, it
follows that in a practical discourse (including a legal discourse),
universally accepted (i.e. being a part of tradition which has been
160
CHAPTER 4
accepted and regarded as valid) standards (topics), practices and
customs should be taken into account.
16. Economic argument can be related above all to the principle of argumentative economy expressed in rule 7. To put it simply, a legal discourse should be maximally efficient, it should involve the minimum
effort necessary, and generate a rational and fair decision which
yields maximum benefits. This argument can also be helpful in constructing a criterion for evaluating the economic efficiency of the
results of a legal discourse. Posner, one of the founders of the school
of economic analysis of law, asserts that the law in its essence is – or
at any rate should be – economically efficient. Subjects who apply
the law (judges, officials, all the participants in a legal discourse)
must be directed by an economic calculation of the costs of the legal
decision available to them. A legal decision may be described as
good, rational and efficient in the economic sense only if it contributes to the maximization of social wealth.
Legal principles. A second group of legal topics embraces legal principles, which can be interpreted broadly or narrowly. According to
Dworkin, a legal principle is a standard which should be observed, not
because it improves or protects some desired state of affairs, but because
it is a requirement of justice, honesty or some other dimension of morality.32 Ultimately, Dworkin distinguishes legal principles from other standards (policies) and legal rules. He asserts that the basic difference
between rules and legal principles lies in how they function, not in their
content. Legal rules are those norms that are applicable in an all-ornothing fashion, which implies that legal rules are either satisfied or not
satisfied. Legal principles, by contrast can be applied to varying scope
and degree (they are graded). We support a broader concept of legal
principles, one which embraces all standards that are not directly
reflected in the provisions of valid law (principles, policies, norms of customary law, general theses about the law and the methods of its examination formulated by legal dogmatics and legal philosophy and theory,
legal “sayings”) as well as at least some general legal rules (with the passage of time some legal principles become rules/norms of valid law). In
our view, controversy over the frontiers separating legal principles from
other standards and legal rules cannot be resolved. One may consider as
a legal principle every general statement about the law that has “entered”
a given legal tradition and been universally accepted. Concrete solutions
to hard cases are most often underpinned by legal principles. Therefore,
legal principles are a result of problem-thinking, and they are reached
ARGUMENTATION
161
through an inductive generalization of particular theses formulated with
a view to deciding concrete cases. It follows that the level of validity of
legal principles in a legal discourse is not constant. According to the
assumptions made here, one may encounter legal principles which are
directly expressed in the provisions of valid law, legal principles which
rest on tradition (their roots often extending as far back as Roman law)
and are recognized by lawyers as valid, and, finally, principles which seldom appear in a legal discourse (if they do appear, they are questioned).
As a result, it is hardly possible to draw up a catalogue (code) of such
principles. All the more so because some previously accepted principles
have fallen out of use, and new standards and customs have appeared
which more and more frequently play the role of principles in a legal discourse. In our view, the division of principles into general, interpretative
and special ones, proposed below, makes possible a rational classification
of the basic types of legal principle33:
1. General principles are connected with “unspecialized places” in a legal
discourse, which is why they are appealed to without limit and – in
addition – are applied in other kinds of practical discourse. These are
some examples of the most important rules of this type: (1.1) Pacta
sunt servanda, i.e. agreements ought to be observed (put differently:
agreements are valid). (1.2) Lex neminem cogit ad impossibilia, i.e. a
statute (law) may not require anyone to do what is impossible. We
have already written about this principle in presenting argument 6,
a rerum natura. (1.3) Exceptions should be interpreted strictly and are
admissible only in special cases. Rule 7, points 1, 2, 4, and 5 concerned
this principle. (1.4) Nemo iudex indoneus in propria causa, i.e. no one
can be a proper judge in her own case. (1.5) Res iudicata pro veritate
accipitur, i.e. an adjudication ought to be accepted as true. The principle of “the validity of an adjudication” is commonly accepted in the
law. Legal validity is the normative counterpart of truthfulness. (1.6)
Audiatur et altera pars, i.e. one ought to hear the other side. This topic
is closely connected to the principle of freedom and equality
expressed in rule 3. (1.7) Nemini permittitur venire contra factum proprium, i.e. one cannot oppose her own position. Recall that rule 7,
point 3 introduced a ban on self-contradiction. This principle is connected with another important topic for a legal discourse: patere
legem, quam ipse tuleris, i.e. submit to a law which you have established. (1.8). In obvious cases proceedings ought to be brief. This topic
is closely related to the principle of argumentative economy – rule 7.
2. Principles of interpretation concern more specific problems connected
with the interpretation of valid law and play an important role in a
162
CHAPTER 4
legal discourse. Again, only a few examples will be presented here:
(2.1) Clara non sunt interpretanda, i.e. what is clear does not need to be
interpreted. This well-known principle is nevertheless controversial: if
practical, legal discourse is conducted in full scope only in hard cases
(rule 5), then – at least according to some theorists of law – the provisions of valid law must always be interpreted irrespective of how clear
these provisions are. (2.2) Lex retro non agit, i.e. law does not have
retroactive force. Another version of this principle says: lex prospicit
non respicit, i.e. law looks forwards not backwards. This topic, having
its source in Roman law, became one of the fundamental principles of
the contemporary state of law. Yet even though it is commonly
accepted in lawmaking practice, one may encounter legal regulations
that consciously violate it. (2.3) Ignorantia iuris nocet, i.e. ignorance of
the law is no excuse. Thus, if ignorance of the law cannot be treated as
an excuse, by implication people are required to possess a minimum of
legal competence. This topic is closely related to the principle: iura
scripta vigilantibus, i.e. law has been written for those exhibiting due
care. Negligence, like ignorance, may not constitute any excuse in a
legal discourse. (2.4) Lex non obligata nisi promulgata, i.e. law which
has not been promulgated is not valid. According to Fuller, the
requirement that legal acts be promulgated is one of eight conditions
which make the law possible.34 (2.5) Lex superior derogat legi inferiori,
i.e. law (a legal act) of a higher degree annuls law (a legal act) of a
lower degree. In systems of continental law this principle is considered
absolute. (2.6) Lex posterior derogat legi priori, i.e. later laws annul earlier laws. This principle is to be applied when two or more legal acts
come into play. (2.7) Lex specialis derogat legi generali, i.e. a special law
annuls a general law. If it is not possible to choose between topics 2.5
and 2.6, then one should apply the “substantial principle”. (2.8) Lex
posterior generali non derogat legi priori speciali, i.e. a general law
does not annul an earlier special law. Appeal can be made to this
principle when it is necessary to choose between topics 2.6 and 2.7.
This principle is then a derogation rule of second degree.
3. Special principles are applied in “specialized places” of a legal discourse, because they concern concrete areas of law, especially civil
and penal law. The following principles are relevant to civil law: (3.1)
Trust deserves to be protected so far as, for instance, possession in
good faith is concerned. (3.2) Impossibilium nulla obligatio, i.e. an
obligation to do an impossible act is invalid. (3.3) It is prohibited to
conclude contracts which impose obligations on the third party. (3.4) Ne
ultra petita, i.e. one should not adjudicate beyond a claim. (3.5) One
ARGUMENTATION
163
ought to remedy a caused damage and give back what has been gained
without legal basis. (3.6) Nemo plus iuris ad alium transferre potest,
quam ipse habet, i.e. no one can transfer more law on the other than one
herself has. Both in civil and penal law two further principles are to be
applied: (3.7) Quisquis praesimitur bonus, i.e. the presumption that
everyone acts in good faith (is innocent). (3.8) In dubio pro reo or in
dubio pro libertate, i.e. where doubt exists a decision should be made
in favor of the accused (defendant), or in favor of freedom. Let us
now give some examples of principles relevant to criminal law. (3.9)
Law should not yield to the violation of law. This principle corresponds
to topics known in Roman law: vim vi repeller licet, i.e. force can be
repelled by force, and vim vi repeller omnia iura permittunt, i.e. all laws
allow the use of force to suppress violence. (3.10) Lawlessness is forbidden. (3.11) Nullum crimen, nulla poena sine lege poenali anteriori,
i.e. an action is not criminal and cannot be punished if it was not a
crime under penal law at the moment when it is was carried out. This
principle is usually explicitly stated in law – in the Polish Penal Code
it is contained in article 1.
4.3.3
Applications
The above analysis implies that the claim to universality raised by the
method of argumentation is rather specific and is subject to serious limitations. Methodology worked out by a practical, legal discourse can be
directly applied only in the field of normative reasoning. Besides, this
discourse should be fully performed only in hard cases. In other situations the methods of a practical, legal discourse are used to a limited
degree; it remains the fact, though, that appeal is always made to some
solutions (procedural or topical–rhetorical) typical of this methodology.
A practical, legal discourse will clearly be most fully used in legal practice, especially in the process of legal interpretation, justification of interpretative decisions, and – at least to a certain degree – lawmaking. Theses
formulated in the theory of legal dogmatics, as well as in the philosophy
and theory of law by means of tools worked out by a practical discourse,
can be justified provided that these theses are normative in character.
Should a thesis be descriptive (theoretical), then theoretical discourse
will decide whether it is true or false and, accordingly, whether it should
be accepted and included in a set of theorems of a legal discipline.
In conclusion, let us repeat once again that it is only necessary to
appeal to methods of argumentation if the potential of “harder” methods
(logic and analysis) has been exhausted in practical discourse (let us add
that this potential is limited in the normative sphere) and one does not
164
CHAPTER 4
wish to appeal to the “soft” (intuitive) methods established in phenomenologically oriented hermeneutics.
NOTES
1. G. Kalinowski, Introduction à la logique juridique, Paris, 1965, pp. 163–164.
2. K. Szymanek, Sztuka argumentacji. Slownik terminologiczny [The Art of
Argumentation. Terminological Dictionary], Warszawa, 2001, pp. 286–288;
M. Korolko, Sztuka retoryki. Przewodnik encyklopedyczny [The Art of Rhetoric.
Vademecum], 2nd edition, Warszawa, 1998, p. 19, 3 ff.
3. Ch. Perelman, Logika prawnicza. Nowa retoryka [Legal Logic. New Rhatoric],
Warszawa, 1984, p. 145.
4. M. Korolko, The Art . . ., op. cit., p. 64 ff.
5. Aristotle, Topiki. O dowodach sofistycznych [Topics. On Sophistic Arguments],
Warszawa, 1978, books I–VIII, pp. 3–235.
6. Aristotle, Topics . . ., op. cit., books I–VIII, p. 264.
7. Aristotle, Topics . . ., op. cit.; A. Schopenhauer, Erystyka czyli sztuka prowadzenia sporów
[Eristics or the Art of Discussion], p. 45 ff; K. Szymanek, The Art . . ., op. cit., p. 45.
8. J. Stelmach, Code . . ., op. cit., p. 33.
9. R. Alexy, Theory . . ., op. cit., p. 51.
10. J. Habermas, Vorstudien und Ergänzungen zur Theorie der kommunikativen Handelns,
Frankfurt am Main, 1984, p. 160, p. 174 ff.
11. See R. Alexy, Theory . . ., op. cit., p. 273.
12. J. Habermas, Wahrheitstheorien, in H. Fahrenbach (ed.), Wirklichkeit und Reflexion,
Pfullingen, 1973, p. 239 ff.
13. See N. Luhmann, Legitimation durch Verfahren, 4th edition, Frankfurt am Main
1983.
14. R. Alexy, Theory, op. cit. . . ., pp. 33–38, 263–272, 356–359; J. Stelmach, Code . . ., op.
cit., pp. 26–31.
15. Ch. Perelman, Legal . . ., op. cit., pp. 160–162.
16. As for arguments, Perelman appeals to the classification of G. Tarello presented in
the article “Sur la spécificité du raisonnement juridique”, Archiv für Rechts-und
Sozialphilosophie 7 (1972), pp. 103–124, whereas as regards the classification of legal
principles, he draws on a juxtaposition made by G. Sruck and described in Topische
Jurisprudenz. Argument und Gemeinplatz in der juristischen Arbeit, Frankfurt, 1971,
pp. 20–34.
17. Ch. Perelman, Legal . . ., op. cit., pp. 129–139; J. Stelmach, Code . . ., op. cit.,
pp. 85–106.
18. Ch. Perelman, Legal . . ., op. cit., pp. 145–147.
19. Th. Viehweg, Topik und Jurisprudenz, 5th edition, München, 1974, pp. 19–30.
20. Ch. Perelman, Fünf Vorlesungen über die Gerechtigkeit, München, 1967, pp. 158.
21. Ch. Perelman, Justice, Law and Argument, Boston, 1980, p. 73.
22. R. Alexy, Theory . . ., op. cit., pp. 225–232.
23. Ibidem, pp. 234–235.
24. Ibidem, p. 285.
25. A. Aarnio, R. Alexy, A. Peczenik, Grundlagen der juristischen Argumentation, in
W. Krawietz, R. Alexy (eds.), Metatheorie juristischer Argumentation, Berlin, 1983, p. 42.
ARGUMENTATION
165
26. A distinction between adequate, lame and leaping theories was introduced by
.
.
Petrazycki. See L. Perazycki, Wste˛p do nauki o prawie i moralności [Introduction to
the Theory of Law and Morality], Warszawa, 1959, pp. 128, 139, 153.
27. See a statute from June 6 1997, Polish Code of criminal law (Book of Statutes, No
89, position 555 with later changes), art. 74, 175, 182, and 186.
28. J. Stelmach, Code . . ., op. cit., pp. 47–50.
29. Rules 4 and 5 are mentioned by R. Alexy in Theory . . ., op. cit., pp. 234–235.
30. R. Alexy, Theory . . ., op. cit., p. 255.
31. See J. Stelmach, Kodeks . . ., op. cit., pp. 72–86; also, Ch. Perelman, Logika . . ., op.
cit., pp. 90–95.
32. R. Dworkin, Taking . . ., op. cit., p. 22. See also J. Stelmach, R. Sarkowicz, Philosophy . . .,
op. cit., p. 50 ff.
33. See G. Struck, Topische Jurisprudenz . . ., op. cit., pp. 20–34; also J. Stelmach, Kodeks . . .,
op. cit., pp. 86–105.
34. L. L. Fuller, Moralność prawa, Warszawa, 1978, p. 68.
CHAPTER 5
HERMENEUTICS
5.1
INTRODUCTION
Hermeneutics is one of the oldest and most disputed of all the philosophies
of interpretation. There exist equally eminent advocates and opponents of
this philosophy. Its opponents, at least in contemporary philosophy,
were most frequently adherents of analytical philosophy. They perceived
hermeneutics as a threat to their own philosophical autonomy, and, in consequence, attacked it, its fundamental assumptions and, in their view, its
unclear and imprecise language with surprising vehemence. They often
acted, in the spirit of the principle derided by Gellner, whereby if you cannot prove the convictions of a rival are false, you should declare that they
make no sense.1 Mutual tensions lost their significance with the passage of
time, one reason being that there appeared different “frontier hermeneutics”, in particular analytical hermeneutics and hermeneutics understood as
a theory of communication.
5.1.1
The Beginnings of Hermeneutics
The term “hermeneutics” comes from the Greek word έρµηνέιν which
denotes the art of prophesying, translating, explaining, interpreting.
Over time, the meaning of this term was enriched and supplemented.
The genesis of the notion of hermeneutics is also associated with the
name of the messenger of the gods, Hermes, who, as is well known, was
believed to have created language and writing. In philosophy this word
appeared in Aristotle’s Peri hermeneias. The term he used – “hermeneia” –
expressed a connection between interpretation and understanding, since
heremeneia is a meaningful enunciation which says “something about
something” and grasps reality by means of expressions. It should be
stressed, though, that the above mentioned work of Aristotle was no systematic exposition of hermeneutics but only a part of Organon – a part
comprised of a description of a certain kind of logical grammar. This
logical grammar was to deal with analysis of the structure of language –
the structure of propositions – without being limited to the examination
of their truthfulness. In modern times the term “hermeneutics” appeared
167
168
CHAPTER 5
in the title of a whole work of Dannhauer – Hermeneutica sacra sive
methodus exponendarum sacrarum litterarum (published in 1654).
Until the nineteenth century, hermeneutics was being developed mainly
in the form of particular “theories” formulated in the fields of theology,
philology and jurisprudence. It was only due to Schleiermacher and
Dilthey that general philosophical (humanistic) hermeneutics arose. As
a result, hermeneutics no longer denoted only the art of interpretation
and understanding of texts: Dilthey elevated it to the objective, universal
methodology of the humanities, which he called “the methodology of
understanding”, whereas Husserl, Heidegger and Gadamer interpreted
hermeneutics as the ontology of understanding. As mentioned, there
appeared also “frontier” accounts of hermeneutics. At the same time, the
scope of application of hermeneutical philosophy was on the increase.
Apart from the disciplines already mentioned – theology, philology and
jurisprudence – hermeneutics was also to find its application in history,
sociology, psychology, political sciences and even economics.
The first period of development (until the end of the eighteenth century) of the philosophy of interpretation under discussion can be illustrated by presenting three particular hermeneutics: biblical, philological
and legal.
Biblical hermeneutics. Biblical hermeneutics was already known to the
authors of Halakha and Aggadah. Hermeneutics was then understood
as an art of the exegesis of the Biblical text. The art of exegesis, explanation and interpretation of the Scriptures was perfected by successive
rabbinic generations. Over time, a conception of revealed, inspired and
prophetic understanding also emerged, which was to play an important
role in late Christian hermeneutics based on Jesus’ activity.
As a result of “the Christ’s event” there arose an extremely important
and fast developing trend in biblical hermeneutics, whose main objective
was to explain the whole of the Scriptures. In the first centuries of
Christianity there existed a sharp conflict over interpretation, which gave
rise to an urgent need to construct a uniform theory of Scriptural interpretation. This theory was to ensure uniform understanding of the whole
biblical tradition. “New hermeneutics” was to make possible the choice
of a proper theory of biblical interpretation. The first works devoted to
the interpretation of the scriptures – works that established rules of interpretation enabling internal coherence to be reached in the understanding
of the Old and New Testament – arose as early as the second and third
century A.D. At that time two hermeneutic schools, the Alexandrian
and Antiochian, concerned with explanation of the Scriptures were
HERMENEUTICS
169
already active. Among the main representatives of the Alexandrian
school are Justin, Tertulian, and Origen; this school developed an allegoric interpretation of the Bible. According to Origen, one can assign
three dimensions of meaning to the Scriptures: corporal, i.e. historical
and available to all the faithful, psychological, and pneumatic (spiritual), which can be reached only by a few scholars analyzing the scriptures. One may obtain spiritual insight only by means of the allegorical
interpretation. The Antiochian school was founded later than the
Alexandrian one in the fourth and fifth centuries by Theodoros and
Diodoros. This school promoted a literal interpretation of the Scriptures.
The allegorical method preferred by the Alexandrian school was
replaced with a critical method of exegesis based on philological and
historical research.
Yet, at that period, St. Augustine was author of the most influential
and coherent theory of biblical interpretation. In book three of his
De doctrina christiana, he presented his view of the role and function of
biblical hermeneutics and conducted a philosophical analysis of the
process of understanding. In particular, he undertook an analysis of
the notion of a sign, which he defined as a medium of thought, stressing
at the same time that every theory of exegesis must have its own theory
of sign and of meaning. He developed a conception of the reasons for
the Bible’s incomprehensibility and described the core rules governing
interpretation of the Scriptures. Having made due allowance for the
significance of historical and philological research, he went much further,
undertaking an analysis of the phenomenon of understanding as something which is conditioned by faith. Thus, he attained the concept of a
mystical – illuminated – understanding underlying the real (inspired,
revealed) interpretation of the Scriptures.
Of course, in the discussed period, many other philosophers and theologians (for instance, Eucherius from Lyon in the fifth century, Julius
African in the fifth century, Cassiodor in the sixth century) also tackled
the problem of biblical exegesis.
In modern times, a work of special significance was Flacius’ Clavis
Scripturae Sacra (1567), which is the author’s attempt to set forth rules
for interpreting the Bible in the form of a systematic set. The application
of these rules was to enable the accomplishment of universally valid
understanding of the Scriptures. Flacius also formulated a general
hermeneutical principle, according to which a part of a work can be
understood only if it is related to the whole work, and to its other parts.
Thus, presumably for the first time, the principle of a hermeneutical
circle was spelled out. In Flacius’ view (before Flacius a similar view was
170
CHAPTER 5
held by Melanchton), every written work is created in accordance with
the same rules and principles. Accordingly, one may build a universally
valid theory of the interpretation and understanding of a text (in this
case the Bible) without appealing to tradition. (N.B. this conviction was
incompatible with the position of the Trident Council.) Under this
account, interpretation becomes a sort of logical game allowing one to
reconstruct the structure of an analyzed text. In sum, Flacius’ conception undoubtedly constitutes an important stage in the process of building
general humanistic hermeneutics.
In the seventeenth and eighteenth centuries there appeared new works
concerning the interpretation and understanding of biblical texts. Good
examples are the works of the following authors: Dannhauer, the author
not only of the afore-mentioned Hermeneutica sacra sive methodus exponendarum sacrarum litterarum, but also of Idea boni interpretis;
Chladenius, the author of Einleitung zur richtiger Auslegung vernünftiger
Reden und Schriften, and Baumgarten, the author of the five-volume
work Nachrichten von einer Hallischen Bibliothek. Three scholars from the
eighteenth century – Semler, Michaelis and Ernesti – wrote in the same
hermeneutical vein – called theological rationalism – to which Baugmarten
belonged. Semler undertook an analysis of grammatical and historical
interpretation, proposing a specific rational variety of theological
hermeneutics, the outline of which is contained in his work Vorbereitung
zur theologischen Hermeneutik. In principle, Michaelis works along the
same lines, grounding his hermeneutical method in philological and historical research. In his view, in the process of interpreting the Bible, one
should allow for both historical context and common sense. Finally,
Enesti, the author of Interpretes dealt with philological aspects of the
Scriptures’ interpretation. His philological hermeneutics was to ensure
the harmony of biblical revelation and rational thinking.
Thus, two hermeneutics – biblical and philological – begin to form
one whole, thereby providing the foundations for general humanistic
hermeneutics. Philological hermeneutics becomes only one method available to hermeneutics (first biblical, later general hermeneutics).
Consequently, it is no longer necessary to separate both varieties of
hermeneutics (biblical and philological), because, according to Meier –
the author of Versuch einer allgemeninen Auslegungskunst – there exists
one general hermeneutical theory, which formulates rules to be taken
into consideration while interpreting all kinds of signs.
Philological hermeneutics. The origins of philological hermeneutics, in
turn, reach back to the beginnings of philosophy in ancient Greece. It
HERMENEUTICS
171
began with attempts to interpret Homer’s poetry, which were undertaken
mainly for didactic purposes. Philological interpretation enabled the text
to be understood from a grammatical and literary perspective; this interpretation was in fact a sort of game between an interpreter and a text.
More solid grounds for philological hermeneutics were provided at the
moment when it was allied with rhetoric – a discipline concerned, not
only with oratory art, but also with more detailed issues, such as, for
instance, the composition of a literary text, the principles of creating
rhymes, or the conditions under which metaphor may be used. An
important contribution to the development of philological hermeneutics
in ancient times was made by the Alexandrian school of philology (during the second century B.C.). According to its main representatives –
Aristarch and Hipparch – philology is a discipline based on a profound
understanding of language, an art of refined critique and interpretation
of a literary text. In a somewhat different direction research developed in
the Pergamon school, whose main representatives included Crates from
Mallos and Aryston from Chios. Pergamon philology was concerned,
amongst other things, with stylistics and rhetoric, developing the principle of allegoric interpretation – known already to the Stoics – which was
to play a significant role in later philological hermeneutics.
In modern times there appeared many works devoted to philological
hermeneutics. The classically philological treaties of authors such as
Scoppius, Clericus and Valesius were already, in fact, complete expositions of hermeneutical theories. In the first part of these works, one can
find catalogues of interpretative rules, as well as a discussion of applicable philological methods, which make possible the critical analysis and
interpretation of literary texts. As we have already pointed out, somewhat
later, a number of works were published that were devoted both to philological and biblical hermeneutics (these are already mentioned works of
Dannahuer, Baumgarten Semler, Michaelis, Ernesti and Meier).
Philological hermeneutics played an especially important role in nineteenth- and twentieth-century general philosophical hermeneutics. In
many ways the conviction that the primary form of hermeneutics is
philological was confirmed. Everything begins with, and not infrequently also ends with, language; understanding can be articulated only
through language – there exists an exceptional agreement between representatives of different hermeneutical trends as far as this point is concerned. According to Schleiermacher, all that is presupposed by
hermeneutics is language. Dilthey spoke about hermeneutics as a theory
of the art of understanding the manifestations of life which are fixed in
language; accordingly, he asserted that literary critique is inseparably
172
CHAPTER 5
connected to hermeneutical process (immanently belonging to it).
According to Heidegger, “language is the home of being”; Gadamer, in
turn, asserted that “the only being that can be understood is language”;
in his view, language is the only medium through which we can accomplish both successful communication with a partner and understanding
of things themselves.2 Thus, it should be remembered that both early biblical hermeneutics and contemporary philosophical hermeneutics rested
on philological hermeneutics.
Legal hermeneutics. Legal hermeneutics is somewhat different. At least
until the eighteenth century it developed separately, forming an integral
part of the methodology of jurisprudence. Of course, one cannot claim
that legal hermeneutics was completely isolated. Even pragmatically disposed Roman jurists, who dealt primarily with concrete cases, appealed to
certain ontological, axiological and methodological conceptions worked
out in the fields of general philosophy and particular hermeneutics: biblical and – above all – legal hermeneutics. Undoubtedly, the foundations of
the science of legal interpretation were built by Roman jurists. It is possible to question further whether legal hermeneutics existed in the methods
cultivated by Roman jurisprudence. The answer to this question will hinge
on what definition of “legal hermeneutics” is assumed. If one considers as
hermeneutical each theory of interpretation and understanding of a text
(in this case – a legal text), then, of course, it will be possible to say that
legal hermeneutics had already been developed within the framework of
Roman jurisprudence. In the early period, there was a preference for the
literal interpretation of a legal text, which – with the passage of time –
developed to assume the form of grammatical, philological and historical
interpretations. Use was also made of philosophical methods: rhetoric and
Aristotelian topics, amongst others; as for the latter method, it was most
likely used first by Cicero for the purposes of legal interpretation.3
In modern times many works devoted to legal interpretation were produced, yet for the most part they were systematizing, i.e. their objective
was to present catalogues of universally valid methods of interpretation
with a view to making a “proper” interpretation of the law possible. In
this context one can list such works as Hermeneutica iuris, recensuit perpetuisque notis illustravit of Eckhardi, Principia et subsidia hermeneutica
iuris of Wittich and Hermeneutik des Rechts of Sammet.
It was not until the nineteenth and twentieth centuries that an essential change occurred in how hermeneutics was conceived and cultivated.
Schleiermacher, and later Dilthey, put forward a new, universalistic
account of hermeneutics, whose task – in their view – was to work out a
HERMENEUTICS
173
methodological basis for all humanistic disciplines, including jurisprudence. Still another kind of ontological and methodological universalism was brought into play with the phenomenologically oriented
hermeneutics of Husserl, Heidegger, Gadamer and Ricoeur. The influence
of this variety of hermeneutics on jurisprudence can be described as
follows. Jurisprudence believes that philosophical hermeneutics offers the
opportunity to solve jurisprudential problems (which are mainly of a
methodological nature). The acceptance of entire hermeneutical conceptions, and of its particular theses, has been accomplished. The older,
technically understood legal hermeneutics continues to lose significance,
becoming in fact a mere object of historical research. The binding link
between old legal hermeneutics and more contemporary – philosophical –
versions is the conception of Savigny, expounded in his work Juristische
Methodenlehre. On the one hand, Savigny defended the methodological
autonomy of jurisprudence, but on the other hand, he availed himself of
solutions proposed by Schleiermacher. Thus, it was not “a pure reception” (which, by the way, will become characteristic for legal hermeneutics inspired by philosophical conception). The thinkers who appealed to
the hermeneutical tradition of were Coing and Betti, and to the
hermeneutical tradition of Heidegger and Gadamer – Reinach,
Maihofer, and Kaufmann. However, it is difficult, if not impossible, to
classify unambiguously the views of many contemporary representatives
of legal hermeneutics. To give an example: how should the views of
Larenz or Esser be classified? Their versions of hermeneutics were modified and transformed to a high degree (mainly for the purposes of a dispute over the method of jurisprudence pursued in the German science
of law at the time). We shall return to these issues in Sections 5.2–5.4
whilst considering different versions of philosophical and legal
hermeneutics.
5.1.2
What Do We Not Know About Hermeneutics?
Now we would like to deal briefly with the most hotly disputed issues
connected with every possible lecture on philosophical hermeneutics.
Criticism of these issues usually constitutes the point of departure for all
approaches that reject hermeneutics as a philosophy of interpretation.
We shall successively present eight of the most frequently formulated
objections against hermeneutics. (1) It has often been emphasized that
no single, and acceptable to all adherents of the hermeneutic approach,
definition of hermeneutics exists. (2) As a result, it is very difficult to set
the boundaries between particular hermeneutical conceptions (this
concerns internal boundaries), and between those positions which are
174
CHAPTER 5
hermeneutical and those which are no longer so (this concerns external
boundaries). As a result, the concept of “hermeneutics” has often been
abused, because it has been used to assess different interpretive philosophies (of the analytical, structuralist and argumentative types), which do
not necessarily have much in common with hermeneutics. (3) A similar
situation exists in the case of other – for hermeneutical philosophy –
basic concepts: understanding, pre-understanding and the hermeneutical
circle. Unambiguous definitions of these notions are lacking; besides,
these notions are given fundamentally different interpretations in different hermeneutical conceptions. (4) Ultimately, we are not able to determine what is meant by the term understanding: a form of cognition, a
form of existence of an individual being, or perhaps both. (5) Even if we
assume that understanding is also (besides representing a form of being)
a form of cognition, we are not in a position to determine what kind of
cognition it is: direct or indirect. As shall be argued more extensively in
the last section of this chapter, plausible arguments exist for both alternatives. (6) In assuming that understanding is a form of direct-intuitive
cognition, we are confronted with a further problem, namely that of
determining what type of intuition we are ultimately dealing with: psychological, analytical or, rather, phenomenological.4 (7) The thesis of
hermeneutical universalism is not entirely clear. Dilthey’s defense of this
thesis, based on a division of naturalistic and anti-naturalistic methodologies adopted is quite weak, for the division itself gives rise to serious
doubts. Moreover, it represents a particular type of universalism,
because it is limited solely to the field – difficult to define unambiguously –
of humanistic disciplines. It is also risky to defend the phenomenological thesis of hermeneutics’ universalism as “the first science”. This argument can be refuted by reversing it, i.e. by claiming that hermeneutics
finds its main, if not exclusive, application specifically in humanistic disciplines and not in the pure and natural sciences. (8) Ultimately, we do
not know exactly how hermeneutics can be applied usefully in interpreting law, or the potential and admissible scope of its application, considering that the object of understanding and interpreting law may often be
the regulations (norms) of valid law. This problem is closely connected
with another – the freedom of interpretation, which is restricted, at least
in continental legal systems, by a prohibition on making interpretations
contra legem.
Do the above objections provide grounds for rejecting hermeneutics as
a method, or philosophy of interpretation? We would answer this question in the negative: we include hermeneutics in a group of the basic
methods of jurisprudence. Yet we must be mindful of the specificity of
HERMENEUTICS
175
hermeneutics as a method of jurisprudence. Its specificity lies in the fact
that it is not a method in the strict sense of this word. Hermeneutics is
different from logic, analysis and argumentation in that its structure,
assumptions, procedure, and inter-subjective criteria for acknowledging
and rejecting solutions can hardly be reconstructed. Hermeneutics – as a
philosophy of interpretation “without the Archimedean starting-point” –
does not offer a method, but, rather, intuition – understanding, that is:
something which is “softer” and deprived of formal structure. A call for
this kind of philosophy of interpretation appears at the level Ricoeur
terms reflexive-existential, especially with reference to cases of interpretation that are usually called “hard”, where more formal methods are
simply insufficient and useless.5 In order to decide such cases, the interpreter cannot help appealing to unconventional methods. In addition to
legal values, they must appeal to social, economic and political phenomena,
or – ultimately – to some sort of intuition enabling the understanding of
a difficult case.
Moreover, one must not forget that both in general philosophy and in
the philosophy and theory of law, most assumed views and theses are
also “soft”, even if they arise in the field of “harder” methodologies. It
is easy to oppose such views and theses (manifesting a similar level of
precision) which will have the same or even better justification. It is to be
noted, however, that the problem of defining basic notions is encountered not only in the case of hermeneutics, but also in the case of other
methods. Similarly to the defenders of hermeneutics, supporters of the
application of logic, analysis and argumentation in legal interpretation
claim the universality of their methods. Furthermore, a dispute over the
usefulness and the scope of application is pursued not only in hermeneutics, but also in remaining methods. Thus, notwithstanding the many
fundamental differences between hermeneutics and other methods, they
also have many points in common, though these, unfortunately, are often
controversial.
5.2
HERMENEUTICS AS EPISTEMOLOGY
We begin our presentation with a discussion of methodological current in
philosophical (Section 5.2.1) and legal (Section 5.2.2) hermeneutics. At
this point, we wish to enter the reservation that the division here introduced into epistemological and ontological approaches is neither sharp
nor unambiguous with reference to some views. The problem of classification will emerge with reference to such philosophers as, for instance,
Habermas, Apel, and to such lawyers as – to give two examples – Larenz
176
CHAPTER 5
and Esser. This is why no special or ultimate significance should be
attached to the divisions and systematization proposed below.
5.2.1
Methodological Current in Philosophical Hermeneutics
The new – epistemological – tradition in hermeneutics was initiated by
Schleiermacher. Another notable representative of the methodological
current in hermeneutics was Dilthey. We shall discuss each of their views
in turn.
Schleiermacher. The scientific climate that prevailed in Germany at the
turn of the eighteenth century was favorable to Schleiermacher’s projects: on the one hand a theory of the interpretation of works of art, proposed by Wickelmann, was developing; on the other hand, there arose
the idea of “entering into the spirit of something” (epochs, peoples) in
the work of Herder, as well as new philological conceptions of interpretation, developed by Heyne, Wolf and his disciples (especially Heindorf).
A breakthrough occurred as far as the attitude of philosophy towards
history is concerned. There also appeared a strong need for a “secondary” understanding of the historical world – this need was revealed in
the views of scholars from this period such as Hegel, Böckhe, Dissen,
Rank, Savigny. A thinker who exerted a considerable influence on
Schleiermacher was Schlegel, who encouraged him to take up the task of
translating Plato into German.6
The sources of Schleiermacher’s hermeneutics are complex. They
embrace Plato’s philosophy (which fascinated Schleiermacher), biblical
hermeneutics, philology, literary criticism, philosophy of history, and –
finally – psychology. Schleiermacher makes an attempt to create universal
(at least from the viewpoint of the humanities), philosophical hermeneutics grounded in philology and psychology. In his view, hermeneutics is
not – strictly speaking – theoretical knowledge, but rather a practical art
of interpreting and understanding all kinds of text (though mainly written
ones).7 As a practical art, it must be – and it is – closely connected with
criticism. Schleiermacher started from philological research, yet he did not
confine himself – as his predecessors had done – to drawing up catalogues
of universally valid rules of interpretation. He went much further, because
he embarked upon an analysis of the process of understanding, which
underlies every interpretation. In Schleiermacher’s view, a degree of the
artistry of interpretation is directly dependent on a degree of understanding; he distinguished two kinds of understanding: clairvoyant and comparative, based on material and grammatical-historical cognition. Both
kinds of understanding are complementary – they operate together.
HERMENEUTICS
177
The process of understanding possesses, in Schleiermacher’s view, historical and psychological dimensions. In order to reconstruct this process,
we must recreate the historical and psychological situations in which the
author of an interpreted work found himself. We must endeavor to understand him better than he understood himself.8 Understanding is a relative
and never-ending process. It has a circular character. Schleiermacher
devoted much attention to the problem of the hermeneutical circle (which
is expressive of the nature of the process of understanding and interpretation), stating that “(. . .) the unity of the whole can be understood on
the basis of single parts, and the value of single parts can be established
(understood) on the basis of the unity of the whole”.9 Thus, when taking
up an interpretative activity, one must begin with a cursory view of the
whole work that one wishes to interpret (translate). This initial understanding will be a necessary condition for further interpretation.
Schleiermacher gave up the division – assumed by his predecessors – into
grammatical, historical, aesthetic and material interpretation. He developed several canons of interpretation, and illustrated their function,
using the exegesis of the New Testament as an example.
Thanks to Schleiermacher hermeneutics became a philosophical problem. According to his account, hermeneutics is a universal theory of the
cognition of the products – expressed in (written) language – of human
words, or, to put it differently, a universally valid method for the humanities. This method appeals to three types of analysis: critical-philological,
psychological and historical.
Dilthey. Thus, Dilthey, whilst formulating his conception of
hermeneutics, already had at his disposal a prepared theory of humanistic interpretation connected with a philosophical theory of understanding. He was fully aware of this fact, and this conviction found
expression in his 1900 essay The Arising of Hermeneutics. The objective
that Dilthey set himself, though, was different – more far-reaching – to
that pursued by Schleiermacher: he attempted to create “a methodology of understanding” for the humanities. The novelty of this, however,
is not to be overestimated, because Vico, in his Scienza Nova, had
already written about the understanding of science, which was to constitute an alternative to the Cartesian model of science based on mathematics. A point of departure for Dilthey was the opposition between
natural sciences and human sciences. This anti-naturalistic division was
introduced into the methodology of the humanities by Droysen, who
distinguished and opposed two aspects of our cognitive reality –
namely, nature and history. The purpose of the natural sciences is,
178
CHAPTER 5
according to Droysen and Dilthey, explanation, whereas the purpose of
the humanities is understanding.
In order to capture the essence of the process of understanding, it is
necessary to appeal to psychology, which Dilthey treats as a descriptive
discipline. In his opinion, what is at issue in the humanities is not a
methodical knowledge of psychological processes, but a repeated experience of these processes, that is: an understanding of them. The purpose
of the humanities is to know the objectified products of human life, and
this knowledge simply amounts to understanding. Thus, understanding
is a fundamental category of Dilthey’s hermeneutics – the most typical
activity encountered in the humanities. This is a process through which
one achieves cognition on the basis of those manifestations of psychic
life which are given to senses. And even though the manifestations of
psychic life which are given to the senses are immensely variable, the very
purpose of this sort of cognition ensures that the understanding of these
manifestations must have common features. Dilthey’s conception of
understanding appealed to the psychological principle of the identity
of human nature. In particular, understanding can be reduced to the
psychological operation of “putting oneself ” (Hineinversetzen) in the psychological situation of the person whose work is being interpreted.
However, in his later works, Dilthey departed from these strongly
psychological assumptions.
Under the unquestionable influence of Schleiermacher, Dilthey considered those expressions of human life (spirit) which are fixed in writing
to be the privileged objects of understanding. Accordingly, he defined
hermeneutics as a theory of the art of understanding the manifestations
(which are fixed in writing) of life – of the traces of human existence
contained in language. What underpins this definition is his conviction
that it is only through language that human inner life (spirit) finds full
expression. This is also why Dilthey asserted that the primary, profound
meaning of hermeneutics is philological. An examination of these –
fixed in writing – manifestations of life, if compatible with rules, is called
interpretation. Understanding, which is attained through the medium of
interpretation, is objective, because its subject is the whole human species
(one can easily notice here a reference to the Kantian concept of the
transcendental subject). Understanding is ultimately the most fundamental activity encountered in the humanities. Dilthey also assumed the
principle of the hermeneutic circle – described already by Flacius, Ast
and Schleiermacher – stating that “(. . .) From what is singular – the
whole, from the whole – again what is circular. The whole of a work
requires proceeding towards the individuality of the author and towards
HERMENEUTICS
179
literature with which it [its individuality] is connected. Only comparative
activity enables me ultimately to understand each and every single
work – and even each single sentence – more profoundly than I understood it before. Thus, understanding is realized on the basis of the whole
and yet the whole is realized on the basis of what is singular (. . .)”.10
Ultimately, though, “a methodology of understanding” cannot be
reduced to the hermeneutics of a text. Its objects are all the products of
human life (spirit). Thus, hermeneutics was elevated to the status of an
epistemology of the humanities. Dilthey wanted to provide the humanities with a method equal in objectivity to that at the disposal of the natural sciences. For precision, we must add that the sharp opposition
between understanding (characteristic of the humanities) and explanation (characteristic of the natural sciences) which he initially defended
was mitigated in his later works. In these works, he asserted that understanding and explanation are two complementary research steps: explanation is usually the initial step, which often takes so long that few
succeed in achieving full understanding.
It is easy to notice that Dilthey’s hermeneutics also has more complex
sources. One can undoubtedly number among them philology and
broadly understood literary criticism (for, in Dilthey’s view, hermeneutical process is inseparably connected to literary criticism, which immanently belongs to this process), as well as descriptive psychology and
anti-naturalistic methodology, which grew from an unfortunate opposition between understanding and explanation. It is difficult to resist the
impression that this conception of hermeneutics is not free from contradictions. On the one hand, Dilthey conceives of hermeneutics in an individualistic manner – as “the art of ingenious interpreters”, and of
understanding as a psychological operation of “putting oneself in” the
position, or psychological situation, of the author of an interpreted
work; on the other hand, though, hermeneutics is to be an objective and
universal method of the humanities, appealing to such categories as
“life”, “human species”, and “history”.
Receptions of Dilthey’s hermeneutics. All the above remarks notwithstanding, Dilthey’s conception became an important source of inspiration for many authors continuing studies in the field of hermeneutics
(also its legal variety) and, interestingly enough, informal logic. There
was even a Diltheyan school founded in Götingen, in which leading roles
were played by Misch, Lipps, König and Plessner.
According to Misch, the world we live in is the world of expressions.
Only this world – of language – is universal. Dilthey was still speaking of
180
CHAPTER 5
a strict connection between experiences, expressions and understanding.
According to Misch, we move only in this second world, i.e. in the world
of expressions. The “logic of expressions”, informally understood, is to
ensure a proper interpretation and understanding of this world.
Dilthey’s hermeneutical philosophy is also a point of departure for
Lipps, who supplements his philosophy with assumptions from existential philosophy and philosophical anthropology. As with Misch, Lipps’
logic is informal and this logic aims to enable analysis of the category of
“speech”, which is fundamental from the point of view of hermeneutical
cognition. This category is transcendental, which guarantees that logic
used for the purposes of its analysis may be objective, making – in addition – possible “a return to the sources” of our knowledge, i.e. to preunderstanding (Vorverständnis). This pre-understanding is a point of
departure for every possible further cognition (understanding).11
Still another contemporary continuation of Dilthey’s thought was
given by Betti. Under his account, hermeneutics is both a theory of cognition and a methodology of the humanities. The method of the humanities is reduced to interpretation, mainly the interpretation of a text. Betti
is the author of one of the most extensive works devoted to a theory of
humanistic interpretation. (N.B. this work bears the meaningful title
Allgemeine Auslegungslehre als Methodik der Geisteswissenschaften.) In
systematizing and arranging the results of the efforts of many generations of philosophers, philologists and lawyers, this work brought a
period of evolution in hermeneutical theories to a close: the vein which
we have termed methodological. The work was started by, inter alia,
Flacius, Meier, Schleiermacher and Dilthey was thus – at least to some
extent – brought to an end.
Analytical hermeneutics. The variety of hermeneutics described as analytical requires separate treatment. This immediately gives rise to doubt
over whether hermeneutics can be analytical, given that the most serious
objections levied against hermeneutics were formulated by the representatives of analytical philosophy. If the possibility of such a combination
is admitted, then all boundaries – even between competing philosophies
of interpretation – are entirely reduced in significance. This may have the
following – momentous – consequences, which are in accordance with
the spirit of postmodernism, a theory which we find hardly acceptable
because there are no boundaries, no methodological paradigms (projects),
everything, or almost everything is allowed, and one type of reflection
turns into another. It poses the threat of anarchy, of the deconstruction
of everything that made sense in methodology and led – partly at
HERMENEUTICS
181
least – to acceptable results; postmodernism makes it difficult – or even
impossible – either to make a correct exposition of, or to learn a method.
The authors of this book are decidedly against such simplification. On
the other hand, though, we must be aware of these mutual influences and
therefore unavoidable obscurities, since this will enable us to free ourselves of the necessity of pursuing never-ending disputes over the
method of the humanities.
Everything seems to turn on the definition of such notions as “analysis”, “language”, “logic”, “objectivism”. According to Bocheński,
whether or not a philosophy is acknowledged as analytical will depend
on how these four notions have been defined. At the same time, we would
like to stress that all these notions found a place in the lexicon of
hermeneutical philosophy (including the variety which we have termed
epistemological – methodological). The representatives of ancient
hermeneutics and, later, Schleiermacher and Dilthey spoke about philological analysis. The program of language research was introduced into
hermeneutics through the medium of philology. As was emphasized both
by Schleiermacher and Dilthey, analysis (interpretation) of a written
(oral) text is a pre-condition for fully understanding it. Misch and Lipps
also wrote about hermeneutical logic. Dilthey devoted much attention to
the issue of the objectivity (universality) of hermeneutic cognition (i.e. of
the method of understanding). Thus, apart from well-known differences
between analysis on the one hand, and phenomenological hermeneutics
on the other, there are many similarities between these views. This fact
was clearly realized by the analytical philosopher Bocheński: he noted
that both views imply that analysis is necessary, and “want to” proceed
objectively (“to the things themselves” – “zu den Sachen selbst”); moreover, both views stress the analysis of language. Symptomatic of this is
Bocheński’s assertion that analytical philosophers should admit the existence of other types of analysis (for instance hermeneutical – our
remark: J. S., B. B.) besides their own – rather radical – version of it.
Taking into account all the above remarks and reservations, it is possible
to speak of analytical hermeneutics. We mention it during our presentation of the epistemological trend, because analytical hermeneutics is a
method of textual interpretation. In twentieth-century hermeneutics one
can find numerous references to linguistics and analytical philosophy.
The thinker who exerted a deep influence on this kind of thinking was
“the second” Wittgenstein. In his Philosophical Investigations, he presented the problem of interpretation as a sort of language game played
between an interpreter and a text. This account of the problem of interpretation is appealed to particularly willingly by representatives of the
182
CHAPTER 5
philosophy of law (who will be discussed in the following section, which
is devoted to the incorporation of hermeneutical methods into legal
interpretation). Hermeneutics has also been cultivated analytically in the
terrain of contemporary theology: thanks largely to Fuchs, Ebeling and
Robinson, biblical hermeneutics has become “a believing science of
language”.
Hermeneutics as a theory of communication. Still another conception of
hermeneutics was put forward by Habermas and Apel. Under their
accounts, hermeneutics is a theory of the processes of ordinary language
communication – a theory that possesses from the start some epistemological sense.
In the hermeneutics proposed by Habermas, references can be found to
methodological hermeneutics (both its classical and analytical versions),
phenomenological hermeneutics (especially the assumptions of Husserl
and Heidegger), and even to psychoanalysis. For instance, Habermas
appeals to the notion (previously used by Husserl) of Lebenswelt
(Habermas speaks in this context about soziale Lebenswelt), but also to
Heidegger’s account of understanding (Verstehen) and individual existence (Dasein). He is, however, skeptical about Gadamer’s philosophy of
understanding, because, in Habermas’ view, hermeneutics fulfils rather
limited tasks in the sphere of human practice. These tasks are limited to
the explanation and description of the processes of ordinary language
communication. In other words, hermeneutics is the art of understanding
sense which can be communicated through the medium of language (this
account is analogous to Dilthey’s). Of paradigmatic significance for
hermeneutics thus conceived, would be such notions, or categories, as:
communicative action (kommunikatives Handeln), communicative competence (kommunikatives Kompetenz) and communicative community
(Kommunikationsgemeinschaft). Ultimately, the task of hermeneutics
amounts to an examination of the structure of these communicative
actions. The communicative action itself is defined by Habermas as the
mutual influence people exert on each other by means of symbols. This
influence has to be compatible with valid social norms, which are understood and accepted by at least two actors interacting with each other.
Hermeneutics would ensure the possibility of communication between
these actors; it would at the same time condition the self-determination
of social groups as well as the process of individualization of members
of these groups. According to Habermas, the art of communication has
not been acquired by all the members of a given communicative
community. For that reason use must be made of hermeneutics, the task
HERMENEUTICS
183
of which is to work out the principles of this art. These principles are
meant to improve the process of transmitting tradition and communication between actors. Hermeneutics is meant to fulfill both an important
practical function, thereby becoming a concrete skill of language communication, an art of communication acquired by at least some members
of a group (communicative community), and a theoretical and metatheoretical function – thereby becoming a theory of every common
process of language communication. Hermeneutics, understood as a theory of the process of language communication, makes references to particular positive sciences, yet hermeneutics itself is not a positive science.
Thus, ultimately, we may speak about hermeneutical consciousness of
science, but not about hermeneutics as occupying – analogously with
physics, chemistry, biology etc. – its own separate field of research.12
A similar account of hermeneutics was proposed by Apel. In his view
too hermeneutics is a theory of the process of language communication,
and one of its fundamental aspects is the notion of the communicative
community. Unlike Habermas, however, Apel builds his interpretation
on the basis of the philosophy of language, which is – additionally –
analytical in nature. Apel could justifiably be included in the list of
representatives of hermeneutics that we have called analytical. Apel’s
analyses, contained especially in the second volume of his Transformation der Philosophie, serve to confirm this thesis.
Of course, it is neither possible nor necessary to make a sharp and
unequivocal distinction between the two hermeneutical traditions – epistemological and ontological. It is true that the ontology of understanding (which appealed to phenomenological philosophy) separated itself
from the methodological tradition of the nineteenth-century hermeneutics of Schleiermacher and Dilthey, yet it remains the case that
hermeneutics always had both an ontological and an epistemological
sense. The only thing that changed the distribution of accents and the
general approach: the approach of phenomenological hermeneutics was
decidedly anti-psychological. This fact was realized by, for instance,
Ricoeur, because he spoke of hermeneutics as an epistemology of interpretation, at the same time as attempting to solve the conflict between
psychological and phenomenological hermeneutics. Thus, the problems
of vague divisions and the “softness” (defeasibility) of formulated theses
emerge once again.
5.2.2
Legal Receptions
Only in the nineteenth century did an essential change in the cultivation
of legal hermeneutics take place. This was associated with the rise of
184
CHAPTER 5
general philosophical (humanistic) hermeneutics, i.e. the epistemological
trend described above. Successive attempts were made to use philosophical hermeneutics for the purposes of legal interpretation. The process of
incorporation, though, was by no means homogenous. In some cases
direct appeal was made to concrete hermeneutical conceptions, especially to Schleiermacher and Dilthey, in others particular solutions
offered by general hermeneutics were used and subsequently incorporated into the broader context of discussions concerning, for instance,
methods of jurisprudence, the legal decision-making process, adjudication and discovery of the law (Rechtsfindung, Rechtsgewinnung,
Rechtsverwirklichung).
Savigny. The binding link between older, eighteenth-century legal
hermeneutics and its contemporary twentieth-century versions was provided by Savigny, whose conception of hermeneutics refers to
Schleiermacher’s philosophy in many essential respects. A sort of paradoxical aspect of Juristischen Methodenlehre, proposed by Savigny, was
that he aimed to defend the methodological autonomy of jurisprudence
by means of methods that were “external” to the science of law – namely,
hermeneutical and historical. Let us recall Schleiermacher’s assumption
that, in the case of comparative understanding (which is the second –
besides clairvoyant – type of understanding), we must appeal to material,
as well as grammatical-historical, cognition.
He distinguished four basic canons of interpretation: objectivity
(autonomy), unity, genetic and technical interpretation. Savigny, by contrast, emphasized that interpretation of the text of a statute, which aims
to recreate (reconstruct) the intention of a legislator, should embrace
four elements (levels): grammatical, logical, historical, and systematic.
Savigny, like Schleiermacher, was an advocate of comparative understanding as well as the grammatical–historical understanding of the act
of interpretation. Likewise, he adopted the thesis that the act of
hermeneutical cognition (interpretation) is objective in character, and its
purpose is to achieve unity between the work of an interpreter and the
will of a historical legislator. It was a specific kind of objectivity, which –
analogously with Schleiermacher’s conception – was to be justified historically and psychologically. Finally, Savigny attached great importance
to criticism, the so-called “higher criticism” in particular, whose objective was to restore (rebuild) the meaning of a distorted (incomprehensible) text. This “higher criticism” was to consist of the same elements as
any other interpretation. Thus, according both to Schleiermacher and
Savigny, the primary form of hermeneutics is philological.
HERMENEUTICS
185
Coing. An important attempt to apply – above all Schleiermacher’s –
hermeneutics to the needs of the contemporary science of law was made
by Coing. He set forth his proposals in an essay entitled Die juristischen
Auslegungsmethoden und die Lehre der allgemeinen Hermeneutik, in
which he examined the possibility of applying the general canons of
interpretation formulated by Schleiermacher for the purposes of legal
interpretation. He discussed each canon in turn, making various additions and changes of his own: the canons of objectivity (autonomy of
interpretation), unity, genetic interpretation, interpretation of factual
meaning (Schleiermacher spoke in this case of technical interpretation),
and, finally, the canon of comparison. This approach led Coing to the
conclusion that all general canons of interpretation find their confirmation and application in jurisprudence, which ensures the universally valid
interpretation and understanding of a legal text. Hermeneutics teaches us
how to critically assess an interpretation of the law, while at the same time
showing that the science of law as an interpretative discipline uses not
only exclusively deductive procedures, but also other approaches, even
topical ones. Within the framework of this “interpretative discipline”
jurisprudence enjoys a specific status, for it is an example of an “applied
interpretative discipline”. As a result, the ‘legal’ method, both universal
and objective in character, could constitute part of a universal and objective humanistic methodology, while legal hermeneutics would be just one
example of the application of general humanistic hermeneutics.13
Betti. The conception of legal hermeneutics developed by Betti clearly
has its roots in Dilthey’s thought. According to him, hermeneutics is simply the science of interpretation. Interpretation is in fact the only method
that the humanities possess. In Allgemeine Auslegungslehre als Methodik
der Geisteswissenschaften, Betti discussed the most important kinds of
humanistic interpretation, including legal interpretation. Like Dilthey
and Coing, Betti understood hermeneutics to be a universally valid and
objective method of the humanities. Betti linked cognitive objectivism
with axiological objectivism. Legal hermeneutics cannot ultimately be
separated from humanistic hermeneutics in general, which assumes an
objective and universally valid (for all specific variants of hermeneutics)
theory of interpretation and philosophy of understanding.
Larenz and Esser. Representatives of modern philosophy and legal theory make numerous references to a methodological understanding of
hermeneutics, especially the German strand referred to as Methodenlehre. References to hermeneutics, though, are not, as a rule, systematic;
186
CHAPTER 5
rather, they serve to justify particular theses. The hermeneutical conceptions of Larenz and Esser constitute an exception to this. It is worth noting, however, that Gadamer’s philosophy was a source of inspiration for
the conceptions of some authors. This thesis, though, can be disputed on
the following grounds. First, these authors make numerous references to
Gadamer’s views, yet their references are for the most part critical;
accordingly, they advance their own proposals and solutions to
“Gadamerian problems”. Second, even though they reject the Diltheyan
account of hermeneutics as an objective method of the humanities, they
nevertheless assume – unlike Gadamer – a purely methodological understanding of legal hermeneutics. This is the final argument for covering
the views of Larenz and Esser in the section devoted to the epistemological
account of hermeneutics.
Larenz seeks justifications for his axiologically oriented theory of
types within the framework of legal hermeneutics. Like Gadamer, he
rejects the objective conception of understanding, advocated by Coing
and Betti. At the same time Larenz does not assign to understanding an
ontological meaning: he distinguishes clearly between understanding as
a type of cognition and the ontology of law. Thus, ultimately, Larenz
does not accept the Gadamerian interpretation of hermeneutics. For
instance, the concept of application to him has a different meaning than
in Gadamer’s theory, and it is not synonymous with the processes of
understanding and interpretation. Consequently, in Larenz’s view, legal
hermeneutics does not possess “an exemplary meaning”. Rather, it is a
special case on account of a specifically dogmatic interest of “the legal
understanding of a statute”. Differences between Larenz and Gadamer
are also strongly marked over the issues of “pre-judgment” (Vorurteil)
and “pre-understanding” (Vorverständnis). Larenz distinguishes the
meaning of both terms, even placing them in opposition to each other.
The function of the Gadamerian Vorurteil in legal cognition is purely
negative – this is “prejudice” (Aberglaube) rather than “pre-judgment”.
“Pre-understanding” (Vorverständnis), by contrast, designates a certain
kind of interpretative hypothesis, which can be confirmed later by
“a successful interpretation”; the hypothesis in question can also be
referred to as “the expectation of sense” (Sinnerwartung). Ultimately,
pre-understanding is a preparation for “adequate understanding”, thus
it is also a condition for understanding what law is.14
Esser represents a different example of “the mixed reception”. He
attempts to determine a philosophy of interpretation that would be suitable for the needs of legal interpretation. In his view, hermeneutics prepares a social and ideological-critical analysis of the reality of law’s
HERMENEUTICS
187
application. Esser, again unlike Gadamer, conceives of the process of
interpreting positive law as something dogmatic – rather than historical –
in character. It is true that legal hermeneutics makes allowance for the
results of philological and historical research; it remains the fact,
though, that it uses them in the process of “logical interpretation” only
in a dogmatic sense.15 According to Esser, the process of interpretation
can be understood as a kind of practical activity; accordingly, legal
hermeneutics can be defined as the science of action (Handlungswissenschaft). The process of the interpretation or – more broadly – the application of law is creative in character, for it may “produce the content of
norms”. However, Esser defines pre-understanding (Vorverständnis) like
Gadamer, i.e. as a condition for the possibility of understanding. Yet he
associates many different intuitions and meanings with this concept, linking it, for instance, with such terms as “interest”, “attitude”, “motive”,
“expectation”, “background”, “the image of a future decision”, “initial
choice or assessment”, and, finally, “prejudice”. In the course of building his conception of hermeneutics, Esser – like Larenz – stopped at the
level of the methodology of legal understanding. In point of fact, his
hermeneutics can be reduced to a theory of law’s application (Rechtsanwendung) and “finding” a legal decision (Rechtsfindung).
Other representatives of the methodological strand of legal hermeneutics included Forsthoff, Engisch, Müller and Kriele.16 Only occasionally,
however, did these authors appeal to the theses of general philosophical
hermeneutics, which is why one cannot speak, even generally, of a full
reception of this kind of hermeneutics as far as their views are
concerned.
5.3
HERMENEUTICS AS ONTOLOGY
Owing to phenomenology a new, ontological, aspect of the problem of
understanding developed. Understanding is no longer simply conceived
as a method of humanistic cognition, but is also – or, rather, above all –
regarded as a property (form) of the existence of being (to which Husserl
assigned the name Lebenswelt and Heidegger – Dasein). Thus,
hermeneutics has become the phenomenologically oriented ontology of
understanding. Hermeneutics is, however, also a method – after all it has
to be. This duality cannot be eliminated even on the grounds of phenomenological philosophy. This fundamental ambiguity is arguably
the most serious inconsistency in phenomenology. The rejection of the
methodological objectivity of older hermeneutics, accompanied by the
assertion that understanding is a form of the existence of an individual
188
CHAPTER 5
being (Dasein) does not seem very plausible or clear, all the more so
because phenomenology did not abandon its claim to universality
(though it did abandon the claim of objectivity). Hermeneutics is universal, because it is a point of departure for all cognitive activity, it is
“the first science” – “without the Archimedean starting-point”, which
means – besides other things – that hermeneutics is simply a method,
albeit a method that is very difficult to interpret. One may argue that
hermeneutics is a method of direct cognition, because, built on the basis
of phenomenology, it aspires to be “the first science”. Yet if one takes
into account the fact that, on the grounds of hermeneutics, “things
themselves” can be known only through the medium of thoroughly
analyzed language, as well as interpretative operations such as actualization, concretization, appeals to pre-understanding and the hermeneutical circle, then one may conclude that hermeneutics is in fact a
method of indirect cognition. We shall return to this intricate issue in
Section 5.4.
The next point to be stressed is that hermeneutics, like the whole of
phenomenology, is anti-psychological. This stance can be justified in
relation to phenomenological philosophy and the philosophy of consciousness (a variant of the philosophy of cogito – the thinking self), but
in relation to hermeneutics, justification is not so easy. The assertion that
the process of understanding takes place only at the level of “pure consciousness” is a consequence of numerous shady and highly speculative
philosophical assumptions. Whether we interpret certain aspects of the
process of understanding (such as, for instance, pre-understanding) in a
psychological or phenomenological way will depend on our philosophical convictions and habits, as well as the specificity of a concrete case,
rather than on universally valid philosophical truths. Indeed, only the
conflict between phenomenological and psychological hermeneutics
seems objective.
5.3.1
Ontology of Understanding
Husserl. The foundations for new hermeneutics were provided by
Husserl’s philosophy. The variant of hermeneutics which he proposed
was an alternative to Dilthey’s methodology. Husserl criticized both the
naturalistic and anti-naturalistic varieties of methodological objectivity.
As a result, he rejected the conception of hermeneutics as the epistemology of interpretation. He sought other grounds for its justification, finding them in ontology, in which, in his view, the fundamental meaning has
the category of “the world of life” (Lebenswelt). This category is primary
in relation to the objective, cognitive relations “subject – object”. In
HERMENEUTICS
189
other words, life itself, being the primary source of understanding, is
prior to objective cognition. Thus, understanding is no longer a method
of cognition, having become a mode of being.
Heidegger. In Heidegger’s conception, philosophy is equated with the
phenomenology of Dasein. Hermeneutics is, then – neither more nor less
than – the phenomenology of Dasein, i.e. the phenomenology of the
individual existence possessing the capacity for understanding.17
Understanding is no longer one of many psychic activities, or a method
of interpreting a text; it becomes a mode of being – a characteristic – of
individual existence. This is so because Dasein possesses the capacity of
self-understanding and self-interpretation; thus, understanding must not
be reduced to purely cognitive categories. It is true that Heidegger
assumes that understanding is realized in language, yet he adds that in
thinking, being turns into language (language is the home of being).18
Heidegger also gave a consistently ontological interpretation to other
notions typical of hermeneutical philosophy like, for instance, “the
hermeneutical circle”: in his view “the circle” does not describe the structure of the process of understanding, but expresses “the existential
pre-structure of Dasein itself ”.19
Gadamer. A special place in the process of the development of phenomenologically oriented hermeneutics is occupied by Gadamer. His
work, Wahrheit und Methode, ended a stage of development in humanistic hermeneutics, simultaneously confirming the presence and importance of hermeneutical issues in philosophy.
The philosophical roots of Gadamer are complex. He defines himself
as a Platonist posing a Kantian question about the transcendental conditions for the possibility of understanding. He answers this question in
the spirit of Heidegger, whom in fact he treats as his main philosophical
predecessor. Finally, when classifying his own philosophy, he places it
between phenomenology and dialectics.20 Gadamer realized that for
hermeneutics to be a real philosophy of understanding, it cannot confine
itself either to humanistic epistemology (Schleiermacher, Dilthey), or to
fundamental ontology (Husserl, Heidegger). In his opinion, hermeneutics must remain open, since only then can it preserve its claim to universality. This openness means in particular that hermeneutics links
notions which otherwise seem unconnected: general and concrete, theoretical and practical, constructive and critical, whilst at the same time
abandoning the traditional quest for truth and objective cognition.
When seen in this light, the different definitions of hermeneutics found
190
CHAPTER 5
in Gadamer’s work come as no surprise. Some examples of these
definitions include: theoretical knowledge of the conditions for the
possibility of all understanding, the continuation of Plato’s theory of
beauty – new universal aesthetic, and a practical art of understanding
and communication.21
Yet hermeneutics is above all knowledge about understanding. In the
course of building hermeneutics, we begin from the Kantian (normative)
question, to which we do not yet provide a Kantian answer. Ultimately,
hermeneutics does not stipulate what understanding should be like, but
merely describes the conditions under which understanding is at all possible. As for understanding itself, it is a phenomenon of a special kind.
Its essence lies in the fact that it is a process – it is something without a
definite beginning or end, it is “the very process of happening” during
which we reiterate our effort to realize the general in the concrete, and
the theoretical in the practical. As is emphasized by Gadamer, “the
hermeneutical problem” always embraces three inextricably linked
moments: understanding (subtilitas intelligendi), explanation (subtilitas
explicandi) and application (subtilitas applicandi). For understanding is
realized through the act of interpretation, and the essence of interpretation
is expressed in its practical application.22
Abandoning the traditional way of presenting the question of truth,
Gadamer poses it in such a way that it may also concern the humanities.
Within the humanities, the question of truth becomes a question about
the conditions for the possibility of understanding. Hermeneutics
enables objective, scientific experience to be joined with individual life
experience. Thus, the division between the objective and subjective elements of our experience of the external world is reduced in significance.
Gadamer – like Heidegger – ultimately assumes that truth is “the disclosure of being” (Unverborgenheit des Seienden), which subsequently turns
into the openness of language statements. In this context, the dialectic
principle of the primacy of question is in force. Thus, truth acquires its
own situational and temporal structure.
The historicity of understanding is elevated to the status of one of two
fundamental hermeneutical principles (the second one is its language
character). According to Gadamer, considerations of the process of
understanding lose their sense – especially as regards the issue of
hermeneutical application – if they are deprived of their historical perspective (horizon). The process of interpretation pursued beyond the historical horizon of understanding would become anew an abstract and
theoretical knowledge of the general principles and rules of interpretation. Our hermeneutical consciousness acts on, develops and is rooted in
HERMENEUTICS
191
history. This historical perspective enables one to be open not only to the
past, to the voice of tradition, but also to the present, to what is “here
and now” as well as, finally, to the future, for every historical act of
understanding contains some projection of a new sense.
The principle of the historicity of understanding is also inter-linked
with other components of hermeneutical experience – namely, the
hermeneutical circle and pre-judgment (Vorurteil). Gadamer examines
both the older, formal–methodological principle of the hermeneutic circle
and its newer, phenomenological–ontological, version. In particular, this
principle captures the connection between the general and the concrete,
between the earlier and the present, and between pre-understanding,
understanding, interpretation and application.
Gadamer emphasizes that the hermeneutical circle is neither subjective
nor objective – it is rather an attempt to describe understanding as a
game (mutual influence) between “the movement of tradition” and “the
movement of the interpreter”. Thus, “the circle of understanding” is not
a methodical circle (a method of cognition), but, rather, a description of
the ontology and structure of the process of understanding. Also of
momentous significance for the historically understood process of understanding, is the concept of pre-judgment (Vorurteil), which refers to
something that exists before (in a temporal sense) both our knowledge
and ignorance.
The second fundamental principle (besides the historicity) of hermeneutical experience (understanding) is its linguistic character. Language is a
sort of medium linking all the elements of the process of understanding in
one whole. Gadamer makes a reference to Schleiermacher, who insisted
that only can be assumed and investigated within hermeneutics is language
(he wrote that the only being that can be understood is language).
However, Gadamer does not assert that the above thesis requires that
hermeneutics be limited to language research, as was suggested by the
older, philologically oriented hermeneutics. For language is “the primary
equipment of man”, with which he comes to the world, and which
expresses his possession of this world. It is thanks to language that we cannot only speak, think and interpret, but also – or, rather, above all – understand. Understanding is, in turn, something more than merely speaking,
thinking and interpreting – it is also a mode of being of man.
Gadamer’s hermeneutics is open – it has no point of departure that
could be determined. Yet at the same time it sets up the claim to be universal, just as the problem of understanding and language is universal.
Moreover, the hermeneutical problem cannot be confined solely to the
methodology of the humanities: “(. . .) Hermeneutics is not simply a
192
CHAPTER 5
science of the method of the humanities, but, rather, an attempt to
determine what the humanities in fact are, the attempt, which is not to be
limited to a reflection over the methodological self-consciousness of the
humanities. (. . .) Hermeneutics should show us the relationship
between the humanities and our whole experience of the world (. . .)”.23
Gadamerian hermeneutics is dialectic (because – in the spirit of
Socrates it gives priority to the question), phenomenological (because it
describes the phenomenon of understanding as the phenomenon of
being), and in addition, it is the philosophy of unity (because it removes
the divisions between the general and the concrete, subject and object,
language and the material world). This philosophy is realized in the
process of communication, in which all historically developed human
communities take part.
Ricoeur. As we have already pointed out, the philosophy of interpretation proposed by Ricoeur is a special “boundary” case, from the standpoint of both ontological and epistemological hermeneutics. According
to Ricoeur, a specific property of hermeneutical interpretation is its
reflectivity. This is because reflection always appeals to symbolic speech,
which, in turn – automatically – necessitates interpretation. Ricoeur’s
hermeneutics appeals to many sources: primarily to phenomenological
philosophy, but also to the philosophy of language, theology, and even
psychoanalysis. Hermeneutics, in Ricoeur’s view, should fulfill three
functions. First, it should be an epistemology of interpretation (metatheoretical function). Hermeneutics must fulfill this function when “a
conflict of interpretations” (for instance, phenomenological and psychoanalytical) emerges. Such a conflict makes it necessary to build “the
hermeneutics of all hermeneutics”, within the framework of which one
can attempt to reconcile otherwise opposed viewpoints. Second,
hermeneutics is simply a theory of the interpretation of symbolic language. Third, and finally, hermeneutics is the practical art of interpreting and understanding this symbolic language, which means that it is its
own application (this kind of hermeneutics is used by Ricoeur, for example, in his work La Symbolique du Mal). Yet in each of these contexts,
hermeneutics is primarily an epistemology, and only secondarily –
through semantic analysis and reflection – an ontology (the ontology of
understanding). Ricoeur, just like his predecessors Heidegger and
Gadamer, asserted that language constitutes the medium of hermeneutical experience. In his programmatic essay L’ existence et l’hermeneutique
he says that all understanding – both ontical and ontological – first finds
expression in speech.24 Language, or, more accurately, symbolic language,
HERMENEUTICS
193
gives rise to the need for interpretation, which, in turn, does not proceed
in a temporal void, but is always “built” into some tradition – it has its
own history.
Symbolic language is universal. Symbol is present even in ordinary
language, which – by its very essence – seems to have little to do with
myth.25 This suggests that hermeneutics is not limited to analyzing particular types of language (for instance, biblical), but concerns itself with
the interpretation of all symbolic structures of meaning. According to
Ricoeur, the structure of meaning can be called symbolic if its direct,
original, and literal sense determines some other – indirect, secondary
and metaphoric – sense which can only be captured through the
medium of the former.26 Symbolic signs are not transparent, they
require interpretation, in other words, they force one to think (ils donnent à penser). Thus, the purpose of the interpretative process ultimately
boils down to understanding a symbol. Symbol and interpretation are
two correlated notions: anywhere there is a multiform sense, i.e. a symbol, there is an interpretation. Ricouer speaks about three stages of
understanding symbol: phenomenological, which consists in understanding the symbol through some other symbol or through all symbols; hermeneutical, the stage at which the proper interpretation of the
symbol takes place (only thanks to interpretation can we again hear and
understand), and existential, the stage at which “thinking enters the
symbol” thereby making possible existential (ontological) interpretation
of the symbol.
The concept of hermeneutical experience ultimately embraces three
inter-linked elements: a text, i.e. language, in which symbolic meaning
structures appear, interpretation and tradition. This “chain”: text –
interpretation – tradition may be read in all directions, because a text
always consists in entering some tradition, and interpretation, in turn,
consists in entering some text. Thus the circle of understanding, interpretation and tradition closes.
5.3.2
Legal Receptions
Reinach. One very interesting attempt to apply the phenomenological
philosophy of Husserl to the needs of jurisprudence was made by
Reinach. Reinach believed that phenomenological analysis lies at the
basis of both statements concerning the ontological essence of the law,
and statements with a methodological character. He set out his ideas in
Zur Phänomenologie des Rechts. Die apriorischen Grundlagen des bürgerlichen Rechts, published in 1986. Reinach conceived of the law as an
a priori category, which we are able to know only thanks to our intuition.
194
CHAPTER 5
Let us recall that according to Husserl, phenomenology is supposed to
enable us to capture what is directly given and evident in the cognitive
process. It is not confined solely to the analysis of notions, but attempts
to reach – “through these notions” – the a priori essence of reality, i.e.
“the things themselves”. Such cognition is possible only by means of
intuition, for it is precisely intuition that expresses the capacity to know
what is directly given and evident. Thus, intuition becomes “the principle of cognitive principles”, the first and irreplaceable source of all cognition, rather than only one of its forms. Husserl enumerated many kinds
of intuition – the number of them is the same as the number of direct
kinds of data. Thus, the following kinds of intuition exist: rational, irrational (used in emotional acts), capturing phenomena in their concreteness, and capturing the essence of a phenomenon.27 The last kind of
intuition, which enables knowledge and understanding of the essence of
law, would presumably be of special significance for Reinach. Positive
law is in a constant state of flux and development. Such contingency, and
the tendency to change make it difficult, if not downright impossible, to
know the a priori essence of law. Thus we must penetrate further and
deeper, through positive laws to the “thing itself ”, to the nature and man
with his needs, desires, will and actions.
The essence of these essential presentations is expressed in a priori sentences, which at the same time are also statements (axioms) of a phenomenologically oriented science of law.28 In this way, besides mathematics and
pure natural science, we are dealing with a case of a “pure, in the phenomenological sense, legal science”. Next, Reinach analyzed the relationship – crucial for an a priori science of law, between the notion of a claim
(Anspruch) and the notion of obligation (Verbindlichkeit). He found the
source of this relationship in the notion of a promise (Versprechen).
A promise creates a particular relationship (connection) between two
persons, by virtue of which one person may require something, and
the other is obliged to fulfill this requirement, or at least to see that it is fulfilled in the future.29 Thus, ultimately, the law has not only an a priori but
also a dialogical nature, because it implies that for every question (claim),
there must be a corresponding answer (obligation) from a second person.
Husserl. Another philosopher of law, G. Husserl, the author of the work
Recht und Zeit also referred to Husserl’s phenomenological philosophy.
He was interested above all in the problem of time in law. He analyzed it
using at least some theses from phenomenology and hermeneutics. He
assumed that every legal system represents a certain phase in the history
of mankind. Thus, legal orders have their own history and they are
HERMENEUTICS
195
themselves history. His distinction between objective time and historical
time roughly corresponds to the distinction made within hermeneutics
between “objective history” and “subjective historicity” (Geschichlichkeit). Thus, a non-linear – phenomenological – conception of time
was supplemented with a phenomenological-hermeneutical analysis of
“the experience of the law”. G. Husserl argued that the essence of legal
cognition is the reduction of legal ideas to an ontological level, on which
“what the law is” can be revealed. This process of reduction is at the
same time a process of actualizing the law itself – applying it in concrete
situations – because the act of reduction uncovers the basic (original)
structure of every possible law – the structure which has an a priori character. In consequence, according to G. Husserl, legal order is the basic
phenomenon of the social world.30
Maihofer. In turn, Maihofer’s ontology of law, outlined in his study
Recht und Sein, has distinctly Heideggerian roots. Maihofer attempted to
transfer “fundamental ontology” to the terrain of philosophical–legal
reflection. The hermeneutics of Dasein (the being capable of self-understanding) is intended to enable Maihofer to construct an existential
ontology of the law. Maihofer emphasized that phenomenology enables
the discovery of a new dimension of being and an order inherent in it –
the order which law is also a part of.31
Gadamer. Gadamer also expressed directly his opinion about legal
hermeneutics. As was mentioned, Gadamer asserted that there is one –
general and universally valid – philosophy of understanding, which
strives to answer questions about conditions for the possibility of understanding in general, and thereby questions about conditions that make
understanding of the law possible. Legal hermeneutics at best might possess an “exemplary meaning” (exemplarische Bedeutung) for other particular hermeneutics, as well as for general hermeneutics. According to
Gadamer, the distance between humanistic hermeneutics and legal
hermeneutics is not as large as it is usually considered to be. Accordingly,
legal hermeneutics is in fact not a special case, but it does make the scope
of problems to be tackled within historical hermeneutics as broad as it
was in the past; in consequence, one witnesses a return to an old-time
unity of the hermeneutical problem – one may say that the lawyer and
the theologian meet anew the philologist. If the lawyer – acting in his
capacity as judge – endeavors to interpret the text of a statute (to reconstruct the original sense of the text and enable its application), then he
acts exactly as he would in the course of any other understanding.
196
CHAPTER 5
Understanding is most strictly connected with interpretation and application. More accurately, interpretation and application are integral parts
of the hermeneutical experience, i.e. of the process of understanding.
The point to be stressed is that application is inherent in (pervades) all
forms of understanding – it is not to be construed as a later (secondary)
stage at which one applies something general, that has been earlier
understood, to a concrete case. Thus, application is the actual understanding of the general, which the interpreted text provides for us.32
Kaufmann. Gadamer’s work in the philosophy of law was continued
above all by Kauffman, who provides an ontologically oriented conception of legal hermeneutics. The key problems in his considerations are
the understanding, interpretation and application of law, and the relationship between law and statutes.
According to Kauffman, the law emerges (is constituted) during the
hermeneutical act of understanding.33 Thus, the law does not exist before
interpretation – it has to be found, created. The interpretation of law, the
process of making legal decisions, and even the whole administration of
justice constitute a fully creative process. Underlying this process is the
distinction between the law and a statute. A statute has its source in a legislator’s authority. As for the law, it really exists – it stems directly from
the being; it has its source in the natural order of things, its existence is
therefore primary – independent of any authority. A statute is only one
aspect of law’s realization. The relationship between the law and a statute
is just like the relationship between an act and potential, or reality and
possibility. Ultimately, both a statute itself, and abstract ideas of law such
as the concept of just law are, in Kaufmann’s view, only possible forms of
law. The essence of the law is located in the fact that it is unavailable
(unverfügbar), concrete and historical. Its positive character constitutes its
existence, and rightness – its essence. The law expresses “the primary
analogy” – correspondence – between “is” and “ought”. Legal cognition
should take account of this analogy.
Kaufmann reckons that only through hermeneutics – the ontology of
understanding – will it be possible to overcome the one-sidedness which
has encumbered both conceptions of the law of nature and positivistic
conceptions. For this reason he rejects the ontology of substance
assumed in classical theories of the law of nature. Such ontology would
create the danger of, inter alia, objectivity which is undesirable from the
perspective of the hermeneutic approach. Objectivity has been replaced
with “the historicity of the understanding of law”. Law emerges through
the act of understanding, it comes into existence, or happens at a specific
HERMENEUTICS
197
temporal (historical) moment; it is not a state but rather an act. To put it
differently, the law is a relationship between a norm (which is usually
general) and a concrete case, a relationship which is finally manifested in
the person. In this way the ontology of substance is replaced with the
existential ontology of relationships.34
The law is therefore the result of a process Kaufmann calls “legal realization” (Rechtsverwirklichung). “Concrete law” is the “product” of the
hermeneutical process of a sense’s development and realization
(Sinnentfaltung – Sinnwerwirklichung). The essence of interpretation is
ultimately the act of understanding – the capacity to interpret is synonymous with the capacity to understand. Three degrees (stages) can be distinguished in the process of legal realization (Rechtsverwirklichung). The
starting-point is abstract – extra-positive and extra-historical – legal principles (ideas). Then we pass through the general – formal-positive –
norms contained in a statute (the second stage) to the concrete – materialpositive – historical law (the third stage). This concrete historical law is
established in the process called “finding the law” (Rechtsfindung), which
consists in “establishing coherence” and “seeking correspondence”
between actual states of affairs (Lebenssachverhalte) and norms or, to put
it differently, “bringing them closer to each other”.35 The establishment of
a legal decision – the act of finding a legal solution – is achieved through
a historical act of understanding which appeals directly to an original
analogy contained in the concept of the law. For the act of understanding brings together subject and object, duty and being, norms and an
actual state of affairs.36 The syllogistic model of the interpretation and
application of law has been decidedly rejected by Kaufmann. For the
understanding of a legal text designates a certain ambivalent – creative –
productive process. The hermeneutical understanding of a text is not
something receptive; it is rather a practically shaped act in the course of
which concrete historical law emerges.
According to Kaufmann, legal hermeneutics is a special example of
the philosophy of understanding, in which there is – besides ontological –
a methodological and practical moment. Proceeding, hermeneutics
implies, is practical in character. Thanks to hermeneutics it is possible to
transform jurisprudence and the philosophy of law into a theory of
action (Handlungswissenschaft). The ontology of understanding has
been transformed by Kaufmann into the ontology of relationships,
which, in turn, has enabled him to build a personalized philosophy of
law. The concept of law thus leads us to the concept of analogy and this
in turn directs us to the concept of relationships, and once again to the
concept of the person. The person is not a substance, but a relation, or,
198
CHAPTER 5
more accurately, the structural unity of what may be defined as “relatio”
and “relata”. The idea of law is the idea of the personally understood
man, neither more nor less. In recognizing (understanding) the essence of
the human person, we recognize (understand) the very essence of law.37
The concept of the person is inter-linked in Kaufmann’s theory with
other components of hermeneutical experience such as, for example, preunderstanding (Vorverständnis) and the hermeneutical circle. These elements of hermeneutical cognition are justified by the ontological
assumption that they are grounded in the structure of individual existence – the person.
Numerous references to the conceptions of Gadamer and Kaufmann
can be found in contemporary jurisprudence – both within legal dogmatics and the philosophy of law. An attempt to apply hermeneutics – in
the form proposed, among others, by Kaufmann – to the needs of penal
law was made by Hassemer in his work Tatbestand und Typus.
Untersuchungen zur straftrechtlichen Hermeneutik. Hinderling, in turn
examines the possibility of using Gadamerian hermeneutics for the purposes of constitutional law; the results of his research are presented in
Rechtsnorm und Verstehen. Die methodischen Folgen einer allgemeinen
Hermeneutik für die Prinzipien der Verfasunsauslegung. Finally, the
hermeneutical account of the philosophy of law can be found in
Stelmach’s work Die hermeneutische Auffassung der Rechtsphilosophie.38
Representatives of the classical current Methodenlehre, especially Larenz
and Esser (discussed earlier), also made reference to Gadamerian
hermeneutics. Occasional remarks on phenomenologically understood
legal hermeneutics were made by such authors as Kriele, Baeyer, Leicht,
Rottleuthner, Fikentscher, Haba, Fuhrmann, Ellscheid, Neumann,
Schroth, Haft, Philipps, Schild, Scholler, Müller-Dietz, Hegebarth,
Brito, Calera, Ollero, Saavedra, Zaccaria and Alwart.39
5.4
THE UNDERSTANDING OF THE LAW
What is not known about hermeneutics has already been outlined in
Section 5.1.2. Now the time has come to attempt to answer at least some
of the questions we posed there. Two problems seem particularly interesting in the context of considering the frontiers of the application of
hermeneutics in legal argument. Those are the claim to universality made
by the philosophy of understanding and the nature of hermeneutic cognition (the core of the latter problem is whether hermeneutic cognition is
direct or indirect).
HERMENEUTICS
5.4.1
199
Claim to Universality
Starting from Schleiermacher, all modern hermeneutics aspired to be in
some sense universal – at least as a method of the interpretation and
understanding of texts, as the methodology of the humanities or – finally –
as the ontology of understanding. This claim of hermeneutics (also legal
hermeneutics) can also be justified by the kind of problems the philosophy deals with. Hermeneutics is universal, because the problem of language (the philological aspect of each interpretation, history and
“things” which are objects of hermeneutic cognition such as, for
instance, life, spirit, culture or being understood as individual existence)
is universal. Hermeneutics is also universal because hermeneutic experience is of a specific nature: within this experience, in transcending
boundaries between general and concrete, theoretical and practical, and
between understanding, interpretation and application, one achieves
cognitive unity. Finally, hermeneutics (at least its phenomenologically
oriented variant) is universal, because it is “the first science” – the startingpoint and essential element of every possible cognitive process no matter
what the object of that process.
The claim to universality was formulated more narrowly by hermeneutics understood as a kind of humanistic epistemology. The anti-naturalistic
thesis which holds that two otherwise objective methodologies –
the methodology of explanation (used by the natural sciences) and the
methodology of understanding (used exclusively by the humanities) –
entails that the universality of hermeneutics is to be limited to the field
of humanistic cognition. Those lawyers who were on a quest to find their
methodological identity often made use of anti-naturalistic hermeneutics. Were they right in doing so? In our view, they were not. The division
of methodologies into the methodology of explanation and the methodology of understanding is entirely groundless, which was clearly realized
by representatives of the variant of hermeneutics built on phenomenological bases. Not only do we lack one acceptable definition of understanding, but there are also numerous controversies concerning the
process of explanation. What’s more, in humanistic, as in every other
type of cognition, appeal is made to both explanation and understanding. It is also difficult to defend theses that there exist specific “humanistic objects” (like, for instance, the law). Even if their ontological
character could justify identifying them as specific, this would not be sufficient reason to use only one type of method to examine them.
The choice of a method will depend on the complexity and the nature of
an interpretative case, the habits of an interpreter, and the interpretative
200
CHAPTER 5
tradition, rather than on – ultimately entirely arbitrary – methodological
decisions. Thus, in the same case, one interpreter will appeal to logic,
another to analysis or argumentation, and still another to hermeneutics.
Consequently, in view of the above reservations, it must be conceded that
attempts to prove the separateness and methodological autonomy of the
humanities (including jurisprudence) on the grounds of anti-naturalistic
hermeneutics is unfeasible.
The concept of universalism was understood differently, and more
broadly, by hermeneutics that was understood as a kind of ontology.
Hermeneutics is universal because it tackles the fundamental problem of
the understanding of individual being (Dasein) and appeals to – intuitive –
methods that enable one to know the very essence of this being.
Ultimately, hermeneutics thus understood may constitute the startingpoint of every cognitive process, without excluding the possibility of
applying other methods (say, logical and analytical) in further stages of
the process. Thus, were we to have the universally valid hermeneutics,
which is both the ontology and epistemology of understanding, we
should be glad and consider the earlier problems of hermeneutics as
resolved. Yet even then, the following dilemma would arise: either the
process of hermeneutic cognition is regarded as purely intuitive, which
gives rise to questions about the sources of this cognition and the criteria for its verification, or the process of hermeneutic cognition is
regarded as a kind of indirect cognition, i.e. one which appeals to numerous prior theses and assumptions which may fail to be unquestionable. In
the first case, one is threatened either by relativism (because one cannot
know for sure that conclusions reached in the pure act of phenomenological cognition are ultimate and indefeasible) or – at best – by psychologism (because one will have to justify phenomenological conclusions by
appealing to some type of introspective psychology). In the second case,
one is bound to get entangled in virtually unsolvable controversies concerning the conditions of hermeneutic cognition (which are: the linguistic character of understanding, historicity, pre-understanding and the
hermeneutical circle).
5.4.2
The Nature of Hermeneutic Cognition
The issue of the nature of hermeneutic cognition – its interpretation and
essence – probably gives rise to the highest level of controversy. In all
modern hermeneutical conceptions, understanding was conceived as a
kind of fundamental and primitive cognitive process, conditioning
“deeper” insights into the nature of examined objects. Furthermore, phenomenological hermeneutics treats understanding not only as a cognitive
HERMENEUTICS
201
capacity (competence), but also as a property of the individual being –
human existence (Dasein). However, the issue becomes complicated
when attempts are made to clearly establish the nature of this cognition –
especially whether it is direct or indirect.
Understanding as direct cognition. The most elementary associations
lead us to regard understanding as a kind of primitive capacity which
makes no appeal to prior knowledge (experience), theses or assumptions.
Hermeneutics was understood in this way by Socrates, who thereby
built his conception of philosophy as “deprived of the Archimedean
starting-point”; St. Augustine, who wrote about inspired understanding;
Schleiermacher, who distinguished the clairvoyant type of understanding; and by the representatives of phenomenology, for whom
understanding was the primitive capacity, conditioning the knowledge
of “things themselves”.
According to the representatives of phenomenology, understanding is
a kind of intuitive cognition which enables one to capture a phenomenon
in its concreteness, as well as its a priori essence. Let us recall that Husserl
distinguished many kinds of intuition corresponding to kinds of direct
data: rational, irrational (used in emotional acts) and, of course, phenomenological (capturing the concreteness, as well as the essence, of a
phenomenon). The representatives of phenomenological hermeneutics,
however, were reluctant to pronounce on the nature of the process of
understanding. Even the most careful reading of Gadamer’s work
Wahrheit und Methode does not allow unequivocal theses about the
essence of understanding to be constructed. In particular, it is not clear
whether understanding is a form of direct cognition. A positive answer
to this question would give rise to the subsequent questions: which faculties make this kind of cognition possible? what kinds of intuition ultimately compose the capacity called understanding? Phenomenological
epistemology may be accepted or not, yet its deeper meaning and significance cannot be denied. Phenomenological hermeneutics engenders
many more problems: its representatives decline – almost in a programmed response – to give answers to most questions concerning the
nature of the process of understanding; they point out that they are
moved by the Socratic reluctance to construct philosophies built on trust
in previously accepted theoretical beliefs.
In our opinion, understanding is a type of cognition which may possess an intuitive character. In the process of understanding, one uses
both intuition which may be described as rational, and “pure” phenomenological intuition. It is the latter type of intuition that opens the way
202
CHAPTER 5
to direct cognition. It is commonly used by lawyers, even when they
know nothing about phenomenology and hermeneutics. This kind of
intuition usually constitutes a starting-point for the process of interpretation; without this intuition it would be impossible to recognize the
a priori essence of the phenomenon of law and the fundamental relationships (e.g., claim – obligation) which exist in law. Another argument
supporting the thesis that understanding is a direct cognition is found in
the view – strongly emphasized by Gadamer – of the unity (simultaneity)
of processes of understanding, interpretation and application. This view
can be reasonably defended only if the very essence of a thing (phenomenon) being examined is reached.
Thus, understanding is a type of direct – intuitive – cognition. But is it
anything more than that? Phenomenology is a philosophy of consciousness, that is, a variety of the philosophy of cogito, i.e. rationalism. Rational
intuition appeals to general notions and earlier accepted theorems and
definitions. At this stage of hermeneutical cognition “language enters
understanding” and “hermeneutical logic” is substituted for primitive
hermeneutical intuition.
Understanding as indirect cognition. The overwhelming majority of
early hermeneutical theories (including those proposed in the nineteenth
century) conceived of hermeneutics as an indirect method of cognition.
Hermeneutics was simply the art of interpreting and understanding
texts. This understanding and interpretation of texts is made possible by
universally valid rules of interpretation. In some cases, it is not only
through the medium of rules of interpretation, but also through psychological facts that we are able to reach understanding and an interpretation. This dual conception of the process of understanding (as direct and
indirect cognition) was not suppressed by phenomenological hermeneutics. It was Gadamer who devoted particularly great attention to such
properties of the process of interpretation and, consequently, hermeneutical cognition, as its linguistic character and historicity. The issues of
pre-understanding and the hermeneutical circle will also reappear continually. The point to be stressed is that all these properties confirm in
some way the theory that hermeneutical cognition is indirect in nature.
The linguistic character of understanding. Phenomenologically oriented
hermeneutics attached particular importance to the theory of the
linguistic character of the process of understanding Let us recall that
both Schleiermacher and Dilthey highlighted the primitive character
of the philological aspect of all hermeneutics, and of every process of
HERMENEUTICS
203
understanding. According to Gadamer, the only being that can be understood is language. Knowledge of the world is possible only through the
medium of language; language, in addition, determines the horizons of
hermeneutical ontology. In Gadamer’s view, language is not only the
means through which we experience the world and the equipment with
which we enter this world, but also an expression of our possession of
the world. The boundary between language and the material world,
which is distinct within other philosophies, loses its sharpness within
phenomenological hermeneutics. Without going into the details of this
relationship, we shall confine ourselves to observing that everything
given to us in the process of understanding (in hermeneutical experience)
is given through the medium of language. This thesis is accepted both by
“old” and “new” hermeneutics. When seen from this philological (in others’ view – analytical) perspective, hermeneutical cognition turns out to
be fully discursive and thereby indirect.
A similar situation holds in relation to legal hermeneutics (it is worth
noting that legal hermeneutics rarely goes beyond the traditional
hermeneutics of texts – Kaufmann’s philosophy of legal understanding
is an exception). The understanding, interpretation and application of
law always concern some entities of language (deontic sentences, rules or
norms). At the level of interpretation, “what law is” is no more than a
certain linguistic expression. A starting-point of the lawyer’s work is in
principle linguistic interpretation, though it is a matter of dispute
whether this interpretation is made in a hermeneutical or analytical way
(assuming that the division of these two approaches is at all practically
viable). In either case, though, one has to appeal to a language, principles
(semantic, syntactic and pragmatic) of that language, and different levels
of rules of interpretation, which either already exist and are universally
accepted and applied in similar cases, or which have to be formulated for
the needs of an interpretative case. Thus, the assumption that the understanding of law is realized through the medium of language is equivalent
to the assumption that hermeneutic cognition is indirect.
The historicity of understanding. The process of understanding has not
only a linguistic but also a historical character. Gadamer stresses that
hermeneutical experience, if examined beyond its historical context,
would be a mere philosophical abstraction, and the very philosophy of
understanding would be nothing more than a continuation of an earlier
metaphysical philosophy which was questioned by phenomenology. The
theory of the historicity of understanding is another argument in
support of the view that hermeneutical cognition may be – and most
204
CHAPTER 5
frequently really is – indirect. This is because, in this process, appeal is
made to numerous historical assumptions, prior knowledge, and tradition, which should be applied (i.e. concretized and actualized) to a concrete interpretative case. According to Ricoeur, interpretation and
tradition are two sides of historicity, and the chain: tradition – text –
interpretation can be read in all possible directions.
The problem of historicity is of particular significance in the context
of legal interpretation. It is connected with the important issue of the
role of tradition in legal interpretation, and with the issue of “legal
application” which is an integral component of hermeneutical experience, along with understanding and interpretation. Each process of
understanding and application becomes an element of interpretative tradition. It is hard to imagine legal interpretation entirely beyond its historical context – beyond the tradition (shaped in the case of our
continental system of law by two millennia), whose elements are constantly present in each interpreter’s consciousness. An interpreter enters
“an interpretative situation” with previously acquired legal knowledge,
intuitions, prior conceptions of basic legal institutions, her whole legal
pre-understanding.
In the process of understanding and interpretation, an interpreter
must apply this general historical knowledge to a concrete case, that is,
she must concretize and actualize it. Ultimately, the essence of legal
thinking (the understanding of law) will always be the process of concretizing, i.e. the application of an interpreted (understood) general text
(general norm) to a concrete case. However, hermeneutical concretization cannot be reduced – at least in Gadamer’s view – to a deductive
operation. Legal hermeneutics emphatically abandons “the syllogistic
model of legal application”, and Gadamer himself stresses that understanding is concretization, which yet is connected with preserving
hermeneutical distance, whatever that means.40
Besides concretization, the second condition of the understanding and
application of law is actualization. It is true that appeals to the past (tradition) are made in the process of actualization, yet such appeals are
intended to modify and change it, i.e. to adapt it to a concrete case,
which “happens in the present”. Legal hermeneutics supports dynamic
accounts of the process of legal interpretation and application, and
rejects psychological – subjective – accounts. Legal interpretation should
result in “adapting the law to the requirements of life”. The velocity of
changes taking place in social and economic reality makes the operation
of actualization necessary. The interpretation of law which is to assist in
reconstructing – in the name of legal constancy and safety – the will of
HERMENEUTICS
205
a historical legislator is not acceptable from the standpoint of legal
hermeneutics.
Ultimately, the fact that understanding has the property of historicity
entitles us to assert that in many, or most, cases hermeneutic cognition
will fail to have a direct – purely intuitive – character.
Pre-understanding and the hermeneutical circle. Further justification for
the view that hermeneutic cognition is indirect appears to flow from considerations devoted to pre-understanding and the hermeneutic circle. It
is important not to formulate fixed opinions as far as the problem of preunderstanding is concerned: on the one hand, one may appeal to this
concept in order to prove that hermeneutical cognition is direct, provided that one defines pre-understanding as an intuitive capacity to
know the very essence of things – capacity making no use of any prior
knowledge; on the other hand, one may appeal to it in order to prove
that hermeneutic cognition is indirect, provided that pre-understanding
is understood as a historically conditioned, prior knowledge.
One should also refrain from formulating firm opinions about to the
hermeneutic circle. If the methodological version of the principle of the
hermeneutical circle is assumed, one may assert that this principle confirms unequivocally the thesis that hermeneutic cognition is direct. This
principle is an interpretative principle – an element of a broader and
informally understood humanistic and legal logic. The situation will
look different if one assumes the phenomenological version of this principle: in this version, the hermeneutic circle is a description of an ontological and structural moment of understanding, rather than a
methodological principle capturing a cognitive moment; in Heidegger’s
terminology, it is expressive of the pre-structure of Dasein. In this
account, the hermeneutical circle, like pre-understanding, is at best one
of those properties of the process of understanding which decide about
the direct character of hermeneutic cognition.
5.4.3
Applications
One reason why legal hermeneutics may preserve its claim to universality is that it may constitute a starting-point – or at least an element – of
each cognitive process connected with the law, irrespective of the level
of generality of theses formulated during this process. By means of
hermeneutics one may perform practical tasks, i.e. make legal interpretations, make legal decisions and justify those decisions. Hermeneutics
can also be applied in legal dogmatics (a theory of dogmatics), making it
possible to formulate and justify theses which may be reasonably
206
CHAPTER 5
defended only when “the hermeneutic approach” is accepted. Finally,
hermeneutics enables one to build a certain type of legal philosophy or
theory. The possible presence of hermeneutics on all levels of a lawyer’s
cognitive activity (practical, dogmatic and theoretical) seems ultimately
to confirm the claim to universality raised by this philosophy of interpretation.
Let us start from the level of practical applications. We shall not
appeal to concrete cases for that would require value judgments to be
made regarding which cases are more, and which less typical of
hermeneutics. Besides, the facts of a concrete case and an interpreter’s
attitude, rather than some prior “hermeneutical intention” often determine “the hermeneutical character of interpretation”. Legal hermeneutics, like other humanistic hermeneutics, has always been presented as a
theory (art) of textual interpretation and understanding. Until the eighteenth century it was hardly possible to separate hermeneutical “theories” from other theories. Only in the nineteenth and twentieth centuries
did legal hermeneutics become a more specific philosophy of interpretation. Within this philosophy, the problem of understanding is emphasized, and understanding (at least within phenomenologically oriented
hermeneutics) is identified or treated as synonymous with interpretation
and application. Legal thinking also becomes enriched by “new”
hermeneutic concepts, such as the hermeneutic circle, concretization and
actualization. Hermeneutics becomes one of the basic points of reference for all the problems that appear in contemporary theories of interpretation. Linguistic interpretation can be performed either in an
exclusively analytical (linguistic), or in an analytical–hermeneutical way,
and in our view, it is difficult to conceive of any third possibility. Whilst
making systematic or functional interpretations, in turn, one uses, albeit
often unconsciously, hermeneutical methods.
Thus, the principle of the hermeneutical circle, upon which systematic
interpretation is in fact based, is used and appeal is made to the
hermeneutical account of the processes of concretization and actualization. The only reasonable alternative for an interpreter is resort to the
analytical (analytical-positivistic) approach. The point to be stressed,
though, is that the difference between this approach and the hermeneutical one is not as substantial as it may at first seem to be. Legal
hermeneutics too has many interesting things to say about the issue of
application. Application is regarded, first, as synonymous with interpretation and understanding, and, second, as reducible to the following two
operations – concretization and actualization. In this context, hermeneutics provides truly original and at the same time really interesting insights:
HERMENEUTICS
207
hermeneutical concretization has nothing in common with the positivistic conception of syllogism, neither does actualization reflect the dynamic
(objective) theories of interpretation advanced, for instance, by adherents
of legal realism. These facts underlie the originality of the hermeneutical
account of the legal decision making process (Rechtsfindung, Rechtsgewinnung), law’s application (Rechtsanwendung, Rechtsapplikation), and
realizing the law (Rechtsverwirklichung). Finally, hermeneutics enables an
interpreter to make an interpretation on an “existential level”. A need for
this kind of interpretation arises in hard cases, where standard methods do not suffice to make an acceptable legal decision; in such cases,
an interpreter can only appeal to ontological analysis, i.e. reach to the
very essence of the process of understanding – to individual existence
(Dasein).
Hermeneutical interpretation can also be used in the course of formulating and justifying the tenets of the theory of legal dogmatics. As mentioned earlier, such attempts were made many times in the contemporary
science of law within different legal disciplines. The phenomenological
analysis of such concepts as, for instance, “claim”, “obligation”, and
“promise” made by Reinach can be used in particular in a theory of civil
law. Hermeneutics, as proposed by Esser and Larenz, can be successfully
applied in different dogmatic disciplines, especially in civil and penal law.
Hinderling examined the possibility of applying general – Gadamer’s –
hermeneutics in the interpretation of constitutional law, and Hassamer
suggested building hermeneutics which might be applied above all in
penal law.41
Finally, hermeneutics was also used – on a theoretical level – as a
method of philosophy and legal theory. In some cases, it was a method,
in others a kind of cognitive attitude, whilst in still others it was an
entirely autonomous type of legal philosophy. Hermeneutics was
described by Betti as a method of the humanities and jurisprudence, and
many representatives of Methodenlehre (e.g., Engisch, Larenz, Esser)
treated it as a cognitive attitude. It was conceived of as a type of legal
philosophy by Reinach (the author of the purely phenomenological philosophy of law), Gadamer (who also tackled the problem of legal
hermeneutics) and – above all – by Kaufmann.
In relation to the hermeneutical philosophy of law, there arose many
misunderstandings and absurd opinions stemming both from the lack of
clarity in hermeneutics itself, and from the lack of competence of those
who attacked it thoughtlessly and aggressively. Hermeneutics introduced
into legal thinking and legal philosophy many new elements (deserving
of more careful examination than has thus far been carried out by
208
CHAPTER 5
various authors) and new questions which most probably cannot be
given definitive answers. Thus the views and conceptions of such authors
as Reinach, Esser, Larenz, Kaufman, cannot be dismissed as superficial
in contemporary philosophies of law. Reinach’s excellent phenomenological analysis of the concept of law and other fundamental concepts of
the science of law, and Kaufmann’s hermeneutical account of law as an
ontologically complex object may confirm this thesis.
In conclusion, whether hermeneutics is liked or otherwise, it has
become one of the most important, most broadly discussed and, at the
same time, most controversial methods of contemporary humanities and
jurisprudence. Hence the reason this specific philosophy of interpretation could not be discounted in the present discussion of jurisprudential
methods.
NOTES
1. See E. Gellner, Words and Things. An Examination of, and an Attack on, Linguistic
Philosophy, London, 1979, p. 17.
2. J. Stelmach, Die hermeneutische . . ., op. cit., p. 50 ff.
3. H. Coing, Die juristischen Auslegungsmethoden und die Lehre der allgemeinen
Hermeneutik, Köln-Oppladen, 1959, p. 8.
4. J. Stelmach, “Die inntuitive Grundlagen der Jurisprudenz”, in H. Bauer, D. Czybulka,
W. Kahl, A. Vosskuhle (editors), Umwelt, Wirtschaft, und Recht, Tübingen, 2002,
p. 161 ff.
5. P. Ricoeur, Egzystencja i hermeneutyka. Rozprawy o metodzie (Existence and
Hermeneutics. Discourses on Method), Warszawa, 1985, p. 195 ff.
6. W. Dilthey, Pisma estetyczne (Writings on Aesthetics), Warszawa, 1982, p. 304.
7. F. E. D. Schleiermacher, Hermeneutik aund Kritik, Frankfurt am Main, 1977, p. 75.
8. The representatives of phenomenological hermeneutics criticized this view, treating
it as a manifestation of unjustifiable psychology. See P. Ricoeur, Existence . . ., op.
cit., p. 314.
9. F. E. D. Schleiermacher, Hermeneutik . . ., op. cit., p. 187.
10. W. Dilthey, Writings . . ., op. cit., p. 316.
11. H. Lipps, Untersuchungen zu einer hermeneutischen Logik, 4th ed., Frankfurt am
Main, 1976, p. 20 ff, p. 115 ff; also O. F. Bollnow, Studien zur Hermeneutik, vol. II,
Zur hermeneutischen Logik von Misch und Lipps, München, 1982.
12. J. Stelmach, Die hermeneutische . . ., op. cit., pp. 44–45.
13. H. Coing, Die juristischen Auslegungsmethoden . . ., op. cit., p. 22 ff.
14. K. Larenz, Methodenlehre der Rechtswissenschaft, 5th ed., Berlin/Heidelberg/New
York/Tokyo 1983, p. 185 ff.
15. J. Esser, Vorverständnis und Methodenwahl in der Rechtsfindung. Rationalitätsgrundlagen richterlicher Entscheidungspraxis, Frankfurt am Main 1970, 93 p. ff, p. 136 ff.
16. E. Forsthoff, Recht und Sprache. Prolegomena zu einer richterlichen Hermeneutik, 2nd
ed., Darmstadt, 1964., K. Engisch, Logische Studien zur Gesetzesanwendung, 3rd ed.,
Heidelberg, 1963, K. Engisch, Einführung in das juristische Denken, 8th ed.,
HERMENEUTICS
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
209
Stuttgart, 1983, F. Müller, Juristische Methodik, 2nd ed., Berlin, 1976, M. Kriele,
Theorie der Rechtsgewinnung, 2nd ed., Berlin, 1976.
M. Heidegger, Sein und Zeit, 9th ed., Tübingen, 1967, p. 39.
M. Heidegger, Über den Humanismus, Frankfurt am Main 1949, p. 5.
E. Coreth, Grundfragen der Hermeneutik, Freiburg–Bael–Wien, 1969, p. 99.
H. G. Gadamer, Wahrheit und Methode. Grundzüge einer philosopischen Hermeneutik,
in Gesammelte Werke, vol. I, Tübingen, 1986, p. 312 ff.
H. G. Gadamer, Wahrheit und Methode. Ergänzungen. Register, in Gesammelte Werke,
vol. II, Tübingen, 1986, p. 3 ff.
H. G. Gadamer, Wahrheit und Methode. Grundzüge . . ., op. cit., p. 312 ff.
H. G. Gadamer, Wahrheit und Methode. Grundzüge . . ., op. cit., p. XXIX.
P. Ricoeur, Existence. . ., op. cit. p. 190.
P. Ricoeur, Symbolika zl~a (Symbolism of Evil), Warszawa, 1986, p. 12.
P. Ricoeur, Existence . . ., op. cit., p. 19 ff.
J. Stelmach, “Intuicja prawnicza” [Legal Intuition], in Dziedzictwo prawne XX wieku,
Kraków, 2001, pp. 35–36.
A. Reinach, Zur Phänomenologie des Rechts. Die apriorischen Grundlagen des
bürgelichen Rechs, 2nd ed., München, 1953, pp. 11–12.
Reinach, Zur Phänomenologie des Rechts . . ., op. cit., p. 21.
G. Husserl, Recht und Zeit. Fün rechtsphilosophische Essays, Frankfurt am Main,
1955, pp. 7, 14, 23, 42, 76.
W. Maihofer, Recht und Sein. Prolegomena zu einer Rechtsontologie, Frankfurt am
Main, 1954, p. 1 ff.
H. G. Gadamer, Wahrheit und Methode. Grundzüge . . ., op. cit., vol. I, pp. 334 and 346.
A. Kaufmann, “Problemgeschichte der Rechtsphilosophie”, in A. Kaufmann/
W. Hassemer (editors), Einführung in Rechtsphilosophie und Rechtstheorie der
Gegenwart, 4th ed., Heidelberg, 1985, p. 122.
A. Kaufmann, “Vorüberlegungen zu einer juristischen Logik und Ontologie der
Relationen. Grundlegung einer personalen Rechtstheorie”, in Rechtstheorie 17
(1986), 258 ff; A. Kaufmann, “Die Geschichtlichkeit des Rechts unter rechtstheoretisch-methodologischem Aspekt”, in ARSP Beiheft II (1988), p. 114 ff.
A. Kaufmann, “Analogie und ‘Natur der Sache’, Zugleich ein Beitrag zur Lehre vom
Typus”, in A. Kaufmann, Rechtsphilosophie im Wandel, 2nd ed., Köln/Berlin/Bonn/
München, 1984, p. 282 ff.
Kaufmann, “Über den Zirkelschluß in der Rechtsfindung”, in A. Kaufmann,
Beiträge zur juristischen Hermeneutik, Köln/Berlin/Bonn/München, 1984, p. 75 ff.
Kaufmann, “Über die Wissenschaftlichkeit der Rechtswissenschaft”, in ARSP
LXXII/ IV (1986), p. 442.
W. Hassemer, Tatbestand und Typus. Untersuchungen zur strafrechtlichen Hermeneutik,
Köln/Berlin/Bonn/München, 1968, H. G. Hinderling, Rechtsnorm und Verstehen. Die
methodischen Folgen einer allgemeinen Hermeneutik für Prinzipien der Verfassungsauslegung, Bern, 1971, J. Stemach, Die hermeneutische . . ., op. cit., Ebelsbach, 1991.
J. Stelmach, Die hermeneutische . . ., op. cit., p. 78 ff.
H. G. Gadamer, Wahrheit und Methode. Grundzüge . . ., op. cit., vol. I, p. 335 ff.
See A. Reinach, Zur Phänomenologie des Rechts . . ., op. cit.; J. Esser, Vorverständnis
und Methodenwahl in der Rechtsfindung . . ., op. cit.; K. Larenz, Methdenlehre der
Rechtswissenschaft . . ., op. cit.; H.G. Hinderling, Rechtsnorm und Verstehen . . ., op.
cit.; W. Hassemer, Tatbestand und Typus . . .., op. cit.
CHAPTER 6
METHODS OF LEGAL REASONING FROM A
POSTMODERN PERSPECTIVE
6.1
A SUMMARY
At the end of the book we would like to make an attempt – hopefully a
justified one – to summarize and review our analyses. First, we would
like to look at our considerations from a broader perspective which
includes those stances that have not been given due attention above.
Second, we would like to provide additional justification for the choices
we made and guard against some obvious objections. Finally, we would
like to offer some conclusions concerning the methodology of law that
follow, we believe, from our considerations.
Arbitrariness
Let us begin with the objection that stresses the arbitrariness of the choice
of methods we present. We have already dealt with this issue in Chapter 1.
Here, we would like to restate those observations adopting a considerably
broader perspective. Observe that the choice of methods presented –
although arbitrary to certain extent – is not accidental. If we were to
compose a list of philosophical conceptions of thinking (not necessarily
legal, but thinking in general), it would most certainly include logic,
analysis, argumentation and hermeneutics. Moreover, it is difficult to
think of any other philosophical approach that would make an obvious
addition to the list. Anyway, any such addition would be objectionable.
Metatheoretical Perspective
This general remark is closely tied up with two further problems. First,
the four described methods should not be considered as “technologies of
interpretation”. We have rather presented the possible bases on which a
coherent theory of law and a method of interpretation can be constructed. Our claim is not the simple one that lawyers or theoreticians of
law can use logic, analysis, argumentation or hermeneutics. This thesis is
true only in the sense that the four stances are “platforms” on which a
coherent and applicable method or technique of interpretation can be
built. Second, we claim that all such techniques are ultimately reducible
to one of the four perspectives or to a combination thereof.
211
212
CHAPTER 6
The acceptance of such assumptions leads to the result that our analyses are not confined to the philosophy and theory of law. We believe that
the presented methods determine the boundaries of any possible theory
of humanistic interpretation. Naturally, the examples we offered as well
as the problems we addressed concerned mostly legal discourse. We tried
to show, however, that the legal applications of logic, analysis, argumentation and hermeneutics are based on more general conceptions, which
in turn are applicable to all humanistic disciplines. It is noteworthy that this
can be shown in connection to considerations concerning legal methods. It
stems from the fact, we believe, that among the humanistic methodologies, methodology of law is particularly well developed. Heated debates
concerning the very existence of a method in law have been raging at
least since the nineteenth century. In such circumstances “the methodological consciousness” of lawyers and legal theoreticians is particularly
sensitive.
Relationships Between the Methods: The Three Theses
The next important issue to consider is the problem of the relationships
between logic, analysis, argumentation and hermeneutics. We have
addressed it in detail in Chapter 1. Here, a more systematic account will be
offered. The issue can be resolved with one of the following three theses.
Let us begin with the weakest thesis. One can maintain that logic,
analysis, argumentation and hermeneutics “have something in common”
because they aim to account for the same phenomenon, i.e., the phenomenon of thinking. This is a very simple solution, and does not
explain much. It shows, however, why some operations of intellect may
be treated as manifestations of the application of two different methods,
e.g., argumentation and hermeneutics. Such an account results in serious
difficulties. For instance, as we have tried to show in Chapter 2, logic cannot provide a basis for constructing a complete theory of legal reasoning, as it is concerned only with one aspect thereof, i.e., the formal
aspect. Moreover, a conception that explains the interconnections
between logic, analysis, argumentation and hermeneutics by the fact that
all four concern the same phenomenon, does not contribute to the
understanding of what those interconnections consist in.
The second thesis reads: logic, analysis, argumentation and hermeneutics are complementary theories of legal reasoning. The complementariness thesis can be interpreted in two ways. First, one can maintain that
the four enumerated methods deal with different aspects of legal reasoning. Such an interpretation is justified, e.g., by the relationship that holds
between logic and analysis or logic and argumentation. For instance, in
METHODS OF LEGAL REASONING
213
Alexy’s theory of argumentation the rules of classical logic are fully
observed but the theory amounts to more than that. The postulate of
applying the laws of logic is only one of the rules of rational discourse,
all of which have to be observed to obtain a rational practical decision.
However, not all of the four basic methods are complementary in the
analyzed sense. For example, hermeneutics aims at describing complete
human cognitive activities, not only their aspects.
Second, the complementariness thesis can be understood as follows:
different methods are applicable to different legal cases. On this account
the simplest (“algorithmic”) cases are solved with the use of logical and
analytic methods, more difficult cases with the use of argumentation
techniques, while the hardest require hermeneutic intuition. The obvious
weakness of such a solution is the need to justify why different legal cases
are solved with the use of different standards of reasoning. Moreover,
the distinction between simple and hard cases is problematic – or, anyway, is a matter of degree.
The third thesis concerning the interconnections between logic, analysis, argumentation and hermeneutics says that the tools offered by the
four philosophical stances should be combined to construct a coherent
method of legal reasoning. In particular, the combination of logic,
analysis and argumentation seems to suit such a construction. This idea
is realized, at least to certain extent, in Alexy’s theory as well as in the
conception presented in Chapter 4. The problem is that if we assume that
there exists only one correct “hybrid” method, we must then ask, what
the criterion is for establishing its correctness.
We are far from advocating one of the three theses, although the first
seems too trivial, and the third – at least from the general philosophical
perspective – too strong. We propose, instead, to look at the four methods as possible bases for constructing concrete conceptions of legal reasoning. One can wonder, of course, what criteria should be met by such
a conception. We claim that it is relatively easy to name at least two of
them: first, the conception should be coherent; and second, it should
serve its function, i.e., determine when the minimal requirement of
rationality of legal reasoning is fulfilled.
The general aim of our analyses made it difficult to base the presentation of the four methods on particular examples of legal reasoning.
Making use of examples – especially as regards argumentation and
hermeneutics – would be difficult for numerous reasons, not least
because it is relatively easy to choose examples justifying any thesis. On
the other hand, it is impossible to treat every type of legal case separately.
The strength of the methods we presented lies in the fact that they can
214
CHAPTER 6
help us to deal with hard cases, i.e., those that are unlike any known case
and, consequently, are difficult to imagine. In other words, our aim was
to present a set of tools and criteria for choosing between them in order to
show how to deal with any legal problem; we did not try to argue that a
given method suits a given class of legal cases.
Let us repeat: it was not our task to construct a coherent “technology
of interpretation”. We aimed to present a general philosophical framework within which such “technologies” can be constructed. Because of
that, we were unable to analyze all the bottom-line consequences resulting from adopting one of the four conceptions.
6.2
DILEMMAS OF THE CONTEMPORARY
PHILOSOPHY OF LAW
Our analyses were therefore metatheoretical in character. We believe that
such analyses are of extreme importance as they are often neglected in
the legal-theoretic considerations. The contemporary philosophy of law,
particularly the Anglo-Saxon version, is usually based on a set of tacit
assumptions, both methodological and ontological. It suffices to say that
British and American legal theorists confine themselves very often to the
methods and techniques of analytic philosophy. We do not claim that
this is a mistake. We would stress that it is not the only possible approach
and that a certain level of “methodological consciousness” requiring
metatheoretical reflection can only help us in legal-theoretic undertakings. The situation of the “continental” philosophy of law is even worse:
with minor exceptions, it is practiced either modo analitico, in connection
to the Anglo-Saxon conceptions, or with the use of the tools offered by
postmodernism, broadly understood. This is a very limiting alternative.
Contemporary Positivism
In order to substantiate the above stated claims let us have a closer look
at some of the contemporary debates in the philosophy of law. We believe
that many of the problems with which legal theorists are preoccupied
today result from the lack of methodological rigor. It seems that the most
hotly debated subject of the contemporary philosophy of law is the
soundness of positivism. In this context, the first problem is the very definition of legal positivism. One can maintain, however, at least with some
degree of adequacy, that the contemporary positivists defend the following three theses: (1) the so-called social sources thesis, which says that it is
social practice or social facts that constitute the sources of law, (2) the socalled conventionality thesis, according to which criteria of legal validity
METHODS OF LEGAL REASONING
215
are conventional, and (3) the so-called separability thesis, which claims
that there is no necessary connection between law and morality, i.e., moral
criteria do not constitute a “test of validity” of legal norms.1
All those theses have their strong opponents and advocates. Among the
opponents one should mention Dworkin who argues that any legal system consists not only of rules but also certain standards, of which some
(principles) are of moral pedigree. Dworkin shows that hard cases, such
as Riggs vs. Palmer described in Chapter 2, cannot be explained within
the positivistic conceptual scheme. The court that decided not to grant
Elmer Palmer rights to his inheritance acted according to a legal principle, which had not been explicitly enacted by the legislator and has an
obvious moral provenance.2
The soundness of Dworkin’s arguments is questioned by the so-called
“soft” positivism. According to this account, the separability thesis says
only that there is no necessary conceptual relationship between legal and
moral rules. The thesis says nothing, the advocates of soft positivism
argue, about their actual relationship. The distinction of two levels – the
conceptual and the factual – leads, however, to serious troubles. One of
them is pointed out by Raz.
According to Raz there is one feature of law that makes the distinction between the conceptual and the factual untenable. He argues in the
following way. At the outset he distinguishes between two kinds of reasons: those that justify beliefs and those that justify actions. The latter
concern the sphere of the practical and are reasons “for a person to perform an action when certain conditions obtain”.3 Among reasons Raz
distinguishes between first-order (reasons for action) and second-order
(reasons for acting for a reason) reasons.4 The second-order reasons can
be of two kinds: positive (reasons to act for some reason) and negative
(reasons not to act for some reason). The latter are called exclusionary
reasons.
Within this conceptual scheme Raz defines rules: they are a combination of a first order reason to act and an exclusionary reason. They
instruct us, therefore, to act in a certain way and to ignore other reasons
for action.5 Raz maintains, furthermore, that law has a claim to authority. This metaphorical expression means that law is a social institution
that consists of rules. In other words, the authority of law displays itself
in the fact that it provides us not only with reasons for acting but also
with exclusionary reasons not to act in the opposite way. It follows from
it, that for law to remain law, i.e., to realize its claim to authority, it has
to be understood as a social fact independent (both conceptually and
actually) from moral norms.6
216
CHAPTER 6
Conclusions
We believe that the contemporary debates concerning positivism are, to
a great extent, futile. There are different arguments justifying this opinion. First, the participants of the debate are trying to answer an ontological question (what is law?) without assuming any particular ontology.
The discussions are carried out without taking into account what the
contemporary ontology has to offer. Moreover, some philosophers, e.g.,
Raz, claim that their aim is to capture “the nature of law”. Only a very
peculiar interpretation of the term “nature” can save us here from serious philosophical problems, as looking for the “nature of law” presupposes both a kind of intellectual intuition capable of grasping “essences
of things” and the essences themselves. Such a solution is, naturally,
acceptable but only within a concrete philosophical stance, e.g., phenomenology. But Raz and his advocates do not go that far.
Second, the very idea of asking ontological questions may be put into
doubt. Although one can maintain that ontological decisions are important for developing a coherent conception of legal reasoning, adopting
the question “What is law?” as the starting point of any philosophizing
concerning law may seem a bad choice.
6.3
THE EPISTEMOLOGICAL APPROACH
Ontological debates lead, ultimately, nowhere. The strength of arguments backing rival conceptions is very often equal or incommensurable.
Furthermore, there is no commonly valid metatheoretical criterion of
determining “the right” ontology of law. The same may be said of legal
axiology. Some values are contrasted or compared with others but the
discussion inevitably leads to ontological dilemmas. It seems natural and
justified, therefore, to turn towards epistemology. When it is impossible
to construct a commonly acceptable philosophy of law (or, in other
words, answer to the question “What is law?”), there remain only epistemological considerations. But two questions follow immediately: is the
choice of epistemology which is released from the chains of ontology
fully free or even possible? Consequently, should we assume that the
process of legal cognition is indeterminate, dynamic and creative and
hence it is impossible to determine the limits thereof ?
The Limits of Legal Cognition
A positive answer to both questions posed at the end of the previous
paragraph leads to serious consequences, which sound unintuitive for
traditionally trained lawyers. Such a solution makes us believe that
METHODS OF LEGAL REASONING
217
theory of law (or jurisprudence) is a discipline without the
“Archimedean point of departure” (no matter how we understand it),
which in turn leads to cognitive or interpretational relativism. This
stance is accepted, at least to a certain extent, in some of the conceptions
presented in Chapter 1, namely the theories of Kirchmann, Hutcheson,
some phenomenological versions of hermeneutics and Critical Legal
Studies. For Kaufmann, a main representative of “ontologically oriented”
legal hermeneutics, legal cognition is completely intuitive, dynamic and
creative. Similar theses are advocated within some postmodern conceptions, which question all kinds of ontological or epistemological aprioricity. Is this approach justified?
We believe it is not justified for the following reasons. First, it is impossible to separate completely legal epistemology and ontology. Even if the
radical Kantian assumption concerning the unknowability of the “thing
in itself ” (Ding an sich) is accepted, we have to admit that this thing has
at least an indirect influence on our cognitive acts. In consequence there
is no pure (i.e., free from ontological assumptions) legal epistemology.
Moreover, no one would question the thesis that almost all people have
very strong, “archetypical” convictions concerning “the nature of law” as
well as a number of ontological intuitions which influence the choice of
legal methodology. Second, the choice of epistemology results in a limitation of the number of applicable methods or techniques of interpretation. Therefore, after an epistemological (methodological) choice is
made, it is not justified to speak of a total interpretational freedom.
Finally, we believe that there exist at least two easily distinguishable – if
not opposite – discourses in law: theoretical and practical. Their epistemological peculiarity limits severely our freedom of choosing methods
and techniques of interpretation.
Two Discourses
The limits of legal thinking or, in other words, of any possible legal epistemology are determined by two discourses: theoretical and practical.
The theoretical discourse is based on the criterion of truth, while the
practical takes advantage of “softer” normative criteria: rationality, reasonableness, justice, validity or efficiency. Both discourses interlace, but
they do not interfere with one another. We make transitions from
the theoretical to the practical or vice versa. The lack of recognition
of the fundamental differences between the two discourses is, so we
claim, the source of many futile debates in the philosophy of law. The
thesis that the findings of theoretical discourse are the sole basis for
normative decisions is mistaken and leads to some forms of cognitivism.
218
CHAPTER 6
The opposite thesis is likewise false. Theoretical discourse enjoys certain
influence on the practical. It is not, however, a logical connection.
Practical discourse has to adjust to the results of the theoretical (the theses
formulated within practical discourse must be consistent with the
theses established by theoretical discourse). In other words, theoretical
discourse sets the limits of any possible practical discourse. It is clear,
therefore, that the relationship between the discourses is asymmetrical.
Practical discourse (i.e., its limits and structure) depends on the findings
of the theoretical but not vice versa.
Theoretical discourse. Theoretical theses are used both in legal practice
and philosophy. The theses are formulated within the legal theoretical
discourse, which is autonomous in relation to the practical. This autonomy results from the fact that the theoretical discourse is “governed” by
the criterion of truth, “stronger” and more precise than the criteria of
evaluating the findings of practical discourse. Theoretical discourse is,
furthermore, purely “scientific”. The contemporary philosophy of science shows that it does not mean complete certainty. Nevertheless, the
successes of science make us believe that the theses of theoretical discourse at least “approach the truth”.
Practical discourse. Practical discourse is not completely autonomous,
as it is carried out within the limits imposed by the theoretical discourse.
It is likewise not purely “scientific”, as it uses criteria other than truth
(i.e., rationality, reasonableness, etc.). The particularity of practical discourse results from two distinct, but interrelated facts. First, normative
statements (directives, norms, rules) have a different epistemological status from the descriptive expressions: if they have meaning at all, they are
not sentences which can be evaluated as true or false.7 Second, the socalled hard cases have a very peculiar feature: they can have, we believe,
more than one “right answer”. The criterion of truth is replaced in this
context with other, “softer” or less determinate normative criteria. The
statements formulated within practical discourse cannot therefore meet
the requirements which are set for scientific theories. Ultimately, it is the
case at hand, skills and backgrounds of the participants of the discourse,
level of interpretational complication of the case together with the
social, political and economic contexts which decide – or at least influence – the choice of the method or technique of interpretation.
These remarks are not designed to show that within the practical discourse there are no methodological constraints. As we have tried to show,
some of those constraints result from the fact that we are dealing with a
METHODS OF LEGAL REASONING
219
practical – and not theoretical – discourse. Furthermore, once certain
choices are made – regardless of how we arrive at them – the chosen
method should be applied coherently and consequently.
The epistemological version of the philosophy of law consists therefore in distancing oneself from ontological debates, while insisting on
conceptual discipline and realizing the importance of the distinction
between practical and theoretical discourses.
6.4
UNFINISHED PROJECTS
The state of the post-modern (i.e., contemporary but not necessarily
postmodern) philosophy and theory of law and a diagnosis thereof
results in turning towards legal epistemology. One may ask: why is it so?
We would like to suggest a somewhat provocative answer to this question. We believe that all the stages in the development of legal theory (or
even of humanities), i.e., the “classical”, the “modern” and the “postmodern”, produced no more than certain unfinished projects.
Consequently, the general “science” of law becomes slowly a relict – a
discipline which is useless and usually misunderstood. Not needed by
either lawyers or philosophers, it is still there through the long tradition
and academic inertia. Most of its debates remind one of the scholastic
discussions concerning the number of evils or angels that can be placed
on the head of a pin. That is the reason why we have decided to present
four basic methods – or methodological stances – which are still important for the humanistic methodology and, more importantly, are applied
in actual legal cases.
We would like to conclude our analyses with a short reflection on three
most important, albeit never completed, projects developed within the
modern and contemporary philosophy and theory of law.
Classicism
The term “classicism” is rich in meanings, but occurs very rarely in
legal-theoretic discussions. Let us use it to designate the “project” of the
eighteenth and nineteenth centuries, founded on “classical philosophy”,
i.e., on scholasticism (in particular, Thomism), Enlightenment (Hobbes,
Rousseau, Kant) and idealism (“Fichte”, Schelling, Hegel). These philosophical conceptions share a belief in a supra-positive law (divine law or
natural law). Naturally, there is more that divides those conceptions than
unites them. It suffices to mention the sources of law, its justification etc.
As a result, the project could not have been completed for three reasons.
First, there were serious internal tensions within the project; in other
220
CHAPTER 6
words, the diversity of ideas within it was irreconcilable. Second, it was
philosophically “one-sided”, as it paid its attention to the concept of natural law. Because of that, the project got old fairly quickly in face of the
developments and changes in European social, political and economical
reality of the nineteenth century. Third, at the same time there emerged
rival conceptions, the majority being ostensibly critical towards the classical project and formulated in the spirit of modernity. In this context
one should mention the German historical school and, especially, legal
positivism.
Modernism
The modernistic breakthrough was a reaction to the political and economical changes in Europe in the nineteenth and at the beginning of the
twentieth century. In such a context, legal theory’s aim was to provide
justification for the development of democratic societies, the systems of
public security within the limits of the rule of law and the principles protecting the free market. Those ends were to be realized with the use of
new legal-theoretic conceptions, which followed strict ontological and
methodological rigors. In this way the tasks of the philosophy and theory of law were understood within legal positivism (both the continental
in its various incarnations – Gesetzespositivismus, Begriffsjurisprudenz,
normativism, and the analytically oriented positivism of Austin and
Hart). The same may be said of legal realism, some schools of philosophical analysis and some conceptions of legal argumentation. There
are many reasons of the crisis of the “modernistic project”. First, one
should mention the ontological extremism of the mentioned conception,
the “fetishisation” of “one and only” account of what is law. Second, the
methodological rigor of the modernistic project was too severe. Third,
the high level of complication of at least some of the conceptions of
interpretation developed in the twentieth century made them inapplicable in practice. Fourth, there were no rational criteria for choosing
between rival theories of law.
Postmodernism
“Postmodernism” is a very fashionable word, but it is hard to say what
signifies in relation to the philosophy and theory of law. It can even be
asked whether there is such a thing as a “postmodern project” in those
two disciplines. Postmodernism is, arguably, a reaction to modernism, an
attempt to deconstruct, to use another fashionable word, all paradigms,
particularly ontological ones. Instead of limitations and methodological
rigors we get “free epistemology” in which “anything goes”. One can use
METHODS OF LEGAL REASONING
221
any method which suits some social or individual ends. Interpretational
decisions are made within an open, unlimited “narration”. The elements
of this way of thinking can be found in several contemporary conceptions of law and legal reasoning, such as Critical Legal Studies,
Kaufmann’s hermeneutics, Luhman’s and Teubner’s system theories and
Habermas’ discursive theory of law. Although this project is still alive –
and even “fashionable” – we believe it will inevitably get the label “unfinished”. There is a plethora of reasons for this. First, the project is placed
in an ontological and methodological vacuum. It is based on a philosophy without any “Archimedean point”, without any initial assumptions;
hence, it is neither justifiable nor refutable. Second, postmodernism is a
set of – sometimes totally – different, incoherent conceptions, so that
one can even doubt whether it is a project (even in the very loose sense
of the word which we apply here), or rather a number of ideas that
can answer any question whatsoever. Third, some of the postmodern
theories are formulated in very obscure and complicated language which
disguises old conceptions and makes them appear fresh and original.
NOTES
1. Cf. S. Bertea, “On Law’s Claim to Authority”, Northern Ireland Legal Quarterly,
vol. 52, no. 4, p. 402.
2. Cf. R. Dworkin, Taking Rights Seriously, op. cit., passim.
3. J. Raz, Practical Reason and Norms, London, Hutchinson, 1990, p. 19.
4. Ibidem, pp. 39–40.
5. Ibidem, pp. 39–48.
6. Cf. J. Raz, Ethics in the Public Domain, Oxford, Clarendon, 1994.
7. After 100 years of the debate concerning the meaning of normative statements, we
believe that it is moderate noncognitivism that has the strongest argument behind it.
Both cognitivism and extreme noncognitivism (e.g., emotivism) seem less persuasive.
INDEX OF NAMES
Aarnio, A. 13, 164n
Ajdukiewicz, K. 65n, 74
Alchian, A.A. 92
Alchourron, C. 3
Alexy, R. 13, 15n,107n, 112, 123, 132,
141-144, 151, 152, 164n, 165n, 213
Allen, D.W. 109n
Alwart, H. 198
Apel, O. 129, 175, 182, 183
Aristarch 171
Aristotle 30, 65n, 70, 114, 115, 116, 117,
118, 119, 124, 125, 133, 135, 136, 140,
148, 164n, 167
Arnauld, A. 71
Augistine, St. 169, 201
Austin, J. 8, 9, 79, 80, 81,107n, 220
Austin, J.L. 75, 76, 77, 79, 83-89, 90,
107n, 108n, 122
Ayer, J.L. 122
Baeyer, A. 198
Baier, K. 6, 122, 142
Baumgarten, A.G. 170, 171
Beaney, H. 107n
Becker, G. 104, 109n
Becker, O. 3, 31
Belnap, N. 43, 66n
Bentham, J. 4, 5, 92, 100
Bergbohm, K. 9
Bertea, S. 221n
Betti, E. 113, 173, 180, 185, 186, 207
Bierling, E.R. 9
Binding, K. 9
Black, M. 75
Boche ński, J.M. 71, 72, 73, 107n, 181
Böckh, A. 176
Bolzano, B. 31
Boole, G. 20
Brito, J.S. 198
.
Broz ek , B. 108n
Bulygin, E. 3
Buridan, J. 71
Calabresi, G. 92
Calera, L. 198
Carmom J. 65n
Carnap, R. 70, 73, 107n
Cassiodor 169
Chisholm, R. 33-35, 66n
Chladenius, J.M. 170
Cicero 117, 118, 133, 136, 172
Coase, R. 92, 103-104, 109n
Coing, H. 173, 185, 186, 208n
Comte, A. 4, 5
Coreth, E. 209n
Cresswell, M.J 65n
Dannhauer, J.C. 168, 170, 171
Davidson, D. 74
Descartes, R. 69, 71, 106n, 107n
Dilthey, W. 6, 113, 168, 171, 172, 174,
176, 177-179, 180, 181, 182, 183, 184,
185, 188, 189, 202, 208n
Diodoros 169
Diogenes Laertios 116
Dissen, L. 176
Dray, W.H. 13
Droysen, J.G. 177, 178
Dubislav, W. 31
Dummet, M. 74
Dworkin, R. 8, 57, 67n, 76, 83, 107n, 138,
152, 160, 165n, 215, 221n
Ebeling, G. 182
Eckhardi, H. 6, 172
Eide, E. 108n
Ellscheid, G. 198
Engisch, K. 137, 187, 207, 208n
223
224
INDEX OF NAMES
Ernesti, J.A. 170, 171
Esser, J. 137, 173, 176, 185-187, 198, 207,
208, 208n, 209n
Etchemedy, J. 65n
Eucherius from Lyon 169
Evans, G. 74
Fahrenbach, H. 164n
Fichte, J.G. 219
Flacius 169, 170, 178, 180
Flew, A. 66n
Forrester, J.W. 66n
Forsthoff, E. 187, 208n
Frege, G. 20, 21, 71, 73
Fuchs, E. 182
Fuller, L.L. 162, 165n
Fuhrmann, M. 198
Gabbay, D. 65n, 67n
Gadamer, H.G. 6, 113, 117, 168, 172, 173,
182, 186, 187, 189-192, 195-196, 198,
201, 202, 203, 204, 207, 209n
Geach, P. 40, 66n
Gellner, E. 167, 208n
Geny, F. 3
Goble, L. 65n, 66n
Gorgis 117
Grabowski, A. 108n
Grobler, A. 107n
Haba, E.P. 198
Habermas, J. 6, 122-123, 125, 127, 129,
141, 142, 151, 164n, 175, 182, 183, 221
Hacking, I. 107n
Haft, F. 198
Hägerström, A. 4
Hanson, J.D. 108n
Hare, R.M. 31, 122, 142
Hart, H.L.A. 3, 8, 9, 49-50, 66n, 76, 77,
78-82, 138, 152, 220
Hart, M.R. 107n, 108n
Hartmann, N. 66n
Hassemer, W. 198, 207, 209n
Hegel, G.W.F. 8, 176, 219
Heidegger, M. 6, 168, 172, 173, 182, 187,
189, 190, 205, 209n
Heidorf, L.F. 176
Herder, J.G. 176
Heyne, Ch.G. 176
Hilpinen, R. 65n
Hinderling, H.G. 198, 207, 209n
Hintikka, J. 12, 15n, 65n, 74, 106n
Hipparch 171
Hobbes, T. 219
Höfler, A. 31
Hofstadter, A. 31
Holmes. O.W. 4, 14n, 92, 108n
Homer 171
Horty, J.F. 45-46, 66n
Hruschka, J. 13
Hughes, G.E. 65n
Husserl, E. 31, 168, 173, 182, 187,
188-189, 193, 194, 201
Husserl, G. 194-195, 209n
Hutcheson, J.C. 1, 11, 14n, 217
Jhering, R. 9
Jörgensen, J. 31, 40, 47-49, 63, 65n
Jones, A.J.I. 65n
Julius Africanus 169
Kalinowski, J. (G.) 3, 31, 65n, 66n, 115,
157, 164n
Kanger, S. 65n
Kant, I. 8, 14, 71, 123, 125, 141, 147, 149,
219
Kantorowicz, H. 3
Kaplow, L. 108n
Kaufmann, A. 14, 113, 117, 173, 196-198,
203, 207, 208, 209n, 217, 221
Kelsen, H. 5, 8, 9, 11, 136, 138
Kirchmann von, J.H. 1, 14n, 179, 217
Korolko, M. 164n
Kotarbiński, T. 74
Krawietz, W. 66n, 164n
Kriele, M. 137, 187, 198, 209n
Kripke, S. 30
Lapie, E. 31
Larenz, K. 137, 173, 175, 185-186, 187,
198, 207, 208, 208n, 209n
Leibniz, G.W.F. 30, 71
Leicht, R. 198
Leśniewski, S. 74
Lewis, C.I. 29
Lipps, H. 113, 179, 180, 181, 208n
INDEX OF NAMES
Liszt, F. 9
Litewski, W. 15n
Llewellyn, K. 4, 14n
Lorenzen, P. 6, 122
Luhmann, N. 5, 129, 164n, 221
Lukasiewicz, J. 74
Lundstedt, V. 4
Machiavelli, N. 92
Mackaay, E. 108n
Maihofer, W. 173, 195, 209n
Mally, E. 31
Maturana, H. 5
McKinsey, J.C.C. 31
Medema, S.G. 109n
Meier, G.F. 170, 171, 180
Meinong, A. 66n
Melanchton, F. 170
Menger, E. 31
Merkl, A. 9
Michaelis, J.D. 170, 171
Mill, J.S. 5, 92
Misch, F. 113, 179, 180, 181
Montesquieu 92
Moore, G.E. 4, 71, 73, 75
Müller, F. 187, 209n
Müller-Dietz, H. 198
Neumann, U. 198
Neurath, O. 73
Nicole, P. 71
Olbrechts-Tyteca, L. 123
Ollero, A. 198
Opalek, K. 15n, 76, 83, 107n
Origen 169
Pappus 69, 70
Pawlusiewicz, K. 108n
Pearson, H. 108n
Peczenik, A. 164n
Peirce, C.S. 20
Perelman, Ch. 13, 112, 117, 118, 123, 130,
131, 133-135, 136, 138, 139-141, 148,
151, 153, 157, 164n, 165n
.
Petraz ycki, L. 3, 12, 14n, 165n
Philipps, L. 198
Plato 70, 116, 119, 124, 176, 179, 190
225
Posner, R. 5, 14n, 92, 160
Pound, R. 4, 14n
Prakken, H. 67n
Priest, G. 65n
Puchta, G.F. 1
Putnam, H. 74
Quine. W.V.O. 23, 24, 65n, 74
Quintilian 117, 118, 133
Rand, R. 31
Ranke, L. 176
Raz, J. 76, 215-216, 221n
Reinach, A. 113, 117, 173, 193-194, 207,
208, 209n
Remes, U. 106n
Restall, G. 65n
Ricoeur, P. 113, 173, 175, 183, 192-193,
204, 208n, 209n
Robinson, J.M. 182
Ross, A. 3, 4, 14n, 33, 83, 107n
Rottleuthner, H. 198
Rousseau, J.J. 219
Russell, B. 20, 21, 69, 71, 72, 73, 81, 82,
106n, 107n
Ryle, G. 75, 76
Salomon, M. 136
Sammet, J.G. 6, 172
Sarkowicz, R. 14n, 91, 108n, 165n
Savigny von, F.C. 1, 6, 8, 173, 176, 184
Schelling, F.W.J. 219
Schild, W. 198
Schlegel, F. 176
Schleiermacher, F.E.D. 6, 113, 168, 171,
172, 176-177, 178, 180, 181, 183, 184,
185, 189, 199, 201, 202, 208n
Schlick, M. 73
Scholler, H. 198
Schopenhauer, A. 116, 117, 119, 120,
164n
Schroth, U. 198
Schwemmer, O. 6, 122, 151
Searle, J. 75, 76, 83,88, 89, 108n
Segerberg, K. 66n
Semler, J.S. 170, 171
Shavell, S. 108n, 109n
Socrates 70, 115, 192, 201
226
INDEX OF NAMES
Stebbing, S. 75
Stelmach, J. 14n, 15n, 164n, 165n, 198,
208n, 209n
Stevenson, Ch. 6, 122
Strawson, P.F. 75, 76, 77, 78, 81, 82, 90,
107n
Stroll, A. 107n
Struck, G. 134, 138, 164n, 165n
Szymanek, K. 164n
Tarello, G. 134, 164n
Tarski, A. 17, 18, 53, 65n, 74
Tertulian 169
Teubner, G. 5, 14n, 221
Theodoros 169
Thon, A. 9
Toulmin, S. 6, 122
Twardowski, K. 74
Ulen, Th.S. 108n
Valdes, E.G. 66n
Varela, F. 5
Vico, G.B. 177
Viehweg, Th. 123, 133, 135-137, 164n
Vreeswijk, G. 67n
Waismann, F. 73
Walter, R. 66n
Whitehead, A.N. 20, 73
Winch, P. 13
Windscheid, B. 9
Wisdom, J. 75
Wittgenstein, L. 3, 6, 13, 73, 75, 83, 122,
181
Wittich, C.F. 6, 172
Wójcicki, R. 107n
Woleński, J. 3, 65n, 66n, 107n
Wolf, F.A. 176
Wright von, G.H. 3, 13, 31, 42, 66n
Wróblewski, J. 28, 65n
Zaccaria, G. 198
Zerbe, R.O. 109n
Zimmerling, R. 66n
Zippelius, R. 138
SUBJECT INDEX
actualization 188, 195, 204-207
analysis: 12-15, 69-109
- as decomposition 70, 71, 72
- as search for logical reasons 70, 72
- as translation 70, 71, 105
- conceptual scheme thesis 105
- economic 5-6, 92-104
- hard (formal) 2-3, 73-74
- linguistic 2-3, 75-91, 145, 181
- soft (anti-formal) 2-3, 72-91
analytic philosophy 2-3, 72-75
analytic philosophy of law 2-3, 78-89
argumentation 6, 12-15, 111-165
argumentation from paradigms 80
arguments (legal) 156-160
- a cohaerentia 158
- a completudine 159
- a contrario 13, 134, 143, 156-157
- a fortiori 13, 115, 131, 134, 143, 157
- a loco communi 158
- a loco specifici 158
- a rerum natura 158
- a simili 161
- ab exemplo 134, 157-158
- economic 160
- historical 159-160
- psychological 159
- sociological 159
- systematic 159
- teleological 159
cognitivism 48, 49, 217
common places (loci communes) 116, 119,
132, 133, 134, 144, 148, 156, 158
constatives 85-87
context of discovery 19, 63
context of justification 19, 20, 63, 64
criteria of practical discourse: 123-130
- efficiency 128-130
- rightness 127-128
- truth 124-127
Critical Legal Studies 1, 2, 5, 11,
217, 221
Dasein 182, 187, 188, 189, 195, 200, 201,
205, 207
defeasibility 49-50
definition:
- in linguistic philosophy 79
- of law 78, 79, 80
discourse: 6, 20, 31, 47, 49, 50, 63, 64, 83,
91, 106, 111ff, 212, 213, 217-219
- legal 20, 50, 63, 64, 106, 112, 119, 123,
126, 128, 130-164, 212
- practical 31, 64, 83, 91, 111, 112,
114, 115, 118, 119, 121, 122,
123-130, 131, 132, 134, 135,
139-161, 218-219
- theoretical 6, 50, 123, 125, 141-142,
146, 152, 163, 217-218
economic analysis: 5-6, 92-104
- Coase’s Theorem 103-104
- descriptive 93-94
- normative 93-94
- stages 97
- the basic thesis 92
efficiency 81, 82, 93, 97, 102-103, 104,
111, 112, 118, 120, 121, 124, 125,
126, 127, 128-130, 140, 150, 217
epistemology of understanding 200
hard cases 54, 57, 64, 113, 146, 151-152,
153-154, 160, 162, 163, 207, 213, 214,
215, 218
hermeneutic circle 178, 191, 205, 206
227
228
SUBJECT INDEX
hermeneutic experience: 190, 199
- subtilitas applicandi 190
- subtilitas explicandi 190
- subtilitas intelligendi 190
hermeneutics (general) 167-168, 173-183,
187-193
hermeneutics (specialized):
- biblical 168-170
- legal 172-173, 183-187, 193-208
- philological 170-172
Historical School 1, 7-8, 92, 220
historicity of understanding 190, 191, 195,
196, 200, 202, 203-205
hunch 1, 11
intuition: 1, 18, 19, 26, 30, 35, 40, 63-64,
70, 89, 91, 104, 111, 122, 142, 174,
175, 193, 194, 201, 202, 213, 216
- phenomenological 111, 174
- psychological 111, 174
- rational (analytic) 111, 174
Jørgensen Dilemma 46-49, 40, 63
justice 80, 93, 102, 103, 104, 105, 126,
138, 160, 196, 217
justification: 19, 53, 63, 112, 113, 114, 124,
141, 142, 143, 163, 186, 188
- external (externe Rechfertigung) 19,
124, 143, 152
- internal (interne Rechtfertigung) 19,
124, 143
Kirchmann’s accusation 1
Lebenswelt 182, 187, 188
legal argumentation 145-163
legal dogmatics 8, 10, 91, 106, 160, 163,
198, 205, 207
legal realism: 3-4, 5, 10, 92, 145, 207, 220
- American 1, 4, 5, 92, 129
- psychologism 3
- Scandinavian 4
- School of Free Law 3
- sociological jurisprudence 4
legal syllogism 26, 27, 29, 35, 136
legal topoi: 7, 13, 156-163
- arguments 156-160
- legal principles 160-163
linguistic analysis: 2-3, 75-91, 145, 181
- evolutionary argument 76-77
- conceptual scheme 76-78
- general directive 89-90
- maximalist 77
- minimalist 77
linguistic philosophy 75-92
linguistic turn 71, 72
logic: 17-67
- classical 20-29
- defeasible 49-63
- deontic 29-40
- of agency 40-49
- of induction 64
- nonformal 64
- nonmonotonic 50-51, 53-54
- probability 64
logical consequence 17, 18, 27, 53, 60, 61, 63
logical form 18
Methodenlehre 6, 8, 133, 135, 136, 137,
173, 184, 185, 198, 207
methodological autonomy of jurisprudence
6-9, 136, 173, 184
methodological heteronomy of
jurisprudence 2-6, 11
modalities: 29
- aletic 29-30
- deontic 30-33
- forbidden 32
- obligatory 31
- ought-to-be 40-41, 42, 46
- ought-to-do 40-41, 42, 46
- permitted 32
modus ponens 21, 22, 25, 28, 50, 51, 52,
57, 58, 60
noncognitivism 48, 221
ontology of understanding 188-193
ordinary language 2, 3, 12, 73, 75, 76, 77,
78, 79, 80, 84, 87, 88, 89, 90, 91, 105,
106, 151, 155, 182, 193
paradoxes: 23-24, 33-35, 40, 63, 64
- CTD 33-35
- Chisholm 33-35
- of deontic logic 33-35
SUBJECT INDEX
- of material implication 23-24
- Ross 33
performatives: 83-87, 89
- happy 83
- unhappy 83
philosophies of argumentation: 112-123
- dialectics 115-117
- eristic 120-121
- hermeneutics 113, 167-209
- logic 17-67, 113-115
- rhetoric 117-118
- sophistry 120-121
- topics 118-120
positivism (legal) 3, 4, 5, 7, 8-9, 10, 26,
133, 145, 214-215, 216, 220
presuppositions 81-82, 86, 90, 91, 106
pre-understanding (Vorverständnis –
Vorurteil) 163, 174, 180, 186, 187,
188, 191, 200, 202, 204, 205
principles of legal discourse: 160-163
- general 161
- of interpretation 161-162
- special 162-163
Roman jurisprudence 1, 6, 7, 136, 138,
156, 161, 162, 163, 172
rules of practical discourse: 149-155
- basic 149-154
- of passage 154-155
semantics: 18, 19, 29-30, 31, 35, 40, 44, 45,
46, 64
- of classical logic 21
- possible worlds 29-30, 40,
Sonderfallthese (Special Case Thesis): 112,
131, 132, 142, 143, 154
- Additionsthese 132, 143
229
- Integrationsthese 132, 143
- Sekundaritätsthese 132, 143
special places (loci specifici) 133, 134, 135,
148, 156, 158
speech acts: 82-89
- act of utterance 88
- Austin’s theory 82-88
- illocutions 88-89
- locutions 88-89
- perlocutions 88-89
- propositional acts 88
- Searle’s theory 88-89
STIT 43-46
syntax 18, 19
systems theory 5
theories of hermeneutics: 175-198
- analytic hermeneutics 180-182
- hermeneutics as epistemology 175-187
- hermeneutics as ontology 187-198
- hermeneutics as theory of
communication 182-183
theories of practical discourse: 130-145
- procedural 139-145
- topical-rhetorical 132-139
tradition 170, 183, 191, 193, 200, 204
typology 84, 89
understanding: 167-168, 174-182, 187-193
- clairvoyant 176, 184, 201,
- comparative 176, 179, 184
- direct 174, 188, 198, 200, 201-202
- indirect 174, 188, 198, 201-202, 203,
204, 205
understanding of the law 195, 198-208
vagueness 85, 87, 90, 91
Law and Philosophy Library
1.
E. Bulygin, J.-L. Gardies and I. Niiniluoto (eds.): Man, Law and Modern Forms of Life. With
an Introduction by M.D. Bayles. 1985
ISBN 90-277-1869-5
2.
W. Sadurski: Giving Desert Its Due. Social Justice and Legal Theory. 1985
ISBN 90-277-1941-1
3.
N. MacCormick and O. Weinberger: An Institutional Theory of Law. New Approaches to Legal
Positivism. 1986
ISBN 90-277-2079-7
4.
A. Aarnio: The Rational as Reasonable. A Treatise on Legal Justification. 1987
ISBN 90-277-2276-5
5.
M.D. Bayles: Principles of Law. A Normative Analysis. 1987
ISBN 90-277-2412-1; Pb: 90-277-2413-X
6.
A. Soeteman: Logic in Law. Remarks on Logic and Rationality in Normative Reasoning,
Especially in Law. 1989
ISBN 0-7923-0042-4
7.
C.T. Sistare: Responsibility and Criminal Liability. 1989
ISBN 0-7923-0396-2
8.
A. Peczenik: On Law and Reason. 1989
ISBN 0-7923-0444-6
9.
W. Sadurski: Moral Pluralism and Legal Neutrality. 1990
ISBN 0-7923-0565-5
10.
M.D. Bayles: Procedural Justice. Allocating to Individuals. 1990
ISBN 0-7923-0567-1
11.
P. Nerhot (ed.): Law, Interpretation and Reality. Essays in Epistemology, Hermeneutics and
Jurisprudence. 1990
ISBN 0-7923-0593-0
12.
A.W. Norrie: Law, Ideology and Punishment. Retrieval and Critique of the Liberal Ideal of
Criminal Justice. 1991
ISBN 0-7923-1013-6
13.
P. Nerhot (ed.): Legal Knowledge and Analogy. Fragments of Legal Epistemology, Hermeneutics and Linguistics. 1991
ISBN 0-7923-1065-9
14.
O. Weinberger: Law, Institution and Legal Politics. Fundamental Problems of Legal Theory
and Social Philosophy. 1991
ISBN 0-7923-1143-4
15.
J. Wróblewski: The Judicial Application of Law. Edited by Z. Bańkowski and N. MacCormick.
1992
ISBN 0-7923-1569-3
16.
T. Wilhelmsson: Critical Studies in Private Law. A Treatise on Need-Rational Principles in
Modern Law. 1992
ISBN 0-7923-1659-2
17.
M.D. Bayles: Hart’s Legal Philosophy. An Examination. 1992
18.
D.W.P. Ruiter: Institutional Legal Facts. Legal Powers and their Effects. 1993
ISBN 0-7923-2441-2
19.
J. Schonsheck: On Criminalization. An Essay in the Philosophy of the Criminal Law. 1994
ISBN 0-7923-2663-6
20.
R.P. Malloy and J. Evensky (eds.): Adam Smith and the Philosophy of Law and Economics.
1994
ISBN 0-7923-2796-9
21.
Z. Bańkowski, I. White and U. Hahn (eds.): Informatics and the Foundations of Legal Reasoning. 1995
ISBN 0-7923-3455-8
22.
E. Lagerspetz: The Opposite Mirrors. An Essay on the Conventionalist Theory of Institutions.
1995
ISBN 0-7923-3325-X
ISBN 0-7923-1981-8
Law and Philosophy Library
23.
M. van Hees: Rights and Decisions. Formal Models of Law and Liberalism. 1995
ISBN 0-7923-3754-9
24.
B. Anderson: "Discovery" in Legal Decision-Making. 1996
25.
S. Urbina: Reason, Democracy, Society. A Study on the Basis of Legal Thinking. 1996
ISBN 0-7923-4262-3
26.
E. Attwooll: The Tapestry of the Law. Scotland, Legal Culture and Legal Theory. 1997
ISBN 0-7923-4310-7
27.
J.C. Hage: Reasoning with Rules. An Essay on Legal Reasoning and Its Underlying Logic.
1997
ISBN 0-7923-4325-5
28.
R.A. Hillman: The Richness of Contract Law. An Analysis and Critique of Contemporary
Theories of Contract Law. 1997
ISBN 0-7923-4336-0; 0-7923-5063-4 (Pb)
29.
C. Wellman: An Approach to Rights. Studies in the Philosophy of Law and Morals. 1997
ISBN 0-7923-4467-7
30.
B. van Roermund: Law, Narrative and Reality. An Essay in Intercepting Politics. 1997
ISBN 0-7923-4621-1
31.
I. Ward: Kantianism, Postmodernism and Critical Legal Thought. 1997
32.
H. Prakken: Logical Tools for Modelling Legal Argument. A Study of Defeasible Reasoning
in Law. 1997
ISBN 0-7923-4776-5
ISBN 0-7923-3981-9
ISBN 0-7923-4745-5
33.
T. May: Autonomy, Authority and Moral Responsibility. 1998
ISBN 0-7923-4851-6
34.
M. Atienza and J.R. Manero: A Theory of Legal Sentences. 1998
ISBN 0-7923-4856-7
35.
E.A. Christodoulidis: Law and Reflexive Politics. 1998
ISBN 0-7923-4954-7
36.
L.M.M. Royakkers: Extending Deontic Logic for the Formalisation of Legal Rules. 1998
ISBN 0-7923-4982-2
37.
J.J. Moreso: Legal Indeterminacy and Constitutional Interpretation. 1998
ISBN 0-7923-5156-8
38.
W. Sadurski: Freedom of Speech and Its Limits. 1999
39.
J. Wolenski (ed.): Kazimierz Opalek Selected Papers in Legal Philosophy. 1999
ISBN 0-7923-5732-9
40.
H.P. Visser ’t Hooft: Justice to Future Generations and the Environment. 1999
ISBN 0-7923-5756-6
41.
L.J. Wintgens (ed.): The Law in Philosophical Perspectives. My Philosophy of Law. 1999
ISBN 0-7923-5796-5
42.
A.R. Lodder: DiaLaw. On Legal Justification and Dialogical Models of Argumentation. 1999
ISBN 0-7923-5830-9
43.
C. Redondo: Reasons for Action and the Law. 1999
44.
M. Friedman, L. May, K. Parsons and J. Stiff (eds.): Rights and Reason. Essays in Honor of
Carl Wellman. 2000
ISBN 0-7923-6198-9
45.
G.C. Christie: The Notion of an Ideal Audience in Legal Argument. 2000 ISBN 0-7923-6283-7
ISBN 0-7923-5523-7
ISBN 0-7923-5912-7
Law and Philosophy Library
46.
R.S. Summers: Essays in Legal Theory. 2000
ISBN 0-7923-6367-1
47.
M. van Hees: Legal Reductionism and Freedom. 2000
ISBN 0-7923-6491-0
48.
R. Gargarella: The Scepter of Reason. Public Discussion and Political Radicalism in the Origins
of Constitutionalism. 2000
ISBN 0-7923-6508-9
49.
M. Iglesias Vila: Facing Judicial Discretion. Legal Knowledge and Right Answers Revisited.
2001
ISBN 0-7923-6778-2
50.
M. Kiikeri: Comparative Legal Reasoning and European Law. 2001
51.
A.J. Menéndez: Justifying Taxes. Some Elements for a General Theory of Democratic Tax
Law. 2001
ISBN 0-7923-7052-X
52.
W.E. Conklin: The Invisible Origins of Legal Positivism. A Re-Reading of a Tradition. 2001
ISBN 0-7923-7101-1
53.
Z. Bańkowski: Living Lawfully. Love in Law and Law in Love. 2001
54.
A.N. Shytov: Conscience and Love in Making Judicial Decisions. 2001
ISBN 0-7923-6884-3
ISBN 0-7923-7180-1
ISBN 1-4020-0168-1
55.
D.W.P. Ruiter: Legal Institutions. 2001
ISBN 1-4020-0186-X
Volumes 56–63 were published by Kluwer Law International.
56.
G. den Hartogh: Mutual Expectations. A Conventionalist Theory of Law. 2002
ISBN 90-411-1796-2
57.
W.L. Robison (ed.): The Legal Essays of Michael Bayles. 2002
58.
U. Bindreiter: Why Grundnorm? A Treatise on the Implications of Kelsen’s Doctrine. 2002
ISBN 90-411-1867-5
59.
S. Urbina: Legal Method and the Rule of Law. 2002
60.
M. Baurmann: The Market of Virtue. Morality and Commitment in a Liberal Society. 2002
ISBN 90-411-1874-8
61.
G. Zanetti: Political Friendship and the Good Life. Two Liberal Arguments against Perfectionism. 2002
ISBN 90-411-1881-0
62.
W. Sadurski (ed.): Constitutional Justice, East and West. 2002
ISBN 90-411-1883-7
63.
S. Taekema: The Concept of Ideals in Legal Theory. 2003
ISBN 90-411-1971-X
64.
J. Raitio: The Principle of Legal Certainty in EC Law. 2003
ISBN 1-4020-1217-9
65.
E. Santoro: Autonomy, Freedom and Rights. A Critique of Liberal Subjectivity. 2003
ISBN 1-4020-1404-X
66.
S. Eng: Analysis of Dis/agreement – with particular reference to Law and Legal Theory. 2003
ISBN 1-4020-1490-2
67.
D. González Lagier: The Paradoxes of Action. (Human Action, Law and Philosophy). 2003
ISBN Hb-1-4020-1661-1
68.
R. Zimmerling: Influence and Power. Variations on a Messy Theme. 2004
ISBN Hb-1-4020-2986-1
ISBN 90-411-1835-7
ISBN 90-411-1870-5
Law and Philosophy Library
69.
A. Stranieri and J. Zeleznikow (eds.): Knowledge Discovery from Legal Databases. 2005
ISBN 1-4020-3036-3
70.
J. Hage: Studies in Legal Logic. 2005
ISBN 1-4020-3517-9
71.
C. Wellman: Medical Law and Moral Rights. 2005
ISBN 1-4020-3751-1
72.
T. Meisels: Territorial Rights. 2005
ISBN 1-4020-3822-4
73.
G.W. Rainbolt: The Concept of Rights. 2005
ISBN 1-4020-3976-X
74.
O. Ezra: Moral Dilemmas in Real Life. Current Issues in Applied Ethics. 2006
ISBN 1-4020-4103-9
75.
N.T. Casals: Group Rights as Human Rights. A Liberal Approach to Multiculturalism. 2006
ISBN 1-4020-4208-6
76.
C. Michelon Jr.: Being Apart from Reasons. The Role of Reasons in Public and Private Moral
Decision-Making. 2006
ISBN 1-4020-4282-5
77.
A.J. Menendez and E.O. Eriksen (eds): Arguing Fundamental Rights. 2006
ISBN 1-4020-4918-8
78.
J. Stelmach and B. Brozek: Methods of Legal Reasoning. 2006
ISBN 1-4020-4936-6
springer.com